Journal of Semitic Studies, Volume 55, Number 1, Spring 2010

92897_JOSS_2010/1_Cv 08-03-2010 11:22 Pagina 1 EDWARD LIPINSKI 1 ROBERT S. KAWASHIMA ‘“Orphaned” Converted Tense ...

Author: Editorial Committee

126 downloads 2790 Views 4MB Size Report

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

DOWNLOAD PDF

92897_JOSS_2010/1_Cv

08-03-2010

11:22

Pagina 1

EDWARD LIPINSKI

1

ROBERT S. KAWASHIMA

‘“Orphaned” Converted Tense Forms in Classical Biblical Hebrew Prose’

11

BAHAA AMER AL-JUBOURI

‘Nouvelles inscriptions araméennes du temple de Nannay à Hatra’

SPRING 2010

‘Le Gérondif en Phénicien’

ISSN 0022-4480 (PRINT) ISSN 1477-8556 (ONLINE)

37

GERRIT BOS

53

AARON D. RUBIN

‘The Development of the Amharic Definite Article and an Indonesian Parallel’

103

ABDULRAHMAN S. AL SALIMI

‘Identifying the (Iba∂i/Omani) Siyar’

115

‘Rhythm and Beat: Re-evaluating Arabic Prosody in the Light of Mahri Oral Poetry’

163

SADOK MASLIYAH

‘The Folk Songs of Iraqi Children: Part One’

183

JAMES DICKINS

‘Basic Sentence Structure in Sudanese Arabic’ REVIEWS SHORT NOTICES

237 263 311

Semitic Studies

SAM LIEBHABER

JOURNAL OF

‘Medical Terminology in the Hebrew Tradition: Shem Tov Ben Isaac, Sefer ha-Shimmush, Book 30’

JOURNAL OF

Semitic Studies VOLUME LV. NO. 1 SPRING 2010

VOLUME LV. NO. 1

92897_JOSS_2010/1_Cv

08-03-2010

11:22

Pagina 2

SUBSCRIPTIONS A subscription to Journal of Semitic Studies comprises 2 issues. Prices include postage by surface mail, or for subscribers in the USA and Canada by airfreight, or in India, Japan, Australia and New Zealand, by Air Speeded Post. Airmail rates are available on request. Annual Subscription Rate (Volume LIII, 2 issues, 2008) Institutional Print edition and site-wide online access: £166//249/US$315 Print edition only: £152//228/US$289 Site-wide online access only: £138//207/US$262 Personal Print edition and individual online access: £54//81/US$103 Please note: £ Sterling rates apply in UK, / rates apply in Europe, US$ elsewhere There may be other subscription rates available, for a complete listing please visit www.jss.oupjournals.org/ subinfo. Full prepayment, in the correct currency, is required for all orders. Orders are regarded as firm and payments are not refundable. Subscriptions are accepted and entered on a complete volume basis. Claims cannot be considered more than FOUR months after publication or date of order, whichever is later. All subscriptions in Canada are subject to GST. Subscriptions in the EU may be subject to European VAT. If registered, please supply details to avoid unnecessary charges. For subscriptions that include online versions, a proportion of the subscription price may be subject to UK VAT. Personal rate subscriptions are only available if payment is made by personal cheque or credit card and delivery is to a private address. The current year and two previous years’ issues are available from Oxford University Press. Previous volumes can be obtained from the Periodicals Service Company, 11 Main Street, Germantown, NY 12526, USA. Email: [email protected]. Tel: +1 (518) 537 4700. Fax: +1 (518) 537 5899. For further information, please contact: Journals Customer Service Department, Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, UK. Email: [email protected]. Tel (and answerphone outside normal working hours): +44 (0)1865 353907. Fax: + 44 (0)1865 353485. In the US, please contact: Journals Customer Service Department, Oxford University Press, 2001 Evans Road, Cary, NC 27513, USA. Email: [email protected]. Tel (and answerphone outside normal working hours): 800 852 7323 (toll-free in USA/Canada). Fax: 919 677 1714. In Japan, please contact: Journals Customer Services, Oxford University Press, 1-1-17-5F, Mukogaoka, Bunkyo-ku, Tokyo, 113-0023, Japan. Email: [email protected]. Tel: (03) 3813 1461. Fax: (03) 3818 1522. Methods of payment. (i) Cheque (payable to Oxford University Press, to Oxford University Press, Cashiers Office, Great Clarendon Street, Oxford OX2 6DP, UK) in GB£ Sterling (drawn on a UK bank), US$ Dollars (drawn on a US bank), or EU/ Euros. (ii) Bank transfer to Barclays Bank Plc, Oxford Group Office, Oxford (bank sort code 20-65-18) (UK), overseas only Swift code BARC GB 22 (GB£ Sterling to account no. 70299332, IBAN GB89BARC20651870299332; US$ Dollars to account no. 66014600, IBAN GB27BARC 20651866014600; EU/ Euros to account no. 78923655, IBAN GB16BARC20651878923655). (iii) Credit card (Mastercard, Visa, Switch or American Express). Journal of Semitic Studies is published two times a year in April and October by Oxford University Press, Oxford, UK. Annual subscription price is £166//249/US$315. Journal of Semitic Studies is distributed by Mercury International, 365 Blair Road, Avenel, NJ 07001, USA. Periodicals postage paid at Rahway, NJ and at additional entry points. US Postmaster: send address changes to Journal of Semitic Studies (ISSN 0022-4480), c/o Mercury International, 365 Blair Road, Avenel, NJ 07001, USA. PERMISSIONS For information on how to request permissions to reproduce articles/information from this journal, please visit www.oxfordjournals.org/permissions. ADVERTISING Inquiries about advertising should be sent to Helen Pearson, Oxford Journals Advertising, PO Box 347, Abingdon OX14 1GJ, UK. Email: [email protected]. Tel: +44 (0)1235 201904; Fax: +44 (0)8704 296864. DISCLAIMER Statements of fact and opinion in the articles in Journal of Semitic Studies are those of the respective authors and contributors and not of Journal of Semitic Studies or Oxford University Press. Neither Oxford University Press nor Journal of Semitic Studies make any representation, express or implied, in respect of the accuracy of the material in this journal and cannot accept any legal responsibility or liability for any errors or omissions that may be made. The reader should make his/her own evaluation as to the appropriateness or otherwise of any experimental technique described. © The University of Manchester 2008 All rights reserved; 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, or otherwise without prior written permission of the Publishers, or a licence permitting restricted copying issued in the UK by the Copyright Licensing Agency Ltd, 90 Tottenham Court Road, London W1P 9HE, or in the USA by the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923. Typeset by Drukkerij Peeters, Herent, Belgium Printed by Bell & Bain Ltd, Glasgow, UK

INFORMATION FOR AUTHORS Contributions, books for review, and other editorial communications should be addressed to: The Editors Journal of Semitic Studies Middle Eastern Studies School of Languages, Linguistics and Cultures University of Manchester Manchester M13 9PL UK Article Length Editorial policy is normally to favour shorter articles, i.e. no longer than 10,000 words. Typescript layout Articles and reviews should be typed on one side of the paper (A4 or 8.5 x 11 inches) with double spacing and generous margins on one side of the paper only. Authors of articles should submit TWO COPIES, and disk, if available. Non-Roman script is to be written as clearly as possible and used as sparingly as possible. Notes should be numbered consecutively, double spaced and placed at the end. The Editors will not accept for publication any typescripts which do not conform to these standards. For detailed instructions please see the style sheet. Abstracts Authors of articles are required to provide an abstract, which should be a single paragraph, not exceeding 150 words. Submission We will also normally require the final revised version to be supplied in hard copy and electronically. Whilst we can accept most wordprocessor formats the preferred option is Word. The files should be submitted in the wordprocessor's native format and as an RTF (Rich Text Format). Illustrations Electronic submission of illustrations. We can accept 3.5-inch floppy disks, CDROMs and email attachments. High-quality print-outs of all figures are still essential, as they will be scanned for reproduction in the event of problems with the electronic files. Figures must be saved as high-resolution TIF files (at least 300 pixels per inch for photographs, and 1200 pixels per inch for black and white line drawings). Offprints and Proofs Authors of articles will receive one copy of the issue containing their article and twenty-five offprints of their article free of charge. Alteration on the proofs, other than the correction of printer's errors, may be disallowed. Copyright and Permissions Manuscripts submitted will be expected to contain original work and should not have been published in abridged or other form elsewhere. It is a condition of publication in the Journal that authors grant an exclusive licence to The University of Manchester. Any requests from third parties to reproduce articles are handled on behalf of the journal by Oxford University Press. This will also allow the article to be as widely disseminated as possible and will protect the rights of the author and OUP. In granting an exclusive licence, authors may use their own material in other publications provided that the journal is acknowledged as the original authority for publication, and Oxford University Press is notified in writing and in advance.

92897_JOSS_2010/1_vw

08-03-2010

11:11

Pagina 1

JOURNAL OF

Semitic Studies V OLUME LV. NO. 1 SPRING 2010

EDITORIAL COMMITTEE P. S. ALEXANDER G. J. BROOKE R. BUCKLEY A. CHRISTMANN J. F. HEALEY P. C. SADGROVE

92897_JOSS_2010/1_vw

08-03-2010

11:11

Pagina 2

JOURNAL OF

Semitic Studies Editorial Committee P.S. Alexander G.J. Brooke R. Buckley A. Christmann J.F. Healey P.C. Sadgrove

Editorial Advisers C.E. Bosworth S.P. Brock K.J. Cathcart R. Gleave M. Kropp Y. Matras A.R. Millard M. Morgenstern M.E.J. Richardson P. Schäfer A. Shivtiel G. Rex Smith Y. Suleiman E. Ullendorff

The Journal of Semitic Studies publishes a book supplement series. For more information see www.jss.oxfordjournals.org. The editors welcome proposals for further titles in the series. For more details please write to the editors, Journal of Semitic Studies, Middle Eastern Studies, School of Languages, Linguistics and Cultures, University of Manchester, Manchester M13 9PL, UK

92897_JOSS_2010/1_vw

08-03-2010

11:11

Pagina 3

JOURNAL OF

Semitic Studies V OLUME LIV. NO. 1 SPRING 2010

Contents articles EDWARD LIPINSKI, ‘Le Gérondif en Phénicien’ ROBERT S. KAWASHIMA, ‘“Orphaned” Converted Tense Forms in Classical Biblical Hebrew Prose’ BAHAA AMER AL-JUBOURI, ‘Nouvelles inscriptions araméennes du temple de Nannay à Hatra’ GERRIT BOS, ‘Medical Terminology in the Hebrew Tradition: Shem Tov Ben Isaac, Sefer ha-Shimmush, Book 30’ AARON D. RUBIN, ‘The Development of the Amharic Definite Article and an Indonesian Parallel’ ABDULRAHMAN S. AL SALIMI, ‘Identifying the (Iba∂i/Omani) Siyar’ SAM LIEBHABER, ‘Rhythm and Beat: Re-evaluating Arabic Prosody in the Light of Mahri Oral Poetry’ SADOK MASLIYAH, ‘The Folk Songs of Iraqi Children: Part One’ JAMES DICKINS, ‘Basic Sentence Structure in Sudanese Arabic’

1 11 37 53 103 115 163 183 237

reviews Rainer VOIGT (ed.), From Beyond the Mediterranean: Akten des 7. Internationalen Semitohamitistenkongresses (VII. ISHaK), Berlin 13, bis 15, September 2004 (Aaron D. RUBIN) Elizabeth FROOD, Biographical Texts from Ramesside Egypt (Karl JANSENWINKELN) William M. SCHNIEDEWIND and Joel H. HUNT, A Primer on Ugaritic: Language, Culture and Literature (Wilfred G.E. WATSON) James W. WATTS, Ritual and Rhetoric in Leviticus: From Sacrifice to Scripture (Walter J. HOUSTON) Christophe NIHAN, From Priestly Torah to Pentateuch (Calum CARMICHAEL)

263 265 266 269 270

92897_JOSS_2010/1_vw

08-03-2010

11:11

Pagina 4

Jill MIDDLEMAS, The Templeless Age: An Introduction to the History, Literature and Theology of the ‘Exile’ (Ralph W. KLEIN) Archie T. WRIGHT, The Origin of Evil Spirits (Philip R. DAVIES) Kay PRAG, Excavations by K. M. Kenyon in Jerusalem 1961–1967: Volume V. Discoveries in Hellenistic to Ottoman Jerusalem (A. Asa EGER) Peter E. PORMANN (ed.), Rufus of Ephesus: On Melancholy (Oliver KAHL) P.S.F. VAN KEULEN and W.Th. VAN PEURSEN (eds), Corpus Linguistics and Textual History: A Computer-Assisted Interdisciplinary Approach to the Peshitta (Jerome A. LUND) Christa MÜLLER-KESSLER, Die Zauberschalentexte in der Hilprecht-Sammlung, Jena und weitere Nippur-Texte anderer Sammlungen (Matthew MORGENSTERN) Avi SAGI, The Open Canon: On the Meaning of Halakhic Discourse (Azzan YADIN) Shmuel SAFRAI lèz, Zeev SAFRAI, Joshua SCHWARTZ, and Peter J. TOMSON (eds), The Literature of the Sages, Second Part: Midrash and Targum, Liturgy, Poetry, Mysticism, Contracts, Inscriptions, Ancient Science and the Languages of Rabbinic Literature (Ephraim NISSAN) Timothy EDWARDS, Exegesis in the Targum of the Psalms: The Old, the New, and the Rewritten (David M. STEC) Mariano GÓMEZ ARANDA, Dos Comentarios de Abraham ibn Ezra al Libro de Ester: Edición crítica, traducción y estudio introductorio (Aaron D. RUBIN) Dvora BREGMAN, The Golden Way: The Hebrew Sonnet during the Renaissance and the Baroque (Arie SCHIPPERS) Leora BATNITZKY, Leo Strauss and Emmanual Levinas: Philosophy and the Politics of Revelation (A.H. LESSER) Khaleel MOHAMMED and Andrew RIPPIN (eds), Coming to terms with the Qur’an. A volume in honor of Professor Issa Boullata (Stefan WILD) John A. MORROW (ed.), Arabic, Islam and the Allah Lexicon: How Language Shapes our Conception of God (Gerhard BOWERING)

271 273 274 278

280 282 291

293 297 298 300 304 306 308

short notes Pekka LINDQVIST, Sin at Sinai: Early Judaism Encounters: Exodus 32 (Mila GINSBURSKAYA) Bernard M. LEVINSON, ‘The Right Chorale’: Studies in Biblical Law and Interpretation (Walter J. HOUSTON) James T. SPARKS, The Chronicler’s Genealogies: Towards an Understanding of 1 Chronicles 1–9 (Brian E. KELLY)

311 312 313

92897_JOSS_2010/1_vw

08-03-2010

11:11

Pagina 5

Fiona C. BLACK, The Recycled Bible: Autobiography, Culture, and the Space Between (Philip R. DAVIES) Andrew D. GROSS, Continuity and Innovation in the Aramaic Legal Tradition (Stephen A. KAUFMAN) Giancarlo TOLONI, La sofferenza del giusto: Giobbe e Tobia a confronto (George J. BROOKE) Mark J. BODA, Daniel K. FALK and Rodney A. WERLINE (eds), Seeking the Favor of God. Volume 1: The Origins of Penitential Prayer in Second Temple Judaism; Volume 3: The Impact of Penitential Prayer beyond Second Temple Judaism (George J. BROOKE) Melvin K.H. PETERS (ed.), XIII Congress of the International Organization for Septuagint and Cognate Studies: Ljubljana, 2007 (George J. BROOKE) Sidnie White CRAWFORD, Rewriting Scripture in Second Temple Times (Ariel FELDMAN) Lorenzo DITOMASSO, The Dead Sea New Jerusalem Text: Contents and Contexts (Charlotte HEMPEL) Antti LAATO and Jacques VAN RUITEN (eds), Rewritten Bible Reconsidered: Proceedings of the Conference in Karkku, Finland, August 24–26 2006 (George J. BROOKE) David T. RUNIA and Gregory E. STERLING (eds), The Studia Philonica Annual: Studies in Hellenistic Judaism Volume XX.2008 (George J. BROOKE) Frances FLANNERY, Colleen SHANTZ and Rodney A. WERLINE (eds), Experientia, Volume 1: Inquiry into Religious Experience in Early Judaism and Early Christianity (George J. BROOKE) Bart D. EHRMAN, Whose Word is It? The Story Behind Who Changed the New Testament and Why (George J. BROOKE) Julius Heinrich PETERMANN, The Great Treasure or Great Book, commonly called “The Book of Adam”, the Mandaeans’ work of Highest Authority (Matthew MORGENSTERN) Benjamin RICHLER, Hebrew Manuscripts in the Vatican Library Catalogue (Renate SMITHUIS) Seth S. SANDERS (ed.), Margins of Writing, Origins of Cultures (Alasdair LIVINGSTONE)

314 315 315

316 317 318 318

319 320

320 321

321 322 323

LE GÉRONDIF EN PHÉNICIEN EDWARD LIPINSKI UNIVERSITY OF LEUVEN

Abstract A single Latin manuscript, Berne codex 123 (folio 7a), preserves a statement by Isidorus Hispalensis who tells of twelve parts of speech in Phoenician, consisting of the usual eight with the addition of the article, the ‘impersonal mode’ (having no set person or number), the infinitive and the ‘gerund’. This paper proposes identifying the gerund with the infinitive followed by a personal pronoun, a construction attested in Phoenician inscriptions from Byblos, Zincirli, Karatepe and Çineköy, and occurring sporadically in Ugaritic, Hebrew and South-Arabian. It is argued that it represents an earlier stage or a variant of the Ethio-Semitic ‘gerund’ denoting an action simultaneous or anterior to the one expressed by another verb in the perfect or the imperfect. Its absence in Aramaic, North-Arabian and Classical Arabic once again shows the weakness of the hypothesis of a ‘Central Semitic’.

Isidore de Séville (c. 560–636), connu en latin sous le nom d’Isidorus Hispalensis, fut évêque de Séville pendant une trentaine d’années (c. 602–36), mais ses charges épiscopales ne l’empêchèrent point de devenir l’un de plus importants chaînons du savoir entre l’Antiquité classique et le Moyen-Âge. Isidore n’a pas l’habitude de mentionner ses sources, mais il est évident qu’il a recueilli ses informations dans nombre d’ouvrages d’auteurs antiques, dont certains sont irrémédiablement perdus. C’est à une telle source qu’il doit aussi sa connaissance des douze catégories grammaticales du phénicien, parmi lesquelles il mentionne l’infinitif suivi du gérondif.1 Isidorus dicit: Phoenices XII partes esse dixerunt, articulum nonam partem putauerunt, qui cum Graecis octaua pars est; inpersonalem 1 Son texte est préservé dans un seul manuscrit, le Codex Bernensis 123, fol. 7a: H. Hagen, ‘Anecdota Helvetica’, dans H. Keil, Grammatici Latini VIII, Leipzig 1870 (réimpr. Hildesheim 1961), CCLV–CCLVI. C’est un extrait, bien sûr, et nous n’avons pas la certitude qu’il reproduit le texte d’Isidore d’une manière exacte et complète.

1

LE GÉRONDIF EN PHÉNICIEN

modum decimam partem putauerunt, quia non habet certam personam et certum numerum, nisi addas pronomen ut ‘legitur a me, a te, a nobis’, et infinitiuum undecimam putauerunt, quia uim nominis habet, idest nominatiui, accusatiui quoque et ablatiui; gerendi modum duodecimam partem putauerunt, quia habent verba gerendi uim et formam nominum, ut ‘amandi, amando, amandum, amatum, amatu’, et ideo Phoenices dixerunt XII. «Isidore dit: Les Phéniciens ont dit qu’il y a douze parties (du discours). Ils considéraient que l’article est la neuvième partie, celle qui chez les Grecs est la huitième partie. Ils considéraient que le mode impersonnel (participe) est la dixième partie, qui n’exprime ni la personne ni le nombre, sauf si l’on ajoute un pronom, par exemple ‘lu par moi, par toi, par nous’. Ils considéraient que l’infinitif est l’onzième (partie), vu qu’il possède des fonctions du nom, à savoir le nominatif (sujet), l’accusatif (objet) et l’ablatif (adverbe). Ils considéraient que le gérondif est la douzième partie, vu que les verbes au gérondif possèdent des fonctions et une désinence à l’instar des noms, comme amandi, amando, amandum, amatum, amatu. C’est pourquoi les Phéniciens ont dit qu’il y avait douze parties (du discours)».

Il y a cinquante ans J.M. Solá-Solé a déjà attiré l’attention sur ce texte important,2 préservé dans un manuscrit de l’Ars de Clément l’Écossais, maître de grammaire à la Cour de Charlemagne. SoláSolé a cependant compris le gérondif des Phéniciens d’une manière très différente de celle qui sera exposée ci-dessous. Il y voyait l’infinitif introduit par une préposition, ce qui ne semble pas pouvoir correspondre à la distinction établie dans la source ultime d’Isidore. L’infinitif construit du phénicien avec préposition équivaut sans doute au gérondif latin, mais il est atemporel et impersonnel, c’està-dire qu’il ne peut guère être considéré comme une catégorie grammaticale distincte de l’infinitif. «Gérondif», c’est par contre le nom attribué par les éthiopisants à l’infinitif suivi du suffixe pronominal indiquant le sujet.3 Cette forme 2 J.M. Solá-Solé, ‘Sur les parties du discours en phénicien’, Bibliotheca Orientalis, 14 (1957), 66–8. 3 Le substantif verbal constituant la base du gérondif a été appelé «infinitif» par F. Praetorius, Grammatik der Tigriñasprache in Abessinien (Halle 1871), 336, et A. Dillmann, Grammatik der äthiopischen Sprache 2 (Leipzig 1899), 235, suivis par la plupart des chercheurs, notamment M. Cohen, Le système verbal sémitique et l’expression du temps (Paris 1924), 50–1, §26. Cf. aussi O. Kapeliuk, ‘Reflections on the Ethio-Semitic Gerund’, Proceedings of the 13th International Conference of Ethiopian Studies (Kyoto 1997), 492–8.

2


s’emploie en guèze,4 tigrigna,5 amharique,6 argobba7 et dans les dialectes occidentaux du gouragué.8 Elle exprime la circonstance sans détermination de moment et peut se traduire en français, suivant les phrases, soit par un présent, soit par un passé. Elle est utilisée le plus souvent pour signifier l’antériorité de l’action par rapport à l’action indiquée par le verbe principal. Ce nom verbal spécial, auquel on a donné à bon escient le nom de gérondif, est comparable à l’infinitif phénicien suivi du pronom personnel9 et exprimant, lui aussi, une circonstance antérieure à l’action signifiée par le verbe principal. Il y a une dizaine d’années j’ai donc appelé cette construction «gérondif»10 en suivant, je pense, l’exemple d’Isidore de Séville qui devait disposer d’une source latine se référant à la grammaire des textes littéraires puniques, spécialement de nature narrative.11 4 F. Praetorius, Aethiopische Grammatik (Karlsruhe-Leipzig 1886 [reprint, New York 1955]), 62, §77; A. Dillmann, op. cit. (n. 3), 235; M. Chaîne, Grammaire éthiopienne 2 (Beyrouth 1938), 35; Th.O. Lambdin, Introduction to Classical Ethiopic (Geëz) (HSM 24, Ann Arbor 1978), 135–6. 5 R.M. Voigt, Das tigrinische Verbalsystem (Marburger Studien zur Afrika- und Asienkunde A/10, [Berlin 1977]), 140–220. 6 E.G. Titov, Le gérondif dans la langue amharique, dans Afrikanskiy etnografitcheskiy sbornik III (Moskva-Leningrad 1959), 184–204 (en russe); Id., Quelques cas de l’emploi de l’ainsi dénommé gérondif et de la construction avec le gérondif dans la langue amharique, dans Semitskiye yaziki I, (Moskva 1963), 82–9 (en russe); J. Hartmann, Amharische Grammatik (Äthiopistische Forschungen 3, Wiesbaden 1980), 198–201, 369–70, 373; U. Maass, Das Gerundium im Amharischen und im Tigrinya. Eine vergleichende Funktionsanalyse, Dissert. (Univ. Leipzig 1990); W. Leslau, Reference Grammar of Amharic (Wiesbaden 1995), 355–90; U. Maass, ‘On Grammatical Functions of the Gerund in Amharic’, dans P. Zemánek (éd.), Studies in Near Eastern Languages and Literatures. Memorial Volume of Karel Petrácek (Praha 1996), 343–57. 7 W. Leslau, Ethiopic Documents: Argobba. Grammar and Dictionary (Wiesbaden 1997), 52–5. 8 W. Leslau, Gurage Studies: Collected Articles (Wiesbaden 1992), 135–6, 443– 58. 9 J.M. Solá-Solé, L’infinitif sémitique (Paris 1961), 110–18; F. Bron, Recherches sur les inscriptions phéniciennes de Karatepe (Genève 1979), 143–6; J. Friedrich et W. Röllig, Phönizisch-punische Grammatik 3 (Roma 1999), 192–3, §267. 10 E. Lipinski, Semitic Languages: Outline of a Comparative Grammar (OLA 80, Leuven 1997) (2e édition, 2001), §42.12. 11 L’existence de cette littérature est encore attestée au Ve siècle par S. Augustin. Cf. F. Vattioni, ‘Sant’Agostino e la civiltà punica’, Augustinianum 8 (1968), 434– 67. On trouvera une bibliographie plus complète dans E. Lipinski (éd.), Dictionnaire de la civilisation phénicienne et punique (Turnhout 1992), 49b.

3


L’usage du gérondif a connu un grand développement dans les langues éthio-sémitiques, celles du nord aussi bien que celles du sud, mais il faut se référer surtout à des exemples simples qui offrent les meilleurs parallèles grammaticaux aux phrases concernées des inscriptions de Byblos, de Zincirli, de Karatepe et de Çineköy. Le gérondif est marqué en guèze par une désinence -ä ajoutée à l’infinitif et suivie du pronom suffixe. Celui-ci désigne très souvent le même sujet que celui de la proposition principale, mais peut aussi se rapporter à un sujet différent. C’est le même sujet que l’on rencontre, par exemple, dans la traduction en guèze du passage de l’Évangile de S. Matthieu 2, 3 se référant à l’arrivée des mages de l’Orient. Nous y lisons: wä-sämiö Herod¢s dängä∂a, «et (l’)ayant appris, Hérode fut alarmé». Le gérondif sämiö est constitué de l’infinitif sämi¨ du verbe «entendre», suivi du -o qui provient du suffixe -ä-hu de la troisième personne. Le gérondif exprime en l’occurrence l’antériorité de l’information à la crainte éprouvée par Hérode. Mais la construction s’emploie aussi pour signifier la simultanéité et le gérondif peut suivre la proposition principale, comme dans le passage suivant de la Chronique du négus Zarä Ya¨qob (1434–68):12 konu }äyhudä Ìädigomu krest¢nnahomu, «Ils devinrent Juifs, abandonnant leur christianisme». Le gérondif Ìädigomu est constitué ici de l’infinitif Ìädig, suivi de -ä-homu, devenu -omu, le suffixe de la 3e pers. masc. plur. Le gérondif s’emploie de la même manière en tigrigna qui rattache le suffixe pronominal directement à l’infinitif, sans désinence vocalique, mais n’a pas conservé le h du suffixe de la troisième personne. Ainsi lit-on dans l’Histoire de l’Éthiopie de Yay¢nsät Gäbrä-}Egzi}ab¢Ìer, écrite en tigrigna:13 Mahdaw¢yan }¢ww¢n }¢nkab Sudan talaïlom n¢ Îtyoππya wåräuwa, «Mais les Mahdistes, ayant pénétré du Soudan, ont pillé l’Éthiopie». Le gérondif est constitué ici de l’infinitif talaïl à préformante ta-, «pénétrer», et du suffixe -om de la 3e pers. masc. plur. Deux gérondifs peuvent se suivre avant la proposition principale, comme dans une autre phrase du même livre:14 nab S¢mä†ru Ra}si }Alula l¢}ikom särawitom }aktitom nab kwinat gäÒom täbäggäsu, «Ayant envoyé Ras Alula à Sematru et ayant rassemblé son armée, il se rendit 12 J. Perruchon, Chronique de Zar}a Yâëqob et de Ba}eda Mâryâm (Bibliothèque de l’École des Hautes Études 93, Paris 1893), 97. 13 Yay¢nsät Gäbrä-}Egzi}ab¢Ìer, Tarik }Ityoππya (Asmara 1962), 109, ligne 9, cité par R.M. Voigt, op. cit. (n. 5), 150. 14 Yay¢nsät Gäbrä-}Egzi}ab¢Ìer, op. cit. (n. 13), 101, ligne 19, cité par R.M. Voigt, op. cit. (n. 5), 150.

4


au combat». Les deux gérondifs l¢}ikom et }aktitom sont constitués des infinitifs l¢}ik et }aktit suivis du suffixe pronominal -om de la 3e pers. masc. plur., car l’auteur utilise ici le pluriel de majesté pour désigner l’empereur de l’Éthiopie, tout comme il le fait dans la proposition principale dont le verbe täbäggäsu est au pluriel, ainsi que l’indique la finale -u. Le gérondif s’emploie de la même manière en amharique, argobba et gouragué. À la différence des langues éthio-sémitiques, le phénicien n’utilise pas comme sujet le suffixe pronominal, mais le pronom personnel indépendant. La construction et le sens sont, pour le reste, identiques. Comme il s’agit d’une proposition subordonnée, qui n’alterne pas avec des parfaits conjugués, mais en dépend, une traduction plus nuancée des inscriptions phéniciennes s’impose dans chaque cas. Malgré sa lourdeur, c’est souvent le participe qu’on pourra employer ici en français, mais on préférera, dans certains cas, une proposition subordonnée, introduite par «comme», «vu que». À Zincirli, dans l’inscription de Kulamuwa, on traduirait ainsi wskr }nk ¨ly mlk }sr,15 «et comme j’ai pris à solde contre lui le roi d’Assyrie, il donna une jeune fille pour une brebis et un homme pour un vêtement», c’est-à-dire: pour un prix dérisoire. Les inscriptions phéniciennes de Karatepe contiennent toute une série de gérondifs, dont l’usage multiple s’explique tout simplement par le fait que ce sont les seuls textes phéniciens comportant une section narrative de quelque ampleur. Il ne s’agit nullement d’une dialecte différent. Voici une traduction de Phu/A I, 3–1316 qui tient compte de l’emploi du gérondif: Baal a fait de moi un père et une mère pour les Danouniens, vu que j’ai fait vivre (yÌw }nk) les Danouniens, que j’ai élargi (yrÌb }nk) le pays de la plaine d’Adana du soleil levant au couchant. Et il y eut en mes jours toute sorte de bonheur pour les Danouniens, abondance et bienêtre, vu que j’ai rempli (ml} }nk) les greniers de Pahar. Ayant accumulé (p¨l }nk) cheval sur cheval, bouclier sur bouclier et armée sur armée, … j’ai brisé les rebelles, détruisant (trq }nk) tout le mal qui était dans le pays. Ayant établi (y†n} }nk) la maison de mon seigneur dans le bien, fait (p¨l }nk) du bien à la descendance de mon seigneur, l’ayant installée (ysb }nk) sur le trône de son père et fait (st }nk) la paix avec chaque roi, chaque roi m’a vraiment tenu pour un père à cause de ma justice, de ma sagesse et de la bonté de mon cœur. 15

KAI 24, 7–8; TSSI III, 13, 7–8. W. Röllig, ‘The Phoenician Inscriptions’, dans H. Çambel (éd.), Corpus of Hieroglyphic Luwian Inscriptions II. Karatepe-Aslanta≥ (Berlin 1999), 50–81 (voir p. 50); KAI 26, A, I, 3–13; TSSI III, 15, A, I, 3–13. 16

5


L’inscription de Çineköy comprend deux gérondifs employés avec un sujet différent de celui des propositions principales, comme c’est le cas, par exemple, dans la courte phrase en guèze näÒära qäwima du Synaxaire éthiopien,17 une collection de vies de saints pour chaque jour de l’année liturgique: «il (la) regarda, tandis qu’elle se tenait debout». Le gérondif est constitué ici de l’infinitif qäwim suivi de suffixe pronominal de la 3e pers. fém. sing. -a-ha, devenu -a, alors que le verbe principal est à la 3e pers. masc. sing. Les gérondifs de Çineköy apparaissent dans des passages très semblables à ceux du texte cité de Karatepe:18 Et comme j’ai accumulé (p¨l }nk) cheval sur cheval, armée sur armée, le roi d’Assyrie et toute la maison d’Assyrie devinrent pour moi comme un père et une mère … Et comme j’ai bâti (bn }nk) des forteresses, huit à l’est et sept à l’ouest, elles furent quinze.

Un passage narratif avec quelques gérondifs se retrouve dans l’inscription de Yehawmilk, roi de Byblos. Voici une traduction de ces lignes, tenant compte de tous les gérondifs:19 Comme j’ai invoqué (qr} }nk) ma maîtresse, la Dame de Byblos, elle a écouté ma voix. Comme j’ai fait (p¨l }nk) pour ma maîtresse, la dame de Byblos, cet autel d’airain …, comme je l’ai fait (p¨l }nk) … pour ma maîtresse, la Dame de Byblos, quand j’ai invoqué ma maîtresse, la Dame de Byblos, elle a écouté ma voix et m’a fait du bien.

Il est intéressant de relever ici le parallélisme entre les lignes 2–3, où l’on trouve le gérondif qr} }nk, et les lignes 7–8, où la conjonction de subordination km}s est suivie du parfait: km}s qr}t, «quand j’ai invoqué». L’emploi phénicien du pronom indépendant écarte la possibilité de l’usage d’un double suffixe avec l’infinitif, celui du sujet et celui de l’objet, comme en arabe.20 Seul le suffixe pronominal désignant

17 I. Guidi, Le Synaxaire éthiopien III. Mois de Nahasè et de Pâguemên, dans Patrologia Orientalis IX (Paris 1913), 237–476 (voir p. 379, ligne 3). Traduction du contexte par S. Grébaut: «Lorsque son mari Tsagâ-za-}Ab sortit du temple avec l’encensoir, il la regarda, tandis qu’elle se tenait debout, alors qu’elle était parée». 18 E. Lipinski, Itineraria Phoenicia (Studia Phoenicia XVIII; OLA 127, Leuven 2004), 127–8, lignes 5–12. L’editio princeps exige certaines mises au point: R. Tekoglu et A. Lemaire, ‘La bilingue royale louvito-phénicienne de Çineköy’, CRAI 2000, 961–1007. 19 KAI 10, 2–8; TSSI III, 25, 2–8. 20 W. Fischer, Grammatik des klassischen Arabisch (PLO, n.s. XI, Wiesbaden 1972), 127, 271, Anm. 1.

6


l’objet peut être attaché au verbe. L’emploi du pronom indépendant signifie aussi que l’on pourrait trouver un substantif à la place du pronom dans la même fonction de sujet de l’infinitif. Cette construction ne peut être identifiée dans les textes phéniciens actuellement connus, mais elle est attestée dans la poésie ugaritique. Ceci permet d’écarter des exemples anachroniques d’un statif utilisé comme parfait. On la rencontre également en hébreu biblique et un emploi comparable de l’infinitif se retrouve en sud-arabique.21 Bien qu’une étude ultérieure soit ici nécessaire, il est utile de présenter quelques exemples de cet usage. Il a certainement une base commune, mais connut un développement différent selon les régions. La parenté du sud-arabique avec l’éthio-sémitique suggère de présenter d’abord des exemples rencontrés en sabéen. L’infinitif absolu y est donc employé de manière analogue, mais il n’est pas accompagné d’un pronom suffixé ou indépendant, du moins dans l’état actuel de notre documentation, le sujet étant chaque fois le même que celui du verbe principal. L’emploi occasionnel de l’infinitif à désinence -n ne permet pas de douter que les formes en question sont bien des infinitifs, tandis que la séquence des verbes utilisés dans certaines phrases indique qu’ils n’expriment pas la suite des actions, mais signifient une circonstance antérieure à l’action exprimée par le verbe principal, tout comme le gérondif éthiopien. Ainsi, la phrase sabéenne b¨dw / whb¨ln / whrg / ws1by / wgnm / wmtlyn22 signifie: «ils emportèrent, ayant saisi, tué, capturé, dévasté, pillé». Il est évident que l’action de saisir, tuer, capturer et piller ne fait pas suite à l’enlèvement du butin, mais le précède. Les infinitifs en question expriment donc des circonstances antérieures à l’action du verbe principal b¨dw. Cependant, la même construction peut marquer la simultanéité: tnÌyt / wtn∂rn,23 «elle confessa faisant pénitence». Les infinitifs des exemples sabéens suivent le verbe au mode personnel, alors que le gérondif éthiopien précède souvent le verbe de la proposition principale.

21 A.F.L. Beeston, Sabaic Grammar (Manchester 1984), 22, §8:2. Cf. aussi M. Höfner, Altsüdarabische Grammatik (PLO 24, Leipzig 1943), 63–5, §54, avec plusieurs exemples, également en qatabanite et minéen. 22 A. Jamme, Sabaean Inscriptions from MaÌram Bilqis (Publications of the American Foundation for the Study of Man 3, Baltimore 1962), no 631, ligne 8, citée par A.F.L. Beeston, loc. cit. 23 CIS IV, 532, 2; K. Conti Rossini, Chrestomathia Arabica Meridionalis Epigraphica (Roma 1931), 54, no 46, ligne 2.

7


Ceci semble être aussi le cas de certains infinitifs ugaritiques24. Ainsi ngs }ank et ¨dbnn }ank précèdent le statif Ìt}u dans KTU 1.6, II, 21–3: Comme j’ai assailli le Très Puissant Baal, comme j’en ai fait un agneau dans ma bouche, c’est dans l’orifice de ma gueule qu’il est englouti.

Le pronom n’est pas employé si le sujet de l’infinitif absolu est le même que celui du verbe principal, comme dans le cas de qm et de ndd en KTU 1.3, I, 2–11, où il précède des prétérits yíqtul: Pardamenni, valet du Très Puissant Baal, laboureur 25 du Prince, seigneur de la terre, s’étant levé (qm), prépara le dîner 26 et lui servit à manger. Il découpa une poitrine devant lui avec un couteau salé, un morceau de (bête) grasse. S’affairant (ndd )27, il se donna de la peine28 et l’abreuva, il mit une coupe dans sa main, un hanap dans ses deux mains.

L’emploi de l’infinitif absolu en début de phrase ou après des formes verbales personnelles se rencontre aussi en hébreu biblique très tardif.29 On a pu citer un exemple en Dan. 9:5, où quatre parfaits «nous avons péché, nous avons commis l’iniquité, nous avons fait le mal, nous avons trahi», sont suivis de l’infinitif absolu sor. Celui-ci ne constitue pas la suite de l’énumération, mais signale en quoi ces fautes ont consisté: «en nous détournant (sor) de tes commandements et ordonnances».

24

Cf. J. Tropper, Ugaritische Grammatik (AOAT 273, Münster 2000), 492–3, §73.531. 25 Le nom s}id doit se comprendre à la lumière de s3}d en sabéen: «superficie cultivée», «terre labourée», «labour». Cf. A.F.L. Beeston, M.A. Ghul, W.W. Müller et J. Ryckmans, Dictionnaire sabéen (anglais-français-arabe) (Louvain-la-Neuve — Beyrouth 1982), 137. 26 On rapprochera yt¨r de l’hébreu mishnaïque s¨r, «préparer le dîner», et du guèze särää, avec métathèse. 27 Le verbe ndd correspond à l’hébreu et à l’araméen ndd, «être affairé», «s’enfuir». 28 Il convient de rapprocher le verbe y¨sr de l’arabe äsura, «être pénible», d’où «se donner de la peine». 29 W. Gesenius et E. Kautzsch, Hebräische Grammatik 28 (Leipzig 1909), 359– 60, §113z. Voir aussi P. Joüon, Grammaire de l’hébreu biblique 3 (Rome 1965), 356–8, §123u–y; P. Joüon et T. Muraoka, A Grammar of Biblical Hebrew II (Roma 1996), 429–32, §123u–y. 8


Deux autres exemples de cette construction se rencontrent en Qoh. 4, 2 et Esth. 9:1.30 La forme verbale indique qu’il ne peut s’agir du parfait, bien que le contexte montre qu’elle exprime un résultat acquis. La phrase n’est pas subordonnée à une proposition principale et ne peut donc être analysée comme les exemples attestés par les inscriptions phéniciennes des 8e–5e siècles av.n.è. C’est de nouveau l’éthiopien qui fournit la clef de la solution, plus exactement le tigrigna et le dialecte amharique de la province de Godjam,31 où le gérondif peut recevoir pleine indépendance et signifier à lui seul le résultat d’une action. Par exemple, en tigrigna, }eskä gälä do sämi¨ki,32 «Dis, as-tu appris quelque chose (et le sais-tu)?». L’infinitif sämi¨, «entendre», est suivi ici du suffixe pronominal de la 2e pers. fém. sing., suivant les règles du gérondif. Dans Qoh. 4:2, suite à son expérience désabusée, l’Ecclésiaste déclare: w¢-sabbe aÌ } anî …, «Et j’ai estimé (et j’estime) les morts qui sont déjà morts».33 Le cas d’Esth. 9:1 est semblable, sauf que le sujet de l’infinitif est un pronom de la 3e pers. masc. sing., détail important, puisque tous les exemples phéniciens actuellement connus sont attestés avec le pronom de la 1re pers. sing. Esth. 9:1 se réfère à la situation des Juifs persécutés: w¢-nah apok hu} } aser, «et elle s’est retournée de manière que …». La situation s’est renversée et est restée telle quelle, les Juifs ayant écrasé leurs ennemis. Il existe d’autres cas où le sujet de l’infinitif n’est pas pronominal, comme dans Esth. 3:13, où l’infinitif absolu est indépendant. Cet usage de l’infinitif fonctionnant à lui seul marque un développement qui n’est certainement pas influencé par l’emploi du gérondif en tigrigna et dans le dialecte de godjam. Il suppose l’existence antérieure, en hébreu, d’un gérondif limité à une fonction subordonnée. Peut-être le trouve-t-on en Lév. 6:7; Nomb. 15:35 et

30 R. Lehmann, ‘Who needs Phoenician?’, M. Witte – J.F. Diehl (éds), Israeliten und Phönizier (OBO 235, Fribourg-Göttingen 2008), 1–37, en particulier p. 27–31, propose d’ajouter à ces textes Gen. 31:5 et Is. 48:13, en y lisant l’infinitif absolu à la place du participe. La structure syntaxique des phrases en question n’est cependant pas la même. 31 D’après D.L. Appleyard, ‘New Finds in the 20th Century: The South Semitic Languages’, IOS 20 (2002), 401–30, en particulier p. 421–2, le développement du gérondif dans ces deux régions serait indépendant, mais elles ne sont pas très éloignées l’une de l’autre et il faut tenir compte de déplacements possibles de populations. 32 Mark Twäyn, N¢}¢s¢nnät Tom Såwyär (Asmara 1967), 181, ligne 15. 33 A. Schoors, The Preacher Sought to Find Pleasing Words I (OLA 41, Leuven 1991), 91, 170, discerne ici an infinitif absolu utilisé à la place d’une forme conjuguée, sans noter quelque particularité.

9


Deut. 15:2.34 Une étude systématique des cas semblables aboutira peut-être à des distinctions chronologiques, mais elle exigera une mise au point de nombreuses traductions qui ne tiennent pas compte des différences grammaticales. Ces exemples montrent que le texte narratif des anciennes compositions ouest-sémitiques35 n’était pas composé de phrases simplement juxtaposées. Les éléments du récit étaient combinés entre eux de manière à exprimer une pensée cohérente. Les formes verbales et les conjonctions de subordination servaient à créer une unité claire et distincte, tout comme l’usage occasionnel de l’infinitif absolu. Les langues du groupe cananéen, l’ugaritique, le sud-arabique et l’éthio-sémitique attestent ainsi l’usage d’un gérondif. En revanche, je n’ai pas trouvé les traces d’un emploi comparable de l’infinitif en araméen, en nord-arabique et en arabe classique. Ceci montre une nouvelle fois, soit dit en passant, que la théorie d’un «sémitique central» ne résiste pas face aux faits grammaticaux, aussi bien que phonologiques. Address for correspondence: [email protected] Ave. Adolphe Lacomblé 50/11, 1030-Brussels, Belgium

34 Les cas de l’infinitif absolu avec sujet nominal en Gen. 17:10; Ex. 12:48; Ps. 17:5; Prov. 17:12; Job 40:2, signalés par W. Gesenius et E. Kautzsch, op. cit, (n. 29), 361, §113gg, paraissent moins clairs. 35 L’emploi du terme «gérondif» dans une grammaire de l’akkadien (G. Buccellati, A Structural Grammar of Babylonian [Wiesbaden 1996], 481, §86.3) ne se réfère pas à un phénomène grammatical comparable.

10

‘ORPHANED’ CONVERTED TENSE FORMS IN CLASSICAL BIBLICAL HEBREW PROSE ROBERT S. KAWASHIMA UNIVERSITY OF FLORIDA

Abstract From Saussure’s distinction between diachronic and synchronic linguistics, it follows that Classical Biblical Hebrew wayyiqtol and weqatal are, grammatically speaking, ‘converted’ tenses. The converted tenses, as such, are not mere verbs, but verb phrases, consisting of a conjunction (conversive waw) and a verb. But verb phrases can be syntactically manipulated: the converted tenses might be realized in alternate forms. In fact, I will analyse numerous anomalous occurrences of both qatal and yiqtol as instances of tense ‘conversion’, in which conversive waw has been replaced by a syntactic equivalent or displaced by a syntactic insertion, and is thus separated from the converted verb — a phenomenon I refer to as ‘orphaned’ converted tense forms. Most notably, I will account for }az yiqtol in this manner. The adverbial }az has replaced waw, so that the converted imperfect appears without the conjunction, but with the same tense value as wayyiqtol.

Introduction A century ago, Ferdinand de Saussure laid the foundation for that movement eventually known as structuralism by setting forth in systematic fashion the concepts, axioms and principles of what he called ‘general linguistics’.1 As Emile Benveniste, a key figure in French structuralism, recognized already in 1954, one of the crucial innovations he bequeathed to linguistics was his distinction between the synchronic and diachronic dimensions of language: ‘The novelty of the Saussurian point of view … was to realize that language in itself 1 Saussure, Course in General Linguistics, trans. Roy Harris (La Salle 1986); hereafter cited as Course. First published posthumously in 1916, Course is based on students’ class notes dating as far back as 1906. For incisive commentary on Saussure’s thought, see Jean-Claude Milner, For the Love of Language, trans. Ann Banfield (New York 1990), 81–97; idem, Le Périple Structural (Paris 2002), 15–43.

11

‘ORPHANED’ CONVERTED TENSE FORMS

does not admit of any historical dimension, that it consists of synchrony and structure, and that it only functions by virtue of its symbolic nature’.2 Thus, according to Saussure: synchronic linguistics ‘will be concerned with logical and psychological connexions between coexisting items constituting a system, as perceived by the same collective consciousness’; and diachronic linguistics ‘will be concerned with connexions between sequences of items not perceived by the same collective consciousness, which replace one another without themselves constituting a system’ (Course, 98; see the general discussion in 79–98).3 It is in reference to Saussure’s programmatic remarks that one should understand Benveniste’s analysis of the tense system in French: ‘The question is, then, to look in a synchronic view of the verbal system in modern French for the correlations that organize the various tense forms’.4 He specifically analyses an apparent ‘defect’ (redundancy) in the language, namely, the coexistence in French of two preterites, the passé simple (aorist) and passé composé (perfect). ‘According to the traditional interpretation’, he notes, ‘these are two variants of the same form … and their co-existence points to a transitional phase in which the early form (il fit) is retained in the written language, which is more conservative, while the spoken language anticipates the substitution of the competing form (il a fait) which is destined to become the only one’ (205–6). This strictly diachronic explanation, which ‘reduc[es] the phenomenon to the terms of a successive development’ (206), he rejects 2

Benveniste, ‘Recent Trends in General Linguistics’, in Problems in General Linguistics, trans. Mary Elizabeth Meek (Coral Gables 1971), 4. 3 Saussure provides an apt illustration of this principle, viz., the mutual irrelevance of diachronic facts and synchronic facts: ‘The Latin word crispus (‘wavy, curly’) supplied French with a stem crép-, on which are based the verbs crépir (‘to rough-render’) and décrépir (‘to strip the plaster from’). Then French at a certain stage borrowed from Latin the word decrepitus (‘worn by age’). This became in French décrépit, and its etymology was forgotten. Nowadays, it is certain that most speakers connect un mur décrépi (‘a dilapidated wall’) and un homme décrépit (‘a decrepit man’), although historically the two words have nothing to do with each other. People often speak of the façade décrépite [‘decrepit façade’–RSK] of a house. That is a static fact, because it involves a relationship between two terms coexisting in the language. But in order to bring it about, certain evolutionary changes had to coincide. The original crisp- had to come to be pronounced crép-, and at the right moment a new word had to be borrowed from Latin. These diachronic facts, it is clear, have no connexion with the static fact which they brought about. They are of quite a different order’ (Course, 83). 4 Benveniste, ‘The Correlations of Tense in the French Verb’, in Problems in General Linguistics, 205. 12


in favour of his now famous synchronic distinction between two linguistic ‘systems’: narration (histoire) and discourse (discours). The passé simple, on the one hand, ‘today reserved to the written language, characterizes the narration of past events…. Events that took place at a certain moment of time are presented without any intervention of the speaker in the narration’ (206). The passé composé, on the other hand, ‘creates a living connection between the past event and the present in which its evocation takes place…. Like the present, the perfect belongs to the linguistic system of discourse, for the temporal location of the perfect is the moment of the discourse while the location of the aorist is the moment of the event’ (210). In other words, there exists in French a synchronic opposition between two modes of tense: discourse, a subjective, deictic system centred on the speaker (‘I’); narration, an objective, non-deictic, ‘unspeakable’ system.5 In fact, as I have argued elsewhere, the same grammatical distinction also exists in Classical Biblical Hebrew (CBH) — which will prove relevant further on.6 To summarize in simplified form: wayyiqtol is a non-deictic preterite, locating events without reference to a speaker, and thus occurs primarily (though not exclusively) in narration (histoire); qatal is a deictic preterite, centred on the linguistically identifiable here and now of the speaker, and thus occurs primarily (though not exclusively) in direct discourse (discours). Saussure, however, had little to say about syntax. For syntax — the concatenation of signs — is a function of parole (language in use), precisely what is excluded from the structural analysis of langue (language as an abstract system of elements).7 This study will therefore proceed within the research program of generative linguistics, which has, as Jean-Claude Milner observes, ‘refuted’ the structuralist ‘empirical doctrine of language’.8 The stated aim of this linguistics, 5 See Ann Banfield, Unspeakable Sentences (Boston 1982), 140–80. In pronouncing the passé simple ‘unspeakable’, Banfield means to indicate that native speakers of French judge it ‘unacceptable’ in the spoken language. 6 See Robert S. Kawashima, Biblical Narrative and the Death of the Rhapsode (Bloomington 2004), 35–76. 7 At most, structuralist linguistics classifies surface structures rather than analysing syntax as such (Banfield, Unspeakable Sentences, 1–21). 8 Milner, Introduction à une science du langage (Paris 1989), 642; see also 36–40, 63–6, 142–4, 639–42. Conversely, text-linguistics has claimed to move beyond Chomsky’s sentence-based paradigm, by analysing larger linguistic structures, i.e., ‘texts’ or ‘discourse’ — an interesting empirical claim which I cannot deal with here. What is clear, however, is that, generally speaking, text-linguistics continues to operate within the limited aims of the structuralist enterprise, viz., the classification of linguistic surface structures.

13


inaugurated by Noam Chomsky, is to represent formally what he calls ‘linguistic competence’: the knowledge native speakers possess that enables them to generate grammatical sentences and reject ungrammatical ones.9 This competence consists of two separate components: a dictionary and a grammar. The ‘mental dictionary’, which Joseph Emonds describes as ‘the faculty of human linguistic memory and culture’, contains those brute facts speakers must memorize — most prominently, words.10 The ‘mental grammar’ is the set of internalized rules — syntactical, morphological, phonological — by which the native speaker can, at least in principle, generate all and only the sentences of a language. Thus, every linguistic realization — word, phrase, sentence, etc. — is either contained in the dictionary or generated by the grammar out of elements from the dictionary. Chomskyan, no less than Saussurean, linguistics studies language as a synchronic object. For neither the grammar nor the dictionary makes reference to diachronic derivation, which is to say, a native speaker’s linguistic competence does not entail knowledge of the history of the mother tongue — such knowledge may, but need not be exploited. If it is nonetheless true that everything in a language is the result of linguistic evolution — what else could it be? — the real question is, then, whether a given linguistic element has been incorporated within the synchronic system, or whether it can only be understood in terms of diachronic linguistic evolution. Within the generative model, this question amounts to whether a given linguistic form, construction, etc. pertains to the dictionary or to the output of the grammar. For if the grammar accounts for everything that is synchronically generated by rules, the dictionary, as a list of memorized facts, accounts for that part of language which has simply been 9 See, e.g., Chomsky’s foundational text, Syntactic Structures (The Hague 1957). He gives his most accessible account of generative linguistics in Language and Mind 3 (Cambridge 2006). N.B., I am not concerned here with a particular version of his theory — extended standard, government and binding, minimalist — but with the overarching research program of generative grammar which can be said to join these various theories: see Chomsky, New Horizons in the Study of Language and Mind (Cambridge 2000), 3–18. 10 Emonds, Lexicon and Grammar (Berlin 2000), vii. The distinction between components corresponds to the process of native language acquisition. Grammatical rules are never encountered as such. Rather, native speakers (primarily as children) encounter an extremely poor set of data (actual speech), and from these acquire those highly abstract rules capable of generating the infinite number of possible grammatical strings for a given language. Since the entries in the dictionary — words, e.g. — cannot be so generated — the sign is ‘arbitrary’, as Saussure observed — each must be encountered as such and then memorized.

14


transmitted, such as it is, from the past. In order to capture the nongenerative nature of the dictionary, some linguists have adopted the term ‘listeme’ — on analogy with ‘morpheme’, ‘phoneme’, etc. — which Steven Pinker glosses as ‘an element of language [morphemes, word roots, irregular forms, collocations, and idioms] that must be memorized because its sound or meaning does not conform to some general rule’.11 In such cases, as Saussure himself admitted, a linguistic element ‘can be explained only in historical terms, by appeal to relative chronology’ (Course, 96).12 The study of CBH, one foot still planted in nineteenth-century philology (diachronic linguistics), has been slow to take full cognizance of the synchronic nature of language.13 Witness the sense of embarrassment frequently surrounding the very concept of ‘converted’ tenses, whose origin in traditional (pre-modern) grammar makes it seem hopelessly unscientific, especially in light of the discovery that the converted imperfect derives from an archaic preterite.14 Thus, Jacob Weingreen, citing G.R. Driver, proclaims: ‘All attempts to explain this at first strange phenomenon, whereby two tenses apparently exchange functions, on logical grounds, have failed, but the historical development of the Hebrew language readily accounts for it…. [T]here are two different systems, drawn from different sources, merged in the Hebrew scheme of tenses’.15 What concerns me here is not the historical thesis itself, which I have omitted, but 11

Pinker, Words and Rules (New York 1999), 292. The term was coined in Anne Marie Di Sciullo and Edwin Williams, On the Definition of Word (Cambridge 1987). 12 Saussure’s admission in no way violates — or ‘complicates’ or ‘nuances’ — the distinction between the diachronic and synchronic dimensions of language. Rather, listemes should be conceived of as tracing diachronic arcs through time, intersecting the synchronic state (plane) of a language at a geometrical point without synchronic extension. 13 There are exceptions. See Galia Hatav, ‘Anchoring world and time in biblical Hebrew’, Journal of Linguistics 40 (2004), 491–526, esp. 492–4 and references there. That field variously known as ‘text-linguistics’, ‘discourse analysis’, etc. is an indirect heir of Saussurean structuralism and, as such, synchronic: see, e.g., Robert D. Bergen (ed.), Biblical Hebrew and Discourse Linguistics (Dallas 1994); Walter R. Bodine (ed.), Discourse Analysis of Biblical Literature (Atlanta 1995); and David Allan Dawson, Text-Linguistics and Biblical Hebrew (Sheffield 1994). 14 For an historical survey of studies of the CBH verbal system, including premodern views, see Leslie McFall, The Enigma of the Hebrew Verbal System (Sheffield 1982). The now standard designation ‘consecutive tenses’ seems to function as a fig leaf, covering up the shame of the medieval notion of tense ‘conversion.’ 15 Weingreen, A Practical Grammar for Classical Hebrew2 (Oxford 1959), 252– 3; see also 90–1. 15


the willingness to freely replace a synchronic explanation ‘on logical grounds’ with a diachronic one based on the ‘historical development’ of CBH. Such Hebraists, in Saussure’s words, have yet ‘to distinguish clearly between states and sequences’ (Course, 98). In fact, while the converted imperfect (wayyiqtol ) descends, as is well known, from the archaic preterite (*wa + yaqtul ), the converted perfect (weqatal ) followed a more complicated path.16 It apparently began as the resultative (*wa + qatala) found in the apodosis of conditional clauses. Only later, on analogy with wayyiqtol, did the converted perfect assume the various tense-aspect values of the imperfect — qatal: wayyiqtol:: yiqtol: weqatal.17 In other words, the converted tenses result from the ‘synchronic reinterpretation’ of previously established linguistic forms.18 As undeniably important as the discoveries of diachronic linguistics have been, they do not change the fact that wayyiqtol and weqatal were, in the eyes of native speakers, converted tenses. Various morphological anomalies — wayyiben, wattasar, instead of *wayyibneh, *wattasîr — would not have presented themselves to native speakers as linguistic problems requiring a diachronic explanation — that they are vestiges of *yaqtul — but would have seemed as harmless as an irregular verb in English — ‘ran’ not ‘runned’. From the synchronic organization of the converted tenses follow several crucial but largely unrecognized linguistic consequences. Inasmuch as grammar conceives of wayyiqtol and weqatal as the converted imperfect and converted perfect, respectively, they are not actual 16

In this discussion, I summarize Ronald S. Hendel, ‘In the Margins of the Hebrew Verbal System: Situation, Tense, Aspect, Mood’, ZAH 9 (1996), 153 n.5. See also G. Bergsträsser, Hebräische Grammatik (2 vols, Hildesheim 1962 [1929]), II, §3 (hereafter Bergsträsser); T.L. Fenton, ‘The Hebrew “Tenses” in the Light of Ugaritic’, in Proceedings of the Fifth World Congress of Jewish Studies (5 vols, Jerusalem 1969), IV, 31–9. On the origins of wayyiqtol, see Mark S. Smith, The Origins and Development of the Waw-Consecutive (Atlanta 1991), 1–15; Anson F. Rainey, ‘The Ancient Hebrew Prefix Conjugation in the Light of Amarnah Canaanite’, HS 27 (1986), 4–19; idem, ‘The Prefix Conjugation Patterns of Early Northwest Semitic’, in Tzvi Abusch, John Huehnergard and Piotr Steinkeller (eds), Lingering over Words: Studies in Ancient Near Eastern Literature in Honor of William L. Moran (Atlanta 1990), 407–20. On the origins of weqatal, see William L. Moran, ‘The Hebrew Language in its Northwest Semitic Background’, in G. Ernest Wright (ed.), The Bible and the Ancient Near East: Essays in Honor of William Foxwell Albright (Garden City 1965 [1961]), 74–5; Bruce K. Waltke and M. O’Connor, An Introduction to Biblical Hebrew Syntax (Winona Lake 1990), §32.1.2. 17 See Saussure’s relevant discussion of linguistic analogy in Course, 160–71. 18 Hendel, ‘Hebrew Verbal System’; see also Angel Sáenz-Badillos, A History of the Hebrew Language, trans. John Elwolde (Cambridge 1993), 69. 16


verbs, but verb phrases consisting of a conjunction (the so-called conversive waw) and a verb: wayyiqtol = waw + converted yiqtol; weqatal = waw + converted qatal.19 Which is to say that the grammatical concept of converted tense is independent of, cannot be reduced to, its paradigmatic forms, wayyiqtol and weqatal. On the contrary, since phrases, unlike words, are subject to internal syntactic manipulations, it follows that the converted tenses can, in principle, be realized in alternate forms. In fact, a number of anomalous occurrences of both qatal and yiqtol constitute instances of tense ‘conversion’, in which conversive waw is replaced by a syntactic equivalent or displaced by a syntactic insertion, and is thus separated from the converted verb — a phenomenon I refer to as ‘orphaned’ converted tense forms. Most notably, I will account for }az yiqtol — but not †erem yiqtol — in this manner.20 The adverbial }az has replaced waw, so that the converted imperfect appears without the conjunction, but with the same tense value as wayyiqtol. The conditions of replacement and displacement indicate that I am not merely providing an ad hoc explanation for grammatical difficulties.21 I am not the first to identify orphaned converted tense forms. Beat Zuber, in particular, anticipates this study in important respects. Instead of providing a properly syntactic explanation, however, he theorizes that a certain power (Umkehrfunktion) resides in various particles (Tempusverhalten verschiedener Partikel ).22 Several generations earlier, a few scholars recognized various instances of these converted tenses, as well — though, sadly, their insight seems to have been largely forgotten by subsequent scholarship.23 Their examples, 19

One might object that wayyiqtol and weqatal are written as single words — at least when word breaks are represented. Typographical word divisions, however, do not necessarily coincide with underlying (abstract) linguistic word divisions: see Steven Pinker, Language Instinct (New York 1995), 126–57. 20 While scholars have often treated }az yiqtol and †erem yiqtol as a matched pair, there is no a priori reason for doing so: see Bergsträsser II, §7c,g; Galia Hatav, ‘The Modal Nature of [†erem] in Biblical Hebrew’, HS 47 (2006), 23–47; and Jan Joosten, ‘The Long Form of the Prefix Conjugation Referring to the Past in Biblical Hebrew Prose’, HS 40 (1999), 18–19. 21 I thus reject the negative view which maintains that tense forms in CBH freely change tense value: see Alexander Sperber, Hebrew Grammar: A New Approach (New York 1943), 181–207. His discussion, however, is valuable for identifying a number of interesting, because anomalous, occurrences of various tenses (192–8). See also Ziony Zevit’s analysis of ‘non-past yaqtul’ in ‘Talking Funny in Biblical Henglish and Solving a Problem of the Yaqtul Past Tense’, HS 29 (1988), 25–33. 22 Zuber, Das Tempussystem des biblischen Hebräisch (Berlin 1985), 156–84. 23 See Julius Friedrich Böttcher, Ausführliches Lehrbuch der Hebräischen Sprache, 2 vols (Leipzig 1866–8), 2.205; Heinrich Ewald, Ausführliches Lehrbuch der hebräischen 17


however, while interesting, are mostly prosodic, which raises problems beyond the scope of this study.24 Conversely, no less a scholar than S.R. Driver dismissed this theory, insisting that without conversive waw, the appropriate unconverted tense is invariably employed.25 Driver, however, betrays here a lack of linguistic clarity all too common in the field. In effect, he takes linguistic surface structure as the sole fact to be accounted for, a false premise common to traditional grammar as well as structuralist methods. Thus, all occurrences of the surface forms yiqtol and qatal are analysed as the imperfect or perfect, respectively. If a particular instance of yiqtol and qatal defies the standard assignment of tense value — and such anomalies are legion — a suitable semantic nuance is invented in ad hoc fashion in order to rescue the surface form and restore grammatical order.26 Generative linguistics, however, has provided powerful arguments for positing beneath the linguistic surface an abstract grammatical structure — I do not even refer here to ‘deep structure’. An adequate description of a linguistic surface structure must capture these underlying relations. Consider the two uses of the present tense form in ‘It is [now] cloudy’ and ‘Pi is [*now] equivalent to the ratio of a circle’s circumference to its diameter’. The former is the present tense proper (co-temporal with the moment of speech), and so it can co-occur with the deictic Sprache des alten Bundes 8 (Göttingen 1870), §346b; and Ferdinand Hitzig, Der Prophet Jeremia 2, (Leipzig 1866), 334 (Jer 44:22); idem, Das Buch Hiob (Leipzig 1874), 23 (Job 3:25), 39 (Job 5:19–20). I also draw from S.R. Driver’s summary of Hitzig’s relevant comments, apparently scattered in additional sources, in A Treatise on the Use of the Tenses in Hebrew and Some Other Syntactical Questions 3 (Oxford 1892), §§85 Obs, 109 Obs. 24 In general, the evidence for this study comes from the prose portions of Genesis-Kings. 25 Driver, Tenses, §§85 Obs, 109 Obs. 26 The contingence of its encounter with the real of language distinguishes linguistics proper — i.e., modern, scientific — in its various forms, from traditional grammar and certain structuralist methods. Thus, traditional grammar, as Milner observes, ‘constructs an image’ of language (pertaining to Lacan’s order of the ‘imaginary’), one that is ‘total’, since nothing need escape grammatical description (For the Love of Language, 75). And various structuralist methods, as Banfield points out, ‘even when they pretend to be empirical … find their data too easily, the notion of argument being absent’ (‘Preface to Phrases sans parole’, trans. Thelma Sowley, in Robert S. Kawashima, Gilles Philippe and Thelma Sowley (eds), Phantom Sentences: Essays in Linguistics and Literature Presented to Ann Banfield [Bern 2008], 463). In marked contrast, the contingence of the linguistic discovery, as part of ‘the order of science’, requires that linguistic knowledge take the form of ‘fragments’ (Milner, For the Love of Language, 76). 18


adverbial ‘now’. The latter is the ‘generic’ present (related to no particular moment), and so it cannot. The hypothetical insertion of ‘now’, a type of syntactic experiment, precludes an explanation of these two sentences based on mere interpretive sleight of hand (semantic nuance). Rather, one must posit beneath a single surface form the abstract but ‘real’ existence of two distinct tenses. Just so, one cannot simply dismiss my theory for positing beneath the surface forms yiqtol and qatal a second pair of tense values. In fact, my theory has considerable explanatory power, accounting for a multitude of grammatical anomalies, while economically replacing a cumbersome set of ad hoc semantic nuances with a single grammatical principle, firmly established on independent grounds, viz., tense conversion. On }az yiqtol as an ‘orphaned’ converted imperfect The most common form of the orphaned converted imperfect is }az yiqtol. In order to account for the surprising fact that the prefix conjugation following }az (‘then’) often refers to past events, many scholars resort to a purely historical explanation: it is a vestigial occurrence of the archaic preterite *yaqtul.27 It must immediately be observed that this diachronic hypothesis is premised on a widespread but largely repressed synchronic perception that, in terms of tense value, }az yiqtol is equivalent to *yaqtul, which in CBH means wayyiqtol.28 As 27 See Bergsträsser II, §7g; Edward L. Greenstein, ‘On the Prefixed Preterite in Biblical Hebrew’, HS 29 (1988), 8; Joosten, ‘Prefix Conjugation’, 22 n.17, 24–5; Alviero Niccacci, Syntax of the Verb in Classical Hebrew Prose (Sheffield 1990), 194. By comparing CBH with Arabic, Waltke and O’Connor effectively offers a diachronic explanation of }az yiqtol (An Introduction, §31.6.3a). Recall that Saussure’s ‘diachronic’ linguistics was originally called ‘comparative’, because whether it compared successive states of a language in order to trace linguistic change, or parallel linguistic facts from different languages in order to posit a common predecessor (such as Indo-European or proto-Semitic), the comparison served the diachronic analysis of language (see Milner, Périple structural, 19–20). 28 For this reason, †erem yiqtol and }az yiqtol require two separate explanations. In the former, yiqtol retains the normal tense value of the (unconverted) imperfect. For as Hendel observes, †erem (‘not yet, before’) ‘requires a relative future verb by its inherent meaning’ (‘Hebrew Verbal System’, 160; he builds on earlier versions of the relative tense theory of CBH; see esp. Bergsträsser II, §§6–7. For an historical survey, see McFall, Enigma, 21–4, 33–4, 41–3, 177–9). It is only because †erem yiqtol is a relative clause, whose translation into English (and other Indo-European languages) typically requires the shifted past, that scholars have often misidentified the Hebrew construction as a form of the past tense. In }az yiqtol, however, the imperfect has been converted into a genuine preterite. What is more, the adverbial

19


we will see, the diachronic hypothesis is almost certainly false. But as we have already seen, even if it were true, it would not, in and of itself, preclude the existence of a synchronic pattern or rule. Rather, one must produce arguments for assigning }az yiqtol to the dictionary as a listeme. Two possibilities immediately present themselves.29 One might be tempted to analyse }az yiqtol as fixed ‘idiom chunks’ preserving the archaic preterite *yaqtul.30 Idioms have two defining linguistic features: unpredictable (idiomatic) meaning, and non-productive (fixed) syntactic structure. To be more precise, Andrew Radford defines idioms as expressions ‘which have an idiosyncratic meaning that is not a purely componential function of the meaning of their individual parts’.31 Apropos of their syntax, he continues: ‘There seems to be a constraint that only a string of words which forms a unitary constituent can be an idiom. Thus, while we find idioms … which are of the form verb + complement … we don’t find idioms of the form subject + verb where the verb has a complement which isn’t part of the idiom’.32 Consider expressions such as: ‘kick the bucket’; ‘bite the bullet’; ‘break the ice’. The meaning of these expressions may have made some sense in their original context, but they survive unchanged as odd relics from the past. They must therefore be memorized individually, like words themselves. Note as }az, unlike †erem, locates the event, not in some relative future, but at a definite point in time, viz., ‘then’. Thus, it has been argued that }az marks the insertion of archival material into Kings, and replaces a date in the source text (James A. Montgomery, ‘Archival Data in the Book of Kings’, JBL 53 [1934], 49; Martin Noth, The Deuteronomistic History [Sheffield 1981], 128 n. 45). On the relation between temporal adverbs and verbal tense and aspect, see Hamida Demirdache and Myriam Uribe-Etxebarria, ‘The Syntax of Time Adverbs’, in Jacqueline Guéron and Jacqueline Lecarme (eds), The Syntax of Time (Cambridge 2004), 143–79. For a general overview of }az, see Martin Jan Mulder, ‘Die Partikel [}az] im biblischen Hebräisch’, in K. Jongeling, H.L. Murre-Van den Berg and L. van Rompay (eds), Studies in Hebrew and Aramaic Syntax: Presented to Professor J. Hoftijzer on the Occasion of his Sixty-Fifth Birthday (Leiden 1991), 132–42. 29 The following arguments also apply, mutatis mutandis, to †erem yiqtol, which is to say, †erem yiqtol is not a listeme, but is generated by the grammar. 30 Joosten calls }az yiqtol a ‘fixed phrase’ preserving an ‘older grammatical function’ (‘Prefix Conjugation’, 25), while Niccacci identifies it as a specimen of ‘certain fixed constructions’ preserving the ‘archaic use of YIQTOL’ (Syntax, 194). Neither provides a precise linguistic definition of this ‘fixed’ phrase/construction with which one might have tested their hypothesis. Montgomery similarly proposes that the construction is a ‘stylism’ resulting from the use of archival data — though he does not discuss the problem of tense per se (‘Archival Data’, 49). 31 Radford, Syntactic Theory and the Structure of English (Cambridge 1997), 322. 32 Radford, Syntactic Theory, 322–3, emphasis his. 20


well their fixed collocation of a particular verb with a particular complement. As fixed expressions they do not even tolerate the insertion of modifiers, at which point they lose their figurative (idiomatic) meaning. *He kicked the green bucket. *Sooner or later you’ll need to bite the silver bullet. *It’s time we broke the hard ice.

It is immediately clear, however, that }az yiqtol fails both criteria. It has a productive syntax, indiscriminately combining subjects, verbs and complements, and it does not in and of itself have an unpredictable idiomatic meaning.33 Alternatively, one could simply analyse }az yiqtol as a ‘strong’ or irregular verb form, which preserves the remnant of an archaic morphological pattern. As Pinker notes apropos of irregular English verbs — drink-drank, sink-sank, shrink-shrank, etc. — they ‘are mere fossils of [Proto-Indo-European] rules’.34 In other words, while the ‘weak’ forms evolve in conformity with those regular synchronic patterns generated by the grammar, the ‘strong’ forms, immune for whatever reason to certain linguistic changes, are left behind by history in various archaic states, which speakers perceive as grammatical irregularities. In CBH, the short forms of wayyiqtol constitute just such an irregular pattern, and they thus provide important evidence for the historical relation between wayyiqtol and *yaqtul. This hypothesis, however, fails to account for }az yiqtol, which is conspicuously regular in terms of the imperfect tense. As is well known, one predictably encounters the long form of the prefix conjugation in these constructions.35 1 Kgs 11:7: }az yibneh selomoh bamâ likmôs — not yiben, as in wayyiben Then Solomon built a high place for Chemosh.

If one still insisted on tracing yiqtol in such constructions back to *yaqtul, one would then have to admit that it is no longer a mere fossil, but an element that has been reabsorbed by the grammar and 33 Pinker identifies another class of listemes similar to idioms, viz., ‘collocations’: ‘string[s] of words commonly used together: excruciating pain; in the line of fire’ (Words and Rules, 290, emphasis his); see also Geoffrey Nunberg, Ivan Sag, and Thomas Wasow, ‘Idioms’, Language 70 (1994), 491–538. I will not consider this case separately, however, since }az yiqtol is not a collocation for the same reasons that it is not an idiom. 34 Pinker, Language Instinct, 138. 35 See also Exod. 15:1; Num. 21:17; Deut. 4:41; Josh. 8:30; 2 Kgs 12:18.

21


resumed evolving as a regular linguistic form — in other words, that }az *yaqtul has been synchronically reinterpreted as }az + imperfect. It may be true, as Edward Greenstein has observed, that ‘in one instance [of }az yiqtol] the Masoretic vocalization reflects the shorter *yaqtul rather than the longer *yaqtulu form’:36 }az yaqhel (‘Then [Solomon] assembled’; 1 Kgs 8:1) — not yaqhîl. But does this justify deriving }az yiqtol from the archaic prefix preterite? One could just as easily point to irregular long forms of wayyiqtol — e.g., wattibkeh (1 Sam. 1:7; 2 Kgs 22:19) and wayyibneh (Josh. 19:50; 1 Kgs 18:32). Following Greenstein’s logic, one should derive wayyiqtol from *yaqtulu.37 As he himself argues, however — correctly this time — the more common short forms (wayyaqom, wayyiben, etc.) argue for its derivation from *yaqtul instead.38 In both cases, then, the exceptional forms merely prove the rule. But they are not without significance. On the one hand, wattibkeh and wayyibneh though quite possibly ungrammatical, indicate the writers’ perception of the ‘conversive’ relationship between wayyiqtol and yiqtol — a type of linguistic Freudian slip.39 In 1 Kgs 8:1, on the other, a biblical writer (or Masorete), aware of how irregular it is for yaqhîl (marked as imperfect) to function as a preterite, ‘corrected’ it by analogy with wayyiqtol. In doing so, this individual betrayed the grammatical equivalence he perceived between }az yiqtol and wayyiqtol. The preponderance of evidence, however, suggests that this example is an instance of hypercorrection. Not only does }az yiqtol not behave like a listeme, we have positive evidence that it has nothing to do with the preterite *yaqtul in the first place. For all the occurrences of }az yiqtol in direct discourse (Benveniste’s discours) refer to events in the future: Gen. 24:41; Exod. 12:44, 48; Lev. 26:34, 41; Deut. 29:19; Josh. 1:8; 20:6; 36

Greenstein, ‘Prefixed Preterite’, 8. Joosten rightly pronounces it a ‘grammatical anomaly’ (‘Prefix Conjugation’, 25 n. 24). 37 That wattibkeh in 1 Sam. 1:7 probably continues an iterative past sequence does not change the fact that it is, morphologically speaking, a (nonstandard) form of wayyiqtol. See the unconvincing analysis of these long forms in Paul Joüon and Takamitsu Muraoka, A Grammar of Biblical Hebrew (2 vols, Rome 1991), II, §79m. 38 Greenstein, ‘Prefixed Preterite’, 8–9. 39 This sort of error — producing a regular form (‘runned’) where an irregular one is called for (‘ran’) — is typical of non-native speakers (and young children). Perhaps, then, these non-standard long forms are due to dialectal variation (including that due to diachronic evolution), as opposed to mere carelessness on the part of a native speaker of ‘normative’ CBH. 22


1 Sam. 6:3; 20:12; 2 Sam. 5:24; 2 Kgs 5:3.40 If this construction were a vestige of *yaqtul, it should refer uniformly to the past.41 Instead, we find that this single construction can refer to either the past or the future. This is a grammatical problem that any adequate theory must solve — more on which below. Finally, diachronic evidence within the Bible itself suggests that }az yiqtol (with past tense meaning) did not evolve from }az *yaqtul or even }az *yaqtulu, but from }az qatal. For it is precisely this last construction that we find in archaic poetry.42 Exod. 15:15: }az nibhalû }allûpê }edôm Then were dismayed the chieftains of Edom.

On the basis of this limited but telling evidence, it appears that }az yiqtol, far from being an archaism in BH, first appears in CBH prose, when yiqtol simply replaces qatal.43 To summarize: }az yiqtol is not a listeme (a vestige of *yaqtul ); rather, it is generated by the grammar in the form of the imperfect. And yet, it does not take the tense value of the imperfect (presentfuture), but rather of the preterite. How can this be? Exod. 15:1: }az yasîr moseh ûbenê yisra}el }et hassîrâ hazzo}t layhwh Then Moses sang, and the Israelites, this song to the Lord. Judg. 5:1: wattasar debôrâ ûbaraq ben }abînoäm bayyôm hahû} And Deborah and Barak son of Abinoam sang on that day …

Consider the nearly identical syntax and context of Exod. 15:1 and Judg. 5:1. Even the temporal adverb, }az, in the former has its equivalent in the latter — bayyôm hahû}. But for the change in subject, these two sentences have virtually the same meaning, including their tense value.44 So much so that I maintain that }az yasîr and 40

In direct discourse, }az qatal is used to refer to the past — more on which below. 41 See McFall, Enigma, 222–3. 42 See also Gen. 49:4; Judg. 5:11, 13, 19, 22. 43 It is true that in Ugaritic }idk (‘then’) can be followed by the prefix conjugation: }idk l ttn pnm ¨m }il, ‘then she verily sets face toward El’ (Daniel Sivan, A Grammar of the Ugaritic Language [Leiden 1997], 181). But the evidence of biblical poetry is more pertinent here. 44 Contra those who give the construction the tense-aspect value of the imperfect: Carl Brockelmann, Hebräische Syntax (Neukirchen 1956), §42a; Driver, Tenses, §§26–7; E. Kautzsch (ed.), Gesenius’ Hebrew Grammar 2, trans. A.E. Cowley (Oxford 1910), §107c (hereafter GKC); Isaac Rabinowitz, ‘}Az Followed by Imperfect Verb-Form in Preterite Contexts: A Redactional Device in Biblical Hebrew’, 23


wattasar — i.e., }az yiqtol and wayyiqtol — are syntactically parallel. Apparently, then, }az has replaced the conversive waw as a type of syntactic equivalent,45 while the now orphaned converted imperfect retains its preterite tense value.46 So reasoned the writer or Masorete in 1 Kgs 8:1, who, as noted earlier, apparently corrected — or hypercorrected — the ostensibly irregular long form of the preterite, yaqhîl, replacing it with yaqhel. What is surprising (and neglected in the literature) is that }az yiqtol has past tense meaning only in narration. In direct discourse, it retains the future tense value of the imperfect, so that the alternate construction }az qatal must be used instead to refer to past events: Josh. 22:31; 2 Sam. 5:24; etc.47 Tense conversion elegantly accounts for this striking fact: }az yiqtol is converted in narration, unconverted in discourse. Furthermore, the synchronic distribution of converted VT 34 (1984), 53–62; E.J. Revell, ‘The System of the Verb in Standard Biblical Prose’, HUCA 60 (1989), 11; Christo H.J. van der Merwe, Jackie A. Naudé, and Jan H. Kroeze, A Biblical Hebrew Reference Grammar (Sheffield 2002), §19.3.2. One should probably include in this group Frithiof Rundgren, who proposes the rather baroque notion of an ‘imparfait de rupture’ that ‘neutralizes’ the opposition between the ‘kursive’ (imperfect) and ‘non-kursive’ (perfect) tenses (‘Erneuerung des Verbalaspekts im Semitischen: Funktionell-diachronische Studien zur semitischen Verblehre’, Acta Universitatis Upsalensis; Acta Societatis Linguisticae Upsalensis n.s. 1/3 [1963], 88–9). 45 Their syntactic equivalence is suggested by their complementary distribution: }az (not counting compound forms with min) is attested 118 times in the Masoretic Text (MT); we}az occurs only 4 times. Of the latter, three are found in the apodosis of a conditional statement, which motivates the presence of we; the fourth (Jer 32:2) constitutes the only unmotivated attestation of we}az. The infrequency of we}az is striking, given the frequency of sentence-initial waw in CBH. It means that the presence of }az precludes we (and vice-versa), which suggests that their combination was felt to be redundant. 46 According to Wolfgang Schneider, }az in narrative ‘ersetzt das TempusZeichen “wa”’ (Grammatik des biblischen Hebräisch 8 [München 1993], §48.4.3.4); see again Zuber, Tempussystem, 163–5. Rainey vaguely suggests that ‘the use of the imperfect as a narrative tense [is probably] made possible by the combination with the adverb, [}az]’ (‘Further Remarks on the Hebrew Verbal System’, HS 29 [1988], 35). McFall offers the still weaker hypothesis (incorrectly attributed to †erem as well): ‘The so-called conversive force of [}az] and [†erem] depends on the fact that a corresponding tense (the past) precedes, with which it is co-ordinated’ (Enigma, 223). Joüon and Muraoka correctly observe that yiqtol here has ‘no iterative or durative aspect, and thus [has] the value of qatal, which would be the expected form’ (A Grammar, II, §113h), but fails to offer an adequate explanation. While Ewald’s theory is based on an incorrect diachronic derivation, we + }az → wa (§231a), it correctly relates }az yiqtol to the converted imperfect (§§233b, 346d). 47 While }az qatal also occurs in narration (Exod. 4:26; Josh. 10:33; 22:31; etc.), this does not affect my argument here. 24


and unconverted }az yiqtol corroborates, in turn, the distinction I have maintained elsewhere between narration (histoire) and direct discourse (discours) in CBH: like wayyiqtol, converted }az yiqtol is characteristic of narration, not discourse.48 The matching diachronic distributions of }az yiqtol and wayyiqtol further corroborates their grammatical equivalence, albeit indirectly. It has been suggested that }az yiqtol begins dying out in Late Biblical Hebrew (LBH), eventually to be replaced by }az qatal 49 — just as it originally derived, I have argued, from the latter. It should be admitted that already in CBH, }az qatal apparently provided a viable alternative to }az yiqtol in narration: Gen. 4:26; Exod. 4:26; Josh. 10:33; Judg. 8:3; 13:21; 2 Sam. 21:17, 18; 1 Kgs 8:12; 9:24; 22:50; 2 Kgs 14:8.50 Nevertheless, the relative frequency of their appearances in Chronicles demonstrates a decisive shift in LBH: }az qatal occurs seven times (1 Chron. 15:2; 16:7; 20:4; 2 Chron. 6:1; 8:12; 8:17; 24:17); }az yiqtol (with past tense meaning) occurs only twice, and both of these derive from the parallel passage in Kings (2 Chron. 5:2 = 1 Kgs 8:1; 2 Chron. 21:10 = 2 Kgs 8:22).51 This construction is arguably replaced by baët hahî} qatal as well, which apart from a single appearance in Josh. 5:2, is restricted to a cluster of occurrences in Kings: 1 Kgs 14:1; 2 Kgs 16:6, 18:16, 20:12, 24:10. The life span of }az yiqtol thus coincides with that of wayyiqtol, which likewise flourished in CBH.52 And this correlation of their diachronic evolution supports my hypothesis that the two are synchronically related. Finally, as a type of linguistic thought experiment, one might consider various near synonyms for }az. Why does the orphaned converted imperfect co-occur only with }az, and not with these alternatives? CBH possesses two prepositional phrases that constitute near synonyms of }az: bayyôm hahû} (‘in that day’) and baët hahî} (‘at that time’). 48

Kawashima, Biblical Narrative, 35–76. See Brockelmann §42; Waltke and O’Connor, An Introduction, §31.6.3 50 I exclude here occurrences found in direct discourse (discussed earlier), where qatal is the expected preterite: Josh. 22:31; 2 Sam. 2:27 (counterfactual); 5:24; 2 Kgs 13:19 (counterfactual). It is far from clear that }az qatal is ‘ungrammatical’ as Montgomery claims (‘Archival Data’, 49). Neither do I perceive a clear semantic difference between }az yiqtol and }az qatal. I view occurrences of }az qatal in narration as the result of dialectal variation or of genuine uncertainty regarding the use of tense with }az. 51 See Arno Kropat, Die Syntax des Autors der Chronik (Giessen 1909), 17. 52 See esp. Sáenz-Badillos, History of Hebrew, 112–29; Smith, Waw-Consecutive, 27–33. 49

25


Gen. 15:18: bayyôm hahû} karat yhwh }et }abram berît In that day, the Lord cut a covenant with Abram.53 Josh. 5:2: baët hahî} }amar yhwh }el yehôsua¨ At that time, the Lord said to Joshua …54

These two prepositional phrases, while similar in meaning and function to }az, have a different syntactical status. They occur only rarely in sentence-initial position — recall, e.g., the sentence-initial position of }az in Exod. 15:1 and the sentence-final position of bayyôm hahû} in Judg. 5:1. And in those few cases when they are in a position to replace conversive waw, they seem disqualified from doing so by virtue of their being a full prepositional phrase, rather than a mere particle. Thus, we never encounter bayyôm hahû} yiqtol or baët hahî} yiqtol, where the imperfect has been converted into the preterite. They are followed by either yiqtol as relative future (Lev. 22:30) or the preterite qatal (see previous paragraph). More frequently, however, they follow the verb and therefore cannot interfere syntactically with wayyiqtol to begin with.55 Other Orphaned Converted Imperfects Aside from these occurrences of }az yiqtol with past tense meaning, there are only a few additional cases of the orphaned converted imperfect, but these nonetheless provide important corroboration of the linguistic analysis just proposed. An exceptional case is found in Judg. 2:1 (}aäleh, ‘I brought [you] up’), where conversive waw has simply been dropped for no discernable reason. The subsequent verb sequence, however, leaves no doubt as to its tense value: wa}abî}… wa}omar, ‘And I brought … and I said’. More typically, conversive waw is displaced from its prefixal position. 1 Kgs 20:33: weha}anasîm yenaÌasû waymaharû wayyaÌle†û And the men divined and hurried and took [his meaning].

In effect, the grammatical subject (‘the men’) is inserted between conversive waw and converted imperfect. Again, the subsequent wayyiqtols leave no doubt as to its tense value. 53

See also Josh. 4:14 and 1 Kgs 8:64. See also 1 Kgs 14:1; 2 Kgs 16:6; 18:16; 20:12; 24:10. 55 To take just a few examples, see: Gen. 30:35; 33:16; 48:20 (bayyôm hahû} ); and Josh. 6:26; 11:10, 21; Judg. 3:29 (baët hahî} ). 54

26


1 Sam. 1:10: wehî} marat napes wattitpallel äl yhwh ûbakoh tibkeh56 And she was bitter of soul, and she prayed to the Lord and wept deeply.

Here, the infinitive absolute (bakoh), functioning as a manner adverbial (‘deeply’), has been embedded within the underlying verb phrase, thus displacing conversive waw from its prefixal position.57 One can restore the underlying syntax by moving the infinitive after wayyiqtol, where it more properly belongs: wattitpallel äl yhwh wattebk bakoh (cf Gen. 31:15). True, one expects the syntactic interruption, which precludes clause-initial wayyiqtol, to result in qatal: ûbakoh baketâ. At least certain biblical writers, however, thought it grammatically acceptable to insert the adverbial between the conversive waw and the converted imperfect. And just as the grammar generates the long form tibkeh according to regular rules, so it generates conversive waw not as wa- but as û. The same syntactic phenomenon seems to be in evidence in the following examples as well, but now an entire phrase separates conversive waw from orphaned converted imperfect. Exod. 8:20: wayyaäs yhwh ken wayyabo} ärob kabed… ûbekol }ereÒ miÒrayim tissaÌet ha}areÒ mippenê heärob And the Lord did so. And a great swarm came… And in all the land of Egypt the land was ruined because of the swarm.

Analogous to the previous example, the adverbial phrase (bekol }ereÒ miÒrayim) inserts itself between the waw and the orphaned converted imperfect (tissaÌet). I interpret the underlying syntactic relations thus: *wattissaÌet ha}areÒ bekol }ereÒ miÒrayim mippenê heärob.58 56

We find a striking potential parallel in the Deir Allah inscription: wyqm bl¨m mn mÌr … ubkh ybkh, ‘And Balaam rose the next day … and wept bitterly’ (Combination 1.3–4; see Shmuel Ahituv, Handbook of Ancient Hebrew Inscriptions, [Jerusalem 1992], 266 [Hebrew]). Given the textual gap, it is impossible to analyse the syntax precisely. If wbkh ybkh continues the preterite sequence of wyqm, I would analyse it in the same way as 1 Sam. 1:10. For a radically different construal of the inscription, see Klaas A.D. Smelik, Writings from Ancient Israel (Louisville 1992), 83. 57 Contra Waltke and O’Connor’s semantic explanation (An Introduction, §31.2c). On the infinitive absolute and its interaction with tense, see Kawashima, Biblical Narrative, 52–6. 58 Even Driver describes this imperfect as ‘of an exceptional character’ (Tenses, §27g). Giving the verb an imperfective tense value — e.g., ‘the land was being ruined’ — makes less sense semantically than the simple preterite reading. Note, restoring wattissaÌet results in the repetitive phrase, ha}areÒ bekol }ereÒ miÒrayim. The writer may have fronted the prepositional phrase, making an orphan of the converted imperfect, in order to avoid this inelegant concatenation. 27


Num. 35:20: we}im besin}â yehdapennû }ô hislîk älayw biÒdiyyâ wayyamot And if in hatred he pushed him or hurled something at him, lying in wait, and he died…

In Num. 35:20, note how we}im … yehdapennû is syntactically parallel to }ô hislîk (qatal ) and continued by wayyamot (wayyiqtol ).59 Moreover, the other parallel cases considered in this passage are all presented as past events (qatal ): e.g., we}im … hikkahû (Num. 35:16, 17, etc.). Deut. 2:12: ûbese¨îr yasebû haÌorîm lepanîm ûbenê ësaw yîrasûm wayyasmîdûm wayyesebû taÌtam And in Seir had dwelt the Horim previously, but the sons of Esau dispossessed them and destroyed them and dwelt there in their place. 2 Kgs 20:14: mâ }amerû ha}anasîm ha}elleh ûme}ayin yabo}û }elêka What did these men say to you, and from where did they come to you?

Now in both of these examples, one must seriously contend with the possibility that the sentence is simply ungrammatical — in spite of the fact that we can no longer consult a native speaker. I for one would replace yîrasûm and yabo}û with their qatal equivalents, yaresû and ba}û (cf. Gen. 42:7) — not as textual emendations, but as grammatical corrections.60 Even so, I interpret the actual data as indirect evidence of the writers’ mental grammar. What function did each writer intend for the unexpected imperfect? In both cases, the converted imperfect.61 Thus, yîrasûm in Deut. 2:12 is continued not by 59 This verse comes from casuistic law and is sometimes interpreted with imperfective/future meaning, but see n. 70 below. 60 Chomsky defines linguistic ‘competence’ in opposition to ‘performance’, which is subject to individual variation and error. Like Saussure’s distinction between langue and parole, this reduction of the linguistic object to an idealized form is the epistemological act whereby linguistics constitutes itself as a science. Inasmuch as one seeks to analyse the linguistic competence of CBH speakers, not all the data contained in MT can simply be accepted as grammatical. For even if it were purified of all textual corruption, MT would still pertain by definition to linguistic performance. Thus, although we can no longer consult native speakers, one must, however cautiously, attempt to separate the grammatical from the ungrammatical. This is no different from the venerable practice of text criticism, which proceeds by sorting through variant readings and even emending the text when necessary. The refusal to admit the existence of ungrammatical constructions (and textual corruptions) in MT is akin to the doctrine of scriptural inerrancy. 61 Greenstein lists Deut. 2:12 and 2 Kgs 20:14 (among less convincing examples) as cases where: ‘Without special pleading, it is simplest to analyze the prefixed

28


weqatal, but wayyiqtol; and in 2 Kgs 20:14, Isaiah’s question (yabo}û) is answered in the perfect tense (ba}û), the event being treated as a fait accompli.62 So why does the converted imperfect appear in orphaned form? Again, due to the displacement of conversive waw. I would thus grammatically ‘restore’ these sentences as follows: *Deut. 2:12: ûbese¨îr yasebû haÌorîm lepanîm wayyîresûm benê ësaw wayyasmîdûm wayyesebû taÌtam *2 Kgs 20:14: mâ }amerû ha}anasîm ha}elleh wayyabo}û me}ayin }elêka

In the first case, however, we lose the disjunction between the first and second clauses; in the second, the interrogative needs to be in clause-initial position. For these reasons, neither sentence actually occurs in these forms. In terms of tense sequence, however, these hypothetical reconstructions are the best approximation of the grammatically acceptable syntactic form each writer was aiming for. Finally, one finds a cluster of orphaned converted imperfects caused by the insertion of the negative particle. 2 Sam. 2:28: wayyitqa¨ yô}ab bassôpar wayyaämdû kol ha äm welo} yirdepû ¨ôd }aÌarê yisra}el welo} yasepû ¨ôd lehillaÌem And Joab sounded the trumpet and all the troops halted, and they no longer pursued Israel and they no longer continued to fight. 1 Kgs 8:8: wayya}arikû habbaddîm wayyera}û ra}sê habbaddîm min haqqodes äl penê haddebîr welo} yera}û haÌûÒâ And the poles extended out, and the ends of the poles were seen from the sanctuary in front of the inner shrine, but they were not seen outside. Jer 44:22: welo} yûkal yhwh ¨ôd lase}t…wattehî }arÒekem And the Lord could no longer bear … and so your land became …

In all three cases, the most natural reading of yiqtol is as the converted imperfect.63 Note in particular the parallel in 2 Sam. 2:28 between welo} yirdepû ¨ôd (yiqtol ) and welo} yasepû ¨ôd (qatal ), both presumably having the same tense value. In fact, welo} qatal is the standard form: Deut. 34:10; Josh. 2:11, 5:1,12; 1 Kgs 10:5; 2 Kgs 2:12. verbs in these pairs as preterites’ (‘Prefixed Preterite’, 11). What he has in mind, however, is a diachronic derivation from *yaqtul. 62 Contra Zuber, Tempussystem, 134–5. This past tense context differentiates 2 Kgs 20:14 from similar examples in present tense contexts: Josh. 9:8; Judg. 17:9; 19:17. On these latter examples, see also McFall, Enigma, 84–5; Waltke and O’Connor, An Introduction, §31.3b. 63 Contra Revell’s analysis of 1 Kgs 8:8 (‘System of the Verb’, 12). 29


In order to account for these anomalous occurrences of yiqtol, one must yet again choose between inventing an ad hoc semantic nuance and invoking an established syntactic principle, viz., tense conversion. Orphaned Converted Perfects The converted perfect (weqatal ),64 as a verb phrase, is subject to the same syntactical manipulations as the converted imperfect (wayyiqtol ). That is, conversive waw can either be replaced by a syntactic equivalent, or displaced by an embedded element, leaving behind an orphaned converted perfect. Since these linguistic principles should by now be clear, the remaining discussion will primarily consist of the presentation of evidence, and this often in highly elliptical form. But in the interest of firmly establishing the orphaned converted perfect as a genuine syntactic phenomenon, the number of examples presented will not be any less exhaustive. In one instance, conversive waw is unceremoniously dropped — similar to }aäleh in Judg. 2:1 (discussed above).65 Isa. 8:8: weÌalap bîhûdâ sa†ap weäbar äd Òawwa}r yaggîa¨ And it will sweep into Judah; it will overflow and pass through; up to the neck it will reach.

The sequence of weqatal and yiqtol forms suffices to establish the future tense value of sa†ap. More commonly, we find orphaned converted perfects where conversive waw has been replaced by a syntactic equivalent. Compare the prodoses in the following two examples, two versions of the oath Judah makes to his father. Gen. 43:9: }im lo} habî}otîw }elêka wehiÒÒagtîw lepanêka weÌata}tî leka kol hayyamîm If I do not bring him back to you and set him before you, I shall be guilty before you forever. Gen. 44:32: }im lo} }abî}ennû }elêka weÌata}tî le}abî kol hayyamîm If I do not bring him back to you, I shall be guilty before you forever.

Rather than propose some overly subtle semantic distinction between }im lo} habî}otîw and }im lo} }abî}ennû, one should simply admit 64

For a recent survey of the various uses of the converted perfect, see Robert E. Longacre, ‘Weqatal Forms in Biblical Hebrew Prose’, in Bergen (ed.), Biblical Hebrew and Discourse Linguistics, 50–98. 65 See already Böttcher, 2.205. 30


that they have the same tense value, and identify the former as an orphaned converted perfect.66 Interestingly enough, we find a regularly attested construction — analogous to }az yiqtol — in which the conversive waw in weqatal is replaced by a syntactic equivalent: }ô qatal.67 In these examples, we find a set of parallel possibilities: an initial case (weqatal ) followed by one or more alternatives, each introduced by }ô (‘or’). Thus, }ô qatal is syntactically parallel (equivalent) to weqatal. Moreover, }ô qatal typically occurs in a sequence of imperfects and/or converted perfects, further confirmation of its tense value. Num. 11:7–8: wehamman kizra¨ gad hû} we¨ênô ke¨ên habbedolaÌ sa†û haäm welaqe†û we†aÌanû bareÌayim }ô dakû bammedokâ ûbisselû Now manna, it was like coriander seed, and its appearance was like the appearance of bdellium. The people would walk around and gather and grind it in a mill or pound it in a mortar, and boil it.

In Num. 11:7–8, sa†û is an habitual past — it should read wesa†û (or perhaps yasu†û).68 The habitual past is continued by a sequence of converted perfects: welaqe†û, we†aÌanû, }ô dakû, ûbisselû. Thus, the manna was processed in one of two possible parallel fashions: the Israelites either ‘would grind’ it (we†aÌanû) ‘or would pound’ it (}ô dakû). Converted }ô qatal appears most frequently in the prodosis of casuistic law.69 Here it becomes part of a parallel structure of two or more legal conditions. Since legal stipulations generally refer to future contingent events, }ô qatal should again be analysed as an orphaned converted perfect, in order to obtain the logically required tense value.70 For this reason it functions, in the following examples, as the 66 Contra Driver, Tenses, §138. C. van Leeuwen assigns future tense meaning to }im qatal in Gen. 43:9, but does not provide a convincing syntactic explanation for it ( ‘Die Partikel [}im]’, OTS 18 [1973], 21 and n.3). 67 See Zuber’s insightful discussion of }ô qatal in Tempussystem, 157–9. 68 Again, I am not proposing a textual emendation, but a grammatical correction. Note, in 1 Sam. 2:13, the habitual past sequence begins with a participial phrase, which I would compare to the two verbless clauses at the beginning of this passage. Cf. Driver’s forced explanation (Tenses, §114a). 69 On }ô in conditional clauses, see Driver, Tenses, §138 Obs; GKC §159cc; Joüon and Muraoka, A Grammar, II, §167 — none of whom comment on the enigmatic yet regular use of qatal here. 70 In a telling contrast, these examples can be distinguished from others where }ô qatal requires a relative past reading. In such cases, it refers to a circumstance markedly antecedent to the legal case under consideration: Exod. 21:36 (an ox previously known to be prone towards aggressive behaviour); Lev. 5:1 (a potential witness who had earlier seen or learned about some disputed matter); Num. 35:18, 20, 21, 22 (the prior killing which resulted in the killer’s fleeing to a city of refuge

31


syntactic equivalent of either yiqtol or weqatal. In the list of examples below, I provide only the relevant parallel verbal constructions.71 Exod. 21:37: û†ebaÌô (slaughters) | | }ô mekarô (or sells) Exod. 22:9: ûmet (dies) | | }ô nisbar (or is injured) | | }ô nisbâ (or is carried off) Exod. 22:13: wenisbar (is injured) | | }ô met (or dies) Lev. 4:22–3: we}asem (realizes guilt) | | }ô hôda¨ (or is made known) Lev. 4:27–8: we}asem (realizes guilt) | | }ô hôda¨ (or is made known) Lev. 5:21–2: wekiÌes (deceives) | | }ô äsaq (or defrauds) | | }ô maÒa} (or finds) Lev. 25:48–9: tihyeh lô (will have) | | }ô hissîgâ (or prospers) Num. 5:14: weäbar älayw (comes over him) | | }ô äbar älayw (or comes over him) Num. 30:3: yiddor (makes a vow) | | }ô hissaba¨ (or swears an oath) Num. 30:4, 11–12: tiddor (makes a vow) | | we}im … nadarâ (but if … makes a vow) | | }ô }aserâ (or takes a pledge)

In all these cases, }ô replaces conversive waw, leaving behind an orphaned converted perfect. The equivalence of }ô qatal to weqatal is further emphasized by the tense sequence in Lev. 5:21–2 (}ô maÒa} … wekiÌes … wenisba¨ ), Num. 5:14 (}ô äbar… weqinne}), and Num. 30:11–12 (we … nadarâ … }ô }aserâ … wesama¨ ): in these examples, }ô qatal is continued by weqatal within the prodosis. In a telling contrast, analogous passages sometimes employ }ô yiqtol, but only when a syntactic element intervenes between the conjunction and the verb. It indicates the equivalence: }ô …yiqtol = }ô qatal. Apparently, the interposed element prevents }ô from replacing the conversive waw of weqatal, resulting in yiqtol instead of converted qatal.72 1 Sam. 26:10: Ìay yhwh kî} im yhwh yiggapennû }ô yômô yabô} wamet }ô bammilÌamâ yered wenispâ As the Lord lives, the Lord will strike him down, or his day will come and he will die, or he will go down to battle and perish.

In casuistic law, we find the parallel structure: kî yiqtol … }ô kî yiqtol.73 or his execution at the hands of the victim’s kin). Note, Num. 35:16–23 comprises several prodoses (vv16, 17, 18, 20, 21, 22f), each having the tense sequence qatal/ wayyiqtol (except for the orphaned converted yehdapennû in v 20); the imperfect/ converted perfect only appear in the corresponding apodoses. 71 In two cases we find }ô qatol (infinitive absolute) — Lev. 25:14 (}ô qanoh) and Deut. 14:21 (}ô makor) — and in both cases I would change (emend, this time) qatol to qatal (orphaned converted perfect). 72 See also Exod. 19:13. 73 See also Lev. 5:3; 13:16; 15:25; Deut. 24:3. 32


Exod. 21:33: wekî yiptaÌ }îs bôr }ô kî yikreh }îs bôr welo} yekassennû wenapal sammâ sôr }ô Ìamôr And when a man opens a pit, or when a man digs a pit and does not cover it, and an ox or a donkey falls into it …

The available evidence suggests that, without the syntactic interruption between }ô and yiqtol, we would find }ô qatal instead: }ô ba} yômô … }ô yarad bammilÌamâ; }ô karâ bor. In a related set of cases, conversive waw is not replaced by a syntactic equivalent, but rather displaced, separated from the converted perfect, by an embedded construction — weqatal, like wayyiqtol, again being treated as a verb phrase, not a verb. Gen. 17:16: ûberaktî }otah wegam natattî mimmennâ leka ben ûberaktîha wehayetâ legôyim malkê ämmîm mimmennâ yihyû And I will bless her, and indeed I will give you a son from her. And I will bless her, and she will become nations. Kings of peoples will be from her.

As Zuber, apparently alone, has recognized: ‘waw allenfalls auch über eine Partikel hinweg seine Umkehrfunktion ausüben kann’.74 The adverbial gam, like the infinitive absolute in 1 Sam. 1:10 (discussed above), has interposed itself between conversive waw and converted perfect. Remove the adverbial and one uncovers a typical weqatal chain: ûberaktî, wenatattî, ûberaktîha, wehayetâ. Analogous to the long form of the orphaned converted imperfect (see discussion of 1 Sam. 1:10 above), the verb phrase wenatattî (with ultimate stress), once it has been separated into its constituents by the syntactic insertion, loses the final stress characteristic of weqatal: wegam natàttî (with penultimate stress). 1 Sam. 2:16: lô} 75 kî ättâ titten we}im lô} laqaÌtî beÌozqâ No, you must give it now! And if not, I will take it by force.

Correct analysis of this conditional phrase must begin with the well established fact that the apodosis should begin with the converted perfect, as in 1 Sam. 6:9: we}im lo} weyada¨nû (‘And if not, then we will know’).76 In other words, laqaÌtî has the same tense value 74 Zuber, Tempussystem, 162. Major grammars — Driver, GKC, Joüon and Muraoka, Waltke and O’Connor — do not comment on this verse. One might also interpret natattî as a performative — ‘I hereby give’ — but the tense sequence argues for an orphaned converted perfect. 75 Reading with qere. 76 Recall that weqatal probably originated in this very syntactic function.

33


as weyada¨nû; i.e., it is an orphaned converted perfect.77 Correct analysis of this orphaned converted perfect requires that one restore the full version of the prodosis, which has been partially elided in surface structure: we}im lô} titten (‘And if you do not give’). Apparently, when the verb in the prodosis (titten) underwent deletion, the waw in the prodosis was able to take over the function of tense conversion from the conversive waw originally contained in the apodosis, which would have read welaqaÌtî. The final result: }im lo} separates conversive waw from the orphaned converted perfect. Finally, on several occasions, a particle, especially }im, has the same syntactic effect in the prodosis of casuistic law.78 For this last set of cases, I merely list several pairs of verb phrases. First, I provide several examples of the orphaned converted perfect. Then, in each case, I corroborate its tense value by offering a syntactic equivalent (yiqtol and/or weqatal ), selected from a closely related passage. I leave it to the reader to further confirm my analysis, by verifying in each case that the orphaned converted perfect occurs in a sequence of yiqtol and/or weqatal. we}im ra}â (Lev. 13:56) = we}im yir}eh (Lev. 13:53) — ‘he sees’ we}im †aharâ (Lev. 15:28) = wekî yi†har (Lev. 15:13) — ‘she becomes clean’ we …lo} hebî}ô (Lev. 17:3–4) = lo} yebî}ennû (17:8–9) — ‘he does not bring’ we}im lo} maÒe}â (Lev. 25:28) = ûmaÒa} (Lev. 25:26) — ‘he does (not) find’ we}im mak (Lev. 27:8) = yamûk (Lev. 25:25) — ‘he is poor’ we}im makar (Lev. 27:20) = yimkor (Lev. 25:29) — ‘he sells’ we}im henî} (Num. 30:6) = yanî} (Num. 30:9) — ‘he forbids’

As with analogous examples of the orphaned converted imperfect, it is possible that these are, strictly speaking, ungrammatical in ‘normative’ CBH. Even so, the notion of the orphaned converted perfect, a single linguistic principle, enables us to account for these anomalous occurrences of qatal without positing yet more semantic nuances. Conclusion In this way, seemingly disparate particular phenomena become part of a general grammatical order. What once required ad hoc sub77

Contra Driver, Tenses, §136g. On }im in conditional clauses, see Driver, Tenses, §§136,138; GKC §159 l–v; Joüon and Muraoka A Grammar, II, §167; van Leeuwen, ‘Partikel’, 16–27. On the possible future reference of }im qatal, see n. 66 above. 78

34


categorizations of qatal and yiqtol are now accounted for by a single grammatical principle — tense conversion. At the same time, the very hypothesis that wayyiqtol and weqatal are, grammatically speaking, converted tenses, receives unexpected confirmation — not a case of circular reasoning, but the corroboration of a theory previously established on independent grounds. True, I cannot explain why a converted tense is retained in orphaned form in some cases and not in others. In many cases, the orphaned converted tense appears to be an acceptable stylistic variant. In others, however, it most likely betrays the influence of non-normative CBH. In either case, the solution offered here, which explains a considerable number of unexpected tense forms, many in constructions that certainly are grammatical (}az yiqtol and }ô qatal ), seems preferable to the ad hoc, impressionistic, and overly subtle semantic explanations frequently offered, in an attempt to reconcile conspicuously odd examples of yiqtol and qatal with their superficially apparent tense values. And even in those cases where an orphaned converted tense form most likely constitutes an actual grammatical error, it betrays the intuition of at least certain biblical writers that wayyiqtol and weqatal are indeed the converted imperfect and converted perfect, respectively. Beyond the particular thesis, one must again insist on the larger theoretical point: the diachronic development of language as such does not enter into grammar. This is not to deny the importance of diachronic linguistics. However, as Saussure, again, predicted: ‘Having paid too much attention to history, linguistics will go back now to the static viewpoint of traditional grammar, but in a new spirit and with different methods’ (Course, 82). What I have suggested, but not been able to pursue here in detail, is that his prediction was fulfilled not by structuralism, but by Chomskyan linguistics. While there has been no need here for analyses based on linguistic ‘trees’, simply to appeal to the biblical writers’ linguistic ‘competence’, to seek to eliminate linguistic ‘performance’, and to employ concepts such as the ‘idiom’ that are defined in terms of generative rules for what is and is not acceptable, is already to enter into another linguistic universe.79 Address for correspondence: [email protected]

79

I am grateful to Ron Hendel and Mark Smith for commenting on an earlier version of this paper. Its basic arguments were first presented in the 2004 Annual Meeting of the Society of Biblical Literature. A substantially revised version was presented at the 2007 Annual Conference of the Association for Jewish Studies. 35

NOUVELLES INSCRIPTIONS ARAMÉENNES DU TEMPLE DE NANNAY À HATRA BAHAA AMER AL-JUBOURI UNIVERSITY OF BAGHDAD

Abstract The inscriptions discussed here were discovered during excavations undertaken by the Iraq Antiquities Service in the summer of 1992 in the remains of a building situated 90 m to the south-west of the enclosure of the Grand Temple at Hatra. The building, constructed above a raised pavement reached by three staircases, is composed of two broad rooms, each with three similar entrances, and in the middle of the west wall of the second room an entry leads to a square stone cellar. The building is similar to the third sanctuary, which is situated in the eastern part of the town, and, according to the author, would be the fourteenth sanctuary in Hatra and was consecrated to the goddess Nannay. Here, we take up the study of these new inscriptions which Dr Jabir Kh. Ibrahim published and translated in the journal Sumer. The numbering of the inscriptions is retained and continues the established numbering of the Hatra inscriptions.

Ces inscriptions étaient découvertes dans les fouilles entreprises par le Service Iraquien des Antiquités et du Patrimoine dans les décombres d’un édifice situé à une distance de 90 mètres du côté sud-ouest de l’enceinte du grand temple de Hatra en été 1992. L’édifice, construit au-dessus du dallage élevé escaladé par trois escaliers, se compose de deux salles larges. Chacune a trois entrées comparables et au milieu du côté ouest de la deuxième salle une entrée se conduit à une chambre carrée et voûtée de pierre. L’édifice est analogue au sanctuaire troisième, situé dans la partie orientale de la ville, et d’après l’auteur il serait le sanctuaire quatorzième de Hatra, consacré à la déesse Nannay. Nous reprenons ici l’étude de ces nouvelles inscriptions que Dr Jabir Kh. Ibrahim a publié et traduit en arabe dans la revue Sumer.1 1 J. Kh. Ibrahim, ‘Unpublished Inscriptions from the Temple of Nannay in Hatra’, Sumer 51 (2001–2), 200–16 (en arabe).

37

NOUVELLES INSCRIPTIONS ARAMÉENNES DU TEMPLE DE NANNAY

La numérotation des inscriptions est celle de Sumer et continue la numérotation de la série d’inscriptions précédente.2 4633 Inscription de deux lignes de caractères enfoncés, gravée sur la face latérale du seuil en marbre de Mossoul et situé à longe du dallage de l’entrée de la chambre sacrée. Longueur 220 cm, largeur 70 cm et hauteur 17 cm. La fin de la première ligne est très abîmée. Longueur du reste de la première ligne 195 cm.

1. bsnt 400+10+2 }qymyt }n} ¨qby kmr} dy nny rb} br syl} mn bny tmny qsys} }? n??? } …?? lnny 2. mrt tg} drgyt} dy Ì†r} ¨l Ìyyhy w¨l Ìy} †hmh kwlh kmr} dy nny w¨l slm Ì†r} klh 1. En l’an 412 (=100/101 de notre ère) j’ai érigé, moi, ¨qby, le grand prêtre de Nannay, fils de syl}, des Bene tmny, les symposiarques … pour Nannay de l’escalier, 2. la femme de la couronne, de Hatra, pour sa vie et la vie de sa tribu tous, les prêtres de Nannay, et pour la paix de Hatra tous.

L.1: bsnt. Il faut signaler, tout d’abord, l’absence de byrÌ, ‘au mois de’. Cette formule de datation nous est connue autrement seulement dans deux inscriptions, nos 214: 1 et 290: 1. }qymyt. Ibrahim lit }qymt, mais le yod est très net sur le facsimilé. Voir no. 288a: 7. ¨qby. C’est un hypocoristique attesté aussi sous la forme ¨qbw: 411: 8b; 450: 2. nny. Cette divinité féminine est attestée une seule fois à Hatra auparavant: 4: 1. 2 J.Kh. Ibrahim, ‘Unpublished Inscriptions from the Eastern Gate in the (City) Wall of Hatra’, Sumer 50 (1999–2000), 166–84 (en arabe), qui contient les numéros 424–62. Les numéros 417–23 seront publiés prochainement par le même auteur. Pour tous les numéros précédents, voir K. Beyer, Die aramäischen Inschriften aus Assur, Hatra un dem übrigen Ostmesopotamien (Göttingen 1998). 3 Cette inscription a été étudiée pour la première fois par l’auteur dans al-Jubouri, ‘Unpublished Aramaic Inscriptions from Hatra’, Journal of the Academy of Sciences (Syriac Corporation) 17 (1999), 92 (en arabe).

38


kmr} dy nny rb}. C’est la première attestation de l’existence à Hatra d’un grand prêtre de Nannay. D’ailleurs, rb} pour Ibrahim qualifie Nannay, mais Nannay est une déesse. Cette épithète appartient à kmr}, ‘le prêtre’, même s’il vient après le nom de la déesse. qsys}. J’y vois le pluriel qassise, ‘les symposiarques’, appartenant aux Bene tmny. syl}. Nom propre nouveau à Hatra. On le trouve aussi dans les anciennes inscriptions édesséniennes: syl} 4. tmny. Nom de tribu à quel appartienne ¨qby. Il est attesté à Palmyre comme nom propre: tmny}5. Ibrahim lit }zn ¨dn} sans commentaire. Malheureusement la fin de la ligne est défectueuse, ce qui nous empêche d’en proposer une lecture satisfaisante. L.2: mrt. Ce mot est déjà connu à Hatra, cf. no. 404:1. tg}. Ce mot est attesté ici pour la première fois à Hatra. On peut y reconnaître l’équivalent de tg} du palmyrénien6 et du syriaque taga 7 qui a le sens de ‘couronne’. Ce mot se présente en arabe sous la forme taj. D’autre part, mrt tg} semble être l’épithète de Nannay. drgyt}. Dans l’article cité plus haut, nous avions lu ce mot rbyt}, ‘l’intendant’. drgyt}, attesté ici pour la première fois à Hatra, est bien connu épigraphiquement dans le monde ouest-sémitique où drg} désigne ‘escalier’. Voir l’araméen d’Empire et le judéo-araméen8. Au syriaque le même mot traduit l’hébreu ma¨ alo† de 2 Rois 9:139. drgyt} est une forme adjectivale au féminin, ‘de l’escalier’, avec référence à la déesse Nannay. Le même mot se trouve encore une fois dans l’inscription no. 473 ici plus bas. Ìy}. est une forme du status constructus (en absence de la particule de liaison d- ou dy, ‘de’). Cette orthographie est bien attestée à Hatra, cf. nos 4: 5; 28: 4; 34:6, etc.10. 4

A.H. Al-Jadir, A Comparative Study of the Script, Language and Proper Names of the Old Syriac Inscriptions (University of Wales Ph.D. thesis, 1983), 407. 5 J. Stark, Personal Names in Palmyrene Inscriptions (Oxford 1971), 55, 117. 6 J. Hoftijzer et K. Jongeling, Dictionary of the North-West Semitic Inscriptions, vol. 2 (Leiden 1995), 1203. 7 J. Payne Smith, A Compendious Syriac Dictionary, (Oxford 1903), 604. 8 J. Hoftijzer et K. Jongeling, op. cit, vol. 1, 259–60. 9 R. Payne Smith, Thesaurus Syriacus, I (Oxford 1879–1901), cols 944–5. 10 Pour d’autres attestations, voir: R. Degen, ‘Die Genitivverbindung im Aramäischen der Hatra Inschriften’, Orientalia 36 (1967), 76–80. 39


†hmh. Il s’agit d’un nom nouveau attesté pour la première fois dans l’épigraphie nord- et ouest-sémitique. Ce mot est conservé en syriaque comme †uhma 11, en mandéen: †uhma12, et en arabe: †uhm13. slm. Ce mot n’était pas encore attesté à Hatra, mais on le trouve en araméen d’Empire, en nabatéen, en palmyrénien, en judéoaraméen14 et en syriaque. 464 Inscription incrustée de plombe gravée sur la face supérieure du seuil précédent.

bsnt 400+10+2 }qymyt }n} ¨qby kmr} dy nny rb} br syl} br ¨qby br syl} br bswn br tqwn En l’an 412 (=100/101) j’ai érigé, moi, ¨qby, le grand prêtre de Nannay, fils de syl} fils de ¨qby fils de syl} fils de bswn fils de tqwn

bswn: Il s’agit d’un nom propre nouveau à Hatra. Nous ne lui trouvons pas d’étymologie probante. tqwn: Nom propre nouveau à Hatra. Comme parallèles nous pouvons citer le nom propre assyrien Taquni15. 465 Inscription de trois lignes gravée sur le socle d’une statue trouvé dans la deuxième salle. La fin de l’inscription est incomplète. Longueur de ligne 65 cm.

11 12 13 14 15

J. Payne Smith, op. cit., 168. E. Drower et R. Macuch, A Mandaic Dictionary (Oxford 1963), 177. Ibn ManÂur, Lisan al-ärab, vol. 15 (Cairo 1882–91), 265. J. Hoftijzer et K. Jongeling, op. cit., vol. 2, 1147. K.L. Tallqvist, Assyrian Personal Names (Helsingfors 1914), 230. 40


1. 2. 3. 1. 2. 3.

bsnt 400+20+3 }qymtny nny kmr} bsnt 400+20+10+2 }qjmt }n} brnny br … En l’an 423 (=110/111) Nannay a érigé pour moi, prêtre. En l’an 432 (=120/121) j’ai érigé, moi, brnny fils de…

L.3: brnny. Le nom propre est bien connu à Hatra, cf. nos 1: 1; 2: 2; 106: 2b, etc. 466 Inscription de deux lignes gravée sur le socle d’une statue en marbre de Mossoul, trouvé près de l’angle sud-ouest de la deuxième salle. Longueur 68 cm et hauteur 15 cm. Le début et la fin de l’inscription sont perdus. Longueur de la première ligne 60 cm.

1. 2. 1. 2.

[Òlm} dy….] br brnny d}qym lnpsh ¨l Ìyyhy w¨l Ìy} … [Statue de …] fils de brnny qu’il a érigé pour lui-même pour sa vie et pour la vie de …

L.1: Òlm} dy. Nous proposons de restituer au début de la ligne la formule Òlm} dy, ‘statue de’. L’érection d’une statue pour lui-même est un phénomène connu à Hatra, voir les nos 20:2; 35:5. 467 Inscription de deux lignes gravée sur le socle d’une statue trouvé à gauche de l’entrée à la première salle. Le début de l’inscription est perdu. Longueur de la première ligne 40 cm.

41


1. mqymsms w¨bd¨gylw 2. bnyh 1. mqymsms et ¨bd¨gylw 2. ses fils

L.1: mqymsms. Ibrahim lit mqwmsms. mqymsms serait attesté deux fois aux nos 60 et 61. ¨bd¨gylw. Ce nom propre est très fréquent à Hatra, cf. nos 37: 7; 80: 1, 7, 8; 141: 1, etc. 468 I Inscription de deux lignes gravée sur le socle d’une statue en marbre de Mossoul. Longueur 51cm, largeur 34 cm et hauteur 13 cm. Longueur de l’inscription 48 cm.

1. mqymsms rbyt} 2. br wrwd rbyt} dmrtn 1. mqymsms, l’intendant, 2. fils de wrwd, l’intendant de Martan

L.1: wrwd. Ibrahim lit zr(d)zd(r). wrwd est bien connu à Hatra, cf. nos 60; 61; 102, etc. D’ailleurs mqymsms br wrwd rbyt} semble être le même personnage qui nous est connu par l’inscription no. 60. L.2: rbyt} dmrtn. Cette fonction nous est connue par le no. 364:4. 469 Inscription de quatre lignes gravée sur une frise de banc incomplet situé sur l’autel. La fin de l’inscription est endommagée. Longueur 67 cm et largeur 43 cm. 42


1. ¨bd wqrb gdyhb [br] }by rby[t}] 2. dbrmryn }lh} …… }l[kwd] 3. mlk} ¨l Ìyyh w¨l [Ìy}…] 4. [wmn dy] rÌym lh klh 1. A fait et a offert gdyhb [fils de] }by, l’intendant 2. de Barmaren le dieu …… }l[kwd] 3. le roi, pour sa vie et pour [la vie de …] 4. [et] tous ceux qui l’aiment.

L.1: ¨bd wqrb. L’interprétation d’Ibrahim comme des noms propres n’est pas adéquate dans le contexte précise de notre inscription. Ils s’agirent de deux verbes de troisième personne au passé qui signifient ‘a fait et a offert’. gdyhb. Nom propre attesté plusieurs fois à Hatra, cf. 4: 2–3; 13:2; 23:2, etc. }by. Ibrahim traduit ‘mon père’. gdyhb br }by est connu par le no. 415:3a. rby[t}] dbrmryn. La fonction rbyt} dbrmryn, ‘l’intendant de Barmaren’ est connue par le no. 224:5. L.2: }l[kwd]. On pourrait lire }lkwd, théophore connu à Hatra, cf. 79: 5, 6, 7, 8; 80:2, 9; 113:1, etc. L.3: ¨l [Ìy}.On pourrait restituer Ìy} et lire ¨l Ìy}, formule usuelle si largement attesté dans les dédicaces hatréennes. L.4: [wmn dy]. On pense dans le contexte hatréen de restituer wmn dy, ‘celui qui’, devant rÌym lh klh, formule bien connu à Hatra, nos 20: 4; 35: 8; 52: 5. 470

Inscription de deux lignes, peinte en encre rouge, gravée sur le socle d’une statue de femme en marbre de Mossoul, dont il ne reste que les pieds, trouvée dans la deuxième salle. Longueur 50 cm, largeur 35 cm et épaisseur 16 cm. 43


1. 2. 1. 2.

Òlmt} dy mlk} brt nbwdyn dmytt Statue de mlk} fille de nbwdyn qui est morte

L.1: Òlmt}. Le mot désignant la statue est au féminin pour une effigie de femme, cf. 5: 1; 30: 1; 34: 2, etc. mlk}. Nom propre nouveau à Hatra. Il est attesté à Palmyre comme nom masculin16. L’anthroponyme Malkah est encore en usage chez les populations de la langue arabe. L.2: nbwdyn. Nom propre connu par les inscriptions nos 279: 1; 310; 416: 3; 446: 1. 471 Inscription de trois lignes en caractères grands gravée sur une pièce de bloc de construction ou pièce d’une base de statue en marbre de Mossoul. Longueur 1,10 mètres et largeur 54 cm. Le début et la fin sont perdus. Longueur du reste 53 cm.

1. [Òlm} dy … br b]rsms kmr} 2. [dy …… Ìpy]zy mlk} …?? 3. ………………w? mlh

16

J. Stark, op. cit., 32, 95. 44


1. [Statue de … fils de b]rsms le prêtre 2. [de …… Ìpy]zy le roi …?? 3. ………………???

L.1: [Òlm} dy]. Ici on pourrait restituer tout simplement Òlm} dy, ‘statue de’, suivi d’un nom propre. b]rsms. Ibrahim propose de lire [¨b]d. brsms est attesté aux nos 393: c; 432: 1. L.2: Ìpy]zy. Nous sommes tenté de restituer Ìpy devant zy et lire Ìpyzy (232: 4b). On pourrait aussi proposer Ìp}zw (59; 118; 121; 126) ou rp}zw (332: 2). L’espace vide qui précédait ce nom propre était occupé par le nom du dédicant suivi de br, ‘fils de’. 472 Inscription de deux lignes gravée sur la base d’une statue représentant un personnage pieds-nus en marbre de Mossoul, trouvée dans la deuxième salle. Longueur 60 cm, largeur 30 cm et épaisseur 16 cm. Le début est abîmé. Longueur de la deuxième ligne 34 cm.

1. 2. 1. 2.

[bsnt] 400+20+20+20+10? Ò[lm} dy }qym] [qym]t kmr} dnny [En l’an] 470? (=158/159) [le statue qu’a érigé] qymt, le prêtre de Nannay

L.1: [bsnt]. L’espace libre devant ‘400+20+20+20+10’ nous permet de restituer la formule qui précéde le chiffre des centaines comme bsnt, ‘en l’an’. D’ailleurs, les inscriptions de Hatra datées jusqu’à ici ont atteint 31 textes. Ò[lm} dy }qym]. Nous proposons de restituer Ò[lm} dy }qym], seule possibilité valable dans le contexte hatréen. On doit avoir le verbe }qym, ‘a érigé’, suivi du nom Òlm}, ‘statue’. 45


L.2: [qym]t. Le nom propre doit probablement être complèté [qym]t : comparer no. 474: 1. Il s’agit du même personnage. Ce nom propre se retrouve à Hatra sous la forme qymty : no. 164: 117. 473 Inscription de deux lignes gravée sur la base d’une statue en calcaire, longueur 73 cm, largeur 57 cm et hauteur 15 cm. Les derniers mots de la deuxième ligne sont disparus. Longueur de la deuxième ligne 65 cm.

1. 2. 1. 2.

}qym Ìyws} kmr} dy nny drgyt} dy [Ì†r} ……] A érigé Ìyws}, le prêtre de Nannay de l’escalier de [Hatra……]

L.1: Ìyws}. Pour ce nom propre à Hatra, cf. nos 180, 407, 412:2a. L.2: [Ì†r}]. On peut restituer [Ì†r}] grâce à l’inscription 463: 2. Pour drgyt} voir le no. 463 ci-dessus. 474 Inscription de deux lignes, dont le début est perdu, gravée sur une pièce de linteau en calcaire (longueur 80 cm ≈ largeur 25 cm), située au dessus de colonnes supérieures de façade du temple, trouvée parmi les décombres à la côté est de l’édifice. Longueur de la première ligne 75 cm.

1. ……]b w¨l Ìy} qymt kmr} wbny ddh 2. ……] w}bgd }rdkl} 17 Pour d’autres attestations, voir: S. Abbadi, Die Personennamen der Inschriften aus Hatra (Hildesheim 1983), 159–60.

46


1. ……] et pour la vie de qymt, le prêtre, et ses cousins 2. ……] et }bgd, le maçon

Le début de l’inscription a été effacé, ce qui nous empêche d’en proposer une lecture soutenable. L.1: Ìy} : voir le no. 463. L.2: }bgd. Nom propre attesté sous deux autres formes }bygd (107: 1, 2; 245:1) et }bygyd (301: 1; 363: 1). 475 Inscription d’une ligne, dont la fin est perdue, gravée sur socle d’une statue en marbre de Mossoul, dont il reste que les pieds. Longueur 64 cm, largeur 38 cm et épaisseur 13 cm, trouvée à la côté droite de la deuxième salle. Elle était mise sur une base ou une étagère fixée au mur du temple.

1. nwdn br qymt kmr} dy [nny] 1. nwdn fils de qymt, le prêtre de [Nannay]

nwdn. Ce nom propre nous est connu ici pour la première fois à Hatra. Il nous mettrait en présence d’une forme fawäl. qymt. Ibrahim lit qwmt. [nny]. Nous proposons de restituer [nny] parce que les textes de ce groupe proviennent du sanctuaire de Nannay. 476 Inscription de trois lignes, d’après Ibrahim, était plus de trois lignes, gravée sur une frise ou base de statue en calcaire trouvée dans la deuxième salle. Longueur de la deuxième ligne 38 cm.

47


1. ……]mt nny 2. …… }p]Ìw }spp†} 3. …… n]sryhb mry} 1. ……]?? Nannay 2. …… }p]Ìw, le préfet 3. …… n]sryhb, le seigneur

L.2: }p]Ìw. Nous serions tenté de restituer }p devant Ìw et lire }pÌw (10: 3c, 3d; 390: 2). }spp†}. Ce nom de fonction militaire, qui n’est attesté qu’une seule fois à Hatra: no. 382, signifie ‘le chef de guerre’18. L.3: n]sryhb. En effet, nsryhb le seigneur (mry}) nous est connu par les nos 274: 3–4; 346: 3; 351: 1–2; 356: 2–3; 361: 2. 477 Inscription gravée sur le support de pierre situé à gauche de l’entrée moyenne de la deuxième salle. Longueur 1,10 mètres. La partie supérieure est érosive, et le reste de ces mots est très abîmé. Longueur de reste de la deuxième ligne 28 cm.

1. 2. 3. 1. 2. 3.

…… kmr} [dy …… ……¨l] Ìyyhy dy …… ……]¨ bnyhy …… le prêtre de…… …… pour la vie de …… …… ? ses fils.

478 Inscription d’une ligne, en long 40 cm, gravée sur la face d’une base de statue en marbre de Mossoul, trouvée dans la deuxième salle. Longueur 46 cm, largeur 40 cm et hauteur 10 cm. 18 Pour d’autres interprétations, voir: B. Aggoula, ‘Remarques sur les inscriptions de Hatra’, Syria 63 (1986), 368–70.

48


1. Òlmt} dy ¨ty 1. Statue de ¨ty

¨ty. Il s’agit d’un hypocoristique nouveau à Hatra, mais bien connu en palmyrénien comme un nom masculin.19 On le voit aussi comme un nom masculin dans le Bible: Ättay, I Chron. 2:3520. On peut citer pour ¨ty les noms hypocoristiques assyriens A-te-} et A-ti-21. 479 Inscription gravée au dos d’une statue féminine en marbre de Mossoul sans tête, trouvée à la proximité de l’entrée droite conduisant à la deuxième salle.

19

J. Stark, op. cit, 46, 108. Pour d’autres attestations, voir F. Brown, S.R. Driver and C.A. Briggs, A Hebrew and English Lexicon of the Old Testament (Oxford 1952), 774. 21 K.L. Tallqvist, op. cit, 47. 20

49


1. gnyt} 2. d(?)sms 3. glp} 4. ¨bdgd} 5. br 6. ¨g} dkyr 7. l†b 1. La génie. 2. de sms, 3. le sculpteur. 4. ¨bdgd} 5. fils de 6. ¨g}, qu’il soit commémoré 7. en bien.

L.1: gnyt}. Ce mot était auparavant attesté une seule fois, qualifiant hdyrt (no. 410: 1). L.2: Ibrahim lit d- avant sms, main nous ne pensons pas que çela est certain. En effet, il n’y a rien sur le fac-similé qui permet un dalat qui précède le mot sms. La ligne verticale n’est que bien vaguement dessiné. Comparer avec les autres dalat comme dans ¨bdgd} et dkyr. sms. Il y a tout lieu de penser que sms désigne un nom propre, et qu’il ne faut pas y voir un nom divin, comme Ibrahim. A mon sens, il s’agit peut-être du nom divin employé comme nom propre de personne. D’autres parts, notre traduction de les trois première lignes diffère de celle d’Ibrahim sur quelques points: il a considéré gnyt} la génie du dieu Samas et ce n’est pas raisonnable. De plus le mot glp}, ‘le sculpteur’, appartient à sms, parce que les fonctions ne précèdent jamais les nom propres dans les inscriptions de Hatra. L.4: ¨bdgd}. Nom propre connu à Hatra par no. 27: 7. L.6: ¨g}. Ce nom propre se rencontre plusieurs fois dans les inscriptions de Hatra: nos 5: 1, 2, 4; 13: 3; 48: 2, etc. 480 Inscription incisée sur le dallage de la troisième salle, en long 10 cm.

50


dkyr }sw l†b Que soit commémoré }sw en bien.

ˆsw. Ibrahim lit ˆsw/b, qui est traduit ‘l’exorciste’. La tentative de l’auteur de l’expliquer, ˆsw/b, par l’assyrien asipu, ‘qui fait des incantations’, ne peut pas se justifier facilement. Pour nous }sw est un nom propre nouveau. Il s’agit d’une variante de ˆs} attesté au no. 30: 2. Ce nom propre rappelle le nabatéen ˆsw 22. Address for correspondence: Dept of Syriac Language, Faculty of Languages, University of Baghdad, Iraq

Plate 1

22

S. Abbadi, op. cit, 82–3. 51


Plate 2

Plate 3

52

MEDICAL TERMINOLOGY IN THE HEBREW TRADITION: SHEM TOV BEN ISAAC, SEFER HA-SHIMMUSH, BOOK 30 GERRIT BOS MARTIN-BUBER-INSTITUT FÜR JUDAISTIK, UNIVERSITÄT ZU KÖLN

Abstract The terminology in medieval Hebrew medical literature is virtually lacking in standard Hebrew dictionaries. The only dictionary with a certain number of medical terms is that composed by Ben Yehuda. The secondary literature dealing with medieval Hebrew medical terminology is extremely limited and only covers a few medical terms. However, in recent years research into medieval Hebrew medical terminology has broken new ground. The translation technique of one major translator, namely ZeraÌyah Ben Isaac Ben She}altiel Îen, and his vocabulary, have been studied, and hitherto anonymous translations have been ascribed to him. Another medieval translator whose medical terminology is currently being analysed is Shem Tov Ben Isaac, the translator of al-Zahrawi’s Kitab al-taÒrif. As part of his translation he compiled two independent glossaries of medical synonyms. However, since these glossaries are only partial, and since they do not give these terms in a specific context, further analysis of the novel medical terminology is necessary to properly define his technical vocabulary. Thus, this study is devoted to an analysis of technical terminology in book 30, which deals with surgery. The terms, arranged alphabetically, are compared with those of the Arabic edition and English translation by Spink-Lewis, and with those used by Nathan ha-Me}ati, ZeraÌyah Îen, and Moses Ibn Tibbon.

Introduction The terminology in medieval Hebrew medical literature, both original works and translations, has been sorely neglected by modern research.1 Moreover, it is virtually lacking in the standard dictionaries for the Hebrew language, such as Ha-Millon he-Ìadash composed by 1 The following survey does not take into consideration the field of medieval plant names and remedies.

53

MEDICAL TERMINOLOGY IN THE HEBREW TRADITION

Abraham Even-Shoshan.2 The only medieval medical work to which Even-Shoshan refers is the Hebrew translation of Ibn Sina’s K. alQanun by Nathan ha-Me}ati;3 and even those references are indirect, having been borrowed from the dictionary composed by Ben Yehuda. Ben Yehuda’s dictionary is the only one which does contain a certain number of medical terms.4 However, it needs to be revised since it does not make use of even the limited sources registered in the introduction. The only dictionary exclusively devoted to medical terms, both medieval and modern, is that by Masie, entitled Dictionary of Medicine and Allied Sciences.5 However, like the dictionary by Ben Yehuda, it only makes occasional use of the sources listed in the introduction and all too infrequently differentiates between the various medieval translators. Further, since Masie’s work is alphabetised according to the Latin or English term, it cannot be consulted for checking on a Hebrew one. Beyond these general dictionaries, the secondary literature dealing with the subject of the medieval Hebrew medical terminology is extremely limited and generally only deals with a small number of medical terms.6 An early study is Joseph Hyrtl’s ‘Das Arabische und Hebräische in der Anatomie’ which was published in 1879.7 Although it was a pioneering work, it is of little use for scholars today for the following reasons: (1) It only deals with a small number of terms; (2) Most of these terms derive from one source only, namely a printed edition of the Hebrew translation of Ibn Sina’s K. al-Qanun fi al-†ibb by Nathan ha-Me}ati, which was published in Naples in 1491–2;8 (3) It does not specify particular translators. Thus, Hyrtl states about the almagabani (fauces), i.e. larynx, that it was translated in the Hebrew Avicenna as 2

The edition I consulted is that in five volumes, Jerusalem 2000. Ibid., vol. 1, p. 16. 4 Eliezer Ben Yehuda, Millon ha-Lashon ha-Ivrit. Thesaurus Totius Hebraitatis et Veteris et Recentioris. 17 vols. Berlin-Tel Aviv 1910–59. Repr. Tel Aviv 1948–59. 5 A.M. Masie, Dictionary of Medicine and Allied Sciences. Latin-English-Hebrew. Edited by S. Tchernichowski, Jerusalem 1934. 6 The following survey does not pretend to be complete or exhaustive. 7 Joseph Hyrtl, Das Arabische und Hebräische in der Anatomie (Vienna 1879). 8 For the Hebrew translations of the K. al-Qanun fi al-†ibb by Nathan ha-Me}ati, ZeraÌyah Ben Isaac Ben She}altiel Îen, Joshua Lorki and an anonymous translator, and the manuscripts of these translations see Chayyim Rabin, ‘Toledot Targum Sefer ha-Qanun le-Ivrit’, Melilah 3/4(1950), 132–46 (Manchester 1950). B. Richler, ‘Manuscripts of Avicenna’s Kanon in Hebrew translation; a revised and up-to-date list’, Koroth 30, vol. 8 (1982), 145–68; Lola Ferre, ‘Avicena Hebraico: La traducción del Canon de Medicina. The Hebrew translation of Avicenna’s Canon’, BIBLID 52 (2003), 163–82. 3

54


‫לוֹע‬ ַ from ‫לוּע‬ ַ (to devour). He fails to specify to which of the different

translations of Avicenna he refers. Shortly after Hyrtl, David Kaufmann’s monograph on the five senses, entitled ‘Die Sinne: Beiträge zur Geschichte der Physiologie und Psychologie im Mittelalter aus hebräischen und arabischen Quellen’ was published in Budapest 1884.9 The work is admittedly very useful even today; however, the medical terminology he discusses is limited to that of the physiology of the five senses. A more general study dealing with medieval medical terminology was undertaken by Hermann Kroner, a Rabbi practicing in Bopfingen, southern Germany, and published in 1921 under the title ‘Zur Terminologie der arabischen Medizin und zu ihrem zeitgenössischen hebräischen Ausdrucke’.10 However, it only discusses a relatively small number of terms since it is based primarily on the Hebrew translations of some of Maimonides’ minor works, and only distinguishes between two translators, ZeraÌyah Ben Isaac Ben She}altiel Îen and Moses Ibn Tibbon. It also suffers from several mistakes, sometimes resulting from the fact that Kroner only had access to corrupt manuscripts. Thus the term ‫ منهرم‬featured on p. 55 and translated as ‘Apathischer (Lässiger)’ should be corrected to ‫ منهزم‬meaning ‘defeated, vanquished’. And ibidem ‫( تسجيع‬Rhythmus (des Herzens) geben) should be read as ‫تشجيع‬ meaning ‘strengthening’ which was translated by Ibn Tibbon as ‫לתת‬ ‫גבורה‬. In 1945 Asher Goldstein published an article entitled Ha-Refu}ah we-ha-Lashon ha-Ivrit (Medicine and the Hebrew Language) in which he discusses different ways in which a novel medieval Hebrew medical terminology was created. However, he only discusses a few terms. Moreover, he seems to have had a certain bias against the Arabic medical terminology since he denies the important role it played in the formation of the medieval medical terminology in general, as he states explicitly: I allow myself to remind [the reader] of my warning published in ‘Ha-Rofe ha-Ivri’, I (1927), p. 18, about the danger of using the Arabic [for the innovation of Hebrew medical terms], for it does not have any scientific value in our days. And also in the Golden Age of Arabic 9

It was published as part of the Jahresbericht der Landes-Rabbinerschule in Budapest für das Schuljahr 1883–84 and reprinted in David Kaufmann, Die Spuren al-Ba†aljûsi’s, (Budapest 1880), and Studien über Salomon Ibn Gabirol (Budapest 1899). With an introduction by Louis Jacobs, Farnborough 1972. 10 H. Kroner, Zur Terminologie der arabischen Medizin und zu ihrem zeitgenössischen hebräischen Ausdrucke. An der Hand dreier medizinischer Abhandlungen des Maimonides (Berlin 1921). 55


medical science, in the Middle Ages, when it left its mark on medicine world-wide, [Hebrew] authors, doctors and translators of Arabic medical works were careful not to use Arabic medical terms. Only a few medical terms in Arabic infiltrated the Hebrew language.11

Accordingly when discussing the term ‫( חולי הפיל‬elephantiasis) as it features in the Sefer Åori ha-Guf by Nathan Ben Jo}el Falaquera,12 Goldstein simply states that it is a translation of Latin elephantiasis.13 However, since the work is based on Arabic sources it would have been more appropriate to suggest that it is a loan-translation from the Arabic ‫داء الفيل‬. Again, in an article entitled ‘MunaÌim refu}iyyim be-Ivrit mi-beÌinah historit’ (Medical nomenclature in Hebrew from an historical point of view), which was published in 1967, Goldstein’s discussion of some medical terms, like hernia, duodenum, cirrhosis, and their Hebrew counterparts, is primarily based on the Hebrew translation of Ibn Sina’s K. al-Qanun fi al-†ibb and some of Maimonides’ medical writings, such as the Pirkei Moshe (= Medical Aphorisms).14 However, he does not specify which of the Hebrew translations he used. For instance, when discussing the disease called ‘Hemorrhagia’ (i.e. bleeding) he mentions as Hebrew equivalents from Maimonides’ writings and Ibn Sina’s K. al-Qanun: ‫ תשפוכת‬,‫שטף דם‬ ‫ רעף דם‬,‫דם‬15 without providing sources. Following Goldstein, research into medieval Hebrew medical terminology was generally limited to a study of the anatomical terms featuring in Vesalius’ Tabulae Anatomicae Sex, which was published in 1538, and his De Humani Corporis Fabrica Libri Septem, commonly known as Fabrica and published in 1543. As Vesalius himself knew no Hebrew, the Hebrew (and Arabic) equivalents and their transliterations for the Tabulae derive from an anonymous friend; for the Fabrica, book one on osteology, he consulted his friend Lazarus de Frigeis of Venice who relied, in turn, on the Naples edition of Ibn Sina’s K. al-Qanun fi al-†ibb mentioned above. However, the Hebrew 11 Asher Goldstein, ‘Ha-Refu}ah we-ha-Lashon ha-Ivrit’, Harofé haivri. The Hebrew Medical Journal, 2 (1945), 88–96, p. 95 (trans. from the Hebrew by Gerrit Bos). 12 See Gerrit Bos and R. Fontaine, ‘Medico-philosophical controversies in Nathan b. Jo}el Falaquera, Sefer Zori ha-Guf’, Jewish Quarterly Review, 90 (1999), 27–60. 13 Goldstein (ibid., p. 91) actually reads: ‫אלשופנאטטיאזיס‬. 14 Asher Goldstein, ‘MunaÌim refu}iyyim be-Ivrit mi-beÌinah historit’, Koroth, vol. 4, 5–7 (1967), 452–62; vol. 4, 8–10 (1968), 625–36, and vol. 4, 11–12 (1968), 773–86. 15 Ibid., vol. 4, 5–7 (1967), 459.

56


terminology in the Fabrica edition of 1543 is very corrupt. De Frigeis’ knowledge of Hebrew and Arabic may have been sketchy; the transcriber and the typesetter, who were evidently unfamiliar with Hebrew, introduced many typographical errors.16 Mordecai Etziony studied the Hebrew material in two articles, published in 1945 and 1946. The first article deals with the Hebrew anatomical terminology featuring in Vesalius’ Tabulae,17 while the second article covers the Hebrew terms featuring in the Fabrica.18 The value of both articles lies primarily in their elucidation of the Hebrew terminology; the author does not analyse the medieval source(s), but only refers to Rabbinic parallels on the basis of Jastrow’s dictionary,19 as in the case of the term ‫( שלבים‬see below). In addition to Etzioni, Charles Singer and C. Rabin studied the Hebrew material in Vesalius’ Tabulae in their monograph entitled: ‘Prelude to Modern Science: Being a Discussion of the History, Sources and Circumstances of the “Tabulae Anatomicae Sex” of Vesalius’ which was published in 1946 as well.20 The authors trace the Semitic terminology in the Tabulae to contemporary oral usage in Arabic, Hebrew and Romance.21 They state explicitly that this work was not influenced by the printed Hebrew Avicenna.22 The study is valuable insofar as the authors extensively discuss these Hebrew terms, compare them with the terminology in the Fabrica and with that of the different medieval 16

See Andreas Vesalius, De humani corporis fabrica, translation and annotation by Daniel Garrison and Malcolm Hast (Vesalius.northwestern.edu), bk. 1, ch. 40, p. 166, esp. n. 5. See as well: Benjamin L. Gordon, ‘Review of Charles Singer and C. Rabin, Prelude to Modern Science: Being a Discussion of the History Sources and Circumstances of the ‘Tabulae Anatomicae Sex’ of Vesalius (Cambridge 1946)’, The Jewish Quarterly Review, New Ser., 38/2 (1947), 201– 3, p. 201. 17 Mordecai Etziony, ‘The Hebrew-Aramaic element in Vesalius’ Tabulae Anatomicae Sex. A Critical Analysis’, Bulletin of the History of Medicine 18 (1945), 413–24. 18 Mordecai Etziony, ‘The Hebrew-Aramaic element in Vesalius. A Critical Analysis’, Bulletin of the History of Medicine 20 (1946), 36–57. 19 Cf. the author’s statement in ‘The Hebrew-Aramaic element in Vesalius’, p. 38: ‘References are given only in the case of some Hebrew terms. The occurrence of those equivalents which are common use in old and modern Hebrew has not been traced to any particular place in literature. Practically all references quoted are those found in Jastrow’. 20 Charles Singer and C. Rabin, Prelude to Modern Science: Being a Discussion of the History Sources and Circumstances of the ‘Tabulae Anatomicae Sex’ of Vesalius (Cambridge 1946), esp. lxxv–lxxxvi. 21 Singer-Rabin, Prelude to Modern Science, lxxvii. 22 Singer-Rabin, ibid. 57


Hebrew translations of Ibn Sina’s K. al-Qanun fi al-†ibb. Take for example, their discussion on p. 24, n. 131, regarding the term ‫זרוע‬: ZEROA¨ is biblical, being mostly applied to the forearm. It is thus employed by Meathi and Lorci. The printed Hebrew edition of the Canon of Avicenna (1491), however, following Graciano (= ZeraÌyah Îen), used it for the upper arm. Doubtless on account of this confusion, Vesalius or Lazarus in the Fabrica give the phrase of the printed Hebrew Avicenna QENEH HA-ZEROA¨ = shaft of the upper arm.

One more study was devoted to the Hebrew elements in Vesalius’ Fabrica by Juan Jose Barcia Goyanes and published under the title ‘Los terminos osteologicos de la Fabrica y la evolucion del lenguaje anatomico Hebreo en la edad media’ in 1982.23 As the title indicates the author, unlike Etzioni, dealt with the origin of medieval Hebrew terminology featuring in the Fabrica, book one. Thus he consulted the Hebrew translations of Ibn Sina’s K. al-Qanun fi al-†ibb, by Nathan ha-Me}ati, ZeraÌyah Ben Isaac Ben She}altiel Îen, and Joshua Lorki. The following comparative table clearly shows the differences in approach by both scholars: Goyanes Ibid.: Suturae… ‫ שלבים‬scelauim: The proper transliteration is shlabim meaning mortised boards, steps of a ladder, plural of ‫ שלב‬shalv or shalav, of the root ‫שלב‬, joint.24 N (= Nathan): ‫ שלב‬y ‫ שלבים‬salaß, selaßîm. L (= Joshua Lorki): idem, Az.25: idem; Z: ‫ חוליה‬Ìulyâ, member, joint, vertebra. Av (= Avicenna)

Etzioni Fabrica, ed. 1543, p. 166, l. 20: Suturae… ‫ שלבים‬scelauim: The proper transliteration is shlabim meaning mortised boards, steps of a ladder, plural of ‫ שלב‬shalv or shalav, the derivative of the root ‫ שלב‬join, fit in with mortise and tenon. Compare the synonym with ‫ מחוברים‬mechubarim, Tabula VI, 1. For ‫ שלב‬see Jastrow, Taanith Yerushalmi, IV 68. 23

Juan Jose Barcia Goyanes, ‘Los terminos osteologicos de la Fabrica y la evolucion del lenguaje anatomico Hebreo en la edad media’. Sefarad 42 (1982), 299–326. 24 For this term see the extensive discussion in H. Rabin, ‘Toledot Targum Sefer ha-Qanun le-Ivrit’, Melilah 3/4 (1950), 132–46, p. 146. 25 Az. is the abbreviation used by Goyanes to refer to Azriel Ben Joseph of Gunzenhausen, who with his father Joseph Ben Jacob printed the Hebrew edition of the K. al-Qanun, Naples 1491–2. However, this edition is, according to Richler (Manuscripts of Avicenna’s Kanon, p. 148, based on the translation of Nathan for Books II–V, while the edition of Book I is based mainly on Lorki’s translation, which is a revision of that by Nathan. Thus, Az. is nothing else but N. In an earlier comparative study Rabin (Toledot Targum Sefer ha-Qanun le-Ivrit, p. 137) came to the conclusion that this edition is mainly based on Nathan’s translation, but that the editor sometimes prefers the version of Joshua Lorki or ZeraÌyah. 58


uses the terms ‫ درز‬and ‫دروز‬, which are transliterated by Alpago26 as adorem and feature in Vesalius as direzan, adorem. AH27 has: seleß, pl. selaßîm, pl. c. silßê. The term darz used by Avicenna is a translation of the Greek Åofß as used by Galen…

In a second article entitled ‘Medieval Hebrew Anatomical Names: A contribution to their history’, published in 1985,28 Goyanes studied nineteen (mainly anatomical) terms in the previously mentioned translations of Ibn Sina’s K. al-Qanun fi al-†ibb, and in Shem Tov Ben Isaac’s Hebrew translation of al-Razi’s K. al-Mansuri.29 The usefulness of the published studies is unfortunately very limited because they discuss so few terms. Thus, they do not contribute significantly to our knowledge of the medieval medical terminology in general and do not make it easier to read the pertinent texts. Moreover, none of them is devoted to the technical vocabulary or translation technique of one translator in particular. These eclectic studies do not help the reader of medieval Hebrew medical literature solve the most common problem, namely, that many of the extant medical texts do not name their translator. However, in recent years research into medieval Hebrew medical terminology has broken new ground. The translation technique of one major translator, namely ZeraÌyah Ben Isaac Ben She}altiel Îen, and his vocabulary have been studied by Gerrit Bos in his edition of ZeraÌyah’s translation of Aristotle’s De anima.30 ZeraÌyah was active in the city of Rome as a translator of philosophical and medical works from the Arabic into Hebrew in the last quarter of the thirteenth century.31 Amongst the medical works translated by him are Maimo26

I.e. Andrea Alpago (sixteenth century), who revised the Latin translation by Gerard the Cremona. 27 AH, i.e. The Academy of the Hebrew Language. 28 Juan Jose Barcia Goyanes, ‘Medieval Hebrew Anatomical Names: A contribution to their history’, Koroth, vol. 8, no. 11–12 (1985), 192–202. 29 For Shem Tov Ben Isaac see below. For this translation see Moritz Steinschneider, Die hebräischen Übersetzungen des Mittelalters (Berlin 1893), 725–6. 30 Aristotle’s De Anima. Translated into Hebrew by ZeraÌyah ben Isaac ben She}altiel Îen. Edited with Introduction and Linguistic Analysis (Leiden 1993), 23–43. 31 On ZeraÌyah see H. Vogelstein and P. Rieger, Geschichte der Juden in Rom, 2 vols. (Berlin 1895–6), vol. 1, 271–5, 409–18; M. Steinschneider, Die hebräischen Übersetzungen des Mittelalters und die Juden als Dolmetscher (Berlin 1893, repr. Graz 59


nides’ Medical Aphorisms and On Sexual Intercourse, Ibn Sina’s K. al-Qanun fi al-†ibb (the first two books only) and Galen’s De causis et symptomatibus and Katagenos. The newly available knowledge about ZeraÌyah’s translation technique and technical vocabulary has led to the ascription of four hitherto anonymous medical treatises to him. Mauro Zonta identified ZeraÌyah as the translator of Hippocrates’ De superfoetatione extant in an unique manuscript in Parma, Biblioteca Palatina,32 I identified ZeraÌyah as the translator of Maimonides’ On Hemorrhoids, extant in MS Parma 2642, De Rossi 354, Richler 1531;33 of On the Regimen of Health, extant in MS Paris BN hébr 1127 (a fragment only); and of On Poisons, extant in MSS Munich 43 and 280 (both fragmentary).34 I was able to identify ZeraÌyah as the author of these translations because of the critical editions I prepared of these works as part of the Maimonides’ project which aims at providing critical editions of his medical works in the original Arabic and medieval translations.35 In particular, the compilation of Arabic-Hebrew glossaries and separate alphabetical indices to the different Hebrew translations of Maimonides’ medical works proved to be very useful for the purpose of identification, as it provided me with the technical terminology typical of the major translators of these works, namely Moses Ibn Tibbon, Nathan ha-Me}ati and ZeraÌyah. Another medieval translator whose medical terminology is currently being analysed is Shem Tov Ben Isaac of Tortosa. While in Marseilles Shem Tov translated the famous medical encyclopaedia entitled Kitab al-taÒrif li-man {ajiza {an al-ta}lif (The Arrangement of Medical Knowledge for One Who is Not Able to Compile a Book for Himself), which was composed in the tenth century by the Andalu1956), 766; A. Ravitzky, Mishnato shel R. ZeraÌyah b. Isaac b. She}altiel Îen, (Doct. diss., Jerusalem 1977), 69–75; Bos, Aristotle’s De Anima, 1–4; M. Zonta, ‘A Hebrew translation of Hippocrates’ De superfoetatione: Historical Introduction and Critical Edition’, Aleph. Historical Studies in Science and Judaism, 3 (2003), 97–143, pp. 104–9. 32 See Zonta, ‘A Hebrew translation’. 33 See B. Richler, Hebrew Manuscripts in the Biblioteca Palatina in Parma. Catalogue. Palaegraphical und codicological descriptions (Jerusalem 2001). 34 Maimonides, On Poisons and the Protection against Lethal Drugs. A New Parallel Arabic-English Translation by Gerrit Bos with Critical Editions of medieval Hebrew translations and Latin translations by Gerrit Bos and Michael McVaugh. (Provo 2009). 35 Forthcoming in the series entitled ‘The Medical Works of Moses Maimonides’ published by Brigham Young University Press. 60


sian physician Abu l-Qasim Khalaf ibn Äbbas al-Zahrawi, known in the western world as Abulcasis.36 In addition to the Kitab al-taÒrif Shem Tov translated Abu Walid MuÌammad ibn Rushd’s Middle Commentary on Aristotle’s De Anima,37 Abu Bakr MuÌammad ibn Zakariyya al-Razi’s medical encyclopaedia K. al-ManÒuri,38 and Hippocrates’ Aphorisms with Palladius’ commentary.39 Shem Tov started his translation of the Kitab al-taÒrif, which he called Sefer ha-Shimmush, in 1254 and completed it at an unknown date. Instead of translating Zahrawi’s glossary of medical terms in book 29 Shem Tov compiled two independent lists of medical synonyms, the first in Hebrew-Arabic-Romance and the second in Romance-Arabic- and sometimes Hebrew. These lists are being edited, translated and annotated as part of a project initiated by Gerrit Bos and Guido Mensching.40 A striking feature of Shem Tov’s translation technique is that in several cases he created a novel41 Hebrew 36 For Shem Tov’s life and works see Gerrit Bos, ‘The Creation and Innovation of Medieval Hebrew medical terminology: Shem Tov Ben Isaac, Sefer ha-Shimmush’, in Anna Akasoy and Wim Raven (eds), Islamic Thought in the Middle Ages: Studies in text, Transmission and Translation in Honour of Hans Daiber (LeidenBoston 2008), 195–218. 37 Cf. Steinschneider, Die hebräischen Übersetzungen des Mittelalters und die Juden als Dolmetscher, p. 148; Averroës. Middle Commentary on Aristotle’s De Anima. A Critical Edition of the Arabic Text with English Translation, Notes, and Introduction by Alfred L. Ivry (Provo, Utah 2002), xxviii–xxix, 150. n. 69. 38 Cf. Steinschneider, Die hebräischen Übersetzungen des Mittelalters, 725–6. 39 His commentary is no longer extant in Greek, but it has recently been rediscovered by Hinrich Biesterfeldt and Y. Tzvi Langermann, who hope to publish soon a preliminary study of Palladius’ commentary, to be followed by a full edition and analysis. 40 The project is dedicated to the edition and the analysis of various unedited scientific texts written in Middle Hebrew that belong to the area of medico-botanical literature. Within this project the Cologne group, consisting of Gerrit Bos and Martina Hussein, is responsible for the Hebrew-Arabic linguistic material, while the Berlin group, consisting of Guido Mensching and Frank Savelsberg, is in charge of the LatinRomance material. First results of the research carried out in the context of the project are: Bos-Mensching, ‘Shem Tov Ben Isaac, Glossary of Botanical Terms, nrs 1–18’, Jewish Quarterly Review, 92 (2001), 1–20; Bos-Mensching, ‘Hebrew Medical Synonym Literature: Romance and Latin Terms and their Identification’; Aleph, Historical Studies in Science & Judaism, vol. 5 (2005), 11–53; Bos-Mensching, ‘A 15th Century medico-botanical synonym list (Ibero-Romance-Arabic) in Hebrew characters’, in Panace@), vol. VII, no. 24 (December 2006); see: http://www.medtrad.org/panacea/Indice General/n24_tribunahistorica-bos.mensching.pdf; Bos, The Creation and Innovation of Medieval Hebrew medical terminology. 41 With novel terms I mean one of three things: 1. terms that do not feature in the current dictionaries at all; 2. terms which can be found in current dictionaries but not in the sense they have in our text; 3. terms which can be found in current dictionaries but are not registered as medieval.

61


medical terminology which was, in some cases adopted by subsequent authors such as Nathan ha-Me}ati and ZeraÌyah Îen. An example is the entry He 11 in our edition of the first glossary: ‫המעדת המעים ב“ה‬ ‫זלק אלאמעא‬. The Hebrew HM{DT HM{YM, which is not attested in secondary literature, may have been coined by Shem Tov as a Hebrew loan translation of the Arabic zalaq al-amä} ‘Dysenteria spuria’. The same Hebrew term features subsequently in Nathan’s and ZeraÌyah’s Hebrew translations of Maimonides’ Medical Aphorisms (XXII, 36; XXIII, 80, 90, 93, 94). However, since these glossaries do not cover all the technical terms featuring in the Sefer ha-Shimmush, and since they do not give these terms in a specific context, further analysis and discussion of the novel medical terminology employed by the author is necessary to facilitate the reading of his translations in general, to ensure recognition of his technical terminology in future dictionaries of the Hebrew language, and to define properly his technical vocabulary. With this end in view the following study is devoted to an analysis of a selection of the technical terminology of book 30, which deals with surgery, and was by far the most popular and most influential part of this vast medical encyclopaedia. Translated into Latin by Gerard de Cremona in Toledo in the second half of the twelfth century and into Occitan in the fourteenth century, it was a major source for the European treatises on surgery composed subsequently, foremost that by Guy de Chauliac (d. 1368) who quotes it no less than 157 times. Thus it played a significant role in the development of the art of surgery in Europe.42 While the original Arabic text of book 30 has been published in a critical edition and English translation by Spink-Lewis43 and the Occitan translation has been edited by Grimaud-Lafont,44 the Hebrew text is still unedited. The analysis of the technical medical terminology of the Hebrew text is based on MS Paris, BN héb. 1163 which is the only manuscript to have preserved book 30 and which was copied in a sephardic script in the fourteenth century.45 The text appears in double columns 42

Cf. Danielle Jacquart and Françoise Micheau, La médicine arabe et l’occident médiéval (Paris 1990), 150–1. 43 M.S. Spink and G. Lewis, Albucasis. On Surgery and Instruments. A definitive edition of the Arabic text with English translation and commentary (Publications of the Wellcome Institute for the History of Medicine. New Series. Volume XII, London 1973). 44 J. Grimaud and R. Lafont, La chirurgie d’Albucasis, texte occitan du XIVe siècle (Montpellier 1988). 45 For the manuscript cf. H. Zotenberg (ed.), Catalogues des Manuscrits Hébreux et Samaritains de la Bibliothèque Nationale (Paris 1866). 62


on fols. 201a–239a and has been illustrated with many drawings of the surgical instruments recommended by the author, copied from an Arabic Vorlage. In my study the terms, arranged alphabetically, are compared throughout with those of the Arabic edition and English translation by Spink-Lewis. I will refer to parallel terminology used by other major translators, namely Nathan ha-Me}ati, ZeraÌyah Îen, and Moses Ibn Tibbon. Of these translators both Nathan and ZeraÌyah were active at a later date than Shem Tov Ben Isaac; Nathan worked in Rome between 1279 and 1283 and ZeraÌyah worked in the same city between 1279 and 1291. It is possible that both used part of the novel terminology invented by Shem Tov Ben Isaac. In the case of Moses Ibn Tibbon, however, it is hard to determine who influenced whom, as he was active as a translator between 1240 and 1283 and some of his translations are earlier than those by Shem Tov.46 The sources consulted for these comparisons are the translations of Maimonides’ medical works which are being published as part of the Maimonides’ Medical Works project mentioned above. I also consulted Moses Ibn Tibbon’s Hebrew translation of the medical encyclopaedia Zad al-musafir wa-qut al-Ìadir (Provisions for the Traveller and Nourishment for the Sedentary), composed by Abu Ja¨far AÌmad ibn Ibrahim ibn Abi Khalid ibn al-Jazzar (tenth century), a practising physician from Qayrawan.47 This translation was composed in 1259 and is still unedited; of the many manuscripts testifying to its popularity in Jewish circles I consulted MS Munich 19 which was copied in 1552.48 In addition to these primary sources I refer to secondary ones, primarily Masie, Dictionary of Medicine and Allied Sciences and Ben Yehuda, Millon ha-Lashon ha-Ivrit for Nathan’s Hebrew translation of Ibn Sina’s K. al-Qanun prepared in Rome in 1279. The dictionaries consulted for this study are those by Ben Yehuda, Even Shoshan, Masie, and the online Dictionary of the Academy of the Hebrew Language. The Bar-Ilan Responsa CD-Rom is used for attestations from sources which are primarily halakhic. 46 For an extensive discussion of the question of the authors consulted by Shem Tov and the authors influenced by him see Gerrit Bos, The Creation and Innovation of Medieval Hebrew medical terminology: Shem Tov Ben Isaac, Sefer ha-Shimmush (forthcoming). 47 For Ibn al-Jazzar’s bio- and bibliographical data see Bos, Ibn al-Jazzar on Sexual Diseases and their Treatment, pp. 5–7, and the relevant literature mentioned there. See as well Bos, Ibn al-Jazzar on Skin Diseases and Other Afflictions of the Outer Part of the Body. Critical edition of the Arabic text, Hebrew translation by Moses Ibn Tibbon and modern English translation (forthcoming). 48 See M. Steinschneider, Die hebräischen Handschriften der K. Hof- und Staatsbibliothek in München, 2nd rev. enl. ed., (Munich 1895).

63


A final introductory note concerns the faithfulness of the translator in adhering to the original text. Hebrew translations of medical texts in general closely follow the original text. Only rarely does one find additions of a personal nature. A remarkable example of such a personal addition and witness to the religious identity of the translator can be found in Book 1, ch. 47 where Shem Tov translates the Arabic ‫( عند نهاية الشعر‬about the hairline) as ‫במקום הנחת תפלין‬ (where one places the Tefillin).49 List of terms ‫ אבר‬:‫ = האבר המושל‬Arab. ‫العضو الرئيس‬: ‘the major organ’; cf. II:40 (SP fol. 215a; SL 321,13): ‫ויהיה הצמח לפעמים קרוב מאבר מושל שהמאחר לדקרו וממתין עד‬ ‫( שיתבשל יזיק לאותו האבר המושל‬Or if the tumour be near a major organ, if you delay the opening of it until it ripens you will damage that major organ)50. In addition to ‫האבר המושל‬, Shem Tov has ‫( האבר הראשי‬II:86; SP fol. 225a; SL 555, 24), and ‫( אחד מן האיברין השרים‬for ‫;عضو رئيس‬ SP fol. 225b; SL 557, l. 52). N translates Arabic ‫ العضو الرئيس‬as ‫האבר‬ ‫ הראשי‬and Z as ‫( האבר השרי‬MA 25:70). M (BIZ 9:8) translates the plural ‫ الأعضاء الرئيسة‬as ‫האברים הראשים‬. ‫ אגודה‬:‫אגודות‬. This term features with the following meanings: 1. Arab. ‫عقد‬: ‘nodules’; cf. II:27 (SP fol. 202b; SL 269, 1): ‫בהוצאת האגודות אשר תקרנה‬ ‫( בשפתים‬on the extraction of nodules occurring on the lip). The same Arabic term is translated as ‫ קשרים‬by M (BIZ 23:1); 2. Arab. ‫تعقّد‬: ‘callus’; cf. III:20 (SP fol. 236b; SL 781,1): ‫ברפואת האגודות המתהוות בעקבות קצת‬ ‫( השברים‬On the treatment of the callus that remains from a fracture); 3. Arab. ‫غدد‬: ‘buboes’; cf. L 2231, s.v. ‫غدّة‬: ‘A ganglion; i.e. any hard lump in the tendinous parts’; cf. II:36 (SP fol. 214a; SL 301, 3): ‫לפעמים‬ ‫( תקרנה בגרון אגודות דומות אל האגודות אשר תקרנה מחוץ‬Sometimes there occur in the throat buboes [called ‘tonsils’], which resemble the buboes occurring externally. The Arabic ‫ غدد‬is translated by N as ‫ גדרים‬or ‫גידים‬ and by Z as ‫ גלנדולי‬or ‫( גרנגולי‬MA 1:8); 4. Arab. ‫سلع‬: ‘cysts’ (= ‫אגודות‬ ‫ ;)הבשר‬cf. II:41 (SP fol. 215a; SL 329, 3): ‫יקרו בעור הראש צמחים קטנים‬ ‫( והם ממיני האגודות יקיפום קרומות הם להם נאדות דומים לזפק התרנגולת‬Small swellings form in the scalp, which are of the various kinds of cyst, contained within membranes which form a capsule to them like the crop of a chicken); see SG Alef 38. -: ‫ = אגודות הבשר‬Arab. ‫سلع‬: ‘cysts’; cf. 2:45 (SP fol. 216a; SL 343, 1): ‫שער‬ ‫( מ“ה בבקיעה על מיני אגודות הבשר‬Chapter forty-five. On incision for various kinds of cyst). Cf. SG Alef 38 and below s.v. ‫תלולית‬. ‫ ִאכּוּל‬:‫ = איכול‬I. Arab. ‫آكلة‬: ‘gangrene’ (cf. D 1:31, s.v. ‫أكلة‬: ‘gangrène, chancre, ulcère’; cf. I:52 (SP fol. 208a; SL 155, 2–3): ‫האיכול אמנם הוא הפסד מתפשט‬ ‫( באבר ויאכלנו כאכילת האש העצים היבשים‬Gangrene is a creeping corruption 49 50

See below s.v. ‫חלל העורף‬. The English translation is that by Spink-Lewis unless indicated otherwise. 64


of a limb, consuming it as fire consumes dry wood); II. Arab. ‫‘ أكال‬irritation’; cf. II:96 (SP fol. 231a; SL 657, 20):‫ומהם מי שימצא חכוך בפניו ובמצחו‬ ‫( וחשיכות ואיכול בעיניו‬some have an itching in their face and forehead and a dimness and irritation in their eyes). Hebrew ‫ איכול‬is only attested in Ben Yehuda (BM 205) as occurring in medieval literature in the sense of ‘itching’. It features in N for Arab. ‫( أكلة‬corrosion/ canker/ cankerous sore), whereas Z has ‫ אוכלת‬or ‫( חולי‬MA 3:109; 7:60; 9:105; 12:32; 16:7). See as well MD 188, s.v. ‘corrosion’. ‫ = אליה‬Arab. ‫ألية‬: ‘buttock’; cf. 4:14 (SP fol. 235b; SL 757, 14–16): ‫אך הדוקו‬ ‫ראוי להיותו על זה התאר והוא לכרוך על השבר במצנפת גסה שלש כריכות או ארבע‬ ‫( ולהעדיף ממנה מותר ואחר כך לקפול השוק עד שוב העקב אצל שרש האליה‬As to

the bandaging, you should wrap a firm broad sash two or three times round the fracture [of the femur], leaving some over; then bind the leg up so that the heel is brought to reach the root of the buttock). Hebrew ‫אליה‬, i.e. fat tail (BM 241) in the sense of ‘buttock’ is a non-attested semantic borrowing from the Arab. ‫ ألية‬which has both meanings (cf. L 87). ‫ א ֶֹמץ‬:‫ = אומץ המעים‬Arab. ‫اعتقال الطبع‬: ‘constipation’; cf. II:81 (SP fol. 223b; SL 515, 31–2): ‫( והפלחים יקרו הרבה מצד נגוב הצואה ואומץ המעים‬Fissures often arise from dryness of the faeces and from constipation). Hebrew ‫ אומץ המעים‬is a non-attested term derived from ‫( לאמץ את המעים‬to constipate) which is attested in Maimonides, Hilkhot Deöt 6:4; cf. BM 287– 8. N and Z translate the Arabic synonym ‫ احتباس البطن‬as ‫עוצר הבטן‬ (MA 9:55). M translates the Arabic synonym ‫ احتباس الطبيعة‬as ‫עוצר הטבע‬ (MZ fol. 91b). ‫ ִא ָסּר‬or ‫ = ֱא ָסר‬Arab. ‫رباط‬: ‘ligament’; cf. II:88 (SP fol. 226b; SL 583, 2–5): ‫כשנתחדש צמח במקצת האיברין הבשריים וארך זמנו עד שקבץ מוגלא ואחר נתבקע‬ ‫או שדקרוהו ויצא כל מה שהיה בו מן המוגלא ונשאר המקום ריקן ככיס והעור אשר‬ ‫עליו כבגד שכבר נתדקדק ולא הגיע אל ההפסד לגמרי לעשות רושם בעצם ולא בעצב‬ ‫( ולא באסר‬When an abscess occurs in any of the fleshy parts and becomes

chronic and eventually collects pus, and then it breaks open or is perforated, and all the contained matter comes out and the site is left hollow, like a vessel, and the overlying skin thinned like a rag, but the suppuration has not gone so far as to involve bone or tendon or ligament). Hebrew or Aramaic ‫ אסר‬is only attested in Rabbinic literature in the sense of ‘band, chain, vow of abstinence’; cf. JD 57, and DA 32 for the Aramaic term). In addition to ‫ אסר‬Shem Tov uses ‫ קשירה‬for ‫;رباط‬ cf. II:86 (SP fol. 225b; SL 563, 103). N and Z translate Arab. ‫ رباط‬as ‫ קשירה‬,‫ קשורים‬,‫( קשור‬MA 1:8, 9-11; 3:21, 52; 7:33; 15:29, 40, 47, 66, 69, 70; 23:18; 25:36, 51, and M as ‫( קשרים‬BIZ 15:5). ‫ אפר‬:‫ = אפר העינים‬Arab. ‫رمد‬: ‘ophthalmia’; cf. II:95 (SP fol. 228b; SL 629, 34–6): ‫והשני ורידין אשר בשתי הרקות תועלת הקזתם לכאב הצלחתא הישנה‬ ‫והכאב הקשה ואל אפר העינים המתמיד והגרת המותרות הנשפכות החדות אל העינים‬

(The section of the two arteries in the temples gives relief for chronic migraine and severe headache and constant ophthalmia and the flow of acrid superfluities into the eyes). Hebrew ‫אפר העינים‬, a loan translation of the Arab. ‫رمد‬, is not attested in the current dictionaries. Both N and Z 65


did not have a Hebrew equivalent for the Arabic term, as N transcribed it as ‫רמד‬, and Z used both ‫ רמד‬and the Romance equivalent ‫ לגניא‬or ‫ַלגְ נְ יַ יא‬ (MA 9:31; 12:22; 19:16; 22:39; 23:70; 24:20). See as well KS 114–15. ‫ אצבע‬:‫ = אצבע קטנה‬Arab. ‫خنصر‬: I. ‘little finger’; cf. IV:12 (SP fol. 235a: SL 747, 13–14):‫( ולהיות האצבע הקטנה למטה מכל האצבעות‬and the little finger downmost); II. ‘little toe’; cf. II:95 (SP fol. 230b; SL 655, 284–5): ‫ואם לא ימצא ולא יראה לחוש כלל ראוי להקיז קצת סעיפיו והם הנראים בגב הרגל בין‬ ‫( אצבע קטנה והשני לו‬and if you do not find it [i.e. the sciatic vein] or it is

in no way palpable, then venesect one of its branches, which will be seen on the surface of the foot between51 the little toe and the fourth toe). Hebrew ‫ אצבע קטנה‬is not attested in the current dictionaries. ‫ = אצילה‬Arab. ‫إبط‬: ‘axilla’; cf. I:25 (SP fol. 205b; SL 77, 1): ‫שער כה‘ בכוית‬ ‫( האצילה‬Chapter twenty-five. On cauterization of the axilla). The Hebrew term features in the Bible in the sense of 1. joint and 2. cubit (cf. KB 81–2). Ben Yehuda also gives several references to its occurrence in medieval literature in the last sense only (BM 368). N translates the Arabic ‫إبطان‬ as ‫ אצילים‬and Z as ‫( שחי‬MA 7:21; 10:15). M (MZ fol. 90a) translates Arabic ‫( إبطان‬BZ 147, l. 678) as ‫השחיים‬. ‫ = אצילי‬Arab. ‫ٕبطي‬ ّ ‫ا‬: ‘axillary’; cf. II:95 (SP fol. 228b; SL 627, 13–14):

‫והבאסליק הוא בזרוע למטה מן הגיד האמצעי לעמת הלב ויקרא גם כן האצילי‬ ‫( ויקראוהו עם הארץ גיד הבטן‬then the basilic vein; this is the one situated

on the inner side and is termed also the axillary vein, but popularly it is called the ‘belly-vein’). Hebrew ‫ אצילי‬is attested in BM 369 as featuring in Nathan’s translation of Ibn Sina’s K. al-Qanun. ‫ אשך‬:‫ = האשך הבשרי‬Arab. ‫الأدرة اللحم ّية‬: ‘a fleshy hernia’; II:63 (SP fol. 219b; SL 435, 1):‫( שער ס“ג בבקוע על האשך הבשרי ורפואתו‬Chapter sixty-three. On cutting for a fleshy hernia and its treatment). -: ‫ = האשך הזמורי‬Arab. ‫الأدرة التي مع دالية‬: ‘hernia with varix’; II:64 (SP fol. 220a; SL 439,1): ‫( שער ס“ד ברפואת האשך הזמורי ר“ל אשר יהיה עם זמורה‬Chapter sixty-four. On the treatment of hernia with varix). Cf. entry ‫ זמורה‬below. -: ‫ = האשך המימי‬Arab. ‫الأدرة المائية‬: ‘a watery hernia’; cf. II:62 (SP fol. 219a; SL 425, 2). -: ‫ = האשך המעיי‬Arab. ‫الأدرة المعائ ّية‬: “intestinal hernia”; II:65 (SP fol. 220a; SL 441,1). -: ‫ = האשך הרוחיי‬Arab. ‫الأدرة الريح ّية‬: ‘flatulent hernia’; II:66 (SP fol. 220a; SL 447,1). The term ‫ אשך‬only features in the current dictionaries in the sense of ‘testicle’; cf. BM 416. Another term for ‘hernia’ used by Shem Tov is ‫( פיתקא‬see below). Both N and Z have ‫ בקיעה‬for Arabic ‫( أدرة‬MA 9:123). Masie (MD 351) mentions the following synonyms for ‘hernia’: ,‫שבר‬ ‫ בקיע‬,‫ פרץ‬,‫שברון‬. ‫ בדק‬:‫ = בדק עצמו‬Arab. ‫تب ّرز‬: ‘to open one’s bowel’; i.e. to defecate; cf. II:81 (SP fol. 2223a; SL 513, 9–11): ‫ורפואת הטחורים אשר יהיו מבפנים לצוות אל‬ 51 ‘between the little toe and the fourth toe’: ‘towards the fourth toe and the little toe’ SL.

66


‫( החולה לבדוק עצמו ולהרגיל העיצום עד שיבלוט פי הטבעת ויראו היבלות‬The treat-

ment of internal piles: bid the patient open his bowel and bear down until the anus opens out and the swellings are disclosed to you). ‫בדק עצמו‬ is a non-attested Hebrew term coined after the Aramaic ‫;בדיק נפשיה‬ cf. SD 187. N translates Arab. ‫ تب ّرز‬as ‫ הוציא בציאה‬and Z as ‫יצא מנקב פי‬ ‫( הטבעת‬MA 9:64). ‫ בית‬:‫ = בית הפרשות‬Arab. ‫فضاء‬: ‘perineum’; cf. II:80 (SP fol. 220a; SL 503, 8–9): ‫ולפעמים יהיה מהם טחורים כשיהיו בבית הפרשות מפולשים אל כיס מקוה‬ ‫( המים ואל מעבר השתן‬Sometimes also there are fistulae occurring in the perineum which penetrate to the urinary bladder and to the urethra). Hebrew ‫ בית הפרשות‬is not mentioned in the current dictionaries. It is attested in BIR in a quotation from Sefer ha-Orah (Part 2, [67], Din Niqqur, beginning with: Heshiv R. (ed. S. Buber, 1905): ‫בית הפרשות שלו‬ ‫( שקוראים קודי‘‘ל‬Its perineum, that is its QWDYL).52 In addition to this term Shem Tov uses ‫ ריקות‬to render Arab. ‫ ;فضاء‬cf. below. ‫ = בליטה‬Arab. ‫نتو‬: ‘protuberance’; cf. II:16 (SP fol. 211b; SL 233, 31–3): ‫אמנם בליטת בשר ראש העין אם תהיה הבליטה מזקת העין נזק מכוער ראוי לתלות‬ ‫הבליטה ההיא בחכה ולחתוך ממנה קצתה מבלי רבוי חתוך פן תתחדש הגרת הדמעות‬

(As for a protuberance of flesh in the angle [of the eye], if it causes great hurt pick it up with a hook and cut part of it away, with not too big an incision lest there be a flux of tears). Hebrew ‫ בליטה‬only features in a medical sense in BM 549 in a quotation from Meir Aldabi, Shevilei Emunah which was completed in 1360. ‫ בעבע‬:‫ = נתבעבע הצמח‬Arab. ‫تق ّرح‬: ‘to break out in open ulceration’; cf. Introduction (SP fol. 201b; SL 5, 32–3): ‫וראיתי רופא אחר דקר צמח סרטני ונתבעבע‬ ‫( הצמח אחר ימים‬And I saw another doctor incise a malignant tumour; after some days the place broke out in open ulceration). The root ‫בעבע‬ only features in the dictionaries in the sense of ‘to bubble’; cf. BM 569. N translates Arabic ‫ تق ّرح‬as ‫ התחבל‬or ‫ השחין‬and Z as ‫ התחבל‬,‫ חבל‬or ‫התנגע‬ (MA 6:72; 23:46); M (BIZ 13:3) translates the Arabic ‫( أقرح‬to ulcerate) as ‫לנגע‬. See also ‫ צמח‬below. ‫ = ֶבּ ֶתק‬Arab. ‫فتق‬: ‘opening; rupture’; cf. II:65 (SP fol. 220a; SL 441, 2–4): ‫ התחדש זה האשך יהיה מבקוע יקרה בקרום הנמתח‬.‫שער ס“ה ברפואת האשך המעיי‬ ‫על הבטן לעמת עיקרי הירכים וישפך המעי מן הבתק ההוא אל אחד מן הבצים‬

(Chapter sixty-five. On the treatment of intestinal hernia. This hernia is due to a split occurring in the membrane stretched from the hypogastrium over the belly in the region of the groin. Through this opening the bowel descends upon one of the testes). Hebrew ‫ ֶבּ ֶתק‬is only attested in Ma}agarim (

Journal of Semitic Studies, Volume 55, Number 1, Spring 2010

Journal of Semitic Studies, Volume 55, Number 2, Autumn 2010

Journal of Semitic Studies, Volume 54, Number 1, Spring 2009

Journal of Semitic Studies, Volume 54, Number 2, Autumn 2009

Asian Music - Journal Volume 42, Number 1, Winter - Spring 2011

Jewish Social Studies - Volume 16, Number 3, Spring Summer 2010

New Testament Studies, Volume 55, Number 1 (January 2009)

Journal of Biblical Literature, Vol. 129, no. 1 (Spring 2010)

Journal of Cuneiform Studies (JCS 55) 2003

JWSR Volume XVI, Number 1, 2010

East-West Journal of Mathematics (Volume 1, Number 1)

New Testament Studies, Volume 56, Number 1 (January 2010)

Canadian Journal of Philosophy, Volume 40, Number 4, December 2010

New Testament Studies, Volume 55, Number 2 (April 2009)

New Testament Studies, Volume 55, Number 3 (July 2009)

New Testament Studies, Volume 55, Number 4 (October 2009)

Journal of Workplace Learning, Volume 17 Number 1 2 2005

Asian Music - Journal Volume 41, Number 2, Summer - Fall 2010

Journal of Forensic Sciences vol 55, issue 6, Nov 2010

Journal of Forensic Sciences vol 55, issue 5, Sep 2010

Journal of Chiropractic Medicine- Volume 9(1); March 2010

Journal of Modern Greek Studies Volume 28, Number 2, October 2010

Journal of Cuneiform Studies (JCS 62) 2010

October, journal No.1 Spring (1976)

Textual Practice, Volume 6, Number 1 (Textual Practice Journal)

Textual Practice, Volume 6, Number 1 (Textual Practice Journal)

Cell Journal, Volume 144, Number 1, January 7, 2011

2DArtist Issue 55 July 2010 issue 55

New Testament Studies, Volume 56, Number 4 (October 2010)

Cultural Studies: Volume 3, Issue 1 (Cultural Studies Journal)

New Testament Studies, Volume 56, Number 2 (April 2010)

Journal of Semitic Studies, Volume 55, Number 1, Spring 2010

Journal of Semitic Studies, Volume 55, Number 2, Autumn 2010

Journal of Semitic Studies, Volume 54, Number 1, Spring 2009

Journal of Semitic Studies, Volume 54, Number 2, Autumn 2009

Asian Music - Journal Volume 42, Number 1, Winter - Spring 2011

Jewish Social Studies - Volume 16, Number 3, Spring Summer 2010

New Testament Studies, Volume 55, Number 1 (January 2009)

Journal of Biblical Literature, Vol. 129, no. 1 (Spring 2010)

Journal of Cuneiform Studies (JCS 55) 2003

JWSR Volume XVI, Number 1, 2010

East-West Journal of Mathematics (Volume 1, Number 1)

New Testament Studies, Volume 56, Number 1 (January 2010)

Canadian Journal of Philosophy, Volume 40, Number 4, December 2010

New Testament Studies, Volume 55, Number 2 (April 2009)

New Testament Studies, Volume 55, Number 3 (July 2009)

New Testament Studies, Volume 55, Number 4 (October 2009)

Journal of Workplace Learning, Volume 17 Number 1 2 2005

Asian Music - Journal Volume 41, Number 2, Summer - Fall 2010

Journal of Forensic Sciences vol 55, issue 6, Nov 2010

Journal of Forensic Sciences vol 55, issue 5, Sep 2010

Journal of Chiropractic Medicine- Volume 9(1); March 2010

Journal of Modern Greek Studies Volume 28, Number 2, October 2010

Journal of Cuneiform Studies (JCS 62) 2010

October, journal No.1 Spring (1976)

Textual Practice, Volume 6, Number 1 (Textual Practice Journal)

Textual Practice, Volume 6, Number 1 (Textual Practice Journal)

Cell Journal, Volume 144, Number 1, January 7, 2011

2DArtist Issue 55 July 2010 issue 55

New Testament Studies, Volume 56, Number 4 (October 2010)

Cultural Studies: Volume 3, Issue 1 (Cultural Studies Journal)

New Testament Studies, Volume 56, Number 2 (April 2010)

Recommend Documents