Babel und Bibel 6: Annual of Ancient Near Eastern, Old Testament, and Semitic Studies 9781575066653

Citation preview

Babel und Bibel 6

Volume XLIII

Российский государственный гуманитарный университет

Выпуск XLIII

Вавилон и Библия 6

Древнеближневосточные, библейские и семитологические исследования

Papers of the Institute of Oriental and Classical Studies Volume XLIII

Babel und Bibel 6 Annual of Ancient Near Eastern, Old Testament, and Semitic Studies

Edited by

L. Kogan, N. Koslova, S. Loesov, and S. Tishchenko

Published for the Russian State University for the Humanities by Eisenbrauns Winona Lake, Indiana

2012

ISBN 978-1-57506-228-0 ISSN 1938-5668

The paper used in this publication meets the minimum requirements of the American National Standard for Information Sciences—Permanence of Paper for Printed Library Materials, ANSI Z39.48-1984.♾™

Contents

Contents

i

I. Articles Ancient Near Eastern Studies I. Arkhipov. Les véhicules terrestres dans les textes de Mari. II. Le harnachement et l’équipement de chars en cuir et en tissu J. Keetman. Akka von Kiš und die Arbeitsverweigerer I. Khait. The Old Babylonian Omens in the Pushkin State Museum of Fine Arts, Moscow M. Krebernik. SF 50 // 52: Zwei bislang unerkannte Duplikate aus Fāra S. Loesov. The Suffixing Conjugation of Akkadian: In Search of Its Meaning S. Loesov. IPTANARRAS Can Render the “Basic” Present (Akkadian Sentences about the Present Time. III/1) E. Markina. Akkadian of the Me-ság Archive R. Nurullin. Philological Notes on the First Tablet of the Standard Babylonian Gilgameš Epic R. Nurullin. The Name of Gilgameš in the Light of Line 47 of the First Tablet of the Standard Babylonian Gilgameš Epic T. Oshima. Another Attempt at Two Kassite Royal Inscriptions: The Agum-Kakrime Inscription and the Inscription of Kurigalzu the Son of Kadashmanharbe S. Parpola. Sumerian: A Uralic Language. II

5 15 31 61 75 149 169 189

209

225 269

Old Testament Studies Y. Eidelkind. Intended Lexical Ambiguity in the Song of Songs J. Joosten. Abram and Sarai in Egypt (Genesis 12:10–20)

325 369

ii

Babel und Bibel 6

Semitic Studies M. Bulakh. Negative markers *!ay-, *!i- and *!alin Ethio-Semitic S. Loesov. A New Attempt at Reconstructing Proto-Aramaic. I P. Stein. Ein weiteres arabisches Syntagma in der altsüdarabischen Epigraphik

385 421 457

II. Short Notes Ancient Near Eastern Studies D. Gromova. Idanda Archive and the Syrian Campaigns of Suppiluliuma I: New Pieces for the Puzzle L. Kogan, E. Markina. Marginalia Sargonica. I L. Kogan, M. Worthington. Accusative casus pendens: Some Further Examples

473 479 487

Old Testament Studies M. Seleznev. Exodus 33:7 and Different Syntactic Patterns for Linking Participial Conditional Clauses with the Main Clause

495

III. Reviews Ancient Near Eastern Studies G. Neumann. Glossar des Lykischen. Überarbeitet und zum Druck gebracht von Johann Tischler (A. Kassian) Recent Studies of Hurrian (A. Kassian) H. Waetzoldt, H. Hauptmann (Hrsg.). Von Sumer nach Ebla und zurück. Festschrift für Giovanni Pettinato (L. Kogan, N. Koslova)

503 517

523

Old Testament Studies I. Young, R. Rezetko, with the assistance of M. Ehrensvärd. Linguistic Dating of Biblical Texts. An Introduction to Approaches and Problems (J. Joosten)

535

Semitic Studies H. Gzella, M. L. Folmer (eds.). Aramaic in Its Historical and Linguistic Setting (E. Barsky et al.)

545

Contents

G. Goldenberg, A. Shisha-Halevy (eds.). Egyptian, Semitic and General Grammar. Studies in Memory of H. J. Polotsky. (M. Bulakh) C. A. Ciancaglini. Iranian Loanwords in Syriac (L. Kogan, S. Loesov) Sh. Talay. Die neuaramäischen Dialekte der Khabur-Assyrer in Nordostsyrien: Einführung, Phonologie und Morphologie (A. Lyavdansky) G. Takács. Etymological Dictionary of Egyptian. III (A. Militarev) Abbreviations of Periodicals, Reference Works, Series, and Sources

iii

565 575

587 593

611

Les véhicules terrestres dans les textes de Mari. II. Le harnachement et l’équipement de chars en cuir et en tissu* Ilya Arkhipov Institute of World History, Russian Academy of Sciences

L’ouvrage récent de J.-M. Durand comporte la (ré)édition de documents importants pour l’étude de la terminologie mariote du harnachement et de l’équipement des véhicules1. Plusieurs termes s’y référant ont été étudiés par l’auteur dans le contexte plus large des habits et des tissus. Le présent article tend à compléter cette étude sur certains points et à présenter un bilan sur la question, lequel on pourrait comparer aux données archéologiques et iconographiques2, ainsi qu’aux témoignages philologiques d’autres lieux et époques3. Aucun texte de Mari ne contient malheureusement de description d’un harnachement ou d’un équipement de char complets. Les objets relevant de ces deux catégories sont le plus souvent enregistrés de manière

* Cet article a été préparé dans le cadre du projet MK-2112.2011.6 financé par le Conseil Présidentiel pour les projets de recherche. Je remercie vivement Antoine Jacquet qui a révisé le texte français de l’article et a fait des suggestions importantes sur son contenu. 1 Durand, J.-M. La nomenlature des habits et des textiles dans les textes de Mari (ARM 30 = MDBP 1). Paris, 2009. Sauf indication du contraire, les textes cités ci-dessous par leur numéro de musée ont été édités dans ce volume. Les textes publiés dans la série des ARM sont cités compte tenu des joints et d’autres corrections proposés ibidem. Pour les moyens de transport à Mari en général et pour leurs parties en bois et en métal, voir désormais ARM 32 = MDBP 3 (sous presse), chapitre «Véhicules». 2 On les trouve réunies dans la synthèse classique de M. Littauer et J. Crouwel, Wheeled Vehicles and Ridden Animals in the Ancient Near East (HdO VII/1.2.B.1). Leyde, 1979. 3 Il semble que la dernière synthèse philologique sur la terminologie du harnachement soit toujours celle de A. Salonen, Hippologia accadica. Helsinki, 1955, pp. 93–166, réunissant surtout des données de textes tardifs ou lexicaux. Hormis Mari, les témoignages les plus importants sur le sujet proviennent du célèbre hymne sumérien à Išme-Dagan (Civil, M. Išme-Dagan and Enlil’s Chariot. JAOS 88 (1968):3–14) et des archives d’Ebla (Conti, G. Carri ed equipaggi nei testi di Ebla. QuSem 19 (1997):23–71). Cependant, il n’existe que très peu de points communs entre la terminologique de l’hymne, celle d’Ebla et celle de Mari.

6

Articles: Ancient Near Eastern Studies

mélangée, souvent à côté d’articles aux fonctions complètement différentes. La quantité des objets n’est jamais mise clairement en relation avec un nombre de véhicules ou d’animaux. Il n’est pas possible non plus de distinguer les types de harnachement en fonction de l’animal de trait4. À titre indicatif, les données des listes d’articles de ce genre les plus complètes sont réunies dans le tableau suivant, qui contient la plupart des termes étudiés ci-dessous.

appatum gabagallum gutappum ¶ullum kammakum naplasum rikis kussîm šabatum šulummum

ARM 7, 161

ARM 18, 45

ARM 21, 294

ARM 22, 317

ARM 22, 324 ii 14–19

M.8140+ ii 1′–12′

7 — 15 10 paires 5 4 paires 3 — x

4 2 — 4 paires 2 1 paire 1 — —

2 — x 2 paires 2 2 paires 2 2 x

2 — 6 2 paires 2 2 paires 2 — 1

3 7? — 7? paires x — — 5 —

— x 2 — 1 3 1 — 1

appatum AHw. 59 appatu(m) II «Zügel» (bab., m/nA); CAD A1 181–182 appatu A «reins, leash (?)» (OB, Mari, EA, MA). Attestations à Mari5: ARM 7, 161:7; ARM 7, 243:5′; ARM 21, 294:5′; ARM 22, 177+ vi 10′; ARM 22, 317:6; ARM 22, 324 ii 17, iv 4; ARM 23, 41:7; M.12812 iv 96; S. 143, 176:5 (ap-pa-tum); ARM 18, 22:12; ARM 18, 30:12; FM 9, 31 (ap-pa-tim); FM 6, 64: 6 (ta-pa-al ap-pa-tim); ARM 18, 45:5 (ap-pa-tum SAG); M.6710:2′ (ap-pa-at na-ap-li-si-im). Initialement, les rênes, une seule par animal, ne permettaient que de frêner et s’attachaient aux anneaux de nez qui étaient combinés dès l’époque très ancienne avec divers éléments de la têtière7. Au début du IIème millénaire, les anneaux de nez étaient toujours utilisés mais un système plus évolué a commencé à se répandre. Il comprenait deux rênes

4

À Mari, divers équidés ou bovidés étaient utilisés dans ce but (voir ARM 32, s. v. rakabbum et mâyaltum). 5 M.6654+ ix 12′ enregistre 1 kušpa-ap-tum ša 5 qa-né-e. J.-M. Durand (ARM 30, p. 258) considère la séquence PA AP TUM non attestée par ailleurs comme une mégraphie pour appatum (pa:ap-tum). Cependant, ce dernier mot n’est jamais écrit avec le déterminatif KUŠ à Mari; il faut d’ailleurs noter que le signe lu KUŠ devant PA ressemble beaucoup plus à un ZU. 6 ARM 32. 7 Littauer, M.; Crouwel, J. Op. cit., pp. 30, 44.

I. Arkhipov, Les véhicules terrestres dans les textes de Mari …

7

par animal, lesquelles se rejoignaient à la têtière et rendaient possible le contrôle directionnel8. En akkadien, le mot pour «anneau de nez» existe bel et bien mais son usage en rapport avec les animaux est limité aux listes lexicales et à l’épopée de Gilgameš; le terme désigne le plus souvent un objet porté par les gens ou s’emploie dans un sens métaphorique (CAD Ô 134–137, ´erretu A)9. C’est surtout son équivalent sumérien eškiri qui apparaît dans le contexte du harnachement, dans les documents de la pratique d’Ebla10 et d’Ur III. L’absence de l’objet dans les listes de harnachement à Mari indique qu’il n’était déjà plus en usage. En effet, l’expression appat naplasim «rêne d’œillère» (M.6710:2′) montre que les rênes étaient susceptibles de s’attacher aux oeillères. Il semble d’ailleurs que l’on continuait à utiliser une seule rêne par animal, puisque l’objet, à une seule exception près, n’était pas compté par paires. L’item était fabriqué en laine d’après ARM 18, 30. gabagallum J.-M. Durand, ARM 30:34, 598 «protection, décoration pour la poitrine»; rajouter ARM 21, 254:6 (GABA.GÁL). Le terme est à l’heure actuelle attesté en des lieux et époques très divers. On en distingue assez nettement trois usages différents. 1) Dans les listes lexicales et quelques textes littéraires11, gabagallum est une partie du devant ou du fond de char. Le déterminatif, lorsque il est présent, est toujours GIŠ, c’est-à-dire que la pièce se fabriquait normalement en bois. On rangerait ici une des attestations de Mari, à savoir ARM 21, 254:6 où le terme entre dans la description d’un char12. 2) Dans un document néosumérien d’Ur, gabagallum est un objet en or, sans doute un bijou13. 3) Dans la majorité des textes de Mari, ainsi qu’à Isin paléobabylonienne14, gabagallum est une pièce en cuir ou en tissu. À Mari, l’item est 8

Ibid., pp. 60–61. Pour Mari, cf. ARM 32, s. v. ´erretum. 10 Conti, G. Op. cit., pp. 44–47. 11 Civil, M. Op. cit., p. 10; Cooper, J. C. The Return of Ninurta to Nippur (AnOr 52). Roma, 1978. Pp. 64–65; Durand, J.-M. Un Commentaire à TDP I, AO 17661. RA 73 (1979):167; George, A. Babylonian Texts from the Folios of Sidney Smith. Part Two: Prognostic and Diagnostic Omens. Tablet I. RA 85 (1991):162. 12 Cf. Durand, J.-M. Textes administratifs des salles 134 et 160 du palais de Mari (ARMT 21). Paris. 1983. P. 285. 13 Sallaberger, W. Eine reiche Bestattung im neusumerischen Ur. JCS 47 (1995):15–17. 9

8

Articles: Ancient Near Eastern Studies

souvent rangé à côté de pièces de harnachement, mais apparaît aussi dans des contextes où sa destination n’est pas précisée. Cela laisse supposer un emploi plus diversifié de l’objet, par exemple comme élément de tapisserie, de vêtement ou de décoration pour un équidé15. Cette dernière fonction fait penser aux poitrails bien attestés dès le début du IIIème millénaire16. gutappum J.-M. Durand, ARM 30, p. 182, 598 «longe»; rajouter ARM 22, 324 i 29 (túggú-tap-pu); M.6654+ ix 11′ (túggu-tap-pu); ARM 21, 294:1′ (gú-tap-pu). Les listes lexicales indiquent que le gutappum est une pièce de harnachement, mais rien ne permet d’en connaître la fonction. L’étymologie sumérienne du mot met l’objet en rapport avec le cou de l’animal. Cela en fait un candidat pour désigner le «collier» de harnachement (cf. s. v. ¶ullum). ¶ullum Les dictionnaires n’enregistrent pas pour le mot le sens d’une pièce de harnachement17, avec lequel il est amplement attesté à Mari: ARM 18, 30: 12 (¶u-ul-li); ARM 7, 161:9; ARM 21, 294:4′; ARM 22, 301+:20′; ARM 22, 324 ii 16; ARM 24, 192:3; M.8279:4; S. 143, 176:4 (ta-pa-al ¶u-ul-li); ARM 18, 22:11 (ta-pa-al ¶u-ul-li-im); ARM 18, 45:4; ARM 22, 317:5; M.5156:19; M.5756 iii 22; M.12812 iv 918 (ta-pa-al ¶u-li); ARM 22, 324 iv 3 (ta-pa-al ¶u-li-im); ARM 18, 22:7–8 (¶u-ul-li ÚS la u¶-¶u-zu-tim). Le plus souvent, le terme ¶ullum sert en akkadien pour les ornements de cou de forme annulaire et en matières précieuses19. À Mari, le même nom est employé pour une pièce de harnachement20, qui se présente toujours par paires, est fabriqué en laine (ARM 18, 30) et susceptible de porter des placages (ARM 18, 22). Le mot est utilisé de manière semblable à Ebla21. 14 Van De Mieroop, M. Crafts in the Early Isin Period (OLA 24). Leuven, 1987, p. 136; cf. le compte-rendu par D. Charpin dans RA 84 (1990):90. 15 Durand, J.-M. ARM 30, p. 34. 16 Littauer, M.; Crouwel, J. Op. cit., p. 30. 17 À el-Amarna, le mot canaanéen ¶ullu est attesté comme équivalent de l’akkadien nīrum «joug». Ce sens est bien différent de celui discuté ci-dessus. 18 ARM 32. 19 ARM 32, s. v. 20 Cf. déjà Rouault, O. Mukannišum. L’administration et l’économie palatiales à Mari (ARMT 18). Paris, 1977. P. 145. 21 Conti, G. Op. cit., pp. 40–42.

I. Arkhipov, Les véhicules terrestres dans les textes de Mari …

9

Compte tenu du sens principal du mot («anneau de cou») et de ses parents sémitiques («joug»), il est difficile d’identifier l’objet à autre chose qu’au «collier» qui servait à attacher l’animal au joug22. Cependant, le caractère pair de ¶ullum à Mari constitue pour cette interprétation un obstacle difficilement surmontable23. kammakum J.-M. Durand, ARM 30, p. 50, 158 «bâche»; rajouter M.11729:5 (kam-makum); M.15076: 3–424 (kam-ma-ki-im ša GIŠ.GIGIR); M.12179: iv 3 (kam-makum ¶i-ru). J.-M. Durand (loc. cit.) a démontré que le terme désignait la bâche d’un char, en laine ou en lin. Un argument important est fourni par le document ARM 21, 255 où le kammakum est la pièce principale d’un char à côté des roues25. On pouvait décorer la bâche avec des étoiles en argent (ARM 7, 116; M.15076). kussûm (ša immerim) AHw. 515 kussû(m) 4 «Packsattel» (aA, jB, Bo.); CAD K 587–593 kussû 6 «saddle (for a donkey)» (OA, OB, Bogh., SB). Attestations à Mari: M.12812 iv 1026 (ku-sú-ú ANŠE LA.GU); ARM 22, 110:7; ARM 22, 111:7; FM 3, 72:3; A.55:5227; M.6479:19; M.8279:8; M.15106:10′28 (GIŠ.GU.ZA ANŠE LA.GU); M.18229:1 (GIŠ.GU.ZA ANŠE LA.GU MA.GAₓ.AN.NA); FM 4, 50:11′ (GIŠ.GU.ZA ANŠE LA.GU qa-du-um ¶i-il-la-tim); ARM 22, 199:2–3 (ki-pa-tim ša GIŠ.GU.ZA LA.GU); ARM 7, 161:5; ARM 22, 317:3 (gadri-ki-ìs GIŠ.GU.ZA); ARM 18, 45:3; ARM 21, 294:3; M.8140+ ii 6′ (gad¶i-rum ri-ki-ìs GIŠ.GU.ZA). La selle d’âne se présentait à l’époque comme une construction en bois, avec un dossier et un repose-pieds, sur laquelle on s’asseyait de tra-

22

Pour ces objets en archéologie, voir Littauer, M.; Crouwel, J. Op. cit., p. 6, 26. Il est vrai que les animaux de trait étaient toujours attachés par paires à l’époque (Littauer, M.; Crouwel, J. Op. cit., p. 35). On pourrait donc imaginer que les colliers de joug étaient toujours livrés par paires, mais comment expliquer que toutes les autres pièces de harnachement ne le fussent pas? 24 ARM 32. 25 Cf. déjà Durand, J.-M. ARM 21 (1983), p. 292, avec un parallèle archéologique. 26 ARM 32. 27 FM 1 (1992):53–55. 28 ARM 32. 23

10

Articles: Ancient Near Eastern Studies

vers29. La construction comprenait des cerceaux en bronze (ARM 22, 199), des attaches riksum et éventuellement des couvertures ¶illum30. naplasum AHw. 739 naplastu(m) «Scheukappe» (a/jB, Bo.), naplāsu 2 «Scheukappe» (Mari); CAD N1 305–306 naplastu 1 «blinkers» (Mari, SB)31, naplasu 2 «blinkers» (Mari). Attestations à Mari: M.12804+ i 4′32 (kušna-ap-la-su); ARM 7, 192:15 (na-ap-la-si-im); ARM 7, 161:8; ARM 18, 45:6 (ta-pa-al kuš na-ap-la-si); ARM 21, 294:6′; ARM 22, 301+:22′; ARM 22, 317:8; M.8279:5; M.15195 ii′ [3′] (ta-pa-al na-ap-la-si); ARM 24, 288+ ii 33, 3633 (ta-pa-al na-ap-la-si-im); ARM 21, 223:14 (ta-pa-al na-ap-la-si ša KÙ.BABBAR); ARM 21, 223:15–16 (ta-pa-al na-ap-la-si ša KÙ.BABBAR i-natu-šu KÙ.GI); ARM 22, 301+:23′; M.11761:6; M.18192:2 (na-ap-la-ás ANŠE LA.GU); ARM 24, 192:2; M.6654+ v [18′] (na-ap-la-su ša ANŠE LA.GU); M.11761:2′ (na-ap-la-ás anšepa-re-e); ARM 18, 22:13 (na-ap-li-si); M.6710:2′ (ap-pa-at na-ap-li-si-im). Les œillères, attestées dans l’iconographie uniquement à partir du Bronze Récent34, étaient déjà largement utilisées à Mari35. L’objet était fabriqué en cuir, d’après le déterminatif, et parfois décoré d’argent (ARM 7, 192:14–15; ARM 21, 223:14) ou d’un ornement nommé īnum «œil» en or (ARM 21, 223:15–16). parūtum AHw. 837 parūtum II «eine Art von Köcher?» (a/jB); CAD P 211–212 parūtu B «a type of leather» (OB, Mari). Attestations à Mari36: ARM 21, 302:9 (kušpa-ru-tum); M.6799:7 (KUŠ ša pa-ru-tim); ARM 22, 317:12;

29

Voir le bilan dressé par Dercksen, J. Sattel. RlA 12:90–93; cf. Littauer, M.; Crouwel, J. Op. cit., p. 66 et fig. 38. 30 Pour ces deux termes, voir Durand, J.-M. ARM 30, pp. 44 et 163. 31 L’exemple de Mari est à rayer, cf. Durand, J.-M. Documents épistolaires du Palais de Mari. T. III (LAPO 18). Paris, 2000. Pp. 253–254. 32 ARM 32. 33 Ibid. 34 Littauer M.; Crouwel, J. Op. cit., p. 90. 35 Cf. Soubeyran, D. Archives administratives de Mari. I (ARMT 23). Paris, 1984. P. 406. 36 Il faut éliminer les exemples de ARM 23, 104:19–20, donnés par le CAD: les signes en question sont endommagés et les lectures proposées, sans parallèles.

I. Arkhipov, Les véhicules terrestres dans les textes de Mari …

11

M.8279:1; M.10061: [8′]37 (GIŠ.GIGIR ša kušpa-ru-tim); M.15195 ii′ 1′, 5′ (GIŠ.GIGIR ša pa-ru-tim); M.12223:138 (GIŠ.GIGIR pa-ru-tim). Le terme est le plus souvent employé comme qualificatif du char; l’objet de ce nom n’est attesté indépendamment qu’un fois (ARM 21, 302: 9)39. Il doit s’agir d’une sorte particulière de bâche puisque ce serait la seule partie en cuir assez importante pour déterminer le type de char. Les autres attestations du mot, dans les listes lexicales et dans un texte paléobabylonien (en sumérien), indiquent seulement que l’objet était fabriqué en cuir. On pourrait, faute de mieux, rapprocher le terme avec le nom de la pierre parūtum40 ou avec pāru «skin, hide» (CAD P 209). puqdum AHw. 880 puquttu 4b «Metall-Dorn (als Werkzeug)» (mA, Nuzi); CAD P 515–516 puquttu d «thorn, barb (as a weapon)» (OB, MA, Nuzi, NA, SB)41. Attestation à Mari: ARM 22, 317:10 (pu-uq-da-tum). Le document de Mari enregistre l’objet parmi l’équipement du char. S’agit-il de l’aiguillon42? šabatum J.-M. Durand, ARM 30, p. 114 «toile blanche?»; rajouter ARM 22, 177+ vi 8′ (ša-ba-tum). Dans la plupart des cas, les objets de ce nom sont enregistrés parmi des éléments de harnachement et d’équipement de char, mais rien ne permet de comprendre le sens du terme. Pour une interprétation étymologique, voir J.-M. Durand, loc. cit. šulummum J.-M. Durand, ARM 30, p. 171 «boite où l’on rangeait le fouet et autres ustensiles du chariot»; rajouter ARM 21, 254:10 (kuššu-lum-mu); M.15195 ii′ 2′ (šu-lum-mu); ARM 31, 24 iv 7 (kuššu-lu-mu SAG). 37

ARM 32. Ibid. 39 L’interprétation de M.6799:7 est incertaine. 40 Cf. Durand, J.-M. ARM 21, pp. 368–369. 41 Cf. aussi AHw. 879 puq/kdu «eine Teil des Pfluges» (jB); CAD P 512 puqdu «part of a plough» (lex.). 42 Cf. l’équivalence, dans les listes lexicales, de puqdum avec karallum, traduit «prick, goad» dans CAD K 200. 38

12

Articles: Ancient Near Eastern Studies

J.-M. Durand a mis le mot, spécifique de Mari, en rapport avec le sumérien esulumak (écrit normalement é-su-lum-ma ou é-zú-lum-ma) dont l’équivalent akkadien est la¶arušhum (avec des variantes)43. Il existe en effet des témoignages très explicites que ces derniers termes sumérien et akkadien désignent une boîte de rangement attachée au char44. Cependant, l’emprunt de esulumak sous forme šulummum serait étonnant. Le mot sumérien est à analyser comme «boîte de sulum»; c’est plutôt ce dernier terme qui aurait donné šulummum à Mari. Or, sulum signifie d’après M. Civil (loc. cit.) «tethering rope». Il est donc tentant d’identifier šulummum à une sorte de longe en cuir. D’ailleurs, le document ARM 21, 254 semble enregistrer l’objet parmi les parties de char et non pas de harnachement. Termes non spécifiques Plusieurs articles en cuir ou en tissu enregistrés parmi des parties de harnachement ou de chars avaient en réalité une fonction plus large, même s’ils étaient utilisés épisodiquement pour les moyens de transport ou les animaux de trait; il est d’ailleurs possible que la présence de certains objets dans ce contexte soit due au hasard. – ¶ayyûm «tapis, tapisserie»45. Des tapis feraient partie de l’équipement du char (ARM 21, 294:10′; ARM 22, 317:4; S. 143, 176:2). – ¶īrum «bande pour renforcer ou couvrir»46. La «bande» en lin servait en particulier pour attacher la selle (voir s. v. kussûm). Les autres cas sont moins clairs: en ARM 21, 254, un ¶īrum, qualifié de l’appositif nakbasum «carpette, tapis (qui recouvre le plancher de char)»47, fait partie de l’équipement d’un char; en M.12179 iv 3, le mot sert à son tour de qualificatif pour kammakum «bâche». – marīnum «sac outre?»48. Des textes d’huile montrent que des marīnum pouvaient faire partie d’un palanquin (FM 3, 54:4–5; 43

Durand, J.-M. ARM 21, pp. 286–287. Civil, M. Op. cit., pp. 8–9; CAD L 40. 45 Durand, J.-M. ARM 30, pp. 42–44, 598. 46 Ibid., pp. 156–158, 598. 47 Ibid., p. 73. Cf. les compte-rendus du CAD N1 par K. Deller et W. R. Mayer dans Or 56 (1987):193, 199. 48 Durand, J.-M. ARM 30, p. 178, 598. 44

I. Arkhipov, Les véhicules terrestres dans les textes de Mari …

13

60:29; 125:25′–26′)49. Selon toutes les autres attestations, il s’agirait d’une sorte de couverture en cuir50. – na¶laptum «drap-housse?»51. Des na¶laptum noires (GÚ GE6) sont enregistrées en ARM 22, 317:11 parmi du matériel de char. – nanaptum «une attache»52. L’article jouait un rôle dans la fabrication des chars selon ARM 21, 420. – riksum «pansement, attache de meuble, de paquetage»53. ARM 24, 181 enregistre la livraison d’une étoffe barkarrûm pour des riksum d’«ânes de char» (ANŠE GIŠ.GIGIR). Les «attaches» feraient donc partie de l’attelage. Cf. aussi l’usage du terme pour désigner les attaches de selle (s. v. kussûm). – ´imdum ša ¶umāšim «bandeau de force»54. Le document ARM 22, 298 montre clairement que le ´imdum est un accessoire des «lutteurs» ša ¶umāšim55. Il est d’autant plus étonnant de voir l’objet enregistré parmi les pièces de harnachement et d’équipement de véhicules (ARM 7, 161:10; ARM 21, 294:[8′])56. – ´ubātum «tissu, coupon d’étoffe»57. Des tissus TÚG de type barkarrûm «grossier» et išarum «écru»58 pouvaient faire partie du matériel associé au char (ARM 7, 161:12; ARM 21, 294:9′; M.8279:6–7). Dans le dernier cas, il est précisé que les tissus couvrent les chars (ša GIŠ.GIGIR ku-ut-tu-mu). Les «cordes» (ÉŠ.ÚÁ) de ARM 22, 317:9 seraient plutôt des tissus divers (TÚG.ÚÁ). – zigurītum «tissu d’ameublement, un coussin»59. Une «natte longue» (zi-gu-ri-tum ar-ku-um, noter le masculin!) se trouvait à l’intérieur d’un char mayyaltum60 d’après M.12804+ ii 26′–28′. 49 Pour nūbalum, voir Arkhipov, I. Les véhicules terrestres dans les textes de Mari I: le nūbalum. CRRAI 53 (2010):405–419. 50 Groneberg, B. Marīnu. NABU 1990/23; eadem. ARM I, 17. NABU 1990/62. 51 Durand, J.-M. ARM 30, pp. 67–69. 52 Ibid., p. 185, 601. 53 Ibid., p. 163, 602. 54 Ibid., p. 95. 55 Pour cette catégorie de personnel de divertissement, voir Durand J.-M.; Guichard, M. Les rituels de Mari. FM 3, p. 51. Il est d’ailleurs surprenant de constater la présence dans le palais de Mari de 60 «amuseurs» ša ¶umāšim (ARM 22, 298:2–3). 56 Cf. d’ailleurs l’existence supposée de l’élément de harnachement ´imdu à Ebla (Conti, G. Op. cit., pp. 43–44). 57 Durand, J.-M. ARM 30, pp. 96–97. 58 Ibid., pp. 97–99, 101–104. 59 Ibid., p. 139, 604.

14

Articles: Ancient Near Eastern Studies

* * * Bien que l’interprétation de plusieurs des termes examinés reste problématique, les textes de Mari permettent de se faire une idée sur le harnachement de l’époque, laquelle correspond assez bien aux données archéologiques. Le système permettant la fixation de l’animal de trait au joug comprendrait des éléments gutappum et ¶ullum, dont les fonctions exactes sont encore à trouver. Ils étaient peut-être accompagnés du poitrail gabagallum. Différents moyens de contrôle sont attestés, tels que les rênes appatum, les longes šulummum et l’aiguillon puqdum; l’absence de fouet serait due au hasard de la documentation. Malheureusement, aucun terme ne se laisse identifier aux parties de la têtière, à l’exception des œillères naplasum auxquelles s’attachaient apparemment les rênes. Les termes ¶īrum et riksum servaient à désigner des bandes et des attaches. La «selle» kussûm ressemblait à l’époque plutôt à un siège en bois sur lequel on s’asseyait de travers. Outre deux sortes de bâche, kammakum et parūtum, l’équipement du char comprenait surtout divers tapis et couvertures (¶ayyûm, marīnum, nakbasum, na¶laptum, ´ubātum, zigurītum), ainsi que des items qui restent à identifier (šabatum, nanaptum, ´imdum).

60

Pour ce type de char, voir ARM 32, s. v.

Akka von Kiš und die Arbeitsverweigerer Jan Keetman Istanbul

Dass Leute murren, wenn sie in ihren Augen zu viel arbeiten müssen, ist allgemein menschlich, doch in der altorientalischen Literatur ist es sogar göttlich. Als die Götter noch der Mensch waren,1 als es ihnen noch ging wie den Menschen, da murrten selbst sie, bis schließlich eine andere Lösung gefunden wurde. So wissen wir es jedenfalls aus Atram-¶asīs. Das Thema der Arbeitsverweigerung der Menschen selbst kommt in der Literatur des Alten Orients nur einmal vor, nämlich in der von uns Gilgameš und Akka genannten Erzählung (GA).2 Der Kern der Handlung besteht daraus, dass ein König etwas befiehlt, doch die Leute weigern sich letztlich erfolgreich. Damit wird postuliert, dass monarchische Gewalt nicht unbegrenzt ist. Sicher ein Thema mit dem man in einer Monarchie vorsichtig sein sollte. Aber mit den Mitteln literarischer Verfremdung lässt sich auch dieses Thema so vortragen, dass kein offener Affront gegen den Herrscher daraus wird. Die Geschichte spielt vielleicht nicht zufällig in weiter Ferne. Sie knüpft an den eher sagenhaften Teil der Sumerischen Königsliste an.3 Doch da es Atram-¶asīs I 1 i-nu-ma i-lu a-wi-lum wird als Nominalsatz gedeutet und awīlum hier als Gattungsbezeichnung. Möglicherweise wird dies durch ein Lautspiel unterstützt (-m)a ilū ~ a(w)īlu(m) oder ilū ~ īlu, als sei der Gott in den Menschen eingebaut. Andere Deutungen der Zeile sind aber ebenfalls möglich. Vgl. Loesov 2004. 2 Erste Bearbeitung von GA Kramer 1949, weitere Bearbeitungen Römer 1980 und Katz 1993. Übersetzung Jacobsen 1987:345–355; Römer 1993:349– 359; ETCSL. Einige wichtige Rezensionen und Einzelbeiträge: Jacobsen 1957; Falkenstein 1966; Lambert 1980; Cooper 1981; Michalowski 1982; Vanstiphout 1986; 1987; Wilcke 1998. Siehe auch die weiteren Anmerkungen. 3 Die bisher älteste Fassung der Königsliste, die, wenn man sich an ihre letzte Zeile hält, auf Sulge zu datieren ist (zur Lesung Sulge statt Šulgi Keetman 2010: 26, Anm. 50), erwähnt zwar Akka (geschrieben aka) und seinen Vater Enmebaragesi als Könige von Kiš doch nicht den Konflikt mit Gilgameš, der im erhaltenen Teil gar nicht auftaucht. In den Augen des Autors kann man auf diese Weise aber weder die Historizität, noch die Ahistorizität der Episode belegen. Die Königsliste enthält sowohl echte Informationen, wie die Existenz eines frühdynastischen Kö1

16

Articles: Ancient Near Eastern Studies

sich bei GA noch offensichtlicher als bei der sumerischen Königsliste um ein literarisches Produkt handelt, können wir unterstellen, dass es in GA nicht darum geht, Geschichte nachzuerzählen, es sei denn die Fakten oder Legenden unterstützen den Plot der Erzählung. Wie Claus Wilcke nachgewiesen hat, trägt der Text unverkennbare Merkmale des Sumerischen der fortgeschrittenen 1. Hälfte der altbabylonischen Zeit.4 Eventuell könnte man sogar noch ein paar Generationen weiter gehen als Wilcke es tut.5 Wilckes Beobachtungen beziehen sich zwar unmittelbar nur auf die bekannten Niederschriften des Textes,

nigs (En-)Mebaragesi von Kiš, als auch mit Sicherheit Erfindungen. Am Anfang nur Erfindungen, am Ende Informationen, die wir als richtig überprüfen können. Dass Gilgameš Akka mit der Waffe besiegte, steht im mittleren Teil, wo sich echte Informationen mit Sagengestalten mischen. Dass diese Episode in der “Ur III Rezension” ausgelassen wurde, mag damit zusammenhängen, dass die Liste eindeutig das Königtum von Kiš bevorzugt, was seltsam ist, wenn man besagter letzter Zeile Glauben schenkt. In jedem Fall können wir nichts nachprüfen. Es ist recht wahrscheinlich, dass nicht nur die Autoren der Sumerischen Königsliste (SKL), sondern auch der Autor von GA, sich ebenso sehr frei überkommener Namen und der eigenen Phantasie bedienten, wie der Verfasser des TIMEArtikels aus dem Claus Wilcke zum Auftakt seiner Analyse der SKL zitiert: “Berosus, the high priest of Babylon, would climb the spiral ramp of the great ziggurat at night and ask the stars, if the time was ripe to move against the Assyrians” (16. 05. 1988). Anstatt nach Fakten zu suchen, deren Authentizität wir in aller Regel nicht überprüfen können, sollte man lieber, so weit als möglich, danach fragen, mit welcher Absicht diese Elemente in der Geschichte kombiniert werden. Zur SKL Jacobsen 1939. Unter den zahlreichen Interpretationsversuchen sei hier nur Wilcke 1989 genannt. Zur fraglichen Authentizität des oft mit der SKL kombinierten Tummal-Chronik Michalowski 2006. Zur erwähnten “Ur III Rezension” Steinkeller 2003. 4 Wilcke 1998. 5 Wilcke 1998:480f. Wilcke weist darauf hin, dass die “für die Zeit von WaradSîn und Rīm-Sîn typische Verwechslung der Sach- und Personenklasse bei den Possessiv-Suffixen” noch nicht festzustellen ist. Diese Unsicherheit gilt jedoch nicht für alle Texte. Z. B. enthalten die Hymnen Rīm-Sîn A und G keine Fehler (zu diesen Texten jetzt Brisch 2007, vgl. auch Steible 1975). Von daher könnte der Text also sogar noch jünger sein, zumal alle Textzeugen aus Nippur stammen, wo man das Sumerische besonders pflegte. In Gilgamešs Tod werden hingegen die Possessivsuffixe -a-né und -bé verwechselt, was für eine noch jüngere Abfassung spricht. Dies könnte erklären, warum Gilgamešs Tod nicht in den Katalogen sumerischer Texte in Nippur erscheint. Außerdem passt es zur Erwähnung Ziusudras, was bereits eine Anspielung auf das akkadische Epos sein könnte. Vgl. die Ausführungen bei Cavigneaux–Al-Rawi 2000:10.

J. Keetman, Akka von Kiš und die Arbeitsverweigerer

17

doch dafür dass ein voraltbabylonischer sumerischer Text so rundum modernisiert wurde, gibt es kein Beispiel. Akka von Kiš6 schickt Boten zu Gilgameš nach Uruk. Gilgameš ist offenbar als sein Vasallenkönig zu denken.7 Das Anliegen hören wir nicht aus dem Mund von Akka oder eines Boten, sondern Gilgameš wiederholt es vor der Versammlung der ‘Stadtväter’, ab-ba-iri. Gilgameš trägt eine Redewendung vor und es bleibt unklar, ob die Botschaft wiedergegeben oder ihr Inhalt pointiert umschrieben wird. Dafür dass Gilgameš die Botschaft auslegt, spricht jedenfalls Zeile 4: inim ba-an-ŋar inim ì-kin-kin-e ‘Die Angelegenheit legte er zur Entscheidung vor (und) untersucht sie’8 oder ‘sucht die Worte aus’. Jedenfalls scheint es, dass Akka etwas ungehöriges, vielleicht sogar sinnloses verlangt, das mit Fronarbeit zu tun hat: túl til-le-da túl kalam til-til-le-da túl níŋ-bàn-da kalam til-til-le-da túl bùru-da éš-lá til-til-le-da ‘Die Brunnen sollen ausgeschöpft werden, die Brunnen des Landes Sumer9 sollen ganz ausgeschöpft werden! 6

Zur möglichen Identität von Akka (so konventionell, die Namensform Aka hat durchaus Argumente für sich) mit dem auf einer Lapislazuli-Perle bezeugten aka lugal GIŠ.KUŠÚki und einer Person auf der Ušumgalstele (ELTS No. 12) Selz 2003. Das später nicht mehr belegte Namenselement aka (falls so zu lesen) und auch die Kurzform aka, sind in Fāra, Tell Abū Ôalābīkh und Adab noch einige Male belegt (z. B. NTSŠ 256Vs. ii 2; 444Vs. i 5; IAS 552 ii′ 5 passim). Der Name ist also nicht so ausgefallen, wie er vielleicht aufgrund der späteren Überlieferung scheinen mag. Schon von daher sollte man mit Identifikationen sehr vorsichtig sein. Wüssten wir über die altbabylonische Zeit nicht mehr als über die frühdynastische Epoche, so könnten wir auch leicht die verschiedenen Sîn-iddinams, Išme-Dagans und gar Hammurāpis verwechseln. Immerhin ist es bemerkenswert, dass sowohl Gilgameš als auch Akka typische Namen vom Beginn der FD-Zeit tragen. Natürlich reicht diese Beobachtung nicht aus, um einen historischen Kern von GA anzunehmen. 7 Dies ist in Übereinstimmung mit der altbabylonischen Version der Sumerischen Königsliste, nach der es immer nur einen souveränen Herrscher in Babylonien gab und das Königtum durch eine Niederlage Akkas an Uruk überging. Also kann Gilgameš vor dem in der Erzählung später beschriebenen Krieg mit Kiš nur ein Vasallenkönig gewesen sein. 8 Vgl. inim kin-kin = amāta šite’û, was in CAD vielleicht zu še’û 4 e ‘investigate’ zu stellen ist. Cf. CAD Š2 355b u. 361. 9 Wilcke (1998:476) weist darauf hin, dass nur Text c den Genitiv kalam-ma schreibt und nur am Anfang. Es gab also auch eine Auffassung, nach der Z. 5 (= 11; 20) ‘Die Brunnen sollen ausgeschöpft werden, die Brunnen – das Land Su-

18

Articles: Ancient Near Eastern Studies Die kleinen Brunnen des Landes Sumer sollen ganz ausgeschöpft werden, Die tiefen Brunnen mit Zugvorrichtung sollen ganz ausgeschöpft werden.’

Wilcke deutet diesen Teil der Rede des Gilgameš ganz anders, nämlich als Aufforderung bis zum letzten zu kämpfen.10 Doch dies führt dazu, dass der casus belli völlig im Dunkeln bleibt. Außerdem wird dann der Appell bis zum äußersten zu kämpfen auch in der Antwort der Honoratioren wiederholt, obwohl er in ihrer Rede sinnlos ist. Schließlich ist das Bild auch etwas gezwungen. Man kämpft eher bis zum letzten Blutstropfen als bis einem das Wasser ausgeht.11 Nur die Jungen, diejenigen welche selbst arbeiten müssten, revoltieren.12 Sie antworten Gilgameš: gub-gub-bu-dè tuš-tuš-ù-dè dumu lugal-la da ri-e-dè ¶áš anše dab5-dab5-bé-e-dè a-ba zi-bé mu-un-tuku-e-še ‘Ständig aufzustehen, ständig sich hinzusetzen, für den Prinzen das Kindermädchen zu spielen,13 die Schenkel des Esels (im Hinterherjagen?) ständig zu packen, wer sagt, er habe den Atem dazu?’14

mer, sollen ganz ausgeschöpft werden’ zu interpretieren wäre und entsprechend die nächste Zeile. Diese Interpretation bringt noch klarer zum Ausdruck, was gemeint ist und sie ist gegenüber dem sich aufdrängenden Genitiv eine mehr überdachte Interpretation. Allerdings fehlt die Quintessenz gerade in der abschließenden Zeile, weswegen diese Interpretation nicht ganz sicher ist. 10 Wilcke 1998:476f. 11 Es drängt sich natürlich auch eine andere Assoziation auf, nämlich die an die frühdynastisch gut bezeugten Auseinandersetzungen um Wasserrechte. Zuletzt Selz 2003:505, Anm. 25 mit weiterer Literatur und ebd. S. 512. Doch bei diesen Auseinandersetzungen geht es um Kanäle, nicht um Brunnen. Das Ausschachten von Brunnen eignet sich eher als Beispiel für eine besonders unangenehme und nicht zur normalen Fronarbeit gehörende Arbeit. Ganz sinnlos wird es, wenn man annimmt, dass das Ausschöpfen von Brunnen gemeint ist. 12 Diese Interpretation ist auch schon bei Selz 1998:316 angedeutet. 13 Für da ri sind die Gleichungen našû ša ´e¶ri und našû ša almatti belegt (CAD N2 81a). Das zweite kann kaum etwas anderes als die Unterstützung der Witwen meinen. Von da aus lässt sich raten, dass die erste Gleichung die Fürsorge für ein kleines Kind meint. Vgl. in schwierigem Kontext dumu lú úr-ra da an-ri-ri ‘sie nimmt das Kind des Mannes auf den Schoß’ (?); Römer 2001:113, dazu mit weiterer Literatur ibid. 134f. 14 Z. 25–28. Vgl. Katz 1993:5.

J. Keetman, Akka von Kiš und die Arbeitsverweigerer

19

Die Deutung der vorletzten Zeile ist geraten, doch aus dem Zusammenhang wird jedenfalls klar, dass die Jungen sagen wollen, dass ihnen zu viel Arbeit abverlangt wird. Die Aussage wird dadurch abgemildert, dass nicht vom König, sondern vom ‘Sohn des Königs’ die Rede ist. Vielleicht ein Hinweis darauf, dass Akka noch jung und unreif ist.15 Wie dem auch sei, jedenfalls ist es ein König, der durch seine Nachfrage die Klagen der Leute zu hören bekommt und schließlich selbst die Revolte anführt. Das monarchische System wird nicht in Frage gestellt und der gute Patron hat nichts zu fürchten. Außerdem war die Forderung Akkas nicht nur unsinnig, sie stand auch außerhalb dessen, was normalerweise an Fronarbeit abverlangt wird. Jedenfalls legt die von Gilgameš gebrauchte Umschreibung der Forderung dies nahe. Die Leute sind auch keineswegs faul. In Uruk wird unter Gilgameš sehr wohl gearbeitet. Wenn die Arbeiter in poetischer Sprache die berühmten Bauten von Uruk preisen, so geben sie für die Arbeit zwar den Göttern den Verdienst, doch jeder weiß, dass man zum Bauen von Tempeln und Stadtmauern auch selbst die Hand anlegen muss. Dies haben die Urukäer auch reichlich und geschickt getan, sonst sähe es nicht aus wie das Werk der Götter. Offenbar geht die hinter der Formulierung dieses Abschnittes stehende Idee von der wörtlichen Interpretation des Namens für den berühmten Tempelkomplex é-an-na ‘Tempel des Himmels’ aus. Dies wird dann so verstanden, als sei Eanna im Himmel gemacht und danach auf die Erde gesetzt worden und ebenso die berühmte Stadtmauer des Gilgameš. Wie immer man sich dies nun konkret vorstellt, ob es nur um die Verwirklichung der von den Göttern im Himmel entworfenen Pläne geht 15 Die folgende, etwas spekulative Interpretationsmöglichkeit soll nicht ganz unerwähnt bleiben: gleich in Z. 1 und dann noch einmal in Z. 49 wird Akka als Sohn Enmebaragesi’s vorgestellt, während Gilgameš ganz ohne Filiation erscheint. Soll mit der Bezeichnung ‘Sohn des Königs’ in der verächtlichen Rede der jungen Männer auch darauf hingewiesen werden, dass es eben nicht reicht, als König geboren zu werden? Im zweiten Teil der Rede wird dann Gilgameš als ein König beschrieben, der sich die Achtung seiner Untertanen verdient hat. Zu vergleichen ist vielleicht Gilgameš P 154–163, wo ein Passant davon erzählt, wie Gilgameš die Bräute vor dem Bräutigam beschläft, und dies mit seiner Geburt als König verbindet. Am Ende der Tafel verzichtet dann Gilgameš auf das noch immer mögliche Eindringen in das Haus des Schwiegervaters und wird darauf von Enkidu als König anerkannt. Das wäre dann der gleiche Gedanke: die Macht hat der König von Geburt, doch erst indem er darauf verzichtet, sie nach Lust und Laune zu gebrauchen, gewinnt er die ganze Anerkennung durch seine Untertanen. Vgl. dazu Keetman 2008:163–167.

20

Articles: Ancient Near Eastern Studies

oder ob auf eine mythische Vorstellung hingewiesen wird, wonach der Tempel im Himmel gebaut und auf der Erde nur verankert wird (man denke an die den Fundamenten beigegebenen Tonnägel und den Mythos Inana raubt den großen Himmel16), in jedem Fall haben auch die Leute von Uruk mit Hand angelegt. Dies folgt schon daraus, dass Gilgameš sich darum gekümmert hat und dass sie daraus seine Befähigung ableiten, sie als König zu führen.17 unugki ŋiš-kin-ti diŋir-re-e-ne-ke4 é-an-na é an-ta e11-dè diŋir gal-gal-e-ne me-dím-bé ba-an-ak-eš-àm bàd gal muru9 ki ús-sa-a-ba ki-tuš ma¶ an-né ŋar-ra-a-ba saŋ mu-e-sì za-e lugal ur-saŋ-me-en ‘Uruk ist die Gießerei18 der Götter! Da ist Eanna – der Tempel, der vom Himmel herabsteigen soll(te) – die großen Götter haben seine Teile gemacht. Als die große Mauer wie ein Nebel die Erde berührte,19 16

Siehe van Dijk 1998. Beachte, dass sich die Formulierungen ganz speziell auf Uruk und wegen der Erwähnung der berühmten Mauer auf Gilgameš beziehen. Es handelt sich also nicht um eine allgemeine Formel, mit der man einen König oder militärischen Führer bestätigt, sondern eine Formulierung, die nur in unserem Kontext Sinn ergibt. Nicht die Form der Aussage, sondern nur ihr Inhalt kann auf einen anderen König übertragen werden. Eine Ausnahme wäre ein Herrscher wie Anam, der sich im 19. Jahrhundert preist, die von Gilgameš errichtete Mauer restauriert zu haben. Vgl. George 2003:92. Ungefähr in diese Zeit kann man mit Wilcke auch die Niederschrift datieren. Doch erscheinen mir diese Argumente zu dünn, um die Entstehung von GA auf Anam zu datieren. Bemerkenswert ist auch, dass es in unserem Text um den rechten lugal geht, Anam diesen Titel, bzw. seine akkadische Übersetzung šarrum, aber nicht gebraucht. 18 Jacobsen 1987:349; Katz 1993:41 und ETCSL übersetzen ‘handiwork’. Man sollte jedoch eher níŋ-ŋiš-kin-ti oder nam-ŋiš-kin-ti erwarten. Wie Wilcke 1998 überzeugend demonstriert hat, ist das Sumerische von GA stark vom Akkadischen beeinflusst. Doch akkadisches kiška/ettû(m) hat nicht die Bedeutung ‘Handwerk’, in welchem Sinn auch immer. Bleibt die Übersetzung ‘Handwerker’. Etwas trickreich wandelt sie Römer in ‘Handwerker(haus)’ um (Römer 1980:55f.; 1993:552). ‘Handwerker’ ist nicht ausgeschlossen, die Übersetzung ‘Gießerei’ passt jedoch besser zum folgenden. Sie stützt sich auf die in CAD s. v. kiškattû 1 als ‘furnace, kiln, oven (of the smith, potter, etc.)’ angegebene Bedeutung. Vgl. auch George 2003:201 mit der Übersetzung ‘forge’ zu Gilgameš Y 161–163. Gedacht ist wohl an einen Vergleich zwischen einer Gießerei, wo die gewünschten Formen ausgegossen werden und Uruk, wo die Pläne der Götter unter Leitung des Königs in die Realität umgesetzt werden. 17

J. Keetman, Akka von Kiš und die Arbeitsverweigerer

21

als An seinen Wohnsitz gründete, hast du dich darum gekümmert, du bist König und Held!’20

Im göttlichen Auftrag Tempel bauen zu lassen und für die Gemeinschaft sinnvolle Arbeiten, wie die Errichtung einer Stadtmauer zu verlangen, wird dem Herrscher nicht verübelt. Im Gegenteil, es zeigt seine Befähigung zum Königtum. Auf den der sich durch solche Arbeiten die Freundschaft der Götter verdient hat, ist dann auch im Krieg Verlass. Das Bild wird auch nicht dadurch getrübt, dass man als bekannt voraussetzen kann, dass die Errichtung der Mauer um Uruk und der Bau von Eanna ein enormes Arbeitspensum dargestellt haben. Die Möglichkeit, auf solche berühmte Riesenwerke hinzuweisen, könnte durchaus für die Wahl des Schauplatzes und der Person des Gilgameš Bedeutung gehabt haben. In dem Sinne, dass selbst die Erbauer einer solchen Riesenmauer, Leute waren, die nur taten, was man von ihnen in rechtschaffener Weise verlangen konnte. Kaum haben sich die Urukäer dazu entschlossen, lieber zu kämpfen als der Forderung Akkas nachzugeben, da verschafft ihnen Akka auch schon die Gelegenheit dazu. Akka beginnt mit der Belagerung Uruks. Zunächst reagieren die Urukäer verschreckt, ihr Verstand wird verwirrt (Z. 50). Gilgameš fragt, ob ein Freiwilliger hinausgehen will, um den Verstand Akkas zu verwirren. Während Akka die Arbeit verweigert wurde, ist Gilgameš um einen Freiwilligen für die gefährliche Mission nicht verlegen. Es meldet sich sein Leibwächter Bir¶urture. Bir¶urture wird gleich am Tor gefangen genommen, geschlagen und Akka vorgeführt. Doch der Gefangene kündigt dreist den Zusammenbruch von Akkas Armee an. Als Grund nennt er das Erscheinen des Gilgameš. Heimpel hat dies als Ablenkungsmanöver gedeutet. 21 Dieser Trick mag durchaus eine Rolle spielen, aber uns erscheint es unwahrscheinlich, dass die Entscheidung nur durch einen einfachen Trick herbeigeführt wird. Auch sachlich ist diese Lösung schwierig, denn Akka steht nicht direkt am Tor und wie Dina Katz bereits eingewandt hat, 19

Diese und die folgende Zeile sind nach Text a gelesen, weil er das auffallende Bild des Nebels enthält. Außerdem macht die Konstruktion als pronominale Konjugation mit temporalem Lokativ mit dem Folgenden gut Sinn. Die anderen Texte drücken sich konventioneller aus und stellen keine so deutliche Verbindung zu der entscheidenden Z. 35 her. Für Z. 33 lautet die Alternative bàd gal bàd an-né ki ús-sa ‘Die große Mauer ist eine Mauer, die An auf die Erde gesetzt hat’ und am Ende von Z. 34 wird ŋar-ra-né eingesetzt. 20 Z. 30–35. 21 Heimpel 1981.

22

Articles: Ancient Near Eastern Studies

stürmt die Armee von Uruk nicht gleich heraus, sondern es ist zunächst nur Enkidu alleine, der das gefährliche Wagnis eingeht.22 Die Aussendung des Bir¶urture folgte als Reaktion auf die Verwirrung der Urukäer beim Erscheinen von Akkas Armee. Nun soll Akka durch das Auftreten eines einzigen Mannes ebenfalls verwirrt werden. Was ihm wirklich Sorgen bereiten könnte, ist die Treue der Urukäer und ihr Vertrauen auf Gilgameš, die Bir¶urture durch sein Herausgehen und seine kühnen Worte offenbart. Wir vermuten, dass damit auch der verborgene Sinn der Rede Bir¶urtures zusammenhängt. Bir¶urture schildert vor Akka ausführlich wie seine Armee zusammenbrechen wird. Diese Schilderung wird später wörtlich als Bericht über das wirkliche Geschehen wiederholt. Doch diese wörtliche Wiederholung verdeckt die Tatsache, dass sich das Geschehen nicht so abspielt, wie man es nach Bir¶urtures Schilderung erwarten würde. Als Gilgameš tatsächlich auf die Mauer tritt, wirkt sein Erscheinen nicht auf Akkas Leute, sondern auf die Urukäer, die in der Gasse am Tor zum Ausfall bereitstehen: egir zabar-déb unugki-ga-ke4 dbìl-ga-mes bád-šè im-me-e11-dè ab-ba di4-di4-lá kul-aba4ki-a-ke4 me-lám bí-ib-šú-šú ŋuruš unugki-ga-ke4 ŋištukul-mè šu-ne-ne b[í-in-si] ŋiš ig-abul-la-ka sila-ba bí-in-gub en-ki-du10 abul-la dili ba-ra-è ‘Nachdem Mundschenk von Uruk stieg Gilgameš auf die Mauer, Alt und Jung von Kulaba bedeckte der Glanz, die jungen Männer von Uruk ließ er die Kriegswaffen zur Hand nehmen. Er ließ (sie) auf der Gasse am Stadttor hintreten. Enkidu ging alleine aus dem Stadttor hinaus.’23

Die ‘Alten’ hier dürften auch die ab-ba-iri einschließen. Das kann man vielleicht als einen Hinweis darauf verstehen, dass in Uruk trotz des Anfänglichen Widerspruches der ‘Stadtväter’ kein ernsthafter Konflikt herrscht. Gilgameš wird dem ungerechten Akka als idealer Herrscher gegenübergestellt. Also darf es in seiner Stadt keine schwerwiegenden Probleme geben. Mag sein, dass die am Anfang geschilderten Nachfragen bei den verschiedenen Versammlungen24 hervorheben sollten, dass der gute Herrscher nicht nur auf einflussreiche Leute hören, sondern sich 22

Katz 1993:7, Anm. 24. Z. 84–88. 24 Zu den Versammlungen vgl. Katz 1987. 23

J. Keetman, Akka von Kiš und die Arbeitsverweigerer

23

auch um die Sorgen der kleinen Leute kümmern soll. Rein kompositorisch betrachtet bilden die Anfragen des Gilgameš bei den beiden Versammlungen das Gegenstück zu den beiden Fragen Akkas nach Gilgameš. Zurück zum Geschehen an der Stadtmauer von Uruk. Nach den Worten Bir¶urtures würden wir annehmen, dass der Feind bei Gilgameš’s Anblick zusammenbricht. Dass ein Feind beim Anblick des Schreckensglanzes zusammenbricht, wäre möglich. Doch von dieser Möglichkeit wird in den sumerischen Epen sonst kein Gebrauch gemacht.25 Wenn Bir¶urture Gilgameš beschreibt, so klingt aus seinen Worten mehr seine eigene Ehrfurcht als die Furchtbarkeit der Erscheinung des Gilgameš: lú-še lugal-ŋu10 in-nu lú-še lugal-ŋu10 ¶é-me-a saŋ-ki ¶uš-a-né ¶é-me-a igi alim-ma-ka-a-né ¶é-me-a su6 za-gìn-na-ka-a-né ¶é-me-a šu-si sa6-ga-né ¶é-me-a šár-ra la-ba-an-šub-bu-uš-àm šár-ra la-ba-an-zi-ge-eš-àm ‘Dieser Mann ist nicht mein König! (Was) dieser Mann soll mein König sein?! Seine zornige Stirne soll es sein?! Sein Wisentstierauge soll es sein?! Sein Bart von Lapislazuli soll es sein?! Seine wohlgeformten Finger sollen es sein?! (Aber) unzählige sind nicht gefallen, unzählige sind nicht aufgestanden!’26

Zwischen Furcht und Ehrfurcht wird im Sumerischen und Akkadischen ohnehin nicht streng unterschieden. Der Schrecken kann auch Anerkennung bedeuten. Wohl deshalb gehört der Besitz von me-lám zu den Grundvoraussetzungen eines Königs. Die lange Fluchformel am Ende des Kodex Hammurapi beginnt damit, dass Anu dem Übeltäter das me-lám des Königtums wegnehmen soll. Erst danach folgt, dass Anu auch sein Szepter zerbrechen und sein Geschick verfluchen soll.27 Nach diesen drei Sanktionen durch den höchsten Gott folgen die Sanktionen durch andere Götter, beginnend mit Enlil. In Gilgameš und Huwawa, setzt Huwawa seine ní-te gegen Gilgameš ein. Doch Huwawa ist kein menschliches Wesen und der Einsatz der ní-te führt keine Wende im Kampf herbei. 26 Z. 70–76. 27 CH Rs. xxvi 48–52. Weitere Belege CAD M2 11, 2′ u. c. 25

24

Articles: Ancient Near Eastern Studies

Das alles macht vorzüglich Sinn, wenn man mit me-lám hier in erster Linie die Ehrfurcht der eigenen Untertanen verbunden sieht, die die Voraussetzung für das Königtum ist. Akka kann den König der Urukäer nicht von seinem Mundschenk unterscheiden (eine verborgene Spitze?). Er muss einen Untertanen des Königs fragen. Das me-lám des Gilgameš wirkt offenbar nicht auf einen Fremden.28 Als Gilgameš wirklich auf die Mauer tritt ist Akka von seinem Anblick keineswegs physisch überwältigt. Sein Heer bricht aber in dem Moment zusammen, in dem Enkidu sagt, dass der Mann auf der Mauer sein König ist. lú-še lugal-ŋu10 ì-me-a bí-in-du11-ga-gin7-nam šár-ra ba-an-šub-bu-uš-àm šár-ra ba-an-zi-ge-eš-àm ‘Dies ist mein König! Wie er das gesagt hatte, fielen unzählige, standen unzählige auf.’29

Die Anerkennung durch seine Untertanen verleiht einem König alle Macht, auch über seine Feinde. Es ist das was Gilgameš hat und was Akka sich mit seiner Forderung verscherzt hat und nun vergeblich versucht mit Gewalt zurückzubekommen. Daneben mögen wir uns einen Ausfall der Leute von Uruk vorstellen. Mit der Bewaffnung und dem Aufmarsch der Urukäer am Tor wird Spannung erzeugt. Dann geht zur Überraschung des Lesers oder Hörers nur Enkidu hinaus. Während man gespannt entweder auf die Gefangennahme Enkidus oder den Beginn der Schlacht wartet, geschieht einen Moment nichts dramatisches, bis Enkidu das entscheidende Bekenntnis zu seinem König spricht. Der Zusammenbruch von Akkas Armee wird dann in Bildern geschildert, die das konkrete Geschehen nur erahnen lassen. Das Entscheidende ist nicht der Kampf, sondern der esprit de corps, während die technischen Einzelheiten der Schlacht keine Rolle spielen. 30 28

Der Umstand, dass Akka Gilgameš nicht kennt, hätte eigentlich schon lange stutzig machen sollen, denn einer Idee von Jacobsen folgend nahmen die meisten Bearbeiter an, es gebe eine Vorgeschichte, in der Gilgameš als Flüchtling von Akka aufgenommen wurde. Siehe Jacobsen 1987:345f. 29 Z. 92–94. 30 Aus der Rede Gilgameš’s an Enkidu in Z. 43–47 und der Schilderung von Akkas Gefangennahme (s. u.) kann man sogar vermuten, dass Akkas Krieger einfach vor Furcht auseinanderlaufen. Doch auch in dieser Rede des Gilgameš ist es nicht sein eigenes Erscheinen, sondern das Auftreten seines bewaffneten Untertan Enkidu, das mit der Furcht der Feinde verbunden wird: ne-šè ŋiššu-kára á mè

J. Keetman, Akka von Kiš und die Arbeitsverweigerer

25

Bir¶urture hat nicht direkt gelogen, es ist tatsächlich das Erscheinen des Gilgameš, das die Entscheidung herbeiführt, aber nicht weil es auf den Feind wirkt, sondern weil es auf seine eigenen Leute wirkt. Während die Urukäer für Gilgameš keine Gefahr scheuen, nützt einem ungerechten König à la Akka sein Heer gar nichts. Akka wird ‘inmitten seiner Truppe’ gefangen genommen.31 Das Thema des despotischen Herrschers, der zumindest anfangs kein Maß im Umgang mit seinen Untertanen kennt, ist innerhalb der Gilgameš-Epen noch mindestens zweimal zu belegen. In Gilgameš, Enkidu und die Unterwelt reitet Gilgameš auf den Söhnen der Witwen. 32 Im Akkadischen Epos mischt sich Gilgameš in ungehöriger Weise in das Privatleben seiner Untertanen ein, vielleicht indem er ein jus primae noctis für sich beansprucht.33 Die Unterdrückung ist jedes Mal etwas abstruses, was in einem realen Staat vermutlich gar nicht vorkommt. Wir denken, dass eben dies Teil der literarischen Verfremdung ist, weil ein offener Affront vermieden werden soll. Dies geschieht noch auf andere Weise. Die Unterdrückung steht jeweils am Anfang, ehe Gilgameš weise geworden ist und erfahren hat, dass auch er nur ein sterblicher Mensch ist. Die Unterdrückten sind das Volk von Uruk, von dem jeder weiß, dass es die berühmte Riesenmauer gebaut hat, also Leute, die für einen gerechten König durchaus zupacken. Schließlich gewinnt der König auch einen Untertanen zum Freund, nämlich Enkidu. In Gilgameš und Huwawa gerät dieser Enkidu darüber, dass ihn Huwawa einen ‘Mietling’ lú-¶un-ŋá nennt, so in Zorn, dass er ihm den Kopf abschlägt. Gilgameš ist nicht nur ein großer König, er hat auch selbstlos dienende Untertanen, wenn er sie nicht aus Lust und Laune überfordert. Auch in GA wird die Ungeheuerlichkeit, dass einem König schlicht die Arbeit Verweigert wird, durch Beispiele großer Treue zum Herrscher ausbalanciert. Da ist zunächst die Treue der Urukäer zu dem sa ¶é-em-mi-gi4 / ŋištukul-mè á-zu-šè ¶é-em-mi-gi4 / ní-gal me-lám-ma ¶é-emdím-dím-me / e-ne ŋen-a-né-ta ní-gal-ŋu10 ¶é-eb-šú ‘Nun sollen die Šukara(-Waffe), der Arm der Schlacht und das (Fang-)netz zurückkehren. / Die Kriegswaffe soll an deinen Arm zurückkehren. / Mit dem Schreckensglanz sollen sie große Furcht machen. / Nachdem er gekommen ist, bedeckt ihn (dann) große Angst vor mir’ (Z. 43–46). 31 Siehe Wilcke 1998:462 zu dem Akkadismus šà erén-na-ka-né (Z. 99). 32 Dazu Keetman 2007. 33 Hierzu Keetman 2008.

26

Articles: Ancient Near Eastern Studies

König, der sie fordert, wo es nötig ist, aber vor Überforderung schützt. Zudem werden noch zwei Einzelbeispiele von großer Treue vorgeführt. Bir¶urture, der trotz Folter und Lebensgefahr sich auch als Gefangener zu Gilgameš bekennt und dem ein erheblicher Teil des Textes gewidmet ist. Sodann Enkidu, der trotz der Gefangennahme Bir¶urtures ebenfalls alleine aus dem Tor geht. Am Schluss verhält sich Gilgameš gegenüber seinem bisherigen Oberherrn Akka zumindest nobel, indem er ihn nicht wie dieser zuvor Bir¶urture quält, sondern sogar freilässt.34 Gilgameš hat nicht vergessen, dass sich Akka als Herrscher auch Verdienste erworben hat. Bisher wurde angenommen, dass die Freilassung Akkas aus Dankbarkeit für eine Wohltat erfolgte, die er früher Gilgameš persönlich erwiesen hatte. 35 Doch es gibt gewichtige Gründe gegen diese Deutung. Die angebliche Vorgeschichte taucht unversehens wie ein deus ex machina in der Rede des Gilgameš auf und das in poetischer Umschreibung, die an sich keinen Hinweis auf ein spezielles Ereignis enthält. Akka kennt Gilgameš offenbar nicht persönlich. In Z. 110 sagt Gilgameš, dass die Stadt ihm seine Wohltat vergolten hat (šu-ŋu10 …gi4, s. u.) und in der folgen34

In Text b wird die Rede des Gilgameš zu Akka durch die Zeile [ak]-kà-aš DI mu-na-ab-bé eingeleitet. Wilcke (1998:480) hat sich zu recht gegen sá du11 ‘erreichen’, ‘ereilen’ ausgesprochen und plädiert für di du11 ‘prozessieren’. Dies passt dazu, dass die Rede des Gilgameš schließlich mit einer Freilassung endet und dass der Richtergott Utu genannt wird. Der ideale Herrscher würde hier auch noch in seiner Funktion als gerechter Richter vorgeführt, der auch die Verdienste des Schuldigen wägt und ihm am Schluss am eigenen Beispiel einen Rat gibt anstatt zu strafen. Man kann aber auch an silim(-šè) du11 ‘(an)preisen’; ‘lobhudeln’ denken (nicht zu verwechseln mit silim-ma du11 ‘grüßen’). Z. B. ka-tuku kušlu-úb-a-né silim im-du11 / gal-gal-di kušlu-úb sù-ga ša-mu-un-de6 / silim-du11 enim sù-ga šaba-ni-ib-ŋar ‘Das Großmaul preist seinen Beutel, / doch der Großtuer bringt einen leeren Beutel, der Lobhudler setzt ein leeres Wort hin’ (Instructions of Šuruppak 106–108), siehe Alster 2005:75–76 mit vielen Varianten. Sulge besteht darauf, dass sein Selbstpreis nichts mit silim du11 zu tun hat: ní-ŋu10 silim-šè-àm ba-ra-ab-du11 ‘Ich habe mich gewiss nicht übertrieben gepriesen!’ (Sulge A 84), siehe Klein 1981:200 mit einigen Varianten. Es fällt auf, dass silim du11 nicht die normalen Worte für ‘rühmen’, ‘preisen’ nâdu, dalālu als Gleichungen hat, sondern nur šarā¶u Dt (siehe CAD Š2 37a, šarā¶u A lex. section; CAD M2 287a, muštarri¶u lex. section). Obwohl die negative Nuance deutlich ist, kann hier wie beim Selbstpreis des Enki in Enki und die Weltordnung, das Verbum nicht ganz darin aufgehen, wenngleich man die folgende Rede des Gilgameš als nahe an der Lobhudelei sehen kann. Aber ist es möglich zu huldigen ohne zu lobhudeln? 35 Siehe Anm. 28.

J. Keetman, Akka von Kiš und die Arbeitsverweigerer

27

den Zeile ist von der Vergeltung früherer Wohltaten die Rede (šu u4-béta …gi4), nicht von der Vergeltung ‘deiner früheren Wohltat’ (šu-zu u4bé-ta …gi4), wie man als Parallele erwarten könnte. Vor allem aber ist die Verbalform in Z. 111 nicht die 1. Person marû, sondern eine 3. Person ¶am¢u: ‘man vergilt dir die früheren Wohltaten’. Akka wird von Gilgameš am eigenen Beispiel belehrt, dass die Untertanen erwiesene Wohltaten nicht vergessen. Von einer ihm persönlich erwiesenen Wohltat spricht Gilgameš hingegen nicht. In seiner abschließenden Rede wiederholt Gilgameš auch die Rede der Arbeiter, mit der sie ihn aufgrund seiner Führung bei den für die Gemeinschaft notwendigen Bauten als König anerkannt haben. Daraus wurde geschlossen, dass es sich um eine allgemeine Formel für die Anerkennung als König handelt und dass sie nun Gilgameš gegenüber Akka erneut gebraucht. Doch diese Sicht übergeht den Inhalt der Aussagen, nämlich die Bauten in Uruk, insbesondere von dessen berühmter Stadtmauer. Das kann sich nur auf Gilgameš selbst beziehen. Gilgameš zählt noch einmal beispielhaft auf, was ihn selbst als König legitimiert. Die Schlussfolgerung, die die Arbeiter aus seinen Verdiensten gezogen haben ‘du bist König und Held’ wiederholt Gilgameš aber nicht, sondern spricht vom Vergelten von Wohltaten.36 unugki ŋiš-kin-ti diŋir-re-e-ne-ke4 bàd gal muru9 ki ús-s[a-a-b]a ki-tuš ma¶ an-né ki ŋar-ra-a-ba saŋ mu-e-sì šu-ŋu10 im-mi-ib-gi4 igi dutu-šè šu u4-bé-ta e-ra-an-gi4 ak-kà kiški-šè šu ba-ni-in-ba d bìl-ga-mes en kul-aba4ki-a-ke4 zà-mí-zu du10-ga-àm ‘Uruk ist die Gießerei der Götter! Als die große Mauer (wie) eine Wolke die Erde berührte, als An seinen erhabenen Wohnsitz gründete,

36

Katz (1993:6) wendet sich in Anlehnung an Jacobsen gegen die Auffassung, dass es hier um eine Formel geht, mit der Gilgameš als militärischer Führer für eine gewisse Zeit anerkennt wird, was der ursprünglichen Bedeutung von lugal entsprechen soll: “Since the last episode took place after the war had ended, and lugal is probably a war time leader, there is no point in nominating a lugal.” Durch Wilckes Datierung in die altbabylonische Zeit, in der der Titel lugal seit langem den permanenten Herrscher bezeichnet, ist diese Diskussion heute ohnehin überholt.

28

Articles: Ancient Near Eastern Studies habe ich mich um sie gekümmert – sie (die Stadt Uruk) hat meine Wohltat vergolten.37 Vor Utu vergilt man dir die frühere Wohltat.38 Akka ließ er nach Kiš frei. Gilgameš, En von Kulaba, dein Loblied ist gut.’

37

Vgl. Wilcke 1998:480. Hier steht ¶am¢u zum Ausdruck einer zeitlosen Aussage. Es geht nicht darum, dass dies im Augenblick oder gleich geschieht, sondern es handelt sich um eine allgemeine Regel. Utu wird erwähnt, weil er der Gott der Gerechtigkeit ist. Der Bezug auf Utu unterstreicht noch einmal den allgemeinen Charakter der Aussage. 38

J. Keetman, Akka von Kiš und die Arbeitsverweigerer

29

Bibliographie Alster 2005 Brisch 2007

Cavigneau–Al Rawi 2000 Cooper 1981 Falkenstein 1966 George 2003 Heimpel 1981 Jacobsen 1939 Jacobsen 1957 Jacobsen 1987 Katz 1987 Katz 1993 Keetman 2007

Keetman 2008

Keetman 2010 Klein 1981 Kramer 1949 Lambert 1980 Loesov 2004 Michalowski 1982 Michalowski 2006

Römer 1980 Römer 1993

Alster, B. Wisdom of Ancient Sumer. Bethesda. Brisch, N. M. Tradition and the Poetics of Innovation. Sumerian Court Literature of the Larsa Dynasty (c. 2003–1763) (AOAT 339). Münster. Cavigneau, A.; Al-Rawi, F. N. H. Gilgameš et la mort. Textes de Tell Haddad VI (CM 19). Groningen. Cooper, J. S. Gilgamesh and Akka: A Review Article. JCS 33:224–239. Falkenstein, A. Zu Gilgamesch und Agga. AfO 21:47–50. George, A. R. The Babylonian Gilgamesh Epic. Oxford. Heimpel, W. A Note on Gilgamesh and Agga. JCS 33: 242–243. Jacobsen, Th. The Sumerian King List (AS 11). Chicago. Jacobsen, Th. Early Political Development in Mesopotamia. ZA 52:91–140. Jacobsen, Th. The Harps That Once… Sumerian Poetry in Translation. New Haven. Katz, D. Gilgamesh and Akka: Was Uruk Ruled by Two Assemblies? RA 81:105–114. Katz, D. Gilgamesh and Akka. Groningen. Keetman, J. König Gilgamesch reitet auf seinen Untertanen. Gilgamesch, Enkidu und die Unterwelt politisch gelesen. BiOr 64:5–31. Keetman, J. Der Kampf im Haustor. Eine der Schlüsselszenen zum Verständnis des Gilgameš-Epos. JNES 67: 161–173. Keetman, J. Enmerkar und Sulge als sumerische Muttersprachler nach literarischen Quellen. ZA 100:15–31. Klein, J. Three Šulgi Hymns: Sumerian Royal Hymns Glorifying King Šulgi of Ur. Ramat-Gan. Kramer, S. N. Gilgameš and Agga. AJA 53:1–18. Lambert, W. G. Akka’s Threat. Or 49:339–340. Loesov, S. I-nu-ma-i-lu-a-wi-lum. B&B 1:183–202. Michalowski, P. Rezension Römer 1980. BSOAS 45:557– 578. Michalowski, P. The Strange History of Tumal. Michalowski, P.; Veldhuis, N. (Hrsg.). Approaches to Sumerian Literature. Studies in Honour of Stip (H. L. J. Vanstiphout) (CM 35). Leiden–Boston. Pp. 145–166. Römer, W. H. Ph. Das sumerischen Kurzepos ‘Bilgameš und Akka’ (AOAT 209/1). Neukirchen-Vluyn. Römer, W. H. Ph. Mythen und Epen. I (TUAT III/3). Gütersloh.

30 Römer 2001 Selz 1998

Selz 2003

Steible 1975 Steinkeller 2003

van Dijk 1998

Vanstiphout 1986 Vanstiphout 1987 Wilcke 1989

Wilcke 1998

Articles: Ancient Near Eastern Studies Römer, W. H. Ph. Hymnen und Klagelieder in sumerischer Sprache (AOAT 276). Münster. Selz, G. Über Mesopotamische Herrschaftskonzepte. Zu den Ursprüngen mesopotamischer Herrscherideologie im 3. Jahrtausend. Dietrich, M.; Loretz, O. (Hrsg.). dubsar anta-men. Studien zur Altorientalistik. Fs. für Wilhelm H. Ph. Römer zur Vollendung seines 70. Lebensjahres mit Beiträgen von Freunden, Schülern und Kollegen (AOAT 253). Münster. Pp. 281–344. Selz, G. Who is Who? Aka, König von Øiš(š)a: zur Historizität eines Königs und seiner möglichen Identität mit Aka, König von Kiš. Selz, G. (Hrsg.). Festschrift für Burkhart Kienast zu seinem 70. Geburstage dargebracht von Freunden, Schülern und Kollegen (AOAT 274). Münster. Pp. 499–518. Steible, H. Rīmsîn, mein König (FAOS 1). Wiesbaden. Steinkeller, P. An Ur III Manuscript of the Sumerian King List. Sallaberger, W. et al. (Hrsg.). Literatur, Politik und Recht in Mesopotamien. Fs. Claus Wilcke (OBCh 14). Wiesbaden. Pp. 267–292. van Dijk, J. J. A. Inanna raubt den ‘großen Himmel’. Ein Mythos. Maul, St. (Hrsg.). tikip santakki mala bašmu. Festschrift für Rykle Borger zu seinem 65. Geburtstag am 24 Mai 1994 (CM 10). Groningen. Pp. 9–38. Vanstiphout, H. L. J. Towards a Reading of “Gilgameš and Agga”. Pt. II. Construction. OLP 17:33–50. Vanstiphout, H. L. J. Towards a Reading of “Gilgameš and Agga”. AuOr 5:129–141. Wilcke, C. Genealogical and Geographical Thought in the Sumerian King List. Behrens, H. et al. (Hrsg.). DUMU-E2-DUB-BA-A. Studies in Honor of Åke W. Sjöberg (OPSNKF 11). Philadelphia. Pp. 557–572. Wilcke, C. Zu “Gilgameš und Akka”. Überlegungen zur Zeit von Entstehung und Niederschrift, wie auch zum Text des Epos mit einem Exkurs zur Überlieferung von “Šulgi A” und von Lugalbanda II. Dietrich, M.; Loretz, O. (Hrsg.). dubsar anta-men. Studien zur Altorientalistik. Fs. für Wilhelm H. Ph. Römer zur Vollendung seines 70. Lebensjahres mit Beiträgen von Freunden, Schülern und Kollegen (AOAT 253). Münster. Pp. 457–485.

The Old Babylonian Omens in the Pushkin State Museum of Fine Arts, Moscow∗ Ilya Khait Russian State University for the Humanities, Moscow

The Pushkin Museum cuneiform collection hosts eight OB omen compendia. Two of them, published by Vladimir K. Schileico in AfO 5 (1929), are well known to Assyriologists. The remaining six were edited with the Russian translation by Yury A. Saveliev in a difficult-to-obtain Pushkin Museum bulletin.1 Both publications are now outdated. The present paper provides a re-edition of seven of these texts with philological commentaries. The eighth one2 has been edited in Khait 2011. No. here 1 2 3 4 5

Museum No. I2б 1724 I2б 1723 I2б 1325 I2б 1664 I2б 1663+1661

Size (cm) 2.1 × 4.3 3.4 × 6.9 2.2 × 5.9 2.5 × 5.5 2.8 × 7.8

6 7 8

I2б 1660 I2б 1662 YBC 11102+I2б 1653

3.8 × 7.0 1.8 × 5.4 3.3 × 8.3 (3.3 × 2.5 + 3.3 × 5.8)

Primary publication Schileico 1929, No. 1 Schileico 1929, No. 2 Saveliev 1985, No. 2 Saveliev 1985, No. 1 Saveliev 1985, Nos. 3 and 6 Saveliev 1985, No. 7 Saveliev 1985, No. 5 YOS 10, 30; Saveliev 1985, No. 4

Provenance and orthographic peculiarities The exact provenance of the texts is unknown. Both amūtu-texts published by Schileico (Nos. 1 and 2) are listed as belonging to the private ∗ This article is dedicated to the memory of Yury Alexandrovich Saveliev, the curator of the Pushkin Museum cuneiform collection who passed away in April 2008. I am deeply indebted to Dr. Leonid Kogan and Dr. Sergey Loesov for informative discussions, suggestions and valuable assistance during my work on this article. The research has been supported by RFBR grant No. 11-0600418-а. 1 Saveliev 1985. 2 It is the fragment I2б 1653, a join to YOS 10, 30 (YBC 11102).

32

Articles: Ancient Near Eastern Studies

collection of N. P. Likhachyov (purchased at the beginning of the past century from antique dealers in Paris and London, nationalized in 1930, one year after the publication of Schileico’s article, and divided between the Hermitage Museum and the Pushkin Museum). The Museum’s cuneiform inventory has a record about one more arrival from this collection, viz. one of the two naplaštu-texts (No. 3) that was earlier sold by Likhachyov and entered the Pushkin Museum in 1924. The origin of the remaining tablets is not specified in the inventory. There are reasons to believe that all the tablets once belonged to Likhachyov’s collection. This follows from the origin of No. 1 and 2 (registered as Museum arrivals of 1937, donated by Schileico’s widow3), as well as from the general information about the sources of the Museum’s cuneiform collection.4 I believe that No. 3 was the first tablet sold by Likhachyov whereas Nos. 4–8 entered the Museum after the nationalization of Likhachyov’s collection in 1930, while the last two tablets, Nos. 1 and 2, are originally from Likhachyov’s collection as well. The omen compendia of the Pushkin Museum have a common provenance. This is clear from the fact they follow the same scribal tradition; even the ductus seems to be the same. All the tablets were exposed to fire5 and have an oblong rectangle shape. Two texts are dealing with amūtum (Nos. 1 and 2), two more, with naplaštu (Nos. 3 and 4), and others deal with abullum (No. 8), padānu (No. 5) and pi¢ru (No. 6). The feature referred to in the last fragment (No. 7) is unknown, as its protases are badly damaged. Each compendium comprises a comparatively small number of records: one (Nos. 1, 3 and 4), two (?) (No. 7), three (No. 2), six (No. 5), nine (No. 8) and fourteen (No. 6). The present writer has established that the tablet I2б 1653 is a distant join to YOS 10, 30 (the composite text, published in Khait 2011, has been described as No. 8 in this introduction). This fact allows one to suppose that other tablets from the Yale collection may have the same provenance. All of the eight texts are small scholarly tablets with short exercises which may be related to other known tablets of this type: YOS 10, 6 and YOS 10, 3

Vladimir K. Schileico passed away in 1930, the year N. P. Likhachyov was arrested, so some of the tablets from this collection that remained in Schileico’s hands were handed to the Museum in 1937 by his widow, Vera K. Schileico. 4 See Perlov 2000. 5 All of them are half-baked and black, except for No. 6 that seems to be well-baked.

I. Khait, The Old Babylonian Omens in the Pushkin Museum …

33

12 (OB Izbu); YOS 10, 21; YOS 10, 27; YOS 10, 28; YOS 10, 37; YOS 10, 43; YOS 10, 50 and OBE 12.6 The texts display the same South OB orthographic conventions7 that are typical of the bulk of the YOS 10 tablets. The texts employ the PI sign to render /pi/. The sequence /¢V/ is expressed by DU, TE,8 and DA (?).9 Noteworthy are the South OB dialectal forms i-na-an-di-in and in-na-andu. An interesting phonological peculiarity is the use of k-signs at the beginning of Sumerian loanwords starting with g (ki-ir-ri, ka-ba-ra-ḫu). The combinations “dental + sibilant” vs. “sibilant + sibilant” are rendered by different signs, viz. ZU (wa-ar-ka-sú) vs. SU (re-su, i-ri-su, te! (wr. PI)-pu-su)10. The use of CVC signs as phonetic complements (lúKÚR-rum) is, within Goetze’s typology, also a Southern feature. A few less common CVC signs are in evidence (¶ar, ¶um!, qar, ´um, ¢ar), which may point to a relatively late date of the texts. Frequent omission of mimation points in the same direction. Logographic writings of a few fundamental concepts are attested in both the protases (gišTUKUL, KÁ É.GAL, ŠU.SI, MURUB4, GABA) and the apodoses (lúKÚR, DINGIR, LÚ, KASKAL, gišGIGIR, NÍG.ḪUL, also the theonyms d UTU, dGÌR.UNUG.GAL, dNIN.GIŠ.ZI.DA). The conjunction šumma at the beginning of the protasis is usually written logographically (DIŠ). The only exception is text No. 6, where it is written syllabically (šum-ma).

6

See the OB omen compendia typology in Glassner 2009:5–6. Goetze 1945:146–147. 8 Note that the usage of ¢e4 is traditionally considered a northern feature (Goetze 1945:146). 9 See commentary on text No. 5, l. 1. 10 Goetze 1958. 7

34

Articles: Ancient Near Eastern Studies

Texts and commentaries 1 Obverse (1–2) Reverse (3–4) §1

1

wa-ar-ka-at a-mu-/tim \ 2 šumma warkat amūtim šinā kakkū ša-ak-nu-ma 2li-ib-bi KÁ šaknū-ma libbi bāb ekallim i¢¢ulū Ll. 1–4 É.GAL i¢-¢ù-lu 3ša-al-gu še-a-am išalgu še’am ima¶¶a´ šarram ina ma-/¶a\-a´ 4šar-ra-am i-na li-ib-bi \ libbi ālīšu idukkūšu a-li-šu i-du-ku-ú-šu ‘If at the rear side of the Liver there are two “Weapons” pointing towards the “Palace Gate”: snow will destroy the barley; the king will be killed in his own city.’ DIŠ

giš

TUKUL

§ 1 Ll. 3–4 For the first apodosis cf. YOS 10, 25:45 ša-al-gu še-a-am i-ma-¶a-a´-ma e-buúr ma-tim ú-/ul i-še-er ‘Snow will destroy the barley and the crop of the land will not prosper.’ The second apodosis has many variants. Cf. e. g. YOS 10, 31 v 10–12 (šar-ru-um i-na li-bi É.GAL-li-š/u i-du-ak ‘The king will be killed in his own palace’); YOS 10, 46 iv 20–22 (šar-ra-am i-na li-ib-bi É.GAL-šu ú-sà-ru-ú-šu!-ma i-du-uk-ku-šu ‘The king will be imprisoned and killed in his own palace’); YOS 10, 22:20 (šar-ra-am i-na ŠÀ-bi KÁ É.GAL-šu i-du-uk-ku-ú-šu ‘The king will be killed within his own palace gate’); Scheil B 23 (be-lam i-na šu-ub!-ti-šu gišTUKUL i-da-\ak]-šu ‘A weapon will kill the lord in his own dwelling place’). 2 Obverse §1

1

DIŠ a-mu-tum ša-ar er-bi-ti-ša ši-pešumma amūtum šār erbettīša tim ma-/li-a-at 2a-mu-ut dGÌR.UNUG. šēpētim maliat amūt Nergal u Ll. 1–2 GAL ù dNIN.GIŠ.ZI.DA \ ša mu-ta-ni ù Ningišzida ša mūtānī u maqlâtim ma-aq-la-tim ‘If the four sides of Liver are filled with “Footmarks”: the omen of Nergal and Ningišzida (which is) of plagues and fires.’

§2

3

DIŠ a-mu-tum e-li a-mu-tim ra-ak-bašumma amūtum eli amūtim rakbat at 4¢e4-e-em ma-tim i-ša-an-ni dUTU ¢ēm mātim išanni Šamaš šarram ša L. 3–5 šar-ra-am 5ša ra-ma-ni-šu a-na ma-ti- ramānīšu ana mātīšu išakkan šu i-ša-ka-an ‘If the Liver lays above (another) Liver: the loyalty of the land will change, Šamaš will set his own king for his land.’

I. Khait, The Old Babylonian Omens in the Pushkin Museum …

35

6

§3

DIŠ a-mu-tum ši-li sà-a¶-ra-at-ma ù šušumma amūtum šīlī sa¶rat-ma u te-eb-ru 7a-mu-ut šar-rum-ki-in ša ek-le- šutebrû amūt Šarrumkīn ša ekletam L. 6–7 tam i-i¶-bu-tu-m/a \ nu-ra-am i-mu-ru i¶butu-ma nūram īmuru ‘If the Liver is filled with holes, and they reach all the way through: the omen of Sargon who had been passing through darkness and saw the light.’

§ 1 Ll. 1–2 Schileico reads ŠE-PI-TIM as ši-we-tim and understands it as a variant of šimtu (cf. his translation: ‘Die Leber ist … mit Marken übersät’). Apparently, the word is šēpu, a well-attested extispicy feature. The duo of Nergal and Ningišzida is also attested in YOS 10, 49:1: DIŠ GAG.TI

i-di e-´e-em-´e-ri-im ti-i´-bu-ta-ma i-di GABA a-ḫe-e i-bi-ša i!-kal \ dNIN.GIŠ.ZI! (wr. NAM).DA ma-tam i-qá-al-lu ‘If the False Ribs are joined at the Spine, but separated at the Breastbone: the God (i. e. Nergal) will eat, Ningišzida will burn the land.’ DINGIR-lum

Also remarkable is the sequence of apodoses in YOS 10, 33 iv 51–52 and YOS 10, 33 iv 54, in which Ningišzida and Nergal (causing conflagrations11 and plague respectively) are mentioned successively. No other attestations of these gods as a duo are known to me.12 § 2 Ll. 3–5 The apodosis is remarkable. Despite the fact that the concept of the divine origin of the kingship and the necessity of the divine patronage for the king was trivial for ancient Mesopotamians and is well-attested in texts of many genres, the present example is almost unique in omen compendia. The closest parallels to this omen are probably found in apodoses involving the expression palê DN, such as YOS 10, 56 iii 12–13:

11

The association of Ningišzida with conflagrations is also reflected in YOS 10, 45:21, 27, 53 and YOS 10, 33 iv 50–51. 12 This pairing might have something to do with the duo of Meslamtaea/Lugalirra (Dingirgirra), a well-attested pair of chthonic gods (see Lambert 1989–1990). Cf. the following apodoses: a-mu-ut dLU[GAL.IR9.RA] ù dMES.LAM.TA.È.A ša mu-ta-[nu] i-na m[a-tim] ‘The omen of Lugalerra and Meslamtaea, which (predicts) plague in the land’ (YOS 10, 15:23–24), a-mu-ut dLUGAL.IR9.RA ù d MES.LAM.TA.È.\A] gišTUKUL ra-bu-um i-na i-di um-ma-\ni]-ka i-la-[ak] ‘The omen of Lugalirra and Meslamtaea; a mighty weapon (i. e. a plague) will help your army’ (YOS 10, 15:21–22), a-mu-ut DINGIR.GÌRìr.RA ù dME[S.LAM.T]A.È.A ša i-na i-di um-mani-\ka i]-la-\ku] ša-nu šum-šu mu-ta-a-nu i-\na ma-tim] i-ba-aš-{ŠI}-šu ‘The omen of Dingirgirra and Meslamtaea, who will help your army; second meaning—there will be a plague in the land’ (YOS 10, 17:37).

36

Articles: Ancient Near Eastern Studies

BAL dEN.LÍL

ma-tum i-ši-ir LUGAL šu-mi-ri-im ma-[tam] i-bé-el ‘The reign of Enlil; the land will prosper, the king of Sumer will rule the land.’13 Also noteworthy is another OB omen that mentions Šamaš in connection with the Land: dUTU ma-as-sú a-\n][a-am (?)] i-ip-pa-al ‘Šamaš will give his land a positive answer (?)’ (YOS 10, 33 iv 4–5). The omen evidently predicts the change of reign (perhaps even dynasty) through the agency of Šamaš, which could amount to either a rightful rule in general or to the rule of someone enjoying Šamaš’s patronage (maybe even the hegemony of Sippar). § 3 Ll. 6–7 This historical omen is known in several variants:14 a-mu-ut šar-ru-ki-in ša ek-le-tam el-li-ku-ma nu-ru-um ú-´i-aš-šu-um ‘The omen of Sargon who had been walking through darkness and the light came out for him’ (Scheil B 16–17). BA3-ut LUGAL.GI.NA šá ana kurMar-¶a-ši GEN-ma dXV ina MÚ KUKKU2 UD-šú È ‘The omen of Sargon who went to the land of Mar¶aši and Ištar appeared for him in a burst of light while he was passing through darkness’ (TCL 6, 1rev.:1). For the reflection of this motive in Res Gestae Sargonis see Goodnick Westenholz 1997:69–71. Cf. also the Old Assyrian Sargon Legend, lines 40–43.15 3 Obverse §1

1 DIŠ i-na i-ši-id na-ap-{\la?]}-la-a[ššumma ina išid napla[štim] šullum tim] 2šu-ul-lum a-wi-lam DINGIR-šu awīlam ilšu šut[eqrubam (?)] irrissu Ll. 1–3 šu-t[i?-eq-ru-ba-am (?)] 3i-ri-su ‘If at the base of the “View” (there is) a wart: the god of the man demands from him repeated petitions (?).’

13

For a general interpretation of this expression s. Jacobsen 1943:170–172. For other apodoses including this expression, cf. Leichty 1970 Izbu V 94 (Enlil), id. II 22 and 66; CT 38, 35:54; CT 20, 32:78; TCL 6, 1:16 etc. (Nergal). In general, the rule of Enlil seems to be associated with prosperity and the South Mesopotamian hegemony, whereas the rule of Nergal with enemy attacks and military collapse. On palê Ištar in the Chronicle of Early Kings s. Grayson 2000: 46, fn. 32. 14 See also Goetze 1947:255–256. 15 See Dercksen 2001; Dercksen 2005:109, 113f.; Cavigneaux 2005:599; Alster– Oshima 2007:9–10, 14.

I. Khait, The Old Babylonian Omens in the Pushkin Museum …

37

§ 1 L. 2 Identification of the sign fragment after ŠU is debatable. Saveliev apparently reads GAL (ilīšu rabû), but such a reconstruction is not convincing paleographically (only one lower horizontal wedge can be seen, the expected upper wedge is missing) or typologically (what one usually finds is either the personal god of a man or a great god with no pronominal suffix). My reconstruction is based on a similar apodosis in YOS 10, 52 iii 3 (ìlí LÚ šu-te-eq-ru-ba-am i-ri-iš ‘the god of the man demands repeated petitions,’ and see the duplicate text YOS 10, 51 iii 3). Since the extant traces do not resemble the small Winkelhacken at the left extremity of TE, one is forced to assume a variant spelling with TI. In this case the line would have to continue on the reverse. 4 Obverse §1

1

šumma imitti naplaštim ana šinā DIŠ i-mi-ti na-ap-la-aš-tim \ a-na ši-na pa-li-iš 2 lúKÚR-rum a-lam i-na pališ nakrum ālam ina pilšim Ll. 1–2 pi-il-ši-im \ \ú]-še-er-ri-da ušerrida ‘If the left (side) of the “View” is pierced two times: the enemy will bring the city down by (making) a breach (in its wall).’

§ 1 L. 1 The expression ana šinā pališ is strange. It is hardly synonymous to adi N pališ ‘pierced N times’16 because of the difference in prepositional usage. It is probably a mistake for the frequent expression ana N pa¢er ‘split into N parts.’17 Still another possibility would be to understand it in the sense of piercing through, alongside more frequent hendiadys expressions with that meaning, like ipluš-ma uštebri or ilpuš-ma ū´i (s. CAD P 60–61). L. 2 This usage of warādu Š seems to be unique in omina, although warādu G is attested in apodoses with suitable meanings. The CAD distinguishes between two different—to some extent, even opposite—meanings of the verb in omen apodoses: ‘to go down’ (i. e. ‘to raid’ in military contexts) and ‘to abandon (a fort).’18 16

See, for instance, TCL 6, 1rev.:37f. Cf. YOS 10, 26 iii 25–26: BAR sí-pi šu-me-el KÁ É.G[AL] a-na er-bi-i-šu pa-¢e4-e[r] a-\al] na-ak-ri-im i-na pi-il-ši-im ta-´a-ba-[at] ‘If the left “Doorframe” of the “Palace Gate” is split in four: you will conquer the enemy’s city by (making) a breach.’ 18 For a similar usage of the Biblical Hebrew ‫ ירד‬in Dt 20:20 and perhaps in La 1:9 see BDB 433, mng. 1k (courtesy L. Kogan). 17

38

Articles: Ancient Near Eastern Studies

The first meaning is manifest in YOS 10, 36 iii 4–5: ši-ip-ka a-na maa-at lúKÚR ur-ra-ad-ma wa-´i-a-am i-le-eq-qé ‘Your expedition will go down to the enemy land and capture the emerging (enemy force)’ (cf. also ibid. 6–7, 8–9 as well as YOS 10, 42 iv 20).19 The second meaning is, to the best of my knowledge, unattested in OB, but the translation ‘to capitulate’ seems to be compelling in a few post-OB examples (CAD A2 217). As for the passage under scrutiny, the easiest way of interpretation is to consider the verbal form as causative to the basic meaning of warādu: ‘to force to go down’ > ‘to make capitulate.’20 Within an alternative solution, warādu Š could be understood as more or less synonymous to the intransitive use of erēbu Š, well attested in similar contexts from OB omens with the meaning ‘to raid’21. In my view, this possibility is to be seriously considered, but in view of its unorthodox nature I preferred to keep the first option in my translation. 522 Obverse §1

1

DIŠ pa-da-an i-mi-tim re-su a-na ešumma padān imittim rēssu ana l[e-nu e-k]i-im i-mi-tum i-mi-tam i- el[ēnu ek]im imittum imittam īkim L. 1 ki-im k/a-ba-ra-¶u kabara¶¶u ‘If the tip of the “Path” of the right is taken upwards (and) the right took the right: panic.’ 2

DIŠ pa-da-nu i-na MURUB4-šu ašumma padānu ina qablītīšu ana na e-l[e-nu e-ki-i]m (eras.) te-šu-um el[ēnu eki]m tēšûm ummān ana L. 2 um-ma-an a-na / KASKAL i-la-ku \ ḫarrān illaku inninnâ-ma iturra in-ni-na-ma i-tu-ra ‘If the “Path” in its middle is taken upwards: panic; the army that will go to the campaign will be substituted and return.’

§2

19

Another passage with warādu from the very same text is as follows: a-al u4-ma-am ur-ra-da-am (YOS 10, 36 i 49). CAD oddly relates it to the second meaning and translates ‘The enemy city will capitulate today.’ In this particular OB text, the problem can be avoided by considering urradam as first person form and translating the prediction as ‘Today I will raid the enemy city.’ 20 Note the usage of warādu Š with direct object (usually, šallatu) in royal inscriptions in the meaning to ‘bring back (booty)’ (CAD A2 217–218). 21 See CAD E 273. 22 The identification of these two fragments as joins belongs to the present writer (in Saveliev 1985, they are treated as unrelated pieces). lú

KÚR-im

I. Khait, The Old Babylonian Omens in the Pushkin Museum … §3

3

§4

4

39

DIŠ pa-da-an šu-me-lim re-su a-na šumma padān šumēlim rēssu ana e-l[e-nu e-ki]-im i-mi-it-tum šu-me- el[ēnu ek]im imittum šumēlam ītekim L. 3 lam i-te-ki-/im \ [a]-la-ak-ti ˹lú˺KÚR- [a]lakti nakrīka taša¶¶i¢ ka ta-ša-¶i\-i¢ ‘If the tip of the “Path” of the left is taken upwards (and) the right took the left: you will attack your enemy’s expedition by surprise.’ DIŠ pa-da-an i-mi-tim re-su a-na š[ašumma padān imittim rēssu ana ap-la-nu] \e]-ki-im šu-me-\lum] i- š[aplānu] ekim šumēlum imittam īkim L. 4 mi-tam i-ki-im \ \a]-la-ak-ta-ka alaktaka nakrum iša¶¶i¢ lú KÚR-rum i-ša-¶i-i¢ ‘If the tip of the “Path” of the right is taken downwards (and) the left took the right: the enemy will attack your expedition by surprise.’ 5

šumma padānu ina qablītīšu ana DIŠ pa-da-nu i-na MURUB4-šu ana ša-ap-[la-nu e]-ki-im a-la-ak-ti šap[lānu e]kim alakti šarri [nakrum] L. 5 šar-ri \ ˹l˺[úKÚR-rum] i-ša-¶i-i¢ iša¶¶i¢ ‘If the “Path” in its middle is taken downwards: the enemy will attack the king’s expedition by surprise.’ §5

Reverse §6

6

DIŠ pa-da-an šu-me-lim re-su a-na šumma padān šumēlim rēssu ana ša-[ap-la-nu e]-ki-im šu-me-lum šu- ša[plānu e]kim šumēlum šumēlam L. 6 me-lam i-ki\-im [ka]-ba-ra-¶i lúKÚR-im īkim [ka]bara¶¶i nakrim ‘If the tip of the “Path” of the left is taken downwards (and) the left took the left: panic among the enemy.’

The arrangement of this compendium appears highly symmetrical (see Appendix I). It comprises two sections with three entries in each. In the first section, parts of the ‘Path’ (padānu)—namely, the tip of the right side (rēš padān imittim), the middle part (padānu ina qablītīšu) and the tip of the left side (rēš padān šumēlim)—are ‘taken’ upwards (ana elēnu ekim), while in the second section the same parts of the ‘Path’ are ‘taken’ downwards (ana šaplānu ekim). The descriptions of the middle part (padānu ina qablītīšu) as ‘taken’ (ekim) are not followed by additional information, but when the tips of the ‘Path’ are described, one finds additional notes to the effect that one of the sides (imittum or šumēlum) ‘took’ (īkim, once ītekim) the other side or itself. The side that ‘was taken’ (in the accusative) always corresponds to the side of the ‘Path’s’ tip, while the side that ‘took’ (in the nominative) corresponds to the direction the main part ‘is taken’ to (in the stative): ‘right’ for ‘downwards’ and ‘left’ for ‘upwards.’ The latter may suggest that the liver could be turned through –90° during hepatoscopy (or at least the point of view was changing correspondingly), see Appendix I.

40

Articles: Ancient Near Eastern Studies

As a technical term of divination, ekēmu ‘to take away’ is traditionally interpreted as ‘stunted,’ ‘atrophied’ in the stative (ekim) and ‘to absorb’ in the preterit and the perfect (īkim and ītekim). This interpretation can hardly be applied to imittum imittam/šumēlum šumēlam īkim. I assume that the verb ekēmu (in any form) rather designates some kind of shifting or bending of one feature towards the other (i. e. one feature ‘takes away’ the other). This suggestion is probably corroborated by CT 20, 29rev.:12 where itkumū is glossed as ana imitti u šumēli maqtū.23 Given the fact that this problem is still far from being definitively resolved, it seems wise to stick to the literal meaning ‘to take’ in the translation. § 2 L. 2 The form in-ni-na24 appears to be the 3 sg. present of enû N, which in this case represents a passive to the meaning ‘to substitute’ (cf. CAD E 176)25. No other examples of enû N in this meaning are known to the author.26 § 3 L. 3 The sequence of omens in §§ 3–5 is a unique example of ša¶ā¢u used in combination with alaktu. One wonders to what extent this combination is synonymous to ša¶ā¢u in combination with ¶arrānu.27 Note that the perfect i-te-ki-im probably introduces a caesura between the two sections of the compendium (courtesy S. Loesov). 23

For other protases which include similar constructions with ekēmu see YOS 10, 33 iv 14–16 (l YOS 10, 34rev.:40–43); YOS 10, 33 iv 17–19; YOS 10, 33 iv 24–25; YOS 10, 33 iv 26–27; YOS 10, 33 iv 28–30; YOS 10, 40:5–6; YOS 10, 40:7–8; RA 38, 80:4; KAR 429 i 17–18; KAR 422:28; CT 20, 23:5; CT 30, 42:19–20; CT 3, 4:65–67. 24 Note the contraction *inninniam > inninnâ(m). 25 Cf. ARM 1 20:5–7 (in the Gt-stem): ´a-bu-um ša iš-tu Qa-ta-nimki il-la-kam ani-i¶ ki-i i-na bi-ir-tim úš-ša-ab it-nu-ú li-te-nu-ú ‘The army that will march from Qatanum is tired—how will it stay in the fort? Let them alternate with one another.’ 26 According to CAD E 177, negated forms of enû N ‘to be revoked’ are used about words, commands, orders in the first millennium texts. This sense, however, does not apply here, as there is no negation and the subject is the army itself. 27 Cf. YOS 10, 25:8 (with a parallel in YOS 10, 26 ii 10): ni-šum KASKAL i-ša¶i-i¢ ‘a lion will suddenly attack the expeditionary force.’ Note, however, that in YOS 10, 52 iii 6 (with a parallel in YOS 10, 51 iii 6) ne-šu-um i-na KASKAL-im i-a¶-¶i-i¢ ‘a lion will suddenly attack at the road,’ where the preposition plainly suggests that ¶arrānu designates the place of the attack, not its object. If, by implication, the (virtual) accusative in YOS 10, 25:8 is interpreted as adverbial, this omen cannot be taken as a true parallel to one under scrutiny.

I. Khait, The Old Babylonian Omens in the Pushkin Museum …

41

§ 5 L. 5 Another possible reconstruction of the apodosis: a-la-ak-ti šar-ri \ ˹l˺[úKÚRam] i-ša-¶i-i¢ ‘the king’s expedition will attack the enemy by surprise.’ There are arguments in favor of both reconstructions, so it is difficult to choose which of them is better. 6 Obverse § 1, l. 1

\ ]

[šu] m -ma pi-i¢-ru 1 pi-¢a!-ir lúKÚR

1

[šu]mma pi¢ru ištēn pi¢ir nakrim

‘If there is one split: departure of the enemy.’ 2\

]

\ ]

šum -ma pi-i¢-ru 2 pu-um kinu-um ‘If there are two splits: reliable word.’ § 2, l. 2

šumma pi¢rū šinā pûm kīnum

3

šum-ma pi-i¢-ru 3 ar-bu-ut um-ma- šumma pi¢rū šalāšat arbūt ummān an lúKÚR-/im nakrim ‘If there are three splits: routing of the enemy’s army.’

§ 3, l. 3

4 šum-ma pi-i¢-ru 4 ki-ir-ri lúKÚR-im šumma pi¢rū erbet kirri nakrim \ ] te-re-di gišGIGIR lú K [ÚR / t]u-ta-ra tereddi narkabti na[krim t]utarra ‘If there are four splits: you will pursue the enemy’s military expedition and you will come back with the chariot of the enemy.’

§ 4, l. 4

šum-ma pi-i¢-ru 5 ni-bi-ir lúKÚR-ka šumma pi¢rū ¶amšat nēber nakrīka ta-da-/ak tadâk ‘If there are five splits: you will destroy the enemy’s ferry.’ § 5, l. 5

5

§ 6, l. 6

6

šum-ma pi-i¢-ru 6 7 gi-ip-šu lúKÚR-im

šumma pi¢rū šeššet sebet gipšū nakrim

‘If there are six (or) seven splits: extension of the enemy(’s army).’ 7

šum-ma pi-i¢-ru 8 9 ti-bu-ut um- šumma pi¢ru samānat tišêt tibūt \ ] ma-an ma-an -/da ummān-manda ‘If there are eight to nine splits: invasion of the ummān-manda-hordes.’ § 7, l. 7

šum-ma pi-i¢-ru 10 šu-bu-ur(sic!) šumma pi¢ru ešret šubur māt nakrim ma-at lúKÚR-im [ip]-/pa-¢ar \ ¶u-ša- [ip]pa¢¢ar ¶uša¶¶u ša šalamtum ¶u ša ša-lam-tum in-na-an-du i-n[a innandû in[a mātim] ibbašši ma-tim] / ib-ba-ši ‘If there are ten splits: the enemy’s land will get out of order (?); a famine of corpses lying unburied will occur in the land.’ § 8, l. 8

8

§ 9, l. 9

9

šum-ma pi-i¢-rum ´a-li-ip ni-ip-¶u-ú

‘It the split is slanted: nip¶u-predictions.’

šumma pi¢rum ´alip nip¶ū

42 § 10, l. 10

Articles: Ancient Near Eastern Studies 10

šum-ma pi-i¢-rum ki-ma pi-lu-ur- šumma pi¢rum kīma pillurtim tēšûm tim te-šu-um e-li \ um-ma-nim i-ma- eli ummānim imaqqut qú-ut ‘If the split has the form of a cross: panic will seize the army.’ Bottom Edge § 11, l. 11

11

šum-ma pi-i¢-rum i-na li-ib-bi pi-i¢- šumma pi¢rum ina libbi pi¢ri nip¶um ri ni-ip-¶um! (wr. TUM) ‘If the split is inside (another) split: nipḫu-prediction.’

Reverse § 12, l. 12

12

šum-ma pi-i¢-rum-ma wa-ar-ka-sú šumma pi¢rum-ma warkassu šīlum ši-lum na-di \ du-ur pa-a¢ lúKÚR-ka nadi dūr pā¢ nakrīka ta´abbat-ma ta-´a-ba-at-ma ta-na-qar-šu \ ta-sa- tanaqqar-šu tasa¶¶ar-ma teppussu ¶ar-ma te! (wr. PI)-pu-su i-na bi-it a- ina bīt awīlim mar´um iballu¢ bal¢um \ ] wi-lim mar-´um i -ba-l[u]-\u¢ \ ba- imât \ ] al-¢ù-um i-ma- at ‘If there is a split, and behind it there is a hole: you will seize the frontier fort of your enemy and destroy it, (but then) you will rebuild it; at the man’s household: the sick person will recover, the healthy person will die.’ 13

šum-ma pi-i¢-rum e-le-nu i-sí-il5-ti šumma pi¢rum elēnu isilti šumēlim šu-me-lim ŠU.SI pa-[ ¢e4]-er n/i-ip-¶u ubānam pa[ ¢]er nip¶ū ‘If there is a split on the “Finger” above the left sphincter: nipḫu-predictions.’

§ 13, l. 13

14

\

]

šum-ma ki -´i-ir-ti šu-me-lim pa- šumma ki´irti šumēlim pa¢rat ki´i[r \ ] a¢-ra-at ki-´i-[ir Š] À DINGIR a-na lib]bi ilim ana awīlim p[a¢e]r / LÚ \ p[a-¢e4-e]r ‘If the ridge to the left has a split: the god’s wrath against the man is dispelled.’ § 14, l. 14

(§ 15), l. 15

17

(§ 16), l. 18

18

šum-ma pi-i¢-rum ka-pi-i´ ‘If the split is bent over: (no apodosis) …’ \

šumma pi¢rum kapi´

]

(eras.: { šum -[ma]})

(Eras.: ‘If ’). Coloph., l. 19

19

14 mu.bi

Fourteen lines

This text has parallels in a number of other OB omen texts dealing with pi¢ru and šēpu, namely YOS 10, 44 and 50 as well as KAR 150. KAR 150, a NA copy of аn OB original, has a section dealing with the number of

I. Khait, The Old Babylonian Omens in the Pushkin Museum …

43

pi¢ru-marks (7, 13–15). The šēpu-tablet YOS 10, 44 seems to be a copy of at least two separate compendia. It has two sections of similar omens, reflecting two different traditions (YOS 10, 44:1–11 and 63–72). The first section is closely related to YOS 10, 50 (another šēpu-compendium). A full account of these parallels can be found in Appendix II. § 1 L. 1 The sequence PI-DA-IR is troublesome. The sign transcribed here as DA is very slipshod, so I cannot be confident whether it is indeed DA or rather the more expected DI or TE.28 The reading DA is, however, more likely (compare DI in l. 4, TE in l. 4 and DA in l. 5). If the reading DA is accepted, one is faced with a gebrochene Schreibung for pi¢ru, a well-known nomen actionis from pa¢āru.29 In the present context, one might think of ‘departure’ (cf. pa¢āru G and N ‘to depart’), but no parallel examples are known to me.30 The unique appearance of pi¢ru both as a technical term in the protasis and a prediction in the apodosis is similar to the well-known usage of šēpu, often attested within one omen both in the protasis (as an extispicy term) and in the apodosis with the meaning ‘approach, attack.’31 If pi¢ru is indeed to be understood as ‘departure, retreat’, its usage here may be exactly the opposite of šēpu. This observation, if correct, provides one

28

Saveliev 1985:13 reads pi-¢i-ir and translates ‘избавление от неприятеля (или: раскол у неприятеля?).’ 29 The sequence a-i (or a-e) might represent some kind of special vocalic quality, apparently conditioned by the following r. However, none of the examples collected in B. Groneberg’s well-known study on the subject displays a-i instead of the expected i, let alone that the very possibility of gebrochene Schreibungen rendering phonetic peculiarities of vocalic phonemes is doubtful (see Groneberg 1980:161ff.). 30 On the whole, pi¢ru displays a wide range of meanings, mostly corresponding to different semantic nuances of pa¢āru. Note, in particular, such meanings as ‘undoing’, ‘separation’ and ‘ransom’ (CAD P 499ff.). 31 Cf., for instance, šēp irrubakkum tadâkši ‘you will defeat the expedition that will arrive to you’ (YOS 10, 44:30), šēp Elamtim ‘attack of the Elamites’ (YOS 10, 20:22), šēp ¶uša¶¶i ‘approach of famine’ (YOS 10, 25:51), šēp nukurtim ‘arrival of enmity’ (YOS 10, 44:40), šēp bussurtim ‘arrival of news’ (YOS 10, 44:36), šēp rī´i ‘arrival of help’ (YOS 10, 50:2), šēp salīmim ‘arrival of peace’ (YOS 10, 50:10), šēp dumqim ‘arrival of good’ (YOS 10, 26 iii 53), etc. In most of these examples, šēpu can be alternatively interpreted with an abstract meaning ‘sign,’ ‘omen’ (as it happens with amūtu and kakku). For more examples and discussion see Richter 1994:240–244.

44

Articles: Ancient Near Eastern Studies

more interesting link connecting the present text with some šēpu-compendia (see Appendix II). § 2 L. 2 Cf. YOS 10, 44:63: DIŠ i-na šu-me-el ta-ka-\al]-tim AŠ pu-um ki-nu-um ki-\´i]ir ŠÀ DINGIR pa-¢e-er-\kum] ‘If on the left side of the kidney (there is) a single “Footmark”: a reliable word; the god’s anger with you is calmed down’32 (note pa¢āru in the apodosis). § 3 L. 3 Cf. YOS 10, 44:64: DIŠ 2 AŠ ar-bu-ut lúKÚR lúKÚR ti-la-tu-šu i-p[a-¢]a-ra-a-šu ‘If (on the left side of the kidney there are) two “Footmarks”: routing of the enemy; the enemy’s auxiliary troops will desert him.’ Note again pa¢āru in the apodosis. § 4 L. 4 Saveliev 1985:13 reads ke-er-ri nakrim te-ri-ba narkabtu (?) [nakrim …] and translates ‘поход (?) неприятеля ты …, колесницу (?) [неприятеля] ты захватишь?’ Cf. an almost exact parallel in YOS 10, 44:66 (the same apodosis is also found in l. 26): DIŠ 4 AŠ gi-ir-ri lúKÚR-ka te-er-te-ni-ed-di na-ar-ka-ba-at lú KÚR-ka tu-t[a-ar] ‘If (on the left side of the kidney there are) four “Footmarks”: you will pursue the enemy’s military expedition and you will come back with the chariots of the enemy.’ Note that the construct state form na-ar-ka-ba-at, usually understood as a plural (cf. CAD N2 355), can also be parsed as singular (GAG § 64h). § 5 L. 5 This is a highly peculiar entry, as no other attestations of a ferry in omen apodoses are known to me. The noun can be used about either a ford or a ferryboat,33 but what is described is, at any rate, a military operation aimed to prevent the enemy’s army from crossing the river. § 6 L. 6 This apodosis is attested twice elsewhere in related contexts, viz. YOS 10, 44:68 (DIŠ 6 AŠ gi-ip-šu lúKÚR ‘If there are six “Footmarks”: extension (or 32

The transcription of the end of this line follows a collation by A. Winitzer (Winitzer 2006:556). 33 Boats in general, however, are well attested in omens, particularly with dâku, as in CT 20, 50rev.:7: šumma ina nāri elippāt nakri šumma ina tābali ummān nakri tadâk ‘Either you will defeat the ships of (your) enemy on the river, or the army of the enemy on the dry land’ (translation from CAD E 92).

I. Khait, The Old Babylonian Omens in the Pushkin Museum …

45

groupings) of the enemy(’s army)’) and KAR 150:18 (BE-ma šu-me-el ZÍ 4 5 6 a-di ma-du-ti pu-u¢-¢u-úr gi-pí-iš um-ma-an lúKÚR ‘If the left side of the gall-bladder is seized with four, five, six or more cracks: extension of the enemy’s army’) (see also Appendix II). The word gipšu is a nomen actionis from gapāšu, meaning ‘extension,’ ‘reinforcement’ or ‘grouping.’ Cf. already Schileico’s rendering ‘Wenn der Spalten 6 oder 7 sind: Ansammlung des Feindes’34 as well as the translations by the authors of CAD (“Mass of the (army of the) enemy”)35 and by Saveliev ‘множество неприятеля’ (1985:13). Cf. gapšu ‘huge’ describing armies, as in AKA 78 vi 1 (Tigl. I): itti ummānātēšunu gapšāte … lū amda¶i´ ‘I fought with their huge armies’ (translation from CAD G 45). The writing gi-ip-šu in two cases out of three cannot be accidental and must represent a plural form, probably intended to emphasize the plurality of the grouping forces. § 8 L. 8 Saveliev 1985:13 reads and translates: šum-ma pi-i¢-ru 10 ŠU (?) … māt nakrim [ta?]-pa-¢ar ¶u-ša-¶u ša ša-lam-tum in-na-an-du ina [ma-ti (?)] ib-ba-ši ‘если трещин 10 — … страны неприятеля ты прекратишь (?) (расколешь?); голод (такой), что будут лежать брошенными трупы, в [стране (?)] случится.’ The sign UR is probably corrupt (maybe a corrected RI: there is a Winkelhacken in the middle, but two (!) horizontal wedges at the bottom). The expression šubur mātim ippa¢¢ar is unique and troublesome. A close parallel is probably attested in the OB omen compendium TIM 9, 80:34, where the apodosis reads šu-bu-\ur]-ti ma-tim na-ka-ar-tim \x x x (x)].36 The expression (literally ‘the anus of the country’) may well be synonymous37 to another rare expression occurring in two apodoses, viz. 34

Schileico 1929:217, fn. 2. CAD G 85. 36 In the light of the passage under scrutiny, šuburtu seems to be a unique \ ] feminine by-form of šuburru ‘anus.’ The emendation to šu-bu- ul! -ti in Richter 1994:221 is thus probably unwarranted. 37 Close association between the anatomical terms qinnatu ‘buttocks,’ šuburru ‘anus’ and isiltu ‘sphincter ani’ is transparent in Šumma Izbu XVII 45′–48′ (text and translation according to Leichty 1970:167–168): 45′ šumma izbu qinassu pe¶iat (var.: pe¶ât) ¶uš[a¶¶u …] If an anomaly’s buttocks are closed: famine [...] 46′ šumma izbu qinnata lā īšu šarru ina ekallīšu ikkammi […] If an anomaly has no buttocks: the king will be captured in his palace […] 47′ šumma izbu šuburrašu pa¢er […] 35

46

Articles: Ancient Near Eastern Studies

isilti mātim/ ummānim ippa¢¢ar.38 This phrase, obviously depicting diarrhea, seems to be used as a metaphor for political and/or social disorder.39 It is not clear whether we are faced with one single apodosis consisting of two subsections or rather with two different apodoses, one of which refers to the enemy’s land and the other, to the land of the diviner. The latter option is quite unlikely, since the hypothetic second interpretation is not introduced by any corresponding formula (such as šanû šumšu, ūmam rēqam, etc.). Besides, the supposed alternative apodoses are of comIf an anomaly’s anus does not contract: […] šumma izbu isiltašu pa¢rat […] If an anomaly’s sphincter ani does not contract: […] Unfortunately, the apodoses are almost completely lost; otherwise they could have given valuable hints pertaining to the interpretation of the omen under scrutiny. \ ] 38 YOS 10, 36 ii 28–30: DIŠ MUR MÚRGU ša ša-ap-la-nu-um pa- ri -is i-sí-il-ti maa-tim ip-pa-a¢-¢a-ar [bu-t]u-uq-tum ib-ba-at-[t]a-aq ‘If the lower “shoulder”of the lung(s) is divided: the country will get out of order (?); a breach of a dike will occur’; KAR 150:21: BE-ma ru-qú ša i-mi-ti ZÍ a-na i-di ZÍ pa-¢e4-er ´i-it ka-ab-tim ana gišTUKUL i-si-il-ti um-ma-ni-ia ip-pa-¢a-ar ‘If the thin part of the right side of the gall-bladder is separated towards the side of the gall-bladder: departure of an important person; for the warfare—my army will get out of order (?).’ 39 CAD L 193 translates isiltu in these apodoses as ‘control (?).’ This translation seems to be broadly correct, although one cannot exclude that this term has some more concrete meanings. In both extant examples (see previous fn.), this expression is connected with a breach of a dike and a departure of an important man, and in the text under consideration, with famine. These apodoses can be grouped together according to their general negative meaning and the motive of breaching and loosing: both are quite expected for omens with words like pi¢ru, pa¢āru and parāsu in the protasis. In this connection, cf. also two omens with isiltu pa¢rat in the protasis, which have bad apodoses connected with a military campaign: DIŠ EŠ iš-da-ša šu-me-lam ki-na-ma i-sí-el-ti i-mi-tim pa-a¢-ra-at ru-bu KASKAL i+la! (wr. I+A)- ú-ul i-tu-ra ‘If the left side of the base of the gallbladder is fine and (its) right sphincter does not contract: the prince will not return from the campaign he wends his way to’ (Scheil B 31), DIŠ EŠ iš-da-ša i-mi! (wr. LIMMU) -tam ki-na-ma ù i-sí-el-ti i-mi-ti pa-a¢-ra-at um-ma-nu KASKAL i-la-ku ú-ul itu-ra-am ‘If the right side of the gall-bladder is fine and (its) right sphincter does not contract: the army will not return from the campaign it wends its way to’ (Scheil B 32–33). Possible typological parallels to šubur mātim are abunnat mātim, zibbat mātim and kutal mātim, each of them attested in military contexts in OB divination compendia (cf., for instance, YOS 10, 33 iii 41–42 l YOS 10, 34:13–16; YOS 10, 33 iii 44–45 l YOS 10, 34:19–22 and YOS 10, 46 iii 48–49). Note that abunnat mātim seems to be exactly parallel to the famous Biblical Hebrew expression ‫ טבור הארץ‬ÑmfalÕj tÁj gÁj (see BDB 371; courtesy L. Kogan). 48′

I. Khait, The Old Babylonian Omens in the Pushkin Museum …

47

pletely opposite nature: a bad omen referring to the enemy’s land and a bad omen referring to the land of the diviner.40 The first option seems therefore more likely, but in this case one must assume that, uniquely, mātu alone (rather than the usual māt nakrim) refers to the land of the enemy. The spelling in-na-an-du is a South OB feature. Note the lack of the expected plene writing for the contracted vowel, a topic discussed at length in a recent article by M. Worthington (2010).41 This type of apodosis is extensively dealt with in Tchapygina 2004: 357–358: the relative particle ša does not display its usual meaning ‘which,’ but rather ‘so that.’ § 9 L. 9 The verb ´alāpu in the protasis occurs in two SB omen texts only. In both of these passages it is used about šēpu.42 CAD’s translation ‘slanted’ is likely,43 but since the verb is poorly attested and its meaning cannot be deduced from the context, the alternative translation ‘to cut off, to erase,’ attested in lexical lists,44 could also be considered. The nip¶u-prediction is noteworthy: it is probably connected with dishonesty and treachery which are also inherent connotations of ´alāpu and its derivates.45 § 11 L. 11 The emendation of TUM to ḪUM follows the parallel texts, where nip¶u is attested throughout as the apodosis for this type of protasis (see Appendix II).46 The reading NI-IP-TUM = neptûm ‘opening, breach’ does not seem to give any acceptable sense as a complete apodosis, notwithstand-

40

To be sure, the introductory formula separating two different apodoses can be omitted, although such cases are quite rare, cf. YOS 10, 17:37; 10, 20:7; 10, 25:49–50; 10, 34:7–11, 13–16 and 19–22. 41 According to Worthington, subordinative forms of verbs with the third weak radical are written without plene in 10.3% of the examples attested in AbB. 42 KAR 454:28: šumma ina šumēl marti ina x šēpu-ma ana panīša ´alpat ana arkīša turrat ‘If at the left of the gall bladder in the … there is a “footmark” and it is slanted towards its front, turned towards its back’ (translation from CAD Ô 86), CT 31, 11 i 8: šēpu pa´lat ‘the “footmark” is slanted’ (in broken context). 43 Cf., for instance, the derivate ´iliptu ‘diagonal’. 44 See CAD Ô 71, lexical section. 45 See CAD Ô 86 (´alpu), CAD Ô 72ff. (´aliptu). 46 Cf. Saveliev 1985:13–14, where ni-ip-tum is translated as ‘свет (блеск, вспышка, пламя?).’ This understanding presupposes the reading nip¶u.

48

Articles: Ancient Near Eastern Studies

ing the fact that this lexeme does sometimes occur in both apodoses and protases.47 § 12 L. 12 The interpretation of this omen in Saveliev 1985:13–14 is rather confusing.48 Emendation of PI to TI is unavoidable: the sequence PI-PU-SU which ought to represent a verbal form, obviously cannot make any good sense, whereas teppussu as the opposite to tanaqqaršu fits well the context. In this hendiadys construction, tasa¶¶ar does not express a repeated action in the strict sense:49 the ruler, here addressed in the second person, rebuilds the fort, but obviously does this for the first time. The meaning of the auxiliary verb is thus ‘to return something to its previous state.’50 (§§ 15–16) Ll. 15–16 These two unfinished omens underscore the fact that the text is a school-tablet, probably copied from a larger compendium. It seems like the scribe start– ed to copy these omens from the original tablet, but did not finish copying them due to lack of space or some other reason. Note the colophon which counts only 14 omens and therefore not includes these last two entries. 7 Obverse §1 Ll. 1–4

[…]\i]-na MURUB4 pa-li-iš \x] […] [šumma …] ina qablītim pališ … [BA?] ŠI 2[…]\I] ŠI ú(-)ša-ap-la-aš (or: ušaplaš (or šaplānu?) […] ezu¶ (?) nu?) 3 […] e-zu?-u¶ BU UD? \xx] […] ŠI … ummānam … [x(x)] \x] / A ÚU? NA? 4[…] um-mana-am i-\x][…] \BU] [x(x)] ‘If … is pierced in the middle below (?) … is girt (?) … will … the army.’

47

1

See CAD N2 171. Cf. his text: šum-ma pi-i¢-rum-ma wa-ar-ka-su ši-lum na-šu du-ur pa-a¢ nakri-ka ta-´a-ba-at-ma ta-na-kap(?)(kar?)-šu ta-sa-¶ar-ma PI(= tal?)-pu-su i-na bi-it a-we-lim mar-´u(?) ba(?) … ba-al-¢um i-ma-qut (или i-ba-aš-[ši]?) ‘Если трещина с оборотной стороны (в дальней части?) имеет углубление, пограничную крепость (?) неприятеля ты захватишь, разрушишь (разобьешь, окружишь, разгромишь (?)) ее (ее разгром …?), в доме человека больной? … живой (и здоровый) упадет (?) (или: будет жив и здоров?).’ 49 For this understanding see CAD S 40f. 50 A similar semantic nuance can be observed in Biblical Hebrew, where the common hendiadys construction with ‫ שוב‬can mean not only ‘to do for a second time,’ but also ‘to return to a previous state’ (Jos 2:23, Ps 71:20, 85:7, Deut 23:14, etc.), see BDB 998, mng. 8 (courtesy L. Kogan). 48

I. Khait, The Old Babylonian Omens in the Pushkin Museum …

49

[…] i-na \MURUB4 na?-si?]-[i¶ …] / [šumma …] ina qablītim [na]si¶ (?) x 6[(…)]\lú]KÚR-rum ú-ZA-ru-nim ši- [(…)] nakrum … šē[p (?) …] [ip (?) …] 7[…] i-ir-ru-ba-am irrubam ‘If … is pierced in the middle … the enemy will … an attack of … will approach us.’ §2 Ll. 5–7

Reverse § 3 L. 8

5

8

[(…)] (eras.: xxxxxx)

Since the text is badly damaged, my readings of the remaining fragments are very uncertain. Due to possible structural similarity to No. 5, the feature under examination is probably padānu.

50

Articles: Ancient Near Eastern Studies

Appendix I (To text No. 5)

4

šumma erbe šēpētum šēp elamtim

3

šumma šittā šēpētum nakrum tappâm i[rašši (?)] 2 šumma šalāš šēpētum šēp rī´[i]

1

(missing)

5 Five features

šumma ¶amiš šēpētum šarrum dūrānīšu inaqqar 5 šumma šediš šēpētum 6 āl dūrīka ¶arbūtam Six features [illa]k 6 šumma sebe šēpētum 7 umm[ā]nka ana pani Seven features umm[ā]n nakrīka kakkīša itabbak

4 Four features

3 Three features

2 Two features

1 One feature

A YOS 10, 50

šumma sebe šēpētum rabûm kabtum ana mātīka i¢e¶¶iakkum

69

šumma šediš šēpētum gipšū nakrim ...

68

6

šumma šediš [šēp]ētum āl dūrim ¶arbūtam [illak] 7 šumma sebe šēpētim […]

šumma ¶amiš šēpētum nip¶um

67

šumma erbe šēpētum girri nakrīka terteneddi narkabāt nakrīka tutâr

66

šumma šalāš šēpētum rākib imēri i¢e¶¶iakkum

65

C YOS 10, 44 (II) 63 šumma ina šumēl tākaltim šēpum pûm kīnum ki´ir libbi ilim pa¢erkum 64 šumma šittā šēpētum arbūt nakrim nakrum tillātūšu ipa¢¢arāšu

šumma ¶amiš [šēp]ētum tibût šub[arîm (?)]

5

šumma šalāš šēpētum nark[um] ... rē´u (?) u till[atu (?) ...] 4 šumma erbe [šēp]ētum šēp elamtim

3

šumma šittā šēpētum š[ēp] (?) tabs[irtim (?)]

2

B YOS 10, 44 (I) 1 šumma ina šumēl tākaltim šēpum ...

Appendix II (To text No. 6)

(missing)

E KAR 150

(missing)

(missing)

(missing)

13 šumma šumēl šumma pi¢rū šinā pûm martim adi šinā pa¢er kīnum pûm kīnum 14 šumma šumēl 3 šumma pi¢rū šalāšat martim adi šalāš pa¢er arbūt ummān nakrim nip¶u 4 šumma pi¢rū erbet kirri nakrim tereddi narkabti na[krim] 15 šumma šumēl tutarra martim erbe ¶amiš 5 šumma pi¢rū ¶amšat šediš adi mādūti nēber nakrīka tadâk pu¢¢ur gipiš ummān nakrim 2

D Text No. 6 1 [šu]mma pi¢ru ištēn pi¢ir nakrim

šumma šēpum ina libbi šēpim nip¶i nakrim

13

šumma ešer šēpētum šēp s[al]īmim

10

C YOS 10, 44 (II) 70 šumma samāne šēpētum awīlum awīlūtam illak-ma mārmārīšu immar 71 šumma tiše šēpētum salīmu ina mātim iššakkan mātum aburrī ušša[b] 8

šumma pi¢ru samānat tišêt tibût ummānmanda

7

D Text No. 6

Note the parallels that go with shift (marked with arrows): 1C=2D=2E 2B=3C 2 A = 3 B? 9 C = 10 A = 10 B

šumma ina šumēl martim pi¢rum ina libbi pi¢ri ālum ša ana lawîšu tallaku innaddi-ma itta´´i nip¶u pûm lā kīnum

7

(missing)

(missing)

E KAR 150

6 C = 4–6 E 9 C = 10 A = 10 B 11 A = 11 B = 11 C = 11 D = 11 E

šumma pi¢ru ešret šubur māt nakrim (missing) [ip]pa¢¢ar ¶uša¶¶u ša šalamtum innandû in[a] [mātim] ibbašši 11 šumma šēpum ina 72šumma šēpum ina libbi 11šumma pi¢rum ina libbi šēpim nip¶ū šēpim nip¶um ummānka libbi pi¢rim nip¶um! nakrim ana ālim rēqim irrub

šumma ešer šēpētum mātum aburrī ušša[b]

10

B YOS 10, 44 (I) 8 šumma samāne šēpētum u[mmān d]ūrīka (?) ana pani nakrim kakkīša [itabbak] 9 šumma tiše šēpētum amūt Šarkašša[rrī]

The obvious parallels between the texts are the following: 1C=2D=2E 4A=4B 2 A = 3 B? 4C=4D 2C=3D 6A=6B

11 One feature inside another

10 Ten features

9 Nine features

8 Eight features

A YOS 10, 50 7 šumma samāne šēpētum ummān dūrīka ana pani ummān nakrīka kakkīša itabbak 8 šumma tiše šēpētum attā u nakirka 9 ta´´abbatā-ma a¶um a¶[a]m ušamqat

I. Khait, The Old Babylonian Omens in the Pushkin Museum …

1

Obv.

3

Rev. 1. I2б 1724

53

54

Articles: Ancient Near Eastern Studies

1

5

2. I2б 1723

1

3. I2б 1325

1

4. I2б 1664

I. Khait, The Old Babylonian Omens in the Pushkin Museum …

5. I2б 1663 + 1661

55

56

Articles: Ancient Near Eastern Studies

1

5

Obv.

10

Bottom Edge

12

15

Rev. 6. I2б 1660

I. Khait, The Old Babylonian Omens in the Pushkin Museum …

1

5

Obv.

8

Rev. 7. I2б 1662

57

58

Articles: Ancient Near Eastern Studies

References Alster–Oshima 2007 Cavigneaux 2005

Dercksen 2001 Dercksen 2005 Glassner 2009 Goetze 1945

Goetze 1947 Goetze 1958 Goodnick Westenholz 1997 Grayson 2000 Groneberg 1980 Jacobsen 1943 Khait 2011

Lambert 1989–1990 Leichty 1970 Perlov 2000

Richter 1994

Alster, B.; Oshima, T. Sargonic Dinner at Kanesh: The Old Assyrian Sargon Legend. Iraq 69:1–20. Cavigneaux, A. Les soirées sargoniques des marchands assyriens. Kolde, A.; Lukinovich, A.; Rey, A.-L. (eds.). Korufa…J ¢ndr…. Mélanges offerts à André Hurst (Recherches et Rencontres, Publications de la Faculté des Lettres de Genève 22). Genève. Pp. 595–602. Dercksen, J. G. The King that Walked in Darkness. NABU 2001/100. Dercksen, J. G. Adad is King. The Sargon Text from Kültepe. JEOL 39:107–129. Glassner, J.-J. Écrire des livres à l’époque paléo-babylonienne: le traité d’extispicine. ZA 99:1–81. Goetze, A. The Akkadian Dialects of the Old-Babylonian Mathematical Texts. Neugebauer, O.; Sachs, A. J. (eds.). Mathematical Cuneiform Texts (AOS 29). New Haven. Pp. 146–151. Goetze, A. Historical Allusions in Old Babylonian Omen Texts. JCS 1:253–265. Goetze, A. The Sibilants of Old Babylonian. RA 52:137– 149. Goodnick Westenholz, J. Legends of the Kings of Akkade. Winona Lake. Grayson, A. K. Assyrian and Babylonian Chronicles. Winona Lake. Groneberg, B. Zu den “gebrochenen Schreibungen.” JCS 32:151–167. Jacobsen, T. Primitive Democracy in Ancient Mesopotamia. JNES 2:195–172. Khait, I. A Newly Discovered Distant Join to YOS 10 30 in the Pushkin Museum of Fine Arts, Moscow. JCS 63: 69–72. Lambert, W. G. Lugal-girra and Meslamta-ea. RlA 7. Pp. 143–145. Leichty, E. The Omen Series Šumma Izbu (TCS 4). Locust Valley. Perlov, B. I. The Cuneiform Collection of the Pushkin Museum of Fine Arts (Klinopisnaya kollekciya Gosudarstvennogo muzeya izobrazitelnykh iskusstv im. A. S. Pushkina). VDI 2000:236–247. Richter, T. Untersuchungen zum Opferschauwesen II. Zu einigen speziellen Keulen-Markierungen. AoF 21: 212–246.

I. Khait, The Old Babylonian Omens in the Pushkin Museum … Saveliev 1985

Schileico 1929 Tchapygina 2004

Winitzer 2006 Worthington 2010

59

Saveliev, Yu. A. Babylonian Omen Texts in the Pushkin Museum of Fine Arts (Vavilonskie teksty prednamenovaniy (omina) v sobranii GMII im. A. S. Pushkina). Yureneva, Yu. P. (ed.). Tezisy dokladov nauchnykh sessiy, posvyaschennykh itogam raboty Gosudarstvennogo muzeya izobrazitel’nykh iskusstv im. A. S. Pushkina za 1983 god. Moscow. Pp. 8–18. Schileico, W. G. Ein Omentext Sargons von Akkad und sein Nachklang bei römischen Dichtern. AfO 5:214–218. Tchapygina, V. A. Observations on the Use of the Subjunctive in the Inscribed Liver Models from Mari. B&B 1:355–360. Winitzer, A. The Generative Paradigm in Old Babylonian Divination. PhD. Diss. Harvard University. Worthington, M. i-ba-aš-šu-ú vs. i-ba-aš-šu from Old to Neo-Babylonian. B&B 4/II:661–708.

SF 50 // 52: Zwei bislang unerkannte Duplikate aus Fāra Manfred Krebernik Unversität Jena

Aage Westenholz zum Geburtstag gewidmet SF 50 und SF 52 gehören zu einer Gruppe kleinformatiger, nahezu runder Tafeln, die kurze literarische und lexikalische Texte enthalten.1 Obwohl diese beiden “Schultexte” literarischen Inhalts bereits 1923 von Pater A. Deimel als SF 50 und SF 52 publiziert wurden, scheint bislang unbemerkt geblieben zu sein, daß SF 50 den mittleren Teil von SF 52 dupliziert2; dies lag vielleicht daran, daß in Deimels Publikation die Vorder- und Rückseite von SF 52 vertauscht sind. SF 52 mißt 7,8 × 7,9 × 2,4 cm. Die Tafel ist vollständig beschrieben und enthält auf beiden Seiten je 4 Kolumnen mit 5 bis 7 Fächern (im folgenden “Zeilen” genannt), die durch einfache oder doppelte, mit Fischgrätenmuster gefüllte Linien voneinander abgegrenzt sind. SF 50 mißt 6,2 × 6,2 × 1,7 cm. Die Tafel enthält auf der Vorderseite 3 vollständig beschriebene Kolumnen mit je 5 Zeilen, auf der Rückseite ist nur der obere Teil der ersten Kolumne beschrieben. Die Zeilen sind stets durch einfache Linien voneinander abgegrenzt. Wenn auch der Inhalt der Tafeln nur annäherungsweise bestimmt werden kann, so sind sie doch aus schriftgeschichtlichen Gründen von Interesse: Während auf der Tafel SF 52 Doppellinien den Text in einzel1

Vergleichbare Tafeln literarischen Inhalts sind: SF 30 mit Duplikat SF 50*1; SF 311; SF 45; SF 49*; SF 54 // TSŠ 170 und SF 71 (Beschwörungen)1; SF 65 (Exzerpt aus der ansonsten in Fāra nur durch die großen, rechteckigen Tafeln SF 26 + TSŠ 124 // SF 27 + TSŠ 327 + NTSŠ 294 repräsentierten frühdynastischen Sprichwortsammlung)1; SF 78. Andere Gattungen1 enthalten: SF 2 (lexikalisch: theophore Personenamen)1; SF 44 (Duplikat zu SF 28 // 29, Typ “Names and Professions List”)1; SF 48 (ED Lú D)1; SF 70 (ED Lú B)1; SF 76 und TSŠ 431 (Exzerpte aus der Liste ED Lú A1, die in Fāra ansonsten nur durch die rechteckigen größeren Tafeln SF 33 // SF 34 // SF 35 // SF 75 sowie das Istanbuler Fragment Š 329-11 vertreten ist); SF 79 (lexikalisch?). 2 Beispielsweise erscheinen beide noch isoliert bei Krebernik (1998:341) im Katalog der Fāra-Texte.

62

Articles: Ancient Near Eastern Studies

ne Abschnitte gliedern, weist SF 50 an den entsprechenden Stellen Interpunktionszeichen in Gestalt halbkreisförmiger Griffeleindrücke auf, die jeweils am Ende eines Faches stehen; sie unterscheiden sich von dem ansonsten formgleichen Zahlzeichen für “1” (DIŠ) dadurch, daß sie schräg gestellt sind. In der nachfolgenden Umschrift werden sie mit À wiedergegeben. Was die Textgattung betrifft, so lassen die kurzen, auf beiden Tafeln in derselben Reihenfolge erscheinenden Abschnitte an eine Folge von Redensarten oder Sprichwörtern denken, die zu Übungszwecken niedergeschrieben wurden. Wie die meisten literarischen Fāra-Texte, so setzt auch dieser einer kohärenten und detaillierten Interpretation große Schwierigkeiten entgegen; sie sind vor allem durch die noch freie, d. h. nicht mit der Aussprache kongruierende, Zeichenanordnung innerhalb der einzelnen Fächer sowie durch die erst rudimentär ausgebildete Notation grammatischer Morpheme bedingt. Grundlage für weitergehende Interpretationsversuche ist die strukturelle Analyse, insbesondere die Beobachtung möglicher Parallelismen. Diese werden in der folgenden Umschrift soweit wie möglich visualisiert: Lemmata und Ausdrücke, die in verschiedenen Abschnitten wiederkehren, sind fett gesetzt, parallele Ausdrücke innerhalb der Abschnitte übereinander angeordnet. Hierbei wird eine Kette graphischer und semantischer Assoziationen sichtbar, die der Komposition des Textes offenbar zugrundelagen. SF 52 (VAT 12758) 0(1) Vs. 01.01 IM+ sikil 01.02 {LÚ?} ra T V še ezen 01.03 01.04 ma-bir5-bir5 (oder nam-ma-bir5) ================ (2) 01.05 išib NÍG GÍD 01.06 šu ×ál ================ (3) 01.07 KU NI KI NI {RU} 02.01 AN-DU.DU 02.02 RU šu ×uruš ×ál 02.03 esirx(KAD5) DU ×uruš ×ál 02.04 šu mu-zu 02.05 ×ìri mu-zu 02.06 unug TUŠ DÙ é 02.07 {x} AN-DU.DU 03.01 nagar {x} nam-du11 ================

M. Krebernik, SF 50 // 52: Zwei … Duplikate aus Fāra (4)

03.02

(5)

03.03 03.04

(6)

03.05 03.06 04.01

(7)

04.02

(8)

04.03 04.04

(9)

04.05 Rs. 05.01

(10)

05.02 05.03

(11)

05.04 05.05 05.06

(12)

06.01 06.02 06.03

(13)

06.04 06.05 06.06 07.01 07.02 07.03 07.04f. 07.06 08.01f. 08.03 08.04 08.06

63

============ tir LAGAB zá (VAT 12747) ================ SF 50 UMBISAG šà-lá Vs. 01.01 UMBISAG šà-lá TAR 01.02 UMBISAG nam [?] À UMBISAG nam TAR ================ LAÚ5.LAÚ5 01.03 LAÚ5.LAÚ5 01.04 SI lú AL SI lu RU AL GAR 01.05 tur mu-du11 À tur mu-du11 ================ a-ba4-lá ir ×ar 02.01 a-ba4-lá ir ×ar À ================ šu-ku6 KA LU + SI DU 02.02 šu-ku6 KA LU 02.03 SI DU {x} À {…} ================ ad-DU LU IGI MU 02.04 ad-DU LU IGI DU.DU À DU.DU ================ ad-DU šu m[u]-du11 02.05 ad-DU šu mu-du11 03.01 gi-gíd ti À gi-gíd mu-ti ================ kur-kur 03.02 kur-kur mu-NÍGIN 03.03 mu-NÍGIN T V 03.04 AMA LU na-nam {MU} À AMA LU na -nam ================ ad-DU 03.05 ad-DU Rs. 04.01 gu4 igi gu4 igi sa6-sa6 04.02 sa6-sa6 04.03 še gu7 IB APIN À še gu7 IB APIN ================ a ŠILIG.ŠILIG da LÚ × “AŠ” {x}

UMBISAG

T

UR PÚ? NÍG

V

SUG

AN GURUŠ LU

úr BAR LU.LU APIN GÚR.GÚR tab-tab ga-DU ; ga-DU ga-ab-zu UR ÚÚB ; šu ti UR ÚÚB šu ti UR ÚÚB du11-ga-ni šu-ni UR PÚ? nu-zu

Kommentar (1) In Fāra werden zwei Zeichenformen differenziert: die hier vorliegende und auch später übliche mit zwei den waagrechten Keil durch-

64

Articles: Ancient Near Eastern Studies

kreuzenden senkrechten Keilen (IM+) und die einfache Form ohne diese (IM–).3 Ausgehend vom Begriff ‘Getreide’ (še) könnte man vermuten, daß ezen und bir5 hier normalorthographischem išin ‘Halm’ bzw. bir ‘(zer)streuen’ entsprechen. Für IM+ liegt in diesem Zusammenhang die Bedeutung ‘Wind’ näher als die Bedeutung ‘Ton’. Aus diesen Prämissen ergibt sich die tentative Übersetzung: ‘Ein reiner Wind, der “schägt”, zerstreut die Halme des Getreides’. (2) Die Zeichengruppe ME BU GAR läßt im Prinzip mehrere Unterteilungen und Lesungen zu, der folgende Ausdruck šu ×ál impliziert jedoch mit einer gewissen Wahrscheinlichkeit eine Personenbezeichnung, die man in ME = išib finden kann. NÍG.GÍD wird in den lexikalischen Texten aus Ebla mit ma-ša-lu-um, ma-za-lu, zi-lum wiedergegeben, womit ein Gefäß gemeint sein dürfte.4 Daraus ergäbe sich die Übersetzung: ‘Der išib-Priester hält ein …-Gefäß in der Hand’. (3) Dieser Abschnitt knüpft offenbar durch den Ausdruck šu ×ál an den vorigen an. Deutliche Parallelismen zeigen die Zeilen RU

šu

esirx(KAD5)

DU

šu ×ìri

×uruš ×uruš

×ál ×ál

mu-zu mu-zu

Im ersten Zeilenpaar entsprechen sich die Lemmata RU und esirx(KAD5) sowie šu und DU. ×uruš ist beiden Zeilen gemeinsam; das betreffende Zeichen ist frühdynastisch noch von KALAG verschieden.5 Das Zeichen KAD5 (= LAK 173), das sich zu GI4 × GI4 = GILIM und KAD4–5.KAD4–5 entwickelt, repräsentiert das später syllabisch als e-sír notierte Wort für ‘Sandale’.6 Die Assoziation von DU mit esirx ‘Sandale’ sowie der Parallelismus mit šu ‘Hand’ lassen vermuten, daß das Zeichen DU, das ja einen Fuß darstellt, hier logographisch in diesem Sinne verwendet ist. Hinweise auf diese Funktion finden sich auch in Ebla.7 DU repräsentierte wohl ein anderes Wort für ‘Fuß’ als das mit dem Zeichen ANŠE = ×ìri geschriebene, welches im folgenden Zeilenpaar im Parallelismus mit šu ‘Hand’ erscheint; man könnte etwa an die Lesung DU = ša4 denken, die lautlich 3

Vgl. Krebernik 1984:27f., wo diese Transliterationsweise eingeführt ist. VE 83; vgl. Conti 1990:75f. mit weiterer Lit. 5 Vgl. Krebernik 1998:277 mit Lit. 6 Vgl. Steinkeller 1981–1982. 7 Die klarsten Beispiele sind VE 14f.: NÍG.LÁ.DU – NÍG-LÁ.SAG; VE 363: GIŠ.DU.DU = du-bù DU // 2-DU; VE 997: DU-2. 4

M. Krebernik, SF 50 // 52: Zwei … Duplikate aus Fāra

65

gut mit šu korrespondieren würde (Alliteration, gleiche Silbenstruktur). muß in diesem Kontext einen Gegenstand bezeichnen, der in der Hand gehalten und ‘geworfen’ (RU = šub) wird, also eine Waffe oder ein Sportgerät. In Betracht kommende Lesungen sind illar = pukku ‘Kugel, Ball’, tilpānu ‘Bogen’ aber wohl auch eine Art ‘Wurfholz’ sowie ×ešpu = kippatu ‘(Arm-)Reif ’. Der Passus könnte somit bedeuten: ‘Das Wurfholz, das sich in der Hand des Jünglings befindet, die Sandale, die sich am Fuß des Jünglings befindet – (es) kennt (seine) Hand, (sie) kennt (seinen) Fuß’. Die restlichen Teile des Abschnittes sind weniger durchsichtig. In den ersten beiden Zeilen könnte man theoretisch ki ‘Erde’ und an ‘Himmel’ bzw. NI.NI und DU.DU parallelisieren: RU

KU

ki an

NI.NI DU.DU

Näher liegt es jedoch, die sich am Ende des Abschnitts wiederholende Zeichengruppe AN DU.DU als finite Verbalform mit Präfix AN- (an- oder am6-) aufzufassen. Die nominalen Bestandteile des zweiten Satzes, deren Reihenfolge und Syntax unsicher sind, weisen in die Sphäre von ‘Haus(bau)’ und ‘wohnen’: unug ‘Wohnstätte’, é ‘Haus’, TUŠ ‘sitzen, wohnen’, DÙ = gag ‘Pflock’ oder dù ‘bauen’ (letzteres könnte mit é im Sinne von späterem é-dù-a ‘bebautes Grundstück’ verbunden werden). Damit korrespondiert möglicherweise ki ‘Ort’ im ersten Satz. Das Zeichen läßt sich aber nur nach Emendation mit dem vorangehenden zu ki-tuš! ‘Wohnsitz’ verbinden, denn die hier vorliegende Zeichenform wird in Fāra konsequent von der in 02.07 vorliegenden unterschieden: erstere steht für KU/DAB5, letztere für TUŠ/DÚR. Für substantivisches KU kommt die Lesung tukul ‘Keule’, ‘Waffe’ in Betracht. Die beiden NI lassen sich als Possessivsuffixe mit tukul bzw. ki verbinden (tukul-ni ki-ni AN-DU.DU ‘seine Waffen bringt/brachte er an seinen Ort’) oder – weniger wahrscheinlich – als redupliziertes Adjektiv (etwa mu5 = banû ‘schön’)8 mit ki-tuš!. Der Schluß des Abschnitts scheint mit der Berufsbezeichnung nagar ‘Zimmermann’ inhaltlich an den vorigen Satz anzuknüpfen. Die anderen beiden Zeichen lassen sich als finite Verbalform nam-du11 ‘er sprach’ lesen. Der Inhalt der Rede des ‘Zimmermanns’ wäre dann wohl der gesamte erste Teil des Abschnittes. Weniger wahrscheinlich ist, daß hier das in Ebla bezeugte Lemma KA.NAGAR vorliegt,9 da das verbleibende NAM dann unerklärt bliebe. 8 9

Vgl. den in Ebla üblichen Lautwert NI = bu16. VE 194 (mit unvollständig erhaltener Glosse).

66

Articles: Ancient Near Eastern Studies

(4–5) Das allen drei Zeilen gemeinsame Zeichen ŠID drückt wohl das Thema der beiden Abschnitte aus. Wahrscheinlich ist wieder von einer Person die Rede, wofür die Lesungen umbisag ‘Schreiber’ und sa××a ‘Tempelverwalter’ in Frage kommen. In ersterem Sinne wird das Zeichen regelmäßig in den Kolophonen literarischer und lexikalischer FāraTexte verwendet. Das durch tir LAGAB zà ausgedrückte, vermutlich nominale Prädikat von Abschnitt (4) ist mir unklar. Die Differenzen zwischen beiden Textzeugen in Abschnitt (5) erklären sich entweder dadurch, daß in SF 50 das zu nam gehörige TAR aus Platzgründen in die vorangehende Zeile plaziert wurde, oder aber dadurch, daß TAR in SF 52, 03.03 versehentlich ausgelassen wurde und in der kleinen Lücke von SF 50, 01.02 zu ergänzen ist: also entweder 03.03 03.04

UMBISAG

03.03 03.04

UMBISAG

UMBISAG

šà-lá nam TAR

01.01 01.02

UMBISAG

šà-lá nam TAR

01.01 01.02

UMBISAG

UMBISAG

šà-lá TAR nam- À

oder UMBISAG

UMBISAG

šà-lá TAR nam [TAR] À

Nimmt man die zweite Möglichkeit an, die ein auf dem gemeinsamen Element ku5 basierendes Wortspiel impliziert, so könnte der Abschnitt etwa bedeuten: ‘Ein Schreiber/Tempelverwalter, dessen Barmherzigkeit (šà-lá) abgeschnitten ist (ku5), ist ein verfluchter (nam-ku5) Schreiber/ Tempelverwalter’. (6) Das Thema des Abschnitts dürfte in LAÚ5.LAÚ5 zu sehen sein, das man als ‘Plünderer’/‘Erbeutete’ interpretieren könnte. Der Abschnitt endet mit einer finiten Form des zusammengesetzten Verbums tur – du11. Es kommt später als Antonym von ma¶ – du11 ‘Bedeutendes sprechen’ = ‘bedeutend, ehrwürdig sein’ vor.10 Eher ist jedoch hier die in Ebla bezeugte lexikalische Gleichung ní×-tur-du11-ga = pa¢ārum ‘lösen’ heranzuziehen.11 Der mittlere Teil ist im Textzeugen SF 50 offenbar unvollständig notiert. Interessant sind die orthographischen Varianten lu (SF 52) und lú (SF 50). SI-lu/lú(-RU) erinnert an si-lu-ur4, ein in den lexikalischen Texten aus Ebla belegtes Logogramm.12 SI könnte aber auch, wie oft in Fāra, für MÁ ‘Schiff ’ stehen. Falls AL und GAR zusammengehören, wäre al-×ar wohl eher eine Verbalform (‘ist/hat gesetzt’) als ein Musikin-

10

Vgl. Attinger 1993:732f. VE 95; vgl. Krebernik 1984:41 und Conti 1990:77. 12 VE 1121 (ohne Gleichung). 11

M. Krebernik, SF 50 // 52: Zwei … Duplikate aus Fāra

67

strument (al-×ar = alû). Denkbar wäre aber auch, daß AL mit Lesung máḫ in Opposition zu tur steht. (7) Der kurze Abschnitt scheint das Wort a-ba4-lá = naruqqu ‘Ledersack’ zu enthalten, das in Ebla in der Form A.LÁ.BA4 = a-ru12-gu bezeugt ist.13 ir ist hier wohl nicht als ‘Schweiß’ (zūtu) o. ä. zu verstehen,14 sondern als ‘Haltering, Haken’ (erru). Der Abschnitt würde somit bedeuten: ‘Ein an den Haken gehängter Ledersack (ist er/sie?)’. Es handelt sich wahrscheinlich um eine Metapher, wie ja auch in “Ištars Höllenfahrt” der in der Unterwelt befindliche Leichnam der Göttin als ‘Schlauch’ (¶alziqqu) bezeichnet wird. (8) Der Abschnitt scheint von einem ‘Fischer’ (šu-ku6) zu handeln. In SF 52 folgt ein Zeichen, das wie RA aussieht. Nach Ausweis von SF 50 handelt es sich dabei um die in Ligatur geschriebenenen Zeichen LU und SI, die dort auf zwei Zeilen verteilt sind. Es kann allerdings auch nicht ganz ausgeschlossen werden, daß die Aufspaltung in SF 50 fehlerhaft ist. Die Zeichen KA und LU dürfen jedenfalls nicht mit dem in Ebla bezeugten Kompositum KA.DIB15 identifiziert werden, da LU und DIB in Fāra wie auch in Ebla noch unterschiedliche Formen aufweisen. Möglicherweise ist hier ka udu ‘Maul eines Schafes’ gemeint. (9) Für die Zeichenkombination ad-DU, die hier und in den Abschnitten 10 und 12 offenbar das Thema bezeichnet, ist aus jüngeren Quellen die Lesung ad-ša4 = nasāsu bekannt, ein Schallverbum der Bedeutung ‘jammern, klagen’. Man würde allerdings wie in den vorangehenden Abschnitten als Thema eine Personen- bzw. Berufsbezeichnung erwarten, also etwa ‘Klagesänger’ oder ‘Klageweib’. In dieser Bedeutung ist ad-ša4 jedoch nicht üblich. Somit stellt sich die Frage, ob wir es mit einer unorthographische Schreibung ad-gub für den gewöhnlich AD.KID geschriebenen ‘Rohrflechter’ zu tun haben. Überzeugende kontextuelle Hinweise auf die Bedeutung sehe ich nicht, gi-gíd in Abschnitt (10) reicht nicht aus, um ad-DU als ‘Rohrflechter’ zu identifizieren. Die Zeichen LU und IGI könnte man zu lu-lim ‘Hirsch’ verbinden, was aber durch den Kontext nicht gestützt wird. Eher ist IGI als igi(-šè) ‘vor(an)’ mit folgendem DU.DU ‘gehen (lassen)’ zu kombinieren. Für LU bliebe dann wohl nur die Lesung udu ‘Schaf ’ übrig. Das in SF 50 fehlende MU ist entweder über die Zeilengrenze hinweg als Verbalpräfix mit DU.DU zu verbinden, oder es stellt das Possessivsuffix der 1. Person (-×u10) dar 13

VE 601; vgl. Conti 1990:167 mit älterer Lit. VE 1041: ir = šu-du-um /¯u"tum/; vgl. Krebernik 1983:37. 15 VE 226; vgl. Conti 1990:101. 14

68

Articles: Ancient Near Eastern Studies

und gehört dann am ehesten zu igi. Eine mögliche Interpretation des Abschnitts wäre somit: ‘Dem ad-DU gehen Schafe voran’ oder ‘Jammer! Schafe gehen (mir) voran!’. (10) Der Abschnitt besteht aus zwei Sätzen. Der erste enthält die finite Verbalform šu mu-du11, etwa ‘er/sie betätigte die Hand’, d. h. ‘streckte die Hand aus’ o. ä.16 Im zweiten Satz ist danach wohl šu als Bestandteil des zusammengestzten Verbum šu – ti ‘nehmen’ zu ergänzen: ‘Der ad-DU streckte die Hand aus und nahm ein gi-gíd’ oder ‘Jammer! Er/sie streckte die Hand aus und nahm ein gi-gíd’. Der Terminus gi-gíd ‘langes Rohr’ ist mehrdeutig, es kann wörtlich gemeint sein oder im Sinne von ‘Flöte’. (11) Der Abschnitt enthält vielleicht ein Rätsel: ‘Es umkreist die (Berg-) Länder. Das ist fürwahr die Mutter des Schafes’. Auffällig ist, daß der Abschnitt einen durch ad-DU in den Abschnitten (9), (10) und (12) konstituierten Sinnzusammenhang unterbricht. (12) Der Abschnitt nimmt das Thema ad-DU wieder auf. Eine finite Verbalform ist nur erkenntlich, wenn man IB als Verbalpräfix (ib-) interpretiert; eine solche Schreibung für (später) übliches íb- wäre allerdings ungewöhnlich. Mit gu4 igi sa6-sa6 ‘Rind mit schönen/freundlichen Augen’ ist wohl še gu7 ‘Getreide fressen (lassen)’ zu verbinden. Mit dem ‘Getreide’ wiederum lassen sich IB (falls uraš ‘Erde’ zu lesen) und APIN (apin ‘Pflug’, engar ‘Pflüger’, uru4 ‘pflügen’) assoziieren. Eine eindeutige, semantisch und syntaktisch befriedigende Gesamtinterpetation drängt sich allerdings nicht auf. Möglich wäre etwa: ‘Der ad-DU gibt dem freundlich blickenden Rind das Getreide des Pflügers zu fressen (še engar ib-gu7)’ oder ‘Jammer! Der Pfüger (engar) gibt dem freundlich blickenden Rind Getreide zu fressen (še ib-gu7)’. (13) Dieser Abschnitt ist länger als alle anderen. Eine Verbindung zum vorangehenden ist kaum zu erkennen; das gemeinsame Zeichen APIN ist wohl zu weit vom Beginn entfernt, um als Brücke zu dienen. Auffällig sind reduplizierte Zeichen und Wiederholungen von Syntagmen bzw. Sätzen im zweiten Teil. Ein zusammenhängender Sinn erschließt sich nicht, vielleicht handelt sich um eine lexikalische Übung. a ŠILIG.ŠILIG: Daß a hier eher ‘Wasser’ als ‘Vater’ bedeutet, wird durch PÚ? ‘Brunnen’ (?) und SUG ‘Teich, Sumpf ’ in der übernächsten Zeile nahegelegt. Für ŠILIG kommen die Bedeutungen ‘aufhören, versiegen’ und ‘gewaltig’ (šagapūru) in Betracht.

16

Zu šu – du11 und seinem weiten Bedeutungsfeld vgl. Attinger 1993:698f.

M. Krebernik, SF 50 // 52: Zwei … Duplikate aus Fāra

69

da LÚ × “AŠ”: Das Zeichen LÚ × “AŠ” steht in UD.GAL.NUN-Orthographie für die Postposition -šè.17 Es ist wohl eine verkürzte Form von LÚ × “KÁR”, das u. a. die Lesung še29 besitzt. LÚ × “KÁR” ist wohl eine Zusammensetzung aus LÚ ‘Mensch’ und EŠE ‘Strick’ (“KÁR” = EŠEtenû). In Ebla ist es durch tabālum ‘wegführen’ wiedergegeben, später sind die Lesungen bzw. Bedeutungen še29 = šagšu, kamû ‘überwältigt, gefangen’, ša×a = ¶ablu ‘be-, geraubt’, ¶eš5 = ¶ablu, kalû ‘geraubt, gefangen’ und er(e)13 = ¶anāqu ‘erwürgen’ bezeugt.18 Möglicherweise ist das Zeichen an unserer Stelle jedoch eine Variante von LÚ × ÚŠ = ADDA, AD6 ‘Leichnam’, wozu da als phonetisches Komplement fungieren könnte. T V UR PÚ? NÍG SUG: Die Zeile würde mit PÚ ‘Brunnen’ und SUG ‘Teich, Sumpf ’ zwei bedeutungsverwandte, wohl miteinander korrespondierende Termini enthalten, falls statt PÚ nicht GIGIR! zu lesen ist. Die Bedeutung bzw. Lesung des in der Folge mehrmals wiederkehrenden Zeichens UR ist unklar, am ehesten kommt wohl ur ‘Hund’ in Frage. Sollte NÍG damit korrespondieren, könnte dieses für nig ‘Hündin’ stehen. AN GURUŠ LU: AN und GURUŠ können theoretisch dLAMMA gelesen werden, doch welche Funktion hätte daneben das Zeichen LU? úr bar: Beide Zeichen können Körper- oder Pflanzenteile bezeichnen, nämlich ‘Schoß, Wurzel’ bzw. ‘Äußeres, Haut, Rinde’. lu-lu APIN GÚR.GÚR tab-tab: Die Zeile enthält drei reduplizierte Zeichen bzw. Formen, deren Zuordnung, Reihenfolge und Bedeutung unklar sind. ga-DU ga-DU ga-ab-zu: Hier wird anscheinend mit kohortativen Verbalformen gespielt, wobei sich hinter DU verschiedene Verbalbasen verbergen dürften, also etwa ga-×en ‘ich will gehen’, ga-túm/de6 ‘ich will bringen’ oder ga-gub ‘ich will stehen’. Höchst interessant wäre, wenn so zu lesen, die Schreibung ga-ab-zu ‘ich will es wissen’: sie würde nämlich mit ga-ab für /gab/ die syllabographische Wiedergabe einer geschlossenen Silbe mittels CV-VC enthalten, wie sie in Fāra ansonsten noch nicht üblich ist.19 Möglicherweise hat hier das mit denselben Zeichen AB und ZU dargestellte Substantiv abzu eingewirkt. UR ÚÚB šu ti: Die Zeichenkombination UR ÚÚB kommt in 3 aufeinander folgenden Sätzen vor, wobei die ersten beiden identisch sind. Wie schon oben zu wiederholtem ga-DU vermutet, könnten die Wiederholun17

Vgl. Krebernik 1984:276f.; 1998:300. Vgl. Steinkeller 1991:230f. (auch zu VE 1283) sowie Tinney 1996:180f. und Attinger 2001:141, Anm. 34. 19 Für einige ähnlich frühe Beispiele aus Tell Abū Ôalābī¶ s. Krebernik 1998:271. 18

70

Articles: Ancient Near Eastern Studies

gen auch hier unterschiedliche Lesungen repräsentieren. Da in Fāra die Zeichen TUKU und ÚÚB miteinander variieren, könnte UR ÚÚB für téštukux ‘Scheu habend, ehrfürchtig’ (ša būštam īšû) stehen. Falls ur ‘Hund’ gemeint ist, kämen für ÚÚB hingegen die Lesungen tu11 ‘niederschlagen’ oder ¶úb ‘springen, tanzen’, ‘Akrobat’ (¶uppû) in Betracht. In jedem Falle ist die Verbindung mit šu – ti ‘nehmen’ schwer zu deuten. Da kein Verbalpräfix geschrieben ist, dürfte eine infinite oder eine imperativische Form intendiert sein. UR ÚÚB du11-ga-ni šu-ni UR PÚ nu-zu: Der Schlußpassus scheint in du11-ga-ni und šu-ni das Possessivsuffix der Personenklasse zu enthalten. Wenn es sich auf UR ÚÚB bezieht, muß der Ausdruck wenigstens an dieser Stelle eine Person (oder ein als Person stilisiertes Tier) bezeichnen. Eine zumindest syntaktisch mögliche Interpretation wäre: ‘UR ÚÚB [topic] – nicht kennt UR PÚ? seine Rede (und) seine Hand’.

M. Krebernik, SF 50 // 52: Zwei … Duplikate aus Fāra

VAT 12747

71

72

Articles: Ancient Near Eastern Studies

VAT 12758-A

M. Krebernik, SF 50 // 52: Zwei … Duplikate aus Fāra

VAT 12758-B

73

74

Articles: Ancient Near Eastern Studies

Bibliographie Attinger 1993

Attinger 2001 Conti 1990 Deimel 1923 Krebernik 1983 Krebernik 1984

Krebernik 1998

Steinkeller 1981–1982 Steinkeller 1991

Tinney 1996

Attinger, P. Eléments de linguistique sumérienne. La construction de du11/e/di ‘dire’ (OBO Sonderband). Fribourg– Göttingen. Attinger, P. Rezension zu Tinney 1996. ZA 91:133–142. Conti, G. Il sillabario della quarta fonte della lista lessicale bilingue Eblaita (MisEb 3 = QuSem 17). Firenze. Deimel, A. Die Inschriften von Fara. II. Schultexte aus Fara (WVDOG 43). Leipzig. Krebernik, M. Zu Syllabar und Orthographie der lexikalischen Texte aus Ebla. Teil 2 (Glossar). ZA 73:1–47. Krebernik, M. Die Beschwörungen aus Fara und Ebla. Untersuchungen zur ältesten keilschriftlichen Beschwörungsliteratur (TSO 2). Hildesheim–New York. Krebernik, M. Die Texte aus Fāra und Tell Abū Ôalābī¶. Attinger, P.; Wäfler, M. (eds.). Mesopotamien. SpäturukZeit und Frühdynastische Zeit (OBO 160/1). Fribourg–Göttingen. Pp. 235–427. Steinkeller, P. On the Reading and Meaning of GI4 × GI4 and LAK-173. AfO 28:140–141. Steinkeller, P. The Reforms of UruKAgina and an Early Sumerian Term for ‘Prison’. AuOr 9 (= FS M. Civil). Pp. 227– 233. Tinney, S. The Nippur Lament. Royal Rhetoric and Divine Legitimation in the Reign of Išme-Dagan of Isin (1953–1935 B. C.) (OPSNKF 16). Philadelphia.

The Suffixing Conjugation of Akkadian: In Search of Its Meaning Sergey Loesov Russian State University for the Humanities, Moscow

Were a language ever completely “grammatical,” it would be a perfect engine of conceptual expression. Unfortunately, or luckily, no language is tyrannically consistent. All grammars leak. Sapir 1921:38

1. Introduction This study is an inquiry into the functions of the Akkadian Stative.∗ It starts from where BK left off: PARIS formed from telic roots is a finite verb form that expresses a trivial state resulting from the action designated by the ∗ Acknowledgements and conventions. This research has benefited immensely from my ongoing conversations with Ilya Arkhipov (IA) and Bert Kouwenberg (BK), both of whom supported the initial insights behind this undertaking and have always done their best to answer my innumerable questions. They have also taken the time to read several drafts of this paper. IA commented on most of the OB Mari examples and so has spared me numerous mistakes. In August 2010, BK sent me his “List of Old Assyrian Verbs,” an alphabetic list of verbs with their attested forms, textual references, and philological notes. He composed this List (some 150 pages long) in the course of his preparatory work towards a new Grammar of OA. For the kind of study I am involved in, this reference tool means more than I can tell. I am grateful to Maria Bulakh who read various drafts of mine related to the present study and made important theoretical suggestions. Ilya Khait shared with me his readings and interpretations of YOS 10. Olga Borovaya discussed with me fine points of English morphosyntax important for this study and helped me improve my writing. I acknowledge the financial support of the Russian Foundation for Humanities within the project “Early stages of Akkadian: the synchrony and the diachrony,” 09-04-00235а, 2008–2010. In this paper, SMALL CAPS stand for universal semantic notions (e. g. FUTURE, PERFECT, RES[ULTALTATIVE]). Terms for language specific morphological categories are written with first capitals: the (German or Akkadian) Preterite, the (English) Present Progressive, etc. The morphological shapes of Akaddian finite verb forms are written in ALL CAPS: IPARRAS, IPRUS, PARIS, etc., or are italicized: paris, etc.

76

Articles: Ancient Near Eastern Studies

respective root. However, a closer examination of PARIS of telic roots shows that non-resultative examples are perhaps no less frequent than the resultative ones and therefore they cannot be treated as “atypical” or “exceptional.” This evidence begs two questions: 1) what are the conditions that allow the resultative reading of PARIS? 2) What are then the functions of non-resultative PARIS forms? The article mainly answers the first question, whereas the second one is tackled only parenthetically. I have gathered the data for this study in the course of a lexicon-based analysis of PARIS, whereby PARIS forms of each root were studied separately. A few entries of this kind constitute the last section of the present contribution. This enterprise is an offshoot of my research into “Akkadian Sentences about the Present Time” (= ASPT).1 The ASPT venture aims at answering a semantic question: “How is the present-time domain encoded by the Akkadian verb?” The question may sound strange because most Assyriologists do not see here a problem and simply assume that for the G-stem both PRESENT and FUTURE are expressed by IPARRAS, for the D-stem they are expressed by UPARRAS, etc. I believe, by contrast, that for the G-stem verbs in the PRESENT there exists a division of labour among IPARRAS, PARIS, and IPTANARRAS. This division of labour is determined by the “Aktionsart” of a given verb, i. e. ultimately by its lexical meaning.2 Unlike ASPT, this study proceeds, in the conventional way, from a formal point of departure. It seeks to describe the semantic side (“the meaning,” “the usage,” or “the functions”) of just one member of the Akkadian finite paradigm—the Stative, called here “the Suffixing Conjugation” (= the SC), and sometimes PARIS/paris. The corpus of the study consists of OA and OB. Since my goal is to learn whatever can be learned about the “genuine” verbal morphosyntax of the two vernacular dialects,3 I am forced to treat the OB Mari with certain caution. In zero approximation, this corpus follows the OB literary variety based on the vernacular of Ešnunna (see most recently Durand 2004:113f.). For my purposes, the epistolary corpus of OB Mari can be best subdivided as follows:4 1

It is reflected in three papers published so far: Loesov 2005, 2006, 2010. For a concise formulation of this hypothesis, see Loesov 2011. 3 In other words, the aim is reconstruction of certain aspects of spoken Akkadian dialects. 4 The classification was suggested to me by IA after I had shared with him the idea of the present study. He also indicated to me the relevant secondary litera2

S. Loesov, The Suffixing Conjugation of Akkadian …

77

1) Documents authored by writers whose native tongues were various dialects of Akkadian. Here belong the major part of the correspondence of Samsi-Addu epoch (cf. Charpin–Ziegler 2007), texts written by native speakers of Akkadian from Central or Southern Mesopotamia and found in the Mari archives, e. g. those authored by Ibal-pi-El of Ešnunna (Charpin 1991), by Hammurapi of Babylon (ARM 6, 51–54; 28, 1 and 3–11), and by Hammurapi’s representatives in Mari (FM 2, 120; 6, 10–17).5 2) “Amorite” letters, especially letters of local governors and administrators of Zimri-Lim, e. g. ARM 27. They constitute the bulk of the published OB Mari epistolary corpus. Mistakes against the OB norm can be found in almost each letter of this group, it is safe to assume the influence of non-Akkadian Semitic mother tongues of the authors. Here also belong letters from Amorite kingdoms of Western Syria and Upper Mesopotamia. In this “Amorite” corpus, the command of the OB norm varies considerably from writer to writer. 3) Letters of “Barbarians,” i. e. writers of Elamite and Hurrian extraction, etc. This triple division may apply also to OB Tell Rimah and Shemshara. Hence, if, within the OB corpus, a certain morphosyntactic fact happens to obtain only in the “Amorite” group, I treat it separately as a suspect of foreign influence. One understands that in practice the borderline between groups (1) and (2) is no iron curtain. E. g., we cannot exclude that among Zimri-Lim’s scribes or other palace folks there were native speakers of OB or genuine bilinguals. Yet the present study cannot do without some such distinction. 1.1. Research History Unfortunately, there is not much to say in the way of a literature review as far as the semantics of the SC goes. I will start with the erstwhile influential essay of A. Goetze “The so-called Intensive of the Semitic Languages” (Goetze 1942). On p. 5, Goetze suggests the following tripartite semantic classification of the SC: “(1) the d u r a t i v e stative. It denotes an inherent quality of a person or a thing.” Examples are ¢āb ‘is good,’ šalim ‘is well,’ qerub ‘is near,’ rapaš ‘is wide,’ etc. This kind of the SC “is identical with the predicative ture. My reading experience in OB Mari makes me subscribe to IA’s subdivision of the corpus. 5 Curiously, there are at least two letters written in OB by a writer who must have been a speaker of OA (Durand 2001).

78

Articles: Ancient Near Eastern Studies

form of the adjective.” In the ASPT papers, I call this variety “semantically deadjectival.” “(2) The p e r f e c t stative. It denotes a condition which results from the subject’s own action with reference to a person or a thing.” It “constitutes an independent category of forms,” which means that for Goetze ´abit ‘he possesses’ belongs to a category different from both the semantically deadjectival kind of the SC (¢āb) and the forms like i´bat/i´´abat. Goetze lists here ´abit and other taking-hold-of transitive forms (with a reference to Ungnad 1918), as well as ša¢er ‘has by written deed,’ šakin ‘has placed,’6 waldat ‘has born (children),’ and other transitive tokens of the SC. He also counts here some telic7 intransitive verbs: maqit ‘is on the ground,’ ¢ebi ‘is sunk,’ etc. These SC tokens of intransitive verb “denote the rest after some movement.” “(3) The p a s s i v e stative. It denotes a state of affairs which results from another person’s action … Any transitive verb may have such a form at its side.” “[T]he presence or absence of an object” makes possible the distinction between active (= A) and passive (= P) tokens of the SC formed from transitive verbs. The illustrative examples are part of those already cited for the previous category, now they receive P glosses, e. g. a¶iz ‘(is) held,’8 ´abit ‘(is) seized,’ ma¶er ‘(is) received,’ naši ‘(is) lifted,’9 šakin ‘(is) placed.’ The treatment of the SC in GAG is inferior to the earlier insights of Ungnad and Goetze. The SC is infelicitously called “Konjugiertes Nomen” (124a). The first half of von Soden’s two-page description is dedicated to the denominal forms of the SC, which are rare outside literary texts. The analysis of the deverbal SC is marred by its dialectic identification with the Verbal Adjective (= VA): the SC is the conjugated VA, while the VA is the declined SC, “beide Kategorien sind also identisch” (GAG

6

It does not seem to be attested in OB, though Goetze refers to his forthcoming Grammar of OB as the source of his material. 7 Goetze describes the telic meaning in the following words: “only if the intransitive by itself points to the termination of the action which it denotes”. He opposes such verbs to “cursive intransitives like ‘walk,’ ‘run,’ ‘shout’,” which do not form the SC. 8 Goetze remarks that this type can be called “a passive participle in predicative use,” hence the bracketed ‘is.’ 9 We will see in due course that for some of this taking-hold-of verbs the SC P is not attested in OB, hence part of Goetze’s examples seem to be made-up.

S. Loesov, The Suffixing Conjugation of Akkadian …

79

125d). In spite of this dialectics, von Soden makes several valuable observations of both individual and general nature (GAG 126):10 – –

– –

in OB, the SC of kašādum ‘to reach’ often has future time reference;11 the SC of telic verbs is RES (‘bezeichnet der St. den sich aus der Verbalhandlung ergebenden Zustand’), the SC of telic transitive verbs is diathesis-neutral;12 the SC of atelic verbs is rare (all three von Soden’s examples of atelic verbs with no SC in the first edition of GAG are intransitive); the availability of the VA also somehow depends on the meaning of the verb, i. e. there is a pattern of “the SC with no corresponding VA” (von Soden offers a few hints but no general rule for this).

Von Soden’s description reveals an unbridgeable gap between an inadequate theoretical frame and a number of far-reaching insights. GAG (1952) had appeared before the general linguistics elaborated such now commonplace notions as situation (or lexical) aspect, 13 the RES verb forms, or transitivity as a semantic rather than a purely syntactic concept. Yet, von Soden made some important semantic distinctions in the spirit of these later theories. The first attempt at describing the semantics of the SC in some detail is Rowton’s “The Use of the Permansive in Classic Babylonian” (1962). As Kouwenberg 2010:170, fn. 32 observes, “It is a well-documented study, exemplary as to how a grammatical problem should be addressed and pioneering in its inclusion of competing categories such as the t-perfect and the passive. However, his semantic interpretation of many statives is highly questionable …”14 With some 600 classified examples, Rowton 1962 is still the best documented semantic study of the SC. Unfortunately, his corpus is ill-chosen (“Classic Babylonian” = OB + SB), his semantic criteria are impressionistic, his terminology is partly coined by the author and not strictly defined. K. Hecker’s description of the SC in GKT 118–121 (1968) is insightful and (unlike GAG) offers quite a few important examples of the SC A, 10

Some of them may have been taken from the earlier scholarship, which is irrelevant when one discusses a reference grammar. 11 A fact that I am dealing with in some detail on p. 116 below. 12 This point was penetratingly formulated in Ungnad 1918:281, “müssen wir doch stets im Auge behalten, daß der Permansiv seiner Grundbedeutung nach denjenigen bezeichnet, an dem eine Handlung zum Abschluß gek o m m e n i s t .” 13 In particular the notion of (a)telicity, which my Forschungsgeschichte uses anachronistically. 14 I endorse completely BK’s evaluation of Rowton’s work.

80

Articles: Ancient Near Eastern Studies

building on Rowton’s semantic classification. Unfortunately, it did not really influence our thinking about the SC, probably because this grammar of OA was meant as a non-ambitious practical addendum to GAG. The year 1968 saw the publication of G. Buccellati’s 12-pages essay “An Interpretation of the Stative as a Nominal Sentence,” whose thrust is adequately rendered by its title. The most daring feature of this paper is that it brings no single example of the SC formed from a genuine verbal root, it hardly ever mentions their existence, while every reader of Akkadian knows that such forms are the only productive ones. Buccellati prefers to deal mostly with made-up examples, some of which may be grammatically wrong (see Kraus 1984). Buccellati’s thesis won the day for the next 30 years. The interpretation of the SC as a nominal clause (verbless clause, “predicative construction,” etc.) was accepted by several generations of scholars in the field, see in particular Huehnergard 1986, 1987 and 1997, Reiner 1970, Streck 1995 (NB the literature review pp. 177–186), Tropper 1995. At an early stage of my work on the problem, BK agreed, in a p. c., that “nobody published anything fundamental on the function of the stative since Rowton. At least one reason, I think, is Buccellati’s 1968 article, which for some reason (incomprehensible to me) was so influential that it may have deterred people from sticking out their neck on this subject.” One of the few dissenting voices was F. Kraus: “Ebenso wortreich wie ergebnisarm krankt der Artikel an schweren Fehlern verschiedener Art … [E]ine detaillierte Auseinandersetzung mit einer so gearteten Arbeit untunlich, wo nicht unmöglich wäre …” (Kraus 1984:39f.).15 It is probably not accidental that the protest was voiced by F. Kraus, the editor of numerous texts and publisher of AbB, who knew the language firsthand but whose interests were other than a “scientific” interpretation of Akkadian grammar. From the notes in the text editions by Kraus it is evident that his interest in grammar was primarily philological, related to the problems of literal understanding of text. Meanwhile, most students of the grammar of Akkadian in the 20th century were not particularly curious to learn what exactly “tenses” in general and the SC in particular do in text. As Kraus points out, it is some kind of (ill-equipped) 15

D. Cohen in his important 1984 monograph on the emergence of new verb forms in Semitic from “la phrase nominal” also seems to follow the lead of Buccellati 1968, but this is not that vital for his study of grammaticalization. D. Cohen was the first one to compile a list of Akkadian “active statives,” which is useful for the student of Akkadian grammar.

S. Loesov, The Suffixing Conjugation of Akkadian …

81

diachronic excitement that was “oft der eigentliche Anlaß zur Beschäftigung moderner Gelehrter mit dem Stativ” (p. 39).16 It stands to reason that the SC was grammaticalized in prehistoric times from the nominal clause consisting of the resultative VA + pronominal subject in the Nom. to its right.17 In itself, this fact has nothing particular about it, since “today’s morphology is yesterday’s syntax” (Talmy Givón). As for the burning historical question, the relationship of the Akkadian SC to the WS “Perfect,” it remains as unsolved as it used to be a century ago. For what it is worth, I would bet that the two paradigms are genealogically unrelated, in particular the nominal base of the WS one is a predicative adjective other than *paris-, i. e. most probably *paras-. On this scenario, *paras- would be soon lost by the WS as a productive attributive participle (or substantive). This is the fate of the PS agent noun *pāris- in various Neo-Aramaic languages: *pāris- became the base of the new present-future forms and was hereby discontinued as a nominal stem. Another possible analogy is the Proto-Slavic perfective participle with the suffix -l (Meillet 1934, § 281f., cf. Vaillant 1948, § 164). In historical times, it is used almost exclusively within analytical tenses that include various finite forms of the verb ‘to be,’ yet at some (pre-Slavic?) stage it must have been a productive noun or adjective that was used in nominal (non-predicative) slots, its appearance as a predicative adjective can hardly be explained in a different way.18 Kouwenberg’s article “Nouns as Verbs: The Verbal Nature of the Akkadian Stative” (2000) rehabilitated the SC as a finite form. As for the SC tokens formed directly from fientive verbal roots (or, as BK puts it, “from the verbal paradigm”),19 BK assigned them an honourable slot within the 16 I do not summarize here Rowton’s and Kraus’s theoretical considerations on the SC, and this is for two reasons: (1) both writers did not aspire to produce the theory; (2) their general ideas are now partly obsolete, while numerous insightful observations they offer are still a challenge and will be useful for any future description of the Akkadian verbal morphosyntax. On the contrary, the nominal sentence theory of the SC did not provide anything of use for our understanding of Akkadian. 17 All the stages and details of the grammaticalization process will be probably never fully recovered. 18 Among the recent opinions on the problem, see in particular Voigt 2002– 2003 and Kouwenberg 2010 § 7.4. 19 This variety, the only one productive in Akkadian, is the subject of the present research. Genuine conjugated nouns (both substantives and adjectives) are not considered here. I believe that this predicative status of nouns is an inner-Akkadian analogical development vis-à-vis the SC as a verbal “tense,” cf. Loesov 2005:142ff. and 2010:763f.

82

Articles: Ancient Near Eastern Studies

standard typological set of tense-aspect meanings: the SC is now REin the sense of Nedjalkov (ed.) 1988 (= TRC).20 As everybody knows, the verb in ancient West Semitic languages had very limited possibilities to express tense-aspect meanings via inflectional morphology. It is usually assumed that the WS verb at the earliest stage of its life could only oppose PERFECTIVE to IMPERFECTIVE (or PAST to NON-PAST). Against this background, Akkadian looks conspicuously well-equipped: it had four tense-aspect conjugations plus productive derivational means for coding the aspectual meaning “iterativity,” i. e. tan-stems that could be formed from each of the four basic stems. Cross-linguistically, Akkadian verb is interesting due to the combination of two features: it has different inflectional forms for PERFECT (iptaras) and RES (paris), while RES of transitive verbs can get P or A readings though it does not inflect for diathesis.21 SULTATIVE

1.2. RES and the Akkadian SC. An overview and criticism of BK’s position The previous paragraphs concluded the introductory story. I will now summarize BK’s views on the functions of the SC, as expressed in Kouwenberg 1997, 2000 and 2010, because this is the juncture where the present study starts. 20

I do not produce here a summary of BK’s argumentation, among other things because he did this himself in Kouwenberg 2010, 7.3. To my knowledge, in the past ten years no attempt has been made to refute Kouwenberg 2000, in particular by the scholars who had bought into the Buccellati 1968 nominal sentence idea. It is therefore a pity that certain popularizing and/or typologically oriented descriptions of Akkadian keep misleading the linguistic readership about the place of the SC in the grammar of Akkadian and consequently about the very structure of Akkadian verb. Thus, Sh. Izre’el maintains that the SC is a “nominal predicative” that “can be substantival or adjectival” (Izre’el–Cohen 2004:49). Streck 2005:51 notes that “Die Analyse des Stativs als nominal oder verbal ist umstritten,” which does not mean anything for the outsider; the only important thing is that the SC was excluded from his morphosyntactic description of the Akkadian verb. 21 According to TRC, A readings of RES formed from trans. verbs are cross-linguistically rare (this follows naturally from the semantic core of RES), while identical morphological marking of P, A and intransitive (= S) meanings is much more frequent than otherwise. Nedjalkov 2001 observes that ‘[i]dentical marking has a functional explanation: it is due to the fact that in all the three cases the derived subject is identical with that underlying constituent which is the most affected by the previous action. In the case of S-resultatives, it is the only constituent, and in P- and A-resultatives it is the most affected one’ (p. 932). This is exactly what Ungnad 1918 tells us.

S. Loesov, The Suffixing Conjugation of Akkadian …

83

BK’s perspective on the grammatical meaning of the SC is consistent and uncompromising: since 1997 up to the present he has been insisting on the purity of the STATIVE meaning of the SC, against all kinds of “syntax-and-semantics” or “target-language-of-translation” oriented attempts to weaken or relativize this basic tenet. Consider BK’s criticism of Goetze’s semantic classification: “[T]he distinction which Goetze makes between three types of statives reflects a semantic distinction between different types of verbs which has no grammatical relevance for Akkadian. From the point of view of Akkadian, the stative has a completely uniform grammatical function, namely, to denote a state, i. e., the absence of any action or change. All differences between individual statives, such as whether a stative expresses a “pure” state (e. g., šalmāku “I am well”), or a state resulting from a previous event (e. g., ¶alqat “she [has escaped and] is [now] on the run”) … are determined by the lexical meaning of the verb and by the context” (Kouwenberg 1997:14).

This thesis may have been influenced by our knowledge of the etymological (or primary) stative-resultative meaning of the SC. Let us narrow down the discussion to dynamic verbs as the most frequent ones and the natural representatives of the VERB as a part of speech. As I believe, the VA of telic transitive verbs in predicative use may have been the genuine historical kernel of the SC. After the verbalization of this predicative construction and its spread to intransitive telic verbs had taken place, the young SC must have been impeccably resultative, but this was no longer the case in historical times. To prove this, we first have to formulate restrictions on the manifestation of the resultative sense, i. e. we have to single out syntactic environments incompatible with RES. To be able to do so, we must put forward a strict working definition of RES. According to TRC 6, “The term resultative is applied to those verb forms that express a state implying a previous event. The difference between the stative and the resultative is as follows: the stative expresses a state of a thing without any implication of its origin, while the resultative expresses both a state and a preceding action it has resulted from.”22

22

According to Lyons 1977:483, “A static situation … is one that is conceived of as existing, rather than happening, and as being homogeneous, continuous and unchanging throughout its duration. A dynamic situation, on the other hand, is something that happens …”.

84

Articles: Ancient Near Eastern Studies

This definition, as well as J. Lyons’ definition of the static situation over against the dynamic one, is not sufficient for our purpose. In our case, the contrast between the categories SIMPLE PAST (or “narrative tense”), PERFECT, and RES can be best formulated in terms of relationship between Situation Time (= SitT), Speech Time (= SpT), and Reference Time (= RefT).23 SIMPLE PAST: PERFECT: RESULTATIVE:

SitT is its own RefT, it precedes SpT.24 SitT precedes RefT (which in case of PRESENT PERFECT coincides with SpT).25 RESULTATIVE: The event’s resultant state exists at RefT, which in the default case coincides with SpT.26

One conclusion that follows from these definitions is the non-narrative nature of PERFECT and RES. Verb forms with these meanings do not belong in narrative chains. Unlike SIMPLE PAST, PERFECT views a past event from a temporal vantage point external to this event. In the natural setting of utterances, i. e. in oral dialogue, both RESULTATIVE and STATIVE can become what they are supposed to be if they are contemporaneous with SpT as the default RefT. From the above there follow four commonsense syntactic restrictions on the appearance of RES encoded by a special linguistic shape,27 and the list is most probably not exhaustive:28 23

See Reichenbach 1947:287–298. Reichenbach’s insights have long become part of the standard procedures in describing tense-aspect, see e. g. Smith 1997. 24 Thus, SpT and therefore the speaker participate in the definition of SIMPLE PAST. In case of narrative verb forms (e. g. the Akkadian IPRUS), RefT “is determined by the point in time at which the last event related in the preceding context took place” (Dahl 1985:112). 25 For Östen Dahl, PERFECT is characterized by “a point of reference (in Reichenbach’s sense) which is different from the ‘point of event’ ” (Dahl 1985:133). In Loesov 2004, I tried to show (following Goetze 1936) that the basic meaning of the Akkadian IPTARAS is PERFECT. I defined this basic meaning as “now extended past-wise” (p. 172), which is materially tantamount to the formal definition “SitT precedes RefT” and to Dahl’s words. The famous “resultative component” of PERFECT, which I also embraced in Loesov 2004, now seems to me no less metaphorical than the “current relevance” approach. Therefore I accept the criticism of Kouwenberg 2010:141, fn. 8. 26 Since the target of these considerations is the SC of Akkadian, this statement presupposes that the resultative form does not inflect for tense, unlike e. g. the German Zustandspassiv. 27 The stipulation about a special linguistic shape (“non-combined,” in the language of TRC) is essential because every perfective verb form can have re-

S. Loesov, The Suffixing Conjugation of Akkadian …

85

1. RES does not occur in non-subordinate narrative sentences. 2. RES is incompatible with past (and probably future) time adverbials but can be modified by temporal adjuncts that include the SpT: “Is rowan atkinson married? Yes, He is married since 1990 (to present) and has 2 children” (http://www.trueknowledge.com), “kevin ya está casado desde el sábado 19” (http://jonasbrothers96.obolog.com). 3. RES is incompatible with place adverbials, unless the verbal root itself denotes spatial relationships, e. g. a position in space.29 The incompatibility of RES with place adverbials that do not point to the location of speech act plausibly follows from the above definition of RES. Yet in various European languages the hic adverbials are also incompatible with resultative constructions: *I am married here, *Ich bin hier verheiratet, *Estoy casado aquí, *Я здесь женат etc. are all hardly acceptable. Thus, I have to posit that deictic localization also possesses a semantic element incapable of coexisting with RES in some languages.30 4. Logically, genuine RES should not square with negations. Within the deictic register of speech (RefT = SpT), negation of the trivial resultant state may presuppose that the corresponding telic or punctual event has failed to take place, therefore there is a chance that {NEG + RES} will sultative readings. As we will see below, the nearest example at hand is the Akkadian Pret.: some basic telic verbs (both transitive and intransitive) do not have the productive SC, therefore their Pret. and the t-Perf. appear in resultative contexts as well. 28 I do not pretend to believe that all these restrictions are 100% universally valid. RES is by definition a complex and therefore unstable notion (it is its semantic depth and ambiguity that make it a worthy object of attention). Its grammatical shape in a given language lends itself synchronically to other uses (thus, the English ‘it is done’ has a resultative and a habitual reading). Besides, much depends on the etymology (or “inner form”) of the respective verb form in a given language. My arrangement of restrictions is meant to follow in the order of diminishing universality (or frequency), but this is also a guess. Needless to say, all four restrictions have been mentioned in typological literature. 29 Examples are verbs with the meanings ‘to sit,’ ‘to lie,’ ‘to hang down,’ etc. In case of such verbs, a place adverbial may fill an obligatory valence. The deletion of place adverbials in resultative constructions of non-location verbs is explicitly mentioned in TRC 54, with a German example: Sie sind dort gefesselt worden → Sie sind (*dort) gefesselt. There is a note ibid. to the effect that cross-language the compatibility with place adverbials helps tell dynamic verb forms from resultative ones. 30 In typological literature, I have found no comprehensive discussion of the problem based on a representative sample from world’s languages. TRC has only occasional remarks pro et contra in the chapters on RES in individual languages.

86

Articles: Ancient Near Eastern Studies

be either impossible or will function as the negation of a dynamic pasttime verb form, this {NEG + RES} may have the PERFECT reading.31 I am now going to adduce a selection of examples from OB and OA where the SC defies these restrictions and has past-time dynamic rather than resultative readings. I have already briefly discussed the problem in Loesov 2005:133f., with references to previous literature, to which I have to add Kraus 1984:31–32. Most recently, Kouwenberg 2010, 7.3.3 introduced the past-time use of the SC under the heading “Marginal and secondary uses of the stative” and provided a dozen OB and OA examples, some of which include the SC of dâkum ‘to kill’ and mâtum ‘to die.’32 BK’s reason to consider these examples non-resultative is syntactic: “[T]he inclusion of a specification of the time and/or the place of death … shifts the attention from the present state to the previous event since it (the specification.— S. L.) can only refer to the event itself.” BK explains this evidence simply by suggesting that Akkadian “was not completely immune” to the diachronic development RES → PERFECT, well-known in the world’s languages. In his opinion, “[t]he Akkadian stative did not become a productive perfect … because Akkadian already had one in the iptarVs form.” To my mind, this explanation does not hold water, and this is for two reasons: (1) As we will see in the course of this study, dynamic tokens of the SC in OB and OA (being of course diachronically secondary) are by no means “marginal,” they are frequent. (2) The SC did not succeed in becoming a regular PERFECT in the rest of the lifespan of Akkadian, in spite of the fact that in the corpora of MB and MA iptarVs acquired the meaning of SIMPLE PAST/PERFECTIVE in the foreground sentences.33 This means that one will have to look for a different explanation of the evidence.

31 See Kozinskij 1988:511f. for an analysis of data collected in TRC. The problem may be relevant for Akkadian, since in OB the t-Perfect is incompatible with negations in most syntactic contexts. The standard negative alloform of iptaras in main clauses is ul iprus, yet the literature reports cases of negated paris with allegedly dynamic force (e. g., Kraus 1984:11, Veenhof in a footnote to AbB 14, 150: 11ff.). 32 BK also introduces another secondary use of the SC, the inflected forms of šakin and nadi functioning as copular predicates. This fact is also relevant for my study. 33 This last statement is based on current reference tools, including Kouwenberg 2010.

S. Loesov, The Suffixing Conjugation of Akkadian …

87

* * * Let us now preliminarily survey some evidence for the SC violating the four constraints on RESULTATIVE formulated above.34 Non-subordinate sentences that relate past events (“narrative” in the widest sense), the SC is often but not necessarily followed by the sequencing -ma: (1)

GU4.ÚI.A ša PN ¶a-al-qú-ma ina qāti PN1 u PN2 DÀM.GAR i´batūšunu-ma umma šunu-ma nādinānū ša iddinūnâši ibaššû ‘Oxen of PN had been lost, and (afterwards) they found them in possession of PN1 and PN2, a merchant. This is what they said: There are certain sellers who sold us (the oxen). [Their names are …]’(Goetze 1958, No. 28:4–11, OB kingdom of Ešnunna).35

The SC ¶alqū-ma stands in the narrative chain and constitutes part of the crime description: alpū … ¶alqū-ma ina qāti … i´batū-šunu. BK tells me, “I doubt if ¶alqū-ma is part of the narrative chain; I rather think it is a circumstantial backgrounded clause, which is very common with stative + -ma at the beginning of a passage. One could also call it ‘virtually subordinate.’ I wonder if there is not some specific kind of rule why the stative is used here instead of the preterite: why not i¶liqū-ma, and what would be the difference if i¶liqū-ma had been used?” This is similar to the way M. B. Rowton understands these paris-ma-initial chains: “[T]he capacity the permansive has to speak of the background of other events. In paratactic syntax it often speaks of the circumstance in which other action occurred, and in that sense it is ‘circumstantial’ ” (Rowton 1962:235, the examples are on pp. 271–278). Rowton tries to answer the question about the distribution of chain-initial paris-ma and iprus-ma: “In OB the enclitic particle (-ma.—S. L.) usually denotes no more than sequence of events when it is affixed to the preterite or the perfect. But when affixed to other tenses, it usually denotes a logical connection of some sort between the two sentences it joins” (p. 272). To escape circular reasoning, I prefer for the moment to keep here the term “narrative chain.” After all, even in the Pret1-ma … Pret2-ma … Pretn-ma chain the conjunction -ma secures the effect of what BK calls “virtual subordination” (see Patterson 1970).36 It may well be that the chain-initial SC-ma was used as a pluper34

As we will learn little by little, this violation has a pattern. Goetze notes in the Edition: “This letter illustrates a law concerning lost animals which is analogous to § 9 CH: Lost animals have been found in somebody’s possession; the possessor must prove that he acquired the animals lawfully.” 36 Building on the above observations of Rowton and BK and on the evidence collected in the course of my research, I will allow myself a near-at-hand guess: in narrative passages, the relationship between paris-ma iprus and iprus-ma iprus 35

88

Articles: Ancient Near Eastern Studies

fect of sorts (see below), but PLUPERFECT is by no means RES with a pasttime reference point (= the RefT in the speaker’s past): PLUPERFECT is dynamic (fientive), while the non-resultative nature of our ¶alqū-ma is the only thing that matters for us at this point.37 (2) PN ¶a-li-iq-ma ina GN … wašib ‘PN has left (the ranks) and stays … in GN’ (AbB 6, 171:6–10). (3) PN MU 5.KAM e-ri-iš-ma warkānum PN1 ina emūqim īkim-šu ‘PN cultivated (the field) for five years, afterwards PN1 took (it) from him by force’ (AbB 4, 160:27′ff.). (4) ittī-kunu bā!erūtam e[pē]šam ku-ul-lu-mu-ma taprik-am-ma ana bā!erūtim epēšim [ul t]addin-aššunūti ‘They were assigned to do the fishing with you, but you were a hindrance and did not allow them to do the fishing. ’ (AbB 11, 112:18–22).

Kullumū is part of a narrative sequence and has an adjunct ittī-kunu ‘with you.’ (5) PN bí-it ana šarrim i¢¶a-ma ‘PN spent the night and then had an audience with the king’ (AS 22, No. 19:3ff., tr. Whiting; archaic OB).

The SC sentences in (1)–(5) can be interpreted as backgrounding/ “virtually subordinate” (though formally paratactic) predications, yet the only formal correlate of this claim (however plausible) is the usage of the SC shape. Thus, the problem needs further research. Complement and relative clauses where the SC has the reading of the pluperfect type (the situation coded by the SC is prior to a past-time reference point expressed by the predicate of the main clause): (6) ¶īšam ša ana šītat kaspim ez-bu-ši īrišū-ši-ma

chains is that of markedness. The chain paris-ma iprus may suggest explicitly that there is something “backgrounding” in its first link, while iprus-ma iprus is a default expression. 37 Note that below (in connection with BK’s perspective on the functions of the SC) I will try to cast doubt on the traditional formulation regarding “the lack of tense distinctions” in the SC. Simply put: in so far as the function of the SC is indeed RES, only the present-time domain is unproblematic. For past and future, we have to specify linguistic environments that allow (or forbid) the resultative meaning of the SC. All these niceties are due to the complex and fragile structure of RES.

S. Loesov, The Suffixing Conjugation of Akkadian …

89

‘They asked her (to produce) the binding agreement which had been made in her favor for the remainder of the money’ (TCL 1, 157:34, text and tr. CAD E 422b, OB leg.). (7) kīma … ina sūnī-ki nīlu-ma aqtīp ‘I have been led to believe that he slept with you’ (TCL 1, 10:23, text and tr. CAD Q 93a, OB letter). (8) ¢up-pá-am ša ku-nu-ki-a ša a-ni-ša-am wa-ba-la-am qá-bi-a-tí-ni u4-maam lá ta-ta-ab-lam ‘Today you have not brought me the tablet with my seal that you promised to bring here’ (Michel 1991, No. 263:9ff.).

PARIS with adverbials of time and place: (9) a-šu-mì 25 TÚG ku-ta-ni ša 〈tù〉-šé-bi4-lá-ni-ni i-na GN ´a-áb-tù a-na É.GAL-lim né-li-ma ‘As for the 25 k.-textiles that you (pl.) sent me, they have been confiscated in GN. We went up to the palace and …’ (CCT 4, 19c:15– 19, OA).

The SC ´abtū has the reading of (present) PERFECT: confiscation of the textiles is the “hot news” and the rationale of the letter. (10) ištu ITI.1.KAM šerrum ina libbīya mi-it-ma ‘The child in my womb died a month ago’ (ABIM 15:10). (11) ina panītim ana LUGAL.MEŠ šunūti awātam kīam ´a-ab-ta-ak-šu-nu-ši-im ‛Some time ago, I addressed these kings the following speech’ (A.1025:5ff. = MARI 6, 337ff.). (12) 5 ma-na 10 GÍN AN.NA iš-tù Za-al-pá a-dí Kà-ni-is na-dá-ku 23 ma-na 10 GÍN AN.NA gám-ru-um ša A-sà-nim ‘5 minas and 10 shekels tin I have spent between Zalpa and Kaniš. 23 minas and 10 shekels tin are the expenses of Asānum’ (BIN 4, 116, OA, and cf. the tr. of Ulshöfer 1995:272).

This is a travel expenses note, cited here in its entirety. The sentence whose predicate is nad!āku speaks about the author’s expenses during the travel rather than about his own resultative state.38 (13)

38

KÙ.BABBAR 1 MA.NA 4 ½ GÍN iš-tí DUMU sà-a¶-ru-ta-na-im e-ri-iš-ma šébi-lam šál-ma-am i-na a-limki ša-qá-lam qá-bi

Contrast ŠU.NIGIN x KÙ.BABBAR na-ad-a-tù-nu ‘You have deposited, all in all, x silver’ (BIN 4, 33:15f., OA). Here the SC A of nadā!um, in a different lexical meaning, does describe the resultative state of the addressees, see 2.1.2 below and the analysis of nadûm in the forthcoming part of this study.

90

Articles: Ancient Near Eastern Studies ‘Demand from the son of the Sahrutanean silver (in the amount of) 1 mina 4 ½ shekel and send (it) to me. In the City, he has promised to pay in full’ (Prag I 472:32–36, a letter from Assur to Kaniš).

Negated forms of the SC:39 (14) ul ina pilši ka-aš-da-a-ku ‘I have not been seized in burglary’ (AbB 2, 83:32).

Note also the place adverbial. (15) aššum pīka lā ša-ma-ku ul ma-ag-ra-ak-šu-nu-ti u anāku ul appal ‘Because I have not yet learned your decision, I have not granted their request, since (lit. “and”) I myself cannot answer them’ (AbB 14, 150:11ff., Veenhof ’s tr.).

In a fn. to his translation of ul ma-ag-ra-ak-šu-nu-ti, Veenhof suggests that here “the stative serves as perfect tense.” Veenhof may well be right, this is probably a negative alloform of the t-Perf.40 (16) ištēn alpum ul na-d[i-i]š-šum ‛Not a single ox has been given to him’ (Goetze 1958, No. 1:22). (17) šumma A.ŠÀ šaddagdam lā e-ri-iš-ma nadi ‘Wenn das Feld im vorigen Jahre nicht bestellt worden ist und brachliegt’ (AbB 2, 92:15f.).

The SC is used here in the “natural” slot of the N-stem Pret. The time adverbial šaddagdam secures the past-time (non-resultative) interpretation perhaps more powerfully than the negation. (18) awīlû ul wa-´ú-ú lā tušadda-šunūti ‘The men are not (yet) gone. Do not prevent them (from leaving)!’ (AbB 8, 87:10f.).

This example is important for our understanding of what the SC is all about. The verb wa´ûm belongs to the basic vocabulary, it is extremely common in the corpus. To my knowledge, awīlû ul wa´û is the only OB example of wa´i for the literal telic meaning of the root ‘to go/come out’ (cf. Loesov 2006:139).41 This evidence calls for explanation. The one that 39 According to the above, {negation + the SC} may be forbidden to be resultative, but there is no reason why it should be past-time and nothing else, and actually it is not. I adduce here only past-time examples simply to make this point comparable with the other ones. 40 Note that the form lā šamāku stands in a subordinate clause, and this environment may be in itself unfavorable for RES. 41 To the material collected in Loesov 2006, add ARM 26/2, 449:42: šumma-

S. Loesov, The Suffixing Conjugation of Akkadian …

91

follows from the present study is two-fold: 1) as we will see below (2.2), intransitive telic motion verbs that are unergative (i. e. their subject is semantically AGENT-like)42 usually do not form the resultative SC for the basic meanings of the respective roots;43 2) Akkadian can tolerate negated SC forms of unergative verbs with past-time readings. In other words, a past-time ul wa´i ‘he has not gone out’ might sometimes have been OK, while wa´i *‘he is gone’ was unacceptable. * * * I will now review BK’s perspective on the meaning of the SC as presented in Kouwenberg 2010. At the outset of his discussion (7.3, p. 163), BK restates his now classical thesis: The grammatical function of the stative is the expression of a state. It is used indiscriminately for all kinds of states, whether permanent or transient, whether a “pure” state or a state resulting from a previous event. As such, it is opposed to the fientive members of the verbal paradigm, which express events.44

With Kantian integrity, BK deduces from this basic premise all the logically compelling semantic conclusions (all the quotations are from 7.3, boldface is added): – “[T]he lack of tense distinctions: the stative may refer to the present, the past, and—much more rarely—the future.”

This statement, well-known from earlier (theoretically unambitious) literature, is nonetheless not self-evident within BK’s approach. For the SC with future reference, BK refers to the material collected in GAG § 77d* and Leong 1994:244. In these sources, reliable OB examples are limited to the SC of kašādum in the motion sense ‘to come,’ and wašābum man … bēlī ina libbī-ka wa-´í. The lexical meaning is figurative (lit. ‘were … my lord to go out of your heart’ ~ ‘were you to forget my lord’), the grammatical reading of wa´i is dynamic (i. e., there is no question of the trivial resultant state of “the lord” as the one who has gone out). The form is suspect of adstratum influence. 42 The notion contrary to “unergative” is “unaccusative”: an unaccusative verb is an intransitive verb whose subject is semantically PATIENT-like. The verb wa´ûm is unergative, while e. g. mâtum ‘to die’ is unaccusative. 43 Cf. also such basic verbs as tebûm ‘to get up,’ târum ‘to return,’ erēbum ‘to enter’ (Loesov 2006:139f.). 44 In a fn. to this paragraph, BK notes, “Only in very specific, exceptional circumstances are some statives capable of referring to events,” with reference to 7.3.3 (“Marginal and secondary uses of the stative”), with which we are already familiar.

92

Articles: Ancient Near Eastern Studies

‘to stay.’ Now inflected forms of kašid with this meaning happen to be an exceptional case, for some reason they always have future reference (Loesov 2006:147, No. 37). Besides, they invariably have specifications of goal, the coming time is also often indicated: ana U4 2.KAM ka-aš-da-ka ‘in two days, I will come to you’ (AbB 12, 42:14), so they seem to be not resultative by BK’s criteria (see also No. 39 below). The forms of wašib are not confined to future, but they have a virtually obligatory valency for place specifications; the two future-time examples reported in GAG and Leong 1994 (AbB 9, 117:11; OBTR 114:7) are 1st p. sg. (wašbāku) and have future-time temporal adjuncts (ištu inanna U4 5.KAM; U4 3.KAM). Since wašābum is a locative verb (see p. 85 above), the resultative sense of its SC may have survived in this environment, but this cannot be presumed and requires a special lexical study. Thus, in view of the above cross-linguistic constraints, the conditions under which the SC can get the meaning {PAST + RES} and {FUTURE + RES} have to be studied on their own and described in general terms.45 – “[I]n terms of semantic transitivity, statives have “zero transitivity,” since they do not indicate a change in the state of the world. Accordingly, they cannot have an agentive subject, since agentivity implies a conscious volitional act on the part of the subject and is therefore only applicable to actions.” – “[S]ince statives do not envisage the termination of the state … they are by nature atelic, whereas the fientive verb itself must be telic in order to have a stative at all.” – “[S]tatives derived from verbs are neutral for voice: they can be “active” or “passive.” This is caused by the fact that they only refer to the result of an event and do not indicate how it came about; it is therefore immaterial whether its subject was the agent or the patient of the event.”

We are now passing to section 7.3.2 “Statives derived from verbs.” Let us adduce those of its theses that are especially relevant for us:46 – “Verbal statives … are the most common and most complex type. … [T]hey denote the state which results from the event expressed by the fientive forms of the verb (GAG § 77e). … The occurrence of resultative 45 I can add a future-time example of the SC from a divination apodosis:r imas-at-ma ina lā ālišu qebir ‘he will die and will be buried not in his own city’ (TIM 9, 79:4; text and interpretation as in Metzler 2002:189). Once more, qebērum (at least in resultative readings) is a location verb. For two more allegedly future-time examples of the SC, see Loesov 2005:127, exx. No. 40 and 41. 46 The book is meant to be a reference tool, so some of these statements seem trivial, yet for the purposes of the present study we have to keep in mind all of them.

S. Loesov, The Suffixing Conjugation of Akkadian …

– –

–

–

–

–

–

–

93

statives is therefore restricted to verbs denoting telic events, which culminate in a state. Atelic verbs for activities … do not normally have a stative.” “[W]e can distinguish three kinds of verbal statives: intransitive statives, passive statives and active statives.” “The most typical cases47 are statives of intransitive verbs of telic movement, such as ¶alāqu ‘to get lost,’ kamāsu ‘to kneel down,’ pa¶āru ‘to come together,’ qerēbu ‘to approach,’ rabā´u ‘to crouch, lie down,’ tebû ‘to stand up’ and wašābu ‘to sit down,’ and change of state verbs, such as belû ‘to go out (of fire),’ mâtu ‘to die,’ pašā¶u ‘to become calm’ and šebû ‘to become satisfied.’ ” “If the verb is transitive, either the subject or the direct object of the underlying fientive clause may become the subject of the stative, depending on the context and on semantic and pragmatic factors outside the stative itself, which … is neutral with regard to voice.” “If the subject of the underlying transitive clause retains its subject position in the stative, we have an “active stative,” which describes the state of the subject after the completion of the event.” “Statives of high-transitivity verbs are almost always passive. This is due to the fact that it is the degree of affectedness of the participants which determines whether the resultant state is more likely to be predicated of the subject (the agent) or of the object (the patient).” “Statives of low-transitivity verbs, on the other hand, can be both active and passive, because the direct object of low-transitivity verbs is not or not significantly affected by the action (see 3.4), so that their subject becomes proportionally more salient and therefore a more likely candidate to be described as having performed the action (Kozinskij 1988:517– 521). Therefore, most active statives come from transitive verbs with a low degree of transitivity.” “The fact that the stative describes an entity in terms of the result of a previous event makes the active stative eminently suitable for legal (con)texts to describe the legal status of a person (Rowton 1962:292– 294).” “[A]ctive statives have a particular tendency to develop a lexicalized meaning.”

* * * In BK’s semantic description of the SC, each of the statements is in itself correct. Yet the description is not gapless, because in fact so many tokens of the SC of dynamic verbs have readings other than RESULTATIVE (they are not necessarily past-time). If we sift the evidence, using among other things the above constraints on RES, we will see that the totality of non-re47

I. e., of the SC of intransitive verbs.

94

Articles: Ancient Near Eastern Studies

sultative readings is so prominent that they cannot be simply lumped together and explained away as “marginal.” Therefore, students of the Akkadian verb have to deal with this material and to find out what were the tense-aspect functions of these non-resultative PARIS forms in text. As we now know, Assyriologists have been since long ago aware that the SC of dynamic verbs is nontrivially sensitive to the lexical semantics of the root (i. e. more so than other “tenses”). Three elements of lexical meaning have been mentioned in the previous studies: (1) Valency (one- or two-place verbs, i. e. syntactic [in]transitivity). (2) “Aktionsart,” also known as “lexical aspect” or “actionality”: atelic verbs do not normally form the SC, and this stands to reason in view of the resultative meaning of the SC. (3) The degree of semantic transitivity (≈ dynamicity) of two-place verbs: high-transitivity verbs tend to have only the SC P, low-transitivity verbs can produce the SC with both A and P readings.48 This difference also stands to reason: it is due to transparent pragmatic factors following from the essence of RES (as was explained in the above quotations from Kouwenberg 2010).

By the same token, since Goetze 1942 atelicity has been the only known lexical restriction on the formation of the SC shape paris (with whatever meaning). Relative frequency of the SC of basic telic (and punctual) verbs has been never discussed, although every reader of Akkadian will tell you that the SC of šakānum ‘to place’ is very popular, while off the top of his head he will probably confess that he has never come across the SC of wabālum ‘to bring,’ although the two words seem to belong in the same semantic class “verbs of telic transitive motion.” Same is true of verbs of telic intransitive motion. E. g., the SC of maqātum ‘to fall’ is frequent, the SC of tebûm ‘to get up’ is rare, while that of erēbum ‘to enter’ is hardly attested at all. There must be a reason that is responsible for this state of affairs. * * * So far, we have put forward two observations, to be supported by further evidence:

48 To my knowledge, this last criterion was first explicitly suggested by D. Cohen (1984, with different terminology) and elaborated by BK. Its kernel probably goes back to A. Ungnad.

S. Loesov, The Suffixing Conjugation of Akkadian …

95

(1) The application of commonsense constraints on RESULTATIVE allows us to detect a sizable number of the SC forms with dynamic readings.49 (2) The SC morphological shapes (with whatever grammatical readings, resultative or otherwise) of basic telic verbs display very uneven token frequencies,50 from very common to zero.

At first sight, the two observations seem to be unrelated, yet we will see that they touch different sides of the same reality “où tout se tient.” If one stayed with the syntactic limitations only [= observation (1)], the matter would appear quite simple: PARIS of dynamic roots is resultative unless the environment blocks its resultative force.51 One would then have to ask why Akkadian needs PARIS in these kinds of environment at all, and one would probably surmise something like this: since syntactic constraints secure complementary distribution between resultative and dynamic tokens of PARIS, the latter were reemployed as pluperfect/background forms, which was a sensible thing to do, given that Akkadian verb cannot render this meaning morphologically, i. e. without syntactic support. Yet the lexicon-based description of the SC forces on us observation (2), i. e. it poses the question about basic telic verbs with no (or almost no) SC, hence one has to look for the reasons. As we will see, this search leads to a third observation: for telic and punctual verbs, both the availability of the SC and whether it is resultative or not further depends on certain fine-tuned parameters of meaning of the given lexeme, to be described below. A fourth lexicon-based observation starts from inflection-morphology evidence: the rarer is PARIS of a given frequent telic lexeme, the more is the chance it will be used in nonresultative senses. Consequently, we may chance to come across PARIS forms of telic verbs that cannot have resultative readings.

49

As I mentioned above, some such constraints have been hinted upon time and again in the literature (at least since Rowton 1962 and Kraus 1984), but they have been never applied systematically. 50 Token frequency is the number of times a given form of a given word is used, in this study it is always the sum total of the SC examples for a given Akkadian verb. It is opposed to type frequency, i. e. the number of existing lexemes with a given property (e. g., the totality of frequent high-transitivity verbs with no SC). 51 This is how I understand BK’s position, although he underestimates the amount of non-resultative examples.

96

Articles: Ancient Near Eastern Studies

2. Conditions for the RES reading of the SC of telic/punctual verbs In an attempt to supplement BK’s semantic description, I will now try to formulate additional rules that are responsible for relative token frequencies of the SC and for its grammatical readings (i. e. resultative or otherwise). The assumed rules will be supported by illustrative examples. 2.1. Transitive verbs Conditions for the resultative reading of the SC are different for P and A tokens. 2.1.1. The SC P RESULTATIVE For the SC P to be resultative, it has to denote a resultant state of the former patient physically observable at the reference time. A note on the observability concept Loesov 2006 produces some Akkadian evidence that first prompted the present writer to advance this criterion. Since then I learned that in TRC observability is considered a frequent cross-linguistic condition that favours the formation of RES (to the exclusion of non-observable situations), see especially Kozinskij 1988, where this very catchword “observable” is used consistently, while in the body of TRC the main term for the feature is “specific resultative meaning,” as opposed to “general resultative meaning.” According to TRC, the criterion of reversibility is cross-linguistically no less relevant than that of observability (see especially the diagram on p. 505). Certain languages of the sample build “non-combined” resultative forms (i. e. those having no other tense-aspect meanings) only to code observable and reversible states resulting from telic events. Note how S. Jaxontov describes constraints on the formation of the resultative form in Chinese: “The resultative is formed from a limited number of verbs. These have to satisfy certain general semantic requirements. First of all, these verbs have to be either static or telic. These latter have to code a physical action with an easily observable result. The resultant state has to be reversible. For this reason, the resultative is not formed from verbs that express creation or destruction of an object” (Jaxontov 1983: 69, my translation from Russian). To illustrate, Jaxontov adduces a few common telic (or punctual) verbs that do not form the resultative, in particular those meaning ‘to melt (intr.),’ ‘to go out (about fire),’ ‘to sell,’ ‘to build,’ ‘to kill,’ ‘to come.’ Consider also my pre-theoretical reflections

S. Loesov, The Suffixing Conjugation of Akkadian …

97

on the SC of ma¶ā´um ‘to hit, to wound, to kill, to strike’: “In OB and OA, the SC of this verb is not used for those of its senses that imply destruction of the object, loss of its identity” (Loesov 2006:137), which is tantamount to saying the resultant state is reversible. Yet a preliminary analysis of the Akkadian evidence allows one to hope that we probably do not need reversibility in addition to observability. For an adequate description, it seems sufficient to pose that Akkadian considered the (irreversible) state of destructed objects to be unobservable. Vladimir Nedjalkov, the editor of TRC, authored an encyclopedic description of RES in a recent handbook of language typology, where he offered a good working definition of observability: “A specific-resultative meaning implies that an observable state of an entity allows us to deduce a particular action or process that has brought it about. For instance, if something is ‘cooked’ or ‘tied’ we can deduce that someone has cooked or tied it first” (Nedjalkov 2001:934). He distinguishes this type from the “general-resultative meaning”: “[T]he speaker describes the state of an entity through an action which he has witnessed or deduced. Thus, we can assert that a person is killed only if we know that someone has killed him, while the body may bear no signs of inflicted death. This meaning is characteristic of resultatives derived from verbs of “non-physical” actions and those that result in destruction or disappearance of the patient. Thus, Das Geld ist gestohlen, means “The money is not where it should be, and I know (or at least suppose) that someone has stolen it” (ibid). Note that in TRC and Nedjalkov (2001), “reversible” and “locational” are practically speaking varieties of “observable.”52 Consequently, within 52 According to Nedjalkov 2001:935, “If a language has resultatives of unobservable states, it also has resultatives of observable states. Within the latter type, the implicational hierarchy corresponds to the order of considerations (in the description of Nedjalkov 2001.—S. L.): 1) observable states → 2) reversible states → 3) position in space.” If we reformulate the “implicational hierarchy” in terms of relative frequencies, then the above passage claims that in the world’s languages the observable RES is more frequent than the unobservable RES (+ if in a given language there is the latter kind, this language is supposed to have the former kind as well). Further, as for observable resultatives, the “position in space” RES is better attested cross-language than the “reversible states” RES (+ if in a given language there is the “reversible states” RES, this language has the “position in space” RES as well). By way of illustration, cf. a more intuitively appealing language universal: “If the verb in the language X has the inflectional dual, it also has the inflectional plural, however the opposite is not true.” This universal also says that in the world’s lan-

98

Articles: Ancient Near Eastern Studies

the distinction “specific vs. general” the former term is roughly equivalent to “observable.” Like so many other distinctions in language, this one does not look terribly “entweder–oder,” yet its application to Akkadian will help us explain the evidence on non-resultative forms of the SC P and that on its uneven token frequency. * * * As we have seen, RES has to be contemporaneous with the reference time (by default, with the speech time). This means that RES is comparable to deictic tense. Deictic grammatical categories (those pointing to the hic et nunc of an oral utterance and to its speaker) are often re-interpreted in written communication,53 which means that more or less drastic semantic shift of a given deictic category takes place. Therefore it is safer to start the argumentation about the relevance of observability with the negative evidence supporting (indeed, prompting) this criterion: the evidence is statistically significant and its unfolding does not involve a hair-splitting semantic analysis to which we have to recur in some parts of this study. According to BK’s commonsense view expounded above, the higher is the degree of the patient’s affectedness, the higher is the chance for the SC P to be formed. This would entail that semantically basic (and demonstrably frequent) verbs with the highest degree of transitivity (in the sense of Hopper–Thompson 1980) will have the highest token frequencies of the SC P. To test this thesis, one has to compile a list of Akkadian representatives for prototypically transitive verbal notions, i. e. verbs encoding events in which volitional agents do killing, destroying, smashing, etc. of concrete, referential, definite, preferably animate patients. Then one will have to look at the SC of these verbs.54 Since ‘to kill’ is perhaps the best representative of the prototypically transitive verbs, we can have a look at the ‘kill’ verbs in Akkadian: dâkum ‘to kill’; nêrum ‘to strike, kill’; šagāšum ‘to kill, slaughter’; šumūtum ‘put (s. o.) to death’ (the glosses follow CDA). The basic Akkadian verb for ‘to kill’ is dâkum. For its SC, three tokens were found in the core OB (all of them in AbB), and about ten in Mari. guages plural is believed to be more frequent than dual. The “implicational hierarchy” formalism is then dual → plural. 53 Cf. my suggestion that the ventive pointing to the location of the addressee is a secondary feature that arose in the milieu of written communication (Loesov 2006a), and the interpretation of the Akkadian “epistolary perfect” as the temporal projection to the addresee’s “now” (Loesov 2004). 54 A first attempt to do this is reflected in Loesov 2006:136–138.

S. Loesov, The Suffixing Conjugation of Akkadian …

99

In OA, dēk was found in two documents, VAS 26, 26:9 and Kay. 183055 (in the latter one, there are four stereotyped occurrences referring to the same real-life event). The other three killing verbs have no SC for this meaning whatsoever. Now, all the tokens of dīk/dēk in its basic meaning found in the corpus are not resultative, they occur in narrative parts of letters, most of them are modified by place adverbials. Consider the illustrative examples: (19) [a¶]ī i-na GN di-[i]k-[m]a ‘My brother was killed in GN (and I informed PN about this)’ (AbB 10, 19:7f.).56 (20) 6 ANŠE.ÚI.A ù ´ú-¶a-ru-a i-na GN de8-ku ‘six donkeys and my servants have been killed in GN’ (VAS 26, 26:4–9, OA).

Kouwenberg 2010, 7.3.3 highlights among his “marginal and secondary uses of the stative” the cases of dīk/dēk57 and mīt/mēt ‘he is dead.’ On p. 175 (fn. 46), he notes that “dâku does not seem to have an N-stem perfective (= the Pret.— S. L.) or t-perfect; there is only an N Impfv iddâk (earlier iddūak), familiar from the Old Babylonian law codes. So it is conceivable that the stative dīk serves to fill an awkward gap, since a passive perfective of the verb ‘to kill’ seems hard to dispense with.”

In other words, the P paradigm of dâkum is suppletive: dīk for PAST, iddâk for FUTURE. Now we can add to the picture the fact that dīk did not function as the G-stem resultative in OB and OA,58 yet this morphological shape is represented in the corpus by some fifteen tokens (very unevenly 55

It was published in Hecker 1995:150f. The relevant words are badly damaged, but da-i-ka-an a-¶i-ia ‘the murderer of my brother’ in line 10 fully justifies the restoration. 57 BK was at the time not yet aware that they are not attested with the “mainstream” resultative reading at all: in 7.3.2, he explains the basic resultative meaning of the SC P with the help of the following made-up example: šarru nēša idūk ‘the king killed the lion’ ⇒ nēšu dīk ‘the lion is dead as a result of having been killed.’ 58 The reason has to be a semantic one. As I have just suggested, for Akkadian, the state of destructed objects was probably unobservable. Be it as it may, in typological literature ‘is killed’ is a stock example of a problematic and difficultly formed resultative form. Nedjalkov (2001:934) hints that ‘is killed’ is usually implicational knowledge, i. e. it has to do with evidentiality: “We can assert that a person is killed only if we know that someone has killed him, while the body may bear no signs of inflicted death” (italics added). In other words, for Nedjalkov ‘is killed’ belongs to the “general-resultative” class. 56

100

Articles: Ancient Near Eastern Studies

distributed). It is therefore likely that the prohibition to use dīk as RES was not pristine in Akkadian,59 it is not pre- or proto-Akkadian but rather appeared in the course of the history of this language. After the prohibition had arisen, dīk was reemployed as the past passive form, and it probably ousted the N-stem Pret. *iddīk from this slot.60 I will now reformulate my thesis as follows: The availability and the productiveness of the resultative SC P do not depend on the prototypical transitivity, but rather on the observability of the trivial resultant state of the former patient.

To introduce more negative evidence corroborative of this thesis, we will look at the SC of frequent “High transitivity verbs with a D-stem” listed in Kouwenberg 1997:104.61 The first part of the list is headed “Verbs which entail a partial or total destruction of the object.” In what follows, I will briefly discuss the most frequent items of this list. abātum ‘to destroy, ruin.’ Among the inflectional forms of the G-stem, CAD A1 41a mentions the SC shape abit but I have found no examples of it in the entry. It does not seem to be attested in the whole of the corpus.62 batāqum ‘to cut off, pierce, break.’ In the core OB, I have found one example of the SC. The context is damaged, but the meaning of this P token is clearly figurative rather than literally destructive: (21) ul a-na-ku-ú-ma ba-at-qá-ku ina a¶¶īya ša warki šarrim illikū-nim naziq mannum kīma yâti ‘Bin ich nicht “abgeschnitten”? Wer von meinen Kameraden, die mit dem Könige hergekommen sind, ist betrübt wie ich?’ (AbB 5, 273:2′–7′).63

59 Otherwise dīk would have been attested more marginally or not attested at all, as is the case with the SC of so many frequent telic verbs. 60 But notice that the available evidence (Kouwenberg 2010, 16.5.3.5) suggests that the N-stem Pret. of hollow roots may have never existed in Akkadian, for some yet unknown reason. So in case of dâkum the morphological slot for PAST PASSIVE may have been empty before dīk filled it, though this of course sounds strange. 61 The verb dâkum is not part of this list because for some reason it does not have the D-stem. 62 IA informs me that Durand’s translation of a-bi-it in ARM 26/1, 18:40 as ‘elle (= la troupe.—S. L.) était anéantie’ is wrong, the form says ‘I stayed overnight.’

S. Loesov, The Suffixing Conjugation of Akkadian …

101

In OA, batiq is common in secondary business-related senses, cf. glosses and examples in AHw. 114. Part of the examples may be semantically deadjectival: rather than being derived from BK’s verbal paradigm, they are predicative-state forms of the VA batqum, cf. CAD B 166b–167, Veenhof 1972:404ff. and the references for batāqum and batiq in Veenhof ’s Index (p. 468). Note in particular the following considerations of Veenhof: “[W]e cannot fully understand the meaning of the verb (batāqum.—S. L.) without taking into account the permansive batiq and the accompanying verbal adjective batqum, which have acquired a meaning of their own. The permansive is used in connection with calculations, counts and checks, and signifies that something is missing” (p. 404f., bold type added). In OA, batiq is also known in more or less non-technical readings, which are also not destructive, as in the following example: (22) iš-tù a-wu-tum ba-at-qá-at-[ni] KÙ.BABBAR PN né-ri-iš um-[ma] šu-ut-ma ‘After the matter had been settled, we demanded the silver from PN. This is what he said …’ (TCL 4, 20:14ff.).

¶epûm ‘to break, demolish.’64 I know of only one “destructive” example in OB, it obtains in a court record: (23) PN šāpir GN u dayyān GN ikšudū awâtī-šunu īmurū-ma ermum ša ¢uppi ¶i-pí-ma ¢uppa-ša išrumū-ma ana pī ¢uppī-ša labiri X É ukinnū-ši ‘They (= the parties to the conflict)65 approached PN the governor of Sippar and the city judge; (the latter two) investigated the facts of their case. The envelope of the (plaintiff ’s) tablet happened to be broken, so they (= the judiciary) broke open her (the sued party’s) tablet, and confirmed her X house lot according to her older tablet’ (RA 9, 22:19–25).

In Loesov (2006:136, fn. 9), I suggested that in this exceptional example “the broken envelope is physically present at the moment of observation,” and this still seems to be true: the officials see that the envelope is broken, and this is legally relevant; most probably this means that the evidence of the respective document cannot be accepted, therefore they prefer to base

63

Kraus comments on batqāku: ‘Mangels Kontext konkrete Bedeutung unerfindlich.’ The sign string ba-ti-iq in ARM 5, 67:35 is unclear (cf. LAPO 17, 672; AHw. 114b; CAD B 167a). This token is not destructive on any of the available interpretations. 64 This verb was also on the list of destructive verbs in Loesov 2006:136 (No. 5). 65 The plaintiff is male, the sued party is female.

102

Articles: Ancient Near Eastern Studies

their decision on the sued party’s document, which is properly sealed. Thus, the syntactic relationships of ermum ša ¢uppi ¶epi-ma ¢uppa-ša išrumū only superficially resemble those of alpū PN1 ¶alqū-ma ina qāti PN2 i´batūšunūti (ex.1).66 The ¶epi is a rare example of the genuinely resultative SC of a destructive verb. This reading arose due to a strong contextual pressure that overrode the general rule according to which Akkadian counts the SC of destructive verbs to those unobservable. As to the rest of the SC tokens found in OB, they are all67 related to documents and say that a tablet ‘is invalid,’ i. e. the meaning of the SC is always figurative.68 It is no chance that the picture is essentially similar to that of batāqum: (1) both verbs normally use the SC in derived (non-destructive) senses; (2) the SC of both verbs tends to appear in purely stative (i. e. nonresultative, descriptive, qualitative) readings. The only serious difference is as follows: batāqum has a productive VA batqum at its side (especially for the meaning ‘deficient, missing, in short supply,’ CAD B 166f.), while ¶epûm ‘broken’ is rare and, according to CAD Ú 170b, not attested in the corpus.69 nakāsum ‘to cut off, fell.’ In OB, I have found three examples of the SC P in letters, and three more in divination protases (YOS 10, 52 ii 34, dupl. 51 ii 35; 42 i 35; and YOS 10, 31 to be cited presently).70 The divination examples ultimately presuppose real-time observations of animal entrails, therefore the described states of morphological features are by definition observable, e. g.: (24) šumma martum ina qablīša na-ak-sà-at ‘If the gall bladder is cut in the middle’ (YOS 10, 31 v 33; xii 16).

66

The “virtual subordination” is of course also palpable (‘since the envelope happened to be broken’), but observability remains a necessary feature of the situation with whatever translation. 67 Save one example in an epic text, OB Atra-Hasis: [¶]i-pí-i-ma libbašu ‘his heart was broken’ (Lambert–Millard 1999:92, 47). The complete list of examples will be presented and analysed in a sequel to this paper. 68 In OA, this idea is rendered by another destructive verb, duākum ‘to kill,’ see CAD D 41a for prefixing forms and TCL 21, 264:11 for the SC dì-i-ik ‘it (the tablet) has been “killed”.’ 69 The only possible example comes from a literary text, I found it in BK’s OA verb list. Here is the entry: VA sf. gen. (kīma karpitim) ¶a--e-tim (RIMA 1, 21:42). 70 A relevant dictionary entry has been found: lú.šà.PA.TAG = ša li-ib-ba-šu naak-su (OB Lu B V 54, text CAD N1 172a). The meaning is probably figurative, ‘the one with broken heart.’

S. Loesov, The Suffixing Conjugation of Akkadian …

103

Technically, nakis tokens in still life protases are “derived statives,”71 and this perfectly accords with the throughout metaphorical morphosyntax of the genre: observable “primitive” (= not created) states are depicted in terms of dynamic Becoming (iprus, iptaras, and iparras) or resultative Having-Become (paris). In other words, if the gall bladder is said to be naksat, this does not imply it was supposed to have been cut in the first place. All three occurrences of the SC in letters are observable: (25) ana GIŠ.TIR.ÚI.A-ku-nu lā tēgiā … urram ina amārī-ya ana 1 GIŠ sikiltim ša na-ak-sa-at awīlam bēl pī¶atim ul uballa¢ ‘Do not be careless about your forests! Tomorrow, during my inspection (of the woods), for every tree that is illegally felled, I will not let the responsible person stay alive!’ (AbB 4, 111:12ff., Hammurapi’s words addressed to foresters).

Clearly, Hammurapi refers to something (like a felled trunk) he may happen to observe during his inspection (= amārum, lit. ‘seeing’). (26) inanna GIŠ.MA.NU na-ki-is-ma šakin ‘Now the MA.NU-wood is felled and put together (get me boats so that I take it along)’ (AbB 12, 194:8f.).

In real life, the wood was first felled and then piled up, but the writer’s words are no narrative, in spite of the sequencing -ma. I. e., nakis is not used here instead of innakis.72 This is proven by the sentence-initial inanna ‘now’ which qualifies each of the two predicates. The writer’s message is that a fresh supply of wood is now ready for transportation, and he renders this message by depicting both facts (being-felled and beingpiled) as resultative states. In Akkadian, this is easily done because the states in question are observable, i. e. the respective SC P tokens can be freely formed. (27) [mā]r irtim ša šaddaqdim waldu [ina] me¶ret sakkanim labirim [ša] elēnu piātim šapiltim[ina ¢]e4-e¶ nārim nadi-ma ´e¶rum šū [ina q]ablī-šu na-ki-

71 Since at least Nedjalkov (1988), DERIVED STATIVE is a verb form that denotes a state that needs not to result from some event, though in a given language the same morphological shape can have the resultative sense. This is how DERIVED STATIVE is used in the present paper, although of course the term has other senses in linguistic literature. 72 The Pret. N of this root is attested in OB with the very meaning (‘to be felled’) that would be needed here were this sequence a narrative, see CAD N1 179.

104

Articles: Ancient Near Eastern Studies ìs-ma [išt]u irtī-šu adi qaqqadī-šu šakin [bal]u qaqqadī-šu adi šepī-šu [laššu] lū zikar [lū s]inniš mannu[um] [lū id]e ištu qablītī-šu [adi ša]pliš ul ibašši73 ‘(Immediately after the letter-head) A baby that was born last year was [found] lying in front of the Old Palace (which [is situated] above the lower city-limit), near the river. The kid was cut across at its middle: the body from the breast up to the head was there, but it had neither the head nor (the lower part including) the feet. Whether it used to be male or female, there is no knowing. There is nothing preserved below the middle part [of the body]’ (ARM 6, 43:5–15).

This is a letter of Bahdi-Lim, the governor of the Mari district during the reign of Zimri-Lim, so it belongs to the “Amorite group” in the sense of IA. Certain turns of phrase in this text may indeed sound clumsy, but our nakis is in perfect accord with the core OB usage: Bahdi-Lim describes the remains of the baby from the vantage point of a synchronic observer who stands on the river bank and examines the mutilated body. Thus, all the available examples of nakis in the corpus can be considered both resultative (given the artificial nature of still-life omens protases) and observable. nasā¶um ‘to tear out, remove.’ Akkadian counted the meaning ‘it is torn’ with the observable ones, because it implies literal seeing.74 The following three examples for the basic meaning, the only ones I have found, are not without blemish (the first is marred by negation, the second is probably a derived stative, and the third one comes from a lexical list), yet I believe that taken together they prove the point: (28) šumma … bītum lā pališ sippu lā ¶ališ aptum lā na-as-¶a-at ‘If … the house has not been broken into, the threshold has not been scraped off, the window has not been torn out’ (LE § 36).

Goetze 1956:96, Yaron 1988:65, and Roth 1995:64 interpret these clauses as semantically circumstantial: ‘the house not having been broken into,’ ‘even without evidence that the house has been broken into.’ It is understood that the meaning of these SC forms depends on the literal

73

The text reproduces all the collations and restorations of LAPO 18, 236f. See this edition for philological details. The interpretation partly follows the Russian translation of IA. 74 Cf. “The last time we saw you you looked so much older // Your famous blue raincoat was torn at the shoulder” (Leonard Cohen).

S. Loesov, The Suffixing Conjugation of Akkadian …

105

“moment of observation” at which the competent authority visually establishes that the house is intact, while the deposit is missing. (29) šumma izbum ullânum-ma imittašu ša imitti na-as-¶a-at ‘If the malformed animal’s right shoulder is torn off “from the start” [= from the moment of its premature birth?]’ (YOS 10, 56 i 10f.). (30) ša inā-šu nas¶ā ‘whose eyes have been torn out’ (OB Lu B IV 48, text and tr. CAD Ú 60a).

Note a metaphoric example in OA, still close to the basic meaning: (31) mīnum annītum ša ¢uppam 2 ubān šulum-kunu lā tušebbalā-nin-ni a´´ēr ana ešrī-šu ¶alqāku-ni-ma u libbī na-as-¶u u attunu šulum-kunu lā tušebbalānin-ni ‘What is this that you (pl.) never send me a letter of (at least) two inches (length) about how you are doing? Besides that I am ten times at a loss and my heart is torn, and (why is this that) you never write me about how you are doing?’ (ICK 1, 17b:4–12).

In the corpus (mostly in OB), there are dozens of P nasi¶ tokens with non-literal meanings. All of them are technical terms of divination (nasi¶ vs. kīn to express a morphological peculiarity of animal entrails, CAD N2 6a), of law and administration (‘has been deported’; ‘has been removed [from a list]’; ‘[the claim] has been rejected’; [a certain amount of something] ‘has been drawn/deducted’ [as payment]); [a person] ‘has been/has to be disinherited’ (e. g. ICK 1, 12b:36, OA).75 According to my prediction, such P forms have to be non-resultative, which will be (at least sometimes) corroborated by the above syntactic criteria. šebērum ‘to break.’ For the basic sense of the root, the SC P has been found only in ARM 13, 40:29f., represented by two tokens in contiguous lines. Though the exponents of subjects are damaged in both of them, the meaning of [š]ebi-ir (l. 29) and še-eb-ru (l. 30) is doubtless observable and resultative: ‘[a beam?] is broken, [several beams?] are broken.’76 Note also the following example that plays with the literal meaning: 75 Most of these SC forms have at their side prefixing forms with the same lexical meanings. 76 The observable meaning of these verb forms follows from the contents of the passage as well. The letter was written shortly after Zimri-Lim’s takeover of power (p. c. of IA, and see Arkhipov forthcoming = ARMT 32, s. v. nūbalum), its author is Yasim-Sumu (at the time, a high official of ZL), who has inspected the royal palace and is now reporting to the king what is to be done for its restoration.

106

Articles: Ancient Near Eastern Studies

(32) eqlētim u kirâm ša i-[di-nu] ekim-šu-ma ana sanāqim appašu lu-ú še-bi-ir ‘Take from him the fields and the garden that he has [given out], may his nose be broken should he transfer (them again)’ (AbB 14, 31:28ff.).

Veenhof explains on p. 205 of the Edition, “I take appašu lū šebir (31: 30), ‘may his nose be broken,’ not literally , but, also in view of the stative, as an idiom expressing a threat which has to prevent an action.”77 In OB letters šebir P is productive in an agricultural meaning ‘(the field) has been broken up’ (and/or ‘harrowed for the second time’?—cf. CAD Š2 248), see examples and glosses in CAD Š1 114; Š2 249, in particular AbB 8, 130:4′; 9, 151:17). This “less destructive” meaning may be observable as well, in spite of its technical nature. Summing up: resultative tokens of the SC P are not attested (or only exceptionally attested) for the basic meanings of the following frequent high-transitivity verbs, selected more or less randomly for the above overview: dâkum ‘to kill,’ nêrum ‘to strike, kill,’ šagāšum ‘to kill, slaughter,’ abātum ‘to destroy, ruin,’ batāqum ‘to cut off, pierce,’ ¶epûm ‘to break, demolish,’ ekēmum ‘to take away (by force),’ naqārum ‘to tear down, destroy,’ ma¶ā´um ‘to hit, smash’ (for evidence on some of these verbs, cf. Loesov 2006:136f.). This fact has been explained by two suggestions: (1) in Akkadian, the SC P has to be observable in order to be resultative; (2) Akkadian counted the state of destructed entities to those unobservable. Note that there is one thing that has not been explained: why then Akkadian possesses the SC shape of certain destructive verbs at all? So far,78 it has been only vaguely surmised that some of these SC tokens may have served as pluperfect/background forms or fillers of paradigm gaps. My analysis of individual verbs shows that observability cross-cuts the distinction “low” vs. “high” semantic transitivity of two-place verbs: observability represents a different semantic parameter. This follows from two types of evidence:

77

BK tentatively suggests me in a p. c. that lu-ú-še-bi-ir could also be rendered as “I shall break.” He adduces EA 356:5 (“Adapa and the South Wind”), where an identically spelled form is traditionally understood as ‘I want to break.’ “Or is this also a stative?” he asks. The OB spelling leaves of course little chance for the former verb form to be a D-stem precative. Yet, frankly, I do not know what Veenhof means by his “also in view of the stative.” 78 See the above discussion of syntactic constraints on RES, especially exx. (1), (6)–(8).

S. Loesov, The Suffixing Conjugation of Akkadian …

107

(1) On the one hand, as we have just seen, the resultative SC P of verbs with the highest transitivity is not productive, because it tends to be non-observable. (2) On the other hand, the observability criterion explains why e. g. the SC P tokens of leqûm ‘to take’ and ma¶ārum ‘to accept, receive’ are very rare (as we will see, the P ma¶ir is virtually unattested).79 The only reliable example of the P leqi in the whole of OB80 happens to be resultative, and this is due to the extraordinary pressure of the context: (33)

É.Ì.DUB ina nēreb KÁ i-[b]i-ru pate¶-ma še-um le-qí amtam u´´ir-ma umma amtum-ma 0.1 ŠE TA.ÀM šinīšu 0.2 ŠE il-qí ‘The granary at the entrance of the Ibiru-gate is broken open and the barley is taken (out). I interrogated81 a slave-girl. She told me: He (= the one who broke into the granary) took out twice 60 qa barley, (i. e.) 120 qa barley’ (AbB 6, 219:14–19).

Within the letter, this is a piece of reported speech, cited here from the beginning (the slave-girl’s words are embedded into it). In a description of a forced storehouse, pate¶ looks appropriate as an observable resultative P form. In the clause našpakum ina nēreb GN pate¶, the prepositional phrase ‘ina nēreb GN’ is an analytical attribute of našpakum rather than a place adverbial modifying the verb form, so ‘ina nēreb GN’ does not disqualify pate¶ as RES. At the moment of observation, vividly reproduced by the writer of the reported message, the granary is demonstrably broken open and empty, i. e. še!um leqi. We have to put the argumentation the other way round, as a prediction: If a telic transitive verb happens to build a P token of the SC with an unobservable meaning, this verb form will not have the resultative reading.82 Indeed, the rest of the P tokens of leqi in the corpus (‘is taken,’ etc., about five examples) live up to the prediction: 1) the verb form’s meaning is unobservable (usually, the “having-been-taken” leaves no physical 79

This goes against the well-known assumption that the SC P of “taking-holdof ” verbs is freely formed (e. g. Huehnergard 1997:395). Due to the wrong premises, the near-absence of the P leqi and ma¶ir has not been noticed. The assumption about their existence can lead to mistakes in text interpretation. 80 See Loesov 2010:769. 81 Cf. AHw. 1439b, CAD Š3 39a. 82 The inescapable questions are then why the language builds such verb forms at all and what they are needed for. It is possible to address both of them only after all the evidence has been passed in review. I suspect that part of the “Why” will be diachronic.

108

Articles: Ancient Near Eastern Studies

impression); 2) syntactic constraints exclude the resultative interpretation. The following sentence illustrates:83 (34) 6 TÚG ku-ta-ni i-na sà-e-tim a-šu-mì PN ù PN1 lá-qí-ú ‘6 kutānu-textiles have been taken from the sa!utum in the name of PN and PN1’ (OAA 1, 111:13ff., tr. Larsen).

In the meantime, discussing the negative evidence pro the observability criterion, we have seen quite a few good examples of the SC P RES formed from relatively highly transitive verbs which however are not in the highest position on the “transitivity cline,”84 such as nakis ‘is cut/felled,’ ¶epi ‘is broken open/damaged,’ nasi¶ ‘is torn off,’ šebir ‘is broken.’ I hope these examples make it clear what is meant here by “observability” of the SC P. In addition, consider SC P tokens of two locative verbs, šakānum (the basic meaning ‘to place’), and nadûm (the basic meaning ‘to throw, to lay down’). Such verbs have the obligatory valency for the participant LOCATION. As noted above, place adverbials do not stand in the way of the resultative reading for the SC of locative verbs:85 (35) (a) 17 ma-na AN.NA i-mu-ta-tí-im ša-ki-in ‘(PN told me:) 17 minas of tin have been placed in the half-pack (take this for yourself)’ (OAA 1, 102:15f.).86 (b) a-ma-lá té-er-ti-kà ANŠE-ru-kà a-na-áb-ri-tim na-du ša-am-kà-kà ša-lim ‘According to your order, your donkeys are put out to pasture; your personnel are doing well’ (BIN 4, 31:43–46).87 (c) KÙ.BABBAR iš-tù u4-mì-im a-mì-im kà-ni-ik-ma i-na Kà-ni-ìš na-dí ‘Since that day, the silver has been sealed and deposited in Kaniš’ (CCT 4, 14a:22ff.).

In (35c), kanik and nadi are P resultative forms,88 both are observable. Note that kanākum ‘to seal’ is a low-transitivity event eagerly terminating in an observable state of the object. 83

See also an analysis of leqi/laqi in Loesov 2010:768–772. The term goes back to Hopper–Thompson 1980. 85 I would like to draw the reader’s attention once more to the relative crosslinguistic frequency of RES depending on lexical semantics. An attempt to calculate “the universals of resultative formation” was made by I. Kozinskij who worked with the data collected in TRC (see Kozinskij 1988:505). According to his calculus, reversible “bivalent locational” resultatives are the most common subtype within the “observable” group. See also fn. 51 above. 86 Compare a very similar verb phrase with the head in the Imv.: kaspam … ina muttātim ištêt šuknā ‘Put (Imv. pl.) the silver into one half-load’ (TCL 4, 16:28). 87 See also CAD N1 79b; N2 148a (discussion section). 84

S. Loesov, The Suffixing Conjugation of Akkadian …

109

2.1.2. The SC A RESULTATIVE Logically, the next point of the agenda is “What is the prerequisite for the SC A of transitive verbs to be resultative?” We do not expect the lexical feature “observability” to play a role here, since an observable resultative state is usually the fate of the former patient.89 From the descriptions of the “transitive resultatives”90 in TRC it follows that in this variety “[T]he result of the action is relevant for the agent rather than for the patient” (TRC 516, I. Kozinskij).91

This is actually the commonsense lexical constraint on the formation of the SC A RES in Akkadian: the telic event should be able to lead to a resultant state affecting the former agent. Consonantly with the “to marry” example in the last footnote, the SC A RES in the corpus typically (but not exclusively) codes what can be loosely called “a socially relevant status.” As we already know, this is exactly the opinion of BK, most recently confirmed in Kouwenberg 2010 and cited above: the SC A is formed mainly from “transitive verbs with a low degree of transitivity” and is “eminently suitable to describe the legal status of a person.” Here as elsewhere, I disagree with BK on a different ground: for BK, all PARIS forms of telic verbs (including two-place ones) are by default resultative, while I believe that this should be (dis)proven piece by piece, or better still: the grammatical meaning of PARIS is an open question. At this turn of our study, we can only assert that the SC has several grammatical readings 88

Note that according to TRC 56, “reversible resultatives” (unlike irreversible ones) are easily compatible with SINCE-prepositional phrases indicating the time span between the coming about of the state and the reference time. 89 M. Haspelmath (1990:40) says that “only the patient can be characterized by means of the result of an action, because in general only the patient is affected by an action,” and this stands to reason. 90 I. e., morphological resultatives that have both a subject (coreferential with the agent of the previous action) and a direct object (the former patient). In grammatical studies, finite verb forms with both subject and direct object are sometimes called “two-place predicates,” to avoid the semantically more pregnant term “transitive.” E. g., since RES codes a kind of state, both Kozinskij and BK are not quite happy with the dynamic term “transitive” as applied to two-place resultative sentences, and use it with an explicit caveat. 91 Cf. a very similar formulation in Nedjalkov (2001:928): “The previous action affects the agent rather than the object.” A good example is ‘to marry,’ adduced by I. Kozinskij, “If John marries he not only makes some lucky girl his wife but also makes himself her husband” (TRC 520).

110

Articles: Ancient Near Eastern Studies

(i. e., more than just A RES and P RES), they are to a certain degree predictable and depend on lexical semantics and syntactic environment. The criterion “relevance for the former agent” is a rigid one, to the extent that in a living language two-place A RES complying with this criterion may exist as a closed-list class. TRC does mention languages in which RES is in fact formed for relatively few verbs, due to lexical semantics restrictions. Yet, if in a given language the morphological category in question is not completely “frozen” (e. g., in the way of certain conjugation types of French or Italian), it will be able to produce time and again new and semantically non-trivial tokens, provided there is a strong pragmatic need. My very tentative Akkadian examples of the non-trivial SC A RES are as follows: (36) (a) [a]nāku-ma kabsāk-šunūti ‘It is I who is going to trample upon them’ (ARM 26/1, 195:16).

This is a prophesy uttered on behalf of a certain goddess Hišamitum. The trivial results of trampling can affect only the patient, yet in this prophesy the goddess asserts herself as one who is stronger than ZimriLim’s enemies ([LÚ.MEŠ b]e-el a-wa-ti-ka), therefore this SC token is probably meant as something like the assumed BH Perfectum Propheticum and the property of the agent emanating from its divine status. Or is this very unusual sentence somehow due to the “Amorite” influence (as IA suggests to me)? (b) [m]īnû awātum-ma PNnom. [a]rdīacc. bītīya … rakis ‘How is it that PN keeps the slaves of my household bound (by contract)?’ (AbB 12, 72:13–16, late OB).

The writer focuses on the condition of PN, not on that of household servants, hence the unusual semantic configuration of the SC A. Did the two-place SC A RES constitute a closed list in the spoken Akkadian vernaculars of ca. 2000–1700?92 Though the SC A has been felt as something “special” (sometimes even disturbing) since at least Ungnad 1918, grammatical literature has not provided us with anything like a syntactic and semantic analysis of two-place A forms of the SC.93 Today, the only well-known facts about the SC A are the following ones: 92

Note that the logical stress in this question is placed on RES, I am not asking just about any kind of the SC A. 93 D. Cohen (1984:257f.) offers a list of roughly one hundred verbs for which the SC A is attested (as it seems, in the whole of the recorded Akkadian, but unfortunately he never defines his corpus explicitly), including verbs that do not

S. Loesov, The Suffixing Conjugation of Akkadian …

111

1. It is rare relative to the SC P (Rowton 1962:235; Cohen 1984:257). 2. It is usually formed for low-transitivity verbs (Cohen 1984:260; Kouwenberg 2010 7.3.2.). 3. It “describes the state of the subject after the completion of the event” (Kouwenberg 2010, 7.3.2, following Kouwenberg 2000). Sure enough, in this case for BK “subject” ≈ “agent,” i. e. a personal, willing being. It can hardly be anything else, e. g. an instrument or an inanimate cause, as in ‘The cyclone has damaged sugarcane crops in the Mackay and Proserpine region.’ 4. The SC A is common (Rowton 1962:235; Cohen 1984:258) only for three nearly-synonymous taking verbs ´abātum ‘to seize,’ leqûm ‘to take,’ ma¶ārum ‘to receive’;94 the fourth one in this set of frequent tokens is našûm ‘to lift, carry’ (Rowton 1962:235) that renders a nuance of taking as well.

I will now produce a few pieces of lexically consistent (and therefore not fortuitous) evidence showing that (contra BK) the SC A cannot be considered resultative by default. In other words, it does not necessarily describe “the state of the subject after the completion of the event,” it can render this service only under certain (and very demanding) conditions. First, let us have a look at a syntactic minimal pair: (37) (a) PN annikīam sà-ni-iq-ni-a-ti-ma alākam ana ´ēri-ka ulā nile!!e ‘PN is checking? us here, so we cannot come to you’ (AbB 9, 88:6– 11, tr. of the Ed.). (b) anāku annikīam sa-an-qa-ku-ma ul alli[kam] (Goetze 1958, No. 46: 4ff.).95

In (37a) saniq is A, in (37b) it is P. Given the rest of the relevant data,96 one has to acknowledge that (37b) is a passive counterpart of (37a), the take direct object in the prefixing tenses for their basic meanings, e. g. alākum ‘to go.’ Kouwenberg 2010:172, fn. 36 notes that this list “could be augmented substantially.” The list, if sifted through, will constitute a good starting point for a synchronic syntactic study of the SC A. 94 In the literature, the value of this piece of evidence has been overestimated out of any proportion. In particular, to my knowledge no attention has been paid to the fact that the SC A of these three verbs is only marginally represented or completely unknown in the OA corpus. This most probably means that ´abtāku ‘I have the usufruct (of a piece of real estate)’ and similar usages of the SC A of the other two taking verbs are OB “bureaucratic” expressions, probably alien to the contemporary spoken OB. 95 See Loesov 2010:779ff. for a detailed analysis of saniq, and already Loesov 2005:124f. for some claims regarding this form. In my previous discussions of the lexical meaning of saniq in the letters from Ešnunna I overlooked the remarks in Kraus 1987:43.

112

Articles: Ancient Near Eastern Studies

facts of life behind both sentences being more or less the same. In spite of the SC A form in the sentence ‘PN annikīam saniq-niāti,’ it is unlikely that (37a) describes the state of PN after he had completed the sanāqum. Hence, logical reasoning and the adjunct annikīam exclude the resultative reading for both sentences. In Loesov 2005:124–133, this example was used as a piece of proof to show that the SC A of telic verbs can code ongoing present-time events. BK rejected this idea in personal communications and in Kouwenberg 2010:91, fn. 8, where he says my examples “are all in some respect problematic and are certainly insufficient to prove such a far-going claim.” I agree wholeheartedly with the latter statement, and I partly agree with the former one. In particular, an important shortcoming of (37) is that the lexical meaning of sanākum is opaque (see Loesov 2010:780 for a discussion). Yet what are we to make out of (37) morphosyntactically (= in terms of the grammatical meaning of saniq), with whatever root meaning in this context? It is clear that (37a) is not resultative because it is relevant to the patient rather than to the agent, and because of annikīam. The example (37b) is not resultative either if it a passive transform of (37a), which is likely. Now, the t-Perf. of sanāqum is represented in AbB 1–14 with some 35 tokens, almost always as a stereotyped futurum exactum for the lexical meaning ‘to arrive, to reach’ in the temporal clauses of the kīma/inūma issanq-akkum kind, lit. ‘as soon as he will have reached you’ (the verb phrase is inflected for subject and personal goal).97 If saniq-niāti in (37a) is semantically PERFECT, why then not *issaniq-niāti? Since we do not really understand how the Akkadian verb functions, to exclude the present-time meaning of this sentence would seem a bit dogmatic. Consider another relevant set of examples, syntactically similar to the above one: (38) (a) É.GAL is-ra-an-ni-ma ana ma¶rī-ka alākam ul ele!!i ‘The Palace is exacting (payment from) me [or ‘keeps me under pressure’?], so I am not able to come to you’ (AbB 10, 73:10f.). (b) ilkum is-ra-an-ni-ma naparkâm ul ele!!i ‘The service keeps me busy, so I cannot stop (working)’ (AbB 14, 43:8f.).

96 For stereotyped P sentences of (37b) kind in OB letters from the kingdom of Ešnunna, see Loesov 2010:780f. 97 The data are from http://www.klinopis.cz and http://klinopis.cz/nobtc.

S. Loesov, The Suffixing Conjugation of Akkadian …

113

(c) kīma tīdû ana amtim ¶i-ša-a-am ēzibpret.-ma adān kaspim šaqālim iktašdperf. anni-ma tamkārum is-ra-an-ni ‘As you know, I drew uppret. a debt-note (obliging me to pay) for the slave-girl, and (now) my term to pay the silver has comeperf. and the creditor is exactingSC (payment from) me’ (AbB 2, 94:9). (d) kīma is-ri-ku ul tīde ‘Don’t you know that I am being pressed (for payment?)’ (AbB 2, 96:36).

The argumentation about (37) applies here as well, except that according to the dictionaries the root meaning in this case is more transparent.98 My claim in Loesov 2005 about the progressive meaning of esranni in (38 a–c) may be not 100% proven, but one has to insist that esranni is not RES, if RES A (with typological literature and BK) implies a change in the state of the (former) agent.99 On the other hand, the evidence collected for this study makes it necessary to suggest (see already some of the above examples and analyses) that the non-RES SC hardly ever appears randomly (i. e., with no reason at all) in the legitimate slots of the Pret. or the t-Perf.100 Faute de mieux, for (37a) and (38a, b) the progressive reading is a tolerable way out. Needless to say, it fits the respective contexts even better than a past-time interpretation. BK informs me regarding my interpretation of exx. (37) and (38): “I would translate 37b ‘I am kept busy/tied down/detained here’ and 37a as ‘PN has tied down me here.’ For esērum, I doubt very much whether ‘press for payment’ is correct, it rather seems to be more or less synonymous to sanāqum, cf. also ussurum ‘to imprison, to take captive,’ so in 38a and 38c ‘the palace/the creditor has detained me’ (like sikkam ka!!ulum in OA?), 38b ‘the work has kept me busy/tied down’ (in the CAD, these instances should have been placed under esērum B). In terms of their stative, these verbs behave like katāmum, etc.: ‘to have covered’ > ‘to keep covered,’ especially 37a. The editions and the dictionaries are unreliable in this respect, and the whole semantics of these verbs has to be reviewed. Is 98

The gloss for the CAD entry esēru A is ‘to press for payment due, to collect, to put a person under pressure.’ This set of examples was mentioned in Loesov 2005:125f. 99 Cross-language, there seems to be nothing unusual about resultative-progressive polysemy of the same marker and even about a diachronic shift from RES to PROGRESSIVE, see e. g. Nedjalkov 2001:938, with references to Chinese, Japanese, Uzbek, Balkar, Mongolian, and Iroquoian. 100 I am extremely grateful to IA who was the first one to come up with the idea that one has to look for rules in the non-resultative use of the SC.

114

Articles: Ancient Near Eastern Studies it possible to extend the definition of the active status so that it may also picture the state of the patient (rather than agent) after the event if the meaning of the verb favours this? This is what comes to my mind for several of these examples.”

BK is most probably right that the lexical semantics of these verbs needs revision. In this study of grammatical meanings, it is best to avoid examples that are lexically (or otherwise philologically) problematic. Yet, by way of exception, I decided to keep this important material and to leave my above interpretations unchanged, in spite of BK’s critical observations. This is because in my opinion BK’s remarks do confirm the only claim that is vital at this point: none of the examples in (37)–(38) is RES. My argumentation runs as follows. We have got four two-place A sentences with paris predicates (37a, 38a–c). It is agreed that these four cannot picture RES states of the respective grammatical subjects. BK’s question also implies that the informational focus in each of the four utterances is the respective direct object.101 In dynamic transitive sentences with the default word order, the direct object naturally belongs to the focal part. For the genuine SC A RES, this universal rule cannot work, because such resultative sentences picture the resultant state of the former agent; in anticipation of more examples below, consider my interpretation of (36a), anāku-ma kabsāk-šunūti, where anāku-ma secures the focal role of the subject. In an attempt to save the resultative meaning of (37a) and (38a–c), BK wonders if these A sentences may be allowed to picture the state of the direct object. Yet such a strong claim is not needed: since BK grants that in numerous other cases the SC A is in fact dynamic (= non-RES), we can safely admit this for tamkārum esr-anni etc. as well. What we get in exchange is the trivial information structure of the dynamic transitive sentence, and this is preferable to the typologically exotic scenario of a syntactically A resultative sentence depicting the state of the patient. Summing up: the four above A sentences are not resultative, their tense value can be either past or present, depending on our judgment about the lexical semantics and the contexts. Let us now follow BK’s comparison of ‘PN annikīam saniq-niāti’ and ‘ekallum esr-anni’ to the OA ‘X sikkam ša Y ukāl’ (‘X is detaining Y,’ CAD K 509; S 255)102 and to the SC of katāmum-like verbs, e. g. še!um girram parik 101

It is the speaker (‘me’) in three cases, and -niāti ‘us’ in 37a. Note that it is also the speaker that is the subject in the two P instances (37b, 38d). 102 The verb kullum/ka!!ulum ‘to hold’ has no SC and consequently is one of the best known “prefixing statives” of the traditional grammar (GAG 127b).

S. Loesov, The Suffixing Conjugation of Akkadian …

115

‘the grain is blocking the road’ (CT 52, 84rev.:10, OB letter, tr. CAD P 155b). For the sake of argument, we can render the former two sentences as ‘PN keeps me detained here’ and ‘the palace does not let me go,’ i. e. more or less with BK. All four situations are present-time and stative; the only major difference between them is that the former three are agentive, unlike the latter one.103 Now, a verb phrase irâm ‘(s)he loves (something)’ is equally stative (typically, nothing happens to the participants of this situation), and the verb râmum has no SC, as well as kullum/ka!!ulum.104 The choice of paris or iparras in these stative sentences about the Present Time is determined lexically,105 while none of them is RES, simply because none of them describes the post-factum state of the subject. The last two paragraphs call for a reminder: for the purposes of this study, we have to distinguish two senses of the word “stative.” The first one is morphological: STATIVE (and RESULTATIVE as its variety) is the etymologically primary function of an Akkadian “tense,” the SC paris. In other words, STATIVE as a possible inflectional meaning of the verb is the raison d’être of the SC paris, and this is also true of various morphological and analytical verb “tenses” in the world’s languages (see TRC). The second sense of “stative” has to do with lexical semantics and syntax rather than with morphology. Roughly, it is an Aktionsart, i. e. one of the four Vendlerian semantic types of predicates (Loesov 2005:107). Thus, “In God We Trust” is a stative predication (both the subject and object of trusting remain as they are) independently of its morphological shell in individual languages, e. g. the Simple Present in English or the SC taklānu in OB (Loesov 2005:141f.). At this point, what is important for us is that there are common verbs whose SC A is both frequent and usually (or always) non-resultative. In addition to exx. (37)–(38) which may be somewhat problematic, I can ad-

103

This difference may turn out to be grammatically relevant. Thus, we do not know whether a sentence *Enkidu bābam parik ina šēpī-šu ‘E. is blocking the doorway with his feet’ was acceptable in OB (cf. Gilg. P. vi 12). It may well be that katim ‘it covers’ (+ Acc.) and parik ‘it blocks’ (+ Acc.) can only denote a spatial position of the inanimate subject relative to a landmark coded by the Acc. 104 Consider a curious example: aššum ša kīam tašpurī umma attī-ma KU6 TUR.TUR r s ša PN i-ra-am-mu ušābil-akkum akkīma mutu-ki PN ina GN u GN2 KU6 TUR.TUR la-am-du anāku ištu pana ina GN3 GN4 GN5 u GN6 KU6 GAL a-ra-am ‘Concerning what you wrote saying: “I have sent you small fishes which PN loves”—just as your husband PN knows little fishes from Qa¢ara and Karana, I myself have from of old loved the big fish from Šubat-Enlil, Ekallatum, Mari and Babylon (OBT Tell Rimah 42:6–20).’ In this text, lamid and irâm are synonymous. 105 See Loesov 2005 and Loesov 2011.

116

Articles: Ancient Near Eastern Studies

duce the SC of kašādum ‘to reach, to come, to arrive’ in OB. Among the meanings of kašid (Loesov 2006:147), the one derived immediately from the literal motion meaning of the root (‘to reach, to come’) is always used as future with an additional reference point (“after a certain event,” “in two days,” etc.). Consider three of the six examples from AbB listed in Loesov 2006:147: (39) (a) šeam ama¶¶ar-ma ka-aš-da-ak-ki ‘I will receive the barley and (then) come to you’ (13, 87:8′f.). (b) ana U4 2.KAM ka-aš-da-ka ‘within two days I will reach you’ (12, 42:14). (c) awâtum É.GAL ka-aš-da ‘(By the time you meet the lord) the news will have reached the palace’ (3, 48:32f.).

Sentence (39c) is not resultative (the reaching of the palace does not affect the news), and so probably are the former two, since they are not really different from the latter one. The Pres. ikaššad is used in the same future environment in OA, core OB, and OB Mari (CAD K 273–274), consider an example: (40) adi ištu inanna UD.2.KAM a-ka-ša-ad-ka ‘I will meet you in two days from now’ (TCL 17, 22:9; OB, text and tr. CAD K 272b).

This evidence reminds one the pluperfect function of certain SC forms as discussed above, exx. (6)–(8): in both cases there is a reference point not identical to the speech time. Another instance is the SC A of the OA qabā!um ‘to say, speak.’ This SC token is common in OA. Consider the examples: (41) (a) šīm 4 GÚ AN.NA kīma i-Šalatuar u Wa¶šušana ša ELLATat I-dí-Ku-bi4im innaddin-u ša ana DAM.QAR šaqālam qá-bi4-a-ku-ni 3 GÚ AN.NA kunukkī-ya ana Wa¶šušana ana ´ēr ša kīma yâti ušēbil ‘(Immediately after the letter-head) The price of four talents tin as it is being sold from the Iddin-Kubum’s caravan in Šalatuar and Wahšušana: what I had promised to weigh out to the merchant, i. e. three talents tin, I sent to Wahšušana under my seals care of my representative’ (Prag I 442:2–9).

The relative clause ša ana tamkārim šaqālam qabiāku-ni does not describe the resultative state of the speaker as the one who is under the self-imposed duty to pay, semantically this is a dynamic PLUPERFECT, but the stereotyped OA usage of the SC verbal phrase šaqālam qabi (inflected for subject) is derived from RES: šaqālam qabiāku = ‘I have promised to pay,

S. Loesov, The Suffixing Conjugation of Akkadian …

117

and I am now obliged to pay,’ the event of promising affects the one who has promised. (b) ¢uppam ša kunukkīya ša annišam wabālam qá-bi-a-tí-ni ūmam lā tattabl-am ‘Today you have not brought me the tablet with my seal that you promised to bring hither’ (Michel 1991, No. 263:9ff.).

Same comment applies here, except that this is not the ubiquitous šaqālam qabi phrase and the meaning is probably more like SIMPLE PAST than PLUPERFECT. What then about examples of the genuine SC A RES? Since this study aims at reconstructing certain features of the tense-aspect in spoken Akkadian, we are not going to abuse the best-known candidates, to wit OB leqi ‘he has got,’ OB ma¶ir ‘he has (received),’ OB ´abit ‘he has (a piece of real estate in usufruct),’ OA adim ‘he has invested, he owns a share,’ OA šapik ‘he has invested,’ because they are stock phrases of administration or business, not necessarily reflecting the natural usage. As I have mentioned above, ´abit A is absent from OA, laqi and ma¶ir A are very rare in OA.106 On the other hand, *adāmum is reliably attested only in OA and only in the guise of the SC A. The following examples, taken preferably from letters, look as appropriate illustrations of what it means to be the SC A RES, though their token frequency is in most cases low: (42)

DUB.SAR-tám

wa-dí lá-am-da-ni e-pá-tá-am a-na um-mì-a-ni-a šu-bi-lam ú ma-tí-ma lá ta-aq-bi um-ma a-ta-ma 1 GÍN KÙ.BABBAR a-na me-er-i-a lu-šé-bi4-il5 ‘Sure enough, we have mastered the scribal art. Send an eppatumtextile to my teacher. And (why then) have you never decided, “Let me send a shekel silver to my son”?’ (CCT IV, 6e:4–14).107

(43) (a) 1 GEME2 ša … ištiššu šinīšu-ma waldat ana utlī-ya leqe-am ‘Bring me for sex a slave girl who has given birth once or twice’ (ABIM 20:82).

106

In response to my question, BK tentatively suggests that “OA uses the preterite instead, cf. especially lists of expenses where the writer writes down what he received or paid, and I think they always use ilqe, alqe, išqul, ašqul, etc., never a stative or a perfect. This makes it even more interesting to learn when the Old Assyrians do use the corresponding stative forms.” BK agrees that the very high frequency of the three “taking-hold-of ” forms of the SC A in OB is probably due to their being “legal jargon.” 107 CAD L 55b translates, ‘as you know, we are learning to write.’ Michel 1998: 250 has ‘Assurément, nous apprenons l’art du scribe.’ Why? By contrast, CAD T 163 translates ‘as you know, we know the scribal craft,’ which seems nearer to the mark.

118

Articles: Ancient Near Eastern Studies

In this sentence, the direct object is implicit in the adverbials. It surfaces in a protasis from CH (§ 158:28), once more in a relative clause: (b) ‘If a man, after his father’s death, has been caught in the lap of his (= the defunct father’s) wife ša mārī wa-al-da-at who has born children (= is mother).’108

In both examples (the only ones I have found), the relative clause ‘(a woman) ša … waldat’ is a way of saying ‘a woman having given birth,’ ‘a parous woman.’ As in the case of lamdāni, the resultative state waldat is an acquired quality that possesses social relevance.109 The case of lamdāni (No. 43) is not really different. Note also my contextual interpretation of anāku-ma kabsāk-šunūti in (36) above as a property of the goddess. (44) (a) awīlum abi dMAR.TU ša a¶-zu-ki ‘The honourable Father of the Amorites who is married to you’ (AbB 9, 15:21, tr. Stol).

Note that a¶iz stands in a relative clause, as in (43 a–b). (b) ina bullu¢ ummi-ki-ma lāma ana sikkātim nu´!u PN ina kišaršim innidi ištu i-šu qātī-šu igmuru u nu-a-um ša ummi-ki ē¶uzu (PN mer!assu a-¶iiz) nu-a-um gamram ša kišaršim igmur-ma ‘While your mother was still alive, before we left for the expedition, Iliya had been thrown into prison. After he had spent all his cash, the indigenous man who had married your mother (Iliya is married to his daughter), (this) indigenous paid the expenses of the jail’ (OA, Kt 91/k 423:24–32, tr. adapted from Veenhof 2008:107).

The other two tokens that have been found are negated, one of them obtains in a relative clause: (c) a¶ū-ni ´e¶rum aššatam ul a-¶i-iz ‘Our youngest brother has not taken a wife’ (i. e., is still unmarried)’ (AbB 3, 2:11, tr. BK). 108 The lawgiver’s point here is that the widow is a woman having given birth, but not the culprit’s mother (the latter situation is taken care of in the preceding norm, with a different sanction). 109 One is almost tempted to suspect that wālittum of our dictionaries is probably in part of the cases rather walittum, especially since no SC P walid ‘is born’ in the resultative reading has been found in the corpus, and the VA waldum ‘(the one) born’ does not really exist (AHw. 1458a). Note that the contexts of AbB 14, 93:11 (wa-al-du) and ARM 10, 106:10 (wa-al-da-at) are illegibly broken, while ARM 6, 43:5 [mā]r irtim ša šaddaqdim wa-al-du cited above as (27) is of course not RES, it is used with the force of the N Pret. (ša šaddaqdim) iwwaldu, which is attested in OB (AHw. 1458a).

S. Loesov, The Suffixing Conjugation of Akkadian …

119

(d) ana a¶ī-šunu ´e¶rim ša aššatam lā a¶-zu ‘To their young unmarried brother (they shall apportion the silver equivalent of the bridewealth)’ (CH § 166:67, tr. Roth 1995).

In the above examples, the relation of negated tokens (44 c–d) to the non-negated ones (44 a–b) looks like the binding of a variable in the static sentence: aššatam ul a¶iz ‘he is unmarried’ vs. mer!assu a¶iz ‘he is married to his daughter,’ a¶iz-ki ‘he is married to you.’ This leads one to think that in Akkadian the natural way to say ‘he is married’ was *aššatam a¶iz, this is once more a resultative state interpreted as an acquired feature (or “quality”) that possesses social relevance. This would mean that aššatam ul a¶iz (45c–d) is what Kozinskij (1988:522) calls “anti-resultative,”110 a state that in real life does not result from an action but nonetheless is not considered by the language to be a primitive, non-created, “natural” property.111 The case of the SC of nadānum/tadānum ‘to give’ is instructive. Although the state of having-been-given is unobservable, the SC P is attested by dozens of stereotyped 3rd person tokens: they are “bureaucratic” in the official correspondence of the first Babylonian dynasty (‘is

110

By way of illustration, I would compare this evidence to Spanish constructions with the auxiliary verbs ser ‘be’ and estar ‘≈ come to be’: este hombre (no) es guapo vs. este hombre (no) está casado, while *este hombre no es casado is unacceptable. A bit simplifying, one can say that ser + adjectives forms purely stative sentences, while estar + past participle forms resultative sentences. The tertium comparationis between Spanish and Akkadian is that on the surface both no está casado and aššatam ul a¶iz look like resultative sentences (because of estar and the direct object aššatam, respectively), though actually they are “derived statives.” 111 The present writer is doing his best to take BK’s RES seriously, and this attitude leads to questions at every turn. By definition, negated tokens of the SC cannot be resultative, while in text they are quite common. What to do with them? Theoretically, we can assign them to either pure (“anti-resultative”?) states or to negations of past-time events (negative [plu]perfects?), and for the moment I do not see a third possibility. In certain cases, the past-time option is inescapable: ištūma KÙ.BABBAR ana qātī-ka lá ma-aq-tá-ni [maqt-an-ni] ‘If the silver has not indeed comevent. into your hand’ (CCT 4, 30b:14f., the syntax excludes the RES interpretation: note the ventive coreferential with the Endpoint/Recipient ana qātīka). Contrast a vivid example of a purely descriptive usage: ul ina pilši ka-aš-da-aku ‘I have not been seized in burglary’ (AbB 2, 83:32). The letter is the author’s cry for help from prison, the burglary language is here nothing more than façon de parler, the message of this sentence being simply “I am no criminal, I have done nothing bad.” IA suggests that at this point I am being oversubtle and too “notional,” while Akkadian is not that sensitive here, it sees no problem at all about producing negated paris forms.

120

Articles: Ancient Near Eastern Studies

given,’ about the state property),112 and technical (‘is sold’) in OA (e. g. TPAK 3:9; Prag 1, 520:6). These P tokens belong to professional jargon, therefore they do not need to comply with the observability criterion. By contrast, the SC A is relatively rare, it appears mostly in the 1st p. sg. (and sometimes 2nd p. sg.). These examples are often (and perhaps always) RES, because they refer to a socially relevant state of the agent: (45) (a) 2 GÍN kaspam na-ad-na-ak-šum ‘I will have two shekels silver given to him’ (AbB 4, 149:10).

The context of AbB 4, 149 makes it clear that it is the addressee who has to pay the silver on behalf of the author (i. e., in the author’s future), therefore F. Kraus notes (p. 99 of the Edition) that the contextual reading is ‘ich habe ihm zwei Sekel Silber zuerkannt/versprochen.’ (b) kaspam gamram lū na-ad-na-ku ‘I have paid all the silver’ (YOS 8, 150:21f., a legal document).

The author is now free from an obligation. (c) ina libbi 3 ´ubātī-ka ištēn na-ad-na-a-ti u šaniam anāku annikīam attadin ištēn-ma ´ubāt-ka u¶¶ur ‘(As regards the three garments, your impost,) of these three garments of yours you have given one, a second one I myself have now given here, (so that) only one garment is (still) due from you’ (AbB 14, 119:6–9).

Kouwenberg (2010:171, fn. 34) claims that nadnāti “implies that the addressee is now free of his legal obligation,” whereas anāku attadin “indicates the news value of the message for the addressee and/or its recentness and relevance.” This looks theory-dependent, since the payment of the threegarments nēmettum by joint efforts of the writer and the addressee is the only concern of the letter. Yet in the corpus there is no such thing as a nonmotivated alternation of the SC A and the t-Perf.113 My explanation is as follows. The form nadnāti is (with BK) RES ‘you have (one piece) given.’ Had the writer simply wanted to say ‘I have another one given,’ he would have said nadnāku, yet he needed to specify to his business partner that he had made the payment annikīam, where he was penning his letter. As we know, the place adverbial annikīam is incompatible with the RES read112

E. g., AbB 4, 6:4–10; 55:9ff., both times about land tenure: the land is or should “be given” to somebody. 113 This follows from my observations in the course of this study and is in agreement with the tenor of Kouwenberg 2010.

S. Loesov, The Suffixing Conjugation of Akkadian …

121

ing of the SC, so the writer selected the second-best option, the t-Perf. attadin. By contrast, annikīam squares with the lexically-determined nonRES readings of the SC, as we have seen in (37a–b): PN annikīam saniqniāti and anāku annikīam sanqāku. The SC of the OB verb ¶abālum ‘to harm, wrong’ is perhaps even more telling. The SC is formed mostly for the specialized legal meaning ‘to deprive,’ i. e. ‘to take something from somebody illegitimately’ (with two accusatives). The SC P appears in AbB in “bureaucratic” contexts, e. g. šumma eqel-šu … ¶a-bi-il ‘if he is (indeed) deprived of his field (give it back to him!)’ (13, 43:20, a letter of Hammurapi). The P form is unobservable (though it is meant to be RES), and this is small wonder, since we already know that the chancellery of the Amorite dynasty was fond of abusing the SC. The important thing is that the SC A ‘he is the one who has deprived (somebody of something)’ is also attested, although the trivial result of harming/depriving is supposed to affect the patient rather than the agent. Yet ¶abil could be used as a charge, an accusation, i. e. to describe the former agent. Consider the following examples: (46) (a) É ¶a-ab-la-an-ni ‘he has deprived me of my house’ (AbB 2, 111:13). (b) eqlam ša ¶a--lu-ni-in-ni ‘the field of which they have deprived me’ (AbB 2, 111:29). (c) PN … ¶a-ab-la-an-ni-a-ti ‘(PN) has wronged us’ (AbB 2, 74:12, beginning of a suit).

To recapitulate: semantically, the “acquired-quality” meaning of the SC A RES resembles the grammatical meaning of predicative adjectives, in spite of its verbal two-place syntax, and in this respect it is radically different from the non-RES A tokens, which are by definition dynamic (or “fientive”). Thus, what is described by the sentences ‘(a woman) ša marī waldat’ and ‘PN mer!assu a¶iz’ are properties of the syntactic subjects, while for ‘PN annikīam saniq-niāti’ or šeam ama¶¶ar-ma kašdāk-ki this interpretation hardly makes a lot of sense. If we ask, “What does it mean for the SC A to be RES?”—The answer will be, “To code an acquired quality of the former agent.” As we have seen, this answer implies a lexical restriction on the formation of the SC A RES. It is understood that the “social relevance/legal status” mentioned above is a variety of the “acquired quality” concept. For obvious reasons, the former is especially well represented in our records of Akkadian. The negative evidence pro this thesis is the fact that e. g. the SC A of basic verbs of telic transitive motion is rare and probably never RES. In

122

Articles: Ancient Near Eastern Studies

particular, this is true of the verbs wabālum ‘to bring,’ šūbulum ‘to make bring, to send,’ tabālum ‘to take along, to carry off,’ šapārum ‘to send (a message),’ ¢arādum ‘to send off, dispatch,’ while for the other finite forms most of these verbs are represented by thousands of examples. This is because Akkadian does not consider “having sent”/“being-sender” etc. to be an acquired quality. In part 3 of this paper, we will look at the respective evidence. 2.2. The SC RES of intransitive telic/punctual verbs Now, what does it mean for the SC of intransitive verbs to be RES? If we start with the above list in Kouwenberg 2010 (7.3.2) which comprises eleven admittedly frequent tokens of the SC formed from “intransitive verbs of telic movement and change of state verbs,” we will see that not all these verbs yield resultative SC forms. An example is tebûm ‘to stand up’: tebi is well attested, yet it hardly ever means literally ‘he is [now] standing (= having stood up).’ For the basic meaning ‘to stand up,’ no single example of the SC has been found in OB and OA letters.114 On the other hand, tebi is attested for several derived meanings in OB literary texts, and especially in SB; see AHw. 1342f.; CAD T 317b.115 All the OB tokens seem to be non-RES. Consider an example from an OB extispicy report: (47) ašar mukīl rēši kakku šumēla te-bi ‘At the “support of the head,” a “weapon-mark” protruded to the left’ (JCS 11, 99, No. 8:20, text and tr. CAD T 317b).

The token is stative and by definition observable, but it is clearly not resultative. Both its lexical and grammatical meanings are metaphorical, the latter feature constituting the norm in still life divination texts. By contrast, the SC of another “intransitive verb of telic movement” in BK’s list, ¶alāqum ‘to get lost,’ behaves differently. For the basic sense of

114

For that matter, tebi with whatever meaning has not been found in OA and core OB letters. In my corpus, the closest approximation to the basic sense of the root would be Enkidu te-bi-i-ma Š. ir¶i ‘[For six days and seven nights] E. was sexually aroused and mated with Šamhat’ (Gilg. P. ii 7, if interpreted correctly), and see CAD T 317f. for more examples of tebi with this meaning in later literary texts. Anyway, this one looks like a technical (and probably euphemistic) usage, since the grammatical subject is never the penis. 115 I am grateful to Rim Nurullin for discussing with me the usage of tebi in literary texts.

S. Loesov, The Suffixing Conjugation of Akkadian …

123

the root, it is attested in OB and OA letters (and OB laws) by dozens of tokens, often in RES readings.116 Consider some RES examples: (48) (a) PN ¶a-li-iq šumma illak-ak-kināšim pādā-šu ‘Dan-ilišu has disappeared. If he shows up at your place (2 f. pl.), detain him!’ (AbB 7, 29:11ff.; the message is quoted in its entirety, save for the letter-head and the seal legend). (b) ‘I found and detained in Isin PN ša ištu … MU.3.KAM ¶a-al-qú who has been missing for three years’ (AbB 14, 186:9). (c) ¢uppam ērišū-ma umma PN ¶a-lá-aq ‘They demanded (from PN) the debt-note, but PN said it is lost’ (JCS 14, 1:7, OA).

Why then ‘PN ¶aliq’ is a good Akkadian sentence while ‘PN tebi’ is not? We cannot explain the contrast in the behaviour of ¶aliq and tebi using notions applied above to transitive verbs. To claim that ‘having-been-lost’ or ‘having-disappeared’ is an observable resultative state (on a par with ‘having-been-broken’) is to overload the concept “observability” by using it as a metaphor too far removed from its literal meaning. As we have already seen, observability as a heuristically useful notion has to be applied literally. Hence, to explain the above evidence we need to look for something else. As already mentioned above, the SC of telic intransitives can be productive and RES if the respective verb is unaccusative, i. e. if its subject is patient-like. According to this prediction, the SC of unergative intransitives (i. e. those whose subjects are agent-like) will be non-RES and not productive for the basic root meanings. Keeping to verbs of movement, we can now account for the virtual non-existence of the SC of agentive motion verbs, e. g. tebûm ‘to stand up,’ erēbum ‘to enter,’ wa´ûm ‘to go out,’ warādum ‘to go/come down,’ elûm ‘to go/come up.’117 By contrast, the SC of maqātum ‘to fall down’ is repre116 In OB, it is also well-known in ¶aliq-ma chains, as in No. 1 above, and see additionally e. g. AbB 6, 171:8; 9, 62:12. 117 Kouwenberg (1998:183) and BK’s OA Verb List gloss this verb as ‘to go/come up; to be high,’ and cf. Loesov 2006:147, No. 15. Von Soden divides the AHw. entry elûm in two parts: “A. St. (hoch sein) selten; B sonst” (AHw. 206b), and suggests that the rare SC forms (OB and SB) in his A part of the entry are deadjectival from elûm I ‘hoch’ (ibid. 205b). This decision of von Soden looks reasonable, if we agree that it makes sense at all to speak time and again about the SC tokens derived not directly from a verbal root (or BK’s “paradigm”) but rather from a deverbal adjective. Sometimes, I do make such claims when I describe a certain SC token as “semantically deadjectival” (cf. e. g. the above story about batiq, ex. No. 21).

124

Articles: Ancient Near Eastern Studies

sented by dozens of tokens in various lexical meanings. In particular, for the basic unaccusative sense of the root ‘to fall down, to collapse’ it is attested by the following RES examples (they are the only tokens of maqit for the basic sense of the root I have found so far): (49) (a) ašamme-ma urkat bētim ma-qí-it ‘As I know, the rear part of the house has collapsed” (Prag I 577: 4f., OA). (b) š[u]m-[m]a É-tum e-le-nu-úš-šu im-qú-ut-ma úš-šu-šu da-an-nu x [o o o] É {x x x x} da-an š[um-ma uš-šu]-šu ma-aq-tu-ma e-le-nu-úš-šu ša-lim m[im-ma e]-pí-is-sú ú-ul i-le-i i-ma-aq-qú-ut ‘Soit une maison dont le haut s’est effondré mais dont les fondations sont solides: le […] de la maison est solide. Mais si ses fondations s’effondrent alors que son haut est en bon état, son maçon ne pourra rien faire et elle s’effondrera’ (ARM 26/2, 392:45–48, text and tr. of the Edition).118

The morphological difference between šumma bītum elēnuššu imqut (‘if the roof of the house has fallen down’) and šumma uššūšu maqtū (‘if its foundation [pl.] has collapsed’) seems to be a free variation. In Akkadian, maqtū in this context is RES (and therefore the marked form), while the Pret. imqut is an unmarked form: being essentially PERFECTIVE PAST/ AORIST, it can appear in the semantic slots of PERFECT and RES as well.119 A derived meaning of maqātum refers to the loss of cattle, maintaining the unaccusative sense of the root. In (50b), the SC used in this meaning is clearly RES. (50) (a) (b)

118

ÁB.ÚI.A ekallim i-ma-qú-ta ‘The livestock of the palace keeps falling’ (ARM 1, 118:21).120 ANŠE.ÚI.A-[i]a

ma-[a]q-tu ar-ka-s[ú] ap-ru-ús-ma ANŠE.[ÚI.A-šu] m[a]-aq-[tu] ‘ “Mes ânes ont subi des pertes.” J’ai fait une enquête. Ses ânes ont bien subi des pertes’ (ARM 26/1, 63:12ff., text and tr. of the Edition).

This is a parable the king Hammurapi of Babylon told an envoy of Zimri-Lim. When I dub maqtū “the marked form,” I want to say that its RES reading is encoded morphologically. In this text, imqut is “an unmarked form” because it is synonymous with maqit and therefore has the RES reading as well, but its morphological shape does not encode RES, it is non-committal as to whether the contextual meaning is RES or otherwise. Cf. also the above observation of BK (p. 117, fn. 106) according to which OA can use ilqe where OB uses the A leqi. 120 See CAD M1 243b for more examples of prefixing forms with this meaning. 119

S. Loesov, The Suffixing Conjugation of Akkadian …

125

This is once more the only example with this meaning that I have found. In Mari, maqit is attested in yet another metaphorical meaning, clearly derivable from the unaccusative notion of physical “falling-down”: (51) mu-ut-ta-at ma-a-tim a-na NP ù mu-ut-ta-at ma-a-tim a-na a-¶i-šu maaq-[t]a-at ‘La moitié du pays a fait soumission à NP, et l’autre moitié à son frère’ (ARM 26/2, 359:6f.).

The token, as interpreted in the Edition, is RES, and so probably are a few similar usages of maqit in Mari, see ARM 26/1, 5:48 and the note (s) on p. 85 of that volume. In OA, maqit is frequent in the motion meaning ‘to reach (a landmark), to arrive,’ said about humans and merchandise. This derived meaning of maqātum is quite common in the prefixing tenses as well, and I believe it keeps the initial semantic element of non-agentive (= “unaccusative”) movement. Consider some examples: (52) (a) šu-ma ´ú-¶a-ru-um a-na li-bi4-kà ma-qí-it ‘If the (above-mentioned) servant has arrived at your place, (let us know)’ (AKT 3, 114:16ff.). (b)

TÙG.ÚI-tí-šu a-šar ma-aq-tù-ni ú-šu-ru-šu- ‘ sollen sie seine Stoffe, wohin sie auch gelangt sind, ihm loslassen’ (Prag I 681: 30f., text and tr. of the Edition).

(c) umma šūt-ma KÙ.BABBAR ana qātī-ya lā maqit KÙ.BABBAR ana qātī-ya imaqqut-ma kasap-šu ušebbal-šum umma anāku-ma ištūma KÙ.BABBAR ana qātī-ka lā ma-aq-tá-ni [= maqtanni] ammala našperti warkitim 20 ma-na KÙ.BABBAR u ´ibāti-šu mala ana ´ibtim ilqe-ak-kun-ni lubil-šumma u šittam warka šēbil ‘He told (me): “The silver has not (yet) come into my hand. If the silver comes into my hand, I will send him his silver.” ’ ‘I answered (him): “Since the silver has not comevent. into your hand, then according to (his) last message 20 mines of silver and their interest, as much as he took at interest on your behalf,—I shall bringJuss. (it) to him, and you bringImv. the rest afterwards” ’ (CCT 4, 30b:8–22).

All the maqit tokens in (52a–c) are RES (some of them are “anti-RES,” if we insist on taking negations in earnest), i. e. none of them occurs in syntactic contexts that cancel the RES force of the SC.121 Note the (+ vent.) clause ištūma KÙ.BABBAR ana qātī-ka lā maqtan-ni that appears in free alternation with the (– vent.) clause KÙ.BABBAR ana qātī-ya lā maqit. To my 121

As we already know, ENDPOINT is the admissible (or even necessary) participant of the RES clauses with verbs of telic motion.

126

Articles: Ancient Near Eastern Studies

knowledge, the directional ventive is used with the SC of motion verbs very sparingly.122 In this case (ištūma kaspum ana qātī-ka lā maqt-am), the ventive probably foregrounds the endpoint. We are now passing to BK’s intransitive “change of state” verbs that form the SC. The sample includes four items, belû ‘to go out (of fire),’ mâtu ‘to die,’ pašā¶u ‘to become calm,’ and šebû ‘to become satisfied.’ Of these four, the SC is attested in the corpus with any noticeable frequency only for mâtum ‘to die.’123 This is not because the writers in our corpus used mâtum a lot more frequently than any of the other three verbs: the numbers that can be gleaned from our dictionaries do not support this hypothesis. I suggest that the SC of mâtum ‘to die’ is productive and often RES not because of “change of state,” and not just because of the importance of the concept ‘to die,’ but due to the patient role of the subject. There is a non-trivial syntactic isogloss between ¶aliq ‘he is lost’ and mīt ‘he is dead’: the two forms are productive in both RES and non-RES slots.124 Consider a few examples taken from my more detailed description of mīt (Loesov 2010:774f.): (53) (a) lā libbi ilim-ma aššitī me-ta-at ‘alas, my wife has died’ (TCL 4, 30:4, OA). (b) šummamin me-ta-ku ‘if I were dead’ (TCL 19, 32:30, OA). (c) ½ TÚG i-nu-mì DUMU.MUNUS-sú me-ta-at-ni a-na PN ú-šé-bi4-il5 ‘I sent half a textile to PN after his daughter had died’ (BIN 4, 141: 1–3, OA). (d) ištu ITI.1.KAM šerrum ina libbīya mi-it-ma ‘The child in my womb died a month ago’ (ABIM 15:10).

122

Besides the naš!-am clause in OA (‘he carries [smth.] to where you are’), I am aware of only two additional examples: ana eqlim šuāti mamman lā ¢e4-¶i-a-am ‘nobody has to approach this field!’ (AbB 4, 125:14); kaspum … šé-bu-lam ‘Has the silver been sent to you?’ (VAS 26, 69:6f.). 123 In AbB 1–14, the SC of šebûm is attested once: ‘except for them, there is no civil servant ša eqlam lā ´abtu u lā še-bu-ú that has not got a field and is not satisfied (= provided with livelihood)’ (AbB 3, 74:25). It has not been found in OA. The SC of belûm is known from OB oil omens in the meaning ‘to burst (said of bubbles),’ see CAD B 73b. The SC of pašā¶um has not been found in the corpus, it is attested in later periods, see CAD P s. v. As for mīt/mēt, we will now see that in OB and OA letters it is productive for the basic sense of the root. 124 See also Kouwenberg 2010 (7.3.3) on the non-RES use of mīt.

S. Loesov, The Suffixing Conjugation of Akkadian …

127

Examples (53a–b) are RES, ex. 53c is PLUPERFECT, while ex. 53d is Note that the RES tokens of mēt can hardly be considered observable. According to suggestions in Loesov 2010:774f., mēt in OA tends to be used in most past-time slots (i. e., instead of the t-Perf. and the Pret.), while in OB letters the t-Perf. imtūt is often used instead of the Pret. imūt (and possibly instead of mīt, though this is difficult to prove). If this is true, both facts represent unusual developments and require further study. PRET/SIMPLE PAST.125

2.3. What is the tertium comparationis between the three diathetic varieties of the SC RES? Now, we have concluded that the SC P is RES if it is observable; the SC A is RES if the action is believed to affect the agent (= personal and volitional subject) and culminate in a quality of the latter, and the SC Intr is RES if its subject is patient-like. The shared feature of the three varieties is obvious: it is the communicative importance of the resultant state. One can christen this feature “(high) cognitive relevance.” Our cognitive relevance (or “salience”) is, so to speak, an umbrella notion that can be neither induced directly from the linguistic evidence nor applied to it directly. For this reason it is appropriate to dub this feature “cognitive” rather than “linguistic” relevance/salience. Indeed, there is no a priori (= prior to a detailed examination) criterion that would tell us what is more “relevant”: nakis ‘it has been felled/cut across’ or ma¶ir *‘it has been received’; mārī waldat ‘she has children’ or nadi *‘he has thrown down (a thing)’; ¶aliq ‘he has disappeared’ or wa´i *‘he has gone out.’ At this abstract level, the very question about “relevance” may sound bizarre. Yet, in the course of a piece-by-piece study of certain basic verbal concepts, we have come to learn that for Akkadian the trivial RES state of the former patient is worth mentioning (= coding with a special “tense”) if it is observable; the RES state of the former agent is encoded morphologically if it is an acquired feature of the agent (i. e., signals the agent’s “change of state”); the RES state of the Intr subject is expressed if the subject looks like a patient, though the respective state does not have to be observable. 125

In anticipation of the research on non-RES functions of the SC of telic roots, we can observe that so far we have seen PLUPERFECT/BACKGROUND as such a function much oftener than PRETERIT. The latter seems to be exceptional. The most prominent example I am aware of is dīk *‘he is killed’ used instead of the non-existent N-stem Pret., but this fact was explained in terms of morphonology rather than lexical semantics.

128

Articles: Ancient Near Eastern Studies

The shared feature of “observability” (patient), “acquired property” (agent) and “patient-likeness” (intr. subject) is the maximum change in the world’s state of affairs. This is another man’s word for “cognitive relevance.” 3. Retrospect and Prospect: The lexicon-based description of Akkadian morphosyntax In this paper, I have attempted to describe conditions that are essential for the SC of telic verbs to be RES. I hope I have made some progress, yet certain vital questions remain unanswered, and others even unasked. When I started writing this article, IA urged me not just to hunt for nonRES tokens, but to ask what they are doing in the language, i. e. to look for patterns of their use.126 This challenge has not been taken up here and remains part of the agenda for future research. Furthermore, so far we have dealt in some detail mostly with verbs whose SC can have RES meanings, because the respective SC forms do follow the stipulations I have suggested. Now, one may ask: what then about the SC of telic transitives which, due to their lexical meanings, can be neither “observable” nor “acquired property”? And what about the SC of unergative (= agentive) intransitive verbs? Are there common examples of these types? What are their functions? Moreover, contra our reference works, from GAG to Kouwenberg 2010, we do come across the SC of certain atelic verbs, and some tokens seem to be productive and intrusive into the present-time domain as habitual predicates. Consider an example: (54) (a) [in]a mimma annîm [a]k-la-ku ‘I am entitled to the use of (lit. “I eat/I may eat”) all this’ (AS 22, 26:5, early OB). (b) annakam emārū ak-lu-ú ‘The donkeys “eat” (= receive fodder, are being fed) here’ (ATHE 46:18, OA). (c)

ANŠE-ku-nu

lu ak-lu ‘let your donkeys get fodder’ (Prag I 718:25).

(d) [u]šbat ak-lá-at u paššat ‘she is entitled to stay, eat and get ointment’ (Albayrak 2004, 12:14f.). 126

BK also writes me, “You show clearly that describing the stative as resultative and nothing else is not the whole picture. What I still miss … is the rest of the picture, in other words, what is the overall function of the non-resultative active statives? Are they secondary developments based on the resultative (like OA qabi)? Or instantiations of the verbal adjective used predicatively?”

S. Loesov, The Suffixing Conjugation of Akkadian …

129

To complicate things even more, the SC tends to be lexicalized the way other tenses are definitely not. In other words, the SC sometimes appears with the meanings which do not really have matches in the prefixing tenses. An extreme example is the OB ¶abtāku (lit. ‘I have been robbed’) that ended up becoming an interjection: it is used as a call for help, or with similar forces. In this function, ¶abtāku combines with appropriate forms of šasûm ‘to cry,’ followed by an articulated utterance of the one crying, introduced by umma: (55) (a) ¶a-ab-ta-ku issi umma šī[ma] amtum ša Bēlšunu anāku bēltī iqīš-anni ‘ “Help!” she cried. She yelled (umma šī-ma): “I am a slave-girl of Bēlšunu. My mistress gave me (to him)” ’ (AbB 1, 27:17ff.).127

For OB Mari, numerous tokens of ¶abtāku (mostly from unpublished manuscripts) were gathered by Fr. Joannès in ARM 26/2, p. 299.128 The syntax is essentially the same as in the core OB examples, the stereotyped rendering is “Quel scandale!” Fr. Joannès points out the idiomatic nature of this verb phrase, not quite described in the dictionaries: “L’emploi constant de la première personne du permansif traduit l’aspect figé qu’a pris cette expression en akkadien.” The lexicalization is perhaps more prominent in the Mari examples than in the AbB ones: (b) sugāgū … itbû-ma ištēniš ¶a-ab-ta-ku issû-ma umma šunu-ma ‘les cheikhs se sont levés, et ensemble ont crié “Quel scandal!”, en disant …’ (M.5437+, cited by Joannès ibid.).

And cf. an even more eloquent example: (c) awâtim šināti eštemme-ma ¶a-ab-ta-ku assi-ma ‘(when I arrived at Saggaratum) I kept hearing about these things and saying “What a shame!” ’ (ARM 26/1, 5:18).

I have set out to elucidate these complexities by describing the SC from its “grassroots,” i. e. from the syntactic behaviour of the SC of a given frequent verb, comparing its SC (wherever appropriate) with other members of its morphological paradigm. This procedure is adequate for a synchronic tense-aspect study of a living language, while for an ancient literary language its validity is restricted at best and sometimes even du127

See also u ¶a-ab-ta-ku ištanassi umma šū-ma … (10, 81:6′f.), ¶a-ab-ta-ku ašsi-ma umma anāku-ma … (10, 184:2f.), and slightly different syntactically: umma šunu-ma wardum ša dAMAR.UTU-na-´ir umma ¶a-ab-ta-ku issi-ma (AUCT 4, 89:6f.). 128 The reference, along with the explanation that this usage is idiomatic, is courtesy IA.

130

Articles: Ancient Near Eastern Studies

bious.129 The truth is my morphosyntactic “shots” presented below depend on the corpus of published texts which is contingent on what can be ultimately called “chance discoveries” in the soil of today’s Iraq, Turkey, and Syria. This means that we will never learn everything we want about the Akkadian verb. On the other hand, the bulk of the available OA and OB corpora consists of a huge amount of non-literary texts (some 31,000 texts in OB, see www.archibab.fr),130 while the verbal roots under investigation are very frequent, many of them are attested in our corpus by thousands of inflectional tokens. All in all, this makes 2000–1700 B. C. Akkadian a rewarding field for a fine-tuned study in verb’s grammatical semantics, perhaps more so than any other ancient textual corpus. I made a first attempt of this kind in Loesov 2010.131 The analyses reflected in that paper had been carried out before my views on the meaning of the SC took the exact shape expounded above. Therefore certain decisions may have become obsolete, yet the evidence collected and commented upon can be useful for future research anyway. In what follows I go on with the lexicon-based investigation of the SC. The aim is to test the value of the above criteria, to try to falsify some of them, and, more generally, to advance the discussion by bringing to the reader’s attention certain facts and observations. The availability and meaning of the SC are related not to the “dictionary verb” as a whole (e. g. nadûm or šapākum), but rather to its individual senses. For this reason, wherever appropriate, I start with the SC for the basic root meaning and then pass to its other meanings, analyzing them 129

With mixed feelings, I am now reading in the Introduction to J. Blau’s new grammar of BH, “Synchronic investigation is not congenial to Biblical Hebrew, since the corpus does not reflect a closely knit linguistic system but is profoundly multilayered. It is not useful to analyze such a language with the tools of synchronic linguistics” (Blau 2010:5; it is understood that here “such a language” = “such a corpus”). Blau’s judgment may sound “unscientific” and blunt, especially given the high number of doctoral dissertations on the synchronic (morpho)syntax of BH and other, much smaller, Semitic corpora. Yet Blau is essentially right. 130 It stands to reason that their value as linguistic evidence is uneven depending on the genre. 131 In Loesov 2010, the following verbs were discussed: kullumum ‘to show, assign’; labāšum ‘to put on clothing’; lapātum ‘to touch, affect, write down’; lupputum ‘to touch, smear, write’; lawûm ‘to move in a circle, surround’; leqûm ‘to take’; ma¶ārum ‘to accept, receive’; mâtum ‘to die’; qerēbum ‘to be near, approach’; sanāqum ‘to arrive, reach, check, put pressure on’; ¢e¶ûm ‘to come near, approach’; ¢u¶¶ûm ‘to conduct, bring near, hand over, add to.’

S. Loesov, The Suffixing Conjugation of Akkadian …

131

one by one. Among other things, this will hopefully permit us a fresh look at lexicalized readings of the SC, i. e. those that have no matches in the prefixing tenses. It goes without saying that by default the data of OB and OA are being treated separately. In the rest of this contribution, I will study three basic verbs of transitive motion whose SC is only marginally attested. In the forthcoming part of this work, we will have a look at frequent transitive verbs whose SC is productive and reveals considerable polysemy. wabālum ‘to bring.’ The SC is rare, given the enormous frequency of prefixing tenses. The VA is not attested at all. In OB, I have found five examples for the basic meaning (both P and A), two of them obtain in literary texts. (1) ‘18 shekels of silver ša ana itbārūtim ba-ab-lu that were brought for an i.-partnership’ (UCP 10, 167, No. 98:24f. text CAD I/J 295a). (2) ‘Has [the silver] that you took ul ba-bi-il-ma-a not been brought?’ (AbB 9, 279:2). (3) [k]u-up-ru ba-bi-il i-pé-e¶-¶i ba-ab-šu ‘Pitch was brought that he might close up his door’ (OB Atra-Úasis, Lambert–Millard 1999, 92 iii 2:51). (4) [kīma bēl]ī īdû 1 LÚ parzillam ba-bi-il a¢¢aras-su ‘As my lord knows, a certain man used to bring iron; I have (now) dispatched him’ (AbB 9, 82:25ff.).

Rowton 1962:276 translates the last sentence ‘He has always been the only one to bring in the iron.’ Stol rejects Rowton’s translation and proposes in the Edition, ‘As my lord knows, one man carries (the) iron.’ BK suggested me a non-habitual solution in a p. c., ‘I don’t think anything is wrong with “the man has brought iron” (the context is fairly obscure).132 This is supported by the Sargonic instances, which clearly refer to a specific delivery (if we take these as statives, which I find attractive).’133 Thus, we have PAST HABITUAL (‘used to bring/has been bringing’), ‘has brought’ (PERFECT or RES), and ‘carries’ (PRESENT HABITUAL) as suggested solutions. The decision is difficult; see the discussion presently. (5)

d

UTU

na-ši-ku-um me-e ÍD.IDIGLAT ù ÍD.BURANUN ša iš-tu ša-di-im ù GIŠ.ŠU.ÚR.MÌN a-na ka-ši-im ba-ab-lu-ú

GIŠ.EREN

132 Indeed, Stol admits in the Edition that he does not understand how the paragraph in question relates to the rest of the letter in terms of contents. 133 See these examples below in this entry.

132

Articles: Ancient Near Eastern Studies ‘O Šamaš! (Here) I have for you water from the Tigris and the Euphrates which brings134 you cedar and cypress from the mountains’ (OB diviner’s prayer; JCS 22, 26:18ff., the collated text as in Steinkeller 2005:31).135

The tense-aspect reading of both našêk-kum and bablū is difficult to pinpoint on non-linguistic grounds because “the conceptual and mythological setup of Babylonian extispicy” (cf. Steinkeller 2005) is not fully understood. My rendering is meant to suggest that našêk-kum is a stative predication, while bablū has habitual force; i. e. the two tokens of the SC display different nuances of the present-time stative reading. Yet a past-time rendering ‘water … which (has) brought to you’ is difficult to exclude. Given the above uncertainties, I decided to present here some of the few examples I had found outside the corpus. Two tokens with the shape wabil are probably attested in Sargonic Akkadian (OAIC 36, an economic text)136: (6) 18. 30 [SIKI] MA.NA 19. in NA4 KU3.BABBAR 20. Bi2-za-num 21. wa-bil2-śi-in 22. 5 Bi2-za-num 23. [w]a-bil2-śu

‘30 minas of [wool] (measured by) the “stone of silver,” Bizanum is their porter. 5 (GUR of barley), Bizanum is its porter.’

As we will now see, Gelb’s interpretation of wa-bil2 as a participle (i. e., agent noun wābil-) may be wrong. This noun is amply attested in OB Mari,137 Tell Rimah,138 and in OA,139 it is used almost exclusively as the 134

The P translations of CAD Š1 54a (‘water from the Tigris and Euphrates, brought for you from the cedar and cypress mountain’) and Š3 350a (‘water … is carried to you’) are a shame, since ša-di-im GIŠ.EREN is not ‘the cedar mountain,’ while Goetze in the Edition (1968) had correctly put forward an A rendering, ‘water … which has carried.’ 135 Steinkeller translates: ‘O Šamaš! I bring you water from the Tigris and the Euphrates which has carried for you cedar and cypress from the mountains.’ 136 The text is that of http://web.uni-marburg.de/altorientalistik/sommerfeld_ oaic.htm. It is not materially different from Gelb’s reading. The translation facing the corresponding lines is that of Gelb. 137 E. g. ARM 2, 141:4; 28, 136:12, 173:5. The form in Mari and Tell Rimah is always wābil-. Curiously, this word has not been found in the core OB, with whatever phonological shape. 138 OBTR 85:9′; 165:1′. 139 See GKT 156e, Veenhof 1972:440f. For the analysis of our alleged nominal clause in Sargonic, it is worth noting that in OA wābil- is used as subject/topic in particular in the identifying nominal clause wābil ¢uppim šūt tamkārum ‘the bearer

S. Loesov, The Suffixing Conjugation of Akkadian …

133

head of noun phrases wābil ¢uppim, wābil ¢uppīya annîm, wābil kaspim (e. g. VAS 26, 70:10), and the like.140 This noun phrase is always topical: it is never the predicate of a nominal clause, as is supposed to be the case in OAIC 36. The overall syntax of OAIC 36 (verbal preterite-final clauses) and the shape of object bound pronouns in Sargonic Akkadian (Hasselbach 2005: 150) favour the interpretation ‘Bizanum has brought them/it.’ The transitive sentence structure will be similar to 1 LÚ parzillam ba-bi-il ‘a certain man used to bring/has brought iron’ (AbB 9, 82:25ff.), as cited above. In the 1st millennium, the SC does not seem to be frequent either. The few tokens I have found come mostly from the SB literary texts. P readings are attested in omina apodoses, consider an example: (7) kakkū ša ana libbi lā bablū ana rubê i¢e¶¶û-ni ‘Weapons that were not brought inside will attack the prince’ (KochWestenholz 2000:187, the critical edition of the SB padānu, a similar example from šumma izbu is cited ibid. fn. 537).

I have found one A present-stative (~ habitual) example in a SB prayer to Marduk: (8) [dumm]uqa ba-ab-la-ta ‘you bring welfare’ (CT 44, 21 i 11, cf. 13, the restoration of AHw. 1451a).141

Thus, for the basic sense of wabālum, the SC is only marginally attested. We can explain this by the fact that, according to the above criteria of RES readings, wabil cannot be RES in either A or P diathesis. Why then does it appear in the corpus at all? I believe wabil did sometimes show up in the of (this) letter is the creditor (in question)’ (e. g., UF 7, 1:22–25; CCT 1, 1a:35f., and see Veenhof 2008a:218, fn. 69 for the explanation that in this stereotyped clause wābil ¢uppim refers to “a person who had acquired a debt-note from the original creditor” [ref. BK]). 140 CAD A1 53a and AHw. 1454b cite only two examples of this noun, both happen to be OA ones, ni-a-um wa-bi-lu (BIN 4, 79:2′, in a partly broken context), and wa-bi4-lá-ni (TC 3, 38:20). Regarding the latter one, BK informs me that it has to be read wa-bi4-la-né ‘porters,’ acc. pl. of (w)ābilānum ‘porter/bringer.’ As additional evidence for this new OA noun, BK adduces ‘PN ābilān ¶usārīki’ ‘PN, the bringer of your (fs) lapis lazuli’ (Kt n/k 10:26, see Veenhof 2008a:218, fn. 69), and [a-n]a a-bi-la-an a-wa-ti-a ‘to the bringer of my case’ (OIP 27, 15:5; reading of Dercksen 2001:53, fn. 75). 141 T. Oshima (NABU 2001/15:16) offers the following text: ina ar-ni gíl-la-ti n[a-ás-¶ur-k]a ba-ab-la-ta ‘Even in the most sinful crimes, you bring [your benevolent attenti]on.’ It is based, as the author explains in fn. 9, on AfO 19, pl. 12:13– 20; CT 44, 21, i 6′–13′; LTBA 1, 68:2–4.

134

Articles: Ancient Near Eastern Studies

spoken Akkadian dialects, and this was due to the interplay of three factors: 1) the RES paris was productive for numerous telic roots; 2) wabālum ‘to bring’ was a common transitive verb of telic motion; 3) in spoken Akkadian, paris was often used in non-RES readings, whether past-time (for numerous telic verbs) or as a habitual predicate for some atelic verbs (see ex. 54 above). These conditions gave rise to a morphological analogy, a sporadic use of wabil for the basic motion sense. Its P readings would be normally past-time, because of the natural association of PASSIVE with PAST PERFECTIVE. By contrast, its A readings would be predominantly present habitual. This alleged state of affairs was also reflected in our written records, as discussed above. Mutatis mutandis, a comparable fact is the sporadic appearance of the SC of bašûm ‘to be’ and edûm ‘to know,’ which are semantically superfluous (see GAG 127b + *b). My suggestion regarding the present-habitual nature of certain A readings is supported by the only other usage of wabil known in Akkadian. This is the idiom pan- X wabālum ‘to show favour,’ the beneficiary of favour (= X) being coded by the noun phrase headed by pan- in the accusative case. Below are all three (or four) examples I know: (9) awīlum an-[nu-um] panī-ya ma-di-[iš] ba-bi-il ša ar¶iš nadān[im] epuš ‘This man is very nice to me,142 arrange to deliver (the barley) as soon as possible’ (Goetze 1958, No. 11:12–16).

In the OB Mari, M. Birot restores a uniqie A shape *wabil within the idiom panam wabālum: (10) ù PN LÚ GAL.MAR.TU ša RN … pa-an143 LÚ Èš-nun-naKI wa-[bi-il] ‘And Silillum, the rab Amurrim of Bina-Eštar, ... is in favour of the king of Ešnunna’ (~ ‘he is likely to defect from Bina-Eštar and join the king of Ešnunna’) (ARM 14, 106:20′ff.).144

142

Cf. CAD A1 18b: panu + abālu ‘to show preference, to favor.’ Goetze translates in the Edition, ‘The man in question very much deserves indulgence on my part’, which now does not seem a likely solution. CAD A1 19a translates ‘this man has shown me great favor.’ In view of the other examples, a present-time stative interpretation is the most likely one, and it does not contradict the context. 143 In my sample, this is the only example of the idiom where pan- doubtless has the sg. shape. 144 Be the restoration correct, this SC token (wabil rather than babil) will be an isogloss with the OA (see presently). Since ll. 22′ and l. 23′ are broken at the end, it is thinkable that wa- represents the Inf. acc. wabālam + a modal/phasal verb to its right (‘is going to become a follower of the king of Ešnunna’ or something like this).

S. Loesov, The Suffixing Conjugation of Akkadian …

135

BK’s OA Verb List brings two examples of wabil, both happen to be used within our idiom pan- X wabālum, and my search has brought no additional results: (11) pá-ni PN wa-áb-lá-tí-ma KÙ.BABBAR ša-áš-qú-lá-am lá ta-mu-a-šu ‘You favour PN (over me), therefore you do not want him to pay me (back) the silver’ (VAS 26, 3:14–17). (12) ù šu-nu pá-ni-a wa-ab-lu-ma ú-lá i-ša-pu-ru-ni-kum ‘They are being nice to me, yet they do not want to write to you’ (AKT 3, 67:28f.).

This is an A present-time usage.145 The pan-Acc. wabālum verbal phrase uses the Pret. for past time references, and the Pres. for future and injunctions (v. CAD A1 18f.). Consider an example: (13) kīma pa-ni la tu-ub-lu tašpur-am ina ubbubī-ka pa-ni-šu-nu la tu-ub-ba-al ‘You wrote me that you had treated them fairly … Treat them fairly (= “show them no preference”) ’ (ARM 1, 82:6–16f.).

This idiom goes back to a weakened, atelic sense of wabālum: ‘to carry’ rather than ‘to bring.’ Etymologically, this verb phrase means ‘to carry (somebody’s) face.’ “Face” is here a pars pro toto designation of a person. An interesting typological parallel is a nearly identical verb phrase in BH, nā»ā! pānīm ‘to receive someone in a friendly manner, be favourably disposed towards someone’ (HALOT 725). It is attested for all three temporal domains (cf. Mal 2:9 for PRESENT, where the A predicative participle is used). The only difference with pan- X wabālum is that the basic meaning of the BH nā»ā! is atelic. According to HALOT 724, it is ‘to carry,’ though I would not rule out the possibility of its being ‘to hold (something) uplifted,’ i. e. nearer to the primary stative meaning of this root in Akkadian, which is probably ‘bei sich haben.’146 145

One could label it “semantically stative,” but this is of course true of the idiom in general. 146 Anticipating the forthcoming part of this study, I would like to mention an important suggestion of BK in a number of p. c.: with much caution, he grants that wabil may have had present-habitual force (the SB, No. 8 above being for him the most cogent example). BK then compares this evidence to grammatically similar usages of naši, e. g. bilassunu na-šu-ni-iš-šum ‘they bring their tribute to him’ (RIME 4.6.8.2:66, Yahdun-Lim of Mari). He concludes, “The instances of naši and babil I mentioned still seem to be a special class of their own of which I do not know any other examples.”

136

Articles: Ancient Near Eastern Studies

I will now compare the shapes of the SC and the verbal nouns of wabālum in various Akkadian corpora, in order to complement the data of GAG 175i–j and Veenhof 1972:440. SC

Part.

wabil

OA, Mari (?), Sarg.

babil

OB (SB)

wābil- Ur III, OA, Mari, Tell-R. bābil-

Inf.

not attested (in particular, not in the core OB)

wabāl- OA,147 Mari babāl- OB

No example of the Part. has been found in the core OB. According to Hilgert 2002:322, 529, the predicative form of wābil- is attested at least once in Ur III Akkadian, within the proper name DINGIR-wa-bil2. The Inf. wabāl- may be attested in Mari at least twice, in broken contexts: i-na wa-ba-[li-im] (ARM 10, 14:10), a-na wa-[ba-lim] (ARM 26/2, 306: 27). In Mari, I know of no inf. babāl-, while this shape is represented by some 13 tokens in AbB, where the wabāl- shape is unknown (the AbB data are from http://www.klinopis.cz/nobtc). Thus, the table shows an isogloss between OA and Mari as against OB: wabāl- vs. babāl-. CAD T 11a claims, “note that in OA tabālu and tabil serve as inf. and stative of wabālu,” yet I have found no corroboration of this idea either in the CAD entry on tabālu or in the searchable corpus of OA (see also the entry on tabālum). tabālum ‘to take along,’148 ‘to carry off.’

147

Cf. e. g. Michel 1991, No. 263:10, cited above as No. 41b (the examples of Inf. were still missing in Veenhof 1972:440). 148 See Kouwenberg 2005:89–93. Using the results achieved in Kouwenberg 2005, BK comments on the drafts of this entry and the preceding one: “CAD starts from the idea that tabālum basically means ‘to take away’ and grudgingly admits that sometimes ‘to take along’ is inescapable. The order is the reverse one: ‘to take away’ is a secondary development arising from ‘to take along.’ As to wabālum/ tabālum, the two partly overlap (cf. the imperatives bilam and tablam), but tabālum is more specific: it is more or less exclusively used with concrete objects and does not have any of the numerous figurative meanings of wabālum. The typical use of tabālum without Vent. is when people acquire things and then ‘take them along’ (to their home or whatever); of course, they also take them away, but I think the focus is not on the separation from their initial locus but on the

S. Loesov, The Suffixing Conjugation of Akkadian …

137

The SC of tabālum is rare, the VA is not attested in the corpus.149 All four tokens of tabil I have found are P, three of them come from OA: (1) i-tù-wa-ar KÙ.BABBAR-pu-šu-nu ta-ab-lu té-er-tum wa-ar-kà-num i-likam um-ma šu-nu-ma ‘[the left-hand context of the letter is broken] … it will return. The amounts of silver belonging to them were carried off. Afterwards their message arrived here, as follows …’ (Prag I 738:5′f.).

The SC form tablū appears in a clear-cut preterital slot. (2) ŠÀ-ba 5 TÚG ur-du-nim 12 TÚG ta-áb-lu ‘[Out of my seventeen garments that had reached the Palace, only] five garments arrived here, twelve garments were taken’ (TMH, 1 19b:12f., a legal decision).

The asyndetic coordination of urdū-nim and tablū favours the same grammatical reading for both verb forms. (3) 6 ša-ru-qú a-na É-tí dA-šur e-ru-bu-ma ša-am-ša-am ša GUŠKIN ša i-ir-tí d A-šur ú pá-a¢-ra-am ša dA-šur 12 [x (x)] a ú mì-šu-ru-um kà-lu-bu-ú sàam-ru-a-tum ù kà-ta-pu ta-áb-lu É-tum lá-qú-ut mì-ma lá e-zi-[bu]150 šari-qí ni-šé-ma lá ni-mar ‘6 Thieves entered Aššur’s temple and 12 [stole?] the golden sun on Aššur’s breast and Aššur’s dagger. The Justice-emblem, the kalubu, the samruātum, and the katappu—(all these) were taken away. The temple has been picked clean, they have left nothing. We keep searching for the thieves but cannot find them’ (Bab 6, 186f., No. 7: 6–19; text of Larsen 1976:261f.).

It is likely that the author of this vivid complaint meant tablū to be RES and observable: the belongings of the god are conspicuously not where they have to be. Consider my discussion of an exceptional case in No. 33 above, našpakum … pate¶-ma še!um leqi ‘the granary … is broken open and the barley is taken’ (AbB 6, 219:14ff.). (4) 0,2.3 ŠE ù 0,0.1 1 SILA3 Ì.GIŠ ta-bi-il! ‘(The writer inspected the condition of his master’s household and found out:) 150 qa barley and 11 qa oil have been taken away’ (MHET 1, 89:13; late OB). connection with their new owner, which is juridically the more relevant one, and this accords with its original indirect reflexive meaning.” 149 In SB, NA and NB, tablu is a substantive ‘stolen goods’ (CAD T 29b). 150 The reading mì-ma lá e-zi-[ib] (with the Ed. pr.) ‘nothing has been left’ can be also defended on morphosyntactic grounds: there are a few P tokens of ezib in OB (Loesov 2006:144) and in OA.

138

Articles: Ancient Near Eastern Studies

The evidence on tabil is compatible with that on wabil in its P readings, except that (as BK notes in fn. 148) tabil demonstrates no polysemy. The marginal existence of tabil is due to the “morphological analogy,” as explained in the previous entry. The dictionaries (AHw. 1297b, CAD T 20b) have no example of the N-stem of tabālum for either the core OB or OA. BK’s Verb List has one token of the N pret. 3ms. i-tí-bi4-il5 (TC 1, 15:29 = OAA 1, 48), analysed differently in CAD N1 277a. All in all, OA tabil appears in the semantic slot of the SIMPLE PAST P for tabālum. Filling a paradigmatic gap once more? šūbulum (OB)/šēbulum (OA) ‘to send.’ Šūbulum is a common causative verb of motion. It has four participants: the Causer (in this case, the sender), the Theme (that which is sent), the Recipient,151 and the Causee (in this case, the one who does the bearing/bringing). All four of them can be expressed in the same sentence whose predicate is a prefixing form of šūbulum. Consider the examples from OA and OB: (1) 1 GÚ 25 MA.NA URUDU ša ište PN alqeu ana PN1 ú-šé-bi-il5-ku-nu ‘One talent and 25 minas of copper which I received from PN, I made youpl. acc. bring to PN1’ (OAA 1, 71:9–12, ref. and tr. BK). (2) (after a list of goods) mimma annîm ana PN nu-šé-bi-il5-kà ‘We have made you sg. acc. bring all this to PN’ (AKT 3, 78:6–11, ref. and tr. BK). (3) ¢uppam šâti ūlū-ma PN ūlū-ma PN1 ana GN šu-bi-il ‘Send this letter to Ešnunna with PNacc. or PN1 acc.’ (ARM 1, 27:12ff., ref. IA, cf. 18, 8:17ff.). (4) luqūtam išti panîm wā´îm ana ´ērī-ni šé-bi4-lá-nim ‘Sendpl. us (ana ´ēr- + vent.) the merchandize with the first (transport/messenger) to depart’ (TCL 19, 69e.:2ff.). (5) anumma 2 MA.NA URUDU uš-ta-bi-la-ak-kum ‘Look, I am sending you two minas of copper’ (AbB 6, 89:4f.).

The Theme is encoded by the Acc., and this is often true of the Causee, if it is expressed at all: in (1)–(3) the Causee is represented by the Acc. of pronouns and nouns, in (4) it is expressed by a prepositional phrase, in (5) it is left unmentioned. Speculatively, it would be possible to form three varieties of the SC of šūbulum, one A (partly preserving the 151

It is understood that the Recipient can be metonymically substituted by the Endpoint of sending (to send to GN = to send to PN), if this is required by context.

S. Loesov, The Suffixing Conjugation of Akkadian …

139

argument configuration of the prefixing tenses) and two P ones, with the Theme (‘something has been sent’) or the Causee (‘somebody has been made to carry/bring’) as the subject. In fact, I have not found satisfactory examples of the SC A in the whole of the corpus. In AbB, the only promising token is not fully reliable because the text is damaged: (6) bukāšum … itti zikir šumi ša LÚ GN [a-na] GN2 šu-bu-l[u x] [… mārī] šipri ¢arādam ukīn ‘The prince decided to dispatch messengers with the gift that the ruler of Aleppo had sent to Babylon’ (AbB 6, 24:9′–14′).

These are the restoration and interpretation offered in the Edition and accepted in CAD T 52a. Yet, as IA suggests to me, a text itti zikir šumi ša LÚ GN [ša a-na] GN2 šu-bu-l[u …] ‘with the gift of the ruler of Aleppo that was sent to Babylon’ would be no worse. The SC P is rare, its subject is always the Theme (an exception is No. 18 at the end of this entry). Below are all five examples found in OA:152 (7) ku-sí-tám a-ta tal-qé um-ma a-ta-ma a-na a-bi-a šé-bu-lá-at ‘You took the kusītum-textile (yourself ), saying: “it has been sent to my father” ’ (POAT 7:12ff., tr. BK). (8) 36 TÚG … šé-bu-lu-ni-kum ‘thirty-six garments … have been sent to you’ (AMMY 2000, pl. IV, No. 4 = kt e/k 68:2–7, CDLI No. P361204). (9) TÚG ša a´´ēr PN u DUMU PN1 šé-bu-lu-ni piqis-sunūti ‘The textile that has been sent to PN and the son of PN1—hand (it) over to them!’ (TCL 4, 13:8ff.). (10) 2 MA.NA KÙ.BABBAR … ša PN ana PN1 … iddin-u-ma ana Ālim ana ši!amātim šé-bu-lu-ma DAM.GÀR wa-du-ú ‘Two minas of silver … that PN gave to PN1 and (that) have been sent to the City to make purchases—the owner (of the silver) has been indicated (on the parcel?) [I am the owner, I have a case-enclosed tablet with PN’s seal, the tablet says that the silver is my silver]’ (TCL 4, 22:3–9).153 (11) i-na té-er-tí-a KÙ.BABBAR a-na na-an-du-im šé-bu-lam

152

I am grateful to BK and J. G. Dercksen who did their best to help me understand these texts. 153 On the agreement rule, see GKT 197e. An A interpretation ‘silver … that PN gave to PN1 and sent [= made (him = PN1) bring (it)] to the City’ cannot be excluded. In the light of the data collected in this entry, it is rather unlikely, however.

140

Articles: Ancient Near Eastern Studies ‘Has the silver been sent to you (vent.) on my instruction in order to let it lie idle?’ (VAS 26, 69:6f., tr. BK).154

In this case, we can explain the evidence by something more tangible than my speculative idea about the “morphological analogy”: the P šēbul is used instead of the Št1 Pret. of wabālum which is not attested in OA (GKT 158; AHw. 1453b; BK’s List) and perhaps in the whole of Akkadian. Moreover, BK points me out in a p. c. that all the alleged instances of Št1 Pres. of wabālum listed in AHw. 1453b have to be relegated to Št2 ‘to do business with, to trade with.’155 This means that šēbul does a good job: it fills an important paradigmatic gap as best it can, coding the idea of “being sent” for the PAST. This fact is a formal correlate of the semantic claims in the theoretical part of this study which make us predict that the P šēbul will not be RES. The situations depicted by the above five tokens of šēbul are hardly observable; these SC forms govern prepositional phrases that code Recipient/Endpoint, which is ill compatible with the idea of RES. The fact they occur in past-time slots that need to be filled anyway squares well with these two non-RES features.156 In the core OB, I found an indisputable token of the SC P only in § 112 of the CH: (12) mimma ša šu-bu-lu ašar šu-bu-lu lā iddin ‘(if) he (= the person entrusted with delivery) did not dispatch (lit. ‘did not give’) what had been sent to where it had been sent,’ the 154 BK informs me, “The sentence is a (rhetorical and indignant) question: the sender wants the silver to be used to make a profit. Veenhof has translated this passage in MOS Studies 1, p. 64f. as ‘Was it my instruction to have the silver sent in order to let it lie idle?,’ which looks as if he took it as an infinitive while at the same time realizing that this is difficult syntactically (he gives no comment).” Hecker (http://www.hethport.uni-wuerzburg.de/altass/) emends the text in order to interpret šé-bu-lam as an infinitive: KÙ.BABBAR a-na na-an-du-im / šé-bu-lam . The emendation is unnecessary. Note that šēbul-am is a rare token of the SC with the directional ventive. 155 BK also informs me that in the OA corpus there is no single token of Št with a P meaning. Does this mean that the stems of the Akkadian verb are no longer a “system” at the beginning of the 2nd millennium B. C.? Note that according to BK’s List, Št lexical is attested in OA for more than ten roots and has a high token frequency. 156 In other words, the writers of OA letters had to render the idea “X was sent to PN,” and they found a way to do so. In agreement with this line of reasoning, the subject of šēbul is invariably the Theme, while the classical passive-of-causative has the Causee for subject. Indeed, it stands to reason that “he was made to bring” is less needed by the language than “it was brought.”

S. Loesov, The Suffixing Conjugation of Akkadian …

141

owner of the sent goods ina mimma ša šu-bu-lu-ma lā iddin-u ukân-šu ‘shall prove that he did not dispatch whatever had been sent’ (CH 25:60–69).

The P šūbul is well-known in OB Mari (all the ref. IA), in stereotyped contexts of subordinate (mostly relative) clauses: (13) ¢uppi PN ša ana ´ēr bēlīya šu-bu-ul ištu GN ikšud-am ‘A letter of PN that has been sent to my lord arrived here from Kurda’ (ARMT 27, 73:6ff.).157 (14) aššum ¢uppi LÚ.ME[Š pā]¢erī adi inanna ana ´ēr bēlīya la š[u-b]u-lu ‘As to the fact that the list of the absentees has not up to now been sent to my lord …’ (ARMT 151:23f.). (15) umma anāku-ma aššum mīnim ¢uppa-ka ša ana ´ēr bēlī-ya šu-bu-lu a-kaal-la ‘I answered him, “Why should I keep your tablet that was sent to my lord?” ’ (ARM 14, 48:44f.). (16) ¢uppi bēlī-ya ša ana GN … šu-[bu]-lu ‘The tablet of my lord that had to be sent to GN’ (ARM 3, 59:5f., and cf. a very similar text 9, 124:10).

Thus, the Mari usage of šūbul is similar to the OA one. Note also the following Mari text (ref. IA): (17) ša adīni šu-ba-al-ta u GEŠTIN.ÚI.A sí-mi-im ša tušābil-am ša ina kunuk PN kank-u ¢āb ‘What you have sent? up to now and the wine of Simum that you sent me, sealed with the seal of PN, is good’ (A.4723:12ff., published as No. 19 in FM 7).

J.-M. Durand thinks (FM 7:50) that “šu-ba-al-ta tient la place d’un šûbulâta attendu,” his translation being A (‘tu m’as fait porter’). In the light of the data of this entry, I do not expect the SC A *šūbulāta ‘you have sent’ to show up in a good OB letter. Hence, šu-ba-al-ta does not represent a phonetic distortion of *šūbulāta158 and, therefore, is no Akkadian verb form at all.159 157 The sporadic lack of subjunctive marker (šūbul rather than šūbul-u) is a feature current in Mari archives (Finet 1955:261). This is doubtless a grammatical error. 158 J.-M. Durand tentatively suggests šubol(a)ta. 159 One solution would be to consider šu-ba-al-ta an active Š-causative (2 m. s. sg. suffixing preterite) formed from the PS root wbl in a language with Š-causative (like e. g. Ugaritic). Since in this case the u-vocalism of the causative prefix (*šuqtala) is

142

Articles: Ancient Near Eastern Studies

Finally, we will have a look at a non-trivial OB example: (18) ullânum inūma bēlni uwa!!er-anniāti abbū ERIN2.ÚI.A-ma ana ´ērī-ka erēbam-ma ¢ēm šitūlī-ka lamādam uwa!!er-anniāti. ERIN2.ÚI.A [= ´ābamacc?] ana libbi ālim šūrubum ul qabi. Ina warkat dūrim ina sa?-we-e ERIN2.ÚI.A ibīat-ma ina šērtim ERIN2.ÚI.A ēma alākim iššer. Birātim ša i-te-né-ru-bu160 u mimma zikir šumim ša šu-bu-la-a-nu ina āl dūrāni nizzibam-ma ištu annīam u annīam nītapl-am tayyartani ana ´ērīka-ma ‘At that time, when our lord instructed us, he instructed us that the army commanders had to approach you and to learn your considerations.161 It was not ordered to bring the troops into the city. The troops shall spend the night on the other side of the city wall in the desert, and in the morning they will head to wherever they have to. All the fortresses that they will enter and whatever gifts we have been sent with we will leave in the stronghold,162 and, after we have discussed among ourselves this and that, our return to you (will take place)’ (BaM 2, 56:12–22; within the letter, this is a piece of reported speech quoted here in its entirety).

difficult to account for (I would expect a-), one could alternatively envisage a borrowing of the Akkadian inf./SC stem šūbul- into a WS language: the newly “backformed” root is šbl, the foreign first-syllable [u] being preserved. The G-preterite receives the dynamic stem-vowel [a], in the way of qatal, qatalta. A partial parallel would be še´i < ušē´i and šezib < ušēzib in Aramaic (HALOT 1993). IA proposes me a third solution: šūbalta < *šawbalta, in the hope that convincing examples of monophthongization will show up. 160 This is K. Veenhof ’s reading, as BK informs me. A. Falkenstein in the Edition has x-bu-ni-a-bu?, ‘Die Festungen des … buni!abu?.’ 161 BK suggests me the following translation of erēbam-ma … lamādam uwa!!eranniāti ‘… he instructed us to let the abbū ´ābim enter in order to learn …’ In terms of overall meaning, this suggestion suits the immediate context pretty well, yet it yields an even more complex syntactic arrangement than the above translation: -anniāti becomes the subject of an implied causative verb, which latter in its turn governs a very complex infinitive clause. As for the inf. complement clause itself, ‘he instructed us that [abbū ERIN2.ÚI.A-ma ana ´ērī-ka erēbam-ma ¢ēm šitūlī-ka lamādam],’ it seems to have no parallel in OB. Structurally, this inf. clause is more complex than its OB second-best dannum enšam ana lā ¶abālim … awâtīya … aš¢ur ‘I wrote … my words … in order that the mighty one not oppress the weak one’ (CH 47:59ff.), which latter inf. clause is no obligatory complement of aš¢ur. Since the former inf. clause is unique, one can only wonder if it is “correct” that its subject is in the Nom. (abbū), while the two infinitival predicates are in the Acc. 162 To save the literal text, BK (p. c.) assumes a zeugma, ‘we will leave all the fortresses they (= ´ābum) will enter, and (we will leave) in the stronghold whatever gifts we have been ordered to bring.’ This then will be a zeugma complicated by a pun based on different senses of ezēbum. My translation presupposes an anacoluthon: the writer inadvertently switched from one grammatical subject to another.

S. Loesov, The Suffixing Conjugation of Akkadian …

143

The SC form šu-bu-la-a-nu was interpreted as A by A. Falkenstein in the Edition (‘und allen guten Ruf, den wir tragen’) and in CAD M2 76a ‘everything which we are transporting’. This is wrong first of all because šūbulum means ‘to send’ rather than ‘to carry, to bring,’ and because the SC A of šūbulum does not exists. Thus, šu-bu-la-a-nu will be the only known P token with Causee as the subject. I believe that BaM 2, 56ff. is a fictitious letter, a purely literary composition. Note that the cohesion of this very long letter is weak, its pragmatic purport is not obvious, which is typical of literary letters.163 This guess would help us digest the syntactic features of BaM 2, 56 that are absent from everyday Akkadian letters, including šūbulānu ‘we have been sent.’ In the forthcoming part of this study, I will analyse, by way of contrast, certain common verbs whose paris shape is very frequent in text, in particular našûm ‘to carry,’ nadûm ‘to throw,’ šakānum ‘to place.’ The burning questions here are the polysemy of respective paris forms and explanations of their lexicalized meanings. I will compare OB and OA data on the SC of the three famous “having” verbs (´abātum, leqûm, and ma¶ārum). I will also address non-resultative tokens of the SC of telic verbs and will elaborate a theory about their functions. We will then have a look at the SC of atelic verbs (cf. the data on akil, No. 54 above) and ask what they are doing in Akkadian. References Albayrak 2004

Arkhipov forthcoming Blau 2010 Buccellati 1968 Buccellati 1996 Charpin 1991

163

Albayrak, I. ‘She Will Live, Eat and Be Anointed together with Them’ ušbat aklat u paššat ištīšunu. Dercksen, J. G. (ed.). Assyria and Beyond. Studies Presented to Mogens Trolle Larsen. Leiden. Pp. 9–20. Arkhipov, I. Le vocabulaire de la métallurgie et la nomenclature des objets en métal (MDBP 3 = ARM 32). Blau, J. Phonology and Morphology of Biblical Hebrew. An Introduction. Winona Lake. Buccellati, G. An Interpretation of the Akkadian Stative as a Nominal Sentence. JNES 27:1–12. Buccellati, G. A Structural Grammar of Babylonian. Wiesbaden. Charpin, D. Un traité entre Zimri-Lim de Mari et Ibâlpî-El II d’Ešnunna. Charpin, D.; Joannès, F. (eds.). Marchands, diplomates et empereurs. Études sur la civilisation mésopotamienne offertes à Paul Garelli. Paris. Pp. 139–166.

On OB fictitious letters see Sasson 1985:250f. (ref. IA).

144

Articles: Ancient Near Eastern Studies

Charpin–Ziegler 2007 Cohen 1984 Dahl 1985 Dercksen 2001

Durand 2001

Durand 2004

Finet 1955 Goetze 1936 Goetze 1942 Goetze 1956 Goetze 1958 Haspelmath 1990 Hasselbach 2005 Hecker 1995 Hilgert 2002 Hopper–Thompson 1980 Huehnergard 1986 Huehnergard 1987 Huehnergard 1997 Izre’el–Cohen 2004 Jaxontov 1983

Charpin, D.; Ziegler, N. Amurritisch lernen. WZKM 97: 55–77. Cohen, D. La phrase nominale et l’évolution du systéme verbal en sémitique. Etudes de syntaxe historique. Leuven–Paris. Dahl, Ö. Tense and Aspect Systems. Oxford. Dercksen, J. G. “When We Met in Úattuš.” Trade according to Old Assyrian texts from Alishar and Boğazköy. Van Soldt, W. H. et al. (eds.). Veenhof Anniversary Volume. Leiden. Pp. 39–66. Durand, J.-M. Une alliance matrimoniale entre un marchand assyrien de Kanesh et un marchand Mariote. Van Soldt, W. H. et al. (eds.). Veenhof Anniversary Volume. Leiden. Pp. 119–132. Durand, J.-M. Peuplement et sociétés à l’epoque amorrite (I). Les clans bensim’alites. Nicolle, Chr. (ed.). Nomades et sedentaires dans le Proche-Orient ancient (CRRAI 46). Paris. Pp. 111–197. Finet, A. L’accadien des lettres de Mari. Gembloux. Goetze, A. The t-Form of the Old Babylonian Verb. JAOS 56:297–334. Goetze, A. The So-called Intensive of the Semitic Languages. JAOS 62:1–8. Goetze, A. The Laws of Eshnunna (AASOR 31). New Haven. Goetze, A. Fifty Old Babylonian Letters from Harmal (reprinted from Sumer, vol. 14). Baghdad. Haspelmath, M. The Grammaticization of Passive Morphology. Studies in Language 14:25–72. Hasselbach, R. Sargonic Akkadian. Wiesbaden. Hecker, K. Rechtlos in der Fremde? AMMK 1995:145– 159. Ankara. Hilgert, M. Akkadisch in der Ur III-Zeit. Münster. Hopper, P. J.; Thompson, S. A. Transitivity in Grammar and Discourse. Language 56:251–99. Huehnergard, J. On Verbless Clauses in Akkadian. ZA 76:218–249. Huehnergard, J. Stative, Predicative, Pseudo-Verb. JNES 46:215–232. Huehnergard, J. A Grammar of Akkadian. Atlanta. Izre’el, Sh.; Cohen, E. Literary Old Babylonian. München. Yaxontov, S. Y. Resultative in Chinese. Nedjalkov, V. P. (ed.). Typology of Resultative Constructions. Moscow. Pp. 67– 60 (Yahontov, S. J. Rezul’tativ v kitajskom jazyke. Nedjalkov, V. P. (red.). Tipologija rezultativnyh konstrukcij).

S. Loesov, The Suffixing Conjugation of Akkadian … Koch-Westenholz 2000

Kouwenberg 1997 Kouwenberg 1998 Kouwenberg 2000 Kouwenberg 2005 Kouwenberg 2010 Kozinskij 1988

Kraus 1984 Kraus 1987 Lambert–Millard 1999 Larsen 1976 Leong 1994 Loesov 2004 Loesov 2005 Loesov 2006 Loesov 2006a Loesov 2010 Loesov 2011

Lyons 1977 Meillet 1934 Metzler 2002 Michel 1991 Michel 1998

145

Koch-Westenholz, U. Babylonian Liver Omens: The Chapters manzāzu, padānu and pān tākalti of the Babylonian Extispicy Series Mainly from Assurbanipal’s Library. Copenhagen. Kouwenberg, N. J. C. Gemination in the Akkadian Verb. Assen. Kouwenberg, N. J. C. Review of Buccellati 1996. BiOr 55: 172–185. Kouwenberg, N. J. C. Nouns as Verbs: The Verbal Nature of the Akkadian Stative. Or 69:21–71. Kouwenberg, N. J. C. Reflections on the Gt-stem in Akkadian. ZA 95:77–103. Kouwenberg, N. J. C. The Akkadian Verb and Its Semitic Background. Winona Lake. Kozinskij, I. Š. Resultatives: Results and Discussion. Nedjalkov, V. P. (ed.). Typology of Resultative Constructions. Moscow. Pp. 497–525. Kraus, F. R. Nominalsätze in Altbabylonischen Briefen und der Stativ. Amsterdam. Kraus, F. R. Sonderformen akkadischer Parataxe: die Koppelungen. Amsterdam. Lambert, W. G.; Millard, A. R. Atra-¶asīs. The Babylonian Story of the Flood. Oxford. Larsen, M. The Old Assyrian City-State and Its Colonies. Copenhagen. Leong, T. F. Tense, Mood and Aspect in Old Babylonian. PhD. Diss. UCLA. Loesov, S. T-Perfect in Old Babylonian: The Debate and a Thesis. B&B 1:83–181. Loesov, S. Akkadian Sentences about the Present Time (I). B&B 2:101–148. Loesov, S. Akkadian Sentences about the Present Time (II/1). B&B 3:133–148. Loesov, S. Marginalia on the Akkadian Ventive. B&B 3: 101–132. Loesov, S. Akkadian Sentences about the Present Time (II/2). B&B 4:759–785. Loesov, S. IPTANARRAS Can Render the “Basic” Present (Akkadian Sentences about the Present Time III/1) B&B 6:149–168. Lyons, J. Semantics. Cambridge. Meillet, A. Le Slave commun. Paris. Metzler, K. Tempora in altbabylonischen literarischen Texten. Münster. Michel, C. Innāya dans les tablettes paléo-assyriennes. Paris. Michel, C. Les marchands et les nombres: l’exemple des Assyriens a Kaniš. Prosecky, J. (ed.). Intellectual Life of the Ancient Near East (CRRAI 43). Prague. Pp. 249–267.

146 Nedjalkov 1988 Nedjalkov 2001

Oshima 2001

Patterson 1970

Reichenbach 1947 Reiner 1970 Roth 1995 Rowton 1962 Sapir 1921 Sasson 1985

Smith 1997 Steinkeller 2005

Streck 1995 Streck 2005 Tropper 1995

Ulshöfer 1995 Ungnad 1918 Vaillant 1948 Veenhof 1972 Veenhof 2008

Articles: Ancient Near Eastern Studies Nedjalkov, V. P. (ed.). Typology of Resultative Constructions. Amsterdam. Nedjalkov, V. P. Resultative Constructions. Haspelmath, M. (ed.). Language Typology and Universals. An International Handbook. Vol. 2. Berlin–New York. Pp. 928–940. Oshima, T. Two Notes on Expressions with ina pī … “in the mouth of …” in “Prayer to Marduk 1.” NABU 2001/15:15–18. Patterson, R. D. Old Babylonian Parataxis as Exhibited in the Royal Letters of the Middle Old Babylonian Period and in the Code of Hammurapi. PhD. Diss. UCLA. Reichenbach, H. Elements of Symbolic Logic. New York. Reiner, E. Akkadian. Sebeok, T. A. (ed.). Linguistics in South West Asia and North Africa. The Hague. Pp. 274–303. Roth, M. Law Collections from Mesopotamia and Asia Minor. Atlanta. Rowton, M. B. The Use of the Permansive in Classic Babylonian. JNES 21:233–303. Sapir, E. Language: An Introduction to the Study of Speech. New York. Sasson, J. M. Yarim-Lim’s War Declaration. Durand, J.M.; Kupper, J.-R. (eds.). Miscellanea Babylonica (FS Maurice Birot). Paris. Pp. 237–255. Smith, C. The Parameter of Aspect. Dordrecht. Steinkeller, P. Of Stars and Men: The Conceptual and Mythological Setup of Babylonian Extispicy. Gianto, A. (ed.). Biblical and Oriental Essays in Memory of William L. Moran. Roma. Pp. 11–48. Streck, M. Zahl und Zeit. Grammatik der Numeralia und des Verbalsystems im Spätbabylonischen. Groningen. Streck, M. P. Akkadisch. Streck, M. P. (ed.). Sprachen des Alten Orients. Darmstadt. Pp. 44–79. Tropper, J. Die semitische “Suffixkonjugation” im Wandel. Dietrich, M.; Loretz, O. (eds). Vom Alten Orient zum Alten Testament. Festschrift für Wolfram Freiherrn von Soden. Neukirchen-Vluyn. Pp. 491–516. Ulshöfer, A. M. Die altassyrischen Privaturkunden (FAOS. Bh. 4). Stuttgart. Ungnad, A. “Haben” im Babylonisch-Assyrischen. ZA 31:277–281. Vaillant, A. Manuel du vieux slave. Vol. 1. Grammaire. Paris. Veenhof, K. R. Aspects of Old Assyrian Trade and Its Terminology. Leiden. Veenhof, K. R. The Death and Burial of Ishtar-lamassi in Karum Kanish. Van der Speck, R. J. (ed.). Studies in the Ancient Near Eastern World View and Society (FS M. Stol). Bethesda. Pp. 97–119.

S. Loesov, The Suffixing Conjugation of Akkadian … Veenhof 2008a

Voigt 2002–2003 Yaron 1988

147

Veenhof, K. R. Communication in the Old Assyrian Trading Society by Caravans, Travelers and Messengers. Michel, C. (ed.). Old Assyrian Studies in Memory of Paul Garelli. Leiden. Pp. 199–246. Voigt, R. Die beiden Suffixkonjugation des Semitischen (und Ägyptischen). ZAH 15/16:138–165. Yaron, R. The Laws of Eshnunna. Jerusalem–Leiden.

IPTANARRAS Can Render the “Basic” Present (Akkadian Sentences about the Present Time. III/1)* Sergey Loesov Russian State University for the Humanities, Moscow∗

For Bert Kouwenberg In this note, I am trying to show that the present-time sense of the G-stem verbs can be rendered by one of the three exponents: IPARRAS, PARIS, or IPTANARRAS, depending on the verb’s lexical meaning. The corpus of my study in progress consists of both OB and OA. I will now put forward a preliminary observation that is important in order to understand what follows. We do not know for sure which verb forms were used in spoken OB and OA to say “He is now crossing the river,” or “He is now felling a tree,” because the epistolary sources did not preserve enough instances of clearcut telic events in progress that are contemporaneous with the speech time or another temporal reference point. This is partly due to the nature of the sources: for communication via letters, “quick” actions are irrelevant; and partly because Akkadian does not have a FUTURE conjugation neatly opposed to a PRESENT conjugation (unlike e. g. te lo digo vs. te lo diré or te lo voy a decir in Spanish: Spanish has two unambiguous futures, while Akkadian has none). Consequently, nearly all the examples of present-time sentences available to us in the Akkadian corpus are of a more or less semantically “stative” nature: there is no change, no action under way just when the writer is penning his sentence about the present time. ∗ In this note, SMALL CAPS stand for universal semantic notions (e. g., PRESENT). Terms for language specific morphological paradigms are written with first capitals: the (German or Akkadian) Preterit, the (English) Present Progressive, etc. The morphological shapes of Akaddian finite verb forms are written in ALL CAPS: IPARRAS, PARIS, IPTANARRAS, etc. The SC = the Suffixing Conjugation. This is my label for the Akkadian Stative. I do not use the latter term because the word “stative” often appears here as a semantic notion. I am grateful to Ilya Arkhipov and Bert Kouwenberg who read various drafts of this note and made numerous useful suggestions. The research has been supported by RFBR grant No. 11-06-00418-а.

150

Articles: Ancient Near Eastern Studies

At this early stage of research into the meaning of tan-stems,1 I tentatively propose that a given G-verb usually (but not always) has only one exponent for the present-time sense (IPARRAS, PARIS, or IPTANARRAS), unlike e. g. in English: “He is building a house these days” vs. “Each time he gets divorced he builds a house.” Thus IPTANARRAS is arguably the only morphological hardware to render the present-time sense for certain verbal notions. This suggestion will help explain the high frequency of Gtn. In the whole of Akkadian corpus, the ratio of G-stem to Gtn-stem is almost 4:1, 1316 G-verbs vs. 312 Gtn-verbs in AHw., as reported in Kouwenberg 2010:246. (Kouwenberg observes that the Present IPTANARRAS is the most frequent among the morphological shapes of Gtn.) 2 Building on the time-honoured tradition, Kouwenberg defines the function of tan-stems as “verbal plurality,” or “pluractional” sense, most often materialized as habitual and iterative readings (Kouwenberg 2010, §§ 10 and 14). After I came to the idea that IPTANARRAS can render the G-stem PRESENT, I suspected that the traditional interpretation is vulnerable from the statistical point of view: a fine-tuned aspectual nuance like verbal plurality should not require so much derivational effort from Akkadian, because the G Present IPARRAS itself can have the very same force (i. e., habitual/iterative) in the non-future domain.3 1

In this preliminary report, I speak about the Gtn-stem only, simply because this is the natural starting point. Kouwenberg 2010:416 observes that “the Gtnstem … is by far the most frequen tan-stem, but the other tan-stems have the same function vis-à-vis the corresponding primary stem.” 2 He states that this is also true of other tan-stems: “A noteworthy feature of the tan-stems is that the imperfective (= the Present.—S. L.) is far more frequent than we would expect on the basis of the normal frequency rates among the tenses. [T]he great majority of tan forms are imperfective forms” (Kouwenberg 2010:417). For Kouwenberg, the reason is a semantic one: the Preterit “has a strong association with one-off, real, and completed events in the past, which poorly matches the pluractional function of the tan-stems and causes a tendency to switch to the imperfective, which, after all, can be used in past contexts to underline the non-completed or repetitive nature of the event” (ibid., boldface added). 3 See also the previous footnote on “imperfective … used in past contexts.” The iterative-frequentative derivational rules do exist in certain well-understood languages, such as Russian (ходил—хаживал) and Latin (traho—tracto; dico—dicto). These rules are/used to be productive for some semantically defined parts of the verb lexicon. An immediate comparison of these word-formation patterns in Russian and Latin with the Akkadian tan-stems is hardly viable. The inflectional and derivational morphologies of the three languages are very different as far as the semantic side goes. In particular, both Russian and Latin (unlike Akkadian) do

S. Loesov, IPTANARRAS Can Render the “Basic” Present …

151

This is not to deny that Gtn can express “verbal plurality,” along with other tan-stems. Yet my current task is to suggest that the picture may be more complex than we are accustomed to believe. Now, I will briefly put down three kinds of observations that prompted me to forward this claim. 1. For certain G-verbs, IPARRAS usually has future reference, while IPTANARRAS has the present-time meaning. An example is šakānum in the meaning ‘to store.’ The following OA text will demonstrate my point: miššu ša ina bē[tīya] tibnam u e´´ī taštanakkun-u bētātī-kunu lā tadaggalā ‘Why is it that you (ms) store straw and wood in my house? Don’t you (pl) own houses of yourselves?’ (BIN 6, 119:17ff.).

To avoid misunderstanding, I can paraphrase: ‘Why is it that you keep straw and wood stored in my house?’ I believe that a sentence miššu ša ina bētīya tibnam tašakkun-u would hardly render this idea. The verb is extremely frequent in the corpus, and all the G-stem Present forms I have checked have the future value. Thus, miššu ša ina bētīya tibnam tašakkun-u would probably mean ‘Why are you going to store?’ From the above reasoning it follows that the morphological contrast of taštanakkan (Gtn) and tadaggalā (G) ‘you own’ in BIN 6, 119:17ff. corresponds to the fact the two verbs belong to different semantic (= Aktionsart) classes, i. e. šakānum in the meaning ‘to store’ is probably considered by OA as more dynamic than dagālum in the meaning ‘to own,’ though at this stage of research I cannot pinpoint the culprit parameter of lexical meaning. This would mean that for certain semantic classes of verbs IPTANARRAS was perceived by the language as more expressive (and therefore appropriate) exponent of the present-time sense (within the G-stem paradigm) than IPARRAS or PARIS. In a preliminary way, I suggest the following hypothesis. If there is a likelihood that a present-time situation encoded by IPTANARRAS is not thought of as cyclic (i. e., regularly repeated = “immer wieder” of AHw. or “iterative” of CAD) but rather durative (denoting a continuing action), then there is a chance that this IPTANARRAS token, semantically, is part of the G-stem paradigm. This likelihood increases if IPARRAS of this root always refers to the future, while its IPTANARRAS is not attested with future-time readings and other Gtn forms of this root are appreciably not use their basic-stem Present (Kouwenberg’s “imperfective”) in the past-time domain with the iterative force.

152

Articles: Ancient Near Eastern Studies

rarer than IPTANARRAS (especially, if IPTANARRAS is frequent while the Precative and Imperative are rare or not attested).4 The principal reason of this shift may have been the desire to renew (and to make more semantically sensitive) the morphological expression of the present-time meaning. Concomitantly, the language came across a way to oppose FUTURE to PRESENT for a part of the verb lexicon (išakkan vs. ištanakkan). As Ilya Arkhipov tells me, the assumed inclusion of the Present of the tan-stems into the paradigms of their respective motivating stems has a partial parallel in the history of Akkadian: consider the t-Perfect IPTARAS which is formally identical to the Preterit of the Gt-stem, etc. The differences are of course also obvious. In particular, the t-Perfect is a brand-new “tense,” probably an East-Semitic innovation vis-à-vis the Proto-Semitic, and it is supposed to be formed freely for all verbs of G-, D-, and Š-stems, with no serious semantic limitations.5 As for the G-stem IPTANARRAS of certain roots, as a present-tense form it is grammatically synonymous to both IPARRAS and PARIS of some other roots, and it is lexically restricted. Yet if we think through this striking parallel, it will probably shed new light on the historical ways of the Akkadian verb. Note that both IPTARAS and, admittedly, IPTANARRAS became elements of the G-stem paradigm due to semantic developments (one is almost tempted to say, “semantic decisions”) rather than as a result of phonological processes or other diachronic shifts that are blind to the meanings of morphological shapes. Now consider an example from an OB diagnostic text: šumma mar´um … libbašu iš-ta-na-¶i-i¢-ma qātam i-sà-ki-ip mar´u šū ul iballu¢ ‘If … a patient’s heart is fluttering and he is thrusting aside his hand, this patient will not get well’ (TLB 2, 21:11ff.; cf. Heeßel 2000:97f.).

4

Clearly, this hypothesis can be applied only to frequent roots with well-attested Gtn. 5 Kouwenberg 2010:141, fn. 8 notes that “the t-perfect is not, like the stative, restricted to telic verbs: prototypically atelic verbs, such as to dance and to walk around, are also used in the t-perfect.” There are of course a few well-known exceptions in the basic verb lexicon (Goetze 1936:317). It would be worth-while to compile a list of basic roots denoting properties (‘to be good/bad,’ ‘to be wide’), as well as a list of common stative and atelic dynamic roots, and then to look at their t-Perfect.

S. Loesov, IPTANARRAS Can Render the “Basic” Present …

153

Let us assume that the above English glosses (found in CAD) are appropriate in terms of lexical semantics. The situations of “fluttering” and “thrusting aside” are simultaneous to the moment of observation. Most probably, none of them was thought of as conspicuously cyclic, “immer wieder.”6 My only point here is that this co-occurrence of IPARRAS and IPTANARRAS in TLB 2, 21:12 (as in so many other texts) is something that needs explanation, while most of the time we manage to overlook the problem by maintaining conventional Assyriological translations, something like ‘keeps fluttering’ (CAD Š1 92a) vs. ‘is thrusting aside.’7 Now, the sakāpu A of CAD (S 70–74)8 does not have tan-stems at all. According to CAD, it has G, Gt, D, Dt, and N-stems (AHw. 1011 does not mention Gt), all of the derived stems are only marginally attested in OB and in the whole of our Akkadian records.9 To my knowledge, the root is not attested in Assyrian.10 Kouwenberg 2010:252 thinks that all G-verbs “can have a pluractional Gtn-stem, although it is attested only for a minority of them.” If this suggestion is generally speaking correct, it would be worthwhile to explain why a given root that is frequent in the G-stem has no Gtn (or D as another possible pluralic derivate for transitive roots, see presently).11 The data about ša¶ā¢u A (CAD Š1 88–92) are quite different. Its basic meaning is probably ‘to jump.’ It is almost unknown in the OA corpus, but well-represented in OB. In OB corpus, the distribution of iša¶¶i¢ (future) and ištana¶¶i¢ (non-future) roughly agrees with my zero approximation prediction. Consider the examples:

6

Both the patient’s heart beating and something like the tremor of his hand cannot be either cyclic or otherwise. 7 They are “conventional” because if we say ‘is fluttering’ vs. ‘keeps thrusting aside,’ nothing will really change from the point of view of English. 8 Its basic meaning is something like ‘to push.’ All the G-stem meanings are transitive. 9 E. g., the D-stem has only one reliable (and semantically transparent) token ‘to refuse, to reject’ (AbB 1, 29:22) out of the three possible ones (CAD S 73f.). 10 In this note, much of my data on the OA verb depend on Kouwenberg’s unpublished “List of OA Verbs” (see also Loesov 2011:75–147). 11 When I say “If this suggestion is generally speaking correct,” I mean that as a matter of fact we are not going to take Kouwenberg’s words all too literally. If a property verb like damāqu ‘to be(come) good’ or a stative verb like edû ‘to know’ has no Gtn, this comes as no surprise. By contrast, if a number of frequent dynamic roots do not have a pluractional derivate (Gtn or D), this leads to questions both about the nature of verbal plurality in Akkadian and the functions of Gtn.

154

Articles: Ancient Near Eastern Studies F u t u r e: nēšum ana tarba´ awīlim i-ša-a¶-¶i-i¢ ‘a lion will raid the man’s fold’ (YOS 10, 25rev.:70; for additional OB omen apodoses with this verb form, see CAD Š1 90). N o n - f u t u r e: wa-ar-ki ši-i¶-¢i4-im ša a-lim Nu-sa-arKI iš-¶i¢ú LÚ´a-bu-um a-na sa-ak-bi-im it-ta-na-Ya´-´úZ ù a-na Ka-ra-na-aKI {x} i-tu-ur-ra-am [DU]MU ši-ip-ri-im ša Iš-me-dDa-gan il-li-kam-ma [u]m-maa-mi ú-ul ni-nu ni-iš-¶i-i¢ mi-id-de sà-ar-ra-ru iš-ta-na-a¶-¶i-[¢ú] ‘[Now,] after they (= the inimical forces in question) raided the town of Nusar, the troops go out for patrolling12 and come back [iturr-am] to the town of Karana.13 A messenger of Išme-Dagan came here with the following news, “It is not us who did the raiding. Perhaps it was certain criminals who were raiding” ’ (ARM 26/2, 515:4–9; Iddiyatum to Zimri-Lim; for additional OB examples, see CAD Š1 91f.).14

From the letter’s contents, it is likely that both verb forms, niš¶i¢ and ištana¶¶i¢ū, refer to the same real-life event in the writer’s past. The only difference is the speaker-dependent viewpoint: for certain semantic classes of verbs, Akkadian allows to interpret the very same past-time fact as either punctual (niš¶i¢) or possessing an internal duration (ištana¶¶i¢ū). In the latter case, both IPARRAS and IPTANARRRAS can be used, once more depending on the root’s meaning. It lies near at hand to ask if ša¶ā¢u A is frequentative by the very nature of its semantics, with the result that OB deemed it useful to employ this root as a Gtn-tantum in the present-time domain. In this case, one would have to speculate (perhaps with a good reason) that this frequentative semantics of the root was cancelled in the future-time sphere. Some such conclusion, if it be correct, would contribute to our understanding of the verb in Akkadian anyway. In particular, this would mean that ša¶ā¢um ‘to jump, to attack, to raid’ and e. g. šapārum ‘to send a message’ belong to different Aktionsart classes, which follows from the difference in their morphosyntactic behaviour: unlike in the case of ša¶ā¢um, the G-stem of šapārum can be used as frequentative in the non-future domain. Consider the examples: aššum mīnim ana šeim i-ša-ap-pa-ru-ni-ku-ma šeam lā tanaddin ‘Why do they write you concerning barley but you do not give any barley?’ (Goetze 1958, No. 12:4ff.). 12

See LAPO 18:583 with further references. Cf. CAD S 22f. The writer’s location is supposed to be this very town of Karana. 14 The reading, including the emendation , follows the Edition, see ARM 26/2, p. 487, note a. 13

S. Loesov, IPTANARRAS Can Render the “Basic” Present …

155

panānum awīlû ša lā īdû-ninni a-ša-ap-pa-ar-šu-nu-ši-im-ma ´abiātīya ippušū ‘Formerly, (even) people who did not know me—I had (only) to write to them, and they used to act according to my wishes’ (AbB 10, 1:13–16).

One then would have to ask why the Present Gtn of šapārum is also productive in OB (see AHw. 1171). I am not ready to compare right now all the non-future išappar tokens to all the ištanappar ones, because the evidence is both huge and perhaps not really very dramatic: both forms are probably frequentative, as one would expect. Instead, I will try to produce a putative example of IPARRAS and IPTANARRAS of the same root in the non-future domain, both shapes being non-frequentative and either synonymous or possessing a certain fine-tuned differentiation that is difficult to grasp with any certainty for this dead language. The verb is adāru B ‘to fear’ of CAD A1 108, adāru B ‘(sich) fürchten’ of AHw. 11.15 Consider two well-known passages from the OB Gilgamesh (tr. A. George): a-wi-lu-tum-ma ma-nu-ú u4-mu-ša mi-im-ma ša i-te-né-pu-šu ša-ru-ma at-ta an-na-nu-um-ma ta-dar mu-tam ‘As for man, his days are numbered whatever he may do, it is but wind Here are you, afraid of death!’ (George 2002:200, 142ff.). i-na-an-na sa-bi-tum a-ta-mar pa-ni-ki mu-tam ša a-ta-na-ad-da-ru a-ia a-mu-ur ‘Now, ale-wife, I have seen your face, but I would not see death, that ever I fear’ (George 2002:278, 12′f.).

This a-ta-na-ad-da-ru seems to be the only Pres. Gtn token of the root attested in the core OB (AHw. has a few more tokens, all of them are jB, there are no examples in OA). I am tempted to suspect that the reason of its appearance may be metre-related: in the surrounding text, the number of syllables pro line is 11–14, being exactly eleven more often than otherwise (and compare von Soden’s transcriptions in ZA 71:184f.). It looks like mū-tam ša ad-da-ru ay-yā-mur (nine syllables only, if ay-yā-mur is the correct rendering) would be too short a line.16 15

For the etymologies of adāru A ‘to be worried’ and adāru B ‘to fear,’ see Kogan 2001:279f. 16 Generally speaking, isosyllabism is not considered to be a relevant feature of metre in the Akkadian poetry (see Wasserman 2003:159–162, with a literature overview). I am grateful to Rim Nurullin for discussing this point with me.

156

Articles: Ancient Near Eastern Studies

The only other example comes from the OB Mari (courtesy Ilya Arkhipov): ´ābum šū … bēlī i-ta-na-ad-da-ar ‘These soldiers … fear my lord’ (Charpin 1993, 200:53).

If we do not discard the Gilgamesh example as poetry and the Mari example as a barbarism,17 and if we are not ready to conceal the problem by simply translating ‘fear very much,’ ‘experience great fear,’ etc., then I would tentatively suggest that ītanaddar is different from iddar in that the former represents the situation of fearing as more dynamic and progressive.18 Another manifestation of the fact that “the problem of Gtn” does exist is the lexicalization of Gtn. An example is nagāšu of CAD N1 108. Its G-stem Preterit is attested a couple of times in SB with the meaning ‘to leave, to go away.’ (The only other form of the G-stem mentioned in this CAD entry is the Infinitive appearing in a Sumero-Akkadian lexical list.). The Gtn-stem of this root is attested by a dozen of tokens in the Present and Preterit, mainly in OB and SB. The CAD glosses are ‘to wander around, to run about, to rove, stroll.’ The best-known example runs as follows: at-ta-na-ag-gi-iš ki-ma ¶a-bi-lim qá-ba-al-tu ´e-ri ‘I was wandering like a criminal in the midst of the steppe’ (George 2002:278, 11′).

The sense is most probably not cyclic (i. e., unless the speaker presents himself as the one who used to run to and fro along the same path); it is simply durative, fully in agreement with the plot of the Gilgamesh-epic. Consider another example (courtesy Ilya Arkhipov): lā bēl isqim ana kussi bīt abīya īrub u anāku at-ta-na-ag-gi-iš ‘The one who has no share (= no inheritance right) entered the throne of my father’s estate, and as for me I keep wandering [≈ I 17 This letter of the provincial governor Yaqqim-Addu has a nice example of the adstrat-conditioned verb morphosyntax: ´a-bu-um … e-pé-ri a-li-ku-tim a-na alimKI i-ša-ap-p[a]-ak ‘the soldiers … are heaping up a mound that is advancing towards the city’ (ibid. l. 25). The participle ālik- here is not an agent noun (as it should be in good Akkadian) but rather something very similar to the Hebrew or Aramaic participial Present. 18 Kouwenberg 1997:293 says that adāru B ‘to fear’ “is basically a fientive verb,” unlike adāru A ‘to be dark, worried.’ (Kouwenberg’s “fientive” is my “dynamic.”) Note, however, that at least the English verb ‘to fear’ is stative in terms of Vendlerian classes: one can say “Right now, I fear him,” but it is wrong to say *“Right now, I am fearing him.”

S. Loesov, IPTANARRAS Can Render the “Basic” Present …

157

am still dispossessed]’ (Charpin–Durand 2004:100, 27ff.; see a historical commentary ibid. 106f.).

To sum up: it would be important to understand by which factors the choice of form (IPARRAS vs. IPTANARRAS) in the present-time domain is determined. Ideally, if we want our picture of Gtn to be more than impressionistic and textbook-dogmatic, we need to work with a complete list of verbs that have both G and Gtn + to register systematically the attested forms of Gtn (is there the Imperative? the Precative? the Prohibitive? the Preterite?) + to identify wherever possible the IPTANARRAS forms with future reference. We will probably come across a certain number of minimal pairs IPTANARRAS = present-time reference, IPARRAS = future reference. To complicate things even more, there exist frequent G-verbs that use PARIS rather than IPARRAS or IPTANARRAS to render the dynamic present-time sense.19 An OB example is qerēbum ‘to be near; to get nearer; to approach.’ Its Present iqerrib is used only for future references,20 while the real-life situations ‘it is near’ and ‘it is getting nearer’ are rendered indiscriminately by the SC qerub (OA qurub). Consider an example where the real-life situation is probably dynamic: d

19

UTU-ši-dIM

u ummānātū-šu qé-ru-ub ¢ēmum annûm li¶mu¢

See Loesov 2005 and 2010 for various attempts to single out such verbs. The only OB token I know of comes from an apodosis in YOS 10, 11 ii 14– 17: tu-ru-ku-tum ana šarrim i-qé-er-ri-bu-nim ‘the Turukkeans will approach the king.’ See my analysis of the morphosyntax of qerēbum in Loesov 2010:775–779. Note that according to my description, the finite forms of qarābum in OA are limited to two tokens of the Precative + five tokens of the SC (Kouwenberg’s list of OA verbs has seven tokens of the SC but no additional prefixing forms). The Preterit of qerēbum is also in very short supply in OB (ibid. 777). In Loesov 2010, I try to explain these remarkable data by the following suggestion: qerēbum is a hopelessly atelic root, therefore it is ill compatible with the prefixing tenses which by default process motion events from the perfective aspectual viewpoint. It looks like even the intransitive D-stem (‘to approach’ + ana and adi to introduce the endpoint of intransitive motion) does not really help: according to CAD Q 239b, its prefixing tenses appear but twice in the corpus, both times in Mari letters: ana GN ú-qé-ri-ba-am (ARM 4, 26:29), ša adi GN ú-qa-ar-ra-bu (ARM 2, 134:15). Note that even the prefixing indicative tenses of damāqum ‘to be(come) good’ (a property root that expresses evaluation) are represented in the letters from the corpus better than those of qerēbum, which is generally agreed upon to be a motion verb. The instances I have found are as follows: i-da-am-mi-qú (AbB 14, 165:9); ad-mi-iq (AbB 3, 39:19); OA: ni-dam-mi-iq (RA 60:120, 28); i-da-am-qa (CCT 5, 1b:11); admi-qa-ku-ma (ATHE 44:26). (The OA data are from Kouwenberg’s Verb List.) 20

158

Articles: Ancient Near Eastern Studies ‘Šamši-Adad with his troops is approaching/is near; (therefore) this message is to be delivered rapidly’ (AbB 8, 15:40–45).

An OA example is the SC of našā!um + vent. ‘to bring,’ discussed in Loesov 2005:128f.: naš!am/naš!akkum = ‘he is bringing to you/he is supposed to bring to you.’ Thus, ištanakkan, idaggal, qerub are the only exponents of the presenttime sense for the corresponding G-verbs in their meanings ‘to store,’ ‘to own,’ ‘to approach.’ 2. As we have just seen, the meanings of tan-stems of the Akkadian verb remain to be described in detail,21 so we do not know for a fact whether the Present of the tan-stems invariably expresses the “verbal plurality” as opposed to a “non-plurality” of the Present of the respective simple stems for the same roots (i. e., for the same verbal concepts). Our translations in text editions, textbooks, dissertations, and dictionaries are based on an ingenuous faith in this plurality. We hardly ever try to ask ourselves if this semantic element is indeed there in a given case. Regarding the assumed productivity of the Gtn-stem in the pluralic sense, I observe that all common verbs that show the regular “competition” of the Present and the SC in the present-time domain22 do not really have the Gtn-stem in the corpus. This is hardly incidental to the Akkadian verb. By way of illustration, consider a textual example of what I call “competition” between PARIS and IPARRAS, selected from the data gathered in Loesov 2006:142, where the morphosyntax of akālum ‘to eat’ was analysed: PARIS: [in]a mimma annîm [a]k-la-ku ‘I am entitled to the use of all this,’ lit. ‘I am consuming all this’ [or, better still though stylistically impossible, ‘I habitually consume …’] (AS 22, 26:5; “archaic” OB).

21

The best semantic analysis of the verbal stems in Akkadian is now Kouwenberg 2010, yet the purpose of this work is encyclopedic, the book is meant to be a reference tool for some time to come, i. e. a balances description of the Akkadian verb. It includes formal and semantic, synchronic (wherever possible, separately for OB and OA), diachronic inner-Akkadian (roughly 2500–700 B. C.), and comparative analyses (against the Semitic and Afroasiatic background), with the aim of a ProtoSemitic reconstruction. A comprehensive semantic description of the verb stems for OB and OA with methods of the corpus linguistics has not yet been undertaken. This kind of research would certainly deepen our understanding of the language. 22 For a provisional list of such verbs see Loesov 2010:761, fn. 7. I thank Bert Kouwenberg who suggested me (in his p. c. back in 2006) the elegant word “competition” as a label of this phenomenon.

S. Loesov, IPTANARRAS Can Render the “Basic” Present …

159

annakam emārū ak-lu ‘The donkeys get fodder here’ (ATHE 46:17f., and a few more times in OA). IPARRAS: eqlam … i-ik-ka-al ‘He supports himself with this field’ (AbB 4, 79:19, and often in Hammurapi-letters).23

We will now discuss the data related to the Gtn and, wherever necessary, to the D-stem of this group of verbs. I will first have to quote one of the conclusions that Kouwenberg reached in his study of the Akkadian derived stems: “In all second millennium dialects, the Gtn-stem is fully productive. It is especially frequent in two kinds of verbs: intransitive action verbs,24 in particular motion verbs such as alāku ‘to go/come,’ elû ‘to go/come up,’ erēbu ‘to enter,’ ebēru ‘to cross,’ etēqu ‘to pass,’ maqātu ‘to fall,’ and wa´û ‘to go/come out,’ and transitive verbs with a low degree of transitivity, e. g., amāru ‘to see,’ akālu ‘to eat,’ apālu ‘to answer,’ leqû ‘to receive,’ nadû ‘to drop, leave behind, lay down,’ našû ‘to lift, carry,’ šakānu ‘to place,’ šemû ‘to hear,’ wabālu ‘to bring, carry,’ and warû ‘to bring, lead.’ These are exactly the same types of verbs that do not normally have a D-stem (see § 11.5, p. 279). So there is a complementary distribution between Gtn and D in its function of underlining verbal plurality: Gtn takes the area of intransitive and low-transitivity verbs, D that of high-transitivity verbs” (Kouwenberg 2010:421, italics added).

Following the lead of Kouwenberg, we will check the D-stem of transitive roots from my list as well (i. e., in addition to the Gtn-stem), though of course we do not expect high-transitivity verbs to reveal the competition of PARIS and IPARRAS. Here are the data. – akālum ‘to eat,’ ‘to receive victuals/fodder.’ No Gtn in OA; in OB it has been found twice within the idiom kar´ī akālum ‘to slander’ (ARM 2, 115:11; YOS 10, 41:56), once with another metaphorical meaning, ‘to hurt (of ailing body parts)’—in TLB 2, 21:23 (an OB diagnostic manual). Thus, the “pluractional” or habitual eating (consumption, getting sustenance) is expressed by the Present and the SC of the G-stem of akālum rather than by its Gtn-stem (see the examples above and in 23

The differences in the government of akālum in the three examples are doubtless significant, but we cannot discuss them here. 24 “Action verbs” is Kouwenberg’s word for agentive verbs (see p. 256f.). To my mind, intransitive verbs from this list are not always impeccably agentive, especially maqātu ‘to fall’ is not.

160

Articles: Ancient Near Eastern Studies

Loesov 2006:142). Note that ikkal and ītanakkal ‘it hurts’ were likely used synonymously in OB medical texts: šumma mar´u qātā-šu u šēpā-šu ikkalā-šu šitassâm mimma lā pa¢er ‘If the patient’s hands and feet ache and he never stops screaming’ (TLB 2, 21:19, tr. CAD Š2 164).25 šumma mar´u […] libbīšu ītanakkalā-šu (i-ta-na-ka-la-a-šu) ‘if the patient’s [somethings = fem. pl. substantive] inside him hurt him all the time’ (TLB 2, 21:23).

There is no D-stem. – baqārum ‘to claim’26 has no Gtn in the whole of Akkadian. Its D-stem, glossed in CAD P 133b ‘to raise a claim, to lay claim to,’ is used “passim in NB sales contracts and kudurrus” (CAD ibid.); besides, CAD P 134a enters two MB tokens with the same meaning. It further lists three OB tokens of the D-stem meaning ‘to contest, challenge’ (a legal text TCL 1, 157:49; AbB 12, 166:13; 1, 58:16); this meaning is also attested in later periods. Thus baqārum, a verb common in OB, has in OB neither Gtn- nor D-stem corresponding to its basic meaning ‘to claim (a property title).’27 – ¶ašā¶um ‘to need,’ ‘to desire’ (AHw. ‘brauchen’, ‘begehren’)28 has no Gtn in the whole of Akkadian. Its D-stem is represented in AHw. 333a and CAD Ú 136a by two SC passive (!) tokens (OB and SB), glossed ‘to deprive, take away’ (AHw. ‘in Bedürftigkeit bringen’), which is irrelevant for our discussion.

25 Note also the periphrasis šitassâmInf. Gtn mimma lā pa¢erSC ‘he never stops screaming’ rather than a straightforward ištanassiPres. Gtn. In the context, lā pa¢er has a “pluractional” reading. 26 For the competition between the Pres. and the SC of this verb in the present-time domain, see Loesov 2006:142. 27 One would suspect that CAD divorced ‘to claim’ (G) from ‘to contest, challenge’ (D) simply because the stems are different, but one of these three examples does deviate from the well-known use of baqārum G in the official OB letters: aššum kunukkī-ša ú-ba-aq-qí-ru ‘because she contested her own sealed documents’ (TCL 1, 157:49, tr. CAD P 134b). AbB 1, 58:16 has an Imperative bu-uq-qú-ur-šu in a badly broken context. AbB 12, 166:13 tu-ba-qá-ra-an-ni ‘you claimed from me (the aforementioned field)’ is not really different from the G-stem examples with two participants. 28 For the examples of both the Pres. and the SC in the present-time domain, see Loesov 2005:137f.; 2006:145.

S. Loesov, IPTANARRAS Can Render the “Basic” Present …

161

– kalûm ‘to hold, detain.’29 This verb is common in both OB and OA, its Gtn is registered in the dictionaries with two identical SB tokens from the same medical text, and there is also one instance from NA. The CAD (K 102b) gloss for the Gtn is ‘to stop repeatedly, to hold up.’ For the D-stem *kullû ‘to hold back,’ CAD K 102–103 has only four tokens (from OA, OB, MB and SB), all of them are the Pres. forms. The OB example tu-ka-la-šu (Sumer 14, 73, No. 48:8) stands in a difficult context,30 therefore the case (dat. or acc.) and antecedent of the -šu pronoun are problematic. Goetze translates in the Edition, ‘[But (as to) getting water on my time—] you will prevent it’, CAD K 102b has ‘[but at the moment it is irrigated,] youm. sg. will block up (the water) for him.’ On any interpretation, this sign string is plausibly construed as the Pres. of kullum, the near-synonymous D-verb, common in both OA and OB: tukall-aššu, -aššu can be analysed as a OB allomorph of the bound dative pronoun 3ms ‘to/for him,’ or as the ventive + acc. -šu (‘you will prevent it for your own sakevent.’).31 The OA example, PN lá tù-kà-lá ‘dosg. not hold PN back’ (CCT 4, 18b:15), can be interpreted as the Pres. of kullum as well (‘do not withhold PN for yourselfvent.’) if we accept, with GKT 54c, the orthographic rendering of the ventive without -m as a possibility for OA: lā tukall-avent.32 Thus, depending on one’s judgement, the D-stem of kalûm is represented in the corpus marginally or not attested at all. Ockham’s razor favors the latter solution, since for *kullûm no Preterit has been found, i. e. the tense form where the orthographic opposition between the D-stems of verbs IIweak and III-weak cannot be neutralized in both OB and OA, because the respective Preterit bases (unlike the Present ones) have orthographically undeletable phonological differences, whatever be affixed

29

For the competition between the Pres. and the SC of this verb in the present-time domain, see Loesov 2006:141. 30 Ll. 7ff.: ù a-na it-ti-i ša-tu-ú tu-ka-la-šu me-e [t]a-ta-ba-la-šum. The last two words probably mean ‘youm. sg. will take the water away from him.’ 31 I. e., in the spirit of “the reflexive benefactive ventive on transitive taking hold-of verbs” (Loesov 2006a). 32 Bert Kouwenberg suggests me a different solution, which looks more likely (p. c. of March 3, 2011): “I prefer to interpret the OA form lá tù-kà-lá as an imperfective pl. In many OA letters, second person sg. and pl. alternate erratically, according to whom the writer has in mind for a particular statement. This is more probable than a ventive, which in the intended meaning does not further occur with this verb, as far as I know, and -a instead of -am is unusual (though possible).”

162

Articles: Ancient Near Eastern Studies

to their right. Consider the comparative table of the 3 sg. Present and Preterit forms for kullûm and kullum in both dialects.33

OB

Preterit ukalli

OA

ukalli

kullûm Present ukalla ukalla

kullum Preterit ukīl/ukillū

Present ukāl/ukallū

uka!!il/uka!!ilū

ukâl/ukallū

– nazāqum ‘to worry.’34 No Gtn in the corpus, it is attested a few times (‘to have constant worries, to squeak constantly’) in SB omens (AHw. 772a, CAD N2 138a). – parādum ‘to be afraid, to care.’35 No Gtn in the corpus, it is attested a few times in SB medical texts (AHw. 827b, CAD P 143a: ‘is constantly fussing’). – takālum ‘to trust.’36 No Gtn in the whole of Akkadian. – wašābum ‘to sit’, ‘to dwell.’37 No Gtn in the core OB and OA corpora. An exception is ‘the chamber [š]a qerbuššu ni-it-ta-aš-ša-bu-nim in which we used to sit’ (MIO 12, 54rev.:17, “OB love lyric” according to CAD R 377b). Note that the compatibility of the subjunctive and the ventive (-nim = qerbuššu?) in this form is also most unusual. Another exception (a Participle) also comes from an OB literary text: mu-ta-ašši-ba-at askuppāt awīlê ‘(she is) the one constantly sitting on the thresholds of men’ (van Dijk 1953:92, 2). According to the dictionaries (AHw. 1483a, CAD A2 386a), Gtn appears a few more times in SB. Thus all eight common verbs (found so far) that use both the Pres. and the SC for coding the present-time sense do not really have pluractional stem forms (whether Gtn or D) in the corpus. 33

Bert Kouwenberg finds this presentation too tortuous: “I think it would be better to state right after mentioning the D forms of CAD that you think they do not belong to kalûm at all.” He thinks it would be enough to simply refer to AHw., where all the relevant examples “are assigned to kullu, as they should be.” 34 For the competition between the Pres. and the SC of this verb in the present-time domain, see Loesov 2005:137. 35 For the competition between the Pres. and the SC of this verb in the present-time domain, see ibid. 137. 36 For the competition between the Pres. and the SC of this verb in the present-time domain, see ibid. 141f. 37 For the competition between the Pres. and SC of this verb in the presenttime domain, see ibid. 140f.

S. Loesov, IPTANARRAS Can Render the “Basic” Present …

163

According to Kouwenberg, lowly dynamic transitive agentive38 verbs are supposed, as a rule of thumb, to have the pluractional Gtn-stem. In my eight-verb list there are three very frequent verbs that meet these criteria: akālum ‘to eat,’ baqārum ‘to claim,’ kalûm ‘to hold,’39 and all three have neither Gtn- nor D-stem at their side. On the face of it, these notions are dynamic and easily “pluralized,” so the absence of either Gtn- or D-stem needs an explanation. Further, such notions as ‘to worry’ (nazāqum), ‘to be fearful about smth./smb.’ (parādum), ‘to desire’ (¶ašā¶um) are quite susceptible to a pluralic interpretation. Only ‘trusting’ (takālum) and ‘sitting, dwelling’ (wašābum) are concepts whose pluralic derivates are perhaps less needed, due to their pronounced stative character. My preliminary explanation of this evidence runs as follows: IPTANARRAS can code not only “plurality” in the narrow sense, it is also involved in the coding of the “basic” present-time sense (probably in its different nuances) within (or for) the G-stem paradigm, along with the G-stem Present and the SC. As I have already suggested, the distribution of the three present-tense shapes is lexical. In particular, from the point of view of Akkadian, the above eight verbs have the same Aktionsart (= “situation aspect” of Smith 1997), though takālum and wašābum usually do not have “accusative” arguments. Whatever their government and valences, these eight are all low-transitivity verbs, therefore their presenttime reading is taken care of by both PARIS and IPARRAS (in “competition”), and they do not need IPTANARRAS for this function. Why they do not use the Gtn-stem for a genuinely “pluralic” sense, I cannot explain for the moment. In addition, there is a group of frequent lowly dynamic transitive verbs that do not form the SC, to be joined to the “prefixing statives” of the traditional grammar. The most common of them are râmum ‘to love’ and zêrum ‘to hate’; most importantly, both have no Gtn. I would add to this group e. g. the OB na´ārum ‘to guard, to keep,’ in e. g. šê ina´´ar ‘he keeps my barley’ (AbB 4, 40:20). Its SC na´ir is rare and always passive, the verb has no Gtn in the corpus. To come back to the roots that use PARIS (and no IPARRAS) to render the present-time sense: two very common verbs of this group are labāšum

38

Note that in the above Kouwenberg’s list of lowly dynamic transitive roots with admittedly frequent Gtn all the items are meant to be agentive. 39 Whether the verb ¶ašā¶um is agentive, i. e. whether its subject has free will, is not clear (cf. von Soden 1959:438).

164

Articles: Ancient Near Eastern Studies

‘to wear’ (labšāku ‘I wear’) and qerēbum ‘to get nearer’ (qerub ‘is near’/‘is approaching’).40 Once more, these roots have no Gtn in the corpus. For these facts, I suggest the same explanation as for the above eight verbs with the PARIS/IPARRAS competition: the respective roots do not form Gtn because for some reason they do not need it for the coding of the basic-stem PRESENT. Since the frequency of the roots and the basic character of the concepts neutralize the factor of chance to a certain degree,41 the only sensible alternative that comes to mind would be the following: Akkadian is supposed to process all these concepts/roots as radically stative and therefore not readily susceptible to “plurification,” with the result that they will not need the Gtn-stem. This claim is stronger and more daring than mine, in particular because it clashes with Kouwenberg’s conclusion (Gtn is productive for intransitive agentive verbs and “transitive verbs with a low degree of transitivity”), which has a lot to commend itself. My paragraph (2) constitutes a sort of ex silentio argument (why is there no Gtn for this or that common root?), but, given the absence of native speakers and a well-formed Akkadian National Corpus, this kind of reasoning and the questions it leads to are hopefully legitimate and useful for the better understanding of the language. 3. In the protases of OB omens, we come across synonymy of PARIS and IPTANARRAS for certain verbs.42 In particular, this is true of rabi´ vs. irtanabbi´ and maqit vs. imtanaqqut, while the G Pres. forms irabbi´ and imaqqut are not attested in these protases. Consider the examples: šumma qutrinnum ana ereb šamši ma¶râtūšu ra-ab-´a arkassu šaqât ‘If the front parts of the (smoke of the) incense are settling down toward the west but its rear part is ascending?’ (Or NS 32, 383:11). šumma qutrinnum ir-ta-n[a]-b[i]-i´ ‘If the (smoke of the) incense keeps settling down’ (UCP 9, 369:37).

40

The latter has been already discussed above; for a morphosyntactic analysis of the former, see Loesov 2010:765. 41 Consider the fortuitous character of our corpus of this long-dead language with no written tradition reaching into the present day through an uninterrupted chain of transmission. 42 Much of what will be said in this section is the Gemeingut of Ilya Khait and the present writer. We discussed the evidence and the interpretations during several sessions in summer 2010.

S. Loesov, IPTANARRAS Can Render the “Basic” Present …

165

The translations are non-committal as to the aspect/dynamicity and taken from CAD. Both rab´ā and irtanabbi´ as descriptions of the behavior of smoke refer to the same real-life situation. In other words, it is hardly the case that rabi´ describes a condition of incense smoke, while irtanabbi´ describes its iterative (or otherwise “pluralic”) movement (e. g., in the texts there are no hints to the effect that the observation was repeated, etc.). The usage of verb forms in divination protases has not yet been fully understood,43 yet it is well-known that this usage is “artificial,” i. e. obviously different from the natural linguistic code. It includes certain literary conventions proper to the genre. (E. g., dynamic preterit forms like i¢¢ul ‘he looked/he had a look’ can describe a condition of still exta, etc.) In the case of rabi´/irtanabbi´, there was probably an uncertainty as to how one should depict the behavior of smoke, in static or dynamic terms, yet for the authors the common denominator of both verb forms was that the two inflectional paradigms (PARIS and IPTANARRAS) could as a matter of principle belong to the G-stem and code the “basic” present-time idea. E. Cohen (2010:713) says that linguistically “beast omens, unlike those concerning extispicy, may answer to a different set of rules: beasts actually do things while exta basically do not.” Yet, surprisingly enough, it looks like the creators of the linguistic rules of the divination genre did not really take seriously this facts-of-life difference (still objects vs. moving ones), since both varieties of protases possess essentially the same repertoire of finite paradigms.44 Only the linear order of “tenses” in the chaining constructions of protases may have been sensitive to the criterion of the real-life stativity vs. dynamicity (see presently). The problem is in much need of research. One thing is clear: we cannot take the morphosyntactic data of the OB divination protases as a prima facie evidence for the tense-aspect of the spoken Akkadian contemporary to the creation of YOS 10 and other OB divination compendia. Both maqit and imtanaqqut can be used to describe a condition of still exta rather than dynamic events of movement: ú-ba-nu ša-al-ma-at MÁŠ [= ´ibtum] ma-aq-ta-at (JCS 21, 231:27, an extispicy report = ARM 26/1, 100bis:50) ‘Le Doigt etait en bon etat. L’Excroissance tombait (?)’ (Nougayrol’s translation in JCS). 43 Metzler 2002 and Cohen 2010 do not address the kind of questions mentioned here. 44 And it is not really different from that of the CH protases.

166

Articles: Ancient Near Eastern Studies

vs. šumma rēš [ubānim] qûm pe´ûm šakim-ma u im-ta-[na]-aq-qú-ut ‘if on top of the “finger” there is a white “filament” in horizontal? position’ (YOS 10, 33rev. iv 35ff.). [erištum] nadiat-ma im-ta-na-qú-ut ‘there is [a feature of liver] in horizontal? position’ (CT 44, 37:21).

In descriptions of still exta, imtanaqqut (rather than maqit) a l w a y s a p p e a r s i n t h e n o n - i n i t i a l s l o t, precedeed by a SC form, whether existential (as in YOS 10, 33rev. iv 35ff. and CT 44, 37:21 above) or descriptive, as in the following example: ‘if the “weapon” … sali¶-ma im-ta-na-qú-ut is covered with drops and lies horizontally?’ (YOS 10, 46 v 16).

This purely syntactic distribution (reminiscent of the OT literature, both prose and poetry) may be due to the just mentioned “conventional linguistic code” of the divinational literature. I. e., the second-slot imtanaqqut is a feature of literary consecutio temporum, idiosyncratic for this genre (more specifically, for still-life protases).45 Once more, maqit and imtanaqqut do refer here to the same real-life facts, and their alternation (while imaqqut is absent from this slot) may be ultimately explained by the suggestion that both finite paradigms (i. e., PARIS and IPTANARRAS) can functionally belong to the G-stem. Plus we have to remember that the originators of Akkadian extispicy decided to describe the still life, at least partly, in terms of becoming, des Werdens. Why then did not they use in the above examples imaqqut in the slot of imtanaqqut?—According to my hypothesis, this is because in spoken Akkadian imaqqut cannot be contemporaneous to the moment of observation (= PRESENT) but rather has to refer to FUTURE. Most importantly for my thesis, in divination protases both maqit and imtanaqqut (but not imaqqut!) can describe real-life dynamic present-time events as well: šumma qutrēnum ana imittī-šu ik´ur-ma šumēl-šu ma-aq-ta-at ‘If the smoke concentrated to the right and (then) its left starts going down?’ (UCP 9, 377rev.:44, see ibid. 48, and cf. CAD M1 244b, B 122b, and Pettinato 1966, 319:29).

45

Incidentally, other tan-forms appear in still life extispicy in the same slot (the SC-ma the TAN Pres.), as in the following example: šumma martum nas¶at-ma it-ta-na-aG-ra-ar ‘If the gull bladder has been torn out and turned upside down?’ (YOS 10, 31 iii 41–42, Ntn of the intransitive verb g/qarāru ‘to writhe, grovel’).

S. Loesov, IPTANARRAS Can Render the “Basic” Present …

167

In this smoke divination protasis, maqtat is hardly static as far as facts of life go. šumma awīlum inūma ´allu ālum im-ta-na-qú-ta-šum u i-¶a-az-zu-ma išemmû-šu ‘If a man, when he is asleep, (dreams that) the city is falling? upon him, and he groans and someone hears him’ (AfO 18, 67 iii 31ff.).

The text suggests that what is meant is a telic event of falling rather than a recurring nightmare. uz-na-šu im-ta-0na-qú-ta146 ‘if the sheep [while it is being slaughtered] is twitching its ears’ (YOS 10, 47:4). DIŠ UDU

The situation of ear-twitching is probably frequentative by its nature, so this example is indecisive. This cumulative evidence seems to indicate that the Akkadian Sprachgefühl did not permit imaqqut as a vehicle for the expression of PRESENT ACTUAL. To sum up, my preliminary conclusion boils down to the starting-point of my inquiry into the meaning of the Akkadian Stative (Loesov 2011 and previous essays mentioned in the present note): in order to understand what the Present of tan-stems was doing in Akkadian, we have to start with a list of OB and OA verbs that are frequent enough in the corpus to allow a deduction about the functions of Gtn in spoken varieties of Akkadian. References Charpin 1993 Charpin–Durand 2004

Cohen 2010 George 2002 Goetze 1936 Goetze 1958 Heeßel 2000 46

I. Khait’s reading.

Charpin, D. Données nouvelles sur la poliorcétique à l’époque paléo-babylonienne. MARI 7:193–203. Charpin, D.; Durand, J.-M. Prétendants au trône dans le Proche-Orient amorrite. Dercksen, J. G. (ed.). Assyria and Beyond (FS Larsen). Leiden. Pp. 99–115. Cohen, E. Conditional Structures in the Old Babylonian Omens. B&B 4/2:709–727. George, A. The Babylonian Gilgamesh Epic. Oxford. Goetze, A. The t-Form of the Old Babylonian Verb. JAOS 56:297–334. Goetze, A. Fifty Old-Babylonian Letters from Harmal. Reprinted from Sumer XIV. Baghdad. Heeßel, N. P. Babylonisch-assyrische Diagnostik. Münster.

168 Kogan 2001 Kouwenberg 1997 Kouwenberg 2010 Loesov 2005 Loesov 2006 Loesov 2006a Loesov 2010 Loesov 2011 Metzler 2002 Pettinato 1966 Smith 1997 van Dijk 1953 von Soden 1959 Wasserman 2003

Articles: Ancient Near Eastern Studies Kogan, L. *ġ in Akkadian. UF 33:263–298. Kouwenberg, N. J. C. Gemination in the Akkadian Verb. Assen. Kouwenberg, N. J. C. The Akkadian Verb and Its Semitic Background. Winona Lake. Loesov, S. Akkadian Sentences about the Present Time (I). B&B 2:101–148. Loesov, S. Akkadian Sentences about the Present Time (II/1). B&B 3:133–148. Loesov, S. Marginalia on the Akkadian Ventive. B&B 3: 101–132. Loesov, S. Akkadian Sentences about the Present Time (II/2). B&B 4:759–785. Loesov, S. The Suffixing Conjugation of Akkadian: In Search of Its Meaning. B&B 6:75–147. Metzler, K. Tempora in altbabylonischen literarischen Texten. Münster. Pettinato, G. Libanomanzia presso i babilonesi. RSO 41: 303–327. Smith, C. A. The Parameter of Aspect2. Dordrecht. van Dijk, J. J. A. La sagesse suméro-accadienne. Leiden. von Soden, W. Zu A. Haldar, The Akkadian Verbal System. Or 28:437–442. Wasserman, N. Style and Form in Old-Babylonian Literary Texts. Leiden.

Akkadian of the Me-ság Archive* Ekaterina Markina Russian State University for the Humanities, Moscow

0. Introduction This study is the second in a series of articles aiming at collecting Akkadian data scattered through Sargonic economic archives and assessing the relevance of this material for Akkadian linguistics.1 The article analyses 149 texts that form a large Sargonic archive associated with Me-ság, a high-ranking official who was active during the reigns of Narām-Sîn and Šar-kali-šarrī. Two suggestions regarding Meság’s identity have been made. P. Steinkeller has argued that the chief administrator of the Me-ság archive was identical with Me-ság, the governor of Umma under the last two Sargonic kings2 well attested in the corpus of mu-iti texts.3 Conversely, B. Foster insists on the multiplicity of Me-ságs in the Umma area and supposes that the ensi of Umma and the head of the Me-ság estate were two different persons.4 On the basis of their paleography the texts are conventionally placed into the later phase of the Sargonic period, more specifically, into the reign of Šar-kali-šarrī.5 S. Bridges attributes the archive to an earlier date, i. e. to the latter part of Narām-Sîn’s reign extending into the early years of Šar-kali-šarrī.6 Date formulae7 and seal impressions found in the texts do not help to calibrate the date of the archive, the former lacking any association with a * This article was prepared with the financial support from the Russian Foundation for the Humanities (project No. 09-04-00235a). The author wishes to express her gratitude to this institution. 1 The first one is “Observations on Gasur Akkadian” (Markina 2011). 2 See Steinkeller–Postgate 1992:8, fn. 36. 3 See Foster 1982b:155. 4 Foster 1979:161, fn. 56; Foster 1982a:52. 5 Foster 1982a:52; Steinkeller–Postgate 1992:8. 6 Bridges 1981:18–19. 7 Two date formulae are attested in the archive: GinH 1 MU [I7]-GEDENH-NA G ALH BA-DÙ-A “In the year when the Eden-canal was dug” (BIN 8, 117:5f.) and MU É GIŠ-GKINH-TI “the year of the workhouse” (BIN 8, 137). They are registered

170

Articles: Ancient Near Eastern Studies

specific king, the latter belonging to Boehmer’s “Akkadisch III” glyptic phase, which extends from the beginning of Narām-Sîn’s reign to that of Šu-Durul.8 However, the available philological evidence (in particular, the use of the bilingual formulary) speaks in favor of the earlier dating suggested by Bridges.9 Since the texts come from illicit diggings,10 their exact provenience is unknown. However, there are strong reasons to believe that the texts originated in an economic establishment located somewhere between Umma and Girsu.11 The fact that the business and administrative activities registered in the archive are closely connected with Lagaš rather than with Umma led Bridges to the conclusion that the estate supervised by Me-ság was in the vicinity of Lagaš.12 This interpretation was then developed by Steinkeller, who proposed Sagub, a town in the Lagaš province, as a possible location of the administrative headquarters of the estate.13

in the list of Sargonic date formulae in Gelb–Kienast 1990 as D-53 Anonym 13 and D-50 Anonym 9 respectively. 8 Bridges 1981:14. 9 “The script and the seal impressions suggest a classical Sargonic date for the archive. The mention of Lugal-ušumgal and the dumu-lugal-me in BIN 8, 214! (text: 134) indicates, more specifically, that the archive existed during the latter part of Narām-Sîn’s reign. … This date is supported by a reference in BIN 8, 134 to the king’s journey in the south and other texts which may, ultimately, relate to the ‘royal progress’ through Sumer by Narām-Sîn and his family” (Bridges 1981: 17–19). According to Foster, Šar-kali-šarrī was the king whose journey is mentioned in BIN 8 (see Foster 1980:39f.). After this article was completed, two additional date formulae (both belonging to the reign of Narām-Sîn) have been made known to me by E. Salgues, to whom I must extend my sincere gratitude. They occur in the unpublished documents RBC 2631 ([in MU Na-ra]-am-[d]EN.ZU ŠUBURki SAG.GIŠ.RA ‘In the year (that) Narām-Sîn conquered Subartu’) and RBC 2664 (in 1 MU Nara-am-dEN.ZU Ar-ma-namki SAG.GIŠ.RA BÀD.[BÀD?] u-na-[qì-ir] ‘In the year (that) Narām-Sîn conquered Armanum (and) destroyed its walls’) and are to be discussed in detail in Salgues 2011. 10 Bridges 1981:3. 11 Foster 1982a:52. Among the features pointing to an Umma–Lagaš locale are: similar notation of area measures; field toponymy common with the Umma– Lagaš region; onomastic similarities. 12 Bridges 1981:17. 13 Steinkeller–Postgate 1992:9.

E. Markina, Akkadian of the Me-ság Archive

171

1. Research history of the Me-ság tablets The article is based on 149 published Me-ság texts.14 The majority of these tablets (135 in total) are now kept in the Yale Babylonian Collection of the Yale University. The Iraq Museum houses a few more documents belonging to the archive, of which 13 are published.15 One tablet is owned by the Bryn Mawr College (Pennsylvania). From among the Yale tablets, 102 items were published in cuneiform copies in BIN 8 by G. Hackman and F. Stephens in 1958. Later on, they were treated by S. Bridges, who discovered 30 additional Me-ság texts in the Yale collection and made them available in transliterations as an appendix to her dissertation.16 The Iraq texts were edited by P. Steinkeller and N. Postgate in ThirdMillennium Legal and Administrative Texts in the Iraq Museum, Baghdad (MC 4) in 1992.17 Prosopographical data show that these documents (labeled by the editors as “Sagub tablets” according to their assumed place of origin) belong to the Me-ság archive. In the above publication, Steinkeller also pointed out that 9 so-called “udu á”18 tablets published in BIN 819 but excluded by Bridges from her corpus, should also be treated as part of the archive.20 The Bryn Mawr College tablet was published by M. deJ. Ellis21 who, on the wake of M. Lambert, proposed Zabalam as its place of origin.22 The attribution of the tablet to the Me-ság archive was made certain by Foster.23

14 An additional lot of 384 documents from the Yale Babylonian Collection (quoted as RBC throughout this article) are treated by E. Salgues in her forthcoming dissertation The Mesag Archive: Rural Administration in Early Empire. 15 The Iraq Museum houses 70 more unpublished documents from the archive (p. c. W. Sommerfeld). 16 Bridges 1981:446–484. 17 Steinkeller–Postgate 1992, Nos. 33–45. 18 This term probably refers to a type of rental payment, see discussion in Steinkeller–Postgate 1992:70. 19 More specifically, only 6 of them are published in BIN 8 (Nos. 185, 186, 187, 197, 328 and 342), while 3 remain unpublished (NBC 6930, 6951, 7026). 20 Steinkeller–Postgate 1992:9. 21 Ellis 1979, No. 15. 22 Ibid. 32. 23 Foster 1993:176.

172

Articles: Ancient Near Eastern Studies

2. The language of the Me-ság tablets The Me-ság texts can be subdivided into four groups according to their language. The first group consists of texts written in Sumerian (Sumerian case endings are employed, verbs are fully conjugated).24 The second group comprises tablets where Sumerian and Akkadian forms are used side by side.25 Documents belonging to the third group are written in Akkadian with the use of Sumerian logograms (no Sumerian case endings or conjugational elements are attested).26 Finally, the fourth group includes texts with no grammatical/lexical features that would allow attributing them to any of the aforementioned groups.27 Coexistence of texts using Sumerian, Akkadian and mixed formulary28 within one archive might indicate that the archive itself came into being during some kind of transitional period, when the state bureaucracy had to abandon old recording practices (Sumerian) in favor of the new ones (Akkadian). If this assumption is correct, an earlier date for the archive (most probably, Narām-Sîn’s reign) should be surmised, since such a transition is unlikely to be assigned to the years of the last king of the dynasty. 3. Glossary of Akkadian lexemes found in Me-ság texts The following list includes all Akkadian forms found in the published Me-ság corpus. The lexemes are listed according to their basic forms as 24 This group includes BIN 8, 118, 131, 142, 147–149, 152, 165, 191–193, 195, 199, 200, 202, 206, 208, 214, 229, 233, 245, 248–250, 259, 262, 272, 274, 276, 281, 283, 285; TTIM 33, 38, 40; NBC 5920, 6926, 6935, 6967, 6983, 6986, 6996, 6997, 7004, 7022, 7039, 7043, 10272, 10301, 12319. 25 BIN 8, 117, 122–124, 130, 132, 133, 137, 141, 182, 184, 204, 230, 231, 236, 251, 265, 267, 268, 273, 278, 280, 291; TTIM 36, 42, 44; NBC 6874, 6969, 6970, 6994, 6998. 26 BIN 8, 126–129, 131, 134–136, 138, 140, 145, 146, 183, 188, 191, 194, 196, 201, 205, 209, 215, 216, 220, 226, 244, 247, 254, 263; TTIM 35, 37, 39, 41; NBC 6999, 7009, 8960, 10196, 10300. 27 These are usually brief texts with an extensive use of Sumerian logograms and no Akkadian forms at all: BIN 8, 219, 223, 225, 240, 243, 255, 269. In view of the widespread use of Akkadian formulary in this archive, the language underlying such logographic writings is most probably Akkadian. 28 Sargonic texts displaying both Sumerian and Akkadian conjugated/declined forms are also attested in Girsu (e. g. CT 50, 123) and Gasur (e. g. HSS 10, 205). As an additional example of this kind, ECTJ 51 immediately comes to mind, but in fact the similarity is rather superficial: ECTJ 51 as such uses Sumerian formulary, Akkadian forms being inserted to render the words of an Akkadian-speaking person.

E. Markina, Akkadian of the Me-ság Archive

173

registered in AHw. The texts published in BIN 8 and TTIM are quoted according to their publication number. The texts published in Bridges’s doctoral dissertation are referred to by the NBC numbers. BMC stands for the Bryn Mawr College tablet. a¶āzum

‘to take’

al alākum

‘on’ ‘to go’

i-¶u-uz29 BIN 8, 146obv.:6

al TTIM 44 i 8, ii 5, iii 4′ a-li-ku30 BIN 8, 135rev.:2 i-li-ku (subj.)31 BIN 8, 129rev.:1 i-li-kam32 BIN 8, 146rev.:3 i-li-k[à-n]i (3 m. s. vent. subj.) BIN 8, 146:11 i-li-kà-ni (3 m. du. subj.) BIN 8, 265obv.:8 All these forms are attested in dating expressions referring to trips of Me-ság and/or other important officials to various destinations33. The pattern of the dating remark is usually ì-nu PN a-na GN i-li-ku. BIN 8, 135 and BIN 8, 146 both mention a journey to Akkade, and, since in BIN 8, 146 Me-ság is the subject of the clause, it is tempting to surmise that BIN 8, 135, featuring the verb form in the first person, was written by Me-ság’s own hand. The use of the first person verbal forms is well attested in the Sargonic economic documentation, although it seems to be restricted to private archives.34 In this respect, BIN 8, 135 offers a rare example of an administrative document with a personal touch.35 For the use of the subjunctive markers in these forms see 4.3 below.

29

According to E. Salgues, there are 9 other attestations of this form in the unpublished Me-ság material treated by her, as well as one attestation of the dual i-ḫu-za in RBC 2631. 30 a-li-ku is found in 4 other RBC texts, each time in the phrase ì-nu a-na A-gaki dè a-li-ku ‘when I went to Akkade’ (reference courtesy E. Salgues). 31 There are 5 other attestations of this form in the RBC texts, all in the dating remark ì-nu PN a-na GN i-li-ku (information courtesy E. Salgues). 32 The form also occurs once in an unpublished document from Yale in the same dating remark (i-nu Me-ság in A-ga-dèki i-li-kam ‘when Mesag came from Akkade’; reference courtesy E. Salgues). 33 Sumerian dating remarks of this type are also attested in the archive, cf. Me-ság Iri-sag-̃ rig7ki-ta im-ma-ge ̃ n-na ‘when Mesag came from Irisagr̃ ig’ (BIN 8, 137:4f.), Me-ság A-ga-dèki-šè du-ni ‘(At the time of) Me-ság’s trip to Akkade’ (BIN 8, 169:3–5+). 34 Other examples include CT 50, 72:4 (Sippar, Qurādum’s archive) and OAIC 8:13, 21:7, 35:11 (all from the Diyala region). 35 DPA 47:10 from Girsu is, probably, an additional example.

174 ana balu¶¶um

barûm

bašûm ¶ubbutum

36

Articles: Ancient Near Eastern Studies ‘to, for’ a-na passim ‘a type of aromatic’ ba-lu-¶um TTIM 43:4′ The term balu¶¶um designates a resinous plant often used, as later texts suggest, for medicinal purposes (CAD B 74–75). Beside this occurrence, it is also attested in Sargonic texts from Girsu and Susa, where it is spelled with ba5 (KA×IM): ba5-lu-¶um (ITT II/2, 4461, 5766; ITT V, 9293, all Girsu), ba5-lu-¶u-um (MDP 14, 89, Susa). In a Sargonic text from the Diyala region (MAD 1, 286) this word is probably attested as ba-lu-kum.36 ‘to inspect’ ib-rí BIN 8, 183rev.:5; 273 iv 8; NBC 6847 iv 10 In all these instances ib-rí refers to an inspection of the estate’s animals by a supervising authority.37 The same term is used in a Sargonic document from Tutub, which registers a similar event— an inspection carried out in this city by its ensi, Narām-Sîn’s son Nabī-Ulmaš (Tutub 65). ‘to be, to exist’ i-ba-šè BIN 8, 138obv.:6, rev.:15; 140rev.:3; 145rev.:4; TTIM 44 i 9, ii 6, iii 5′ ‘to debit’ ¶u-bu-ut BIN 8, 141rev.:5 This term is attested in BIN 8, 141 in the following context: 270 ADDA KUŠ ù UDU È-A a-na An-na SIPA na-sí-i¶ in DUB-śu ú-la ¶u-bu-ut (lines 2–9 of the reverse).38 Bridges describes ¶u-bu-ut as one of the “well-attested accounting terms,”39 but this qualification is rather unclear: the term is quite uncommon both in Sargonic and in later periods (cf. CAD Ú 11). In addition to BIN 8, 141, the following attestations of ¶bt in administrative contexts are known from the Sargonic sources: (1) AIA 8 i 8 from Pugdan, where the remark la ¶u-bu-ut refers to barley and emmer, the yield from 5 fields:40 872;0.0 GUR A-ga-

See Steinkeller–Postgate 1992:57. In Steinkeller’s view, the equation balu¶¶um = balukkum is rather unlikely since both terms co-occur in MAD 1, 286 (lines 3 and 5 respectively). This objection is no longer valid, as Sommerfeld’s collation in the framework of his online TKTA project shows. The sequence of signs previously identified as balu¶¶um (25 ŠIM GSUH ba-lu-¶um in MAD 1, 286:3) is now read 25 šim[sú]-pá-lulum (Sommerfeld, TKTA 286:3; supālum also designating a variety of aromatic plant). 37 In the case of BIN 8, 273 and NBC 6847 (which appears to be a duplicate of the former) the supervising authority is BÀD-mu-pi5, who, according to the legend of his seal, was the princess’ (?) scribe (DUB-SAR ru-ba-tim). 38 The phrase in DUB-śu ú-la ḫu-bu-ut occurs once more in an unpublished RBC text (information courtesy E. Salgues). 39 Bridges 1981:323. 40 See Foster 1982c:22–23. Foster reads ḫu-bu-tú, which seems unlikely because UD is not used with the syllabic value tú in Sargonic (von Soden–Röllig 1991, No. 221).

E. Markina, Akkadian of the Me-ság Archive

175

dèki 80;0.0 ZÍZ GUR (names of 5 fields) la ¶u-bu-ut in 1 MU NINDINGIR En-líl-la. (2) MAD 5, 79:5, also from Pugdan, labels two deliveries (?) of grain as ¶u-bu-tum: 2;0.0 ŠE SAG-GÁL Pù-a 1;0.0 ŠE SAG-GÁL Ú-KA-li ¶u-bu-tum. The term ¶u-bu-ut in BIN 8, 141 has been differently interpreted as a D-stem stative of ¶abātum with the meaning ‘to debit’41 or as ḫubuttum ‘loan’42 in the predicative state. As the mention of ṭuppum suggests, the term must refer to an accounting practice.43 This passage is the most informative one as it throws some light on the accounting mechanism. The text deals with cattle, both living and dead. The latter are said to be first na-sí-i¶ (‘deducted’) for the shepherd, but then ‘not ḫu-bu-ut’ on his tablet. Since the final operation mentioned in the text is nukkusum ‘to balance an account’, it is indeed tempting to follow Foster in interpreting ¶u-bu-ut as a stative of ¶ubbutum with the meaning ‘to debit; to register in a(n accounting) document.’ Both CAD and AHw. suggest that ¶ubbutum may be connected with ¶abātum ‘to borrow’ (CAD Ú 11; AHw. 304), well attested in later periods. However, such a meaning does not plausibly fit the administrative context,44 which forces the editors of CAD to modify it as ‘to debit (?).’

41 ‘Expended for Anna the shepherd, not debited on his tablet’ (Foster 1982c:24); ‘to Anna the shepherd charged, (but) from his tablet, not debited’ (Bridges 1981:322). Foster compares ¶ubbutum with Sumerian zi(g) ‘to issue’ (Foster 1982c:24). 42 ‘270 of his (sheep) carcasses and expended sheep were deducted for DINGIRna, the shepherd, from his tablet, which is not a ¶ubuttu-loan’ (Steinkeller–Postgate 1992:9–10). While the ¶ubuttum-loan is not attested in Sargonic documents, a related term ¶u-bu-tá-tum is found once in a text from the Diyala region: 1 KIRX (SILA4.SAL) iś-te4 Kà-lí-iś-DU10 ¶u-bu-tá-tum (OAIC 32:4). The term ¶ubuttatum ‘interest-free loan’ is well known from later periods (CAD Ú 221), and its meaning could hardly be different in Sargonic times. 43 Both ¢uppum and ¶ubbutum may also be mentioned in the Ur III document CST 19. This tablet records issues of small quantities of sesame (GIŠ-Ì) to various persons. The concluding lines of the document read: a-na DUB ME ḪI la ¶u-bu-tu ‘not debited (?) on the tablet of evening sacrifices (?)’ (the question marks are mine; for the translation see Foster 1982c:24 and cf. ana ON lā ḫu-bu-tu ‘sind sie nicht schuldig’ in AHw. 304 sub ¶abātum II). In any case, this Ur III passage differs from the Sargonic texts under consideration in what concerns the referent of ¶ubbutum: in the Sargonic texts the singular stative form ¶ubbut refers to the commodity, whereas in CST 19 we are faced with the plural stative form ¶u-bu-tu (¶ubbutū), most probably referring to the persons mentioned on the obverse of this tablet. 44 The term refers to documents that describe operations between offices within a household. In this strict sense, “administrative documents” are opposed to private records that describe transactions between individuals.

176

Articles: Ancient Near Eastern Studies It is worth observing that, while ¶abātum ‘to borrow’ is a Gstem verb, all the relevant Sargonic forms rather belong to the Dstem.45 As the D-stem is commonly used to produce denominative verbs (GAG § 88g),46 one is inclined to consider the verb ¶ubbutum ‘to debit’ as derived from a noun, possibly attested in MAD 5, 79:5. MAD 5, 79 is an account of barley that records two amounts labeled ¶u-bu-tum, which are associated with two different persons (lines 1–5). Besides, the text records a delivery of barley by a sagiofficial. As the formulary suggests, the purpose of this delivery were barley-rations (cf., e. g., TTIM 39); it was probably made in three installments. In a sequence like this, one expects to find two similar administrative operations, which makes the interpretation of ¶u-bu-tum as a type of loan (unattested elsewhere in the Sargonic documentation) unlikely. On the contrary, the meaning ‘debit’ or ‘debited (grain)’ (i. e. grain delivered and registered) seems plausible in this context.

in

‘in, inside’

in passim

īnu (conj.)

‘when’

ì-nu BIN 8, 129obv.:5, 134obv.:8, 135rev.:1, 146obv.:9, rev.:1, 265obv.:5

išpikū

iś-pi5-ku47 BIN 8, 205:2, 226:2; NBC 6970obv.:5, rev.:7, 7009rev.:6′, 10300obv.:4 The morphological pattern of išpikū is ipris- (GAG § 56a). Other ipris-formations attested in the Sargonic sources are iškinū ‘additional payment’48 and ip¢erū ‘redemption.’49 Both išpikū and iškinū are always used in the plural in Sargonic. In the Me-ság archive the term always refers to grain (barley, emmer and wheat) provided from an allotment of land by its holder. Bridges suggests that išpikū refers to the rent, i. e. the yield of an allotment going to the lessor.50 Interestingly, the grain total is

45

‘yield’

The only attestation of ¶abātum in the G-stem in the Sargonic corpus is in HSS 10, 201, which mentions 3 missing servants of the king (IR11 LUGAL): ŠU+ NÍGIN 3 GURUŠ IR11 LUGAL in GN ši ´ābim ¶abtū ‘Total: 3 males, servants of the king, missing from the town of Bàd-lugal, (the headquarters) of the workforce’ (or ‘from the guarded town of Bàd-lugal’). 46 This would not be the only example of such derivation in Sargonic: see Kutscher 1989:31 for the possible association of the verb ṣubbûm ‘to form an army’ with ´ābum ‘army.’ 47 There are over 20 additional attestations of this lexeme in the unpublished RBC texts (information courtesy E. Salgues). 48 MAD 1, 45rev.:4, 50:4, 51:7, 52rev. i 5. 49 Used as a personal name only, cf. MAD 1, 158+328:4 and 8. 50 See Bridges 1981:118 for the discussion.

E. Markina, Akkadian of the Me-ság Archive

177

roughly the same in all the instances (180 to 200 gur), which might indicate that the land was leased in standard-size plots. The only occurrence of this term outside the Me-ság archive is in the famous Gutian letter Gir 19:38 (ù at-tá MAŠ.ANŠE ù-la tá-na´a-ar iś-pi5-kí kí-nu-tim a-rí-iś-ka). Kienast and Volk’s suggestion that in this case išpikū may refer to the profit from the sale of the animals51 is rather unlikely in view of the Me-ság parallels. ište kukkurum

‘with’ iś-te4 passim a type of aromatic ku-ku-ru-u[m] TTIM 43obv.:7′ This is the least common form of the lexeme normally attested as kukrum in the Sargonic sources52. Outside the Me-ság archive kukkurum is found only in Umma texts (see BIN 8, 300:6; 319rev.: 5, both Umma C). For the pronunciation kukkurum (as against kukurum in CAD and AHw.) see Steinkeller–Postgate 1992:79.

kullum ma¶ārum

‘to hold’ ‘to receive’

mašlûm

‘skin-bucket’ maš-lí-um BIN 8, 267 i 8′; 280 i 4′, 8′ All Me-ság occurrences follow the standard Sargonic orthography for this lexeme (with LÍ).53 Deviant spellings attested in texts from the Diyala region and Girsu include maš-li-a-tum (OAIC 7:10) and kuš A-GÁ-LÁ maš-lì-um (RTC 239 ii 6).54

mušālum

‘mirror’ mu-ša-lum BIN 8,145obv.:2 mušālum is a by-form of mašālum, the common Sargonic term for a bronze mirror.55 Note that outside the Me-ság archive the form is attested only in Girsu (ITT V, 9262, 9302),56 which might be regarded as a lexical argument for the Me-ság–Girsu connection.

u-kà-al BIN 8, 291rev.:10 im-¶ur BIN 8, 117obv.:3, 122 i 4, 10; ii 4; 123 i 3; 124rev.:2; 126obv.:5; 127obv.:6; 128obv.:6; 129rev.:3; 130rev.:5; 131obv.:9; 132 iv 13; 133rev.:7; 134obv.:7; 135obv.:6; 267 iv 12; 280 iv 11; TTIM 42 i 3, ii 4; BMC 15obv.:5 tám-¶ur (prt. 3 f. s.) BIN 8, 124obv.:4

51 “Es dürften aber eher die Erträge gemeint sein, die sich aus der Verwertung oder dem Verkauf der Tiere ergeben würden” (Kienast–Volk 1995:93). 52 ITT I, p. 13, No. 1214; p. 21, No. 1330; ITT II/2, 4461, 4587, 5766; ITT V, 9293 (all Girsu) 53 See examples in MAD 3, 270. 54 The reading of IGI as lì is assured by the parallel kušA-GÁ-LÁ maš-lí-um (TCL 5, 6052 i 13, Ur III Umma), with a regular spelling. 55 Since this word is usually found in Sumerian contexts, it should probably be interpreted as an Akkadism in Sumerian rather than a true Akkadian lexeme. 56 Where the form mašālum is also attested (ITT II/2, p. 2, No. 2839). This form is also present in a Sargonic document of unknown provenience, with an unusual spell-

178

Articles: Ancient Near Eastern Studies

makāsum

D ‘to balance an u-na-ki-is BIN 8, 141rev.:9; 182 iv 10 account’ The verb nakāsum is poorly represented in the Sargonic corpus. To my knowledge, it is completely unattested in the G-stem. Four attestations in the D-stem refer to the conclusion of an administrative operation: 1) (sheep) … in Sag-ubki ši Lagaški Me-ság u-na-ki-is (BIN 8, 141rev.:9); 2) (grain) … Ama-bára u-na-ki-is (BIN 8, 182 iv 10); 3) (various items) ... a-na Ur-dNin-tu u-na-ki-i[s] (Tutub 46 iv 10); 4) (malt) … šu DUB ma¶-rí-im la-ma nu-ku-us (Glassner 1983:211, text 1rev.:5).57 Foster has shown that in this type of contexts nukkusum should be interpreted as an accounting term, most probably with the meaning ‘to balance an account.’58 His suggestion that nukkusum with this meaning is a D-stem denominative from nikkassum ‘balanced account’ seems likely, especially in view of a possible parallel in ¶ubbutum < ¶ubuttum discussed above.

nasāḫum

‘to deduct’

na-sí-i¶ BIN 8, 141rev.:4

In the Sargonic corpus nasā¶um is commonly found in curse formulae of the royal inscriptions with the meaning ‘to tear out (foundations)’.59 BIN 8, 144, where the verb is used in connection with a large number of dead (ADDA) and expended (È-A) animals, is the only Sargonic occurrence of nasā¶um with the meaning ‘to deduct, subtract’, well attested for this verb in later periods (from OA on, see CAD N2 11).60 In the Me-ság text it is preceded by ana, which introduces the destination of the deduction: 270 ADDA KUŠ ù UDU È-A a-na An-na SIPA na-sí-i¶ ‘270 of (sheep) carcasses, skins and expended sheep were deducted for Anna, the shepherd.’61

ing ma-ša4-lum (Sommerfeld et al. 2005:206, No. 13:2). To my knowledge, the use of not attested anywhere else in the Sargonic corpus. 57 Antroponyms and toponyms that occur in this text suggest its provenience from Gasur. 58 Foster 1989:91. 59 See examples in Kienast–Sommerfeld 1994:254 (under nasā¶um). 60 The verb nasā¶um is often attested in connection with cattle as a term for physical transferring (see CAD N2 8). However, in BIN 8, 141 its use among terms designating various bureaucratic procedures suggests that nasā¶um must describe here an administrative action rather than a physical act, so the meaning ‘to transfer (cattle)’ was rejected in favor of ‘to deduct.’ 61 An alternative interpretation is suggested in Steinkeller–Postgate 1992:9f.: ‘270 of his (sheep) carcasses (ADDA-śu) and expended sheep were deducted for DINGIR-na, the shepherd.’ DU with the syllabic value ša4 is

E. Markina, Akkadian of the Me-ság Archive

179

This use is paralleled by the OB passage ana Á 150 ERÍN na-si-i¶ ‘(barley) was drawn as a payment for 150 men’ (VS 9, 22:9). nāsapum

dug a container na-sà-pu BIN 8, 267 i 3′ 15′ ii 1′ To my knowledge, this term is not found elsewhere in the Sargonic corpus. It is tempting to interpret it as nāsapum, an early form of nēsepum ‘a container’ (< *!sp), well attested in later dialects such as OB, NA, and NB (CAD N2 183).62 The absence of the ecoloring in the Sargonic form is an archaism, with well-known precedents in rāśum ‘head’ (< *ra!š-) and ´ānum ‘small cattle’ (*´a!n-).

šu

‘that’

šukūsum tarûm

‘allotment of land’ ‘to lead away’

šu passim ši BIN 8, 141rev.:7 šu-ut BIN 8, 273 ii 11; NBC 6847 ii 11 ša-at BIN 8, 273 ii 2; GšaH-[at] NBC 6847 ii 2 The use of the plural forms in BIN 8, 273 (as well as in its duplicate NBC 6847) is peculiar from the point of view of gender/number agreement. This text records an inspection of the estate’s animals (bovines and equids). Two herds consisting of cows, breeding bulls and calves (ÁB, GU4-ÁB and GU4) are mentioned in the text, along with the name of the official under whose care they are placed: ŠU+NÍGIN 16 ÁB ša-at Ba-ḫar ‘16 cows of Bahar’ (BIN 8, 273 ii 1–2; NBC 6847 ii 1–2) and ŠU+NÍGIN 25 ÁB šu-ut BÀD-mu-pi5 ‘25 cows of BÀD-mu-pi5’ (BIN 8, 273 ii 10–11; NBC 6847 ii 10–11). In both cases the relative pronoun refers to ÁB, a collective term denoting bovines of both sexes and all ages (CAD L 217). The Akkadian equivalent of ÁB in this meaning—liātum—is formally feminine plural.63 The form of the relative pronoun in BIN 8, 273 ii 1–2 (ša-at, f. pl.) is in agreement with this, but in BIN 8, 273 ii 10–11 we find a masculine plural form šu-ut.

ù

‘and’

wabālum

‘to bring, to carry’

62

ŠUKU-sà-tum BIN

8, 194rev.:5; 196rev.:17′ it-ru BIN 8, 146rev.:10; 241rev.:6′; 254obv.:6, rev.:4; it-[ru] NBC 6998rev.:9′ ù passim

u-ba-lam TTIM 44 i 9, ii 6, iii 5′ u-bìl BIN 8, 220obv.:7; 245obv.:5; [u-b]ìl BIN 8, 215rev.:6′; [u]-bìl BIN 8, 216rev.:7′ G u-ubH-lu (prt. 3 m. pl.) BIN 8, 122 ii 2, 13

.

An alternative interpretation is suggested by Sallaberger, who interprets nassapum as a rare by-form of na´pûm ‘Einweich-Gefäß (Flasche?),’ probably deriving from ´apûm ‘to soak’ (Sallaberger 1996:115). 63 For the collective ‘cattle’ (written ÁB.GU4¶i.a in Old Babylonian texts) and its Akkadian equivalent see Stol 1995:173.

180

Articles: Ancient Near Eastern Studies

warādum warûm

‘to go down’ ‘to lead, to bring’

zibibiānum

‘black cumin’

u-ur-da-ni (3 m. sg. vent. subj.) BIN 8, 134rev.:1 u-GruH BIN 8, 146obv.:8

zi-bí-bí-a-nu BIN 8, 123 i 11; 128obv.:4 zi-bí-bí-a-núm BIN 8, 132 iii 6–7; 267 ii 1′ zi-bí-GbíH-[a-núm] NBC 10196obv.:14 Note: zi-zi-GbíH-bí-a-núm in BIN 8, 276 ii 7 Two by-forms of this lexeme are attested in the Sargonic sources. To my knowledge, zibibiānum is only found in the Me-ság archive, whereas zizibiānum occurs in a number of Girsu documents.64 A unique mixed form zi-zi-GbíH-bí-a-núm is attested in one Me-ság text, perhaps resulting from a scribal correction. The absence of mimation in BIN 8, 123 and 128 is intriguing, but a closer look at the texts reveals a clear distributional pattern. When zibibiānum as a commodity is measured with the standard capacity measures (as in BIN 8, 123 and 128), it is written without mimation. In all other instances, when zibibiānum is measured approximately with pots (DUG or dugna-sà-pu as in 132, 276 and 267 respectively) or baskets (GURDUB as in 132) the mimation is present.65 The term is related to zibûm ‘black cumin’66 with which it shares the ideogram (Steinkeller–Posgate 1992:77).

Several forms attested in the archive are unclear, which is due either to the bad preservation of the text or to the lack of parallels elsewhere in the Sargonic corpus. (1) The restoration of the form in BIN 8, 146rev.:4 is uncertain. Signs visible in Hackman’s copy include the sequence i-ru-[…],67 of which, judging from the CDLI photo of the tablet68 only I is visible nowadays. Unfortunately, the right edge of the tablet is damaged, so it is impossible to say how many signs (if any) were destroyed. Assuming that Hackman’s copy is reliable, the sequence i-ru-[…] can hardly be anything else than a G-stem Akkadian verbal form (the more so in view of the fact that BIN 8, 146 uses Akkadian formulary). The text 64

ITT I, p. 8, No. 1137; II/2, p. 8, No. 2943. Interestingly, zizibiānum is the common form in the Ur III sources, v. references in MAD 3 332. 65 This distribution seems to become out of use by the Ur III-period. In the Ur III texts zizibiānum is measured out only with the standard capacity measures, but the lexeme is never written without mimation, cf. MAD 3 332. 66 zibûm is the standard term for this spice in MB, NB and SB, see CAD Z 103. 67 BIN 8, pl. LX, No. 146 68 The photo of the tablet can be found on http://cdli.mpiwg-berlin.mpg.de/ search/result.pt?id_text=P212692&start=0&result_format=single&-op_id_text= eq&size=100.

E. Markina, Akkadian of the Me-ság Archive

181

begins with a list of cattle labeled máš-da-ri-a ‘regular daily offerings’ in l. 5 of the obverse and continues as follows: BIN 8, 146 Obv. (list of cattle)

Rev.

06. An-na i-¶u-uz 07. Da-da SAGI 08. u-GruH 09. ì-nu Me-ság G H 10. in A-ga-dè ki 11. i-li-k[à-n]i 12. 1 UDU.NITA 1 SILA4 13. 1 KIR11 GAG-[…]69

01. ì-nu Me-ság 02. GinH A-ga-dèki 03. i-li-kam Ur-Mes 04. a-na An-na i-ru-[…] 05. 2 UDU.NITA 1 SILA4 06. 1 MÁŠ 07. šu Nam-ti-la-ni. 08. Ur-mes 09. a-na Lagaški 10. it-ru

The position of dating remarks in the text is ambiguous. As the parallels from other Me-ság texts show, they can be taken as referring either to the preceding or to the following transaction.70 However, since the dating remark never precedes the subject of the transaction, it is more likely that lines 9–10 of the obverse describe the delivery of cattle by Dada to the estate (l. 1–8), while lines 1–3 of the reverse refer to the preceding list of cattle (obv. l. 12–13). A structure like this requires a transitive verb in line 4 of the obverse (with Ur-mes being its subject and the cattle in l. 12–13 being its object): 6

Anna took 1–5(the cattle). 7–8Dada, the cupbearer brought (it to the estate) when Me-ság came from Akkade. 12–13One sheep, one lamb and one … female lamb— rev. 1–2when Me-ság came from Akkade— Ur-mes 4… (them) to Anna. 5–7Two sheep, one lamb and one goat belonging to Namtilani 8–9Ur-mes led away to Lagaš.

Unfortunately, I cannot offer any plausible restoration for the verb in l. 4, because none of the semantically suitable lexemes matches the beginning i-ru-[…].71 69

This term should probably be interpreted as a designation of the lamb (KIR11). 70 The dating remark follows the verbal form that describes the transaction in three cases (BIN 8, 134, 135, 137), while in one instance it precedes it (BIN 8, 129). 71 Identification of i-ru-[…] with the preterit of erēbum is unlikely, and not only because this verb is used only intransitively in the economic documentation of the period). According to E. Salgues (p. c.), RU at the end of the tablet is barely visible nowadays and there is hardly any room for anything else after it, so probably

182

Articles: Ancient Near Eastern Studies

(2) The term nu-tu-tum occurs in BIN 8, 291rev.:6 as a summary notation for the total areas of leased (SA10.A) and sustenance (ŠUKU) land. The epistolary passage Ga 3:23, dealing with seeding issues, may be relevant for the interpretation of the term in question: šumma erāšiś na¢u zēram līzib ‘If the field is suitable for plowing, let him leave the seed-barley (to the cultivators).’ If this comparison is justified, we may be faced with a purust- formation of na¢ûm ‘fitting, appropriate, suitable’ (CAD N2 130), in which case nu¢ûtum could mean ‘land suitable (for cultivation).’ Alternatively, the form may be interpreted as a masculine plural adjective from the same root, i. e. *nu¢¢ûtum ‘cultivable (plots of land).’ 4. Me-ság Akkadian 4.1. Orthography and phonology The table below gives an overview of signs used syllabically in the Me-ság texts. BA

BÍ

Ci (= NE)

Ce –

Cu

bp

Ca

aC –

iC

dt¢ gkq

DA

DÈ

(=NE)

–

GA

KI

– –

ID

KU

– –

l

LA

LI

–

–

AL

– –

MU

RU

– –

IM

BU

eC –

uC UD

–

– –

–

–

–

–

– –

UM

IB

UB

–

LÍ

r m

– –

RÍ

n š

NA

NI

–

–

–

SU

ṯ zs´

ŠA

ŠI

ŠÈ

ŠU

ZA

ZI

–

–

¶ ġ

– –

– –

– –

ḪU

! "

A

Ì

–

– –

Ù

–

h µ

– –

– –

y w

– –

I

–

–

UR

– –

IN IŠ

– –

– –

– –

– IZ (= GIŠ)

– –

– UZ

– –

VḪ

–

– –

– –

–

– –

– –

– –

– –

– –

– –

– –

– –

– –

– –

– –

U

– –

– –

– –

– –

NU

–

–

nothing is missing at the end of the line. This is supported by a clear attestation of the form i-ru in a similar Me-ság text (RBC unpublished).

E. Markina, Akkadian of the Me-ság Archive

183

4.1.1. The vocalic opposition Ci vs. Ce As one can see from the table, Ce signs (BI GI LI SI11 ŠÈ, as opposed to BÍ KI LÍ SI ŠI that are used for Ci)72 are poorly represented in the corpus. The attested sign pairs of this type are LI vs. LÍ and ŠÈ vs. ŠI. The only attestation of LÍ (outside personal names) is in maš-lí-um ‘bucket’ (BIN 8, 267 i 8′+). LI is only found in forms of alākum: a-li-ku (BIN 8, 135rev.:2), i-li-ku (BIN 8, 129rev.:1), i-li-kam (BIN 8, 146rev.:3), i-li-kà-ni (BIN 8, 265obv.:8+). As is well known, this verb is consistently written in this way throughout the Sargonic corpus.73 There are good reasons to suspect that this orthography actually renders the e-coloring somehow conditioned by the etymological *h, but the details of this intriguing development remain to be elucidated. For the contrast between ŠÈ and ŠI, cf. i-ba-šè (BIN 8, 138obv.:6+) vs. ši (BIN 8, 141rev.:7). In the pairs BI—BÍ and GI—KI, only the Ci signs are attested: zi-bí-bía-núm (BIN 8, 123 i 11+), iś-pi5-ku (BIN 8, 205:2+), u-na-ki-is (BIN 8, 141rev.:9). The use of these signs is consistent with the standard Sargonic orthography.74 4.1.2. Syllable initial /y/ As is well known, the Sargonic syllabary makes use of special signs (I, È and U) to mark the initial y- (notably, in the 3 m. verbal prefix), as opposed to Ì, E and Ú/Ù which represent the (’)V syllables. The syllabary of the Me-ság texts is no exception to this rule. The sign I is regularly used for the 3 m. verbal prefix in the G-stem (for the pertinent forms see a¶āzum, alākum, barûm, bašûm, ma¶ārum and tarûm in the glossary above). The only form written with the contrasting vowel-initial sign is ì-nu (BIN 8, 129obv.:5+). The sign U renders the 3 m. verbal prefix in the D-stem (the forms are listed under kullum and nakāsum in the glossary). Besides, it is also used for the 3 m. prefix of the verbs Iw in the G-stem, as in wabālum, warādum 72 For signs rendering syllables with an e-Auslaut as opposed to signs reflecting syllables with an i-Auslaut see Sommerfeld 1999:18 and Hasselbach 2005:39ff. 73 Hasselbach 2005:46. 74 Note that in the Girsu letter iśpikū is spelled with BI (iś-pí-kè, Gir 19:38), which suggests the reading /iśpekē/. Such an orthography can probably be explained by the lowering of /i/ into /e/ before the oblique plural ending /ē/ (vowel accommodation), see Hasselbach 2005:44.

184

Articles: Ancient Near Eastern Studies

and warûm. The forms beginning with the U-UC- sequence (u-ur-da-ni in BIN 8, 134rev.:1, Gu-ubH-lu in BIN 8, 122 ii 2, 13) are not without precedent in the Sargonic corpus, cf. u-ub-lam (Ad 3:8), u-ub-lu (MAD 1, 169 iii 12).75 4.1.3. Representation of the sibilants Each of the three sibilant series of the Sargonic syllabary (ŠV, SV and ZV) is in use in the Me-ság texts, with no apparent deviations from the normative Sargonic orthography. The ŠV-series is represented by four signs (ŠA, ŠI, ŠÈ and ŠU) that consistently reflect the etymological *¬, as in mu-ša-lum (BIN 8, 145obv.:2; *m¬l), i-ba-šè (BIN 8, 138obv.:6; *b¬y), the relative pronoun šu (passim) and its forms ši (BIN 8, 141rev.:7), šu-ut (BIN 8, 273 ii 11) and ša-at (BIN 8, 273 ii 2). The signs of the ZV-series, reserved for s, ṣ, and z, are found in the following forms: ŠUKU-sà-tum (BIN 8, 194rev.:5+), na-sí-i¶ (BIN 8, 141rev.:4), zi-bí-bí-a-nu (BIN 8, 123 i 11+). Besides, the sign UZ is attested in the preterit form of a¶āzum (*!¶¯): i-¶u-uz (BIN 8, 146obv.:6).76 The SV-series, used for the reflexes of *š and *ŝ in the Sargonic syllabary, is poorly attested in the Me-ság texts. In fact, the only extant example is -SU representing the 3 m. sg. pronominal enclitic -śu: in DUB-śu (BIN 8, 141rev.:5). 4.1.4. Representation of the gutturals The Me-ság texts do not yield any evidence pertinent to the fate of the etymological gutturals in Sargonic. The only lexeme with a laryngeal in the prototype is the preposition al (TTIM 44 i 8+), written with the sign AL as elsewhere in the Sargonic corpus. 4.2. Morphology The Me-ság texts provide a relatively modest amount of data pertinent to the nominal and verbal morphology of Sargonic Akkadian. 75 From Narām-Sîn on, this orthography is also attested in the royal inscriptions. The relevant verb is kamûm ‘to put in fetters’: i-ik-mi (Nar 1:19, Nar A 1:15; *Nar C 24:20.44), i-ik-mi-ù (Nar C 5:75). 76 There are several Sargonic attestations of aḫāzum spelled with the sign uzx (EŠ), a sign otherwise unattested in the Sargonic syllabary: li-¶u-uzx (Gir 3:9), a-¶uuzx (MAD 5, 8:12+). It has been suggested that such spellings may reflect the etymological *¯ (cf. Hasselbach 2005:73).

E. Markina, Akkadian of the Me-ság Archive

185

The oblique plural marker /-ē/ is explicitly written with the sign E: ì-nu a-na ŠE.BA ENGAR-e Me-ság ù Ama-bára DUB.SAR i-li-kà-ni (BIN 8, 265:5– 8).77 This attestation can be added to similar examples collected in Sommerfeld 1999:19. The same passage (BIN 8, 265:5–8) displays the only verbal form in the dual attested in the archive: i-li-kà-ni /yillikāni/ ‘(when) they both went.’ 4.3. The subjunctive Two types of subjunctive markers are used in the Me-ság texts: -u and -ni. All attestations are found in temporal clauses introduced by īnu ‘when.’ The morphological distribution is in agreement with the reconstruction proposed in Hasselbach 2005:208: -u was originally used after consonants and -ni, after vowels. (1) -u after a consonant: G ì-nuH Me-ság a-na ŠE.BA Ga-naH Lagaški i-li-ku (BIN 8, 129:5f.); G H ì-nu Me-ság in A-ga-dè ki a-li-ku (BIN 8, 135rev.:1–2); (2) -ni after a vocalic ending: ì-nu a-na ŠE.BA ENGAR-e Me-ság ù Ama-bára 8, 265:5–8).

DUB.SAR

i-li-kà-ni (BIN

In two cases we find -ni attached to the ventive marker -am (as regularly in Assyrian from OA on, Hecker 1968:134): G H

ì-nu Me-ság in A-ga-dè ki i-li-k[à-n]i (BIN 8, 146:9–11); ì-nu LUGAL u-ur-da-ni (BIN 8, 134:6f.).

The subjunctive markers -u and -ni do not co-occur in one and the same verbal form in the Me-ság texts, although such use is known from other Sargonic sources.78 4.4. The ventive The ventive marker -am indicating movement towards the speaker is attested in the archive with alākum, warādum and wabālum. The most interesting examples come from the difficult document BIN 8, 146, tentatively translated above. Three verbs of movement are used in this text, namely alākum, warûm and tarûm. The latter two form a contrasting pair, warûm indicating movement towards the speaker and tarûm indicating movement away from 77 Bridges takes -E as a defective spelling of the Sumerian plural marker -ene and emends this form to engar-e- (Bridges 1981:349). 78 See Hasselbach 2005:206 and Kogan–Markina 2006:578, fn. 55.

186

Articles: Ancient Near Eastern Studies

the speaker. The direction is thus rendered lexically and the ventive marker is not used. Conversely, both occurrences of alākum in this document are provided with the ventive marker, suggesting that a return trip is meant:79 G H

ì-nu Me-ság in A-ga-dè ki i-li-k[à-n]i (BIN 8, 146obv.:9–11); ì-nu Me-ság GinH A-ga-dèki i-li-kam (BIN 8, 146rev.:1–3).

In phrases that describe Me-ság’s journeys to other places, no ventive ending is attached to alākum, e. g. Gì-nuH Me-ság a-na ŠE.BA Ga-naH Lagaški i-li-ku (BIN 8, 129:5f.) and ì-nu a-na A-ga-dèki a-li-ku (BIN 8, 135rev.:1–2) It may seem unusual that the starting point of the return trip is introduced with in (otherwise mainly attested as a locative preposition with the meaning ‘in, inside’), especially since iśtum ‘from’ is well attested in the Sargonic corpus.80 However, the separative use of in can be illustrated by several other Sargonic examples where, remarkably, the ventive marker is also involved. The most transparent one comes from the inscription of Rīmuś: 14,100 GURUŠ.GURUŠ in IRI.IRIki Šu-me-ri-im u-śu-G´í-amH-ma a-na kà-ra-ši-im iś-kun ‘He expelled 14,100 men from the cities of Sumer and put them into a forced labor camp’ (Rim C 1:32). Another probable case is (x GUR of barley) in a-lí-im Ì-lí-iś-tá-kál ú-šu-ri-dam ‘(x gur of barley) Iliśtakal delivered from the city’ (MAD 4, 10:2, unknown provenience)81. As far as alākum is concerned, our text likely bears on the interpretation of the date formula in MU … LUGAL in !À-mar-nu-um i-li-kà-am (MAD 5, 76rev.:1–3, Pugdan). Frayne (1993:86) does not hesitate to translate ‘the year the king went on a campaign in Amarnum’, but the use of in to introduce the destination of a campaign is, to say the least, unexpected—as certainly realized by Gelb and Kienast (1990:58), who tentatively translate ‘im Jahre, nach dem …, der König, nach (?) Amarnum gezogen ist’. In fact, the event described is probably the king’s return from Amarnum.

79 This interpretation is especially attractive in view of a Sumerian parallel found in the archive: Me-ság Iri-sa×-rig7ki-ta im-ma-×en-na ‘when Me-ság came from the town of Irisa×rig’ (BIN 8, 137:4f.). 80 Kienast–Sommerfeld 1994:221 sub ištum. Throughout Akkadian, the starting point of a journey is rarely indicated in passages that employ alākum with the ventive marker (the destination point, introduced by ana, is specified more often, see CAD A1 306). The available (although scarce) evidence suggests that ištu was indeed used in such phrases, cf. anāku ištu GN allikam-ma PN a´bat-ma ‘I came from GN and seized PN’ (BIN 6, 188:9, OA). 81 Note that the verbal form is orthographically deviant in at least two points: ú-šu-ri-dam instead of the expected *u-śu-rí(or ri)-dam.

E. Markina, Akkadian of the Me-ság Archive

187

References Bridges 1981 Ellis 1979 Foster 1979 Foster 1980 Foster 1982a Foster 1982b Foster 1982c Foster 1989 Foster 1993

Frayne 1993 Gelb–Kienast 1990 Glassner 1983 Hasselbach 2005 Hecker 1968 Kienast–Sommerfeld 1994 Kienast–Volk 1995

Kogan–Markina 2006 Kutscher 1989 Markina 2011 Salgues 2011 Sallaberger 1996

Bridges, S. J. The Me-ság Archive: A Study of Sargonic Society and Economy. PhD. Diss. Yale. Ellis, M. de J. Cuneiform Tablets at Bryn Mawr College. JCS 31:30–55. Foster, B. R. New Light on the “mu-iti” Texts. Or 48: 153–162. Foster, B. R. Notes on Sargonic Royal Progress. JANES 12:29–42. Foster, B. R. Administration and Use of Institutional Land in Sargonic Sumer (Mesopotamia 9). Copenhagen. Foster, B. R. Umma in the Sargonic Period (Memoirs of the Connecticut Academy of Arts and Sciences 20). Hamden. Foster, B. R. An Agricultural Archive from Sargonic Akkad. ASJ 4:7–47. Foster, B. R. Old Akkadian nukkusu(m) ‘balance an account.’ NABU 1989/115. Foster, B. R. Select Bibliography of the Sargonic Period. Liverani, M. (ed.). Akkad, the First World Empire. Structure, Ideology, Traditions (HANES 5). Padova. Pp. 171–182. Frayne, D. The Early Inscriptions of Mesopotamia: Sargonic and Gutian Periods (RIME 2). Toronto. Gelb, I. J.; Kienast, B. Die altakkadischen Königsinschriften des dritten Jahrtausends v. Chr. (FAOS 7). Stuttgart. Glassner, J. J. Texts and Fragments. JCS 35:209–212. Hasselbach, R. Sargonic Akkadian. A Historical and Comparative Study of the Syllabic Texts. Wiesbaden. Hecker, K. Grammatik der Kültepe-Texte (AnOr 44). Roma. Kienast, B.; Sommerfeld, W. Glossar zu den altakkadischen Königsinschriften (FAOS 8). Stuttgart. Kienast, B.; Volk, K. Die sumerischen und akkadischen Briefe des III. Jahrtausends aus der Zeit vor der III. Dynastie von Ur (FAOS 19). Stuttgart. Kogan, L.; Markina, E. Review of Hasselbach 2005. B&B 3:555–588. Kutscher, R. The Brockmon Tablets at the University of Haifa. Royal Inscriptions. Haifa. Markina, E. Observations on Gasur Akkadian. Fs. Aa. Westenholz (in press). Salgues, E. Naram-Sin’s Conquests of Subartu and Armanum. Fs. Aa. Westenholz (in press). Sallaberger, W. Der babylonische Töpfer und seine Gefäße nach Urkunden altsumerischer bis altbabylonischer Zeit sowie lexikalischen und literarischen Zeugnissen (MHEM 3). Ghent.

188

Articles: Ancient Near Eastern Studies

von Soden–Röllig 1991 Sommerfeld 1999 Sommerfeld et al. 2005 Steinkeller–Postgate 1992

Stol 1995

von Soden, W.; Röllig, W. Das akkadische Syllabar (AnOr 42). Roma. Sommerfeld, W. Die Texte der Akkade-Zeit. 1. Das DijalaGebiet: Tuttub (IMGULA 3/1). Münster. Sommerfeld, W.; Markina, K.; Roudik, N. Altakkadische Texte in der St. Petersburger Eremitage. B&B 2:185–232. Steinkeller, P.; Postgate, J. N. Third-Millennium Legal and Administrative Texts in the Iraq Museum, Baghdad (MC 4). Winona Lake. Stol, M. Old Babylonian Cattle. BSA 8:173–213.

Philological Notes on the First Tablet of the Standard Babylonian Gilgameš Epic∗ Rim Nurullin Russian State University for the Humanities, Moscow

1. Lines 15–16 ´a-bat-ma gišsimmilta(KUN4) ša2 ul-tu ul-la-nu // qit-ru-ub ana E2.AN.NA šu-bat dIštar(15) ‘Take the stairway, which is from of old, // approach Eanna, the dwelling of Ištar.’ The reading of the logogram gišKUN4 (l. 15) as simmiltu ‘staircase’ is widely accepted in the recent translations of the Epic1 and seems to be preferred to the more traditional gišKUN4 = askuppatu ‘threshold.’2 According to George, “what is meant is a stairway on the wall, which the reader is invited to climb so that he can go up on to it” (George 2003:781). However, it is not clear how one, having ascended the city wall, could come nearer to Eanna. George explains this as follows: “Though E-anna is situated in the middle of Uruk, the topography of the town is such that there are stretches of city wall that take one nearer to the temple area” (ibid.). But as a glance on the plan of Uruk shows (v., for instance, Roaf 2000: 60), there are hardly any stretches of the wall which could be considered significantly closer to the temple area than the others. George’s interpretation of this passage depends closely on his treatment of the previous lines as dealing with the wall of Uruk: “The next couplet introduces Gilgameš’s one great concrete achievement, the wall of Uruk, which the epic thus holds up as an enduring monument of his fame (11–12) The conceit is that, rapt in admiration for the wall, one will climb on to it (13–21) …” (George 2003:446).

∗ I wish to thank L. Kogan who discussed with me many of the problems dealt with in the present article. I am also grateful to B. Alexandrov who provided me with some research materials otherwise inaccessible to me. My thanks go to РГНФ/RFH for its financial support (project No. 09-04-00234а). 1 ‘Prends donc l’escalier’ (Tournay–Shaffer 1994:40); ‘mount the stone stairway’ (Foster 2001:3); ‘take the stairway’ (George 2003:539); ‘nimm doch die Treppe’ (Maul 2008:46); ‘nimm doch die Treppe’ (Röllig 2009:35). 2 For the details v. Tournay–Shaffer 1994:41; George 2003:781.

190

Articles: Ancient Near Eastern Studies

However, lines 11–12 are commonly taken to mean that Gilgameš built the wall of Uruk and the wall of Eanna: [up-pi]š dūra(BAD3) ša2 uruk(UNUG)ki su-pu2-PriQ // ša2 E2.AN.NA qud-du-ši šu-tum4-mi el-PlimQ ‘He built3 the wall of Uruk, the sheepfold, // (and) of the holy Eanna, the pure storehouse.’4 3 The form uppiš in a Kuyunjik tablet (MS B3) is restored by means of the Babylonian manuscript h: Pu2Q-pi-šu2 (v. George 2003:779f.). According to George, in this case the D-stem may have an intensive function: “perhaps the building of a city wall, which would best be begun in several places at once, was in itself an intensive activity” (George 2003:780). George also cites another passage where dūru ‘wall’ is an object of the verb uppušu: BAD3 ša RN1 RN2 RN3 RN4 RN5 DUMU RN6 ab-ba-ia u2-up-pi2-šu-ni e-na-a¶-ma ‘The wall which RN1, RN2, RN3, RN4, RN5, son of RN6, my forefathers, had built had become dilapidated’ (Grayson 1987:101, ll. 5–8). The D-stem uppušu in this passage is commonly explained as expressing plurality of the subject (CAD E 232a; Kouwenberg 1997:148); note that the Gstem of epēšu is used in the same inscription when the subject is in the singular: ištu uš-še-šu a-di ša-ap-ti-šu e-pu-uš ‘I rebuilt (the wall) from top to bottom’ (Grayson 1987:102, ll. 10–11); ru-ba-u2 ur-ki-u2 e-nu-ma e-ep-pu-šu ‘a future prince, when he rebuilds (the wall)’ (ibid., ll. 12–14). Kouwenberg notes that the use of the D-stem in this passage “is connected with the remarkable structure of the subject, a series of five coordinated nouns, and has distributive meaning: it underlies that the action is performed by different subjects on successive occasions” (1997:148). Similarly, the D-stem has a tendency to indicate plurality of the object “if the plural constituent is not simply a plural noun, but consists of a series of coordinated nouns” (ibid. 170). The following example cited by Kouwenberg (ibid. 148, No. 91) illustrates such a use of epēšu D: ni-ik-ka-as-si2 LU2.LU2meš E2.GALlim u3 LU2meš mu-uš2-ke-nim u2-up-pi2-iš-ma ‘I have settled the accounts of the personnel of the palace and of the muškênu’s’ (ARM 27, 100:22f.). Kouwenberg explains the use of the D-stem in this passage as follows: “its use here might be caused by the fact that two different actions are involved, one for each category of personnel” (ibid. 148). Now, the presence of epēšu D in l. 11 of the first tablet of the Epic may find a similar explanation: it is used with two coordinated direct objects (the wall of Uruk and the wall of Eanna) and describes two different actions (the construction of each of these walls). 4 Cf. for instance ‘He built the wall of Uruk, the enclosure, // Of holy Eanna, the sacred storehouse’ (Heidel 1949:16); ‘Die Mauer um Uruk-Gart lies er bauen, // Um das heil’ge Eanna, den strahlenden Hort’ (Schott–von Soden 1988:15); ‘Er baute die Mauer von Uruk, der Hürden(umgehten), // die des hochheil’gen Eanna, des reinen Schatzhauses’ (Maul 2008:46). Note, however, the following comment on these lines in Gardner–Maier 1984:62: “ ‘of holy Eanna’ is a variation of ‘Uruk of the Sheepfold’ in the previous line and so designates the whole of the city.” One may wonder, therefore, whether the translations cited above should also be understood in this sense. In Foster’s translation, the second object of uppušu ‘to build’ is šutummu ellu ‘the lustrous treasury’: ‘He built the walls of ram-

R. Nurullin, Philological Notes on the First Tablet …

191

George’s rendering of these lines is probably based on the interpretation put forward by P.-A. Beaulieu. In Beaulieu’s view, E2.AN.NA (to be read as ayakku in Akkadian) in l. 12 “stands in apposition to Uruk, as an alternative name for the city Ayakku is evidently a metonymy for Uruk” (Beaulieu 2002:39). If Beaulieu’s hypothesis is accepted, it is indeed conceivable that the wall around the temple area is not mentioned at all in the Epic of Gilgameš. Beaulieu’s treatment of E2.AN.NA(= ayakku) in our passage is based on the evidence from the Basetki inscription of Narām-Sîn (Frayne 1993: 113f.), where Eanna’s name clearly substitutes the name of Uruk. This inscription lists the gods which were asked by the citizens of Akkade to make Narām-Sîn the god of their city. The name of each god is accompanied by the name of the city where he/she was worshiped. The only exception is Ištar whose name heads the list: instead of the name of the city of Uruk we find Eanna (v. Beaulieu 2002:38f.). It is possible, however, that this usage is an ad hoc peculiarity of the Basetki inscription. As Beaulieu himself correctly observes, Narām-Sîn was deified because he succeeded to subdue the Great Revolt. Now, one of the main centers of this revolt was Uruk ruled by Amar-girid. In the Basetki inscription, as well as in other inscriptions of Narām-Sîn, the victory is imputed to the interference of Ištar (v. Beaulieu 2002:38). It seems reasonable to suppose that the author of the Basetki inscription did not want to associate the name of the goddess with the rebellious Uruk and replaced it by the name of her temple. Thus, the substitution of Eanna for Uruk in the Basetki inscription can hardly be used as supporting evidence for a new interpretation of the present passage from the Epic. An interpretation similar to Beaulieu’s has been recently proposed by A. Zgoll (2010:453ff.). She is also of the opinion that Eanna in the passage under discussion is metonymically used for Uruk, but her arguments are different from those put forward by Beaulieu. Zgoll’s conclusion is based primarily on the fact that in the prologue to the Epic the name of Uruk and that of Eanna are found either in parallel passages or just comparatively close to each other: “Der Anfang des Gilgamesch-Epos spricht zweiparted Uruk, // The lustrous treasury of hallowed Eanna!’ (Foster 2001:3). However in three out of four extant manuscripts šutummu ellu is in the genitive: šutum4-mi e[l-lim] (B3); […] el-PlimQ (F3); e[l?]-PlimQ (d1) (cf. ša2 E2.AN.NA qud-du-šu šutum3-mu […] in MS h). It seems more likely, therefore, that šutummu ellu ‘the pure storehouse’ is an attribute of E2.AN.NA quddušu ‘the holy Eanna,’ which is in the genitive because of the preceding nota genitivi ša.

192

Articles: Ancient Near Eastern Studies

mal vom Ištartempel Eana, einmal in 1:12 und einmal in 1:16. An beiden Stellen zeigt sich der Tempel Eana begleitet und in Parallele zur Stadt Uruk” (ibid. 453). In the first case (l. 12), where Uruk and Eanna are mentioned in syntactically parallel passages, Eanna may—but certainly need not—be taken as a synecdoche for Uruk. However, there is definitely no obvious link between the name of the temple in l. 16 and that of the city in l. 18, and the mere fact that they occur not too far from one another is not sufficient to support Zgoll’s hypothesis. A more important piece of evidence, strangely disregarded by both Beaulieu and Zgoll, comes from another passage from the first tablet of the Epic, where Šamḫat invites Enkidu to go to Uruk (ll. 209–210): al-ka lu-[t]ar-ru-ka ana libbi(ŠA3)bi uruk(UNUG)ki su-pu2-ri // a-na bīti(E2) Pel Q-lim mušab da-nim u diš-tar ‘Come, let me lead you to Uruk-the-Sheepfold, // to the pure temple, the dwelling of Anu and Ištar.’5 In this passage, it is indeed quite conceivable that Eanna is used as an alternative designation of the city. This suggestion becomes even more probable when one realizes that in Enkidu’s response to Šam¶at (l. 217) only the temple’s name is mentioned. All in all, the assumption that Eanna in the prologue to the Epic may be equivalent to the city of Uruk as a whole is probably not completely groundless. However, it does not help us to solve the problem raised in the beginning of this note: was it indeed possible ‘to approach Eanna’ by climbing ‘the wall of Uruk’?6 As long as the traditional reading of l. 12 is accepted, the next two lines (13–14) can be easily interpreted as referring to the wall of Eanna: a-mur du-ur-šu2 ša2 ki-ma qe2-e ni-ib-¢[i?] // i-tap-la-as sa-me-ta-šu ša2 la u2-mašša2-lu mam-ma ‘Behold its (i. e. Eanna’s) wall which is like a shining copper,7 // gaze at its parapet which nobody can replicate.’8 In that case, the 5

Note that MS cc from Uruk adds two more lines (ll. 210a–210b in George’s edition): […]-ma ana libbi(ŠA3) uruk(UNUG)ki su-pur // [ana E2.AN.N]A qud-du-šu2 muša2-bu ša2 dištar(15) ‘[…] to Uruk-the-Sheepfold, // [to] holy [Eann]a, the dwelling of Ištar.’ 6 Note also that within Beaulieu’s interpretation the use of the name of Eanna turns out to be inconsistent in this—very short and structurally homogeneous— segment of the prologue: in l. 12 Eanna substitutes Uruk, but in l. 16 it stands for its own. 7 For this reading see the following note. 8 To be sure, this interpretation of ll. 13–14 is not entirely new. One can mention Schott’s discussion about outer (dūru) and inner (samītu) walls of Eanna (Schott 1934:93f.); cf. also Wiseman’s account of these verses: “The reader is in-

R. Nurullin, Philological Notes on the First Tablet …

193

stairs (l. 15) are probably to be considered as leading up to the top of the ziggurat of Eanna. Only after having visited the sanctuary (ll. 15–17), the reader is supposed to climb on the city wall and walk around (l. 18). Within such a reading, the fragment under discussion is organized chiastically. First, the wall of Uruk is mentioned ([uppi]š dūra ša Uruk, l. 11), after that, the wall of Eanna ([uppi]š dūra … ša E2.AN.NA, l. 12). In the following lines the attention of the reader is drawn first to the wall of Eanna (ll. 13–14) and to the temple itself (ll. 15–17) and then to the wall of Uruk (ll. 18–21) and to the city as a whole (ll. 22–23).9

vited to look at the walls of the unrivalled sanctuary of Eanna and at the outer walls of the city of Uruk which Gilgameš as king of Uruk had built” (Wiseman 1975:157). Among recent translations, this reading is adopted by Maul, as far as one can judge from his note on l. 11: “Denn mit der in I, 11 und 18–21 gerühmten Mauer von Uruk bot Gilgamesch den in der Stadt lebenden Menschen den notwendigen Schutz vor äußeren Feinden …” (Maul 2008:154). 9 Chiastic structure can be recognized in two other passages from the prologue. The first passage describes the wall of Uruk (ll. 19–21): te-me-en-nu ¶i-i¢-ma (A) libitta(SIG4) ´u-ub-bu (B) // šum-ma libitta(SIG4)-šu2 la a-gur-PratQ (B) // u uš-šu2-šu2 la id-du-u2 7 PmunQ-tal-ku (A) ‘Examine the foundation document (A), inspect the brickwork: (B) // Is not its brickwork of kiln-fired brick? (B) // And did not Seven Sages lay its foundations? (A).’ For temmennu in l. 19 as ‘foundation-deposit,’ ‘foundation-document’ cf. Wiseman 1975:157f. (i. e., upon reading the foundation inscription one may learn that the foundation of the wall of Uruk was laid by the Seven Sages). The second passage (ll. 31–34) comes from the “old” prologue. It describes Gilgameš’s ability to attack an enemy and also to defend his people. These qualities are first depicted in rather straightforward terms (ll. 31–32): ašaridu ‘leader’ (A), tukulti a¶¶ēšu ‘the trust of his brothers’ (B). After that, a few metaphoric descriptions are used and the order of the qualities is reversed (ll. 33– 34): kibru dannu ‘mighty bank’ (B), agû ezzu ‘violent flood-wave’ (A). For agû ‘wave’ used metaphorically to describe a destructive force v. Streck 1999:105, No. 124; ibid. 112, No. 143, with further references. ‘Bank’ (kibru) as a metaphor for protection is poorly attested. George explains it as follows: “the imagery is drawn from riverine navigation, in which the bank offers safe haven in a storm or other difficulty” (George 2003:783). He compares it with kibru in personal names (e. g. Ilī-kibrī ‘My god is my bank’). However, kibru in these names may rather be related to the West Semitic root k-b-r ‘to be big, great’ (v. Streck 1999:105; 2000: 267, 269). A more interesting parallel favoring George’s interpretation comes from the hymn to Nabû, where agû ‘wave’ and kibru ‘bank’ are also used side by side. In this text, the victim of Nabû’s fury is likened to a person who is drowning in the stormy water, the safe riverbank being too far from him: ina gi-piš e-de-e nadi-ma a-gu-u2 e-liš it-ta[k-kip] // kib-ri ru-uq-šu2 ne2-si-iš na-ba-l[u] ‘He is swept away by the onrush of the tide, a wave overwhelms him; // the bank is remote from him, the dry land is at distance’ (von Soden 1971:52, ll. 49f.).

194

Articles: Ancient Near Eastern Studies

2. Line 13 a-mur du-ur-šu2 ša2 ki-ma qe2-e ni-ib-¢[i?] ‘Behold its wall which is like a shining copper.’ The second half of this line has been variously interpreted. Ever since Ebeling 1932–1933, the last word in the line was generally read as NI-IB¶[u-šu], identified with either nēbe¶u ‘belt,’ ‘frieze’ (with a phonetic variant neb¶u)10 or nip¶u ‘flash,’ ‘blaze,’ ‘rising (of celestial bodies).’11 As George (2003:780) points out, this reading seems very unlikely for two reasons. First, on a Babylonian tablet (MS h) there is space for only one sign after NI-IB. Second, the traces of this sign in the manuscript F3 seem to exclude the reading ¶[u] (as well as ¶[a] or ¶[i]). The rendering of the word in question thus crucially depends on one’s interpretation of the last sign in the sequence NI-IB-PxQ. The preceding word is to some extent problematic as well, for one has to choose between two homonyms: qû ‘thread’ or qû ‘copper.’ George restores the last word of the line as nipšu (ni-ip-š[u?]) ‘combed wool,’ ‘tuft of wool,’ ‘fluff ’ (CAD N2 248)12 and opts for qû ‘thread’: ‘See its wall which is like a strand of wool’ (George 2003:539). George’s interpretation implies that a cord or belt could be used as an image of the wall. For George, the evidence for such imagery comes from an inscription of Nabopolassar where the wall of Babylon is described as e-bi-i¶ dannum ‘a mighty cincture’ (Al-Rawi 1985:10 ii 41; George 1991a; 1991b; 2003:781). However, with Farber 1991, it seems more convincing to interpret e-bi-i¶ in this passage as the name of the mountain Ebi¶.13 Another piece of evidence is thought to come from Lugalbanda Epic II where “the wall of Uruk is compared with the drawstring of a bird snare

Cf., for instance: ‘… deren nib¶u (Fries) wie Kupfer ist!’ (Ebeling 1932– 1933:226); ‘… qu’entoure une frise pareille au cuivre’ (Tournay–Shaffer 1994: 40). The following translations try to clarify the meaning of the simile by explaining the tertium comparationis: ‘… deren Friese wie Bronzeschalen scheinen!’ (Schott–von Soden 1988:15); ‘… whose facing gleams like copper’ (Foster 2001:3). 11 Cf. Heidel’s translation: ‘… whose brightness (ni-ip-¶[u-šu]) is like (that of ) copper!’ (Heidel 1949:16). 12 George defines this word as follows: “The word nipšu appears to signify one of the strands that are the result of pulling apart (napāšu) a tuft of wool (itqu), and thus a stage in the process of turning raw wool into woolen thread” (George 2003:781). 13 An ultimate association between the mountain’s name and the word ebī¶u ‘belt’ has been surmised, however (Farber 1991; Vanstiphout 1991). 10

R. Nurullin, Philological Notes on the First Tablet …

195

stretched out over the plain” (George 2003:781). But this comparison is conditioned by the specific context of that passage, where the wall of Uruk is described as a barrier against the Amorite invasion: ki-en-gi kiuri ni×in2-na-a-ba mar-tu ¶u-mu-zi bad3 unuki-ga gu-mušen-nagin7 eden-na ¶e2-ni-la2-la2 ‘In all Sumer and Akkad the Martu-people had arisen, (but) the wall of Uruk stretched out on the plain like a bird snare’ (Wilcke 1969:118, ll. 303–305; 124, ll. 369–371). Thus, the point of the simile is that the wall of Uruk acted as a snare and blocked the invasion.14 An image of this kind is quite unlikely to be used to describe the wall in the prologue of the Epic, especially if we assume that the line under discussion refers to the wall of Eanna rather than to the wall of Uruk (see the preceding note). Last but not least, the semantic link (tertium comparationis) between the wall (dūru) and the thread of (combed/plucked) wool (qê nipši) is rather hard to imagine, and it is by no means surprising that George himself regards this image as “not wholly convincing.” I propose to read the last sign of the line as DI15 and to interpret the word in question as nib¢u ‘shine.’ The genitive in the phrase qê nib¢i is to 14

Note that a physical similarity between the wall and the bird snare is also conceivable: “The image is of a long net, of a type used for snaring low-flying birds, extended across the open country. The point here is the length of the wall, and its effectiveness, which are interpreted as a symbol of Unug’s security and power” (Black 1998:105). It is highly important, however, that the wall described in Lugalbanda Epic is probably not to be identified with the city wall of Uruk. As rightly acknowledged by Black, “historically it is an image appropriate to the period of the Third Dynasty of Ur (ca. 2168–2060 B. C.), when such long cross-country walls were actually built to keep the incursive Mardu (Amorite) tribes out of Sumer. It seems that the image is drawn from such a crosscountry wall, and not from a wall encircling a city” (ibid. 105f.). 15 On the basis of Haupt’s copy (1884–1891:6, No. 1f.) as well as on his own impression (George 2003, pl. 41, MS F3) George describes the traces of this sign as follows: “a sign beginning with two horizontals, the lower preceding the upper” (ibid. 2003:780). To my eye, the lower “horizontal” wedge may also be taken for the beginning of an angle wedge, while the upper one may be the top of a vertical wedge, so that the whole is to be identified with the first two elements of the sign DI. A similar conclusion was reached by R. C. Thompson. On his copy (Thompson 1930, pl. 1) the angle wedge and the lower part of the vertical wedge are seen (the sign was read by Thompson as ŠI). The remains of the sign on a Babylonian tablet (George 2003, pl. 48, MS h, absent from Pinches’s copy, CT 46, No. 17) could be interpreted as the beginning of a lower angle wedge which characterizes the sign DI and similar signs in this manuscript, cf., for example, DI in lines 102 ([ ¢i]-Pi¢ Q-¢i), 286 (i-de), and KI in lines 11 (UNUGki), 293 (ki-ma ki-´ir).

196

Articles: Ancient Near Eastern Studies

be understood as an attribute of qû ‘copper’: ‘the copper of shining’ meaning ‘the shining copper.’16 Thus the wall of Eanna is depicted as shining like copper.17 Shining edifices are frequent in Mesopotamian building inscriptions.18 Cf. for instance the following passage from Nabonid’s inscription where the root n-b-¢ also occurs: ekurra(E2.KUR) šu-ati ul-tu te-me-en-ni-šu a-di gabadibbî(GABA.DIB)bi-šu e-eš-ši-iš ab-ni-ma kaspa(KU3.BABBAR) u3 ¶urā´a(KU3.SI22) igārātī(E2.GAR8)meš-šu u2-šal-biš-ma u2-ša2-an-bi-i¢ dšaššāniš(UTU.ŠI)niš “I rebuilt this temple from its foundation to its parapet I coated its walls with silver and gold and made (the walls) shine like the sun” (Schaudig 2001:419, Text 2.12 11 ii 4ff.).19 3. Line 220 a-na-ku lu-ug-ri-šum-ma da-an-n[iš] lu-gab-Pbir?-šu2?Q ‘I, myself, shall fight him and overpower him by force.’ 16

It remains to be determined whether there are any restrictions on the use of attributive genitive in Akkadian. According to von Soden, the choice between an adjectival noun phrase and a genitive one is based on stylistic grounds (GAG3 § 136f and 186c). For some observations on this question see further Buccellati 1976. 17 Brightness is undoubtedly one of the prototypical qualities of copper in the ancient Mesopotamian literary language. In the lexical lists qû ‘copper’ is coupled with namru ‘bright’; ebbu ‘polished,’ ‘shining’; ellu ‘clean’; mušālu ‘mirror’ (v. CAD Q 291a qû, lex. section). There are also passages in which a man’s purity is compared to the polished copper. Cf. for example: amēlu(LU2) mut-tal2-li-ku ina ni-iq ree-me u šul-me ki-ma qe2-e maš-ši lim-ma-šiš ‘May the restless man be polished like polished copper through a sacrifice to obtain mercy and well-being’ (Borger 1969: 15, ll. 271f.). One cannot exclude that qê nib¢i ‘shining copper’ in our passage also refers to polished copper. 18 A bright coating material (like silver or gypsum) is often mentioned as the reason for the shining (cf. the passage from Nabonid’s inscription cited below). 19 Finally, it should be noted that the interpretation suggested here is consonant with a rendering of the line under discussion by a few other scholars. To the renderings of this line collected in fnn. 10 and 11 one may add Maul’s recent translation: ‘Sieh an dessen Mauer, die wie Kupfer glänzt!’ (Maul 2008:46, the italics are Maul’s). It is not clear to me what reading of the end of the line stands behind Maul’s translation. Obviously, it differs from ni-ip-¶[u-šu] (v. fn. 11), as well as from ni-ib-š[u] (suggested by M. Kovacs, v. George 2003:780), since both these readings have been convincingly rejected by George (ibid. 780f.). Cf. also the rendering of the end of the line by P. Haupt (1901:12): ‘his … which [shines] like brass’ (one may observe, in passing, that Haupt also managed to restore [iš]dī māti in the first line of the Epic, a reading which was recently confirmed by the discovery of the fragment Rm 956, v. George 2003: 392, 778).

R. Nurullin, Philological Notes on the First Tablet …

197

Ever since Jensen 1900, the verbal form in the end of the line is commonly restored as lu-qab-[bi-ma].20 The form is preserved in the last line of the fourth column of a Kuyunjik tablet (MS P in George’s edition). To judge from George’s copy (George 2003, pl. 45), the tablet has suffered deterioration, and as a result the last signs of the line are now almost indiscernible. But on the earlier copies the signs LU and GAB are clearly visible.21 The restoration lu-qab-[bi-ma] is questionable. The form luqab[bi] is generally parsed as 1st person precative of the D-stem of qabû ‘to say, speak,’ translated as ‘I shall speak’ (cf. the renderings of the line collected in fn. 20). Within this interpretation, one is forced to disregard the fact that the D-stem of qabû is virtually unattested: the only relatively reliable example comes from the Old Babylonian prayer to Anūna ([x x] x un-ne2e-ni ša-nu-¶i-iš u2-q[a2?]-ab-ba ‘Exhausted he laments …,’ cited from Lambert 1989:327, l. 97).22 From this passage one might infer that qubbû meant something like ‘to lament,’23 but such a meaning would scarcely fit the context of the line under scrutiny.

20 a-na-ku lu-uk(£)-ri-šum-ma da-an-[niš] lu-£ab-[bi-ma] ‘Ich will ihn laden und will mächt[ig] ruf[en und] …’ (Jensen 1900:128f. iv 47; according to Jensen (ibid. 128, fnn. 7 and 8), one may also restore da-an-[nu] ‘Gewaltiger’ and lu-qab-[bi-šu] ‘ihm (zuschreien?)’). Cf. Thompson’s reading of the line: a-na-ku lu-u£-ri-šumma da-an-n[iš l]u-£ab-[bi-ma] (Thompson 1930:14 iv 47). Cf. also: ‘I, I will summon him and [will] speak bold[ly]’ (Heidel 1949:22); ‘Moi, je veux le provoquer, lui lancer un défi’ (Tournay–Shaffer 1994:57, with the reading lu-ug-ri-šum-ma daan-[niš] lu-qab-[bi-ma], ibid. fn. 61); ‘I myself will challenge him, [I will speak out] boldly’ (Foster 2001:9). 21 V. Haupt 1884–1891:12, No. 3 iv 47; Thompson 1930, pl. 5 iv 47. Note that in George 2003:550 the second half of the line is read as da-an x x x x. George does not take into account the reading lu-qab-[bi-ma], but makes the following remark on the reading lu qab-[lu] suggested by von Soden (1959:222): “The traces at the end of the line (MS P) do not appear to allow lu qab-lu” (George 2003:800). 22 Note that according to the copy (PBS 1/1, 2 iii 46) the verb is to be read u2-qa2-ab-PbaQ. 23 This meaning may find a confirmation in the well-attested substantive qubbû ‘lamentation.’ Note also Moran’s reading of EA 11:5 (and cf. l. 11): [ultu aššat] abīka qu-ub-ba-tu4 ‘After your father’s wife have been mourned’ (Moran 1984:301; 1992:22, nn. 1 and 5). The form tu-qab-ba in a grammatical text (5R, 45 viii 32) is not pertinent for establishing the meaning of qubbû. On the possible relation between the meanings of qabû (‘to say, speak’) and qubbû (‘to lament’) v. Kouwenberg 1997:178 (sub nabû D), 283.

198

Articles: Ancient Near Eastern Studies

An alternative restoration was proposed by W. von Soden: lu qab-[lu] ‘(dann) sei Kampf!’ (von Soden 1959:222).24 However, the occurrence of the first-person precative (lu-ug-ri-šum-ma) in the first hemistich plainly suggests that the signs LU-GAB must also belong to a precative verbal form. In my view, the most probable restoration is lu-gab-Pbir?-šu2?Q ‘I will overpower him.’25 The verb gubburu (guppuru) ‘to overpower’ is rare, but its meaning is more or less uncontroversial.26 What seems to be particularly attractive in this restoration is that in the previous line (219) we find the verb gutaššuru.27 Now, in Explicit malku = šarru I 146f. the verbs guppuru and gutappuru are equated with guššuru and gutaššaru (v. Hrůša 2010:158, 435).28 Thus, Enkidu intends to treat Gilgameš (lugabbiršu) in the same way as Gilgameš himself treats (ugdašsar) the menfolk of Uruk. The line in question presents still another difficulty. The adverb danniš29 in this passage is commonly interpreted as an adverb of manner (‘mächtig,’ 24

Cf. also: ‘… heftig tobe der Kampf!’ (Schott–von Soden 1988:22). The traces of an angle wedge seen in George 2003, pl. 45 (in Thompson 1930, pl. 5 it looks like the beginning of a horizontal wedge) may be interpreted as the beginning of the sign BIR, while the upper part of the vertical wedge seems to be identified with ŠU2. 26 Thus, in an incantation against Lamaštu gubburu occurs in parallelism with nuppu´u ‘to crush, smite’: [g]u-ub-bu-ru u2-gab-bar [n]u-up-pu-´u u2-nap-pa-´u ‘She (Lamaštu) indeed overpowers, she indeed smites’ (PBS 1/2, 113 i 15, cited from CAD G 118a). 27 u3 ki-i rīmi(AM) ug-da-aš2-ša2-ru eli(UGU) Pe¢lūti(GURUŠ)Qmeš ‘And like a wild bull he (Gilgameš) acts overbearingly towards men.’ On the meaning of gutaššuru v. Kouwenberg 1997:322. Note that this line clearly echoes I 64: ug-da-aš2-ša2-ar rima-niš ša2-qu-u2 re-e-[ši] ‘He acts overbearingly like a wild bull, head held high.’ Both lines seem to allude to athletic contests in which the menfolk of Uruk is forcibly involved (v. Tigay 1982:185; Klein 2002:190). 28 Cf. also Explicit malku = šarru I 140f.: ga-ab-rum, ga-ap-rum = ga-aš-[rum] (Hrůša 2010:158, 435, on the repetition of ga-ab/p-ru v. ibid. 284, with further literature); Explicit malku = šarru I 144: mu-un-ga-ab-rum = mu-ug-da-[aš-rum] (ibid. 158, 435). Note also a similar (reciprocal) use of gutaššuru and gutabburu in the following two passages: šar-¶a nišū(UG3)meš ug-da-ša2-ra a-na a-ma-ri ka-a-ta ‘proud people vie with each other to look at you’ (Mayer 1976:492, l. 8; for the restoration šar-¶a v. Kouwenberg 1997:326); e¢-lu ug-ta-ap-pa-ru ki ša (text: ša ki) ar-ka-tim ‘the men vie with each other like lance-bearers’ (VAS 10, 214 iii 1f. (Agušaya A); the reading is adopted from Metzler 2002:730, for a different interpretation of this passage v. Groneberg 1981:119f.). 29 Note that in George’s view there is probably “too much space on MS B, between da-an and the trace, to read da-an-n[iš] with Thompson, and such a reading is in any case unlikely on orthographic grounds” (George 2003:800). George’s argu25

R. Nurullin, Philological Notes on the First Tablet …

199

‘boldly,’ etc.), but such a use of danniš is hardly ever attested elsewhere:30 normally, this word functions as an adverb of degree (‘very’). That means that danniš can only be used with verbs whose meaning can be intensified, but this is not the case of either gubburu ‘to overpower’ or qubbû (if the meaning to ‘speak’ is accepted). There seem to be three possible approaches to the interpretation of danniš in the present line. Perhaps the easiest solution is to stick to an adverb of manner, as actually done in the renderings collected in fn. 20. Another possibility is to abandon the very reading danniš and to restore instead a form of the adjective dannu ‘strong,’ ‘mighty,’ as proposed by Jensen (v. fn. 20).31 But from the copies of Thompson and George it is clear that the sign following da-an- begins with an angle wedge.32 Hence, the readings da-an-n[u], da-an-n[a], da-an-n[i] as well as da-an-n[am] must be ruled out. What is left, then, is da-an-n[um], but in the Standard Babylonian version of the Gilgameš Epic the sign NIM is hardly ever used to render case endings.33 The third possibility is to assume that Enkidu’s speech needs not be grammatically blameless. At this point of the narrative, Enkidu is on the way to becoming a civilized man, but he is still more a wild creature than a fully civilized human. It is, indeed, the first time that he says something at all and it is conceivable that his speech could be intentionally depicted ments are undoubtedly weighty, but probably not sufficient to reject the reading daan-n[iš]. Thus, in another line of the same manuscript (l. 241) there is also some space between the signs DINGIR and UD, but the reading dŠamaš(UTU) seems indisputable. What is likely meant by “orthographic grounds” is that in Akkadian texts of the first millennium danniš (as well as dannu, dannatu, etc.) is regularly spelled with the sign KAL(dan). Indeed, in late texts such spellings as dan-nu or dan-niš became prevalent. However, the writings with da-an- are also found: cf. da-an-nu (SBH 77, 44:15); da-an-ni (KAR 119:14); da-an-na-ta-a-nu (ABL 633:15). 30 To my knowledge, outside the present line from the Epic there is only one instance of danniš functioning as an adverb of manner: ta-an-ni-iš tētepšanni fPN u mimmama lā taddinanni ‘You have mistreated me by not giving PN or anything at all to me’ (HSS 14, 27:3, cited from CAD D 92b). Needless to say, this example from a Nuzi document can scarcely be taken into consideration in order to ascertain the good Akkadian usage of this adverb. 31 Cf. also Ungnad’s translation: ‘Ich selbst will ihn fordern, will den Mächtigen herbeirufen’ (Ungnad 1921:73, the italics are Ungnad’s). 32 Thompson 1930, pl. 5 iv 47; George 2003, pl. 38; absent from Haupt’s copy (Haupt 1884–1891:12, No. 3 iv 47). 33 To be precise, there is no evidence of NIM being used with the values ni7 or nu3, but cf. a-mur-din-nim-ma in XI 248, with mimation preserved before -ma. Note also that NIM is commonly used to write the divine name da-num/da-nim.

200

Articles: Ancient Near Eastern Studies

as not yet fully developed: Enkidu intends to say that he will beat Gilgameš by force, but chooses the wrong word (danniš) to express his idea.34 There is hardly any need to emphasise that our hypothesis, presupposing that the author of the Epic could purposely introduce grammatical and stylistic infelicities into Enkidu’s speech (ll. 216–223), must be taken with utmost caution. In order to justify such an assumption, this segment of the text has been scanned for further possible parallels. Enkidu’s first speech runs as follows35: 216. 217. 218. 219. 220. 221. 222. 223.

al-ki fšam-¶at qi2-ri-i[n-n]i ia-a-ši a-na bīti(E2) el-lim qud-du-ši mu-šab da-nim dPiš-tarQ a-šar dGIŠ-gim2-maš git2-ma-lu e-mu-qi2 u3 ki-i rīmi(AM) ug-da-aš2-ša2-ru eli(UGU) Pe¢lūti(GURUŠ)Qm[eš] a-na-ku lu-ug-ri-šum-ma da-an-n[iš] lu-gab-Pbir?-šu2?Q [lul-tar(?)]-ri-iḫ ina libbi(ŠA3) uruk(UNUG)ki a-na-ku-mi dan-nu […]-um-ma ši?-giš?-tu2 u2-nak-kar [ša2 i-n]a ´ēri(EDIN) i’-al-du [da-a]n i-mu-qi2-i-šu

216. 217. 218. 219. 220. 221. 222. 223.

‘Come, Šam¶at, invite me to the pure temple, the holy dwelling of Anu and Ištar, where Gilgameš is perfect in strength, and like a wild bull acts overbearingly towards men. I, myself, shall fight him and overpower him by force! [I shall vaunt] myself (?) in Uruk, (saying) “I am the mightiest!” [The migh]ty will change (the course of) the fighting, [the one] born in the wild is mighty in strength!’

Strictly speaking, only five of the eight lines belong to Enkidu (ll. 216, 220–223), whereas the remaining ones merely repeat the words addressed to him by Šamḫat (ll. 207–212).36 Except for danniš, no further mistakes can be reliably detected in the speech of Enkidu.37 Still, there is As long as the wrong use of danniš in Enkidu’s speech is admitted as a possibility, it need not be taken as an ad hoc literary device invented by the writer: this kind of mistake may well be familiar to him (and other native speakers of Babylonian) from the speech of foreigners. 35 A few philological notes on ll. 216, 222, 223 are collected in the Appendix. 36 217 // 210 (cf. also l. 210b in MS cc); 218–219 // 211–212. One may wonder whether these lines were added to the present passage by the compiler(s) of the first-millennium version. His/their inclination to the monotonous repetitions can best be illustrated by the comparison of the dream episodes in the Standard Babylonian (I 245–297) and Old Babylonian (OB II 1–43) versions of the epic (v. Cooper 1977; Tigay 1982:82ff.; this part of the Epic is now also known from a tablet of the Middle Babylonian period, v. George 2007). 37 The use of qerû in l. 216 seems peculiar (see Appendix), but this is probably due to our limited knowledge of the Akkadian lexicon. 34

R. Nurullin, Philological Notes on the First Tablet …

201

nothing to detract from the fact that the speech is stylistically rather idiosyncratic. Thus (with Gardner–Maier 1984:80, 83), one may observe a fairly dense concentration of the first person independent pronoun (ll. 216, 220, 221), apparently intended to put emphasis on Enkidu’s “I” (GAG3 § 41a). Another point of interest is the change from the first to the third person in l. 223.38 While similar examples are not completely unknown from Akkadian literature,39 their significance has never been ascertained, although it is hard to believe that they are pragmatically neutral. The expressiveness of Enkidu’s speech can be further recognized against the context of his dialogue with Šam¶at. Enkidu’s reaction to her words (ll. 207–212) is not fully adequate, as she did not actually invite him to rival Gilgameš.40 In what follows, Šam¶at apparently tries to correct Enkidu, declaring that he does not yet know life (l. 233) and that Gilgameš, the god’s favorite (ll. 241–242), is stronger than him (l. 238). In such a context, Enkidu’s speech can easily be perceived as baseless boasting. A certain redundancy of Enkidu’s style is thus in agreement with his excessive vanity. Similarly, the wrong use of danniš might be intended to reveal at once that Enkidu is actually no more than a barbarian whose pretentious words sound rather ridiculous. From l. 240 one may also infer Šam¶at’s ironic attitude toward Enkidu’s ardor: den-ki-du3 nu-uk-ki-ra še-ret-ka ‘O Enkidu, change your sinful intention.’ Whatever the exact meaning of šērtu may be,41 the D-stem of nakāru may well be seen as a mocking echo of unakkar in the speech of Enkidu (l. 222): […]-um-ma ši?giš?-tu2 u2-nak-kar ‘[The migh]ty (i. e. Enkidu himself) will change (the

38

Or in l. 222, see Appendix. Cf. the following passage from the Old Assyrian Sargon Legend (ll. 4–7): Adad dannūtam iddiššum-ma ištu ´īt šamšim adi erāb šamšim mātam a´bat-ma ‘Divine Adad gave him strength and as a result I took possession of the land from East to West’ (translation from Dercksen 2005:108). 40 As observed by George (2003:451), Enkidu’s response is also hardly compatible with the statement that ‘his wise heart was seeking a friend’ (mūdû libbašu iše’’a ibra, l. 214). 41 šērtu normally means ‘misdeed, sin’ or ‘punishment.’ None of these meanings seems to fit the context of the line. In the modern translations of the Epic šērtu in this passage is generally taken to designate the evil intent of Enkidu: ‘… get rid of your sinful intention’ (George 2003:553); ‘… lasse ab von deinem frevelhaften Plan!’ (Maul 2008:54). However, other interpretations are also possible, cf. ‘… wandle deinen Zorn!’ (Jensen 1900:131, with commentary on pp. 340f.); ‘… change your boorish ways’ (CAD Š2 324b). 39

202

Articles: Ancient Near Eastern Studies

course of ) the fighting.’42 Šam¶at seems to advise Enkidu that it would be better for him to change his wrongful thinking.43 Appendix: Philological notes on ll. I 216, 222, 223 Line 216. This line corresponds to l. 209, where Šam¶at address Enkidu: al-ka lu-[t]ar-ru-ka ana libbi(ŠA3)bi uruk(UNUG)ki su-pu2-ri ‘Come, let me lead you to Uruk-the-Sheepfold.’ In view of the parallelism itarrû (l. 209)44 || qerû (l. 216), many translators render qerû as ‘to take along,’ ‘to escort.’45 Such a meaning of qerû is, however, otherwise unattested,46 whereas the ordinary meaning of this verb (‘to call,’ ‘to invite’)47 does not fit the context very well, as it is not clear why it is necessary for Enkidu to wait for Šam¶at’s invitation to come to Uruk (besides, she has, in fact, already invited him in ll. 209ff.). No obvious solution for this problem suggests itself at present. Line 222. The first word of this line is commonly restored either as the first person singular pronoun ([a-na-ku]-um-ma)48 or as a form of erēbu ‘to enter.’49 One may suggest, alternatively, that Enkidu is speaking of 42

See further below in the Appendix. The pun was probably understood by M. Kovacs (1989:11), who translates: ‘Enkidu, it is your wrong thoughts you must change!’ Cf. her translation of l. 222: ‘Lead me in and I will change the order of things’ (ibid. 10). 44 The verbal form in l. 209 is usually interpreted as a G-stem form of (w)arû, cf. lu-[u]-ru-[u]-ka (Jensen 1900:128 iv 36); lu-u(?)-ru-ka (Thompson 1930:14 iv 36); lu-Pu2Q-ru-ka (CAD A2 314a). According to George (2003:799), the most probable reading of the broken sign is TAR (lu-[t]ar-ru-ka, presumably from (w)arû Gt). 45 Cf., for instance: ‘Come, O prostitute, take me …’ (Heidel 1949:22); ‘Viens, fille de joie, escorte-moi …’ (Tournay–Shaffer 1994:57). 46 According to CAD Q 242f., qerû with the meaning ‘to take away, along, to lead away’ can also be found in the passages where ilū ‘gods’ appear as its subject, more concretely, in the phrase PN ilū-šu/ša iqterû-šu/ši ‘PN—his/her gods have called him/her’ (a euphemism for dying). This interpretation is based on the comparison with ūm šīmātūša itarrâši ‘when her fates lead her away’ in UCP 10, 173, No. 105:7f. (cited from CAD Š3 17b). However, the supposedly parallel use of tarû ‘to lead away’ in this passage is not sufficient for establishing the meaning of qerû in the formulaic phrase under scrutiny, which can still be easily read as ‘(im Tode) abberufen’ (AHw. 918b). 47 Cf. ‘Auf, Freudenmädchen! Lade (mich,) mich …’ (Jensen 1900:129, the italics are Jensen’s; cf. also his commentary on p. 430); ‘Komm, Hure, lade du mich ein!’ (Schott–von Soden 1988:22); ‘Komm, Schamchat, mich lade ein …’ (Maul 2008:53). 48 Cf., for instance, Thompson 1930:14. Note that anāku with non-etymological mimation before -ma is also found in SB Gilg. XI 123. 49 Cf. [er-ru]-um-ma ‘ich werde eintreten und’ (von Soden 1959:222; cf. already Jensen 1900:128); [lu-ru]-um-ma ‘I will enter’ (CAD Š3 13a). 43

R. Nurullin, Philological Notes on the First Tablet …

203

himself in the third person, as is undoubtedly the case in the following line. The damaged word ([…]-um-ma) may then be interpreted as a noun in the nominative case (with mimation preserved before -ma), by means of which Enkidu describes himself. As far as the second word is concerned, ever since Jensen 1900:128 the reading ši-ma-tu2, generally interpreted as šīmātu ‘fates,’ ‘customs,’50 has been usually taken for granted.51 However, a look at the copies of Haupt (1884–1891:5, l. 22) and George (2003, pl. 38, MS B1, fragment K 2756) reveals that this reading is far from certain: on Haupt’s copy one can find the signs PI GIŠ UD,52 whereas on George’s copy the signs IGI53 GA2 UD are seen. Evidently, the passage is in need of collation, but a new—of necessity preliminary—interpretation of the line can still be ventured. Combining the evidence of the two copies, the difficult word can be read as šigištu (ši-giš-tu2), whose only attestation recognized hitherto comes from § 47 of the Laws of Eshnunna: šumma awīlum ina šigištim awīlam IG-te-el 10 šiqil kaspam išaqqal ‘If a man has wounded (?) a man in a fight (?), he shall pay 10 shekels of silver.’

This section belongs to the provisions dealing with bodily injuries. It is preserved on a tablet from Eshnunna (Tablet A in Goetze’s edition, v. Goetze 1956:3) as well as on a fragment from Tell Haddad (Al-Rawi 1982, v. also Roth 1990). Both the noun šigištu and the verbal form IG-teel are difficult to interpret. While the root of the verbal form is not attested anywhere else,54 šigištu is with all probability derived from šagāšu ‘to murder, to slaughter.’ However, the meaning ‘killing, slaughter’ does not fit the context: the mildness of the penalty (10 shekels), as well as the occurrence of risbatu ‘brawl (?)’ in the next section of the fragment from

50 Cf., for instance, [lūr]umma ši-ma-tu2 unakkar ‘I will enter (Uruk) and change the customs’ (CAD Š3 13a). However, as rightly observed by George (2003:800), “the spelling ši-ma-tu2, if correctly read,” may also be “taken as an example of a literary singular, šīmatu.” 51 Thus, the copies of Thompson (1930, pl. 6 v 2) and Parpola (1997:7, l. 205) both show a clear MA. 52 Cf., however, Haupt’s note to the copy: “Könnte allenfalls auch IGI MA UD sein.” 53 The initial angle wedge of the sign IGI is probably placed too high. 54 IG-te-el seem to represent the perfect of a middle weak verb, cf. Al-Rawi 1982:120; Roth 1990:70f.

204

Articles: Ancient Near Eastern Studies

Tell Haddad,55 make it likely that § 47 of the Laws of Eshnunna should be seen as a parallel to § 206 of the Laws of Hammurapi: šumma awīlum awīlam ina risbatim imta¶a´-ma simmam ištakanšu awīlum šū ina īdû lā am¶a´u itamma u asâm ippal ‘If a man struck a man in a brawl (?) and inflicted upon him a wound, that man shall swear, “I did not strike purposely,” and he shall pay the physician.’

It seems probable that both texts refer to a mitigating circumstance of the injury. Given the fact that risbatu is commonly interpreted as a designation of fight of which both parties are equally guilty,56 also šigištu is likely to be taken to mean ‘fight, brawl.’57 With due caution, l. 222 can be read as follows: […]-um-ma ši-giš-tu2 u2nak-kar ‘[The migh]ty (i. e. Enkidu) will change (the course of ) the fighting.’ What is meant is probably not merely a brawl, but rather athletic contests between Gilgameš and the menfolk of Uruk which were reported to Enkidu by Šamhat (l. 212//219, v. fn. 27).58 As is well known, these contests are portrayed in the Epic as oppressing the citizens of Uruk (Tigay 1982:184ff.; Klein 2002:190f.; George 2003:449). So far, Gilgameš could permanently triumph over his subjects, but now the course of the events may, for the first time, be changed.59 Line 223. Here we are faced with a unique combination of two stock phrases. The first half of the line ([ša in]a ´ēri i’’aldu) recalls the frequent description of Enkidu as born in the wild, cf. wa-al-da ´ērim(EDIN) Pmi-itlu-kam i Q-l [e]-i ‘The wild-born was able to give counsel’ (OB Harmal2 47); wa-al-dam ´e!-Pri-imQ mi-it!-lu-ka-am i-le-’i3 ‘The wild-born was able to give counsel’ (OB IM 19); [š]a2 i’-al-dam-ma ina ´[ēri(EDIN) mit-lu-ka i-le-’i] ‘[The one] who was born in the w[ild was able to give counsel]’ (SB IV 26, and cf. IV 107). In all these passages, it is Enkidu’s cleverness that is associated with the fact that he was born in the wild (v. George 2003:819). In our case, however, it is rather his strength that is in issue. 55

This section is absent from the Tablet A. It is, however, more likely that it was omitted from this source rather than added to the fragment from Tell Haddad (Yaron 1988:28f.). 56 V., for example, Driver–Miles 1952:412 and Westbrook 1992:61ff., where similar provisions from Exodus and the Hittite Laws are also discussed. 57 Cf. Al-Rawi 1982:118 (‘brawl’), Roth 1990:70 (‘fray (?)’). 58 As regards the provisions from the Laws of Hammurapi and the Laws of Eshnunna cited above, one may wonder whether the meaning ‘wrestling, contest’ for risbatu and šigištu could fit the context better than ‘brawl, affray.’ 59 Cf. also the interpretation of I 216–223 in Tigay 1982:186 and Klein 2002:190.

R. Nurullin, Philological Notes on the First Tablet …

205

The epithet dan emūqīšu ‘mighty in strength’ is found in the Epic as a part of the stock couplet ina māti dan emūqīšu // kīma ki´ri ša Ani dunnunā emūqāšu ‘In the land he is mighty in strength, // his strength is as mighty as a rock from the sky’ (v. George 2003:793). The spelling e-mu-qi2-i-šu in tablets from Kuyunjik (I 223, MS P; I 269, 292 MS B1) is usually taken to comprise emūqu ‘strength’ and išû ‘to have’: emūqī īšu.60 This interpretation is corroborated mainly by the spelling e-mu-qa i-[šu(2)] on a Babylonian tablet (I 292, MS o). It may also be indirectly confirmed by the fact that emūqu is often attested in combination with išû.61 A piece of conflicting evidence comes from another Babylonian tablet where we read e-mu-qi2-šu2 (I 269, 292, MS h). According to George (2003:803), this spelling “may be a witness to a different tradition, rather than an example of crasis or a mistake.” Read in this way, emūqīšu is to be combined with the stative form dan. The phrase dan emūqīšu seems to represent the tamyīz-construction recently discussed at length by N. Wasserman (2003:29ff.). Wasserman’s definition of tamyīz is as follows: “the specific bicomponental construction comprising a noun in the accusative case accompanied by a predicative stative The noun in the accusative denotes a POSSESSUM to which a specific, inherent and inseparable quality is attributed by the stative form, i. e. QUALIFIER although T. is formally a bipartite construction, logically it involves three constituents in the following collocation: ‘As for X [POSSESSOR], his Y (POSSESSUM)—which relates to him inalienably—is qualified as Z (QUALIFIER)’ ” (ibid. 29f.). Thus, dan emūqīšu can be rendered as ‘he (Enkidu) is mighty in strength.’ The spelling emu-qi2(-)i-šu in the Kuyunjik manuscripts B1 and P is therefore ambiguous. I prefer to read emūqīšu, since this reading has a striking parallel in Gilg. MB Nippur1 3, where a singular form of emūqu is used: lu da-an e-mu-qa ‘let him (Gilgameš’s rival, i. e. Enkidu) be mighty in strength.’62

60

Cf., for instance ‘Der Stärkste im Land ist er, Kräft hat er’ (Schott–von Soden 1988:19); ‘Mightiest in the land, he possesses strength’ (George 2003:545). 61 Cf., for instance dan-na e-mu-qa e-li-ka i-ši ‘he is stronger than you’ (SB Gilg. I 238); e-ru-u2 ki-ma ne2-ši-im na-e-ri e-mu-qa2-am i-šu ‘the eagle gained strength like (that of) a raging lion’ (Haul 2000:113f. vi 3′–4′). 62 e-mu-qa i-[šu(2)] in MS o (I 292) is thus to be considered a secondary development, apparently under the influence of passages like Gilg. SB I 238 cited above.

206

Articles: Ancient Near Eastern Studies

References Al-Rawi 1982 Al-Rawi 1985 Beaulieu 2002 Black 1998 Borger 1969

Buccellati 1976 Cooper 1977

Dercksen 2005

Driver–Miles 1952 Ebeling 1932–1933 Farber 1991 Foster 2001 Frayne 1993 Gardner–Maier 1984 George 1991a George 1991b George 2003 George 2007 Goetze 1956 Grayson 1987 Groneberg 1981 Haul 2000 Haupt 1884–1891

Al-Rawi, F. N. H. Assault and Battery. Sumer 38:117–120. Al-Rawi, F. N. H. Nabopolassar’s Restoration Work on Imgur-Enlil at Babylon. Iraq 47:1–13. Beaulieu, P.-A. Ayakkum in the Basetki Inscription of Narām-Sîn. NABU 2002/36. Black, J. Reading Sumerian Poetry. Ithaca. Borger, R. Die erste Teiltafel zi-pa3-Beschwörungen (ASKT 11). Dietrich, M.; Röllig, W. (eds.). Lišān mit¶urti. Festschrift Wolfram Freiherr von Soden. Kevelaer– Neukirchen-Vluyn. Pp. 1–22. Buccellati, G. On the Akkadian “Attributive” Genitive. Afroasiatic Linguistics 3:19–27. Cooper, J. S. Gilgamesh Dreams of Enkidu: The Evolution and Dilution of Narrative. Ellis, M. de J. (ed.). Essays on the Ancient Near East in Memory of Jacob Joel Finkelstein. Hamden. Pp. 39–44. Dercksen, J. G. Adad is King! The Sargon Text from Kültepe (with an appendix on MARV 4, 138 and 140). JEOL 39:107–129. Driver, G. R.; Miles, J. C. The Babylonian Laws. Vol. 1. Legal Commentary. Oxford. Ebeling, E. Review of Thompson 1930. AfO 8:226–232. Farber, W. The City Wall of Babylon: A Belt-cord? NABU 1991/72. Foster, B. R. The Epic of Gilgamesh. New York. Frayne, D. R. Sargonic and Gutian Periods (2334–2113 B. C.). Toronto. Gardner, J.; Maier, J. Gilgamesh, Translated from the Sînleqi-unninnī Version. New York. George, A. R. Seven Words. NABU 1991/19. George, A. R. The City Wall of Babylon: A Belt Cord. NABU 1991/101. George, A. R. The Babylonian Gilgamesh Epic. Introduction, Critical Edition and Cuneiform Texts. Oxford. George, A. R. The Civilizing of Ea-Enkidu. An Unusual Tablet of the Babylonian Gilgameš Epic. RA 101:59–80. Goetze, A. The Laws of Eshnunna. New Haven. Grayson, A. K. Assyrian Rulers of the Third and Second Millennia B. C. (to 1115 B. C.). Toronto. Groneberg, B. R. M. Philologische Bearbeitung des Agušayahymns. RA 75:107–134. Haul, M. Das Etana-Epos. Ein Mythos von der Himmelfahrt des Königs von Kiš. Göttingen. Haupt, P. Das babylonische Nimrodepos. 2 vols. Leipzig.

R. Nurullin, Philological Notes on the First Tablet … Haupt 1901 Heidel 1949 Hrůša 2010 Jensen 1900 Klein 2002

Kouwenberg 1997 Kovacs 1989 Lambert 1989

Maul 2008 Mayer 1976 Metzler 2002 Moran 1984 Moran 1992 Parpola 1997 Roaf 2000 Röllig 2009 Roth 1990 Schaudig 2001 Schott 1934 Schott–von Soden 1988 Streck 1999 Streck 2000 Thompson 1930 Tigay 1982 Tournay–Shaffer 1994 Ungnad 1921 Vanstiphout 1991

207

Haupt, P. The Beginning of the Babylonian Nimrod Epic. JAOS 22:7–12. Heidel, A. The Gilgamesh Epic and Old Testament Parallels. Chicago. Hrůša, I. Die akkadische Synonymenliste malku = šarru (AOAT 50). Münster. Jensen, P. Assyrisch-babylonische Mythen und Epen. Berlin. Klein, J. A New Look at the “Oppression of Uruk” Episode in the Gilgameš Epic. Abusch, T. (ed.). Riches Hidden in Secret Places. Ancient Near Eastern Studies in Memory of Thorkild Jacobsen. Winona Lake. Pp. 187–201. Kouwenberg, N. J. C. Gemination in the Akkadian Verb. Assen. Kovacs, M. G. The Epic of Gilgamesh. Stanford. Lambert, W. G. A Babylonian Prayer to Anūna. George, A. R.; Finkel, I. L. (eds.). DUMU-E2-DUB-BA-A. Studies in Honor of Åke W. Sjöberg (OPSNKF 11). Philadelphia. Pp. 321–336. Maul, S. M. Das Gilgamesch-Epos. Neu übersetzt und kommentiert. Vierte, durchgesehene Auflage. München. Mayer, W. R. Untersuchungen zur Formensprache der babylonischen “Gebetsbeschwörungen.” Rome. Metzler, K. A. Tempora in altbabylonischen literarischen Texten. Münster. Moran, W. L. Additions to the Amarna Lexicon. Or 53: 297–302. Moran, W. L. The Amarna Letters. Baltimore. Parpola, S. The Standard Babylonian Epic of Gilgamesh. Helsinki. Roaf, M. Cultural Atlas of Mesopotamia and the Ancient Near East. Oxford. Röllig, W. Das Gilgamesch-Epos. Stuttgart. Roth, M. T. On LE §§ 46–47A. NABU 1990/92. Schaudig, H. Die Inschriften Nabonids von Babylon und Kyros’ des Großen. Münster. Schott, A. Zu meiner Übersetzung des Gilgameš-Epos. ZA 42:92–143. Schott, A.; von Soden, W. Das Gilgamesch-Epos. Stuttgart. Streck, M. P. Die Bildersprache der akkadischen Epik. Münster. Streck, M. P. Das amurritische Onomastikon der altbabylonischen Zeit. Münster. Thompson, R. C. The Epic of Gilgamish. Oxford. Tigay, J. H. The Evolution of the Gilgamesh Epic. Philadelphia. Tournay, R. J.; Shaffer, A. L’épopée de Gilgamesh. Paris. Ungnad, A. Die Religion der Babylonier und Assyrer. Jena. Vanstiphout, H. L. J. A Further Note on Ebiḫ. NABU 1991/103.

208 von Soden 1959 von Soden 1971 Wasserman 2003 Westbrook 1992 Wilcke 1969 Wiseman 1975 Yaron 1988 Zgoll 2010

Articles: Ancient Near Eastern Studies von Soden, W. Beiträge zum Verständnis des babylonischen Gilgameš-Epos. ZA 53:209–235. von Soden, W. Der grosse Hymnus an Nabû. ZA 61:44–71. Wasserman, N. Style and Form in Old Babylonian Literary Texts. Leiden. Westbrook, R. The Trial Scene in the Iliad. Harvard Studies in Classical Philology 94:53–76. Wilcke, C. Das Lugalbandaepos. Wiesbaden. Wiseman, D. J. A Gilgamesh Epic Fragment from Nimrud. Iraq 37:157–163. Yaron, R. The Laws of Eshnunna. 2nd rev. ed. Jerusalem– Leiden. Zgoll, A. Monumentum Aere Perennius—Mauerring und Ringkomposition im Gilgameš-Epos. Shehata, D.; Weiershäuser, F.; Zand, K. (eds.). Von Göttern und Menschen. Beiträge zu Literatur und Geschichte des Alten Orients. Festschrift für Brigitte Groneberg. Leiden–Boston. Pp. 443–470.

The Name of Gilgameš in the Light of Line 47 of the First Tablet of the Standard Babylonian Gilgameš Epic∗ Rim Nurullin Russian State University for the Humanities, Moscow

Line 47 of the first tablet of the Standard Babylonian Gilgameš Epic reads as follows: dGIŠ-gim2-maš1 ul-tu u4-um i’-al-du na-bu šum-šu2. Ever since this line was recovered on a tablet from Nimrud (Wiseman 1975; MS g in George’s edition), it was generally treated in three alternative ways. (1) Some scholars followed C. Wilcke’s proposal to read šu!-pu instead of na-bu: ‘Gilgamesch—seit dem Tag, an dem er geboren wurde, ist sein Name berühmt’ (Wilcke 1977:205).2 Given the fact that the combination of šumu ‘name’ and nabû ‘to name, to summon’ is more than natural (and, indeed, fairly common),3 this emendation is scarcely appealing. (2) J. Tigay explains this line in the following way: “The section describing Gilgamesh’s creation begins with his name being called—that is, destined by the gods for kingship—on the day of his birth. This theme is connected with one frequently found in royal inscriptions, the gods’ calling the king’s name (for kingship), often before his birth, while he is still in the womb or in his childhood” (Tigay 1982:153). Within Tigay’s interpretation, the present line may be rendered as ‘Gilgameš—from the day

∗ I am grateful to L. Kogan for discussing with me many of the issues raised in this paper. My thanks go to B. Alexandrov for providing me with some research materials unavailable in Moscow. I want to thank РГНФ/RFH for its financial support (project No. 09-04-00234а). 1 I follow George (2003:84f.) in reading dGIŠ.GIN2.MAŠ as dGIŠ-gim2-maš (probably to be rendered as dgilgigim2-maš). 2 With the following comment: “Wiseman, l. c., liest na-bu; dies scheint auch der Text zu schreiben. Die Korrektur zu šu!-pu erfordert einen weiteren Keil” (ibid. 201). Cf. also ‘Gilgamesch, seit dem Tage, an dem er geboren wurde, ist sein Name herrlich’ (Schott–von Soden 1988:17); ‘Gilgamesh (was) named from birth for fame’ (Dalley 1989:51). Note that the emendation of na-bu to šu!-pu was also suggested (but subsequently rejected) by Wiseman (1975:163). 3 na-bu in the text must be an error for nabi (v. George 2003:784).

210

Articles: Ancient Near Eastern Studies

he was born his name was summoned.’4 Indeed, in the context of the two preceding lines (45–46), where Gilgameš is described as an unrivalled king, Tigay’s interpretation is by no means implausible: man-nu it-ti-šu išša2-an-na-nu a-na šarrū(LUGAL)-ti // u3 ki-i dGIŠ-gim2-maš i-qab-bu-u2 a-na-ku-ma šarru(LUGAL) ‘Who is that who5 can be compared with him in kingship, // and can say like Gilgameš, “I am king indeed”?’. Still, it is worth pointing out that parallels from royal inscriptions referred to by Tigay6 are not quite exact. As a rule, these passages mention a particular god (or gods) who summoned the king; it is also common to find the purpose of this act being specified. Cf. for instance ša2 Aššur(AN.ŠAR2) ab(AD) ilāni(DINGIR.DINGIR) ana šarrū(LUGAL)-ut māt(KUR) Aš-šurki šakkanakkū(GIR3.NITA2)-ut māt(KUR) Šumeri(EME.GI7) u Akkadî(URI)ki im-bu-u ni-bit šumī(MU)-ia ‘(I am Esarhaddon …) whose name Aššur, the father of the gods, called to kingship over Assyria and rulership over Sumer and Akkad’ (Borger 1956:119, § 101, ll. 10ff.); ša a-na za-na-nam ma-ḫa-zi u3 ud-du-šu e-eš-re-e-tim dMarduk (AMAR.UTU) be-li2 ra-bi2-u3 ib-bi2-u3 ni5-bi-it-su ‘(Nebuchadnezzar) whose name Marduk, the great lord, called to provide for the sanctuaries and to renew the temples’ (Winckler 1887:169, ll. 5f.; VAS 1, 39 i 5f.).7 Within Tigay’s interpretation the line under scrutiny turns out to be somewhat too laconic.8 4

Casus pendens. Cf. Tigay’s own translation: ‘[Gilg]amesh was summoned by name from the very day of his birth’ (ibid. 142). Cf. also Tournay–Shaffer 1994:44 (‘Gilgamesh, dont le nom fut prononcé dès le jour de sa naissance’), with fn. t on p. 45 (“Son nom fut prononcé par les dieux qui fixèrent ainsi sa destine. Le theme est bien attesté pour les rois mésopotamiens …”). 5 For the restoration of the relative pronoun ša in l. 45 v. George 2003:784. 6 The passages are collected in Seux 1967:176–179, 370–371. 7 The superficially similar passage from Tiglathpileser’s annals has probably nothing to do with the idea of the king being summoned by the gods to kingship. Rather, it describes the fame of the king: rēyû(SIPA)iu ke-e-nu ša si-qir-šu eli(UGU) ma-li-kimeš ne2-bu-u2 ‘Faithful shepherd whose name is proclaimed over (the names of) the (other) rulers’ (Grayson 1991:13, ll. 34f.); ‘Tiglathpileser whose name is more famous (?) than (the names of ) the (other) rulers’ (CAD N1 35b). Cf. also the use of nabû in the Š-stem in the following passage from Assurbanipal’s hymn to Ištar: u2-šab-bi-u zikir šu[mīya] eli kalīšunu malikī ‘(The great gods) had my name proclaimed over (the names of) all the (other) rulers’ (von Soden 1974–1977:46, l. 19). In view of the obvious similarity between these two passages, I find unlikely von Soden’s correction of u2-šab-bi to u2-šaq!-qu!-u as well as the “irregular spelling for šūpû” tentatively supposed in CAD N1 38b. 8 As recognized by Tigay himself, cf. his note to the transliterated text: “Perhaps restore the beginning of the line, instead of [dGIŠ.GI]N2.MAŠ, either [ana šarrū(LUGAL)]-ti!, ‘for kingship,’ or [ana da-ra-a]-ti! ‘forever’ ” (Tigay 1982:264).

R. Nurullin, The Name of Gilgameš …

211

(3) In a few recent translations, this line has been taken to mean that the hero of the Epic was named Gilgameš since the very day he was born: ‘From the day he was born his name was called “Gilgameš”.’9 At first sight, this grammatically blameless interpretation may appear rather pointless in the context of the hymn to Gilgameš (ll. 29–62): there is nothing peculiar about the fact that Gilgameš received his name the day he was born.10 However, this reading becomes more attractive if we assume that the clue to the interpretation of the line is furnished by the meaning of the name of Gilgameš—that is, the name was probably taken here to designate some remarkable quality of the hero. As is well known, the question of what the name of Gilgameš means is a very complex one.11 In fact, two different problems are to be solved: (I) the reconstruction of the original form of the name and its meaning, and (II) the exegetic interpretation of the later forms of the name of Gilgameš as supposedly arrived at by learned scribes.12 It is the second question that I consider of vital importance for a correct understanding of the line under scrutiny. The first question has no immediate bearing on it,13 but a brief summary of recent views on this issue can nevertheless be useful in the present context.

Needless to say, these restorations cannot be maintained anymore in view of the explicit evidence from a Babylonian manuscript (MS h in George’s edition). 9 ‘Gilgameš was his name from the day he was born’ (George 2003:541); ‘ “Gilgamesch” ist er seit dem Tage, da er geboren, mit Namen genannt’ (Maul 2008:47); ‘Den Namen Gilgamesch trägt er seit dem Tag seiner Geburt’ (Röllig 2009:36). 10 As recently argued in Radner 2005:27f., it was normal for every Mesopotamian child to receive his name soon after birth (our line is cited by Radner among other supporting examples), even if not necessarily on the very day he was born (cf. Stol 2000:178). 11 My inquiry into the meaning of Gilgameš’s name is largely based on George’s brilliant survey of the problem (George 2003:71–90). 12 On hermeneutical practices in ancient Mesopotamia v. Lieberman 1987; Cavigneaux 1987; Lambert 1999; Maul 1999; Selz 2002; Frahm 2004. 13 Admittedly, the meaning of the Sumerian form of the name, as it was interpreted by Falkenstein (‘der “Alte” ist ein junger Mann,’ v. Falkenstein 1957–1971: 357a), was supposed by some scholars to be alluded to in the eleventh tablet of the Gilgameš Epic (XI 299), v., among others, Böhl 1957–1971:370; Tournay– Shaffer 1994:9; Foster 2001:xii; Röllig 2009:15f. and, critically, George 2003:75, fn. 29; ibid. 525, fn. 288.

212

Articles: Ancient Near Eastern Studies

I. The original form of the name of Gilgameš and its meaning The original Sumerian form of Gilgameš’s name is commonly thought to be pa-bilga-mes.14 Within such a reading, it consists of pa-bilga, a term denoting an elder paternal uncle or grandfather (v. Edzard 1960:254f.; Sjöberg 1967: 212ff.) and mes ‘(young) man’ (often translated as ‘hero’). The name itself is usually understood as a nominal sentence: ‘the forebear (was) a hero.’15 For several reasons, such a rendering is open to criticism. First, there is no reliable evidence for the meaning ‘hero’ of Sum. mes. It was recently argued by G. Marchesi (2004:192f.) that mes “is related to the idea of manliness and strength,” but I do not find Marchesi’s arguments sufficiently convincing. Thus, the equation mes = zi-ka3-ru (MSL 14, 97, 187:5) can be explained by the confusion between ×eš(3) ‘male, man’ (= zikaru) and mes ‘(young) man’ (= eṭlu), cf. the interchangeable use of the signs MES and GIŠ in Proto-Lu and Lu2 = ša: sul, mes, ×uruš (MSL 12, 44:311–313);16 sul, ge-eš×eš, gu-ru-uš×uruš = e¢-lum (MSL 12, 126: 77–79). Ninsianna’s epithet mes-sa× (translated by Marchesi as ‘leading mes’) in an inscription of Rīm-Sîn (Frayne 1990:297, l. 8) may as well be read as dub-sa× ‘former; leader’ (cf. Alster 1997:111, 3.188). Finally, in the passages cited by Marchesi (Wilcke 1969:102, l. 110; Sjöberg– Bergmann 1969:32, l. 266; Klein 1991:302, l. 19) the rendering of mes as ‘vigorous male’ or ‘strong one’ is hardly compelling—in each case the neutral meaning ‘man’ would easily fit as well. Now, the rendering of Gilgameš’s name as ‘the ancestor (was) a (young) man’ is, to my taste, rather pointless (cf., however, Glassner 2003:211). Moreover, as pointed out by Krebernik (2002:15), in the early Sumerian onomasticon mes alone is not used as a predicative element. Thus, for example, it cannot be used in the manner of ur-sa× ‘hero’ in the name a-ur-sa× ‘the father is a hero’ (UET 2, 143 i 6). It is, therefore, quite likely that pa-bilga-mes is an abbreviated form of the name (Krebernik 2002:15). The only other name which seems comparable to it comes from Archaic Ur: pa4-bilga(GIŠx)bil-mes-utu-pa3-da17 ‘the ancestor 14

For the variant spellings of pa-bilga v. George 2003:72ff. George 2003:74, cf. also Diakonoff–Jankowska 1990:104; Tournay–Shaffer 1994:9; Sallaberger 2008:42; Röllig 2009:15. 16 In MSL the sign in line 312 is read as meš, but in Chiera’s copies there are clear mes signs. V. Chiera 1929, No. 102 i 8 (followed by ×eš in l. 9); No. 238rev. ii 7; No. 240rev. i 40. 17 The word pa-bilga is transliterated according to the system developed by George (2003:72ff.), for alternative transliterations v. Krebernik 2002:14, fn. 39; Marchesi 2004:196. Note that according to Krebernik (2002:15) the reading of this name is far from certain. 15

R. Nurullin, The Name of Gilgameš …

213

(was) the man which was chosen by Utu.’18 It is tempting to suppose that Gilgameš bore the same or similar name (cf. Krebernik 2002:15; for a different interpretation v. Marchesi 2004:195ff.). II. The exegetic interpretation of Gilgameš’s name The Ancients’ first attempts to make sense of Gilgameš’s name might have been triggered by the loss of the element /pa/. According to George (2003:74ff.), this loss occurred after the Fara period. This thesis is supported by the following arguments. After the Fara period the name became spelled without PAP, its first element being written either with GIŠ.BIL2 or, less frequently, with GIŠ.BIL. In the late third millennium, these sign-groups were sometimes used as syllabograms for /bil/ and /pil/ (v. Gelb 1961:84f.; von Soden–Röllig 1991:30). The are also examples of the first syllable of Gilgameš’s name written with BIL2 or BIL(NE).19 However, the Ur III spellings dpa-GIŠ.BIL2-ga-mes and dpa-GIŠ.BIL-ga-mes (cited by Marchesi 2004:196) clearly suggest that the old form of Gilgameš’s name was also retained and apparently coexisted with the new one (bilga-mes) at least during the Ur III period. As soon as the abbreviated form of Gilgameš’s name had become prevalent, a need might have been felt to give a meaning to it. It seems possible that the ancient treatment of dbil3/4-ga-mes was similar to some of its modern interpretations. It was suggested by Klein (1993:94, fn. 5) that dbil3/4-ga-mes (according to Klein, apparently the original form of Gilgameš’s name) is to be parsed as bilga ‘offshoot, offspring’ + mes, a designation of a kind of tree (v. Powell 1987:149; 2003:17b). The whole could thus be rendered as ‘the offshoot (of) the mes-tree.’20 Klein’s interpretation is based primarily on a passage from the Marriage of Martu, where the city of Ninab seems to be de-

18 This name is attested once in an administrative text (UET 2, 281obv. ii 5) that records the delivery of cakes: on the reverse it is stated that 127 cakes (gug2) were delivered by engiz(EN:ME:GI, on this office v. Zgoll 2006:417, with further literature). According to Marchesi (2004:196), pa4-bilga(GIŠ)bil-utu-pa3 in Deimel 1923, No. 29rev. i 19 (list of personal names) may be “an abbreviated or corrupted form of the same name.” 19 Note especially the spelling dbil2-ga-mes on an Ur III mace-head, v. George 2003:77, No. 4c. 20 Cf. Parpola 1998:325, fn. 38. Note also that according to Marchesi (2004: 197, fn. 253) the older form pabilga-mes may mean ‘the progenitor (?) is (as fruitful as) a mes-tree.’

214

Articles: Ancient Near Eastern Studies

scribed as bil2-ga ×išmes-me-en ‘you are an offshoot21 of the mes-tree’ (cited from Klein 1997:110, l. 7). In Klein’s opinion, this phrase “is ambiguous and can also be translated ‘you are Gilgameš,’ it may be a word play on the part of the author of the myth” (Klein 1993:94). However, as rightly observed by George (2003:75, fn. 30), it is difficult to see “what such an allusion to would add to the story.” The resemblance of this phrase to Gilgameš’s name may well be accidental. One must acknowledge, nevertheless, that in Sumerian hymns the mes-tree is indeed a frequent metaphor for a king, v. Klein 1993:94, fn. 5; Parpola 1998:325, fn. 39.22 All in all, Klein’s suggestion that ‘the offshoot (of ) the mes-tree’ is the original meaning of Gilgameš’s name can hardly be accepted. There is little doubt that the original form of this name had pa-bilga, a term of kinship, and not bilga ‘offshoot’ as its first element. Besides, “nowhere is the name identifiably a genitive compound” (George 2003:75). Nonetheless, the shortened form of the name (bilga-mes) may have been interpreted as a combination of bilga ‘offshoot’ and mes-tree, but for the time being such a hypothesis cannot be taken as proven. Attempts were made to see the meaning of Gilgameš’s name as encoded in the standard first-millennium spelling dGIŠ-gim2(GIN2/TUN3)-maš. According to F. M. Th. Böhl, the name of Gilgameš may allude to the axe seen by Gilgameš in his second dream about the coming of Enkidu. He reads dGIŠ.GIN2.MAŠ as giš-tun3-maš and renders it as “ ‘Zwillingsaxt’ oder ‘Doppelaxt’ (Labrys): ein Bild unverbrüchlicher Freundschaft” (Böhl 1957–1971:370).23 C. Saporetti sees in this spelling the Sumerian expression tun3 bar ‘to cleave (wood) with an axe.’ Accordingly, the name of Gilgameš, being spelled as dGIŠ.GIN2.MAŠ (= dGIŠ.TUN3.BAR), would mean ‘colui che taglia (con l’ascia) il legno/gli alberi’ (Saporetti 1984:404). In Saporetti’s view, the allusion is to the episode of the chopping down of the cedar trees in Úuwawa’s forest as well as to the felling of the ¶alub-tree described in the Sumerian poem “Gilgameš, Enkidu and the Nether World.” A much more sophisticated interpretation was put forward by S. Parpola. In his opinion, the spelling dGIŠ.GIN2.MAŠ conceals the follow21

Note that bil2-ga in this passage is often translated as ‘ancestor’ (v., for instance, PSD B 153a). 22 In addition to the passages adduced by the authors one may add Klein 1981:34, l. 15, with commentary on p. 24, fn. 122; Civil 1996:163 ii 3′; Römer 1965:51, l. 244. 23 Cf. also Beckman 2003:50: “Perhaps the choice of GIN2 reflects a clever play on its alternate value TUN3 = pāšum, in allusion to the axe of which Gilgamesh dreams.”

R. Nurullin, The Name of Gilgameš …

215

ing esoteric meaning: (Gilgamesh) ‘(who) matched the tree of balance’ ([ša] i´´i šitqulti umaššil[u]) (Parpola 1998:324).24 The name of the hero thus portrays him “as an embodiment of the sacred tree.”25 As one can see, the interpretations of Gilgameš’s name suggested so far are notoriously divergent from each other, being, it seems, primarily grounded on their authors’ views of what the name should mean. This is not to say that an exegetical approach to the name of Gilgameš is unwelcome (cf. George 2003:87). However, any treatment of this kind is deemed to remain highly uncertain unless supported by ancient sources,26 the more so since the spelling dGIŠ.GIN2.MAŠ does not seem to have been intentionally coined by the Ancients to load the name of Gilgameš with a hidden meaning.27 George’s conclusion that “the writing GIŠ-gim2-maš has its origin in old traditions of spelling” (ibid. 87f.) is very well founded. The primary goal of the present study is to find out how the name of the hero was understood by the author(s) of the “old” prologue to the Epic. I propose the following steps to solve the problem. The first is to seek for ancient evidence of the name of Gilgameš being interpreted in a hermeneutical way. Thanks to the passage under discussion, we are also able to make the next step, namely to check whether the exegetic meaning obtained on the basis of this evidence fits the context of this particular 24

The correspondences used by Parpola are as follows: GIŠ = i´u ‘tree’; GIN2 = šiqlu ‘shekel’ (according to Parpola, this logogram might also be read as šitqultu ‘balance’); MAŠ = mašālu ‘to be like.’ 25 Parpola’s interpretation is based on his perception of the Epic of Gilgameš as “structured after the sacred tree, its individual tablets corresponding to the divine powers making up the tree and marking progressive stages in the hero’s quest for immortality and spiritual perfection” (ibid. 325; for a detailed discussion on this topic v. Parpola 1993:192–195). 26 A parade example of such a corroboration is the name of the demon Šulak as explained in a commentary of a medical text: šumma amēlu lišānšu eb¢et-ma ana bīt musâti lā irrub Šulak ima¶¶assu Šulak ša iqbû : ŠU : qātu : LA : lā : KU3 : ellu : ana bīt musâtu irrub qātāšu lā ellā ana mu¶¶i qabi ‘If a man’s tongue is swollen (this man) should not enter the lavatory, (otherwise) Šulak will strike him. (The name) Šulak, as they said, (should be interpreted as) ŠU “hand,” LA “not,” KU3 “clean”; he enters the lavatory and his hands are not clean, it is because of this that he is called so’ (Hunger 1976, No. 47, ll. 1ff.; see further Cavigneaux 1987:250; Parpola 1998:319; Glassner 2003:155; Selz 2002:661). 27 So explicitly in Parpola 1998:318: “The fact that the spellings dGIŠ.GIM.MAŠ d and GIŠ.GIN2.MAŠ are close to the actual pronunciation of the name but slightly deviate from it suggests that they were indeed introduced to make a specific logographic reading possible.”

216

Articles: Ancient Near Eastern Studies

line. In my opinion, the fact that the line under scrutiny occurs in the context of the hymn (ll. 29–62) suggests that the name ‘Gilgameš’ was intended to describe some admirable quality of the hero. This quality might have characterized Gilgameš from the very first days of his life (ultu ūm i’’aldu), but it is also possible that the name foretells the heroic features to be acquired by Gilgameš in his more mature years.28 Turning now to the meaning of the name of Gilgameš as understood by ancient scribes, one has to recognize that the most important piece of evidence comes from the so-called “Group Vocabulary 3” (“série šarru” according to Cavigneaux 1980–1983:638a): kal-ga imin = dbil3-ga-mes kal-ga imin = muq-tab-lu kal-ga imin = a-lik pa-na (CT 18, 30 iv 6–8)

In the first line of this lexical segment, it is stated that kal-ga imin is an explanation of the name of Gilgameš. In the next two lines, the meaning of the equation kal-ga imin = dbil3-ga-mes is clarified: it describes the hero as a ‘warrior’ (muqtablu) and as a ‘leader’ (ālik pana). The Sumerian word kal-ga ‘strong’29 seems to refer to the first element of the name of Gilgameš. This element was apparently pronounced in a variety of ways. The spellings that interest us here are those with /k/ in Anlaut as well as those with /a/ in the first and second syllables: ki-il-ga-mes;30 dgal-ga-mi-iš; d gal-ga-mi-šu-ul;31 gal-ga-meš.32 In George’s opinion, the use of the sign GAL in the last three examples “can best be explained by reference to a western orthographic tradition whereby signs CaC can have the value CiC (and vice 28 The latter possibility has been suggested to me by L. Kogan. One cannot rule out that it also predicts some of his future heroic achievements (such as killing Úuwawa), but this is, in my view, considerably less likely. 29 Sumerian KAL-ga is usually transliterated as kalag-ga or kala-ga. This reading is supported by the spellings like kal-la-ga (Alster 1997:92, SP 3.67, MS A) as well as by the syllabic writings ka-la-ka and ka-la-ak-ka (v. CAD D 93a dannu, lex. section), v. also Römer 1965:103 with further literature. However, there is also evidence for the variant pronunciation with the vowel in the second syllable syncopated, cf. gal-ga in a syllabically written Sumerian text from Ebla (Edzard 1984: 43, A iii 1), which corresponds to KAL in the standard orthography (ibid. B ii 1). Cf. also ur-sa× ga-ar-ga in KAR 158 iii 22, which seems to stand for ur-sa× kal-ga (v. Falkenstein 1950:103, fn. 7; van Dijk 1960:115, fn. 3). 30 George 2003:81, No. 20 (Hh XXII, Emar). 31 Ibid. 79, No. 11 (Hurrian version of the Gilgameš Epic). 32 Ibid. 81, No. 18 (Hh XX–XXII, Ugarit).

R. Nurullin, The Name of Gilgameš …

217

versa). The sign gal can thereby be given the value gilx” (George 2003:79). However, to my knowledge, there are no other examples of such a use of GAL. Note that a few pages later George takes a rather different view: “the sign kal in GIŠ-kal-TUK recalls the western spellings gal-ga-mi-iš (No. 11) and gal-ga-meš (No. 18); perhaps it glosses GIŠ” (ibid. 86). Tournay and Shaffer (1994:10) also refer to the spelling kal-ka-meš, said to be attested on a cylinder seal of the Old Akkadian period (with no exact reference). The association of the first element of Gilgameš’s name with kal-ga ‘strong’ may be further corroborated by the following passages: kal-ga dGIŠ.BIL2-ga-mes DI ¶a-ra-a[n…] ‘May the mighty Gilgameš …’ (Sjöberg 1983:315, l. 96); kal-ga dGIŠ.BILga-mes mi2 du11-ga (var. za3-mi2) ‘The mighty Gilgameš has been praised (?) (var. Praise be to the mighty Gilgameš)’ (Edzard 1991:232, l. 201). Since in Sumerian an adjective normally follows the noun modified, one may wonder whether, in this case, kal-ga ‘strong’ placed before dGIŠ.BIL(2)-ga-mes was supposed to allude to the first element of the name. Also remarkable is the fact that in the hymn to Gilgameš which constitutes a part of Šulgi O, the strength of the hero is frequently mentioned. Cf., for example, šul-gi d GIŠ.BIL2-ga-mes nam-kal-ga-na (var. nam-ur-sa×-×a2-na) mu-ni-in-i-i kal-ga me3-a ‘Šulgi praised Gilgameš in his might : “Mighty one in the battle …” ’ (Klein 1976:276ff., ll. 49–53). It is, therefore, possible that in the passages just mentioned the first element of Gilgameš’s name (spelled with the sign-group GIŠ.BIL(2)-ga) was read as /kalga/ (or /galga/). Such a reading, though nowhere explicitly attested, may be confirmed by the occurrence of the name of Gilgameš in the Song of the Hoe. As is well known, in this composition the syllable /al/ plays an important role: the verbal prefix al- as well as a number of words and names which contain the syllables /al/ or /Cal/ (and also /ar/ or /Car/) are frequently used, obviously to make a pun on the Sumerian word for ‘hoe’ (×išal), v. Wilcke 1972:36f. It is tempting to suggest that Gilgameš, who is involved in lines 73–78 of the Song (v. Civil 1969:70, fn. 1), is referred to because his name was pronounced (at least by the author of the Song) with /al/ (i. e., galgames or the like). Thus, it seems likely that d GIŠ.BIL(2)-ga-mes here (and, probably, elsewhere) is a historical writing of the name which has little bearing on its actual pronunciation. The role of imin ‘seven’ is less clear. George interprets it as “a symbolic number” and translates kal-ga imin as ‘strongest of all’ (George 2003: 86),33 noting that “whatever the interpretation of imin, it is Gilgameš’s physical prowess that is at issue” (ibid. 87). 33

This translation is based firstly on the equation imin = kiššatu which is found in the same list (for details see George 2003:87). Note that in this list the

218

Articles: Ancient Near Eastern Studies

That Gilgameš’s name may allude to the strength of the hero can also be deduced from the spelling GIŠ-KAL-TUK,34 apparently interpreted by the Ancients as ‘one who has strength’ (KAL = danānu; TUKU = išû).35 This spelling seems to be derived from the more common GIŠ-TUK(-maš).36 In George’s opinion, the latter writing “suggests deliberate manipulation of the spelling,” since “spellings that include the signs GIŠ and TUK might have been favoured (if not coined) as a scholarly exercise in revealing a characteristic of the hero Gilgameš for which he became celebrated in the Babylonian epic tradition: he was a man of extraordinary wisdom” (George 2003:86). The reason for this conclusion is that “the signs giš-tuk evoke the Sumerian word geštug, written gišPItug2 etc., meaning ‘ear; understanding, wisdom’ (uznu, ¶asīsu) and also ‘wise, perceptive’ (¶assu)” (ibid. fn. 81).37 According to George, “the spelling dGIŠ-kal-TUK offers the possibility of an additional exegesis that makes allusion to the hero’s strength as well as his wisdom.” As regards the abbreviated form GIŠ-TUK, it might indeed have been interpreted as an allusion to the wisdom of the hero. But the fuller spelling, GIŠ-TUK-maš, cannot be easily treated in this way, for, within such a treatment, maš would be left unexplained. Returning to the spelling GIŠ-KAL-TUK, it seems doubtful whether it can allude to both the strength and the wisdom of the hero. More likely, what is at issue here is the strength alone. If the assumption that KAL glosses GIŠ (kalga(GIŠ)kal-TUKU, v. fn. 35) is accepted, then the writing GIŠTUK-maš may be read as kalga(GIŠ)-TUKU-maš ‘the … (maš) which has number ‘seven’ is also involved in the commentary on the name of Enkidu: a-ra2 imin = en-gi-du3 (CT 18, 30 iv 10). In view of the correspondence a-ra2 = māliku, George renders a-ra2 imin as “counsellor par excellence” (ibid. 143). 34 George 2003:86, No. 21 (An = Anum, MA copy; Hh IV). 35 Cf. George 2003:86, fn. 82. The sign KAL should probably be regarded as a gloss (George 2003:86). Note that it seems to indicate both the pronunciation (kalga(GIŠ)kal) and the meaning (KAL = dannu, danānu) of the name. 36 George 2003:81, No. 15, 17 (Hh XXII, Emar; conventional Sumerian column of the Emar version of the Poem of Early Rulers; early NA sources of the Epic of Gilgameš). The origin of this spelling is probably due to misreading. As observed by George (2003:84), “the signs TUK and KIN (qi) can easily be confused at Emar, Ugarit, Hattusa and elsewhere in the West.” Thus, GIŠ-TUK-maš is probably “descended from a Middle Assyrian spelling *GIŠ-qi-maš or an Old Babylonian spelling *GIŠ-kin-maš. Both would be simple variants of the attested GIŠgim-maš and GIŠ-gim2-maš.” 37 George also draws attention to the Sumerian compound verb ×iš ~ tuku ‘to hear.’

R. Nurullin, The Name of Gilgameš …

219

strength.’ There are, therefore, two features which distinguish GIŠ-TUKmaš from GIŠ-KAL-TUK: first, the gloss is lacking; second, the last element of the name (maš) is preserved. The crucial problem is how to interpret this element. The clue to this question seems to be found in the passage from the Group Vocabulary cited above. In my view, the equation kal-ga imin = a-lik pa-na suggests that the last element of Gilgameš’s name might have been associated with the Sumerian word mas-su(2), borrowed into Akkadian as massû (with phonetic variants maššû, mansû, mensû, massu’u). In both languages, this word appears to have two principal meanings: (1) ‘expert’; (2) ‘leader.’38 The latter meaning seems to agree well with the explanation of Gilgameš’s name as ālik pana in the Group Vocabulary 3.39 38 It was proposed by Landsberger (MSL 2, 104, fn. 1) that there are two Akkadian words massû, each having a different Sumerian etymon: massû I < maš2sa× ‘Leitbock’ (this Sumerian word can also mean ‘leader,’ v. Heimpel 1968:18) and massû II < mas-su3 ‘mit fernem (weitem) Blick’ (note that mas-su3 is a late spelling, up to the Old Babylonian period this word was written as either mas-su or mas-su2, thus Landsberger’s rendering of this word seems rather doubtful). Landsberger was subsequently followed by the majority of scholars dealing with this question (v., for instance, Lambert 1960:310; Römer 1965:243f.), including the standard dictionaries of Akkadian. The former term is generally rendered as ‘leader’ (CAD M1 327b), ‘Anführer’ (AHw. 619a), whereas the latter as ‘expert’ (CAD M1 327a), ‘Sachkundiger’ (AHw. 619b). Note, however, that in CDA 200a there is only one entry massû ‘leader, expert.’ Doubts were cast upon the validity of this distinction also by Lieberman (1977:389, fn. 631). Indeed, on the one hand, the phoneme /×/ (maš2-sa×) is not expected to disappear in Sumerian loanwords into Akkadian, cf. uršānu ‘warrior’ (< ur-sa×) or geršānu ‘kind of leek’ (< ga-raš-sa×). On the other hand, mas-su(2) seems to mean both ‘leader’ and ‘expert,’ ‘counselor,’ cf. na-gada mas-su ni×2-zi-×al2-la-kam ‘(Enlil) is a herdsman, the leader of all living creatures’ (Enlil A 94, cited from Reisman 1969:51); dumu an-na mas-su2 inim pa3-de3 ‘(Bau,) daughter of An, a counselor who chooses words (carefully)’ (Römer 1965:236, l. 5). Thus, it seems likely that there is only one word massû ‘leader; expert,’ borrowed from Sumerian mas-su(2) with the same meanings. Note also that in CAD M1 327a massû ‘expert’ is listed as an adjective. This seems to be based first of all on the passage from an inscription of Esarhaddon: d[Sîn(30) u dŠam]aš(UTU) ilāni(DINGIR)meš maš-šu-u2-te (Borger 1956:2 i 31). Borger, however, interprets it differently: ‘Sin und Šamaš, die “Zwillingsgötter” ’ (ibid., cf. also AHw. 631a), perhaps having in mind an otherwise unattested nisba adjective maššû (< māšu, maš(š)û ‘twin’). 39 Note that ālik pana could hardly be derived directly from kal-ga imin. In my opinion, these two explanations of Gilgameš’s name are of different origin. The former is the traditional interpretation of the name which stemmed, long before the Group Vocabulary 3 was composed, from the need to clarify its last element.

220

Articles: Ancient Near Eastern Studies

The line under discussion may therefore be read as follows: ‘From the day he was born his name was called “Gilgameš” (that means “the mighty leader”).’40 Within this interpretation, one must assume that the name of the hero was identified with the Sumerian phrase kal-ga mas-su(2).41 In other words, the name was not explained on the graphemic level (as usually assumed), but was rather subjected to a kind of popular (or, rather, learned) “Sumerianizing” etymology.42 An inherently related question is how the name of Gilgameš was pronounced by the author of the “old” prologue of the Epic (the date of its composition is uncertain, but early OB is possible). If, as proposed here, it was understood as kal-ga mas-su(2), then a form with the vowel /a/ in all syllables is expected (i. e. /galgāmaš/ or the like). Such a form is so far unattested. However, this may be due to the scarcity of the extant evidence. As is well known, the name of the hero was generally pronounced either with /i/ in the second syllable and /a/ in the third one (Gilgimmaš) or with /a/ in the second syllable and /e/ (or /i/) in the third one (Gilgāmeš). To my knowledge, there are only two exceptions. The first is dGIŠ.BIL(2)-gim2mes43 (George 2003:75, No. 2), the Old Sumerian spelling of the name current in the city-state of Lagaš. More interesting for our purpose is the second exceptional spelling, dGIŠ-ga-maš (ibid. 78, No. 9), which comes

As for the latter, much more sophisticated kal-ga imin, it may have been coined ad hoc by the compiler of the list and then combined mechanically with the former variant. 40 One may wonder whether line 31 of the first tablet of the Epic of Gilgameš may be considered as an allusion to the name of the hero: [i]l-lak ina pa-ni a-ša2-rid ‘Going ahead he was the leader.’ 41 Judging from the equation ma-aš2MAŠ = a-ša2-ri-du (MSL 14, 227:93), MAŠ alone might have been identified with the word mas-su(2) (for a different interpretation of this entry see George 2003:82, fn. 65). Note also that in the version of the Poem of Early Rulers from Emar mdGIŠ-TUK-m[aš] (George 2003:81, No. 15) in the regular Sumerian column is rendered as ki-iš-mas-su in the syllabic Sumerian column (ibid. No. 16; according to George (ibid. 85) it is to be read as ki-iš-mas-su). 42 The spellings GIŠ-TUK-maš and GIŠ-KAL-TUK discussed above were probably coined (perhaps by an accident of misreading, v. fn. 36) and interpreted against the background of this kind of thinking of the name of Gilgameš. GIŠ-TUK-maš thus may be rendered as kalga(GIŠ)-TUKU-maš ‘the mighty leader’ (lit. ‘the leader who has strength’). Note, however, that the reading GIŠ-TUKU(= /×eštu.g/)-maš ‘the wise expert’—a development of George’s idea mentioned above—cannot be entirely ruled out. 43 On this reading v. George 2003:75f., with further literature.

R. Nurullin, The Name of Gilgameš …

221

from Elam of the sukkalma¶s.44 The sign GIŠ does not allow us to detect the reading of the first syllable of the name.45 George (2003:82) proposes to render it as bilga(GIŠ)ga-maš. However, if we accept that Gilgameš’s name could also be pronounced with /a/ in the first syllable, the alternative reading galga(GIŠ)ga-maš (or the like) can hardly be excluded.46 References Alster 1997 Beckman 2003

Böhl 1957–1971 Borger 1956 Cavigneaux 1980–1983 Cavigneaux 1987 Chiera 1929 Civil 1969 Civil 1996 Dalley 1989 Deimel 1923 Diakonoff–Jankowska 1990

Alster, B. Proverbs of Ancient Sumer. I–II. Bethesda. Beckman, G. Gilgamesh in Úatti. Beckman, G.; Beal, R.; McMahon, G. (eds.). Hittite Studies in Honor of Harry A. Hoffner Jr. on the Occasion of His 65th Birthday. Winona Lake. Pp. 37–57. Böhl, F. M. Th. de Liagre. Gilgameš. B. Nach akkadischen Texten. RlA 3:364–372. Borger, R. Die Inchriften Asarhaddons, Königs von Assyrien. Graz. Cavigneaux, A. Lexikalische Listen. RlA 6:609–641. Cavigneaux, A. Aux sources du Midrash: l’hermeneutique babylonienne. AuOr 5:243–255. Chiera, E. Sumerian Lexical Texts from the Temple School of Nippur. Chicago. Civil, M. Review of CT 44. JNES 28:70–72. Civil, M. Literary Text about Ur-Namma. AuOr 14:163– 167. Dalley, S. Myths from Mesopotamia. Oxford. Deimel, A. Die Inschriften von Fara. II. Schultexte aus Fara. Leipzig. Diakonoff, I. M.; Jankowska, N. B. An Elamite Gilgameš Text from Argištihenele, Urartu (Armavir-blur, 8th century B. C.). ZA 80:102–123.

44 George (2003:83) notes that this spelling is found in “essentially well-written Babylonian texts,” and hence “there is no reason why this has to be a local coinage.” 45 The writings of the name of Gilgameš with the sign GIŠ may be explained as a survival of old orthography. Originally it was used to render the second part of the word pa-bilga, a term of kinship which played an important role in the early Sumerian onomasticon (v. George 2003:71ff.). As part of the later spellings of the name of Gilgameš, the sign GIŠ (as well as GIŠ.BIL(2), see above) was apparently read in a variety of ways, depending on how the name was pronounced within one or another tradition. 46 Falkenstein’s proposal that a-ba-al-ga-maš, the name of the ruler of Marhaši (Parahšum) conquered by Nāram-Sîn, contains the name of Gilgameš (Falkenstein 1957–1971:357; 1949:9, fn. 1) is intriguing, but difficult to prove.

222

Articles: Ancient Near Eastern Studies

Edzard 1960 Edzard 1984 Edzard 1991 Falkenstein 1949 Falkenstein 1950 Falkenstein 1957–1971 Foster 2001 Frahm 2004 Frayne 1990 Gelb 1961 George 2003 Glassner 2003 Grayson 1991 Heimpel 1968 Hunger 1976 Klein 1976 Klein 1981 Klein 1991

Klein 1993 Klein 1997 Krebernik 2002

Edzard, D. O. Sumerer und Semiten in der frühen Geschichte Mesopotamiens. Genava N.S. 8:241–258. Edzard, D. O. Hymnen, Beschwörungen und Verwandtes aus dem Archiv L. 2769. Rome. Edzard, D. O. Gilgameš und Huwawa A. II. Teil. ZA 81: 165–233. Falkenstein, A. Grammatik der Sprache Gudeas von Lagaš. I. Schrift und Formenlehre. Rome. Falkenstein, A. Sumerische religiöse Texte. ZA 49:80–150. Falkenstein, A. Gilgameš. A. Nach sumerischen Texten. RlA 3:357–364. Foster, B. R. The Epic of Gilgamesh. New York. Frahm, E. Royal Hermeneutics: Observations on the Commentaries from Ashurbanipal’s Libraries at Nineveh. Iraq 66:45–50. Frayne, D. R. Old Babylonian Period (2003–1595 B. C.). Toronto. Gelb, I. J. Old Akkadian Writing and Grammar. Chicago. George, A. R. The Babylonian Gilgamesh Epic. Introduction, Critical Edition and Cuneiform Texts. Oxford. Glassner, J.-J. The Invention of Cuneiform Writing in Sumer (transl. and ed. by Z. Bahrani and M. van de Mieroop). Baltimore–London. Grayson, A. K. Assyrian Rulers of the Early First Millennium B. C. (1114–859 B. C.). Toronto. Heimpel, W. Tierbilder in der sumerischen Literatur. Rome. Hunger, H. Spätbabylonische Texte aus Uruk. I. Berlin. Klein, J. Šulgi and Gilgameš: Two Brother-Peers (Šulgi O). Kramer Anniversary Volume. Cuneiform Studies in Honor of Samuel Noah Kramer. Neukirchen-Vluyn. Pp. 271–292. Klein, J. The Royal Hymns of Shulgi King of Ur: Man’s Quest for Immortal Fame. Philadelphia. Klein, J. The Coronation and Consecration of Šulgi in the Ekur (Šulgi G). Eph’al, I.; Cogan, M. (eds). Ah, Assyria… Studies in Assyrian History and Ancient Near Eastern Historiography Presented to Hayim Tadmor. Jerusalem. Pp. 292–313. Klein, J. Additional Notes to “The Marriage of Martu.” Rainey, A. F. (ed.). kinattūtu ša dārâti. Raphael Kutscher Memorial Volume. Tel Aviv. Pp. 93–106. Klein, J. The God Martu in Sumerian Literature. Finkel, I. L.; Geller, M. J. (eds.). Sumerian Gods and Their Representation. Groningen. Pp. 99–116. Krebernik, M. Zum Struktur und Geschichte des älteren sumerischen Onomastikons. Streck, M. P.; Weninger, S. (eds.). Altorientalische und semitische Onomastik (AOAT 296). Münster.

R. Nurullin, The Name of Gilgameš … Lambert 1960 Lambert 1999

Lieberman 1977 Lieberman 1987 Marchesi 2004 Maul 1999

Maul 2008 Parpola 1993 Parpola 1998 Powell 1987 Powell 2003 Radner 2005 Reisman 1969 Röllig 2009 Römer 1965 Sallaberger 2008 Saporetti 1984 Schott–von Soden 1988 Selz 2002

Seux 1967 Sjöberg 1967

223

Lambert, W. G. Babylonian Wisdom Literature. Oxford. Lambert, W. G. Babylonian Linguistics. Van Lerberghe, K.; Voet, G. (eds.). Languages and Cultures in Contact. At the Crossroads of Civilizations in the Syro-Mesopotamian Realm. Leuven. Pp. 217–231. Lieberman, S. J. The Sumerian Loanwords in Old-Babylonian Akkadian. I. Prolegomena and Evidence. Missoula. Lieberman, S. J. A Mesopotamian Background for the So-Called Aggadic ‘Measures’ of Biblical Hermeneutics? HUCA 58:157–225. Marchesi, G. Who Was Buried in the Royal Tombs of Ur? The Epigraphic and Textual Data. Or 73:153–197. Maul, S. M. Das Wort im Worte. Orthographie und Etymologie als hermeneutisches Verfahren babylonischer Gelehrter. Most, G. W. (ed.). Commentaries / Kommentare. Göttingen. Pp. 1–18. Maul, S. M. Das Gilgamesch-Epos. München. Parpola, S. The Assyrian Tree of Life: Tracing the Origins of Jewish Monotheism and Greek Philosophy. JNES 52:161–208. Parpola, S. The Esoteric Meaning of the Name of Gilgamesh. Prosecký, J. (ed.). Intellectual Life of the Ancient Near East. Prague. Pp. 315–329. Powell, M. A. The Tree Section of ur5(= ÚAR)-ra = ¶ubullu. BSA 3:145–151. Powell, M. A. Obst und Gemüse. A. I. Mesopotamien. RlA 10:13–22. Radner, K. Die Macht des Namens. Altorientalische Strategien zur Selbsterhaltung. Wiesbaden. Reisman, D. Two Neo-Sumerian Royal Hymns. PhD. Diss. University of Pennsylvania. Röllig, W. Das Gilgamesch-Epos. Stuttgart. Römer, W. H. Ph. Sumerische “Königshymnen” der Isin-Zeit. Leiden. Sallaberger, W. Das Gilgamesch-Epos. Mythos, Werk und Tradition. München. Saporetti, C. tun3.bar = ‘tagliare’. Cagni, L. (ed.). Il bilinguismo a Ebla. Napoli. Pp. 403–404. Schott, A.; von Soden, W. Das Gilgamesch-Epos. Stuttgart. Selz, G. J. “Babilismus” und die Gottheit dNindagar. Loretz, O.; Metzler, K. A.; Schaudig, H. (eds.). Ex Mesopotamia et Syria Lux. Festschrift für Manfred Dietrich zu seinem 65. Geburtstag. Münster. Pp. 647–684. Seux, M.-J. Épithètes royales akkadiennes et sumériennes. Paris. Sjöberg, Å. W. Zu einigen Verwandtschaftsbezeichnungen im Sumerischen. Edzard, D. O. (ed.). Heidelberger

224

Articles: Ancient Near Eastern Studies

Sjöberg 1983 Sjöberg–Bergmann 1969 Stol 2000 Tigay 1982 Tournay–Shaffer 1994 van Dijk 1960 von Soden 1974–1977 von Soden–Röllig 1991 Wilcke 1969 Wilcke 1972 Wilcke 1977 Winckler 1887 Wiseman 1975 Zgoll 2006

Studien zum Alten Orient, Adam Falkenstein zum 17. September 1966. Wiesbaden. Pp. 201–231. Sjöberg, Å. W. The First Pushkin Museum Elegy and New Texts. JAOS 103:315–320. Sjöberg, Å. W.; Bergmann, E. The Collection of the Sumerian Temple Hymns. Locust Valley. Stol, M. Birth in Babylonia and the Bible. Its Mediterranean Setting. Groningen. Tigay, J. H. The Evolution of the Gilgamesh Epic. Philadelphia. Tournay, R. J.; Shaffer, A. L’épopée de Gilgamesh. Paris. van Dijk, J. J. A. Sumerische Götterlieder. II. Heidelberg. von Soden, W. Zwei Königsgebete an Ištar aus Assyrien. AfO 25:37–49. von Soden, W.; Röllig, W. Das akkadische Syllabar. Rome. Wilcke, C. Das Lugalbandaepos. Wiesbaden. Wilcke, C. Hacke. B. Philologisch. RlA 4:33–38. Wilcke, C. Die Anfänge der akkadischen Epen. ZA 67: 153–216. Winckler, H. Studien und Beiträge zur babylonisch-assyrischen Geschichte. ZA 2:148–178. Wiseman, D. J. A Gilgamesh Epic Fragment from Nimrud. Iraq 37:157–163. Zgoll, A. Traum und Welterleben im antiken Mesopotamien. Münster.

Another Attempt at Two Kassite Royal Inscriptions: The Agum-Kakrime Inscription and the Inscription of Kurigalzu the Son of Kadashmanharbe* Takayoshi Oshima Universität Jena

The Agum-Kakrime Inscription and the inscription of Kurigalzu, the son of Kadashmanharbe, are supposed to be later copies of Kassite royal inscriptions. Yet the authenticity of these documents and their reliability as historical sources have long been debated by modern scholars. The present article is another attempt to edit these well-known inscriptions based on a collation in the British Museum. Part I. The Agum-Kakrime inscription The Agum-Kakrime Inscription is an Akkadian unilingual text and composed in a first person narrative form. It recounts the return of the statues of Marduk and Zarpanitu from exile and the refurbishment of their cult images and Kasulimma, the inner cella of Esagila, Marduk’s temple in Babylon. This inscription is known only from two late copies from Kouyunjik, Nineveh, K.4149+4203+4348+Sm 27 and Rm 505. The first manuscript, K.4149+, preserves about 340 lines of an eight column tablet, however, a large section from middle portion of the tablet is missing. Presumably we have recovered around 4/5 of what appears to be roughly 450 lines of text when complete. This manuscript was published by Pinches as 5R 33 in 1880.1 Rather interestingly this exemplar bears

* We would like to thank The Trustees of The British Museum for their kind permission to collate the manuscripts and to publish the photographs of the tablets from their collections. While K.4190+4303+4348+Sm 27 were photographed by the author, the photographs of other tablets, namely Rm 505 and BM 108982, were taken by Dr. Uri Gabbay. The present writer is grateful for his kind assistance. This article was written during the author’s stay in Universität Leipzig with a generous grant from the Alexander von Humboldt Fund. They are also thanked. Chris Conlan is also thanked for editing English. 1 A fragment K.4149 was previously published by Norris, in 2R 38, No. 2.

226

Articles: Ancient Near Eastern Studies

two different colophons (col. viii 23–33 and col. viii 34–42)—the first one is mūdû mūdâ likallim2 while the second is that of Assurbanipal’s library. This fact clearly indicates that it is a copy of an older manuscript. The second exemplar of this inscription, Rm 505, is a small fragment and preserves 37 lines on its obverse and reverse (ii 13–28; iii 12–19; vi 27– 31; vii 30–35). It was published by Campbell Thompson among the manuscripts of the Gilgamesh Epic in 1930. 3 Both manuscripts are written in Neo-Assyrian script and one may even speculate that these were prepared around the same period based on the sign forms.4 Each line is very short and a sentence often runs through several lines, which is very common in Old Babylonian inscriptions. The inscription is divided into at least 15 sections by horizontal lines. The length of each section varies5 and it seems that each horizontal line marks the end of a particular subject. Incidentally, a horizontal line was not drawn after each individual line. There are two previous full text-editions and several translations of this Agum-Kakrime Inscription. The first full edition is Jensen 1892:134–153 and the last edition was prepared by P. Stein (2000:150–165). Brinkman also announced that a new edition of this inscription was in preparation in 1976.6 The last translation of this text was prepared by B. Foster in his collection of ancient texts in English translation (Foster 2005:360–364). Further references are found in Sommerfeld 1982:172, n. 4; Stein 2000: 150; Foster 2005:364; and Pruzsinszky 2009:184. The authenticity of this text has been challenged a number of times by modern scholars on the basis that ancient documents only partially corroborate the accounts that are recounted in this inscription. However, except for the script, nothing in this inscription suggests a late date of composition with certainty, i. e. clearly marking it as a fake. Thus, the opinions of modern scholars concerning its authenticity are still divided and no consensus has been reached.7 Below, we discuss several points that raise doubts on the reliability of this text as an historical source. 2

For this colophon, see Lenzi 2008:201–202. Campbell Thompson 1930, pl. 36. 4 Yet, these were definitely written by two different hands. 5 Given this fact, Jensen speculated that the inscription was originally written on a statue, not on a cylinder or a tablet (1892:134, n. 1). See also Jaritz 1958: 228. Stein offers a stele as an original form of inscription based on the narrow columns (2000:150). 6 Brinkman 1976:97. 7 E. g., Landsberger 1928–1929:312; 1954:67–68; Gelb 1949:348, n. 12; Grayson 1969:108; and Borger 1971:17, note for col. i 13–38 of The Marduk Prophe3

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions …

227

Point A. The identity of the narrator of the Agum-Kakrime inscription This inscription, written in a first person narrative form, has always been credited to Agum-Kakrime since its publication by Pinches (1880).8 The first line of K.4149+, where we expect the name of the narrator, is damaged and preserves the last four signs, KA AK RI ME. As the narrator later refers to himself as šarru agum, ‘the king Agum’ (vi 40; vii 11), or šarru damqu agum, ‘the favourable king, Agum’ (vii 29), kakrime was almost certainly either a title9 or a second element of his name. The narrator presents himself as the king of the Kassites and the Akkadians (by which he probably meant the Babylonians)10 and the king of the land of Babylon (i 31–34). He claims his hegemony over the lands of Eshnunna, Padan, Alman, and the barbaric Gutians (i 35–39). As Eshnunna was not occupied after the destruction by Hammurabi in his regnal year 31,11 it seems that Agum actually claimed his sovereignty over the Diyala region, north-east of the Tigris, rather than the city itself.12 However, we have no way of verifying his claim because we have no further textual evidence for the political geography of the early Kassite period.

cy (Longman 1991:83ff.). Brinkman is more cautious and states: “In its present form, the inscription may date from a later time; but this needs further study” (1976:97, n. 8). Others believe its authenticity: Gadd 1973:226; Weidner 1959–1960:138; Astour 1986:328–329; Dalley 1997:169–170; Eder 2004. For further references, see Foster 2005:364 and Pruzsinszky 2009:183–186. 8 Pinches 1909:2, “Text of Agû-kak-rimê.” 9 Balkan 1954:157: “Bedeutung ‘zweiter’ daraus zu erschliessen, dass Agum I. in dieser Inschrift Agum ma¶rû, ‘der erste Agum’, genannt wird.” Diakonoff offered a different interpretation, kak rēme, ‘sword of mercy’ (1956:126). Astour also took it as Akkadian and rendered: ‘Weapon of Thunder,’ by taking it kak rimme (1986:331). 10 Note, for example, URU KÁ.DINGIR.RAki ša KUR ak-ka-di-i (Grayson 1991:41, 6); URU KÁ.DINGIR šá [KUR ak]-ka-di-i (ibid.: 104 iv 38). 11 Hammurabi’s year 32 claims victory over the army of Eshnunna, Subartu, and Gutium and conquest of the land Mankisum, see Horsnell 1999 II 143–144, l. 134. 12 Diakonoff (1956:126–127) believed that Agum’s claim over his territory befitted the actual situation of the early Kassite period. On the other hand, Sassmannshausen (2004:64) argues that “the titles of Gandaš and Agumkakrime ‘king of the four world quarters’ and ‘king who causes the four world quarters to exist permanently’ in their inscriptions of disputed authenticity do not fit into the early Kassite period and this speaks against the authenticity of these inscriptions.”

228

Articles: Ancient Near Eastern Studies

The narrator states that he is the son of taš-ši-gu-ru-maš (i 2; 13). He also presents a genealogy of four generations. Agum Rabû (i 19) | Kaštilyašu (i 17) | Abir[attaš] (i 15) | Taššigurumaš (i 2; 3) | Agum (the narrator) The genealogy between the first Agum and Tazzigurumaš in The Agum-Kakrime Inscription can be supported by two king lists (Babylonian King List A, BM 33332 = Rm 3, 5;13 Synchronistic King List, Ass. 14616c = now A.117).14 However, there is no other reference to ‘Agum, the son of Tazzigurumaš’ in the extant cuneiform sources.15 Initially Weidner had offered the ninth king of the Kassite dynasty in the Synchronistic King List as our Agum, although this person was the third generation from Tazzigurumaš in the king list.16 F. Kraus collated the first column of the original tablet in 1949 and reported to Weidner that Weidner’s reading a-gu-[u]m could not be sustained. 17 Weidner withdrew his reading and offered x-gu-x. Brinkman collated the original tablet as well as the excavation photo and reported in 1976: “… the next sign begins with a definite horizontal wedge. The supposed -gu- is not certain; and there are no definite traces, as distinguished from scratches, for the final sign.”18 One can offer three possible solutions for this problem in light of the evidence in The Synchronistic King List. The first, and probably the most likely suggestion, is to take the ninth Kassite king in The Synchronistic 13

Gadd, CT 36, 241. Brinkman 1976:9–11. 15 Gasche et al. (1998:88–89) linked Agum, the son of Kashtiliash, to our Agum-Kakrime. The Chronicle of the Early Kings states: ‘Agum the son of Kashtiliash mustered his army and marched to the Sealand. He seized Dur-Enlil and destroyed Egalgasheshna, the temple of Enlil in Dur-Enlil’ (Grayson 1975a:156, ll. 15–18). However, based on the genealogy offered in the Agum-Karime incription, their suggestion seems less likely. 16 Weidner 1921:14; 1926:68. 17 Weidner 1959–1960:138. 18 Brinkman 1976:98. Cf. Grayson 1980–1983:121. 14

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions …

229

King List as our Agum-Kakrime. Based on the copy of this line in Weidner 195–160, one may read this line, WagX-gu-[m]e, a possible alternative orthography of a-gu-um. 19 The second suggestion is to speculate that one of the successors of Tazzigurumaš was indeed Agum but he changed his name during his reign. If this scenario is true, AgumKakrime was most probably the throne name of one of the successors of Tazzigurumaš, with the Synchronistic King list recording only his old name. The last possibility is that The Synchronistic King List does not list this Agum-Kakrime at all.20 Due to a lack of further evidence for the existence of Agum-Kakrime in cuneiform texts, we cannot come to any conclusion on this subject. Point B. The likely date of the events recounted in this inscription As there is no reference to the restoration work of Esagila by an early Kassite king in other extant cuneiform sources, it is very difficult to date the refurbishment of Esagila as recounted in the Agum-Kakrime Inscription. If we believe the genealogy presented in this inscription, he must have ruled Babylon in the early Kassite period, probably very close to the fall of the First Dynasty of Babylon. A Hittite source,21 as well as a Babylonian Chronicle,22 recounts the raid of the Hittites on Babylonia, which apparently brought an end to Hammurabi’s dynasty.23 In addition, The Marduk Prophecy reports that Marduk left for the land of the Hittites and stayed there for 24 years (i 17).24 One may take the allusions to Marduk’s divine headgear and his image (i 31–32) in a broken context later in the text as a reference to Agum’s restoration work on the cult statue of Marduk. Given the statements in these three different ancient Brinkman 1976:11, n. 15 offers m┌ag-gu ┐-. However, he refuses to take it as our Agum-Kakrime because the name Agum is customarily spelled a-gu-um in all the texts known to us, ibid. 13, n. 24. Astour offered m┌kak ┐-┌ri┐-i[m-m]e based on the traces copied by Weidner (Astour 1986:330). Nonetheless, his “improved” copy of this line in question does not really agree with Weidner’s copy in Weidner 1959–1960:138. 20 Brinkman 1976. 21 The Edict of Telepinu (TUAT I/5, 465–466). 22 The Chronicle of Early Kings (Grayson 1975a, chr. 20 B, rev.:11). It states that “at the time of Samsu-ditana” the Hittites raided Babylonia. Richardson (2002:50) suggests that the collapse of the First Dynasty of Babylon had already taken place during the time of Ammi´aduqa. 23 Brinkman 1976:12. 24 Borger 1973:5. 19

230

Articles: Ancient Near Eastern Studies

texts, one can suggest a date of ca. 1570 (according to the Middle Chronology) for the likely return of Marduk to Babylon and the restoration of Esagila by Agum-Kakrime.25 There are several objections to this suggestion. The first issue to be discussed is the place from where the statue of Marduk was retrieved. The narrator of The Agum-Kakrime Inscription states that he sent his messengers to KUR ¶a-ni-i, ‘the land of the Haneans’ (ii 9–10), in order to recover the statues of Marduk and Zarpanitu. However, The Marduk Prophecy recounts that Marduk travelled to the land of the Hittites (The Marduk Prophecy i 13–14). As The Edict of Telepinu and The Chronicle of Early Kings speak of the Hittites as the aggressors, it is reasonable to assume that the statues of Marduk and Zarpanitu were taken to Hattusha rather than Hana.26 Some scholars have offered different interpretations to cope with this problem.27 For example, Podany, in her study of “the kingdom of Hana,” takes this allusion to the land of Haneans literally and speculates that “the people of Hana assisted the Hittites and rebelled against Babylonian suzerainty.”28 One may also speculate that the scribe miscopied a sign and offer an emendation, ¶a-ti-i instead of ¶a-ni-i,29 although we normally expect ¶a-at-te/i or ¶at-ti for the GN Hatti.30 The second problem is the reliability of the claims in The Marduk Prophecy, as these statements cannot always be corroborated by other ancient texts. The best example is the account of Marduk’s “journey” to Assyria. The Marduk Prophecy claims that Marduk left for Assyria but returned to Babylon before the invasion of the Elamites and his departure to Elam.31 This journey to Assyria must be an allusion to 25

Note also the estimated 1530 for the accession of Burna-Buriash I (the tenth king of the dynasty) by Rowton (1973:207). Also Brinkman 1976:30–32, n. 87, and Sassmannshausen 2004:65. For the problem of the date of the fall of Babylon, see now Pruzsinszky 2009 (passim, esp. 93–101). 26 Gadd 1973:226. 27 Gadd ibid.; Goetze 1964:98; Weidner 1952–1953:6. All references are in Podany 1991–1993:58, n. 48. 28 Podany 1991–1993:60. She further suggests that the Haneans played an important role in the Kassites’ attempts to retrieve the statues of Marduk and Zarpanitu as the middlemen between the Hittites and the Babylonians. 29 Landsberger 1954:65, n. 160; Dalley 1997:165. Cf. also Brinkman 1972:281. 30 For the writing of Hatti see Nashef 1982:123–124. Another possible explanation is to take māt ¶anî as the indication of the northwestern Mesopotamian regions as well as Anatolia. See Albright 1952:25, n. 3. For this issue, see below. 31 Borger 1971:6–7 i 1′–17′.

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions …

231

Tukulti-Ninurta’s invasion to Babylonia recounted in Chronicle P i 2– 13.32 According to Chronicle P, Tukulti-Ninurta I (1243–1207) returned to Babylon and destroyed the city wall, he slew the residents and sacked the temple Esagila. He removed the statue of Marduk and sent it to Assyria as booty. The exact length of Marduk’s stay in Assyria is not preserved in Chronicle P.33 However, it could not have been earlier than the mid twelfth century B. C., i. e., after the invasion of the Elamites led by KudurNahhunte during the reign of Enlil-nadin-ahi (1157–1155), the thirty-sixth and the last king of the Kassite Dynasty,34 because Chronicle P iv 12–13 states that Bel stayed in Assyria ‘until (the time of) (Ninurta)-tukulti-Assur (1134).’35 The statue of Marduk that was removed by the Elamites could not be the same statue as that which was taken by Tukulti-Ninurta in the thirteenth century B. C., but rather a replacement made by the Babylonians after the conquest of Tukulti-Ninurta (1225). 36 Clearly The Marduk Prophecy was based on the idea that there had been only one image of Marduk, and that every time it “left” Babylon it returned. The author of The Marduk Prophecy presumably “altered” the historical facts in order to fit his view. In other words, we cannot rely solely on the accounts in The Marduk Prophecy in order to establish the historical facts either. Point C. The style of the Agum-Kakrime Inscription As already stated above, the Agum-Kakrime Inscription is composed in an Akkadian unilingual form. Akkadian royal inscriptions in the Kassite period form a minority as most of the Kassite building inscriptions known to us were composed in Sumerian.37 In addition, the first-person narrative form for a royal inscription in this period is also uncommon.38 32

Grayson 1975a:175–176. Boese (1982:20–21) proposed 86 years while Finkel and van der Spek offer 66 years (see http://www.livius.org/cg-cm/chronicles/chron00.html). 34 See Brinkman 1969:88–89. 35 For the regnal year of Ninurta-tukulti-Assur, see Brinkman ibid., pl. II (after 76). 36 Dalley 1997:166. The so-called Adad-shuma-usur Epic recounts a revolt of the priest of Babylon against the king (Grayson 1975b:64ff.). It is very likely that, at least according to the later tradition, Marduk’s statue had been “reinstalled” by the time of Adad-shumausur because this text refers to the statue of Bel, i. e. Marduk (ii 24). It seems, however, that the main subject matter of this text is Adad-shuma-usur’s restoration of Esagila as it refers to craftsmen and bricks in col. iii. 37 Grayson 1980:159. 38 Brinkman 1976:63, n. 227. 33

232

Articles: Ancient Near Eastern Studies

Other examples of the first-person Akkadian royal inscriptions are: a first-millennium copy of an inscription of Gandaš (BM 77438),39 who is credited as the first generation of the Kassite dynasty in the king lists, and a royal inscription of Kurigalzu the son of Kadašman-Harbe, which is also known only as a late copy.40 Brinkman, for instance, doubts the authenticity of the inscriptions of Gandaš and Agum-Kakrime.41 Nonetheless, these two facts do not rule out the possibility that an early Kassite king left his inscription in an Akkadian first person narrative form. The royal inscriptions from the kings of the First Dynasty of Babylon were composed both in Sumerian and Akkadian but not in bilingual formulae. The Sumerian texts and the Akkadian texts are normally written on separate objects and never appear together. In other words, these texts principally form unilingual texts. The royal inscriptions were written in both the first and the third persons during the Old Babylonian period. When it is in the first singular, the verbs also appear in the affirmative but without subjunctive mode. For example, Hammurabi (Frayne 1990, E.4.3.6.12): lu ù-ul-li [17]; lu uš-ta-ás-¶i-ir-šu [19]; lu a¶-re-a-am [22]; Samsu-iluna (ibid., E.4.3.7.5): lu ú-´i-ir [61]; lu ú-ša-tir-šu [61]; lu iš-ru-ku-nim [83]. The same is true also in Assyria of this period. For example, Šamšī-Adad I (Grayson, 1987, A.0.39.1): lu am-ta-¶a-ar [80]; lu-ú ašku-un [87]; A.0.39.2: lu-ú aš-ku-un [iii 5]; lu-ú iš-ru-kam [iii 9]. Given this, we can still place our Agum-Kakrime inscription in the early Kassite period. Possible Forgers? We need to ask an inevitable question. Is this account of Agum-Kakrime’s recovery of the statue of Marduk and the restoration work of Esagila a fiction? Although many modern scholars have argued that it is a later apocryphal writing, as we have already discussed, nothing in this text points out with any certainty that it is a later fiction. However, it is also admittedly true that we cannot prove its authenticity either. If this Agum-Kakrime Inscription is proven to be an ancient forgery, who would go through all of the trouble to invent an imaginary king and fake an inscription? It is possible to speculate that this Agum-Kakrime In39

For references and an edition, see Stein 2000:149–150. See also Eder 2004: 214–221. Eder placed Gandaš during the reign of Samsuditana based on the Gandaš Inscription (ibid. 214–217). See also Pruzsinszky 2009:99. 40 For the edition see below. For further references, see Brinkman 1976:209 and Foster 2005:366. 41 Ibid. 97 and 127.

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions …

233

scription was based on the ideology of the priests and thinkers of Esagila and that there was only one statue of Marduk which always returned to Babylon by Marduk’s will, as stated in The Marduk Prophecy. As they were responsible for the archive(s) of Esagila, they must have had access to the records of the deeds of the ancient Babylonian kings. In other words, they could have falsified a building inscription by combining different pieces of information which were at their disposal, probably found in the archive of Esagila. If we look for the person who could have been behind the hoax, the priests and intellectuals of Esagila were the only people who had a motive, an opportunity and the skills to execute it. 42 Nevertheless, in order to prove or disprove any of the speculations raised here we would need further evidence. Edition of the Agum-Kakrime Inscription (composite text)43 Manuscripts MS A: K.4149+4203+4348+Sm 27 (5R 33: photos, pp. 258–259) MS B: Rm 505 (Campbell Thompson 1930, pl. 36: photos, pp. 260–261) Transliteration Col. i 1. [ … u]m-ka-ak-ri-me 2. WDUMU taz-ŠI X-gu-ru-maš 3. NUMUN el-lum 4. ša dšu-qa-mu-nu 5. ni-bi-it da-nim u d+en-líl 6. dé-a u dAMAR.UTU 7. d30 u dUTU 8. eṭ-lum da-an-nu 9. ša diš-tar qá-rit-ti 10. i-la-a-ti a-na-ku ——————————— 11. LUGAL mil-ki u ta-šim-ti 12. LUGAL téš-me-e u Wsa-li X-me 13. DUMU taz-ŠI-g[u-ru-ma]š 14. li-ib-li-i[b-bi (x)] 15. ša a-bi-r[at-taš] 42

E. g., Borger 1971:17: “Tatsächlich könnten interessierte Kreise im Marduk-Tempel von Babel sie im Zusammenhang mit der Erhöhung Marduks unter Nebukadnezar I. oder sogar noch später verfasst haben.” 43 The asterisk in the transliteration indicates signs which were copied by Pinches (1909) but are nor visible any more because of deterioration of the condition of the tablets.

234

Articles: Ancient Near Eastern Studies 16. qar-ra-W duX e[ṭ-lum] 17. DUMU kaš-til-[i]a-šu 18. IBILA r[eš-t]u!-Wú!X 19. ša a-gu-um ra-Wbi-iX 20. NUMUN el-lum NUMUN LUGAL-ti 21. [t]a-mi-i¶ ´er-ri-ti 22. Wni X-[i?]-ši Wre-é-úX 23. [g]a-áš-ru Wa-na X-ku 24. re-é-i 25. UN.MEŠ DAGAL.MEŠ-tim 26. qar-ra-du 27. re-é-a-um 28. mu-ki-in 29. SUḪUŠ gišGU.ZA a-bi-šú 30. a-na-ku ——————————— 31. LUGAL kaš-ši-i 32. ù ak-ka-di-i 33. LUGAL KUR KÁ.DINGIR.RAki 34. ra-pa-aš-tim 35. mu-še-ši-ib 36. KUR áš-nun-na-ak UN.MEŠ 37. DAGAL.MEŠ-tim LUGAL KUR pa-da-an 38. u al-ma-an LUGAL KUR gu-ti-i 39. UN.MEŠ sak-la-a-ti 40. LUGAL muš-ta-aš-ken 41. kib-rat ar-ba-a’-i 42. mi-gir DINGIR GAL.GAL.LA 43. a-na-ku ——————————— 44. i-nu dAMAR.UTU 45. E[N] é-sag-íla 46. WùX KÁ.DINGIR.RAki 47. [DINGI]R? GAL.GAL.LA 48. [i-n]a pi-i-šu-nu el-lim 49. [a-n]a KÁ.DINGIR.RAki 50. [ta]-ar-šú iq-bu-u 51. Wd XAMAR.UTU ana TIN.TIRki 52. [DÚR?-t]i?-[š]ú? pa-ni-šú iš-ku-na 53. [xxx(x)]-x-[m]a dAMAR.UTU 54. [xxxxxxx(x)]-WúX-a-a Col. ii 1. ak-pú-ud-ma at-ta-id-ma 2. a-na le-qé-e dAMAR.UTU 3. a-na KÁ.DINGIR.RAki 4. pa-ni-šu áš-kun-ma

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions … 5. tap-pu-ut dAMAR.UTU 6. ra-im BALA-e-a 7. al-lik-ma ——————————— 8. LUGAL dUTU ina SILA4 lúḪAL a-šal-ma 9. a-na KUR ruq-ti a-na KUR ¶a-ni-i 10. lu-ú áš-pur-ma ŠU dAMAR.UTU 11. ù dzar-pa-ni-tum 12. lu i´-ba-tu-nim-ma 13. dAMAR.UTU ù dzar-pa-ni-tum 14. ra-im BALA-e-a 15. a-na é-sag-íla 16. ù KÁ.DINGIR.RAki 17. lu ú-tir-šu-nu-ti 18. i-na É šá dUTU 19. i-na pa-ra-as EGIR 20. ú-kin-na 21. lu ú-tir-šu-nu-ti-ma 22. DUMU.MEŠ um-ma-ni 23. lu ú-še-šib-šu-nu-ti 24. TABIR KÙ.DÍM u BUR.GUL 25. lu Wú X- [ša]-Wa´-bit? X- ma 26. […]-šu-nu 27. lu(-)WúX(-)[ … ] x 28. LÍM GÚ WKALX [KÙ.GI S]A5 29. a-na lu-bu-u[š-ti] 30. dAMAR.UTU ù dzar-pa-ni-tum 31. lu-ú ad-di-nu-ma 32. lu-bu-uš-ta ra-bi-ta 33. lu-bu-uš-ta KÙ.GI SA5 34. dAMAR.UTU ù dzar-pa-ni-tum 35. lu ú-lab-bi-šú-šú-nu-ti-ma 36. na4ZA.GÌN.KUR.RA na4ZÚ mar-¶a-ši SIG7 37. na4NÍR IGI na4NÍR.MUŠ.GÍR 38. na4NÍR.BABBAR.DILI na4NÍR.NANNAR.MIN5 39. na4IGI.MEŠ me-lu¶-¶a 40. na4GIŠ.NU11.GAL 41. na4SAL.LA aq-ra u NA4 SIKIL 42. ša ina KUR-šu na-as-qu 43. a-na áš-rat dAMAR.UTU 44. ù dzar-pa-ni-tum 45. lu-ú ad-di-nu-ma 46. mu-u¶-¶i lu-bu-uš-ti 47. i-lu-ti-šú-nu 48. ra-bu-tim 49. lu ú-za-’i-i-nu-ma 50. a-ge-e qá-ar-ni

235

236

Articles: Ancient Near Eastern Studies 51. ´i-ra-a-ti 52. a-ge-e be-lu-ti 53. si-mat i-lu-ti 54. ša ša-lum-ma-ti 55. ma-la-t[a?] Col. iii 1. WšaX [n]a4ZA.GÌN ù KÙ.GI 2. i -WnaX SAG.DU-šu 3. lu-ú Wáš X- ku-nu-ma 4. i-na UGU S[A]G a-ge-šu 5. na4NÍR IGI [MU]Š.GÍR 6. NA4 me-ni-[š]ú-ti 7. lu-ú áš-ku-nu-ma 8. na4NÍR MUŠ.[G]ÍR 9. na4KA.MARku6 IGI na4ZA.[GÌN] 10. na4NÍR.BABBAR.DILI 11. ina UGU a-ge-š[u] 12. lu ú-za-i-nu-[ma] ——————————— 13. MUŠ.ḪUŠ-a šu[b-tu] 14. i-lu-ti-šú G[AL-tim] 15. KÙ.GI lu Wú X- [še-pi?] 16. ki-gal-la r[a-ba] 17. šu-bat(-)x(-)[ … ——————————— 18. lu-ú [ … 19. aš-k[un?(-) … 7 lines missing 26. [xxxx] xx [ … 27. lu(-)ú(-) [ … 28. lu ú-šib* x* [ … 29. a-bu-us-sa*-a[t* … 30. ta-am-l[i-ti … 31. lu aš-ku-n[u]-Wši X- n[a?-ti] 32. na4NÍR na4M[UŠ].[G]ÍR 33. na4NÍR [n]a4[x]x 34. É.KIŠIB.MIN.KAM giš[E]R[EN?] —————————————— 35. mu-¶i ir-ti dA[MAR?.UTU?] 36. lu ú-za-i-n[u-ma] 37. ¶i-iš KÙ.GI n[a4x x] 38. na4ZÚ na4ZA.[GÌN] 39. i-na na-piš-[ti-šú-nu] 40. lu-ú áš-k[u-nu-ma]

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions … 41. šu-kut-ti [xxxx] 42. šu-kut-ti [xxxx] 43. na4IGI.MEŠ na4Z[A.GÌN] 44. na4NÍR M[UŠ?.GÍR?] 45. na4ZÚ n[a4xx] 46. na4NÍR.DILI n[a4NÍR.ḪAL] 47. i-na WIGI.MIN.MEŠX-[šu-nu] 48. lu Wú X- [za-i-nu]-šú 49. WùX [ … -t]i? Col. iv 1. lu-WúX [u]¶-¶i-zu-ma 2. i-na UGU šub-tì-šú 3. šu-bat gišERE[N] 4. lu uš-zi-zu-ši-ma 5. a-di aš-ra-ti 6. i-lu-ti-šú-nu 7. ra-bi-ti 8. ú-šá-a´-bi-tu-ma ———————————— 9. [DUMU u]m-ma-ni i-pu-šú 10. [a-na] Wgiš XEREN gišŠUR.MÌN 11. [x]-x-LAM ana KUR KÙ 12. [x]-x-su ¢a-a-bu 14. [lu]-WúX áš-pur-ma 15. [xxxx K]Ù.M[EŠ] 14 lines missing 29. [l]u(-)ú-[ … 30. ina GIŠ.MEŠ [ … 31. Wšu! X- a-t[u … 32. gišIG.MEŠ GAL.GAL.MEŠ 33. gišIG.MEŠ gišEREN 34. tu-’a-Wma X- ti 35. [l]u ú-še-pí-šu-ma 36. ina pa-pa-¶a-a[t] dAMAR.UTU 37. ù dzar-pa-ni-tum 38. lu-ú ú-k[i]-WinX 39. ina a-šá-at ZA[BAR*] 40. [ni]k-ku-r[i] 41. lu-ú áš-pu-ši-na-t[i] 42. [š]a-gam-mi-ši-n[a] 43. i-na ku-ru-us-s[i] 44. ša URUDU mi-si-WiX 45. lu ú-´ab-bit 46. ba-aš-me WlචX-me

237

238

Articles: Ancient Near Eastern Studies 47. ku-sa-rik-kum 48. u4-gal-la 49. UR.IDIM 50. [k]u6-lú-u17-lu Col. v 1. [SUḪ]UR.MÁŠku6 2. [i-n]a na4ZA.GÌN 3. [n]a4DUḪ.ŠI.A 4. [n]a4GUG 5. [n]a4GIŠ.NU11.GAL 6. lu ú-šá-am-lu-ši-na-a-ti 7. te-lil-ta-šu-nu 8. lu ú-ša-áš-ki-in-ni-ma 9. gišIG el-li-tim 10. i-na pa-pa-¶a-at 11. dAMAR.UTU 12. ù dzar-pa-Wni X- tum 13. lu ú-kin-ši-na-a-ti ——————————————— 14. ù é-sag-íla 15. mit-¶a-riš MUŠ.LAḪ4 16. lu-ú ul-li-lu-ma 17. iš-tu te-lil-ti 18. Wbi X*-tim mit-¶a-r[iš] 19. [MAŠ*.MAŠ …] 16 lines missing 35. [… B]I [… 36. [xx]-WúX ruk-k[us] 37. [i-n]a ká-su-l[im]-WmaX 38. pa-pa-¶a-at dAMAR.UTU 39. lu ú-še-rib-šu-nu-ti 40. ni-ga-ti-šú-nu 41. ra-ba-a-ti 42. lu-ú e-pu-uš 43. ée-da-di-¶é-gal 44. lu ú-šá-ap-qid 45. a-na be-lí u be-el-ti 46. qí-šá-at!-šu-nu lu ad-din Col. vi 1. Wú? X- [u]l [… 2. x [… 3. Wlu(-)úX(-)[… 4. é-è[š?- … —————————————————

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions … 5. [G]AL KÙ.G[I] [GA]L [KÙ.BABBAR] 6. [G]AL na4ZA.[GÌN] 7. ta-ri-im-te KÙ.B[ABBAR] 8. ra-bi-ti 9. a-na dAMAR.UTU lu ad-din 10. qí-ša-a-ti 11. šá KÙ.BABBAR u KÙ.GI 12. a-na DINGIR.DILI šá é-sag-íla 13. lu ad-din ul-tu 14. é-sag-íla 15. ¢a-a-bu u dam-qa 16. ap-qí-du ————————————————— 17. a-na ši-ma- š[á … 18. [a-n]a! DINGIR.G[AL.GAL] 19. […] xx [… 7 lines missing 27. x [… 28. a-na I[GI? … 29. dAMAR.UTU xx [… 30. lu(-)Wú X(-)[… ————————————————— 31. mZALÁ-d[ … 32. a-di [É-šu] 33. A.ŠÀ-šu WùX [gišKIRI6-šu] 34. mNÍG.BA-dA[MAR?.UTU?] 35. MAŠ.MAŠ a-di WÉ X- [šu] 36. A.ŠÀ-šu ù gi[šKIRI6-šu] 37. mdAMAR.UTU-mu-bal-l[iṭ]? mNAGAR 38. a-di É-šu A.ŠÀ-šu 39. gišKIRI6-šu ———————————————— 40. a-na LUGAL a-gu-um 41. ša pa-pa¶ dAMAR.UTU 42. i-pu-šu 43. é-sag-íla ud-di-šu Col. vii 1. dAMAR.UT[U] 2. a-na šub-ti-šú 3. ú-še-ri-[b]u 4. qí-ša-a-tim 5. DUMU.MEŠ um-ma-ni 6. šu-nu-ti 7. qà-du É A.ŠÀ u gišKIRI6 8. a-na dAMAR.UTU

239

240

Articles: Ancient Near Eastern Studies 9. WùX dzar-pa-ni-tum 10. ú-zak-ki-šu-nu-ti ——————————————— 11. ša LUGAL a-gu-um 12. U4.MEŠ-šú lu-u ar-[k]u 13. MU.MEŠ-šú li-ri-ka 14. BALA-šú ina dum-qí 15. lu bu-ul-lu-ul 16. ´er-re-et AN-e 17. rap-šu-ti 18. li-ip-pé-ta-šú 19. er-pé-tu 20. zu-W un X- na 21. l[i-šá-az-ni-in] 22. […] 23. dA[MAR.UTU] 24. ´i-pat gi[šKIRI6] 25. a-na da-riš [U4.MEŠ] 26. in-ba ṭa-a-[ba] 27. li-te-el-li-[pa] 28. a-na LUGAL dam-q[í] 29. a-gu-um 30. ša pa-pa-¶a-at dAMAR.UTU 31. i-pu-šu 32. DUMU.MEŠ um-ma-ni 33. ú-zak-ku-ú —————————————————— 34. da-nu u an-tum 35. ina AN-e lik-ru-bu-šu 36. d+en-líl u dnin-líl 37. i-na é-kur ši-mat TI.LA 38. li-ši-mu-šu 39. dé-a 40. u ddam-ki-na 41. a-ši-ib ABZU GAL 42. TI.LA U4.MEŠ 43. GÍD.DA.MEŠ 44. lid-di-nu-šu 45. DINGIR.MAḪ dbe-let KUR GAL.GAL Col. viii 1. [NUM]UN el-lam (text QA!) 2. [l]i-šak-lil-šu 3. d30 dNANNAR AN-e 4. NUMUN NAM.LUGAL.LA 5. a-na U4.MEŠ GÍD.DA.MEŠ 6. lid-di-iš

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions … 7. eṭ-lum dUTU 8. e-tel AN-e 9. ù KI-tim 10. SUḪUŠ gišGU.ZA 11. LUGAL-ú-ti-šú 12. a-na U4.MEŠ 13. GÍD.DA.MEŠ 14. li-ki-in 15. dé-a 16. EN IDIM 17. né-me-qam 18. li-šak-lil-šu 19. dAMAR.UTU ra-im BALA-šú 20. be-el na-ag-bi 21. ¶é-gál-la-šu 22. li-Wšak X- lil-šu —————————————————————— —————————————————————— 23. MU.[SAR].R[A] 24. [šá a-g]u-um 25. [ZU]-ú ZU-a 26. [li]-kal-lim 27. [NU] ZU-ú 28. [NU] im-mar 29. [NÍG.G]IG dPA 30. WùX dLUGAL 31. dUTU ù dIM 32. DINGIR.MEŠ ´e-ru-ti 33. EN.MEŠ MÁŠ 34. ¢up-pi mdaš-šur-DÙ-IBILA 35. LUGAL ŠÚ-ti LUGAL KUR aš-šurki 36. šá a-na AN.ŠÁR u dNIN.LÍL GI 37. NIR.GÁL.ZU NU TÉŠ 38. LUGAL DINGIR.MEŠ AN.ŠÁR 39. man-nu šá TÚM-lu MU-šú GIM MU-ía SAR-ru 40. AN.ŠÁR u dNIN.LÍL ag-giš MIR-iš 41. [lis]-ki-pu-šu-ma 42. [MU-š]ú NUMUN-šú ina KUR ḪA.A.MEŠ

Translation Col. i 1–10 [Agu]m-Kakrime, the son of Tazzigurumash, the pure seed (i. e., descendant) of Shuqamunu, nominated (to the kingship) by Anu and Enlil, Ea and Marduk, Sin and Shamash, the strong young man of Ishtar, the most heroic among the goddesses, am I.

241

242

Articles: Ancient Near Eastern Studies Col. i 11–30 The king of counsel and wisdom, the king of listening to (prayers) and peace, the son of Tazzig[uruma]sh, the offspr[ing] of Abir[attash], the warrior, the y[oung man], the son of Kashtil[y]ashu, the pr[eemine]nt son of the great Agum, the pure seed, the seed of kingship, the one who holds the nose-ropes of the people, the powerful shepherd, am I. The shepherd of the wide-spread people, the warrior, the shepherd, the one who establishes the foundation of the throne of this father, am I. Col. i 31–43 The king of the Kassites and the Akkadians (i. e., the Babylonians), the king of the wide land of Babylon, the one who settles the land of Eshnunna, the wide-spread people, the king of Padan and Alman, the king of the land of the Gutians, the barbarous people, the king who subjugates the Four-Corners (of the world), the favourite of the great gods, am I. Col. i 44 – ii 7 When Marduk, the lord of Esagila and Babylon (and) the great [gods] commanded his return through their holly command, Marduk set his face towards Babylon, [h]is [dwelling-pla]ce. [....(.)]. Marduk, I have planned [......(.)] of mine and paid attention. I urged Marduk to (allow me) to take him to Babylon. For the aid of Marduk, the one who loves my reign, I marched. Col. ii 8 – iii 12 I asked the king Shamash by means of the lamb of the diviner. I sent a message to the distant land, to the land of Haneans. Marduk and Zarpanitu indeed seized the hand and I indeed returned them —Marduk and Zarpanitu, those who love my reign—to Esagila and Babylon. I indeed brought them back to the building which Shamash confirmed during the examination of the facts. I indeed settled the craftsmen (in the temple). I indeed [settled] metal workers, goldsmiths, and seal-cutters. I indeed […] their […]. I indeed gave four talents of precious red-gold for the garments of Marduk and Zarpanitu. I indeed dressed Marduk and Zarpanitu with the great garment(s), their red-gold garment(s). I indeed donated lapis-lazuli of the mountain, green obsidian of mar¶ašu, ¶ulālu-beads, muššatu-stone, pappardilû-stone, papparmīnustone, beads from melu¶¶a, alabaster, precious sēlu-stone, and the pure stones, which were chosen in its mountain, to the sanctuary of Marduk and Zarpanitu. I indeed overlaid onto the garments of their great divinity. I indeed placed the headgear of the lordship, the symbol of the divinity with full awesome luminosity, made of lapis-lazuli and gold on his head. On the top of his headgear, I

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions … placed ¶ulālu-stone, muššaru-beads, and menišutu-stone. I overlaid ¶ulālu-stone, muššaru-(beads), kamāru-stone, lapis-lazuli-beads, and pappardilû-stone onto his headgear. Col. iii 13–18 I in[deed created] Muš¶uššu-dragon, the dwel[ling-place] of his g[reat] divinity from gold. The gr[eat] dais, the dwelling-place […] indeed […] I p[laced … 7 lines missing Col. iii 26–36 […] … [… I indeed […] I indeed settled […]. The forelock […]. Completi[on …] I indeed set t[hem]. The ¶ulālu-stone, muššaru-stone, and ¶ulālustone, […]-stone, the second seal-storage-box made of cedar onto the breast-(plate?) of M[arduk], I indeed overlaid. Col. iii 37 – iv 8 I placed a band (made) of gold, [xx]-stone, obsidian, la[pis-lazuli] around [their] nec[k?]. I overlaid the adornment of […], the adornment of […], with the beads of l[apis-lazuli?], ¶ulālu-stone, mu[ššatustone], obsidian, [… s]tone, pappardilû-stone, [papparmīnu-stone], in [their] eyes. And, I indeed mounted […]. I set it up on his dais, the dais made of cedar, and I (indeed) installed (it) together within the sanctuaries of their great divinity. Col. iv 9–15 I [indeed] sent […] for cedar (and) juniper to the holy mountain, the pleasant […]. [H]oly […] 14 lines missing A line is too damaged to read. Col. iv 28 – v 13 I indeed produced the great doors, the twin doors of cedar, from the trees […] thes[e …]. I indeed set (them) up at the cella of Marduk and Zarpanitu. I indeed wrapped them with reins of nikkurru-bronze. I made their upper door hinge fast with bands of pure copper. I indeed filled them (the doors) with bašmu-dragon, la¶mu-man, kusarikku, ugallu, Mad Dog, kulullu, su¶urmaššu, (made) of lapis-lazuli, du¶šû-stone, carnelian and alabaster. I indeed carried out their (quasi-divine beings’) purification and I indeed set them up, the pure doors, in the cella of Marduk and Zarpanitu. Col. v 14–19 And the snake-charmers indeed purified Esagila everywhere. After the overall purification of the temple, [the incantation-priest …

243

244

Articles: Ancient Near Eastern Studies 16 lines missing Col. v 35 – vi 4 […] […]. […] was bound. I indeed brought them [i]n Kasul[im]ma (The Gate of Radiance), the cella of Marduk. I indeed did great festivals for them (lit. their festivals). I indeed gave (these offerings) to the care of Edadihegal (The House of Abundant Tribute). I indeed gave to the lord and the lady their offerings. [I did n]o[t …] […], I indeed […] Ee[š… Col. vi 5–16 I indeed gave a goblet made of gold, [a gobl]et [made of silver], and a goblet made of lapis-[lazuli], a great present of gold. I indeed gave gifts of silver and gold to the gods of Esagila. After I took care of Esagila pleasantly and favourably (lit. pleasant and favourable) … Col. vi 17–19 For the decree o[f …] [t]o the g[reat] gods […] […] […] 7 lines missing The following 4 lines are too damaged to read. Col. vi 31–39 (I exempted) Nūr-[DN, the snake-charmer?] as well as [his house], his field, a[nd his orchard]; Qišat-M[arduk], the incantation-priest, as well as [his] h[ouse], his field and [his orchard]; Marduk-muball[i¢], the carpenter, as well as his house, his field (and) his orchard. Col. vi 40 – vii 10 (A prayer) for the king Agum, who made the cella of Marduk, renewed Esagila, and brought back Marduk to his abode, exempted them (from tax for) the gifts of these workers as well as the house, field, and orchard for Marduk a[nd] Zarpanitu: Col. vii 11–33 May the days of the king Agum be long. May his years become long. May his reign be smeared with prosperity. May the nose-rope of the wide-spread heavens be opened for him. May the clouds [bring down] rains. […] May M[arduk] always multiply the fruit [orchard], sweet fruit, for distant [days] for the favourable king Agum, who made the cella of Marduk (and) exempted the workers. Col. vii 34 – viii 22 May Anu and Antum bless him from the heavens. May Enlil and Ninlil determine the destiny of (good) health to him. May Ea and Damkina, those who dwell in the great Apsû, give him long days of (good) health. Dingirmah, the lady of the great mountains, provide him with a pure [s]eed. May Sin, the luminosity of the heavens,

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions …

245

renew the seed of the kingship for long days. May the young man, Shamash, the potentate of the heavens and the earth, establish the foundation of the throne of his kingship for long days. May Ea, the lord of the deep spring, provide him with wisdom. May Marduk, the one who loves his reign, the lord of the deep spring, provide him with his abundance. Col. viii 23–33 The in[scrip]ti[on] [of Ag]um: [The one who know]s [shall] show (only) the one who knows, the one who [does not] know shall [not] see. The taboo of Shullat and Hanish: Shamash and Adad, the lords of divination. Col. viii 34–42 The tablet of Assurbanipal, the king of the entire world, the king of the land of Assur, who trusts Assur and Mullissu. Do not bring shame (on me) who trusts in you, the king of gods, Assur. If anyone will take out (my name) and write his name instead of my name, Assur and Mullissu shall furiously rage and [r]eject him. May they make his [n]ame and his offspring disappear from the earth.

Commentary i 2. As for the reading of the name of this king, we follow Grayson (1980– 1983:134). Needless to say, it is also possible to read Uršigurumaš. His name is normally written UR-zi-gu-ru/guru12-ma-áš/maš in the king lists (King List A and The Synchronistic King List)44 instead of UR-ŠI-gu-rumaš in The Agum-Kakrime Inscription i 2 and 13. As other texts have ZI instead, we take the sign UR to be taz, not taš. If our reading is correct, this combination of ZŠ in our text may present the consonant z̄, the sound shared with Hurrian. For this consonant, see Brinkman (1976–1980:473). i 27. re-é-a-um: a is most probably used instead of ja (= PI) which was often used instead of ’a in OB and MB. Although we admit that it is less common in the Old/Middle Babylonian dialects, the sign PI for a halfvowel or guttural is also attested in Old Babylonian literary texts (e. g., Babylonian “Man and his God”).45

44

Brinkman 1976:320. The Babylonian “Man and his God,” ll. 25 (ra-bi-PI-ku, for rabiāku), 50 (qá-biPI-at, for qabiāt), 51 (di-PI-am, for di’am), and 66 (pa-ti-PI-et, for patiēt) but te-le-’i (l. 51). For the edition, see Lambert 1987:188–202. 45

246

Articles: Ancient Near Eastern Studies

i 36. For the writing of Eshnunna, see Jacobsen 1934:1ff., esp. 15f. and Nashef 1982:41. This line probably refers to Agum’s sovereignty over the Diyala region, north-east of the mid-Tigris. i 37–38. Padan and Alman. Alman can be the same as Arman and Halman, see Nashef 1982:15. ii 9. The identification of ‘the land of the Haneans’ is difficult. Based on other ancient documents (see above in the introduction), we here expect ‘the land of the Hittites’ instead of the Haneans. During the Old-Babylonian period and onwards, the land of Hana/Haneans normally indicated either the Hanâ kingdom or the regions around the middle Euphrates and the Habur river (Groneberg 1980:89). However, by the 13th century, this term was also used for the north-western mountain regions of Mesopotamia, around the land of Lullumê/Lullubê.46 For example, RIMA 2, 103:17–18 (an inscription of Aššur-bēl-kala): KUR ¶a-a-na šid-di KUR lu-lume-e ù KUR.MEŠ-ni47 šá KUR.KUR na-i-ri, ‘the Land of Hâna, in the district of the land of Lullumê and the mountain-regions of the lands of Na’iru.’ This is further testified to by an ancient lexical text, The Miscellaneous Geographical List (MSL 11, 55obv. ii 26–29):48 KUR ḪÉ-A-NAki

= KUR bi-ta-lá = KUR ¶a-né-e KUR LU-LU-BIki = [KUR] lul-lu-bé-e

KUR ḪÉ-A-NAki

Interestingly, ḪAR-gud Recension E, rev.:8–11 (MSL 11, 35) refers to the Land of the Haneans together with the land of the Hittites: KUR GU-TI-UMki

= KUR gu-ti-i = KUR ab-da-da-ni = KUR ¶a-ni-ì = ka[š]-šú-u LUGUD2.DA.MEŠ49 giš ki KUR ÉRIN = KUR e-ri-ni = KUR ¶at-ti KUR MÁ-RÍki = ŠU = KUR ¶at-ti ku-ur ¶a-an

KUR ḪÉ-A-NAki

It is clear from this toponym list that the term ‘the land of the Hittites’ did not mean the Anatolian mountain regions, the territory of the former Hittites, but modern eastern Syria.50 Given this, one may speculate that there was no significant difference between ‘the land of Haneans’ and 46

Nashef 1982:188–189. This should be read either šadânu or ¶uršānu. 48 Note also MSL 11, 15, Urra = ¶ubullu XXI, sec. 4, 44–46. 49 ‘The Short Kassites.’ 50 Hawkins 1972–1975:152. Note also The Sargon Geography which describes the empire of Sargon the Great. Grayson 1974–1977:56–64 and Horowitz 1998:79f. 47

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions …

247

‘the land of the Hittites’ for the Neo-Assyrian scribe who copied the manuscript.51 ii 22. Mārū ummâni in this text must indicate craftsmen rather than temple scholars as the following lines allude to workers, such as metal workers, goldsmiths, and seal-makers. Judging from the reference to the production of the clothing and the headdresses of the divine pair with different precious metals and jewels, they were apparently involved in refurbishing the cult-statue of Marduk and Zarpanitu as well as other decorations of their cella in the temple Esagila. It seems that seal-makers were employed for shaping jewels rather than in the production of seals probably due to their skill to cut precious and semi-precious stones. ii 28. Note that, according to the Marduk Ordeal, Marduk was dressed with red wool. Apparently it was red because it was dyed with Marduk’s blood. See Livingstone 1989:82, l. 15 (Assur Version) and 88, l. 23 (Nineveh Version). ii 36. The orthography na4ZA.GÌN.KUR.RA, lit. ‘lapis-lazuli of mountain,’ against na4ZA.GÌN ‘lapis-lazuli’ in other lines clearly suggests that there were two different kinds of ‘lapis-lazuli.’ While this ‘lapis-lazuli of mountain’ was the lapis-lazuli found in a mountain, i. e., genuine lapis-lazuli, na4 ZA.GÌN was most likely man-made lapis-lazuli, i. e., glass. For further discussion, see Oppenheim et al. 1970:9–14, esp. 13. See also CAD Š1 53f. šadû A, 1 d) and e). iii 13. Muš¶uššu is a horned snake-dragon whose identity was secured on the basis of the snake-dragons of the Ishtar-gate of Babylon and the inscription of Nebuchadrezzar II. The images of Marduk and Nabû as well as their divine symbols often appear on the back of this horned snakedragon. As this section concerns the dais, most likely of Marduk and Zarpanitu, it is very plausible that this Muš¶uššu was a part of the cultic platform, on which the statues of Marduk and Zarpanitu were placed. Incidentally, the horizontal line before iii 13 is not visible in Rm 505. iii 29. Von Soden offers a restoration a-bu-us-sa-a-at [bīti?] (AHw. 9 abūsu, abussu).

51

Nonetheless, the GN Úâna was used for the regions west of Babylonia in the Neo-Assyrian and Neo-Babylonian documents, see Zadok 1988:151.

248

Articles: Ancient Near Eastern Studies

iii 30. Von Soden offers ta-am-[l]i?[-i?] (AHw. 1316 tamlīu, tamlû). The reading LI is secured by our collation. iii 32–36. Interpretation of lines 32–36, particularly line 34, is difficult due to lacunae. The first two signs of line 34 are É MES, not É GAL as copied by Pinches in 5R 33. Here we propose to read É.KIŠIB, probably a container for the cylinder-seal of Marduk (rather than bīt kunukki ‘storehouse’). We know of two examples of cylinder-seals belonging to Marduk. The first one was dedicated by Marduk-zakir-šumi to Marduk himself52 while the second seal was dedicated to Adad of Esagila.53 Clearly this Adad is one of the Fifty Names of Marduk, rather than the god Adad.54 One might wonder if it is also possible to read É-MES instead. Emes is known as a sanctuary of Marduk in the Canonical Temple List 230,55 otherwise the temple of Lil, the son of Bēlet-ilī.56 However, if we understand this section correctly, this was an object placed on Marduk’s chest, and it is hard to believe that there was a miniature shrine of Marduk placed on his chest. giš [E]R[EN?]: Based on the visible traces, one may also reconstruct giš [TUK]UL.M[EŠ?] ‘weapons,’ which could be taken as Marduk’s abūbu ‘The Flood-Weapon.’ The Flood-Weapon is known as one of the Marduk’s weapons in Enūma Eliš IV 49: iš-ši-ma be-lum a-bu-ba gišTUKUL-šú GAL-a ‘The lord (Marduk) wielded the Deluge, his great weapon.’

Marduk himself is also addressed abūb kakki in BMS 12:23.

52

We know from the inscription of this kunukku of Marduk that it was dedicated by Marduk-zakir-šumi and hung around the neck of the statue of Marduk. Yet, the identity of the dedicator is a matter of debate as there are two Babylonian kings who bore this name. Wetzel suggested that it is Marduk-zakir-šumi I from the ninth century B. C. (Wetzel et al. 1957:37). On the other hand, Lambert believes that it is Marduk-zakir-šumi II who reigned in Babylon for only one month in 703 B. C. (Lambert 1985:89–90). For further discussion on Marduk and the snake-dragon, see Oshima 2006. 53 Wetzel et al. 1957:36–37. Although it bears an inscription of Esarhadon, it is very likely that this kunukku was made much earlier, possibly during same period as the second kunukku of Marduk discussed above judging from the iconography and the sign forms. 54 Note that KAR 142 iii 19′ identifies Enbilulu as Adad of Babylon. Nonetheless, Enbilulu is also known as one of the Fifty Names of Marduk in Enūma Eliš (Talon 2005:72, ll. 57–60). 55 George 1993:126, No. 800. 56 Ibid. No. 799.

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions …

249

iii 35–36. Although these lines are separated by a ruling from the previous section, we take lines 35–36 as a part of previous section because lines 32–34 lack a verb. v 19. We restore here MAŠ.MAŠ ‘incantation-priest,’ based on what Pinches copied. Later in this text (vi 35–36), Agum claims that he exempted Qišat-M[arduk], an incantation-priest, from taxes. As we know from different ancient texts, incantation-priests were also involved in the purification of newly built temples and it is very likely that this section of the Agum-Kakrime inscription also refers to the purification rite carried out by an incantation-priest prior to the re-installation of the cult statues of Marduk and Zarpanitu.57 The seventh century Assyrian king Sennacherib states that he sent an āšipu-priest and a kalû-priest to carry out a purification ritual for a newly dug canal. Furthermore, he claims that he dedicated offerings (such as carnelian, lapis-lazuli, muššaru-stone, ¶ulālu-stone, pappardilûstone, gold, fish, herbs, and probably silver as well) to Ea and Enbilulu.58 These precious metals and precious/semi-precious stones parallel the gifts given to the sanctuary of Marduk by Agum in ii 36–45. Nebuchadrezzar I also claims in his inscription that he and his subjects offered lavish offerings when he brought back Marduk from exile in Elam.59 v 39. It seems that this section refers to the entrance of the cult-statues of Marduk and Zarpanitu to their cella, Kasulimma of Esagila, when complete. See commentary on col. v line 19 above. vi 4. Judging from The Agum-Kakrime inscription, Edadihegal was located inside of Esagila and was not another name for Akītu-temple or more precisely Esiskur. The offerings that were brought to deities during the Akītu-festivals are better known but we also know from cuneiform texts that the ancients used to bring offerings on other occasions. For example, different cuneiform inscriptions suggest that the ancients brought gifts to the Esagila during/after the cleansing rituals for their sins in order to thank Marduk. Probably the best example is Ludlul IV (ll. 76–100), 57

Cf. Walker–Dick 2001:76f., 60b–65a and Ambos 2004:59. Luckenbill 1924:81, 27–30. 59 4R 2, pl. 20, No. 1, ll. 21–35. This inscription is also only known from a late copy found in the library of Assurbanipal. For the edition, see Frame 1995:30. 58

250

Articles: Ancient Near Eastern Studies

wherein the narrator approaches Esagila with prayers and offerings to thank Marduk and Zarpanitu for his salvation from hardship. Note also that a prayer to Marduk of more than 200 lines in length, which was published as The Prayer to Marduk No. 1, in Lambert 1959–1960 promises offerings in return for Marduk’s forgiveness to the sufferer and his salvation (159–166): 159. mu-¶ir kàd-ra-šú le-qí pi-de-e-šu ina qaq-qar šul-mì mah-ra-WkaX lit-tal-lak ḪÉ.NUN ¶i-i´-ba pa-rak-k[a-k]a li-šá-Waz X- nin bi-tuš-ka za-na-na-[š]u lu ka-a-a-WanX šam-na ši-ga-ri-ka GIM A.MEŠ li-šar-mì-ik ì-¶é-nun-na-ku li-¢a¶-¶i-da sip-pi-ka 165. li-šá-a´-li-ka za-’i-i e-re-ni ar-man-ni ri-šá-a-ti ¢u-pu-uš áš-na-an 159. ‘Accept his offering, take his present, On the soil of well-being, in front of you, may he walk. May he rain down abundant agricultural products on your dais, Let his provision of foods be done regularly in your temple. May he drench your bolt bar with oil like water, May he drip precious oil on your door-jambs plentifully. 165. May he burn for you cedar resin, Apricots of celebrations abundance of grain.’60

Another prayer to Marduk, KAR 25 ii 3–26 (Syncretistic IncantationPrayer to Marduk) also refers to offerings to Marduk (17–22): 017. šu-un-du-ul-tú KI-tim BARÁ.MEŠ-Wku X- nu be-WluX šá ina É uduSISKUR.MEŠ šá-[ru]-Wú/tumX u´-bu-[ni] f al-mat-tu ina ZÌ..GÁ šá-ru-[ú] Wi X- na? UDU SU i-qar-ru-b[u]-ku-nu-ši ina UGU ak-li ù A.MEŠ šá pa-li-¶i-ku-nu 022. gu-šá-a-ni 017. ‘The outspread earth is your dais’, Lor[d]?, Where, in the temple, the rich offerings are increase[d]. A widow with madga-flour and a ric[h (man) wi]th sheep Are approaching you (for offering). On the bread and the water of the one who fears you, 022. Come quickly here!’

Very interestingly, a Šuila-prayer to Madanu refers to the god as pa-qid éda-di-¶é-gál [IG]I.SÁ-e ‘the one who takes care of Edadihegal of gifts’ (Mayer 1976:469, l. 7). The god Madanu is listed as the throne-bearer

60

For further discussion on the subject of offerings to a god in the process of remission of sins, see T. Oshima’s forthcoming Babylonian Prayers to Marduk.

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions …

251

(guzalû) of Marduk in An-Anum II 253,61 and his temple stood adjacent to Esagila.62 In addition to the Agum-Kakrime inscription, Edadihegal is known from several cuneiform sources, for further references see George 1993:73, No. 135. é-è[š?-: Pinches copied É AN but the traces of the second sign clearly indicate that it is AB rather than AN. There are several temples and sanctuaries that were located in/adjacent to Esagila and begin with é-èš or èš: èšbàn-da (the seat of Nanše at Kamah of Esagila);63 èš-ma¶ (the seat of Ea at Kamah of Esagila);64 èš-šà-abzu (a chamber in Esagila).65 In addition, one may think of èš-gú-zi, which is known as another name for Esagila in The E-sagila Commentary.66 It is also possible to read, é-a[p-si], which could be the temple and the cosmic domain of Ea (a variant of èš-abzu).67 vi 40. Foster has offered an emendation and read a-na-, see Foster 1993:277. This reading was also followed by Stein 2000:162, note 149. Here, on the other hand, we take it as the beginning of the epilogue, probably a prayer for the sake of Agum.68 Although a prayer as an epilogue of a building inscription is better known from the NeoBabylonian period,69 requests for blessings for the royal construction works are attested already in the Old Babylonian period. See, for example, Frayne 1990, E.4.3.7.7, ll. 128–138 (Samsu-iluna).70 This also explains the shift of the mode from the first person to the third person in the following section.71

61

Litke 1998:97. George 1992:90–91. 63 George 1993:83, No. 265. 64 Ibid. 85, No. 282. 65 Ibid. No. 287. 66 George 1992:80, 31–34. 67 George 1993:65, Nos. 30 and 83, No. 264. 68 Jartiz 1958:229. 69 For this, see Shao 1989. 70 Normally Old-Babylonian royal inscriptions give statements that gods granted blessings to the kings for their works. 71 Note that while the account on the construction works of Ebabbar is told in the first person, the subject of the sentences in the epilogue shifts to the third person in an inscription of Samsu-iluna (Frayne 1990, E.4.3.7.3, ll. 107–123). 62

252

Articles: Ancient Near Eastern Studies

viii 1. The last sign is QA as copied by Pinches in 5R 33. As no combination of EL and QA yields a good sense, we offer an emendation and read el-lam based on i 3 and 20 (NUMUN el-lum). viii 16. Ea as the lord of the deep spring, see Tallqvist 1938:50. The Inscription of Kurigalzu the son of Kadashmanharbe This inscription of Kurigalzu, the son of Kadashmanharbe,72 is known from two exemplars, a clay cylinder, BM 108982 (CT 36, pls. 6–7) and NBC 2503 (BIN 2, No. 33). It is very difficult to date the former to a specific period on the basis of the script. Given the sign forms, in particular RU in col. ii 17 and 25, we believe that it is a copy from the Neo/Late-Babylonia period, however, there is no conclusive point that excludes a Middle Babylonian date with certainty. Judging from the fact that phrases in NBC 2503 are separated by two Winckelhacken, unlike the London manuscript, the former is clearly a later copy. There are a number of variations between the two and Drower (1973:466) implied that these bear different inscriptions. Yet, these variants are mostly orthographical differences, i. e., a Sumerogram vs. Akkadian phonetic writing, different signs for the same syllable. Judging from the contents, it is clear that these are duplicates.73 This text concerns a royal grant to Ishtar in Uruk. Kurigalzu devoted a large section of land to her temple. In addition, he also established a large number of offerings (bread, fine-beer, dates, sesame oil, and cakeingredients) as a daily offering to the goddess.74 Like the previous text of Agum-Kakrime, modern scholars have challenged its authenticity. Specifically, the large amount of daily ration dues

72

There is confusion about the identity of Kurigalzu, or identities of Kurigalzus. As Brinkman convincingly argued it seems that there were two different Kassite kings named Kurigalzu and this Kurigalzu the son of Kadashmanharbe was Kurigalzu I who lived around 1400 B. C. For discussion, see Brinkman 1969: 320–327; id. 1976:205; and Pruzsinszky 2009:117. 73 Brinkman 1976:209. 74 Foster implies that it is an indication of “a later forgery” (Foster 2005:365). However, there are eleven bricks commemorating a certain Kurigalzu’s building activities of Eanna, Anu and Ishtar’s temple in Uruk. See Clayden 1996:119. Given the reference of offerings to Marduk and Zarpanitu in the Agum-Kakrime Inscription, one may consider a possibility that the gifts recorded in the present inscription were to celebrate Kurigalzu’s work on Eanna.

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions …

253

and the vast lands granted to the temple often cast doubt on the authenticity of the inscription and the claim for the gifts made by Kurigalzu.75 There are three other texts referring to Kurigalzu, the son of Kadashmanharbe: baked clay cones from the time of Kadashman-Enlil (BM 91036 = BBSt, No. 1 // BM 135743 i 4–6) and a legal text from the reign of Nazi-Maruttash (CBS 12914 = BE XIV 39, 8).76 Below is the edition of the inscription of Kurigalzu the son of Kadashmanharbe. The variants are noted as footnotes. Manuscripts MS A: BM 108982 (CT 36, pls. 6–7: photos, pp. 262–264) MS B: NBC 2503 (BIN 2, No. 33) Edition Ungnad 1923:29–36. BIN 2, 50–51. Translations Foster 2005:365–366. Longman 1991:224–225. Reference Brinkman 1976: 209. Transliteration Col. i 1. mku-ri-gal-zu LUGAL GAL-ú 2. LUGAL dan-nu LUGAL KIŠ 3. mi-gir da-nim u d+en-líl 4. ni-bit EN DINGIR.MEŠ a-na-ku 5. LUGAL šá ina kal LUGAL.MEŠ AD.MEŠ-šú 6. GABA.RI NU TUKU-ú 7. IBILA šá mk[ad]-dáš-man-¶a[r]-be 8. LUGAL la šá-na-an 9. [m]u-šak-lil BÀD ki-Wsu? X- ri 10. [ga?]-mir é-kur-ra 11. [za]-[n]in URI2ki u UNUGki 12. [mu-ša]l-lim pa-ra-a´ NUNki 13. [e-p]iš é da-nim u dINNIN

75

Gelb 1949:348, n. 12; Longman 1991:90–91; Powell 1982:111–112. On the other hand, Brinkman (1976:209) suggests that there is no reason not to believe in its genuineness. 76 Brinkman 1976:147.

254

Articles: Ancient Near Eastern Studies 14. mu-šal-lim sat-tuk-ku WGAŠANX77 15. [šá DIN]GIR.MEŠ GAL.MEŠ 16. da-num AD WDINGIR X. MEŠ GAL.MEŠ78 17. ina ku-um-mi-šú ´i-i-ri 18. [ú-ša]r-ma-a šu-bat-su 19. ana diš-tar GAŠAN šur-bu-ti79 20. a-li-kàt i-di-ia80 21. mu-šal-li-mat ERIN2-ni-ia81 22. re-’i-a-ti UN.MEŠ-ia82 23. sa-ki-pat la ma-gi-ri-’-ia83 24. TA urua-da-at-ti84 25. šá GÚ ídBURANUNki 26. a-di-i uruman-gi-i´-´ì 27. ÚS.SA.RÁ A.GÀR 28. dur-an-ki na-ram den-líl 29. ul-tu URU GAŠAN-ia é-dGAŠAN-AMA-KAL-LA85 Col. ii 1. a-di-i BULUG URU gír-su 2. 60 ŠÁR86 šeNUMUN 10 IKU BANEŠ 3. DIŠ-KÙŠ GAL-t[i?] 4. ana diš-tar GAŠAN-ia ad-din 5. 3 GUR NINDA¶i.a 3 GUR KAŠ-SAG 6. 387 PI NINDA.Ì.DÉ.ÀM88 7. BANEŠ ZÚ.LUM.MA.Ì.TUKUki 8. BANEŠ Ì.GIŠ BÁRA.GA89 9. [š]á u4-mu 3 UDU.NÍTA.MEŠ 10. sat-tuk-ka-šá90 ú-kin91 kàl u4-mu 11. e-liš u šap-liš ú-kin ku-dúr-ri92

77

MS A i 14: [mu-šal-li]m SÁ.DUG4. MS B 3′ omits GAL.MEŠ, instead, at-ti. 79 MS B 4′: šur-bu-tú. 80 MS B 4′: -ía. 81 MS B 4′: ERIN2->-ni-i[a]. 82 MS B 5′: -ía. 83 MS A 23: le-em-mi-ia. 84 MS B 6′ omits -da-. 85 MS B 8′: TA URU dGAŠAN É-dga-ma-ma-k[al]-W d+EN? a-di┐ mi-´i-ru gi-ir-s[i]ki. 86 MS B 9′: šá-a-ru. 87 MS B 10′: 2. 88 MS B 10′: mi-ir-si. 89 MS B 11′: har-´u. 90 MS A ii 10: SÁ.DUG4. 91 MS B 11′: ú-kin-nu. 92 MS B 12′: ú-kin-nu NÍG.DU. 78

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions … 12. ú-šal-lim ki-si-ru93 13. URU.MEŠ94 A.ŠÀ.MEŠ ÍD.MEŠ u na-ba-lu 14. šu-bat e-du-re-e-[š]u95 15. ana diš-tar GAŠAN-ia ad-din 16. a-a-ú EGIR-ú šá E11-ma 17. DÙ.MEŠ-ia ú-nak-ka-ru 18. qí-bit DUG4.GA ú-bal-ú 19. ku-dúr-ri-ia ú-na-sa-¶u 20. ki-sur-re-e-ti ú-sa¶-¶u-ú 21. URU.MEŠ A.ŠÀ.MEŠ ÍD u na-ba-lu 22. šu-bat e-du-re-e 23. li-mi-ti URU UNUGki 24. i-tab-ba-lu ú-šat-ba-lu 25. lu-u ana NAM ú-tar-ru 26. diš-tar GAŠAN šur-bu-ti 27. ina ŠEN u MÈ Á-a-šú a-a DU-ik 28. a-bi-ik-ti BAD5.BAD5 29. KI.SU.LU.ŠÈ.NÍG.ME-šú GAR-ma 30. UKKIN-šú li-par-ri-ir

Translation Col. i 1–18 Kurigalzu, great king, strong king, king of all, the favourite of Anu and Enlil, nominated (to the kingship) by the lord of the gods, am I. The king who has no rival among all the kings, his fathers. The first born of K[a]dashman[h]arbe, king with no equal. [The one who] completed (the work) of the wall of the boundary (of Uruk?), [the one who finis]hed (the work) of Ekur, [the provid]er of Ur and Uruk, t[he one who ke]eps the sanctuary in Eridu in order, [the one who b]uilt the temple of Anu and Ishtar, [the one who ma]intains the regular offerings [of the] great [g]ods. Anu, the father of the great gods, made his dwelling place established in his magnificent sanctuary. Col. i 19 – ii 4 To Ishtar, the lady of greatness, the one who walks at my side, the one who protects my men, the one who leads my people, the one who overthrows those who do not obey me, I dedicated an area of 216,000 barrels using a ratio per surface unit of 1 keg of seed barley, measured by the large cubit, from Adatti, which is located on the bank of the Euphrates, as far as Magi´´u, on the border of the meadow, Duranki, the beloved (place) of Enlil, from the city of my MS A 12: ┌ki-sur-ri┐. MS B 12′: 10 URU.ME. 95 MS B 13′: e-zu-re-šú. 93

94

255

256

Articles: Ancient Near Eastern Studies lady, Bit-Gashan-ama-kala (The House of the Mother of All), as far as the border of Girsu, to Ishtar. Col. ii 5–10 I have established 3 barrels of bread, 3 barrels of fine beer, 3 smallbarrels of cake-ingredients (dates, sesame, oil etc.), a keg of dates from Dilmun, a keg of refined sesame-oil, 3 sheep per day, as her regular offering (for) each day. Col. ii 11–15 I established the border-stones all over (lit. above and below). I maintained the boundary. I dedicated the cities, fields, watercourses, and the dry lands, the rural settlements to Ishtar. Col. ii 16–30 Whoever ascends (to throne) after me; (and either) shall change my deeds, shall alter the commands of my mouth (lit. speech), shall pull out my boundary stones, shall change the boundaries, the cities, fields, watercourses, and the dry lands, the rural settlements, shall remove or shall allow someone to remove the enclosing wall of the city of Uruk, or shall convert (them) to the state lands; Ishtar, the lady of greatness, shall walk at his side neither in a battle nor in a war; she shall inflict (heavy) loss and defeat on his army, (and) she shall scatter his assembly.

Commentary i 9. As col. ii 23 refers to the enclosing wall of Uruk, one may wonder if this wall of boundary also refers to that of Uruk. Given the fact that this text concerns Kurigalzu’s donation to Ishtar, this seems to be very reasonable. i 10. There are several stamped bricks bearing a Sumerian inscription commemorating the building activities of a certain Kurigalzu. See Clayden 1996:117–118. In addition, a year name of Kurigalzu I also refers to a construction of Ekurigibarra (i. e., Ekur).96 i 16. For the collation, see Brinkman 1976:209, n. 14. i 28. For Duranki as Nippur, see Nashef 1982:88–89.

96

Brinkman 1976:402. For Ekurigibarra, see George 1993:117, No. 682.

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions …

257

i 29. MS B preserves É-dga-ma-ma-k[al]-Wd+EN?X (to be read k[al]-WlaX?). It must be an erroneous syllabic writing of é-gašan-ama-kala. ii 5. Our translation ‘barrel’ for GUR (= kurru ‘Kor’) is meant to convey the sense of a large amount. 1 kurru was 300 qû, and 1 qû of the Middle Babylonian period was about 1 liter. In other words, Kurigalzu claims that he daily offered 900 Kor of bread and 900 Kor of fine-beer to the temple of Ishtar in Uruk. These numbers alone well demonstrate the wealth of Babylonia under Kassite rule if this inscription is authentic. For the measurement units of the Middle-Babylonian period, see Sassmannshausen 2001:448–449. ii 7. BANEŠ = ´imdu = ca. 25 liters, see RlA 7:522–525.

258

Articles: Ancient Near Eastern Studies

Pl. 1. Agum-Kakrime MS A: K. 4149+ obv.

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions …

Pl. 2. Agum-Kakrime MS A: K. 4149+ rev.

259

260

Articles: Ancient Near Eastern Studies

Pl. 3. Agum-Kakrime MS B: Rm 505obv.

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions …

Pl. 4. Agum-Kakrime MS B: Rm 505rev.

261

262

Articles: Ancient Near Eastern Studies

Pl. 5. Kurigalzu MS A: BM 108982, pt. 1

Pl. 6. Kurigalzu MS A: BM 108982, pt. 2

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions …

Pl. 7. Kurigalzu MS A: BM 108982, pt. 3

Pl. 8. Kurigalzu MS A: BM 108982, pt. 4

263

264

Articles: Ancient Near Eastern Studies

Pl. 9. Kurigalzu MS A: BM 108982, pt. 5

Pl. 10. Kurigalzu MS A: BM 108982, pt. 6

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions …

265

References Albright 1952 Ambos 2004 Astour 1986 Balkan 1954 Boese 1982 Borger 1971 Brinkman 1969 Brinkman 1972 Brinkman 1976

Brinkman 1976–1980 Borger 1971 Campbell Thompson 1930 Charpin 1995 Clay 1906 Clay 1919 Clayden 1996 Dalley 1997 Diakonoff 1956 Drower 1973

Eder 2004 Foster 1993 Foster 2005

Albright, W. F. A Note on the Chronology of the Second Millennium B. C. BASOR 126:24–26. Ambos, C. Mesopotamische Baurituale aus dem 1. Jahrtausend v. Chr. Dresden. Astour, M. C. The Name of the Ninth Kassite Ruler. JAOS 106:327–331. Balkan, K. Kassitenstudien. 1. Die Sprache der Kassiten (AOS 37). New Haven. Boese, J. Burnaburiaš II., Melišipak und die mittelbabylonische Chronologie. UF 14:15–26. Borger, R. Gott Marduk und Goot-König Šulgi als Propheten. Zwei prophetische Texte. BiOr 28:3–24. Brinkman, J. A. Ur: “The Kassite Period and the Period of Assyrian Kings.” Or NS 38:310–348. Brinkman, J. A. Foreign Relations of Babylonian from 1600 to 625 B. C.: The Documentary Evidence. AJA 76:271–281. Brinkman, J. A. Materials and Studies for Kassite History. Vol. I. A Catalogue of Cuneiform Sources Pertaining to Specific Monarchs of the Kassite Dynasty. Chicago. Brinkman, J. A. Kassiten (kaššu). RlA 5:464–473. Borger, R. Gott Marduk und Gott-König Šulgi als Propheten: Zwei prophetische Texte. BiOr 28:3–24. Campbell Thompson, R. The Epic of Gilgamesh: Text, Translation, and Notes. Oxford. Charpin, D. À propos des rois de Hana. NABU 1995/23. Clay, A. T. Documents from the Temple Archives of Nippur Dated to the Reigns of Cassite Rulers (BE 14). Philadelphia. Clay, A. T. Empire of the Amorites (YOS 6). New Haven. Clayden, T. Kurigalzu I and the Restoration of Babylonia. Iraq 58:109–121. Dalley, S. Statues of Marduk and Enūma eliš. AoF 24: 163–171. Diakonoff, I. M. The History of Media. Moscow–Leningrad (Istorija Midii. Moskva–Leningrad). Drower, M. S. Chapter X. Syria c. 1500–1400 B. C. Edwards, I. E. S. et al. (eds.). The Cambridge Ancient History. Vol. II/1. History of the Middle East and Aegean Region c. 1800–1380 B. C. Cambridge. Pp. 417–525. Eder, Ch. Assyrische Distanzangaben und die absolute Chronologie Vorderasiens. AfO 31:191–236. Foster, B. R. Before the Muses: An Anthology of Akkadian Literature. Vol. I. Archaic, Classical, Mature. Bethesda. Foster, B. R. Before the Muses: An Anthology of Akkadian Literature. 3rd ed. Bethesda.

266

Articles: Ancient Near Eastern Studies

Frame 1995

Frayne 1990 Gadd 1973

Gasche et al. 1998 Gelb 1949 George 1992 George 1993 Goetze 1964 Grayson 1969

Grayson 1974–1977 Grayson 1975a Grayson 1975b Grayson 1980 Grayson 1980–1983 Grayson 1987

Grayson 1991 Groneberg 1980 Hawkins 1972–1975 Horowitz 1998 Horowitz–Oshima 2006 Horsnell 1999

Frame, G. Rulers of Babylonia: From the Second Dynasty of Isin to the End of Assyrian Domination (1157–612 B. C.) (RIMB 2). Toronto–Buffalo–London. Frayne, D. Old Babylonian Period (2003–1595 B. C.) (RIME 4). Toronto–Buffalo–London. Gadd, C. J. Chapter V. Hammurabi and the End of His Dynasty. Edwards, I. E. S. et al. (eds.). The Cambridge Ancient History. Vol. II/1. History of the Middle East and Aegean Region c. 1800–1380 B. C. Cambridge. Pp. 176–227. Gasche, H. et al. Dating the Fall of Babylon: A Reappraisal of Second-Millennium Chronology. (MHEM IV). Ghent–Chicago. Gelb, I. J. The Date of the Cruciform Monument of Maništušu. JNES 8:346–348. George, A. R. Babylonian Topographical Texts (OLA 40). Leuven. George, A. R. House Most High. Winona Lake. Goetze, A. The Kassites and Near Eastern Chronology. JCS 18:97–119. Grayson, A. K. Assyrian and Babylonian King Lists: Collations and Comments. Röllig, W. (ed.). lišān mit¶urti, Festschrift Wolfram Freiherr von Soden zum 19.VI.1968 gewidmet von Schülern und Mitarbeitern (AOAT 1). Kevelaer– Neukichen-Vluyn. Pp. 105–118. Grayson, A. K. The Empire of Sargon of Akkad. AfO 25: 56–64. Grayson, A. K. Assyrian and Babylonian Chronicles (TCS V). Locust Valley. Grayson, A. K. Babylonian Historical-Literary Texts. Toronto–Buffalo. Grayson, A. K. Assyria and Babylonia. Or NS 49:140–194. Grayson, A. K. Königslisten und Chroniken: Akkadisch. RlA 6:86–135. Grayson, A. K. Assyrian Rulers of the Third and Second Millennia B. C. I (to 1115 B. C.) (RIMA 1). Toronto–Buffalo– London. Grayson, A. K. Assyrian Rulers of the Early First Millennium B. C. I (1114–859 B. C.) (RIMA 2). Toronto–Buffalo–London. Groneberg, B. Die Orts- und Gewässernamen der altbabylonischen Zeit (RGTC III). Wiesbaden. Hawkins, J. D. Úatti: The 1st Millennium B. C. RlA 4: 152–159. Horowitz, W. Mesopotamian Cosmic Geography. Winona Lake. Horowitz, W.; Oshima, T. Cuneiform in Canaan. Jerusalem. Horsnell, M. J. The Year-Names of the First Dynasty of Babylon. Vol. I. Chronological Materials, the Year-Name System and the Date-Lists. Vol. II. The Year-Names Reconstructed and Critically Annotated in the Light of Their Exemplars. Hamilton.

T. Oshima, Another Attempt at Two Kassite Royal Inscriptions … Jacobsen 1934 Jaritz 1957 Jaritz 1958 Jaritz 1960 Jensen 1892

King 1912 Lambert 1959–1960 Lambert 1985

Lambert 1987

Landsberger 1928–1929 Landsberger 1954 Lenzi 2008 Litke 1998 Livingstone 1989 Longman 1991 Luckenbill 1924 Mayer 1976 Nashef 1982 Oppenheim et al. 1970

Oshima 2006 Pinches 1909 Podany 1991–1993

267

Jacobsen, Th. Philological Notes on Eshnunna and Its Inscriptions (AS 6). Chicago. Jaritz, K. Die kassitischen Sprachreste. Anthropos 52: 850–898. Jaritz, K. Quellen zur Geschichte der Kaššu-Dynastie. MIO 6:187–265. Jaritz, K. Die Kuturreste der Kassiten. Anthropos 55:17–84. Jensen, P. Inschrift Agum-kakrimi’s, d. i. Agum’s des Jüngeren. Abel, L. et al. (eds.). Sammlung von assyrischen und babylonischen Texten in Umschrift und Übersetzung (KB III/1). Leipzig. King, L. W. Babylonian Boundary-Stones and Memorial Tablets in the British Museum. London. Lambert, W. G. Three Literary Prayers of the Babylonians. AfO 19:46–66. Lambert, W. G. The History of the muš-¶uš in Ancient Mesopotamia. Borgeaud, Ph. (ed.). L’animal, l’homme, le dieu dans le Proche-Orient ancient. Geneva. Pp. 87–94. Lambert, W. G. A Further Attempt at the Babylonian “Man and His God.” Rochberg-Halton, F. (ed.). Language, Literature, and History: Philological and Historical Studies Presented to Erica Reiner (AOS 67). New Haven. Pp. 187–202. Landsberger, B. Das “gute Wort.” MAOG 4:294–321. Landsberger, B. Assyriche Königliste und “Dunkels Zeitalter”. JCS 8:31–45, 47–73, 106–133. Lenzi, A. Secrecy and The Gods: Secret Knowledge in Ancient Mesopotamia and Biblical Israel (SAAS 19). Helsinki. Litke, R. L. A Reconstruction of the Assyro-Babylonian GodLists, AN : dA-NU-UM and AN : ANU ŠÁ AMĒLI. New Haven. Livingstone, A. Court Poetry and Literary Miscellanea (SAA 3). Helsinki. Longman, T. Fictional Akkadian Autobiography: A Generic and Comparative Study. Winona Lake. Luckenbill, D. D. The Annals of Sennacherib (OIP 2). Chicago. Mayer, W. R. Untersuchungen zur Formensprache der babylonischen “Gebetbeschwörungen” (StP SM 5). Rome. Nashef, Kh. Die Orts- und Gewässernamen der mittelbabylonischen und mittelassyrischen Zeit (RGTC V). Wiesbaden. Oppenheim, A. L. et al. Glass and Glassmaking in Ancient Mesopotamia: An Edition of the Cuneiform Texts which Contain Instructions for Glassmakers with a Catalogue of Surviving Objects. London–Toronto. Oshima, T. Marduk, the Canal Digger. JANES 30:77–88. Pinches, Th. G. A Selection from the Miscellaneous Inscriptions of Assyria and Babylonia. London. Podany, A. H. A Middle Babylonian Date for the Hana Kingdom. JCS 43–45:53–62.

268

Articles: Ancient Near Eastern Studies

Powell 1982 Pruzsinszky 2009

Richardson 2002

Rowton 1958 Sassmannshausen 2001

Sassmannshausen 2004

Shao 1989 Sommerfeld 1982

Stein 2000

Tallqvist 1938

Talon 2005 Ungnad 1923 Walker–Dick 2001

Weidner 1921 Weidner 1926 Weidner 1952–1953 Weidner 1959–1960 Wetzel et al. 1957 Zadok 1988

Powell, M. A. Metrological Notes on the Esagila Tablet and Related Matters. ZA 72:106–123. Pruzsinszky, R. Mesopotamian Chronology of the 2nd Millennium B. C.: An Introduction to the Textual Evidence and Related Chronological Issues. Wien. Richardson, S. F. C. The Collapse of a Complex State: A Reappraisal of the End of the First Dynasty of Babyon, 1683– 1597 B. C. PhD. Diss. Columbia University. Rowton, M. B. The Date of Hammurabi. JNES 17:97–111. Sassmannshausen, L. Beiträge zur Verwaltung und Gesellschaft Babylonians in der Kassitenzeit (BaF 21). Mainz am Rhein. Sassmannshausen, L. Babylonian Chronology of the 2nd Half of the 2nd Millennium B. C. Hunger, H.; Pruzsinszky, R. (eds.). Mesopotamian Dark Age Revisited: Proceedings of an International Conference of SCIEM 2000. Wien. Pp. 61–70. Shao, J. T. A Study of Akkadian Royal Hymns and Prayers. PhD. Diss. Ann Arbor. Sommerfeld, W. Der Aufstieg Marduks: Die Stellung Marduks in der babylonischen Religion des zweiten Jahrtausands v. Chr. (AOAT 213). Kevelaer–Neukirchen-Vluyn. Stein, P. Die mittel- und neubabylonischen Königsinschriften bis zum Ende der Assyrerherrschaft: Grammatische Untersuchungen (JBVO 3). Wiesbaden. Tallqvist, K. Akkadische Götterepitheta. Mit einem Götterverzeichnis und einer Liste der prädikativen Elemente der sumerischen Götternamen (StOr 7). Helsingforsiae. Talon, Ph. The Standard Babylonian Creation Myth Enūma Eliš (SAACT 4). Helsinki. Ungnad, A. Schenkungsurkunde des Kurigalzu mâr Kadašman-Harbe. AfK 1:19–23. Walker, C.; Dick, M. The Induction of the Cult Image in Ancient Mesopotamia: The Mesopotamian Mīs Pî Ritual (SAALT 1). Helsinki. Weidner, E. F. Die Könige von Assyrien: Neue chronologische Dokumente aus Assur (MVAG 26/2). Leipzig. Weidner, E. F. Die grosse Königsliste aus Assur. AfO 3: 67–77. Weidner, E. F. Das Reich Sargons von Akkad. AfO 16: 1–24. Weidner, E. F. Die älteren Kassiten-Könige. AfO 19:138. Wetzel, F. et al. Das Babylon der Späterzeit (WVDOG 62). Berlin. Zadok, R. Geographical Names According to New- and Late Babylonian Texts (RGTC VIII). Wiesbaden.

Sumerian: A Uralic Language. II∗ Simo Parpola University of Helsinki

Appendix 1. Sumerian‒Ural-Altaic etymologies in Lenormant 1875 Sumerian a ‘eau’ ad, adda ‘père’

Ural-Altaic Fin. jä, Esth. ea, jea, Mordv. je, Tchér. ej ‘glace’ Lap. attje, Vot. ataj, Tchér. ätjä, Ost. ata, Mag. atya, Turc. ata

a¶an-du ‘vomir’ ai ‘père’

Fin. oksenn-, Ost. aχt-, Mag. okád- ‘vomir’ Fin. äijä ‘vieillard, grand-père,’ Lap. aija, Zyr. aj ‘père,’ Vog. jaj

an ‘ciel’ bar ‘moitié’

Zyr. Perm. jen ‘dieu,’ Votj. in ‘ciel,’ in-mar ‘dieu’ Fin. puoli, Mag. fel ‘moitie’

dingir ‘dieu’ gál ‘habiter’

Turc. tänri, Mongol tagri ‘dieu’ Zyr. ol, Ost. val ‘habiter’

ge6 ‘nuit’ ¶a, ku6 ‘poisson’

Fin. yö ‘nuit,’ Esth. ö, Zyr. voj, Votj. üj, Md. vä, Tchér. ju-t Fin. kala, Lap. guolle, Vot. kala, Mordv. kal, Tchér. kol, Ost. χut’, Vog. kul, χul, Mag. hal, Sam. Your. hālea, Kam. kola, Mongol kal Fin. kusi, Ost. kos etc. ‘urine’

kàš ‘urine’ kiši7–9 ‘fourmi’

Fin. kusi-ainen ‘fourmi,’ Lap. got-kka, Zyr. kot-koźu, Perm. köźil, Votj. kuz-i-lï, Tchér. kut-ka, Samoy. Ost. keča, keca

kul ‘semence’ kur ‘montagne’

Fin. kylv- ‘semer,’ Esth. külv-, Liv. küll-, Lap. gilv-, Ost. kirTchér. korok, Vot. gurez, Perm. kerös ‘montagne’

kur ‘s’élever’ ma ‘pays’

Fin. korkea ‘élevé,’ korko ‘élévation,’ Veps. korged ‘élevé,’ Esth. körge ‘haut,’ Vot. kôrkea ‘élevé,’ Liv. kuord ‘haut,’ Ost. kereš ‘haut’ Fin. maa etc. ‘pays’

nim ‘s’élever’ sug ‘marais’

Vog. num-an ‘élevé’ Fin. suo, Esth. sō, Liv. súo ‘marais,’ Turc. su, Tchouv. šu ‘eau’

šurin ‘grillon’ ur ‘mâle’

Fin. sir-kka, Esth. sir-e, Perm. čir-k ‘grillon’ Mag. ur ‘monsieur, seigneur’

úr ‘racine’ urud ‘cuivre’

Fin. jūri, Veps., Esth., Liv. jūr, Mordv. Ost. jor ‘racine’ Fin. rauta, Esth. raud, Lap. route ‘fer’

∗ The first part of this paper was published in B&B 4, Vol. I/2 (2010):181– 210. For the abbreviations, see ibid., first footnote on p. 181 and pp. 198–199. For references, see ibid., pp. 200–209.

270

Articles: Ancient Near Eastern Studies

Appendix 2. Selected Sumerian—Uralic etymologies proposed after 1950 Sumerian ab(.ba) ‘father, elder’ a.gàr ‘meadow’ ág ‘to measure (out)’ àš ‘six’ bàd ‘wall; dam’ bàn(.da) ‘child; wild’

Finnish appi ‘father-in-law’

biz ‘to trickle, drip’

pato ‘dam’ peni, pentu ‘puppy, brat’ pisara ‘drop’

bu7 ‘to winnow’ bur ‘meal, food’

puhu- ‘to blow, speak’ puuro ‘porridge’

búr ‘to release, undo’

purka- ‘to dismantle’

bùru ‘to drill, perforate’ buru4 ‘crow’ dag ‘dwelling’ diri(g) ‘(to be) excessive’ dun ‘to throng, push in’ e ‘this’ eme ‘tongue’ ème ‘female donkey’ en ‘high priest, lord’ gan ‘to bear’ géme ‘slave girl’ gir5, gir7 ‘to run quickly’ gùd ‘nest, home’ gul ‘to destroy; perish’ gúr ‘disc, hoop’ gur7 ‘harvest’ gen, mèn ‘to go’ ¶áb ‘malodorous’

Hungarian apa ‘father’ ugar ‘fallow’ ad ‘to give’ hat ‘six’ fal ‘wall’ fene ‘wild’

fúj ‘to blow’

fúr ‘to drill’ varis ‘crow’ lakka ‘attic, loft’ tyrty- ‘have too much of ’ tunke- ‘to crowd, push’

emä ‘dam, mother’

varju id. lak ‘dwelling’

Gostony 1975 Chong 2003 Chong 2003

dug ‘to push’

Hakola–Assadian 2003 Gostony 1975 Gostony 1975 Bobula 1951 Gostony 1975 Hakola–Assadian 2003 Gostony 1975 Hakola–Assadian 2003 Hakola–Assadian 2003 Chong 2003 Gostony 1975 Gostony 1975 Hakola–Assadian 2003 Hakola–Assadian 2003

e (ez) ‘this’ nyelv id. eme ‘female’ vén ‘elder’

kanta- ‘to bear’ gim ‘female’ kiri- ‘to spurt’ kota ‘teepee,’ koti ‘home’ kuol- ‘to die’

men- ‘to go’ hapan ‘sour, acid, rotten’

Reference Gostony 1975 Gostony 1975 Gostony 1975 Gostony 1975 Chong 2003 Hakola–Assadian 2003 Hakola–Assadian 2003 Bobula 1951 Hakola–Assadian 2003 Hakola–Assadian 2003 Bobula 1951

ház ‘house’ hal ‘to die’ kör ‘circle’ góré ‘silo’ men- id.

S. Parpola, Sumerian: A Uralic Language. II Sumerian ¶ar ‘to scratch’ ì ‘oil’ im ‘wind, rain’ kàd ‘to knot, weave’ kar ‘to flee away’ kiš(i)4 ‘half’ kúm ‘hot; fever’ kur4 ‘sick’ líl ‘air, breeze, ghost’ lúgud ‘short’ me ‘we’ me- ‘what?’ mes ‘young man, man’

Finnish harat- ‘to drag, harrow’ voi ‘butter, (K) oil’ ilma ‘(bad) weather’ kuto- ‘to weave’ karat- ‘to run away’ keski ‘middle’ kuuma ‘hot,’ -e ‘fever’ kurja ‘miserable, poor’ löyly ‘heated air’ lyhyt ‘short’ me ‘we’ mi- ‘what?’ mies ‘man’

Hungarian

vaj ‘butter’ köt id.

köz ‘interval’ hév ‘heat’

lélek ‘ghost’

mi ‘we’ mi ‘what?’

271 Reference Hakola–Assadian 2003 Gostony 1975 Hámori 2005 Gostony 1975 Hakola–Assadian 2003 Gostony 1975 Chong 2003 Hakola–Assadian 2003 Bobula 1951 Hakola–Assadian 2003 Gostony 1975 Chong 2003 Chong 2003

Appendix 3. Distinctive features of Sumerian Cf. B. Comrie, “General features of the Uralic languages,” in Sinor 1988: 451–477; Klesment et al. 2003. Features shared by Sumerian and Uralic are indicated by [=]. 1. Phonology 1.1. Vowel harmony (Poebel 1931; Keetman 2005) [=] 1.2. Vowel system with at least 8 and possibly 9 vowel qualities, probably including the front-rounded vowels [œ] and [y] (Bobrova–Militarev 1989; Keetman 2005:16; cf. Smith 2007) [=]1 1.3. Binary quantitative opposition of vowels (Edzard 2003:13–14) [=] 1 The presence of front rounded vowels is a highly distinctive phonological feature. In the sample of 562 languages surveyed in Maddieson 2008, there are only 37 languages (about 6.6%) in which one or more such vowels are recognized as part of the vowel inventory. The majority of these (29 of the 37) are found in the more northerly parts of the Eurasian land mass. Outside of this area there are only a few widely scattered languages reported with front rounded vowels, and the reports are not always sufficiently detailed to be relied on with confidence. High and mid front rounded vowels ([œ] and [y]) are found in 23 of the surveyed languages, including Finnish, Cheremis, Hungarian, Tawgi, Chuvash, Bashkir, Turkish, Kirgish, Tuvan, Dagur, Manchu, Azerbaijani, Tibetan, Yakut, Norwegian, German, French, and Breton.

272

Articles: Ancient Near Eastern Studies

1.4. Quantitative consonant opposition [=]2 1.5. Voice phonemically non-distinctive (Thomsen 1984, § 20) [=] 1.6. No word-initial consonant clusters (Edzard 2003:22) [=] 1.7. Consonant lenition (‘gradation’) at morpheme juncture (cf. provisionally Bauer 1975) [=] 1.8. Stress on the first full vowel of the stem (see, provisionally, Krecher 1969 and the examples in Parpola 1975:256, 1.21; cf. Thomsen 1984, § 10). In most Uralic languages (Finnic, Lappish, Mordvin, Ziryene, Ostyak, Vogul, Hungarian, Samoyed) the place of main stress is on the first syllable. However, there are many exceptions: – In Mordvin, stress can shift to the second syllable, if the first vowel is u, i or ȇ : 3 and a full vowel (usually a or ä, sometimes o, e) follows (Bartens 1999:28, 30) – In Cheremis, stress is on the last full vowel (E) or on the penultimate vowel (M), provided that this is not etymologically short and preceded by an etymologically long vowel (Alhoniemi 1985, § 3; Collinder 1957:251) – In Votyak, stress is usually on the last syllable (Collinder 1957:276) – In some dialects of Ziryene, the vowel a will attract the stress to the detriment of the high vowels (u, y, i) (Collinder 1957:300) – In the Tavda dialect of Vogul, stress is often on the second syllable (Collinder 1957:322) 1.9. Basic maximal syllable structure CVCC(V) in native words (cf. the list of CVC(V) and more complex words in Halloran 2006, version 3) [=] 2. Morphology 2.1. No article [=] 2.2. Use of postpositions instead of prepositions [=] 2.3. No grammatical gender [=] 2.4. Opposition animate : inanimate [=] 2.5. Extensive use of compounding in word formation [=] 2.6. Nominal and verbal categories largely expressed by means of synthesis [=] 2.7. Reduplication as morphological and lexical device to indicate 2.7.1. superlative (sag9-sag9 ‘very beautiful’ ~ M mazi-mazi id., U śe̮d-śe̮d ‘blacker than black’) 2

E. g., kar /kār/ → Akk. kāru ‘quay,’ as against gàr /karr/ → Akk. karru ‘knob.’

S. Parpola, Sumerian: A Uralic Language. II

273

2.7.2. plurality, inclusiveness (Thomsen 1984:61, e. g. kur-kur ‘all the mountain lands’ ~ F kukkura/kukkula ‘hill, peak; brim’) 2.7.3. iteration, repetition (Thomsen 1984:126, e. g. gíd-gíd ~ F pitkittä- ‘to extend’) 2.8. Onomatopoetic or expressivistic reduplication with vocalic alternations (Civil 1966, e. g. dubul-dabal ~ F liplatta- ‘to bubble, ripple’; si¶5-sa¶4 ‘confusion’ ~ F sekasotku id., sikin sokin ‘helterskelter’) 2.9. Nine productive grammatical cases (cf. Finnic 12–15, Lappish 8, Mordvin 12, Cheremis 7, Votyak 15, Ziryene 17, Ostyak 10, Vogul 7, Yurak 7, Selkup 12) 2.10. Use of possessive suffixes [=] 2.11. Conditional mood (Thomsen 1984:280; Edzard 2003) [=]3 2.12. Complex verbal prefix chains (see Part I, n. 29) 3. Syntax 3.1. Basic unmarked word order is SOV (Thomsen 1984, § 44) [=] 3.2. Attribute follows main word (see Part I, n. 29) 3.3. Adjective and main word form a noun phrase functioning like a compound word [=] – In Finnic, adjective attribute agrees with the main word in number and case – In all other Uralic languages, it is invariable and does not take case endings

3

The Sum. conditional particle gišen is equated with Akkadian irrealis suffix -man in OBGT I 460–462, NBGT I 460 and bilingual texts (BWL 244rev. iv 42– 45). Etymologically, it corresponds to the Ziryene conditional particle veśke̮, veśke̮n (L 12:323) and Finnish ehkä, ehken (NSS I 191a) ‘perhaps,’ which are possibly ossiefied 1st p. sg. prs. forms of the Uralic verb *kwe̮śk3- ‘to believe, trust’ (F usko-, L osko-, M vaśka-, C išane-, ušane-, U oskȋ-, Z eski-, O 3šȧn-, 3šǡn3-, H hisz-, cf. Sum. ešdé ‘to believe, entrust,’ giškim-ti ‘to trust,’ gišgal ‘stand, step,’ F uskalta- ‘to dare, trust,’ Z vośkolt- ‘to venture a step’). In finite verbal forms, gišen appears as an infix (-giš-) inserted, like the Uralic conditional infix *-kwśi- (F -isi-, K -iži-, E -ksi-, L -š-, -vč-, -pč-, -pc-, -M -iks(el’)-, U -sal-, Y -bs-, -bś-, -bt’-), between the verb and the personal suffix, cf. níg-tur sig7-a inim bí-in-dug4-ga 119 igi-dù-dù-a na-an-gig-gagiš-še-en 120 ki ad-da-gá ì-kin-gá-giš-še-en ‘I (-en) should not have a hard time in explaining the little things of the creation (story), the words said in it, if only I (-en) had frequented the place of my father!’; á mu-e-da-ág-giš-še-en 44 á-ág-gá ma-ab-sum-mu-un-e-še ‘Should I give you an order, you (shout): You are giving me orders!’ (Scribe and his perverse son, lines 118–120 and 43–44).

274

Articles: Ancient Near Eastern Studies

3.4. Comparison expressed syntactically by means of the ablative case (Balke 2006:133–134) [=]4 3.5. ‘to be’ instead of ‘to have’ (cf. Aalto 1999) [=]5 3.6. Extensive use of concatenation in syntax (“nominal chain”) 3.7. Split ergative construction of transitive verbs (see Part 1, n. 32) 3.8. Use of non-finite verbal constructions/nominalization as equivalents of subordinate clauses (Gragg 1973; Thomsen 1984, § 44; cf. Klesment et al. 2003:7) [=] 4. Lexicon 4.1. Copular verb -àm ‘is’ (Gragg 1968; Thomsen 1984:275–278; see Part I, n. 28) [=] 4.2. Negation nu-, na-, nam- (Thomsen 1984:190–195; Honti 1997) [=] – K ni-, C ńi, n3, nu, U no, Z ńe ‘not,’ ńem, ne̮m ‘nothing,’ O än ‘not,’ nèm ‘nothing,’ V nē-, nēm- ‘not,’ H nem ‘not,’ ne ‘don’t,’ Y ńo, ńon, Yn ńe, Km nåγ(å), nå ‘nicht’. Not in F, L, M. Appendix 4. Sumerian–(Ob-)Ugric isoglosses Cf. Honti 1979; Kálmán 1988:396–404 1. Phonology 1.1. Change k > χ in front of velar vowels (cf. Sum. gaz ~ ¶az ~ ¶aš ‘to break’; gul ~ ¶ul ‘to destroy’; gur8.sag [PSg] ~ ¶ur.sag ‘mountain (range), foothills,’ as against kur ‘mountain’; guz ~ ¶uz ‘to cut off,’ šerkan ~ šer¶an ‘ornament’) 1.2. Change č > š (see Part I, n. 27) 1.3. PU *kakta ~ *käktä ‘two’ > kat (Sum. kát-a-bi ‘secondly’ (NBGT IV 35) ~ V kit ‘two,’ O kät, kat, kĕt ‘two,’ kitpa ‘twice,’ H két ‘two’) 2. Morphology 2.1. Suffix order Px + Cx (see App. 6) 2.2. Lack of genitive ending -n (found in F, L, M, C and S; also lacking in U and Z)

4 E. g., abzu-ta ma¶ [Abzu-ABL. lofty] ‘loftier than Abzu’; cf. M t’el’e-ďe paro [winter-ABL. good] ‘better than winter’ (Bartens 1999:109); C ačat-δeč kuγu [father-ABL. great] ‘greater than the father’ (Alhoniemi 1985:79–80); Z ponj-įś įdžįd [dog-ABL. great] ‘bigger than a dog’ (L 12:370); V pūm-n3 teɛlk [grass-ABL. low] ‘lower than grass’ (UrLang 155). 5 E. g., Sum. an-da gál ‘he had/has,’ lit. ‘there is/was at/with him’ (Izi A iii 12′–13′) ~ F hänellä oli id.; mu-da gál ‘I have/had’ (OBGT III 59) ~ F m(in)ulla oli ‘I had.’

S. Parpola, Sumerian: A Uralic Language. II

275

2.3. Determinative suffix -pi used to indicate comparative, superlative and contrast (Sum. bar ‘side,’ bar-bi ‘extra, additional’; dùg-ga-bi ‘best’ (Gudea Cyl. A x 11); mìn ‘two,’ mìn-a-bi ‘both,’ cf. O num-pi ‘upper side; higher’; t’ŏm-pi ‘the other side’; äj-pi ‘the smallest,’ neγoi-pi ‘the whitest’; H alább ‘further down,’ nagyobb ‘bigger’ (also in F and Y) 2.4. Second person singular verbal suffix -en (Sum. -en ~ O, V -3n; also Z -n) 2.5. Infinitive ending *-tAk, *-δak (Sum. -V-da, -V-dè ~ O -taʿ, -tä, -Da, -tāγɜ, H -t; also in F and L) 2.6. Verbal prefixes (Sum. mu-, bí- ~ H meg- ‘back’ etc., be- ‘in’) 3. Syntax 3.1. Split ergativity (O, V) 3.2. Word order SOV 4. Lexicon and semantics – Invariable negative particle (see 4.2 above) – Possessive verb (tuk, tuku ~ O tŏj-, tăj-, taja- ‘to own, have’) – Sum. aj ‘water’ ~ O j3ŋk ‘water’ (cf. Part I, Table 7) – Sum. nar ‘musician’ ~ O nār3s-, ner- ‘to play a musical instrument’ (in other Uralic languages, nar- ~ nir- means ‘to rub’) – Sum. dnin-urta ‘divine prince and hero, son of Enlil, defender of celestial order, cosmic judge, holder of tablet of destinies’ ~ O wort, ort, urt ‘a Khanty prince, a god’s son … a hero warrior spirit, the defender of people, the upholder of justice’ (Kulemzin et al. 2006:144, 148; Paasonen–Donner 1926:2826; Karjalainen–Toivonen 1948:83) – Sum. duraš (= Ninurta, see An I 227) ~ O war3š ‘a high deity, who up in the air monitors people’s doings and reports them to the supreme god’ (Paasonen–Donner 1926:2901) – Sum. ég ‘small canal, ditch’ ~ O jēγ3̑, jįγį ‘small river,’ V jē, jā ‘river’ (as against F joki, M jov, C joγ3̑, U ju, Z ju(g), S jaχa, joha, PU *joke ‘river’) Appendix 5. Sumerian—Finnic isoglosses 1. Phonology 1.1. Change t, ć > s, ś – in final position: aš < *e̮ktä (*ikte, *ükti) ~ F yksi, Z e̮t, e̮ťi ‘one’; àš < *kwie̮ktä (*kutte) ~ F kuusi ‘six’; diš ‘perfect’ ~ F täysi, C tić ‘full’; èreš ‘sweat, resin, fragrance’ ~ F erite ‘secretion,’ O eräs ‘birch sap’; eš10 ~ F vesi ‘water, tear’; geš, kišib ~ F käsi, M keď ‘hand’; gúz ~ F köysi, L kievđa, Z gez, H gúzs ‘rope’; kuš ~ F kesi, M keď ‘skin’; é.meš ‘sum-

276

Articles: Ancient Near Eastern Studies

mer’ ~ F kevät, Veps. keväź ‘spring’; mùš ‘looks, appearance’ ~ F muoto ‘looks, appearance,’ muoti ‘model,’ muotti ‘form, shape’6; šeš < *ćečä ‘brother’ ~ F setä, seti ‘uncle’ – in middle position: bur.zid ‘bowl’ ~ F purtilo, pursilo ‘trough’; udu, ezi, izi ‘sheep’ ~ F utu, uti, ući ‘sheep, ewe’; esiš, isiš ~ F, E vetist-, vesist- ‘to weep’; gú.zal ‘scoundrel, rascal’ ~ F kutale ‘rascal, worthless person,’ kʿät̜’ʌ, kēt’l ‘worthless’; ¶enzer ‘weak, slight’ ~ F hintelä, hinterä ‘tiny, slender, weak’; šerzi ‘sunshine, radiance’ ~ F helle, helte‘hot weather, heat’; (-t- > ś ) ušum < *e̮ktim ~ F yksin ‘alone’; gašam ‘clever, skilled’ ~ F kaseva, kätevä ‘clever, skillful’ – in initial position: zi.da < *cistaj ‘true, proper, decent’ ~ F siisti/a ‘clean, proper, decent,’ M čistaj ‘clean,’ O s/tĭstam, H tiszta ‘clean, proper, blameless’;7 zizna, zazana ‘spawn, roe’ ~ F tahna ‘paste’; zí¶ ~ F siku ‘piglet,’ M tuva ‘pig’; za ~ F sä, M ton ‘you’; -zu ~ F s(in)un ‘your’; -zé.en ~ F -tte ‘you’ 1.2. Change č-, ć- > ¶- (F h-)8 – č/š > ¶: ¶áb ‘malodorous, sour’ ~ F hapan ‘sour,’ contrast M čapamo, C šopo ‘sour’; ¶al ‘to divide’~ F halko- ‘to split,’ M čalga- ‘to break’; ¶al ‘to meander’ ~ F häily-, M čara- ‘to waver’; ¶al ‘to pour out’ ~ F hol(j)ot-, O sŏʌ3ʌt3̑- ‘to pour out; ¶ar ‘crest’ ~ F harja ‘crest, mane,’ M čeŕ, C šar ‘mane’; ¶ar ‘to scratch’ ~ F haro- id., contrast M čirkaď-, šorkaď- id.; ¶u.ba ‘weak, feeble(-minded)’ ~ K hupa ‘weak, small,’ höpö ‘stupid,’ M čovalaj ‘simpleton’; ¶u.bak ~ F hupakko ‘fool’; ¶u× ‘to rest, take breath’ ~ F huoka- ‘to sigh, take a rest,’ L s/šuokke- ‘to sigh, rest,’ H sóhaj ‘sigh’; ¶um ‘sludge, slush’ ~ F hyhmä, M šušmo, čušma id. – ć/ś/s > ¶: ¶al ‘secret’ ~ F sala, M salava, Z će̮l ‘secret,’ H csalás ‘deception’; ¶al ‘legs’ ~ F haara, saara ‘branch, fork, (pl.) legs,’ L suorge ‘fork, branch,’ O śărĭ ‘hips’; ¶a.an ‘weak’ ~ F hieno, h́en ‘fine, thin, weak,’ H csün- ‘to pine away’; ¶e.en.zèr ‘weak, slight’ ~ F hinterä 6

The Finnish words also have Germanic etymologies, muoto being regarded as an early, perhaps even pre-Germanic loan (Kylstra et al. 1991:270–271). 7 Explained in SSA 3 177 as a loan from Russ. чиcтый (OCSl. čistъ ‘pure’; Derksen 2007, s. v. či°stú), but the wide distribution of the word in Uralic, the initial consonant in O and H, the final vowel of the Uralic words, and the lack of a convincing IE etymology of the Slavic word (cf. Pokorny 1959:920, s. v. sk^i‘schneiden, trennen, scheiden’) indicate that the latter is rather an old loan from Uralic. 8 Cf. the change š- > F h-, e. g. Sum. šab ‘bowl, cup,’ F haavo ‘bowl,’ M čava, šava ‘bowl, cup,’ C šoβaš, šaβaš ‘beer keg,’ V χāp, O χaβ ‘trough’ (PU *šapa).

S. Parpola, Sumerian: A Uralic Language. II

277

‘slender, weak,’ Z kiźe̮r, kiźer ‘weak, thin’; ¶u.nu ‘weak, ill’ ~ F huono ‘ill, bad,’ H csúnya ‘bad’ 1.3. Change -s > -¶ ~ Ø (F -h ~ Ø): dura¶ ‘ibex’ ~ F teuras, teurah-, teura‘cattle, deer’; ú/ù.lu¶ ~ F oras, orah- ‘young shoot, sprout’ 2. Morphology 2.1. 1st and 2nd p. sg. personal pronoun stems ending in -u: -mu, -zu ‘my, your’ ~ F mun, sun id., mua ‘me,’ etc. 2.2. Translative case composed of Uralic *-k and *-ś latives (-ak-šè, F -ksi, -kse-, M -ks; -ke4-éš ‘because (of ),’ M -kis id.)9 2.3. Comparative suffix -pi used adverbially: ul4.la ‘quick,’ ul4.la.bi ‘quickly’ ~ F väleä ‘quick,’ K väljämbi ‘(more) quickly’10 2.4. Superlative marker -im (also used in verbal derivation): ušum ‘foremost; leader’ ~ F isoin ‘greatest,’ isoinen ‘mighty man,’ M oćuv ‘eldest; chief,’ oćuńä ‘commander, prince,’ šerkan-di ‘to decorate’ ~ F herkentää ‘to render more delicate’ 2.5. Infinitive ending *-tAk, *-δak (Sum. -V-da, -V-dè, F -ta, -da, -tä, -dä; also in O) 3. Lexicon and semantics – ád ~ F ota ‘thorn, prick’; in other Uralic languages the meaning is ‘spear’ – a.rá ‘time, occasion, installment; times; way, journey’ ~ F erä ‘time, occasion, installment; hunting trip’ (not in other Uralic languages; ~ O 3rɜ ‘extra, surplus’) – aš.te(n) ~ F istuin ‘throne, seat, chair’ from istu- ‘to sit down,’ M ašt’ema ‘domicile’ from ašt’e- ‘to sit, live, sit’ (except for L ( j) ištɒ- ‘to sit down’ and O est3- ‘to settle, descend,’ not attested in other Uralic languages) – bùru ‘to drill; sob, wail’ ~ F porat- ‘to drill, weep,’ poru- ‘to weep, wail’; the stem ending in -u is not attested outside Finnish in the meaning ‘to cry’

9 E. g., min-kam-ma-šè [minakam-akśe] ‘secondly’ (Gudea Cyl. A ix 5); ur-sag ug5-ga ì-me-ša-ke4-eš ‘because they are dead heroes’ (ibid. xxvi 15); áš-bal-a-bake4-eš ‘because of these curses’ (UET 1, 294:29). 10 Note also ul4.la ‘rapidly’ (Ur-Namma A 3) ~ Ingrian väljǟ, vällǟ ‘soon, quickly’; ul4.la.aš ‘quickly’ (Šulgi B 69) ~ F väleästi id., ul4.e.en ‘speedily, very quickly’ (SBH 8:82f.) ~ F väleen ‘soon, quickly,’ ul4.la.bi ‘very soon, quickly’ ~ K välembä, väl’jämbi ‘more quickly.’

278

Articles: Ancient Near Eastern Studies

– gar9 ‘cream’ ~ F kuor(r)e, E koor ‘cream’ (not in this meaning in other Uralic languages) – gìb ‘ill’ (Schretter 1990:186) ~ F kipeä ‘ill, sore,’ kipu ‘pain, ache’ (Finnic only, cf. H hiba ‘flaw, fault, blemish’) – gir5 ‘to race’ ~ F kiri- ‘to spurt,’ gur5.ru.uš ‘quickly, hurried’ ~ F kiiru, kiirus ‘hurry,’ kiiruusti ‘quickly,’ M kur3ksta ‘soon, quickly’ (not in other Uralic languages) – mes ‘man’ ~ F mies, E mēs id. (not found in Uralic outside Finnic) – mìn.a.bi ‘both’ (from mìn, man ‘two’) ~ F molempi ‘both’ (not in other Uralic languages; cf. H monna ‘both’) – sìg, si.ki ‘to strike; weave’ ~ F iske-, E iski- ‘to strike’ also ‘to weave, twine’ (in this meaning not in other Uralic languages)

ane, ene menden

menzen enene

3 Pl. 1

2 3

-ani -me

3 Pl. 1 -nne -nsa

-nsa -mme

-ni (mun) -si (sun)

Finnish

te he, ne

hän me(t)

Finnish mä, minä sä, sinä

-dek -sek

-s -mek

-m -d

Lappish

dī, dij sī, sij

son mī, mij

-nk -nzok

-zo, -ze -mok/mek

-m -t, -ćä

-δa, -ta -št

-ž3, -žo -na

Cheremis -m, -em -t, -et, -č

tiń, tä, te siń, nine, n3ne

son, tudo miń, mä, me

Volgaic mon, m3n, m3j ton, t3n, t3j

Mordvin

Lappish mon, mun don

-dį, -tį -zį, -sį

-ez, -įz -mį

-e, -į -ed, -įd

Votyak

ti, tije soos, nije

so, si(je) mi, mije

Permian mon, me ton, te

-nįd -nįs

-įs -nįm

-e̮j -įd

Ziryene

-3n -et

-ä, -3t -w

-em -en

Ob-Ugr.

nen, nēn tän, teγ, leγ

tüw, tew, luγ min, meŋ

Ob-Ugrian äm, mä, män naŋ, neγ

-tok, -tek -(j)uk

-(j)a, -(j)e -nk

-m -d

Hungarian

ti ők

ő, ön, anmi, mí

Hungarian én te

Place of possessive suffix (Px) in relation to plural (Pl) and case (Cx) suffixes: – Sumerian: N + Px + Pl (-ne-) + Cx (dumu-zu-ne-ra ‘son-your-Pl-to’) (Edzard 2003:53) – Finnish: N + Pl + Cx + Px (talo-i-sta-mme ‘house-Pl-from-our’) (Korhonen 1981:234) – Lappish: N + Pl + Cx + Px (nieidâ-i-nâ-m ‘girl-Pl-in-my’) (Korhonen 1981:236)

-zune -nene

-gu10/mu -zu

Sg. 1 2

2 3

Sumerian

0000

Possessive suffixes

Sumerian gá, mà, me za, zé

0000 Sg. 1 2

Personal pronouns

Appendix 6. Personal pronouns and possessive suffixes

-len, -let -den, -det

-t, -tį, -²į -men, -met

-m, -ḿi, -w -l, -lį, -r(3)

Samoyed

tē, tende

tep mē, mende

Samoyed man, mǟ tan, tǟ

Lappish

sī, sij ‘they’ juokke ‘every(one)’ mi ‘what?’

Finnish

(e-) ne(t) ‘they’ se ‘it; that’ joka ‘he who; every (one)’ mi- ‘what?’

M e- ‘this’ M ńe(t) ‘these’ śe ‘that; it’ M tago-, tavo‘anyone’ M me-, C ma id.

O t’i, ś ĭ ‘that’

O mĕj ‘what?’

U ma, Z mij id.

in, iń ‘now’

Ob-Ugrian

Z eta ‘this’ Z ne̮ ‘these’ U so, Z si id.

Permian

mi ‘what?’

aki ‘he who’

ez ‘this, it’

Hungarian

mī ‘what?’

šē ‘that, it’

Samoyed

Also used as the 3rd person (sg. and pl.) inanimate possessive suffix and as a derivational suffix used for adverbs, comparatives and superlatives. To be compared with the Uralic comparative suffix -pi and possibly with the 3rd person sg. suffix -pi of the present tense in Finnic (Häkkinen 2002:107). Cf. Turkic bu ‘this’ (n. 11).

11

Sumerian -bi ‘it, that’11 -e ‘this’ ne ‘this’ še ‘that’ a.ba ‘who; whoever’ me- ‘what?’

Volgaic

N + Cx + Pl (-n-) + Px (moda-d3̑-n-t ‘land-from-Pl-your’) (Bartens 1999:102) N + Px + Pl + Cx (olma-m-βlak-eš ‘apple-my-Pl-into’) N + Pl + Cx + Px (olma-βlak-eš-em ‘apple-Pl-into-my’) N + Pl + Px + Cx (olma-βlak-em-3̑šte ‘apple-Pl-my-in’) (Alhoniemi 1985:76) N + Pl + Px + Cx (pi-jos-įz-leś ‘son-Pl-his-from’) (Wichmann 1954:146) N + Pl + Px + Cx (ćoj-jas-įs-dine̮ ‘sister-Pl-her-to’) (Wichmann–Uotila 1942:369) N + Pl (-n-) + Px (aćįm, aćįd ‘my-, yourself,’ Pl. aćnįm, aćnįd ‘our-, yourselves’) (ibid. 370) N + Pl + Px + Cx (χot-ʌ-am-na ‘house-Pl-my-in’) (Kulonen 1993:20, UrLang 181) N + Pl (-an-) + Px + Cx (kol-an-3w-n3 ‘house-Pl-our-to’) (Kulonen 2007:62) N + Pl + Px + Cx (ház-ai-m-ban ‘house-Pl-my-in’) (Kulonen 1993:20, 70) N + Pl (-ni-) + Px (+ Cx) (loga-ni-m ‘fox-Pl-my’) (Korhonen 1981:234)

Appendix 7. Other Pronouns

– Ostyak: – Vogul: – Hungarian: – Selkup:

– Mordvin: – Cheremis: or or – Votyak: – Ziryene:

Sg. 1 2 3 Pl. 1 2 3 (tr.) 3 (itr.)

Person

Sumerian -(e)n -(e)n -Ø -(e)nden -(e)nzen -(e)ne -eš

Finnish -n -t -Ø -mme -tte -vat -vat

Appendix 8. Personal suffixes of the verb Lappish -m -k -Ø -mek -dek -ik -Ø

Volgaic M -n, C -m -t, -č -ś, -š, -Ø -ma, -ńek -tada, -ďe -jť, -t -śť, -č

Permian -Ø U -d, Z -n -z, -s -mį, -m -dį, -nnįd -zį, -nį -zį, -isnį, -eś

Ob-Ugrian -3m -3n -Ø -īw(3), -3w -īn(3), -t3n -3t -st, -3t

Hungarian -m, -k -d, -sz, -l -Ø, -ott -u(n)k -tok -nak, -(á)k -ak, -ek -wa, -mu -da, -ru -ʔ

Samoyed -m -n, -ŋ

Sum. -a, -u -a,13 -i -àm -a(š), -e(š) -s, -š, -še -d, -l -da, -dè -ed, -ud, -l -e, -u, -Ø -e, -u(¶) -im, -um -im, -n, -me -g, -¶ -k -gi -g, -gu -uš, -u

Meaning deverbal nouns & adjectives nomina agentis, participles ordinal numbers similarity, appurtenance similarities, diminutives

places, localities, distances infinitives, gerunds present participles, adjectives results, products, tools results of action, adjectives superlatives tools, products possessive adjectives appellatives names of plants and animals beings, places, groups nouns of similarity or affinity

-l, -la -t/dA -e(v)a -e’, -ek -e, -eh -in, -im -in, -me -kas -kka -kki -kko -us, -u, uh, -os

Fc -a, -o -ja, -va -nte-as, -es -ća -l’, -la -do, -da -do, -d3̑ -e, -γe -3̑š, -ež (-v) -em -v, -j -ka -ka -ks, -kš

-s, -š

Vc -a -aj, -j -mše-(a)s, -aš -ća, -će

-t -d -â(g) -âs, -es -mus/š -im -kkes -k

L -o -(gg)je -ante-âs, -aš -š

-ka -k -k -eź, -e̮s

-e̮m

-įt, -įd -e -įś

-l

P -a, -o -(j) -eti-įs, -e̮s -čį/-²́, -ś

-es

-k -ki, -kɜ

-l -ta(γ3) -te -γ, -χ -e̮s, -e̮š -im

-aj -m3t-3(š), -a -ťɜ, -ś ĭ

OU

-c, -s

-kos -k

-s

-ete, -z

-l

H -a -ó -mad-a, -asz -s, -zs

-kka -ke -kko

-t -eδa

S x -j -omć-ɒ -će, -še

-ak before vocalic suffixes, cf. lú é dù-a-ke4 ‘the man building the house’ (Gudea Cyl. A xx 24, B xiii 13, xiv 14); šà gù-di é dù-da-ka-na ‘into his heart crying for building the house’ (Gudea Cyl. A xx 2); lú é dù-da-ka nam-ti-la-ni ¶é-sud ‘may the life of the man building the house be prolonged’ (Gudea Stat. C iii 18 – iv 1). -i after stems ending in i, cf. ¶úb ~ F hyppi- ‘to dance,’ ¶úb.bi ~ F hyppijä ‘dancer.’

13

*-a, -o *-aka *-amte *-aš/-eš *-ća, -će *-δa *-δak *-eδ(a) *-ek *-eś *-im *-ime *-kaś *-kka *-kki *-kko *-kśe

0Uralic

Nouns (including nominal forms of the verb)

Appendix 10. Derivational suffixes (cf. Lehtisalo 1936; Hakulinen 1961)

*-ak *-il *-kta *-kśi *-lta *-mta *-śta0000000 *-šta *-ta

Verbs

*-pa *-pi *-ra *-ri *-ut *-uδek

*-la *-le *-ma *-na *-ńe

0Uralic

momentary verbs causative/factitive verbs

imperatives frequentative verbs causative/curative verbs sensitive verbs momentary verbs translative/causative verbs factitive/causative verbs

Meaning adjectives pejoratives verbal nouns, participles (descriptive) verbal nouns diminutive & caritative nouns, adjectives participles, nouns, adjectives comparative, contrast substantives and adjectives actors and professions nouns and adjectives abstract nouns

-a, -ù, -ì -il -tta, -da -g -l, -mt -nd -š, -šd, -d, -z -¶, -š -d

-b, -ba -bi -ra, -r -r -ud, -u -uš, -u

Sum. -l -l -m, -n -n -n

-hta -ta

-ʾ, (-k) -el, -il -tta -ksi -lta -nta -sta

Fc -la, -lo -l(e) -m/va -na -(i)nen, -n -pa, -va -(m)pi -ra, -ro -ri -ut, -o -(u)us

-št -t, -δ

-lt, -äl -aŋ, -em -št

-ld -d -st -st -d

-k, -Ø -l’e, -la -vto/-kt

-či/-t, -c

-Ø, -h -âl -t(t)

-udâk

-b, -p -ŕ(e) -ar

-v, -f -bo, -b́e -(e)r

-(i)ń3

-n, -ń3

-m

Vc -le, -lo -l -mo, -v

L -l

P

-št -t

-lt -nt -st

-Ø -įl, -al -t

-r

-m

-ń

-m, -n

-l’

-t

-lt -m

-t

-jl -nt

-j -l -t

-id

-3t, -3l

-a, -Ø -l -tt3

-bb -r, -rú

-m -ni -ny

H

-pi -rɜ, -r

OU -t -l, -t -m -nį -ŋ

-t

-l

-ʾ -l -ttā

-β, -pa -mpoj -rʾ, -r

-ne

-m3

S

F -sti -sti, -steen -ttain -in, -en -mpi -ešt

L

-3n

Vg -ste -śť -sa

P

OU -(3)ś

ki-5-šè, ki-5-šè-te-en = adi ¶anšīti ‘five times’ (Kagal C 131–132) ~ F viidesti ‘five times,’ viisisteen ‘by fives.’

-ta.àm -um, -n -bi

distributive numbers adverbs modal adverbs

14

Sum. -éš, -eš -ešten14

Meaning modal adverbs ‘n times, in groups of n’

Adverbs

-en

H

S

286

Articles: Ancient Near Eastern Studies

Examples (base word → derivate ~ Uralic parallels) *-a šub ‘to diminish’ → ¶u.ba ‘destitute, weak’ ~ K hupa ‘weak, little’; ¶u.nu ‘weak, ill’ ~ F huono ‘bad, poor, weak, ill,’ H csúnya ‘bad’; sù¶ ‘to confuse’ → sù¶.a ‘blurred’ ~ F sokko ‘blindman’; gur10, búr ‘to harvest’ → guru7, buru14 ‘harvest’; ~ F korjuu ‘harvest’; ul ‘beginning of time’ → u18.lu ‘primal, original (man)’ ~ F alku- ‘beginning, primitive’ *-ak Edzard 2003:127–129 *-aka eme ‘to suckle’ → emeda ~ F imettäjä ‘wet nurse’; gur4, kur4, búr ‘to be round, fat’ → gurušda, kurušta ‘fattener’ ~ F pyöristäjä ‘rounder’; ¶úb ‘dance’ → ¶úb.bi ‘dancer’ ~ F hyppijä ‘dancer’; igi ‘witness’ ~ F näkijä, M ńeji ‘witness’; íl ‘to raise, elevate’ → íl.la ‘high, elevated’ ~ F ylevä ‘high, exalted,’ L âllâg, ɛ̮llaγ, jɛ̮lla̮Gε ‘high’; til ‘to live’ → ti.la ‘living, fresh’ ~ F elävä, eläjä ‘living, alive,’ L jiellej ‘living,’ M eŕi, eŕej ‘living; inhabitant,’ C ( j)il’e ‘fresh, alive,’ U ulep ‘living, alive,’ O j3l3w, jil3p ‘new, fresh,’ H élő ‘living,’ S ďili, ǯili ‘living, alive’; zur ‘to moan’ → zurx, sura10 ~ F *surja ‘wailer’ *-amte min-nam ‘second,’ eš5-àm ‘third,’ i-àm ‘fifth,’ àš-àm ‘sixth’ (Gudea Cyl. A xxi 1–9) *-aš, -eš gír ‘sword, knife’ → kirix(TÙN×KÁR) ‘axe’ ~ F kirves, kervez, L kerwes ‘axe, adze,’ M carvaska ‘battle axe,’ O kèras ‘adze’; kungi ‘onager’ → kuniga ‘royal’ ~ F kuningas, kunikas, kuńigaz, L konogas, gonâgâs ‘king’; kúr ‘strange’ → kúr.ra ‘strange, stranger,’ kiraš ‘slave girl’ ~ Fc vieras, vēraz, võõras ‘guest, stranger’; kúm ‘warm, hot’ → émeš ‘summer, hot weather’ ~ Fc keväš, kevät, čeväd ‘spring,’ C keŋež ‘summer,’ U Z tuvįs, O tǡw3, V tūj3, tūja, H tavasz ‘spring,’ S tåŋɒ ‘summer, spring’; lal ‘team of equids’ → kušlál, kušlá.lá ‘horse harness, strap’ ~ F valjas ‘horse harness,’ Z volįs ‘harness, carrying strap,’ O täwuš ‘harness,’ loγwɒš ‘reins’; síg ‘hair, wool’ → siki ‘wool’ ~ Fc säie, šäigeh, seigiž, säike ‘(wool) fibre, bunch of flax,’ L saigka ‘wool,’ saigâš ‘fleece, hair,’ M śuks, śiks ‘laced ribbon, wool for embroidery’; te ‘support, rack’ → té¶i ‘support, stanchion’ ~ F tu(k)e, Z čeg3̮s, H dúc ‘support, prop’; ug, úg, ug4 ‘beast’ → ugamušen(.s) ‘crow, raven, bird of prey ~ Fc haukka, habuk(az), haugas, L hāpkie, hāpak ‘hawk,’ U čoŋγa ‘a bird, snipe,’ H csóka ‘jackdaw’ *-ća mí ‘(young woman) → munus/š ‘woman’ ~ Slk emn3ćä ‘daughter-in-law’; muš ‘snake’ → muš.¶uš ‘dragon’ ~ Fc vaskitsa,

S. Parpola, Sumerian: A Uralic Language. II

*-δa

*-δak *-eδa

*-ek

*-eś

*-il

287

vaškič, vaśkus ‘a limbless reptile, Anguis’; nin ‘lady’ → nanše, našše, nazi ‘a goddess’ ~ Fc neitsi, neitsy, ńeičoi ‘virgin, girl,’ Slk nēću ‘little daughter’ bur ‘bowl’ → bur.zid ‘(offering) bowl or cup’ ~ F pursilo, purtilo ‘high and narrow drinking bowl’; gišgal ‘stand, step’ ~ F askel, M eśkel’ks, C ašk3̑l, U ućkįl, Z vośkol, V ūs3l ‘step’; gú ‘pulse; bean, pea’ → gú.gal ‘chick pea’ ~ F kukkaro ‘pod, pouch,’ L kuhkkār, gukkar ‘pouch,’ kukol’ ‘wild pea’; kankal ‘uncultivated land’ ~ Fc kangas, kankari, kankaro, kangar, kāŋgaŕ ‘heath,’ U kaŋga ‘dry, hard and barren (soil),’ O konk3̑l ‘dry river-bed’ Edzard 2003:134–137 dag ‘to trample down’ → da×al(a), damal ‘(to be) broad, wide’ ~ F lakea, lavea, L łȧɯ̯je, Z lapįd, H tágas ‘broad, wide’; dugud, zé.be.da ‘heavy, thick, difficult’ ~ Fc tukeva, tykevä, jykevä, d’üged ‘heavy, thick,’ tukela ‘difficult,’ C čokata ‘thick,’ U Z śekȋt ‘heavy, difficult’; gib ‘to bend, twist’ → gibil, gi.gi.ir ‘(to be) new’ ~ Fc kevyt, kepeä, kebjed, kevelä keppelä, kepperä ‘light, nimble, elastic,’ L gæppâd, geppis, M ḱepana ‘light,’ Z kįpįd ‘agile, deft,’ H könnyed, könnyü ‘light, airy, nimble’; sig5 ‘(to be) nice, good,’ si.ga ‘clear, clean’ → sikil ‘pure, clean, virginal’ ~ Fc siveä, čived(a) ‘pure, clean, chaste,’ O suŋl3̑χ, soŋʌĭ ‘clean, clear, innocent’ en4 ‘voice’ → èn.di, èn.du11(g) ‘song,’ F äänne’ (äänte-) ‘sound,’ H ének ‘song’; bar- ‘to split’ → bar ‘chip, cut-off piece of wood,’ Fc päre(h), ṕärag, pirg, L bǣrâ(g-) ‘shingle, chip,’ U pȋrȋ, pȇri, Z pįrįk, pįrįg ‘bit, crumb,’ O p3r ‘bit, bite, crumb; chip, shingle,’ V pār ‘slice,’ H forgács ‘chips, cuttings’; búr ‘to glow, burn’ → bùru ‘foam,’ Fc pore’ ‘bubble,’ U bįre̮ktį- ‘to boil, bubble,’ H buborék ‘foam, bubble’ éd- ‘to come out, sprout’ → gišù.lu¶ ‘shoot’ ~ Fc oras, e̮raz ‘young shoot, sprout’; gu.ul- ‘to get enough’ → ¶i.li ‘luxuriance, abundance, plenty’ ~ F kylle, kyllyys ‘abundance, plenty, satiety,’ L gâllas, ka̮łłäs ‘satisfied, fed,’ Z ke̮lįś ‘wedding’; še.er ‘radiance’ → še.er.zi ‘sunshine, ray’ ~ F helle’ (helte-) ‘heat of sun’; tur ‘young’ → duru5 ‘soft, moist,’ Fc tuore, tuoreh, torez ‘moist, fresh,’ Z ńure̮ś ‘wet,’ H nyers ‘fresh’ gú.gi ‘to test’ → gú.gil ‘to check, try out’ ~ F kokeil- ‘to test, try out’; geštu, muštu ‘ear, memory’ → ×izzal, ×issal, misal ‘to pay attention’ ~ Fc muistele-, muissele- ‘to recall, pay attention,’ L

288

*-im

*-ime

*-kas

*-kka

*-kki *-kko

*-kśe

Articles: Ancient Near Eastern Studies

mušt(e)le- ‘to recall,’ Z kaźal- ‘to pay attention,’ O käšät-, käźät-, V kešäl- ‘to notice’ íl ‘to raise, lift, elevate’ → alim, e.lum ‘exalted, honoured,’ F ylin ‘highest,’ O 3lim ‘greater’; u18.ru ‘high, tall’ → ušum ‘first in rank, leader,’ F isoin ‘greatest,’ M oćuv, oću ‘chief, leader, greatest’ ab ‘opening,’ ba ‘to open’ → apin ‘plough’ ~ F avain ‘key, opener’; aš.te(n) ~ F istuin ‘seat,’ M ašťema ‘domicile, seat’; gir5 ‘to turn’ → girin ‘potter’s wheel,’ F kierrin ‘turn wheel’; gul ‘to cut, break’ → gul.me ‘axe for hewing stone’ ~ F vuolin ‘paring tool’; kár ‘to shine’ → karadin ‘pile of sheaves, bundle of reeds’ ~ F kärväin ‘drying rack for hay,’ Z gįrʒ́im ‘bundle, bale’; umbin ‘wheel; currycomb; chisel’ ~ F vempain ‘gadget’; zi, zid ‘to clean’ → zíz.àm ‘hulled emmer’ ~ F leseme, listime ‘bran,’ U čužjem, ćužjem ‘malt’ ùš ‘womb; afterbirth’ → ú.sag, ú.sug, míú.zu¶ ‘unclean woman’ ~ Fc vačakas, vačoikas ‘potbellied, pregnant,’ M vaćev, vaćej ‘dirty, filthy,’ H mocskos ‘dirty, filthy, soiled, damned’ amaga, amagi ‘ice, snow’ ~ K jähmäkkä ‘stiff, numb with cold’; zal ‘to shine’ → zálag, zalag(.ga) ‘bright,’ zalagku6 ‘a fish’ ~ F helakka ‘bright,’ salakka ‘bleak (Alburnus alburnus),’ Z ćal’ak ‘roach,’ O ťāl3̑k ‘clear, light,’ H csillogó ‘shiny, glistening, lustrous, light’ dìm.ma ‘weakling, castrate’ → dìm.gisar ‘a garden plant’ ~ F lemmikki ‘forget-me-not, Myosotis’ èš, e.zé (PSum. *kwe̮ć3) ‘sanctuary’ → usug, únug ‘shrine’ ~ F pyhäkkö ‘shrine, sanctuary’; nin ‘lady’ → nik ‘bitch’ ~ F naikko ‘wench,’ nakku ‘bitch,’ Km nɯkɛ ‘hag,’ Slk nēku ‘stepdaughter’; nir ‘strand, cord’ → nir.gu ‘fruit string’ ~ F norkko, norko ‘catkin’; šub ‘to diminish’ → ¶u.bak ‘silly, fool’ ~ F hupakko ‘silly, fool,’ L håpāk ‘senile,’ M čavača ‘stupid, silly, fool’ dal ‘to fly’ → kušdal.úš, kušda.lu.uš ‘sling’; gi7, gir15 ‘male, manly’ → guruš ‘adult man, male’ ~ F uros ‘man, male,’ U vere̮s ‘man, husband,’ O kǡr3s ‘adult elk,’ H béres ‘farm hand’; gir5, gir7 ‘to sprint’ → gur5.ru(.uš) ~ F kiire, kiiruh-, kiirus ‘hurry’; kur ‘mountain’ → gur5.ru(.uš) ‘scalp’ ~ F kiiru, kiire(h), L ḱierre, M seŕc ‘top of the head, scalp,’ U gureź, Z kere̮s ‘hill, mountain, elevation,’ O V kèras ‘ridge,’ H bérc ‘rock, mountain hill’; su6 ‘beard’ → sukuš ‘point; bristle’~ Fc suas, šuvaš, sugas ‘point; bristle,’ M śuvoź ‘pointed; sewing needle,’ U zueś ‘bristly’; ùlu,

S. Parpola, Sumerian: A Uralic Language. II

*-kśi *-kta

*-la

*-le

*-lta

*-ma

*-mta

289

u18.lu ‘south’ (= discharge of the Euphrates and the Tigris) ~ F alus ‘lower part, base; lower course of the river,’ L vuolaš, V lo ‘lower course of river, north,’ Z ule̮s, le̮s ‘seat, base, flooring’ ¶ul ‘to hate, detest’ → ¶ul.gig ‘to hate, despise, dislike, avoid’ ~ F hyljeksi- ‘to shun, avoid, despise’ ár ‘glory’ → aratta ‘praise,’ F arvotta- ‘to esteem,’ O orkat- ‘to praise’; gú ‘neck’ → gú.gíd ‘to lean over, bend toward, stretch the neck, crane’ ~ F kokotta-, kurkotta- ‘to crane, stretch the neck,’ H kókad- ‘to wilt, bend down’; súg ‘to stand’ → sukud.da ~ F sojotta-, soikotta- ‘to project’ a.gù ‘brain’ → á.gál(.la), á.mal ‘able, intelligent, clever’ ~ F ovela ‘clever, shrewd, crafty,’ evälä ‘intelligent,’ L oaivalâš ‘able, intelligent,’ C ak3lan ‘wise, reasonable,’ amalan ‘crafty,’ U amalo ‘smart, artful, wily,’ O omāt3ŋ ‘crafty, clever, shrewd’; gib ‘to bend, twist’ → gibil ‘new’ ~ F kevelä ‘lissom, lithe’; su8, súg ‘to go, glide’ → sukkal ‘messenger; minister’ ~ F sukkela ‘swift; smart’ dug4 ‘to do, make’ → tukul ‘weapon’ ~ F tekele ‘unfinished product, bungle’; kud, gudu4 ‘to cut, break (off ) → gú.zal ‘damaged thing; scoundrel, rascal’ ~ F kutale ‘rag, frazzle; rascal, worthless person,’ kʿȧt’̜ ʌ, kēt’l, kȧtkŏt ‘rag, shred; worthless (person, thing)’ bu, bu7 ‘to blow’ → bul ‘to blow (away), winnow’ ~ F puhalta-, C pualte-, U peltį-, Z p3̑lt-, O pŏγ3̑lt-, H fújt-, S pūk3l- ‘to blow (away)’; á.×ál, á.mal ‘to be able, understand’ ~ F oivalta- ‘to understand, grasp,’ L oaiveldi- ‘to think,’ C 3̑ŋg3le- ‘to understand,’ U amaltį- ‘to grasp’ agan ‘nipple, carbuncle’ ~ F äkämä ‘pimple, carbuncle’; gan ‘to carry’ → ganam ‘ewe’ ~ F kantava, O kŏn3̑w ‘pregnant’; ge4 ‘to kill’ → gidim ‘spirit of the dead’ ~ F kalmo ‘dead body, graveyard spirit,’ L galbme, M kulov ‘dead, corpse,’ U Z kulem ‘dead, deceased,’ O χat3m, χăd3̑m ‘dead, corpse,’ Km kolmu ‘spirit of the dead’; imma ‘to be thirsty’ → immin ‘thirst’ ~ M śiḿima ‘thirst’; u5, mu ‘strong current, river’ → ù.mu.un ‘blood vessel, vein’ ~ F (v)uoma ‘river bed, channel’ še.er ‘delicate’ → šerkan-di ‘to decorate’ ~ F herkentä- ‘to render more delicate,’ C šerγemδe- ‘to raise the price,’ H serkent- ‘to stimulate’; u¶, ú¶ ‘cough; spittle, phlegm’ → a.¶a.an, ¶a.an ‘to vomit’ ~ F oksenta-, M uksno-, C ukšinča-, uks3̑nza-, O āχ3̑t-, V ājtoχt-, aiden-, H okád- ‘to vomit’

290

*-na

*-ńe

*-pa

*-pi

*-ra

Articles: Ancient Near Eastern Studies

mu7.mu7 ‘to rumble, mutter’ → mumun ‘noise, rumble, spell’ ~ F mumina ‘mumbling, muttering, murmur’; ba ‘to open,’ ab ‘opening’ → akan, aka4, kan4, ká ‘doorway, gate’ ~ Fc a(a)kkuna, akkona, akan, ake(n), oken ‘window,’ L ɛ̮ɔkan, jeɔ̄kan, M akońśa, aŋkońća, C okna, U ukno, Z e̮kińća ‘window’ ezina, ezinu ‘emmer wheat’ → ašnan ‘grain’ ~ F vehnänen ‘bun’; gašan, kaša(n) ‘lady, queen’ ~ F kaaso, kaasein ‘bridesmaid,’ E kaasa ‘spouse’; ná ‘to lie, sleep with’ → nin ‘lady, queen’ ~ F nainen, M ńi ‘woman,’ O nen, niŋ, ni ‘wife, lady,’ H néne ‘elder sister’; giš, muš ‘tree, wood’ → mušen ‘bird’ ~ mettinen ‘stock dove,’ M mäčin ‘small dove,’ C m3̑ze, muźo ‘hazel grouse,’ V mansin, O manšiŋ, mänčim ‘wood grouse’; pisan ‘box, chest; gutter’ ~ F pesä, pesänen ‘den, nest; bowl; firebox,’ M pizińe ‘small nest,’ C p3žäš ‘nest, den,’ Z poz ‘nest, cage, coop’ erib ‘leatherworker’ ~ F nirhivä, nyrhivä ‘abrading, rubbing off,’ Y ńirraBś ‘scraper’; gub ‘to stand’ → gub.ba ‘standing, regular,’ F käypä, käyvä ‘current, valid’; geš ‘hand’ → kišib ‘hand, fist,’ U Z kȋrȋm, kįrįm ‘hand, fist’; ¶a.lu.úb ‘a variety of elm or willow’ ~ Fc halava, salava, alāb ‘brittle willow,’ H szilfa ‘elm,’ S sāl’pä, sajb3̑, haljbé ‘rowan, mountain ash’; kíri ‘to tie, wrap’ → kilib ‘bundle of reeds,’ F kerp(p)u, kerp(p)o ‘bunch of twigs,’ H ḱ iŕf ‘bundle, bale, roll’; múr(u), úr ‘middle, space between’ → múrub ‘asshole, vulva,’ F välppy ‘slit, crotch’ bar ‘side’ → bar.bi ‘extra, additional’ ~ F parempi ‘better,’ K more,’ L buorēb, pārēp ‘better’; mìn ‘two’ → mìn.a.bi ‘both,’ F molempi, Vt me̮le̮pad ‘both’ à¶-, ¶ád- ‘to dry’ → a.gàr, a.da.ar ‘meadow, grassland’ ~ F ahtera, aher, ōd3̑r ‘fallow (land),’ H ugar ‘fallow land’; ba(d) ‘to open’ → ba9.rá ~ F avara ‘open, wide’; ba.ba ‘porridge’ → bappir ‘beer bread’ ~ F pöp(p)erö ‘pudding made of malt or flour’; ¶a.an ‘weak’ → ¶e.en.zèr ~ F hinterä ‘slender, weak’; nim ‘dusk,’ šem ‘patina’ → ne.mur ~ še.mur ‘ash, potash’ ~ F hämärä ‘dusk, dark’; sù¶ ‘to confuse’ → su¶ur ~ F sykkyrä, sykerö, Z ćuke̮r ‘bunch,’ O š3kǡrɜ, V siqar ‘curl’; tab ‘to do quickly’ → tabira, tebira ‘craftsman working in wood and metal’ ~ F näppärä ‘handy, dexterous,’ M ćebeŕ, C ceβer, ťeβer, U ćeber, čeber ‘beautiful, good, quick,’ O ťepärä, ćeBärä ‘well, properly,’ V śäper ‘good,’ H éber ‘alert’; tuk4 ‘trouble’ → tukur ‘grave, difficult, serious’ ~ F tukala, tukela ‘difficult, tight, troublesome,’ C š3̑γ3̑r,

S. Parpola, Sumerian: A Uralic Language. II

*-ri

*-śta-

*-šta

*-ta

*-uδek

291

šȋγȋr ‘tight, strait, difficult,’ O tǡg3rt, jǟg3rt, ʌǟg3rt ‘heavy, hard, difficult’; umun5 [*we̮ŋ] ‘pool’ → engur, an.gur4 ‘sweet-water ocean’ ~ F onkalo, onkura ‘depth, abyss, cavity,’ C aγur, aβur ‘deep spot in river,’ O ŏŋar, ăŋarĭ ‘spring, well; depth’ àra- ‘to grind’ → arar ‘miller’; pàd ‘bucket’ → bá¶ar ‘potter,’ cf. O pič3- ‘to turn, plait,’ H fágy- ‘to turn, spin’; sìg ‘to strike’ → iškur ‘storm god’ ~ F iskuri ‘striker’; nagar ‘carpenter’ ~ F nikkari ‘carpenter,’ O ńŏγ3̑r-, ńŏŋγ3̑r- ‘to whittle, carve’ eš10 ‘water’ → eseš, isiš, šéš ‘to weep’ ~ F vesistä- ‘to weep’; gú ‘bent, crooked’ → gú.ki.iz ~ F koukista- ‘to bend’; gú.ur5 ‘throat’ → gú.gíd ‘to strangle, throttle’ ~ F kurista-, E kurgista- ‘to strangle’; gur4 ‘fat’ → gurušda ‘fattener’ ar¶uš ‘to have mercy’ ~ F armahta- ‘to have mercy,’ armasta- ‘to love’; zal ‘to glide, pass’ → zala¶ ‘to slip, drift, slink in’ ~ F solahta- ‘to slide, glide, slip’; gú.ur5 ‘throat’ → gù.ra.a¶, gù.ra ‘to bellow, bray, cry, rumble’ ~ F kurahta-, kurat- ‘to croak, (g)rumble, growl, snarl’; gur4, kur4, búr ‘to be round, fat’ → gurušda, kurušta ‘fattener’ ~ F pyöristä- ‘to round (off ), make rounded’ à¶, ¶a5 ‘to be dry’ → ¶ád ‘to dry, burn, chap’ ~ F ahvat- ‘to dry, tan, chap’; ba ‘to divide’ → bad ‘to open’ ~ F avat-, L āvδe-, ǡveδe-, M avto- ‘to open’; bùr ‘hole’ → bùrud ‘to drill’; e4 ‘water (PSum. “ice”)’ → éd-, èd- ‘to cool,’ F jäätä- ‘to freeze,’ jäähty- ‘to cool down,’ L jieksō-, M jakšavto-, C (j)ükše-, O j3γ3tti- ‘to cool’; *eme ‘to suck’ → emeda ‘wet nurse’ ~ F imettä-, L njâmât-, Z ńimet-, O ēm3tt3- ‘to nurse’; gurud ‘to whip’ ~ F kurita- ‘to beat, whip, castigate,’ L gurit- ‘to discipline,’ C k3̑re-, k3reδala- ‘to beat, whip,’ Z bordjį- ‘to beat, castigate,’ O kŏr-, χŏr- ‘to beat, whip’; kéš ‘band, belt’ → kešed, kéš.da ‘to bind, tie, gird’ ~ F kiehto-, kiehtä- ‘to captivate, ensnare,’ L giesto-, kiesti-, ‘to wrap, tie’; til ‘to live’ → tilid ‘to keep alive, revive’ ~ F elättä- ‘to keep alive,’ L jĭĕllate- ‘to refresh, revive,’ M eŕaft3- ‘to enliven, let live,’ C il’ate‘to revive,’ U ulįtį- ‘to let live,’ V ilt-, jelt- ‘to revive,’ H éled- ‘to recover’; .tu(d) ‘to be born’ → tu.tu(d) ‘to give birth, engender, create’ ~ F synnytä-, sündüda-, L ŝε̮DDē-, M čačto-, Z ćužtį-, O tăj-, tăjit-, H szül- ‘to give birth, create’ ar¶uš ‘mercy, compassion’ ~ F armaus ‘tenderness, affection’; esi, eša ‘hard/tough wood’ → usu, uššux ‘strength, force’ ~ F visuus ‘strength, solidity’; gissu ‘shadow, shade’ ~ F mustuus

292

*-ut

Articles: Ancient Near Eastern Studies

‘blackness’; lá, lál ‘to be short in quantity, deficient’ → la(l)’u ‘deficit, shortage’ ~ F vajaus ‘deficit, shortage’ ìsim, ú.šim ‘green shoot, aromatics’ → isimu, isima, usmu, isimud ‘divine plant expert’ ~ F osmo, osma, osmi, ozmut ‘myth. hop and plant grower’; lúgud ‘short’ ~ Fc lyhyt, lyhy, l’ühüd, lühü (lühkü), Lv līt, līki ‘short,’ O ńāγ3t, ńāχ3t, ńōγ3l, H rövid ‘short’

gur11, búr, ur4 ‘to collect, harvest’ guru7, buru14 ‘harvest’ karadin ‘sheaf ’ su7, sug8 ‘grain pile, hay stack’ bu7 ‘to blow, winnow,’ bu ‘to blow’

kylvä- ‘to sow’ kylvö ‘sowing’ itä- ‘to sprout, germinate’ vilja ‘corn; fruit’ vehnä ‘wheat’ vehnänen ‘bun’ haippa ‘stunted corn’

kul ‘to sow, scatter; sowing’ è, éd, i ‘to sprout, germinate’ bíl, bìl ‘fruit, shoot’ ezina, ezinu, ašnan ‘cereal’ še gig, gìb ‘stunted corn, wheat’ ù.lu¶ ‘young shoot’ gú(.gú) ‘pulse; bean’ úkuš ‘cucumber’

kärväin ‘hay rack’ suova ‘(hay) stack, rick’ pui- ‘to winnow’ puhu- ‘to blow’

korjat- ‘to collect, harvest’ korjuu ‘harvest’

oras ‘shoot, sprout’ E kukk ‘bean’ (a)kurkku ‘cucumber’

Finnish kuokka ‘hoe’ siemen ‘seed’

Sumerian gug ‘harrow’ numun ‘seed’

M puva-, C pue‘to blow’

M viš(ä), vešiń ‘spelt, wheat’

M śiḿeń ‘stem, seed, breed’ M kalavt‘to scatter’

Volgaic

Appendix 11. Terms for agriculture and animal husbandry

Z gįrʒ́im ‘bundle’

U ogreć ‘cucumber’ Z e̮gurći ‘cucumber’

U vaź ‘spelt, wheat’

U kiźà- ‘to sow’ Z ke̮ʒ́- ‘to sow’ U ud-, udà- ‘to sprout’

Permian Z kokan ‘hoe’

pu-, pŭw-, pŏw-, pŏγ- ‘to blow’

O kul ‘corn’ kire- ‘to sow’ O et- ‘to sprout, grow’

O sem, śemä ‘seed’

Ob-Ugrian

fú, fúj ‘to blow’

uborka ‘cucumber’

hajt- ‘to sprout, grow’

elvet- ‘to sow, seed’

szem ‘grain, particle’

Hungarian

hevos-, hevočču ‘horse’ lehmä ‘cow’ kuoho ‘gelded animal’

zi.iz, zi.zi ‘horse’

šilam, šelem ‘cow’ gu4, gud, gu4.r‘ox, bull’ sún ‘wild cow, (leader) bull’ u8, us5, uš ‘ewe’ udu, e.zé, i.zí ‘sheep’

ganam, gana4 ‘ewe’

siku ‘piglet’ E sikk ‘goat’ karju ‘hog’ emä ‘dam’ koni ‘(old) horse’

zí¶ ‘piglet’ ze¶, zi¶ ‘she-goat’ gir, kiri ‘pig’ ème ‘she-ass’ kungi ‘onager’

uuhi ‘ewe’ utu, uttu, uti, ući ‘sheep, ewe’ kantava ‘pregnant’

sonni ‘bull’

Finnish lese, leseme, lesty ‘(rye, wheat) bran’ käsikivi ‘quern’ kivi ‘stone, millstone’ jauha- ‘to grind, mill’ jauho, -e ‘flour’ sika ‘pig’

Sumerian zíz, zíz.àm ‘hulled emmer, cereal’ kíkkin, kíkken ‘quern, hand-mill’ àra ‘to grind’ eša, áša ‘fine flour’ ša¶(a) ‘pig’

M kandi ‘bearing, pregnant’

M uča ‘sheep’

M ava, C aβa id. M koń ‘horse, thoroughbred’ M śivoj, śivaj ‘horse’ M l’išḿä ‘horse’

U įž, įź, e̮ź, Z įž ‘sheep’

Z ćańa ‘mare with foal,’ U ćuńi ‘foal’

Z koďź/ś- ‘to geld’

U kuńan ‘calf,’ Z kuk, kukań ‘calf ’

Z če̮g ‘fat, thick; fat, bacon, lard, tallow’

M tuva, tuvo ‘pig’

M śeja ‘she-goat’

Permian Z čuž, ćuž U č/ćužjem ‘malt’ U ki-ke̮ ‘hand-mill’ ke̮ ‘millstone’ izį-, iz- ‘to grind’

Volgaic M čuž, šuž ‘barley’ käď-kev ‘quern’ kev ‘(mill)stone’ M jaža-, eža- id.

O oš, oč, V ōš, ɔ̄s ‘sheep’ O kŏn3̑w, kŏn3̑ŋ ‘pregnant mare’

O kowti, χaβti‘geld(ed)’

kő, köve‘(mill)stone’ őröl- ‘to grind, mill’

O köγ, V küw, käw ‘(mill)stone’ īs-, iz- ‘to grind’ O j3ŋet ‘flour’

juh ‘sheep’

tehén ‘cow’

eme ‘sow’ Slk künď (künn3-) ‘horse’

Hungarian

Ob-Ugrian

tarha, tara ‘pen, yard’

tùr ‘cattle pen, yard’

M vaz ‘calf ’ M varda ‘deer’ M kaźa ‘goat’ C kajźa ‘goat’ M ut’a, C woć ‘duck’

Volgaic

Finnish (Finnic) ohjas, ohjat ‘rein(s),’ ohje ‘direction(s)’ ohjat- ‘to direct, guide, steer’ oja, ojas, aisa ‘shaft’ aja- ‘to drive, practice, execute’ ies, igž, igg3z ‘yoke’

kieri- ‘to roll (o. s.),’ vieri- ‘to roll, roll down’

Sumerian uš, ús ‘guidance, direction’ uš, ús- ‘to guide, steer, drive’ giš á ‘plow arm, beam’ aka, a5 ‘to practice, execute, act’ giš geš ‘yoke’ giš gigir ‘chariot’

gir5(-gir5), gígri ‘to rotate’

ḱeviŕ ‘cart, chariot’ M ḱeveŕe‘to roll, rotate’

Mordvin (v)ožďat, (v)ožjat ‘reins, bridles’ vožjad3̑‘to harness’ ažja, ažďa id. ajďa- ‘to drive (on), run’

Appendix 12. Terms for harnessing and drawn vehicles

uz ‘duck’

Finnish teuras ‘bovine; deer’ tarvas ‘stag, deer’ kili, killa, kilu ‘young goat, kid’ vasa, vasikka ‘calf ’ vohla ‘goat kid’ vuohi ‘goat’ kuttu ‘she-goat’ E vais, väis ‘duck’

Sumerian dura¶ ‘ibex’ dàr.maš ‘stag, deer’ ki.di, ki.da, gílin, gili ‘nimble wild animal’ mas, maš ‘calf, kid’ maš.dà ‘gazelle’ ùz, ud5 ‘goat’

O wožat-, wast3‘to drive, spur’ O uš, wošjuχ id. O äjdät-, äjtät‘to drive, chase’ O ikȯw ‘collar’

U uzatį-, Z įzjed- ‘to direct, urge, spur’ U vajįž, Z vož id. U ujį-, ujà‘to drive, pursue’ U eges ‘hoop, band’ U ťigil’ ‘wheel,’ Z ge̮gįl ‘wheel’ U tigil’a-, Z ge̮gįlt‘to roll, rotate’

V kir-, O käri-, kär3j- ‘to roll’

Ob-Ugrian O woš ‘reins’

O was3(γ), V wasse, wiś id.

pereg ‘to roll, roll down’

iga ‘yoke’ szekér ‘cart, wagon’

Hungarian

vöczök ‘dabchick’

őz ‘deer, roe’

üsző ‘heifer’

szarvas ‘red deer’ gida ‘kid, deer calf ’

O käl3ŋ ‘reindeer’ V wāsi(γ) ‘calf ’

Hungarian

Ob-Ugrian

Permian

U keć, keč ‘goat’ Z ke̮za ‘goat’ U vati, vaći Z ve̮ś/ź ‘duck’ Z dor ‘yard’

Permian

sudin, giššudun ‘collar, yoke’ giš umbin ‘wheel; collar bow’ zú.kéš, zukše(d) ‘to hitch up, bind’

giš

rín ‘binding, yoke’

giš

lal ‘to harness, hitch, tie up, bind’ kuš lalla, kušlá.lá ‘harness, strap’ (kuš) pa ‘reins’

kab ‘horse bit’ giš kab.íl (part of a chariot) lal ‘team of equids’

Sumerian gur4(-gur4) ‘to roll, rotate’ giš gúr, bùr ‘disc’

päitset ‘headstall, bridles’ riimu ‘halter (strap),’ rihma ‘cord’ sidin, sidoin, sidos ‘band, binder’ vempain ‘gadget,’ vempele ‘collar bow’ suitsi(t)- ‘to bridle,’ Lv suiksud ‘bridles’

valjas ‘harness’

Finnish (Finnic) pyöri- ‘to rotate, roll,’ vyöry- ‘to roll down’ kurra, kurru ‘disc,’ pyörä ‘wheel’ kapu ‘calf gag’ kaplas, kapilas (part of a sleigh) valio ‘select (animal),’ valjakko ‘team’ valjasta- ‘to harness’

ŕisḿe, ŕiśḿä ‘chain; rope’ sot3ma ‘binding,’ C šüδ3̑š ‘band’

M ulav ‘team of 1–3 horses’

M k3r3jot ‘circle’

Mordvin M puvŕa- ‘to roll’

Z ve̮ń, voń ‘belt’

U, Z śi(j)es ‘collar’

Z volįs ‘harness, carrying strap’

U val, våw ‘horse,’ Z ve̮l ‘horse’

Permian U bergà-, porjil-, Z bergal- ‘to rotate’ U kurj ‘ring,’ Z kor ‘circle, hoop’

O ḱempar, ćemʙar ‘hoop’

O ʌău̯χ, ʌou̯, loχ, V low, lo ‘horse’ O ʌăχ-, lɒχ-, taχ‘to bind, lace’ O loγu̯ɒš ‘reins,’ täwuš ‘harness’ O päk, V pekh ‘rein, bridle’

Ob-Ugrian V powart- id., O peŋ3rγ3- ‘to roll’ O kör3k ‘circle’

gömb ‘ball’ gömbölyü ‘circular’

fék ‘bridle, rein, halter, brake’

ló, lov- ‘horse’

Hungarian forog, pörög ‘to roll’ kör ‘circle, ring,’ kerék ‘wheel’

emo, emä, emoi ‘mother’ appi, appe ‘father-inlaw’ ätä, ättä, ätti ‘father’

akka ‘old woman, wife, lady’ kave ‘girl, maiden’

ama ‘mother’ aba, ab.ba ‘father; old man, elder’ ad, ad.da ‘father’

égi, ègir ‘queen, princess’ géme, gìm, ga.am,.gi4.in ‘girl, maid’ ku.li ‘friend, comrade’ mes, mèš ‘man, male’ mí ‘(young) woman, bride’

mies, mēs, meź ‘man’ miniä ‘daughter-inlaw’

käly ‘sister-inlaw’

Finnish äijä ‘old man, father, grandfather’

Sumerian aja ‘father’

Appendix 13. Kinship terms

mânnje ‘daughter-inlaw’

mihha ‘man’

akka, akko ‘old woman, grandmother’ gāvâ, kāva, kāβ ‘wife,’ kievva ‘female’ kiella ‘sister-in-law’

ačče ‘father’

vuoppâ ‘father-in-law’

Lappish aggja, ajja ‘old man, grandfather’

aťa, aťä, ača C ‘father,’ M ‘grandfather’ M aka, C äkä ‘elder sister; aunt’ M kopə̑na ‘capercaillie hen’ M kijalo, kel ‘sister-in-law’

M ava, C aβa, aβai ‘mother’ C oβȇ, oβo ‘father-in-law’

Volgaic

U meń, Z moń, muń id.

U akī, ako, aka, ak ‘elder sister’ U ken ‘girl’ Z ke̮n-tar ‘greyhen’ kal’i, Z kel ‘daughterin-law’

ataj ‘father’

Permian U aj ‘father,’ Z aj, aja ‘id., grandfather’

O küli, V kil ‘brother or sister-in-law’ O mańť, mbś ‘Ostyak’ O meń, V miń, mäń ‘daughter-in-law’

V ama, oma ‘mother’ apa ‘father,’ O up, V op ‘father-in-law’ äťi, aś3, ɒči, ǟćī V ‘father,’ O ‘uncle’ V ēk 3, ēk a, ° ° jēk, jēkwä ‘wife, lady’

Ob-Ugrian O j3γ, j3χ, V jeγ, jaj ‘father,’ ojGa ‘old man’

meny ‘daughter-inlaw’

ik, ike, ük ‘grandmother’

atya ‘father’

apa ‘father’ ipa ‘father-in-law’

Hungarian

ḿeje ‘daughter-inlaw’

äkä, akka, eke, jegeč ‘elder sister’ keibe, šejmɒ ‘female; mare’ śēl, śaluŋ ‘brother-inlaw’

aćća, aťa ‘uncle’

āwε, äwä ‘mother’ aba, åbå, āβa ‘father’

Samoyed

um.ma ‘old woman’

nita, ninta, nita¶ ‘man’ nu.siki, nus.gi17 ‘orphan’ sul, šul, šu.lá ‘youth; young man’ šeš ‘brother’

nin9 ‘sister’

Sumerian munus, munuš ‘woman’ nin ‘lady’

æmme, įmme ‘old woman’

čæcce, čieɔccE ‘uncle’

sulho(i), sulha ‘bridegroom, young man’ setä, seti ‘uncle’

ämmä ‘old woman’

læskâ ‘widow’

Lappish

leski ‘widow’

nainen, naińe ‘woman; lady’

Finnish

M ć/śora, C šol’a ‘young man’ čiče, č3č3 M ‘brotherin-law,’ C ‘uncle’

M ńi ‘wife, woman’

Volgaic

U čuž, Z ćož, čož ‘uncle’

Permian

O imi, im3 ‘old woman’

Z ʒ́ul’ej ‘bull, uncastrated boar’ V šäš ‘father’s brother’

O ń3ŋi ‘little sister’ O nēt3, neŋtȧ χoj ‘bachelor’

O neŋ, niŋ ‘wife’

Ob-Ugrian

süly ‘male boar’

néné ‘elder sister’

néni ‘dame; aunt’

Hungarian

imma ‘woman, wife’

čičä ‘uncle’

ńeńa ‘sister, stepsister’

Samoyed emn3ćä ‘daughter-inlaw ńe, ńeŋe ‘woman, wife’

kuusi

àš ‘six’

guttâ, kuttA

vittâ

njælljĕ/ă

golbmâ ‘three’

guokte ‘two’

Lappish âktâ, oktâ, akt, e̮χt(A) ‘one’ aktu, e̮χtɒ̑ ‘alone’ månah ‘a couple, two’

M kolmo ‘3,’ kol’ ‘much,’ C kum ‘3’ M ńil’e, ńil’ä, C n3̑l, nil’ M vet’e, C βič, βić, βiť M koto, C kut

Volgaic M ve, vejke, C ik, ikt3, itte M śkam(o), iśkam ‘alone’ C moń ‘the like, another’ C mol3 ‘the other one’ M kavto, C kok ‘two’ U uśse ‘in 3 days’ U kwiń(m-), Z ku(j)im ‘three’ U ńɯl’, ńįl’, Z ńol U vit’, Z vit U kwat’, Z kvat’

U kȋk, kȋkt-, Z kįk ‘two’

Permian U og, ok-, Z e̮t, e̮t’ik U ogńa-, Z e̮tna- id. U Z mįn ‘many’

O kät, kat, kĕt, V kit,’ kit ‘two’ O kitpa ‘twice’ os ‘further, in addition, still’ O k/χɔl3m, V kōr3m, kur3m ‘three’ O ńĕl3, ńil, V ńil3, ńila O wet, uwȧt, V ät, ǟt, at O kut, χot, χut, V kat, χōt

Ob-Ugrian O ĕj3t, ĭt, ĕj, V ük(ī), äko3γ O ātemna-, V ås ‘alone’

hat

öt ‘five’

négy

három ‘three’

két ‘two’

egyén ‘individual’ mennyi ‘how many?’

Hungarian egy, egyik ‘one’

Y ćēt, ťet, Slk tēt3 jūt, bīʾ, k8t, bet, bi ‘ten’ maʾttE, mukt3t, muktu’ ‘six’

Slk ku3̑(l) ‘too much’

Y śiďē, Slk śit3, M kydy ‘two’

Samoyed Slk okkŭr

PU *ńel’a > *ńew(w)a ~ *nim(m) [cf. nimin ‘forty’] > limma ~ limmu. For the dissimilation n- > l- before -m- cf. min/man.a.bi ~ F molempi, Vt me̮le̮p- ‘both,’ and the dissimilation nu- > l- before -b- (Thomsen 1984:190; Edzard 2003:113); note also EG nimir, nigir ~ ES libir (Schretter 1990:202) and PU *nime ‘name’ > M l’em, l’äm, C l3m ‘name.’

15

neljä, ńel’l’ä, ńel’l’ü viisi, vīž, vijź

usea ‘many,’ usein ‘often’ kolme ‘three’

kaksi ‘two’

moni, me̮ni ‘a couple, many’ molempi ‘both’

Finnic yksi, ükś, ühsi ‘one’ yksin ‘alone’

eš5, peš ‘3’ eš ‘many’ kul ‘much, many’ limmu, limma ‘4’15 i ‘five’

ušum ‘alone’ min, man ‘two, both’ min.a.bi ‘both’ kát.a.bi ‘secondly’

Sumerian aš ‘one’

Appendix 14. Numerals

Finnish -s (-nte-, -nsi-, -nne-)

aivo, aju, oimo ‘brain’

kuusi ‘6’

àš ‘6’

Finnic

a.gù, ugu ‘cranium, brain; top’

Sumerian

guttâ, kutt(A) ‘6’

oaive ‘mountain top, head’

Lappish M uj ‘brain,’ C βuj ‘head, upper end’ M koto, C kut ‘6’

Volgaic

Appendix 15. Words with posited PU labio-velar stop

Sumerian -àm, -akam

Permian

Permian U -eti, Z -t, -d, -ďź

Ob-Ugrian

O kut, χot, V kat, χōt id.

Hungarian

hat ‘6’

Samoyed -omć/t’ei, -3mdž/ďel

Samoyed

Samoyed Y Km Slk ŋäeββa, ūk3 ‘top, aiba head’ ‘head’

Hungarian -mad

hatvan

-ven, -van ‘ten’ (40–70) negyven

Hungarian

agy ‘brain’

Ob-Ugrian O -m3t, V -(3)t (-nt-)

O k̔ōm3n ‘flat of the hand’ O ńĕ̜ʌ3jeŋo, V nälm3n O χotjoŋ, V kot-low

Ob-Ugrian

O uχ, oχ, V oχi ‘head’

U Z -mįn ‘ten’ (30–60) U ńįl’mįn, Z ńel’a-mįn U kwat’t’on, Z kvat’-das

Permian śiźȋm, śiźįm

U kwat’, Z kvat’ ‘6’

kuttA-loɔk

kuusikymmentä

Mordvin M -će, C -mše

M ḱeḿeń, ḱemeń ‘ten’ M ńil’eńģeḿeń, C n3̑lle M kodḱeḿeń, C kud-lu

ḱėǟma(n) ‘flat of the hand’

kymmen(en), čümmē ‘ten’ neljä-kymmentä

Lappish -d (-nte-, -nne-)

Volgaic M śiśem, C šišim, šȋm

Lappish čieǯâ, čiččam, kīďča̮m

Finnic seitsemän, śeič(i)ḿe

Ordinal suffixes

Sumerian imin, umun7 ‘seven’ umun, ún, u ‘ten’ nimin ‘forty’ géš.da, géšta ‘sixty’

parvi ‘swarm, flock, troop, group, shoal’ vaihtaja ‘exchanger’

var(h)ainen, vara(z) ‘early’ vika, vaiva ‘trouble, illness,’ kipu ‘pain’ yö ‘night’

buru5, bir5 ‘flock of birds, locust swarm’ ga.eš(8) ‘agent, merchant’

gar4 ‘early’

ge6.ge6(d) ‘dark, black; to darken’

ge6(.b/g) ‘night’

pime(d)ä ‘dark,’ pimet- ‘to darken’

varo-, varat‘to fear, beware’ E kere ‘body’

bar ‘to fear, beware’ bar ‘body, corpse’

gig.(ga), gib ‘(to be) sick; trouble, pain’

Finnic

Sumerian

iggjâ, jįjjɛ̮, jija ‘night’

vikke, vaive ‘trouble, illness’

parva, parve ‘flock of birds’ vajeɔtte- ‘to exchange’

gorod, gorog ‘body, carcass’

Lappish

M čopuda, C p3ckätä ‘dark’

M ve, vä, vej ‘night’

M vel’e, C ul’e, β3̑l’a ‘beehive’ C βaštalte‘to change’

Volgaic

U vekà- ‘to be ill, suffer,’ kajge ‘trouble’ U uj, oj, ɯj, Z voj, oj ‘night’ U pejmȋt, Z pemįd ‘dark,’ pejmȋ‘darken’

U vošjà-, Z vež‘to change’

Permian

O jĕj, ĕj, V jī, jiγ ‘night’ O p3γt3, p3γ3tɜ, V piti, pit ‘black’

O weχ, wew ‘hardship’

sereg ‘crowd, flock, host, troop’

O χărĭ‘to fear, beware’ O χŏr, V koar, kwar ‘image, form, figure’ O păr3, pŏr, parr3̑ ‘flock, shoal, herd’

fekete ‘dark’ befeketít‘to blacken’

éj ‘night’

fáj ‘to ache’ baj ‘trouble’

kora ‘early’

Hungarian

Ob-Ugrian

peββā‘to be dark’

pī, pʿį ‘night’

bok3 ‘pain’

β’eʌ́ʌɯ ́ ‘flock of birds’

pim(pi)-, ‘to be dark’

pi, pö ‘night’

kekk3 ‘illness, pain’

keru, kɛ̮r, kēr3 ‘flock of birds’

Samoyed Y Km Slk

gìr, kiri5, miri ‘foot’ juolge ‘foot’

guhkki ‘long’

pitkä ‘long, tall’

jalka, ǵaug ‘foot’

kuojjɒ, kįjj(a) ‘man’ ḱiesse‘to pull’

oskeldi‘to dare’

uskalta‘to dare’

hui, huja ‘penis’ vetä- ‘to pull, draw, drag’

buoččâ ‘penis’

vehje ‘gadget, penis’ uskomus ‘belief ’ askel ‘step’

gèš, uš ‘penis, male’ geškim ‘trust, belief ’ gešgal ‘stand, step’ geškim.ti ‘to trust, dare’

gi7, ge15 ‘male, man’ gíd ‘to pull, draw, drag, lengthen’ gíd.da ‘long; length’ gigir ‘chariot’

pėȧrȧ(ǯ) ‘mayfly’

Lappish

perho, perhos‘butterfly’

Finnic

gereš, giriš ‘butterfly’

Sumerian

M ved’a-, C βiδe- ‘to lead, bring’ M k(u)vaka, C kužka id. ḱeviŕ ‘cart, chariot’ M ṕil’ǵe, C jal, jol ‘foot’

C koγerš3 ‘flying moth’ C uzo, uźo, uz3̑ ‘male’ U vaśkams ‘trusting’ M eśkȇlks, C ašk3̑l M eśkil’da‘to step’

Volgaic

U ťigil,’ Z ge̮gįl ‘wheel’ kuk ‘foot’

U gogo, Z kaj ‘penis’ Z votjį‘to pull, draw’ kuźa ‘long’

U kiźan, Z kuzj ‘penis’ U oskon ‘trust’ U ućkįl, Z (v)ośkol ‘step’ Z veśkal‘to head to’

Z tiri-bob3̑ ‘butterfly’

Permian

V kor, kȯr̮ ‘foot’

O χoj ‘man,’ kuj, χɯ̆j ‘male’ O tät-, täϑ-, täl- ‘to pull, draw, drag’ O p3t, V koåś3 ‘long’

V ūs3l, ūsil ‘step’ O wosk3̑l ‘to leap’

O χaχ3̑rsȧ ‘(flying) moth’ O ĕč3, V wäś3j ‘penis’

Ob-Ugrian

szekér ‘cart, wagon’ gyalog ‘on foot’

húz‘to pull, draw, drag’ hosszú ‘long’

húz- ‘to pull, draw, drag’

fasz ‘penis’ hős, hés ‘hero’ bizalom ‘trust, belief ’

kérész ‘mayfly’

Hungarian

βe̮ttā ‘to lengthen’

mir, m3̑r ‘foot’

kütt3- ‘to draw’

āsel‘to step over’

Samoyed Y Km Slk

kieri-, vieri‘to roll’

verka ‘broadcloth’ vihi ‘insight’

pätsi ‘oven, furnace, heat’ metsä ‘forest, wood, logs’

vehje ‘gadget; penis’ kuoho ‘gelded animal’ vuol‘to carve, whittle’

gir5, gígri, diri ‘to roll’

gir(i)5 ‘cloth of ordinary quality’ giš, uš4 ‘ear, understanding’

giš, mu ‘fire’

gìš, uš ‘penis; male, man’

gul ‘to carve, hew’

gu4 ‘ox’

giš, muš ‘tree, wood, timber’

kireä ‘severe,’ vireä ‘strong’

Finnic

gìr.ra, ir9.ra ‘strong’

Sumerian

vuollâ‘to carve’

buoččâ ‘penis’

pieɔ̄ts ‘oven, furnace’ mæcce ‘forest, woods’

gârrâ ‘stern, severe’ vierrâ‘to roll’

Lappish

C pisa ‘penis,’ uźo ‘male’ C βusko ‘gelded’

M ḱeviŕ(ď), ḱiḱiŕ(ď)aid. C β3̑rγem ‘dress’ C ȇš, uš ‘wit, reason’ M ṕiśi ‘heat’

Volgaic

Z koďź/ś‘to geld’ ve̮lį-, Z volō‘to carve, whittle’

U pis ‘wood, log, prop,’ Z pes ‘wood,’ mįǯ ‘prop’ Z kuz, kuzj ‘penis’

U viź ‘wisdom, memory, wit’ U Z pe̮ś ‘heat,’ Z pać ‘oven’

U tigil’a-, Z gḛgil’t‘to roll’

Permian

O ĕč3, V üć3w, wośiγ wäś3j ‘penis’ O kowti(-) ‘geld(ed)’ wolt-, βalt‘to carve, whittle’

m3š, m3č, mȧš, mȧč ‘prop’

O ŏš, V wuz ‘memory, wit’

O kär3j-, käri-, V kir‘to roll’

Ob-Ugrian

fasz ‘penis,’ hős, hés ‘hero’

göcs ‘chunk of wood, log’

ész ‘intellect, reason, wit’

bír ‘to have strength’ pereg ‘to roll’

Hungarian

mać3, māťč3, màǯ ‘woods’ ke̮ć3, kuȯť ‘worker, knight’

ku3̑č3 ‘heat’

Samoyed Y Km Slk

varis, vares, variks- ‘crow’

korjat‘to collect, reap, remove’ kiire(h), K kiiru ‘scalp, top,’ kero ‘fell top’ kiire, kiiru(s) ‘hurry’ varoi ‘crow’

gur10/11, búr, ur4 ‘to collect, remove, reap’ gur5.ru(.uš), búr.ru(.uš) ‘scalp’

gur5.(ru.)uš ‘hurried(ly)’ gurux, buru4 ‘crow, raven’

pierra ‘top, whirligig’

kurra ‘disc,’ pyörä ‘wheel’ pyöri-, vyöry‘to rotate, roll down’

gúr, bùr ‘hoop, disc’ gur4(-gur4) ‘to roll, rotate’

vuorâš ‘crow’

ḱierre ‘scalp,’ čærro ‘fell top’ kiiras, kįras ‘hurry’

bârde-, pordte‘to collect’

buna (name of red mottled cow)

Lappish

puna ‘red, redness’

Finnic

gùn, u.gùn ‘colourful, red; mottled’

Sumerian pona ‘hair, feather, (M) colour’ M k3r3jot ‘circle’ M puvŕa-, C pörta- ‘to roll, rotate’ M p3̑rda‘to collect, rake, put aside’ M seŕc ‘scalp,’ C βar3̑š ‘ridge’ M kur3ksta ‘quickly’ M varaka, varśi ‘crow’ C βara(k)š ‘hawk’

Volgaic

U Z varįš ‘hawk’

U gureź, Z kere̮s ‘hill, mountain’ Z perįt ‘quick’

O war3̑s ‘hawk’

O warγaj ‘crow’

O kèras, V kēräs ‘ridge, hogback’

O kör3k ‘circle’ O peŋ3rγ3-, V po/üwart‘to roll, rotate’

O pun, ‘hair, plumage; colour,’ O k̔ŏn ‘hair’

U gon, Z ge̮n ‘hair, plumage’

U kurj ‘ring,’ Z kor ‘circle’ U bergà-, Z bergal‘to rotate’

Ob-Ugrian

Permian

varjú ‘crow’

bérc ‘rock, mountain, hill’

kör ‘circle,’ kerék ‘wheel’ forog, pörög ‘to roll’

fan, fon ‘(pubic) hair’

Hungarian

warŋäe, bare id.

ku3rus ‘hawk’

ku3rä ‘crow’

Samoyed Y Km Slk

kin, kí× ‘work, task’

kaš, kas ‘beer’

¶a.lච‘to startle’ igi ‘front side’

guz ‘to limp, halt’

guruš ‘man, male’ guz, ¶uz ‘to cut, trim, clip’

guru7 ‘grain heap’ buru14 ‘harvest’ gurud, kuruda ‘to beat, whip’

Sumerian

kalja, kahja ‘week beer’ tinki ‘daily task, day work, job’

konti‘to crawl, creep’ havahtu-, kavahta- id. esi ‘front,’ etelä ‘south’

kurita- ‘to beat, whip, castigate’ uros, koiras ‘male’ veistä- ‘to cut, whittle, chip’

korjuu ‘harvest’

Finnic

vāssâ ‘weak beer’

guanto‘to walk limpingly’

ores, vâres ‘male’

gurit- ‘to discipline’

Lappish

M iḱel’ ‘front side, porch’ kuvas, kvas ‘weak beer’

M ḱeŕa-, C k3reδala- ‘to beat, whip’ M ur3̑ś ‘boar’ M końźeďe‘to cut, clip’ M kuźi-, C küze-, kuče‘to climb’

Volgaic

U śu-kaś ‘rye beer, kvas’

O wȧŋ- ‘to crawl, slog, trudge’

U gȋžȋltȋ‘to go slowly, crawl’ U ku̯al’ektį‘to startle’

O pùsa ‘beer’

O jok, jako, joχjil ‘south’

O kŏr-, χŏr‘to beat, whip’ O kǡr3s ‘adult elk’ O χăn-, kăn‘to cut’

Ob-Ugrian

Z bordjį‘to beat, castigate’ Z vere̮s ‘man,’ kįre̮s ‘he-cat’ U vandȋ-, Z vundį- ‘to cut’

Permian

munka ‘work, labour, task’

dél ‘south’

kúszik‘to climb, crawl’

béres ‘farm-hand’ kaszál‘to cut, mow’

korrigál‘to castigate’

Hungarian

kūās ‘beer’

Samoyed Y Km Slk

6

kur.gimušen ‘crane’

kur9, ku4(r) ‘to enter’

kúr ‘strange(r)’ gir5 ‘guest’ kiraš ‘slave’ kúr- ‘to change (itr.), become different, alter’ kur4 ‘high, eminent’

6

viero, vieras ‘strange(r), guest’ pyörty‘to faint, get confused’ korkea ‘high, eminent’ pyrki‘to aim, try to enter’ kurki ‘crane’

kinnat-, kingat‘to exert o. s., strive,’ tinki- ‘to strive, try’ kiiski, piski ‘ruff ’ katko- ‘to cut, break (off )’ vuori ‘mountain’

kin, kí× ‘to strive, try, exert o. s.’

kiši5ku , péšku ‘a small fish’ kud ‘to cut, break (off )’ kur ‘mountain’

Finnic

Sumerian

guorgâ ‘crane’

bârgâ‘to try’

guorgâd ‘haughty’

vierrâ id., kårrå ‘ridge’ vieres ‘strange’

Lappish

M kargo(s), C turńa id.

C pure‘to enter’

M seŕej ‘high’

C kur3̑k, kuruk ‘hill’

M pok3̑ď3‘to work,’ C peŋešte‘to labour, exert o. s.’

Volgaic

U turj, Z tuŕi, turig ‘crane’

U Z pįr‘to enter’

U śe̮r, sör ‘alien, strange’ Z pe̮r- ‘to change (itr.),’ pe̮rt- ‘to alter’ Z gįrįś ‘high, tall’

U ku̯aćkàt‘to break’ U vȋr, vįr, Z kįr ‘hill’

Permian

O tōrɜ id., V tar3w, tāriγ

O k3r3š, V kör3s ‘high’

O wur, ur, ür, V wor, wur id. O 3̑r3̑ŋ, pēr3χ, pēr3γ ‘strange’

O kŏse, ko̮ṡ e ‘a small fish’

O täk‘to strain, exert’

Ob-Ugrian

daru ‘crane’

fér- ‘to enter’

orr ‘nose, peak, tip’

küsz ‘a small fish’

Hungarian

χārū, kuro id.

pɯrže, M hirge id.

bōr ‘ridge’

kara, kʿåra id.

śēr‘to enter’

pirk3, pürg id.

kūra ‘ridge’

kā(s)sa ‘ruff ’

Samoyed Y Km Slk

karju-, kalju-, kilju- ‘to bawl, yell, bellow’

suukko ‘kiss’

kúš.ù ‘to tire’

kušum(4) ‘to cry, moan, bray’

sub(-dug4) ‘(to) kiss’

kuras ‘knife, baton’ köyhä, keuh ‘poor, needy’

va(a)ksa, voakša ‘span’ väsy- ‘to tire’

kùš ‘cubit’

šag4, šab ‘heart, inside; embryo’ ugur ‘sword, mace’ úku(.r), úku.rá ‘poor, needy’

pyrky ‘aim, endeavour’ tyrsky, hyrsky ‘breaker’ vaski ‘copper, bronze, brass’

Finnic

kur.ku ‘desire, aim’ kur.ku ‘wave, flood’ kù.sig17, guškin ‘gold’

Sumerian

covkte, cuwke- ‘to kiss, smack’ sajge ‘carrion, carcass’ korr ‘knife’

vuokse ‘span’ viessâ‘to tire’ giljo-, kel’l’ɒ- ‘to yell, bray’

veški ‘copper’

Lappish

M tor(o) ‘sword’ M kamaža ‘poor’

M tolkun ‘wave’ M uśke ‘brass,’ C waž ‘ore’ C β3̑jakš ‘cubit’ M śiźe-, śiźi‘to tire’ M parće‘to bellow, bleat, moan’ C šupšala‘to kiss’

Volgaic

U viś, Z veś ‘span’ viś(į)- ‘to be ill, sick’ U keśkȋ‘to cry,’ Z gerjav-, gorzį- ‘to cry’ ćup (karį-) ‘(to) kiss, smack’ U še̮j ‘carrion,’ Z šoj ‘stillborn child’ U port, Z purt ‘knife’ U kɯes ‘bad, poor’

U tulkįm ‘wave, flood’ U az-veś, Z ez-iś ‘silver’

Permian

O ťŏp3̑l-, V śoχ3l‘to kiss’ O t3γ-, ϑ3γ-, j3γ-, t3γpɜ ‘inside’ kùr ‘cooper’s adze’

O mōž3m- ‘to fall ill, sick’ O χeŕ(3)ť‘to roar, bellow, yell’

U korskį‘to splash’ O waγ ‘iron,’ V küš ‘lead,’ woχ ‘copper’

Ob-Ugrian

kard ‘sword, sabre’ kevés ‘slight, scarce, scant’

ordít- ‘to bawl, bellow, bray, cry, howl, yell’ csók ‘kiss,’ csókol ‘to kiss’ szív ‘heart’

vas ‘iron’

Hungarian

śei, śēj, T sa, soa ‘heart’ χar, kura ‘knife’

ẃeśe, båzå ‘iron’

ćokk3mpi ‘to smack’

ku3̑s, kwöze ‘iron’

Samoyed Y Km Slk

oacce ‘shelter’

piha ‘(fenced) yard, mansion, home, village’

säke- ‘to hide, conceal, bury, mislead’

uru, iri ‘(walled) city, town’

zá¶, sa¶6 ‘to hide, bury, run away, steal’

duru5 ‘(to be) wet, fresh, soft’ eme ‘tongue’

Sumerian dur ‘cord, strap, band’

Finnish nuora ‘cord, rope, string’ tuore(h) ‘wet, moist; fresh, raw’ nielu ‘pharynx’ M narvaś id., C nöre‘to become wet’ M ńil’ḿe ‘pharynx,’ C j3̑lm3 ‘tongue’

ńur, ńür ‘moist,’ Z ńure̮ś ‘wet’

Permian ńe̮r ‘twig’

U sog(ȋ)-, Z ʒ́eb- ‘to hide, bury’

Z voǯ ‘weir, pilework’

njuorâs ‘soft’

njalbme, ńälme ‘mouth’

Permian Z ve̮ń, voń ‘belt’

Volgaic

C ümbal ‘horseback’ M oš ‘city,’ C βiča, βiťa, βüča, βüťa ‘fenced yard’ M sopo-, soko- ‘to embezzle’

Volgaic

Lappish nuorra, nūrre ‘cord, string’

Appendix 16. Words with PU palatal nasal

čiekkâ- ‘to hide, bury, conceal’

væmbel ‘hoop’

Lappish

vempele ‘collar bow’

Finnic

umbin ‘collar bow; wheel’

giš

Sumerian

O ńäl3m, V ńil3m ‘tongue’

O ńārɜ ‘moist, raw’

nyelv, nyév, nyőve ‘tongue’

nyers ‘fresh, raw’

Samoyed

sįpk3l‘to hide’

keč, ku3̑čč3 ‘city, town’

Y ńērʾʾE ‘juice,’ Slk ńār, ńεr ‘moor’ Y ńālā‘to swallow’

takkal-, šabdol‘to hide’

βāttie ‘fence, enclosure’

Samoyed Y Km Slk

Hungarian

sikkaszt- ‘to embezzle’

gömb ‘ball’

Hungarian

Ob-Ugrian ńur, nür, ńir ‘cord, strap’

O čä(w)γi-, săwĭ- ‘to hide, bury’

O woš, wos, V ūs, ōš, wūš ‘city, fortress, enclosure’

O ḱempar ‘hoop’

Ob-Ugrian

tìl, ti.la ‘arrow’ tìl ‘to live’ ti.la ‘alive’

urud

nir ‘to pour out, impregnate’ ti8 ‘eagle’

nuoli ‘arrow’ elä- ‘to live,’ elävä, eläjä ‘alive’

miniä, mīńa ‘daughterin-law, bride’ nika/erta‘to carpenter’ noru-, niru‘to trickle, ooze, flow’

näke-, ńägi‘to see,’ näkö ‘eyes, E face’ neljä, ńel’l’ä

igi ‘to see; eye; face; witness’

limma, limmu ‘four’ mí ‘(young) woman, bride’ nagar ‘carpenter’

Finnish imettäjä ‘wet nurse’

Sumerian emeda ‘wet nurse’

njuollâ, ńįll ‘arrow’ jielle- ‘to live,’ jiellej ‘living’

a

njorrâ‘to pour out, trickle’

Lappish njâmât-, ńįmɔtte‘to nurse’ ńieGka-, nieγγē‘to see a dream’ njælljĕ/ă, ńiel’l’(e) mânnje, mańń ‘daughter-inlaw’

M nal, C nölö ‘arrow’ M eŕa-, C ile- ‘to live’

nuŕge‘to drip, trickle, ooze, flow’

M ńil’e, ńil’ä, C n3̑l, nil’

ńeje-, ńiji‘to see,’ ńeji ‘witness’

Volgaic

U ńow, ńil, Z ńįl, ńe̮v id. U ulį-, ul-, Z ol-, ov‘to live’

O ni-, nīw-, nīwγ-, nīwχ‘to be visible’

U aʒ́ȋ-, aďďjȋ-, Z aʒ́į-, aʒ́‘to see’ U ńɯl’, ńįl’ Z ńol U meń, Z mon/uń id.

O ńal, ńoʌ, V ńēl, ńāl id. O jĕl3w ‘fresh, alive,’ V jelt‘to revive’

O ńĕl3, ńil V ńil’ī, ńil3/a O meń, V miń, mäń ‘daughter-inlaw’ ń3γ3r-, ńŏγ3̑r-, ńŏŋγ3̑r‘to carpenter’ O ńaχ3̑r‘to pour out, make flow’

Ob-Ugrian O ēm3tt3‘to nurse’

Permian Z ńimet‘to nurse’

él ‘to live,’ élő ‘living, alive’

nyíl ‘arrow’

meny ‘daughter-inlaw’

négy

néz‘to look, watch’

Hungarian em- ‘to suck’

Slk ńi, ńī, Km ńie ‘arrow’ Y jīl’e-, T ńile‘to live,’ Km ďili ‘alive’

Km ńī ‘eagle’

Km ťåγār3l‘to carve’

Y ćēt, ťet, Slk tēt3 Y ḿeje, Yn mē, mī id.

Samoyed Y ŋaḿā-, Slk ńem-, ńim‘to suck’

Finnish nuori ‘young,’ norja ‘soft, supple’ uni, uńi, uń ‘sleep’

Lappish nuorrâ, nūrr(a) ‘young,’ njuorâs ‘soft’

Volgaic M ćora, ťora ‘youngster,’ C nörγ3 ‘young, fresh, supple’ M on ‘dream,’ C omo ‘sleep’

Permian U ńored ‘thin, weak,’ Z tor ‘little (boy)’ U um, un, Z on ‘sleep’

Ob-Ugrian O ń3r ‘thin,’ ńōr3χ ‘young,’ V ńaur ‘child, young’

Pictogram ZATU 037

ZATU 250

ZATU 614

No. 001

002

003

bulu¶, buru8(¶) dili.dili, didli màn, min5 mug

díš ge15, gì géš mákaš ¶al

Sum. reading aš dili

Meaning ‘one’ ‘single, solitary, perfect’ ‘one’ ‘one’ ‘one’ ‘clamor, uproar’ ‘to split, divide’ ‘legs’ ‘secret’ ‘to meander’ ‘to pour out’ ‘to throw up’ ‘all; completely’ ‘two’ ‘ergot, dregs, tow’

Hungarian nyurga ‘thin, slender, lean’

Slk āŋG, ŋg3, āŋku ‘sleep’

Samoyed Km tʿōri ‘small fish’

Uralic etymology F yksi, L âktâ, akt, e̮χt, Z e̮k, e̮ťi, H egy ‘one’ M ćela ‘whole, intact,’ U čolak ‘solitary, lone,’ Z de̮la ‘whole,’ H teljes ‘whole, perfect’ F täysi ‘perfect,’ L tīvt, C tić, tič, tiť ‘full, whole’ F ken, M ḱi, C kö, U Z kin, H ki ‘anyone who’ E kes, kis ‘anyone who’ F mekastus ‘clamor,’ O moχleś- ‘to be noisy’ F hala-, halko-, L čuollâ- ‘to split, divide’ F haara, saara ‘legs (pl.),’ M śeŕk, śäŕka ‘leg’ F sala, M salava ‘secret,’ Z će̮l-će̮l ‘secretly’ F häily-, M čara-, O sāʌi- ‘to waver’ F holjotta-, O sŏʌγ3ʌt3̑- ‘to pour out’ F pulauta-, C βark3̑ne- ‘to throw up,’ O pŏl3̑χ- ‘to gush out’ M doloj, U dol-dol, Z de̮la, O ťel ‘all; wholly’ F moni, U Z mįn ‘many,’ L månah ‘a couple, two’ F L moska ‘trash, dregs, mash,’ C muš ‘hemp fibre, tow’

Appendix 17. Signs 1–25 of the Sumerian syllabary (Ellermeier 1979–1980:577–584)

ù ‘sleep’

Sumerian tur ‘young; little; weak; child’

ZATU 040

ZATU 485 ZATU 485

ZATU 521

ZATU 046

ZATU 218

006 007

008

009

010

Pictogram

005

No.

ád gír, gíri

bala

bal

zu su, sug6 kuš dur10 šen

ba

Sum. reading

Meaning ‘broadcloth’ ‘to open, gape’ ‘to trim, carve, cut’ ‘to divide’ ‘to know’ ‘skin, flesh; kin’ ‘skin, leather’ ‘axe, hatchet’ ‘cauldron’ ‘clean’ ‘verdigris’ ‘battle’ ‘to pour, libate’ ‘to dig, heap up’ ‘speech, talk’ ‘to curse’ ‘reign, rule’ ‘to turn back’ ‘to (ex)change’ ‘crossbeam’ ‘spindle’ ‘(box)thorn, prick’ ‘sword, dagger’ ‘lightning, flash’ ‘to flash’ ‘scorpion’

Uralic etymology F verka, M muška, C muš ‘broadcloth’ F ava-, auko-, O aŋ3-, V āŋoko- ‘to open, cut open’ O waγ-, waŋk-, V waŋk-, H vág- ‘to hew, trim, cut, carve’ F jaka-, M java-, U Z juk- ‘to divide’ F tut-, M soda-, U tod(ȋ)-, Z te̮d-, H tud- ‘to know’ F suku, L sokkâ ‘kin,’ O sŏχ, sŏγ ‘skin, hide,’ V sow ‘skin’ F kesi, L ku3ššE, M keď, U Z ku, V kot, koät, T kuhu ‘skin’ U tir, Z ćer ‘axe,’ F terä, C tür ‘blade’ F hinkki, hinkka ‘metal milk can’ F hyvä, Z šań ‘good,’ C šeńä ‘fine,’ H csinos ‘clean, pretty’ F hime ‘tarnish,’ M čeḿeń ‘verdigris,’ Z sim, O sämĭ ‘rust’ F hymy ‘din,’ M C šum ‘feud,’ U čeŋešį- ‘to quarrel’ F vala-, L vǡlla-, M valo- ‘to pour,’ C βele- ‘to pour, spill’ M val’a- ‘to dig, fill up’ F vala ‘oath,’ M val ‘word,’ Y βāδa ‘word, speech’ L vŭŏllē- ‘to curse, damn,’ O wēl3k- ‘to abuse, scold’ F valta, voli ‘power, might, rule,’ M vol’a ‘freedom, power’ F palaja-, M vel’a- ‘to turn back’ M polavto-, O pil3γt3-, H vált- ‘to (ex)change’ F palku ‘crossbar,’ L balka ‘beam,’ M ṕal, Y βāl’ ‘pole’ F paalakka, L bāllâk, M palka ‘handloom stick’ F ota ‘prick,’ F C ora ‘prick, thorn, awl,’ M ura, H ár ‘awl’ F kärki ‘point, tip,’ K ‘(knife) blade,’ C kerγ3 ‘sword’ E kirg ‘flame, blaze’ L gârge- ‘to flash, sparkle,’ M veŕgeďe- ‘to flash’ F kirppu ‘flea’

ZATU 064

ZATU 549

012

Pictogram

011

No.

¶aš kud, gudu4 kurum sil sila, sulu tar

búlug du9, dun5 sun5 ušu, ušum ušum

búr

Sum. reading ul4

Meaning ‘to shine, gleam’ ‘to be quick’ ‘field, pasture’ ‘to undo, explain’ ‘to blow’ ‘sack, sail (?)’ ‘to break’ ‘to glow’ ‘to hasten’ ‘authority’ ‘to cut wood’ ‘to dash, prowl’ ‘to be humble’ ‘snake, worm’ ‘leader’ ‘alone’ ‘to cut (off ), break’ ‘to cut, break’ ‘to cut’ ‘to split, cut up’ ‘street, alley’ ‘to cut, incise’ ‘to ask, examine’ ‘reliable, wise’ ‘to tremble, shake’ ‘to strike, lash’

Uralic etymology F välä-, väly-, O wal’i-, V wal’γ- ‘to shine,’ H villog ‘to glint’ F väleä, E väle ‘quick’ F väljä ‘clearing, pasture,’ E väli ‘field’ F purka- ‘to undo, loosen, dissolve, unravel’ F purju-, C purγ3̑kta- ‘to blow, gust,’ U purʒ́į- ‘to whirl’ F purje, L borjâs, M C parus, O paru̯is, pōr3š ‘sail’ F porotta- ‘to break,’ C purγeδ- ‘hack,’ O pɒr3χ- ‘crumble’ F porotta- ‘to glow, boil,’ U bįre̮ktį-, M pura-, H forr- ‘to boil’ F porotta-, porhalta- ‘to go fast, dash’ F porho ‘magnate’ F pilkko-, L pelka-, U pil’(j)-, Z pel’ke̮d-, O pelkit3- ‘to split’ F rynni- ‘to rush, storm, jostle’ K sońńistu-, M šuŋg3̑d’e-, Z śin- ‘to bend,’ C śuke- ‘to bow’ E uss, L ūśka, ūšk, C už ‘snake’ F isoin ‘greatest,’ M oćuv ‘chief, leader, greatest’ F yksin, M iśkam, C ikt3n, Z e̮tna-, O ātemna- ‘alone’ O k̔ȯs- ‘to rip off, break,’ V χusät-, H hasít- ‘to split, rend’ F katko- ‘to cut, break,’ U ku̯aćkà(t)- ‘to break’ F kurmita- ‘to cut, notch’ F säli-, M sal’a-, C šel’a-, Z će̮l-, V sil-, H szel ‘to split, cut’ F sola ‘alley, passage, pass’ F tärkät-, E tärki- ‘to cut, notch,’ Z torjed-, O tart- ‘to cut’ F tarkasta- ‘to check, examine,’ C terγe- ‘to examine, ask’ F tarkka ‘diligent, wise,’ H derék ‘honest, brave, good’ F täris-, tärä-, Z tarźi-, O tar3̑γ-, tari-, V torg- ‘to tremble’ M ďorkaďe- ‘to strike,’ U targà- ‘to whip,’ Z tarke̮d- ‘to flap’

ZATU 271

–

015

016

zú, zúg zu¶ gèle, gìli, méle mu7 mumun se18, šegx sug7, zug4

kir4

ka, kág, kák

e7, en4, i5, inim giri17 gù

dingir, dimir dag5 dug4

an, am6

Sum. reading

Meaning ‘to dig, smoke’ ‘sky, heaven, god of heaven, rain’ ‘god’ ‘to burn of desire’ ‘to speak, tell, recite’ ‘to do, make’ ‘word’ ‘to be radiant’ ‘call, noise’ ‘thunder’ ‘utterance, rumor’ ‘mouth, pit’ ‘nose, snout’ ‘hyena’ ‘tooth, spike’ ‘to steal’ ‘uvula, gullet’ ‘to rumble, mutter’ ‘incantation’ ‘noise, voice, cry’ ‘opening, pharynx’

Uralic etymology M targa- ‘to dig up, smoke,’ H túr ‘to dig’ Fc juma ‘god,’ ilma, ium ‘weather,’ C jumo ‘god, heaven; thunder,’ U in, im, Z jen ‘god, sky, weather,’ O jŏm ‘rain’ Fc jumala, ďumal, L jimmel, U ińmar, immar ‘god’ F tahto-, L datto- ‘to want, wish, desire’ F luke-, L lokkâ-, M lovo-, U lįďʒ́́į-, lįgʒ́́į-, Z lįďďį-, O tŏŋ3t-, lŏŋ3t‘to count, read, recite’ F teke-, L dâkkâ-, M ťeje- ‘to do, make’ F ääni ‘sound, vote,’ L jiednâ, jänä ‘voice,’ H ének ‘song’ E kirgat- ‘to shine, be radiant’ F kohu, M kuš, C kuž, U ku̯až, Z guž ‘din, noise, roar’ F ukko, E uku, O ŏwγ/χ ‘thunder’16 F kaiku, L goaje, Z govk ‘echo, sound,’ U kɯi ‘melody’ F kaivo, kaivu, U kɯjvu, V ke̮je̮, kojo ‘pit, well, excavation’ F kirsu, kärsä, L gærsse ‘snout’ F koira, L goairre, Z kįr ‘dog’ F sokka, K čoga, H szeg ‘peg, pin, spike,’ M suka ‘plough’ M soko-, H sikkast- ‘to embezzle, misappropriate’ F kieli ‘tongue,’ nielu ‘gullet,’ U kȋl ‘tongue,’ ńįlon ‘throat’ F mumis- ‘to mumble, mutter’ F mumina ‘mumbling’ L čuojâ, Z šį, O s3j, V soj, sī, H zaj, szaj ‘sound, noise’ F suu, H száj ‘mouth,’ L čuwwa, S śō, soj ‘throat, mouth’

16 Note gu-u GÙ = rigmu ša ili ‘shout of the god’ (i. e., ‘thunder’) in Nabnitu B:202. gù ‘shout’ apparently goes back to earlier ú.gù, cf. gù.dé.a = šasû ‘to shout’ (Izi E:316) and u-[ug?] UD = rigmu (A III/3:14).

ZATU 031

Pictogram

013

No.

– –

– –

– – ZATU 060

018 018*

019 022

024 024* 025

Sum. reading tu6 uš7 u¶4 uš11 nundun su6, sùl, sum4, sun4 kana6 mùrgu úrgu še25, šed14 musug5, uzug5 sub sub ‘to be gloomy’ ‘anger’ ‘anger’ ‘voice, cry, noise’ ‘unclean’ ‘kiss’ ‘to rub, scrub, polish’

Meaning ‘incantation, spell’ ‘spittle, venom’ ‘foam, phlegm’ ‘spittle, venom’ ‘upper lip’ ‘lower lip, beard’

Uralic etymology F taika, L daigâ, M ťeića ‘magic, charm, incantation’ F viha ‘hate, poison,’ L vašše, U vož, Z vež ‘hate, anger’ F va(a)hto, vahti, Vt vāhsi, vāssi, Z bįg ‘foam’ F viha, L vašše, V vož ‘hate, venom’ O nɯm pēl3k tŏr3̑p ‘upper lip’ F huuli ‘lip,’ M śuva ‘beard,’ C šuu̯, šul ‘bristles,’ U Z zu ‘pig’s bristles,’ H szakáll ‘beard, whiskers’ F kaunat- ‘to bear a grudge’ H méreg ‘anger, poison,’ F myrkky ‘poison,’ myreä ‘cross’ H harag ‘anger,’ V χōrj- ‘to get angry,’ F äreä ‘cross’ F soitto ‘call,’ M śed’a-, C šokte- ‘to ring, sound’ M vaćev, vaćej, H mocskos ‘dirty, filthy’ U Z ćup ‘kiss, smack’ F hi(v)o-, M čova-, C šume- ‘to rub, polish, whet’

This appendix is provided in response to the comments of William H. Jacobsen Jr. and Johannes Knoblauch in Zakar 1971:217–219. The Sumerian column contains the most common Sumerian equivalents of the English words, insofar as known; the other columns the suggested etymological equivalents of the Sumerian words in the Uralic languages. In some cases, more than one Sumerian equivalents have been given for the English word. Unless otherwise stated, the meanings of the Uralic words are identical with those of the Sumerian ones. According to Jacobsen,

Appendix 18. Swadesh list for Sumerian and Uralic languages (100 words) (http://en.wikipedia.org/wiki/Template:Swadesh_list_100_plain)

–

Pictogram

017

No.

English ‘I’ ‘you’ ‘we’ ‘this’ ‘that’ ‘who’ ‘what’ ‘not’ ‘all’ ‘many’ ‘one’ ‘two’

Sumerian gá, mà za me(nde) e še a.ba menu, na(m) gú eš aš man, min

Finnish (Fc) mä, minä sä, sinä me(t) ese, śe, še joka mini- (K) koko usea yksi moni ‘many’ me-/mańe/ńi-, nu-

vejke/ikt3 moń ‘yet another’

aktâ månah ‘two’

Volgaic mon/m3n ton/t3ń miń/mä/e eśe/seδe

juoka mi nipuk

Lappish mon don mī

og/uk, ȋk mįn-/mįn-, mun ‘many’

ma/mįj no-/ne̮m-

Permian mon/me ton/te(n) mi eso/sije̮

mĕj/mεnem/nē(m) χŏś os ‘still’ e̮k, e̮t/3j, ĭt

Ob-Ugrian mä(n)/äm naŋ meŋ/min iťi, śĭ

egy mennyi ‘how many’

mi ne, nem

Hungarian én te(n) mi e-, i-

mī ńī-, ńo, ńon

Samoyed man tan men(de) ešē, sete

Despite the enormous time distance separating Sumerian from the present-day Uralic languages, there are at least 10 accurate CVC agreements in the list, and many more if systematic sound changes are disregarded. In addition, there are several bisyllabic words showing a good agreement. The 111 Sumerian words included have 103 etymological counterparts in Finnish, 63 in Lappish, 54 in Volgaic, 57 in Permian, 66 in Ob-Ugrian, 47 in Hungarian and 35 in Samoyed.

“Swadesh (1954:315–321) calculated that with a test list of 97 words, four consonant-vowel-consonant (CVC) agreements would make a historical connection of two languages highly probable, and even three agreements would be almost satisfactory. Cowan (1962) calculated certain factors in more detail, and showed that only three CVC agreements brought one to a high confidence level. We now have Bender’s important empirical study (1969) involving comparisons among 21 presumably unrelated languages. This emphasizes the importance of the criteria for phonetic resemblance, and shows that with varying degrees of narrowness of criteria for CVC resemblances, it takes for two to seven agreeing items to make a historical relationship highly probable.”

ur u¶ geš ‘tree, wood’ numun ? úr bar kuš uzu uri4, ùri

‘dog’ ‘louse’ ‘tree’

‘fat’ (noun)

‘bone’

‘seed’ ‘leaf ’ ‘root’ ‘bark’ ‘skin’ ‘meat’ ‘blood’

gèr.pad.rá ‘broken leg’ ì, ú, u5

guruš mušen

‘man’ ‘bird’

‘woman’

Sumerian gal, gu.la gíd.da tur ‘small, young’ bàn ‘little’ mí ‘bride’ munus/š

English ‘big’ ‘long’ ‘small’

vuoggjâ

barko ku3ššE oaǯǯe vârrâ

juuri parkki kesi osa, oža veri sirpale ‘splinter’ voi

mæcce ‘woods’ sĭĕmmân

vâres piecce ‘grouse’ goairre

guhkki nuorrâ ‘young’ pienna mânnje

Lappish

uros metso ‘grouse’ koira uhku- ‘itch’ metsä ‘woods’ siemen

pieni miniä

Finnish (Fc) koljo, paljo pitkä nerko ‘small’

oj, vaj’, ü, üj

ked’ uća/oso ver/w3r, wür

jur

śiḿeń

ḿeče/muź3̑ ‘grouse’

Volgaic kol’ ‘much’ k(u)vaka ćora, ťora ‘youngster’

ve̮j, vo/vį(j)

vir

woj, wăj, uj

vaj

vér

gyökér bárka

jor/jēko3r

jur ‘stump’ barka ku

kot, koät woš, ūś wĕr, wur/ vür, vįgr

göcs ‘wood, log’ szem ‘grain’

béres

m3š, m3č ‘prop’ sem, śemä

mänčim/ mansin id. χor

meny

hosszú nyurga ‘thin’

p3t/xoåś3 ńōr3χ ‘young’ meń/miń

Hungarian

Ob-Ugrian

pis/pes ‘wood, log’

kįr

vere̮s

kuźa tor ‘little (boy)’ pińal/pońi meń/muń

Permian

kuho w3̑ć3, waǯe βejje

mać3, māťč3 ‘woods’

ḿeje/mī emn3ćä ‘daughter-inlaw’

tʿōri ‘small fish’

Samoyed

nunuz si

kun

dal ? ku6, kua síg sag, šen ‘head, top’ pa ‘head, tip’ geš, geštu, muštu ‘ear, intelligence’ igi

kun ‘nail’ kir4

sug7, zug4 ‘mouth, throat’

‘egg’ ‘horn’

‘tail’

‘feather’ ‘fish’ ‘hair’ ‘head’

‘fingernail’ ‘nose’

‘mouth’

‘eye’

‘ear’

Sumerian izi

English ‘fire’

kynsi kirsu ‘nose, snout’ suu ‘mouth, opening’

näky, näkö ‘(eye)sight’ gâƺƺâ gærsse ‘snout’ čuwwä ‘gullet, throat’ šu ‘opening’

ḱenǯe/küć

ńeje-, ńiji‘to see’

ȇš, uš ‘wit, memory, reason’

bâj ‘top’

pää moštA ‘memory’

ćoŋga ‘island’ ṕe, ṕä

vihi ‘insight,’ muisto ‘memory’

tolka kal/kol

Volgaic eže- ‘to be warm’ mona/munȇ

dolge guolle

Lappish aššo ‘live embers’ mânne asâ ‘weapon’

sulka kala tukka

häntä

muna ase ‘weapon’

Finnish (Fc) ahjo ‘forge’

gižį/giž

aʒ́ȋ- ‘to see’

viź ‘wit, memory, wisdom’

ćuk ‘tuft’ čaŋga ‘hill, top’ puŋ/pon

tįlį/tįl

Permian e̮zjį- ‘to catch fire’

sūp ‘mouth’

künš/könč

ni-, nīw- ‘to be visible’

ŏš/vuz ‘wit, memory, reason’

śoŋk/säŋkw ‘hill’ pŏn3k/päŋ

č3nč/š3š ‘back’ tŏγ3l kul, χul

Ob-Ugrian wiź3t- ‘to catch fire’ mun/mon

száj, szá ‘mouth’

néz- ‘to look, watch’

ész ‘wit, intellect, reason’

fej

toll hal ciki ‘hairy’ ság ‘hill’

mony

Hungarian izzó ‘fiery’

śō, soj ‘throat; mouth’

pε’, feai

ku3̑l3 tūja ‘hair’

sańńe, χańńe, santu ‘tail’

mona

Samoyed ać3- ‘to glow’

gìr, meri dùg(a)

geš ùš ‘womb’

tig ubur šag4, šab

máš

nag

gu7

bur igi

kul ‘to see, learn’

‘foot’ ‘knee’

‘hand’ ‘belly’

‘neck’ ‘breast’ ‘heart’

‘liver’

‘to drink’

‘to eat’

‘to bite’ ‘to see’

‘to hear’

‘tongue’

Sumerian zú, zúg ‘tooth, spike’ eme

English ‘tooth’

kuul- ‘to hear, learn’

kalu-, kalva‘to gnaw’ purnäke-

naukat-

maksa

käsi vatsa ‘belly, womb’ niska udar

jalka, ǵaug luokka ‘bend’

Finnish (Fc) sokka, čoga ‘peg, spike’

gullâ‘to hear’

borrânieγγē-

sajge ‘carrion’ muokse

ńišḱE

giettâ

juolge luoka ‘bend’

Lappish čokk(a) ‘point’

kul’e-/kola‘to hear’

kalga‘to gnaw’ poŕe-, purańeje-, ńiji-

makso/mokš

odar, woδar

ked’/kit, kizä vaće ‘dirt, defecation’

ṕil’ǵe/jal luk ‘bow’

jilm3

Volgaic suka/šoγa ‘plough’

kȋlȋ-/kįl‘to hear’

pur(į)aʒ́ȋ-/aʒ́į-

ve̮ra/vura še̮j/šoj ‘abortion’ mus

ki

kuk/kok

Permian šį, ši ‘spear, spike’

pŏr-/purnīwγ- ‘to be visible’ kol-/χūl- ‘to hear, learn’

laχ äw3r, äγ3r t3γ-, t3γpɜ ‘inside’ mūγ3t/ maj3t, māt

ńäl3m/ ńil3m kor l’oganɜ ~ t’ŏk3n ‘bend’ ket/kǟt waś ‘belly’

Ob-Ugrian

marnéz‘to look’ hall‘to hear’

nyak öböl szív ‘heart’ máj

kéz

nyelv, nyév, nyőve gyalog

Hungarian szeg ‘spike, pin, peg’

ńūk3l- ‘to lick’

śej, šiej, sa ‘heart’ mūδ, mūt, mįt

nŭkku

kākšē

mir, kwälās

Samoyed

u5 ? dal

kul ‘to go, walk, run’

du ‘to come, go’

lug

tuš, suš ‘to sit, dwell’ gub ‘to stand, step’

‘to swim’ ‘to fly’

‘to walk’

‘to come’

‘to lie’

‘to sit’

‘to stand’

ge4

‘to kill’

úš ‘to die’

ku ‘to sleep, lie, lie with’ ug5 ‘to die’

‘to sleep’

‘to die’

Sumerian zu

English ‘to know’

käy‘to go, step’

seis- ‘to stand (still)’

loju-

tul‘to come’

ui-, ujusulja-, solju‘to glide’ kulke- ‘to go, walk, run’

hukku- ‘to drown, die’

koina‘to copulate’

Finnish (Fc) tut-

gǣvva‘to go’

čouǯǯo‘to stay’

tolla‘to come’

golgâ‘to flow’

vuoggjâ-

hokko- ‘to drown’

Lappish dowdâ-

kaje-, keje‘to go’

šińća- ‘to sit down, stay’

tola‘to come’

uje-, ijasul’ka‘to glide’ kolge‘to flow’

kuja-, kuje‘to lie with’

Volgaic soda-

gįv- ‘to go to and fro’

siʒ́- ‘to settle down’

uj-/ujašilgį‘to soar’ kįlt‘to drift’

ǯokà‘to drown’ įš-/voš-, ve̮š‘to die’ vij-/vi-

kujl(į)‘to lie’

Permian tod(ȋ)-/te̮d-

tet-, jet- ‘to sit down’

čūj3t‘to drown’ woš-, os-/ūs‘to die’ w3l-, wet-/ älüj-/(w)ujt3γ3l-/tiγl-, tēl- ‘to fly’ koγ3l‘to go, walk, run’

χŏj-/kuj- ‘to sleep, lie’

Ob-Ugrian toŋ3̑m-

legel‘to graze’

halad ‘to go, run, drift’

uszikszáll- ‘to fly’

vész‘to perish’ öl-

Hungarian tud-

čečči- ‘to rest’

to-, tǖ-, tul’‘to come’

ūtilćo-, tŏlď3‘to fly’ kål ‘to go, walk’

ku3̑t-, ku̯el-

Samoyed tün3m3-, ćeńńeβa-

utu ‘sun’

itu, id4, ì.ti ‘new moon’ mul(u) ‘star’ a eš10 šèg ‘rain; cold, snow’ zá ‘stone, mineral’ kín(.kín) ‘millstone’

ma, me ‘earth, land’ dungu ‘cloud’

‘sun’

‘moon’

‘earth’

‘cloud’

‘stone’

‘rain’

‘water’

‘star’

Sumerian sum, sì e ‘to say, tell’ dug4 ‘to say, tell’

English ‘to give’ ‘to say’

maa ‘earth, soil, land’ sumu ‘fog’

kivi ‘stone, millstone’

jää ‘ice’ vesi sää ‘weather’ savi ‘clay’

valo ‘light’

ehta ‘new moon’

hautoja ‘warmer’

Finnish (Fc) suo-, soovijutel‘to chat’ luke‘to read, tell’

ciekkâ ‘fog’

šaŋŋ, še̮ŋŋ id. čuwje ‘clay’ ḱĭĕδGE ‘stone’

jiegŋâ id.

Lappish suovvâoakte‘to say’ lokkâ‘to count’

suv ‘fog, mist’

valdo ‘light, gleam’ (j)äj/i(j) id. ved’/βüt či, čä ‘sun, weather’ śovoń/šon, šun ‘clay’ kev/kü, küj ‘(mill)stone’

jet, jüt ‘night’

jovta‘to say, tell’ lovo‘to count, read’

Volgaic

mu ‘earth, land’ čįŋ/čȋn ‘smoke’

ke̮, kö/iz-ki ‘millstone’

śuj/śoj ‘clay’

val’(i) ‘light, shine’ jö, ji ‘ice’ vu/va

šundȋ/ šońgej ‘sun’

lįďʒ́į-/ lįďďį‘to recite’

Permian śi-

m3χ, m3γ/ mā, mē id. čiuχ, čuγ/ sēŋoko ‘fog’

jĕŋk ‘water’ wot/wüt’ čōγ3t, šōγ3t ‘snow’ săγį, sŏwĭ ‘hard clay’ kew, köγ/ küw ‘stone’

Ob-Ugrian sai, soi ‘gift’ ńoχ3̑m- ‘to say, speak’ lŏŋ3t‘to count, read’ suŋ(k) ‘sun,’ 3t‘to warm’ ìtài/ēt’, īt ‘evening’

kő, köve‘stone, rock’

jég ‘ice’ víz ég ‘heaven; weather’ ső ‘clay’

világ ‘light’

este ‘evening’

ejt- ‘to pronounce’

Hungarian

mou ‘earth, soil’ śīńńɯ̄, śɯńńɯ ‘fog, mist’

χäi ‘stone’

jīʾ ‘water’ βi’, βit, üt śįka, šįkka ‘snow, rain’ sü, se ‘clay’

ǖtö ‘evening’

Samoyed

‘night’

‘black’

‘white’

‘yellow’

‘green’

‘mountain’ ‘red’

sig7 ‘yellow, green, pale’ bar6, babbar gigi, gíg, gi6.gi6 ge6(.b/g)

kaskal ‘road, travel’ gìri ‘road, path’ kur ¶uš.a, ruš.a ‘reddish’ sari ‘green’

polta-

bíl (tr.)

‘road’

pala-

bar, bar7

‘to burn’

pime(d)ä ‘dark, black’ yö

valkea, valko

sara ‘sedge’

vuori ruskea ‘brown, red’

polku ‘path’

käyskely ‘wandering’

hämärä ‘dim; twilight’

Finnish (Fc)

Sumerian i.bí še.mur ‘ashes’

English ‘smoke’ ‘ashes’

čappâd ‘black’ iggjâ, jija

vielgâd

vierrâ ruškâd ‘reddish’

boalde-

buolle-

Lappish

čopuda/pič ‘dark’ ve, vej, vek-

valdo/βolγȇδo

sar, sare ‘yellow, gray’ śeŋǵere ‘green’

kuruk ravužo/rakš ‘brown’

košt3̑la‘to wander’

pulta-/pelte-

palo-/pela-

śumura ‘dark, dim’

Volgaic

pejmȋt/pemįd ‘dark’ uj, oj/voj

sari, saŕi ‘fallow’ zaŋgar ‘green, blue’ te̮ďȋ, te̮ď

śures, śȋres ‘road, path’ vįr/kįr

pålt‘to melt’

džomit ‘dark, dim’

Permian

p3γ3t/piti ‘black’ jĕj, ĕj/jī, jiγ

sȧŋkɜ ‘light, bright’

wur/wor r3sk3 ‘reddish’

k̒3sk̒3l- ‘to run a little’

păj3t-/pāl’‘to freeze’

Ob-Ugrian

fekete ‘black’ éj

szőke ‘fair, blond, light’ fehér

orr rőt ‘reddish, red, foxy’ zöld ‘green’

homály ‘dim; twilight’ fagy‘to freeze’

Hungarian

päeĭmmĭe, faemei ‘dark’ pi, pö

şarṕṕe, sarp3 ‘road, path’ kūra, bōr

parā‘to burn’ paraδā‘to burn’

Samoyed

dùg, du10.ga

gur4, búr

šuruga

mu ‘name’ nam ‘fate, office’

‘good’

‘round’

‘dry’

‘name’

Collinder 1957 Thomsen 1984

Additional references

‘full’ ‘new’

‘cold’

Sumerian kúm ‘hot, warm; fever’ še4, šed7, šeg4 ‘cold’ diri(.ga) gibil

English ‘warm’

nâmmâ ‘name’

soarve ‘dry pine’

čođđe ‘rime’

Lappish kubmè ‘hot’

l’em/lüm ‘name; soul’

kuŋra, pokaŕav s3r3k ‘dry’

Volgaic kumoka ‘fever’

ńim ‘name, title’

śurį‘to dry’

topa ‘right, proper’ ge̮gre̮s

tȋr(o)/tįr(a)

Permian pįm ‘hot’

nem, näm ‘name’

toŋ(e) ‘good, fair, right’ kör3k, pöŋrök sōr3̑m ‘dry’

sŏj/sol’ ‘hoarfrost’ t3r3m

Ob-Ugrian pom ‘heat, warmth’

Collinder, B. Survey of the Uralic Languages. Stockholm. Thomsen, M.-L. The Sumerian Language (Mesopotamia 10). Copenhagen.

surkea ‘lean, pitiful’ maine ‘fame’ nimi ‘name’

kevelä ‘lissom’ tokeva ‘right, suitable’ pyöreä

hyy ‘cold, frost’

Finnish (Fc) kuuma ‘hot; heat’

név ‘name, reputation’

száras ‘dry’

kerek

hűs ‘cool’

Hungarian hő ‘hot, heat’

nɯm, ńim, n3m ‘name’

šišš3gɛ, ťēćiďe ‘cold’ tīri

Samoyed

Intended Lexical Ambiguity in the Song of Songs Yakov Eidelkind Russian State University for the Humanities, Moscow

I. Introduction By “intended lexical ambiguity” I mean puns based on homonymy, homography and polysemy. Obviously, there are cases when ambiguity is not “intended,” i. e. when the differing interpretations are only due to our imperfect knowledge of Hebrew, ancient literary conventions etc. Nevertheless, it is widely recognized that ambiguity may serve as a literary device in Hebrew poetry. I am aware of the theoretical difficulties involved in talking about a poet’s “intention,” but still I propose to use “intended” as a conventional term. When a text can be read meaningfully in two or more ways, it is reasonable to suppose an “intended” ambiguity. Intended ambiguity may occur on different levels of the text. For instance, on the level of the plot: what does it mean when the Woman in Song 6:2 says: “My beloved has gone down to his garden”? Has he been with her all the time? Then what should we make of her sad story about his disappearance and her looking for him in 5:2–8? Was it a fantasy, a dream? Or rather his coming to the garden is a fantasy (perhaps a trick to mislead the Daughters of Jerusalem who have become too curious about her lover)? Another kind of ambiguity is when an image is used simultaneously in its literal sense and as a metaphor (for instance, “vineyard” in Song 1:6 is both literal and a metaphor for the Woman). In this article I want to treat only cases of ambiguity occurring on the most literal level of the text; hence: “lexical ambiguity.” Part of them are cases of the so-called Janus-parallelism, when “a polyseme parallels what precedes it with one meaning, and what follows it with a different meaning.”1 Janus-parallelism in the Hebrew poetry has received a good deal of scholarly attention in recent decades: see especially Noegel’s monograph Janus Parallelism in the Book of Job (1996) and the literature cited there. In the Song of Songs several examples of in-

1

Gordon 1982:80*.

326

Articles: Old Testament Studies

tended lexical ambiguity (including Janus-parallelism) have been suggested; I shall add a few more. Below (part II) I try to assemble as many instances of intended lexical ambiguity in the Song of Songs as I can find. Discussing them, I pay attention to the relation between the meanings of the ambiguous word. Sometimes they are equally appropriate (cf. example 1), but in other cases there is one primary meaning, while the other is optional (cf. example 4). Especially interesting are examples of “wrong” (contextually inappropriate) meanings, which, however, are suggested by the text and add important nuances to it (examples 3, 8, 11, 14). Then I discuss those examples, suggested by others, which seem to me dubious or incorrect (part III). At the end, two more general questions emerge (parts IV–V). First, how to explain the concentration of intended lexical ambiguities in some parts of the book? Second, was the poetic device perceived aurally or visually; in other words, was the poetry oral or written? II. Instances of intended ambiguity (1)

WhyPiê tAqåyviN>mi ‘ynIqe’Vy' I `!yIY")mi ^yd) ‘ynIa] hr"AÛ xv. ~Øli v'_ W' ry> tAnàB. rd"qê e ylehä \aK' . `hmol) v{ . tA[ßyrIyKi

‘I am black and5 beautiful, O daughters of Jerusalem, like the tents of Kedar, like the curtains of Solomon’ (1:5).

The “curtains of Solomon” cannot be “the curtains Solomon made for the Temple (2 Chr 3:14), dyed dark with blue and purple and crimson,”6 since this curtain is called in Hebrew pāro−e¬. It is also impossible to interpret y3rī"ō¬ as “the rich tapestries hung on the walls of oriental mansions”7 or as curtains of Solomon’s litter (3:7, 9):8 such things are never called y3rī"ā. In prose y3rī"ā means a piece of material for making a tent (Ex 26:1 etc.), but in poetry (and once in prose: 2 Sam 7:21) it is simply a synonym of !ōhel ‘tent’ ( Jer 4:20, 10:20, 49:29, Hab 3:7, Isa 54:2, Ps 104:2; in all of these cases, except Ps 104:2, the word !ōhel is present in the previous parallel colon). A second feature characteristic of the poetic usage is that !ōhel and its synonyms may denote not only temporal dwellings of nomadic tribes, but also normal houses: cf., e. g., Prov 14:11, Job 8:22 and again Jer 4:20. So “the curtains of Solomon” are a poetic way of saying “Solomon’s palace.”9 Then, however, we face another problem. The palace was perhaps proverbially beautiful; but it was hardly black. Many commentators think that each of the two comparisons refers only to one adjective: ‘black—like the tents of Kedar’ and ‘beautiful—like the curtains of Solomon.’10 This interpretation is superficially attractive; but then the point of the comparisons is to add emphasis to both adjectives. However, in the context of the Woman’s 5

The conventional ‘but’ is undoubtedly better (against Pope 1977:311–318; Keel 1992:54; Provan 2001:267), since in 1:6 the Woman obviously apologizes for her dark skin. Of course, her apology and self-effacement turn out to be ironical: she is really proud of her beauty. 6 Goulder 1986. 7 Jastrow 1921:162. 8 Bergant 2001:14. 9 Zakovitch 2004:118. 10 Ibn Ezra, Ewald 1826:57; Döpke 1829:68; Hitzig 1855:18; Ginsburg 1857: 132–133; Delitzsch 1875:26; Oettli 1889:172; Budde 1898:3; Haupt 1902b:221; Meek 1956:105; Schmökel 1956:51; Gordis 1974:79; Krinetzki 1981:70; Bloch– Bloch 1995:47; Barbiero 2004:69; Zakovitch 2004:118.

328

Articles: Old Testament Studies

apology to the Daughters of Jerusalem, she probably means to produce examples of the compatibility of darkness and beauty. To persuade the audience, she should mention something which is both black and beautiful. A century ago H. Winckler proposed to read šlmh as a name of an Arabian tribe: šalmā.11 This ingenious conjecture has been accepted by many scholars12. The tribe Shalma is known from Nabataean inscriptions13, various Rabbinic sources14, and from an entry in the Ethnica of Stephanus Byzantius: Sal£mioi, œqnoj 'Ar£bwn. s£lama d ¹ e„r»nh. çnom£sqhsan d ¢pÕ toà œnspondoi genšsqai to‹j Nabata…oij.15 Reading šalmā in Song 1:5, we have two examples of things that are black, since the tents were made of goat wool. But their beauty is not as evident and proverbial as that of Solomon’s buildings. The best solution, I think, is to read simultaneously šalmā and š3lōmō. In fact, both readings are suggested by the context: the first, by the parallelism with Kedar; the second, by the fact that Solomon is mentioned in the title of the book and in 3:7–11, 8:11–12. The Woman compares herself to the Bedouin tents: at first sight, they are only black, not beautiful. The same applies to herself: the Daughters of Jerusalem have so far noticed only her blackness. Yet she claims she is not only black, but also beautiful: as šlmh means not only šalmā, but also š3lōmō. In other words, her evidence for the compatibility of darkness and beauty is “linguistical”: she points to the fact that y3rī"ō¬ šlmh may denote something very black— and something very beautiful. (3)

tr v; . ynIaå ]v, ‘ynIWa’rT> -i la; vm,Vh'_ ; ynItp. z;Þ V" v/ , ybi-ª Wrx]nI) yMiaä i ynEBô .

‘Do not gaze at me because I am dark, because the sun has gazed on me. My mother's sons were angry with me …’ (1:6) 11

Winckler 1902:152. Robert–Tournay–Feuillet 1963:71; Loretz 1971:6; Pope 1977:320; Gerleman 1981:99–100; Fox 1985:102; Würthwein 1969:40; Rudolph 1970:122; Ringgren 1981:258; Müller 1992:14–15; Garbini 1992:184; Ravasi 1992:170–171; Gerhards 2010:175–176, n. 97. 13 Cooke 1903:217, 224. 14 See Ja. s. v. In the Targums Hebrew qēnī is frequently interpreted as ‘Shalmaite’ (Onkelos: Gen 15:19, Num 24:21–22; Jonathan: Jud 1:16, 4:11, 17, 21, 5:24, 1 Sam 15:6, 27:10, 30:29). 15 See also PW s. v. Salamii and ABD s. v. Shalma. 12

Y. Eidelkind, Intended Lexical Ambiguity in the Song of Songs

329

The Masoretic vocalization niµărū clearly points to an interpretation of nµrw as a form of nµr (Pi‘el) ‘to snort > be angry,’ cf. n3µirayim ‘nostrils,’ naµar ‘snorting’ (of a horse); Akkadian na¶āru ‘to snore’ etc.16 Qal of nµr is possibly used in Jer 6:29 with the meaning ‘to blow’ (though the text is unclear); but here the verb means ‘to rebuke, show anger.’ This meaning is attested only in Rabbinic Hebrew (Sanhedrin 94a) and in Aramaic (cf. especially nµrw byh ‘they rebuked him’ in Kiddushin 81b, note the same preposition b3). It is also possible to read nµrw as neµĕrū ‘became angry’ (µrh Nip‘al; cf. Isa 41:11, 45:24, both times with b3),17 as it was read by Septuagint, Peshitta, and Vulgate (™macšsanto, !tkšw, pugnaverunt; note that the renderings of µrh Nip‘al in Isa 41:11, 45:24 are similar). A third possibility is to read niµārū or nāµārū ‘were burned up, became charred’ (µrr Nip‘al; cf. Ezek 15:4–5, 24:10, Jer 6:29, Ps 102:4).18 Landy and Exum suggest that what we have here is a pun involving all three of the possible derivations.19 Obviously the first two options (which are almost synonymous) and the third one are not equally possible interpretations. ‘The brothers rebuked me’ and ‘the brothers were angry’ are perfectly “right” as literal interpretations, while ‘the brothers became charred’ is “wrong” (literally impossible, grotesque). However, this grotesque (mis)understanding is part of the meaning of the poem. It adds a hyperbolic quality to the brothers’ anger: they are literally ‘burned up.’ It also creates an unexpected parallel between them and their sister: while she becomes ‘dark’ because of the sun, they become ‘charred’ because of their temper. Below we shall see more cases of literally impossible interpretations which, however, are provoked by the text and usually have a grotesque/comic20 effect (cf. examples 8, 11, 14). 16

Ewald 1826:58–59; Gerleman 1981:99; Würthwein 1969:40; Ringgren 1981: 258; HALOT. 17 Robert–Tournay–Feuillet (1963:74), correctly admitting the revocalization into neµĕrū. It is very common to interpret nµrw as a form of µrh Nip‘al, explaining away or simply ignoring the conflict between this interpretation and Masoretic vocalization (see, for instance, GKC 75x; BDB; Delitzsch 1875:28; Joüon 1909: 134; Gordis 1974:79; Pope 1977:323). 18 It is pointless to explain the form as derived from µrr but meaning ‘were angry’ (Döpke 1829:70; Ginsburg 1857:134; Murphy 1990:126): µrr never has such a meaning. 19 Landy 1983:321 (n. 12); Exum 2005:105. 20 Grotesque is only partially comic, since it always has a terrifying aspect, but I think that in the Songs of Songs the comic and lighthearted mood prevails. On

330

Articles: Old Testament Studies (4)

h[orê p> ; ybekä r. BI . ‘ytiss' lu . `ytiy( [" r. : %ytiyÞ MiDI

‘I compare you, my love, to a mare among Pharaoh’s chariots’ (1:9).

Noegel and Rendsburg rightly note that this verse is linked to 1:7–8 by paronomasia or alliteration: in 1:7–8 the verb r"h ‘to pasture’ is used three times (tir"ē, r3"ī, rō"īm), and in 1:9 we encounter ra"yā¬ī ‘my love (lit. friend)’ and par"ō ‘Pharaoh.’21 In fact, r"yty may be read not only as a noun meaning ‘my friend’ (ra"yā¬ī), but also as a feminine participle of r"h (rō"3yā¬ī, cf. "ō¢3yā in 1:7), which would mean ‘my shepherdess.’22 The form r"yty is used nine times in the Song of Songs. It is a term of endearment, parallel to ¯ō¯ī. In most cases the context doesn’t suggest any pastoral connotations. Therefore ra"yā¬ī ‘my friend’ is the primary interpretation, while rō"3yā¬ī ‘my shepherdess’ is an optional and additional meaning suggested by the play on words in 1:7–9. See also example 10 below. (5)

‘ydIAd hp,yÛ " ‘^N>hi ~y[inê " @a:å `hn"n) [" r] : WnferÞ [> ;-@a;

‘Ah, you are beautiful, my beloved, truly lovely. Our couch is green …’ (1:16)

The noun "ereś occurs eight times in Biblical poetry (in prose: only Dt 3:11) and seems to be a poetic word for ‘bed.’ It is common in Aramaic and has the same basic meaning. In Rabbinic Hebrew "ereś itself is not used, but it had given rise to two derivatives: "ārīs ‘arbor, espalier’ (of grape vines) and "ărīsā ‘kneading-trough; cradle.’ Joüon and Gordis proposed to render "arśēnū in Song 1:16 as ‘our arbor.’23

grotesque in the Song of Songs, with more emphasis on its sinister side, see interesting essays by Fiona C. Black (Black 2000a; 2000b). 21 Noegel–Rendsburg 2009:74; cf. Munro 1995:98. 22 Barbiero 2004:79; cf. also Hess 2005:65. 23 Joüon 1909:149–151 (with Arabic parallels); Gordis 1974:80. Note that Joüon incorrectly refers to "rys ‘tonnelle (de vigne)’ as “aram.” In fact this meaning is attested in Rabbinic Hebrew, not in Aramaic. The form "rys does occur ten times in Targum Pseudo-Jonathan, but only in the expression tšmyš d"rys ‘sexual intercourse,’ in which "rys is simply a by-form of "rs! ‘bed.’

Y. Eidelkind, Intended Lexical Ambiguity in the Song of Songs

331

In the context of 1:16–17 ‘couch, bed’ is entirely appropriate. The lovers, lying on the green grass in the shade of cedars and junipers, sing in turn small stanzas, praising themselves and each other. This is a typically pastoral scene. Cf. Theocritus 5.31–62, where the shepherds look for a shady place with soft grass24 to lie there: only so can they begin their musical competition. At the same time the place described in Song 1:16– 17 is a royal bedroom, the cedars and junipers forming its beams and rafters. Moreover, ‘couch’ is a Leitmotif in this part of the Song: note that ‘the king’ is also ‘on his dining couch’ (m3sibbō) in 1:12, and that he ‘spends the night’ (yālīn) between the Woman’s breasts. As for "ārīs ‘espalier,’ this reading is suggested by the reference to vineyards in 1:14. Like ‘my vineyard’ in 1:6, ‘our espalier’ in 1:16 may be taken both literally and metaphorically. On the literal level, it evokes the topos of love adventure in a vineyard.25 An Egyptian literary text of XIII century is usually cited in this connection: ‘Thou art come into Joppa, and thou findest the meadow blossoming in its season. Thou breakest in to the inside and findest the fair maiden who is watching over the gardens. She takes thee to herself as a companion and gives thee the color of her lap. (But) thou art perceived and makest a confession.’26 The same topos is apparently present in Jud 21:20–23 and in a story about a šōmēr k3rāmīm ‘vineyard-keeper’ who tried to seduce a woman (Jerusalem Talmud Berakhot 26b). On the metaphorical level, ‘our espalier’ is the Woman herself; cf. ‘our vineyards’ in 2:15. Both ‘bed, couch’ and ‘arbor, espalier’ are meaningful interpretations, and Zakovitch rightly sees here an intended ambiguity.27 As in the previous four examples, the ambiguity probably28 could exist only in the written (unvocalized) text, depending on the possibility of different readings: "arsēnū vs. "ărīsēnū.

24 Among the plants growing there kÚpeiroj is mentioned (5.45), cf. kō«er in Song 1:14 (the names are similar, but the plants are different: kō«er = kÚproj is henna, while kÚpeiroj is galingale). Other intriguing points of contact between Theocritus 5.92–95 and Song 2:1–3 were noted by Garbini (1982:43; 1992:195– 197). It should not be excluded that the poet of the Song of Songs read Theocritus. 25 See Hagedorn 2003:345–346 for evidence from Greek vases. 26 ANET 478. 27 Zakovitch 2004:134. 28 According to Ja., "eres sometimes is written instead of "ārīs. This may be a result of contamination, but if it is not (i. e. if "eres ‘espalier’ is genuine), then the ambiguity is also present in the vocalized form "arsēnū.

332

Articles: Old Testament Studies (6)

!yIYh"ë ; tyBe-ä la, ‘ynIa;’ybih/ `hb'h( ]a; yl;[Þ ' Alïgd> wI >

‘He brought me to the banqueting house,29 and his intention toward me was love’ (2:4).

The interpretation of the word de°el is controversial (see Appendix). Traditionally it was understood as ‘banner’ (cf. RSV: ‘and his banner over me was love’), but recently was explained as ‘glance’ (cf. dgl ‘to look’; the NRSV’s ‘intention’ is a less likely variant of the second interpretation). Noegel and Rendsburg rightly suggest that “both meanings could be inherent, given the thread of polysemy that permeates the book.”30 In fact what we have here is a kind of distant Janus-parallelism. Munro notes that “there is an imaginative connection between 2.3 and 2.4, for the description of the woman sitting in the shade of the apple tree in 2.3 anticipates the image of the banner raised over her head in 2.4.”31 At the same time ‘his glance’ parallels ‘his right hand’ and ‘his left hand’ in 2:6.32 Moreover, dgl ‘to look’ anticipates 2:9 (the Man is mašgīaµ min haµallōnō¬ mē´ī´ min haµărakkīm ‘gazing in at the windows, looking through the lattice’). In its classical form, Janus-parallelism ties the Janus line to the lines which immediately precede and follow. In 2:4 a more distant connection is created. The verse 2:3 ends one poem (or episode), while 2:4 begins a second (2:4–7),33 and 2:9 is placed near the beginning of third one (2:8– 29

I think, ‘cellar’ is more probable. Cf. Symmachus and Vulgate: e„j tÕn o„nîna, in cellam vinariam. The Hebrew bē¬ hayyayin (lit. ‘house of wine’) corresponds exactly to the Aramaic bē µamrā!, which clearly means ‘cellar’ in Berakhoth 51b. ‘Banqueting house’ would be called bē¬ mištē hayyayin (Esth 7:8, cf. Dan 5:10 bē¬ mišt3yā!) or simply bē¬ mište (Jer 16:8, Eccl 7:2 and in Rabbinical Hebrew). Of course, we should not imagine the Woman literally visiting the Man’s cellar: ‘wine’ metaphorically means love, ‘to bring to the cellar’ means to give all his love. 30 Noegel–Rendsburg 2009:192; cf. also Landy 1983:216. 31 Munro 1995:36. 32 Ravasi 1992:220. Ravasi states that both ‘banner’ and ‘glance’ suit the context well. Perhaps he understands this as a case of intended ambiguity. 33 I cannot agree with those (Longman 2001:109; Callow 1994:471–473) who treat 2:1–7 as one poem, since 2:1–3 is a dialogue between the Man and the Woman, while 2:4–7 is her monologue addressed to the Daughters of Jerusalem. Moreover, the Man is not present in 2:4–7, since the Woman is longing for him (‘sick with love’). Pope’s notion of “a sort of nymphomania combining surfeit and insatiable desire” (Pope 1977:382) is rather absurd. Would such a nymphomaniac spend her time talking to her friends when embraced by her lover? The poem 2:4–7 is built according to a ABAB pattern (Zakovitch 2004:141). The B verses (2:5, 7) suggest that the Woman speaks to the Daughters of Jerusalem. In the A

Y. Eidelkind, Intended Lexical Ambiguity in the Song of Songs

333

17).34 So ‘his banner’ ties 2:4–7 to the preceding poem, and ‘his glance’ ties it to the subsequent one. Moreover, I think that de°el has also a third meaning, which ties 2:4–7 to an even more distant part of the book. As a ‘sign’ (not only military ‘banner’), it may be understood as parallel to hō¬ām ‘seal’ in the famous phrase: ‘Set me as a seal upon your heart, as a seal upon your arm’ (8:6). Cf. Deut 6:8: ūq3šartām l3!ō¬ "al yā¯e−ā ‘bind them as a sign on your hand.’ Here ‘sign’ is !ō¬, a common word of which de°el is a rare synonym. It is possible to translate 2:4b in the following way: ‘His sign/token on me (i. e. on my body) is love.’ (7)

‘~Øli ;’vW' ry> tAnÝB. ~k,øta. , yTi[B. ;’vh. i hd [*T-. ~aiw>) Wry[iTó -' ~ai `#P'x( T. v, , d[;î hb'hÞ ]ah; -' ta,

‘I adjure you, O daughters of Jerusalem, by the gazelles or the wild does: do not stir up or awaken love until it is ready!’ (2:7 = 3:5)

Normally, one swears or adjures by a god’s name. Several scholars rightly note that ´3®ā!ō¬ ‘gazelles’ is a homonym of Yahweh’s epithet “Sabaoth” (i. e. “Yahweh of Hosts”). At the same time !ayl&¬ haśśā¯é ‘wild does’ sounds like !ēl šaddáy, another name of God. Gordis has convincingly explained this as a colloquial euphemism. He brought up many French, German and English parallels (like “Holy Cow” for “Holy Christ”). Fox noted Rabbinic parallels, such as b3µayyē haqqayi´ ‘as fruits live.’35 At the same time gazelles and does are important metaphors in the Song of Songs (cf. 2:8–9, 17, 4:5). Their presence in 2:7 and 3:5 is not due only to their phonetic similarity to the God’s names. Scholars have demonverses (2:4, 6) she imagines the Man to be present. It may be a recollection of past joys, a wish (understanding hĕ®ī!anī as a precative perfect) or a vision of the future. 34 For its unity see Exum 1973:54–55; Landy 1983:40–43; Heinevetter 1988: 99–104; Murphy 1990:138–142; Ravasi 1992:240–241. 35 Gordis 1974:26–29; Fox 1985:109–110; see also Robert–Tournay–Feuillet 1963:108; Goulder 1986:20; Murphy 1990:133; Loretz 1971:15; Schwab 2002: 42–43, 108 (n. 6); Hess 2005:82. Cf. Septuagint: ™n ta‹j dun£mesin kaˆ ™n ta‹j „scÚsesin toà ¥grou ‘by the forces and powers of the field,’ Targum: bh! ´bwt wbtqpy !r"! dyśr!l ‘by Yahweh of Hosts and the powers of the land of Israel’ (both versions apparently read !ăyālū¬, not !aylō¬).

334

Articles: Old Testament Studies

strated the erotic symbolism of these animals. They belonged to the group of images traditionally connected with goddesses of the Ancient Near East.36 It is important to recall that in Prov 5:19 a wife is called !ayyele¬ !ăhā®īm w3ya"ăla¬ µēn ‘a lovely hind, a graceful doe.’ In the context of 2:7 and 3:5 note also the meaningful contrast created between ‘Jerusalem’ and ‘field.’ The city-dwelling Daughters of Jerusalem are asked to swear by the animals which represent wild nature.37 The strange oath by animals has a very interesting parallel in Longus’ novel Daphnis and Chloe.38 The young shepherd Daphnis swears by Pan to love Chloe. She, however, says: “Pan is an amorous god and not to be trusted.” So she asks: sÝ dš moi tÕ a„pÒlion toàto Ômoson kaˆ t¾n aga ™ke…nhn, ¼ se ¢nšqreye … ‘Swear to me by this flock of goats, and that nanny-goat who nursed you …’

He consents: st¦j e„j mšson tÕ a„pÒlion kaˆ tÍ mn tîn ceirîn a„gÕj tÍ d tr£gou labÒmenoj êmnue ClÒhn filÁsai filoàsan· `H d œcaire kaˆ ™p…steuen æj kÒrh kaˆ nšmousa kaˆ nom…zousa t¦j agaj kaˆ t¦ prÒbata poimšnwn kaˆ a„pÒlwn „d…ouj qeoÚj. ‘Standing in the middle of his flock, and taking hold of a nanny-goat in one hand and a billy-goat in the other, he swore to love Chloe while she loved him … She was delighted and trusted him, for she was but a young girl and a shepherdess and thought the goats and sheep were the special gods of shepherds and goatherds (2.39.4–6).’39

To sum up, ‘the gazelles’ and ‘the wild does’ are, on the one hand, substitutes for God’s names, but on the other hand, they are meaningful in themselves, as erotic metaphors and as images evoking pastoral naïveté (we have here one of the links between the Song of Songs and Greek bucolic literature). This is an instance of intended ambiguity,40 though a special one: yhwh ´3®ā!ō¬ and !ēl šaddáy are only virtually present in the texts, being represented by euphemisms. 36 See especially Keel 1992:92–94 and Abb. 43–45; also Haller 1940:29; Meek 1956:114; Würthwein 1969:44; Krinetzki 1981:97; Pope 1977:386; Müller 1992:26. 37 Barbiero 2004:94, 129. 38 Ibid. 92 (n. 204). 39 Translation by J. R. Morgan (Morgan 2004:81). 40 Cf. Rudolph 1970:131–132; Bloch–Bloch 1995:152; Longman 2001:116; Zakovitch 2004:145.

Y. Eidelkind, Intended Lexical Ambiguity in the Song of Songs (8)

335

#r I ‘~ynICN' hI ; [:yGIh+ i rymiÞZh" ; t[eî `Wncer( a> B; . [m;vî n. I rATàh; lAqïw>

‘The flowers appear on the earth;41 the time of singing has come, and the voice of the turtledove is heard in our land’ (2:12).

The word zāmīr was understood as ‘pruning’ by Septuagint (tomÁj)42, Aquila and Symmachus (kladeÚsewj), Peshitta (ksµ!), Vulgate (putationis). Some modern scholars accept this rendering.43 Apparently it also forms the basis of the allegorical interpretation in Targum ("dn qy¢wp bkr!y ‘the time of the extirpation of the firstborn’). In fact only z3mīrā ‘pruning’ is found in Hebrew (and only in Rabbinic sources), while the form without -ā is unattested. However, it could well exist: cf. µārīš ‘ploughing,’ qā´īr ‘harvest.’ The meaning of zmr in the Gezer inscription is debated: to judge from its position between q´r ‘grain harvest’ and q´ ‘harvest of fruits,’ it should mean ‘vintage’ rather than ‘pruning’; Lemaire suggested ‘vintage’ in Song 2:12 as well.44 Most commentators understand zāmīr as ‘singing’ (of people,45 of birds,46 or both47). Otherwise zmr and its derivatives denote only human singing and music, so this must be the basic meaning here. The objection that zmr refers only to ritual music48 is not valid, cf. Isa 25:5. Rabbi Aqiba condemned those who make a zmr out of the Song of Songs, clearly in a very secular context (t Sanh 12:5). Another objection is that people can sing at any time, so that ‘time of singing’ is unspecific.49 But surely the Man’s point is not to tell the Woman what’s the date today! His intention is to invite her to come out and share with him the vernal joys. The ‘singing’ is their, her and his singing. This is not to deny the parallelism be41

More probably, ‘the blossoms have appeared in the land’ (NJPS), since nē´/ni´´ā otherwise refers only to blossoms of the vine and of the olive tree. 42 Followed in this case by the mediaeval Codex Venetus. 43 Budde 1898:10–11; Joüon 1909:165–166; Jastrow 1921:174; Gerleman 1981:124–125; Murphy 1990:138–139; Müller 1992:30–31. 44 Lemaire 1975:22–26. 45 Hitzig 1855:38; Robert–Tournay–Feuillet 1963:119; Würthwein 1969:45– 46; Rudolph 1970:133; Gordis 1974:6–7; Krinetzki 1981:252; Keel 1992:98; Garbini 1992:204; Loretz 1971:18; Barbiero 2004:109–110. 46 Ibn Ezra; Rashi; Ewald 1826:84; Delitzsch 1875:48; Haupt 1902a:72; Fox 1985:113; Goulder 1986:23. 47 Ginsburg 1857:146; Hess 2005:94. 48 Jastrow 1921:174. 49 Müller 1992:30–31.

336

Articles: Old Testament Studies

tween ‘the time of singing’ and ‘the voice of the turtledove.’ In 2:10–14 there is a whole network of ‘voices’: the voice of people singing, of the turtledove, of the Woman (2:14: ‘let me hear your voice’), and of course, of the Man who sings the serenade. Cyrus H. Gordon proposed to understand zāmīr as intentionally ambiguous: ‘pruning’ is parallel to ‘flowers,’ and ‘singing’—parallel to ‘the voice.’ For this example he first coined the term “Janus parallelism.”50 Many others have accepted this interpretation.51 However, no pruning takes place at the time when the vine blossoms (2:13, 15: s3mādar; the ni´´ānīm in 2:12 are also most probably ‘blossoms’ of vines and trees, not ‘flowers,’ since this word is used only of the vine and olive blossoms, while bā!are´ means ‘in the land’). The time of pruning is January–March, before blossoming; a second, less important pruning takes place in July–August.52 So in Song 2:12 it is too late or too early for the blooming vines to be pruned (and, of course, no vintage can take place when they are blooming, so Lemaire’s hypothesis does not work). Pope and Murphy try to support ‘pruning’ with reference to Isa 18:5, but (1) even there the vines already have green grapes (bōser), and (2) what is described there is not care of a vineyard, but its destruction.53 ‘Singing’ is the only right understanding of zāmīr in Song 2:12, and ‘pruning’ is a misunderstanding. But this misunderstanding may be provoked by the text itself: by the combination ‘time’ + zāmīr in a context referring to vineyards. The intended ambiguity here takes the form of a brief moment of suspense before one realizes that nobody prunes blooming vines. The pun in 2:12 calls attention to an important theme in 2:8–17: the theme of right and wrong time. The Man says the right time has come for ‘singing’ as well as for more sensual joys, of which ‘singing’ is only a metonymy. But the Woman retorts by comparing him to a fox54 coming to a vineyard when there are no grapes there, only blossoms. Evidently he has come too early (2:15). 50

Gordon 1982:80*, with reference to his previous 1978 article. Even before Gordon, Pope had already suggested that “the poet was aware of the possible ambiguity of the zmr in the present passage and used it deliberately for the provocative double entendre” (Pope 1977:396; and cf. also Döpke 1829:104). 51 Landy 1983:42; Fox 1985:113; Paul 1992:150; Ravasi 1992:252; Munro 1995:24; Zakovitch 2004:151; Noegel–Rendsburg 2009:193. 52 Dalman 1987, IV 312–313. 53 Ibid. 331. 54 On the erotic symbolism of foxes see Keel 1992:104–106 with Abb. 56, 57; Hagedorn 2003:339–342.

Y. Eidelkind, Intended Lexical Ambiguity in the Song of Songs (9)

337

@s,k,ê hf'['ä ‘wyd"WM[; bh'zê " Atædy" pir> !m"G+ r" a> ; AbßKr' >m, hb'hê ]a; @Wcår" ‘AkAT

‘He made its posts of silver, its back of gold, its seat of purple; its interior was inlaid with love’55 (3:10).

The passive participle rā´ū« is usually understood as ‘inlaid’ (though some prefer ‘woven,’56 ‘sewed together,’57 etc.) In the Bible only a noun ri´«ā ‘floor’ is attested, but in Rabbinic Hebrew there is a verb r´p, meaning ‘to press, crush; join close, to place in rows; to pave with blocks.’ ‘Inlaid,’ ‘woven,’ etc. are only guesses based on the meaning ‘to pave.’ Since ‘inlaid with love’ seems too abstract, some interpreters take !ahăbā as ‘love scenes,’58 which is very strained.59 Others understand ‘with love’ as ‘lovingly,’60 but in the context of the Song of Songs ‘love’ is Eros, hardly “the attitude of the worker or the loveliness of the work,” as Murphy puts it. Consequently, emendations were proposed (such as !ă®ānīm ‘stones,’61 ho®nīm ‘ebony’62 or even *!ehabbīm ‘ivory,’ a non-existent word

55 The last words of the verse (mibb3nō¬ y3rūšālayim ‘from the Daughters of Jerusalem’) can hardly belong to it syntactically. That the women were responsible for making Solomon’s !appiryōn (Bloch–Bloch 1995:165; Longman 2001:138; Hess 2005:111) is improbable. Many commentators understand b3nō¬ y3rūšālayim as a vocative beginning 3:11. Three Masoretic manuscripts read b3nō¬ without mi(Garbini 1992:56, 62). Keel 1992:126 attaches the m to !ahăbā, reading !ahăbīm. 56 Murphy 1990:148. 57 Gerhards 2010:50. 58 Pope 1977:445; Keel 1992:126. Pope could mention the fact that this interpretation had been advanced already by Hugo Grotius: “media lecticae amorum historias habent” (Grotius 1775:451). Keel adduces examples of Ancient Near Eastern beds decorated in this way (Abb. 38, 71) and cites Sanhedrin 39b for pornographic images in the chariot of Ahab. 59 H. Tawil has recently argued that !ahăbā means here ‘sumptuous decorations,’ comparing the Akkadian idiom lalû/lulû mullû/šumlû, frequently used with reference to building activity (Tawil 2003:268–270). However, such a meaning for !ahăbā is unattested elsewhere. It is hard to accept !ahăbā as a “functional equivalent” of Akkadian lalû/lulû which basically means ‘wish, desire; wealth, happiness’: µe«e´ or µemdā would suit better. 60 Ginsburg 1857:153; Delitzsch 1875:63; Robert–Tournay–Feuillet 1963:152; Murphy 1990:149; Ravasi 1992: 327; Munro 1995, 25; Longman 2001:138; Hess 2005:110; Gerhards 2010, 49. 61 Gerleman 1981:139; Fox 1985:121; Müller 1992:37; Exum 2005:139.

338

Articles: Old Testament Studies

reconstructed on the basis of šenhabbīm63). Driver hypothesized an homonym !ahăbā ‘leather’ (cf. Arabic !ihāb).64 In Rabbinic Hebrew the passive participle rā´ū« can be used in the sense ‘stuffed, full.’ One can be said to be ‘full of commandments (rā´ū« mi´wō¬) as a pomegranate’ (Midrash Rabba on the Song of Songs 4.4). In Syriac r´yp may be said of a bed or a bedroom in the sense ‘prepared, made’; for instance, in Peshitta to Song 1:16 "rsn r´yp). These Rabbinic and Syriac usages are not noticed in the commentaries on Song 3:10, but can help to explain rā´ū« !ahăbā. Solomon’s !appiryōn is ‘stuffed with love,’ and love is, metaphorically, its bedclothes. It is also possible to derive rā´ū« from a verb r´p II ‘illuminate.’ In the Bible re´e« or ri´pā (with a plosive p, as against ri´«ā) means ‘glowing coal.’ As a verb, r´p (Pi"el) is attested in Sir 43:8, where the moon mr´p rqy" mzhyrtw ‘illuminates the sky with its shining.’ Comparing rā´ū« to the Biblical noun, Ibn Ezra interpreted the participle as ‘burning.’ Garbini, taking his cue from Sir 43:8, understood Love as illuminating the !appiryōn’s interior.65 Garbini’s hypothesis is marred by an unnecessary emendation: he reads twkh ry´ph (with y!) !hbh ‘l’interno l’illumina l’Amore.’ There is no need to change the text: r´p II ‘illuminate’ could be used in Qal as well as in Pi"el; or at least a passive participle rā´ū« could be used along with Pi"el/Pu"al forms (cf. bārū− ‘blessed’ along with m3®ōrā−). To the credit of the Ibn Ezra–Garbini interpretation note, first, that silver, gold and purple in the previous lines combine to create a visual impression of something bright and shining. Second, the poem 3:6–11 begins with a verse mentioning ‘columns of smoke’ (tīmāro¬ "āšān). An allusion to ‘burning’ near the end of the poem 3:6–11 creates an inclusio. I understand rā´ū« !ahăbā as having a triple meaning: ‘stuffed with love’/‘covered with love’/‘illuminated by (or burning with) love.’66

62

Haupt 1902b:2; Jastrow 1921:184; Haller 1940:32; Würthwein 1969:49; cf. Graetz 1871:103. 63 Krinetzki 1981:125, following Tur-Sinai 1951:356, n. 8. 64 Driver 1936:111; followed by Meek 1956:120; Gordis 1974:85; HALOT. On Grossberg’s proposal to understand !ahăbā as simultaneously meaning ‘love’ and ‘leather’ see below under 24. 65 Garbini 1992:213. 66 Zakovitch 2004:178 proposed ‘inlaid’/‘burning’; Noegel–Rendsburg 2009: 196: ‘arranged with leather’/‘burning with love.’ The case for ‘leather’/‘love’ ambiguity is considered separately below (example 24).

Y. Eidelkind, Intended Lexical Ambiguity in the Song of Songs (10)

339

hp'yê " %N"hå i ‘ytiy[" r. : hp'yÛ " %N"“hi %tEM+ c' l; . d[;B;Þmi ~ynIAë y %yIny:å [e ~yZI[ë hi '( rd K; ; ‘%yIl[;’ ' %vEaÜ ro !m"G+ r" a> K; ' %vEaß ro tL;dî w: >

Gesenius Thes. II 789; Haupt 1902b:229; Schmökel 1956:67; Meek 1956: 129; Gordis 1974:91; Jastrow 1921:207; Robert–Tournay–Feuillet 1963:215–216; Pope 1977:538–539; Fox 1985:148; Goulder 1986:5; Keel 1992:187 and Abb. 108, 109; Garbini 1992:237; Longman 2001:164; Barbiero 2004:246; Exum 2005:205; Hess 2005:163–164; Zakovitch 2004:224 (but he accepts ‘eingelegt’ as an additional meaning). 107 Delitzsch 1875:92; Oettli 1889:187; Grätz 1871:177; Loretz 1971:36; Ringgren 1981:276; Provan 2001:336 (but he thinks that there is “also an allusion to doves bathing in ample water”). A variant of this approach is to insert šinnāw ‘his teeth’ at the beginning of the line, then millē(!)¬ means the gums in which the teeth are set (Würthwein 1969:57; Rudolph 1970:159; Murphy 1990:166; Ravasi 1992:457–458). 108 Gerleman 1981:174: “Das Baden in Milch soll wahrscheinlich einen Zustand des Überflusses und Wohlstandes andeuten … In die gleiche Richtung weist auch taLem-i l[; tAbv.yO …” 109 Tur-Sinai 1951:379; Kutscher 1976:33. 110 DJBA 679.

Y. Eidelkind, Intended Lexical Ambiguity in the Song of Songs

349

‘Your head crowns you like Carmel, and your flowing locks are like purple’ (7:6a).

The word karmel in 7:6a is usually understood (by all of the ancient versions and by the majority of modern commentators) as a place-name: Carmel, the well-known promontory on the Mediterranean coast. It continues a whole row of toponyms in 7:5: ‘Your eyes are pools in Heshbon, by the gate of Bathrabbim. Your nose is like a tower of Lebanon, overlooking Damascus.’ In poetry the word karmel forms a traditional pair with ‘Lebanon’ (in such cases it may be a place-name as well: Isa 33:9, 35:2, Nah 1:4; or karmel may be used in the meaning ‘garden’ and referring to Lebanon itself: 2 Kings 19:23/Isa 37: 24, Isa 29:17). ‘Your rō!š on you is like Carmel’ is a pun, since rō!š means ‘head’ and ‘top of a mountain’ (cf. rō!š hakkarmel in 1 Kings 18:42, Am 1:2, 9:3).111 At same time karmel sounds like karmīl ‘crimson,’ and ‘crimson’ parallels ‘purple’ (!argāmān), another variety of red dye. A number of scholars prefer to read karmīl ‘crimson,’112 while others defend ‘Carmel’ against them.113 What we have here in fact is a clear example of Janus parallelism114 based on homography. Moreover, both Lebanon and Carmel remind of the Phoenicians, who lived at the nearby coast and made the purple.115 Thus beside the Janus parallelism (schematically: AcB1, B2cC) there are additional links between A and C, B1 and C. (20)

hd I rd:êmS' h. ; xT;Pä i ~ynIA+ MrIh' WcnEhß e `%l") yd:DÞ -o ta, !TEaï , ~v'²

‘… let us go out early to the vineyards, and see whether the vines have budded, whether the grape blossoms have opened and the pomegranates are in bloom. There I will give you my love’ (7:13).

Here is another instance of dō¯īm read as daddayim ‘breasts’ not only by Septuagint (mastoÚj) and Vulgate (ubera), but also by Peshitta (tdy). Again Luther follows them, but the moderns unanimously121 reject this reading for no reason. Note that ‘breasts’ (šā¯ayim) were central in 7:8–9, and that they were imagined to be clusters of grapes. In 7:13, only four verses later, ‘vines’ and ‘grape blossoms’ are mentioned. Is it not enough to suggest ‘breasts’? Of course, not as the only reading, but as a possible one, along with ‘love.’ Note also that the pun continues into 7:14: here the first word is dū¯ā!īm ‘mandrakes,’ playing both on dō¯īm and daddayim. It is interesting to compare one more instance of dō¯īm/daddayim:

!xE-ï tl;[ñ y] w: >) ~ybiªha' ] tl,Yali `~ynIM) ro hI ' WcnEhß e

‘I went down to the nut orchard, to look at the blossoms of the valley, to see whether the vines had budded, whether the pomegranates were in bloom’ (6:11).

129 130

Malul 1997:249. Eidelkind 2006:222–224.

Y. Eidelkind, Intended Lexical Ambiguity in the Song of Songs

355

According to HALOT,131 the word naµal has here an unusual meaning: ‘date-palm’ (naµal II) instead of ‘river valley, wadi’ (naµal I). Other examples of naµal II in HALOT are Num 24:6, Job 29:18 (a conjectural reading, with a question mark), Sir 50:12. Noegel and Rendsburg render ‘palm tree’ and comment: “Both senses are appropriate here—thus polysemy is evident—though we elect the latter in our translation since the context is that of fruit trees.”132 However, since ‘river valley, wadi’ is possible in Num 24:6 and Sir 50:12, there is no unequivocal evidence for the very existence of naµal II ‘date-palm’ in Hebrew. IV. Clustering of puns After reviewing 21 instances of intended ambiguity (excluding dubious examples), it would be interesting to see how they are distributed over the 117 verses of the book. Eight of the instances are concentrated in 4:1–10 (examples 10–17). Three are found in 7:6–13 (examples 19–21), two in 1:5–6 (examples 2– 3) and two in 2:4–7 (examples 6–7). Others do not appear in clusters. The phenomenon of clustering perhaps can be understood as a way of calling the reader’s attention to the feature. Moreover, it is interesting that the main clustering occurs around the description of the Woman in chapter 4. In this chapter there is also an unusually high concentration of paronomasia:133 gāl3šū ‘streamed down’ — gil"ā¯ ‘Gilead’ (4:1); ra"yā¬% ‘my love’ — śa"rē− ‘your hair’ — k3"ē¯er ‘like a flock’; k3"ē¯er ‘like a flock’ — mibba"a¯ ‘behind’ — gil"ā¯ ‘Gilead’; šekkullām ‘all of whom’ — šakkūlā ‘bereaved’ (4:2); šinnayi− ‘your teeth’ (4:2) — haššānī (4:3) ‘scarlet’ — š3nē ‘two’ (4:5) — kišnē ‘like two’ (4:5) — šōšannīm ‘lotuses’ (4:5); talpiyyō¬ ‘layers’ — !ele« ‘thousand’ — tālūy ‘hung’ (4:5); l3®ānōn ‘Lebanon’ (4:8, 11, 15) — libbā®tīnī ‘you have ravished my heart’ (4:9) — l3®ōnā ‘frankincense’ (4:14); tāšūrī ‘look/leap’ — rō!š ‘top’ (4:8); mēharrē n3mērīm ‘from the mountains of leopards’ (4:8); mimm3"ōnō¬ ‘from the dens’ (4:8) — mē"ēnayi− ‘of your eyes’ (4:9) — "ănāq ‘necklace’(4:9); nō«e¬ ti¢¢ō«nā śi«¬ō¬ayi− ‘your lips drip honey’ (4:11); gan ‘garden’ — gal ‘door’ (4:12).

131

Following Rudolph 1970:166; Gerleman 1981:188. The same translation is hesitantly accepted by Müller 1992:69–70. 132 Noegel–Rendsburg 2009:202. 133 Cf. Noegel–Rendsburg 2009:85–88.

356

Articles: Old Testament Studies

Clustering of puns makes the discourse in chapter 4 especially sophisticated and playful. Such a style is appropriate for the Woman’s portrait abounding in strange metaphors, frequently with comical overtones. Segal rightly saw here elements of grotesque which can be explained only as ‘playful banter.’134 V. Oral or written poetry? Many modern scholars still imagine the Song of Songs to be an anthology of folk oral poetry. This romantic idea, which can be traced back to Herder’s Lieder der Liebe, contributed largely to the peasant wedding theory of Budde. The theory was rightly abandoned, because most of the poems in the Song of Songs have nothing to do with wedding. But even now the concept of the Song of Songs as oral poetry is very much alive. Witness the following quotation from a recent book by Noegel and Rendsburg: “If we can imagine a group of ancient Israelites assembled to hear the Song of Songs read (sung) aloud we may rightly ask: how does one continue to engage the attention of the gathered throng as the poetry proceeds through its eight chapters?”135 In the same book there is an interesting chapter on alliteration in the Song of Songs. This poetic device is understood as “befitting of a literature with an oral/aural quality, that is, one that was intended to be read aloud by a single reader before a listening audience.”136 If the oral hypothesis were correct, all the cases of intended ambiguity in the Song of Songs would be oral as well. However, this is not the case. In the total of 21 examples, there are only 8 instances of true homonymy (6, 8, 9, 11–13, 18, 20). These puns could be pereceived aurally as well as visually. But the other 13 examples exploit homography rather than homophony (1–5, 7, 10, 14–17, 19, 21). Perhaps it is possible to argue that "arsēnū/"ărīsēnū (5) and karmel/karmīl (19) could be perceived aurally, but I doubt that a “gathered throng,” imagined so vividly by Noegel and Rendsburg, could hear yašqēnī in yiššāqēnī, šalmā in š3lōmō, šadday in śā¯e, etc. Even in our age poetry is frequently read aloud, sometimes “before a listening audience.” Of course, the Song of Songs was also sometimes recited or “sung.” We even know where and how this was done: on a party, in a modulating (mn"n") voice (Tosefta Sanhedrin 12.5). But I think this 134

Segal 1962:480. Noegel–Rendsburg 2009:107. 136 Ibid. 66–67. 135

Y. Eidelkind, Intended Lexical Ambiguity in the Song of Songs

357

was not the only or the most productive way of engaging with book (so far I find myself in agreement with rabbi Aqiba). In a provocative essay David Clines rightly stated: “A text is not a performance. Whatever else the Song of Song is, it is not a song. A song implies, for its realization, a singer and a hearer—at least one, but often many. A text implies, for its realization, only a reader. It actually requires a single, lone reader .”137 I would add that a text like the Song of Songs requires a very careful and slow reader, frequently pausing to reflect or returning some pages back. The visual puns are only an example (though a very telling one) of literary techniques in the Song of Songs that presume a reader meditating over a written page. This undermines not only the folklore hypothesis, but also the “soft pornography” hypothesis advanced by Clines.138 The Song of Songs is too sophisticated to be pornography. Pornography operates with a few very simple topoi in order to attain a quick effect. It cannot be read (or viewed) slowly, cannot be re-read, re-thought or scrutinized. And of course, it would be very strange to find there Janus parallelism, paronomasia, etc. Such subtleties are as unnecessary in pornography as would be, for example, footnotes and references. Appendix: di°lō in 2:4 1. ‘Banner’ The interpretation of de°el as ‘banner’139 goes back to the medieval Jewish scholars: see quotations from Abulwalides (Ibn-Janah) in Gesenius’ Thesaurus.140 It is reflected in BDB, most translations and many, if not most, commentaries. The military metaphor of ‘his banner’ in Song 2:4 is accounted for in at least two different ways. а) The Woman imagines herself to be a city (cf. 6:4 and 8:10) captured by the Man. He has raised his banner above her as a sign of his victory.141 Or we may say that the banner is “protective.”142 Dominance and protection are the same thing viewed from different points.

137

Clines 1994:6. Ibid. 8, 24. 139 More precisely, a sort of standard: see Keel 1992, Abb. 39–40. 140 Gesenius Thes. I 320. 141 Barbiero 2004:91–92. 142 Döpke 1829:93; Delitzsch 1875:41; Grober 1984:98. 138

358

Articles: Old Testament Studies

b) Or the Woman is “a lover-warrior,” going into the “battle” with a banner in her hands.143 Then her “lovesickness” in 2:5 may be understood as a consequence of being wounded in the battle, cf. Septuagint: tetrwmšnh.144 This reminds one of a topos of the Roman erotic poetry succinctly formulated by Ovidius: Militat omnis amans, et habet sua castra Cupido ‘Every lover makes war, and the Cupid has his own troops’ (Amores 1.9.1).

In Ars amatoria he says: Militiae species amor est; discedite, segnes: non sunt haec timidis signa tuenda viris ‘Love is a sort of war. Be off, you lazy ones! These banners are not for timid men to guard’ (2.233–234).145

Compare also Horatius: Late signa feret militiae tuae ‘He will widely carry your (i. e. Venus’) military banners’ (Ode 4.1.16).

To judge from these parallels, “war” and “banner” probably belong to the pool of erotic topoi common to ancient Mediterranean peoples.146 Therefore ‘his banner’ in the erotic context of 2:4–7 is traditional and perfectly understandable. It remains only to demonstrate that de°el did mean ‘banner’: on this see below. Some modern commentators, seeking a more obvious link between de°el and the ‘house of wine,’ understood it as tavern-sign147 or flag mark-

143 Grotius 1775:450: “milito sub eius vexillo, cui inscriptum est, Amor”; Gesenius Thes. I 321: “sequor vexillum amoris, quod mihi praefert amicus meus, ut milites signum militare sequuntur neque umquam deserunt”; see also Keel 1992:86. 144 There is no need to reconstruct a different Vorlage µllt (Garbini 1992:45, 201; 1983:25–29), since µlh can be used of the sickness of a wounded (2 Kings 8:29). Nor is it necessary to see here an allusion to Eros the archer (already present in the Hebrew text, according to Garbini, or introduced by the Greek translator as an interpretatio Graeca, according to Gerleman 1981:117). 145 The parallels are noted by Hitzig 1855:31. 146 For the concept cf. Hagedorn 2003:337; Hunter 2005:230. 147 Haupt 1902b:232; Loretz 1971:14; Müller 1992:25; HALOT; Bergant 2001:25: “… it was not uncommon then, nor is it uncommon today, that taverns and pubs frequently would display outside of their establishments some kind of a sign, such as on overflowing flagon, indicating the pleasures that transpire inside.”

Y. Eidelkind, Intended Lexical Ambiguity in the Song of Songs

359

ing the house where a banquet is taking place.148 In this case di°lō would mean ‘its (the house’s) sign/flag’ rather than ‘his (the lover’s) banner.’ While formally possible, this interpretation is unnatural: since y3mīnō is ‘his right hand’ and ś3mō!lō ‘his left hand’ (2:6), di°lō must be ‘his de°el.’ Moreover, I think the ‘house of wine’ is neither a tavern nor a private house where a banquet is taking place, but a cellar (see n. 29 above). Cellars surely had no signs. 2. ‘Glance’ In 1969 Gordis suggested that di°lō in 2:4 should be rendered ‘his glance.’149 He was building on his earlier proposal (in the first 1953 edition of his commentary) to understand ni¯gālō¬ in 6:4, 10 as ‘sights,’ cf. Akk. dagālu ‘to look.’ In 5:10 the passive participle dā°ūl was already in BDB explained with reference to this Akkadian verb: ‘looked at, conspicuous.’150 However, BDB kept to the traditional understanding of ni¯gālō¬ as ‘bannered hosts,’ thus unnecessarily creating homonyms dgl I ‘to look’ and dgl II ‘to carry, or set up standard, banner.’ ‘Hosts’ was also an improbable rendering for ni¯gālō¬, a feminine plural participle. One would expect masculine plural ni¯gālīm ‘(men) with banners.’ Finally, in Ps 20:6 b3šēm !elōhēnū ni¯gōl is best understood as ‘we shall trust on the name of our God,’ ‘we shall seek favor with the name of our God’ (literally: ‘look at our God’s name’).151 Pope modified Gordis’ interpretation of Song 2:4, proposing ‘intent’ (cf. ‘wish’ as one of the secondary meanings of Akkadian diglu).152 This is less convincing than “glance”: we cannot be sure that in Hebrew dgl developed this secondary meaning as well. The parallelism with 2:9 (see above) also supports the primary meaning ‘glance.’ A minor grammatical point is worth making. Gordis and others understood di°lō as a form of the noun de°el. In this case we have two de°el’s 148 Gerleman 1981:118; Budde 1898:8; Haller 1940:28; Rudolph 1970:130– 131; Krinetzki 1981:92. 149 Gordis 1969:203–204; 1974:81–82. Before him Meek had already proposed to read di°lū ‘look’ (plural) (Meek 1956:113). Gordis was followed by Murphy 1990:132 and Zakovitch 2004:142. 150 Though in this case “marked with a sign” > “conspicuous” would be also possible (Codex Venetus: shsemaiwmšnoj; Delitzsch 1875:90; Graetz 1871:176; Gerleman 1981:173; Garbini 1992:236). 151 Cf. DCH, s. v. In Akkadian dagālu pani N ‘to wait for, to be subject of a king or god,’ dagālu īnē means ‘to seek favour with a person’ (CAD D 23–24). 152 Pope 1977:376–377; Fox 1985:108; Hess 2005:45.

360

Articles: Old Testament Studies

(or two different meanings of the same noun): ‘glance’ and ‘military unit’ (plus ‘sign, banner,’ see below). It is also possible to interpret di°lō as an infinitive construct of dgl,153 without creating a hapax noun. 3. Doubts raised against de°el ‘banner’ Gordis, Pope and other proponents of the new interpretation of di°lō based on Akkadian did not recognize the intended ambiguity in Song 2:4 and felt themselves obliged to deny the traditional interpretation ‘banner.’ Gordis simply stated that it is “virtually meaningless.” Pope argued that de°el in the Bible, in Qumran and Elephantine Papyri means only ‘military unit.’ “The evidence for taking the primary sense of de°el as banner or standard appears tenuous.”154 Pope was not the first to raise doubts against de°el ‘banner.’ In 1898 Buchanan Gray, having carefully reviewed the evidence, concluded that the only certain meaning of the noun is ‘military unit.’ 4. The word de°el in other parts of the Bible: ‘military unit’ The noun is attested 13 times in the book of Numbers (1:52, 2:2, 3, etc.). Its most obvious meaning here is ‘military unit, division’ (of a tribe):

WhnE±x]m-; l[;( vyaió laer_ f" y. I ynEBå . Wnàxw' > `~t'a( bo c. li . AlßgD> I-l[; vyaiwî > ‘The other Israelites shall camp in their respective regimental camps, by companies’ (Num 1:52).

Most probably !īš "al di°lō means ‘each according to his military unit,’ not ‘each near his banner.’ Note the parallel construction with maµanēhū. In such cases as Num 10:14 it is especially difficult to take de°el in the sense of ‘banner, standard’:

hd"²Why>-ynEb) . hnExô ]m; lg : !yBeämi qqExß omW. ‘The scepter shall not depart from Judah, nor the ruler's staff from between his feet …’

Instead of the MT mbyn rglyw ‘from between his feet’ the Samaritan text reads mbyn dglyw. Here ‘from among his military units’ is a possible understanding, but ‘from between his standards’ would be meaningless. 5. Ancient versions of Song 2:4: ‘military unit’ Almost all of the ancient versions try to understand de°el according to the usual meaning ‘military unit,’ which was apparently the only one known to them. LXX:

Vulgate: Peshitta:

e„sag£getš me e„j okon toà o‡nou t£xate ™p' ™m ¢g£phn (Aquila: œtaxen ™p' ™m ¢g£phn) (Symmachus: ™piswreÚsatš moi ¢g£phn) introduxit me in cellam vinariam ordinavit in me caritatem !"lwny lbyt µmr! ¢ksw "ly rµmt!

Septuagint and Peshitta read the Hebrew dglw as a plural imperative di°lū (t£xate, ¢ksw), in line with their interpretation of hby!ny as hă®i!ūnī ‘bring me’ (plural). This is a misguided attempt to harmonize 2:4 with 2:5, in which there are two plural imperatives. It is misguided, since the ‘wine’ in 2:4 symbolically means ‘love’ (cf. 1:2, 4, 4:10, 8:2); only the Man can bring the Woman into this ‘house of wine.’ Making the verb ‘to bring’ plural, the versions trivialize the situation: we are supposed to imagine the Woman going to a tavern for a cup of wine with a company of other people (probably, the Daughters of Jerusalem, cf. 2:7). The MT vocalization is better.155 I return to the word dglw rendered t£xate, ¢ksw by Septuagint and Peshitta. The first uses the Greek verb t£ssw ‘to draw up in order of battle, form, array, marshal,’ frequently used of troops; the second uses the Syriac ¢ks ‘to order,’ ultimately derived from the same Greek verb (through the noun ¢ks! = t£xij). Most probably both versions understand dgl as a

155 Against Graetz 1871:103–106, 138–139; Jastrow 1921:170–171; Schmökel 1956:111 (all of them read both hby!ny and dglw as plural imperatives, with Septuagint and Peshitta) and Müller 1992:24–25, who reads only hby!ny in this way.

362

Articles: Old Testament Studies

denominative verb derived from de°el ‘military unit’ and meaning ‘to make military units, to divide and order troops.’ In contrast to Septuagint and Peshitta, both Aquila and Vulgate render dglw with a 3 sg. masc. Perfect form: œtaxen, ordinavit. At first sight, one could think these versions had a different Vorlage—namely, dgl without w. Such a variant is attested in one Masoretic manuscript.156 However, it is also possible (perhaps, more likely) that they simply read ‘his military unit on me (is) love’ and then tried to make some sense of this awkward construction.157 The sense, as they perceive it, is that the Man’s troops, which are his love, are marshaled by him against the Woman. Symmachus is the only ancient version which does not derive dglw from de°el ‘military unit.’ Instead, Symmachus equates dgl with the Aramaic dgr ‘to pile up,’ hence: ™piswreÚsatš moi ¢g£phn ‘pile up (plural) love for me,’ a curious but senseless rendering. To conclude, Gray rightly observed that the ancient versions usually understand de°el as ‘military unit’ and that they never translate it with a word meaning ‘banner.’ ‘Military unit’ was also the standard interpretation in the midrashim.158 6. A description of de°el ‘banner’ in Pseudo-Jonathan However, Gray failed to take notice of an interesting description of Judah’s de°el in the Targum Pseudo-Jonathan to Num 2:3.

tlt lbq-lk !ynwwg tlt tlymm hwh hysqyjw qyqx hybw aqwrbw aqwryw aqwms anvwxbd atyylgrm !wlwbzw rkvvy hdwhy ayjbv tlt thmv vrpmw !wqyr[yw $ans !wrdbtyw yyy ~wqy bytk hytw[ycmbw !wwyra rb trwc qyqx hwh hybw $mdq-!m $bbd-yl[b bdnym[ rb !wvxn hdwhy ynbl abrd lwjm And its banner was made of wool of three colors, as there were three pearls in the breastplate: carnelian, chrysolite, and emerald. And on it were clearly written the names of three tribes: Judah, Issachar, and Zebulun. And in the middle of it was written: “The

156

R264 (Garbini 1992:42, 45). In his 1983 article Garbini accepted this reading and rendered ‘ha schierato conro di me Amore’ (Garbini 1983:28–29). Subsequently he proposed to emend dgl to dglh, understanding it intransitively and rendering: ‘Amore si è schierato in battaglia contro di me’ (Garbini 1992:200). 157 This is what two modern scholars have also tried to do: Joüon 1909:155– 156 (‘et son armée contre moi est l’amour’) and Provan 2001:285 (‘The “military force” that looks after her is his love’). 158 Gray 1898:92–101.

Y. Eidelkind, Intended Lexical Ambiguity in the Song of Songs

363

will arise, and his enemies will be scattered, and his foes flee before him.” And there was an image of a young lion on it, for the leader of the people of Judah was Nahshon son of Amminadab. LORD

To be sure, the Hebrew word de°el is rendered here ¢ks! (< t£xij), whose primary is ‘military unit.’ But the description itself suggests that ¢ks! = de°el is a tricolour standard made of wool. At the very least, de°el ‘banner’ is not a scholarly fiction. Was it an invention of the Targumist? 7. Dgl ‘sign’ at Qumran After Gray’s 1898 article, new evidence has appeared, which was unavailable to Gray. Strangely, it is not mentioned by Pope or in any of the recent commentaries on the Song of Songs known to me. In an Aramaic fragment of the book of Enoch found at Qumran (4Q Ena ar II.2–3, 6–7) Enoch summons people to observe dgly qy¢h and dgly śtw!. h

db[b wnn[wbt]aw h[ral [wz]x wzx [!kl] a[z]xtm lkw hynXa[l al ~[]d[n]md hnrx[al hymdq !m] !ym almtt] a[ra [lk]d awtX ylgdbw hyl[ ... [... hjyq] ylgdl !m[ arb ...] !yXybym !hlk [!hyn]lya lkd wzx !ykpX hrjm hnn[[w !ynX tltw !ytrtd[ d[ ...] ymyqtm !hyl[d [!]ynlya rs[ t[bra llj !tnaw hqlXw h[ywk !hb hXmX yd hjyq] ylgdl !kl wzx [!rb[y] hrp[ l[ $rd[mlw htywk a[ra ypna l[] hymdq !m !y[b !yrtsmw !hb !hyl[[!w]cynty !hlk hynly[a lkb ...] !m !wxkXt al hy[pk] l[[]w !yla hydb[ lk wnnwbta[w w]l[lh] hxbXt rd[hl ...] !ypxw !yqrwy hrmm !ydb[ !hlk[w ...] hnX !yla hydb[ lk db[ !yml[d ~l[l awh [...] ‘Ob[serve] the earth and con[sid]er /its/ works [from the first to] the [l]ast, how [n]one changes and everything is vis[ib]le [to you.] Observe the signs of [the summer: …] … above it. And the signs of winter: how [all] the earth [is filled with water and] the clouds pour rain. Observe how all [the] tre[es] wither altogether [and lose all their leaves, apart] from fourteen trees whose leaves remain, [and do not renew their leaves until] two or three years [go by]. Observe for yourselves the signs of [summer: how then (in it) the sun burns] and warms and you look for shade and relief from it [upon the scorching face of the earth] without finding a way of walking on the dust or [o]n the [sto]nes because of [the heat. Observe and understand all the t]rees; on all of them their leaves sprout, turn green, and cover [the trees, and all their fruits are for spl]endid praise. [Pra]is[e and consider all these works [and realise that God, who lives] for everlasting eternity, has made all these works. Year [after year his works do not change, instead] they all carry out his word.’159 159

Translation from García Martínez–Tigchelaar 1999:401.

364

Articles: Old Testament Studies

Here dgl most probably means ‘weather sign,’ ‘token of changes of weather and times.’160 Another interpretation cited by Black, ‘time divisions,’ seems here less appropriate.161 In a Hebrew text from Qumran dgly yr[µym] are mentioned before kwkby ‘stars [of heaven]’ (4Q RitMar 27:3). Baillet translates ‘les troupes des mo[is …]’ and notes that dgl “s’applique en partculier aux anges dans la littérature rabbinique (Jastrow 1921:280). S’agit-il donc ici des groupes d’étoiles qui marquent les mois?”162 One can arrive at the same interpretation (“constellations marking the months”) in a more simple way, assuming ‘sign’ as the basic meaning;163 cf. Greek sÁma. A similar expression dgly µwdšym appears in 4Q Bera 1:9, after šbw"y qwdš ‘holy weeks.’ According to Nitzan, these dgly are “heavenly liturgical hosts who praise God at appointed times.”164 Again it is possible that they are constellations marking the months.165 The same applies perhaps to dgly !wr (4Q PrQuot 7:4, 10:2), dgly lylh (4Q PrQuot 29:11, 19), dgly "rb wbwqr (4Q PrQuot 39:3); cf. 4Q PrQuot 1:4, 64:9.166 The Qumran evidence makes it clear, first, that “military unit” was not the only meaning of de°el. Second, it demonstrates that de°el could mean different types of signs, not only military banners. This prompted me to suggest above that ‘his sign/token’ may also be a legitimate interpretation of Song 2:4. As for the meaning ‘military unit,’ it is best explained as a case of metonymy, cf. Greek shma…a and Latin vexillum. References Barbiero 2004 Bergant 2001 Beyer 1984 Black 1985

Barbiero, G. Cantico dei Cantici (LBPT 24). Milano. Bergant, D. The Song of Songs (Berit Olam). Collegeville. Beyer, K. Die aramäischen Texte vom Toten Meer. Göttingen. Black, M. The Book of I Enoch (SVTP 7). Leiden.

160 Black 1985:111; similarly Beyer 1984:233–234; Uhlig 1984:511; García Martínez–Tigchelaar 1999:401. 161 In 4Q Enastrb ar 28.1 ([l]m"dyhwn lµdšyhwn ldglyhwn) it is less clear whether dglyhwn means ‘their signs’ (Beyer 1984:257; García Martínez–Tigchelaar 1999: 439) or ‘their batallions’ (Black 1985:418; cf. Uhlig 1984:669: ‘Ordnungen’). 162 DJD VII 89. 163 García Martínez–Tigchelaar 1999:997. 164 DJD XI 13. 165 “Signs of the months” in García Martínez–Tigchelaar 1999:645. 166 Baillet understands here dgl as ‘troupe ou groupe d’étoiles’ (DJD VII 107), in this case followed by García Martínez–Tigchelaar 1999:999–1007 (‘companies’).

Y. Eidelkind, Intended Lexical Ambiguity in the Song of Songs Black 2000a Black 2000b

Bloch–Bloch 1995 Brenner 1990

Budde 1898 Callow 1994

Clines 1994 Cooke 1903 Dalman 1987 Delitzsch 1875

Döpke 1829 Driver 1936 Eidelkind 2006 Ewald 1826 Exum 1973 Exum 2005 Fox 1985 Garbini 1982 Garbini 1983

Garbini 1992 García Martínez– Tigchelaar 1999

365

Black, F. C. Beauty or the Beast? The Grotesque Body in the Song of Songs. BibInt 8:302–323. Black, F. C. Unlikely Bedfellows: Allegorical and Feminist Readings of Songs of Songs 7.1–8. Brenner, A.; Fontaine, C. R. (eds.). The Song of Songs (FCBSS 6). Sheffield. Pp. 104–129. Bloch, A.; Bloch, C. The Song of Songs. A New Translation with an Introduction and Commentary. New York. Brenner, A. Come Back, Come Back the Shulammite (Song of Songs 7:1–10). Radday, Y. (ed.). On Humour and the Comic in the Hebrew Bible (Bible and Literature 23). Sheffield. Pp. 251–275. Budde, K. Das Hohelied erklärt. Die Fünf Megillot (KHC 17). Freiburg i. B. Pp. IX–48. Callow, J. Units and Flow in the Song of Songs 1:1–2:6. Bergen, R. D. (ed.). Biblical Hebrew and Discourse Linguistics. Winona Lake. Pp. 462–488. Clines, D. J. A. Why is There a Song of Songs and What Does It Do to You If You Read It? Jian Dao 1:1–27. Cooke, G. A. A Text-Book of North-Semitic Inscriptions. Oxford. Dalman, G. Arbeit und Sitte in Palästina. Hildesheim–Zürich–New York. Delitzsch, F. Hoheslied und Koheleth (Keil, C. F.; Delitzsch, F. Biblischer Kommentar über das Alte Testament. Theil IV. Bd. IV.) Leipzig. Döpke, J. C. C. Philologisch-critischer Commentar zum Hohen Liede Salomo’s. Leipzig. Driver, G. R. Supposed Arabisms in the Old Testament. JBL 55:105–120. Eidelkind, J. Two Notes on Song 4:12. B&B 3:217–236. Ewald, H. G. A. Das Hohelied Salomos. Göttingen. Exum, J. Ch. A Literary and Structural Analysis of the Songs of Songs. ZAW 85:47–79. Exum, J. Ch. Song of Songs: A Commentary (OTL). Louisville. Fox, M. V. The Song of Songs and the Ancient Egyptian Love Songs. Madison. Garbini, G. La datazione del “Cantico dei Cantici”. RSO 56:39–46. Garbini, G. Poesia alessandrina e “Cantico dei Cantici”. Bonacasa, N.; Di Vita, A. (eds.). Alessandria e il mondo EllenisticoRomano. Studi in onore di Achille Adriani. Roma. Pp. 25–29. Garbini, G. Cantico dei cantici. Testo, traduzione, note e commento (Biblica. Testi e studi 2). Brescia. García Martínez, F.; Tigchelaar, E. The Dead Sea Scrolls: Study Edition. Leiden.

366

Articles: Old Testament Studies

Gerhards 2010

Gerhards, M. Das Hohelied. Studien zu seiner literarischen Gestalt und theologischen Bedeutung (Arbeiten zur Bibel und ihrer Geschichte, 35). Leipzig. Gerleman, G. Ruth. Das Hohelied (BKAT 18). 2. Aufl. Neukirchen-Vluyn. Ginsburg, C. D. The Song of Songs. London. Gordis, R. The Root lgd in the Song of Songs. JBL 88: 203–204. Gordis, R. The Song of Songs: A Study, Modern Translation and Commentary. New York. Gordon, C. H. Asymmetric Janus Parallelism. EI 16:80–81. Goulder, M. D. The Song of Fourteen Songs (JSOT Sup 36). Sheffield. Graetz, H.-H. Schir Ha-Schirim oder das Salomonische Hohelied. Wien. Gray, G. B. The Meaning of the Hebrew Word lg *!a- in Epigraphic Geez.

392

Articles: Semitic Studies

morphological context (RIE 187:4; less important is the evidence from RIE 232:9–11, which belongs to a later period), against !ay- registered thrice. This makes the hypothesis of free variation between !i- and !ayundesirable—one cannot fail to notice that !ay- is preferred before the prefix y3- of 3 masc. sg. imperfect/jussive, and !i- elsewhere (nothing can be said, of course, about those verbal forms whose combinations with negative morphemes are not documented in EG, such as 2 pers. imperfect/jussive). In principle, an even stricter distribution between !i- and !ay- could be postulated, restricting the latter to only one position, viz. 3 masc. sg. of the jussive. To make feasible such a hypothesis, one would have to interpret y3tmawwā! in RIE 188:5, RIE 189:4, and RIE 189:6 as jussive rather than imperfect. Such a usage in itself is unproblematic: at least once in the epigraphic corpus (RIE 192B:4) the jussive is actually employed in a similar construction (l-!gzbµr / l-¯-d! / ymt / w-!-ybl = *la-!3gzibµer la-za-da! y3mut wa-!ay-y3bli ‘to God, may he not die or grow old (lit. “who may not die or grow old”)’). In this passage, the unvocalized form ymt can only stay for the jussive (*y3mut). This interpretation would imply that in the phrase ‘which is not to be conquered by the enemy’ the verb was variably used in the imperfect (RIE 187:4) or the jussive (elsewhere). Such a variation is, however, rather unlikely within such a clearly stereotype expression: a slight vocalic change in the negative prefix is much easier to imagine. The negative !ay- seems thus to be primarily connected with the verbal prefix y3- of 3 masc. sg. imperfect/jussive. It is tempting to see in EG a transitory stage from the original negative paradigm with a rigid distribution between !ay- (before y3-) and !i- (elsewhere) to that of Classical Geez where !i- is generalized in all positions. The most likely explanation for this distribution is that the allomorph !ay- is a result of dissimilation of -i- before -y3- (!i-y3-> !ay-y3-)12. It seems likely, therefore, that the emergence of the allomorph !aywas conditioned by the following sequence -y3-. This assumption allows us to reconstruct this marker for each position of the verbal paradigm where y3- is present, viz. 3 masc. sg. and 3 masc. and fem. pl. imperfect/jussive (the pertinent negative forms are not attested in the EG corpus). However, already in the early Geez inscriptions !i- began to penetrate even this position. In later periods, the allomorph !ay- was not in use any more. 12

On the sequence *-īy and various means of avoiding it in Semitic v. Brockelmann 1908:250–251.

M. Bulakh, Negative markers *!ay-, *!i- and *!al- in Ethio-Semitic

393

It is possible that such a distribution had existed already in proto-ES. On a later stage, the allomorph !i- was generalized in Classical Geez and Tigre, and the allomorph !ay-, in Tigrinya. 2.2. General negative markers in SES 2.2.1. Negative markers used with verbs Traditionally, the negative marker *!al- is thought to underlie all negative verbal forms throughout SES and hence reconstructed as the only general negative marker for proto-SES (Cohen 1931:35, Hetzron 1972: 21). Within an alternative approach developed by Hudson (2003), the negative marker *!al- in SES is present only in the perfect, whereas in the imperfect/jussive the underlying negative marker is *!ay-. To be sure, the reconstruction of the negative marker *!ay- for the imperfect/jussive in SES is not improbable. Indeed, in the SES negative paradigm of the imperfect/jussive the element -l- surfaces in 1 sg. only (Amh. al-säbr-3m ‘I do not break’), being allegedly assimilated to the following consonant elsewhere: at(t3)-säbr-3m (2 masc. sg.), at(t3)-säbri-m (2 fem. sg.), ay-säbr-3m (3 masc. sg.), at(t3)-säbr-3m (3 fem. sg.), an(n3)-säbr-3m (1 pl.), at(t3)-säbru-m (2 pl.), ay-säbru-m (3 pl.). More precisely, it is usually claimed that *-l- is palatalized into -y- before *y3- (with a subsequent simplification of -yy3- into -y-, cf. Brockelmann 1908:262) and assimilated to t- and n- in *t3- and *n3- with a subsequent (optional) degemination and syncope. In 1 sg., the development is assumed to be *!al-!3- > *al3- > al-, i. e. loss of the gutturals in agreement with the general trend of SES and subsequent syncope of -3-. As far as the 2/3 person sg./pl. and the 1 person pl. forms are concerned, the choice between *!al- and *!ay- in the underlying form can indeed be considered debatable: there is no immediate reason for the traditional *al-t- > at- to be preferred to Hudson’s *ay-t- > at-. But the 1 sg. form *!al- is, at first sight at least, by far less ambiguous, and it is this form that provides the most serious argument in favor of the traditional reconstruction and against Hudson’s thesis. Being aware of this difficulty, Hudson rejects the traditional understanding of -l- in the negated forms of 1 sg. as part of the negative morpheme *!al-. Instead, he considers it as a marker of 1 sg., identical to the element -l- in the 1 sg. prefix of the jussive (Amh. l3sbär ‘let me break’).13 13

For the SES element -l- in the 1 sg. prefix of the jussive and its parallels elsewhere in Semitic v. Huehnergard 1983:580.

394

Articles: Semitic Studies

Hudson does not deny that l in the jussive need not immediately bear on the diachronic interpretation of the negation in the imperfect. He nevertheless observes that in some SES languages (notably in Argobba) the element l is present also in 1 sg. and pl. of the affirmative imperfect: Arg. 3lsäk3r ‘I am drunk.’14 For Hudson (2003:212), the element -l- in both the jussive and the imperfect is a reflex of the proto-Semitic asseverative particle la-, extensively discussed in Huehnergard 1983. An immediate objection to Hudson’s hypothesis is that l-forms in the affirmative imperfect—unlike l in the jussive and the negative imperfect—are by no means common SES. In view of this circumstance, Hudson’s hypothesis must force one to assume that the reflex of the asseverative *la-, originally present in the 1 sg. affirmative imperfect and jussive, was taken over to the negative imperfect but, for some reason, subsequently disappeared in the affirmative imperfect in most SES languages. Such a development is clearly very unlikely. Moreover, Hudson’s claim that “ ‘asseverative l ’ ... was naturally favored in emphatic negative, 1st person” is not compatible with the evidence of other Semitic languages, where the asseverative particle *la- “was confined to main clause assertions, and did not co-occur with negative particles” (Huehnergard 1983: 592). Thus, Hudson’s reconstruction turns out to be highly artificial in comparison with the more straightforward and economical traditional explanation which presumes that the negative particle *!al- was used in the imperfect as well as in the perfect. Hudson’s other arguments against the traditional view (2003:212) are all concerned with the optional gemination of the consonant of the personal marker in 2 pers. and 1 pers. pl.: Amh. at(t3)-säbr-3m ‘do not2 fem. sg. break’, etc. As just mentioned above, within the traditional approach this gemination is thought to result from the assimilation of *-l- and the forms without gemination are explained as due to secondary simplification. Hudson, on the contrary, considers the geminated forms to be secondary, emerging in order to avoid the contact between the consonant of the prefix and the first radical which risked to result in full assimilation: atsäb3r > *assäb3r. Hudson’s explanation is supported by the fact that secondary gemination of the prefix consonant is attested elsewhere in the Amharic verbal paradigm: consider, for example, the optional gemination of -n- in the 1 pl. affirmative form (3n(n3)säb3r ‘we break’) or in the affirmative subordinate verbs (b3t(t3)säb3r ‘if you break’). However, this 14

After prefixes also in Zway, cf. fn. 48.

M. Bulakh, Negative markers *!ay-, *!i- and *!al- in Ethio-Semitic

395

explanation undermines the reconstruction of *!ay- exactly in the same way as it does with *!al-. Indeed, if the forms with gemination are secondary, the last consonant of the negative marker must have disappeared without leaving a trace, which would be highly atypical for both -l- and -y-. “[T]he reasonable loss of y before consonants” posited by Hudson (2003:215) is difficult to accept in the absence of compelling parallel examples15. In short, the hypothetical shift *!ay-t3säb3r > *!atsäb3r faces exactly the same difficulties as *!al-t3säb3r > *!atsäb3r. At the same time, Hudson’s seven arguments in support of his own analysis (2003:213–217) are by no means flawless. Three of these arguments rely on the general assumption that loss of -y- is more plausible that loss of -l-. For Hudson, fluctuating gemination in 2 pers. and 1 pers. pl. in Amharic (Argument 1) as well as absence of gemination in the corresponding forms in Gafat (Argument 5) and Selti (Argument 7) all suggest that the consonant of the negative morpheme was omitted rather than assimilated and, consequently, point to *!ayrather than *!al- in proto-SES. As pointed out above, this reasoning is questionable: simplification of gemination is at least as plausible as complete and traceless loss of -y-. Argument 2 focuses on the prefixes ay- and ä- used with the negative copula (*dbl) in Amharic and Soddo respectively, as well as on the prefix a- before the perfect forms of *kona in the Argobba variety of Aliyu Amba. As will be shown below (3.3), the Amharic and Soddo negative copulas are 3 masc. sg. verbal forms whose conjugation originally followed the paradigm of the imperfect rather than that of the perfect. In such a context, the use of *!ay- is only to be expected. As for the prefix a- in Argobba, it can be reasonably traced back to *!al- (v. below, 3.4). Argument 3 derives from Hudson’s important observation that the negative marker *!i- is not entirely absent from SES, but can be detected in the paradigm of the locative verb *hallawa. Within Hudson’s approach, the hypothetic *!ay- in the imperfect represents still another (and by far more broadly present) SES remnant of the common ES negative marker *!ay-/*!i-. For Hudson, the presence of the negative marker *!i- in the perfect of *hallawa favors the presence of *!ay- in the imperfect of all verbal lexemes in SES. However, the usage of *!i- in combination with 15

Hypothetic monophtongization of *ay into e cannot be relevant as it would involve change of the vowel quality, which is not the case. Nor does the variation ay ~ a in aydälläm — adälläm ‘he is not’ provide a good analogy: the unique phonetic processes affecting this form are undoubtedly conditioned by its high frequency.

396

Articles: Semitic Studies

*hallawa is a clear example of a lexically restricted negative morpheme and may well have functioned as such already in proto-SES, where— as the cumulative evidence of the attested SES languages plainly suggests— all other verbs employed the general marker *!al-. In other words, the presence of the negative *!i- in the paradigm of one single verb is an important piece of evidence for the proto-ES reconstruction, but it does not tell us anything on the distribution of general negative markers in SES. In his Argument 4, Hudson claims that the prefix an- (instead of the expected al-) in the 1 sg. negative imperfect/jussive in Harari (coinciding with the 1 pl. form) can only result from the combination of the negative a- (ay-) and the reflex of asseverative *l3- (> n3-). In his opinion, the shift l > n in Harari is restricted to the 1 sg. affirmative jussive (n3sbär < *l3sbär). When a similar shift is observed in 1 sg. negative imperfect, it means, for Hudson, that we are faced with one and the same morphological element. However, sporadic change from l to n is found in various contexts throughout ES (cf. Podolsky 1991:48 for Amharic), and its occurrence in two etymologically unrelated morphemes is not so improbable as Hudson believes. At the same time, merger of 1 sg. and 1 pl. in the negative imperfect paradigms in some ES languages is a matter of fact which can hardly be denied: note the negative subordinate paradigm in Selti, where both 1 sg. and 1 pl. display -l- rather than -n-: al-nakt ‘I do not beat,’ al-naktina ‘we do not beat’ (Gutt 1997:923). The same considerations apply to the 1 sg. negative prefix in Soddo (ann3-). Argument 6 deals with the main-verb negative imperfect/jussive paradigms of Soddo and Gafat and the presumably related affirmative paradigm of the subordinate verb in Zway. Although these three paradigms display the element -l- in 1 sg. vs. gemination of t/n in 2 and 1 pl., Hudson convincingly shows that there is no negative morpheme *!al- in the background of such forms. This fact, however, is no argument for reconstructing *!ay- as the negative marker in proto-SES: in each of these languages, the common SES negative paradigm is also in evidence (in Soddo and Gafat restricted to the subordinate clauses). None of the arguments brought forward by Hudson is thus solid enough to postulate *!ay- (rather than *!al-) as the underlying form of the negative marker in the imperfect/jussive paradigm of SES. However, still another potential piece of evidence in favor of *!ay- in at least some of the imperfect forms might be detected in the Argobba variety of àollaha.

M. Bulakh, Negative markers *!ay-, *!i- and *!al- in Ethio-Semitic

1 2 masc. 2 fem. 3 masc. 3 fem.

sg. assäm"3m att3säm"3m att3säm"3m isäm"3m att3säm"3m

397

pl. ann3säm"3m att3säm"um isäm"um

The NEG:3MS/PL prefix in the imperfect is i-, opposed to aC- elsewhere. Does this prefix i- show that the underlying morpheme is *!ay-/!irather than *!al-? Such a proposal, however tempting, must nevertheless be rejected. As long as one accepts Hudson’s reconstruction of *!ay- as the general negative marker in the imperfect/jussive in proto-SES, the cumulative evidence of all SES languages except Argobba of àollaha would point to *!ay- rather than *!i- in the prototype. Moreover, the Argobba picture can hardly be considered archaic: the hypothetic distribution of *!ay- and *!iin the Argobba imperfect paradigm does not match the distribution observed in the corresponding paradigm of Epigraphic Geez (2.1.1), but is rather exactly the opposite: Argobba shows i- in the only position where EG has *!ay-, and displays aC- (potentially traceable to *!ay-) elsewhere, (against *!i- in EG). The i-vowel in Argobba of àollaha must therefore be considered a secondary development from *!ay-.16 As for the underlying consonant of the negative morpheme aC- in the other forms of the imperfect, it could be either *-l- or *-y-: for both of these phonemes, assimilation to the adjacent consonant is easy to imagine 16

Such a development could be tentatively explained by the necessity to avoid the homonymy between the NEG:1S and NEG:3SM morphemes. Since assimilation of the consonantal element of the negative marker in Argobba of àollaha favored the evolution of *!ay- into aC-, the resulting paradigm would have lacked any morphological distinction between 1 sg. and 3 sg. masc. The shift from *!ay- to *!i- blocked the assimilation and made impossible this undesirable development. A similar reasoning could be applied to the negative paradigm of the imperfect/jussive in Epigraphic Geez (2.1.1), which, as just observed above, structurally resembles the negative paradigm of Argobba of àollaha. In Geez, the negative prefix !i- triggers the dissimilation of the 1 sg. prefix !3- into y3-, which becomes homonymous with the 3 sg. imperfect/jussive prefix y3- (!i-y3sabb3r ‘I do not break’/‘he does not break’). Since there are good reasons to believe that this phonological rule was operative in early Geez, the use of a special negative allomorph in 3 masc. sg. could be regarded as an attempt to avoid the ambiguousness of forms like !i-y3sabb3r. However, the very fact that homonymy between 1 sg. and 3 masc. sg. in these forms is tolerated in Classical Geez speaks against this hypothesis.

398

Articles: Semitic Studies

in this variety of Argobba, where the tendency towards assimilation of the consonantal element of the negative marker seems to be stronger than anywhere in ES. There is nothing to suggest that the negative prefixes in the imperfect of proto-Argobba were to any serious extent different from those found in Argobba of Aliyu Amha or in Amharic, that is, *!al- for NEG:1S and *!ay- for NEG:3SM. As soon as this conclusion is accepted, one has to acknowledge that the evidence of Argobba of àollaha cannot be used as an argument in favor of *!ay-/*!i- as the general negative marker of the imperfect in proto-SES. And, conversely, diachronic explanations proposed for the emergence of the negative morphemes in the imperfect paradigm of Amharic turn out to be valid also for proto-Argobba: in both languages, the NEG:3MS element *!ay- can, as we have seen, result from either *!al-y3- or *!ay-y3-. Summing up, the evidence of the SES languages does not provide any persuasive argument in favor of reconstructing *!ay- as the imperfect/jussive negative marker in proto-SES. Rather, the element *!al-, surfacing in 1 sg. throughout SES, speaks against such a reconstruction. However, in view of the evidence of Epigraphic Geez the picture turns out to be more complicated. In section 2.1.1 above I have tried to reconstruct for Epigraphic Geez (more tentatively, also for proto-ES) a complementary distribution between the allomorphs *!i-/*!ay-, the latter appearing before the personal prefix *y3-, and the former elsewhere in the verbal paradigm. Theoretically, such a distribution can also be postulated for proto-SES, assuming that the allomorph *!i- was replaced by *!al-, whereas *!ay- before *y3remained unchanged. In other words, the element *!ay- (NEG:3MS/PL) in the imperfect/jussive in SES would go back to proto-SES *!ay-y3-, whereas for the rest of the verbal paradigm the negative marker *!alwould be reconstructed. Such a hypothesis, while not wholly improbable, can hardly be proved and is deemed to remain a purely theoretical construct.17 17 The negative paradigm of Argobba of àollaha may be thought to support this reconstruction (the distribution between the allomorphs aC- and i- is exactly the same as one between *!i- and *!ay- reconstructed for Epigraphic Geez), but in fact it does not exactly for the same reasons which have prevented us from using it to support Hudson’s reconstruction: one is forced to assume then that the allomorph *!i- was replaced with *!al- in proto-SES, whereas the change from *!ayto *!i- in 3 masc. sg./pl. took place in Argobba of àollaha only. The latter step,

M. Bulakh, Negative markers *!ay-, *!i- and *!al- in Ethio-Semitic

399

A few other general negative markers (such as !3l-, !3lla- in East Gurage or t- in Gafat and Soddo) are innovative morphemes restricted to just one or several ES languages and have no relevance for the reconstruction of the negative markers of proto-ES. 2.2.2. Negative markers used with non-verbal forms A peculiarity of at least some of the SES languages (unfortunately, for many idioms of the subgroup no exact information on this part of grammar is available) is the employment of a special negative marker for negating non-verbal forms (nouns, adjectives, verbal nouns).18 Transparent manifestations of this phenomenon are extant in Amharic, Zway and Wolane.19 From the synchronic point of view, it is only in these languages that non-verbal forms are negated with a special marker *!ala- (which can be optionally expanded with the relative pronoun *ya-). In Harari and Selti, the morpheme *!al- is used with both verbal and non-verbal forms, in the latter case always preceded by the relative pronoun *z3- in Harari. However, the data of Amharic, Zway and Wolane seem to suggest that the pre-nominal negative element *!al- has only secondarily become homonymous with the pre-verbal *!al- and may also go back to the same prototype *!ala-. The syncope of a short unstressed a in an open syllable is not improbable, although I have to admit that I am not aware of a fully reliable comparable example of such a phenomenon in ES. Still, omission of a short -a- in the course of a syllable structure rearrangement is considerably easier to imagine than (apparently, nonmotivated) insertion of -a- independently in several languages. Some evihowever, is equally compatible with the element *!ay- (NEG:3MS/PL) going back to *!al-y3-. 18 For Tigrinya, Leslau (1941:94) reports a special negative marker zäy- before infinitives. The underlying negative element is still *!ay- (*z3- being a relative pronoun), by no means different from the general negative marker of Tigrinya. 19 Nominal lexemes comprising the negative markers !al-/!ay- in Zway, such as !alčāy ‘ignorant’ < !al-čāl-u ‘he did not know’, !aysor£u ‘granary entrance/aperture’ < !ay-sor£-u ‘he does not steal’ (Meyer 2005:247) are not considered here since their negative markers are taken over from the verbal negative paradigm. For deverbal negative adjectives in Selti a phonological distribution between !ay- and !al- has been reported by Gutt (1997:910), the latter form appearing only before vowels (ayÌēno ‘sterile, barren’ < Ìēñe ‘give birth’, al-ēsābo ‘unexpected’ < ēsaba ‘notice’). One wonders whether this distribution is a further development from a system documented in Zway or whether !ay-/!al- in Selti continue the negative marker *!al(a)- of non-verbal forms (for the possibility of such a development v. fn. 36).

400

Articles: Semitic Studies

dence in support of this reconstruction can be found in the lexically restricted negative *!al(a)- (3.1), which can be identified with the negative morpheme under scrutiny. 3. Lexically restricted negative markers In the discussion of the negative markers occurring with copulas, locative verbs and possessive constructions, attention should be paid to the specific verbs or predicative elements with which the lexically restricted negative markers are associated. At least in some cases, the combination of a given negative prefix with this or that root is ubiquitous in ES and can be safely reconstructed for the proto-language. Negative forms of the pertinent predicative elements are examined below in this section, in an attempt at establishing for each of them its original negative marker. 3.1. Possessive expressions involving *ba + possessive suffix are negated with the particle *!al(a)- in NES (Geez, Tigre of Mensa, Tigre of Habab, Dahalik, Tigrinya) as well as in SES (Gafat). An important question is, however, which of the variants is original: *!ala- (Tigre of Mensa, Tigre of Habab, Dahalik, Gafat) or *!al- (Geez and Tigrinya)? The primacy of the former element appears more attractive: the shift *!alabo > *!albo in Geez and Tigrinya can be explained as a result of vocalic syncope (v. 2.2.2), whereas gemination of -l- in Dahalik (y-alle-bū-ni) and Gafat (alläb-am) can be regarded as “strengthening” of the first syllable in order to avoid the syncope. Furthermore, the reconstructed negative *!ala- is compatible with the negative marker *!ala- combined with nominal forms in some of the SES languages (2.2.2).20 Conversely, transition from *!al- to *!ala- is hardly plausible as there is no immediate reason for the hypothetic insertion of -a- between !al- and -bo. In Dahalik, Tigrinya and Gafat the old negative possessive construction has been expanded with new negative morphemes, a process obviously to be accounted for by desemantization of *!al(a)-. In Dahalik, the negative construction is augmented with the general negative marker y-(...-ni): y-alle-bū-ni ‘he does not have.’ As for Tigrinya, an extensive discussion of the whole spectrum of the expanded forms with a special attention to their historical background can be found in Voigt 1977:237– 250. For the present purpose, it is sufficient to say that the most common 20

For this comparison v., e. g., Leslau 1956:149. Note that in the case of the pre-nominal *!ala- the second vowel is absent (presumably syncopated) in Harari and Selti.

M. Bulakh, Negative markers *!ay-, *!i- and *!al- in Ethio-Semitic

401

form of the negation of existential construction is *!albo augmented with the general negative marker *!ay-…-n: *!ay-!albo-n > yäl-bo-n.21 In Gafat, the element -am (also a part of the general negative circumfix) is imposed upon the old possessive construction. Another peculiarity of this language is that *!alab- has been reanalyzed as a verbal stem and is conjugated through the suffixes of the perfect. 3.2. Negative forms based on *hallawa in the expressions of location, possession, or existence are negated with !i- in Tigre and Geez. It seems reasonable to assume (with Hudson 2003:213–214; v. already Praetorius 1879:259) that the corresponding forms in the rest of ES (Tna. y-ällo-n, Amh. y-ällä-m and, with yä- > e/ē/ī, Har. ēl, Zwy. !īl(l)o, Cha. enä, etc.) go back to the same proto-form. The shift from !i- to y- in Tigrinya and SES can be plausibly explained by a morphophonemic rule !VH- > yVH-, synchronically operative in the negated forms of verbs with initial gutturals in Tigre (*!i-µazze > y3-µazze ‘I do not want’/‘he does not want’).22 One can also surmise that the negative markers y3- and yä- recorded for some NES idioms (v. 2.1) result from generalization of the allomorphs of *!ibefore gutturals. 3.3. Negative forms of the copula based on the root *dbl are found in Amharic, Gafat and Soddo.23 In Soddo (Leslau 1968:13), all forms of the negative copula consist of the preformative ä, the stem of the imperfect 21

The negative form of the possessive construction (yäbillu-n) may also go back to the combination *!al(a)- + ba- + pronominal suffix. It can be explained as a result of metathesis of l and b with a subsequent reinterpretation of l as an applicative suffix (hence, possibly, its gemination). Alternatively, it can be seen as a further development (simplification) of the rarely attested yäl-b3llu-n, clearly derived from *!al-b- + pronominal suffix and expanded with a general negative marker and the applicative suffix -l- (for a detailed analysis v. Voigt 1977:242– 243). In Voigt’s view, however, yäbillu-n rather goes back to a regular negative form of the otherwise unattested possessive construction (a combination of the preposition !ab-, the applicative -l- and a pronominal suffix), v. Voigt 1977:247. 22 Saleh Mahmud Idris, communication made in the framework of the International Workshop “History and Language of the Tigre-speaking Peoples (Eritrea and Sudan),” Naples, 7–8 February 2008. A similar diachronic rule probably accounts for such Amharic forms as yazä ‘to take’ < *!a¶aza, although one has to admit that in some non-verbal roots containing two adjacent gutturals (such as *!aµad- ‘one’ or *!u¶t- ‘sister’) no change !- > y- is registered anywhere in ES. 23 Leslau (1956:77) mentions the form adäb3l for Chaha, which is, however, missing from EDG I and EDG as well as from Hetzron 1977. Note also the obviously related expression aldäbäl ‘except, with the exception of’, found in Soddo and Wolane (EDG 39).

402

Articles: Semitic Studies

and an object pronoun (with the exception of the 1 person, where the suffix of the perfect conjugation is used instead of the object pronoun).24 3 masc. 3 fem. 2.masc. 2 fem. 1

Sg. ä-däb3ll (ä-IMPF:OBJ.3MS) ä-däb3lla (ä-IMPF:OBJ.3FS) ä-däb3kkä (ä-IMPF:OBJ.2MS) ä-däb3čč (ä-IMPF:OBJ.2FS)25 ä-däbukk (ä-PERF:SUBJ.1S)

Pl. ä-däb3lläm (ä-IMPF:OBJ.3MP) ä-däb3lläma (ä-IMPF:OBJ.3FP) ä-däb3kk3m (ä-IMPF:OBJ.2MP) ä-däb3kk3ma (ä-IMPF:OBJ.2FP) ä-däb3llänä (ä-IMPF:OBJ.1P)

The paradigm of the Gafat cognate is very similar except for the addition of the innovative morpheme of the negative imperfect t-…-am (Leslau 1956:77).26 3 masc. 3 fem. 2 masc. 2 fem. 1

Sg. t-ä-däb3ll-am (t-ä-IMPF:OBJ.3MS-am) t-ä-däb3llat-am (t-ä-IMPF:OBJ.3FS-am) t-ä-däb3kk-am (t-ä-IMPF:OBJ.2MS-am) t-ä-däb3čč-am (t-ä-IMPF:OBJ.2MS-am)27 t-ä-däb3kkw-am (t-ä-PERF:SUBJ.1S-am)

Pl. t-ä-däb3lläŭm-an (t-ä-IMPF:OBJ.3P-am) t-ä-däb3kkämw-am (t-ä-IMPF:OBJ.2P-am) t-ä-däb3llän-am (t-ä-IMPF:OBJ.2P-am)

The verbal root *dbl is well known in ES with the meaning ‘to add, repeat, join’; parallels with the meaning ‘to collect, unite’ are found in Arabic and Soqotri (v., e. g., EDG 195–196; for its comparison with the negative copula v. Leslau 1956:77, Goldenberg 1968:74, Hudson 2003:213– 214 as well as EDG 14–15 for an overview of alternative proposals). The semantic development is reconstructed as follows: ‘X is not added to Y’ > ‘X is not counted as Y’ > ‘X is not Y.’28 The Amharic negative forms aydällä-m, ay-dollä-m is generally considered to go back to the same verbal 24

Note that this form has no Main Verb Markers, and that the object suffixes of the 2 person are attached to the stem in a way slightly different from that in which they are normally attached to the imperfect stems (namely, without a vowel inserted between the stem and the pronoun: y3gädl-ä-hä ‘he kills yousg.,’ y3gädl-3h3m ‘he kills youpl.’). 25 Normally, the 2 fem. sg. object pronoun is -š, but the form -čč appears in the impersonal conjugation (Leslau 1968:23–25). 26 As in Soddo, no vowel is inserted between the stem and the 2 pers. object pronoun, although such a vowel usually appears when the object pronoun is attached to a regular verb in the imperfect. The forms of the object suffixes of 2 fem. sg. and 2 pl. deviate from the suffixes employed with regular verbs (-š and -hum < *-kum respectively). 27 In Gafat, both -š and -č are present as the exponents of the 2 fem. sg. object pronoun (Leslau 1956:59–60). 28 Cf. the usage of the Wolane verb dä®äl- in the meaning ‘to be counted as’ or even ‘to become’ in the expression täsäb yädä®älä Ì3lō ‘being circumcised,’ lit. ‘a child which is added to the human beings’ (Meyer 2006:140).

M. Bulakh, Negative markers *!ay-, *!i- and *!al- in Ethio-Semitic

403

root *dbl (with the shift *-ab- > *-aw- > -o-). Indeed, the Old Amharic evidence (Getachew Haile 1970:75) favors such a comparison, revealing both the original -o- and the object suffixes rather than the perfect subject suffixes attached to the base. Relying on the actual forms appearing in the old manuscripts, as well as on the comparative data from Gafat and Soddo, one can propose the following paradigm of the negative copula in Old Amharic. 3 masc. 3 fem. 2 masc. 2 fem. 1

Sg. !ay-dol(l)äw-3m (*!al-IMPF:OBJ.3MS-m) !ay-dol(l)at-3m (*!al-IMPF:OBJ.3FS-m) !ay-dol(l)äh-3m (*!al-IMPF:OBJ.2MS-m) !ay-dol(l)äš-3m (*!al-IMPF:OBJ.2FS-m) !ay-dol(l)ähu-m (*!al-PERF:SUBJ.1S-m)

Pl. !ay-dol(l)aččäw-3m (*!al-IMPF:OBJ.3P-m) !ay-dol(l)aččhu-m (*!al-IMPF:OBJ.2P-m) !ay-dol(l)än-3m (*!al-IMPF:OBJ.1P-m)

This reconstructed paradigm differs from the one proposed in Getachew Haile 1970:75 in one position, viz. 1 sg.: the form *!aydolläññ3m ‘I am not’ reconstructed by Getachew Haile is neither found in the manuscripts nor confirmed by data from Gafat and Soddo. It is reasonable to assume that the Old Amharic paradigm of the negative copula which immediately preceded the modern one was similar to that of Gafat and Soddo. Of course, this does not exclude the possibility that the paradigms of Old Amharic, Soddo and Gafat, in their turn, go back to a paradigm which had object suffixes in all positions (similar to the one offered by Getachew Haile). The substitution of the object pronominal suffixes with the perfect subject suffixes which one can thus observe in the history of the negative copula in Amharic is undoubtedly related to the similar replacement in the affirmative copula, postulated by Hetzron (1972:80–81) for the majority of SES languages.29 Interestingly, this replacement did not occur simultaneously in the affirmative and negative copulas. As we have seen, in the negative copula based on the root *dbl this development took place in Amharic only. As for the the affirmative copula, it has preserved the original suffixes in Amharic, Argobba and Gafat, whereas in the rest of SES they were replaced with the perfect subject suffixes.30 29

In other words, the Amharic negative copula provides an additional argument in support of Hetzron’s claim that the object suffixes in the copula are more archaic. 30 The use of subject (perfect) and object pronominal suffixes within one and the same paradigm of a negative copula is also registered for Argobba of Aliyu Amba (Leslau 1997:35–36, v. below, 3.4), where the object suffixes are found in 3

404

Articles: Semitic Studies

It becomes clear that the subject suffixes of the perfect paradigm in modern Amharic represent a relatively recent innovation: originally, as suggested by the evidence of Gafat and Soddo, the morphological form of this copula was that of 3 masc. sg. of the imperfect, which agreed with the logical subject by means of the object pronoun. This reconstruction undermines Hudson’s reasoning according to which the Amharic form ay-dällä-m represents a rare (and, presumably, archaic) case of the negative element ay- combined with the perfect and provides a direct link between SES *!ay- (not traceable to *!al-) and the main negative marker !ayin Tigrinya (Hudson 2003:214). The data presented above suggest that the prefix ay- in the Amharic negative copula is diachronically the regular Amharic negative prefix of 3 masc. sg. imperfect: ay-dällä-m/ay-dollä-m is thus not different from ay-säbr-3m. In the corresponding forms of Gafat and Soddo the prefix *!ay- does not surface, however: the shape of the base under scrutiny is ädäb3l-31 and it can hardly be a coincidence that both languages display -ä- before the first radical, which is difficult to explain synchronically. What is the origin of this vowel? For Hudson (2003:214), the initial element ä- in Soddo is a continuation of *a- < *ay-. Since the alleged shift *a- > ä- is difficult to explain, one may wonder whether the development *ay > *e > ä could provide a plausible alternative. Such a shift, however, would also be unique since the combination ay- is usually preserved in both Soddo (ay-£ärs ‘he does not begin’) and Gafat (ayfär3k ‘he cannotsubordinate’). An irregular simplification of *ay is not that improbable for a grammatical word of high frequency. Indeed, the Amharic form adälläm is to be regarded as an independent example of a similar simplification from a more regular aydälläm. We may conclude that the negative copula based on the root *dbl originally consisted of the imperfect base preceded by a vocalic element (probably going back to a combination of a personal prefix of the imperfect and of a negative marker) and followed by an object suffix which agreed with the logical subject. This may be true for all forms of the negative copula, although in 1 sg. the object suffix is replaced by the subject suffix of the perfect in all attested paradigms. sg. and pl. and the subject suffixes of the perfect elsewhere. In Argobba of àollaha, only subject suffixes are used (A. Wetter, p. c.). 31 The only difference between Gafat and Soddo is that the Gafat form also displays the new negative prefix t-, obviously imposed on the older negative copula as soon as its original negative marker became desemantized.

M. Bulakh, Negative markers *!ay-, *!i- and *!al- in Ethio-Semitic

405

3.4. Negative copulas based on the verb *kona usually attach general negative markers rather than any lexically restricted markers. The only exception is Argobba of Aliyu Amba, where the negative copula has the form a-huney-u, apparently with a negative prefix a- instead of the expected general marker al- or the preformative ay- (NEG:3MS). Hudson (2003:214) seems to interpret this prefix as a continuation of the negative marker *!ay- (rather than *!al-), but the data from Argobba of àollaha permit one to reconsider this question from a different angle. Indeed, the paradigms of the negative copula of Argobba of Aliyu Amba and Argobba of àollaha are structurally similar: the base of the jussive is followed by an object suffix in 3 masc. sg. and pl. and by a subject suffix of the perfect in the rest of the paradigm. In Argobba of àollaha, the negative marker used with this combination is *!al- (-l- sometimes surfacing, but mostly assimilated to the first radical of the verb), in agreement with the general pattern of negation of the perfect conjugation in this idiom. It stands to reason that the element a- in the negative copula of Argobba of Aliyu Amba must go back to the same negative morpheme, *l being first assimilated to the initial radical and then dropped as a result of simplification of geminated consonants (v. fn. 40). 3.5. It is generally believed that the indeclinable negative copula !akko in Geez derives from the verb kona ‘to be, become’ combined with the negative marker *!al- (CDG 14 with further references). This etymology seems to be supported by the Tigre negative copula !i-kon which, as the Geez one, is indeclinable, but preserves the final -n and is thus clearly related to the verb kona. At the same time, it is hard to refrain from comparing Geez !akko with Zway !unku and Selti inko, very similar to it in both form and function.32 The vowels of the first syllable in East Gurage point to original *3,33 which makes it difficult to derive !un-/in- from *!al-. Instead, these elements are to be identified with *!3n-, a marginally attested but doubtless archaic negative marker found in such fossilized forms as Gez. !3n-dā"i, Tna. !3n-d3"i, Amh. 3n-ğa ‘I do not know,’ Gez. !3n-bi, !3n-b3ya ‘no,’ Tna. !3m-bi bälä ‘refuse’ (Leslau 1969b:140).34 If this etymology is correct, one has to choose between *!3n-ko and *!al-ko as two alternative proto-ES reconstructions. 32

The form !unku in Zway is probably to be explained by a mutual assimilation of the first and second vowels (*!3n-ko > !unku). 33 For the variation between 3, i and u in Zway v. Meyer 2005:46. In Gutt’s transcription of Selti, i is the regular reflex of 3. 34 Leslau further compares the element 3n-, combined with 2 person forms of the perfect and expressing prohibition in Muher, Mesqan, and Western Gurage,

406

Articles: Semitic Studies

The replacement of *!3n- by *!al- is indeed not improbable since *!al(a)- in proto-ES apparently functioned as the general negative marker in the same positions where *!3n- probably occurred as a lexically restricted one: before nouns (*!3n-dā"ī) and before the preposition *ba(*!3n-ba+POSS.1S). In such a context, identification of *-ko with kona may seem a major problem: why a verbal form should have used negative markers typically associated with non-verbal forms? Still, one can surmise that the use of non-verbal negative markers in this case was conditioned by desemantization of -ko, which already in proto-ES probably began to be perceived as a non-verbal form. The emergence of negative copulas based on kona in Tigre and elsewhere in ES must then be attributed to independent development. 3.6. It is not to be excluded that the negative copula in Harari (alta-m) and the negative locative verb in Argobba of àollaha (yata-m) are related. The element -t- in the Harari form is certainly the same as in the positive copula (with the basis t-, v. Garad–Wagner 1998:4–5). Given the fact that there is no plausible diachronic explanation for Argobba yata-m, one may hypothesize that it is based on the same element -t-, followed by suffixes of the perfect and probably preceded by the general negative marker *!al-, which, as in some other positions (fn. 38 and 40), could be assimilated to -t- with the subsequent loss of gemination. The negative locative verb in Argobba of àollaha is thus structurally identical to the negative copula of Harari (for the shift from copula to locative verb, v. section 1 with fn. 5). One must admit that the initial y- in Argobba of àollaha remains difficult to explain. One might tentatively analyze it as a relative pronoun, whose appearance with a negative copula or a locative verb would, however, be quite unusual. 4. Reconstruction of the proto-ES negative paradigm 4.1. Table 3 (q. v.) gives a summary of the evidence presented and analyzed above. A comparison between general negative markers and lexically restricted negative markers allows one to draw two conclusions about proto-ES negative markers combined with the perfect and with nonverbal forms. but this is rather unlikely. Instead, the Gunnän-Gurage 3n- can be compared to !3l(a)- as the negative marker of the main imperfect in Wolane and Selti.

M. Bulakh, Negative markers *!ay-, *!i- and *!al- in Ethio-Semitic

407

(1) The element *!i- can be reconstructed as the proto-ES general negative marker of the perfect. This function is preserved in Tigre of Mensa and in Geez, whereas in the rest of ES its traces can be detected in the reflexes of the negative forms of the locative verb *hallawa (which, as is well known, has the inflectional paradigm of the perfect when it refers to the present time). (2) Proto-ES probably used different methods for negating verbal forms and other words. The negative marker of non-verbal forms can be reconstructed as *!ala-, which preserves this function in some of the SES languages, but is restricted to negative constructions based on the preposition *ba- in Gafat, Tigre, Tigrinya and Geez. 4.2. My hypotheses concerning the proto-ES negative markers of the imperfect/jussive are considerably more speculative since the data allow several interpretations. The general negative marker of the imperfect/jussive is !ay- in Tigrinya and !i- in Classical Geez and Tigre, whereas in Epigraphic Geez a complementary distribution between !ay- and !i- is observed. As far as proto-SES is concerned, the general negative marker of the imperfect/jussive is to be reconstructed *!al- or, alternatively, a complimentary distribution between *!al- and *!ay- can be posited, structurally resembling the Epigraphic Geez picture. The most straightforward reconstruction based on these data is to assume that the distribution between !ay- and !i- in Epigraphic Geez faithfully reflects the proto-ES heritage. In such a case, various generalizations in the verbal paradigm must have occurred in Tigre, Classical Geez and Tigrinya, whereas in SES the prefix *!al- (presumably going back to the proto-ES negative marker of non-verbal forms *!ala-) must have replaced the allomorph *!i- (or the morpheme *!i-/*!ay- in general) throughout the verbal paradigm. This reconstruction may seem to face some difficulties. It is indeed hard to imagine how the proto-ES negative marker *!ala-, thought to be restricted to the rather marginal function of negating non-verbal forms in proto-ES, could have been generalized as the main negative morpheme in SES. Still, if one takes into consideration some phonological processes which likely affected the morphemes *!ay- and *!i- on their way from proto-ES to SES, the emergence of *!al- in SES can find a satisfactory explanation. The development of the negative paradigm from proto-ES to protoSES is thought to be influenced by two major factors.

408

Articles: Semitic Studies

The first one is the morphophonemic rule !VH- > yVH- which, as shown above (3.2), can be postulated for Tigrinya and SES. The negative marker was thus likely realized as *y3- before the forms of the perfect of verbs with initial gutturals as well as before the 1 sg. prefix !3- of the imperfect/jussive. In the latter case at least, such a realization must have brought about the clearly undesirable development *!i-!3- > *y3!3- > *y3-, with the resulting homonymy of 1 sg. of the negative imperfect/jussive with 3 masc. sg. of the affirmative imperfect/jussive. It may well be that generalization of !ay- at the expense of *!i- in Tigrinya was triggered by the necessity to avoid this homonymy. One has to reconstruct a similar generalization for proto-SES in order to account for the further development: the replacement of *!ay- by *!al- is likely to be attributed to a phonological process (for which v. immediately below), whereas a direct change from *!i- to *!al- can hardly be explained in phonological terms. The second factor potentially accounting for the shift from *!ay- to *!al- in SES is the alternation between -l- and -y- —a phenomenon well attested in SES, but conspicuously absent from Geez, Tigre and Tigrinya. Indeed, palatalization of *-l- into -y- before front vowels and y (as in Amh. bäy ‘sayimv. fem. sg.’ < *bäl-i) is a regular morphophonemic process in SES. The opposite development, viz. hardening of the original intervocalic -y- into -l-, can thus be regarded as hypercorrect interpretation of each intervocalic -y- as resulting from palatalization. This kind of hypercorrection is admittedly rare and sporadic, but several reliable examples can be detected in some SES languages: Amh. abbälä ‘to break one’s promise, promise falsely,’ Har. abäla ‘to refuse’ as opposed to Gez. !abaya ‘to refuse’35 or Amh. £älla ‘to be red’ vs. Gez. £eµa (*£yµ) ‘to be red’ (CDG 456, Cohen 1939:28, Praetorius 1879:48, 75; Podolsky 1991:36). A particularly interesting case is Gaf. alä ‘brother’/al3t ‘sister’ (Leslau 1956:175), likely traceable to PS *!a¶(w)- (*¶ > *y > l). One can assume that the same hypercorrect development took place whenever the negative *!ay- was followed by a vowel. Although such environments in proto-SES were probably not so numerous as they are in most modern SES languages because of the loss of gutturals, there are good reasons to suppose that the shift was at work at least before the etymological *-!-, likely dropped earlier than other gutturals. The path of the development can be reconstructed as follows: *!ay-!V- > *!ay-V- > *!al-V-. The shift from *!ay- to 35

Contra Cerulli (1936:406) and Leslau (EDH 17), who reject this comparison.

M. Bulakh, Negative markers *!ay-, *!i- and *!al- in Ethio-Semitic

409

*!al- was especially prominent in the 1 sg. imperfect/jussive (which otherwise would have become homonymous with 3 sg. masc. imperfect/jussive) and in the dependent verbal forms (where the exponent of the negative marker was least discernible). Indeed, the shift from -y- to -l- is quite conspicuous in some dependent negative forms of the locative verb, cf. the Ancient Harari relative form of ēl (zi-lēl or za-lēl ‘which is not’ < *z3-yällä), Amharic yä-lellä (< *yä-yällä), Argobba of Aliyu Amba yälella (< *yä-yällä). The same process may be responsible for the emergence of the allomorphs y-/-l- of the 1 sg. prefix in Zway (v. fn. 48). It is thus likely that *!al- in the verbal negative paradigm of SES emerged as a phonetically conditioned allomorph of *!ay-. How can one explain its expansion to preconsonantal environments? The existence of the phonetically similar old negative morpheme *!ala-, appearing before all non-verbal forms, including participles and infinitives, must have contributed in no small way to this expansion. Generalization of *!al- can thus be regarded as a merger of two negative markers of different origin— *!ala-, an archaic element used with non-verbal forms, and *!al-, an innovative allomorph of *!ay-.36 The present reconstruction does not presuppose a common NES generalization of *!i-/*!ay- as the principal negative marker throughout the verbal paradigm. Rather, generalizations of *!i- in Tigre and Classical Geez and that of *!ay- in Tigrinya are seen as independent developments from a common proto-ES stage, at which *!i- and *!ay- functioned as complimentary distributed allomorphs within the verbal paradigm. Moreover, generalization of *!ay- is reconstructed for proto-SES as well, as a stage preceding generalization of *!al-.37 In a similar vein, loss of the negative marker *!ala- confined to nonverbal elements appears as a trivial development which took place independently in Tigre, Tigrinya and Geez (most probably, in a few SES languages as well) and, as such, has no classificatory value. Conversely, generalization of *!al- as the general negative marker of the verb in SES turns out to be a reliable common innovation. 36

It is possible that the morphophonemic distribution of !ay- and !al- in negative deverbal adjectives in Selti (v. fn. 19) is due to a similar merger of the preverbal *!al-/!ay- with the negative marker *!al(a)- of the non-verbal forms. Obviously, the process in Selti should be dated at a more recent stage. 37 In view of the absence of *!ay- in Tigrinya of May Tchew it seems wise to refrain from treating *!ay- as a common innovation of Tigrinya and proto-SES.

410

Articles: Semitic Studies

This reconstruction is in agreement with other pieces of comparative evidence connected with the NES/SES dichotomy: while common SES innovations are by no means a rarity, there are hardly any convincing examples of an innovative morphological feature shared by Geez, Tigre and Tigrinya (Bulakh–Kogan 2010).

Table 1. Negation in Ethio-Semitic languages

y3-...-n(ni)/!i-39

y3-...-n(ni)/!i-

!i-sabb3r

!i-sabb3r

Imperfect 3 masc. sg.

Imperfect 1 sg.

(y)i-...-ni (yi-me!eki-ni ‘he did not come to you’) (y)i-...-ni (yi-!ammer-ni ‘he does not know’) ?

Dahalik

Tna.

!i-y3sabb3r

!i-y3sabb3r

al-säbbärä-m ‘he did not break’

Amh.

!ay-säbb3r-3n

al-säbr-3m

!ay-y3säbb3r-3n ay-säbr-3m

!i-sabara !ay-säbärä-n ‘he did not ‘he did not break’ break’

Gez.

al-säkr-u ‘I do not get drunk’

ay-sädb-u

as-säm"-3m (< *!aC-säm"-3m)40

i-säm"-3m

Arg. of Aliyu Arg. of àollaha Amba al-säddäba-w ‘he a-s3m"a-m (< did not insult’ *!aC-s3m"a-m)38 ‘he did not hear’

The negative form of the perfect in Argobba of àollaha employs the base of the jussive, although the suffixes are still those of the perfect (Wetter 2006:902). It is unclear why gemination disappears in this form (an analogy with the corresponding form of the imperfect?). In other forms of the negative perfect paradigm, the first radical is always geminated: ass3m"ewm (1 sg.), ass3m"ex3m (2 masc. sg.). 39 For the distribution between y3- ... -n(ni) and !i- v. section 2.1. 40 The prefix al- surfaces in some verbs (al-mä¢-3m ‘I will not come’, Wetter 2006:906). Forms with the negative prefix a- and non-geminated first consonant adduced in Wetter 2006:902 have been later re-analyzed by Dr. Wetter as having an underlying *al(A. Wetter, p. c.). Indeed, degemination of originally doubled consonants seems to be relatively common in this variety of Argobba.

38

y3-sabra-n(ni)/ !i-sabra ‘he did not break’

!i-sabra ‘he did not break’

Tgr. (Habab)

Perfect

Tgr. (Mensa)

The sign = shows that this context is not treated as a separate morphological category in the language in question; the sign ? shows that no exact data have been found in the grammatical descriptions available to the author. Whenever gemination of the adjacent consonant is involved, no attempt at reconstructing the consonant of the prefix is undertaken, so that the negative morpheme is referred to as *!aC-. The forms are normally adduced as they are quoted in the grammars, with necessary adaptations of the transcription. The negative morphemes (or elements which result from the fusion of negative morphemes with personal prefixes) are in bold.

42

41

!i-kon/!i-kon-i/ !i-kon-ini (3 pers.); y(3)- + cop. (1 & 2 pers.) !i-halla/ya-hallanni bu !i-kon/ !ala-bu

?

!i-kon (3 pers.), !i- + copula (elsewhere) !i-halla

!ala-bu

!ala-bu

?

Tgr. (Habab)

V. further variants in Voigt 1977:239. V. further variants in Voigt 1977:240.

Negation of the locative verb Negation of poss. constr. (3 masc. sg.) Negation of existential construction

Negative form of the copula

Negative non- !i- + nonverbal forms verbal form

Tgr. (Mensa)

y-alle-bū-ni

?

y-elle-ni (?)

i-tú-ni

?

Dahalik

Tna.

!al-bo

!al-bo

!i-hallawa

yäl-bo-n/ y-ällo-n

y-äbillu-n

y-ällo-n41/ yäl-bo-n42

!akko (in!ay-konä-n decl.)/!i-/!ikona

!i-/!akko + zäy- + non-verb. inf./adjective form

Gez.

y-ällä-m

y-älläw-3m

ay-dällä-m/ ay-dollä-m/ a-dällä-m/ a-doll y-ällä-m

alä-/yalä- + non-verbal form

Amh.

?

?

yelley-u (< *y-älläy-u) (y)ella-w (< *y-älla-w)

neg. forms of the reflexes of *kona: ak-kuna-m (< *!aC-kuna-m) yat-a-m

?

Arg. of àollaha

negative forms of the reflexes of *kona: a-huney-u (y)ella-w (< *y-älla-w)

Arg. of Aliyu Amba ?

Ancient Har. al-boräda ‘he did not reach’43 ay-säbri ‘he does not break’ al-µur ‘I do not go’47 = ay-sibär49 =

ay-nakt

ilaw-nakt

!al-nä£3l-u48

isäbr-um-ē−

!3ll3s-säbb3r (!3ll3C-säbb3r) -āy-bäl ‘he does not eat’

ila-nakt

y3sabr-um-ēl 46

Sel. al-nakata ‘he did not beat’

Zwy. Wol. !al-n3£äl-44 ‘he !al-säbäräm ‘he did not pull did not break’ out’ !ay-nä£3l-u !3l3-säbb3r

Modern Har. al-säbärä-m ‘he did not break’ ti-färk-am ‘he can not’ til-färk-am/ t3l-färk-am ay-fär3k

Gaf. al-fättär-am ‘he did not die’45

44

Wagner 1983:111, Text B, line 153; cf. Cerulli 1936:359. The negative perfect of this verb employs the base of the jussive (Meyer 2005:138, 192). There are, however, other types of verbal conjugation where the base of the affirmative perfect is the same as that of the negative perfect. 45 Leslau (1956:98) posits for the verbs with a final -ä a negative circumfix al-...-m (with a change of final -ä to -a) and a negative circumfix al-...-am for the rest of the verbs. It seems more economical to postulate a single morpheme al-...-am, with the final -ä elided before -a- (for this rule v. ibid. 22, § 11b). The origin of the element -a- in -am is unclear. 46 In Modern Harari, the negative forms of the imperfect in the main clauses go back to combinations of an affirmative imperfect form of the main verb with a negative form of the auxiliary verb (Wagner 1997:596). The negative forms of the “simple imperfect,” quoted in Leslau 1958:22, are, according to Wagner 1997:596–597, not used in Modern Harari and “either formed ad hoc by placing the negative prefixes before the simple imperfect or the informants gave the forms which were known to them from liturgical texts.” 47 Wagner 1983:127, Text B, line 222. 48 According to Meyer (2005:96), the 1 sg. imperfect prefix in Zway has two allomorphs: y- in word-initial position and l- after other prefixes (cf. Zwy. tī-l-nä£3l-u ‘while I take’). Consequently, he does not regard -l- as part of negative prefix !al-. This distribution, however, may be due to reanalysis of *!al- (NEG:1S) as !a-l- (NEG-1S), with l- perceived as an allomorph of 1 sg. prefix in noninitial position (cf. section 4.2 for the shift -y- > -l-). The same explanation can be valid for the element -l- in 1 sg. after the negative marker t- in Gafat. In both cases the present author’s analysis differs from that of Hudson (2003:216), for whom these two cases are clear examples of the original function of l(3)- as the marker of 1 sg. in the affirmative imperfect paradigm. 49 In the negative forms of the imperfect of the subordinate clauses in Modern Harari, the negative prefixes are attached to the jussive rather than to the imperfect stem (Wagner 1997).

43

Imperfect subordinate 3 masc. sg.

Imperfect 3 masc. sg. Imperfect 1 sg.

Perfect

ēl (< *y-äl) ēla-m/yela-m (< *y-äla-m), ēl-ba-m/yel-ba-m (< *y-äl-ba-m)53 ēl (< *y-äl)

ēl (< *y-äl)

ēlā (< *y-älä)

Wol. -āl-bäl

!īl(l)o (< *y-ällo)

!ēlä (< *y-älä)

!alä-/yālä- + !alä-/yālä- + infinitive non-verb. form !unku (indecl.) neg. forms of reflexes of *kona: !al-ōn !īl(l)o !ēlä (< *y-älä) (< *y-ällo) !īleyī !ēley (< *y-äläy) (< *y-äläyī)

Zwy. =

Sel.

ēla (< *y-äla)

ēlay (< *y-äläy)

ēla (< *y-äla)

inko (indecl.)/negative forms of reflexes of *kona: al-ōna

al- + infinitive50

al-nakt

allä-b-am

allä-b-3ğğw-am54

allä-b-am (?)52

t-ädäb3ll-am51

?

Gaf. al-fär3k

51

V. Gutt 1997:962. For the prefixes !al- and !ay- combined with deverbal adjectives and nouns in Zway and Selti v. fn. 19. From the morphological point of view, this is a verbal form of the imperfect followed by object suffixes and preceded by täinstead of the expected ti-/t3- (NEG:3MS, v. Leslau 1956:77). 52 Suffixes of the perfect are attached to the stem alläb- (< *!ala-b-). 53 Negation of the locative verb with a pronominal suffix, with or without the locative suffix -b-. 54 The same base as in the negation of the locative verb, with a pronominal suffix inserted between the stem and the perfect ending of 3 masc. sg. The origin of the element -3ğğ-, appearing in forms of the 3 person, is unclear.

50

Negation of existen- ēl (< *y-äl) tial construction

Negation of the locative verb Negation of possessive construction

Ancient Har. Modern Har. Imperf. subordinate = an-sibär 1 sg. Negative non-verbal al- + infinitive zal- + non-verbal form forms Negative form of the ? al-ta-m, m-al-ta copula

negative forms of the reflexes of *kona

? ?

neg. forms of the reflexes of *kona: an-xänä (< *!alkänä) yännä (< *y-ällä) yännän (< *y-ällän)

Negation of existential yellä (< *y-ällä) construction

Negation of the locayellä (< *y-ällä) tive verb Negation of poss. con- yellänn str. (3 masc.sg.) (< *y-ällänn) ?

?

?

?

yännä (< *y-ällä)

=

=

ann3-£ärs

ädäb3ll (ädäb3l with object pronoun)

=

=

ay-£ärs

ann3-säb3r

e-säb3r (< *!ay-säb3r)

Impf. subord. 3 masc. sg. Impf. subordinate 1 sg. Negative non-verbal forms Negative form of the copula

Imperfect 1 sg.

e-säb3r (< *!aysäb3r) ann3-säb3r

Muh. Msq. an-säbärä an-säbärä (< *!al-säbbärä) ‘he (< *!al-säbbärä) ‘he did not break’ did not break’ = =

t-i£ärs ‘he does not begin’ t-ä£ärs

=

Sod. al-säfärä ‘he did not measure’

Imperfect 3 masc. sg.

Perfect subordinate

Perfect

Enm. ã-säpärä-da (< *an-säpärä + da < !al-säbbärä + da) ‘he did not break’

ã-säβ3r (< *an-säβ3r < *!al-säb3r) ?

?

?

enä-da (< *y-ällä + da)

enän (< *y-ällän)

neg. forms of the re- negative forms of the reflexes of flexes of *kona: *kona: an-xärä (< *!al-känä) ã-xZr͂Z-da (< *an-xZr͂Z + da < *!alkänä + da) enä (< *y-ällä) enä-da (< *y-ällä + da)

?

=

ã-säpärä (< *an-säpärä < *!al-säbbärä) e-räχ3b (< *!ay-räk3b) ay-säβ3r-ka ‘he does not find’ aŋ-käft (< *!al-käft) ã-säβ3r-ka (< *an-säβ3r + ka < *!al‘I do not open’ säb3r + ka) = ay-säβ3r

Chah. an-täkäsä (< *!al-täkkäsä) ‘he did not kindle’ =

RIE 189:13: sobe / l3!ik3yani / !isam"ani ‘when, (even) after I have sent, he did not listen to me’ RIE 189:16–15: wa!3mz3 / gwayu / wa!i£omu ‘and after that they fled and did not stop’ DAE 12:14–15: ...!itarakuni... ‘they did not leave me’58

3 masc. sg. perfect 3 masc. pl. perfect

RIE 189:6: £3dmeya / !ay3£um57 / ¥ar ‘let the enemy not stand before me’

*!ayRIE 188:5–6: za!ay3tmawā! / la[¥a]r ‘which is not to be conquered by the enemy’ RIE 189:4: za!ay3tmawā! / la¥ar id. RIE 189:6: za!ay3tmawā / la¥ar id.

!3- is most likely to be explained as a variant of !i- before -y-. This type of variation (-3y-/-iy- and -3w-/-uw-) is well known from both Epigraphic Geez (v. E. Littmann in DAE, p. 80 and Dillmann 1907:80, 98–99, 348) and early Geez manuscripts (v. Bausi 2005:158). 56 The form is homonymous with 3 masc. sg. However, the verb y3be is likely to introduce the direct speech, as do its cognates elsewhere in ES. 57 Or, rather, !ay£um. 58 In RIE (193:15), most of this inscription is treated as illegible (this particular word being transcribed as .-t-.-kuni).

55

RIE 187:11: !imo£aµnā[hu] ‘we did not put (him) in chains’

!iRIE 187:4: za!3y3tmawā / la¥ar (< *za-!i-y3tmawwā(!) la¥ar) ‘which is not to be conquered by the enemy’55 RIE 232:9–11: zabala"a / ś3gāya / wasatiya / dam3ya / !iy3¢3"3mā / lamot ‘(he) who ate my flesh and drunk my blood will not taste the death’ RIE 189:8–9: sobe / tamakaµa / wa!iy3fal3s56 / !3mtakazi / y3be / !3µ[zā][ba] / nobā ‘when the Noba people boasted and said: “I will not withdraw from Takazi” ’ RIE 189:48: !3nza / !i!3"em3¥ / !3µzāba ‘while I do not oppress the peoples’ ? RIE 189:21–22: za!i[yā]!am3r / ¶w3l£o ‘whose number I do not know’ (the reading !i is not quite certain)

1 pl. perfect

3 masc. sg. jussive

1 sg. imperfect

3 sg. imperfect

Table 2. Negative markers !i- and !ay- in Epigraphic Geez

y3-: general negative marker in Tigre of Habab (also yä- in the Tigrinya variety of May Tchew); allomorph of !i- before verbal forms with initial guttural in contemporary Tigre of Mensa !ay-(...-n): general negative marker in Tigrinya (all verbal forms); allomorph of !i- before the prefix y3- in Epigraphic Geez; in SES, most probably not present at all !al-(...-m): general negative marker for verbs in SES; -lsurfaces in the perfect as well as in 1 sg. of the imperfect/jussive; in Harari and Selti also registered before nonverbal forms *!ala-: general negative marker with nominal forms in Amharic, Zway and Wolane (the same function probably traceable to proto-SES)

General negative markers !i-: general negative marker (perfect, imperfect, nominal forms) in Tigre and Geez; in 3 masc. sg. of the imperfect also in Argobba of àollaha

*!ala-: negative marker before the preposition *ba- in possessive and existential construction (such as Tigre !alabu) and traceable to PES. Cf. also Geez !akko, perhaps to be analyzed as *!al- (< *!ala- ?) + -ko (< *kona ‘to be’) *!3n- in a few fossilized negative expressions (such as Geez !3n-dā"i and !3n-b-) and possibly discernible in the East Gurage negative particle *!3nko

Lexically restricted negative markers *!i-: throughout ES before the verb *hallawa (*!i- > y- in Tigrinya and SES)

Table 3. Negative markers in Ethio-Semitic languages: a summary

418

Articles: Semitic Studies

Language abbreviations Amh. — Amharic; Arg. — Argobba; Chah. — Chaha; EG — Epigraphic Geez; Enm. — Ennemor; ES — Ethio-Semitic; Gaf. — Gafat; Gez. — Geez; Har. — Harari; Msq. — Mesqan; Muh. — Muher; NES — North Ethio-Semitic; PES — Proto-Ethio-Semitic; Sel. — Selti; SES — South Ethio-Semitic; Sod. — Soddo; Tgr. — Tigre; Tna. — Tigrinya; Wol. — Wolane; Zwy. — Zway.

References Bausi 2005 Brockelmann 1908 Bulakh–Kogan 2010

Cerulli 1936 Chamora–Hetzron 2000 Cohen 1931 Cohen 1939 Dillmann 1907 Elias 2005 Faber 1991

Faber 1997 Friedrich–Röllig 1999 Garad–Wagner 1998 Getachew Haile 1970

Goldenberg 1968 Gutt 1983 Gutt 1997

Bausi, A. Ancient Features of Ancient Ethiopic. Aethiopica 8:149–169. Brockelmann, C. Grundriss der vergleichenden Grammatik der semitischen Sprachen. I. Laut- und Formenlehre. Berlin. Bulakh, M.; Kogan, L. The Genealogical Position of Tigre and the Problem of North Ethio-Semitic Unity. ZDMG 160:273–302. Cerulli, E. Studi Etiopici. I. La lingua e la storia di Harar. Roma. Chamora, B.; Hetzron, R. Inor. München. Cohen, M. Études d’éthiopien méridional. Paris. Cohen, M. Nouvelles études d’éthiopien méridional. Paris. Dillmann, A. Ethiopic Grammar. Ed. by C. Bezold, tr. with additions by J. A. Crichton. London (repr. 1974). Elias, D. Tigre of Habab. Short Grammar and Texts from the Rigbat People. Harvard University. PhD. thesis. Faber, A. The Diachronic Relationship between Negative and Interrogative Markers in Semitic. Kaye, A. S. (ed.). Semitic Studies in Honor of Wolf Leslau on the Occasion of His Eighty-fifth Birthday. Wiesbaden. Pp. 411–429. Faber, A. Genetic Subgrouping of the Semitic Languages. Hetzron, R. (ed.). The Semitic Languages. London. Pp. 3–15. Friedrich, J.; Röllig, W. Phönizisch-Punische Grammatik. 3. Auflage. Roma. Garad, A.; Wagner, E. Harari-Studien. Wiesbaden. Getachew Haile. Archaic Amharic Forms. Proceedings of the Third International Conference of Ethiopian Studies. Addis Ababa. Vol. 2. Pp. 61–80. Goldenberg, G. K3stan3ñña. Studies in a Northern Gurage Language of Christians. OrSu 17:61–102. Gutt, E.-A. Studies in the Phonology of Silti. Journal of Ethiopian Studies 16:37–73. Gutt, E.-A. Concise Grammar of Silt’e. Gutt, E. H. M.; Hussein Mohammed. Silt’e-Amharic-English Dictionary (with Concise Grammar by Ernst-August Gutt). Addis Ababa. Pp. 895– 957.

M. Bulakh, Negative markers *!ay-, *!i- and *!al- in Ethio-Semitic Heine 1997 Hetzron 1970 Hetzron 1972 Hetzron 1977 Hudson 2003

Huehnergard 1983 Leslau 1941 Leslau 1956 Leslau 1958 Leslau 1968 Leslau 1969a

Leslau 1969b

Leslau 1981 Leslau 1995 Leslau 1997 Leslau 2004 Meyer 2005 Meyer 2006 Müller 1905 Podolsky 1991 Praetorius 1871 Praetorius 1879 Raz 1983

419

Heine, B. Possession: Cognitive Sources, Forces, and Grammaticalization. Cambridge. Hetzron, R. Review of Leslau 1968. JAOS 90:561–581. Hetzron, R. Ethiopian Semitic. Studies in Classification. Manchester. Hetzron, R. The Gunnän-Gurage Languages. Napoli. Hudson, G. Ethiopian Semitic Negative Nonpast. Bender, M. L.; Takács, G.; Appleyard, D. L. (eds.). Selected Comparative-Historical Afrasian Linguistic Studies, in Memory of Igor M. Diakonoff. München. Pp. 209–218. Huehnergard, J. Asseverative *la and Hypothetical *lu/ law in Semitic. JAOS 103:569–593. Leslau, W. Documents tigrigna (éthiopien septentrional). Paris. Leslau, W. Étude descriptive et comparative du Gafat (éthiopien méridional). Paris. Leslau, W. The Verb in Harari (South Ethiopic). Berkeley– Los Angeles. Leslau, W. Ethiopians Speak. Studies in Cultural Background. III. Soddo. Berkeley–Los Angeles. Leslau, W. Is There a Proto-Gurage? Proceedings of the International Conference on Semitic Studies Held in Jerusalem, 19–23 July 1965. Jerusalem. Pp. 152–171 (repr. in: Id. Gurage Studies, Collected Articles. Wiesbaden 1992. Pp. 226–245). Leslau, W. The Negative Particle !in in Arabic and (!)3n in Ethiopic. Annali dell’Istituto Universitario Orientale di Napoli 19:137–145. Leslau, W. Ethiopians Speak. Studies in Cultural Background. Part IV. Muher. Wiesbaden. Leslau, W. Reference Grammar of Amharic. Wiesbaden. Leslau, W. Ethiopic Documents: Argobba. Grammar and Dictionary. Wiesbaden. Leslau, W. The Verb in Mäsqan as Compared with Other Gurage Dialects. Wiesbaden. Meyer, R. Das Zay. Deskriptive Grammatik einer Ostguragesprache (Äthiosemitisch). Köln. Meyer, R. Wolane. Descriptive Grammar of an East Gurage Language (Ethiosemitic). Köln. Müller, D. H. Die Mehri- und Soqo¢ri-Sprache. II. Soqo¢riTexte. Wien. Podolsky, B. Historical Phonetics of Amharic. Tel-Aviv. Praetorius, F. Grammatik der Tigriñasprache in Abessinien. Halle. Praetorius, F. Die Amharische Sprache. Halle (repr. Hildesheim–New York, 1970). Raz, Sh. Tigre Grammar and Texts. Malibu.

420 Simeone-Senelle 2008

Voigt 1977 Voigt 2006

Wagner 1964 Wagner 1983 Wagner 1997

Wetter 2006

Articles: Semitic Studies Simeone-Senelle, M.-C. The Specificity of the Dahalik Language within the Afro-Semitic Languages. Paper presented at the International Workshop “History and Language of the Tigre-speaking Peoples (Eritrea and Sudan). Naples, 7–8 February 2008 [published as Simeone-Senelle, M.-C. The Specificity of the Dahalik Language within the Afro-Semitic Languages. Lusini, G. (ed.). History and Languages of the Tigre Speaking Peoples. Napoli, 2010. Pp. 127–145]. Voigt, R. M. Das tigrinische Verbalsystem. Berlin. Voigt, R. M. Südtigrinische Dialekte: das einfache und zusammengesetzte Präsens im Dialekt von May-Č̣äw. Uhlig, S. (ed.). Proceedings of the XVth International Conference of Ethiopian Studies, Hamburg July 20–25, 2003. Wiesbaden. Pp. 893–898. Wagner, E. Der Übergang von Fragewörtern zu Negationen in der semitischen Sprache. MIO 10:261–274. Wagner, E. Harari-Texte in arabischer Schrift mit Übersetzung und Kommentar. Wiesbaden. Wagner, E. The Negative Imperfect in Ancient and Modern Harari. Fukui Katsuyoshi; Eisei Kurimoto; Masayoshi Shigeta (eds.). Ethiopia in Broader Perspective. Papers of the XIIIth International Conference of Ethiopian Studies. Vol. I. Kyoto. Pp. 596–600. Wetter, A. The Argobba of T’ollaha—A Comparative Overview. Uhlig, S. (ed.). Proceedings of the XVth International Conference of Ethiopian Studies, Hamburg July 20–25, 2003. Wiesbaden. Pp. 899–907.

A New Attempt at Reconstructing Proto-Aramaic. I Sergey Loesov Russian State University for the Humanities, Moscow

0. Introduction: Non-Trivial Common Drifts This study is an attempt of internal reconstruction of Proto-Aramaic.1 It is meant as a prolegomenon to a future historical grammar of Aramaic and is based on two assumptions to be made plausible in what follows: Proto-Eastern Aramaic (= PEA) left the Common Aramaic fold around 1000 B. C., producing the first known dialectal split in the history of Aramaic.2

1

I thank Maria Bulakh, Claudia Ciancaglini, Eran Cohen, and Stephen Kaufman who have read an earlier version of this paper and pointed out to me numerous errors. I am also grateful to Otto Jastrow for discussing with me numerous questions of Aramaic philology. I have to define the epochs (or periods) of Aramaic presupposed in this paper. From the point of view of this study, “Old Aramaic” ends with the emergence of those spoken Aramaic varieties that, with time, would become the vernacular foundations of the six “Middle Aramaic” literary idioms. Spoken “Middle Aramaic” ends with the appearance of a distinctly “neo-Aramaic” speech type. Linguistic features that justify this division will be discussed in the continuation of this study. No reliable absolute dating is now possible. Impressionistically, “neoAramaic” may turn out to have been the longest period in the documented history of Aramaic, while the classical “Middle Aramaic,” in reality, may have been (at least in the Eastern branch) a comparatively short transition period between the “Old” and the “New” epochs of the language. Abbreviations of languages and corpora: BA = Biblical Aramaic; CPA = Christian Palestinian Aramaic; EA = Eastern Aramaic; ENA = Eastern NeoAramaic; IA = Imperial Aramaic; JBA = Jewish Babylonian Aramaic; MEA = Middle Eastern Aramaic; MWA = Middle Western Aramaic; NENA = NorthEastern Neo-Aramaic; PA = Proto-Aramaic; PEA = Proto-Eastern Aramaic; SA = Samaritan Aramaic; WA = Western Aramaic; WNA = Western Neo-Aramaic. 2 This is to say that we have to part ways with the communis opinio of Englishspeaking scholarship reflected e. g. in the words of F. Rosenthal (1978:85), “all the later Aramaic dialects can be shown to go back basically to Official Aramaic,” or of J. Huehnergard (1995:274), “official Aramaic essentially wipes the preceding Aramaic slate clean.”

422

Articles: Semitic Studies

Aramaic languages in their entirety (= both the Western and the Eastern branch) have been experiencing non-trivial “common drift” not immediately anchored in Proto-Aramaic, i. e., this drift has not been building materially on developments that may have been underway already in Proto-Aramaic.3 The drift speed of Eastern Aramaic (= EA) has been higher than that of Western Aramaic (= WA), while the drift direction has always been the same for both branches. A future historical grammar of Aramaic has to explain the directions of changes (in so far as they are not typologically trivial) and the different development speeds in the two branches. By way of introduction, let us now consider a few examples of what I believe to be “non-trivial common drifts” in the history of Aramaic, and perhaps their relevance to the reconstruction of Proto-Aramaic. 0.1. Resultative does its best to be morphologically sensitive to transitivity Most of the modern ENA languages (Turoyo/Mlahsô and NENA, but not Neo-Mandaic), at a certain stage of their lives, had developed two morphological shapes of RESULTATIVE/PERFECT that later shifted to PRETERIT: (1) {pass. part. qtīl + l + pers. possess. pron. as agent index} for transitive verbs: nšiqāli ‘I [have] kissed her’; nšiqīli ‘I [have] kissed them’ (2) {qattīl part. + pers. subj. pron.} for intransitive verbs: dammīxnā ‘Imasc. [have] slept.’4

3

I use the term “common drift” roughly in the Sapirian sense, but perhaps not quite so, because Sapir has no stringent definition of this suggestive concept. Sapir says, in particular, “The momentum of the more fundamental, the predialectic, drift is often such that languages long disconnected will pass through the same or strikingly similar phases. In many such cases it is perfectly clear that there could have been no dialectic interinfluencing” (Sapir 1921:172). For the purposes of my study, a common drift is neither a purely “typological” phenomenon (like disappearance of case endings in languages belonging to different linguistic families) nor exclusively shared innovations of the Stammbau genealogical classification. It is not simply an independent parallel development (i. e., shared by certain languages of a given genealogical subgroup) either, because this latter can be typologically trivial as well. My informal idea is that the pan-Aramaic drift to be discussed has to do with certain features of Proto-Aramaic. I hope this will become clearer as we proceed. 4 See Hopkins 1989 and the criticism of Goldenberg 1998:662, fn. 63. Cf. also Khan 2008a:106: “Vestiges of the original system in which the ergative inflection is restricted to transitive verbs has survived in a few Jewish dialects on the eastern

S. Loesov, Reconstructing Proto-Aramaic …

423

Given this evidence, one may want to know if there existed a feature in the immediate parent language of Turoyo/Mlahsô and NENA (i. e., in PEA, one of the two firstborns of PA) that is responsible for this development. And how did it come about that Mandaic escaped the common lot and preserved the PEA Preterit *qatal? It may well be that such a feature did exist in PEA in an incipient way, given that periphrases of type (1) are well attested in all three middle EA literary corpora, including Mandaic (Nöldeke 2005:381ff. with references to JBA; Nöldeke 1966:210; Bar-Asher Siegal 2011). By contrast, we do not know much about the qtīl l- periphrasis in the Middle Western Aramaic (= MWA) literary corpora. According to Kutscher 1969:137, it did not exist at all. In the first millennium B. C. NorthWest Semitic, qatūl l-/qatīl l- is attested as an “ergative” (= not-alreadyquite-passive) periphrasis that could sometimes be formed at will.5 Be this as it may, Western Neo-Aramaic (= WNA) has not developed the lpreterit, yet, most amazingly, it does have RESULTATIVE/PERFECT finite forms that are morphologically sensitive to the value of transitivity (Arnold 1990:76): {*pers. subj. pron. + *qtīl} for transitive verbs: ni¢µen ‘Imasc. have ground/milled.’ {*pers. subj. pron. + *qattīl} for intransitive verbs: nnaµµeč ‘Imasc. have descended/gone down.’

periphery of the NENA area (e. g. J. Sulemaniyya, J. Kerend and J. Sanandaj),” with references to earlier literature. Thus *qattīl may have been the productive resultative adjective formed for most intransitive roots in the whole of Middle Aramaic at a certain stage of it history, with the result that dammīx-nā appeared as a past-time form in the parent languages of Turoyo and of all of NENA, as well as in WNA (in the latter, with the preposed personal index, n-dammīx). 5 A BH example of a freely formed expression of this kind is !iššā "ă´ūrā lānū ‘the woman is closed for us’, i. e., ‛we have not been having sex’ (1 S 21:6); a formulaic example from BH is b3rū−ā !at l-yhwh bittī ‘the LORD bless you, my daughter’ (Ruth 3:10, and often in BH). Cf. similar optative/injunctive formulas in IA: brk PN l-!sry ‛May Osiris bless PN’ (TAD D22.13); ydy" yhwy lk ‛let it be known to you’ (TAD A6.10:3), and see Muraoka–Porten 2003:202f. for more examples (all the grammatical interpretations in Muraoka–Porten 2003:202f. are different from mine). Odeberg 1939 II 103 offers an example from Galilean Aramaic (TY, Šqalim III2 11b), "wlpny škyµ ly ‘I have forgotten what I have learned,’ but the context of the story shows that his translation is wrong (and cf. DJPA 550a ‘my learning is within me’, škyµ being an adj. ‘frequent, at hand’).

424

Articles: Semitic Studies

The source of agreement in both cases is the grammatical subject only (there is no object agreement), the personal prefixes are etymologically “in the nominative,” the verb forms are inflected for gender and number via suffixes, but the bases are different depending on the value of transitivity. Now, the split ergativity picture in Turoyo (identical to an earlier phase of all or most of NENA) could have emerged (and most probably did emerge) so to speak “of itself.” When the EA languages (with the exception of Mandaic)6 started working on an active resultative finite form, the natural point of departure for intransitive verbs was qattīl, because in MEA qattīl of dynamic intransitives tended to be resultative (nappīq ‘he has gone out’ as against nāfeq ‘he is going out/he will go out’). Thus, if one wants to understand the history of the Aramaic verb, one has to figure out if (and to what degree) qattīl was productive and semantically predictable in the spoken varieties of Middle Aramaic. For the moment, there are no relevant data available for any of the six literary languages. What Nöldeke said in 1880 about qattīl in his chapter on the nominal word-formation is lapidary but stimulating: “sind sehr viele Adjectiva gebildet, namentlich solche, welche als Verbaladjectiva, resp. Part. perf. neben intransitiven Verben stehn (zum Theil wohl ziemlich junge Bildungen)” (1966:71, boldface added).

Despite the “Part. perf.” label that invites one to include qattīl into the “system” of the Syriac verb, Nöldeke does not legalize qattīl in his Syntax: he limits his account of the regular predicative functions of verbal adjectives to the uses of qātel and qtīl (§§ 269–280). Given the deplorable state of Classical Syriac philology (both grammar and lexicography, especially the non-availability of morphologically tagged corpora), my tentative exegesis of the above Nöldeke’s words in the context of his Kurzgefasste Syrische Grammatik is as follows: qattīl of intransitives is appreciably less frequent than their qātel, while qattīl is more prominent in the late part of the corpus (“ziemlich junge Bildungen”) than before the Arab conquest. In the meantime, let me adduce a few illustrative Syriac examples that square well with our evidence on the Neo-Aramaic reflexes of qattīl. The

6

All of ENA save for Mandaic are/were located in what is now Kurdistan. This isogloss would be the only attractive argument pro the Iranian influence on the emergence of the l-preterit. Yet, for various reasons, this idea is hardly viable (Pennacchietti 1988). It emerges time and again in contemporary scholarship (see e. g. Khan 2008a:105f.), yet it is usually formulated in general terms as something self-evident (or already proven) and with no detailed argumentation.

S. Loesov, Reconstructing Proto-Aramaic …

425

examples are from the NT Pešitta, an admittedly early text7 with reliable vocalization. An important additional advantage is that we have at our disposal the Greek original, while the NT Pešitta is known to be a notslavish translation that strives to be reasonably faithful to the sense rather than to the letter of the Greek text.8 Whether this selected evidence reflects a trend current in the spoken Edessan Aramaic around 300 A. D., we do not yet know. Consider the examples: !bd ‘to perish’ qātel: pa´´ān !ā®dīn-µnan ‘Rescue us! We are perishing’ (Mt 8:25, the Present Actual). qattīl: !e¬ā gēr bręh d-!nāšā d-naµµē meddem da-!bbī¯-hwā ‘For the Son of Man came to restore to life those who had perished’ (Mt 18:11, the Resultative). npq ‘to go out’ qātel: ka¯ nā«qitton men baytā ‘When you are going forth out of the house’ (Mt 10:14). qattīl: w-!ezzla¬ l-®aytāh w-!eškµa¬ bar¬āh ka¯ ramyā b-"arsā wnappīq mennāh šē ! ¯āh ‘And she went back home, and found her daughter lying in bed, and her demon had gone out from her’ (Mk 7:30).

As for dynamic transitives, the only deverbal adjective they could rely upon in the whole of EA for creating RESULTATIVE was the passive qtīl (as is well-known, its aspectual meaning was resultative).9 So, if these verbs needed to get an active RESULTATIVE, the language had to provide, within the new construction, a secure slot for the pronominal agent (in the 7

I. e., it was created in the epoch when the literary Syriac was still quite close to its vernacular basis, the Aramaic dialect spoken in Edessa in the early centuries A. D. 8 I thank my doctoral student Evgeny Barsky for allowing me to use these and other examples of Syriac deverbal adjectives from his dissertation in progress. 9 As follows from the above, the distribution of productive deverbal adjectives (= “participles”) qātil, qattīl and qtīl depending on the root meaning has not yet been seriously studied even for Syriac, to say nothing about other middle Aramaic corpora. Obviously, this distribution has to be understood in detail first if we are to really follow the development of the Aramaic verb from the Middle to the New period. Goldenberg 1998:611 says that “no active q¢īl forms can ever be derived in Syriac from intransitive verbs.” What is for Goldenberg “active q¢īl” as far as intransitive verbs go, is not quite clear from his article (his suggestion to define certain adjectives in terms of “reference directly to the act itself ” rather than to its participants is not really convincing). Note that intransitive (only unaccusative?) verbs do sometimes form qtīl adjectives: d-men tešbuµtā lā zī"īn ka¯ m°ad«īn ‘those who are not afraid of the Glory while they are blaspheming’ (2 Pet 2:10; zw" is ‘to move,’ ‘to tremble,’ ‘to be afraid’; SL 364a).

426

Articles: Semitic Studies

oblique case), and then the language was supposed to reinterpret as soon as possible the “qtīl l- + pronominal agent” construction as a pragmatically non-passive predication, with the consequence that the split-ergativity picture would emerge out of a formally passive construction in a most natural way, “of itself,” only to switch to the accusative syntax as soon as possible.10 So far, so good. Yet what had come to pass between MWA and WNA is more difficult to interpret historically. The standard grammar of CPA, Müller-Kessler 1991, does not mention qattīl as a productive deverbal pattern at all. There is an obvious orthographic difficulty, since CPA has no means of rendering the gemination -tt- in qattīl, in particular no diacritic signs useful for this end. Yet the real hindrance is that Müller-Kessler 1991 has no syntax. We know for sure that the resultative qattīl was productive in the immediate Middle Aramaic ancestor of WNA, as well as in that of Turoyo, so it would have made sense to look for *qattīl in CPA with the help of syntactic criteria, yet this did not happen. Another difficulty is MüllerKessler’s phonological solution (probably inspired by her mentor R. Macuch): she has, as a result of stress retraction to the penult syllable, PA *qatl, *qatal, *qatil, *qatīl and *qātil shifted to q#tel, while her *qattīl shifted to q#ttil, with the outcome that the six morphological patterns are supposed to coincide orthographically in the extant corpus of CPA. In practice, MüllerKessler allots the relevant orthographic tokens to this or that among the six etymological shapes randomly, i. e. she lets herself to be guided by intuitive semantic ideas; nowhere in the Grammar does she formulate them. Müller-Kessler’s CPA does not participate in the trivial Aramaic drift; e. g., it has no short vowel syncope (see Beyer 1995 for a detailed discussion). It does not reveal certain diagnostic features typologically anticipating WNA and leading to it. This means that we are probably in need of a new grammatical treatment of CPA in a historical perspective, a study that would supplement the achievement of Müller-Kessler.11 In the meantime, we have to state that WA, whose verb in its main features is today similar to the EA verb of the mid-first millennium A. D.,12 10

A part of this diachronic path of historical Aramaic follows the commonsense prediction of Estival–Myhill 1988:441, “ergative constructions develop diachronically from passive constructions. This development … consists in the pragmatic, syntactic, and morphological verbalization of a deverbal form.” 11 It is understood that the spoken basis of CPA is probably not a direct ancestor of the WNA we know. Yet it is hard to believe that *qattīl, so important for the WNA verb, had no say whatsoever in the grammar of CPA. 12 In particular, in WNA the prefixing conjugation (i. e., the reflex of the PA *yaqtulu) is JUSSIVE-SUBJUNCTIVE only. Same is partly true of MEA, where the pre-

S. Loesov, Reconstructing Proto-Aramaic …

427

at a certain point also felt that it was impossible to make do with *qatal as the shared exponent of SIMPLE PAST, PERFECT, RESULTATIVE, etc., and set on creating a special deadjectival resultative form. This move is in itself trivial, it is a standard way to renew the past-time domain. The nontrivial common drift that requires an explanation is the fact that the new WA resultative is sensitive to the transitivity semantics of the root in the same way as the EA one is, though Middle WA missed the opportunity to develop the qtīl lī construction. Correll 1978:75ff. tries to reduce the situation in Proto-Western NeoAramaic to the MEA one: q¢īl lī = ‘mir ist getötet’ > ‘ich habe getötet’ = anā q¢īl (p. 77). This putative WA replacement of q¢īl lī with anā q¢īl, i. e. a paradigmatic levelling in favour of intransitive inflectional affixes (the nnaµµeč type) is improbable, and this is for two reasons: 1) The transitive verb is a more prototypical member of the part of speech “verb” than the intransitive verb, so q¢īl lī > anā q¢īl replacement is not likely; 2) In accord with thesis (1), all individual NENA languages had experienced the opposite direction of replacement (or levelling): qayyīmnā > qīmlī ‘I have stood up,’ with both the stem and inflection of the intransitive resultative adjusting to the transitive pattern (see Khan 2007 and Khan 2008a with references to earlier studies).13 D. Cohen 1979:238f. and 1984:491–499 provides a more sophisticated hypothesis. He suggests that the starting point of the development were the famous low-transitivity two-place verbs (in Middle WA?) whose qtīl could be used with the active value. His example is X drīk l-Y ‘X a suivi Y’ (Cohen 1979:239). In the course of time, all lexical constraints on the active diathetic reading were lifted, and as a result the *R1R2īR3 base of transitives is now used to form the resultative/perfect for both diathetic values, e. g. ne"zem (√ "zm) ‘I have invited’ and ‘I have been invited’ (Cohen 1984:493), the distinction being the matter of syntax and semantics only. Consider an example of a high-transitivity root with the active diathetic reading of *qtīl: qsīma ba¢¢ixča ‘sie hatte eine Melone zerteilt’ (Arnold 1999:4). Typologically, a partial parallel to Cohen’s scenario is provided by the functions of the Akkadian “Stative” paris (a resultative verb form with a fixing conjugation is used both as FUTURE and JUSSIVE-SUBJUNCTIVE, while the active predicative participles with future-time readings are also very common. 13 Although the qayyīmnā “absolutive” type is only marginally attested in today’s NENA, there is hardly any doubt about the validity of this scenario.

428

Articles: Semitic Studies

transparent deadjectival etymology) in the early second millennium B. C.: the Stative of transitive verbs used to have either active or passive readings depending on the root meaning. There are two major differences between the resultative verb forms in Akkadian and WNA, however: 1) the Resultative in WNA has two base shapes in the G-stem (*qattīl and qtīl), depending on the value of transitivity, while Akkadian has only one; 2) according to D. Cohen, in WNA, the inflected base *R1R2īR3 is fully productive and can freely get both active and passive resultative readings,14 while for Akkadian transitive verbs the formation and the diathetic readings of paris (active or passive) are subject to severe lexical restrictions (Loesov 2011). To make this solution really cogent, D. Cohen would have to explain how it had come to pass that in this case Marcel Cohen’s déponents internes generalized for all transitive verbs and carried the day. Yet there seems to be no explanation, no more so than in the case of Correll’s hypothesis. Right now, I will not try to explain this remarkable isogloss between ENA and WNA. To explain it means to answer the question, “Why has it been important for Aramaic to create and maintain some kind of morphological distinction between transitive and intransitive verbs in the past-time domain?”15 For the moment, I am only trying to show the relevance of the “non-trivial common drift” notion for our understanding of the Aramaic linguistic history, and perhaps for the description of PA. E. g., the difference between the Spanish he ido and the Italian sono andato (m.)/sono anadata (f.) shows that Spanish gave up the original transitivity-related variation still kept by the more conservative Italian. Cantar de mío Cid reveals the earlier Spanish picture, identical to that of today’s written Italian: “Los mandados son ydos a todas partes; Hydos son los caualleros,” etc. This tendency of Romance (and of NENA, as mentioned 14

For the active reading of qtīl in the dialect of Ma"lula, see especially Bergsträsser 1983:102. The reference works on WNA (including Correll 1978!) do not describe the expression of diathesis as a subject in it own right (no chapter on “the Voice” or something of the kind), yet the nature of collected examples and certain formulations allow one to believe that in WNA (or at least in Ma"lula, the only WNA dialect described by Bergsträsser and Spitaler) qtīl of transitives is productive for both diathetic readings, with no further lexical-semantics restrictions. Spitaler 1938:210 notes that “q¢īl ist volkommen lebendig in seiner passivischen Bedeutung”; according to Arnold 1990:76, “Das alte Partizip passiv erfüllt auch im NWA noch seine Funktion”; Arnold 1989:27: “Das Perfekt kann passive Bedeutung ausdrücken, da es historisch aus dem Partizip passiv hervorgegangen ist. In aktiver Bedeutung dient es zum Ausdruck von Vorzeitigkeit …” 15 Note that most of NENA, as stated above, happened to lose this distinction.

S. Loesov, Reconstructing Proto-Aramaic …

429

above) to eliminate meaning-related morphosyntactic distinctions in favour of the transitive type and to switch from the split-ergative to the accusative alignment16 needs no special explanations, because it is trivial. In our case, on the contrary, both branches of Aramaic managed to create their transitive vs. intransitive distinctions in historical times, after the disintegration of PA. As we have seen, WA and EA have been striving for the same end, but EA started doing so earlier in its life, and it is probably because of this time gap that the two branches availed themselves of somewhat different means. In view of these data, it is reasonable to ask if PA possessed at least the morphological hardware necessary for this development, i. e. the resultative adjective qattīl formed from intransitive verbs, which is a pattern more or less unknown in Akkadian and not really productive in BH. Thus, the data analysed in 0.1 may be relevant for the reconstruction of Proto-Aramaic. By way of a telling individual example of how verbs’ lexical meanings can influence grammar in Aramaic, note the semantically motivated instance of common drift of the (former) resultative form of šm" ‘hear’ in WNA and Turoyo: šamme" (Arnold 1990:76), šam3" (Jastrow 2002:180).17 Given that šmy" ly is our parade example of the new Aramaic ergative resultative in Imperial Aramaic (Muraoka–Porten 2003:202) and in Syriac (Nöldeke 1966:210), the non-qtīl forms in WNA and Turoyo are doubtless innovations: for this low-transitivity root (as well as for several similar ones), both languages switched secondarily to the intransitive pattern *qattīl. Note that the general outline of our story of two Resultatives resembles the development of New (denominative) Presents in Aramaic. The EA (and probably the Proto-Eastern Aramaic) pattern of the New Present is “*qātil + subj. pronoun,” cf. the evidence of the three literary MEA idioms and the whole of ENA. This pattern is identical to the order of elements in the Proto-Semitic nominal sentence with personal pronouns as subjects (cf. in particular Andersen 1970 and Huehnergard 1986: in OB, šarrum anāku is the natural way of saying ‘I am a/the king’). This Prednoun—Subjpers. pron. order had been probably still preserved in PA, and was petrified within the MEA New Present.18 As it happens most of the time, WA started working 16

Cf. the Italian li ho visti, le ho viste, etc., as against the Spanish les/las ho visto. The base of the two forms goes back to šammī", this happens fully in accord with the historical phonology of both idioms. 18 The Prednoun—Subjpers. pron. order is still predominant in Syriac. Otherwise, in the historical Syriac, the organization of nominal clauses is manifold and per17

430

Articles: Semitic Studies

seriously on its New Present later than EA,19 and by then the default word order in the WA noun clause was S—P with whatever subject. This is the linear order of constituents within the WA New Present as well. Consider an example from CPA: h! !n! mšlµ ml!ky qwdm !pyk (Mk 1:2, CCPA IIA, p. 76)

as against the Pešitta: hā mšaddar !nā mal ! a−y q¯ām par´ōpā−

Both versions are translations of the following text: „doÝ ¢postš postšllw tÕn ¥ggelÒn mou prÕ prosèpou sou ‘I am sending my messenger ahead of you’ (NRSV).

As a result, now all of Aramaic has *qātil as the imperfective verb base, with personal affixes at its right in the Eastern subgroup and at its left in the Western subgroup. The tertium comparationis of the two stories is as follows: both the Resultative and the New Present appeared in each of the two subgroups independently, i. e. after the disintegration of PA. They emerged in the course of the common Aramaic drift, i. e. according to the same general plan but with different implementations, and not synchronically. 0.2. Aramaic infinitives and agent nouns A second instance of the non-trivial common drift of EA and WA (once more with probable implications for our reconstruction of PA) is the shape of the productive action noun for the G-stem. As I will claim in the course of this study, the Proto-Aramaic G-infinitive is *maqtal,20 it replaced the PS infinitive *qatāl-. Maqtal was inherited by both daughterlanguages of Proto-Aramaic, i. e. Proto-WA and Proto-EA. Yet *qatāl,

haps poorly understood (see van Peursen 2008:244ff. for the research history). It foreshadows the situation in ENA and Goldenberg’s syntactical revolution. 19 This follows from the fact that WNA has preserved the Prefixing Conjugation as Jussive/subjunctive, while the whole of ENA has no trace of it. Whether Palestian Arabic adstratum did help conserve the Prefixing Conjugation of WNA, I cannot say now. 20 Most probably, it is not *meqtal or *miqtal, as is often assumed, because in Egyptian Aramaic the G-inf. of ybl is written mwbl, that of ytb is mwtb, etc. (Muraoka–Porten 2003:122).

S. Loesov, Reconstructing Proto-Aramaic …

431

having lost its verbal syntax in favour of maqtal,21 had been never completely discontinued. This follows from the middle Aramaic evidence and can be plausibly corroborated by the data of the first millennium B. C. Aramaic as well. Consider some of the possible examples from Egyptian Aramaic: !mr ‘saying,’ !sr ‘oath,’ bkw ‘crying,’ bky ‘weeping,’ ¢ll ‘protection,’ pg" ‘meeting,’ q¢l ‘execution,’ qrb ‘battle’ (see Muraoka-Porten 2003:77ff.). Biblical Aramaic definitely has several semantically fitting examples, e. g. zmār ‘Spiel,’ tqā« ‘Stärke’ (see Bauer–Leander 1927:187). From the first millennium B. C. onwards, the syntax of this *qatāl nomen actionis has been invariably nominal, i. e. the semantic object of qatāl derived from transitive verbs has been appearing as the dependent in respective noun phrases. The extant corpus of the first millennium B. C. Aramaic is too meagre to justify big claims of any kind, yet it is not to be excluded that in that epoch qatāl had been living through a period of temporary decline and search of new identity. One thing is certain: Middle Aramaic had the productive (*qatāl >) qtāl shape for “zahllose Nomina actionis zu den Verben des einfachen Stamms” (Nöldeke 1966:68) at the side of the meqtal infinitive that used to be formed quite regularly as well. In Syriac (and, as it seems, in all of the Middle Aramaic) the major difference between the synonymous productive patterns qtāl and meqtal was the one already mentioned: l-meq¢al l-°a®rā hānnā ‘to kill this man’ vs. q¢ālā d-°a®rā hānnā ‘killing of this man.’ Another difference is the consequence of this “more nominal” nature of qtāl: unlike meqtal, qtāl-nouns tended to develop all kinds of derived meanings. An example is qyām-: ‘standing’; ‘stability’; ‘garrison’; ‘covenant’; ‘vow of celibacy’; ‘monkish orders’ (CSD 504a). With the advent of the Neo-Aramaic epoch, meqtal disappeared almost completely from the whole of Aramaic, Western and Eastern alike,22 which is a manifestation of the common drift. As for qtāl-, materially (in particular, syntactically) it remains where it used to be since the very beginnings of Aramaic (admittedly, in the second millennium B. C.): it keeps being a productive action noun with nominal rection. Yet functionally, it is now the only productive verbal action noun (or “infinitive”) in the whole of 21

I. e., the innovative PA maqtal of transitive verbs governed the syntactic accusative of the patient (l-maqtal !iyyāt PN), exactly as the Akkadian qatāl- used to behave when the direct object appeared to its left (šipram epēšum ‘to do work’). 22 The few m-“infinitives” of Turoyo adduced in Jastrow 2002:135 are not phonologically regular reflexes of the respective Western Syriac infinitives: mazlo ‘Gehen’ vs. WS mīzal, ma¬yo ‘Kommen’ vs. WS mī¬ō, mamro ‘Sagen’ vs. WS mīmar, mahwo ‘Geben’ vs. WS mettal, mŭklo ‘Essen’ vs. WS mē−al. I have not found more examples of m-forms in Turoyo.

432

Articles: Semitic Studies

contemporary Aramaic. Consider the examples from Turoyo: u=dworo daµ=µe¢e ‘das Unterpflügen des Weizens’ (Jastrow 2002:84; 10.11.1); bû-nµōto d-nāµ3t ‘als er heruntergestiegen war’ (Ritter 1990:641; lit. ‘in the descending with which he descended’).23 Obviously, the question is if there exists a Proto-Aramaic “DNA” responsible for the short life of the “more verbal” *maqtal and survival of the “less verbal” *qatāl throughout the history of Aramaic. In other words, how do we explain this trans-Aramaic common drift? I will try to answer this question in the course of the study. We will now sketch, in a preliminary way, a history of Aramaic agent nouns derived from the G-stem verbs. The target of this section is a diachronic comparison (≈ “parallel lives”) of *qātil-, *qattāl-, and *qātōl- as Aramaic agent noun patterns.24 The productive Proto-Semitic agent noun *qātil- (= the Active Participle of the traditional grammar) was still widely used in Middle Aramaic in the syntactic slots of substantives. At the advent of the Neo-Aramaic epoch, it was discontinued in the whole of Aramaic as a productive rule for deriving deverbal nouns or adjectives with whatever meanings. In Neo-Aramaic, *qātil has been completely verbalized and is now the inflected imperfective base (in ENA, it is Jussive/subjunctive as well).25 Proto-Aramaic had the *qattāl- pattern inherited ultimately from ProtoSemitic (cf. Bauer–Leander 1927:191 for examples in BA). Whether it was ever productive (= freely derivable from the G-stem agentive verbs) in the first millennium B. C. Aramaic, we do not know for sure, yet I believe that in the whole of ancient Semitic (starting from Proto-Semitic and including Old and Middle Aramaic), *qattāl- was less frequent and less semantically predictable than *qātil- in the latter’s capacity of agent noun.26 In Syriac, *qātil- in the construct state is freely used as an agent noun (mostly with dependent substantives, see presently): dāµlay šmā− ‘those who revere your name,’ 23 I keep, with minor simplifications, impressionistic (non-phonological) transcriptions of Ritter and do not “normalize” them according to Jastrow’s phonology of the Mid3n dialect. This is in particular because the phonology of the town dialect (Midyat) is different from “Jastrow’s,” and it has not been described as yet. Mundarten of different villages may possess phonological differences as well. 24 We will see that the story is unexpectedly similar to that of the action noun that can now be epitomized as follows: the innovative PA *maqtal had a short life and did not oust the inherited *qatāl that has survived to the present day. 25 We will deal with the history of *qātil- in more detail elsewhere. 26 For the early life of qattal/qattāl (including a comparison of phonological shapes in Akkadian and Hebrew), see Kouwenberg 1997:49–68.

S. Loesov, Reconstructing Proto-Aramaic …

433

dāµlawhy ‘those who revere him = his followers, his faithful ones’ (Nöldeke 1966:212), while qattāl- is well-attested but probably not productive.27 Let us now briefly mention the Syriac data on qattāl-. Nöldeke 1966:70 brings nine illustrative examples of qattāl- associated with the G-stem, followed by an “and so on.” He also has five tokens of qattāl- associated with the D-stem. Stephen Kaufman kindly created for me, on the basis of the Comprehensive Aramaic Lexicon (= CAL), a list of Syriac substantives and adjectives, of whatever origin, whose bases admittedly have the shape CaCCāC-. My doctoral student Maksim Kalinin has analyzed all these data using the available dictionaries and has isolated some seventy-five deverbal agentive nouns (including adjectives, such as e. g. ba¢¢āl- ‘vain; lazy; idle’).28 None of these has a specifically Christian meaning as its basic gloss, and many of them have parallels in other Middle Aramaic varieties. My zero-approximation hypothesis is that most of these seventy-five or so deverbal qattāl-s had been inherited by Syriac from its Old Aramaic parent language rather than freely formed by the speakers of Syriac. In other words, I suspect that Syriac was unable to derive qattāl- action nouns from verbal roots anymore. At this stage, a comparative study of all the pertinent Middle Aramaic material will be most helpful. The etymology of the Middle Aramaic qātōl- is obscure. Nöldeke believes it was derived from qātil- (2005:113; 1966:68). In Central Semitic, qātōl- is hardly known outside Aramaic29 where it had not been really attested before the Middle Aramaic period (Bauer–Leander 1927:191). It is generally believed that in Syriac the agent noun qātōl- can be formed at will of any semantically appropriate verb (Nöldeke 1966:68).30 Nöldeke

27

The semantic description of the pattern in Nöldeke 1966:70 is as follows: “Steigerungsadjectiva, Nomina agentis und Berufsnamen.” 28 Unexpectedly for the present writer, there showed up no less than ten agent nouns derived from Aramaic substantives, e. g. µammārā ‘donkey driver’ (SL 468, with parallels in JBA and Mandaic). 29 By Semitic standards, a deverbal CVCVC- noun with both long vowels looks weird. The [ō] of qātōl- is also unexpected because the only regular source of [ō] in Eastern Syriac is [*aw] in closed syllable (yōm ~ yawmā). 30 Stephen Kaufman informs me in a p. c. (May 10, 2011): “By contrast, there are currently 265 entries listed (in CAL.—S. L.) as nomina agentis of the G stem (i. e. qātōl). The only reason that these particular ones should have been listed as opposed to the remaining thousand or so potential ones (i. e. verbs occurring in the G stem in Syriac = 1223) is that they were thought to have special semantic development by earlier lexicographers. In fact not all do so.”

434

Articles: Semitic Studies

asks the question about distribution of the two productive patterns, qātōland qātil-, and answers it in terms of the surface syntax: – qātil- normally has to have a dependent substantive encoding the semantic object, while bound possessive pronouns in this slot are rarer and “mehr auf bestimmte Wörter beschränkt” (p. 212). – qātōl- is used elsewhere, i. e. (1) without exponents of the semantic object in respective noun phrases; (2) freely with bound possessive pronouns; (3) with the semantic object introduced via d(p. 212ff.; the same distribution holds for active participles and nomina agentis of the derived stems).

Note Nöldeke’s words on p. 212f. that interpret the situation as a complementary distribution of the two (not the three!) nominal patterns, qātiland qātōl: “Für alle die Anwendungen, wo das Part. seltner wird, tritt das Nomen agentis ein.” In Turoyo, an ENA language whose MA parent tongue was quite similar to the Syriac spoken in Edessa around 300 A. D., there remained virtually no deverbal *qātil- nouns (Jastrow 1993:174f.).31 Deverbal *qātōlis marginal.32 By contrast, *qattāl- is productive in its capacity of agent noun, unlike in Syriac and Akkadian. Jastrow 1993:184f. gives a list of some fifty tokens which is not meant to be exhaustive. Consider an example from the town dialect of Midyat with two tokens not listed in Jastrow’s grammar that describes the village dialect of Mid3n: µā-mineye zāmōro u dāyōqo d-kamâna-we u ġalabe mašhûr-we ‘Einer von ihnen war ein Musiker und Geigenspieler und sehr berühmt’ (Ritter 1967:18/105).

31 In the dialect of Mid3n, nominal *qātil- has to have the shape q3tlo. In Jastrow 1993:174f., the only good example of agentive deverbal *qātil- is r3"yo ‘shepherd,’ lexicalized probably already in Proto-Semitic. In text samples of the Midyat dialect, I have come across two more examples: suh¯o ‘witness’ (Ritter 1979:455), probably a cultural loan from Syriac (with Ritter 1979, cf. SL 973), and fu"lo ‘worker.’ 32 Jastrow (1993:188) lists all eleven tokens known to him (= ‘belegt sind’) of qotulo of whatever (not only deverbal) origin. In this list, the only genuine agent noun with Aramaic etymology is no¢uro ‘watchman’ (see also Tezel 2003:25). There are a few lexicalized examples of Aramaic provenance, e. g. noqušo ‘church bell’ (cf. SL 945: ‘musician’; ‘bell clapper’; ‘sound,’ etc.) and koruxo ‘shroud’ (cf. SL 648: ‘band’; ‘weaver’s beam’). Jastrow 1993 brings two agent nouns derived from Turoyo verbs with roots borrowed from Arabic: šohudo ‘witness,’ xodumo ‘servant.’ In the actual speech activity, *qātōl- (> qotulo) seems to be rare indeed. Thus the informants whose records were included in Jastrow’s Lehrbuch (Jastrow 2002) did not use it a single time, while *qattāl- (> qatolo) is frequent in the glossary of the textbook.

S. Loesov, Reconstructing Proto-Aramaic …

435

The agent noun zāmōro is an old Aramaic word, it is attested in BA (Esr 7:24), while dāyōqo is formed of the verb root dyq (doy3q—d3qle) ‘klopfen (an Tür)’, ‘schlagen’, ‘spielen (Karten u. a.)’, ‘spielen (Musikinstrument)’ that is common in Turoyo (Ritter 1990:573ff.; Jastrow 2002:161 has ‘stampfen’). The Turoyo root dyq is the Aramaic adaptation of the Arabic root dqq ‘to strike,’ etc. In Anatolian Arabic, dqq is ‘anschlagen,’ ‘stampfen,’ ‘berühren, antasten,’ ‘spielen (Musik)’ (Jastrow 2005a:51).33 To take another example, ¢aboxo ‘cook’ (Jastrow 1993:185) is no Arabic loan, it is derived from the regular Turoyo verb ¢ob3x—¢b3xle ‘to cook’ (Ritter 1990:68) that was formed of the root borrowed from Arabic. Moreover, Ritter 1990:57 and Jastrow 2002:142 label *qattāl- the participle of intransitive motion verbs.34 It is small wonder that azolo and a¬oyo are used as action nouns ‘the one who goes/comes’ (Ritter 1990:764), but an important vaticinium ex eventu for our theme is that *qattāl- of this group of verbs regularly appears as a predicate, i. e. a progressive verb form (accompanied by a copular element of some kind): this innovation is vividly reminiscent of the Old Aramaic *qātil- that was both an agent noun and the New Present. Consider a couple of examples: m3lle “k-oba"no, d-m3blatli "am rūµux, lî-duk¬o d-izzúx!” — “!ē” m3lle “ono u-dārbei¯i rāµûqo-yo nāµôto-no lî-walāe d-Buġdid.” ‘[X said] “Ich wünsche, daß du mich mit dir nimmst an den Ort, wo du hingehst.” — [Y replied] “Ja, aber mein Weg ist weit, ich ziehe hinab nach der Stadt Bagdad” ’ (Ritter 1969:30/273f.).

How the sense would change if the text read ko-noµatno (i. e., the regular Present) instead of nāµôto-no, I cannot tell. A speaker of the "Iwardo village dialect granted me in a p. c. that the two sentences may say the same thing. kile uw-abro dú-malko !a¬ōyo, nfāqu qāmū¬e ‘Der Sohn des Königs kommt! Zieht aus, ihm entgegen!’ (Ritter 1967:152, 240). 33 In Turoyo, all Semitic hollow and geminated roots (whether Aramaic or loans) are reflected as R1yR3. Syriac has both dwq ‘to crush’ (SL 286) and dqq ‘to pound, crush’ (SL 318), which are cognates of the Arabic dqq, but the examples available in Ritter 1990 and in published texts prove at least the semantic influence of the Arabic root. 34 Cf. Jastrow 1993:184, fn. 1: “Der Strukturtypus C1aC2óC3o entspricht AWS (= Old Western Syriac.—S. L.) qa¢¢ōlō. Wie im AWS umfasst er Berufsbezeichnungen, daneben jedoch auch die nomina agentis zum I. Stamm der dreiradikaligen Verben, die im AWS als qō¢ūlō gebildet werden. Die Nomina agentis werden zum Teil adjektivisch verwendet. Bei einigen Verben in I.N (= intransitive G-stem verbs.—S. L.) dienen sie als Partizipien.”

436

Articles: Semitic Studies

In the context, the message can hardly be anything but ‘the prince is coming.’ Note the conjugated (3ms) presentative particle kile ‘le voici’ (Ritter 1990:38–42; Jastrow 2002:106) that probably renders the postpositive 3s copula -yo superfluous (i. e., either kile a¬oyo or a¬oyo-yo, cf. raµuqo-yo and naµoto-no in the previous example). This presentative particle serves, among other things, to actualize the ka-naš3q/ko-noš3q ‘he is kissing’ Present of ENA (see Khan 2007a:115) that is itself an actualization of the Old Aramaic Present nāšiq.35 Now, one has to ask why the qātōl- agent noun appeared in Middle Aramaic, how the above distribution of qātil- and qātōl- had arisen in Syriac, and how it came about that qātil- as a nominal derivation rule was shortlived in Aramaic. Finally, one has to understand why *qattāl- was promoted to the “active participle” for intransitive motion verbs in Turoyo.36 My explanation runs as follows. In the historical pre-modern Aramaic (as well as in Akkadian,37 perhaps less so in Hebrew), qātil- was not used as a productive attributive adjective.38 Let us consider the vocalized BA corpus. According to the “BibleWorks 7” software, BA has 204 tokens of qātil-.39 Seventeen of them are syntactic substantives, all seventeen being semantically agent nouns. At least one noun (appearing more than once

35

Ko/a- probably goes back to kā ‘here’ (cf. CSD 201a and the BH kō). The standard etymology relates this particle to qā!em (Khan 2007a:115), cf. the modern Mandaic Present qa-gā¢el (Voigt 2007:163). The deictic function of the prepositive particle(s) remains the same, with whatever etymology. 36 Note that in this language G-stem transitive verbs do not have an active participle. 37 See Kouwenberg 2010:203–210. 38 This is what “the Participle” is in the traditional understanding: it is a productive deverbal adjective that possesses certain voice and tense-aspect properties. Ideally, the Participle is related to individual tense-aspect and voice paradigms of the verb rather than to the verb root in general, a good example is Greek with its ten participles for a given transitive verb. One thing that follows from this definition is the ambivalent nature of the Participle: in order to be the Participle it has to be a genuine Adjective, i. e. to behave as the target of agreement in noun phrases in the same way as the other adjectives of the language. If it has tense/aspect and voice properties, it is likely to be a syntactic head governing the accusative case (in a more general way, “the non-genitive case”) of the dependent substantive, which is not a possibility for the garden-variety attributive adjective. A “normal” adjective, if substantivized, will govern the dependent noun in the genitive. 39 I thank Maksim Kalinin who helped me process this material.

S. Loesov, Reconstructing Proto-Aramaic …

437

in the corpus) is definitely lexicalized: gāzrā ‘the diviner’;40 yā"¢ā ‘the adviser/counselor’ is most probably a genuine substantive as well, since finite G-stem forms of the root are not attested in Aramaic. Pālµā ‘the (temple) servant’ (Ezra 7:24) may also belong here, while šāf¢ā ‘the judge’ (Ezra 7:25) is probably a Berufsname formed directly of the Hebrew root according to the then productive pattern. All the lexicalized items happen to appear in the plural (this may be due to the contents of the stories, with no grammatical reason whatsoever), the forms are either unbound (both “long” and “short”), bound with possessive pronouns or construct. “Substantivized participles,” i. e. agent noun qātil- forms that do not deviate from lexical meanings of the respective verb roots, are represented by nine tokens (out of the total of seventeen syntactic substantives).41 All of them but one are in the construct or bound with possessive pronouns (i. e., in agreement with the Syriac usage), e. g. yā¯"ē dā¬ē !ĕlāhā− ‘those who know the laws of your God’ (Ezra 7:25), and see Dan 2:21, 28, 29, 47; 4:16, 32bis; the only free-standing example is Dan 7:16 (µa¯ min qā!ămayyā ‘one of those standing’), and see Bauer–Leander 1927:296. I have found only two impeccable attributive examples of qātil- in the BA corpus: nūr dāliq ‘burning fire’ (Dan 7:7 and six more times in this chapter), nūrā yāqi¯tā ‘flaming fire’ (Dan 3:15).42 The rest of qātil- forms in BA (i. e., some 180 out of 204) are predicative in different kinds of constructions: either “naked” (encoding Present Actual, or roughly synonymous to the past-time “Perfect,” or to the future-time “Imperfect”), or in conjunction with the finite forms of the verb ‘to be,’ or with the existential word ‘there is’ (see also Bauer–Leander 1927:290–296). Thus non-predicative syntax of qātil- is marginal in BA. At this point, I am not ready to discuss the nature of the above partial isogloss between Akkadian and pre-modern Aramaic: on the one hand, the nominal qātil- in both languages is agent noun rather than attributive adjective; on the other hand, in the historical Aramaic qātil- is predominantly the base of the New Present, while in OB and OA letters the predicative use of the Active Participle is virtually unknown (it is rare even in literary texts). The productive predicative use of the PS *qātil- in Central Semitic is doubtless innovative vis-à-vis the PS situation. Whether the PS 40

I use long forms as lemmata and do not write the Biblical šwa mobile to make this material easily compatible with the Syriac one. 41 The division in two groups is in borderline cases uncertain and subjective. 42 I do not count instances of the double accusative construction to “pure” attributive usages; cf. Dan 4:20; 6:12.

438

Articles: Semitic Studies

the PS *qātil- was primarily an agent noun or a deverbal adjective we do not know. Yet it is the attributive (and not the agent noun) usage that has to be the source of qātil- as the emerging New Present. “I am writing” can develop from “I am a writer,” but non-agentive verbs with non-volitional subjects (e. g., ‘to burn,’ ‘to decay’) can hardly have regular agent nouns at their side, while the qātil- Present can be formed of almost any Syriac G-stem verb. So it is simpler to posit the same source of the Present for all verbs rather than to conjecture an analogical development for nonagentives. This makes me suggest that (regardless of the interpretation of the Akkadian evidence) in Aramaic the attributive use of qātil- was discontinued early because it was “absorbed” almost completely by the predicative one. The predominantly verbal identity of qātil- lead to the above syntactic restriction on its use as an action noun: its appearance in the construct had to safeguard its substantival nature. The qātōl- agent noun43 was probably derived from qātil- as its renewal in the changed situation (with Nöldeke), though I cannot explain why this happened via the [i] > [ō] replacement implied in Nöldeke’s words. All this means that we must understand the BA and Syriac evidence on qātil- as stages in the common drift towards Neo-Aramaic where the substantival deverbal qātil- would not survive, having being replaced by other agent nouns.44 Now, since qātōl- appears in both Eastern and Western branches of Middle Aramaic,45 it must have been a feature of the immediate MA ancestor of Turoyo as well,46 though as we saw *qātōl- is only marginally attested in Turoyo. At this stage, the only thing we can do is to state that qātōl- as a

43 It is attested in the whole of Middle Aramaic, see in particular Nöldeke 2005:113, Dalman 1905:154, Müller-Kessler 1991:90, 161. Strange as this sounds, qātōl- may itself have been an instance of common drift, since it cannot be traced down to a proto-language. 44 I thank Aleksey Lyavdansky for checking the data of various NENA idioms, in particular those of Tkhuma, Challa, Mangesh, and Sardarid. 45 Unfortunately, the post-Nöldeke descriptions of MA corpora do not usually ask various questions that Nöldeke had in mind when he was working on his Kurzgefasste Syrische Grammatik. Note that his Syntax (conceived practically and informally, i. e., meant to describe what is special and worth thinking about) occupies more than a half of his Syriac Grammar, while subsequent generations of standard MA grammars usually do not have Syntax at all. Why?—These words are my excuse for not pursuing the story of agent nouns across all of MA in detail. I hope to deal with this theme later. 46 I. e., of an EA dialect close to Edessan but not identical to it.

S. Loesov, Reconstructing Proto-Aramaic …

439

productive agent noun47 was very short-lived and (at least in Turoyo) for some reason replaced by qattāl-, which became productive (probably for the first time in the history of Semitic languages) and, in its turn, shifted to the Participle/Present of a semantically delineated group of verbs, by way of a partial Wiederkehr des Gleichen. At the moment, I cannot explain the reason. 0.3. Grammaticalization of Definiteness Another type of evidence that invites attention is that Aramaic keeps reincarnating into grammar certain semantic values that are not generally speaking the most obvious candidates for grammaticalization. An example is definiteness/referentiality. As is well-known, PA possessed status, i. e. inflection of nouns and adjectives48 for definiteness, partly cumulative with inflectional markers of number (and of gender, in the adjectival paradigm). Below is the reconstructed PA definiteness paradigm of the adjective ¢āb- ‘good’.49 Table 1 gender

status

Sing.

Plur.

Masc.

indefinite

¢āb-Ø

¢āb-īn

definite

¢āb-ā!

¢āb-ay-ā!

Fem.

indefinite

¢āb-ā

¢āb-ān

definite

¢āb-at-ā!

¢āb-ā-t-ā!

47 Curiously, qātōl- is supposed to have been an active participle somewhere in MA as well (Macuch 1982:248). Note that according to Müller-Kessler, the participial *qātōl- had verbal usage in SA and CPA, “alternierend mit qā¢el” (MüllerKessler 1991:161). From the point of view of the present study, all this looks highly significant and provocative, but the lack of syntax and semantics in both reference grammars does not allow one to exploit these hints. A new study of the respective corpora is necessary. 48 Attributive adjectives agree with their heads for number, gender, and state. In what follows I inflect a PA adjective rather than two substantives to save space and because the suffixes of adjectival inflection are completely predictable. 49 I posit the proper PA marker of definiteness to be *-ā! rather than *-a! because in Syriac [*a!] often yields [ē]. Yet choosing *-a! would not affect the gist of the argumentation. The reconstructed paradigm follows the scholarly consensus, it is supported even by the reductionist judgment J. Huehnergard voiced in the last lines of his essay “What is Aramaic?”, “Aramaic in my view is thus considered to begin with the presence of the following three features: a definite article represented by final -!; and the loss of the N stem; and feminine pl. forms in -ān; it is these features that may be said to characterize Proto-Aramaic” (Huehnergard 1995:282).

440

Articles: Semitic Studies

Among the eight members of this paradigm there are no two identical forms, whether we reconstruct fs. indef. as *¢āb-ā or *¢āb-at (which is less likely), and this situation is cross-linguistically rare for nominal declensions.50 This means that the PA adjective from the very start was designed to track down the grammatical values of its head noun. The evidence contradicts the suggestion of Huehnergard 1995:270, “the occurrence of the article is to a large extent syntactically determined, rather than semantically … Given the limited distribution of the article in the earliest Aramaic inscriptions, one may extrapolate backwards and suggest that this seemingly most characteristic of all Aramaic features was, in ProtoAramaic, merely a distinctive phonological reflex of a common Central Semitic phenomenon.” Independently of our judgment about the language of 9th–8th centuries Aramaic inscriptions and its value for the PA reconstruction, the postpositive article was deeply entrenched in the PA nominal declension, as witnesses e. g. the fem. plur. opposition ¢āb-ān (indef.) vs. ¢āb-ā-t-ā! (def.), while, after the pre-Aramaic fall of inflection for case in both sg. and pl. of the noun, the pre-Aramaic all-purposes fp base *¢āb-āt- survived only as the head of noun phrases with dependent substantives. The common drift of Aramaic languages led them to lose the PA definiteness paradigm, i. e., the original morphosyntactic exponent of definiteness. In none of the Neo-Aramaic languages has it been preserved as a meaningful whole. The main reason was the increasing overuse by speakers of the “full” (i. e., definite) forms to the detriment of the short ones. In response, individual Aramaic varieties created new methods of encoding definiteness, including cataphoric pronouns in MEA and prepositive articles in Turoyo and part of NENA (Jastrow 2005; Jastrow 2008, Khan 2008).51 The example of Turoyo/Mlaµsô is the most telling one for 50

Construct forms of PA nouns have nothing to do with this definiteness paradigm, because their appearance is called for by a purely syntactic context: if you are going to expand your noun phrase by a dependent substantive (including pronoun), you have to select the appropriate “prepositive” shape of your head noun, whatever its definiteness value. Morphologically, construct auslauts are frozen structures of pre-Aramaic origin: ms cs = ms indet ( ¢āb), mp cs ¢āb-ay is not derivable from a productive PA nominal inflected form, fs cs ¢ābat is most probably pre-Aramaic, as well as fp cs ¢ābāt (it would be illogical to derive it synchronically from the fp det ¢ābāt-ā!, and it is impossible to derive it from the fp indet ¢āb-ān). 51 While finishing this paper, I came across an abstract of an unpublished talk by Na’ama Pat-El from 2009 (“Cyclic changes and the development of the article in

S. Loesov, Reconstructing Proto-Aramaic …

441

our theme: in the singular, there are two prepositive articles (not copied onto attributive adjectives), agreeing with the head substantive in gender (masculine or feminine); in the plural there is only one article, since both the substantive and the verb in Turoyo (as in most of Neo-Aramaic) do not have gender in the plural. All three prepositive articles go back to MEA cataphoric/proleptic pronouns that used to be lexical and analytical devices to render definiteness: hū ‘he,’ hī ‘she,’ hānnon (or, more likely, hānnēn) ‘these.’52 Consider the Turoyo examples: u=kalbo=rabo ‘the big dog,’ i=qri¬o=rab¬o ‘the big village,’ ak=kalbe=rabe ‘the big dogs,’ a=qriyawo¬e=rabe ‘the big villages,’ ann=abne ‘the sons’; i. e., the etymologically basic form of the plural article ann- gets three different shapes depending on the substantive anlaut. To sum up: in the case of Turoyo, Aramaic managed to break through and to build a grammatical expression of definiteness that is structurally similar to the PA one (in particular, there is no obligatory indefinite article). PA happened to employ to this end an enclitic demonstrative particle, unlike Hebrew and Arabic. When the PA definiteness paradigm had worn out and collapsed, individual Aramaic idioms started putting proclitic pronouns to the service of encoding definiteness (consider the Syriac example in the last footnote). The proclitic strategy was a step enforced by the mess created in the suffixing nominal inflection after the PA paradigm (Table 1) had disintegrated (for details, see e. g. Nöldeke 1966:144–154). At least in the case of Turoyo, the proclitic method was a success: it culminated in a full-fledged definite article, opposed to the zero grammatical expression of indefiniteness. Neo-Aramaic,” http://utexas.academia.edu/NaamaPatEl/Talks). It is germane to my theme: “This paper attempts to point to another connection between Neo-Aramaic innovations and processes which are attested in much earlier dialects of Aramaic. Aramaic shows a cyclic change of Demonstrative order, from the common Semitic postposition in Old Aramaic (Sefire: b-spr! znh ‘in this inscription’), to pre-position in Late Aramaic (Mandaic: halin nišmata ‘these people’) and back to post-position in Central Neo-Aramaic (àuroyo: u-bayt-āno ‘this house’). Aramaic seems to have done so as a result of internal pressure, and not contact. In this paper, I will argue that the word order change is related to the growing use of anticipatory pronouns to mark definite nouns in all dialects since Official Aramaic. Eventually, the preposed demonstrative was used as a definite article in some Neo-Aramaic dialects.” Compare especially Pat-El’s “cyclic” with Jastrow’s (2008) “circular” developments, while both writers refer to the marking of definiteness throughout the history of Aramaic. It looks like the idea of Aramaic common drifts is now “in the air.” 52 Cf. the !emar hū malkā construction (lit. ‘said he-the-king’) that is frequent in Syriac. Hū malkā is an immediate source of the ms article of Turoyo.

442

Articles: Semitic Studies

1. Reconstructing Proto-Eastern-Aramaic In order to gain a glimpse of Proto-Aramaic, it is best to start with the more “progressive” of its two firstborns, i. e. PEA. I suggest that the genealogically relevant features of PEA are as follows: 1.1. At the separation of PEA, l- replaced y- in the Jussive (= the short form of the prefixing conjugation). In the course of early Eastern Aramaic history, l- replaced y- in the respective forms of the indicative (3ms, 3mp, 3 fp), in order to prevent the threatening homonymy of the 1cs and the 3ms indicative forms (with Rubin 2007). Finally, some of the daughter-languages of PEA swapped lfor n-.53 As for the Jussive, EA lost it very early due to the morphosyntactic developments that will be mentioned in the second part of this paper. The whole story looks as follows. The PA paradigm of the two varieties of the prefixing conjugation has to be reconstructed as in Table 2 (cf. Degen 1969:65; Muraoka–Porten 2003:104ff.; Huehnergard 1987).54 Table 2 Indicative 3m 3f 2m 2f 1c

sing. yaqtul(u) taqtul(u) taqtul(u) taqtulīn aqtul(u)

plur. yaqtulūn yaqtulān taqtulūn taqtulān naqtul(u)

Jussive sing. yaqtul taqtul taqtul taqtulī aqtul

plur. yaqtulū yaqtulna taqtulū taqtulna naqtul

My thesis (1.1) means that the Proto-Eastern Aramaic paradigm of the prefixing conjugation is different from the Proto-Aramaic paradigm only due to *laqtul, *laqtulū and *laqtulna in the Jussive. 53

Semitic languages demonstrate a few more cases of l > n in affixes, but I cannot explain the reason of our particular shift. 54 As is well known, both Early and Achaemenid Aramaic (the most ancient witnesses of WA) opposed orthographically yµzh ‘he will see’ to yµzy ‘let him see.’ This fact can be construed in the sense that Proto-Aramaic had yaqtulu rather than yaqtul in the indicative, though this interpretation is not obligatory. The post-positive article of the Proto-Aramaic noun was created after the short case vowels had been lost, so the short vowel apocope may have been a part-of-speech independent shift characteristic of Proto-Aramaic in statu nascendi. According to the latter scenario, *yaµz-ē ‘he will see’ as against *yaµz-ī/yaµz-ay ‘let him see’ were inherited by Proto-Aramaic as unanalysable “endings” of IIIy verbs. In this paper, I sometimes use *yaqtulūn etc. rather than *yaqtulūna etc. as a PA form, but for most of my purposes whichever notation is good.

S. Loesov, Reconstructing Proto-Aramaic …

443

In Aramaic texts, the l-Jussive is attested but in TF;55 within the first millennium NWS, the l-Jussive of TF forms an isogloss only with KAI 214 (Samalian). In TF, all three third-person forms (*laqtul, *laqtulū, and *laqtulna) are attested (Abou-Assaf et al. 1982:49f.). TF has yet more linguistic traits otherwise unusual for the first millennium Aramaic (we will address some of them in due course): (1) Samek to render the PS *¬ (e. g. µds ‘new,’ TF 11). This fact has received no cogent explanation. The most popular guess is that the reason is orthographical (a partial disambiguation of the polyphonic ‫ )ש‬rather than phonological, though of course there is no knowing. (2) There is the consistent syncope of [h] in the bound adnominal pronouns of the 3 pl.: l-mt kln ‘to all the lands’ (TF 3); l!lhyn klm ‘to all the gods,’ nhr klm ‘all rivers’ (TF 4).56 This feature reminds BH (3mp: klm, kwlm, along with klhm; 3fp: klnh, klhnh; see also Garr 1985:54–58 for an analysis of the epigraphic evidence). Against the Aramaic background (including Syriac), this is an unusual morphonological trait. (3) !l ygtzr ‘let it not be cut off ’ (TF 23): t-infixation, as in Ugaritic and the Mesha inscription (this is the only token of a t-stem in TF). (4) The verb forms written as knn in TF 10 (‘he erected’), and lknnh TF 11 (‘let him erect it’) are supposed to belong to the D-stem of √ kwn. For the 1st millennium NWS, this kind of orthography is a Canaanite feature (cf. Garr 1985:133f.). (5) There is no nota accusativi, unlike in KAI 202 (Zkr king of Hamat), KAI 222–224 (Sefire),57 KAI 320 (Bukan).58

As a counterpart to the five above “non-Aramaic” features, TF has at least two important traits that first surfaced as the standard ones in IA and are present throughout the Middle Aramaic varieties: the productive Ginfinitive *maqtal, and zy to introduce both dependent substantives of noun 55

In this paper, I will not use the evidence of admittedly Aramaic and semantically Jussive l-forms within personal names preserved in cuneiform sources. I hope to analyze this material later in my study. 56 On the assumed broken plurals in TF, Sefire and the rest of early Aramaic, see most recently Lipiński 2008. 57 See DNWSI 48 for references to discussions of paleographic and philological problems. 58 In Aramaic inscriptions from the 9th and 8th centuries, !yt does not occur in KAI 310 (Tell Dan), where it is probably not needed in the extant fragments. It is not attested in the Aramaic texts from Zincirli. In these texts, there seems to be at least two contexts where !yt would be appropriate (see Tropper 1993:233, his sentences Nos. 226 and 233).

444

Articles: Semitic Studies

phrases and relative clauses (dmwt! zy PN zy šm qdm GN ‘the statue of PN that he set before GN,’ TF 1). The G-infinitive *maqtal and *¯ī as nota genitivi have no first millennium B. C. NWS parallels outside genuine Aramaic, and they are virtually unknown in the early epigraphic texts in Aramaic outside TF.59 From the point of view adopted in this study, at least *maqtal may well have been present in PA, yet this hypothesis does not explain why zy and *maqtal should appear in TF and be absent from the other earliest witnesses of Aramaic (they use “annexation” and *qatāl-, respectively).60 Thus I believe TF is the earliest available textual witness of Eastern Aramaic. The *-ayyā! of m!ny! zy bt hdd ‘the vessels of the temple of Hadad’ (TF 16f.) does not really stay in the way, in particular because short nominal bases can preserve this PA ending even in Syriac (Nöldeke 1966:49). We will now shortly review the evidence of Samalian. The l-prefix with Jussive force is attested six times (counting reliable restorations) in KAI 214, e. g. p-lktš-h b-!bny ‘and then let themmasc. stone him,’61 i. e. probably laktašū-hu (line 31); p-lktšn-h b-!bny ‘and then let themfem. stone her,’ i. e. probably laktašna-hā (ibid.), ltgmrw … zkrw ‘let all the males come together’ (lines 30–31), hdd µr! lytkh ‘let Hadad pour (K-stem of √ ntk) heat [of wrath] over him’ (line 23).62 In KAI 214, y-forms seem to be used with the same force, e. g. w-yqm wth ‘and let him put him’ (line 28; note the syncope of h in the K-stem). In Samalian, the 2mp and 3mp forms of the prefixing conjugation with indicative readings had no final -n: thrgw ‘you will kill’ (KAI 215:5), y"bdw !rq ‘they used to cultivate the land’ (KAI 214:7). This means that in Samalian (unlike in PA and in Old [Western] Aramaic) there was no y-Jussive opposed to the y-indicative inflectionally, i. e. via suffixes. Two synonymous Jussive forms in the same language (yaqem and laqem ‘let him 59

The only alleged *maqtal infinitive in Sefire is l-mšlµ yd ‘to raise a hand’ (KAI 222B:34). This reading is supported in Fitzmyer 1995:50. Degen 1969:68 works with a different paleographical reading, l-yšlµ, which would give the same sense. The passage is badly damaged and syntactically difficult, so it is hardly useful in this kind of research. Outside TF, the only reliable early-epigraphic token of zy in a noun phrase is kl mlky! zy sµrty ‘all the kings of my vicinity’ (KAI 224:7f.). Zy as a relative particle is well-attested in Sefire (Fitzmyer 1995:201), it is known in Samalian as well (Tropper 1993:193). 60 Note that the TF lexicon seems to be impeccably Aramaic (Kogan 2005, in particular 540–543). 61 On the verb ktš ‘to beat, strike’ in Aramaic, see SL 663b; for the Semitic data, see HALOT sub voce. 62 The readings and interpretations are as in Tropper 1993:82, 94f.

S. Loesov, Reconstructing Proto-Aramaic …

445

establish,’ as against yaqīm ‘he will establish’) are not a likely possibility. This is pace Tropper 1993:237, “In den Texten von Zincirli sind die volitivischen Modi Jussiv (= Präfixkonj.-Kurzform) und Prekativ (lū + Jussiv) bezeugt, wobei der Prekativ als verstärkter, emphatischer Jussiv fungiert.” (As the examples ibid. reveal, Tropper counts w-yqm wth ‘and let him put him’ and other semantically volitive prefixing forms in KAI 214 to morphological Jussives). As Tropper proposes (p. 83f.), the l- of the Samalian “precative” is no modal particle, but a genuine prefix, even in the unique and much-discussed K-stem example lytkh ‘let him pour over him,’63 where according to Tropper (1993:84) /y/ serves as the secondary K-stem marker, given the syncope of the causative affix /h/ in Samalian.64 This picture makes the existence of Tropper’s “emphatischer Jussiv” (i. e., to a certain extent an analytical form) unlikely. It is no incident that “emphatic” forms are attested for 3ms, 3mp, 3fp (all three have y- in the indicative), but not for the 3fs, for which only “non-emphatic” occurrences are attested, e. g. tšty ‘let her drink’ (KAI 214:22).65 Note that in Old Assyrian the relevant Jussive form is lū taprus, e. g. amtum išti PN lu ta-ta-al-kam ‘let the slave-girl come together with PN’ (CCT 3, 14:21). Structurally, there seems to be no reason why the “emphatic” *lū should not appear before consonantal prefixes of the Central Semitic Jussive as well. Thus in Classical Arabic, as our reference grammars explicitly state, the apocopate of all persons normally needs the particle li-/l- to get the positive Jussive force: fa-l-na!¶u¯ ‘so laßt uns nehmen!’ (Fischer 1987:96), fa-l-tuba" ‘[if anything happens to me,] then let it be sold!’ (Wright 1862:24).66 Rubin 2007:17 seeks to explain the EA picture (the one whose origins I am trying to trace back to PEA) roughly at the time level of Middle

63

Cf. in particular Testen 1998:108. I suspect, as probably Tropper does, that *lū in lytk (lVyattik) may be less morphologised than etymologically selfsame element in lktš /laktaš/ ‘let him stone’ (the vocalization la- or la-/li- in compliance with Barth’s law is strongly suggested by cuneiform renderings cited in Beyer 1984:109f., cf. Testen 1998:109). Phonetically, the syncope of /-h-/ is not related to the presence or absence of /l-/, since the trigger of the syncope is the /y-/, i. e. yattik < *yVhantik (Tropper 1993:83), in contrast to the Old Aramaic verb form lhynqn (probably lahayniqna) ‘let them suckle’ (TF 20f., bis), where the syncopated element is /y/ (< *luyahayniqna or the like). 65 In Samalian orthography, the relevant semantically indicative IIIy forms ended in -y as well, unlike in Old Aramaic (Tropper 1993:213f., 224ff.). 66 Wright ibid notes that ‘li… is very rarely omitted, except perhaps in poetry.’ Cf. also Huehnergard 1983:578. 64

446

Articles: Semitic Studies

Aramaic. For him, the trigger of the y- > l- change in the 3ms form is the shift yi- > i-/#__ that took place “before the first attestation of Syriac in the first century CE.” This change was fraught with the coalescence of the 1cs and 3ms in the common form iq¢ul, and this “intolerable” homonymy was overcome in the following way: “In the third person, there existed a rare Jussive form, identical with the imperfect but for the addition of a prefixed l-. Since any imperfect could also serve as a Jussive, this special Jussive form liq¢ul must have come to be thought of as a bi-form of yiq¢ul. Therefore when iq¢ul (< yiq¢ul) was no longer an acceptable 3ms form, liq¢ul was the logical alternative. The third person plural forms were always generated from the 3ms form, and therefore the l- prefix replaced y- there as well” (Rubin 2007:17).

I accept Rubin’s disambiguation thesis as the reason of l-’s appearance in the indicative paradigm, yet his “rare Jussive form” comes into play a little bit like a deus ex machina.67 His typological parallel with Tigré is felicitous,68 except that Tigré has a full-fledged Jussive paradigm with l- in the 3ms, 3mp and 3fp forms, while Rubin’s source of l-/n- in the whole of the MEA Imperfect is a “rare” and optional 3ms Jussive liqtul, somehow available relatively shortly (?) “before the first attestation of Syriac in the first century CE.”69 If one is to understand the MEA evidence, one has to ask first about the emergence and the subsequent life of the l-Jussive in Aramaic. Let us compare the l-Jussives in Aramaic and Akkadian. Akkadian created its precative (= the l-Jussive) paradigm in order to disambiguate the prefixing conjugation. At a pre-Akkadian stage of Eastern Semitic 67 Rubin 2007:13–15 surveys the data on the l-prefix in Central Semitic and Old Aramaic, but he does not embark on the task of justifying its appearance in Aramaic nor enter the ensuing problems of genealogical classification of Aramaic varieties. 68 A. Rubin believes that in Tigré, the Jussive prefix l- appeared in the Imperfect paradigm “to make the third person masculine unambiguous,” i. e. to prevent its coalescence with the 1cs form (Rubin 2007:18). This approach to the Imperfect (or “Present”) paradigms in Tigré has received a detailed development in Bulakh 2011. 69 Rubin does not propose a terminus a quo, yet we do not expect the shift yi- > i-/#__ to be an early development, since the “attenuation” yaC > yiC in the nonstressed syllable had to occur first. Besides, at the turn of the Christian era there was hardly a morphological Jussive left in Aramaic (Beyer 1984:152). In spoken Aramaic varieties, the Jussive must have been lost a few centuries earlier (Muraoka–Porten 2003:200, and see a discussion presently). It is understood that EA lost the Jussive earlier than WA.

S. Loesov, Reconstructing Proto-Aramaic …

447

(and probably in Proto-Akkadian as well), the inherited PS prefixing conjugation used to have both PRETERIT and JUSSIVE readings, and this nontrivial homonymy (or polysemy) came to be felt as unsatisfactory.70 The rest of sufficiently known ancient Semitic languages disambiguated the prefixing conjugation by creating a new PRETERIT qatala71 while keeping the Jussive paradigm without serious morphological and semantic changes for more or less long time.72 As for Akkadian, on the one hand it did not renew its PRETERIT before the end of OB and OA periods;73 on 70

I. e., if we do not buy wholesale into the hypothesis of Hetzron 1969 about the PS Preterit (or “Perfect,” apud Hetzron) yáqtul vs. Jussive yaqtúl. For the Akkadian we know, this hypothesis is irrelevant anyway. As R. Hetzron (1969:5) admits, there is no synchronic Akkadian evidence to support the stress opposition íprus vs. liprús. Pace Hetzron, we do not need this hypothesis for the proto-Akkadian either, i. e. it is not necessary “to explain contraction in the optative and non-contraction in the asseverative” (ibid.), because the contraction of *lū and *yiprus into liprus is not explainable by any synchronic rules we know of (especially since there is no vowel contraction in OA), it is simply the formal expression of the diachronic shift which led to the new paradigm. By the same token, the non-contraction in the asseverative lū iprus (e. g. lū išām ‘er hat (es) gekauft!’, VAB 5, 287:22) is readily explainable by the fact that the asseverative lū is optional for the kind of meaning expressed with its help, it is a real particle rather than a prefix, and cf. GAG 292. See also Kouwenberg 2010:130f. for a detailed criticism of Hetzron 1969. 71 Kouwenberg (2010:126) points in the same direction, if I understand him correctly: “In the historical period, the irrealis use (of *yiprus.—S. L.) was not grammaticalized as a result of the loss of the past tense function, as in West Semitic, but because the particle (l-.—S. L.) became obligatory. From that moment, the irrealis function resided in the particle.” 72 As is well-known, distinct reflexes of the PS Jussive survive in MSA and in modern Ethiosemitic languages. Most importantly for us, in certain individual idioms of both groups l- appears as a prefix of the Jussive (for MSA, see the comparative tables in Simeone-Senelle 1997:405ff.). As we have mentioned above, in Tigré the l-prefix is a feature of 3ms, 3mp and 3fp forms of both the Jussive and the Imperfect. It is always present in these forms of the Jussive, while in the Imperfect l- is in complementary distribution with the zero prefix depending on (morpho)phonological conditioning, see Raz 1983:55f., Voigt 2009, Bulakh 2011. For our theme, there are two differences between Tigré and the attested EA: 1) unlike Tigré, EA preserved an unambiguous reflex of the PS Jussive but in TF; 2) in EA, l-/n- always appears in the three finite forms of the Imperfect. It is a commonplace that the languages with the *yaqattal Imperfective tend to preserve distinct and semantically loaded reflexes of the PS Jussive, while Central Semitic, in the course of its history, lost modal oppositions in the prefixing conjugation. 73 It looks like in spoken varieties iprus was replaced by iptaras in (at least) main clauses towards 1500 B. C.

448

Articles: Semitic Studies

the other hand, all known dialects of Akkadian set out to renew the whole of their Jussive paradigm with the help of the l-prefix early, most probably no later than 2500 B. C., as the joint testimony of Sargonic, OA and OB indicates.74 The Aramaic development must have possessed a different motivation, since PA had *qatala for PRETERIT. Further, unlike in the case of Akkadian, the l-prefixation in 3ms, 3mp and 3fp Jussive was not panAramaic, it affected the whole of EA, while the historical WA had never had the l-prefix and was keeping the PA Jussive paradigm after it had parted ways with EA.75 According to the consensus view,76 the EA prefix n- < *l-.77 This leaves us with three speculatively thinkable accounts of the path along which the MEA evidence came into being (the etymology of l- and the motivation of the development are another two questions, to be discussed below): 1) l- was a feature of the PEA Jussive, as the present writer believes; 2) l-Jussive arose in the course of the common drift of various Eastern Aramaic idioms: they had not inherited this l- from a mother-tongue but developed it independently (or via interdialectal loans, which I believe are impossible to explain); 3) l-Jussive was a feature of PA lost in historical WA idioms.

As we have already seen, possibility (3) is improbable because early WA still possessed the essentially PS Jussive (unlike the PS one, with -na rather than -ā as the fp suffix). Possibility (2) cannot be ruled out, but if there is choice it is safer to explain a shared innovation as stemming from a proto-language, because, other things being equal, it is alternative (and more sophisticated) explanations that have to bear the burden of proof.

74

The precative paradigms of OB and OA can hardly be traced down to a single-parent paradigm in Proto-Akkadian, the shapes of precative in OB and OA may be due to a common drift (= parallel developments in the daughter idioms of Proto-Akkadian). In this case, we can safely pinpoint the trigger, i. e. the PRETERIT-JUSSIVE homonymy in PS and Proto-Akkadian. Cf. Kouwenberg 2010: 215: “Since the outcome of this grammaticalization process is different for Babylonian and Assyrian, it must have started after the Proto-Akkadian period.” 75 The y-Jussive is attested in all known varieties of the first millennium B. C. Western Aramaic (cf. in particular Sefire and IA). This is part of the reason to consider TF a manifestation of Eastern Aramaic. 76 Cf. Huehnergard 1983 and Rubin 2007, with references to previous studies. 77 The Semitic parallels (if not actual causes) for this shift will be discussed in due course in this study.

S. Loesov, Reconstructing Proto-Aramaic …

449

At this point, the etymology of l- has to be considered. I will divide the problem into two questions: (1) whence did PEA get the l-prefix? (2) Why is it limited to 3ms, 3mp, and 3fp? I will start with question (2), the one about reasons of paradigm-internal constraints on the l-form. Aramaic had probably never had analytical lVJussives for the finite forms with t- and n-prefixes along with synthetic formations for 3ms etc. This means that Aramaic proved unable to obviate the phonological constraint of the word-initial “real” consonant and to form something like lV-taqtul ‘she shall kill/let her kill!’ This sounds strange, especially in view of OA or Classical Arabic with their full-fledged Jussive paradigms, analytical as the case may be (cf. the OA 3fs lū taprus, 1cp lū niprus). Did PEA have *lVqtul for the 1cs, like Akkadian and unlike Tigré? In this case, a phonological obstacle is not very likely: since lV + yaqtul contracted to lVqtul at an early stage in Aramaic, lV + !aqtul could probably have done the same. My answer is we have no data to judge about this, since we have no semantically volitive 1cs forms in TF. Yet it is probable that PEA, like Tigré, did not build the 1cs l-Jussive to avoid the homonymy with the 3ms form (in the spirit of Rubin 2007 and Bulakh 2011). As for question (1), the one about the origin of l- in PEA, we know that the volitive l(V)- was ubiquitous in the linguistic milieu from which PEA emerged. In Samalian, I believe that the l-Jussive was giving way to a “normal” form of the Imperfect (in this case, the latter’s inflectional morphology was similar to that of the Central Semitic Jussive, i. e. *yaqtulū rather than *yaqtulūna): the Imperfect was thereby acquiring one more reading.78 This is a progressive feature of Samalian: KAI 214 was probably written shortly before 750 B. C. (Tropper 1993:55), while WA seems to have lost the morphological opposition “indicative vs. Jussive” in the prefixing conjugation no less than a quarter of a millennium later, during the Achaemenid period, i. e. after 500 B. C. This follows from the fact that this opposition became part of the Imperial Aramaic norm, yet it 78

Other things being equal, I do not believe that texts written in ancient literary languages (and a fortiori a royal votive inscription like KAI 214) can easily reflect changes under way in the spoken idioms of their authors. Yet this possibility is not 100% excluded, if what is at stake is verbal morphology rather than orthography/phonology. E. g., in Peshitta, a standard corpus of classical Syriac, both the Imperfect and the predicative participle can render the “indicative” future: the former option is the PA inheritance, the latter one adumbrates the Neo-Aramaic, i. e. the contemporary stage of Syriac speech. If there is a distribution, the choice will depend on syntactic and semantic nuances that have never been studied so far.

450

Articles: Semitic Studies

“was not consistently observed” in the Egyptian corpus (Muraoka–Porten 2003:200).79 Note also lw yqµ mn µyl krm znn ‘let him take from the best (produce) of this vine(yard)’ in Pardee 2009:53, l. 8f. This example is unique in our “Aramoid” corpus and without parallel in Aramaic. The nominal morphology of this inscription from Zincirli is different from the Samalian one (see the discussion in Pardee 2009). In view of the rest of the data, a natural solution is to admit that in this previously unknown NWS language, the precative particle lū had not (yet) undergone crasis with the prefixing 3ms form.80 Finally, there is the TF inscription with its l-Jussives and no y-forms in the injunctive slots of the text.81 And add l-forms of hwy in BA and epigraphic Aramaic texts found (and probably written) in Palestine (Beyer 1984:560f., see the detailed literature overview in Rubin 2007). In addition, there are personal names written in cuneiform which contain verbal l-forms of specifically Aramaic roots, e. g. -la-a¶-zi, allegedly ‘let [the DN] see!’ (Beyer 1984:109f., 148). The main reason to think through these data is the presence of lforms in more voluminous and “regular” Eastern Aramaic corpora, the most ancient of them being Hatran (44 B. C.—238 A. D.).82 With this in 79

In other words, l-Jussive was the only morphological Jussive in Samalian for 3ms, 3mp and 3fp forms. This Jussive was an obsolescent paradigm giving way to the “normal” Samalian imperfect. The development is in itself trivial, it will occur in Aramaic (roughly, in the second half of the first millennium B. C.) and in various spoken Arabic varieties. The interesting thing is its early date. 80 This evidence contradicts an important claim of Huehnergard 1983, the one about the historical path of two PS modal particles, the hypothetical *law and the asseverative *la: Huehnergard suggests that prefixing l-forms of both Akkadian and Aramaic verb go back to *la rather than *law, though (as he admits) this *la is otherwise unknown in both languages (cf. p. 590, fn. 187 for Huehnergard’s plausible explanation of a rare l! in Nabatean and Hatran as a loan from Arabic). From the Assyriological perspective, Huehnergard’s claim about the etymology of the precative was rejected in particular by A. Rainey (1996 III 195f.). Most recently see Kouwenberg 2010:216, where Huehnergard’s view on the destiny of *la and *law in Akkadian is evaluated as “unnecessary complicated.” D. Testen’s (1998) discussion of the problem is marred by his desire to establish a syllabic allophone of *l (*ḷ) in Proto-Semitic. 81 Huehnergard 1983:589 notes that in TF “no non-Jussive plural forms occur, and we are unable to state whether they would be *yaqtulū or *yaqtulūna.” This silence of TF is unfortunate but probably not critical since everybody agrees that its language is a variety of genuine Aramaic. 82 All 3ms Imperfect forms in this corpus have l-, 3fs forms have t-, while other forms of the Imperfect are not attested (Beyer 1998:137f.).

S. Loesov, Reconstructing Proto-Aramaic …

451

mind, I will attempt once more83 to make sense of the above evidence, though it is perhaps impossible to have all the pieces of the puzzle fall together in a gratifying way and without doing violence to these pieces.84 The PS optative particle *law (less likely, its contracted variant *lū)85 was doubtless present in PA, but its only productive continuation in the grammatically describable Aramaic corpora (i. e., from 500 B. C. on) obtains within the frequent particle of counterfactual condition illū/ellū ‘would that’ < hin ‘if ’ + lū, the compound is structurally identical to the Russian ‘если бы’ (Beyer 1984:564;86 SL 48a).87 In IA, its etymology was probably still transparent to the speakers of Aramaic, since it was usually written hn lw without assimilation nl > ll and with a space, and had a variant hnw lw (< *hin-hū lū?), see DNWSI 1:567 and Huehnergard 1983: 571f. Thus, our recorded Aramaic has not preserved a single token of a clear-cut proclitic lū with whatever independent force of its own. In Pardee 2009:53, l. 8, lw yqµ ‘let him take’ is written with no word-divider,88 yet Pardee 2009:57 is most probably right that lw provides an example of a word-final mater lectionis, since internal matres are not attested in this text. (Or is it *law?) Be it as it may, the lw of Pardee 2009:53 is a reflex of the very same PS counterfactual/optative particle *law that in this text is surprised “on the verge of grammaticalization,”—the grammaticalization whose results we witness in OA and OB, in Samalian (KAI 214), in TF and in the whole of Middle Eastern Aramaic. Thus Samalian and PEA felt a need to strengthen their Jussives with the *law particle, and each of them did so on its own, by way of common drift. Why did this happen? The TF—KAI 214 l-isogloss makes the influence of the Akkadian precative on PEA unlikely, because l-Jussive in KAI 83 Cf. Huehnergard 1983, Testen 1998, and Rubin 2007, all of whom ask questions different from mine. 84 The reason may well be that something is missing from our kit. 85 R. Steiner in a witty short note (1987) showed that in Aramaic aw > ū after l-, therefore we do not have to posit for PS two allomorphs of this particle, *law and *lū, pace Huehnergard. 86 Beyer 1984:617 believes that this *lū is a presentative ‘siehe,’ but this is hardly correct (cf. Huehnergard 1983:571, fn. 23). Beyer 2004:386 notes that *lū is not a contractrion of *law, ‘das syr. erhalten wäre!’, but Steiner 1987 probably takes care of this objection as well. 87 Lw is attested in both relatively long Samalian texts (KAI 214:13, 31; KAI 215:11bis), its functions there are not clear, as Tropper 1993:74 notes. J. Tropper tentatively renders lw as ‘wahrlich’ (p. 74), ‘gewiß’ (p. 118). 88 In this text, the word-divider was used 52 times on 13 lines that contain 202 letters (Pardee 2009:53).

452

Articles: Semitic Studies

214 cannot be explained as Akkadian influence anyway. The motivation I am going to propose is less transparent than the above suggestion about the raison d’être of the Akkadian precative, and it has to do with the subsequent history of the prefixing conjugation in Aramaic rather than with its prehistory. The inherited PS Jussive had a shorter life in Aramaic than in Arabic, in terms of absolute chronology. As for Aramaic, EA has been always developing more rapidly than WA. It is almost certain that the reflexes of 3ms forms *yaqtulu and yaqtul merged early in Aramaic (see p. 445 above). The Imperfect *yaqtulūna lost the PRESENT ACTUAL reading in favour of the conjugated *qātil and started acquiring modal usages early in the historical life of Aramaic, as we will see in the course of this study. My speculative suggestion is that both PEA and Samalian once attempted to save what could be saved of the Jussive and added a reflex of *law to the apocopate forms with vocalic anlaut as a new Jussive prefix.89 Yet they were unable to counter the historical progress, the Jussive was doomed to die out. As I proposed above, KAI 214 provides a chain missing in Aramaic, i. e. a transition period where *yaqtulū (< *yaqtulūna) is used in all the canonical slots of the Imperfect (e. g., future indicative and habitual situations in the past) and as a volitive form, while *laqtulū (< law + yaqtulū) is still used as a volitive. To conclude this section: the l-form is a shared innovation of Eastern Aramaic that goes back to PEA and antedates the emergence of IA, the main written language of the Achaemenid Empire. The next part of this study is supposed to see the completion of the Proto-EA reconstruction. References Abou-Assaf et al. 1982 Andersen 1970 Arnold 1989 Arnold 1990 Arnold 1999

89

Abou-Assaf, A.; Bordreuil, P.; Millard, A. La statue de Tell Fekherye et son inscription bilingue assyro-araméenne. Paris. Andersen, F. The Hebrew Verbless Clause in the Pentateuch. Nashville. Arnold, W. Lehrbuch des Neuwestaramäischen. Wiesbaden. Arnold, W. Das Neuwestaramäische. V. Grammatik. Wiesbaden. Arnold, W. Das Verbum im Neuwestaramäischen. Nebes, N. (ed.). Tempus und Aspekt in den semitischen Sprachen. Wiesbaden. Pp. 1–8.

This means that analytical forms of the lw yqµ type had been the immediate predecessor of the stage witnessed by TF.

S. Loesov, Reconstructing Proto-Aramaic … Bar-Asher Siegal 2011

Bauer–Leander 1927 Bergsträsser 1983 Beyer 1984

Beyer 1995

Beyer 1998

Beyer 2004

Bulakh 2011

Cohen 1979 Cohen 1984 Correll 1978 Dalman 1905 Degen 1969 Estival–Myhill 1988

Fischer 1987 Fitzmyer 1995 Garr 1985 Goldenberg 1998 Hetzron 1969

453

Bar-Asher Siegal, E. A. On the Passiveness of One Pattern in Jewish Babylonian Aramaic—A Linguistic and Philological Discussion. JSS 56:111–143. Bauer, H.; Leander, P. Grammatik des Biblisch-Aramäischen. Halle. Bergsträsser, G. Introduction to the Semitic Languages. Text Specimens and Grammatical Sketches. Winona Lake. Beyer, K. Die aramäischen Texte vom Toten Meer samt den Inschriften aus Palästina, dem Testament Levis aus der Kairoer Genisa, der Fastenrolle und den alten talmudischen Zitaten. Göttingen. Beyer, K. Die Aussprache des christlich-palästinischen Aramäisch. Zur neuen Grammatik von Christa MüllerKessler. JSS 40:241–257. Beyer, K. Die aramäischen Inschriften aus Assur, Hatra und dem übrigen Ostmesopotamien (datiert 44 v. Chr. bis 238 n. Chr.). Göttingen. Beyer, K. Die aramäischen Texte vom Toten Meer samt den Inschriften aus Palästina, dem Testament Levis aus der Kairoer Genisa, der Fastenrolle und den alten talmudischen Zitaten. Bd. II. Göttingen. Bulakh, M. The Origin of the Personal Marker *lV- in the Prefixing Conjugation of the Tigré Language. Paper read at the XXV International Conference “Oriental and African Studies in Dialogue of Civilizations,” SaintPetersburg University, April 22–24, 2011 (in Russian). Cohen, D. Sur le système verbal du néo-araméen de Ma"lûla. JSS 24:219–239. Cohen, D. La phrase nominale et l’évolution du systéme verbal en sémitique. Etudes de syntaxe historique. Leuven–Paris. Correll, Chr. Untersuchungen zur Syntax der neuwestaramäischen Dialekte der Antilibanon. Wiesbaden. Dalman, G. Grammatik des jüdisch-palästinischen Aramäisch. Leipzig. Degen, R. Altaramäische Grammatik der Inschriften des 10– 8. Jh. v. Chr. Wiesbaden. Estival, D.; Myhill, J. Formal and Functional Aspects of the Development from Passive to Ergative Systems. Shibatani, M. (ed.). Passive and Voice. Amsterdam. Pp. 441–492. Fischer, W. Grammatik des klassischen Arabisch. Wiesbaden. Fitzmyer, J. The Aramaic Inscriptions of Sefire. Roma. Garr, R. Dialect Geography of Syria-Palestine, 1000–586 B. C. E. Philadelphia. Goldenberg, G. Aramaic Perfects. Goldenberg, G. Studies in Semitic Linguistics. Selected Writings. Jerusalem. Pp. 605–629. Hetzron, R. The Evidence for Perfect *yáqtul and Jussive *yaqtúl in Proto-Semitic. JSS 14:1–21.

454 Hopkins 1989 Huehnergard 1983 Huehnergard 1986 Huehnergard 1987 Huehnergard 1995 Jastrow 1993 Jastrow 2002 Jastrow 2005

Jastrow 2005a Jastrow 2008

Khan 2007

Khan 2007a

Khan 2008

Khan 2008a

Kogan 2005 Kouwenberg 1997 Kouwenberg 2010 Kutscher 1969

Articles: Semitic Studies Hopkins, S. Neo-Aramaic Dialects and the Formation of the Preterite. JSS 34:413–432. Huehnergard, J. Asseverative *la and Hypothetical *lu/law in Semitic. JAOS 103:569–593. Huehnergard, J. On Verbless Clauses in Akkadian. ZA 96:218–249. Huehnergard, J. The Feminine Plural Jussive in Old Aramaic. ZDMG 137:266–277. Huehnergard, J. What is Aramaic? Aram 7:261–282. Jastrow, O. Laut- und Formenlehre des neuaramäischen Dialekts von Mīdin im àūr ‛Abdīn. Wiesbaden. Jastrow, O. Lehrbuch der àuroyo-Sprache. Wiesbaden. Jastrow, O. Der bestimmte Artikel im Aramäischen—ein Blick auf 3000 Jahre Sprachgeschichte. Tropper, J.; Younansardaroud, H. (eds.). Studia Semitica et Semitohamitica. Festschrift für Rainer Voigt. Winona Lake. Pp. 137–150. Jastrow, O. Glossar zu Kinderib (Anatolisches Arabisch). Wiesbaden. Jastrow, O. Old Aramaic and Neo-Aramaic: Some Reflections on Language History. Gzella, H.; Folmer, M. (eds.). Aramaic in Its Historical and Linguistic Setting. Wiesbaden. Pp. 1–10. Khan, G. Ergativity in the North Eastern Neo-Aramaic Dialects. Bar, T.; Cohen, E. (eds.). Studies in Semitic and General Linguistics in Honor of Gideon Goldenberg. Münster. Pp. 147–157. Khan, G. The Morphology of Babylonian Jewish Aramaic. Kaye, A. (ed.). Morphologies of Asia and Africa. Winona Lake. Pp. 107–119. Khan, G. The Expression of Definiteness in North-Eastern Neo-Aramaic Dialects. Gzella, H.; Folmer, M. L. (eds.). Aramaic in Its Historical and Linguistic Setting. Wiesbaden. Pp. 287–304. Khan, G. Remarks on the Function of the Preterite and the Perfect in Northeastern Neo-Aramaic. Khan, G. (ed.). Neo-Aramaic Dialect Studies. Piscataway. Pp. 105–130. Kogan, L. Lexicon of the Old Aramaic Inscriptions and the Historical Unity of Aramaic. B&B 2:513–565. Kouwenberg, N. J. C. Gemination in the Akkadian Verb. Assen. Kouwenberg, N. J. C. The Akkadian Verb and Its Semitic Background. Winona Lake. Kutscher, E. Y. Two “Passive” Constructions in Aramaic in the Light of Persian. Proceedings of the International Conference on Semitic Studies, Jerusalem, 19–23 July 1965. Jerusalem. Pp. 132–151.

S. Loesov, Reconstructing Proto-Aramaic … Lipiński 2008

Loesov 2011 Macuch 1982 Müller-Kessler 1991

Muraoka–Porten 2003 Nöldeke 1966 Nöldeke 2005 Odeberg 1939 Pardee 2009 Pennacchietti 1988

van Peursen 2008

Rainey 1996

Raz 1983 Ritter 1967 Ritter 1969 Ritter 1979 Ritter 1990 Rosenthal 1978 Rubin 2007 Sapir 1921

455

Lipiński, E. Aramaic Broken Plurals in the Wider Semitic Contexts. Gzella, H.; Folmer, M. L. (eds.). Aramaic in Its Historical and Linguistic Setting. Wiesbaden. Pp. 27–40. Loesov, S. The Suffixing Conjugation of Akkadian: In Search of Its Meaning. B&B 6:75–148. Macuch, R. Grammatik des Samaritanischen Aramäisch. Berlin–New York. Müller-Kessler, Chr. Grammatik des Christlich-PalästinischAramäischen. Schriftlehre, Lautlehre, Formenlehre. Hildesheim. Muraoka, T.; Porten, B. A Grammar of Egyptian Aramaic. Leiden etc. Nöldeke, T. Kurzgefasste syrische Grammatik. Darmstadt (1st edition 1880; 2nd edition 1898). Nöldeke, T. Mandäische Grammatik. Eugene (1st edition 1875). Odeberg, H. The Aramaic Portions of Bereshit Rabba, with Grammar of Galilean Aramaic. Lund. Pardee, D. A New Aramaic Inscription from Zincirli. BASOR 356:51–71. Pennacchietti, F. Verbo neo-aramaico e verbo neo-iranico. Orioles, V. (ed.). Tipologie della Convergenza linguistica: Atti del Convergno della Società Italiana di Glottologia. Pisa. Pp. 93–110. van Peursen, W. Language Variation, Language Development, and the Textual History of the Peshitta. Gzella, H.; Folmer, M. L. (eds.). Aramaic in Its Historical and Linguistic Setting. Wiesbaden. Pp. 231–256. Rainey, A. Canaanite in the Amarna Tablets: A Linguistic Analysis of the Mixed Dialect Used by Scribes from Canaan. Leiden–New York. Raz, Sh. Tigre Grammar and Texts. Malibu. Ritter, H. àūrōyo. Die Volkssprache der syrischen Christen des àūr ‫ع‬Abdîn. A: Texte. Bd. 1. Beirut. Ritter, H. àūrōyo. Die Volkssprache der syrischen Christen des àūr ‫ع‬Abdîn. A: Texte. Bd.2. Beirut. Ritter, H. àūrōyo. Die Volkssprache der syrischen Christen des àūr ‫ع‬Abdîn. B: Wörterbuch. Beirut. Ritter, H. àūrōyo. Die Volkssprache der syrischen Christen des àūr ‫ع‬Abdîn. C: Grammatik. Stuttgart. Rosenthal, F. Aramaic Studies during the Past Thirty Years. JNES 37:81–91. Rubin, A. On the Third Person Preformative l-/n- in Aramaic, and an Ethiopic Parallel. ANES 44:1–28. Sapir, E. Language. An Introduction to the Study of Speech. New York.

456 Simeone-Senelle 1997

Spitaler 1938 Steiner 1987

Testen 1998 Tezel 2003

Tropper 1993

Voigt 2007 Voigt 2009 Wright 1862

Articles: Semitic Studies Simeone-Senelle, M.-Cl. The Modern South Arabian Languages. Hetzron, R. (ed.). The Semitic Languages. London. Pp. 378–423. Spitaler, A. Grammatik des neuaramäischen Dialekt von Ma"lūla. Leipzig. Steiner, R. Lulav versus *lu/law: A Note on the Conditioning of *aw > ū in Hebrew and Aramaic. JAOS 107: 121–122. Testen, D. Parallels in Semitic Linguistics. The Development of Arabic la- and Related Semitic Particles. Leiden. Tezel, A. Comparative Etymological Studies in the Western Neo-Syriac (àūrōyo) Lexicon: with Special Reference to Homonyms, Related Words and Borrowings with Cultural Signification. Uppsala. Tropper, J. Die Inschriften von Zincirli. Neue Edition und vergleichende Grammatik des phönizischen, sam’alischen und aramäischen Textkorpus. Münster. Voigt, R. Mandaic. Kaye, A. (ed.). Morphologies of Asia and Africa. Vol. 1. Winona Lake. Pp. 149–166. Voigt, R. Das Präsens im Tigre. Aethiopica 12:155–163. Wright, W. A Grammar of the Arabic Language. Volume II. Cambridge.

Ein weiteres arabisches Syntagma in der altsüdarabischen Epigraphik* Peter Stein Universität Jena

Amiritisch — ein arabisch beeinflußter Dialekt des Sabäischen? Mit ‘Amiritisch’ bezeichnen wir einen in altsüdarabischer Schrift geschriebenen Dialekt, der in den Jahrhunderten um die Zeitenwende im Gebiet des nördlichen Jemen, zwischen den Oasenstädten des Wadi al-Ǧawf im Süden und Nagrān im Norden, gesprochen worden ist (s. zuletzt Stein 2007). Namensgebende Träger dieses Idioms sind die Angehörigen des Stammes der !Amīr, der in den sabäischen Inschriften zum einen durch nordarabische Personennamen, zum anderen durch seine starke Affinität zur Kamelzucht hervortritt. Wie die Sabäer oder gar die Stammesleute selbst ihre Sprache bezeichnet haben, ist ebenso ungeklärt wie die Frage, ob noch andere Stämme außerhalb der !Amīr sich dieses Idioms bedienten. Die Bezeichnung ‘Amiritisch’ ist also konventionell, trägt aber der historischen Situation stärker Rechnung als die früher übliche Benennung ‘Haramisch’ nach einem einzelnen Fundort im Wadi al-Ǧawf. Auch wenn epigraphische Nachweise für den Stamm der !Amīr bis in die altsabäische Zeit zurückreichen, tritt der Dialekt selbst erst relativ spät in Erscheinung. Das früheste Beispiel eines amiritischen Textes ist ein auf Holzstäbchen geschriebener Brief aus der Sammlung der Bayerischen Staatsbibliothek in München, dessen Paläographie in die Mitte des ersten vorchristlichen Jahrtausends verweist (Mon.script.sab. 240, verfaßt

* Der Aufsatz basiert auf einem Vortrag unter dem Titel „Neues zum amiritischen Dialekt“, den der Autor auf dem 4. Arbeitstreffen der Arbeitsgemeinschaft Semitistik in der Deutschen Morgenländischen Gesellschaft am 2. Oktober 2009 an der Universität Leipzig gehalten hat. — Inschriften mit dem Siglum X.BSB sind jetzt bei Stein (2010) publiziert, die ebenfalls zur Sammlung der Bayerischen Staatsbibliothek in München gehörenden Texte mit dem Siglum Mon.script.sab. sind noch unveröffentlicht. Die beiden unter dem Kürzel J zitierten sabäischen Inschriften wurden von Jamme (1962) publiziert.

458

Articles: Semitic Studies

im Duktus Ry IId).1 Amiritische Monumentalinschriften, also in erster Linie in Stein und Bronze verfaßte Widmungen sowie die sogenannten Buß- und Sühneinschriften, sind dagegen erst aus dem 2. Jh. v. Chr. bekannt. Gleichermaßen ungewiß ist das Ende der schriftlichen Überlieferung. Die ausschließlich paläographisch datierbaren Schriftzeugnisse scheinen im ersten nachchristlichen Jahrhundert zu versiegen (vgl. Robin 1992:29), doch mag dieser Befund angesichts der verhältnismäßig geringen Anzahl überhaupt vorhandener Quellen Zufall sein. Es sind nämlich erst zwei Dutzend Texte in amiritischer Sprache bekannt. Dabei stehen knapp 20 Monumentalinschriften2 mindestens sechs Minuskelinschriften, vornehmlich Briefe, gegenüber.3 Es sei dabei nachdrücklich auf den Umstand hingewiesen, daß wenigstens einer der auf Holzstäbchen verfaßten Texte (X.BSB 167) als Schulübung klassifiziert werden kann. Das bedeutet, daß der Dialekt der !Amīr im Nordjemen der Zeitenwende nicht etwa das marginale Schattendasein einer gesellschaftlichen Minderheit fristete, sondern neben dem Sabäischen und dem Minäischen als Verkehrssprache im alltäglichen Geschäftsleben etabliert war. Was zeichnet nun diesen Dialekt vor den anderen genannten, benachbarten Sprachen aus? Es sind im wesentlichen zwei Merkmale, welche die sprachgeschichtliche Verortung des Amiritischen bestimmen: die morphologischen Übereinstimmungen in der Bildung von Personalpronomina und Kausativstamm mit dem Sabäischen (sogenannte ‘H’-Sprachen) sowie der postpositive bestimmte Artikel (-n) auf der einen und eine Reihe nordarabischer Charakteristika auf der anderen Seite. Dazu gehört zu allererst die Bildung der Suffixkonjugation bei erster und zweiter Person. Während alle altsüdarabischen Sprachen, soweit erkennbar, das Personenzeichen mit k bilden (also f"lk, f"lkmw, wie im Äthiopischen), zeigt das Amiritische die aus dem Arabischen (und Nordwestsemitischen) ge1 Weitere frühe Minuskelinschriften sind X.BSB 95 und 167 (paläographische Stufe Ry IIIa) sowie die sprachlich allerdings nicht sicher klassifizierte Mon.script. sab. 663 (Ry IId). Der auffallend häufig bei amiritischen Texten zu beobachtende Duktus Ry IIIb (X.BSB 96 und 97 sowie Mon.script.sab. 672) ist in seiner chronologischen Einordnung bislang nicht gesichert (s. Stein 2010:45–47). 2 Zu den von Stein (2007:32–45) zusammengestellten 18 Nummern (einschließlich zweier unsicherer Zuordnung) kann inzwischen noch die von a´-Ôilwī (2005) veröffentlichte Buß- und Sühneinschrift auf einer Bronzetafel hinzugefügt werden. 3 X.BSB 95–97 und 167 sowie Mon.script.sab. 240 und 672. Unsicher ist die Zuordnung von X.BSB 67, 105 und 188 sowie Mon.script.sab. 165 und 663.

P. Stein, Ein weiteres arabisches Syntagma …

459

läufige Bildungsweise mit t (f"lt, f"ltm). Die übrigen Isoglossen mit nordarabischen Charakteristika betreffen neben einigen lexikalischen Gemeinsamkeiten die reduzierte Reihe der nicht-emphatischen stimmlosen Sibilanten (s und š, gegenüber s, š und ś im sonstigen Altsüdarabischen) sowie auf syntaktischer Ebene die besonders augenfällige Verneinung in der Vergangenheit mit der Negation lm und folgender Präfixkonjugation (vgl. die folgende Tabelle).

Sibilanten Unbestimmter Artikel Bestimmter Artikel Personalpronomina Kausativstamm 2. Person der SK f- zwischen Neg. der Vorzeitigkeit Präposition “von” Konjunktion “als” Konditionalpartikel Satzeinleitungspartikel

Amiritisch Nordarabisch

Minäisch Sabäisch Phonologie s, š, ś

s, š

Morphologie — -m

—

postpositiv (Nunation) -sw, -sm etc. -hw, -hmw etc. sf"l

hf"l f"lk

f"lt

Syntax Hauptsätzen Satzgliedern lhm + SK !l + SK Lexik bn (b-)ywm hn

hn ¯t

— / -un präpositiv -hu, -hum etc. !af"ala fa"alta

Hauptsätzen lm + PKK lam + PKK mn !¯

min !i¯

hn

!in !an

Die Definition des Amiritischen als sabäischer Dialekt mit nordarabischem Einschlag ist sicherlich artifiziell. Die Frage nach der Genese dieses Dialektes, ob es sich eher um ein sabäisiertes Altarabisch oder um ein arabisiertes Sabäisch handelt, wird sich angesichts der spärlichen Überlieferungssituation auf absehbare Zeit wohl nicht lösen lassen. Unbestreitbar ist jedoch die Funktion des Amiritischen als lebendige, gesprochene Sprache. Früher geäußerte Annahmen, es handele sich bei diesem eklektizistischen Sprachgebilde um den Versuch von Angehörigen einer fremden, nämlich nordarabischen, Muttersprache, korrektes Sabäisch zu schreiben, müssen spätestens vor dem Hintergrund einer lebhaften Briefkorrespondenz revidiert werden. Die konsequente Scheidung eines Repertoires an nordarabischen Spezifika von sabäischen Merkmalen, wie sie in sämtlichen Texten zu beobachten ist, würde von einer reinen Kunstsprache wohl nicht erreicht.

460

Articles: Semitic Studies

Ein neuer Brief im amiritischen Dialekt Das sprachliche Erscheinungsbild eines in diesem Dialekt geschriebenen Textes soll beispielhaft an dem folgenden, noch unveröffentlichten Brief aus dem Bestand der Bayerischen Staatsbibliothek in München demonstriert werden. Der achtzeilige Text ist auf die Vorder- und Rückseite einer reichlich 12 cm langen Palmblattrippe geschrieben. Die Buchstabenformen folgen einem Duktus (Ry IIIb), der für eine Reihe amiritischer Briefe als charakteristisch gelten kann, dessen chronologische Einordnung in die paläographische Entwicklung der altsüdarabischen Minuskelschrift aber noch nicht gelungen ist (vgl. oben mit Fn. 1). Die Entstehungszeit des Briefes kann also nur grob in den historischen Rahmen der Präsenz amiritischer Schreibkunst, mithin in die Spanne zwischen dem 5. Jh. v. Chr. und dem ersten nachchristlichen Jahrhundert datiert werden. Trotz leichter Beschädigung des äußerst fragilen Holzes sind die fünf Textzeilen der Vorderseite nahezu vollständig lesbar. Lediglich auf der Rückseite wird die Lesung des Textes durch Oberflächenabrieb stärker beeinträchtigt, doch ist die nur lückenhafte Rekonstruktion des letzten Drittels des Briefes für unser Ansinnen ohne Belang (beschädigte, aber sicher identifizierbare Passagen sind in der folgenden Transkription nicht eigens markiert, unleserliche Zeichen durch x wiedergegeben). Zudem zeichnet sich der Text durch einige Schreibfehler aus; konkret ist die Auslassung mehrerer Buchstaben zu beklagen, was in der Transkription durch spitze Klammern angedeutet ist. Mon.script.sab. 672 (vgl. Abb. 1) Vs.: 1. l"symm / w¶n"m / "mn / "?l?lm / w¯smwy / lkrb 2. nkmy / wlhw / ltµywnn / wftµ / !lb〈b〉kmy / k!l 3. s¢rtmw / f!w / hµd¬tmwhw / µ〈d〉¬kmw / w 4. !yhmw / fnf¥t / ¯t / n¯rtn / fl / tµmdnn / b 5. 〈h〉n / ltst"wr / wlh / tµmdn / b?r?b? Rs.: 6. µn / lhw / tµn?xn? / ? b?f? / !?x!?s? / !?nsm / xxx[xxx] 7. wb / µq?bn / l?ff?n / ltµmdn / w!l? / t?xy?n / x[xx]x 8. w?¯tš?!n / s¢r [(xxxxxxxx) / ]wlkmw / n"mtm Übersetzung 1. An "SYMM und ÚN"M von "LLM(?). D̠-SMWY möge euch 2. segnen. Seid von ihm gegrüßt. Er hat sich euren Herzen gegen.über beklagt, daß ihr weder 3. geschrieben noch ihm eure Neuigkeiten mitgeteilt habt.

P. Stein, Ein weiteres arabisches Syntagma …

461

4. (Was) sie (sc. den Absender und seine Angehörigen) (betrifft), so .hat dieses Zeichen sie in Unruhe versetzt. Nun habt Dank dafür, 5. daß du ein Darlehen erbitten(?) willst. Und sei von ihm bedankt .für die Erhebung 6. von Zinsen(?) zu seinen Gunsten … Menschen […]. 7. Für die Aufbewahrung der Urkunde hab Dank. Und … nicht! 8. Was du wünschst, das schreibe! [(…)] Euch sei Glück (beschie.den).

Zusammengefaßt stellt sich der Inhalt des Schreibens wie folgt dar: der Empfänger beabsichtigt, ein Darlehen aufzunehmen, wofür der Absender des Briefes ihm dankt. Dieser Dank kann nun einerseits dahingehend verstanden werden, daß der Empfänger ein solches Geschäft im Auftrag des Absenders erledigt, oder aber, daß der Gewährer des Darlehens mit dem Absender unseres Briefes selbst identisch ist. Letztere Annahme dürfte die wahrscheinlichere sein, zumal als Nutznießer der in Aussicht stehenden Zinszahlung (sofern unsere diesbezügliche Rekonstruktion von Z. 5f. korrekt ist) allein ebendieser Absender in Frage kommt. Neben diesem geschäftlichen Inhalt sind aber auch die einleitenden Gruß- und Höflichkeitsformeln an diesem Brief von Interesse. So wird aus der an die Empfänger gerichteten Beschwerde in Z. 2–4 ersichtlich, daß sich der Absender und seine Angehörigen ob des Ausbleibens einer Nachricht bereits Sorgen gemacht haben. Doch ganz abgesehen von seinen inhaltlichen Aspekten: was macht nun diesen Brief zu einem amiritischen Text? Da ist zunächst die Klage des Absenders über ausbleibende Post in Z. 2–3: die Verbformen s¢rtmw ‘ihr habt geschrieben’ und hµd¬tmw-hw ‘ihr habt ihm Neuigkeiten berichtet’ zeigen einmal mehr die für diesen Dialekt charakteristische Bildung der 2. Person der Suffixkonjugation mit -t. In einem sabäischen Brief müßten die entsprechenden Formen s¢rkmw bzw. hµd¬km(w)-hw lauten, und auch im Minäischen würden diese Formen mit dem Element -k gebildet. Gut sabäisch an unserem Text sind allerdings die Formen mit h, also der Kausativstamm hµd¬ (minäisch sµd¬) und das suffigierte Personalpronomen -hw (minäisch -s). Ein weiteres Merkmal auf syntaktischer Ebene, die Aneinanderreihung von Hauptsätzen mit der Partikel f-, ist schon weit weniger spezifisch. Zwar sollte ein syntaktischer Neueinsatz wie f-l tµmdnn am Ende von Z. 4 in einem sabäischen Text so nicht zu finden sein, doch ist die Koordination mit f- von Anbeginn im Minäischen produktiv (und deshalb nicht zwingend als Nordarabismus zu werten).

462

Articles: Semitic Studies

Was wieder als typisch “arabisch” angesehen werden kann, ist eine lexikalische Besonderheit im vorletzten Satz des Briefes in Z. 8. Während die durchaus formelhafte Aufforderung zur Fortführung des Briefwechsels mit dem Satz “Was du wünschst, das schreibe!” in sabäischer Korrespondenz mit einer Verbform der Wurzeln RÓW bzw. "KR gebildet wird, steht in amiritischen Briefen (neben dem vorliegenden noch in X.BSB 97:10) eine Form, die zweifelsfrei mit dem arabischen Verbum šā!a ‘wünschen, wollen’ in Beziehung zu setzen ist: w-¯-tš!n s¢r ‘Was du wünschst, das schreibe!’. Morphologisch ist die Form allerdings gut sabäisch: Das auslautende -n entspricht der Endung der Langform der sabäischen Präfixkonjugation im Singular, während in der Mehrzahl eine aus zwei n bestehende Endung angefügt wird, wie in den Formen l-thµywnn ‘seid gegrüßt’ in Z. 2 und l-tµmdnn ‘habt Dank’ am Ende von Z. 4 auch zu sehen. Sabäische (also altsüdarabische) und nordarabische Merkmale gehen in dem Text Hand in Hand. Der älteste Beleg für die arabische Objektpartikel !iyyāAn einer Stelle unseres Briefes begegnet eine syntaktische Konstruktion, die wir aus Inschriften aus Altsüdarabien bislang nicht kennen. Der mit der Konjunktion w- ganz am Ende von Z. 3 beginnende und mit dem Subjekt ¯t n¯rtn in der Mitte der folgenden Zeile endende Satz w-!y-hmw f-nf¥t ¯t n¯rtn ist zunächst nach gut sabäischem Muster in ein vorangestelltes Satzglied und den mit f- eingeleiteten Verbalsatz zu zerlegen. Diese Konstruktion VORANGESTELLTER SATZTEIL-f-VERBFORM ist, wie Norbert Nebes (1995) überzeugend dargelegt hat, ein spezifisches Charakteristikum des Sabäischen, welches sich in dieser Produktivität in keiner anderen semitischen Sprache findet. Als vorangestelltes Satzglied können sowohl Subjekt oder Objekt des Satzes als auch Adverbialbestimmungen dienen; ihre Voranstellung bewirkt dabei eine besondere Fokussierung, eine Akzentuierung im Erzählablauf. Mehr noch als in den von Nebes untersuchten Monumentalinschriften kommt dieses syntaktische Mittel der Gliederung innerhalb sabäischer Briefe zum tragen. Ein besonders markantes Beispiel bietet der folgende, von Stefan Weninger (2002:217– 220) veröffentlichte Brief, in welchem ausnahmslos alle(!) verbalen Hauptsätze das Subjekt, Objekt oder einen Präpositionalausdruck voranstellen (der Text ist hier nach Hauptsätzen untergliedert, die mit den Buchstaben A bis H durchnumeriert sind, die originale Zeilendisposition wird durch die hochgestellten Ziffern angegeben).

P. Stein, Ein weiteres arabisches Syntagma …

463

Mon.script.sab. 68 l-šfnm "mn !rzn An ŠFNM von !RZN. A. w-"¬tr l-h´bµn l-k 2 n"mtm "T̠.TR lasse dir Glück leuchten. B. w-s¢r s¢rk b-"m rµbm f-m¥! 3 w-r¥y Das Schreiben, das du mit RÑBM geschickt (wörtl.: geschrieben) hast, ist angekommen und hat gefallen. C. w-h! f-!l y¶dgn b-h¢bn l-k kl 4 !µ´n-k Er wird nicht versäumen, all deine Angehörigen von dir zu informieren. D. w-h! f-r! k-hysrn bn ´n"w 5 ¶¢ bnt-k k-wkb-hmw rµbm k-y!tyn 6 bn ¥fr Er, siehe, er wird aus Ôan"āw das Schriftstück deiner Tochter schicken, sobald RÑBM sie angetroffen haben wird, wenn er aus Ýafār zurückkehrt. E. w-b wz! s¢r l-hmw l-tµmd 7 n Dafür, daß du fortfährst, ihnen zu schreiben, hab Dank. F. w-hn!m f-µywn l-hw Und HN!M — grüße ihn von ihm! G. w-"lhn f-µywn 8 l-hw Und "LHN — grüße ihn von ihm! H. w-l-k n"mtm Dir sei Glück beschieden.

Von insgesamt acht Sätzen (die diesbezüglich nicht relevante einleitende Adreßzeile klammern wir aus) ist in den ersten vier (A–D) das Subjekt, in den folgenden drei das Objekt (in Satz E in präpositionaler Konstruktion) dem Prädikat vorangestellt. Der letzte Satz (H) ist nominal, mit dem präpositionalen Prädikat in Voranstellung. Unter den sieben Verbalsätzen mit vorangestelltem Satzglied werden allein fünf durch die Partikel f- eingeleitet, die lediglich in A und E entfällt. Daß die Voranstellung der Subjekte bzw. Objekte der syntaktischen Strukturierung des Brieftextes dient, ist unübersehbar. Kehren wir nun zu unserem amiritischen Brief zurück, so stellen wir fest, daß in dem Satz w-!y-hmw f-nf¥t ¯t n¯rtn die Position des Subjekts aufgrund der eindeutigen Kongruenzverhältnisse durch das Nomen (¯t) n¯rtn besetzt ist. Der auf die Partikel f- folgende Satz für sich genommen kann also, in Anlehnung an arabisch nafa¥a ‘schütteln, erschauern lassen’ sowie na¯īra etc. ‘Ankündigung, Warnung’ übersetzt werden ‘dieses Zeichen (nämlich das Ausbleiben einer Nachricht der Angehörigen) hat zittern lassen’. Nun erhebt sich zwangsläufig die Frage nach dem Objekt

464

Articles: Semitic Studies

dieses Verbums — schon aus inhaltlichen Gründen liegt die Lösung praktisch auf der Hand. Das vorangestellte Satzglied !y-hmw ist in das auf den Absender des Briefes und seine Angehörigen bezügliche Pronominalsuffix -hmw und eine Partikel !y aufzulösen, welche in dieser Konstellation zwingend mit der arabischen Objektpartikel !iyyā zu verbinden ist, die ebenso der Kennzeichnung eines pronominalen Akkusativobjektes insbesondere bei dessen betonter Voranstellung dient: w-!y-hmw f-nf¥t ¯t n¯rtn ‘Sie (nämlich die Absender) hat dieses Zeichen (das Ausbleiben einer Nachricht) in Unruhe versetzt’. Es bleibt noch ein kleines Fragezeichen. Trotz der unzweifelhaften Isoglosse mit dem Klassischen Arabischen ist es keineswegs ausgemacht, daß die Objektpartikel !y- im Amiritischen auch tatsächlich in der nordarabischen Einflußnahme auf diesen Dialekt ihren Ursprung hat. Es könnte sich genauso gut um ein Merkmal handeln, welches das Amiritische mit dem Sabäischen teilt. Zwar haben wir oben konstatiert, daß die fragliche Konstruktion (und damit die Partikel !y-) in dem hier besprochenen Brief überhaupt zum ersten Mal in einem altsüdarabischen Text in Erscheinung tritt. Ihr Fehlen in den sabäischen Inschriften könnte aber schlicht dem Umstand geschuldet sein, daß entsprechende syntaktische Kontexte in dem bekannten Textmaterial einfach nicht angelegt sind. Wie wir gesehen haben, ist die Voranstellung des Objektes von Verbalsätzen im Sabäischen gang und gäbe, man vergleiche nur die Sätze F und G aus dem zitierten sabäischen Brief. Allerdings ist die absolute Anzahl epigraphischer Belege relativ gering. Von den beinahe 120 von Nebes (1995) gesammelten Belegen für einem Verbalsatz vorangestelltes Nomen aus sabäischen Monumentalinschriften sind lediglich 20 ein direktes Objekt (vgl. a. a. O. 32f. und 226f.), die überwiegende Mehrheit hingegen stellt das Subjekt des Satzes nach vorn. Daß unter den bislang publizierten Inschriften kein Beispiel für ein vorangestelltes pronominales Objekt zu finden ist, dürfte daher schlicht Zufall sein, zumal auch im Falle vorangestellten Subjekts ein reines Pronomen (wie in Satz C und D unseres Beispiels) außerhalb von Briefen eher selten begegnet. Läßt uns dieses Syntagma für eine Klärung unserer Frage also im Stich, sollten wir nach anderen Konstellationen suchen, in denen im Arabischen die Partikel !iyyā Verwendung findet. Dies sind etwa Fälle, in denen das Pronomen aus satztechnischen Gründen nicht direkt an die Verbform gehängt werden kann, wie in den folgenden, aus der einschlägigen Literatur willkürlich zusammengestellten Beispielen:

P. Stein, Ein weiteres arabisches Syntagma …

465

!anzalahum !iyyāhā ‘Er ließ sie es bewohnen’.4 la-yaq¢a"annī wa-!iyyāka ‘Er wird mich und dich durchstechen’.5 man tad"ūna !illā !iyyāhu ‘Wen wollt ihr anrufen außer ihm?’6

Und hier gibt es in der Tat zwei (und nur zwei) immerhin eindeutige Belege dafür, daß die Konstruktion des pronominalen Objektes in isolierter Form im Sabäischen nicht nach dem arabischen Beispiel, sondern mittels der Obliquus-Formen des Demonstrativpronomens der Ferndeixis erfolgt: s"d¬wn tly !frs mlkn hqny 2 !lmqh b"l!wm ´lmn ¯-¯hbn 3 µmdm b-¯t mt"-hw (b-¶)lf ¯mr hwt w-bn 4 -hw dd!l (J 584:1–4) S"DT̠.WN, der Verantwortliche für die Pferde des Königs, hat !LMQH, dem Herrn von !WM, die(se) Statuette aus Bronze gewidmet zum Dank dafür, daß der ihn gerettet hat in der Umgebung der (Stadt) D̠.MR,7 ihn und seinen Sohn DD!L. w-bn-hw f-wqh-hmw mr!-hmw šm 27 r yhr"š mlk sb! w-¯-rydn hwt w-sb"y wm 28 !t !sdm … … l-tqdm〈n〉 w-twś"n !š 30 "b "km w-¯-shrtm (J 649:26–30) Daraufhin befahl ihnen ihr Herr ŠMR YHR"Š, der König von SB! und D̠.-RYDN, (nämlich) ihm (sc. dem Stifter) und 170 Mann … …, vorzurücken und die Stämme von "KM und von SHRTM anzugreifen.

Der inhaltliche Kontext der beiden Belegstellen ist klar: das direkte Objekt der Verbform (mt" bzw. wqh) wird im Anschluß an ein weiteres Satzglied (eine Ortsbestimmung im ersten bzw. das Subjekt im zweiten Beispiel) wieder aufgegriffen, um das diesbezüglich nicht eindeutige suffigierte Personalpronomen weitergehend zu spezifizieren. Wäre im Sabäischen die Partikel !y- produktiv, stünde sie hier anstelle des selbständigen Demonstrativpronomens hwt zu erwarten (also *!y-hw w-bn-hw dd!l im ersten Beispiel). Da dies nicht der Fall ist, kann die Partikel in der Tat als

4

Reckendorf 1921:285. Brockelmann 1913:324. 6 Brockelmann 1913:324. 7 Jamme (1962:90f.) liest im Anschluß an die Verbform ¯lf / ¯mr, was inhaltlich völlig unklar bleibt (vgl. den Interpretationsversuch von Müller 1978:141). Es handelt sich hier mit Sicherheit um ein Versehen des Schreibers oder Kopisten (eine Fotografie der Inschrift existiert nicht) für die aus zahlreichen anderen Inschriften bekannte Ortsbestimmung b-¶lf (hgrn) ON, die jedenfalls in Gl 1358 = Gr 212:2f. (veröffentlicht von Solá Solé 1964:35f.) in ebensolchem Zusammenhang mit dem Verbum mt" ‘retten’ begegnet: w-l-¯t mt" "bd-hw h"n b-¶lf 3 […] ‘und (zum Dank) dafür, daß er (sc. der Gott T!LB) seinen Diener H"N gerettet hat in der Umgebung von […]’. 5

466

Articles: Semitic Studies

Spezifikum amiritischen Sprachgebrauchs angesehen und unsere Liste nordarabischer Einflüsse in diesem Dialekt dementsprechend erweitert werden. Zugleich ist der Beleg das mit Abstand früheste Zeugnis für die Partikel !iyyā- überhaupt. In den frühnord- bzw. altarabischen Inschriften ist eine entsprechende Form bislang nicht nachgewiesen,8 während andere arabische Merkmale des Amiritischen dort sehr wohl präsent sind.9 Damit ist ein weiteres arabisches Syntagma in der vorislamischen Epigraphik nachgewiesen, und zwar womöglich ein ganzes Jahrtausend vor dem Aufkommen der (vorklassischen) arabischen Literatur.

8

Die diesbezügliche Recherche wird durch freundliche Auskunft von Michael C. A. Macdonald (Oxford) bestätigt, wofür der Autor herzlich danken möchte. 9 Vgl. den Hinweis bei Macdonald (2004:521) auf die Negation lm + Präfixkonjugation im Ôafaitischen.

P. Stein, Ein weiteres arabisches Syntagma …

Abbildung 1. Mon.script.sab. 672 (Faksimile P. Stein)

467

468

Articles: Semitic Studies

Bibliographie Brockelmann 1913 Jamme 1962 Macdonald 2004

Müller 1978 Nebes 1995

Reckendorf 1921 Robin 1992 a´-Ôilwī 2005

Solá Solé 1964 Stein 2007

Stein 2010

Weninger 2002

Brockelmann, C. Grundriß der vergleichenden Grammatik der semitischen Sprachen. Bd. 2. Berlin. Jamme, A. Sabaean Inscriptions from Maµram Bilqîs (Mârib) (AFSM III). Baltimore. Macdonald, M. C. A. Ancient North Arabian. Woodard, R. D. (Hrsg.). The Cambridge Encyclopedia of the World’s Ancient Languages. Cambridge. S. 488–533. Müller, W. W. Die sabäische Felsinschrift von Ma´na"at Māriya. NESE 3:137–148. Nebes, N. Die Konstruktionen mit /fa-/ im Altsüdarabischen. Syntaktische und epigraphische Untersuchungen (Veröffentlichungen der Orientalischen Kommission der Akademie der Wissenschaften und der Literatur Mainz 40). Wiesbaden. Reckendorf, H. Arabische Syntax. Heidelberg. Robin, Ch. Inabba!, Haram, al-Kāfir, Kamna et al-Ñarāshif (Inventaire des inscriptions sudarabiques 1). Paris–Rome. ̬ a´-Ôilwī, !I. M. Naqš gadīd min nuqūš al-i"tirāf al-"alanī (naqš min ma"bad !D̠.NN). Dirāsa fī dalālātihi al-luġawīya wa-d-dīnīya. Sholan, A. M.; Antonini, S; Arbach, M. (Hrsg.). Sabaean Studies. Archaeological, Epigraphical and Historical Studies in Honour of Yūsuf M. "Abdallāh, Alessandro de Maigret and Christian J. Robin on the Occasion of Their 60th Birthdays. Naples–Ôan"ā!. S. 109–121 (arab.). Solá Solé, J. M. Inschriften aus Riyām. Sammlung Eduard Glaser 4 (SAWW 243/4). Wien. Stein, P. Materialien zur sabäischen Dialektologie: Das Problem des amiritischen (“haramischen”) Dialektes. ZDMG 157:13–47. Stein, P. Die altsüdarabischen Minuskelinschriften auf Holzstäbchen aus der Bayerischen Staatsbibliothek in München. Bd. 1. Die Inschriften der mittel- und spätsabäischen Periode (Epigraphische Forschungen auf der Arabischen Halbinsel 5). Tübingen–Berlin. Weninger, S. More Sabaic Minuscule Texts from Munich. PSAS 32:217–223.

Idanda Archive and the Syrian Campaigns of Suppiluliuma I: New Pieces for the Puzzle Daria Gromova Moscow State University

The political history of the Syrian kingdoms in the Amarna Age still has a lot of gaps, though it has been attracting much scholarly attention. Therefore publication of new texts from this period is always greatly welcomed. In 2002, the Royal Palace of Tell Mishrifeh (Qatna) made a fine present to its excavators—a tiny cuneiform archive, containing letters addressed to Idanda, king of Qatna.1 Unfortunately, this archive remains unpublished as a whole, but some of the texts keep being published in fragments by Thomas Richter, the epigraphist of the German archaeological expedition at Tell Mishrifeh. The first substantial set of fragments was published in 2002.2 The article written by Richter in 20083 contains a new one. Both articles are trying to relate the new data to the traditional reconstruction4 of the political history of Syria, thus provoking a number of questions. I have already touched on some of them previously.5 The 1 Richter, Th. Das “Archiv des Idanda”. Bericht über die 2002 in Qatna gemachten Inschriftenfunde. MDOG 135 (2003):167–188. 2 Richter, Th. Der “Einjährige Feldzug” Šuppiluliumas I. von Hatti in Syrien nach Textfunden des Jahres 2002 in Mišrife/Qa¢na. UF 34 (2002):603–618. 3 Richter, Th. Šuppiluliuma I. in Syrien. Der “Einjährige Feldzug” und seine Folgen. Wilhelm, G. (Hrsg.). Hattuša-Boğazköy—Das Hethiterreich im Spannungsfeld des Alten Orients (CDOG 6). Wiesbaden, 2008. Pp. 198–199. 4 I mean, first of all, the extensive studies by H. Klengel and J. Freu (Klengel, H. Geschichte Syriens im 2. Jahrtausend v. u. Z. Berlin, 1965–1970; id. Syria 3000 to 300 B. C. A Handbook of Political History. Berlin, 1992; id. Geschichte des hethitischen Reiches. Leiden–Boston–Köln, 1998; Freu, J. Les guerres syriennes de Suppiluliuma et la fin de l’ère amarniene. Hethitica 11 (1992):39–101; id. La chronologie du règne de Suppiluliuma: essai de mise au point. Taracha, P. (ed.). Silva Anatolica. Anatolian Studies Presented to Maciej Popko on the Occasion of His 65th Birthday. Warsaw, 2002. Pp. 87–107). The main feature of their reconstructions is the connection of the most of the military activities of Suppiluliuma in Syria with the One-year War. 5 Gromova, D. Hittite Role in Political History of Syria in the Amarna Age Reconsidered. UF 39 (2007):286–302.

474

Short Notes: Ancient Near Eastern Studies

present article continues the discussion of the evidence of the text fragments published by Richter in 2008. Richter’s reconstruction of the history of Qatna in the Amarna period can be briefly summarized as follows. Its starting point is Richter’s attractive (although lacking proof) identification of Adad-nirari of Qatna (mentioned in the inventories of the temple of Nin-egal from Qatna and in the fragment of the letter MSH02G-i0380, published by Richter in his 2008 article) with Adad-nirari of Nuhašše (the author of EA 51, figuring also in the historical preamble to the Treaty of Suppiluliuma with Niqmaddu of Ugarit6).7 The main territory of this united state must have been Nuhašše (which explains why Adad-nirari is presented in EA 51 as the ruler of Nuhašše). As for Qatna itself, it is supposed to have been ruled by Adadnirari’s governor lúŠAGINA Lullu (mentioned in the inventories of the temple of Nin-egal from Qatna). The rule of Adad-nirari is dated by Richter to the period preceding the One-year War, during which he must have been deposed by the Hittites, his state being divided into three parts, where the Hittites’s creatures were enthroned: Idanda in Qatna, Šarrupše in Nuhašše and Takip-šarri in Ukulzat. A similar replacement occurred in Nii, where Aki-Teššup was substituted by Takuwa. Nii, however, kept its territorial integrity. Therefore, the One-year War turns out to be a terminus post quem for dating Idanda’s archive. The next period, according to Richter, was characterised by a relative political stability: Mitanni re-established its positions and won its former vassals round again. During this process some of the Syrian kingdoms faced a new change of power. One of the newly displaced rulers was the Hittite protégé Idanda. The final act of the play is the Six-year War, as a result of which the Hittites regained control over Qatna and put on its throne their new protégé—Akizzi, well known from the Amarna archive, whose nearly first step was to organize a revolt against his benefactors. The foregoing reconstruction causes a number of questions, which I would like to discuss below. First of all, there are no good reasons behind the assumption that Adad-nirari of Nuhašše (regardless of his possible identity with Adadnirari of Qatna) was dethroned by the Hittites, and specially for dating this event to the One-year War. The fact is that Adad-nirari is not men6

RS 17.340obv.:2–32. The same idea was proposed already by M. Astour in 1977 (Astour, M. C. Tunip-Hamath and Its Region. A Contribution to the Historical Geography of the Central Syria. Or 46 (1977):57). 7

D. Gromova, Idanda Archive and the Syrian Campaigns …

475

tioned in the sources from the period after the aggression of the coalition of Nuhašše, Nii and Mukiš against Ugarit. But the very dating of this event to the period of the One-year War does not look self-evident.8 Secondly, there is no evidence that Adad-nirari lost his throne by force, nor, moreover, that the Hittites played any part in this. On the contrary, one can find an indirect proof against this hypothesis. Adad-nirari is not at all mentioned in the Treaty of Suppiluliuma with Tette of Nuhašše. But the structure of the Hittite treaties implies that the whole history of relations between the Hittites and the other party of the treaty should be fixed in the historical preamble (in particular, transition from hostility to friendly relations owing to the endeavours of the Hittite king is usually emphasized). So, if Adad-nirari was in fact dethroned by the Hittites during the One-year War, why was this fact not mentioned in the Tette Treaty? Rather, the structure of this treaty indicates that Nuhašše submitted to Hatti only in the time of Šarrupše without any connection with Adad-nirari. Nor does Idanda’s presence on the throne of Qatna in the first years after the One-year War necessarily mean that he obtained it specifically as a result of this war, being appointed instead of his deposed predecessor. The Šattiwaza Treaty tells nothing about Qatna’s resistance to the Hittites, nor about the dethronement of its ruler. Ex hoc silentio we can conclude that there was just no resistance and no change of the ruler, since Idanda, reluctant to risk his small kingdom, could well have decided to collaborate with the Hittites, admitting their suzerainty. It is also interesting that in the Šattiwaza Treaty Suppiluliuma clearly distinguishes between Qatna and Nuhašše: he definitely does not include Qatna into the Nuhašše territories since he considers Qatna submitted (KBo I, 1obv.:37) even before entering the land of Nuhašše (KBo I, 1obv.:38–40). Due to the genre specifics, the Šattiwaza Treaty could not include more than a short synopsis of military operations. Accordingly, a lengthy description of submission first of a part of a state and then also of its main territory would be clearly redundant. The separate mention of Qatna per se speaks in favor of its independence from Nuhašše.

8

From my point of view, it took place earlier (Gromova, D. Hittite Role in Political History of Syria in the Amarna Age Reconsidered. UF 39:281–290). There is also a suggestion to date it to the later period: Freu, J. Ugarit et les puissances à l’èpoque amarnienne (ca. 1350–1310 av. J. C.). Semitica 50 (2000):25–27; id. La chronologie du règne de Suppiluliuma: essai de mise au point. Silva Anatolica, pp. 98–100.

476

Short Notes: Ancient Near Eastern Studies

Hereby I can accept Richter’s idea that Šarrupše and Idanda got the thrones of Nuhašše and Qatna after Adad-nirari (although these could be two different Adad-niraris), but I would rather date this change of power to the period before the One-year War. During the One-year War Šarrupše already lost his throne, which was described in the Šattiwaza Treaty: ‘When I went to the land of Nuhašše, I seized all its territories. Šarrupše escaped, but I captured his mother, his brothers, and his children, and brought them to Hatti. I put Takip-šarri, subject of Šarrupše, on the throne in the city of Ukulzat’ (KBo I, 1obv.:38–40). It is hard to interpret this episode as anything but a description of the deposition of Šarrupše from the throne of Nuhašše and deportation of his family. Note that the victory over Šutatarra of Qadeš is described in the same way in this text (KBo I, 1obv.:40–43). Moreover, a fragment of Hannutti’s letter to Idanda (MSH02G-i0193), published by Richter in 20089, probably corroborates this reconstruction: 1. a-na Iid-a-an-da 2. um-ma I¶a-an-nu-ut-ti 3. lu-ú šul-mu a-na UGU-ka 4. LÚ.MEŠURU-ia-né-na ša URUqàt-na 5. \ pu-uk-lu-uš-te ù ¢up-pu 6. lu-ú il9(LIL)-te-né-mu-šu-nu 7. um-ma dUTUx-ši LUGAL be-li-ia 8. \ da-ni-ia-áš-šu11(SU)-la-an 9. ša Išar-ru-up-še Išar-ru-up-še- 10. LUGAL KUR¶ur-ri \ ¶é-lu-ul-a 11. ù it-tal-kam a-na ia-ši 12. ù ki-a-am iq-bi šu-zi-bá-an-ni 13. \ u¶-ni-da-an aš-pur-šu 14. Išag-ga-pi \ e¶-lu-uš-a-ma-an 15. ù i-na-an-na a-na-ku it-tal-kam 16. \ ¶é-ez-zu-ši-ma-an GIŠIG 17. iš-tu pa-ni-ia i-nu-ma 18. Išar-ru-up-še TÚGqá-an-ni-ia 19. i´-bat-šu […] 34. at-tu4-nu-ma i-du-ku-nu 35. ù KURmi-id-da-ni ¶a-liq 36. ù iš-tu 3 GIŠGIGIR an-nu-ú 37. i-pal-la-¶u-ku-nu ù at-tu4-nu-ma 9 Richter, Th. Šuppiluliuma I. in Syrien. Der “Einjährige Feldzug” und seine Folgen. Wilhelm, G. (Hrsg.). Hattuša-Boğazköy—Das Hethiterreich im Spannungsfeld des Alten Orients (CDOG 6). Wiesbaden, 2008. Pp. 198–199.

D. Gromova, Idanda Archive and the Syrian Campaigns …

477

38. ta-am-mar6(ÚAR)-ku-nu \ wu-ri-da-áš-šu11(SU) 39. ša e-pu-uš-šu-nu \ da-na-áš-te-da-še-na ‘ 1–3 To Idanda. Thus (speaks) Hannutti: Let well-being be over you! 4–6 Gather citizens of Qatna and let them all listen to (this) tablet. 7 Thus (speaks) the Sun, king, my lord: 8–19 What Šarrupše did is that he submitted to the king of Hurri. And now he came to me and said: “Save me!” and because of his grief I sent (Šaggapi), and Šaggapi saved him. And now I came myself, and he opened the gate in front of me, so that Šarrupše seized a rim of my dress.’

Then Hannutti (or Suppiluliuma himself) tries to convince Idanda and the citizens of Qatna in the following way: ‘ 34–39 You know that the land of Mitanni is demolished, but you are afraid of these three chariots! You will see what they will do!’

It is obvious that this text on the whole follows the historical preamble to the Tette Treaty: ‘When the king of the land of Mitanni sought to kill Šarrupše, and the king of the land of Mitanni entered the land of Nuhašše with his select troops and his chariots and oppressed him, Šarrupše sent his messenger to me, King of Hatti, (saying:) “I’m subject of the King of Hatti. Save me!” And I, the Sun, sent troops and chariots to his aid, and they made the king of Mitanni with his troops and chariots go out from the land of Nuhašše’ (KBo I, 4 i + KUB III, 10obv.:2–11). The difficulty is that we do not know what was written between the lines 19 and 34. If we suggest that the letter describes the current events of submission of Šarrupše to the Hittites that should have taken place before on in the very beginning of the One-year War, then this would contradict the statement that Mitanni had been demolished. Another possibility is to date the letter to the period after the defeat of Mitanni’s main troops at the end of the Oneyear War and to suppose that those parts of the tablet which survived offer an excursus to the past, whereas the description of the present events— namely the repression of Šarrupše—is lost in the lacuna. At this point, an important question emerges: what was the aim of sending this letter? Political acts are rarely purposeless, and this letter is undoubtedly a political one. A very important detail is the instruction to read the letter in public, so that all the citizens of Qatna be aware of its contents (lines 4–6). This detail is quite unusual, since most rulers prefer to settle their affairs within their own circle. Addressing the masses is already an act of propaganda.10 Thus, the let10 Cf. the situation in 2 Kings 18:26–35, where Rabshakeh, the messenger of the Assyrian king, addresses not only the officials of the king of Judah in order to persuade them to surrender, but also appeals in public to the citizens of the besieged Je-

478

Short Notes: Ancient Near Eastern Studies

ter was likely intended to justify in public the reprisal over Šarrupše and, incidentally, to warn the citizens of Qatna (and Idanda himself ) against any attempts to oppose the Hittite rule. Hence, we can make the following conclusion: if the king of Hatti thought it necessary to send such a warning, it should have been caused by a certain instability among the inhabitants of Qatna, probably connected with possible advantages of switching to the side of the Hittites’ enemy, the aforementioned master of “three chariots.” It is obvious that, speaking literally, the “three chariots” mentioned in the letter could hardly be a real threat for a big city like Qatna—a threat capable to shake its loyalty to the present sovereign. What we have here is a meiosis, an intentional underestimation of the enemy’s forces used for the purpose of propaganda. This understatement of the potential of a political rival contrasts sharply with the demonstration of the might of the Hittites themselves and the menace of repression colorfully illustrated by the fate of Nuhašše. The last thing to do is to identify the lord of the “three chariots.” Richter denies the possibility of identifying him with Tušratta, arguing that Tušratta must have been already dead at the time. This negative evaluation is due to Richter’s dating of the letter to the period considerably after the One-year War,11 but if it is dated to the final period of the One-year War (as proposed above) Tušratta could well fit the present situation. The letter states that Mitanni is demolished. But the propagandistic style of the letter makes it feasible that we are merely faced with one more figure of speech—an exaggeration of the military success of the Hittites over Mitanni. Such a reconstruction becomes even more probable if we remember that Tušratta of Mitanni had not indeed been completely defeated during the One-year War. After the capture of his capital city of Waššuganni by Supiluliuma, Tušratta fled away (KBo I, 1obv.:27–29), surely not without troops which could be sufficient for trying to regain the control over Qatna. * * * I hope that in this note I have succeeded to show that the data of the newly published texts from the Idanda archive are compatible not only with the reconstruction proposed by Th. Richter, but also with the one I suggested in UF 39. Needless to say, no historical reconstruction of this period can be absolutely convincing. Hopefully, new texts relevant to the problems under scrutiny will appear. rusalem, speaking Hebrew instead of Aramaic, so that common people could understand him as well (thanks to L. Kogan for kindly reminding me of this parallel). 11 Richter, Th. Šuppiluliuma I. in Syrien. Der “Einjährige Feldzug” und seine Folgen, p. 201.

Marginalia Sargonica. I* Leonid Kogan, Ekaterina Markina Russian State University for the Humanities, Moscow

1. nabalkutum The only attestation of this verb in the N-stem is found in the economic document of unknown provenience MAD 4, 10: 3;0.1,20 ŠE GUR.SAG.GÁL in a-lí-im Ì-lí-iś-tá-kál ú-šu-ri-dam ib-ba-al-kì-it-ma a-na 2;0.0,40 ŠE GUR in súti ANŠE è-wi. The form ib-ba-al-kì-it-ma has been translated as ‘he transgressed’ in Hasselbach 2005:277. In view of the prominent negative connotations of nabalkutum,1 it is indeed conceivable that in the present passage this verb refers to a person who, after performing the delivery of grain, “behaved improperly” in this or another way. However, the general context of the document makes this interpretation rather unlikely and prompts one to think of nabalkutum as an accounting term. This is suggested, first of all, by the use of its nominal derivate nabalkattum in a document from Pugdan: (grain of the period of three years) na-ba-al-kà-at DUB Kí-nu-mu-pí (AIHA 8 iv 8). In this passage, nabalkattum in the construct state is followed by DUB (= ¢uppum ‘document’) and, with all probability, describes the result of an administrative action. In MAD 4, 10 ib-ba-al-kì-it-ma must render the same concept in a “verbal” form. A more precise meaning of nabalkattum can probably be elicited from Sumerian sources (including Sargonic economic documents), which use the expression (dub-bé)—bala ‘to transfer (a record) to (another) tablet’ (PSD B 51): En-ig-gal nu-bànda dub-bé e-da-bal gú-na e-ni-×ar ‘Eniggal, the supervisor, transferred (the debit of n quantity of grain) to the tablet * This note has been written in the framework of the project 09-04-00235a supported by РГНФ/RFH. The authors must express their sincere gratitude to this foundation for its financial support. 1 ‘To act against an agreement, to rebel against the authority’ (CAD N1 13). This semantic nuance does not seem to be attested in the Sargonic corpus, but note the use of the Š-stem with the meaning ‘to incite to rebellion’ in the inscriptions of Gutian rulers (Kienast–Sommerfeld 1994:247). Semantically neutral (‘he will remove’) is uś-ba-la-kà-at in the ritual text MDP 14, 90:10.

480

Short Notes: Ancient Near Eastern Studies

and entered it as his debt’ (VAS 14, 121 ii 5 – iii 3). Such a meaning could plausibly fit the context of AIHA 8, which lists amounts of grain obtained by the estate in three consecutive years and then sums them up as ŠE GIŠ.RA šu 3 MU … na-ba-al-kà-at DUB Kí-nu-mu-pí ‘threshed barley of three years … transfer (to) the tablet of Kí-nu-mu-pí.’2 Here the grain seems to be the object nabalkattum refers to, just as in the Sumerian passage quoted above. The Sumerian expression also occurs without dub: guru7 ká ×iš-kí×-ti uru-kù-ka-ka Ur-pú ká-guru7-ke4 e-bal ‘Urpu, the head storekeeper, has transferred (the amount of grain) in stock in the storehouse at the Craftsmen’s Gate at Uruku (to this tablet)’ (DP 566 ii 1 – iii 3). If we assume that nabalkat ¢uppim semantically corresponds to dub-(bé)—bala, ib-ba-alkì-it in MAD 4, 10 may well represent the Akkadian rendering of the same term with ¢uppum omitted.3 The Sumerian verb bala can be used both intransitively and transitively (PSD B 50), thus corresponding to Akkadian nabalkutum and šubalkutum respectively. While in the Sumerian documents discussed above we are faced with the transitive application, in MAD 4, 10 the intransitive one is used, thus referring to grain itself rather than to the person who delivered it: ‘PN delivered (the grain) from the city. It (= the grain) entered (the accounting) tablet.’ As the context of MAD 4, 10 suggests, nabalkutum does not necessarily denote a simple transferring of data to a document, but could also include related accounting operations such as recalculation. In this document ib-ba-al-kì-it is followed by the form è-wi, likely to be analyzed as a 3 m. sg. preterit of ewûm (*hwy) ‘to become, to turn into’ (CAD E 413). The orthography of the verbal prefix (È = /ye/) is reminiscent of è-la-kam ‘he will come’ in Ga 1:6, which also has *h in the prototype. Since the preposition ana is standard for ewûm with the meaning ‘to turn into, to become,’ the sentence a-na 2;0.0,40 ŠE GUR in sú-ti ANŠE è-wi likely means 2

The idea that nabalkattum in this passage can be understood upon the semantic background of Sumerian bala has been expressed already by Foster (1982:24), for whom the underlying concept is physical transfer of a commodity from one person to another. In the light of the expression dub-bé—bala it seems more likely that nabalkat ¢uppim refers to administrative paperwork. 3 While no similar expression involving nabalkutum is attested in later Akkadian, the Š-stem of this verb does frequently appear with the meaning ‘to transfer (letters, merchandize, etc.)’: me¶ram ša ¢uppim … šubalkit-ma šēbilam ‘Transfer the copy of the document and have it delivered here’ (TCL 19, 9:19 and other examples in CAD N1 18).

L. Kogan, E. Markina, Marginalia Sargonica. I

481

‘it (= the grain) became 2 kor and 40 sìla of barley in the ass-seah.’ After being registered in a document, the original quantity of grain became reduced to a different amount measured in a system based on a different capacity unit (sūt ANŠE).4 The whole passage is thus to be interpreted as follows: ‘Iliś-takal delivered (the grain) from the city. It entered (the accounting tablet) and became 2 kor and 40 sìla of barley in the ass-seah.’ 2. barûm The occurrences of the verb(s) barûm in Sargonic economic documents fall into three major groups. (1) The first and best known category comprises passages with barûm ‘to see, to inspect.’ All known instances are in the G-stem. These passages deal with administrative inspections carried out by various Sargonic officials. Only one such example, viz. MAD 1, 220:11–13, is adduced in the glossary of Hasselbach 2005 (p. 266): Na-bí-Ùl-maš in Tu-tuki ib-rí ‘NabiUlmaš made an inspection in Tutub’ (MAD 1, 220:11–13). This attestation must be supplemented by three more structurally similar examples: (animals PN) ib-rí ‘PN inspected (the animals)’ (BIN 8, 183rev.:5; BIN 8, 273 iv 8; NBC 6847 iv 10). (2) In the passages belonging to the second group, barûm (mostly in the D-stem) is used in connection with land: (x GÁNA) … Lu-lu ú-ba-rí (BIN 8, 144 i 3–4), (x GÁNA) šu li-ip-tim [PN] … (y GÁNA) šu bu-ru-im [PN2] šu pá-šari-im (MAD 1, 332obv.:9 – rev.:5′). As far as BIN 8, 144 i 3–4 is concerned, Hasselbach (2005:266) translates ú-ba-rí as ‘he announced?,’ but it is unclear to us how such a meaning could fit the context in that passage. For two reasons, Foster’s identification with barûm ‘to be hungry’ (Foster 1982:19) is considerably more appealing. On the one hand, the use of barûm with the meaning ‘to starve for water (about land parcels)’ in the 4 For sūt ANŠE as a capacity unit and its metrological implications see Chambon 2006:95. In his translation of MAD 4, 10 Chambon interprets nabalkutum as ‘to change one’s mind’ (‘il a changé d’avis’). This interpretation is in agreement with Chambon’s reading of l. 8 (è-’à ‘il va (la) faire sortir’), which, however, is impossible as PI (rather than É) is clear on the available photo (moreover, the very reading underlying the alleged sequence È É—Sumerian or Akkadian?—is obscure and nowhere explained by the author).

482

Short Notes: Ancient Near Eastern Studies

G-stem is well documented in OB5 and is at least once attested in the Sargonic corpus: eqlum a ib-ra ‘May the field not hunger (for water)’ in Di 8:5 (missing from Hasselbach’s glossary).6 On the other hand, the general context of BIN 8, 144 suggests that we are faced with an action making land uncultivable. In the beginning, this document states that a certain official has made a plot of land—the work assignment of 5 ploughing teams—‘hungry’ for water (ú-ba-rí) and hence not suitable for cultivation. The final section of the document says that the same official has leased a parcel of land from the ensi: x GÁNA iś-tum Kí-nu-mu-pí ÉNSI Lu-lu ú-śé-´í ‘Lulu has leased x land from Kí-nu-mu-pí, the ensi.’ Since the areas of the parcel leased and one deprived of water supply are nearly equal, it lies at hand to conclude that the land was leased as a replacement for the land that became temporarily unusable. The document MAD 1, 332 is damaged, but what remains suggests that we are dealing with a list of land plots provided with a brief description and the name of the owner. The last plot in the document seems to have two descriptions at once: šu pá-ša-ri-im cannot refer to all the parcels mentioned before it since in such a case one would expect šu-ut instead of šu. If pašārum in this passage is used in its regular meaning ‘to release, to undo’ (CAD P 236), the interpretation of barûm D as ‘to make hungry for water’ becomes quite plausible: the passage would mean that the land plot was first deprived of water (possibly because some canal work was about to be carried out) and then the normal level of watering was restored to it. The two passages can thus be translated as follows: ‘(x bur of the field) Lulu made hungry (for water)’ (BIN 8, 144 i 3–4); ‘(x bur of the field assigned) for work—Ba"lī-asû …; (y bur of the field previously? assigned) to be made hungry (for water) is to be released (from this state)’ (MAD 1, 332).

(3) The third group consists of one passage only, again unmentioned in Hasselbach’s grammar in spite of the peculiar use of the sign RI, quite uncommon in the Sargonic orthography: (x grain) ba-ri-um (MAD 1, 151:1–4). 5 E. g. mû ipparrasū-ma māt bēlīya iberri ‘If the water is cut off, the land of my lord will be hungry (for water)’ (ARM 1, 3:18). 6 Kienast and Volk (1995:164) understand eqlum a ib-ra in a slightly different way: ‘Das Feld soll nicht brach liegen!’ (first suggested in Westenholz 1974:74). Note that ib-ra (< barûm ‘to be hungry (for water)’) is clearly distinct from ib-rí (< barûm ‘to see, to inspect’) not only semantically, but also morphologically: barûm ‘to be hungry’ (presumably with a guttural as the third radical) has a as the thematic vowel, whereas barûm ‘to see, to inspect’ (< *bry) displays i.

L. Kogan, E. Markina, Marginalia Sargonica. I

483

The document lists two amounts of grain, of which the former is qualified as SIG5 (= damqum ‘good’) and the latter, as ba-ri-um. This adjective can hardly be related to either of the barûm-verbs discussed above in this note. Although the general context is not very informative, the meaning of barium in this context can possibly be deduced from the entry 718 of the bilingual lexical list from Ebla, which reads ¶ul = ba-rí-um (Conti 1990:184). As the Sumerian equivalent clearly suggests, the Eblaite lexeme must mean ‘bad, improper.’ Such a meaning would plausibly fit the context of the Sargonic document, yielding the following translation: 3046;2.0 ŠE GUR ŠE SIG5 132.2.2 [ŠE GU]R ba-ri-um ‘3046;2.0 kor of barley—good grain; 132.2.2 kor of barley—[gra]in of low quality.’

3. Vocative and mimation in RIME 2.1.4.6 I 21″ Narām-Sîn’s account of putting down the revolt of Ip¶ur-Kiš(i) as preserved in the OB copies now housed in Philadelphia and Haifa is fraught with linguistic and philological difficulties which cannot be resolved in the framework of this note.7 Nevertheless, the basic meaning of the sequence AN-UD-SU KIŠ-KI-ŠI-UM in ll. 41′–42′ of the first column of the Haifa tablet has been correctly elicited already by its editor (Kutscher 1989:32): “These lines seem to be the beginning of Naram-Sin’s prayer to Šamaš.” Indeed, the first preserved lines of the second column of the tablet almost certainly belong to the king’s report about the ingratitude of the Kišites who “crossed his way”8 in spite of the favors lavished on them by his forefather Sargon (cf. Goodnick Westenholz 1997:234). In such a context, NarāmSîn’s appeal to Šamaš as the god of justice is quite natural. Kutscher’s normalization dUTU-śu KIŠki-ši-um and the ensuing translation ‘Oh Šamaš! The Kišite…’9 are thus basically correct and have been rightly accepted in Frayne 1993:105. Gelb and Kienast preserve AN.UD-śu in their edition (1990:228, without translation) and try to justify their doubts by the following remark: “der Vokativ nach GAG § 62j endungslos sein sollte” (Gelb–Kienast 1990:240). 7 The possibility of a coherent interpretation of the passage under scrutiny will largely depend on the possibility of a new inspection of the Haifa tablet. 8 i-tá-kir9 śá-pí-a /yittaker śa!pēya/, literally ‘he opposed my feet’ (for this interpretation of SÁ-BI-A v. Markina 2010). A feasible alternative is /yītager śa!pēya/, literally ‘he crossed my feet’ from egēru ‘to cross’ (CAD E 41). The first person pronominal suffix fits eminently the reading of this passage as part of Narām-Sîn’s prayer: direct speech in the first person singular. 9 Referring to the rebellious king of Kiš.

484

Short Notes: Ancient Near Eastern Studies

The absolute state Šamaš would indeed be expected here according to the rules of standard Akkadian grammar, and not just because of the vocative, but also in view of such well-known parallels as La¶ar, Ašnan, Siraš or Šam¶at. One may legitimately wonder, however, whether these rules are fully applicable to the archaic strata of the Akkadian language as represented by Sargonic royal inscriptions. As pointed out in Kogan–Markina 2006:556, this corpus has already yielded one prominent example of a nominal form with a case ending but without mimation, namely ma-¶i-ra ‘opponent’ in the formulaic expression DN mā¶ira lā yiddiśśum ‘the god gave him no rival.’ Lack of mimation in mā¶ira finds a transparent explanation in the syntactic nature of the phrase: as an old marker of definiteness (Kuryłowicz 1950; Dolgopolsky 1991), mimation is incompatible with a prototypically indefinite environment still patent even in modern translations (einen Rivalen, no opponent). A similar explanation can now be applied to dUTU-śu in RIME 2.1.4.6 I 21″. On the one hand, lack of mimation in Śamśu may be conditioned by the vocative, which would find an excellent parallel in Classical Arabic, where nominative without nunation is regularly used after the vocative particle yā: yā muµammad-u ‘Oh Muhammad!’ (Fischer 2002:95). This explanation is well compatible with the use of nunationless dual forms in the vocative context of MAD 5, 8:6–7 (Hasselbach 2005:184): wa-a[r-d]a-tá da-mì-iq-ta /wardat-ā damiqt-ā/ ‘Oh two beautiful maidens.’ On the other hand, mimation may be missing because Śamśu is definite by nature as a proper name. Here again an instructive Arabic parallel is at hand: proper names form a prominent subclass of diptotic nouns, which are incompatible with nunation (Fischer 2002:93). The historical development of the “old absolute state” can be outlined as follows. When mimation ceased to function as the marker of definiteness (presumably well before the Sargonic period), forms with mimation became the standard (unmarked) forms of all Akkadian nouns (the “normal state”). It was just in a few restricted environments that the spread of mimation was blocked, presumably because association with definiteness was still perceived there as undesirable (vocative, prototypical syntactic indefiniteness) or superfluous (proper names). Later on, also this restricted usage was abandoned and the “new absolute state” without case endings was introduced in the corresponding syntactic and semantic slots.

L. Kogan, E. Markina, Marginalia Sargonica. I

485

References Chambon 2006 Conti 1990 Dolgopolsky 1991

Fischer 2002 Foster 1982 Frayne 1993 Gelb–Kienast 1990 Goodnick Westenholz 1997 Hasselbach 2005 Kienast–Sommerfeld 1994 Kienast–Volk 1995

Kogan–Markina 2006 Kuryłowicz 1950 Kutscher 1989 Markina 2010

Westenholz 1974

Chambon, G. La mesure de capacité “ânée” et la marque en forme d’âne. NABU 2006/94. Conti, G. Il sillabario della quarta fonte della lista lessicale bilingue eblaita (QuSem 17). Firenze. Dolgopolsky, A. Two Problems of Semitic Historical Linguistics. Kaye, A. (ed.). Semitic Studies in Honor of Wolf Leslau. Wiesbaden. Pp. 328–339. Fischer, W. A Grammar of Classical Arabic. New Haven– London. Foster, B. R. An Agricultural Archive from Sargonic Akkad. ASJ 4:7–51. Frayne, D. The Royal Inscriptions of Mesopotamia. Early Periods. Vol. 2. Sargonic and Gutian Periods. Toronto. Gelb, I. J.; Kienast, B. Die altakkadischen Königsinschriften des dritten Jahrtausends v. Chr. (FAOS 7). Stuttgart. Goodnick Westenholz, J. Legends of the Kings of Akkade. Winona Lake. Hasselbach, R. Sargonic Akkadian. A Historical and Comparative Study of the Syllabic Texts. Wiesbaden. Kienast, B.; Sommerfeld, W. Glossar zu den altakkadischen Königsinschriften (FAOS 8). Stuttgart. Kienast, B.; Volk, K. Die sumerischen und akkadischen Briefe des III. Jahrtausends aus der Zeit vor der III. Dynastie von Ur (FAOS 19). Stuttgart. Kogan, L.; Markina, E. Review of Hasselbach 2005. B&B 3:555–588. Kuryłowicz, J. La mimation et l’article en arabe. ArOr 48: 323–328. Kutscher, R. The Brockmon Tablets at the University of Haifa. Royal Inscriptions. Haifa. Markina, K. “They Embraced His Feet, Saying…”. śa!pēn a¶āzum and śa!pēn ezēbum as Idioms of Loyalty and Defiance in Sargonic. ZA 100:65–68. Westenholz, Aa. Early Nippur Year Dates and the Sumerian King List. JCS 26:154–156.

Accusative casus pendens: Some Further Examples Leonid Kogan Russian State University for the Humanities, Moscow

Martin Worthington School of Oriental and African Studies, London

1. The Akkadian casus pendens can display either nominative or accusative case markers. This phenomenon was the subject of a recent study by one of the present authors (Kogan 2008), who tried to demonstrate that, in view of it, relative clauses introduced by ša (rather than šu) in Old Akkadian (Sargonic) royal inscriptions can no longer be regarded as inexplicable syntactic oddities (or even mistakes). L. Kogan (2008) provides numerous parallel examples from later (primarily Old Babylonian) sources, but almost all of them come from one single corpus: the OB royal inscriptions. In this article, a few additional examples of the casus pendens displaying the accusative marker in Old Babylonian and Old Assyrian are collected, most of them coming from corpora other than royal inscriptions (esp. letters). This fact is of some importance in itself, as it shows that the feature under scrutiny was not bound to one register only, but occurred in several (including nonliterary ones). After presenting the attestations, we will try to assess what light they can shed on the origin of the accusative casus pendens. 2. The examples collected in this section are structurally the closest to those dealt with in Kogan 2008. Except, of course, for the accusative ending, they represent the pendens construction in its classical form: an extraposed possessor is resumed by an anaphoric pronominal element attached to the possessed. (1) Išu-m[u]-um-li-ib-ši [´]e-e¶-ra ša a-na ma-a¶-ri-ka a¢-ru-dam a-na na´a-ri-šu ù i-ta-ap-lu-s[i]-š[u] ni-di a-¶i-im la ta-ra-aš-[ši] ‘Little Šumum-libši, whom I sent to you—do not be negligent about guarding and watching him!’ (AbB 7, 138:26–29). (2) ¶a-am-mu-ra-pí ru-ba-am na-a!-dam pa-li-i¶ ì-lí ia-ti … šu-mi ib-bu-ú ‘Me, Hammurabi, reverent prince who fears the gods— … they summoned me’ (CÚ I 28–49).

488

Short Notes: Ancient Near Eastern Studies

We understand the entire phrase Úammurāpi rubâm na!dam pa-li-i¶ ì-lí ia-ti as a single accusative casus pendens: it is an extraposed possessor, appropriately resumed by a first person singular possessive suffix attached to the possessed (šumī ‘my name’). 1 This interpretation clarifies the grammar of the sentence, which is otherwise problematic. For either (a) nabû exceptionally has three accusatives; or (if šumī were understood as an adverbial accusative, leaving nabû with the more usual two objects),2 then (b) one would have an unidiomatic formulation.3 The interpretation of the whole phrase up to iâti as an accusative casus pendens dissolves these difficulties. Our interpretation also has some bearing on the general meaning of the passage: with CAD N1 35–36, it seems to belong to meaning 3b ‘to summon, call a person (to exercise a function), to appoint a person to an office’ rather than to meaning 1a ‘to name, to give a name’ (ibid. 33). The emphasis is thus not so much on the gods giving king Hammurabi his name, but—more appropriately in this context— appointing him to his office.4 This interpretation is supported by the infinitive of purpose (mīšaram ina mātim ana šūpîm), which is a characteristic feature of the summoning/appointing passages listed in CAD. (3) i-la iš-mu-ú ri-gi-im-šu ‘The god—they heard his cry’ (Atra-Úasīs III ii 50).

Lambert and Millard (1969:160) suggested as a “mere conjecture” that ila is a hapax conjunction meaning ‘as soon as,’ but this has found no 1 CAD N1 36b quotes only anum u enlil ana šīr nišī ¢ubbim šumī ibbû, translating ‘Anu and Enlil appointed me to make the people happy,’ omitting the foregoing words. The omission may suggest that the editors recognised that the construction involved a casus pendens, but in any case is odd. Ditto AHw. 699b, which quotes only šumī ibbû. 2 This seems to be the interpretation of several scholars, who translate šumī as ‘by my name’ (e. g. Driver–Miles 1955:7; Roth 1995:76; sim. Borger 1982:40 ‘Damals haben mich, Hammurapi, … mit meinem Namen genannt’), but we find it unsatisfactory for the reason given above. Driver and Miles’s remark “literally ‘called my name Hammurabi’ ” is opaque: what they call “literal” looks more like a paraphrase (it does not take account of iâti), whereas their actual translation does seem to be more or less literal. 3 The sentence would run ‘the gods called me, Hammurabi, the pious prince … by my name’ = ‘summoned me’ or ‘appointed me,’ but as far as one can judge from the examples collected in CAD N1 36–37 the second object is usually avoided in summoning-passages, which normally look like ibbûnim ‘they called me’ or šumī ibbû ‘they called my name,’ not *šumī ibbûnim ‘they called me by my name.’ 4 No clear distinction between the two meanings is made in AHw. 699: “(mit Namen) nennen … v(on) Königen (zur Berufung).”

L. Kogan, M. Worthington, Accusative casus pendens …

489

support in subsequent research. A recognition of casus pendens—which is a possessive construction after all—may well underlie Hecker’s somewhat obscure comment on this line (“durch appositionelle Konstruktionen können dann gelegentlich auch genitivische Verbindungen umschrieben werden,” 1974:125). (4) ´ú-ba-tám qá-at-na-am ša tù-šé-bi-li-ni ša ki-ma šu-wa-ti ep-ší-ma ‘The thin textile which you sent me—make (more) like it!’ (TC 3, 17:6–8). (5) a-ba-ar-ni-a-am ša tù-šé-bi-li-ni la ta-tù-ri-ma ša ki-ma a-mì-im la tùšé-bi4-li-im ‘The Abarnian textile which you sent me—do not send me another like that one again!’ (ibid. 23–26). (6) ga-am-ra-am ´ú-ba-ta-am ša té-pí-ší-ni tí-šé i-na-mì-tim lu ú-ru-uk-šu ša-ma-né i-na a-mì-tim lu ru-pu-šu ‘A complete textile which you make—it should be nine cubits long and eight cubits wide’ (ibid. 33–36).

The Old Assyrian letter from which examples (4–6) have been excerpted was carefully edited and commented upon in Veenhof 1972:103– 109. Veenhof makes no explicit statement about the unusual accusative in any of the three passages, although in example (6) the accusative forms are appropriately provided with an exclamation mark, whereas in examples (4–5) the presence of the dash in Veenhof ’s translation almost necessarily presupposes an accusative casus pendens interpretation for the relevant passages. (7) a-la-ší-am ki-ma sí-ni-iš-tim qá-qá-da-tí-šu-nu ak-tù-um ‘The people of Alashia—I covered their heads like a woman’s’ (ArAn 3, 135: 53–55). (8) gu5-tí-tám lu-lu-am ù ¶a-¶a-am ´ú-ba-tí-šu-nu ú-ša-ri-im ‘The people of Gutium, Lullubum and Hahhum—I slit open their clothes’ (ibid. 62).5

These two examples of unusual accusatives in the Old Assyrian Sargon story went unnoticed in its recent detailed treatments (Dercksen 2005; Cavigneaux 2005; Alster–Oshima 2007). The first example is to some extent suspect, as one cannot exclude that Alašiam was the only form of this geographic name, and not declined.6 The second is quite unambiguous, however (for the genitive ¶a-¶i-im in the OA corpus v. Nashef 1991:46). 5

Reading and interpretation after Dercksen 2005:109. For the same reason we thought it wise not to consider Úu-du-ra bi4-bi4-na-tim qá-qá-da-tí-šu-nu áš-ku-un ‘As for the people of Hudura—I arranged a slave’s hairdress on their heads.’ 6

Short Notes: Ancient Near Eastern Studies

490

3. Examples included in this section are different from the foregoing ones, insofar as the explicit pronominal anaphora is lacking. This is undoubtedly due to the fact that the role of the noun in the accusative is different: it is not a possessor, but rather the subject.7 Now, pronominal resumption in the nominative is not necessary (and, indeed, uncommon) in Semitic. In more practical terms, it means that the examples below could have been more or less naturally read as simple linear sentences8 if not for the unexpected accusative, which alone arouses the suspicion that we are faced with a phenomenon similar or identical to the main subject of this note. (9)

ù am-tam ša a-na šu-bu-lim a-na ´é-ri-ka im-tu-ta-an-ni ‘The slave girl who was to be sent to you—she has died on me’ (AbB 2, 87: 10–11).

The unusual accusative was duly observed by Frankena, whose translation interprets amtam as casus pendens (topicalized: ‘(was) die Sklavin (betrifft), ... diese ist mir gerade gestorben’). (10) a-wi-le-e ša aš-šu-mi-šu-nu a-na ša-pí-ri-ia sà-ar-tam iq-bu-ú Ia-¶uum ù na-ra-a[m]-ì-lí-šu eb-bu-tum ša ša-pí-ri [i]š-ku-nu … ‘The men about (or: because of ) whom they told a lie to my superior—A¶um and Narām-ilišu, the inspectors whom my superior appointed …’ (AbB 13, 34:11′–15′).

This instance is of uncertain interpretation, owing to the lacuna which follows. Nonetheless, it seems likely that awīlê is a casus pendens (so clearly van Soldt’s translation in the edition). (11) A.ŠÀ-am ša a-na … in-na-ad-nu ki-ma na-ad-nu-ma na-di-in ‘The field which was given to …—it was given as it was given’ (AbB 4, 6:4–7).

The unexpected accusative was noted by Kraus. (12) a-wi-le-e ša a¢-ru-da-ku mu-uš-ke-nu-ú ‘The men whom I sent to you—they are wretches!’ (AbB 5, 166:6–7).

Since awīlê is the logical object of a¢rudakku, it is possible that its accusative ending was induced by some sort of case attraction (see section 4 below). This was already the view of F. R. Kraus: “falscher Kasus durch Prädikat des Relativsätzchens attrahiert.”

7

The authors are grateful to Maria Bulakh who kindly drew their attention to this very important circumstance. 8 Thus, example (9) could be easily transformed into *ù am-tum ša a-na šu-bu-lim a-na ´é-ri-ka im-tu-ta-an-ni ‘The slave girl who was to be sent to you has died on me.’

L. Kogan, M. Worthington, Accusative casus pendens …

491

(13) a-wi-lam ša gišKIRI6-šu la ka-ap-du a-na gišKIRI6 ša ta-ap-pí-šu i-inšu i-na-aš-ši-ma gišKIRI6-šu ú-ul i-ka-pu-ud ‘A man whose garden is not looked after—he looks enviously on his friend’s garden, and does not look after his (own) garden’ (AbB 14, 69:10).

The unexpected accusative ending was commented on already by Veenhof in his edition.9 This example is, strictly speaking, not identical to the others treated in this section because the possessor (‘a man’) is reassumed by an anaphoric pronoun attached to the possessed (‘his garden,’ ‘his eye’). However, the possessed in this case does not constitute an independent syntactic unit, as it is governed by the same verbal form whose subject is the possessor. In other words, this sentence, too, could easily be transformed into a normal linear one (*a-wi-lum ša gišKIRI6-šu la ka-ap-du a-na gišKIRI6 ša ta-ap-pí-šu i-in-šu i-na-aš-ši … ‘A man whose garden is not looked after looks enviously …’). The unexpected accusative is the only reason for suspecting some kind of topicalizing structure here.10 4. In the majority of examples (1–5 and 7–8) treated in Section 2 (“classical” pendens), the possessed is in the accusative, but even when it is not, its role as a logical direct object is not in doubt. This is in agreement with the bulk of the evidence collected in Kogan 2008 and makes rather feasible the possibility of the accusative casus pendens emerging from the nominative one via case attraction from the possessed. It is only example (6) that deviates from this picture.11 In all examples listed in Section 3 (and, indeed, in most of those from Section 2) the possessor is followed by a relative clause. In view of its obvious prominence, this circumstance is difficult to disregard and it stands 9 Veenhof comments “awīlam must be a mistake, presumably because the scribe originally had a different construction in mind.” In passing, we note that Veenhof ’s translation of the whole sentence (‘A man who, not having planned his own garden, covets his colleague’s garden, will never develop his own garden!’) is syntactically difficult (it presupposes inaššû rather than inašši and disregards the copulative -ma). 10 The authors are grateful to N. J. C. Kouwenberg who, already when Kogan 2008 was in preparation, was kind enough to share with them his observations on this interesting passage, pointing out its specificity with respect to other examples analyzed in Kogan 2008 (and, by implication, in the present note). 11 The syntactic structure of this example (nominal sentence with the precative particle) makes it strikingly similar to one of the few exceptional cases treated in Kogan 2008:19, viz. sa-am-su-i-lu-na na-aš-pa-ri dan-nam la a-ni-¶a-am … lu nu-úršu na-aw-ru-um at-tu-nu-ma ‘Samsuiluna, my mighty and untiring envoy …—may you be his shining light’ (RIME 4.3.7.7:30–36).

492

Short Notes: Ancient Near Eastern Studies

to reason that this syntactic feature also played its role in the emergence of our construction. Such a possibility is most easily conceivable when the possessor is the direct object of the verb of the relative clause, as in example (12): a-wi-le-e ša a¢-ru-da-ku ‘the men whom I sent to you …’ It is hard to say, at present, which of the two factors was more prominent in the history of this construction, but one thing is certain: both in this note and in Kogan 2008 we are faced with some examples which are not explainable by either. One may wonder, therefore, whether a certain trend towards using the accusative case as a kind of focus marker can be postulated for early Akkadian. References Alster–Oshima 2007 Borger 1982 Cavigneaux 2005

Dercksen 2005 Driver–Miles 1955 Hecker 1974 Kogan 2008

Lambert–Millard 1969 Nashef 1991 Roth 1995 Veenhof 1972

Alster, B.; Oshima, T. Sargonic Dinner at Kaneš: The Old Assyrian Sargon Legend. Iraq 69:1–20. Borger, R. Akkadische Rechtsbücher. TUAT 1/1:32–95. Cavigneaux, A. Les soirées sargoniques des marchands assyriens. Kolde, A. et al. (eds.). Korufa…J ¢ndr…. Mélanges offerts à André Hurst. Geneva. Pp. 595–602. Dercksen, J. G. Adad is King! The Sargon Text from Kültepe. JEOL 39:107–129. Driver, G. R.; Miles, J. C. The Babylonian Laws. Oxford. Hecker, K. Untersuchungen zur akkadischen Epik (AOAT Sonderreihe 8). Kevelaer. Kogan, L. Accusative as Casus Pendens? A Hitherto Unrecognised Emphatic Construction in Early Akkadian Royal Inscriptions. RA 102:17–26. Lambert, W. G.; Millard, A. R. Atra-Úasīs. The Babylonian Story of the Flood. Oxford. Nashef, Kh. Die Orts- und Gewässernamen der altassyrischen Zeit (RGTC 4). Wiesbaden. Roth, M. Law Collections from Mesopotamia and Asia Minor. Atlanta. Veenhof, K. Aspects of Old Assyrian Trade and its Terminology. Leiden.

Exodus 33:7 and Different Syntactic Patterns for Linking Participial Conditional Clauses with the Main Clause Michael Seleznev Russian State University for the Humanities, Moscow

1. The problem In Ex 33:7–10 we read (NRSV): ‘ 7 Now Moses used to take the tent and pitch it outside the camp, far off from the camp; he called it the tent of meeting. And everyone who sought the LORD would go out to the tent of meeting (kolm3baqqēš YHWH yē´ē! !el-!ōhel mô"ē¯), which was outside the camp. 8 Whenever Moses went out to the tent, all the people would rise and stand, each of them, at the entrance of their tents and watch Moses until he had gone into the tent. 9 When Moses entered the tent, the pillar of cloud would descend and stand at the entrance of the tent, and the LORD would speak with Moses. 10 When all the people saw the pillar of cloud standing at the entrance of the tent, all the people would rise and bow down, all of them, at the entrance of their tents.’

Exegetes dealing with this pericope used to concentrate on such issues as the function of the ‘tent of the Meeting,’ which appears here for the first time in the narrative, or its relation to the Tabernacle. Another exegetical problem of the text used to be overlooked: what was the reason for ‘everyone who sought the LORD’ to go out to the tent of the Meeting? It is clear from what follows that not ‘everyone’ was allowed to enter the tent or to speak with God. Only Moses was granted these rights. Ordinary people were to watch his entering the tent from afar, struck with fear. The text of Ex 33:7b as it stands in the NRSV (as well as in other versions known to me) does not seem to fit well with the context. I would like to propose two theses concerning this text. First, from the point of view of the Hebrew syntax the conventional translation of Ex 33:7b is not the only one possible. Second, from the point of view of the inner logic of the pericope it is the least probable. I invite the reader of Exodus to look more closely at the syntax of Ex 33:7b.

496

Short Notes: Old Testament Studies

2. Syntactic patterns for linking participial conditional clauses with the main clause The participial clause in this verse (kol-m3baqqēš YHWH) clearly functions as a protasis of a conditional statement. Such clauses (called further ‘participial conditional clauses’) are quite widespread in Biblical Hebrew.1 The apodosis of conditional statements of this type conforms to the general rules governing the structure of conditional statements. In our case, for example, wayyiq¢ōl is used to depict a customary, habitual action. I propose to distinguish four patterns of linking a participial conditional clauses with the main clause (the apodosis). PATTERN A. A participial conditional clause is the subject of the main clause. Gen 4:14 ‘Anyone who meets me (kol-mo´3!î ) may kill me.’

This is characteristic, for example, for numerous laws of the penal code scattered throughout the Pentateuch. PATTERN B. A participial conditional clause is the object of the main clause. Gen 39:3 ‘Everything that he [ Joseph] did [kol !ašer-hû "ōśeh] the prosper in his hands.’

LORD

caused to

PATTERN C. A participial conditional clause supplies a referent for an anaphoric pronoun (a pronoun suffix) in the main clause; this is a case of casus pendens. 1 Sam 3:11 (= 2 Kings 21:12 = Jer 19:3) ‘Whoever hears of it (kol šōm3"ô)—his ears will tingle.’

PATTERN D. A participial conditional clause is neither a member of the main clause, nor supplies a referent for a pronoun in the main clause. The link between the participial conditional clause and the main clause should be established at the level of meaning. The resulting construction resembles to some extent the absolute participial constructions of Greek

1

See Gesenius’ Hebrew Grammar as edited and enlarged by E. Kautzsch (tr. G. W. Collins; rev. A. E. Cowley; Oxford: Clarendon, 1898), p. 521 (§ 159.2i); J. C. L. Gibson, Davidson’s Introductory Hebrew Grammar: Syntax (Edinburgh: T&T Clark, 1994), pp. 156–157 (§ 123, n. 3); Ronald J. Williams, Hebrew Syntax: An Outline (Toronto: University of Toronto, 2nd edn., 1976), p. 85 (§ 513).

M. Seleznev, Exodus 33.7 and Different Syntactic Patterns …

497

and Latin. This is an infrequent case, but there are enough examples to prove its existence. 1 Sam 2:13

rc;Bh; ' lVeb'K] @heKho ' r['n" ab;W jb'z≤ j'bze ú vyaiAlK; wdoyB: ] !yIN"Vhi A' vløv] glezM“ h' w' “ ‘Whoever offered sacrifice (kol !iš zō®ēaµ zē®aµ), the priest’s servant would come, while the meat was boiling, with a three-pronged fork in his hand …’ Num 35:30

j'xre ho A; ta, jx'ry] I !ydi[e ypil] vp,nA< hKemA' lK; ‘Whoever has killed a person (kol makkēh ne«eš), [someone] should kill the murderer, on the evidence of witnesses.’

In this example the subject of the main clause is not explicitly stated and we are to substitute an indefinite subject (‘someone’). Prov 11:27

.WNa,/bt] h[;r; vredwo “ ‘Whoever searches for evil (dōrēš rā"ā), [it] comes to him.’

The subject of the main clause is the object of the participial conditional clause. Jer 23:17

!k,l; hy Ñ and Ġ > ") and Greek Translations (2 Esdras and Judith). JBL 124:229–267. Streck, M. Das amurritische Onomastikon der altbabylonischen Zeit. Münster. Tal, A. The Language of the Targum of the Former Prophets and Its Position within the Aramaic Dialects. Tel-Aviv (in Hebrew). Wright, W. A Short History of Syriac Literature. London. Yamauchi, E. M. Mandaic Incantation Texts. New Haven. Yusmanov, N. V. Assyrian and its Writing (Assiriyskiy yazyk i ego pismo). Pismennost’ i revol’ucija. Moscow– Leningrad.

G. Goldenberg, A. Shisha-Halevy (eds.). Egyptian, Semitic and General Grammar. Studies in memory of H. J. Polotsky. Jerusalem, 2009. 501 p. The volume, dedicated to the memory of the great Semitist and Egyptologist Hans Jacob Polotsky, has been published as the proceedings of a workshop held at the Israel Academy of Sciences and Humanities in Jerusalem on 8–12 July 2001 in order to commemorate the tenth anniversary of Polotsky’s death. Both the scope and the quality of the contributions abundantly testify to the fact that Polotsky’s work is appreciated by many modern scholars, who continue to revitalize and update his theories in their own research. Some papers (as those by A. Niccacci and A. Shisha-Halevy) are directly inspired by Polotsky’s ideas and theories. At the same time, one cannot fail to observe that some of the key aspects of Polotsky’s scholarly activity are somewhat underrepresented in the volume, notably Coptic, Neo-Aramaic and Ethio-Semitic studies. The volume comprises 25 contributions in English and German. Most papers are not provided with summaries, which is partly compensated by the editors’s analytic survey in the introduction. E. Ullendorff ’s introductory article (“The Young (and Not So Young) Polotsky: Scholar and Teacher,” pp. 1–15) is a moving account of the author’s encounters with Polotsky, offering the reader an opportunity of catching some glimpses on the charismatic personality of the scholar. From the rest, four contributions deal with Egyptian, one with Coptic, five consider various aspects of Arabic linguistics and literature, three are dedicated to Modern Hebrew grammar, three are dealing with other Semitic languages (Akkadian, Biblical Hebrew, Tigrinya). One paper is sociolinguistic in its scope (it deals with introduction of literacy in several Gurage languages). Seven contributions treat various theoretical issues, mostly applied to Semitic linguistics. In his contribution “A Comparison between Classifier Languages and Classifier Script: The Case of Ancient Egyptian” (pp. 16–39) O. Goldwasser considers the intriguing question of whether the Egyptian pictograms known as determinatives can be regarded as graphemic classifiers comparable to the classifiers attested in a number of modern languages. He presents a coherent picture of the origin and employment of the determinatives in Egyptian, compares this evidence with that from the classi-

566

Reviews: Semitic Studies

fier languages (drawn from A. Aikhenvald’s well-known typological investigation of 2000) and comes to the conclusion that the Egyptian determinatives can and should be treated as classifiers, even if they might have been restricted to the written language and have never had any phonetic realization. However, linguistic elements and graphic signs can hardly be treated on the same level. Accordingly, the question of whether the determinatives in Egyptian texts were purely graphic or not—which remains unanswered in the paper and probably is not to be answered at all— becomes crucial for the correct interpretation of the evidence. If these elements were purely graphic (as Goldwasser is apparently inclined to think), it seems advisable to draw a proper distinction between classifiers in spoken languages and classifiers in script, in spite of their functional and structural similarities so convincingly demonstrated by the author. W. Schenkel (“Prädikatives und abstrakt-relativisches śÌm.n=f. Beobachtungen an den Verben II.gem. und ult. n im Korpus der Sargtexte,” pp. 40–60) presents the evidence on certain aspects of the Egyptian verbal system as reflected in the Coffin Texts. The investigation by H. Satzinger (“On Some Aspects of jw in Middle Egyptian,” pp. 61–69) discusses the use of the Egyptian particle jw and proposes to analyze it primarily as an existential verb (jw + noun = ‘there is N’), which was grammaticalized into a common means of introducing adverbial sentences.1 The author considers, furthermore, the possibilities of etymological derivation of jw from a verbal root. A. Niccacci’s paper “Polotsky’s Contribution to the Egyptian VerbSystem, with a Comparison to Biblical Hebrew” (pp. 401–465) is an overview of Polotsky’s investigations on the Egyptian verbal system in the context of the subsequent development of Egyptian linguistics. The author also introduces the results of his own investigation of the Biblical Hebrew verbal system based on Polotsky’s method, and then compares the Egyptian and the Biblical Hebrew systems. In his paper “Methodological Issues in the (Morpho)Phonological Description of Coptic” (pp. 70–91), W.-P. Funk addresses an issue familiar to every scholar of a dead language: whether (and how) it is possible to study its phonology and whether it is legitimate to speak of a “phonological” or “morphophonological” description if a language is known from written sources only. The author gives a positive answer to this question and tries to confirm his thesis by offering a list of phonological rules ac1

Cf. also Niccacci’s contribution discussed below, where this element is treated as a verb.

Studies in Memory of H. J. Polotsky … (M. Bulakh)

567

counting for the variation between eight Coptic dialects (thus updating F. Hintze’s seminal work of 1980). The contribution also offers a crossdialectal description of the status pronominalis in Coptic, including a set of phonological rules accompanying suffixation in various Coptic dialects. N. Wasserman’s contribution “The Modal Particle tuša in Old Babylonian” (pp. 149–168) thoroughly investigates the syntax and semantics of this particle as well as its relationship to other modal particles of Akkadian. The author rejects the widespread interpretation of tuša as an irrealis particle and defines it, instead, as “an epistemic modal particle denoting a false assumption … that proves to be refuted.” An extensive lists of examples illustrating the use of tuša in Old Babylonian is provided. D. Taube’s paper “The Passive Participle in Modern Hebrew” (pp. 317– 336) provides ample and interesting material on the use of the passive participle in Modern Hebrew. After a thorough discussion of its various semantic nuances, Taube establishes three major functions of the passive particple: actional (i. e., denoting a present action), resultative and stative. The study by T. Bar (“On Cleft Sentences in Contemporary Hebrew,” pp. 337–355) extensively describes the syntax and pragmatics of the cleft constructions in Modern Hebrew, providing numerous examples from a corpus of ca. 350 sentences. The author pays considerable attention to the difference in usage of the demonstrative pronoun ze and the personal pronoun hu in cleft sentences. The possibility of interpreting hu as a pronominal copula (different from the personal pronoun hu) is implicitly discarded in favor of an alternative way of analysis (for which see M. R. Domènech’s article discussed below). Nevertheless, such an approach could provide some promising avenues of comparison with cleft sentences in languages employing verbal copulas (such as English). Besides, the distinction between the pronominal hu and the copulative hu would provide an explanation for such constructions as ze !ani hu še- … (this-I-COP-who …)/!ani hu ze še- … (‘I-COP-this-who …’), both = ‘it is me who’: in these constructions, the copula not directly following the predicate (*!ani ze hu še- …) would be ungrammatical. Such an approach would also account for the non-occurrence of nouns as cleft constituents in “pure” cleft sentences, the relativized part of which is directly attached to the cleft constituent: !ani še-bati ‘It is I who entered.’ Indeed, these cleft sentences likely involve pronominal copulas rather than personal pronouns (COP:1.sg. REL-enter.1sg.). Since a noun cannot function as a copula, it requires an additional lexeme in order to establish a predicative relationship.

568

Reviews: Semitic Studies

T. Zewi’s article “Content Expressions in Biblical Hebrew” (pp. 302– 316) focuses on the dependent clauses known in the linguistic literature as content clauses (noun clauses, substantive clauses or complement clauses) as well as on functionally similar constructions (notably, the construct infinitives). The author shows that the grammatical difference between content and adjectival (relative) clauses is prominent in Biblical Hebrew of the Classical period: the conjunction kī is used exclusively with content clauses and, conversely, the conjunction !ăšer normally introduces adjectival clauses and only rarely appears in content clauses. As for Late Biblical Hebrew, the author claims that both content and adjectival clauses are usually introduced by !ăšer or še (originally, relative conjunctions). The use of kī in late Biblical texts is not discussed at all, although a few pertinent cases can be found even among the examples quoted in the article (such as w3-rā!ītī kī !ēn ¢ōb ‘I saw that there is nothing better …’ in Qoh 3:22). Accordingly, the author’s statement “In Late Biblical Hebrew the possibility of distinguishing between content clauses and adjectival clauses according to the type of introductory particle is largely eliminated” seems to require additional support. The concluding part of the article deals with clauses which behave syntactically as nomen rectum of the genitive construction (both in Biblical Hebrew and elsewhere in Semitic). In Zewi’s view, such dependent clauses are to be interpreted either as adjectival or as content clauses. The article “Three Related Analyses in Modern Hebrew Morphology” (pp. 277–301) by O. Schwarzwald deals with different types of word formation in Modern Hebrew. According to the author’s statistical analysis, the traditional Semitic pattern of morphological derivation—one involving consonantal roots and vocalic patterns (transfixes)—is still by far the most frequent one as long as one’s counts are restricted to the entries of a dictionary. However, real texts show a much lower percentage of lexemes produced after the root-and-pattern system. This observation testifies to a steady decrease in the use of the transfix derivation in Modern Hebrew, which gives way to alternative derivational means (such as multi-stem compounding). The second part of Schwarzwald’s contribution deals with the complicated question of distinguishing between affixal and pattern derivation (an issue also prominent in Sh. Isre’el’s contribution discussed below). Can a suffix or a prefix be part of a pattern or should such cases be treated as affixal derivation from a consonantal-vocalic stem? Both ways of analysis seem to be acceptable for the author, who postulates “an overlap between base + affix structures on the one hand and root + pat-

Studies in Memory of H. J. Polotsky … (M. Bulakh)

569

tern structures on the other.” In the last section of the article, the derivational type “root + pattern” in Modern Hebrew is analyzed, with a special emphasis on its productivity and the ways of formation of new consonantal verbal roots. Sh. Izre’el’s article “Constructive Constructions” (pp. 106–130) also addresses the problem of morphological devices of Semitic languages, focusing on the dichotomy between linear (affixal) vs. non-linear (pattern) morphemes and their distribution on the derivational-inflectional scale. Izre’el’s investigation is restricted to the verbal systems of two languages (Akkadian and Modern Hebrew), although a more extensive perusal of the evidence (first and foremost, that of Biblical Hebrew) would undoubtedly produce more interesting results. According to his description, vocalic patterns have inflectional functions only (primarily Time/Modus/ Aspect), whereas verbal prefixes are analyzed as means of derivation. The author is well aware of the difficulties entailed by this approach: thus, it remains to be explained why this or that derivational prefix (such as hi-fil or ni-fal) is associated with this or that vocalic pattern. In his opinion, this type of morphological variation belongs entirely to the surface level of morphology and can be easily discarded on a deeper level of analysis. As a result, the traditional meaning of the term binyan is rejected: according to Izre’el, it should be restricted to patterns (in his analysis, mostly inflectional morphemes indicating the TAM meanings) as opposed to verbal stems. In his article “Two Studies in Arabic Tamyīz” (pp. 197–219), R. Talmon concentrates on the syntactic phenomenon designated by the Arab grammarians as tamyīz (or tafsīr, or tabyīn) and labeled “accusative of limitation” (or restriction, or specification) in the Western grammars of Classical Arabic. In the first part of his contribution, Talmon surveys the emergence and the development of the concept of tamyīz in the Arabic grammatical tradition. The second part of the article offers a classification of the tamyīz constructions based on semantic and syntactic criteria. J. Blau’s article “Reconstruction of Neo-Arabic Dialectal Features from Middle Arabic Texts” (pp. 220–229) considers the potential of Middle Arabic for the reconstruction of the history of modern Arabic dialects. Blau brings into discussion some 20 lexical items found in the Middle Arabic corpus which are continued by Neo-Arabic, but either entirely missing from the Classical, or substantially different from the Classical prototypes phonologically and/or structurally. In the concluding part of his article, Blau offers an analysis of the adverbial suffix appearing as -ā

570

Reviews: Semitic Studies

in Middle Arabic texts. According to Blau, the history of the case ending an was different from that of -un and -in: whereas the latter markers were dropped at an early stage of the development of spoken Arabic, the ending -an shifted to -ā. It was much later that this ending disappeared in Neo-Arabic, being preserved only in a handful of fossilized forms. One may wonder, however, whether the shift from -an to -ā involved both the adverbial suffix and the accusative case marker (as Blau assumes) or the disappearance of the -an as the accusative case marker should be dated to the same stage as the loss of -un and -in. It seems that the evidence presented in Blau’s contribution does not exclude the latter possibility. Another contribution dealing with modern Arabic dialects is that by O. Jastrow (“The Arabic Dialects of the Carmel Coast,” pp. 230–238). The author publishes the materials of his 1999 field research season on a dialect group in the area of the Carmel Coast, bordering on the so-called “Triangle” in Central Israel. Jastrow discusses a few phonological and morphological features (including pronominal and verbal paradigms) of four dialects of the Carmel Coast, recorded by him in the villages Fradīs, Mifǧar and Ǧisir izZarga (the latter in two varieties, labeled by Jastrow as Ǧisir I and Ǧisir II). These features are systematically compared to the evidence of the dialects of the “Triangle” (Mu¬alla¬). The main innovation postulated by the author for the Carmel Coast dialects is the introduction of the long ā in the 3 pl. f. forms of the perfect from hollow roots: kālin/gālin ‘they saidfem.’ as opposed to kulin/kalin in the Mu¬alla¬ and qulna in Classical Arabic. In his paper “Some Remarks on Maltese Prepositions of Italian Origin” (pp. 270–276), R. Saari describes four Italian loanwords among Maltese prepositions: favur ‘in favor of ’ (It. favore), kontra ‘against’ (It. contra), rigward ‘considering’ (It. riguardo) and skond ‘according to’ (It. secondo). Phonological, morphosyntactic and semantic features of these prepositions are analyzed upon the background of their Italian etymons. R. Henkin’s study “How Interdialectal is Peripheral Oral Bedouin Poetry” (pp. 239–269) focuses on the phenomenon of interdialect, usually thought to influence the language of poetic compositions in the Arabic-speaking Bedouin societies. In his analysis, based on textual evidence recorded in different areas and at different periods, Henkin considers both phonological and morphological features (short vowels in open syllables; nunation; verbal suffixes of 2 f. sg. and m. pl.). In his opinion, distribution of interdialectal features in Bedouin poetry is uneven, the major conditioning factors being geographical (distance from Najd, the core area of the poetic interdialect), temporal, socio-political and cultural.

Studies in Memory of H. J. Polotsky … (M. Bulakh)

571

R. Voigt’s contribution “Südtigrinische Dialekte” (pp. 169–183) is a study in Tigrinya dialectology. The author gives a survey of the dialect of May-Ìäw in Tigray (Ethiopia) on the basis of his field records in this city (one week in 1998, two weeks in 2001), see further Voigt 2006. The phonological section examines the process of spirantization of velars (which, in this dialect, undergo spirantization in all positions when nongeminated), whereas the behavior of the labial b is discussed only briefly. The morphological section is restricted to the independent personal pronouns and is diachronically oriented. Of interest is Voigt’s new etymology of the standard Tigrinya and Amharic personal pronouns. In both languages (as in some other Ethio-Semitic languages), the personal pronouns are formed by means of attaching the possessive suffixes to a nominal base (n3ss- in Tigrinya and 3rs-/3ss- in Amharic): 3 masc. sg. Tigrinya n3ss-u vs. Amharic 3rs-u/3ss-u. Instead of the traditionally reconstructed *nafs- ‘soul’ (for Tigrinya) and *r3!s- ‘head’ (for Amharic), Voigt postulates *karś- ‘belly’ as the source-word for the base of the personal pronoun in both languages. The article includes a brief overview of Tigrinya sochiolinguistics as well as a summary on previous work on Tigrinya dialectology (to which a few recent studies, such as Fitzgerald 2006, could be added). In his contribution “From Speech to Writing in Gurage-Land” (pp. 184–196), G. Goldenberg presents his observations on the emergence of written culture in previously unwritten languages of Gurageland in Ethiopia. The practice of writing in local vernaculars was a spontaneous process that started on a large scale in the last decade of the 20th century. The adaptation of the Amharic writing system to the needs of the Gurage languages, developed with no supervision from either government or scholars, is judged by the author as generally successful. The paper discusses some features of the Gurage languages as reflected in the newly emerging writing systems, as well as the lexical and syntactic differences between written and spoken language. Copious text samples in Ethiopic script and transcription are mostly drawn from K3stane, a Gurage language also known as Soddo or Aymälläl. A. Shisha-Halevy’s article “On Conversion, Clause Ordination and Related Notions” (pp. 92–105) focuses on the subordinate clause and the related concept of “conversion” (i. e., transformation of a clause into a dependent constituent whose syntactical function is that of a verbal complement, adjectival or adverbial modifier, etc.). The latter term was first introduced by Polotsky for Coptic, and afterwards applied by him and his

572

Reviews: Semitic Studies

followers to other languages. A separate section deals with the evolution of Polotsky’s understanding of “converters,” followed by the author’s analysis of the evidence from various languages for which the presence of “converters” has been postulated (Coptic, Modern Greek, Welsh, Old Nubian, Egyptian). Shisha-Halevy concludes his study with a critical evaluation of the generally accepted approach to this problem, which claims the primacy of the verbal forms and the secondary status of the “converted” forms. His alternative analysis transfers the discussion from the domain of morphology and syntax to that of text linguistics. E. Cohen’s paper “Nexus and Nexus Focusing” (pp. 131–148) explores the focus falling on the predicative link between the main constituents of the sentence. He advocates the use of the term “nexus focusing,” to be preferred to the more widespread label “focus on the truth value” (and similar). The advantage of the expression “nexus focusing” is that it covers even those sentences with marked focus which have no truth value at all (for instance, sentences with modal verbs or verbs that are not in the indicative mood). Having spent some attention to the nexus focusing in English, Cohen proceeds to the description of the same phenomenon in Old Babylonian Akkadian. According to Cohen, nexus focusing in Old Babylonian is usually rendered by the asseverative paradigm and expresses insistence, oath and concession. An interesting feature is the possibility to combine a nexus focus with a contrastive focus (expressed by the encliting particle -ma). The paper by M. R. Domènech (“ ‘Kopula’: Ein ‘zur rechten Zeit gestelltes Wort’?”, pp. 356–362) is a short theoretic discussion of the legitimateness of the term “copula” as applied to the “pronominal copulas” of some Semitic languages. The author argues against the copulative interpretation and suggests an alternative analysis which treats the sentences with pronominal copulas as cleft sentences (or, rather, as sentences with topicalization since no relativization is employed, which is so typical for clefts). S. Hopkins’ article “ ‘That Monster of a Man’ and the Emotive Genitive” (pp. 363–389) deals with the genitive constructions in which the regens semantically qualifies the rectum (labeled by Hopkins “emotive genitive”). The author offers a detailed analysis of this construction in English, with due attention paid to semantics, pragmatics, syntax and even prosodic characteristics of such constructions. Hopkins further analyses the evidence from several other European languages as well as from Turkish and a few Semitic languages. As far as the Semitic-speaking do-

Studies in Memory of H. J. Polotsky … (M. Bulakh)

573

main is concerned, Hopkins’ most solid examples come from modern languages (Neo-Aramaic, Amharic, Maghreb Arabic, Maltese, Modern Hebrew), where the emergence of this construction can be ascribed to foreign influence. The evidence from classical Semitic languages is sparse and partly not very convincing. Thus, Hopkins’ single example from Geez is µassāwe masiµ ‘Antichrist’ (lit. ‘false of Messiah’). This combination seems to be a calque of Greek ψευδZχριστοa (for a similar, albeit structurally not identical calque, cf. r3tu"a hāymānot lit. ‘straight of faith,’ reproducing ὀρθZδοξοa). M. Erdal’s typological study (“First and Second Person Nominal Subject,” pp. 390–400) deals with a cross-linguistically well-attested, but, nevertheless, intriguing feature, viz. the occurrence of nominal subjects with verbs marked for 1st or 2nd person. Erdal’s investigation is based on the evidence from a variety of genealogically unrelated languages, such as Old Turkic, Old Uygur, Ottoman Turkish, Pre-Classical Mongolian, Classical Greek, Latin, Spanish, Biblical Hebrew (represented by one example from Judges 5:7) and Modern Hebrew. The author makes an appropriate distinction between nominal subjects and nouns which are appositions to pronominal subjects, arguing that all of his examples display true nominal subjects rather than appositions to covert pronouns. Still, it remains to be answered to what extent this construction is preferred by pro-drop languages, for which the appositional interpretation cannot be absolutely ruled out. The volume is concluded by P. I. Kirtchuk-Halevi’s ambitious article “Language: A Typological, Functional, Cognitive, Biological and Evolutionary Approach” (pp. 466–501). The author proclaims a new era of linguistics which, in his opinion, is characterized by a special attention to the emergence of the human language, and, as a result, to such “primordial” elements as intonation and deictics. Strangely enough, the article has no references to numerous special studies which not only deal with this issue, but directly address most of the topics raised in Kirtchuk-Halevi’s article (see, among many others, Knight et al. 2000; Givón–Malle 2002; Tallerman 2005). The volume is a carefully edited and beautifully printed book with very few misprints. As such, it definitely meets the high standards expected of a book commemorating such an outstanding Orientalist as Hans Jacob Polotsky. M. Bulakh

574

Reviews: Semitic Studies

References Aikhenvald 2000 Fitzgerald 2006

Givón–Malle 2002 Hintze 1980 Knight et al. 2000

Tallerman 2005 Voigt 2006

Aikhenvald, A. Y. Classifiers. A Typology of Noun Categorization Devices. Oxford. Fitzgerald, C. M. More on Phonological Variation in Tigrinya. Uhlig, S. (ed.). Proceedings of the XVth International Conference of Ethiopian Studies. Hamburg July 20–25, 2003. Wiesbaden. Pp. 763–768. Givón, T.; Malle, B. F. (eds.). The Evolution of Language out of Pre-language. Amsterdam–Philadelphia. Hintze, F. Zur koptischen Phonologie. Enchoria 10:23–91. Knight, Chr.; Studdert-Kennedy, M. G.; Hurford, J. R. (eds.). The Evolutionary Emergence of Language: Social Function and the Origins of Linguistic Form. Cambridge. Tallerman, M. (ed.). Language Origins: Perspectives on Evolution. Oxford. Voigt, R. M. Südtigrinische Dialekte: das einfache und zusammengesetzte Präsens im Dialekt von May-Ìäw (Tigray). Uhlig, S. (ed.). Proceedings of the XVth International Conference of Ethiopian Studies. Hamburg July 20–25, 2003. Wiesbaden. Pp. 893–898.

C. A. Ciancaglini. Iranian Loanwords in Syriac. Wiesbaden: Dr. Ludwig Reichert Verlag, 2008 (Beiträge zur Iranistik 28). XLVII + 315 pp. The book under review was written by a scholar whose principal field is Indo-European (and in particular Iranian) comparative philology rather than Semitics, and this does not happen on a regular basis: most studies of contacts between Semitic and non-Semitic languages were done by Semitists. In the past three years, the book has become a standard tool for the students of Aramaic, and for a good reason. It is to be hoped that Claudia Ciancaglini will study Iranian loanwords in the remaining Middle Aramaic corpora and produce their equally thorough descriptions. The high value of Iranian lexical material in semitischen Nebenüberlieferungen for Iranian philology is well-known. As far as Aramaic studies go, when all the lexical Iranisms in Middle Aramaic are gathered and analyzed, these data will turn out indispensable for scholars of Aramaic: they will permit deeper insights into the history of Aramaic and its early medieval literary varieties, and into the history and culture of Aramaicspeaking communities. One remarkable fact about Ciancaglini’s contribution is that slightly more than the third of the volume (if one does not count Bibliography and Word Index) is occupied by “Part One: Linguistic and Historical Survey” that reflects the author’s interest in the theory of languages’ contacts and in the non-lexical features of Aramaic that may be due to its contacts with Iranian languages. Some of Ciancaglini’s statements in this chapter invite criticism. It is to be noted that part of the assumed errors originate not with Ciancaglini but in the Aramaistic literature she had to use and, being a non-Semitist, to regard with a certain measure of trust. The intended readership of the introductory chapter is in particular students of Iranian languages (cf. Reimond 2009:336) who are not supposed to control Aramaic data and be able to form independent judgements on matters Aramaic. Therefore it is worthwhile to register possible errors and/or to indicate statements that are less communis opinio than Ciancaglini was lead to believe. For the convenience of the user, our observations will follow wherever possible the order of matters in the introductory chapter.

576

Reviews: Semitic Studies

P. 1. “Afterwards the Christian kingdom of Edessa lost its independence and became a Roman colony (213–214 A. D.).” At the time, Edessa was hardly a Christian kingdom. In the Chronicle of Edessa, while describing the flooding of the city in A. D. 201, the annalist says in particular that the waters “damaged the nave of the church of the Christians ("rttā da-−res¢yānē)” (Segal 1970:24). This means the Christianity was not then the official religion of the kingdom (and see Segal 1970 for all the available information on the early days of Christianity in Edessa). P. 3. Ciancaglini notes that Turoyo “also shows features of western Neo-Aramaic dialects,” with reference to Macuch 1976:9 and n. 21. Genealogically, Turoyo is an Eastern Aramaic language simply because it developed out of another Eastern Aramaic language in historical times (like, say, both Italian and Romanian developed out of Latin beyond doubt, and this fact cannot be undone by whatever features they display now). Structurally, Turoyo has no conspicuously Western features, in particular less so than Neo-Mandaic: note that even Macuch or Tsereteli did not hold Neo-Mandaic to be a central or otherwise “intermediate” Neo-Aramaic idiom. P. 7. “The number of Greek loanwords in Syriac is comparable to the proportion of Latin and French loanwords in English; they are often deeply integrated and productive in the receiving language” (with a reference to Schall 1960). This is too high an estimation. R. Voigt’s index to Schall 19601 has about 440 words. Schall 1960 does not strive to be exhaustive. Using the glossary in Lehto 2003, one of the present reviewers found that Aphrahat’s Demonstrations have some 3100 lexemes. In this sufficiently vast and representative early corpus, there are a bit more than 100 Greek loans2 and about 20 Iranian loans (Loesov 2009:622). Among the Greek loans, there are twenty verbs; all of them are denominative by way of inner-Syriac derivation, complying with Ciancaglini’s view that verbs are not usually derived immediately from foreign substantives: these substantives have to be borrowed first. Claudia Ciancaglini informs S. Loesov in a p. c. (June 22, 2011): “Direi che Afraate non è un buon esempio, visto che Brock dice espressamente che lui ed Efrem sono i rappresentanti degli autori siriaci del IV secolo non ancora ellenizzati, mentre la grande massa dei prestiti dal greco si vede negli autori successivi, dal V al VI secolo, e ancora di più nei secoli dopo, quando cominciano a presentarsi anche i calchi, sia lessicali 1 2

Voigt 1998. The token frequencies of each of these loaned lexemes were not counted.

C. A. Ciancaglini. Iranian Loanwords … (L. Kogan, S. Loesov)

577

che morfosintattici (S. Brock. Syriac Perspectives on Late Antiquity. London, 1984. II 1; III 75, 84 ss.; IV 80 ss.; V 107 ss.). E forse anche Schall non va bene, perché si occupa solo dei grecismi nella letteratura siriaca più antica e non di traduzione, cioè quella che ha meno prestiti dal greco (cf. Brock 1984 IV 90).” — All this is quite true, but the main thing one wants to know is what lexical loans from Greek were doing in the spoken vernacular Syriac, and not in relatively late literary works. Note that regarding the older stratum of Iranian loanwords Ciancaglini observes that “these old borrowings … had long been integrated into the Syriac lexicon and probably no longer sounded as foreign words” (p. 27), i. e. some of them became part and parcel of the vernacular Syriac vocabulary. The lexicon of Aphrahat’s theological opera is of course far removed from that of everyday speech, but at least it has less non-integrated graecisms than the later literature. Let us make a note by way of an inner-Middle-Aramaic comparison of graecisms.3 It has recently been found that no less than 20% of entries in M. Sokoloff ’s DJPA are Greek loanwords, many of them look like part of the everyday lexicon, e. g. qdrws ‘cedar’ < kšdroj, !wrz ‘rice’ < Ôruza, !ys¢pnyny ‘carrot’ < staful‹noj, prs£h ‘peach’ < persikÒn, ¢pys ‘carpet’ < t£phj, zoni ‘belt’ < zènh, zog ‛couple’ < zugÒn (Nemirovskaya 2009:558– 560). A historian would like to know as much as possible about AramaicGreek bilingualism in the early centuries A. D. Syria, Palestine and Mesopotamia, so a methodologically thought-through comparative analysis of the available data will have to bring results important for our understanding of history. P. 11. “Notwithstanding the use of different linguistic terms, the scholars agree that the Aramaic of the Achaemenid period is not a homogeneous language: different texts show different dialect features. Consequently, Official Aramaic is simply a chronological label indicating Aramaic language(s) used in documents dating back to the Achaemenid period. This conclusion is relevant for the question of the relationship between Syriac and Official Aramaic, and will be dealt with later.” A noncommittal talk about dialects of Achaemenid Aramaic is indeed common in the literature (Ciancaglini refers to several well-known studies by S. Kaufman, F. Rosenthal, M. Folmer, J. Greenfield, E. Kutscher), but their existence has been never proven. Since Ciancaglini does her very best to take our Aramaic studies seriously, she arrives at a logically com-

3

For bibliography, see now SL:XIX. There is no definitive study as yet.

578

Reviews: Semitic Studies

pelling conclusion that “Official Aramaic” has to be but an umbrella notion for the Aramaic corpus of the Achaemenid times.4 This is of course too unequivocal a conclusion for the school of thought represented for Ciancaglini by the above names, because Ciancaglini’s statement is logical, transparent, and controllable. The bad news is that one cannot have “Official Aramaic” as both the main official written language of the Achaemenid Empire (a proposition that nobody doubts) and “simply a chronological label.” Claudia Ciancaglini writes in a p. c. to S. Loesov (June 11, 2011): “Ho letto e confrontato quegli autori dopo che tu mi avevi messo il dubbio sul rapporto tra aramaico imperiale e siriaco, e mi è sembrato che fossero più o meno concordi nel considerare l’aramaico imperiale un’entità linguistica complessa e disomogenea. Inoltre, mi aveva molto convinta l’affermazione di Kaufman che ho riportato per intero a p. 26.” The cited statement (Kaufman 1974:159f.) runs as follows: “Thus while the characteristic traits of Imperial Aramaic are eastern, it is not Eastern or Mesopotamian Aramaic; nor is it Western or Syrian Aramaic. It must be something in-between. Nor is it necessarily artificial in origin. That is to say, it is reasonable that this dialect mixture arose in the process of normal intercourse between dialect groups and quite possibly even became a native language for some. Thus, quite naturally, each of the characteristics of Imperial Aramaic spread differently through the Aramaic speech community.” It is difficult to say less in so many words. The main question is now, as it has been during the last forty years, whether Imperial Aramaic had a say in the formation of the Middle Eastern Aramaic, simply put: whether it is the ancestor of the Middle Eastern Aramaic. In the sentence Ciancaglini omitted from her quotation, Stephen Kaufman says: “Certainly at Elephantine it is difficult to imagine that the private letters on ostraca, which have free word order, were written in a dialect whose syntax was significantly different from the writer’s native speech.” This might well be true, but note that very few Aramaic texts of any dialect or genre have come down to us from the post-Achaemenid Egypt.5 The Greek conquerors did no serious harm to the local popula4

Note, once more, her very clear statement about “the chronological and non-linguistic nature of the label ‘Official Aramaic’ ” (p. 26). 5 For a short recent state-of-art on the role of Aramaic in the early Ptolemaic Egypt, see Stadel 2009. Most scholars (as e. g. Jan Joosten) believe that Aramaic stopped to be spoken in Egypt soon after Alexander the Great.

C. A. Ciancaglini. Iranian Loanwords … (L. Kogan, S. Loesov)

579

tion, so this evidence (if it is not a result of archeological chance) shows that writers of private letters and contracts in Ptolemaic Egypt did not see Imperial Aramaic as an adequate linguistic vehicle anymore, while it turned out to be quite appropriate in the (almost certainly Arabic-speaking) Nabataean kingdom, as well as in the Aramaic-speaking Palestine, etc. This means that Imperial Aramaic was a written code with a specific cultural Stellenwert (think about reflexes of the Achaemenid norm in the Mandaic orthography). Ciancaglini dubs various features of Syriac relevant to the study of Iranian influence on pre-modern Aramaic as “inheritance of Official Aramaic,” following the above school of thought in the English-speaking Aramaic studies. Practically, two questions are at stake: the early history of qtīl l- and those pre-middle-Iranian loanwords in Syriac that are attested in Imperial Aramaic as well. It is well-known that many of pre-middle-Iranian loans of Syriac are attested in JBA (Telegdi 1935) and Mandaic, less of them appear in JPA, Samaritan Aramaic and CPA. To move the discussion any further, it is necessary in the first place to produce a synopsis of Iranisms in Imperial Aramaic6 and the six Middle Aramaic corpora and to see what happens. So we have to content ourselves with a discussion of qtīl l- (pp. 31–37 in the book under review), which is the burning question in the history of the Aramaic verb. Is it possible to detect, with Ciancaglini and much of previous scholarship, the Iranian trace in the emergence of this verb form? In order to discuss the problem “qtīl l- in Aramaic Studies,” we have to distinguish three sub-topics: (1) *qtīl l- in Old Aramaic; (2) *qtīl l- in Middle Eastern Aramaic; (3) *qtīl l- in Eastern Neo-Aramaic. As for (1), in spite of the impression our secondary literature creates in the mind of an outsider,7 the “ergative” qtīl l- is not attested in Old Aramaic beyond doubt, the discussion has been revolving for decades around one (probably lexicalized) instance represented by two tokens.8 So we will not discuss this point any more. 6

For the list of Old Persian loans in Egyptian Aramaic, see Muraoka–Porten 2003:370–373 (72 items). 7 Ciancaglini observes: “An important morphological feature borrowed from Old Persian into Official Aramaic, which then deeply modified the verbal system of eastern Aramaic, is the q¢yl l- construction; it follows the model of the Old Persian manā kartam ‘I made’ construction” (p. 31). 8 They are as follows: šmy" ly ‘I have heard’ (TADAE A6.10:3), šmy" ln ‘we had heard’ (TADAE A3.3:13). See further Loesov 2006:632 for references to previous

580

Reviews: Semitic Studies

As for (2), *qtīl l- in Middle Eastern Aramaic (hereafter MEA), its being a morphological loan translation of the Middle Persian man kard construction is an unlikely solution (pace Kutscher 1969 and other scholars more or less followed by Ciancaglini). In her theoretical discussion of the morphological borrowing, Ciancaglini says among other things: “The socio-cultural conditions in which a morphological borrowing is possible are: the contact between the two languages must be deep and prolonged in time; the source language must exert a strong cultural pressure on the other one, playing the role of the prestige language; finally, there must be a large number of bilinguals. The main structural constraints are: the morphological borrowing is always preceded by a great number of lexical borrowings …” (p. 29f., italics added). Ciancaglini’s glossary of Iranian loanwords in Syriac has some 760 words. Is this number of borrowings great enough in the sense of her theory? Compare an important piece of information about the verb in Turoyo. According to Ontikov 2011, the corpus of Turoyo verb roots dealt with in Ritter 1990 (the book mentions most of the verb roots of this language) has 603 roots of Aramaic origin, 706 Arabic roots, 98 Kurdish roots, 27 Turkish roots, and 262 roots with yet unidentified etymology. Aramaic roots are a minority both in lexicon and available text specimens, yet all derivational and inflectional patterns of the Turoyo verb are impeccably Aramaic, i. e. there is not a single morphologically undigested Arabism (four-consonant roots, e. g. borrowings of the Arabic VIII-stem verbs, acquire the shape of the Turoyo D-stem, exactly as in Syriac). The state of nominal inflection is different, in particular there are many borrowed adjectives that are indeclinable (Jastrow 2002:98). Most speakers of Turoyo (as well as those of other Neo-Aramaic languages) have been bilinguals for centuries, perhaps for more than a millennium, yet nothing dramatic has happened to the Turoyo verb on account of this bilingualism. The difficulty about this kind of comparisons is that Turoyo is a spoken language (with a good corpus that was formed due to field studies, although the language has no written tradition), while MEA is a discussions. It is generally believed that the letter of the Persian prince Arsames (TADAE A6.10) was written from outside Egypt, i. e. probably from Babylon or Susa. Since Arsames hardly spoke Imperial Aramaic, it lies near at hand to suggest that the šmy" ly is in this case due to the Eastern Aramaic mother tongue of the scribe. It could also be a morphosyntactic loan translation from Old Persian, and more things. This problem is the Shibbolet-Incident of Aramaic studies.

C. A. Ciancaglini. Iranian Loanwords … (L. Kogan, S. Loesov)

581

sum total of various “streams of tradition” in the sense of Leo Oppenheim, i. e. all our data come from copies (= and not autographs/protographs) of literary texts. According to Stephen Kaufman’s p. c. to S. Loesov (June 26, 2011), in the extant Syriac corpus there are “between 20 to 25 thousand , but it is hard to estimate because of what one counts as freely formed. there are between 5 and 7 thousand .”9 In the case of Syriac, morphosyntax (including productive analytical forms) must have been common for all functional spheres, from everyday exchanges (unavailable to us) to scholarly works and poetry, while lexicon was genre-dependent: written texts could harbor words unknown to most speakers of Syriac. All in all, the hypothetical number of Middle Persian lexical borrowings in spoken Syriac does not seem to be big enough to be able to usher important morphosyntactic calques. Is it true that among the speakers of all three known MEA languages there were lots of Aramaic-Persian bilinguals? As far as Upper Mesopotamia and Syria go, the Aramaic-Persian bilingualism was hardly widespread outside certain circles of educated native speakers of Syriac (this is the opinion of Ciancaglini, cf. pp. 14–20). Consider also the findings of J. B. Segal: “In their knowledge of Persian the Christians of Nisibis and the north were backward, perhaps on account of their Syriac-speaking university. … Even as late as the seventh century there were leading Christian theologians at Nisibis who could converse only in Syriac” (Segal 1955:135f.).10 To the west of Nisibis (i. e. outside the Sassanid Empire, e. g. in Edessa), speakers of Aramaic had been probably never exposed to heavy contacts with Iranian in pre-Kurdish times. Before the Arab conquest of the Byzantine territory of Northern Mesopotamia, speakers of Aramaic were a majority in a part of this region, so most of them were probably monolinguals. It looks like the linguistic situation of JBA and Mandaic could have been different, because the speakers of both idioms lived under political control of Arsacids and Sassanids most of the time. Segal 1965:33 makes the following suggestion: “Aramaic was a lingua franca throughout Mesopotamia. In the north, however, it was well-nigh universal; in the south it was only the speech of minorities, Jewish and Christian, and of literati, for there the dominant language was Persian.” For the present reviewers, 9 The difference is of course due to the fact the Syriac corpus is larger than the rest of Middle Aramaic corpora. 10 We thank Nikolai Seleznyov for alerting us about this study.

582

Reviews: Semitic Studies

the key word here is “the speech of minorities,” because language minority groups tend to be bilingual to some extent unless they live isolated lives in a kind of ghetto.11 Yet note that Telegdi 1935, still the most important reference work on Iranian loanwords in JBA, arrives at the following conclusion regarding the Iranian loans in Talmud Bavli: “En comparant ces deux séries on arrive à des resultats d’un très grand intérêt. On constate, d’abord, que les mots de la première série, c’est-àdire les emprunts qui datent sûrement de la première moitié de la periode arsacide ou de l’époque achéménide, sont beaucoup plus nombreux que ceux qui constituent la série des emprunts plus récents. On observe aussi que les mots de la première série jouent en général un rôle plus important dans le vocabulaire des divers dialectes des Juifs, ils désignent des notions plus élémentaires, ils sont employés plus souvent, ils sentent moins l’étranger que notre série 2. En d’autres termes: la partie la plus grande et la plus importante des imprunts iraniens, qu’on trouve dans les divers dialectes juifs, ont été faits avant notre ère” (Telegdi 1935:220). By contrast, Ciancaglini estimates only 90 out of her 760 items as “entrati nell’aramaico prima dell’emergere del siriaco” (a p. c. to S. Loesov). This corresponds to S. Telegdi’s “faits avant notre ère,” while both judgements rely on roughly the same criteria based on the historical phonology of Iranian. Thus these data do not confirm a suggestion that a noticeable part of JBA speakers were bilinguals. It is not self-evident that at the beginning of the Christian era the etymological structure of the Middle Persian man kard was transparent for the assumed bilingual speakers of Aramaic to the degree that they were able to loan-translate it idiomatically into their respective vernaculars. There are no data about a MEA koine or intensive contacts among the members of the three religious (and linguistic) Aramaic communities. Rebus sic stantibus, common sense suggests that independent parallel development (not induced by a contact with Middle Persian) in all known MEA varieties is a sensible solution. This parallel development must have had a shared structural motivation, a kind of inner-language pressure (see Loesov 2011). As for (3), *qtīl l- in Eastern Neo-Aramaic (hereafter ENA), Ciancaglini follows those scholars who attribute it to the influence of Kurdish. It is known for a fact that many (and perhaps most) ENA speakers from 11 What was the majority language in the region of Ctesiphon (roughly, the vicinity of today’s Baghdad) in the years 400–500 A. D., Aramaic or Middle Persian? The literature does not give a clear answer.

C. A. Ciancaglini. Iranian Loanwords … (L. Kogan, S. Loesov)

583

Kurdistan used to speak some Kurdish, yet it is not clear whether the etymologically split-ergativity structure of the Kurdish verb was still visible to the speakers of Aramaic when they came in contact with Kurdish. Thus it is not evident that we have to posit three epochs of Iranian influence to account for *qtīl l- in ENA. We do not need the Iranian influence to explain the evidence.12 The sad fact is that speaking about qtīl l- in Syriac we still most of the time manipulate the handful of examples that were collected and cited by Nöldeke 130 years ago. Kutscher 1969 added six more tokens he had found in the Bible (mostly in the NT), Ciancaglini discusses them on p. 35. Peshitta is important for an obvious reason: if qtīl l- translates a transitive verb in the active voice, one is on the right track. When we have all the pertinent examples from the Bible, we will doubtless be able to learn something new about what qtīl l- was doing in Syriac. On p. 32, Ciancaglini says that the MEA “q¢yl l- construction is employed to express the perfect of transitive verbs (but often of intransitive ones too …), and consists of a past participle in the absolute state (showing non-agreement in gender and number with the patiens) and a pronominal agent preceded by the preposition l- ‘to, belonging to’ ” (italics added). As a matter of fact qtīl l- of intransitive verbs is attested both in Syriac and Mandaic, but it seems to be extremely rare, the few examples that are now known stem from the two grammars of Nöldeke. In MEA qtīl l- of transitives usually does agree with the patient in gender and number, e. g. da-hwaw µšī®īn l-−on b-µubbā yattīrā ‘[we ask you] that youm. pl. esteem themm. pl. in great love’ (Peshitta 1Th 5:13).13 Once more, Nöldeke cites a few (probably rare) Syriac and Mandaic examples where qtīl does not agree with the exponent of patient. P. 33. “Synchronically, in Eastern Neo-Aramaic dialects the old ergative construction q¢yl l- is clearly perceived as an active one, as it is also confirmed by the fact that the participle q¢yl does not agree in gender and number with the patient (the old grammatical subject) and thus may govern a direct object …” Generally speaking, the lack of object agreement 12

See Pennacchietti 1988 (and especially pp. 105ff.) for a review of virtually all logically possible hypotheses, all of which had been formulated in the literature prior to 1988. Nothing new has happened since then. 13 kaˆ ¹ge‹sqai aÙtoÝj Øperekperissoà ™n ¢g£pV, the deponent present infinitive of ¹gšomai (lit. ‘lead, guide’) has a direct object ‘them’ (aÙtoÝj), its semantic subject (‘you’) is supplied from the left-hand context. The Syriac translation is non-slavish both lexically and grammatically, and extremely faithful to the sense. Cf. BDAG s. v.: “In 1Th 5:13 there emerges for ¹. the sense esteem, respect.”

584

Reviews: Semitic Studies

signals that the ergative alignment has 100% converted into the accusative one, yet this is not completely the case with NENA, cf. Khan’s article “Ergativity in the North Eastern Neo-Aramaic Dialects,” where “ergativity” is tantamount to “object agreement” (Khan 2007).14 It looks like Ciancaglini was misled by her reference (Pennacchietti 1994:260, n. 4), where it is indeed said “senza però accordarsi in genere e numero.” The author has to be congratulated for a useful survey of the phonological phenomena pertinent to the reception of Iranial loanwords in Syriac (pp. 63–90). Of great interest are, in particular, Ciancaglini’s observations on the different reflexes of Iranian x, which confirm Telegdi’s brilliant discovery made more than 70 years ago (1935:197–202): in the older stratum of loanwords, x is represented by µ, whereas in the younger one k is found instead. This opposition is well known to be of great relevance for two key problems in the historical phonology of Aramaic, viz. the loss of uvulars and the emergence of spirantization. Note that this aspect of the dilemma went unnoticed in Richard Steiner’s excellent account of 2005. There follow a few remarks on individual etymologies. P. 120. The author’s treatment of
 (LSyr 98, SL 115) is unclear to us. Brockelmann’s translation ‘Pascha’ does not imply that the Syriac word was understood by him as ‘Easter’ or ‘Passover,’ what is meant is certainly a well-known Middle Eastern administrative title. A borrowing from ‫ باشا‬is, therefore, quite obvious and it is hard to say why an Arabic origin (so Sokoloff) should be preferred to the Neo-Persian one. P. 169. An Iranian origin of  ‘basket’ is rather unlikely in view of the presence of zabbīlu already in Neo-Babylonian Akkadian. The Akkadian word has been generally considered an Aramaism (CAD Z 6, AHw. 1501, DNWSI 301), no doubt because of its exclusively late attestation. Kaufman (1974:111) is nevertheless correct to observe that there is no genuinely Aramaic root from which such a noun could be produced, whereas Akk. zabālu ‘to carry’ is well attested throughout the history of Akkadian. It is thus tempting to suppose that we are faced with a late— but genuinely Akkadian—nominal formation subsequently borrowed into Aramaic and, then, spreading to Arabic and Persian (cf. SL 362). Still one cannot exclude that the nominal pattern C1aC2C2īC3, virtually unknown in Akkadian, is due to Aramaic influence.

14

Some remnants of object agreement in Turoyo are well-known as well.

C. A. Ciancaglini. Iranian Loanwords … (L. Kogan, S. Loesov)

585

P. 192. The author is undoubtedly correct to reject the Iranian etymology of
  ‘cake,’ but does not mention the fact that this word is actually borrowed from Akk. kukku, as duly recognized by Kaufman (1974:65, cf. now SL 605 and already LSyr. 326). P. 240. While it is not impossible that
 ‘purslane’ entered Syriac through an Iranian intermediary, the etymological priority of Akk. pappar¶û (CAD P 109, AHw. 824), in its turn possibly borrowed from Sumerian babbar.¶i (PSD B 31), must be acknowledged as certain in view of its broad attestation already in Old Babylonian (cf. now SL 1248). P. 258. As seen already by von Soden,
 ‘hematite’ is inseparable from Akk. šadânu with the same meaning, well attested since Old Assyrian and Old Babylonian on (CAD Š1 36, AHw. 1123).15 This circumstance does not necessarily exclude the possibility of an Iranian intermediary for the Syriac word, but almost certainly means that the emergence of Middle Persian šāh-dānag is due to meta-analysis by popular etymology. L. Kogan, S. Loesov References Jastrow 2002 Kaufman 1974 Khan 2007

Kutscher 1969

Lehto 2003

Loesov 2006 Loesov 2009

Jastrow O. Lehrbuch der àuroyo-Sprache. Wiesbaden. Kaufman, S. The Akkadian Influences on Aramaic. Chicago. Khan, G. Ergativity in the North Eastern Neo-Aramaic Dialects. Bar, T.; Cohen, E. (eds.). Studies in Semitic and General Linguistics in Honor of Gideon Goldenberg. Münster. Pp. 147–157. Kutscher, E. Y. Two “Passive” Constructions in Aramaic in the Light of Persian. Ben-Ñayym, Z.; Dotan, A.; Sarfatti, G. (eds.). Proceedings of the International Conference on Semitic Studies, Jerusalem, 19–23 July 1965. Jerusalem. Pp. 132–151. Lehto, A. Divine Law, Asceticism, and Gender in Aphrahat’s Demonstrations, with A Complete Annotated Translation of the Text and Comprehensive Syriac Glossary. PhD. Diss. Toronto. Loesov, S. Review of H. Gzella. Tempus, Aspekt und Modalität im Reichsaramäischen. B&B 3:627–634. Loesov, S. Classical Syriac (Классический сирийский язык). Belova, A. et al. (eds.). Languages of the World. Semitic Languages. Akkadian. North-West Semitic (Языки мира. Семитские языки. Аккадский язык. Северозападносемитские языки). Moscow. Pp. 562–625.

15 Note that the Old Assyrian form ša-ad-wa-na-am makes rather feasible a derivation from šadûm ‘mountain’ (Sargonic and Old Assyrian šadwum) tentatively hinted at by von Soden.

586 Loesov 2011 Macuch 1976 Muraoka–Porten 2003 Nemirovskaya 2009

Ontikov 2011 Pennacchietti 1988

Pennacchietti 1994 Reimond 2009 Ritter 1990 Schall 1960 Segal 1955

Segal 1965

Segal 1970 Stadel 2009

Steiner 2005

Telegdi 1935

Voigt 1998

Reviews: Semitic Studies Loesov, S. A New Attempt at Reconstructing Proto-Aramaic. B&B 6:421–456. Macuch, R. Geschichte der spät- und neusyrischen Literatur. Berlin–New York. Muraoka, T.; Porten, B. A Grammar of Egyptian Aramaic. Leiden–Boston. Nemirovskaya, A. Jewish Palestinian Aramaic (Иудейско-палестинский арамейский язык). Belova, A. et al. (eds.). Languages of the World. Semitic Languages. Akkadian. North-West Semitic (Языки мира. Семитские языки. Аккадский язык. Северозападносемитские языки). Moscow. Pp. 531–562. Ontikov, A. The Verb in Turoyo: Synchronic and Etymological Aspects. Diploma Thesis. RSUH. Moscow. Pennacchietti, F. Verbo neo-aramaico e verbo neo-iranico. Orioles, V. (ed.). Tipologie della convergenza linguistica: atti del convergno della Società italiana di glottologia. Pisa. Pp. 93–110. Pennacchietti, F. Il preterito neoaramaico con pronome oggetto. ZDMG 144:259–283. Reimond, E. Review of C. A. Ciancaglini. Iranian Loanwords in Syriac. JAOS 129:335–336. Ritter, H. Tūrōyo. Die Volkssprache der syrischen Christen des Tūr ‫ع‬Abdîn. C: Grammatik. Stuttgart. Schall, A. Studien über griechische Fremdwörter im Syrischen. Darmstadt. Segal, J. B. Mesopotamian Communities from Julian to the Rise of Islam. Proceedings of the British Academy 41. Pp. 109–139. Segal, J. B. The Jews of North Mesopotamia before the Rise of Islam. Grentz, J. M.; Liver, J. (eds.). Studies in the Bible Presented to Professor M. H. Segal. Jerusalem. Pp. 32*–63*. Segal, J. B. Edessa, “the Blessed City.” Oxford. Stadel, Chr. The Recovery of the Aramaic Root br! ‘to cleanse’ and Another Possible Aramaising Rendering in the Septuagint. Aramaic Studies 7/2:155–162. Steiner, R. On the Dating of Hebrew Sound Changes (*Ú > Ñ, *Ġ > ") and Greek Translations (2 Esdras and Judith). JBL 124:229–267. Telegdi, Zs. Essai sur la phonétique des emprunts iraniens en araméen talmudique. Journal Asiatique 226:177– 256. Voigt, R. Griechischer Wortindex zu Anton Schalls Studien über griechische Fremdwörter im Syrischen. Symposium Syriacum 7:539–543.

Talay, Sh. Die neuaramäischen Dialekte der Khabur-Assyrer in Nordostsyrien: Einführung, Phonologie und Morphologie. Wiesbaden: Harrassowitz, 2008. xxvii + 479 pp. Prior to the appearance of the work under review, the Neo-Aramaic dialects of the Khabur river valley (northeastern Syria) had not been properly described. The first attempt to describe a Neo-Aramaic dialect in that area was made by Heidi Jacobi whose book was published more than 30 years ago (Jacobi 1973). H. Jacobi could not visit the area for political reasons, so all the material was recorded in Lebanon from the speakers who had moved there from the Khabur region. The undertaking of Shabo Talay is much more profound and extensive. The ultimate aim of his study is to create descriptions of all the Neo-Aramaic dialects spoken in the Khabur river valley. The book under review is a reworked dissertation (Habilitationsschrift) written under supervision of Prof. Otto Jastrow and submitted to Friedrich-Alexander Universität Erlangen-Nürnberg in 2006. It is the first part of the planned three-volume work which is the outcome of many years of field research (1997–2005); the preliminary results were published earlier by Sh. Talay in an article (Talay 1999). The three-volume work includes: 1. The volume under review: Introduction, which is meant as an introduction to the whole work; Phonology; Morphology. 2. Texts in 23 dialects out of 25 described in the first volume, excluding Bnerumta and Čāl (Talay 2009).1 3. Glossary to these texts that is still to be published, as stated by the author in the Introduction (p. 4).

In the following I present notes which are partly critical, partly highlighting the most interesting points. The Introduction (pp. 3–53) includes explanations of important ethnlogical and ethnolinguistic terms such as aturaye and ašuraye, sur3t and lešana a¬uṛaya.The wealth of sociological information about Assyrians in the Khabur valley given in the book is most welcome and integral to this study. It provides additional arguments for the author’s statement (p. 49)

1

Only the introductory pages of this volume were available to me. I plan to review the work as a whole when all the three volumes are at hand.

588

Reviews: Semitic Studies

that the twenty five Khabur-dialects mostly keep their distinctiveness as opposed to the linguistic situation in Northern Iraq, where a sort of Christian Neo-Aramaic koine had emerged. Presently the need for a new classification of NENA is felt as never before because we have many new dialect descriptions. 25 Khabur-dialects are subdivided by the author into five groups or “clusters”: A. Tiyari, B. Txuma, C. Hakkari, D. Šamm3sdin and E. External Dialects (“Äußere Dialekte”). The term “cluster” here must be taken with a grain of salt because this classification is a mixture of a) traditional ethno- and geographical boundaries, and b) a number of phonological, morphological and lexical features. Sh. Talay did his best to support this classification with linguistic features, but the general picture is complicated, since most of the features are not exclusive. Probably the most distinctive profile has Šamm3sdin with its pronominal suffix 3 pl. -u and ergative suffix -lu; the other two features suggested—separate pronouns 3 ms aw3n, 3fs ay3n and nominal plural ending *-a¬a > -ate—are not unique, they are shared by some dialects in the Hakkari cluster. The Hakkari cluster includes eight dialects, each of them characterized by a number of specific features, but all of them have the derivational morpheme for abstract nouns *-u¬a > uwwa, which is unique for Khabur-dialects. Txuma and Tiyari are not so easily definable: both clusters are characterized by the retention of interdentals /¬ / and /¯/, but part of Tiyari (Upper Tiyari) dialects have /*¬ / > /š/ in the position after *ae > ε. The fifth cluster (“external dialects”) which includes Halmun and Lew3n is not a real cluster at all because these two dialects are different in many respects which is shown by the author himself (p. 53). Apparently this category of unclassifiable dialects was named “cluster” for the sake of terminological consistency. P. 58. “Auch nach Konsonanten können die Affrikaten stehen.” The examples that follow (such as kalwa) demonstrate that the author meant “Frikative” or “spirantisierte Allophone.” P. 59. Within the list of examples with /w/ as a realisation of old /®/2 we have the form š3Ðwa ‘soup’ which is a clear loanword, not recorded in CS dictionaries.3 It is a regional word, which is attested in Kurdish, Turkish, Arabic and Persian. Since its (Kurdish?) etymon most probably had /b/, it 2

It would be more correct to write /*®/, /*«/ etc. in this chapter, dedicated to “Realisierung der alten b3°a¯k3«a¬-Phoneme” (italics are mine.—A. L.). 3 According to Tezel 2003:73 šarbo is “recorded already in Bar Ali Syro-Arabic Lexicon, p. 405, ed. Gottheil”; this loanword is etymologically related to š3Ðwa but it has different meaning—‘a little jug, pitcher.’

Sh. Talay. Die neu-aramäischen Dialekte … (A. Lyavdansky)

589

is related to the discussion, but certainly has nothing to do with the realisation of /*®/. In fact this example (š3Ðwa) touches upon the rules of phonological adaptation of loanwords in NENA, an issue which deserves attention, but not discussed in the book. P. 60f. As an appendix to the discussion of the realisation of /«/ the list of forms with /f/ is adduced. All of these forms are loanwords with only one exception4 from Classical Syriac, naxfu¬a. In fact the list is not about the realisation of /«/, it is about /f/ in loanwords. P. 84f. The phenomenon of emphatization, which is sometimes designated by the term “synharmonism,” is understood by the author as velarization rather than pharyngealization. The emphatization is present in Khabur-dialects “only partly” as opposed to Urmia-dialects where it is widespread. It is not told which dialects have it and which ones do not have it. The emphatization may affect a syllable, a word, or a group of words. There are 10 phonemes which may be secondary velarizied: /p/, /b/, /d/, /č/, /z/, /¯/, /m/, /n/, /l/, /r/. Accordingly the phoneme inventory may be expanded by 10 new phonemes. Nevertheless the author lists 12 velarized phonemes, including /´/ and /¢/: “dadurch erhöht sich der Phonembestand durch die Konsonanten /¡/, /¤/, /¢/, /¥/, /Ì/, /´/, /À/, /Ï/, /ṃ/, /Û/, /ḷ/, /Ð/.” Indeed, some of the instances of /´/ and /¢/ are born by emphatization, but these two phonemes are old insiders of the Aramaic phoneme inventory. P. 102. The etymology of reflexive pronoun gana/ǧana is given in fn. 159: gana/ǧana < kurd. can / pers. ǧān ‘Seele, Geist; Selbst.’ On the other hand, on p. 207 it is stated differently: gana/ǧana < kurd. gan. It is all the more important because it is stated (p. 207) that in Neo-Aramaic gana is original while ǧana is secondary. Following the author’s etymological reasoning, one may conclude that, on the other hand, ǧana is original, if it is true that it goes back to the Kurdish can (p. 102). According to Chyet’s dictionary, Kurmanji Kurdish has can, but Zaza Kurdish gan (Chyet 2003:82). Taking into account dialectal variety in Kurdish the question with the etymology of Neo-Aramaic gana/ǧana should be further clarified. P. 130. Here among the examples of the present tense forms appears a form tam3r translated ‘he says.’ Cf. p. 113 am3rra ‘he says to her’; p. 112 am3r ‘(that) he says.’ To my mind the difference tam3r vs. am3r deserves 4 The other example with probable CS etymon, friyya (< *prī"ā) ‘numerous’ is interesting, but problematic: there are no attestations of CS prī"ā in dictionaries of CS; perhaps it was construed as a derivative from CS pr" ‘to produce.’

590

Reviews: Semitic Studies

explanation. On p. 308, fn. 359, the verb form tate is explained as follows: t-ate < *d-ate ‘that he comes.’ It probably contradicts translating am3r ‘(that) he says.’ The lack of the explanation for the verb forms like tam3r is related to the fact that the description of syntax is not included in the book under review. Apparently the absence of any discussion of particle *d is also related to that. P. 220. Noteworthy is the fact that besides Infinitive KKaKa we have two more verbal nouns: KKaKTa (p. 365) and KuKaKa (p. 355f.). The functional distribution may be presented as follows: KKaKa is used in progressive tense forms (“Verlaufsform”); KKaKTa are mostly, but not only, nomina vicis; KuKaKa is nomen actionis (buraxa ‘Trauung,’ ǧuhada ‘Anstrengung,’ xušawa ‘Denken’). The further clarification would be possible only with sentence examples in the context of a syntactic description. P. 438. The dialects Marbišo and SaÐa have the preposition 3ll-/l- (< *"al) along with al (< *"al), which in my opinion may be explained only by the influence from adstrate dialects. P. 439–440. I do not think it is correct to say that prepositions *am und min “fielen zusammen zu m/m3n.” It is possible that *am > m-, but hardly *am > m3n. It is more correct to state that m/m3n assumed the meanings of *am. P. 443ff. In the section 3.6.1, dedicated to prepositions, the author generally follows etymological principle of description: there are entries b-, l-, m-, etc. But in 3.6.1.8 (“Die Preposition ‘für, zu nach, hin’ ”) he substitutes etymological principle for functional principle. Before concluding this review I would like to briefly compare the findings of Shabo Talay with the description by Heidi Jacobi whose work on Txuma was based on the material recorded from the informants from the Khabur region as well (Jacobi 1973). The problem with this comparison is that it is not clear to which variety of Txuma5 belonged two of her five informants (three of them were from Tall Hurm3z). As the following comparative table shows, there are significant differences for some items:

5 Now, thanks to the evidence in the present book of Sh. Talay, we see the distinctive character of five varieties of Txuma: Tall Hurm3z (Txuma Gawaya), MazÐa, Gund3k, G3ssa and Ber3ǧnaye.

Sh. Talay. Die neu-aramäischen Dialekte … (A. Lyavdansky) Jacobi 19736

Talay 20087

Personal pronouns

2m 2f

ā¬ ~ āha¬ ā¬ ~ āhi¬

a¬i, ā¬, a¬3n, ay3¬ a¬i, a¬3n, ati

Copula

3ms 3fs 2ms 2fs 1ms 1fs 3pl 2pl 1pl 2ms 2fs 1ms 1fs 3pl 1pl

hāwe hōya hāw3¬ hōya¬ hāw3n hōyan hāwe hāwi¬u(n) hāwax ~ hāw3x lēw3¬ ~ lē¬ lēwa¬ ~ lē¬ lēw3n ~ lēy3n~ lēn lēwan~ lēyan~ lēn lēle, lēlun/on, lēna lēwax, lēw3x, lēx

hole hola holux, how3¬, hol3¬ holax, howa¬, hola¬ holi, howi, how3n holi, howi, howan holē hóloxu(n), ho¬un, ho¬u holan, howan, howax lew3¬, lel3¬ lewa¬, lela¬ lew3n, leli lewan, leli lelε lew3x, lelax

Copula, negated

591

It is remarkable that some of the non-contracted forms (ahat, ahit; hāwitun) are present in Jacobi’s work but absent from Talay’s. On the other hand, certain contracted forms of negated copula (lē¬, lēn, lēx) are totally absent from Talay’s description. Summing up, the book of Sh. Talay conforms to high standards of quality. The linguistic evidence is presented with care and attention to minute but significant details. The book is a synchronic description of Khabur-dialects, but the author goes into diachronic analysis wherever he deems it relevant. Some of the dialects described in the book appear for the first time on the linguistic map not just as names, but as distinctive idioms whose linguistic profile is demonstrated by ample evidence. All this makes the book of Sh. Talay an important contribution to the research of Neo-Aramaic. A. Lyavdansky References Chyet 2003 Jacobi 1973

6 7

Chyet, M. L. Kurdish-English Dictionary: Ferhenga Kurmancî-Inglîzî. New Heaven–London. Jacobi, H. Grammatik des thumischen Neuaramäisch (Nordostsyrien). Wiesbaden.

The transcription was harmonized with the system of transcription by Sh. Talay. I gathered the evidence for all Txuma varieties in the book of Sh. Talay.

592 Talay 1999

Talay 2009 Tezel 2003

Reviews: Semitic Studies Talay, Sh. Überblick über die neuaramäischen Dialekte der Assyrer am Khabur (Nordostsyrien). Mediterranean Language Review 11:156–175. Talay, Sh. Neuaramäische Texte in den Dialekten der KhaburAssyrer in Nordostsyrien (Semitica Viva 41). Wiesbaden. Tezel, A. Comparative Etymological Studies in the Western Neo-Syriac (àūrōyo) Lexicon with Special Reference to Homonyms, Related Words and Borrowings with Cultural Signification. Uppsala.

Gabor Takács. Etymological Dictionary of Egyptian. Vol. 3. Leiden–Boston, 2008.* To begin with, I evaluate all three published volumes of Etymological Dictionary of Egyptian (further on EDE) by Takács as a Herculean labor for one person and a major step forward in Egyptology and, moreover, comparative Afrasian/Afroasiatic linguistics,1 of which they are the so far largest published up-to-date compendium performed on a methodologically acceptable level.2 Unlike many other linguists dealing with Afrasian * This review was written within the frames of three projects: Featuring early Neolithic man and society in the Near East by the reconstructed common Afrasian lexicon after the Afrasian database (supported by the Russian Foundation for Sciences), Semitic Etymological Dictionary (supported by the Russian Foundation for the Humanities), and Evolution of Human Languages (supported by the Santa Fe Institute). 1 By comparative Afrasian linguistics in this context I mean the field of lexical comparison and establishing etymologies aiming at drawing data from all the Afrasian branches and languages available, and not the individual branches inner comparison (even with optional sporadic allusions to other branches like in an otherwise very helpful Omotic Lexicon and Phonology by the late L. Bender—see Bender 2003), which is a necessary but not sufficient condition for operating with the Afrasian macro-family as a whole. 2 Among other recent works, in main features meeting this level’s requirements, I have to limit myself to very few: G. Takács’ and V. Blažek’s numerous scattered papers, A. Dolgopolsky’s papers and his regrettably still unpublished Nostratic Dictionary, the Afrasian comparanda in SED I and II and the Afrasian Database and individual papers by O. Stolbova and the present reviewer; of course, one comes across good, feasible or trivial individual etymologies and lexical comparisons in other authors, too, but they do not lay a key—or any—role in the field’s progress. It does not mean I accept all, or even most, concrete etymologies, comparisons or reconstructions by the listed authors including my own previous ones, but only that all of them generally comply with the demands of the comparative-historical method. For instance, I disagree with many Dolgopolsky’s etymological decisions in his gigantic and invaluable collection of data accounted for by his forcing the Afrasian material into Nostratic, which, in my opinion (as well as in the opinion of the Moscow Starostin’s group and, still more significant, in that of late Sergei Starostin himself ), is taxonomically on “sister,” not on “mother-daughter” terms with Afrasian thus requiring a more discriminating and cautious comparative analysis. Of less recent studies, this evaluation generally applies to the much-attacked HSED by Orel and Stolbova, which should be recognized, in spite of all its drawbacks, a telling achievement in the comparative Afra-

594

Reviews: Semitic Studies

languages comparison, Takács not only professes the comparative-historical method but keeps to it in his practice. Mainly following the “Russian school” in this method3—while reasonably critical towards many of its particular etymologies—he seems to realize well what painstaking obligations this method imposes: sticking to regular sound correspondences,4 restricting one’s soaring fancy in semantic comparison (an unfortunate trait of certain authors), scrupulously collecting and explicitly presenting all the comparative data available from all the languages involved, without which a solid complete etymology cannot be constructed (the rule not followed by most Afrasianists and even by etymological dictionaries of such best studied Semitic languages as Hebrew), etc., etc. As it was long ago established, each of Egyptian hieroglyphs transliterated as ± (represented by Egyptian vulture), x (flowering reed stalk), y/j (double reed stalks), r (mouth), n (waves of water) and ¯ (cobra) stands for more than one consonant corresponding to more than one Semitic and Afrasian phoneme. In this situation and in view of enough blank spots in Afrasian consonantism, Takács’ choice of strategy is unmistakable and the only one workable: to adduce as many feasible parallels as possible to any phonetic interpretation of each Egyptian lexeme, the shotgun risk of missing the mark naturally being higher than with any other language, to say nothing of a tremendous bulk of extra work. Another feature of the reviewed work which I highly appreciate is what I would call the author’s scholarly generosity. He presents a lot of data seemingly unnecessary for this or that Egyptian etymology he suggests5 but contributing to comparative Afrasian studies whose main problem, same as in Semitics, is an incredible abundance of misleading variant roots. This makes the task of establishing regular sound correspondences and building etymologies on them extremely difficult. Usually Takács does his best to produce all variant roots and, in most cases (regrettably sian field, especially if one looks at this dictionary not as a paragon of etymological work, but as materials for a reconstruction—in accordance with its subtitle (it is not by accident that Takács, in spite of all his sharp criticism of HSED, frequently refers to it in EDE disagreeing with some of its suggestions and accepting others). 3 Cf., for example, EDE I 7–8. 4 Instead of an effortless invention of new disposable ones, unique for every and any suggested comparison as is an unfortunate habit with some of otherwise highly professional descriptive linguists who cannot resist the temptation of etymologizing, lexical comparison or reconstructing proto-forms. 5 It is sometimes necessary in order to contest or dismiss an existing alternative etymology, though for my money it could be done less often than Takács does.

G. Takács. EDE III (A. Militarev)

595

not in all6), juxtapose them after separating from each other instead of making them into an indiscriminate salad as many authors often do. The reviewer, on his part, benefits a lot from this bulk of data collected from many scattered sources. As for the “insider” Egyptologists’ criticism of Takács’ individual etymologies, not being an Egyptologist I am no arbiter in this controversy, but, generally speaking, I am inclined, in many cases, to trust the etymologies based on reasonable comparison with even distantly related languages more than a well-known practice, often by mono-language specialists, to forcibly wring etymologies out of the same language, which results in creating what I would call “scholarly folk etymologies” or “mythetymologies.”7 Another issue where I fully agree with Takács is his acceptance as a whole of “traditional” Egyptian-Semitic consonant correspondences rather than those proposed by O. Rössler and his followers: I am still unaware of examples corroborating their theories which cannot be better or alternatively interpreted by consistent comparison with Semitic and other Afrasian languages on the basis of “traditional” correspondences and those outlined by the “Russian school.” And, last but not least, the author’s overall familiarity with the published and unpublished literature (sometimes excessive, I daresay), comprehensive, nearly exhaustive, bibliography and accurate references to the relevant sources must also be praised. This said, I will allow myself to dwell upon what I consider the weak points or drawbacks of this fundamental endeavor hoping that the au6

As an example: giving on pp. 708–710 the entire spectrum of etymological opportunities for Copt. *mat- (if it continues Eg. *md.t ‘greatly, very’), Takács reconstructs the Afras. proto-form as *m-t ~ *m-d ‘many’ indiscriminately uniting the forms in *-d and *-t in various languages; to separate such roots as accurately as possible (properly treating, of course, regular *d > t or *t > d cases if any) and only then make a cross-reference to them as to variant roots is the only cautious and methodologically correct way of approaching this extremely entangled and evasive phenomenon (see also SED I lx–lxiii). 7 Like the doubtful derivation, current in the Egyptian philology (and not infrequent in Semitics), of m±w.t ‘rays, beams, brightness’ with m- prefix from ±wy ‘to be long,’ which implies the meaning ‘those stretching out (from the sun)’ considered by Takács—in spite of his adducing quite plausible Afrasian parallels meaning ‘bright, shining’—“not impossible” (p. 67), or a similarly indemonstrable derivation of m±w.ty ‘two arms, shoulders’ from the same ±wy ‘to stretch out, reach’ (p. 73; for lack of better solutions I would even consider relating m±w.ty as a metathesis to Afras. *!am-: Sem. *!amm-at- ‘elbow, forearm,’ West Chadic *!am- ‘hand, arm’ and, perhaps, Central Cushitic: Aungi yiŋ ‘claw’).

596

Reviews: Semitic Studies

thor shares my attitude: it is both much more profitable and thrilling to hear professional criticism of one’s own work—preferably in a benevolent or at least politically correct form, of course—than an ovation giving no more than a transient enjoyment. The main deficiency of all the three volumes of EDE is lack of any index of Egyptian lexemes (especially unfortunate in case of Volume One where the lexicon is not arranged alphabetically), suggested Afrasian proto-forms and their meanings, to say nothing of all individual nonEgyptian forms quoted in the dictionary. This lack of indexes gravely complicates any serious work with—rather than just an idle scanning through—Takács’ materials. The author’s choice not to compile such indexes, in view of relative ease of doing it today using one’s computer, remains for me absolutely enigmatic. It can hardly be justified even by such an excusable motivation as economy of time and space since the author, while quoting various etymologies ever proposed for each individual case—sometimes useful or even necessary for the discussion but sometimes exuberant—spares neither time nor space on adducing and even analyzing fantastic etymologies by the same authors.8 There are many etymologies including my own ones that Takács refutes (some of his arguments I thankfully accept—fully or partially, some do not), as well as there are lots of his etymologies with which I disagree or can suggest alternative ones. This would be a normal and indispensable scientific procedure if not for an excessive vehemence and peremptory self-assurance with which some of the comments in EDE III are tinged—unlike Takács’ earlier works showing a more discreet attitude toward his precursors and senior colleagues.9

8

It is quite beyond my understanding how Egyptian or Afrasian studies could benefit from another allusion to L. Homburger’s (comparing, e. g., Eg. mn¶ with Ful ¤ur-de) and some other authors’ amateurish comparisons only to stigmatize them every time as “absurd,” “absolutely false,” “irreal,” “pure fancy” or provide them with “out of question” and “no comments” remarks (let me note parenthetically that such appraisals, especially when applied to qualified specialists’ etymologies, even erroneous, look somewhat rude: the same appraisal can always be formulated in a more polite manner). 9 Not toward the junior ones for lack of them in comparative Afrasianistics: far from young Takács and Blažek are dramatically the youngest (not counting the younger Leonid Kogan and Maria Bulakh in Moscow, who could but would not go beyond Semitics) active doers, the last of the Mohicans, perhaps; if new hands do not fall from heaven, the whole field is doomed to wither away before blossoming out.

G. Takács. EDE III (A. Militarev)

597

A few examples: On p. 52, Takács adduces Eg. m±" ‘to present, offer; sacrifice; slaughter’ with the following comment: “Evidently, we are dealing with an inner Eg. innovation from the root m±" … (meaning ‘richtig, wahr’.—A. M.), not a separate root with an original meaning ‘to offer, give’ (or sim.) as often erroneously suggested in etymological literature.” Further on under the heading “All other etymologies are unacceptable,” he quotes a Lowland East Cushitic parallel meaning ‘presents given to a bride’s family’ and ‘tax’ matching Eg. m±" (if < mr") both phonetically and semantically, but commenting that it “seems tempting …, but, for the reasons outlined above … is here fully improbable.” If it were formulated in a more cautious way—something like “In my opinion, m±" ‘to offer, give’ is more likely an inner Eg. innovation from m±" ‘in order, proper, set aright, direct’ (or sim.)”—it would be not quite convincing for me, but passable as the author’s personal hypothesis leaving certain doubts. The cited categorical overstatement, however, with all its “evidently,” “not a separate root,” “erroneously suggested,” “unacceptable” and “fully improbable” provokes a strong objection. First, the only “reasons outlined above” is the author’s statement with no argumentation at all. Second, the meaning shift from ‘in order, proper, set aright, direct’ to ‘to present, offer; sacrifice; slaughter’ is not only not evident, but rather improbable or at least unprovable. Third, each of the two Eg. m±" roots can be phonetically interpreted in three different ways,10 which implies three opportunities that these two lexemes represent the same or homonymous roots and six opportunities that they represent two phonetically different roots. On p. 150, looking for cognates of Eg. myn ‘today, now,’ the author qualifies Orel and Stolbova’s Cushitic parallels (Bilin imānā ‘time, past,’ Saho amāna, Somali imin, amin ‘time’) as semantically false: I am unable to see what’s false in this comparison. Neither do I understand on what grounds Chadic *b-l ‘thigh’ should be carefully distinguished from Central Chadic: Musgu blá ‘Bein’ and the like (p. 232); or why the equation by Orel and Stolbova of these Central Chadic forms with North Berber: Zayan a-®a¶il ‘camel’s leg’ is unacceptable both phonetically (this is certainly so) and semantically (ibid.); or why the connection of Eg. mn.t ‘(hind) leg, thigh’ (if from *mn, which Takács fully admits) with such East Cushitic forms as Oromo manyē ‘lower part of leg’ seems improbable (p. 233). The only answer I can deduce from these 10

Cf. on p. 48 on the etymology of Old Eg. m±" ‘richtig, wahr’: “Another obstacle is the disputable origin of OEg. -±- (*-r-, *-l-, *-!- ?).”

598

Reviews: Semitic Studies

assertions is that ‘thigh’ should not be compared with ‘leg.’ This semantic purism (which Takács, in most cases, is luckily far from sticking to), unjustified not only in long-range comparison between the Afrasian branches, but even between closely related languages, if accepted, would make any comparative and etymological work meaningless—not less so than fanciful semantic equations or not caring a damn about regular sound correspondences, both phenomena too common in Afrasian studies. The above reasoning fully applies to the author’s confidence in semantic improbability of the comparison (made by the same Orel and Stolbova) between Eg. mn.t ‘swallow’ and regularly corresponding phonetically Chadic forms meaning ‘bird’ and ‘poule des rochers.’ It would be too self-assuring of an etymologist to claim to be an expert in subtle criteria by which prehistoric people could or could not transfer such meanings; what at the most can be said about this equation is that it is hard to prove or disprove.11 Sometimes a comparativist’ discriminating taste surprisingly fails Takács. On p. 61, he gives the following Berber parallels for Eg. m±" ‘place’: (?) Zenaga e-mer (unexpected -r) ‘sein,’ Ahaggar e-mel ‘être dans un lieu’ (and sim.). Even ten question marks, however, can hardly justify the equation of Zenaga e-mer with -r and the meaning ‘breast, bosom’ (!) with other Berber, Cushitic and Chadic forms with -l meaning ‘place’ and thus matching Eg. m±" (if < *ml") ‘place.’12 On p. 94, Eg. m±¶ ‘Korngarbe’ (“most probably … < *mr¶”) is suggested as “cognate with Sem.: perhaps Yemeni Ar. marµ-ah [irreg. -µ-?] ‘heap of corn’ … ||| NAgaw: … Hamir mµr-ā … (a very strange devocalized form.—A. M.) ‘Garbe’ … || HECu.: Gedeo … mir£- [irreg. -£-] ‘to bundle’ … ||| ECh.: cf. Lele mòr³ò [-³- < *-g-?] ‘botte (ensemble)’,” and further on: “A striking parallel appears also in Akk. … bira¶¶u ~ biri¶¶u … ‘Bund (von Gemüse u. ä.)’.” This mess of irregular correspondences can be at best mentioned as a set of variant roots, but as an etymological entry, even hypothetic, it looks strange—as if proposed by another author, gifted with a very specific idea of lexical comparison, whom there are not few in Afrasian studies.

11

In SED II we had a lot of such problems, with bird names especially; by the end of work on that volume we hopefully learned to treat them more cautiously and less categorically. 12 If not for the fact that in part of Cushitic languages quoted, -" is expected to be preserved.

G. Takács. EDE III (A. Militarev)

599

In the entry Eg. mtf¶ ~ f¶ ‘bark (?) of Soker,’ the author asks a question (p. 209): “Any connection to WCh.: Zaar ¤agi ‘to carry load’?” With no traces of phonetic resemblance at all, I see no grounds for even posing this question.13 At times, yielding to the understandable temptation of presenting the entire spectrum of feasible etymologies, Takács goes beyond reasonable limits. On p. 328, adducing possible parallels to Eg. mnq ‘to complete, give effect to, restore, heal, etc.’ he says: “alternatively, cf. perhaps Tna !amengewe [irreg. -g-] ‘interrompere, lasciare à metà un lavoro incominciato’.” The isolated Tigrai verb with a polar meaning ‘to interrupt, leave a job unfinished’ and irregular correspondence has nothing in common with the Eg. verb ‘to complete, etc.’ And further in the same entry: “… in principle, a cognateship with NBrb.: Qbl. … a-melγiγ ~ a-menγiγ ‘1. sommet de la tête, 2. os du crâne’ … is not to be excluded.” The meaning shift from ‘head’ to ‘complete, finish’ is possible of course, if not for the fact that in this particular case, according to the logics of linguistic comparison, it implies the derivation of the Eg. term meaning ‘complete, finish’ from the assumed Afrasian proto-form meaning ‘head’ retained only in Qabyle! I daresay such examples only pollute the otherwise fairly professional author’s work and are not even worth mentioning. There also occur passages which I completely fail to understand. Thus, on p. 300, discussing M. Cohen’s comparison of Eg. nhp ‘to copulate’ with “a certain Brb. *n-b-y”14 id., which Takács discloses as Ahaggar e-nbi ‘monter (une femme), accomplir l’acte sexuel,’ he qualifies this comparison as “out of question” for two reasons: (1) the Berber term with this 13

Let alone the fact that Takács, as many others, criticized HSED exactly for using isolated data (EDE I 7): “… Stolbova and Orel were sometimes (too many times, in fact, but Takács was at that time milder in his appraisals.—A. M.) inclined to use isolated data from one single modern Chadic language.” By the way, I see nothing criminal in properly using isolated data, the more so in materials for a reconstruction. A whole another thing is that an author ought to well realize him/herself—and preferably to convey this understanding to the reader—that, in most cases, comparative data from only one language or several tightly related languages out of a multi-language taxon of a higher level are worth much less than from a selection of languages representative of this taxon. This is true, say, even of such well-studied languages as Aramaic or Hebrew represented alone out of the whole Semitic family, not to mention Chadic, still in the initial stage of comparative cultivation—the most difficult object of Afrasian studies. 14 Implying Afras. *¡, once proposed by H. Müller and H. Grimme and later advocated by I. Diakonoff (for Semitic) and the present reviewer (see SED I cv– cxvi), but not yet “institutionalized” in comparative Afrasian linguistics.

600

Reviews: Semitic Studies

meaning exists only in Ahaggar (which is not true: there is also E. Tawllemmet 3nb3y with the same meaning), and (2) in almost all other Berber languages it denotes ‘to taste.’ The whole comment is concluded with a stunning remark about *nby: “eventually PBrb. *m¥i” referring to—and, hence, accepting—an impossible cognation by M. Kossmann uniting the two synonymous roots, *nby and *m¥i, both meaning ‘to taste’ but etymologically having nothing in common. Beyond my understanding are also such assumptions as Eg. mÂt.w being “perhaps an irregular cognate” of Arabic mi"d-at- ~ ma"id-at- (p. 529), or Eg. m¶±.t being “perhaps an irregular cognate” to Sem. *kyl (p. 500), or that “there may be a further (genetic, not borrowed) cognate to the Eg.Ug. isogloss (A. M.: Eg. mrµ, said “borrowed from Sem. *rumµ-” on p. 437, and Ugaritic mrµ ‘lance’)” in Beja mal£ā (p. 438), etc., etc. An “irregular cognate” is a rather than not a contradictio in adjecto,15 and if Takács accepts the concept of “irregular cognates,” what is all his inspired criticism of other authors’ false etymologies based on irregular correspondences aimed at?16 Why not good-humoredly declare all or most of them, including those by Homburger and the like, just innocent irregular cognates—and close up shop? 15 As opposed to an assumed and consciously singled out variant root, which is entirely another pair of shoes. With my experience of editing S. Maizel’s Ways of Root Derivation in Semitic (see Maizel 1983) and undying interest in root variation, I hate to pose myself as a purist not admitting any exceptions from regular sound correspondences. However, I insist again that every assumption of a concrete root variation case (which may be due to various reasons including those beyond cognation—like contamination/analogy—or be haphazard and, hence, fictitious which we often have no means to distinguish) must be explicitly differentiated from an averment of cognation based on regular sound correspondences and “sound laws.” At the same time, I recognize that there are “border cases” between “normal” cognation and root variation (in Afrasian languages, fortunately, incommensurably less numerous than variant roots). Unlike root variation whose origin poses a great and perplexing theoretical problem and may fully clarify (or may not) only with a radical and still remote progress in comparative-historical linguistics (I would call this a strategic task), these marked “border cases” are usually hard to analyze merely due to our insufficient knowledge of historical phonology or morpho-phonology of a given language and are cognizable in the much more foreseeable future (a tactical task)—anyway, they always have some very concrete raison d’être which must—and can—be eventually brought to light. 16 Cf., e. g., (p. 421) the just appraisal of C. Brockelmann’s and F. von Calice’s relation of Eg. mrw ‘desert’ with Sem. *barr- as phonologically impossible since Eg. m- does not correspond to Sem. *b-: what makes it more “phonologically impossible” than the above listed “irregular cognates”?

G. Takács. EDE III (A. Militarev)

601

Sometimes Takács, usually quite prudent in his semantic comparisons, surprises me with nearly “Ehretian” flight of fancy. On p. 152, Afras. *m-n ‘wet’ is reconstructed uniting Arabic mn! ‘macérer la peau avant de la tanner,’ Oromo mānya ‘ocean’ and Chadic forms meaning ‘rain’ and ‘dew.’ I fail to figure out a common criterion by which a semantic connection is legitimate between ‘soak the hide before tanning it,’ ‘ocean’ and ‘rain, dew,’ or between ‘in order, proper, direct’ and ‘to present, offer; sacrifice; slaughter’ (the right thing to do?), or between ‘breast, bosom’ and ‘stay in a place’ (all above), or between ‘poison’ and ‘tears’ (pp. 704–705), or between ‘word, expression’ and ‘to measure’ (p. 781),17 while the comparison between ‘today, now’ and ‘time’ is regarded as false (above). Several more examples of my disagreement with the author. On p. 49, he compares the above-mentioned Eg. m±! ‘richtig, wahr’ with Ethiopian *mrµ (which is phonologically untenable per se) and Beja melah ‘to lead the way’ with the comment “[-l- < *-r- reg.].” Not only this correspondence is wrong (cf. on p. 438: “Bed. -l- < Cu. *-r- is irreg.,” but Takács even quotes (p. 48) Berber *ă-mVl ‘se diriger vers’ perfectly matching Beja melah. On p. 102, Eg. µmsy ‘sit’ is justly mentioned as not related with m±s ‘knee,’ but it is also said that the presence of the µ- prefix in µmsy cannot be ruled out in view of CCh.: Masa músā ‘sit down’ (I would add WCh.: Warji mas and Diri masa ‘to rest’); cf., however, ECu.: Bayso amac- ‘sit’ (in other records, hama´- and hamaÉ-) and Jiddu huusamo (met.) id., which, unless directly related to Eg. µmsy pointing to Afras. *µ- as the first radical, at any rate deserve to be mentioned in this context.

17 This partly reminds me of some of Chr. Ehret’s pearls of comparison— much more impressive, to be just, than those by Takács (or similar semantic approximations we in Diakonoff ’s Afrasian team were given to, though it was long ago …), but what Ehret is allowed, Takács is not—that I sometimes cannot resist the temptation of citing. Here are the selected ones: ‘armpit’ and ‘to thatch roof ’ (with the comment: “armpit is a covered area of the body”); ‘forest’ and ‘thirst’ (the reconstructed proto-meaning being ‘waterless place, desolation’); ‘widow’ and ‘thief ’ (‘to impoverish’ as the proto-meaning); ‘woman,’ ‘small’ and ‘few’ (reflecting the ancients’ machismo or conveying their complaint of too few real women available?); erotic ‘to marry’ and ‘thumb’; enigmatic ‘pregnant,’ ‘molar tooth’ and ‘to spread out’; ‘to sink,’ ‘knee’ and ‘egg’; ‘calabash’ and ‘to know’; ‘mane’ and ‘callus’; poetic ‘sugar-cane’ and ‘to be sad’; picturesque ‘log’ and ‘old cow’; and such theological revelations as ‘God,’ ‘to brand cow,’ and ‘soot’ (with ‘to scorch, to sear’ as the proto-meaning); and ‘God,’ ‘be, become,’ ‘fresh milk’ and ‘to sprout.’

602

Reviews: Semitic Studies

On p. 159, adducing North Omotic *mayz- ‘liver’ as an undoubted cognate of Eg. myz.t id., Takács includes South-East Ometo forms reconstructed by him as *mayy- with the comment “[reg. from *mayz-].” However tempting it is to equate the two forms, in Blažek’s Lexicostatistical Comparison of Omotic Languages (see Blažek 2008),18 the development *z/ž > y is postulated only for Kafa-Mocha; Bender’s way of solving this conundrum by reconstructing (Bender 2003:118), North-West Ometo *may + z does not help as it implies an unproved suffix or unlikely root extension *-z. Anyway, the development *z/ž > North-West Ometo y, possible in theory, requires proofs; until they are produced, I prefer to see in North Omotic *mayz- and *ma!ay- (besides Zaysse, Gidicho, Ganjule, Koyra, Kachama maayyé, Chara mayya and She mai there are also Zergulla ma!e and Sezo m5!i wherein -!- must reflect Afras. *-"-) two separate roots, the latter having a plausible Afras. etymology: *ma"ay- ‘liver, entrails’ > Akkadian amūtu ‘liver,’ Hebrew mē"ayim (pl.) ‘entrails, intestines,’ Arabic ma"y- id., etc. (see SED I No. 185); more problematic East Chadic: Gadang mùyò ‘liver’ (quoted by Takács on p. 161; an isolated form, but hardly coincidental); and probably East Cushitic: Oromo mo"o ‘loins, back’ and South Cushitic: Gorowa meheme ‘second stomach,’ Ma!a mméme ‘navel’ (in Ehret 1980:157). On pp. 231–232, analyzing Eg. (Pyramid text) mn.t ‘(hind) leg, thigh, buttocks, etc.,’ the author quotes its continuation in Coptic bilti ‘thigh, hips, buttocks’ via Demotic mn(y).t ‘thigh.’ To justify an irregular shift from Eg. m- to Coptic b-, which he accepts without any doubts, Takács (“with respect to the Cpt. reflex”) does not rule out a connection to Chadic *b-l ~ *¤-l asking if the following chain of sound changes should be assumed: Afras. (i. e., Eg.-Chadic) *b-l > Proto-Eg. *bl-t ~ *bn-t > Old Eg. mn.t ~ *ml-t > late Eg. *bl.t ~ *bn.t > Coptic bilti again. From both angles, the historical phonology of Egyptian and the Okham’s razor principle, this “chain of sounds changes” looks extremely strange.19

18

Where consonant correspondence tables by Lamberti-Sottile, Bender, Hayward and Ehret are adduced. 19 I wonder if the Coptic forms could be explained instead from another Eg. (Pyramid text) root, bnw ‘ein Körperteil’ or ‘waist or buttocks,’ according to Faulkner, likely confirmed by what seems its variant stem wbn.w ‘root of tail’ implying that Eg. bn.w is from *bl-w (all the Eg. forms—without Coptic—and feasible Chadic cognates are quoted in EDE II 207), from which they may derive without violating regular sound correspondences; perhaps the assumpion of a secondary -t in Coptic is less untenable than the explanation suggested by Takács.

G. Takács. EDE III (A. Militarev)

603

On p. 296, the comparison of Eg. mnn.w ‘fortress’ with a few Ethio-Semitic forms (Tigre mäna ‘create,’ Selti mäne ‘build,’ etc.) is accompanied with NB 2: “Alternatively, Eth.-Sem. *mny may derive from PSem. *bny (assim.?).” Since b and n are well compatible in Semitic languages in general and in Ethiopian in particular (though there are cases of *nb > mb of course), the quoted Ethiopian words, unless Cushitisms, can be at best regarded as root variants of *bny, but not as derived from the latter. On pp. 319–320, Eg. (Old Kingdom) mnz± (likely < *mnzl with the mprefix of names of instruments) ‘jar for liquids (water, milk, etc.)’ is justly called a perfect match of Akkadian (from Old Babylonian on) mazzālu (< *manzāl-) ‘vessel for pouring oil or water’ with the comment: “… the hypothetic Eg. nz± … may be identical with Sem. *nzl ‘to flow down (or sim.)’,” followed by a comprehensive etymology of common Sem. *nzl (with a few non-Semitic Afrasian parallels). While this explanation looks acceptable, I can offer another one making unnecessary the reconstruction of the hypothetic Eg. verb *nz± (< *nzl)—not attested to in Egyptian texts and thus being a weak point in the above reasoning—namely, that Eg. mnz± is an early borrowing of Akk. mazzālu (< *ma-nzāl-) < nazālu ‘to flow, etc.’20 On pp. 407–408, a few East Cushitic forms meaning ‘to have mercy, pity’ are justly supposed to derive from Ethiopian, particularly from Amharic marä ‘to have pity,’ with a parenthetical remark: and eventually from Eth.-Sem. *mrr ‘to be bitter, sad.’ Not only such a semantic shift is improbable,21 but Eth. *mµr ‘to have mercy, pity’ is well known to be metathetically derived from common Semitic *rµm with the same meaning. On p. 421, O. Rössler’s comparison of Eg. mrw ‘desert’ with Arabic malā id., impeccable if Eg. mrw reflects *mlw, is rejected on the ground that Arabic root mlw primarily means ‘marcher d’un pas vigoureux et rapide, et se mettre à courir de toutes ses forces.’ In view of a well-known 20 Representing a perfect parallelism of Egyptian (Pyramid text) nms.t ‘k. of can, jar’ ~ Akkadian (Old Babylonian) nemsētu ‘washbowl’ derived from mesû ‘to wash, to clean’ < Semitic *msw ‘melt, dissolve, flood’; like in the case of Eg. mnz± ~ Akk. mazzālu, a deverbal origin of the Akkadian term implies an Akkadian loan into Egyptian (see Militarev 2007:139). 21 Here, again, I do not mean to say that rare, odd, unique or exotic meaning shifts do not ever occur—they do, of course—but each of them should be individually and cautiously treated as such; an etymologist ought to explicitly expound every such shift by any reasoning (similar developments, metaphor, plausible mental associations, historical circumstances, etc.) whenever possible, but not just state it in passing as something that goes without saying.

604

Reviews: Semitic Studies

overwhelming homonymy of Arabic roots, one is better be wary of such traps; the only remote association for desert I could fancy is the place to quickly walk/run through/from (strange enough, too: deserts are usually a huge space), but this does not work either as the main and the primary meaning of this root is ‘time’ (sing. malā, dual !al-malawāni ‘the two times,’ i. e. ‘day and night’) with an apparently denominative verb with various ensuing meanings besides the above one: ‘to preserve someone for a long time (of God),’ ‘to enjoy for a long time,’ etc. Desert has hardly anything to do with time if not for profound philosophical associations in ancient Arabs, while Takács suggests on the same page a few Afrasian forms with *ml radicals matching Arabic malā ‘desert’ quite acceptably. On p. 451, comparing Eg. mh.wt ‘family, relatives, etc.’ with Berber *iman ‘person, soul,’ Takács analyzes the latter as *i-m-an translating it as ‘soul(s)’ and providing with the following comment: “pl. affixes *i- … -an attached to the orig. root *m.” While Ch. de Foucauld, in his great dictionary (see Foucauld 1951–1952), marked imân as plural with no singular, iman, though formally coinciding with the plural pattern (cf., however, min in Zenaga with no formal traces of plural), in various Berber languages usually denotes a person and/or a soul with no collective or plural meaning. Since the singular noun pattern *iC1VC2 is well attested to in Berber (cf., e. g., *iγass ‘bone’: Siwa îγ3s, Ayr eγăs, Zenaga i!ssi < *iγVss-, Fodjaha, Qabyle iγass, etc. < Afras. *£(w)ays-), I prefer to relate Berber *iman with Afras. (and “global”) *man- ‘man, person.’ On p. 520, discussing possible parallels to Eg. m¶tb.t ‘an ornament of gold,’ Takács qualifies M. Görg’s comparison of its supposed root (*¶tb) with Sem. *µtb ‘schmücken’ as “phonologically dubios” (misprint for dubious) because Eg. ¶ is not equal to Sem. *¶. All this is a mess. First, Eg. ¶ is of course equal to Sem. *¶ (misprint for µ?). Second, when Takács quotes M. Görg quoting Sem. *µtb, both are wrong—it is not *µtb, but *¶tb: HALOT 307 gives Arabic ¶atiba ‘to be colourfully striped’ (¶- confirmed by BK 1 592–593). The last comment by Takács is also something I do not understand: “In addition, as noted by E. Edel … no m-prefix derivative of this Sem. root is attested.” No hypothesis of the Eg. word having been borrowed from a Sem. source is mentioned (naturally enough as in Hebrew and Aramaic the derivatives of the µtb root mean ‘cloth’ and ‘garment’), in which case this averment would have made sense as an argument against borrowing, while as an argument against cognation with the Sem. root the same averment from a comparative linguist sounds preposterous: why on earth should a word prefix in one

G. Takács. EDE III (A. Militarev)

605

language prevent from comparing this word with its presumed cognate without a prefix in another language provided their radicals regularly correspond to each other? On p. 543, discussing Afras. *m-č, Takács remarks: “A. Ju. Militarev (2005, 114) erroneously compared also Akk. … m/eiz!u ‘(a type of beer and wine)’ … but this derives from Akk. mazû … ‘to press out’.” First, I certainly did not compare Akk. m/eiz!u to -z- with Afras. forms in *-č, but, naturally, analyzed Afras. *miwač- ‘alcoholic drink’ as a separate root (Militarev 2005:114). Second, I think that either the original notion of the Akkadian root is a certain method of producing beer (in CAD M2 439 it is given as mazā!u, mazû ‘to squeeze, produce a liquid (beer, oil, juice)’) or, less likely, there was a contamination of the primary noun denoting beer or wine (cf. Arabic muzz- ‘vin très-agréable au goût; goût un peu acidulé du vin ou d’un fruit, p. ex. de la grenade’ and mzz V ‘déguster le vin’22) with the homonymous verb mazâ!u, mazû ‘to squeeze, press, rape’— anyway, the primary meaning ‘alcoholic drink’ of the Akkadian-Arabic noun is sustained, on the one hand, by such Afras. parallels as Central Chadic: Muzgu maza ‘beer,’ Central Cushitic: Bilin māz, Kemant mīz and East Cushitic: Saho mez-t ‘mead’ and, on the other, by such Semitic forms with the assumed fossilized *-r 23 as Arabic mizr- ‘sorte de boisson faite de millet,’ mazar- ‘sorte de boisson enivrante preparée de froment’ (BK 2 1099), Sabaic mzr-m ‘date-wine,’ and Geez mazara ‘to prepare beer from grain,’ m3zr ‘beer, ale.’ On p. 557, Eg. sy± ‘erkennen’ is equated (after Zyhlarz) with Berber *z-r-y ‘to see’ with the comment: “In principle, PBrb. *z- can derive from … *c- (Mlt. 1991, 242) yielding Eg. s-.” However, the allusion to the reviewer’s study establishing this correspondence does not help: it is not *zry, but *À3r with À in some of the languages (Siwa, Fodjaha, Seghrushen, Senhaja, Shenwa, etc.) and a normal loss of emphasis in some others. Hence, the Berber verb cannot be equated with Eg. sy± but is most likely derived from Afras. *Ìar- ‘to look, guard’: Sem. *nVÎar- (with a fossilized *n prefix) ‘to guard’ and ‘to look’ (in Arabic); West Chadic Hausa Éárè ‘to guard.’ As for Eg. sy± (if < *syr) ‘learn’ (and what seems its variant stem, s±± ‘understand,’ also if < *s!r) it matches well Afras. *car- ‘to know, re22

The verb mzz ‘sucer’ may be denominal or homonymous. Envisaging pending severe criticism, I may say that, as in most similar cases of fossilized (or “frozen”) elements, their function in the root-formation is not clear, which does not mean that this phenomenon is non-existent or is not worth further research (see Militarev 2005). 23

606

Reviews: Semitic Studies

member, teach’: Sem. Akkadian asāru ‘remind,’ Ugaritic wsr, ysr, Hebrew ysr ‘teach, instruct’; Central Chadic: Mofu -s3r- ‘know, remember,’ Mesme sar ‘remember’; East Cushitic: Burji sēr-is- ‘learn,’ etc. In EDE I 194 and 262, Sem. *šm" ‘to hear’ is related with Eg. s¯m id. instead of smt (Pyramid text) id., sm.t (New Eg.) ‘ears’ and smtmt (Book of the Dead) ‘eavesdrop.’ Though, according to EDE III 609, “… Eg. s¯m has been convincingly etymologized from *sm¯ < *sm" ~ Sem. *šm" ‘to hear’ (see EDE I 262),” I did not find anywhere in EDE any arguments— convincing or not—for this comparison; to see in Sem. *-" versus no traces of it in Eg. a secondary root extension (a triconsonantizer, perhaps) better matches the Okham’s razor principle than the many-step explanation of incompatible s and " in Eg., with " for some unknown reason changing to ¯ (reflecting *g? or any emphatic affricate?), the whole process crowned with metathesis at that! On p. 610, the author remarks parenthetically that Afras. *Ì yields Berber *À, which “has been demonstrated by G. Takács 2006, 59–61 contra Mlt. 1991, 242.” In fact, what I put in the table of consonant correspondences in the mentioned source (Aikhenvald–Militarev 1991:242) was two Berber reflexes of Afras. *Ì and *Ã, namely ¥/¢ and À/´.24 On p. 629, Hebrew and Phoenician šmr ‘to guard, watch over, etc.’ are united, without any comments, with Old South Arabian (Sabaic, in fact) ¯mr ‘to protect’ thus following the Semitilogical tradition to combine two different—perhaps to be regarded as two variant or contaminated—Semitic roots *šmr and *¯mr,25 which is strange for a comparativist of the level of Takács. On p. 632, I noticed a confusion (a misprint, perhaps) in the Afras. correspondences: “Eg. -š- ~ Brb. *-z- < AA *-ŝ-.” It is Afras. *-ĉ-, not *-ŝ-, of course. On p. 661, assuming that Berber *nkr ‘to raise, rise’ might be related to Eg. mk±.t ‘support, pedestal’ (a weak comparison, but “mentionable” at least), Takács avers that “an alternative etymology is also available” for Berber *nkr, namely Eg. ng±g± (< *ngrgr?); since Berber *k does not cor-

24

That Takács looked at that table attentively enough to notice it is confirmed by another comment of his (p. 800): “… PBrb. *¥ can go back to PAA *Ã (cf. Mlt. 1991, 242; Takács 2006, 57–59, 62).” 25 Cf., e. g., the comments to Hebrew šmr in HALOT 1581: “… in Amorite and Ugaritic it is striking that š corresponds to ¯ and z.” For *¯mr (without confusing it with *šmr) see DRS 335–336.

G. Takács. EDE III (A. Militarev)

607

responds to Eg. g, which Takács knows perfectly well, there is no way of calling it “an alternative etymology.” On p. 699, one finds East Cushitic *mur£- ‘gristle’ (with a reference to H.-J. Sasse) > Oromo morg-aya ‘tendon, nerve,’ Arbore morgi ‘tendon, sinew,’ Somali muruq ‘muscle,’ Yaaku mor³-i! ‘sinew of neck.’ In fact, these are two different (perhaps, variant) roots erroneously fused into one:26 Sasse (in Sasse 1979:54) quotes under *mur£- ‘gristle’ only Somali muruq ‘muscle,’ Konso murq-a ‘soft bud,’ Gidole mor£-a ‘bone of nose, soft part of ensete’ and Yaaku mor³-i!: inclusion of Oromo morg-aya and Arbore morgi would have contradicted the regular correspondences of consonants in Proto-East-Cushitic he established.27 As for Yaaku mor³-i!, I prefer to derive it from *morg- (see root 2): though Sasse allows of both opportunities (saying that g is sometimes palatalized to ³ and that there are several instances of palatalization of *£ into ³ ibid.), I am unaware of any unequivocal cases of *£ > ³ in Yaaku and doubt they ever exist, while *g > ³ is a natural palatalization paralleled by *k > ç (thus rendered by Heine; phonetically most likely č). The other root is *ma/urg- ‘shoulder, (muscle, tendon of) neck’ discussed on p. 78: East Cushitic: Oromo and Arbore (above), Jiddu mer³a ‘neck,’ Dasenech morg-oč ‘hump of cow,’ Darasa morge!-o ‘tendon (of neck?),’ Yaaku (above); North Omotic: Wolaitta morgiy-a ‘shoulder, hump,’ Gofa morg-e ‘shoulder,’ etc. As for Akkadian murgu ‘Rückgrat, Schiffskiel’ considered a loan of Sumerian murgu or mur7-gú, with gú ‘neck,’ I would not be so confident about claimed Sumerisms in Akkadian, when a suspected Akkadian word has attractive Semitic—or even only Afrasian as in this case (triconsonantal at that)—parallels; as for Arabic mr³ ‘presser et manipuler un membre du corps’ tentatively compared by Takács, what is implied is hardly ‘shoulder, neck or back,’ but quite a different body member, I am afraid. On p. 749, Takács admits the possibility of “a remote etymological connection” between Eg. m¬± ‘to flout, vex’ and what he calls Sem.-Brb. *r-g-m ‘to flout’ with the parenthetical comment: “Eg. m¬± < *mkr ~ 26

Though on p. 79, Takács refers to the “Cushitological lit.” combining these two roots, this is no argument, of course—it only means that various authors (except Sasse 1979—see fn. 28) make the same mistake, the reason steadily being the same: deviation from the comparative-historical technique, somehow justified only in very special situations, with its regular sound correspondences. 27 The reference to Sasse 1982:148 is misleading—that is what he says in his comment to Burji morg-: “Cf. Or[omo] morg-aya … Elsewhere, the word appears as PEC *murk’- (stands for *mur£-.—A. M.) … Because of the irregular correspondence the Bu[rji] word must be taken to be a loan from Or[omo].”

608

Reviews: Semitic Studies

*rkm via met. and an irreg. Eg. ¬ < *k, Sem.-Brb. *g.” Not to mention the facts that Eg. *rkm is non-existent, and Brb. *rgm ‘to flout’ is very likely a loan from Arabic or, as suggested by W. Vycichl, even from Punic,28 Eg. *mkr and Sem. *rgm are too remote to qualify even for variant roots. On p. 800, it is said about C. Hodge’s idea of relating Eg. m¯ ‘ten’ with Berber *tē-mih¥ay (referring to the reconstruction by Prasse) ‘100’ that it “is to be accounted for, although PBrb. *-h- seems to have no match in Eg. m¯.” All Berber forms of this numeral following have no -h-29 and can hardly compromise the elegant cognation by Hodge. On p. 804, Modern South Arabian *maken- ‘many’ is said to derive from *mgn, with *-k- < *-g-. This sound correspondence—unquestionably presented as regular at that—is a pure fantasy; the most plausible etymology of Modern South Arabian *maken- is from *ma-kayn-, cf. Soqotri k33n, k3yh3n ‘many,’ likely related to Arabic k!n ‘be strong, become strong and big’ and perhaps further to East Cushitic: Oromo ikkaan, Dirayta kaan- ‘big’ (unless a chance coincidence), implying Afras. *ka!n- ‘big, many.’ A personal remark pertaining to ethics rather than accuracy: quoting Semitic data from SED, Takács, usually very precise in his bibliographic references, refers to this source as “Kogan” implying that my part of our common work was Afrasian parallels only (he must be well aware it is not so).30 In one place (p. 211), he refers to a certain Shalunov; in Quoted Lit28

A common triconsonantal root with a completely coinciding meaning, fairly marked at that, inherited by two Afrasian branches is rather unusual; the common origin in such cases is especially “fishy” when Berber and Semitic terms are under consideration, with a lot of well-attested Arabic and a few Punic loanwords in Berber, even in Tuareg. 29 With all my due respect to K.-G. Prasse’s Berberological studies, far most advanced in the field, his *h when yielding what he calls “zéro panberbère”, i. e. Ø everywhere, even in Taneslemt, the most conservative Tuareg language (or when the Taneslemt data are missing), remains a speculative construction—based on subtle differences in vocalism only—until outer Afrasian cognates reflecting *h are presented (cf. Prasse 1969:12–13). 30 Amusingly, Takács, without knowing it, anticipated a future development which I could not envisage at that time: I am discontinuing my participation in the Semitic Etymological Dictionary. Not to be misunderstood, I’d like to use this opportunity and explain my motifs. One of them is a too long wait for my coauthor Leonid Kogan, who has been busy with no less important projects, to go on with his part of work on the third volume to be devoted to flora (that volume had been previously announced as fourth, the third one planned to be devoted to Semitic lexicostatistics, but this plan changed and we are publishing the corresponding materials in separate papers)—at his age, he can afford such a procrastination, at mine, I cannot take chances. The other is my being more and more

G. Takács. EDE III (A. Militarev)

609

erature, it is “Shalunov. S.: The Prefix *!V- in Semitic. MS. Handout for the paper presented at the 6th International Hamito-Semitic Congress (Moscow, April 1999).” Let alone the fact that the 6th International Hamito-Semitic Congress, of which I was the convener and Takács a participant, took place in 1994, the reference to Shalunov made me laugh. That noted Semitist was at that time one of my several students at the Jewish University in Moscow whom I was trying to get involved into Semitic or Afrasian studies whenever they showed the least interest; in case of Shalunov it was to no purpose: I asked him to do some simple lexical search (which he did) and, to encourage, included him as a co-author of my paper at the Congress (of which Takács must have been aware, too); after some time the guy dropped out and I have never heard of him again. In conclusion: my list of disagreements, critical comments and alternative interpretations could be continued—the margins of all the three volume copies I’ve got are covered with comments. However, this would be a finite list and the critique of EDE is an accomplishable task—unlike criticism of comparisons and etymologies by many other authors which I prefer not to take on myself: criticism makes sense of only those colleagues who stick to the same general principles in their method as you. I want to stress again that the reviewed book is full of excellent, revealing and simply correct etymologies and must be considered a very helpful and efficient tool in Egyptian and Afrasian studies. I hope the author’s work on forthcoming volumes will be successfully carried through.31 A. Militarev focused on Afrasian comparison and proto-Afrasian reconstruction rather than Semitic. To this I may add that I am confident that Kogan, with his profound knowledge of practically all Semitic languages—not on the vocabulary level as myself, but on the philological one—and his skill in the comparative-historical method, will manage this difficult task brilliantly; I am not sure about Afrasian data, but it is the etymological dictionary of Semitic, after all, and the comparison with non-Semitic Afrasian languages is quite optional. 31 If I am not mistaken, it was myself who once introduced Gabor Takács as a promising young comparative student to my teachers and friends Igor Diakonoff and Sergei Starostin, for years shared with him my data and works (which he mentions in Acknowledgements to EDE II)—as many others did and mutually, of course,—recommended him to various colleagues in Russia and abroad, arranged or helped to arrange for his visits to Russia including the conferences I organized (all of which he does not mention in his long lists of Acknowledgements to each of the three volumes). I am fine without curtseys, and quite satisfied that these modest efforts were not wasted.

610

Reviews: Semitic Studies

References Aikhenvald–Militarev 1991

Blažek 2008

Bender 2003 Ehret 1980 Foucauld 1951–1952 Maizel 1983 Militarev 2005 Militarev 2007

Prasse 1969 Sasse 1979

Sasse 1982

Aikhenvald, A.; Militarev, A. Lybian-Guanche Languages. Diakonoff, I. M.; Sharbatov, G. Sh. (eds.). The Afroasiatic Volume of the Languages of Asia and Africa Series. IV/2. Мoscow. Pp. 148–267 (Livijsko-guanchskije jazyki). Blažek, V. Lexicostatistical Comparison of Omotic languages. Bengtson, J. D. (ed.). In Hot Pursuit of Language in Prehistory. Amsterdam–Philadelphia. Pp. 57–148. Bender, M. L. Omotic Lexicon and Phonology. Carbondale. Ehret, Chr. Historical Reconstruction of South Cushitic. Berlin. Foucauld, Ch. de. Dictionnaire touareg-français. Paris. Maizel, S. S. Ways of Root Derivation in Semitic. Moscow (Puti razvitija kornevogo fonda semitskikh jazykov). Militarev, A. Root Extension and Root Formation in Semitic and Afrasian. AuOr 23:1–2, 83–130. Militarev, A. Akkadian-Egyptian Matches in Lexicon. Miller, C. L. (ed.). Papers on Semitic and Afroasiatic Linguistics in Honor of Gene B. Gragg (SAOC 60). Chicago. Pp. 139–145. Prasse, K.-G. À propos de l’origine de H touareg (tahaggart). København. Sasse, H.-J. The Consonant Phonemes of Proto-EastCushitic (PEC): A First Approximation. Afroasiatic Linguistics 7/1:1–67. Sasse, H.-J. An Etymological Dictionary of Burji. Hamburg.

Abbreviations of Periodicals, Reference Works, Series, and Sources AASF AASOR AB AbB ABD ABIM ABL AfK AfO AFSM AHw. AIA AIHA AJA AJSL AKM AKT 3 ANES ANESS ANET AnOr AO AOAT AoF AOS ArAn ARM ARMT ArOr

Annales Academiae Scientiarum Fennicae Annual of the American Schools of Oriental Research Anchor Bible Altbabylonische Briefe in Umschrift und Übersetzung The Anchor Bible Dictionary. Ed. by D. N. Freedman. Vol. I–VI. New York, 1992 A. al-Zeebari. Altbabylonische Briefe des Iraq-Museums. Münster, 1964 R. F. Harper. Assyrian and Babylonian Letters. Chicago, 1892– 1914 Archiv für Keilschriftforschung Archiv für Orientforschung American Foundation for the Study of Man W. von Soden. Akkadisches Handwörterbuch. I–III. Wiesbaden, 1965–1981 Cuneiform Tablets in the Collections of the Australian Institute of Archaeology (Melbourne) F. Rasheed. The Ancient Inscriptions in Himrin (Himrin 4). Baghdad, 1981 American Journal of Archaeology American Journal of Semitic Languages and Literatures Abhandlungen für die Kunde des Morgenlandes E. Bilgiç; C. Günbattı. Ankaraner Kültepe-Tafeln. III (FAOS Bh 3). Stuttgart, 1995 Ancient Near Eastern Studies Ancient Near Eastern Studies. Supplement J. B. Pritchard (ed.). Ancient Near Eastern Texts Relating to the Old Testament. Princeton, 31969 Analecta Orientalia Antiquités orientales, Louvre Alter Orient und Altes Testament Altorientalische Forschungen American Oriental Series Archivum Anatolicum Archives royals de Mari Archives royales de Mari. Traduction Archiv Orientální

612

AS ASJ ASKT ATD ATHE

AUCT AuOr Bab B&B BaF BaM BASOR BBB BCOTWP BDAG BDB BDTNS BE BET BibInt BIN BiOr BK BKAT BLS BM BN BR BSA BTB BZ

Babel und Bibel 6

Assyriological Studies Acta Sumerologica P. Haupt. Akkadische und sumerische Keilschrifttexte. Leipzig, 1881–1882 Das Alte Testament Deutsch B. Kienast. Die altassyrischen Texte des Orientalischen Seminars der Universität Heidelberg und die Sammlung Erlenmeyer—Basel. Berlin, 1960 Andrews University Cuneiform Texts Aula Orientalis Babyloniaca. Etudes de philologie assyro-babylonienne Babel und Bibel Bagdhader Forschungen Baghdader Mitteilungen Bulletin of the American Schools of Oriental Research Bonner Biblische Beiträge Baker Commentary on the Old Testament Wisdom and Psalms A Greek-English Lexicon of the New Testament and Other Early Christian Literature. 3rd ed. Chicago–London, 2000 A Hebrew and English Lexicon of the Old Testament by F. Brown, S. R. Driver and Ch. A. Briggs. Oxford, 1907 Base de Datos de Textos Neosumerios (http://bdtns.filol.csic.es) The Babylonian Expedition of the University of Pennsylvania. Series A. Cuneiform Texts Beiträge zur Biblischen Exegese und Theologie Biblical Interpretation Babylonian Inscriptions in the Collection of James B. Nies. Yale University, New Haven Bibliotheca Orientalis A. de Biberstein-Kazimirski. Dictionnaire arabe-français. I–II. Paris, 1860 Biblischer Kommentar. Altes Testament Bible and Literature Series Cuneiform Tablets from British Museum Biblische Notizen Biblical Research Bulletin on Sumerian Agriculture Biblical Theology Bulletin Biblische Zeitschrift

Abbreviations of Periodicals, Reference Works, and Sources

BZAW CAD CAH CBS CCPA CCT CDA CDG CDOG CH/CḪ CM CRRAI CSD CST CT CUSAS DAE DCH DJD DJPA DNWSI Dozy DPA DRS DUL

613

Beihefte zur Zeitschrift für die alttestamentliche Wissenschaft The Assyrian Dictionary of the Oriental Institute of the University of Chicago. Chicago, 1956–2011 Cambridge Ancient History Cuneiform Tablets in the University Museum, Philadelphia A Corpus of Christian Palestinian Aramaic Cuneiform Texts from Cappadocian Tablets in the British Museum J. Black; A. George; N. Postgate. A Concise Dictionary of Akkadian. Wiesbaden, 1999 W. Leslau. Comparative Dictionary of Ge"ez (Classical Ethiopic). Wiesbaden, 1987 Colloquien der Deutschen Orient-Gesellschaft Codex Ḫammurapi Cuneiform Monographs Comptes rendus, Rencontre assyriologique internationale J. Payne Smith. A Compendious Syriac Dictionary Founded upon the Thesaurus Syriacus. Oxford, 1903 T. Fish. Catalogue of Sumerian Tablets in the John Rylands Library. Manchester, 1932 Cuneiform Texts from Babylonian Tablets in the British Museum Cornell University Studies in Assyriology and Sumerology E. Littmann. Deutsche Aksum-Expedition. IV. Sabaische, griechische und altabessinische Inschriften. Berlin, 1913 The Dictionary of Classical Hebrew. Ed. D. Clines. Sheffield, 1993ff. Discoveries in the Judean Desert. Oxford, 1955ff. M. Sokoloff. A Dictionary of the Jewish Palestinian Aramaic of the Byzantine Period. Ramat Gan, 1990 J. Hoftijzer; K. Jongeling. Dictionary of the North-West Semitic Inscriptions (HdO 21). Leiden–New York–Köln, 1995 R. Dozy. Supplément aux dictionnaires arabes. T. I–II. Paris, 1927 H. Limet. Etude de Documents de la Période d’Agade. Bruxelles, 1973 D. Cohen. Dictionnaire des racines sémitiques ou attestées dans les langues sémitiques. La Haye, 1970ff. G. Del Olmo Lete; J. Sanmartín. A Dictionary of the Ugaritic Language in the Alphabetic Tradition (HdO 67). I–II. Leiden, 2003

614

EA ECTJ

EDG EDG I EDH EI ELTS

ETCSL FAOS FCBSS FM GAG3 Gesenius Thes. GKC GKT HALAT HALOT HANE/S HAT HdO HK

Babel und Bibel 6

J. A. Knudtzon. Die El-Amarna Tafeln (VAB 2). Leipzig, 1915 A. Westenholz. Early Cuneiform Texts in Jena. Pre-Sargonic and Sargonic Documents from Nippur and Fara in the HilprechtSammlung vorderasiatischer Altertümer, Institut für Altertumswissenschaften der Friedrich-Schiller-Universität, Jena. Copenhagen, 1975 W. Leslau. Etymological Dictionary of Gurage (Ethiopic). Vol. III. Etymological Section. Wiesbaden, 1979 W. Leslau. Etymological Dictionary of Gurage (Ethiopic). Vol. I. Individual Dictionaries. Wiesbaden, 1979 W. Leslau. Etymological Dictionary of Harari. Berkley–Los Angeles, 1963 Eretz-Israel I. J. Gelb; P. Steinkeller; R. M. Withing. Earliest Land Tenure Systems in the Near East: Ancient Kudurrus (OIP 104). Chicago, 1989 J. Black et al. The Electronic Text Corpus of Sumerian Literature (Oxford 1998ff.); http://www-etcsl.orient.ox.ac.uk Freiburger Altorientalische Studien; Bh = Beihefte: Altassyrische Texte und Untersuchungen A Feminist Companion to the Bible. Second Series Florilegium Marianum W. von Soden. Grundriss der akkadischen Grammatik (AnOr 33). Roma, 1995 Gesenius, G. Thesaurus philologicus criticus linguae Hebraeae et Chaldaeae Veteris Testamenti. Lipsiae, 1835–1853 Gesenius’ Hebrew Grammar. Ed. by E. Kautzsch. Transl. by A. E. Cowley from 28th Germ. ed. Oxford, 1910 K. Hecker. Grammatik der Kültepe-Texte (AnOr 44). Roma, 1969 L. Koehler; W. Baumgartner. Hebräisches und Aramäisches Lexikon zum Alten Testament. Leiden, 1991 L. Koehler et al. The Hebrew and Aramaic Lexicon of the Old Testament. Transl. and ed. by M. Richardson. Leiden, 1994–2000 History of the Ancient Near East/Studies Handbuch zum Alten Testament Handbuch der Orientalistik Handkommentar zum Alten Testament

Abbreviations of Periodicals, Reference Works, and Sources

HSED HSS HThKAT HUCA IAS IB ICK 1 ITT

Ja. JAOS JANES JBL JBVO JCS JEOL JNES JQR JSHRZ JSOT Sup JSQ JSS K. KAI KAR KB

KBo KK KEH KHC

615

V. E. Orel; O. V. Stolbova. Hamito-Semitic Etymological Dictionary. Materials for a Reconstruction (HdO 18). Leiden, 1995 Harvard Semitic Studies Herders Theologischer Kommentar zum Alten Testament Hebrew Union College Annual R. D. Biggs. Inscriptions from Tell Abū Ôalābīkh (OIP 99). Chicago, 1974 Interpreter’s Bible. New York B. Hrozný. Inscriptions cunéiforms de Kültepe. Prague, 1952 F. Thureau-Dangin; H. de Genouillac; L. Delaporte. Inventaire des tablettes de Tello, conservées au Musée Impérial Ottoman. I–V. Paris, 1910–1921 M. Jastrow. A Dictionary of the Targumim, the Talmud Babli and Yerushalmi, and the Midrashic Literature. New York, 1926 Journal of the American Oriental Society Journal of the Ancient Near Eastern Society of Columbia University Journal of Biblical Literature Jenaer Beiträge zum Vorderen Orient Journal of Cuneiform Studies Jaarbericht Vooraziatisch-Egyptisch Gezelschap “Ex Oriente Lux” Journal of the Near Eastern Studies Jewish Quarterly Review Jüdische Schriften aus hellenistisch-römischer Zeit Journal for the Study of the Old Testament. Supplement Series Jewish Studies Quarterly Journal of Semitic Studies Cuneiform Tablets from Kuyunjik in British Museum H. Donner; W. Röllig. Kanaanäische und aramäische Inschriften. I–III. Wiesbaden, 1962–1964 E. Ebeling. Keilschrifttexte aus Assur religiösen Inhalts (WVDOG 28; 34). I–II. Leipzig, 1919, 1920–1923 E. Schrader et al. (Hrsg.). Keilinschriftliche Bibliothek. Sammlung von assyrischen und babylonischen Texten in Umschrift und Übersetzung. Berlin, 1889–1915 Keilschrifttexte aus Boghazköi Kurzgefasster Kommentar zu den heiligen Schriften Alten und Neuen Testamentes sowie zu den Apokryphen Kurzgefaßtes exegetisches Handbuch zum Alten Testament Kurzer Hand-Commentar zum Alten Testament

616

KJV Kt KUB LA LAK Lane LAPO LBPT LSyr LTBA I M. MAD MAOG MARV 4 MB MC MDBP MDOG MDP MEE MHEM MHET MIO MSL MVAG MVN NABU NBC NEB NESE NETS NICOT NIVAC NJPS NRVS

Babel und Bibel 6

King James Version Cuneiform Tablets from Kültepe Keilschrifturkunden aus Boghazköi Ibn ManÏūr. Lisānu l-"arabi. Bayrūt, 1990 A. Deimel. Liste der Archaischen Keilschriftzeichen (WVDOG 40). Berlin, 1922 W. Lane. An Arabic-English Lexicon. London–Edinburgh, 1863–1893 Littératures anciennes du Proche-Orient I libri biblici. Primo Testamento C. Brockelmann. Lexicon Syriacum. Halle, 1928 L. Matouš; W. von Soden. Die lexikalischen Tafelserien der Babylonier und Assyrer in den Berliner Museen. Berlin, 1933 Cuneiform Tablets from Mari Materials for the Assyrian Dictionary Mitteilungen der Altorientalischen Gesellschaft H. Freydank et al. (ed.). Mittelassyrische Rechtsurkunden und Verwaltungs-texte (WVDOG 99). Wiesbaden, 2001 Le Monde de la Bible Mesopotamian Civilizations Matériaux pour le Dictionnaire de Babylonien de Paris Mitteilungen der Deutschen Orient-Gesellschaft zu Berlin Mémoires de la délégation en Perse Materiali epigrafici di Ebla Mesopotamian History and Environment. Memoirs Mesopotamian History and Environment. Texts Mitteilungen des Instituts für Orientforschung Materials for the Sumerian Lexicon. Roma Mitteilungen der Vorderasiatischen Gesellschaft Materiali per il vocabulario neosumerico Nouvelles assyriologiques brèves et utilitaires Cuneiform Tablets in the Nies Babylonian Collection New English Bible Neue Ephemeris für semitische Epigraphik New English Translation of the Septuagint New International Commentary on the Old Testament The New International Version Application Commentary New Jewish Publication Society Version New Revised Standard Version

Abbreviations of Periodicals, Reference Works, and Sources

NSGU NTSŠ OAA OAIC OBCh OBE OBO OBTR OIP OLA OLP OP(SN)KF Or Or NS OrSu OTL PAT PBS PDT 1

PDT 2 Prag PSAS PSD PW QuSem RA

617

A. Falkenstein. Die neusumerischen Gerichtsurkunden. I–III. München, 1956–1957 R. Jestin. Nouvelles tablettes sumériennes de Šuruppak au Musée d’Istanbul. Paris, 1957 Old Assyrian Archives I. J. Gelb. Old Akkadian Inscriptions in Chicago Natural History Museum. Chicago, 1955 Orientalia Biblica et Christiana U. Jeyes. Old Babylonian Extispicy: Omen Texts in British Museum. Istanbul, 1989 Orbis Biblicus et Orientalis St. Dalley; C. B. F. Walker; J. D. Hawkins. The Old Babylonian Tablets from Tell al Rimah. Hertford, 1976 Oriental Institute Publications Orientalia Lovaniensia analecta Orientalia Lovaniensia periodica Occasional Publications of the Samuel Noah Kramer Fund Orientalia Orientalia Nova Series Orientalia Suecana Old Testament Library D. R. Hillers; E. Cussini. Palmyrene Aramaic Texts. Baltimore, 1996 University of Pennsylvania, Publications of the Babylonian Section M. Çiğ; H. Kızılyay; A. Salonen. Die Puzriš-Dagan-Texte der Istanbuler Archäologischen Museen. I (AASF B 92). Helsinki, 1954 F. Yıldız; T. Gomi. Die Puzriš-Dagan-Texte der Istanbuler Archäologischen Museen. II (FAOS 16). Stuttgart, 1988 K. Hecker; G. Kryszat; L. Matouš. Kappadokische Keilschrifttafeln aus der Sammlung Karlsuniversität Prag. Praha, 1998 Proceedings of the Seminar for Arabian Studies The Pennsylvania Sumerian Dictionary. Philadelphia, 1984ff. A. Pauly; G. Wissowa. Realencyclopädie der klassischen Altertumswissenschaft. Stuttgart, 1894ff. Quaderni di Semitistica Revue d’assyriologie et d’archéologie orientale

618

RBC REB RGTC RIE

RIMA RIMB RIME RlA Rm. RS RSO RSV RTC RTL SAA SAACT SAALT SAAS SAOC SAWW SBAW SBH SBL SED I SED II SF SKL SL

SMEA StOr

Babel und Bibel 6

Cuneiform Tablets in the Rosen Babylonian Collection (Yale University) Revised English Bible Répertoire géographique des textes cunéiformes E. Bernand; A. J. Drewes; R. Schneider. Recueil de Inscriptions de l’Ethiopie des Periodes pré-Axoumite et Axoumite. I. Les Documents. Paris, 1991 The Royal Inscriptions of Mesopotamia. Assyrian Periods The Royal Inscriptions of Mesopotamia. Babylonian Periods The Royal Inscriptions of Mesopotamia. Early Periods Reallexikon der Assyriologie und Vorderasiatische Archäologie. Berlin, 1932ff. Cuneiform Tablets in British Museum Excavated by H. Rassam Cuneiform Tablets from Ras Shamra Rivista degli Studi Orientali Revised Standard Version F. Thureau-Dangin. Recueil de tablettes chaldéennes. Paris, 1903 Revue théologique de Louvain State Archives of Assyria State Archives of Assyria. Cuneiform Texts State Archives of Assyria. Literary Texts State Archives of Assyria. Studies Studies in Ancient Oriental Civilization Sitzungsberichte der Österreichischen Akademie der Wissenschaften in Wien. Phil.-hist. Klasse G. Reisner. Sumerisch-babylonische Hymnen nach Thontafeln griechischer Zeit. Berlin, 1896 Studies in Biblical Literature A. Militarev; L. Kogan. Semitic Etymological Dictionary. I. Anatomy of Man and Animals (AOAT 278/1). Münster, 2000 A. Militarev; L. Kogan. Semitic Etymological Dictionary. II. Animal Names (AOAT 278/2). Münster, 2005 A. Deimel. Schultexte aus Fara (WVDOG 43). Leipzig, 1923 Th. Jacobsen. The Sumerian King List (AS 11). Chicago, 1939 M. Sokoloff. A Syriac Lexicon. A Translation from the Latin, Correction, Expansion, and Update of C. Brockelmann’s Lexicon Syriacum. Winona Lake, 2009 Studi Micenei ed Egeo-Anatolici Studia Orientalia

Abbreviations of Periodicals, Reference Works, and Sources

SVTP TAD TADAE TAPhS TC TCL TCS TDP TED Th-A TIM ТKTA TLB TM TMH

TPAK TSŠ TTIM

TUAT UAVA UCP UET UF VAB V(A)S

619

Studia in Veteris Testamenti Pseudepigrapha S. Langdon. Tablets from the Archives of Drehem. Paris, 1911 B. Porten; A. Yardeni. Textbook of Aramaic Documents from Ancient Egypt. 4 vols. Jerusalem, 1986–1999 Transactions of the American Philosophical Society Tablettes cappadociennes du Louvre Textes cunéiformes du Louvre Texts from Cuneiforms Sources R. Labat. Traité akkadien de diagnostics et pronostics médicaux. 1– 2. Paris–Leiden, 1951 Th. L. Kane. Tigrinya-English Dictionary. Springfield, 2000 J. Theodor; Ch. Albeck. Midrash Bereshit Rabba. Critical Edition with Notes and Commentary. I–III. Jerusalem, 1965 Texts in the Iraq Museum Transliterationen und Kollationen von Texten der Akkade-Zeit (http://www.uni-marburg.de/altorientalistik/tkta.html) Tabulae Cuneiformes a F. M. Th. de Liagre Böhl Collectae Leidae Conservatae Cuneiform Tablets from Tell Mardī¶ (Ebla) Texte und Materialien der Frau Professor Hilprecht Collection of Babylonian Antiquities im Eigentum der Friedrich Schiller-Universität Jena C. Michel; P. Garelli. Tablettes paléo-assyriennes de Kültepe. I (Kt 90/k). Paris, 1997 R. Jestin. Tablettes sumériennes de Šuruppak conservées au Musée de Stamboul. Paris, 1937 P. Steinkeller; J. N. Postgate. Third-Millennium Legal and Administrative Texts in the Iraq Museum, Baghdad (CM 4). Winona Lake, 1992 Texte aus der Umwelt des Alten Testaments Untersuchungen zur Assyriologie und Vorderasiatischen Archäologie University of California Publications in Semitic Philology Ur Excavations. Texts Ugarit-Forschungen Vorderasiatische Bibliothek Vorderasiatische Schriftdenkmäler der Königlichen (Staatlichen) Museen zu Berlin

620

VA(T) VDI VE VT WVDOG WZKM YBC YOS ZA ZAW ZBAT ZDMG

Babel und Bibel 6

Cuneiform Tablets in Vorderasiatische Abteilung, Tontafelsammlung. Staatliche Museen zu Berlin Vestnik drevney istorii Il Vocabolario di Ebla Vetus Testamentum Wissenschaftliche Veröffentlichungen der Deutschen Orientgesellschaft Wiener Zeitschrift für die Kunde des Morgenlandes Cuneiform Tablets in Yale Babylonian Collection Yale Oriental Series. Babylonian Texts Zeitschrift für Assyriologie und vorderasiatische Archäologie Zeitschrift für die alttestamentliche Wissenschaft Zürcher Bibelkommentare: Altes Testament Zeitschrift der Deutschen Morgenländischen Gesellschaft