Journal of Cuneiform Studies (JCS 62) 2010

Journal of Cuneiform Studies Volume 62 Editor Piotr Michalowski, University of Michigan, Ann Arbor 2010 Associate Ed...

Author: Editor: Piotr Michalowski

51 downloads 1328 Views 20MB Size Report

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

DOWNLOAD PDF

Journal of Cuneiform Studies Volume 62

Editor Piotr Michalowski, University of Michigan, Ann Arbor

2010

Associate Editors Gary Beckman, University of Michigan, Ann Arbor Elizabeth Carter, University of California, Los Angeles Piotr Steinkeller, Harvard University Matthew W. Stolper, The Oriental Institute of the University of Chicago Managing Editor Billie Jean Collins

CONTENTS Massimo Maiocchi: Decorative Parts and Precious Artifacts at Ebla . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Sarood T. Mohammed Taher: Texts in the Sulaimaniah Museum Related to the Turam-ili Archive . . . . . 25 Gonzalo Rubio: Reading Sumerian Names, I: Ensuhkešdanna and Baba . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Thomas E. Balke: The Sumerian Ternary Numeral System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Paul Delnero: Sumerian Extract Tablets and Scribal Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Cécile Michel: Deux textes atypiques découverts à Kültepe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Jan Tavernier: Akkadian Personal Names with paḫāru or saḫāru as Initial Element . . . . . . . . . . . . . . . . . . 81 John P. Nielsen: Three Early Neo-Babylonian Tablets Belonging to Bēl-ētir of the Misirāya Kin Group . . 95 Alice Mouton: Sorcellerie hittite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Jeanette C. Fincke: Zu den akkadischen Hemerologien aus Ḫ attuŠa (CTH 546), Teil II: Die „Opferbrot-Hemerologie“ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

DECORATIVE PARTS AND PRECIOUS ARTIFACTS AT EBLA Massimo Maiocchi (Venice)

1. Introduction The complexity of the third-millennium B.C. Ebla city-state is a well-known topic among scholars, who have analyzed important aspects of precious-metal circulation: the standardization of common artifacts,' the high functionaries involved in their transfers,^ the administrative tools of accountability required to trace goods and persons,^ and their role in international and local trade and gifts.** Despite the fact that the large-scale transfer of precious metal has been studied extensively, revealing important clues for our understanding of the general trends of economic and political systems in third-millennium Syria,^ approximately one hundred records of small deliveries of manufactured goods are yet to be studied in full. Over the years, the analysis of the organization of Ebla has been handicaped by philological difficulties, mainly due to the extensive use of sumerograms in the texts, the omission of grammatical morphemes, and to our incomplete understanding of the local language. Recent improvements in our undertanding of Eblaite writing practices have provided insight into many such issues.^ This progress allows us to reconsider some of the published material to investigate in detail the administrative procedures

A prelimiary draft of this paper was presented to the joint seminar Cultural History of Mesopotamia in the Third Millennium B.C., Venice, 7-9 June 2004, Università Ca' Fosead, Venezia, Dipartimento di Scienze dell'Antichità e del Vicino Oriente; Universität Wien, Institut für Orientalistik. In the transliterations, readings that differ from the actual edition are marked with ""." 1. This practice fulfills various needs: control by the central administration to simplify accounting procedures; mass production to satisfy a great number of requests quickly; and easily trading precious objects, whose weight was approximately known on the basis of their typology. On the standardization of artifacts see Archi 1985a; Milano 1991; Pomponio 1998b; Zaccagnini 1991. 2. Especially noteworthy are the contributions of Tir, Ibrium, and Ibbi-Zikir. See Pomponio 1984 and 1988. 3. Many documents are classified in the colophons as "tablet of incoming goods" (dub-gar mu-kUj,[DU]), "tablet of outcoming goods" (dub-gar è), "tablet of received goods" (dub lu su ba4-ti), "tablet of purchases" (dub-gar nig-sám), see Milano 1980. 4. Precious metals were traded both in non-worked drops of silver and gold, and cast in artifacts, see Pinnock 1991; Archi 1993b. Valuable objects were also given as gifts to foreign delegations from Mari, A.BAR.SAL, Ki5, Armi, Dulu, Dub, Harran, Imar, Kakmium (see, e.g., the ceremonial weapons offered by the Eblaite administration to foreign kings in ARETI 12 o. i 10-ii 6), or presented to the temples of Kura, Ammarik, and Astapi. 5. The attention of modern scholars has been particularly focused on about thirty documents classified on the basis of the colophons as annual accounts of metals, or textiles and metals. These long tablets register huge amounts of silver and gold cast in artifacts brought (mukUjj[DU]) by court personnel. For such reason, they have been considered the starting point in the debate on third-millennium "economy" at Ebla. Through the annual account it was in fact possible to establish some fixed points on the internal chronology of the texts (see Archi 1986; Biga and Pomponio 1990), following the rise of power of the Syrian capital city (see Pomponio 1982; Archi 1985b). 6. In connection with the topic of this article, particularly noteworthy is the correct comprehension of es as "that is, what is left is, then," previously read AB and understood as "(goods) available." For a discussion on this term with previous bibliography see Waetzoldt 2001: 35-36. Y

ICS62(2010)

2

MASSIMO MAIOCCHI

of the capital city, focusing particularly on daggers and decorative disks (DIB), the most commonly mentioned artifacts of the Ebla archive.

2. Daggers (gir) The only excavated example of a third-millennium Syrian dagger is a bronze blade found in level IV A Ebla.^ The central motif is created by the superimposition of two metal layers, a technique epigraphically attested as nizi-mu (written also ne-zi-mu). There is also an iconographie representation on a marble inlay,'* showing a prisoner killed by a figure holding a dagger with a knob, probably to be identified with a gir mar-tu ga-me-u^ "mar-tu dagger with knob" (see below). The blade of the dagger shows also the ni-zi-mu feature, rendered as two parallel lines in the inlay. Daggers often occur together with textiles (ib-lá "ceremonial belt," and si-ti-tum, "sheath" or "pendant")' and are sometimes decorated with precious metals. The bodies of these artifacts are usually made of bronze, with a variable percentage of tin. The fact that their alloys seem not especially hard and that decorative parts are often added to them (see Archi 1993a: 620-21), suggest that they were not designed to be used as actual weapons. In order to better appreciate the amount of precious metal utulized in these artifacts, as well as their circulation among the Eblaite administration, one has to classify them according to their peculiarities.

2.1 Typology of Daggers The term gir, "dagger," may occur alone or in connection with other terms. Besides gir mar-tu, the documents mention gir kun, gir na-ba-hu, gir nidba, gir sum, and gir tur.'" All of these are extremely rare in the Ebla texts. On the contrary, the mar-tu (written also tu:mar and mar) and kun types are very well attested. These artifacts are usually made of bronze, with additional parts made of gold and/or silver."

2.2 gir mar-tu Mar-tu daggers are by far the most frequently mentioned in the Ebla archives. It seems reasonable to assume that this kind of dagger originates in the Martu region, despite the fact that people belonging to Amorites tribes in Nothern Syria are seldom refered to at Ebla.'' As recently noted by Fronzaroli (2003: 179), these ceremonial artifacts were sold by professionals called BAD gir mar-tu, lit. "owner ofa mar-tu dagger" to merchants/messengers (U5). These sellers may be connected with the elusive (semi)-nomadic clans or families under Eblaite control.

7. See Matthiae, Pinnock, and Scandone Matthiae 1995: 334. 8. See Matthiae, Pinnock, and Scandone Matthiae 1995: 275. 9. On the term si-ti-tum see Waetzoldt 2001: 70; Pasquali 2005: 175-76. 10. The term gir kun probahly refers to a metallic part attached to sheath and belt; see below § 2.5. As far as gir na-ba-i¡u is concerned, the only mention of this kind of dagger is found in MEE II16 v. 13, among the gifts (n i g - b a ) for Enna-Dagan. G. Pettinato translates "lucente" and relates the term to Akk. nabi\¡u "(golden) ornament" or napixu "shining" (MEE II commentary ad no. 16 r. I 3, p. 115); but Pasquali (2005: 67-71) convincingly argues that this refers to rock crystal. A golden gir sum, lit. "hacksaw," or "slaughter-knife," is mentioned in MEE X 20 r. XVII 27. 11. The most informative document on the metallurgical composition of m a r -1 u daggers is probably MEE XII 37 r. i 26-iv 3, listing the amount of tin and copper to be melted in order to obtain bronze for the main parts of these artifacts and the amount of silver to be exchanged for gold to make the relative decorations (zi-du). For a discussion and classification of these artifacts see below. 12. See Archi 1987: 12. For a discussion on Amorite presence at Ebla as well as in other third-millennium sources see Buccellati 1992: 83-104, with previous bibliography.

DECORATIVE PARTS AND PRECIOUS ARTIEACTS AT EBLA

3

The value of mar-tu daggers could be increased by adding decorations made of precious materials. Accordingly, subclasses of mar-tu daggers may be listed alphabetically as follows: a) gir mar-tu 'à-ma-da-wiim. The term'à-ma-do-ntim has been interpreted recently by Pasquali (2005:1034) as "di qualità eccellente, pregiato," by comparison with the West Semitic root *bmd. This designation occurs only three times in the Ebla tablets, in connection with daggers made with one or more precious metals." According to ARÊT VII 42 o. II1-4, a handle could be added to these objects: 2kù:babbar / 1 gir 'à-ma-da-num / Vi kù:babbar /â-ma-fl-twm-sù, "2 shekels of silver: 1 excellent mar-tu dagger; V2 shekel of silver: its handle." This may imply that 'à-ma-da-num refers to the quality of the metal (usually an alloy of silver and gold) and not to a specific kind of dagger. b) gir mar-tu 'o-NE-LUM. The etymology of 'a-NE-LUM is unclear. The term is perhaps to be compared with a-NE-LUM, read a-bi-lum in ARETVW 68 o. II1 in relation to a jar (la-ha) of precious metal. The alternation of 'à- and a- in word initial position is a common phenomenon in Eblaite orthography,'" possibly resulting from the dropping of the initial radical. Depending on the context, the term a-bi-lum has been interpreted either as "sostegno (per vasi)" from the root *w¿)/," or as "consumed/drossed (part)," by comparison with Akk. abalu "to dry up" (see Archi 2005: 40). Moreover, on the basis of a possible existing parallel, a-bi-lum has also been compared With, ga-bi-lum, related to Akk. kablu, "leg of a piece of furniture" (see Archi 1999:154 n. 32). As far as daggers are concerned, the term 'à-NE-LUM could be read 'a-bi-lum and related to Akk. lablru, "old, ancient, traditional, inherited, used," with the dropping of initial III and the common shift of/I/ to /r/ in final position. If the sign NE is to be read /¿g (see Krebernik 1982:196-97), one could also consider the reading 'à-li^-lum; see alälu {halälu) "to suspend, to hang," in relation to an additional element designed to hang the dagger and the sheath from the belt. In this case, the term could be the Semitic equivalent of gir kun, see §2.5. Due to the paucity of attestations and to interpretive difficulties, further details (such as the amount of precious metal involved, and its circulation among Eblaite personnel) remain unknown."' c) gir mar-tu ba-du-u^, also written ba-du-Ugandba-du. The additional element is probably a metallic ring used for hanging the sheath on the belt. The etymology oiba-du-u^,^ remains unclear: Waetzoldt (1990: 9) follows Pettinato (MEE II commentary ad no. 20 o. VIII 6, p. 151?) in relating this term the Semitic root *pth "to open;" Archi (1993b: 621) translates "fodero" by analogy with Old-Akk. ba\iyum "a platter or container;" Pasquali (2005: 112) relates ba-du-u^,g to Akk. napdû, "tie, bandage," but agree with Waetzoldt on the function of this element as a link between dagger and belt or garment. According to MEE VII 34 r. XIX 12-14 this decorative part is quite valuable: 7 gin DILMUNkù:babbar ba-du-u,, I gir m ar-tu-5Ù, "7 heavy shekels (of) silver (for) the ring of his dagger." Rings made out of both silver and gold are also attested, but is unclear if we are dealing with metal alloys

13. See/4RETI44o. V 7 - 1 0 : 4 k ù : b a b b a r / s u b i i - m / k ù - s i g i ; / k i n j - a k l g i r mar-tu 'á-ma-í/fl-niim "4 shekels ofsilver to be melted with gold to make 1 excellent mar-tu dagger"; AR£T VII 38 r. I 1-3: 14 kCi-sig,; / 4 kù:babbar / kinj-ak 1 gir mar-tu 'à-ma-da-niim "14 shekels of gold (and) 4 shekels of silver to make 1 excellent mar-tu dagger"; ARET VII 42 o. II 1-4 (for transliteration and translation see below). 14. See for instance the spelling of the toponym n-rUjj-îi''' in AilET III 118 r. I 6, attested also as 'à-ru^-îi'" in ARET I 8 o. XII 10 et passim. 15. See Pasquali 2005: 97. The existence of an object called NE.LUM, possibly a kind of clasp, may lead us to speculate that 'a-NE-LUM could also be a defective writing for 'à- NE.LUM.Tbe reading of NE.LI as dè-li or ti^-li (and consequently of NE.LUM as dè-lum or ti^-lum) "fermaglio" has been suggested by Pettinato in Mander (1982: 236). Contra Waetzoldt (MEE XII: 58) stressing that the term may be understood either as a "Schmuckstück," "Gerät," or "Waffe." 16. Only two instances are known to me: ARETXll 607 II' 5'; 1155 II' 1'.

4

MASSIMO MAIOCCHI

(perhaps as a result of remelting practices; see § 5 below) rather than different metallic parts.'^ Their purchase price fluctuates between 4 shekels (~ 31.32 g) and 9 DILMUN shekels of silver (~ 105.75 g).'^ d) gir mar-tu ga-me-ii, also written ga-me-Ug,'' ga-mi-ù,^" ga-ma-(a)-tum,^^ and ga-me.'-'^ The term has been understood as a metallic part, probably a knob or a decoration in the handle." A couple of texts give us some clues on these additional parts. MEE VII 47 r. IV 10-15 reads: sa-pi gin DILMUN kù:babbar / nui,-za 20 gir mar-tu ti kù:babbar / TAR kù:babbar / su-bal-ak / 6 gin DILMUN kii-sigij/ ga-ma-a-tum-sù "40 heavy shekels of silver (for) the decoration of 20 mar-tu daggers with silver point; 30 (shekels) of silver to be exchanged for 6 heavy shekels of gold (for) their knobs." One infers 0.3 heavy shekels (~ 3.52 g) of gold for the knob of each dagger. More frequently gir mar-tu ga-me-ù kù:babbar, "mar-tu daggers with silver knob," are mentioned. The amount of precious metal involved in this type of artifacts fluctuates between half shekel (~ 3.91 g) and VA heavy shekel (~ 15.27 g)," the latter being the figure mentioned in M££XII 37o. XVIII28-33:18 gin DILMUN nagga / sub si-in 2 ma-na sú + sa 4gin DILMUN a-garj-gar; / 10 lá-2 gir mar-tu /10 gin DILMUN 2 NI / kù:babbar / ga-ma-a-tum-sù, "18 heavy shekels of tin to be melted together with 2 mina 44 heavy shekels of copper: 8 mar-tu daggers; 10% heavy shekels of silver: their knobs." e) gir mar-tu kù:babbar, gir mar-tu ku-sigj^. "Golden" daggers are by far the most frequently mentioned in the documentation. Despite their designation, these are made of bronze (~ 237.94 g according to MEE XII 37 r. II 4-10). The shape and size of the additional parts is unkown. According to the above-mentioned MEE XII37 r. II4-10, the added part is composed by 6 gin DILMUN (~ 70.5 g) of gold. Their price fluctuates between 90 gin DILMUN (~ 1057.5 g) and 10 gin DILMUN (~ 117.5 g). "Silver" daggers are less frequently attested. They usually appear in connection with precious textiles or as gifts related to foreign cities." f) gir mar- tu GIS.SAL. This object is seldom mentioned in the documentation," although there are passages recording a hundred items." The meaning of GiS.SAL is uncertain, being listed in VE 361 without translation. As already pointed out by Waetzoldt (1990: 10), the similarity with Sumerian SAL.US may suggest that we are dealing with a "sheath of a dagger."^* This part is usually made of, or decorated with silver or, less frequently, gold, as showed in M££X 20 r. XI 18-23: 2 gin DILMUN kù:babbar / 1 GiS.SAL / 10 gin DILMUN kù:babbar / su-bal-ak / 2 gin DILMUN kù-sig,^ / nu „- za-sù, "2 heavy shekels of silver: a sheath; 10 heavy shekels of silver to be exchanged for 2 heavy shekels of gold: his decoration." Hence, the above-mentioned sheath hoards in total

17. SeeM££VII29r.IX6'-9':8ginDILMUNkù:habbar nig-aám l g i r mar-tu ba-ííu-U4ku:babbar kù-sigi^ C/-ra-ANmaskimSÙ "S heavy shekels of silver: price of 1 golden mar-tu dagger (with) ring of silver (and) gold: Ura-AN his agent;" MEE VII47 r. vi 15-18: 5 gin DILMUN kù:babbar nig-sám l g i r mar-tu ba-du-U4 kù:babbar-sig,7 Pu-ma-NI Gii-ra-Aru/'", "5 heavy shekels of silver: the price of 1 mar-tu dagger (with) ring of silver (and) gold: Puma-NI (from) Gurakul"; M££ VII47 v. XIII 8-10:10 gin DILMUN kù:babbar nig-sám 1 gir mar-tu ba-du-u,^ kù:babbar-sigi7 E"-"a-N\ lu Bíl-za-Nl, "10 heavy shekels of silver: the price of 1 mar-tu dagger (with) ring of silver (and) gold: Enna-NI the "man" (of) Pilsa-Nl." 18. For the interpretation of gin DILMUN in terms of the so-called "Anatolian" standard of- 11.75 g, and the translation "heavy" for DILMUN see Maiocchi 2005. In the present article I assumed a mina of ~ 470 g = 60 "Eblaite"shekels of- 7.83 g = 50 "Syrian" shekels of - 9.40 g = 40 "Anatolian"shekels of- 11.75 g. Where not specified in the tablets, I assumed that the "Ehlaite" shekel is intended. 19. AilETXII335o.Vl. 20. AÎÊ£X20v.XXIII17. 21. MEE VII 34 r. XIX 12'; 47 r. IV 15. 22. AR£rXII 443 H' 1'; 469 r.' F 1'; 1117 H' 5' (gir ga-me). 23. See Pasquah 2005: 131-32; DAgostino 1996: 218; Waetzoldt 1990: 10-11. 24. For a 0.8 shekels decoration (- 6.26 g) see ARETVll 42 r. III 2-v. I 2: TAR-6 a-gars-garj 5 gir 5 ma-za-ù 4 kù:bahbar gfl-m[fl]-atum-sù "36 (shekels) of copper (for) 5 daggers (and) 5 ..., 4 (shekels) of silver (for) their knobs." 25. See for instance ARETIV 15 o. VIII 7-8. 26. See Ai?£r II 30 o.vi9(where the dagger is not denoted asmar-tu);AilET III 378 r. II 1'; AR£TXII 424 I' 1'; 1491 ir,l';M££XII 37 r. Ill 27, XVII 8. 27. SeeAR£TIII378r. II 1'and also AR£r XII 4241' 1', where the number of daggers is only partially preserved. 28. On the term SAL.US in third-millennium sources see Steinkeller and Postgate 1992: 37-38.

DECORATIVE PARTS AND PRECIOUS ARTIFACTS AT EBLA

5

~ 47 g of both silver and gold." Another instance is provided by MEE XII 37 r. Ill 24-33, where 2Í4 heavy shekels of tin are melted together with 18 heavy shekels of copper to produce the dagger itself (~ 237.94 g of bronze), while lVi shekels of gold (~ 11.75 g) are used for its decoration. g) gir mar-tu ra-'à-tum. The term ra-'à-tum denotes either a decoration made of a material derived from some aquatic animal (presumably shell), or a handle possibly made of this material.'" According to TM.75.G.10236 r. II 6-14, the price of this item is 1 DILMUN shekel (~ 11.75 g of silver). h) gir mar-tu rMi2-ciu-^fl-tMw. The term ru^^-du-ga-tum is related to the West Semitic root *rtq, "to bind." It has been suggested that this term denotes the mounting for decorative parts, usually made of silver." Note especially: MEE XII 35 o. XVII 16-30: \Vi nagga / sub si-in I 2 ma-na TAR a-garj-gar; / 1 gir mar-tu / 1 ma-na sa-pi gin DILMUN kù:babbar / su-bal-ak / sú + sa gin DILMUN kù-sig,; / nu„-za-sw / 1 gin DILMUN kù:babbar / ru^^-du-ga-tum I zú-LAGABxME / 2 si-sù "l'/2 (shekels) of tin to be melted with 2 minas 30 (shekels) of copper: 1 mar-tu dagger; 1 mina 40 heavy shekels of silver to be exchanged for 20 heavy shekels of gold: its decoration; 1 heavy shekel of silver: the mounting of a lahmu-hodk^^ (and) 2 horns (inlaid?)." Note also TM 75.2428 r. I 33-112 (not collated), where the term mar-tu is omitted: Wi kù:babbar / su-bal-ak / V2 kùsig,7 / Vi ZU.PIRIG / 2 gir / 4 gin DILMUN kù:babbar / zi-du-sù II 2 NI kù:babbar / ru,^-du-ga-tum-sù "l'A (shekel) of silver to be exchanged for V2 (shekel) of silver: V2 lion(-decoration)?" (of) 2 daggers; 4 heavy shekels of silver: their decorations; V3 (shekel) of silver: their mountings." Hence, the Eblaite ru^^-du-ga-tum is probably to be considered as Semitic equivalent of Sumerian kesda. i) gir mar-tu ti. This type of dagger is characterized by a precious point, or a precious side part.'" Its purchase price is stated in M££ VII 47 r. VI 7-11: sti+sa-5 gin DILMUN kù:babbar nig-sám l g i r mar-tu ti kù-sigi7 Ti-ra-NI maskim En-mar su-dug in Ma-n'^ "25 heavy shekels of silver: purchase price of 1 mar-tu dagger with golden point: Tira-NI, agent of Enmar, collector/has collected" in the city of Mari." The bodies of these artifacts hoarded ~ 123.33 g of copper, seeM££ VII 34 r. XIX 17: 26 ma-na sa-pi a-garj-gar; UNKENak 1 mi-at mar-tu ti [...] "26 minas 40 (shekels) of copper to make 100 mar-tu daggers with point (...)." The term ti is sometimes denoted as 'à-lum, to be connected with akk. ellu, "clean, pure, holy, sacred,"'* possibly with cultic implications. The purchase price of'à-lum daggers fluctuates between 15 and 20 (heavy) shekels of silver." An additional element denoted as an-dùl is attested in TM.75.G.1560 o. V 8 and TM.75.G.1560 r. VIII 1, both concerning gir mar-tu ti an-dùl kù-sigi^. j)gír mar-tu zu, zú-ak, zú-LAGABxA and variants.'* The term denotes probably a decoration at the end of the blade, perhaps in the shape of a hook or a tooth." The bilingual lists give the translation KA. AN.A.LAGABxHAL

29. The figure is slightly different from the one offered by Waetzoldt (1990:9), who does not take into consideration the fact that the shekels mentioned in the text are referred to as DILMUN (see above note 18). 30. See Pasquali 2005: 72-76; Waetzoldt 1990: 15. 31. See Pasquali 2005: 88-89; Waetzoldt 1990: 11-12. 32. The term zú-LAGABxME is apparently a variant of zú-LAGABxA, see below. 33. On the term ZU.PIRIG see § 2.6. 34. This guess is merely based on the equivalence ti = sétu "side (as a structural part of a manufactured object)," see CAD $ sub fèlu and AHw sub fêlilu. Accordingly, Waetzold (1990: 16) translates gir mar-tu ti as "Martu-Dolch mit Mittelrippe." 35. On the meaning of the term Su-duj see Pomponio 2003. 36. See CAD E sub ellu; Pasquali 2005: 103; Waetzoldt 1990: 14-15. 37. See Waetzoldt 1990: 15 with references, and Table 1. 38. The term zú-LAGABxA is mentioned in ARET I 16 o. XI 2, ARET II 30 o. V 6, and in TM.75.G.1447 o. II 2. Other known variants are: gir mar-tu zu-AN.A.LAGABxHAL (VE 202b), gir mar-tu zú-AN.LAGABxHAL (ARET III 109 o. II 2'), gir mar-tu zúAN.LAGABxHAL.A (ARET III 630 o. Ill 2'; 635 v. IV 5), gir mar-tu zú-LAGABxHAL (MEE XII 37 r. XVII 4), gir mar-tu zú-AN. LAGABxAN (AR£T III 302 o. I 1'; 864 o. Ill 1'), gir mar-tu zú-LAGABxAN (ARETIII 831 o. I 2'), gir ma-tu zú-AN.A.AMA (M££ II 49 o. VII 3; M££Xn 25 v. V 7), gir mar-tu zú-AMA (ARETIV 15 o. V 9), gir AN.A.AMA (ARET VII 9 o. IV 4; M££X1I 25 o. XI 5), gir ZÚ-AN.A.AMA (AR£T VII 16 o. Ill 5). 39. For zú-ak as "hook" see D'Agostino 1995, contra Waetzoldt 1990: 14 n. 76.

6

MASSIMO MAIOCCHI

= st-nu a-ha-mu/m[i] (VE 202b = M££ 4,221) "tooth" (the sign KA hence has to be read zu = sinnu) possibly ofa deity related to water."" Some information on the metallurgical composition of these artifacts is provided in ARET III 630 o. Ill l'-2': 3 ma-na sú-i-sa gin DILMUN an-na / sub 26 gir mar-tu zu LAGABxHAL-a "3 minas 20 heavy shekels of tin: (for?) fusing of 26 mar-tu daggers with a hook." Additional parts denoted as si, "horn(s)," are, on one occasion at least, purchased on the market, as stated in M££ VII 34 o. XVIII 7-11: 1 ma-na nagga / kin5-ak / 'x" mi-at gir mar-tu zabar / ^x"! ma-na ^x^ gin DILMUN kù:babbar / n i g - s á m / si-si-sù. It is unclear if the term si refers to horns or to a horn inlay. The reference to horn in connection with mar-tu daggers is exceptional. Other terms related to daggers are mah, "heavy, big"; hul, "of bad quality"; sa^, "of good quality"; gibil, "new"; and libir, "old." The terms nUn-za (read also sir-za)'" and zi-du are also often mentioned in connection with daggers and other artifacts, denoting a precious ornament usually made of gold. Most probably, both of these terms simply mean "decoration.""^ Hence, they do not denote a particular kind of dagger, as opposed to the ones listed above, but rather an artifact with additional parts."" It is also noteworthy that decorations referred to as nuii-za occur without further reference to the object they modify."" This fact implies that they are evaluated per se, as other standardized items often are, and could circulate independently from the original artifact they were originally attached to. As far as daggers as concerned, some relevant data are given in ARET VIII 534 (= M££ V 14)"5 o. IV 3'-4': 21/2 gin DILMUNbar,:kù / nu^-za 1 gir mar-tu Ib-al¿^ "21/2 heavy shekels: the decoration of 1 mar-tu dagger: the city of Ibal"; note also Aiî£r VIII534 (= M££ V 14) o. IX 5'-12': 9 gin DILMUN kù:babbar / su-bal-ak / 3'' gin DILMUN ku-sigj^ / nu,,-za 1 gir mar-tu 10 kii-SAL / nig-ba / en / ^'A-da I lu halam^ "9 heavy shekels of silver to be exchanged for 3 heavy shekels of gold: the decoration (of) 1 mar-tu dagger (and-) 10 straps:"** gift (of) the king (to) the god 'Ada, the one of Halam."

2.3 The Classification of mar-tu Daggers In order to understand better the circulation of these artifacts among officials and court personnel, we have to classify the various types according to their value, that is according to the amount of precious metals used to make them. The prices (nig-sám) of mar-tu daggers are summarized in Table 2. One notes that the kind denoted as gir mar-tu kù - s ig,^ is by far the most frequently purchased. Its value varies from 60 to 40 shekels of silver, but exceptional values are also attested. Other mar-tu daggers, however, such as 'à-ma-da-num, 'à-NE-LUM, game-ù, GIS.SAL, ra-'à-tum, and ru^^-du-ga-tum are never purchased. This maybe explained partly by the paucity of attestations of some of these varieties. The reason for lack of information on the prices of gir mar-tu ga-me-ù—a. well-attested class of artifact—is not evident. By correlating the figures in Table 2 with the content of some texts listing various types of daggers it is possible to classify some of these artifacts in decreasing order of value, as shown in Table 1:"'

40. Taking a-i¡a-mu as a writing for lahmu or lahamu, see Pomponio and Xella 1984: 27-28. 41. On the reading of this term see Waetzoldt 2001: 51, with previous bibliography. 42. The term zi-du has been connected either to Akk. situm, sädu, or ftdu, see Mander (MEE X, 92). For a possible interpretation as a Semitic equivalent of Sumerian nu n-za see Fronzaroli 1996: 64. Pasquali (2005: 93-94) points out a possible connection with the Semitic root *wsm "to be decorated." 43. This is evident from MEE XII 37 r. I 26-IV 1, where precious parts belonging to various types of daggers (gir mar-tu kù-sigj^, gir mar-tu 'á-/um, gir mar-tu ti, gir mar-tu ba-du-u^,gir mar-tu GiS.SAL) are denoted as zi-du. 44. See for instance M££ VII 2 r. XI 4. 45. M££ Vreadstheinitial figure assusanajj, without further comments. The photo provided in Afi£T VIII actually supports the reading 2. 46. These items are often quoted with harnasses, as already noted by Lahlouh and Catagnoti (2006: 559, with previous references), who translate kù-SAL as "monile." For further references see Civil 2008: 16, 114, 121, 123. 47. See especially ARBTIV 16 r. XIV' 8-XIir 4 (a list of artifacts).; ARET XII335 o. V l'-2', VII3-6, r. 11'-5', VII l'-6', VIII l'-5' (a record


TABLE l. THE CLASSIFICATION OF M A R - T U DAGGERS. Translation

Type

vail

Daggers with golden parts

gir gir gir gir gir gir gir

mar-tu kii-sig,; mar-tu ti kù-sig,^'a-Zwrn mar-tu ti ku-sigj^ mar-tufcfl-£Í«-U4/,kù-sig,^ mar-tu GiS.SAL (kù-sig,;) mar-tu ga-me-ù kù-sig,; mar-tu zu-AN.LAGABxHALkù-sigi^

goiden mar-tu dagger mar-tu dagger with a "pure" gold point mar-tu dagger with a gold point mar-tu dagger with a gold ring mar-tu dagger with a (gold) sheath' mar-tu dagger with a golden handle mar-tu dagger with golden lahtnu-hook Daggers with silver parts

gir mar-tu kù:babbar gir mar-tu ti kù:babbar gir mar-tu {)a-dw-t(4kù:babbar gir mar-tu GiS.SAL gir mar-tu ga-me-ù kù:babbar (gir mar-tu zú-AN.LAGABxHAL) gir mar-tu zú-ak

silver mar-tu dagger mar-tu dagger with a silver point mar-tu dagger with a silver ring mar-tu dagger with a (silver) sheath' mar-tu dagger with a silver handle (mar-tu dagger with a (silver) lahmu-hook) mar-tu dagger with a (silver) hook

Some remarks are in order. First of all, the relative position of gir mar-tu ba-du-u^,,^, GIS.SAL, are merely inferred from ARET II30 o. VI7-11, VIII6-8, r. 19-12,48 and MEE XII37 r. XVII7-9. Second, one has to keep in mind that exceptions may occur especially in regard to golden mar-tu daggers, whose prices fluctuate between ninety and ten heavy shekels. Einally, the decoration called zú-AN.LAGABxHAL attached to the daggers is almost invariably made of gold, while the z ú - a k is presumably made of silver, though this is not explicitly stated in the texts. In other words, there are no attestations of *zú-ak kù-sig,;. Due to paucity of available data, martu daggers denoted as 'à-ma-da-num, 'à-NE-LUM, ra-'à-tum, and ru^2-du-ga-tum are excluded from the listing.

2.4 Patterns in the Circulation of Daggers A dozen Eblaite texts deal with internal (re)-distribution of daggers and other artifacts. Most of them have been discussed by Amadasi Guzzo (1988:123-24) and Pomponio (1998b: 32-39)."" In these documents, precious objects listed in decreasing order of value are allotted to various individuals. Though some variations may occur, the pattern usually followed in these texts may be resumed as follows: 1 gir mar-tu PN,

1 mar-tu golden dagger PN,

of precious artifacts); MEE VII 40 o. I 1-VI 7 (a register of various kind of daggers received by, or allotted to officials). 48. This document is written using abbreviated forms of terms (for instance, (gir] mar, [gir] ga-me, [gir] ba-du). Note also that a (gir) z Ú - a k is mentioned before a(gir)ga-meinr. 111-12. To my knowledge, this fact is exceptional and has not been taken into consideration in the chart provided above. 49. Texts concerning daggers: ARET II 15 r. Ill 1-IV 8; ARET IV 7 r. VII 20-VIII 13; ARET IV 16 r. VII 10-VIII 3; MEE VII 50; TM.75.G.1781 v. VIII 2-13. Texts concerning decorative disks (DIB): ARETX 1 r. XII 7-19; ARETI 10 r. XI 13-24; ARETIV 7 r. VIII 14-IX 4;AREriV8r.V 13-VI4;TM75.G.2281o. II ll-III 18; TM.75.G.1784 r. V 18-VII 2; TM.75.G.1793 v. VII 16-VIII 1;TM.75.G.2242 r. VIII 3-IX 3.

MASSIMO MAIOCCHI TABLE 2. T H E PRICE ( N Í G - S Á M ) OF DAGGERS ( G Í R ) .

Typology

Price (shekels of silver)

Price per item (shekels of silver)

Weight per item

Reference

gir mar-tu kù-sig,; sa, 1 gir mar-tu kù-sig,y sa. 5 gir mar-tu kù-sigi^ sa.

1 ma-na 27 [+N gin (DILMUN)]

127 [-FN (gin DILMUN)]

- 994.41 g

5 ma-na

60

-~470g

AÄfirVIIllo. VI6

MEE II49 r. VI 3

gír mar-tu ku-sigj^ •^6^ gir mar-tu kùSigl7

13 ma-na 20 gin DILMUN

90 gin DILMUN

~ 1057.5 g

MEE X 23 r. IV 4-5

1 gir mar-tu ku-sigj^

54 g in DILMUN

54 gin DILMUN

~ 634.5 g

M££VII34r.XII14"-15"

1 gir mar-tu kù-sig,^

1 ma-na 10 gin DILMUN

50 gin DILMUN

~ 587.5 g

MEE X 22 0. Ill 2-4

1 gir mar-tu kCi-sigi^

50 gin DILMUN

50 gin DILMUN

- 587.5 g

AR£rVm534r.V21

1 gir mar-tu ku-sigj;

50 gin DILMUN

50 gin DILMUN

~ 587.5 g

MEE XII 36 0. XXVII 21-22


50 gin DILMUN

50 gin DILMUN

~ 587.5 g

MEE VII 34 r. XII 6'-7'

1 gir mar-tu kù-sigiy

40 gin DILMUN

40 gin DILMUN

-470 g

AÄ£riI13o. I V l l

1 gir mar-tu kù-sigj^

40 gin DILMUN

40 gin DILMUN

-470 g

AÄ£riII635r. V6'-7'

2 gir mar-tu

2 ma-na

60 (gin)

-470 g

MEE X 29 r. VIII 4-5


1 ma-na

60 (gin)

-470 g

MÊ£ XII 37 o. XXIII 7-8

1 gir mar-tu kù-sigi^

1 ma-na

60(gin)

-470 g

MEE XII 37 0. XXII44-45

1 gir mar-tu kù-sig,7

1 ma-na

60(gin)

-470 g

Af££ XII 36 r. XXVII 1-2


1 ma-na

60(gin)

-470 g

M££XII36r.XII23-XIIIl


1 ma-na

60(gin)

-470 g

MEE XII 36 r. VIII 5-6

1 gir mar-tu kù-sigjy

1 ma-na

60 (gin)

-470 g

MEE XII 36 0. XXXI 20-21


1 ma-na

60(gin)

-470 g

M££ XII 35 o. XXVII 1-2


1 ma-na

60(gin)

-470 g

M££XII36o. V6-7


1 ma-na

60 (gin)

-470 g

M££ XII 36 0. XXVII 10-11


1 ma-na

60(gin)

-470 g

M££ VII 34 o. XVI1-2

1 gir mar-tu ku-sig¡7

1 ma-na

60(gin)

-470 g

M££X20r. XII 18-19


1 ma-na

60(gin)

-470 g

M££X29o. 119-20


1 ma-na

60(gin)

-470 g

AÍ££X29r. VI30-31


1 ma-na

60(gin)

-470 g

AÄErVIII534r.V13


1 ma-na

60(gin)

-470 g

Ai{£r I 45 r.'+IX 3


1 ma-na

60(gin)

-470 g

Af££ XII 35 0. VII45-46



35 gin DILMUN

35 gin DILMUN

-411.25 g

MEE X 29 r. I 25-11 1


40(gin)

40 (gin)

-313.2 g

M££X 20 0. XII 19-20


40(gin)

40(gin)

-313.2 g

M££X 20 0. XVIII 22-23

1 gir mar-tu kii-sigi^

40(gin)

40(gin)

-313.2 g

AÄ£ri44o. IV22


40(gin)

40(gin)

-313.2 g

Aí££ VII47 o.V 15-16

1 gir mar-tu kù-sigj,,

40(gin)

40(gin)

-313.2 g

M££ VII 47 o.V 20-21


40(gin)

40 (gin) •

-313.2 g

Ai££ VII 47 0. XVIII4-5


1 ma-na

30(gin)

-235 g

AR£TVIII534r.XV4'


1 ma-na

30(gin)

-235 g

Aß£TVIII539r. IV11'-12'


40 gin DILMUN

20 gin DILMUN

-235 g

M££ XII 36 0. III 9-10


20 gin DILMUN

20 gin DILMUN

-235 g

Ai££ XII 35 0. II 6-8

1 gir mar-tu kù-sig,;

20 gin DILMUN

20 gin DILMUN

-235 g

M££ XII 35 0.x 14-15


30 (gin)

30 (gin)

-235 g

M££XJ7 37r. IX26'-27'


30(gin)

30(gin)

-235 g

M££XII37o.XVI2-3


30 (gin)

30(gin)

-235 g

M££ XII 37 0. II 6-7


30 (gin)

30(gin)

-235 g

ARET XII 450 i r 4 ' - 5 '


30 (gin)

30 (gin)

-235 g

Ai££ VII 34 0. XVIII 2-3

1 gir mar-tu ku-sigj.

30(gin)

30 (gin)

-235 g _

AJ?£rVIII539r.IX20'


25 (gin)

25 (gin)

- 195.75 g

M££VII47o.VI7-8


20(gin)

20 (gin)

- 156.6 g

M££X 20 0. XII 21-22

1 gir mar-tu kù-sigjy

20(gin)

20 (gin)

-156.6 g

M££ XII 35 r. XXIII12-13

1 gir mar-tu kù-sig,y

10 gin DILMUN

10 gin DILMUN

-117.5 g

M££XII35o. VII7-8

gir mar-tu ti ku-sigj^ 'à-lum 1 gir mar-tu 'à-lum kù-sig,7 1 gir mar-tu ti kùsigi7 / 'à-lum 1 gir mar-tu 'à-lum kù-sig,7

20 gin DILMUN

20 gin DILMUN

-235 g

M££ X 29 r. VII 11-12

20 gin DILMUN

20 g in DILMUN

-235 g

M££X 20 o.V 29-31

15 gin DILMUN

15 gin DILMUN

-176.25 g

M££XII36o. III16-17

43 1/3 gin DILMUN

-509.17 g

AÄ£TII13o.VI8

20 gin DILMUN

-235 g

M££ VII 47 0. IV 3-4

30(gin)

30 (gin)

-235 g

M££X23r.VI2-4

25 (gin)

25 (gin)

- 195.75 g

M££ VII 47 r. VI 9

gir mar-tu ti kù-sig,^ 3 gir mar-tu ti kùSigl7

10 lá-2 gir mar-tu ti kù-sig,7 1 gir mar-tu ti kùSigl7

1 gir mar-tu ti kùsigi7

3 ma-na 10 gin DILMUN 3 ma-na 40 gin DILMUN

10

MASSIMO MAIOCCHI

1 gír mar-tu ti kù1 gír mar-tu ti kùsigi7 1 gír mar-tu ti kù-

M££X29r. III9-10

10 gín DILMUN

10 gín DILMUN

-117.5 g

10 gín DILMUN

10 gín DILMUN

-117.5 g

M££ XII 35 0.1X47-48

10 gín DILMUN

10 gín DILMUN

-117,5 g

M££ X 20 r. XXIV 27-28

Sigl7

Sigl7

gir mar-tu ba-du-u^ 1 gir mar-tu badu-U4 kù:babbar-

9 gin DILMUN

9 gín DILMUN

-105.75 g

AJ?£riII635r.Vl'-2'

8 gín DILMUN

8 gín DILMUN

-94 g

AÄ£riII635r.VI2'-3'

10(gin)

10 (gín)

- 78.3 g

M££ VII 47 r. XIII 9

8 (gín)

8 (gín)

- 62.64 g

M££VII29o.IX7'

5 (gín)

5 (gin)

-39.15 g

M££VII47o.VI17

4(gín)

4 (gín)

-31.32 g

Ai££ VII 34 r. IV 3

Sigl7

1 gir mar-tu badu-U4 kù:babbar kù-sigi7 1 gir mar-tu badu-u^ kù:babbarSigl7

1 gir mar-tu badu-u^ kù:babbar kùSigl7

1 gir mar-tu badu-u^ kùibabbarSigl7

1 gir mar-tu badu-u^

gir mar-tu zú-LAGABxHAL.A 1 gír mar-tu zúLAGABxHAL.A

15 gín DILMUN

15 gín DILMUN

- 176.25 g

Aiî£riII635r. IV4'-5'

gir mar-tu zú-ak 1 gír mar-tu zú-ak

17gín DILMUN

17 gin DILMUN

- 199.75 g

M££ VII 29 r. III 9-10

1 gír mar-tu zú-ak

7 (gin)

7 (gín)

~ 82.25 g

AÄ£T I 44 o.V 3

30 gír mar-tu zú-ak

46 gín DILMUN

~ iy2 gín DILMUN

- 18.01 g

M££VII29o. VlIIlO'-ir

i-na-súm PN, TITLEj es lgir mar-tu ti 'à-/ww kù-si PN2 i-na-súm PN3 TITLE3 es 1 gir mar-tu ti kù-sigj^ PN3 i-na-siim

gave to PN^ TITLE2 then 1 mar-tu dagger with a "pure" gold point gave to PN3 TITLE3 then 1 mar-tu dagger with a gold point PN3 gave to

DECORATIVE PARTS AND PRECIOUS ARTIFACTS AT EBLA PN, TITLE, es l g i r m a r - t u tia-iiw-w^ku-sigiy PN, 1-na-súm PN5 TITLE,

11

PN, TITLE, then 1 m a r - t u dagger with a gold ring PN, gave to PN5 TITLE,

To the documents belonging to these group is to be added ARETXII 313 r. VI 7'-16', whose initial line, transliterated [...] by the authors, is integratable on the basis of existing parallels as follows:

ARET XII313

f

r.

\n

VI

7'. 8'. 9'. 10'. 11'. 12'. 13'.

[I" gir" mar"-tu"] kù-sig,y Bù-da-m

14'. 15'. 16'.

RUj2-zi-^í-lam

lú Ü-gú-sum

es 1 gir mar-tu ti kú-sigi^ lú Bù-da-Nl i-na-súm ses Ra-ù-'tum'^

[1] golden [mar-tu dagger]: Buda-NI the "man" of Ugusum; then 1 mar-tu dagger with a gold point: the "man" of Buda-NI gave to Ruzi-Ilam the "brother" ofRautum.

We see here another instance of a chain of dagger allotments, where the first person receives the most valuable object, and then delivers a less valuable one to the next individual. The people involved belong to the elites of the royal court, as is the case for most of the documents of this group. Buda-NI appears in fact in ARET III 214 r.' IV 3' and ARETlll 458 r. VI12, as well as in ARETl 16 r. V 2 as the dumu-nita^ of Ugusum, who is titled pa^-ses en in the same texts.'" Ruzi-Ilam is also titled pa4-ses en in later texts,^' while in earliest documents he is referred to as ses Ra-ù-tum. The latter is a high-ranking wife (dam) of the king, being mentioned in the first entries in the lists of royal woman (see Archi, Biga, and Milano 1988: 250-59). Hence, this text is to be dated to the late reign of Isar-Damu. Pomponio (1998b: 39) connects this peculiar movement of artifacts with the "promotion" of an individual to a higher status. Thus, this official would get a more valuable dagger suitable for his new position, at the same time giving the one he owned previously to his successor in his old office. This pattern is attested also for circulation of decorative disks and textiles, though the occurrences are few. As mentioned above, the officials involved in these movements of precious objects are mostly pa-ses, en, but a couple of fragmentary texts suggest a similar pattern also outside the royal palace.

50. These observations were already made by Pomponio (1998b: 34), although in relation to the wrong individual. The alleged bu-da-[Nl] in TM.75.G1781 = MEE VII 50 r. VII 18 is actually bu-da-na-im, as correctly read by by Amadasi Guzzo (1988: 123, whose reading pu-tfl[na-im] is apparently to be corrected to pu-tá"-[na-im]) and D'Agostino in MEE VII {pu-tá-na-im, without square brackets). The reading of DAgostino is supported by ARET III 458 o. VI 4, where bu-da-na-im is listed as the recipient of textiles together with ¡l-e-i-iar, su-na-im, naamg-i-giá, and na-am¿-ha-lu, i.e., the same individuals mentioned in TM.75.G1781 = MEE VII 50. In addition, one may note that Buda-I(l) is usually written either bù-da-NÏ or bù-da-ii, that is with bù and not bu. 51. See, e.g., ARET IV 22 r. I 6-7 and ARET IV 23 r. VII 8-9.

12

MASSIMO MAIOCCHI

ARET XII718 r-

II'

I'." 2'." 3'." 4'." 5'." 6'." 7'."

III' r. 2'. 3'. 4'.

[ l g i r mar-tu kù-sigi;] [PN] '"in-na''-s[úm] Is¡i-gur-li-im \ù A-mur-li-im es 1 gir mar-tu [(ti'") kù]-s[ig J [•••] Is^^-gur-li-im su-dug in t/j-r«,^'^

[1 golden mar-tu dagger] [PN] '^gaveM[o] Isgur-Lim the "man" of Amur-Lim; then 1 mar-tu dagger [with] a go[ld point'] [...] Isgur-Lim collector/has collected in the city of Uru.

Isgur-Lim appears as a "man" of Amur-Lim also in MEE XII 35 r. Ill 21-24, in conection with a decorative disk (DIB) of 40 shekels. His activities in the city of U^-ru^^^ are also attested in ARETlll 467 o. VI l'-7'. It is also noteworthy that this individual receives valuable textiles and a dagger related to the anointment of the king of Kakmium in MEE XII 35 o. XX 7-15. Amur-Lim is known from a couple of documents, but the information provided in these texts is not sufficient to outline his office.^^ In any case, his "men" (lu ) seem to carry their functions outside Ebla.

2.5 Additional Precious Parts: gir kun. The term kun is glossed sè-ne-bu^{Nl) wa li-sa-ntim in VE 1372' = MEE IV 337, meaning "tail and tongue" (see Waetzoldt 1990: 19-20; Pomponio and Xella 1984: 28). The term gir kun has been often translated as "curved dagger," in the sense of "shaped as the tail and tongue of an animal."^' This interpretation has been criticized by Waetzoldt (1990: 18-20), who convincingly argues that this is a metal element used to hang the dagger (or better the sheath) from the belt ("Dolch-Aufhängung"). The term kun also occurs alone, referring to a precious object. Though in some context kun may indeed refer to an ornamental tale of an animal or a bull-headed figure, it is rather clear that it may also stand for an abbreviated form of gir kun.'** Further indications may qualify these objects. Accordingly, one can list the following subcategories: a) gir kun kù:babbar, gir kun kù-sigi^, gir kun zabar. They usually come with ceremonial belt (iblá) and sheath (si-ti-tum), with which they form a standard set. The amount of precious metal contained in these artifacts fluctuates between 1 mina 45 heavy shekels and 20 shekels. It is noteworthy that golden sets are mentioned more frequently than the others. The weight of kun daggers is standardized according to the series 60, 40, 30, and 20 shekels, but note that golden and silver sets usually weight 30 or 60 shekels," the ones made of silver-t-gold 20 or 40 shekels (see Table 3).

52. To my knowledge, the PN A-mur-ii-im occurs in ARETl 4 r. XI 9'; ARET I 17 o. IV 8; ARET III 635 r. V 5'; ARET IV 15 o. VII 1; MEE XII 35 r. Ill 23; MEE XII o. XVIII 2. 53. For a Middle Bronze Age representation of such a dagger on a basalt basin see Matthiae, Pinnock, and Scandone Matthiae 1995: 422. 54. As already noted by V^faetzoldt (1990: 19-20; see note 110 for references). 55. As far as silver sets are concerned, note that 40 gin DILMUN is the more frequent value, corresponding to 60 (gin) = 1 ma-na, see above note 18.


13

TABLE 3. G Í R K U N AND TEXTILES.

Reference

Metal hoarded/stored

Artifacts

gir k u n kii-sig,7 1 ib-lá 1 si-ti-tum 1 gir kun 1 ib-lá 1 si-ti-tum 1 gir kun

3 ma-na kù-sigi^ GAXLA 1 ma-na 45 gin DILMUN

kù-sigi7

1 ib-lá 1 si-ti-tum 1 gir kun

lma-na30(gin)kù-sig,7

1 ib-lá 1 si-ti-tum 1 gir kun kCi-

GAXLA 1 ma-na 25 gin DILMUN

Sigl7

kù-sig,7 1 ma-na 20 gin DILMUN kù-

1 i b - l á 1 si-ti-tum 1 g i r k u n

M££VII23r. Vl-2 AilETVII 115 0.1 1 ARET VIII 525 (= MEE V 5) r. XI 6' AR£T VII I I MEE XII 36 r. VIII 3-4

Sigl7 1 i b - l á 1 si-ti-tum 1 g i r k u n k ù -

GAXLA 1 ma-na 10 gin DILMUN

AR£rXII1010o.XIl'-2'


1 ma-na ku-sigjy

M££XII 36 r. VIII10-11


1 ma-na kù-sigi^

Ai££ XII 36 0. XXXI 18-19


1 ma-na ku-sigj^

M££ XII 35 0. VII 39-40


1 ma-na kù-sig,7

AR£riV 18 0.113

1 ib-lá 1 gir kun

1 ma-na kù-sig,^

Aiî£T VII 76 0. Ill

1 ib-lá 1 gír kun

1 ma-na kù-sig[7

M££XII36o. VI 17-18

1 ma-na GAXLA-5Ù

AR£riII 915 0.12


1 ma-na kù-sig,^

A/l£riII732o. III9'


1 ma-na kCi-sigi^

ARET II 4 (= iVÎ££ VII 17) o. V 8-9


1 ma-na kù-sig,.

ARET II4 (= M££ VII 17) r. IV 7


2ma-na kù-sigi^

M££ XII r. XI 5-6


1 ma-na kù-sig,7

ARBT VII 66 0.111


40 gin DILMUN kù-sigi^

AR£riII587r. 12


40 gin DILMUN kù-sig,7

M££ XII 37 r. IX 24'-25'


40 gin DILMUN kù-sig,^

M££ XII 35 0. XXVI 16-17


46(gin)kù-sigi7

AR£r VIII 539 (= Ai££ V 19) r. VII 11'

1 ib-lá 1 gir kun

40(gin)kù-sig,7

AÄ£ri6r. X31


40(gin)kù-sig,7

AÄ£rXII 819 0.13'


40(gin)kù-sig,7

Ai?£rXII796Iiri3


40(gin)kù-sigi7

M££ VII 23 0.1 2


GAXLA 30 (gin) kù-sig,7

Aiî£TXII1047r. i r i '


1 ma-na kù-sig,^

M££ VII 27 r. VII 2-3


[3] ma-na ku-sigj^

M££ VII 34 r. XIII 1-3

13 ib-lá 13 gir kun

6 ma-na 30 [kù-sigjy]

M££XII 37 0. XXV 17-18

701á-lib70lá-l5i'-fi'70lá-lgir-kun

34 ma-na 30 (gin) ku-sig¡7

M££XII27r. 119


30(gin)kù-sigi7

M££ XII 35 0. XXIV 19-20

Sigl7

1 ib-lá 1 si-ti-tum 1 gir kun kùSigl7

14

MASSIMO MAIOCCHI

1 ib-lá 1 gir kun

30(gin)kù-sig,7

M£BXII35r.XIX4-5

1 íb-lá 1 si-ti-tum 1 gir kun

30(gin)kù-sig,7

Ai££ XII 35 0. XIV 29-30


30(gin)kù-sig,7

M££XII35o.XX9-10

1 ib-iá 1 si-ti-tum 1 gir kun

30(gin)kù-sigi7

M££ XII 35 0.1114-15


30(gin)kù-sigi7

M££VII48r. m u


30(gin)kù-sigi7

AJIETXII872O. I l '

1 íb-lá 1 si-ti-tum lgir kun

30(gin)kù-sigi7

AÄ£rXII860ir6'


30(gin)kù-sigi7

AR£rXII708r. V'5'


30(gin)kù-sigi7

ARET Vlll 524 (= MEE V 4) r. VI 21

1 ib-lá 1 si-ti 1 gir kun

30(gin)kù-sigi7

Ai?£TVII54o.IIIl


30(gin)kù-sigi7

AilETIII80o. I l '

1 íb-lá 1 si-ti-tum 1 kun

30(gin)kù-sigi7

AÄ£riII628o. 5 1'


30(gin)kù-sig,7

Aiî£ril2o. V8


30(gin)kù-sig,7

AR£rilr. VIO


30(gin)kù-sigi7

AR£TIV9o.III2

1 ib-lá 1 gir kun

30(gin)kù-sigi7

M££ XII 37 0. VII 3-4


30(gin)kù-sigi7

M££XII 36 r. VII 18-19


30(gin)kù-sigi7

ARET VIII 524 (= Ai££ V 4) r. VI 21



M££XII 36 r. XIX 15-16



M££ XII 36 0. XXVIII 25-26

10 ib-lá lOgir-kun

17(gin)kù-sigi7

Ai££ XII 37 0. XXV 13-14


1 gin DILMUNkù-sigi7

M££ XII 36 r.V 18-19

gir kun kù:babbar 1 íb-lá 1 si-ti-tum 1 gir kun

1 ma-na (kù:babbar)

AR£ri3r.IV12


1 ma-na kù:babbar

AR£T m 265 0. IV 4'


AR£riII168v. I l l '

1 ib-lá 1 gir kun

GÁxLÁ 1 ma-na kù:babbar 10 lá-3 ma-na 10 gin DILMUN kù:babbar 40gin DILMUNkù:babbar

[...] lgir kun

40gin DILMUNkù:babbar

ARErXII724II'l'

1 íb-lá 1 gir kun


M££ XII 35 0. X 8-9

1 íb-lá 1 gir kun

40 gin DILMUN kù:b ab bar

M££ XII 35 0. VIII 12-13

1 ib-lá 1 gir kun

40 gin DILMUN kùtbabbar

M££XII35o.VIIl-2

[1 ib-lá 1 gir kun]

40 gin DILMUN kù:babbar

M££ XII 35 0.1 37-111

1 íb-lá lgir kun


M££ XII 36 r. X 8-9

1 ib-lá 1 gir kun


M££ XII 36 r. XXX 27-28

1 ib-lá 1 gir kun


M££ XII 37 0. XVII 34-35

20 ib-lá 20 gir kun-40


M££ XII 37 o. XXV 21-22

lOíb-lá 13gir kun

M££XII37r. 15-6 AilEriII666o. IV2'


1 íb-lá 1 gír kun

44' (gín) kù:babbar

M££ XII 35 r. XIV 26-27

20íb-lá20kun-40

40 (gín)

1 ib-lá 1 gír kun

40 (gín)ku:babbar


10 lá-2 ma-na kù:babbar

AR£riII42o.VI3' ARET VIII 539 (= M££ V 19) r. VII 11' M££ XII 37 r.X 20-21

21 íb 301á-2 gír kun-sa-pí

15 ma-na 10 gín DILMUN

M££ XII 37 r. X 26-27


30(gín)ku:babbar

M££ XII 35 r. II 42-43

1 íb-lá 1 si-ti-tum 1 gír kun

30(gín)ku:babbar

M££ VII 47 0. VIII 5


30(gín)ku:babbar

M££ VII 34 r. XIII17-18

[...] 4gír kun-30

30 (gín)

AR£rXII1080iri'


20 (gín) kù:babbar

Aí££ XII 35 r. XXIII 6-7

1 íb-lá 1 gír-kun

20(gin)kù:babbar

AR£r VIII 539 (= M££ V 19) r. IX 8'

1 íb-lá 1 gír-kun

12 gín DILMUN kù:b ab bar

M££ VII 34 o. XI 16-17


1 gín DILMUNkùrbabbar

Aí££ XII 35 o. XX 16-17


gir kun kù:babbar MAS.MAS kù-sigjy 40(gin)kù:babbarMAS.MA§kù-

15

AR£TXII 809 0.13

SÍg,7

1 íb-lá l g í r kun 1 íb-lá 1 gír kun bar5:kú MAS. M A S kù-sig,7 LAGABXLA-SÙ 1 íb-lá 1 gír kun M A S . M A S kù-

20(gin)kù:babbarMA§.MAS

AR£ri3r.X14

20 gín DILMUN kù:babbar

Ai{£TIV13o. 12

20(gin)kù:babbar

Aií£TIlr. VIII15

SÍg,7

b) gir kun (kù:babbar) M A S . M A S kù-sig,^. These artifacts have extra decorations made of gold, referred to as M A S . M A S , possibly denoting "stripes" or "bands" of precious metal (see Waetzoldt 2001: 68-70). The accessory parts are presumably made of composite materials, being silver and gold sometimes referred to as decorations (ni-zi-mu) of gir kun.'' The term M A S . M A ^ appears also in connection with spears (gis-gu-kak-gid), straps (kii-SAL), and in the abbreviated' form MA§ with lion-head decorations' (ZU.PIRIC)." c) gir kun (kù:babbar) TAR.TARkù-sig,,. The termTAR.TARmaybe readhas-has in relation to nonworked metal,'' but its reading and meaning in relation to belts and daggers remains unclear. The comparison with the above-seen MA§.MA§ suggests an additional element made of precious metal, attached also to bracelets (gúli-lum). As already noted by Waetzoldt (2001: 73), the TAR.TAR-parts are rather heavy compared to the standard weight of bracelets and belts, being composed by an amount of precious metal fluctuating between 5 and 22 heavy shekels. d) gir kun si. This kind of artifact has some extra decorations, either made of horn or ivory, or in the shape of horns (si). On the one hand, the latter assumption nts well the interpretation of gir kun si as "part of a dagger

56. See for instance MEE VII 27 o. VIII 10-IX 4: 5 gin DILMUN ku-sigj^ / ni-zi-mu I MAS.MAS / 4 ib-lá 1 gir kun M A S . M A S "5 heavy shekels of gold: the decoration of stripes (of) 4 ceremonial belts and 1 pendant." 57. As far as spears are concerned, see Waetzoldt (1990:4). MAS.MAS related to kù-SAL are attested in M££XII 35 r. Ill 36-40 (etpassim): 5 gin DILMUN kù:babbar / § u - b a l - a k / 1 gin DILMUNkii-sigj^/ «i-zi-mu / MAS.MAS4kù-SAL. Stripes? incorrection to lion-head decorations? occur in M££XII 35 r. I 36 ( M A S - Z U P I R I G ) . 58. For a discussion on the meaning and reading of TAR.TAR see Waetzoldt 2001: 71-73; Civil 2008: 54 with n. 94,66.

16

MASSIMO MAIOCCHI

shaped as "tail" (and/or: "tongue") (and) with horn(s)," possibly referring to a dagger with bull or ram features.^' As far as animal shaped figurines are concerned, note also the exceptional set composed by a ceremonial belt, a kun element, and a decoration made of, or in the shape of, feathers (teg""^''")^'' mentioned in ARET III 365 o. II 2'. On the other, one also has to keep in mind that mar-tu daggers and textiles may have horn decorations. In this case, an inlaid made of this material in the handle of the artifacts seems more suitable. The parts denoted as kun and si are also decorated with precious metals, as stated in ARET IV 23 r. Ill 1-5: 7 ku-sigjy / nu„-za 1 kun /13 kùsigi5 / nui[-za 1 si "7 (shekels of) gold: decoration (of) a "tail" (and/or: "tongue"); 13 (shekels of) gold: decoration (of) a horn (inlaid')." The same happens in AilET VIII 528 (= MEE V 8) o. II 5-6: 15 gin DILMUN kù-sigi^ / 1 nu,i-za kun-si "15 heavy shekels of gold: the decoration (of) a "tail" (and/or "tongue") (with') horn(s) (inlaid')." To my knowledge, kun daggers are never purchased on the market.'''

2.6 Other Structural and Decorative Elements of Daggers Other terms related to daggers occur only sporadically in the Ebla archives. These are: gi-ba-du, "metallic ring" (see Pasquali 2005: 149), gu-ma-lum, "hilt" (see Pasquali 2005: 139) made of wool and precious metal, ma-ra-zatum, possibly ma-ra-d-tum, "chain?,"*"^ perhaps to be compared with ma-ri-tum "an accessory ofa pendant or a seal."^^ Also noteworthy is the occurence of the term ZU.PIRIG (and the variant SU.PIRIG) in relation to daggers. This might refer to some form of decoration, either made of a hard stone, similar to the skin of a spotted "lion," or to a technique used to obtain a "leopard skin effect," as suggested by Bonechi (2003: 83-86). In some contexts ZU.PIRIC might also refer to lion shaped figurines (hence the sign sequence is to be considered as a graphic variant ofug). MEE XII35 r. 129-39 reads: 2^2 nagga sub si-ini 16 gin DILMUN a-gars-garj/MAS kù:babbar / 21/2 kù:babbar / su-bal-ak /1/2 kù-sigiy/ MAS ZU.PIRIG / 2 gir / 4 gin DILMUN / zi-du-sù ^Vi (shekels of) tin to be melted with 16 heavy shekels of copper: a silver stripe?; 2y2 (shekels) of silver to be exchanged for Vï (shekel) of gold: a stripe? (for attaching?) the "lions"? (of) 2 daggers; 4 heavy shekels: the relative decoration(s)." A golden lion head figurine dating to the third millennium B.C.E. has actually been excavated in the royal palace of Ebla (see Matthiae, Pinnock, and Scandone Matthiae 1995: 325, 338). Despite the paucity of attestations, these elements could have been reasonably common as additional or structural parts of daggers and other objects. Unfortunately, the information provided in the available texts is not enough to allow general conclusions, or to evaluate properly the amount of precious metal involved in these elements.

3. Decorative Disks (DIB) Together with bracelets {gú-li-lum), decorative disks or plaquettes (DIB) are the artifacts more commonly mentioned in the Ebla archives. Nevertheless, the understanding of the term DIB remains uncertain. A. Archi in ARET VII translates "metal rod or bar," while F. DAgostino in MEE VII writes of a "metal layer." Finally, H. Waetzoldt

59. Battering-rams (g u,- s i - d i 1 i ) are well attested at Ebla and in Mesopotamia. They are composed by various elements, among which the tongue (eme), the tooth (zu), and apparently horns (si), see Pasquali 2005: 38. 60. On this term see most recently Civil 2008: 123-24, where it is noted that this artifact can be used to decorate also some military implement, possibly a helmet, and the yoke of a vehicle, in which case it "does not need to be made actually from a tuft of eagle feathers, it could simply take its original shape and its name from the bird." 61. To the contrary, the prices of mar-tu daggers are often mentioned in the texts. Note that MEE XII 37 r. IX 22'-32', lists a gir kun with the relative weight after the purchase (nig-sám) ofa mar-tu dagger. 62. See MEE XII 36 o. V 19, and Pasquali 2005: 167. 63. See Archi (2002: 193-94), listing the few known occurrences.


17

TABLE 4. SIXTEEN-SHEKEL DECORATIVE DISKS. Manufactured goods

Shekels

Further data

Agent

Reference

textiles, 1 DIB

16

Puzurra from Arugadu

su-dug

AK£riV3r. VIO

textiles, 1 DIB

16

Iti-NI from 'Ada collector of GardaNEdu in Litiba

su ba^-ti

AJlErVIII533o.XII10

lDIB

16

EN-daza agent of Kitir

-

AR£rVII 54 0.112

-

Ai{£TVIII527r.XIIl'

-

Ai?£TVIII534r.XIII18'

gi; é-en

AÄ£rVII54r. Ill

-

AÄ£TVIII533o.XII18

-

AK£ri4r. IXll

'Ane "man" of Taldagar collector in Adum (for) Anisum agent of Ruzi-Malik Abusgu from GN commission (su-ra) in Arbatum

lDIB

16

lDIB

16

lDIB

16

1 set oft., 1 DIB

16

1 set oft., 1 DIB

16

1 set oft., lDIB

16

Adasa "man" of Hara-NI

su-dUg

AR£TVIII533o.VIII4

1 set oft., 1 DIB

16

Kun-Daban "man" of Iddi(n)-Agu collector in Aha-Damu

-

Aiî£T VIII 533 0. XII 4

1 set oft, lDIB

16

Ilum-Bal agent of Ibbi-Zikir collector

-

AÄ£TVIII523o. V6

-

Aiî£rVIII527r. VII13

-

AR£T VIII 524 o.XIIl

-

AilErVIII540o.VIII7

-

AÄ£TVIII524r. VII8

Istama from Dur Iluza-Malik from Adab collector (in) Za'ar A-pusqu from GN commission (su-ra) in Arbatum

Puzurra-Malik agent of Redi collector in Dammilu Gila-Malik "man" of Sura-Karru collector in GN Napha-NI "man" of Ra'izu collector in IliNE Puzurra-Malik from Madanu collector (in) Asuba

1 set oft., 1 DIB

16

textiles, 1 DIB

16

1 set oft., 1 DIB

16

1 set oft., 1 DIB

16

1 set oft., 1 DIB

16

Zikir-NI

su ba^-ti

AR£rVIII536o. V i r

1 set oft., 1 DIB

16

'Ati from Burman collector in NINEzigu

-

AR£T VIII 527 0. XV 23

1 set oft., 1 DIB

16

Damda-NI from Dasaba collector [...]

-

AR£rVIII530r. 114

1 set oft., 1 DIB

16

Massur-ahu

-

AilErVIII 530 r. VIII

1 set oft., lDIB

16

(for) PN from 'Ane

i-na-súm

AilErVIII527o.XVI26

2 sets oft, 1 DIB

16

PN from GN

-

AR£rVIII541r. 110'

16; 20

PN|_2 "men" of PN,,^. collectors in GN,.^

-

AR£rVIII527o.XI14

16; 30

for the anointment of PN,.2

-

Ai?£rVIII540o. V16

t, 1 DIB-20 1 DIB-16 1 DIB TAR 1 DIB16

prefers the translation "Schmuckscheibe, Schmuckplatte" (see Waetzoldt 2001: 77), stressing that decorations of various kind may apply to these objects. In this regard, noteworthy is the mention of a golden "decorative disk without stone(s)" (1 dib / ba-lu-ma na4) in ARETXll 708 r. IV' 2'-3', which suggests that stone elements might be a standard feature of these artifacts. If Waetzoldt's interpretation is correct, one may think of the numerous

18

MASSIMO MAIOCCHI

golden studs belonging to the Middle Bronze Age found at Ebla (see Matthiae, Pinnock, and Scandone Matthiae 1995: 483, 524). In addition, a circular golden disk, found in level IV A, previously analyzed as a shaft handle (see Matthiae, Pinnock, and Scandone Matthiae 1995: 331, 342), may represent what is epigraphically known as DIB. It is uncertain if the cross that is carved on the top refers to its weight, or to its quality (the sign mas was used for an alloy between two equal quantities of metal), or again is to be considered as a rough factory mark, rather than just a sign left by a temporary owner. Its weight is unfortunately not provided in the catalog. On the basis of its dimensions, one may speculate that this object is actually a 16-shekel gold disk, which is a common weight for this kind of artifact." Decorative disks were very carefully standardized (see Table 5). Like daggers, they may have participated in the circulation of silver and gold since in some cases they are decorated with precious metals. Let us consider one interesting example of their circulation:

ARETXll 719

r.

I'." 2'J! 3'." 4'." 5'." 6'." •J' !!

8'." 9'." 10'."

[1 DIBsa-pi-kù-sig,^] [PN] [in-na-súm] [Bar-za-ma-ù]

es lDIBTARkù-sig,, Bar-za-ma-ù in-na-súm En-na-il lú A-ma-ga

[1 decorative disk of 40' shekels of gold] [PN] [gave to] [Barzamau] then 1 decorative disk of 30 shekels of silver Barzamau gave to Enna-Il the "man" of Amaga

The initial lines are restored on the basis oí ARET 110 r. XI13-24. In this text, a decorative disk of 50 shekels, referred to as gibil, is given to Raizu, who gives to Iphur-NI another disk of 40 shekels. Finally, Iphur-NI returns (gÍ4) to the palace a disk of 35 shekels. To my knowledge, almost half of the recipients of decorative disks are officials acting in various cities under Eblaite control. Such officials are said to be collectors (su-dug, dabj-LÚxTIL)**' in charge of gathering taxes and tribute (see Table 4). Coing back to Table 5, one may note that golden decorative disks are by far the more numerous. According to the overall amount recorded in the texts published in ARET I-IX, MEE IV and MEE XII, the gold casted in these artifact is 136;23 minas (~ 64.1 kg), almost the double in comparison with silver (62;35 minas, that is -29.41 kg). Three possible explanations may apply. First, we may consider that the primary function of decorative disks was the storage of precious metals, gold being the most suitable for such a task. Second, it seems that there was a surplus of gold at Ebla. This consideration is connected with the fluctuating exchange ratio between silver and gold. The ratio seems to change from 5:1 in the earlier texts to 3:1 during the final period covered by the archives, but

64. The disk is 32 mm in diamenter, 9 mm in height. The specific weight of gold is 19,25 kg./cm'. Hence, a golden cylinder with the same measurements weights 139.18 gr (nr2-h-19,25), which is about 18,24 Eblaite shekels. The artifact is pillow shaped, which suggests a slightly smaller value. 65. On these officials see Pomponio 2003.


19

there is no full agreement on this topic among scholars.*^^ It is nevertheless possible that the central administration may have stored a part of its surplus in this manner. Third, we may have to reconsider our data. As we will see, administrative documents record several instances of remelting, recycling, and sharing artifacts. This common practice may lead us to overestimate the real amount of metal that was actually circulating, due to a double counting of specific objects, passing among different persons, and therefore there might not be any surplus of gold at Ebla during the time covered by the archives.

4. Selected Decorative Elements Additional parts made of silver and gold are often attached to artifacts and textiles to increase their value. For example, ARET III 630 o. 4'-8' records a certain amount of tin for the fusion of eagle decorations (teg"""*"") as additions (diri) to the ones already present on gurus-objects, probably decorative parts of bracelets (gú-li-lum)." Eagle-decorations are also mentioned in partially broken context after a list of daggers (gir mar-tu ba-du-u^, g i r mar-tu zú-AN.LACABxHAL, gir mar-tu zú-ak) in Aiî£r III 109 o. I l'-4'and AßET III 110 r. V l'-4'. As mentioned above, ARET III 365 o. II 2' lists these items as parts of a set composed by a ceremonial belt and a g i r kun. Hence, it seems that eagle decorations could be attached to various kinds of artifacts. Some decorations seem peculiar to specific objects, though there are exceptions. Thus the "hook" (zú-AN.LAGABxHAL and zú-ak) is almost invaribly associated with mar-tu daggers, but in a few instances one can observe the use of this additional part with kun elements.^* In addition, hooks are sometimes counted per se, although in contexts related to other objects, with which they might be related.® This fact suggests that they are actually removable. This is partly supported by MEE XII 37 o. IX 4-7, where a decorative hook (zu -LAGABxHAL), which is typical of m a r -1 u daggers, is attached to a bracelet {gú-li-lum).''° Another hook (zú-ak) is exceptionally related to a gir kun in ARETW 23 r. V 12. In addition, parts made of horn, commonly found in connection with kun daggers, may sometimes be attached to mar-tu daggers. As mentioned above, kun-elements and horns (si) decorated with gold are also recorded as accessories of valuable textiles.^' Though there is no conclusive evidence to relate these parts as the ones that usually characterize daggers, the above-mentioned evidence seems to hint at this interpretation. Removing precious and semi-precious parts might have been a common practice in the (re)distributive economy of Ebla. The attention of the central administration was in fact focused not only in the artifact itself, but also

66. The discussion is mainly based on the internal chronology of the annual accounts of metals, see Pomponio 1998a. 67. The text reads: sa-pi-5 gin DILMUN an-na / §ub 10 lá-2 teg""*™ / diri / te^"'"''" / gurus-guruá. The interpretation of gurus as an additional element attached to a bracelet {gú-li-lum), possibly peculiar to workers, was already proposed by Archi and Biga (1982: 357). Comparative evidence includes the following texts: ARETI 14 r. VIII 5-12: 2gú-U-lum a-gar5-gar5 kù-sig,^ 1 gurus / 1 gir mar-tu baííu-u^ku-sigi^/lú Is-má-ga-lu / i-na-súm / LAGABxÚS-ra-ar/dumu-nita //fc-â-Í5-/w "2 bracelets of copper and gold (with) 1 guruádecoration, 1 mar-tu dagger with golden knob: the "man" of Iämagalu Gave to PN son of Ibgaislu;" ARETVIII 525 r. XII 8-11: 1 SAL-TÚG 1 ib-III-TÚG DAR Bar-zi nagar UNKEN-ak 1 gurus kah-gunû "1 SAL-garment, 1 triple-colored belt: Barzi the carpenter to make 1 guruä-inlaid-object." The term gurus in connection with bracelets is found also in ARÊTI 14 r. I 6-II l;AR£riV lOo.I 1-5, o. IV 9, o.V 6, o. 1X5; AR£T IV 11 v.II-2-7. 68. See for instance AÄETIV 23 r.V 12: 2 gir rkuni zú-ak. 69. See ARET VIII 528 v. VI 14; AR£TXII 956 V l'-3'; AR£rXII 1178 I' l'-5'; MEE XII 3 r. I 1-7. 70. 5 ma-na TAR kù:babbar / Su-bal-ak / 1 ma-na 6 gin DILMUN / kù-sig,; / 1 gú-li-lum zú-LAGABxHAL-4 "5 minas and 30 shekels of silver to be exchanged for 1 mana and 6 heavy shekels of gold: a quadruple bracelet with lahmu-hook" For a typological subdivision of bracelets see Mander 1995 and note Waetzoldt 2001: 547. 71. See Afi£r I 29 r. I 3; ARET II 4 (= MEE VII 17) r. V 4,6, r. VIII 6-7; ARET II 40 o. Ill 2-3.

MASSIMO MAIOCCHI

20

TABLE

5. DECORATIVE DISKS (DIB).

160140120100806040-

i 1

20-

•JL.

6 10

16 20

30

40

I II , 50

60 Shekels

70

80

90

100

120

I I Decorativedisksexplicitly stated as of gold, ex.: N ma-na kù:babbar su-bal-ak N gin DILMUN NDIB H Decorative disks not explicitly stated as of gold, ex.: N DIB N (gin DILMUN) Sources: ARET I-IX, MEE IV, X, XII

in its typology.''^ From this point of view, decorative objects assume various functions: as ceremonial gifts,'^^ as a standard for "payments" (artifacts having a standard weight), as vehicles for the storage of silver and gold. It is also interesting to compare the data available for the above-seen mar-tu daggers with those relative to extra decorative parts. Very few texts mention both gir mar-tu and gir kun in the same context,^^ but when this happens, gir-kun objects are usually placed before mar-tu daggers. This fact may suggest that the former artifacts are more valuable than the latter ones. Accordingly, the precious metal hoarded in the most valuable daggers probably comes from these additional parts.

72. Precious objects were often evaluated per se, without any further indication of their weight or value in shekels of silver; see for example ARET VII 23: 1 an-zam^ 1 kur^-sw 1 a-za-mu kù-sig,^ si mu-tiim I-bi-zi-kir en in Bù-zu-ga'^ iti I-ba^-sa "1 a.-container (with?) 1 handle? 1 a.-container (decorated with) gold (and) horn: contribution (of) Ibbi-Zikir (to) the king in (the city of) Buzuga. Month: Ibasa." On this topic see also Milano 2003b. 73. See, for example AR£r IV 24 o. V 1-3 and ARETVUl 531 o. I 7-11, both of which mention daggers assigned on the occasion of the anointment of Ibbi-Zikir. Daggers and other artifacts are also mentioned for the ritual called ÉxPAP, perhaps a funerary celebration, see ARET 11 o. VIII 15-IX 7, r. X 13-19. On the term ÉxPAP see D'Agostino (1996: 87). 74. OneofsuchtextsisAJlBTI31,seeo. 11:1 ^-zi-dum'"« 1 zarae'^M ib-«lá»-3 dar TUG sa^ 1 ib-lá si-ii-dum gir kun kù-sig,-, 1 gir mar-tu kù-sigi^ 1 banäur kù-sigi^; o. Ill 1: 2 ib-lá si-ti-dum gir kun kù:babbar 2 gir mar-tu kCi-sig,,; o. IV 1: 1 ib-lá si-ti-dum gir kun kii-sigj^l gir mar-tu kù-sig,^ 1 banäur k


21

5. The Remelting of Metal Objects and the Term TIL As clearly shown by Milano (1991), ARET VII 51 carefully records the recycling of precious metal. In this document, three sets of silver bracelets are melted in order to cast an equal number of low quality golden decorative disks. More specifically, the golden decorations ("supports or additional parts" zi-bù, and "metal layers" ni-zi-mu) are first taken, then the silver of the bracelets is exchanged for low quality gold (the ratio is 3:1). This metal is melted with the golden decorations (presumably all having the same ratio) in order to cast golden decorative disks weighing 16 shekels each. The final section of this document follows the same pattern, but additional gold, referred to as MU.MU.MU, is added, pehaps to compensate for the loss of metal during the casting process. Unfortunately, the evidence for the complete process of remelting metal is scanty. A couple of texts mention the (re)melting (sub) of bronze items to cast new artifacts, but the interpretation of these passages is not completely clear. For instance, ARET II37 o. II 1-r. 13 records the (re)melting ofa decorative disk, possibly to obtain precious metal for the decoration of another disk: súH-sa gin DILMUN kù-sigjy / 1 DIB / ir-az-NI / Ab-su^ I sub / ni-zi-mu 11 DIB-5M "20 heavy shekels of gold: 1 decorative disk: Iraz-NI of Absu: to be mehed with (or: for the melting of) the decoration of 1 decorative disk belonging to him." One may wonder just how common this practice was. We may find some clues by conducting a comparative analysis of the decorative disks mentioned in the published documentation. One notes that the 16 shekels disks, which are unexpected in terms of standardization, are the most attested, the complete series being 10, 16, 20, 30, etc. (see Table 5). Several instances of 16-shekel silver decorative disks are known from many documents in various contexts (see Table 4). Usually they are delivered together with textiles to agents (maskim), "men" (lú) of court officials, or collectors (su-dug) in various cities. It is therefore plausible that the remelting of metal was a widespread practice, even if it is described in detail only in the text quoted above. It is also possible that the common references to "good" ( s a j and "bad" (hul) metal may be related to remelting practices similar to the one oí ARET VII 51. One can get a fairly good idea of the amount of good quality metals involved in such an evaluation system by the following references: 2 [-t-x] ma-na 7 gin DILMUN kú:babbar nig-sám 1 gir mar-tu kù-sig,ySa6,"2 [-i-x] minas (and) 7 heavy shekels (of) silver: purchase (of) 1 excellent mar-tu golden dagger" (MEE II49 r. VI 1-3); 5 ma-na kù:babbar nig-sa,(, 5 gir mar-tu kú-sig,, sa^ ás-ti Kab-luçul^,"5 minas (of) silver: purchase (of) 5 excellent mar-tu golden daggers in (the city of) Kablul" (ARETVll 11 r. VI4-8). Since the values attested here do not differ from standard price of daggers, we may conclude that the metal circulating in administrative context was primarily of good quality, and there is hardly any information on low-quality metal, but note the exceptional amount of 1167 Vi minas (~ 548.72 kg) of low-quality gold attested in ARETVll 69, a monthly account of precious metal and textiles. A few texts mention low-quality mar-tu daggers," but without any specific information. Decorative objects of composite metals may also imply the remelting or reuse of precious parts. Examples of this are bracelets of silver and gold (gú-li-lum kù:babbar-sigi7), gold and copper (a-gars-garj kú-sig,y), and the above mentioned daggers of silver and gold. There is also some evidence for incomplete cycles in the circulation and recycling of silver and gold. The text ARET 110 documents a decorative disk that was returned into the hands of the administration, and this may be interpreted as an instance ofa partially used item. We know from other texts that sometimes metal objects were characterized as TIL (also in the reduplicated form TIL TIL). This term was previously understood as "destroyed" by Pettinato (1988:311), or rather "to end (a commercial trip), to arrive" by Pomponio 1989: 301-5). To the contrary, DAgostino understands the objects characterized by TIL as "lost" (see DAgostino 1996: 48, 78). More recently, L. Milano suggests that in commercial contexts this term has to be understood as "(goods) sold, completely used" (Milano 2003a: 415-16). According to the (re)distributive practices seen above, in a context regarding artifacts this meaning may be clarified further as "worn out."

75. The term hul denoting gir mar-tu is found in ARETIV 16 r. II 6; ARÊTXII 36 IF 3'; 388 I' 1'; Ai££ VII 38 o. X 1.

22

MASSIMO MAIOCCHI

6. Conclusions The total amount of silver and gold cast in small artifacts (pendants, bracelets, containers, decorative disks) delivered to workers, officials, and court personnel over the approximately fifty-year period covered by the archives is small if compared to the large amounts of precious metal exchanged with foreign courts or given as gifts to gods and temples. The (re)distributive patterns described here reveal that the amounts of precious metal actually circulating in the Elba state economy were actually smaller than we previously thought. There can be no doubt that the circulation of valuable artifacts implies a real movement of silver and gold. Nevertheless, the internal (re)distribution seems to be only a small—though significant from the point of view of reconstructing ancient economies—part of administrative practice. Silver and gold have two distinct contexts of circulation: on the one hand, horizontal movement towards foreign cities and gifts to temples, involving heavy and valuable objects; on the other, vertical movement in a (re)distributive context, characterized by practices of sharing, remelting, and recycling small quantities of metal cast in decorative objects. The study of such epigraphic material may provide important keys to an understanding of pre-monetary economies, which were based on princeiples not directly analyzable by means of modern categories such as "market economy," "purchase price," "demand and offer," or "inñation and defiation."'^^ Understanding (re)distributive patterns and the peculiar (re)use of decorative objects is therefore the first step in the difficult research on ancient economic systems, and provides a complementary approach to general conclusions based on the annual accounts from Ebla.

References Amadasi Guzzo, M. G. 1988 Remboursement et passage de propriété d'objects en métal précieux. Pp. 121-24 in Wirtschaft und Gesellschaft von Ebla, ed. H. Waetzoldt and H. Hauptmann. HSAO 2. Heidelberg: Orientverlag. Archi, A. 1985a Circulation d'objets en métal précieux de pouds standardisé à Ebla. Pp. 25-34 in Miscellanea Babylonica. Mélanges offerts à Maurice Birot, ed. M. Durand and J. R. Küpper. Paris: Editions Recherche sur les Civilizations. 1985b Les rapports politiques et économiques entre Ebla et Mari. Mari 4: 63-83. 1986 Die ersten zehn Könige von Ebla. ZA 76: 213-17. 1987 Ebla and Eblaite. Pp. 7-17 in Eblaitica. Essays on the Ebla Archives and Eblaite Language Vol. I, ed. C. H. Gordon, G. A. Rendsburg, and N. H. Winter. Winona Lake, IN: Eisenbrauns. 1993a Bronze Alloys in Ebla. Pp. 615-25 in Between the Rivers and Over the Mountains, Archaeologica, Anatolica et Mesopotamica Alba Palmieri Dedicata, ed. M. Frangipane, H. Hauptmann, M. Liverani, P. Matthiae, and M. Mellik. Roma: Dipartimento di Scienze Storiche e Antropologiche dell'Antichità, Università di Roma "La Sapienza." 1993b Trade and Administrative Practice: The Case of Ebla. AoE 20: 43-58. 1999 The Steward and His Jar. Iraq 61:147-58. 2002 Jewels for tbe Ladies of Ebla. ZA 92:161-99. 2005 Minima ebliaitica 19: na-gu-lum /naqqurum/ "to engrave." N.A.B. U.I4Q. Archi, A., and Biga, M. G. 1982 Testi amministrativi di vario contenuto. Archivio L. 2769: TM.75.G.3000-4101. ARET III. Roma: Università degli studi di Roma "La Sapienza." Arcbi, A.; Biga, M. G.; and Milano, L. 1988 Studies in Eblaite Prosopography. Pp. 205-306 in Eblaite Personal Names and Semitic Name-giving. Papers ofa

76. The debate on pre-monetary systems is a too broad a subject to be discussed here. For a discussion about circulation of standardized objects, the use of silver as value indicator, on alleged technical development as a background for Lydian coinage, and the historical context of metals circulation in the third millennium, see Milano 2003b. For a general overview on household economy and economic history of ancient Mesopotamia see Renger 2001.


23

Symposium Held in Rome, July 15-17, 1985, ed. A. Archi. ARES I. Roma: Università degli studi di Roma "La Sapienza." Biga, M. G., and Pomponio, F. 1990 Elements for a Chronological Division of the Administrative Documentation of Ebla. JCS 42: 179-201. Bonechi M. 2003 Leopards, Cauldrons, and a Beautiful Stone. Notes on Some Early Syrian Texts from Tell Beydar and Ebla. Pp. 75-96 in Semitic and Assyriological Studies Presented to Pelio Fronzaroli by Pupils and Colleagues, ed. P Marassini et al. Wiesbaden: Harrasowitz. Buccellati, G. 1992 Ebla and the Amorites. Pp. 83-104 in Eblaitica. Essays on the Ebla Archives and Eblaite Language, vol. 3, ed. C. H. Gordon, G. A. Rendsburg, and N. H. Winter. Winona Lake, IN: Eisenbrauns. Civil, M. 2008 The Early Dynastic Practical Vocabulary A {Archaic HAR-ra A). ARES IV. Rome: Università degli studi di Roma "La Sapienza." D'Agostino, F. 1995 «zú-aka» = «gancio» nella documentazione epigráfica di Ebla. NABÍ7/13. 1996 Testi amministrativi di Ebla. Archivio L. 2769. MEE VII. Rome: Università degli studi di Roma "La Sapienza", Dipartimento di Studi Orientali. Fronzaroli, P. 1996 A propos de quelques mots ébla'ítes d'orfèvrerie. Pp. 51-68 in Tablettes et images aux pays de Sumer et dAkkad. Mélanges offerts à Monsieur H. Limet, ed. Ö. Tunca and D. Deheselle. Association pour la Promotion de l'Histoire et de l'Archéologie Orientales, Mémoires 1. Leuven: Peeters. 2003 Testi di cancelleria: i rapporti con le città {Archivio L 2769). ARET XIII. Rome: Università degli studi di Roma "La Sapienza." Krebernik, M. 1982 Zu Syllabar und Ortographie der Lexikalishen Texte aus Ebla. ZA 72: 178-236. Lahlouh, M., and Catagnoti, A. 2006 Testi amministrativi di vario contenuto. ARET XIl. Rome: Università degli studi di Roma "La Sapienza." Maiocchi, M. 2005 Sul valore pondérale del siclo "DILMUN" nei testi di Ebla. Kaskal 2: 43-53. Mander, P. 1982 Osservazioni al testo amministrativo di Ebla M££ 11453 (= ARET II13). OA 21: 227-36. 1995 The gú-li-lum (Bracelets) in the Economie Texts from Ebla. OAM 2: 41-87. Matthiae, P.; Pinnock, F.; and Scandone Matthiae, G. 1995 Ebla: Aile origini délia civiltà urbana. Milano: Electa. Milano, L. 1980 Due rendiconti di metalli a Ebla. SEb 3: 1-21. 1991 Circolazione di recipienti d'oro e d'argento in Siria nell'Antico e Medio Bronzo. Scienze dell'Antichità 5: 353-68. 2003a Les affaires de monsieur Gida-Na'im. Pp. 411-29 in Semitic and Assyriological Studies Presented to Pelio Eronzaroli by Pupils and Collegues, ed. P. Marassini et al Wiesbaden: Harrassowitz. 2003b Sistemi finanziari in Mesopotamia e Siria nel III millennio a.C. Pp. 3-58 in // regolamento degli scambi nellantichità {III-I millennio a.C), ed. L. Milano and N. Parise. Bari: Laterza. Pasquali, I. 2005 // lessico dellartigianato nei testi di Ebla. Quaderni di Semitistica 23. Firenze: Dipartimento di Lingüistica, Università di Firenze. Pettinato, G. 1988 Nascita matrimonio, malattia e morte a Ebla. Pp. 299-316 in Wirtschafi und Gesellschaft von Ebla, ed. H. Hauptmann and H. Waetzoldt. HSAO 2. Heidelberg: Orientalverlag. Pinnock, F. 1991 Considerazioni sul sistema commerciale di Ebla protosiriana (ca. 2350-2300 a.C). Parola del Passato 46: 270-84. Pomponio, F. 1982 Considerazioni sui rapporti tra Mari ed Ebla. VO 5: 191-203. 1984 I lugal dell'amministrazione di Ebla. AuOr 2: 127-35. 1988 Gli ugula nell'amministrazione di Ebla. Pp. 317-23 in Wirtschaft und Gesellschaft von Ebla, ed. H. Waetzoldt and H. Hauptmann. HSAO 2. Heidelberg: Orientalverlag.

24

MASSIMO MAIOCCHI

1989 1998a 1998b 2003

Epidémie et revenants a Ebla. t/F 21: 297-305. The Exchange Ratio between Silver and Gold in the Administrative Texts of Ebla. ASJ 20: 127-33. The Transfer of Decorative Objects and the Reading of the Sign DU in the Ebla Documentation. JNES 57: 29-39. La Terminologia Amministrativa di Ebla: su-dug e TUS.LÛxTIL. Pp. 540-59 in Semitic and Assyriological Studies Presented to Pelio Eronzaroli by Pupils and Collegues, ed. P. Marassini et al. Wiesbaden: Harrassowitz. Pomponio, E., and Xella, P. 1984 Ricerche di lessico eblaita I. AfO 31:25-31. Renger, J. 2001 When Tablets Talk Business: Reflections on Mesopotamian Economic History and Its Contribution to a General History of Mesopotamia. Pp. 409-15 in Historiography in the Cuneiform World. Proceedings of the XLV Rencontre Assyriologique Internationale, ed. T. Abusch, P.-A. Beaulieu, J. Huehnergard, P. Machinist and P. Steinkeller. Besthesda: CDL. Steinkeller, P, and Postgate, J. N. 1992 Third Millennium Legal and Administrative Texts in the Iraq Museum, Baghdad. MC 4. Winona Lake, IN: Eisenbrauns. Waetzoldt, H. 1990 Zur Bewaffnung des Heeres von Ebla. OA 29: 1-39. 2001 Wirtschafts und Verwaltungstexte aus Ebla. Archiv L. 2769. MEE XII. Roma: Università degli studi di Roma "La Sapienza," Dipartimento di Studi Orientali. Zaccagnini, C. 1991 Ancora sulle coppe d'oro e dargento nel Vicino Oriente nel Tardo Bronzo. Scienze dell'antichità 5: 369-79.

TEXTS RELATING TO THE ARCHIVE OE TURAM-ILI IN THE SULAIMANIAH MUSEUM Sarood T. Mohammed Taker (Saladin University I Kurdistan-Iraq)

The archive of Turam-ili is an important source for the study of commercial activity in Ur III Mesopotamia.' A number of critical studies have been published on the subject, but at the present time the original find spot of the archive is unknown. Garfinkle (2002: 29 n. 3) suggests that the texts may have come from somewhere north of site of Tell Al-Wilayah, but others have proposed that they may have been found at Wilayah itself (Michalowski 2002: 24-26). Nothing new on this matter can be said until new information comes to light. The two tablets published here were first edited as part of my M.A. thesis. They are currently in the Sulaimaniah museum (S.M.), but they were purchased and therefore have no documented provenience.^ Both are receipts typical of tbe archives associated with Turam-ili, the Mesopotamian merchant who was active during the reigns of kings Amar-Sin, §u-Sin, and the first years of Ibbi-Sin (Garfinkle: 2002: 30). In the first of our texts, he bears the title ugula dam-gàr, "overseer of merchants," a role in which he is also attested five years earlier in another document (Van de Mieroop 1986:11, no. 37, SS 5). In both of the texts the grain received by Turam-ili comes from Ili-asu> who is otherwise undocumented in this archive. Both tablets record transactions that properly should belong to the archive of Ili-asu, the creditor; either Turamili kept duplicates of such records, and Ili-asu lived close by, given the way these archives have appeared on the market. There are only three or four other similar creditor receipts among the published Turam-ili materials.' Noteworthy is the great volume of the cereals advanced to Turam-ili in these two tablets (especially in no. 2), as well as the late date of both of them. The third year of Ibbi-Sin year tbree marks the effective end of most of the state-controlled enterprises of the Ur III state, and also marks the end of Turam-ili's documented activities. Text No. 1:S.M. 1188 4.5 X 3.9 X 1.9 mm. -/-/IS 1 1.

60x2-1-50-h4 se gur níg-sám gud-sè ki dingir-tj-s« x'-ta tu-ra-am-i-li

1. I would like to thank my M.A. supervisor Ahmed Kamil, who facilitated the work on the tablets in the museum, and also the directorate of the Sulaimaniah Museum for their kindly allowing me to use their facilities. I would also like to thank Dr. Jacob Dahl for his kind help during my M.A study period on these tablets. 2. The city of Sulaimaniah lies in the Kurdistan federative region in the north of Iraq, about 112 km from Arbil, the capital of Kurdistan federative region, 330 km from Baghdad. 3. Van De Mieroop 1986: 30 no 36; Garfinkle 2002: 32 nos. 1, 4; 61 no. 27b; 61 no. 28 (?). 25

|CS62(2OIO)

26

SAROOD T. MOHAMMED TAHER

5. ugula dam-gàr au ba-ti mu ''!-Í7!'-''en.zu lugal "Turam-ili, overseer of merchants, received 174 kor of barley from Iliasu, the x, for the purchase of oxen. The year: Ibbi-Sin (became) king (ISl)."

Notes a. The text contains an amount of barley for the making of beer, from Ili-asu, received by Turam-ili. b. The name dingir-a-su is without any doubt Akkadian; see Gelb, MAD 3,68. The name is known as far back as ED II from Kish (Vito 1993:28), frequently in Sargonic texts from Diyala region (Gelb 1952), and finally in the Ur III period, especially from Ur and Drehem (e.g., Hilgert 1998: 376; Vito 1993: 172), but is S.M. 1188 until now undocumented among the Turam-ili texts, as already noted above. Ili-asu's title is difficult to decipher and will require collation at a later time. The sign looks like dub, but what follows does not seem to resemble sar. Note, perhaps, the place name ù-sa-ardingir-a-s« as the location where a grain loan from Nippur is to be repaid (Owen 1982 no. 247: 7, possibly ISl). Text No.2: S.M 1172 7.2x5.7x2.2 mm. -/9/IS 3 1.

3600x2-h60x 10x2-1-60x6-1-30-1-1 äe gur [...] zizgur [...]giggur 5. ki dingir-a-sw-ta fw-ra-am-i-/i'su ba-ti iti kir,,-si-ak mu "î-bi-ên.zu 10. lugal uri.'''-ma-ke4 si-mu-ru-um''' mu-hul "Turam-ili received 8791 kor of barley, [x] kor of emmer, and [x] kor of wheat

S.M. 1172

TEXTS RELATING TO THE ARCHIVE OF TÛRAM-ILÎ

27

from Ili-asu. Month 9, the year: 'IbbiSin, king of Ur, destroyed Simurum' (IS 3)."

a. This is a receipt for barley, emmer, and wheat from Ili-asu, received by Turam-ili. b. The month kir|,-si-ak is one of thirteen month names mentioned in the Turam-ili archive (Widell 2003: 1), apparently number nine in the sequence.

S.M. 1172

References Garfinkle, S. 2002 Gelb, I. J. 1952 1957 Hilgert, M. 1998

Turam-ili and the Community of Merchants in the Ur III Period. JCS 54: 29-48. Sargonic Texts from the Diyala Region. MAD 1. Chicago: University of Chicago Press. Glossary of Old Akkadian. MAD 3. Chicago: University of Chicago Press.

Drehem Administrative Documents from the Reign ofSulgi. OIP115. Chicago: The Oriental Institute of the University of Chicago. Michalowski, P. 2002 Editorial Preface to the Turam-ili Tablets. JCS 54: 25-27. Owen, D. I. 1982 Neo-Sumerian Archival Texts Primarily from Nippur in the University Museum, The Oriental Institute and the Iraq Museum. Winona Lake, IN: Eisenbrauns. Van De Mieroop, M. 1986 Turam-ili: An Ur III Merchant. JCS 38: 1-80. Di Vito, R. A. 1993 Studies in Third Millennium Sumerian and Akkadian Personal Names. Studia Pohl 16. Rome: The Pontifical Institute. Widell, M. 2003

The Ur III Calendars of Turam-ili. CDLJ 2003:2.

READING SUMERIAN NAMES, I: ENSUHKESDANNA AND BABA Gonzalo Rubio (Pennsylvania State University)

A. Ensuhkesdanna or Ensubakesdanna 1. On the Meaning of MUS^The name of the ruler of Aratta with whom Enmerkar engages in various challenges is written e an-na. At least two versions of the name are attested in modern scholarship: Ensuhkesdanna and Ensuhgiranna. This name contains the element MUSj-KESj, which a priori could be read either suh-keSj (or suh-kesda) or suh-gir,,. As will be shown below, the evidence clearly supports a reading with keSj. Moreover, the term suhkesj has often been understood as a kind of headgear, although a pectoral ornament or a low-hanging necklace are perhaps more likely interpretations (Klein 1981: 91; Civil 2008: 67). The word denotes a symbol of en-ship, the same way that a crown is a symbol of kingship (Falkenstein 1959: 96-97). The connection with headgear is based on the fact that this term occurs alongside other symbols of power worn on one's head (men "tiara," aga "crown"). However, suh-keSj may have been a different piece of regalia, as there is a limit to how many things even a deity can wear on her head. In fact, this study will suggest that suh-keSj may have consisted of stones mounted on a pectoral or necklace. Still, any specific interpretation of its meaning beyond its symbolic function remains tentative.' The reading suh of MUSj {Mö^^-gunü) is attested in a couple of lexical items: The verb suh "to pull out; to choose from," often translated as nasàhu; Aa 8/1: 167 {MSL 14,493); Syllabary B Voc. 1: 20 {MSL 3, 97); and reconstructed for Ea 8: 59 {MSL 14,478); see CAD N/2,1. This verb is also written suh; (KU), as in Ea 1:152 {MSL 14,184). This verb is well attested in literary compositions: Uruk Lament 4.29, 5.21 (Green 1984:273-74); Lugalbanda 1294, 309 (Wilcke's ms.); Lugale 297 (van Dijk 1983, 2: 94); Diatribe C 12 (Sjöberg 1972: 107); Ur-Ninurta A 32 (Falkenstein 1950: 108; Sjöberg 1977: 190); The Tree and the Reed 59 (Civil's ms.). The compound verb igi — suh, meaning both "to be angry" {nekelmû, CAD N/2,152) and "to choose" {nasäqu, CAD N/2, 21); see Civil 1999-2000: 185-86; Karahashi 2000: 127-28. Probably the general

I must thank Miguel Civil and an anonymous reviewer for reading an earlier version of this article and making useful suggestions. Needless to say, responsibility for all opinions expressed here rests solely with the author. 1. Of all the passages in which suh-keáj is attested (see a sample in Steinkeller 1998: 93-95), the only one that may actually point to a meaning related to headgear is Lamentation over Sumer and Ur458:sig2-pa n a m - e n - n a suh keäj-da-gaj (Michatowski 1989:183),since sigj-pa refers to hair (Klein 1981: 149; Attinger 1984: 46). To the selected occurrences of suh-keSj listed by Steinkeller, one should add SuSin hymn J 7 (BM 100042; Kramer 1989: 304), Ihbi-Sîn hymn E 8 (suh,„-ke§2, CBS 15158; Sjöberg 1970-1971: 149), and Nanna hymn E 48 (suh,o-ke§2. UET6/1 67; Charpin 1986: 367). 29

JCS 62 (2010)

30

GONZALO RUBIO

meaning of this compound verb is "to make a gesture with one's eyes or eyebrows," which could be understood both as "to frown (wrinkling the eyebrows)" and "to glare or stare (raising the eyebrows in approval)." There exists another word suh, which is also written suh,„ (MUSj), and which is often understood as "crown" and, by extension, "the top of something," such as a building (Attinger 1993: 513; Steinkeller 1998: 93-95; Cavigneaux and al-Rawi 2000: 53 n. 160). This is the suh that occurs in the sequence MU§2-KE§2- It is, nonetheless, not always easy to decide between the readings s u h / s u h ^ and mus2/mus3. The latter has two basic meanings in lexical and bilingual texts: "appearance, look, glow" (zïmu; CAD Z, 119), and "plain, flat area" (mätu; CAD Mil, 414).^ Moreover, one should not confuse the aforementioned architectural term (suh/suhj^) with the use of MUS2 instead of sur, (MUS^ = HIxAS), which alternates with sur^ (KI.GAL) and sur7(KI.KAK). The noun sur3/surj/ sur^ refers to the foundation ofa building, canal, or pit, and normally corresponds to Akkadian süru "ditch" (CAD S, 415), berütu "foundation pit" {CAD B, 213), and even ussu "foundation" {AHw 1442; Borger 2003: 284). The reading sur3 is attested already in the Ebla Sign List (66), where H I X M A S (or rather HIxTIL) corresponds to suru,,-um (Archi 1987: 96; Steinkeller 1992: 16-17 n. 11).='

2. ReadingMUS2.KES2 as suh-kes^ In the Late Babylonian bilingual composition the Exaltation of Ishtar III 83 (Hruska 1969: 485), one finds the following syllabic gloss: '''•'^•'äMUS2-KES2 mah-a... {TCL 6 51 rev. 33). ZgoU (1997:305-6) argues that this su-kiis could be glossing either MUS2.KE§2 as a whole or MU§2 alone. The latter possibility is based on the gloss suku-us for MUS3 (not MVè^) in Nabnitu 9: 104 {MSL 16, 120): MU§3'"''""^KA-KA = si-ip-ra-a-tum {sipru "stone bead"; CAD S, 204-5; Civil 2008:68). Moreover, the reading su-ku (sukuj) is assigned to MUS^ (MU^^-gunû) in Proto-Ea 755 {MSL 14, 60), in alternation with su-uh2; the latter occurs as the reading of MUS2 in the previous line, Proto-Ea 754. In the Exaltation of Ishtar, glosses pertaining to a sequence of signs are as frequent as glosses providing the reading ofa single sign (e.g.. Ill 69, 81, 85, 87; IV 13). Thus, it seems quite likely that, in the Exaltation of Ishtar, su-ki-is is intended to provide the reading of MUS2-KE§2 as a whole. Moreover, the Akkadian translation of the aforementioned entry in Nabnitu, sipratu, plural oí sipru "stone bead, precious stone trim," suggests a connection with the general meaning of MU§2-KEá2 proposed by BQein and Civil. If one assumes that in that Nabnitu entry MU§3 stands for MU$2. then MUS2-KE§2 could be interpreted literally as "stone beads (MUS3/MU§2 = sipratu) bound together (kesda = raksätu)!"' In fact, one should bear in mind that earlier occurrences of MUSj, particularly in the third millennium, often correspond to MU§2 in later texts (Civil 2008: 67). In sum, the gloss su-ki-is in the Exaltation of Ishtar provides the reading of MU§2-KES2 as a whole sequence, which supports the reading suh-keS2 (or suh-kesda) of MU§2.KE§2 in the name en-

2. On muS2/muá3, see Sjöberg 1969: 55-56; 1973: 39; Heimpel 1972:286-87; Klein 1981: 90-91; Attinger 1993: 513 n. 1416. The meaning tûSaru (or tuSaru) of MUS2 given in Erimhus frag, b ii 3' {MSL 17,89) has to do with a kind of reed fence, since it appears in sequence with sippu ("doorjambs, doorframe"; CAD S, 300-303) and zamú (a type of wall; CAD Z, 41), and this is likely the same tuSaru that occurs in the sequence qan tûiari, equated with gi nam-erim2Ín two lexical lists {CAD T, 495). Elsewhere, gi nam-erim2 (lit. "the oath reed") is translated as kiikiiiu, probably a room separated by reed screens, in which the emblems used for oath-taking were stored (Reiter 1989, l99Upice CAD K, 359). It has been assumed that this tusaru = MU§2 is the same tusaru that means "flat land, desert" {CAD T, 495-96); see Sjöberg 1969: 56; Heimpel 1972: 286. However, the context in that fragment of Erimhus and the connection with kilkiium point to a different word. 3. The reading sur3 of HIXAS (MUS4) is further established in Proto-Ea 369 {MSL 14,46), Aa 5/2: 92-99 {MSL 14,417), Ea 5:93 {MSL 14, 399), and Syllabary B Voc. 1: 26 {MSL 3, 98). The readings sur^ of KI.GAL and sur^ of KI.KAK stem from Diri 4: 311-312 {MSL 15, 162-63). See also iirimtu "cutting" {CAD S/3, 105) and Sihittu {CAD S/2,415). 4. Since sipru comes from fepéru, which can mean "to strand (hair, linen)" {CAD Ç, 132), Cohen (1975-1976: 32-33) argued that the ES2 was a sort of braided or wrapped headgear. 5. Civil (2008: 68) argues that the gloss su-ki-i§ in TCL 5 51 comes from *sub-keá2, with simplification of the cluster /bk/. However,

READING SUMERIAN NAMES, I

31

The alternation between Ikl and /h/ is well attested in the readings of MUS2 {M\]Í^-gunu) in lexical lists, but not so in the case of MU$3, as can be seen in Proto-Ea 752-756 {MSL 14,60): mu-us MU§3 mu-us-ku su-ku se-e

MU$2

The reading /suku/ also occurs in Erimhus 5: 245a {MSL 17, 77): "'•''''MUS2 = ma-sà-hu "to flare up, to shine brightly" {CAD M/1,354). Moreover, the catchline at the end of Erimhus 6 is ^"-"''MUSj = ma-sa-hu {MSL 17, 88). However, the reading /suh/ is the most commonly attested, as in the lexical references for suh = nasàhu given above, as well as the list of sign names in Ea 8 Assur MAss excerpt 27 {MSL 14,485). Note also the gloss in Nabnitu O 156 {MSL 16, 291): dug3''"-™"'>MUS2 (= messétu "distance"; CAD Mil, 29).

3. The Signs MUS, and MUS^ (MUS,-gunû) In spite of the aforementioned passage in Proto-Ea (752-756), the sign MUSj occurs instead of MU§3 in a number of instances in the second and first millennium, particularly in the spelling of/suba/ with ZA.MU§3. This can be seen also in the variants of a passage from a canonical lamentation attested in manuscripts from the Old Babylonian (CT 44: 12 [BM 78175] = A) and the Seleucid periods {SBH 57 = C), as well as Sultantepe {STT 156 = B):* 82 Aobv. 13 B obv. 9 C obv. 24 83 Aobv. 14 B obv. 11 C obv. 26

Uj-nu su-bi-a mi du-ga-zu [ ZA] .MU§2 mi^ '"dug,"' -ga- '"a'^ -zu iá X X x-i ú-sa-'âlf-h[u]-ka TE. â^ ZA.MU^^ : su-ku-ut-ta-ki : mi^ dug^-ga-zu su-bat-ka sa el-li-is ku-un-na-a-at '^su-ba-a^ za-gisa-sa-ra-zue2-ne-ba-ta-re ZA.[MUS2 z]a-gin3-n[a] '"em3'^-sar2-[sar2]-ra-zu ina su-bi-i el-[U] ú--qi[n\-k[a] ZA.MU$2 za-gin3-na em3-sar2-sar2-ra-[zu] ina su-be-i u uq-ni-i sa'{RA)-ak-ka-at

Along with this and other examples mentioned throughout this article, one finds additional instances of the exchangeable nature of the signs MUS^ and MUS3 in several second- and first-millennium sources (see also Attinger 1993: 513 n. 1416,619-20): • Most manuscripts of The Hoe and the Plow 113 have suh written with MU§2, but at least one has MUS3. DI (YBC 8959; Civil's ms.); compare, among the published manuscripts: 5£ 31; 50 obv. 14'(Kramer 1940: 252): igi-zu dug gestin-na-gin^ a im-ta-suh(MU§2)-en 3N-T905,10 (SIFN49): [ i]m-ta-suh(MU§,)-e-en

su-ki-iä may also stem from the weakening of/-h/in syllable coda: suh-kes2 > s u - k i - i s . 6. See Civil 1967: 206-7; Cohen 1988: 54; Löhnert 2009: 220-21.

32

GONZALO RUBIO

The erroneous spelling MUS3.DI may stem from an analogy predicated on the readings /sed/ and / sed/ that this sign shares with MUS2,3. MU^j.DI can be read sed4/sedio, in alternation with A.MUS3.DI (sedii), as, for instance, in OB Diri Nippur 343 {MSL 15, 24-25); see Klein 1981: 208. Likewise, MU^^ can be read sed^/sed^, and MUS3 sed^/sedj^ (Borger 2003: 77). • Two Nippur manuscripts of the Kes Temple Hymn have muSj on two lines (14, 26) on which all the others have mus3 (Delnero 2006: 2187, 2191). . ZA.MUS3 = subaôccurs in an Old Babylonian lament: NBC 1315 (RA 16: 208) rev. 16 (cf. Cohen 1988: 291). However, ZA.MUSj = suba is attested in first-millennium manuscripts of an ersemma: SBH 53 rev. 19 // Sm. 954 (Delitzsch AI^ 135) obv. 19 (cf Cohen 1981:131; Oberhuber 1990:456). . In AN ='>Anwm III 249, one manuscript (CT25: 21 [Rm. 11,32] rev. 17) has'^ZA.MUS^-nun-na, whereas the other manuscripts have ''ZA.MUS3-nun-na. This same deity is also spelled ''ZA.MUS3-nun-na in a Neo-Assyrian god list (KAV 48 ii' 3'); see Litke 1998:143. Although MU§3 is usually the early spelling, which gets replaced with MUS^ later on, there are also examples of MUS3 occurring instead of MUSj in the Old Babylonian period, coincidentally in the very writing of suh-keSj as suh,„(MU§3)-kes2: Ibbi-Sin hymn E 8 (CBS 15158; Sjöberg 1970-1971: 149); Nanna hymn E 48 (UET6/1 67; Charpin 1986: 367); Gilgames and Huwawa B 29 (Edzard 1993: 19). The occurrences of MUS^ instead of MUS3 and vice versa in a variety of second- and first-millennium contexts, lexical and otherwise, should not be disregarded as a matter of confusion or simple errors. These are not isolated instances produced by a single scribe or typical of a specific period or school. On the contrary, they correspond to a sufficiently widespread phenomenon that began during the Old Babylonian period.

4. Other Readings From the point of view of the second- and first-millennium lexical traditions, the alternation seen above (section 2) between the readings suh and /suk(u)/ of MUSj would seem to correspond to two different verbs: MUS2 = suh = nasähu "to tear out" (CAD N/2,1); • MUS2 = suku; or suk = masähu "to flare up, to shine brightly" (CAD M/1, 354). However, the interpretation masähu is probably secondary and predicated on the phonetic similarity between that verb and nasähu. Nevertheless, besides the alternation between /h/ and /k/, MUSj is assigned the reading susbuj, s u s b u , o r s u z b u / s u s b u i n t h e name of a cultic functionary that corresponds to Akkadian ramku, as in Aa 8/1: 168 (MSL 14, 493), as well as susbû, as in Erimhus 5: 10 (MSL 17, 67). On this cultic functionary, see CAD R, 126-27; CAD S, 416; Focke 1998: 218-19. A conventional distinction between susbu (MUSj.BU), susbuj (MUSj). susbu3 (MUS3.BU), and susbu4 (MUS3) is kept here (Borger 2003: 77). The spellings with BU are more frequent, but they are probably secondary. Compare the following lexical entries: "-"*-'"' in LU2 4: 83 (MSL 12, 131) 2 ^ ' ' " = [ra-a]m-ku in Nabnitu 23-i-Q: 288 (MSL 16, 220) MU§/"-"=-'''BU = su-us-bu-ú in Erimhus 5: 10 (MSL 17, 67) Whereas in the first two BU is part of the gloss (/susbu/, /suzbu/), in the last one the gloss is /susbi/, to which a spurious BU has been added due to reanalysis of the sequence. This led to the extrapolation of a reading /sus/ for MUS2 and the interpretation of BU as a phonetic complement of the kind found in MUS3,2.DI, which can be read sed/sedip (see above, section 3). This reinterpretation of MUSj as /sus/ found analogical basis in the readings suh and suk.


33

The assignation of a value /sus/ to MUS2 can be linked to its reading susinak (su-si-nak) in Aa 8/1: 170 {MSL 14,493). For the extension of the /sus/ or /sus/ reading to MU$3 early on, see these parallel passages in the god list from Abu Salâbïh: OIP 99 82 obv. iv 5 OIP 99 86 iii' 4'

''n i n -MUS3.EREN(LAK-671 )

Both instances can be read ''n in-su s um, as MU§3.EREN = susum; cf. OB Diri Oxford 462 (MSL 15, 47) and Diri 4: 94 (MSL 15, 152-53). Nonetheless, in the Abu Salàbih god list MU§3.EREN/Ê§4 simply stands for / susinak/. In fact, MU§3.EREN can be read/susin/or/susina/on the basis of syllabic spellings, such as su-s in2 and su-sin2-na (Sollberger 1956: 24; Steinkeller 1984: 140 n. 18). Thus, ''nin-MU§3.EREN is an early spelling of what later on would be written syllabically as in-su-us-na-ak and in-su-si-na-ak (Hinz 1976-1980: 117).' On the other hand, the spelling ''nin-MUS3.SES4(LAK-668) from Abu Saläbih is somehow continued in the Old Babylonian period (RGTC 3, 230) and it resurfaces in some first-millennium god lists.* Note also the spelling of the Sumerogram in the Annals of Assurbanipal, Prism A vi 30 (Borger 1996: 53): '

5. Ensubakesdanna The readings of MU§2/3 with Ihl have Early Dynastic precedents. In the ED Practical Vocabulary A (57-58)— also known as Archaic HAR-ra A, attested in Ebla (MEE 3 nos. 45, 46, 61, pp. 143-55, 247-50) and Abu Saläbih (OIP 99: 30-31, 33, 34, 37, 225-26)—one finds the following entries (Civil 1987: 143; 2008: 28-29, 67-68): ARES 4: EA (// EC) obv. iii 15 KE§2-za ARES 4: EA (// EC) obv iii 16 a

ARES 4: EB rev. ii 2 = sè-ba-dum ARES 4: EB rev ii 3 = su-bù-lu

Eblaitic sè-ba-dum may correspond to the plural siprätu ("stone beads"), which was discussed above (section 2) in connection with Nabnitu 9:104, with a not unparalleled inconsistency in the use of iè to write an emphatic (Civil 2008: 14, 68). Moreover, su-bù-lu is perhaps related to subulu in Mari Akkadian, a term that occurs in a context of jewelry (ARMT 23: 535 rev. i 14; Civil 2008: 68). In both entries, ZA is merely a frozen classifier for names of stones. Thus, Civil (2008: 67) has proposed that, when designating a stone, both MU§3 and ZA.MU§2/MU§3 should be read /suba/ or /subu/, or /suba/ or /subu/.'" However, as Civil notes, in the Old Babylonian period MU§2-KE§2 seems to be have been understood

7. See Sargonic and Ur III texts from Susa; e.g., ''nin-MU53.EREN (MDP 11, pi. 1 obv. i 8), ''MUS3.EREN (MDP 6, 7; pi. 2, I ii 5). Both spellings, with and without n in, occur in Elamite texts (Hinz and Koch 1987:957,1002). The name Puzur-Inauáinak is written Puzur4-''MUS3. EREN and Puzur4-''nin-MUS3.EREN in Sargonic inscriptions from Susa (EAOS 8, 56). 8. See II R 60 obv. i 9 (TuL no. 2, p. 12): ''MU§2.§ES2(SIG2.LAM) sa iu-iá-an'^; CT29 (BM 46550): 46 iii 4: äu-si-nak = ''MUS3.§E§; AN = 'Anum V 286 (Litke 1998: 194): ''MUS3.SES2 = dumu ''tispak(MU§2)-ke4. 9. MUS3.EREN is the common logographic spelling of the toponym Susa in the Early Dynastic period (RGTC 1, 154-55), in Sargonic texts (FAOS 8, 98-99), and in Ur III (RCTC 2, 187-91). It occurs in later periods as well, even as a Sumerogram in Akkadian, from a few Old Babylonian texts (RGTC 3, 230) to the Cyrus Cylinder (30). In addition, it is the logographic spelling in Elamite texts (Hinz and Koch 1987: 957), and MUS2/MUS3.EREN*'' is attested in texts from Susa from various periods (RGTC 11, 270). Note the presence of both ''MUS3.EREN and MUS3.EREN''' in an inscription from Sargonic Susa (MDP 6 7). 10. Note the following conventions regarding indexes: Suba = ZA.MUS2, suba2 = ZA.MUS3, áuba3 = MUS3.ZA or MUS3XZA, and b MUS3 (Borger 2003: 77, 217; Mittermayer 2006: 19, 189). For the possibility of reading/suba/as/suba/or/subi/, see Sjöberg 1988: 172-73 n. 6. On the spellings and readings of/suba/, see Wilcke 1969: 200-201; Sjöberg 1969:113; Steinkeller 1989:202; Rubio 2000:210 n. 28.

34

GONZALO RUBIO

as a sequence noun H- verb; e.g., Ninurta's Journey to Eridu B 12 (STVC 34 obv. ii 22): nam-en-se3 suh zagin3 mu-un-kes2 "for his lordship, he bound/tied the shining necklace/pectoral" (Falkenstein 1959: 82, 96-97; Reisman 1971: 4)." On the contrary, the Ebla attestations would point to a construction noun -^ noun. In fact, MUS2.KES2 is regarded as a noun and equated with tiqnu "adornment, ornament" in lexical lists, and in at least one instance it is clear that such an ornament was made of stones, in HAR-gud to HAR-ra 16:100 (MSL 10, 32): NA4.MUS2.KES2 = tiqnu (Steinkeller 1998: 93; Civil 2008: 67). In Ebla the sequence was taken to be a single word, which lends itself to be analyzed as noun + noun from our lexical perspective. Later on, the sequence was perhaps reanalyzed as a compound, which led to the reinterpretation of the form as noun -h verb. There is evidence to support a /suba/ or /suba/ reading of MUS2 and MUS2.KE§2 in the Old Babylonian period as well: A gloss in a manuscript of Sulgi hymn E 10 (YBC 7152 obv. 10) reads ™-'"'MU§2 za-gin2 kes2-da-meen (Civil 2008: 67-68). Note that Klein (1981: 90) proposed to emend the gloss and read su-uh'. On his part. Civil proposes to emend Antagal A 208, MUS2^''"\SAR = tiqnu (MSL 17, 188; Civil 2008: 68 n. 137), and to read the gloss there as su-ub' instead. In Antagal, MUS2''''"'".SARstands for MU§2''""'"-KE§2, as SAR can occur instead of KES2 in Neo-Assyrian texts (Borger 2003: 308, 359). KES2 and EZEN had merged by the end of the Old Babylonian period (cf Borger 2003: 98). • The whole sequence MUS2.KES2 is glossed /s u b a / in an Old Babylonian ersemma of Dumuzi: u 3 k i - s i kil tur-re »'-'^MU§2.KES2 b a - n i - i n - a k * (CT58: 42 rev. 79; collated by I. Finkel).'' Nevertheless, in most instances, su-bu and su-bu would not seem to gloss MU§2/MU§3 alone, but rather ZA.MUS3/ZA.MUS2; e.g., Diri 4: 47 (MSL 15, 150-51): [s]u-bu : U2.ZA.MUS2.KI = uruk (= U2-suba''') Hh 21 sec. 10: 8 (MSL 11, 19): RIGy"'s-^''-''"ZA.MU§3.KI In fact, ZA.MUS3 should be read suba; cf. Diri 3: 99-110 (MSL 15, 140-41); Klein 1981: 152. As mentioned above, this ZA is a frozen occurrence of the classifier for names of stones, which in the second- and first-millennium lexical traditions was reinterpreted as part of the compound sign. Thus, the reading /suba/ designates a kind of stone. Additional support in favor of a reading suba, for MUS2 could be found in the use of kur suba,2,3,4) as a common epithet of Aratta, as in Enmerkar and the Lord ofAratta (34, 230, 432); see Rubio 2000: 210 n. 28; Mittermayer 2009: 224-25, 234-35. Note, however, elsewhere this may refer to other mythical locations (Sjöberg 1969:113; 1975: 314-15). It is, therefore, quite logical to find the element /s u b a / in the name of the ruler of Aratta: en-suba^^-kesda-an-na,perhaps to be understood as an appositional name, "The ruler (is) the heavenly suba -stone pectoral." Needless to say, this name is no actual person's name, but rather a literary artifice purposely devised to convey the nature of this character within the Aratta compositional cycle. In conclusion, two possible readings of this name are supported by the evidence: e n - s u h - k e s d a - a n - n a and en-suba,-kesda-an-na (oren-suba,-kesda-an-na). Nonetheless, one can find no particular reason to support a reading "en-suh-gir,,-an-na.""

Concerning the suba stone as a real object in general (Akkadian Subü and sabû), particularly used in medical contexts, see CAD S/3, 185-87; Scurlock 2006: 267-70 et passim; Schuster-Brandis 2008: 271-73, 288. The suba stone should be distinguished from the säbu stone, the latter occurring in magical and medical contexts, including childbirth, with both stones often attested in close proximity (CAD S, 5; Stol 2000: 50; Scurlock 2006: 269-70, 481-86). 11. Similar constructions of suh za-gin3(-na) with the verb kes2 occur in Sulgi hymn E 10 (Klein 1981: 90), Enki and the World Order 349 (Benito 1969: 105), Gilgames and Huwawa B 29 (suh,(,, Edzard 1993: 19), and Inanna-Dumuzi B (SRT 31 32; Sefati 1998: 129). 12. The gloss was originally regarded as an illegible sign by Kramer (1980: 8). Alster and Geller copied and read su-ku (CT 58 p. 19); see also Steinkeller 1998: 93 n. 42. However, I. Einkel's collation (courtesy of M. Civil) reads the gloss as su-ba. 13. The reading gir,, for KES2 corresponds to a verb meaning "to tie up, to yoke" (samâdu, Cv4D Ç, 89-90); see Reciprocal Ea A 101 (MSL


35

B. Baba 1. Baba in Eära

The name of Ningirsus spouse is alternatively read ''ba-U2, ""ba-bag, or ''ba-U2. The main basis for the apparently unique reading b a ^ of U2 in this theonym lies in phonetic spellings with the sign b a (b a - b a ). However, in a stimulating and learned contribution, Marchesi (2002) provided a reassessment of the evidence for such phonetic spellings and put forward various arguments in favor of the reading ''ba-U2, as this would reflect the pronunciation of the name in the first half of the third millennium. One of the earliest phonetic spellings occurs in the Early Dynastic god list from Fära {SF1 rev i 13 '). Marchesi (2002: 162) follows here Mander (1986: 84, 99) in reading ''baK^. This ''ba-ku^ would be "the turtle (god)," an interpretation supported by ED Fish list 75 (M££ 3,101): ba - ku^, which would stem from ba - al - gu^ (BA.AL.KU2) and this from ba-al-gi^ {PSD B, 14-15; Englund 1990: 223; Färber 1974: 195-96). However, Manders reading should be corrected, as the third sign in that line of the Fära god list is not HA ( k u j but BA (Krebernik 1986:179). The sign BA can be seen in the CDLI photo, and as such is clearly reproduced in Kreberniks copy.'" Moreover, Joachim Marzahn kindly collated the line for us. Mander was probably misled by a couple of scratches: a small vertical scratch at the beginning of BA, which may make it look like HA, and another scratch that may seem to be an additional wedge at the top of BA. As Marzahn points out (personal communication), "the edge of the tablet is partly broken off but glued together and just in this position one may see some more wedges than originally written" {SE 1 is the result of joining VAT 12760-^12762H-12776). Note that a few lines after ''ba-ba, the Fára god list has ''kuS7-ba-U2 (SF 1 rev i 21'), a deity closely associated with ''ba-U2 (see below in regard to a-ba-ba).

2. Is Baba Short for Ababa?

As in the god list from Fára, third-millennium personal names exhibit an alternation between ur-^'ba-ba and ur-''ba-U2, and the corresponding hypocoristic name ba-ba, which supports the reading ba^ of U2.'' Marchesi (2002: 163) argues that in these names ba-ba does not stand for ba-U2 but rather for ''a-ba-ba or ''a-ba^-ba^, with apheresis of the initial vowel. He refers to parallels such as Inûrta for Ninurta and Ingurisa/Imgursa for Ningirsu (Parpóla apud Radner 1998: xxv). However, such cases of apheresis are not properly a Sumerian phonetic development, but rather the result of dialectal realizations of original Sumerian words that had long been incorporated into the native Semitic lexicon of proper names. It is most certainly in Neo-Assyrian, and not in Sumerian, that those instances of apheresis took place. In Neo-Assyrian, there are many examples of apheresis, particularly of aphesis, that is, the apheresis of an unstressed, generally short vowel: né-e-nu for antnu; mar for ammar; su-ungal-li for usumgallu, etc. (Hämeen-Anttila 2000: 37-38; Luukko 2004: 121-22). In general, apheresis does not commonly occur in Sumerian (Jestin 1965:131-33). The few instances that may come to mind include the initial element nin and would involve loans from Sumerian into other languages, and perhaps from other languages into Sumerian as well.'* A good example is provided by Insusinak, which comes

14, 526); Proto-Ea 762 {MSL 14, 60); Ea 8: 85 {MSL 14, 478); Aa 8/2: 35-36 {MSL 14, 498); Geller 1985: 114-16; 1989: 201. In this word, the sign KES2 occurs along with glosses indicating a reading hir in texts from Meturan (Cavigneaux and al-Rawi 1995: 34-35). Thus, in the OB forerunners of Udug-hui 479 (Geller 1985: 46), one should read '"bai-an-hir'''-hir'''-re-es (Cavigneaux and al-Rawi 1995: 34). There is no reason to assume that this verb occurs either in the noun or in the name discussed above. 14. For the CDLI photo, see http://www.cdli.ucla.edu/dl/photo/P010566.jpg. Krebernik's copy can be found here: http://www.cdli.ucla. edu/dl/lineart/PO 10566_l.jpg 15. See Limet 1968:105,108,383, 536; Di Vito 1993: 30; Sauren 1969:28; Rubio 2001: 272 n. 26. Note that,pace Rubio, the sequence e2 baba-ta {YOS 4: 203 obv. 5) probably does not mean "from the temple of Baba," but rather "from the house of Baba (PN)" (Marchesi 2002: 163). 16. Words such as gis-eren2 or gis-rin^ {gisrinnu "balance, scale"; CAD G, 107; Steinkeller 1989: 95 n. 276) have nothing to do with apheresis, since there gis was actually read. The fall of posttonic vowel is a common phenomenon completely unrelated to apheresis.

36

GONZALO RUBIO

from ''nin-susinak (Hinz 1976-1980; Selz 1991: 31; Malbran-Labat 1995: 190). The case of the name of Inanna, which is commonly assumed to come from /n i n an - ak/, may pose a complication, since this may be merely a folk etymology (Gelb 1960; Jestin 1965: 133; cf. Selz 2002: 662). The idea that b a - b a may be the apheresis version of a - b a - b a presents also a frequency problem. For instance, a search of the CDLI database yields 3,124 texts containing the name ur-''ba-U2 of which 135 are Early Dynastic, seven Sargonic, 475 from Lagash II, and almost 2,500 from Ur III. Without the divine determinative, ur-ba-U2 occurs only 8 times, always in Ur III. The name ur-ba-ba occurs 34 times: 7 in the Early Dynastic, once in the Sargonic and Lagash II periods respectively, and 24 times in Ur III.'' The name with the determinative, ur-''baba, is attested twice in Ur III {TMH n¥ 1/2: 11 and an unpublished Schoyen tablet). These figures correspond to one's expectations: the phonetic spelling ba-ba is substantially less frequent than the spelling ba-U2. On the other hand, the name ur-''a-ba-ba turns up only 19 times in the CDLI corpus, with 16 attestations in Ur III and three in Sargonic texts; and ur-a-ba-ba seems unattested.'^ If a-ba-ba were the original form, one would expect it to be more frequent than the form that would have allegedly undergone apheresis, ba-ba. However, personal names with ba-ba are twice more frequent than names with a-ba-ba. In fact, besides a handful of anthroponyms, âba-ba is very rare in the third millennium and afterwards." In terms of context, in god lists and offering lists, both in Ur III and later on, Ababa normally appears as one of the deities associated with Ninimma (Such-Gutierrez 2003,1: 311; Richter 2004: 93). However, the ''ba-ba that occurs in the Fàra god list is not close to any mention of Ninimma, whose name does appear elsewhere in that list {SEI obv. V 24'). Moreover, in the Fàra god list,''ba-ba {SEI rev. i 13') occursafew lines away from''kus7-ba-U2 {SE 1 rev. i 21). In the god list from Abu Saläbih, ''ba'-U2 occurs immediately before ''kus7-ba-U2 {OIP 99: 82 obv. iv 21-22). The association between '^ba-U2 and ''kus7-ba-U2 is well documented (Selz 1995:157). In sum, there is no reason to suppose that ba-ba is short for a - b a - b a .^°

3. The Problem of Reading ba¿

The third-millennium phonetic spellings of the name of Ningirsu's wife are continued in second and first millennium sources. For instance, a Neo-Assyrian god list from Assur {KAV46 i 14') has [ba] -a-bu : '"'ba"'-U2, which points to a reading bu,, of U2, corresponding perhaps to a Semiticized pronunciation. Concerning the reading of U2 as ba^, the famous god list AN: ''Anum exhibits two variant spellings for the same line (II 268; Litke 1998: 99):'' ms. A (YBC 2401 iv 27 [MAss]) ms. B {CT 24: 28 [K. 4349] iii 72 [MAss]) ms. C (CT24: 16 [K. 4332] iii 17 [NAss])

'*ab-ba ''ab -ba ''ab-U2

All three manuscripts refer to the doorkeeper of the Esagila, the spelling of whose name exhibits an alternation between ''ab-ba and ''ab-U2 (Richter 2004: 117-19, 197). In the three manuscripts, this theonym is followed by

17. Instances of ur-ba-ba in ED: NTSS 165; WF 17, 18, 26, 27, 29, 125. In the Sargonic period: OSP 1 131. In Lagash II: ¡TT 47716. Of the Ur III attestations, see, e.g., NATN 423, 518, 573; TMH nF 1/2: 48, 312, 327; UET3 10; SANTAG 7 26. 18. Spellings of this anthroponym with a-ba4-ba4 seem unattested. In fact, the spelling a-ba^-ba^ seems limited to god lists and lexical texts before the first millennium (Such-Gutierrez 2003,1: 311; Richter 2004: 93). Eor the reading of this theonym, see OB Diri Nippur section U: 4 {MSL 15, 36): ''A.GA2.GA2 = a-ba-ba. 19. ''a-ba-ba occurs in two offering lists from Ur III: MVN 10 144, and TCL 5 6053. See Such-Gutierrez 2003, 1: 311; 2003, 2: 353. 20. The occurrence of ba-ba(?) in the OB Nippur god list, sandwiched between ba-ab-ba and a-ba (Peterson 2009: 47), is of little consequence here. The scribe of SLT122 did not carefully distinguish between BA and KU, so this could be ku-ku instead (Peterson 2009:76). 21. The association between ab-Uj and lugal U2 "lord of plants" is explicit in Enki and Ninhursag 270 (Attinger 1984: 30; Peterson 2009: 49). However, this seems to be a learned folk etymology (Attinger 1984: 46), the sort of phenomenon that Selz (2002) has called Babilismus.


37

''t a -KU-n a. Nevertheless, Marchesi (2002:168-69) believes that these variants respond to two different traditions. In the Middle Assyrian tradition, the doorkeeper of Marduk's temple in Babylon would have been Abba, but this would have been replaced with Abu in the first millennium. Such a substitution would be reflected also in Tintir II 29': (manzäz) ''ab-Uj sa sip-pi [MIN] "the station of Ab-Uj at the doorsill of (ditto)" (George 1992: 54-55, 292-93). The assumption is that the first-millennium scribes would have suddenly resurrected a very obscure third-millennium deity, âh-U^.^^ However, as is the case of compositions with a long textual history, such as Angim and Lugale, this is another instance of first-millennium scribal traditions being closer to earlier ones—here in the use of Uj as ba^-and Middle Assyrian scribes departing substantially from such traditions (Rubio 2009:44)." The evidence for the reading ba^ of Uj is not limited to this god list. In an entry of the ED Geographical List (189), the sign bu3 (KAxGANj-fenú; Rubio 2006: 113) in Ebla corresponds to U^ in Abu Salâbih (MEE 3, 236): MÊÊ 3, 56 obv. ix 5 OIP 99,91 obv. vii' 6 O/P99,103ii'l'

bu3-bu3''' b a 5- b a •'" b[as-b]a'''

Moreover, the word for "ax" and "small hoe," normally spelled h a - b u 3 - d a o r h a - b u - d a , appears as h a - Uj-d a in at least two Early Dynastic texts from Lagash ( VAS 27 74 obv. ii 7) and Umma (BIN 8 87 obv. 1 ), and in two texts from Umma probably dating to the Sargonic period (ß/N8 89,1; 8 99 obv. 2).^" Marchesi (2002: 170 n. 99) advocates against the reading ha-bui,-da of ha-U^-da, and prefers to read ha-Uj-da representing /hawuda/ "from a Semitic etymon *hawüd/ta!' However, such a Semitic root meaning "hoe" or "ax" does not really exist. Moreover, since Akkadian hapütu "small hoe" and ha-U^-da are clearly related, one wonders why there is no spelling ha-wutum in Akkadian texts (CAD H, 86; AHw 322,1559). In this regard, Marchesi argues that some Semitic loans in Sumerian underwent a shift /w/ > /b/ in intervocalic position. However, cases such as hawû, which occur as ha- um and ha-bu-um in Sumerian texts, reflect internal Akkadian phenomena. These are not loanwords (Lehnwörter) properly speaking, but rather Fremdwörter, whose every occurrence in Sumerian reflects the various individual realizations of the word in Akkadian: hawû, habû, ha'û (CAD H, 162-63). Regardless, the term /habuda/ may well have a Semitic origin." Nonetheless, many Mesopotamian terms for agricultural implements and tools are ultimately Wanderwörter and Kulturwörter (Rubio 1999: 5, 8-9).^^

22. On ''ab-Uj, particularly in connection to Kuwara, see Sallaberger 1993, 2: 134; Selz 1995: 115 n. 429; Steinkeller 1995: 277; Such-Gutiérrez 2003, 1: 171 n. 722, 247 n. 1078. 23. Richter (2004: 118 n. 526) wants to read ""ab-ba-U^ as ''ab-'"bag in Vr Lament 24-25. However, all the manuscripts containing those lines have ''ab-ba-Uj, with the exception of two that confuse the uncommon ''ab-ba-Uj and write instead ''ba-Uj (Römer 2004: 15,109). The ubiquity of a phonetic gloss throughout a complete manuscript tradition would be rare. 24. On the Sumerian word for "ax" or "small hoe," see Limet 1960: 243-49, 262-63, et passim; Salonen 1968: 143-47; Wilcke 1972-1975: 34-35. Pace Limet (1960:119, 270) there is no word "uruda-ba^-ba^." In Diri 6 B 84 (MSL 15, 192-93) tbe sequence URUDA.Uj.U2 is given the reading ku-ku-us and equated with Akkadian rätu, which means "channel, runnel, drain, metallic pipe" (CAD R, 219-20; Goetze 1945: 235; Salonen 1964: 124). This should then be read '"»''"kuS3-ku§3 (Uj = kuS3, as in Proto-Ea 231 [MSL 14, 41]). 25. A tantalizing possibility for Akkadian hapütu and Sumerian ha-bu3-da is offered by the Semitic root attested in the Arabic verb habata "to beat, strike," whose potential cognates would exhibit a mild irregularity in the labial second radical (Ge'ez hafata "to scratch, engrave"; Leslau 1987: 260a). Another two semantically close Akkadian words have more likely Semitic cognates; marru "spade, shovel" (CAD M, 287: AHw 612) is related to Arabic marr "iron shovel, spade"—although this may be a Wanderwort (Rubio 1999: 9)—and allu "hoe" (CAD A/\, 356; AHw 37) has a possible cognate in Gurage wällät "forked digging stick" (Leslau 1979: 653). 26. The possibility of reading the anthroponym Uj-U^ as bun-bUn, taken to be a variant spelling of bu3-bu3, cannot be substantiated; see Steinkeller 2003: 621 n. 3.

38

GONZALO RUBIO

4. On the Nature of Glides in Sumerian In some regards, the reading ba^ of U2 is related to its reading wa3. The latter is well attested in Ur III spellings of the name Huwawa as hu-U2-U2 (hu-wa3-wa3), which coexists with the spelling h u - b a - b a in the same period.^' However, the reading wa3 of U2 does not point to the presence of a velar glide (or voiced labiovelar approximant) in the name of Ningirsu's spouse. Although there are specific sequences of syllabograms that seem to point to the existence of palatal and velar glides in Sumerian, there is little reason to assume that they had a true phonological entity or were anything other than secondary or marginal segments. The few possible instances of glides appear to be limited to allophonic variants, particularly junctural realizations in prefix and suffix sequences, as well as a few expressive words, interjections, and perhaps loanwords.'* Huwawa itself, particularly in its literary context, is obviously a motivated name intended to convey onomatopoeic suggestions of monstrosity, strangeness, and foreignness. Few Sumerian lexical morphemes seem to contain traces of glides. It is no coincidence that some of the best evidence for glides in Sumerian words and word sequences comes from Ebla (Civil 1982:6,13-15; 1984: 80), precisely where one expects a Semiticized reading of Sumerian (Civil 1983; Civil and Rubio 1999)." Other Sumerian syllabograms with initial /w/ have more basic and common readings with initial labial stop: BA = wa2, be^; PI = bÍ3, be^, wa/we/wi. This by itself can be regarded as a symptom of the non-phonological and marginal nature of glides in Sumerian. Since our readings of Sumerian syllabograms are filtered through the intermediary lens of Akkadian phonology and Akkadian does have widespread phonological glides, it is difficult to find solid reasons for which glides would not be more obvious and frequently attested in the Sumerian writing interface, if this language actually had non-marginal phonological glides. In fact, the very readings of Sumerian syllabograms with /w/ most likely stemmed from the reinterpretation of signs with labial segments (BA, PI/bÍ3) or back vowels (U2, Ug/waJ, a process that was primarily triggered by the need to write the velar glide in Akkadian.^"

5. Bawu or Baba? While the idea of understanding ''ba-U2 as /bawu/ is not new (Kraus 1951: 67, 69, 74-75; Litke 1998: 173 n. 50), Marchesi (2002) has put forward the most erudite and comprehensive case for such a reading. However, as has been shown here, it is still very difficult to explain away all the evidence, which leads to the following conclusions: (a) The alternation between the spellings m is rare. Akkadian is rather exceptional in this regard. Cross-linguistically, the shift m > w, particularly in intervocalic position, is far more common. Thus, if one were to postulate a hypothetical proto-form, or an underlying representation, the assumption should be en < /*ewen/ < /*emen/, the emesal word being closer to the original (see Schretter 1990: 263-64). 30. The personal name wa-wa-ti occurs a few times in Ur III (MAD 3, 9). As Michalowski (personal communication) points out, wawa-ti occurs mainly in contexts in which other anthroponyms are Semitic (TCL 2 5572; TIM 3 151; TGS 1 346; CST328; Nisaba 16:63; Owen and Young 1971: 98-99, 111 no. 8; Michalowski 1977: 95; Hruska 1980: 20). Thus, wa-wa-ti is probably the Semitic version of the name normally written ba-ba-1 i and, less often, ba4-ba4-ti in properly Sumerian contexts (RIME 3/2,340-42; Limet 1968: 383; Michalowski 1976: 48-56, 217-20; Walker 1983: 93).


39

glide in ''ba-U2 is minimal, if only because of the limited and peripheral evidence for the marginal, non-phonological, and eminently secondary existence of glides in Sumerian. It seems, therefore, more logical to assume the reading ''ba-ba^ (Selz 1995: 26; Richter 2004:118-19 n. 526). Regardless of its ultimate origin, the name Baba exhibits a common pattern attested in other names of deities for which no obvious Sumerian or Semitic etymologies can usually be found, such as Alala, Aruru, Belili, Búlala, Bunene, Igigi, Inanna, Izuzu, Kubaba, Sidada, Zababa, and so forth (Rubio 1999: 3).

References Archi, A. 1987 The "Sign-List" from Ebla. Eblaitica 1:91-113. Attinger, R 1984 Enki et Ninhursaga. ZA 74: 1-54. 1993 Eléments de linguistique sumérienne: La construction de du^^/e/di. OBO Sonderband. Göttingen: Vandenhoeck & Ruprecht. Benito, C. 1969 "Enki and Ninmah" and "Enki and the World Order." Ph.D. dissertation. University of Pennsylvania. Borger, R. 1996 Beiträge zum Inschriftenwerk Assurbanipals. Wiesbaden: Harrassowitz. 2003 Mesopotamisches Zeichelexikon. AOAT 305. Münster: Ugarit-Verlag. Cavigneaux, A., and al-Rawi, E. N. H. 1995 Textes magiques de Teil Haddad (Textes de Teil Haddad li). Deuxième partie. ZA 85:19-46. 2000 Gilgames et la Mort: Texts de Tell Haddad VI, avec un appendice sur les textes funéraires sumériens. CM 19. Groningen: Styx. Charpin, D. 1986 Le clergé d'Ur au siècle d'Hammurabi. Geneva: Droz. Civil, M. 1967 Remarks on "Sumerian and Bilingual Texts" (review of O. R. Gurney and J. J. Einkelstein, The Sultántepe Tablets, vol. 2, London, 1964). INES 26: 200-211. 1976 The Song of the Plowing Oxen. Pp. 83-95 in Kramer Anniversary Volume: Cuneiform Studies in Honor of Samuel Noah Kramer, ed. B. L. Eichler et al. AOAT 25. Kevelaer : Butzon & Bercker. 1982 Studies on Early Dynastic Lexicography, I. OrAn 21: 1-26. 1983 Bilingualism in Logographically Written Languages: Sumerian in Ebla. Pp. 75-97 in // bilingüismo a Ebla, ed. L. Cagni. Naples: Istltuto universitario orientale. 1987 The Early History of HAR-ra: The Ebla Link. Pp. 179-89 in Ebla 1975-1985, ed. L. Cagni. Naples: Istituto Universitario Orientale. 1999Reading Gilgames. Pp. 179-89 in Arbor Scientiae: Estudios del Próximo Oriente Antiguo. Festschrifl for Gregorio 2000 del Olmo Lete on Occasion of his 65th Anniversary, ed. M. Molina, I. Marquez, and J. Sanmartin. AuOr 17-18. Barcelona: Ausa. 2008 The Early Dynastic Practical Vocabulary A (Archaic HAR-ra A). ARETS 4. Rome: La Sapienza. Civil, M, and Rubio, G. 1999 An Ebla Incantation against Insomnia and the Semiticization of Sumerian: Notes on ARET 5 8b and 9. OrNS 68: 254-66. Cohen, M. E. 1975ur.sag.me.sár.ur: A s i r n a m s u b b a ofNinurta. WdO8: 22-36. 1976 1981 Sumerian Hymnology: The Ersemma. HUCA Suppl. 2. Cincinnati: Hebrew Union College. 1988 The Canonical Lamentations of Ancient Mesopotamia. Potomac, MD: CDL. Delnero, P. 2006 Variation in Sumerian Literary Compositions: A Case Study Based on the Decad. Ph.D. dissertation. University of Pennsylvania. Di Vito, R. A. 1993 Studies in Third Millennium Sumerian and Akkadian Personal Names. Studia Pohl 16. Rome: Pontificio Istituto Biblico.

40

GONZALO RUBIO

Edzard, D. O. 1993 "Gilgamesch und Huwawa": Zwei Versionen der sumeriscben Zedernwaldepisode nebst einer Edition von Version "B." Sitzungberichte der Bayerischen Akademie der Wissenschaften, philos.-hist. Klasse, 1993, 14. Munich: Bayerische Akademie der Wissenschaften. Englund, R. K. 1990 Organisation und Verwaltung der Ur Ill-Pischerei. BBVO 10. Berlin: Reimer. Farber, W. 1974 Von BA anderen Wassertieren: testudines sargonicae? JCS 26: 195-207. Falkenstein, A. 1950 Sumerische religiöse Texte. ZA 49: 80-150. 1959 Sumerische Götterlieder, L Abhandlungen der Heidelberger Akademie der Wissenschaften, Phil.-bist. Klasse, Jahrg. 1959. Heidelberg: Winter. Focke, K. 1998 Die Göttin Nin-imma. ZA 88:196-224. Gelb, 1. J. 1960 The Name of the Goddess Innin. JNES 19: 72-79. Geller, M. J. 1985 Porerunners to Udug-hul: Sumerian Exorcist Incantations. FAOS 12. Stuttgart: Steiner. 1989 A New Piece of Witchcraft. Pp. 193-205 in Dumu-e^-dub-ba-a: Studies in Honor ofAke W. Sjöberg, ed. H. Behrens et al. Philadelphia: University Museum. George, A. R. 1992 Babylonian Topographical Texts. OLA 40. Leuven: Peeters. Goetze, A. 1945 The Vocabulary of the Princeton Theological Seminary. JAOS 65: 223-37. Green, M. W 1984 The Uruk Lament. JAOS 104: 253-79. Hämeen-Anttila, J. 2000 A Sketch of Neo-Assyrian Grammar. SAAS 13. Helsinki: The Neo-Assyrian Text Corpus Project. Heimpel, W. 1972 Review of Â. Sjöberg, The Collection of the Sumerian Temple Hymns (Augustin, 1969). JAOS 92: 285-88. Hinz, W 1976Insusinak. RIA 5: 117-19. 1980 Hinz, W, and H. Koch 1987 Elamisches Wörterbuch, I-U. AMI 17. Berlin: Reimer. Hruska, B. 1980 Drei neusumerische Texte aus Drehem. WdO 11: 17-22. 1969 Die spätbabylonische Lehrgedicht 'Inannas Erhöhung.' ArOr 37: 473-522. Jestin, R. 1965 Noies de graphie et de phonétique sumériennes. Paris: Honoré Champion. Karahashi, F. 2000 Sumerian Compound Verbs with Body-Part Terms. Ph.D. dissertation. University of Chicago Klein, J. 1981 Three Sulgi Hymns: Sumerian Royal Hymns Glorifying King Èulgi ofUr. Ramat-Gan: Bar-Ilan University Press. Kramer, S. N. 1940 Langdon's Historical and Religious Texts from the Temple Library of Nippur: Additions and Corrections. JAOS 60: 234-57. 1980 The Death of Dumuzi: A New Sumerian Version. AnSt 30: 5-13. 1989 BM 100042: A Hymn to Su-Sin and an Adab to Nergal. Pp. 303-16 in Dumu-e^-dub-ba-a: Studies in Honor ofAke W. Sjöberg, ed. H. Behrens et al. Philadelphia: University Museum. Kraus, E R. 1951 Nippur und Isin nach altbabylonischen Rechtsurkunden. JCS 3:1-228. Krebernik, M. 1986 Die Götterlisten aus Fära. ZA 76: 161-204. Leslau, W 1979 Etymological Dictionary ofGurage (Ethiopie), I-IU. Wiesbaden: Harrassowitz.

READING SUMERIAN NAMES, I 1987 Limet, H. 1960 1968

41

Comparative Dictionary ofGe'ez. Wiesbaden: Harrassowitz. Le travail du métal au pays de Sumer au temps de la III' dynastie d'Ur. Bibliothèque de la Eaculté de Philosophie de l'Université de Liège, 155. Paris: Les Belles Lettres. Ihnthroponymie sumérienne dans les documents de la 3' dynastie d'Ur. Bibliothèque de la Eaculté de Philosophie et Lettres de l'Université de Liège, 180. Paris: Les Belles Lettres.

Litke, R. L. 1998 A Reconstruction of the Assyro-Babylonian God-Lists, AN: Â-nu-um and AN: Anu sa améli. TBC 3. New Haven: Yale Babylonian Collection. Löhnert, A. 2009 "Wie die Sonne tritt heraus!" Eine Klage zum Auszug Enlils mit einer Untersuchung zu Komposition und Tradition sumerischer Klagelieder in altbabylonischer Zeit. AOAT 365. Münster: Ugarit-Verlag. Luukko, M. 2004 Grammatical Variation in Neo-Assyrian. SAAS 16. Helsinki: The Neo-Assyrian Text Corpus Project. Malbran-Labat, E. 1995 Les inscriptions royales de Suse. Paris: Réunion des Musées Nationaux. Mander, P. 1986 II pantheon di Abu-Sälabtkh (sic). Naples: Istituto Universitario Orientale. Marchesi, G. 2002 On the Divine Name "20 exceeded by 4."" The arithmetical structure of the ternary numeral system is complex. In the numeral sequence 1 to 7, one should, in view of the established universal pattern for numeral systems that base numbers are generally omitted, expect forms such as *gi-bala "4" (1 beyond [3]) without the basic numeral pes, "3." Our texts, however, seem to omit the numeral to be added, e.g., pes-bala. Thus arithmetically pes-bala and simple pes have identical number values, as also have pes-bala-gi and pes-gi. Gonsequently, the given number values in the Akkadian rendering make only sense on the assumption that the auxiliary base bala already implies an extension by "1" (step) so that pes-bala is to be analyzed as "3 (+1) beyond (-t-NUM)." The suffixed -/bi/ in the incantation K 2438, where it is used as a numeral, and not as a possessive pronoun, alternating with gi "1" in other textual sources, may perhaps be explained as a simple dialectal /g/:/b/-shift—taking bi as the Emesal equivalent of g i in the main dialect (e.g., sig4:se-eb "brick(work)," but while this is not impossible, there are arguments against it. For in contrast to /gi/ that is equated with isten (= ge-e) in Ea II 54, 202 (cf MSL 14, 249 and 256), a similar lexical evidence for /bi/ doesn't exist to the present author's knowledge. Therefore this element -bi might well be interpreted as an allograph of the coordinative particle "and," also equated with u in Akkadian, representing a shortened form of the connective particle -/bida/ "and, together with." In view of the presented lexical material, an analysis for both pes-bi-bi, "5," and pes-pes-bi-bi, "8," as /three-and-and/ and /three-three-and-and/, respectively, is in all likelihood to be preferred, even if this would assume a sweeping morphemic erosion like */bid(a)-bid(a)/ >

The Origin of the Ternary Numeral System Although the evidence in the mythical tale Enki and Ninmah already indicates that this peculiar numeral system is not restricted to first-millennium texts, the question of an possibly earlier origin is strictly connected with the etymological and semantic interpretation of the basic lexeme pes and its semantic and lexical differentiation from "regular" es^,, "3." As I will argue in due course, the ternary system and its base unit pes probably originated from an established system of verbal bases, such as gi(4) "to return," tah "to add," taka4 "to leave," and also pes "to make wide/thick,"" that is partially conceived in single entries of the lexical sources, for instance, gi, "1," and tah, "2," in Antagal C 52-53, taka4(-a-bi) "2" in CTMMA II, 61:35 and gi "1" in Enki and Ninmah 59. This verbal system, presented in Table 2 with reference to the semantic concepts of "2" and "3," served to establish an alternative system of numeration for the first three low numerals, 1-3, certainly comparable to the use of tab "to make double" in expressions as m i n - t a b "doubled" or eSj-tab-ba "threefold."''' We can find traces of this system apart from purely lexical and grammatical sources in literary texts as the Farmer's Instructions (ex. 1) and in the introductory part of letters (ex. 2), following the main clause: (1) gis ùr-ra-ab gis gi^-a-ab gis pes-bi-ib "Harrow (once), harrow a second time, harrow a third time!" (Farmer's Instructions 32), and also:

11. See also in line 31: mi-in ni di-ri-ga "45" (< min( + )ni(s) [+ia](+) diri[g]) > 2 x 20, exceededby [5?]. 12. For the coordinative particle -/bida/ see Th. E. Balke 2006: 116-18, especially for its allomorph -bi, p. 117 with n. 500. 13. Attestations of the simple verbal base pes, apart from the well-known compounds SU~PES "to enlarge" and gú~PES "to grow," are only quite rarely found and not always easily to be connected with the assumed primary semantic, e.g., nig(-)na-de5 gij-a e n - n u - ù g sa,a-gin^ sila dagal pes-pes-e "You (= Hendursaga) who enlarges threefold (?) the city's squares by night during the middle of the watch" (Hendursaga hymn A 12; cf. Attinger and Krebernik 2005: 37, 57). 14. See also the lexical evidence in MSL 9, 179: gikid sa m i n - t a b - b a (= sa ana libbi efpu) that occurs together with the synonymous but somehow obscure «"kid sa pes gÍ4-a, "reed mat whose middle part is twined fourth times (?)."

THE SUMERIAN TERNARY NUMERAL SYSTEM

49

(2) ù - n a - d è - p e s "Say it to him a third time" (var.: ù - n a - d e - t a h "Add it for him again" [= a second time]) (Ali, Letter Coll. B 16:7)'^

T A B L E 2. SYSTEM O F V E R B A L B A S E S E X P R E S S I N G R E P E A T E D S E Q U E N C E S OF S Y M B O L I C A C T I O N S .

Concept of "2" Concept of "3"

g i 4, "to return, to repeat" > "second time," "first countable step" (cf second < insequi "to follow") t a h , "to add" > "to add a further step (= 2)" / "to do something a second time" t a k a 4, "to leave" > "one left over (beyond 1 = 2)" (e.g., eleven "U" < *ain-lif "one left [beyond ten]") p e s , "to make thick/wide" > "to enlarge by a (final) third step" (= "3"

Thus the ternary counting system, restricted to the lower numerals of the first ten numbers, might have had a real and unmysterious background in spoken Sumerian long before the first millennium B.C. Within this system, pes, "to make thick/wide," represents the end and limit of the whole counting sequence expressing a last final extension by "1" (step) so that certain symbolic actions would be usually perfectly performed by three steps or attempts (cf. three wishes and tasks in fairytales). In contrast, examples such as pes-a-bi (= sullusu), in 1.16 of the dialogue "Examination Text A" (Sjöberg 1975), that might well be interpreted as traces of a ternary structure, only provide further evidence for the use of the verbal base pes, "triple, threefold," with reference to verbal triplication, but not for an entire numeral system.'* In my opinion, one can find further support for an early semantic differentiation between es,5,, "3," and pes, "to be thick/wide (= pregnant)," in texts from the third-millennium lexical tradition, respectively: e.g., áb es, "three-years old cow" (ED Animal list A 5 = MEE 3 51) vs. áb pes, "thick (= pregnant) cow" (ED Animal hst B 14 = MEE 3 65). To this one might possibly add the name of the deity''ES(5,-p es, "Threefold-thick one(?)," attested in the ED god lists from Eara and Tell Abu Saläblkh (SF 1 xii 9 = ZA 76 180; LAS 83 iii '5')." This name can be linked with the concept of numeration and seems to confirm a lexical (and morphemic) differentiation of es¡5) and pes already in the third millennium B.C. However, it cannot be ruled out that in the divine name either the "regular" numeral eS(5,,"3," functioned as a determinative for PE§ or that pes served as a phonetic compliment for the preceding es (5), and in the end this would corroborate the proposal for a common origin of both words. Nevertheless, the given evidence, especially from the third millennium, clearly indicates a diachronically independent semantic and morphemic relationship of the two cardinal numeral bases eS(5),"3," and pes, "to make thick/wide" x "3," and speaks against an assumed morphological development *ewes/PES > es^, "3," as well as against the hypothesis of the development of the regular cardinal numeral system out of a ternary counting system.

The Cognitive and Cultural Background Due to its limited range of application, it is unlikely that the ternary system was used to quantify lots of discrete objects in an economical environment, but it seems to be mainly restricted to the context of symbolic ritual

15. A similar use of tah "to add (words)" is possibly attested in Inanna and Sukaletuda 48, if correctly completed asinim m u - n i - [ t a h e']. 16. For this text see also Black (1991: 72-74) who interpreted this passage as a possible reference to the grouping in verbal paradigms of forms into threes, each of three persons. 17. For this deity see Mander 1986: 59; Alster and A. Westenholz, 1994: 37; and Such-Gutierrez 2005-2006:16, no. 38. It is also attested in the personal name u r-•"£á ,5)-pes (WF59: 2) and in the Adab toponym iîd"i -""Es-p es - s a4(SIA A 839:7)

50

THOMAS E. BALKE

actions with a strong emphasis on a temporal pattern, as can be seen in the structure of the incantation K 2438-H, where each numeral is preceded by ud, "day" and followed by ud kalag-bi "its strong day," the latter a term that belongs to the specialised language of Babylonian hemerologies.'* From an arithmetical and pragmatic point of view, choosing "3" as the main numeral unit of an independent counting system has psychologically the advantage that "3" is about the limit of subitization, for it allows instant number judgement without counting and counting with only one hand, too. Anthropologically, "3" is the smallest designation of a multitude in so-called primitive societies, and represents often the smallest constellation possible, for instance, in court procedures, for example, the German Schöffengericht consists of at least two lay judges and one professional judge. Furthermore "3"—and also "6"—are components by which higher numerals of the sexagesimal system can be divided without leaving a remainder, what is impossible for "7" or "8." Thus in mathematical calculations working with the standard month length of 30 days, "3" would improve the countability of complex accounts. Moreover, every cardinal number of the sexagesimal system can be interpreted as an integer containing no other prime factors than 2, 3, and 5. On the other hand, if we look at "7" as the usual endpoint of the counting sequence, it is the first non-regular integer according to the standards of Old Babylonian mathematics that is not exclusively divisible by 2, 3, or 5, and connotes a symbolic meaning "innumerable" (= kissatu) within this corpus as well. The symbolic prominence of 7, certainly due to the astronomical relevance of the single moon phases with the crown's half-size on the 7th day, in its designation ofa group's completeness is pretty evident in Sumero-Babylonian literature as well, for example, in Inanna's Descent, where Inanna has to give up her seven me-symbols, her crown, her measuring rod, her wig, etc., when entering through the seven gates of the netherworld, or in Aträ-hasis, when Bêlet-kâla-ili is mentioned to nip off fourteen pieces of clay to produce seven males and seven females, or as in the final example the Fâra literary composition dealing with the seven children of Asnan (cf. IAS 231). Here many more examples could be quoted, but it is obvious, that the numeral 7 generally connoted the propitious completeness of a limited number of steps or members ofa group, respectively." The same symbolic prominence though to a lesser extent is relevant for 12, the endpoint of the counting sequence in K 2438, as well. This number is primarily associated with the total number of months in the cycle of the year, representing a common factor within the sexagesimal structure that can be factorized in the basic digits 2 and 3 as 2 • 2 • 3.^° In conclusion, the ternary numeral system can certainly be traced back at least to the early-second millennium B.c., and its primary setting appears to lie in symbolic ritual contexts. Despite the sporadic evidence and the atypical structure the ternary system must be considered a subsystem of Sumerian numerals that is a remnant of a possibly older means of counting. The cultural background of the ternary counting system and the morphological shape of the "regular" cardinal Sumerian numerals will require further research in the years to come.

Appendix to /merga/ "one, first" While the evidence for the Sumerian verbal bases gi(4) "to return," tah "to add," taka4 "to leave over," and pes "to make thick/wide," connoting the semantic concepts of "r'(?), "2," and "3," is relatively clear, the few occurrences for me/irga, "1," quoted above, do not seem to point to a verbal base as its etymon.^' Though this heterogeneous picture only confirms the universal observation that for the numeral "1" as the most frequently occurring numeral in languages, this higher frequency may often encourage morphophonemic irregularities and idiosyncratic fusions with other morphemes as can be already seen from the multiple English allofams derived somehow from

18. For this aspect of ud kalag, see Livingstone, 1998: 62-64. 19. For the symbolic relevance of this number see in detail Reinhold 2008: 1-10, 65-67, etc. 20. Only the Epic of Gilgames seems to contain traces of the symbolic importance of the numeral 12, for instance, when Enkidu lies on his deathbed till the twelfth day, or when Gilgames follows the course of the sun through the mountain of darkness for twelve "double-hours." 21. In this case one would have to assume a construed compound verb consisting of me-r(i), a dialectal variant of giri, "foot, step," and ga, a dialectal variant of de^/tiim "to bring" (see mús~GA "to cease [working]"), but this is highly speculative.

THE SUMERIAN TERNARY NUMERAL SYSTEM

51

the IE root *oino (c£ one, an, once, only, alone, anon, onion, etc.; see Lujan Martinez 1999: 206-9). In Sumerian we also find strictly-speaking at least six lexemes belonging to the semantic sphere of the numeral concept "1," for example, dis/as—with the dialectal variant di-id/ta—dili, gi,^); bi; usum (cf MSL 14, 502: 203 [ú-su-um : BUR : is-te-en]^^) and the obscure merga. Since number words from one to ten, especially within the range from 1 to 3, are rarely replaced by loan words, what might be indicated by the final -/a/ of the word (e.g., ne-ha "rest" from Akkadian néhtum)," a genuine Sumerian etymology is generally to be presumed. A new aspect is provided by the obvious use as an adverb in Rim-Sin Hymn B 35 (me-er-ga-bé). This may be connected with the use of adverbials as dili-bé, "alone," and proves a deeper semantic affmity to the concept of "1" already in the Old Babylonian period. On the other hand, me/irga does not seem to belong—in contrast with tab "to double," for instance—to the standard inventory of terms used in the corpus of Old Babylonian mathematical texts (cf the list in Hoyrup 2002:43-49). Unfortunately, in the end one can only speculate about the origin of this numeral lexeme, but, according to universal patterns of the "body-part model" by which expressions for single parts of the human body serve as sources for numeral bases (cf "hand" > "5"), it might well be associated with the Sumerian nominal base murgu "shoulder" indicating the beginning of the counting sequence by the metaphorical use of an upper body-part term (cf Lynch 1998: 251).^" A similar use of sag "head" can be observed in expressions as sag-gal/ kal "foremost one" (see RIME 4, Idd. 2:2, Sin. 9:3). Nevertheless, such a semantic and lexical association remains a little bit shaky. In yiew of the presented Old Babylonian occurrences from Ur, UET 193: 3 and UET 613 561: 13, where is occurs together with gi^ "2," one is inclined to interpret this lexeme whose origins seem to lie outside the Old Babylonian Edubbâ school tradition, as a possible lexical(?) variant of murgu first attested in texts from Ur.

References Alster, B., and Westenholz, A. 1994 The Barton Cylinder. ASJ 16: 15-46. Attinger, P., and Krebernik, M. 2005 L'Hymne à Hendursaga (Hendursaga A). Pp. 21-104 in Von Sumer bis Homer: Pestschrifl für Manfred Schretter, ed. R. Rollinger. Münster: Ugarit-Verlag. Balke, T. E. 2006 Das sumerische Dimensionalkasussystem. AOAT 331. Münster: Ugarit-Verlag. 2009 Das sumerische Numeraliasystem: Versuch einer typologischen Einordnung. Pp. 671 -98 in Empirical Dimensions in Ancient Near Eastern Studies, ed. G. ]. Selz, WOO 6. Wien: Lit. Benito, C. A. 1969 Enki and Ninmah and Enki and the World Order. Ph.D. dissertation. University of Pennsylvania. Black, I. A. 1991 Sumerian Grammar in Babylonian Theory. Studia Pohl 12 Roma: Pontificio Istituto Biblico. Borger, R. 1985 Einige Texte religiösen Inhalts. OrNS 54: 14-26 Brisch, N. M. 2007 Tradition and the Poetics of Innovation: Sumerian Court Literature of the Larsa Dynasty (c. 2003-1763 BCE). AOAT 339. Münster: Ugarit-Vedag. Civil, M. 2005 No. 61: Late Babylonian Grammatical Text. Pp. 244-47 in Cuneiform Texts in the Metropolitan Museum ofArt, vol. 2: Literary and Scholastic Texts of the Eirst Millennium B.C., ed. I. Spar and W. G. Lambert. CTMMA 2. New York: The Metropolitan Museum of Art.

22. See also ú-aú = e-dei-su "alone" in MSL 3 142: 169. How this numeral concept can be associated and harmonized with the (primary) meaning "dragon," e.g., in the names of deities as''usum-ur-sag or''uäum-dubur-an-na, remains unclear. 23. See also Sjöberg 1967: 276 no. 3. with reference to d / k - i r - g a "piled up (water)" < icerku(m} (cf AHw 468a). 24. In languages with a fully developed body-part counting system, however, a lexeme "shoulder" albeit attested is usually associated with higher numerals such as "10" {ragpd) in the Papuan language Haruai; see Comrie 1999: 88-91.

52 Comrie, B. 1999

THOMAS E. BALKE

Haruai Numerais and Their Implications for the History and Typology of Numeral Systems. Pp. 81-94 in Numeral Types and Changes Worldwide, ed. I. Gvozdanovic. Berlin: de Gruyter. Diakonoff, I. A. 1983 Some Reflections on Numerals in Sumerian. JAOS 103: 83-93. Edzard, D. O. 2003 Sumerian Grammar. HO 1.71. Leiden: Brill. Englund, R. K. 1998 Texts from the Late Uruk Period. Pp. 15-233 in Mesopotamien—Spaturuk-Zeit und Erühdynastische Zeit, ed. P. Attinger and M. Wafler. OBO 160/1. Fribourg: Vandenhoeck & Ruprecht. Hanke, T 2005 Bildungsweisen von Numeralia. BBL 3. Berlin: Weißensee. Heine, B. and T. Kuteva 2002 World Lexicon ofGrammaticalization. Cambridge: Cambridge University Press. Hoyrup, I. 2002 Lengths, Widths, Surfaces: A Portrait of Old Babylonian Algebra and Its Kin. Sources and Studies in the History of Mathematics and Physical Sciences. Heidelberg: Springer. Hunger, H. 1998 Zur Lesung sumerischer Zahlwörter. Pp. 179-83 in dubsar anta-men. Eestschrift für Willem H.Ph. Römer, ed. M. Dietrich and O. Loretz. AOAT 253. Münster: Ugarit-Verlag. Hurford, I. R. 1999 Artificially Growing a Numeral System. Pp. 7-41 in Numeral Types and Changes Worldwide, ed. ]. Gvozdanovic. Berlin: de Gruyter. Lambert, W. G. 1969 Review of O. R. Gurney and P. Hulin The Sultantepe Tablets II (Ankara-London, 1964). JSS 14: 242-47. Langdon, S. 1917 Syllabar in the Metropolitan Museum. JSOR 1:19-23. Livingstone, A. 1998 The Use of Magic in the Assyrian and Babylonian Hemerologies and Menologies. SEL 15: 59-67. Lujan Martinez, E. R. 1999 The Indo-European System of Numerals from "1" to "10." Pp. 199-219 in Numeral Types and Changes Worldwide, ed. ). Gvozdanovic. Berlin: de Gruyter. Lynch, I. 1998 Pacific Languages. Honolulu: University of Hawaii Press. Mander, P. 1986 II Pantheon di Abu Sàlablkh. Contributo alio studio del pantheon sumerico arcaico. Istituto Universitario Orientale, Departimento di Studi Asiatici. Series Minor 26. Napoli: Università degli Studi di Napoli "L'Orientale." Powell, M. A. Jr. 1971 Sumerian Numeration and Metrology. Ph.D. dissertaton. University of Minnesota. Proust, C. 2008 Les listes et tables métrologiques, entre mathématiques et lexicographie. Pp. 137-53 in Proceedings of the 51st Rencontre Assyriologique Internationale held at the University of Chicago, July 18-22,2005: Lexicography, Philology, and Textual Studies, ed. R. D. Biggs, J. Myers, and M. T. Roth. SAOC 62. Chicago: The Oriental Institute of the University of Chicago. Reinhold, G. G. G. 2008 The Number Seven in the Ancient Near East. Erankfurt: Lang. Sjöberg, Â. W. 1967 Contributions to the Sumerian Lexicon. JCS 21: 275-78. 1975 Der Examenstext A. ZA 64: 137-76. 1993 CBS 11319-H. An Old-Babylonian Schooltext from Nippur. ZA 83: 1-21. 1996 UET 7 no. 93. A Lexical Text or a Commentary? ZA 86: 220-37. Such-Gutierrez, M. 2005Untersuchungen zum Pantheon von Adab im 3.Jt AfO 51:1-44. 2006 Thureau-Dangin, E. 1928 Le système ternaire dans la numeration sumérienne. RA 25: 119-21.

SUMERIAN EXTRACT TABLETS AND SCRIBAL EDUCATION Paul Delnero (The Johns Hopkins University)

In recent years there has been renewed interest in scribal education in ancient Mesopotamia.' Much of this research has focused on the early-second millennium, or the period known as the Old Babylonian period (ca. 2000-1595 B.c.),^ a time for which there is abundant evidence for scribal training. Earlier treatments of this topic, like Kramer's "The Sumerian School: A Pre-Greek System of Education," (1951). Falkenstein's "Die babylonische Schule" (1953) and especially Sjöberg's seminal "The Old Babylonian Edubba" (1975), were based primarily on a group of Sumerian literary compositions about the Eduba, the academy or institution in which scribes were trained. But it is now generally acknowledged that these texts contain idealized accounts of the daily life and activities of scribal pupils, and are not reliable sources for reconstructing the educational practices of this time> By considering additional sources of evidence, such as archaeological context," and correlations in the groupings of different types of compositions on scribal exercise tablets as indications of the sequence of the scribal curriculum,^ new insight has been gained into how scribes were taught the Sumerian language and the cuneiform signs used to write it. One particularly productive approach has been the study of tablet typology. Along with the archaeological contexts in which they were discovered and the quality of their script, the shape and format of tablets containing Sumerian lexical lists, collections of "proverbs," and literary compositions indicate clearly that many, if not all of the sources for these texts are school exercises. The classification of these tablets into different types also sheds

1. I would like to thank J. Cooper and D. Fleming for their helpful comments and criticisms during the preparation of this article. 2. See Velhuis 2000 for the Middle Babylonian period, and Gesche 2000 and Veldhuis 2003 for the Neo-Babylonian period. 3. For a more detailed critique of the so-called Edubba texts for understanding scribal practices and a description of the shift, in recent studies of scribal education, away from the study of these texts toward other types of evidence see Robson 2001: 39. A more critically informed use of these compositions to identify aspects of scribal training can be found in Robson 2002: 348-52. For a discussion that includes a list of all the compositions that have been classified as "Edubba texts" with additional references to editions and previous treatments of these texts as sources for reconstructing scribal education see Delnero 2006: 65-81. 4. Studies of the archaeological contexts for scribal training include Robson 2001 for Nippur; Charpin 1986 and Brusasco 1999-2000 for Ur; and Tanret 2002 for Sippar. For a detailed synthesis of the archaeological evidence for scribal education at these and other cities in Mesopotamia such as Isin, Kish, Babylon, and Uruk with further references see Delnero 2006: 35-64; for a more concise summary of some of this data see Robson 2002: 329. All of the contexts in which scribal exercise tablets were discovered suggest that scribes were trained in small groups in private houses during this period and not in large institutional buildings, like modern schools, as was previously assumed. 5. In a pioneering study using this approach, the sequence of the first, or elementary stage of the scribal curriculum, when scribes copied sign and thematic word lists to learn cuneiform signs, their phonetic values, and vocabulary, was reconstructed by Veldhuis (1997: 40-63). Veldhuis's reconstruction served as the basis for identifying the sequence of texts that were learned in the intermediate and advanced stages of the scribal curriculum, in which literary compositions were copied and studied. These two stages were examined by Tinney ( 1999) and Robson (2001), discussed in more detail below. In all of these studies, the sequence in which specific compositions are grouped together on sources with more than one text or across a series of tablets plays a significant role, along with other factors, such as pedagogical logic and identical groupings in literary catalogues, in determining the order in which specific texts were copied during each stage of the scribal curriculum. For a complete list of the sequence of exercises that were learned in Veldhuis's reconstruction of the elementary phase of the scribal curriculum see Veldhuis 1997: 63, and the modified versions of this list presented in tabular form in Robson 2001: 47 and 2002: 331. 53

JCS 62 (2010)

54

PAUL DELNERO

light on the function of these exercises. By analyzing the physical characteristics and the formal features of scribal exercise tablets, it has been possible to distinguish between the texts and exercises that were copied during the elementary phase of the scribal curriculum from those that were learned at a later stage and to gain insight into the methods that were employed to train scribes. In his study of the sources for the Old Babylonian versions of the lexical lists from Nippur, Givil identified four basic "types" of tablets, grouped according to their shape and format: Type I. Large tablets, cylinders, or prisms, with long sections of the series and no extraneous material. [... ] Type IL The obverse contains a two-column calligraphic exercise; the left column is the instructor's model, the right, rarely preserved, the student's copy (= type Uli). The excerpt from the series is about ten to twenty lines long. The reverse contains an excerpt from [...] or another series in four or five columns (occasionally more) running from right to left (= type II/2). Type III. Small one-column tablets with an excerpt from the series. Type IV. Small lentil-shaped tablets.^ Although Givil's typology was conceived primarily for the purpose of classifying the sources for the lexical lists from Nippur, it can also be applied to the sources for compositions belonging to different textual genres. Similar tablet types are attested for literary texts, the proverb collections, and also, to a large extent, mathematical tables, problems, and model contracts.^ Furthermore, comparable types have been found at Ur, Susa, Sippar, Uruk, Kish, and most of the other places where Old Babylonian exercise tablets have been discovered. Even after taking into account regional differences in the exact shape and format of these types, as well as the presence of a few tablet types that were not in use at Nippur at some sites and the absence of some of the Nippur types at others, this typology remains generally applicable to tablets from all of these sites. Civil's typology, with slight modifications and elaborations, has provided a valuable framework for examining scribal education and determining the relative sequence in which different types of exercises were learned. Veldhuis's study of elementary scribal training at Nippur was one of the first attempts to utilize tablet typology systematically as a source of evidence about the Old Babylonian scribal curriculum (Veldhuis 1997). Studying the sources for thematic lexical lists, and the list of wooden objects specifically, Veldhuis observed that many of the duplicates for these texts were Type II tablets. Since other compositions that can be identified as basic exercises, like the sign list "Syllable Alphabet B," the syllabary "Tu-ta-ti," and the lists of personal names also occur on Type II tablets, it is evident that this tablet format was commonly used during the elementary phase of scribal training (Veldhuis 1997: 32-37, 40-41). In addition to Type II tablets, exercises from this stage of the curriculum are also frequently attested on Type IV tablets. These tablets (also referred to as "lentils") typically contain one to two lines of a particular text, written first as a model in a careful and accurate script, and followed by a copy of the same lines in writing that "varies from bad to excellent" (Falkowitz 1983-1984: 21). The model lines were almost certainly written by the instructor, and then copied by the pupil. The difference in the quality of the script in the model text and the often inferior quality of the copy is one of many indicators that these lines were copied by beginning scribes. Another indication that Type IV tablets were used for elementary exercises is their content. In a survey of 350 Type IV tablets from Nippur, Falkowitz compiled a list of all of the texts that occur on these sources (Falkowitz 1983-1984). Nearly all of these are texts that Veldhuis identified as exercises that were learning during "Phase 1," or the elementary phase of the curriculum (Falkowitz 1983-1984: 21). Type IV tablets, along with Type II tablets therefore seem to have been the primary tablets types that were used during the initial stages of scribal training.

6. Civil 1979: 5. See also Civil 1969: 27-28 and 1995: 2308, which include additional "types" to describe the sources for lists from later periods. 7. For a detailed typology of mathematical texts see Robson 1999: 174-79.

SUMERIAN EXTRACT TABLETS AND SCRIBAL EDUCATION

55

The observation that Type II and IV tablets were used for elementary exercises has led to the identification of four additional texts that appear to have been learned early in the sequence of the scribal curriculum. Tinney has shown that the literary compositions Lipit-Estar B (ETCSL 2.5.5.2),^ Iddin-Dagan B (ETCSL 2.5.3.2), Enlil-bani A (ETCSL 2.5.8.1), and Nisaba A (ETCSL 4.16.1), which belong to a group of texts he called the "Tetrad," were probably also copied near the beginning of scribal training (Tinney 1999). The first text in the Tetrad, Lipit-Estar B, was first identified as an elementary exercise by Vanstiphout, in part by observing that the text seems to consist almost entirely of simple sentences with a relatively large number of basic syntactical and grammatical constructions of various types (Vanstiphout 1978: 51; 1979:121-23), but also because many of the sources for this text are Type II and Type IV tablets (Vanstiphout 1978: 51). Tinney proposed that the Tetrad was learned at an intermediate stage between the elementary and advanced phases of the scribal curriculum since the other three texts also occur on these two tablet types, and the four compositions appear together on a series of four six-sided prisms, presumably written by the same scribe, as well as in partial sequence on collective tablets (Tinney 1999: 162-63 with notes 23-24). With the exception of the compositions in the Tetrad, however, literary compositions are only rarely attested on Type II and Type IV tablets. In contrast to sign lists, thematic lexical lists, and other elementary exercises, literary compositions, which Veldhuis argued were learned in the second or advanced phase of the curriculum, occur instead primarily on Type I and Type III tablets. Type I tablets with literary texts are typically divided into four columns, and contain a single composition in its entirety; though variant tablets of this type with more than four columns, and, in some cases, more than one composition (collective tablets) also exist. On the obverse, columns one and two are oriented from left to right; but on the reverse columns three and four are read from right to left. By contrast. Type III tablets with literary compositions have only one column and contain a single extract from a longer text. These sources are typically inscribed on both the obverse and the reverse, and generally contain twenty- to fifty-line extracts of a given text; though tablets of this type that are only inscribed on the obverse, or that have more than sixty lines are also attested. In colophons Type III tablets are sometimes identified with the Sumerian termim-gid^-da. The formal characteristics of Type I and Type III tablets with literary texts are similar, but not identical to those with lexical compositions. Unlike multi-column Type I lexical tablets, which are generally square-shaped, usually have more than four columns, and contain wider lines, with larger, less compactly written signs, for example. Type I literary tablets tend to be longer than they are wide, and to have narrower lines, with smaller signs that are written more closely together. Similarly, Type III lexical tablets are typically shorter, containing extracts of between ten and fifteen lines, and more rectangular, with wide lines and relatively large signs, whereas literary Type III tablets, while also rectangular, are often pillow-shaped, with rounded edges, and have lengthier extracts with narrower, more compact lines. To account for these deviations, Tinney extends Civil's typology, classifying Type I literary tablets as "multicolumn tablets" (Type M) and Type III literary tablets as "single-column tablets" (Type S; Tinney 1999: 160). In general, however, there appear to be more similarities than differences between these two types of lexical and literary sources. Moreover, many of the slight formal differences probably reflect the type of texts they contain, not the function of these tablet formats. Lexical entries are generally shorter and tend to have fewer signs than the lines in literary texts, which could account for why tablets with these texts are usually divided into more numerous, narrower columns, than the sources for literary texts, which have fewer columns with longer lines. Since assigning different labels to Type I and Type III literary tablets could be thought to imply that there are substantial qualitative differences between the sources for literary and lexical texts, in the absence of evidence for such differences, it is

8. Unless otherwise indicated, all references to the titles and numbers of specific literary compositions are the titles and numbers used to label these texts by the Electronic Text Corpus of Sumerian Literature (ETCSL; online: http://etcsl.orinst.ox.ac.uk).

56

PAUL DELNERO

reasonable to assume that Type I and III tablets had the same function for the two types of texts and to maintain the same designations for both groups when referring to these sources. Although Veldhuis has demonstrated the didactic function of Type II and IV tablets during the elementary phase of the curriculum, less is known about the pedagogical function of Type I and III tablets. Since these two tablet types were the primary formats used during the second or advanced phase of the scribal curriculum, understanding their function would also provide insight into the methods of training that were used at this stage. Furthermore, most of the known duplicates of Sumerian literary compositions occur almost exclusively on tablets of both of these types, a factor that has a direct bearing on why these sources were produced and how scribes were trained, but that is nonetheless frequently neglected in studies of Sumerian literature. In this article, the function of Type III tablets as tools for training scribes will be considered in more detail. I will argue that these tablets were used together with Type I tablets at different stages in the process of learning Sumerian literary works.

The Distribution of Type III Extract Tablets Although many of the duplicates of Sumerian literary compositions are Type III sources, and the use of extract tablets is frequently noted in editions and discussions of these texts, little attention has been devoted to determining their function. One question that has been considered, however, is whether the groupings of lines on these types of tablets reflect meaningful divisions. In an early edition of the text Enlil in the Ekur (ETCSL 4.05.1) Falkenstein argued that many of the Type III sources, which he calls einkolumnig tablets, contain extracts that represent distinct narrative or poetic units (Falkenstein 1959: 9-10). Assuming that this composition was arranged in strophes of two to five lines, he noted that most of the Type III sources that he knew began at the beginning and concluded at the end of a strophe.' He distinguished these sources from the four exceptions that began or ended in the middle of a strophe, which he identified as Schülerabschriften (Falkenstein 1959: 9). While the notion that extract tablets reflect the division of compositions into poetic units has not been widely adopted, the distinction between Type III tablets that were written as scribal exercises and those with less arbitrary extracts was maintained in many of the text editions that followed Falkenstein's study. Hallo and van Dijk identifled several groups of extract sources of The Exaltation of Inana in their edition of this composition (ETCSL 4.07.2). These include four groups of three tablets with approximately fifty lines, and three groups of five tablets with approximately thirty lines from different sections of this 153-line text (Hallo and van Dijk 1968: 38-39). The extract tablets in these groups, which Hallo and van Dijk called "three-tablet" and "five-tablet recensions," are contrasted with extract tablets of thirteen to twenty-one lines that do not seem to have been arranged as a series, interpreted as school exercises with "the daily pensum of advanced students" (Hallo and van Dikj 1968: 39). Similar groups of sources

9. At the time Falkenstein was preparing his edition of Enlil in the Ekur twenty-two exemplars were known to him, of which eighteen are Type III tablets. At present at least eighty-nine duplicates, forty-eight of which are Type III tablets, have been identified. For a list of these sources see Delnero 2006: 2108-14. The additional sources complicate the evidence that the extract tablets for this composition begin and end at the beginnings and endings of groups of lines Falkenstein considered to be strophes. In support of tbis theory Falkenstein counted five sources that begin with the first line ofa strophe, three sources that end with the last line ofa strophe, two sources that end with the first line ofa new strophe, and four exceptions (which he identified as school tablets). When the sources not known to Falkenstein are added there are a total of seven sources that begin with the first line ofa strophe, six that end with the last line ofa strophe, ten sources that end with the first line ofa new strophe, seven sources that do not begin at the beginning ofa strophe, ana four sources that end in the middle ofa strophe (two of which also do not begin at the beginning ofa strophe). The new totals increase the number of sources that are consistent with Falkenstein's theory by only five sources, but increase the number of exceptions to thirteen sources. If Falkenstein is correct, however, about the first lines of strophes occurring at the end of sources being Eoigeweiser (lines that indicate the first line of the next tablet in a series), then the number of exceptions decreases to seven, or three more exceptions than he noted.


57

for Nanse A (ETCSL 4.14.1) and The Curse of Agade (ETCSL 2.1.5) were classified as "editions" by Heimpel and Cooper in their publications of these compositions.'" The identification of a series of connected tablets that collectively contain an entire composition as a "recension," however, has been largely rejected. In more recent editions of Sumerian literary compositions these terms generally are no longer used to classify groups of extract tablets. The primary basis for this rejection is the recognition that all, or nearly all of the preserved sources for Sumerian literary texts, including the extract tablets that seem to be part of a series, are school tablets and were not intended to be definitive versions of particular compositions as such terms imply. This position was first proposed by Civil in his edition of The Farmer's Instructions (Civil 1994; ETCSL 5.6.3). Observing that the number of lines chosen for scribal extract tablets is not consistent and that the partitioning of literary texts does not appear to have been the result of a standard process, he argued that "terminology such as 'three-tablet recension' does not... adequately reñect these ñuid conditions, suggesting as it does some kind of formal, official' partition" (Civil 1994:12). Michalowski presents similar reservations about distinguishing between tablets containing scribal exercises tablets and "official" editions in his edition of The Lamentation over the Destruction of for Sumer and Ur (ETCSL 2.2.3) and by Tinney in his edition of Nippur Lament (ETCSL 2.2.4) who both emphasize that the extract sources for these compositions are school texts, and follow Civil in not referring to tablets of this type as "recensions" or "editions" (Michalowski 1989:18; Tinney 1996: 88-89). While it is now broadly accepted that most extract tablets, including those that seem to be part of a series, are scribal exercises, the function of these types of sources as tools for training scribes has yet to be examined. One reason this question may not have been addressed is that discussions of Type III tablets, including those cited above, are typically confined to individual compositions like The Exaltation of Inana, The Cursing of Agade, The Lamentation over the Destruction of for Sumer and Ur, and The Farmer's Instructions, without considering their function more generally across the corpus of Old Babylonian Sumerian literary compositions as a whole. As a result, the function of scribal extract tablets has not been considered in relation to the other tablet types that were used to copy Sumerian literary texts. One way of approaching the question of how Type III sources were used to train scribes is therefore to consider their distribution with more than one composition and with respect to different tablet types. In addition to Type III tablets and multi-column Type I tablets, literary compositions also occur on prisms. Cenerally prisms with literary texts have between four to six sides, and contain a single composition; though prisms with more than one text are also attested. The number of sides a prism contains seems to be contingent on the length of the text or texts that appear on it. Although prisms are also classified as Type I sources in Civil's typology, they are formally quite distinct from multi-column tablets, which Civil grouped in the same category. The quality of the texts inscribed on prisms can be substantially inferior to the textual quality of the other types of sources with the same composition, indicating that prisms either had a different function or were copied by less advanced pupils." Prisms should therefore be viewed as distinct from multi-column Type I tablets and classified separately, following Tinney, who designates them as "Type P" sources (Tinney 1999: 160).

10. Heimpel 1981: 70-72; Cooper 1983: 44-45. Note, however, that Cooper also raised the question of whether a series of connected extract tablets are school tablets instead of editions, stating: "single-column excerpt tablets of ('The Curse of Agade') and other Old Babylonian Sumerian literary compositions were parts of a series of tablets written by individual scribes or students containing the entire text, or whether they were just random excerpts done as exercises." While Cooper does not directly commit to this position, and recognizes at least one group of sources that might represent a multi-tablet edition of The Cursing of Agade, his raising the question anticipates the current consensus about the function of extract tablets as scribal exercises, discussed in more detail below. It should also be noted that Heimpel and Cooper may use the term "edition" to refer more generally to any complete grouping of Type III sources, without necessarily suggesting that these grouping contain definitive versions of a composition, in contrast to Hallo and van Dijk's usage of the term "recension," which implies a more authoritative redaction of a composition. 11. Klein 1981: 170. For a more detailed discussion of prisms, with additional references to the inferior quality of their content, see Veldhuis 1997: 29-32.

58

PAUL DELNERO

Although almost all of the sources for literary compositions can be classified as Type I tablets. Type III tablets, or prisms, there are a few exceptions. In his survey of Type II tablets from Nippur, Veldhuis identified a relatively small number of literary texts that are attested on sources of this type (see Veldhuis 1997: 65-66 for a complete list). Two of these, Lipit-Estar B and Enlil and Namzitara (ETCSL 5.7.1) occur on more than one tablet with this format, but the instances in which other literary texts are found on Type II tablets appear to be limited. One isolated example is Lipit-Estar A (ETCSL 2.5.5.1) which occurs on the reverse of a Type II tablet and contains an extract from the lexical list "Proto-Aa" on the obverse (CBS 10988). Since according to Veldhuis's interpretation of the pedagogical function of Type II tablets the text on the reverse was learned earlier in the scribal curriculum than the text on the obverse (Veldhuis 1997: 32-37), this source would imply that Lipit-Estar A was learned before "Proto-Aa"—a possibility that is at variance with the assumption that literary texts were learned later than lexical compositions. In the absence of similar evidence of this type, however, it seems more likely that this tablet is anomalous, and not necessarily representative of the sequence of the scribal curriculum as a whole. In his survey, Veldhuis also identifies a source for the composition The Song of the Hoe (ETCSL 5.5.4) as a Type II tablet (CBS 9856; Veldhuis 1997:66). A closer examination of this tablet though, reveals that this source is probably not a Type II tablet, but instead a Type I tablet whose obverse was later reused to copy an extract of "Proto Aa." The reverse contains columns three and four of The Song of the Hoe, and the obverse, which must have originally contained columns one and two of this composition now contains an extract from "Proto Aa" oriented at right angles to the text on the reverse. Thus, when the extract of "Proto Aa" is held as read, and the tablet is turned over in the normal way (around the horizontal axis), the reverse must then be rotated 90 degrees to the left for the columns on the reverse to be running, as they should, from top to bottom, indicating that this extract was probably added after columns one and two of The Song of the Hoe had already been erased. Type IV tablets with literary texts are less common, but are also attested. In addition to the four compositions in the Tetrad, which, as noted above, all occur on Type IV tablets, presumably because they were learned before the advanced phase of the scribal curriculum, there are isolated attestations of other literary compositions on Type IV sources. In Falkowitz's survey of Type IV tablets from Nippur (see above), the only identifiable literary compositions besides the texts in the Tetrad are "Nothing is precious" (Alster 2005:266-87), "An elegy on the death of Nannaya" (ETCSL 5.5.2), and the Emesal hymn Lisin A (unpublished; see Civil 1974-1977: 67). Examples from other sites include a Type IV source from Ur with lines from Enlil in the Ekur, written together with a Sumerian proverb (UET6/2, 371), and a Type IV tablet from Susa (MDP 18,49), with an excerpt from Gilgames and Huwawa (Version A; ETCSL 1.8.1.5). Among the Type IV tablets from the Scherbenloch at Uruk (see Cavigneaux 1996: 1-5), there are also several with unidentified compositions that may be literary texts (Cavigneaux 1996: 97-98; for a list of these sources, see Tinney 1999: 162 n. 21). One of these is a partly preserved extract from a royal hymn, which might be line 21 of Lipit-Estar A.'^ Some others contain phrases that are characteristic of Sumerian royal hymns (see, e.g., texts 227, 230, 231, and 233 in Cavigneaux 1996). With the exception of the texts from the Tetrad, however. Type IV tablets seem to have only infrequently been used for literary compositions. To determine the tablets types and to analyze and interpret the distribution of the sources a small but representative group of Sumerian literary compositions was selected. The texts that were chosen for this study are from a group often compositions known collectively as the Decad. This group comprises the following texts:

12. Text 207 in Cavigneaux 1996. The identification of this source with Lipit-Estar A was suggested by Tinney (1999:162 with n. 20), based on the occurrence of-me-en ("I am") at the end of one of the lines and the traces of the signs before it. But since lines ending with "-me-en" are also characteristic of Sumerian royal self-praise hymns in general, including a group of hymns to the ruler Sulgi comprising Sulgi A (ETCSL 2.4.2.01), Sulgi B (ETCSL 2.4.2.02), Sulgi C (ETCSL 2.4.2.03), and §ulgi E (ETCSL 2.4.2.05), the identification of this source with Lipit-Estar A is not certain. The traces of the signs before "-me-en" do not correspond directly to any of the lines in Lipit-Estar A and too much of the fragment is broken to provide sufficient evidence that this line is from this composition, and not from another composition with a similar line.

SUMERIAN EXTRAGT TABLETS AND SGRIBAL EDUGATION

59

1) SA: A praise poem of Sulgi (Sulgi A); ETGSL 2.4.2.01. 2) LiA: A praise poem of Lipit-Estar (Lipit-Estar A); ETGSL 2.5.5.1. 3) Al: The Song of the Hoe; ETGSL 5.5.4. 4) InB: The Exaltation of Inana (Inana B); ETGSL 4.07.2. 5) EnA: Enlil in the Ekur (Enlil A); ETGSL 4.05.1. 6) KH: The Kes Temple Hymn; ETGSL 4.80.2. 7) ErH: Enki s Journey to Nippur; ETGSL 1.1.4. 8) /£k Inana and Ebih; ETGSL 1.3.2. 9) Nu: A hymn to Nungal (Nungal A); ETGSL 4.28.1 10) GH: Gilgames and Huwawa (Version A); ETGSL 1.8.1.5. In a study of the sequence of the Old Babylonian scribal curriculum, Tinney identified the compositions in the Decad as the first ten texts that were learned at the beginning of the second or advanced phase of the scribal curriculum." The evidence that the Decad was learned at this stage includes the occurrence of the ten texts in sequence in two of the so-called "literary catalogues," the existence of sources with catchlines connecting some of the individual texts in the Decad in sequence, and collective tablets containing groups of these compositions (Tinney 1999:169-70). The prominent place the Decad occupied in the scribal curriculum is also indicated by the relatively large number of preserved sources for these compositions. Each of the texts in the Decad is preserved in an average of fifty to eighty duplicates, and at present a total of over 740 sources for all of these compositions have been identified. Because of the quantity of duplicates for these texts and their frequent use in Phase Two of the scribal curriculum, when literary compositions were copied, the Decad is well suited for a study of the distribution and pedagogical function of the sources for literary texts. The distribution of tablet types among the sources for the ten compositions in the Decad is summarized in Table 1.'" As can be seen from this table, the distribution of tablet types among the texts in the Decad is similar to the general distribution of tablet formats across the corpus of Sumerian literary texts as a whole. Type I and Type III sources are the most common, greatly outnumbering the other tablet types that are attested for these compositions. Out of a total of 634 sources (454 of which are from Nippur) for which the tablet type could be identified with certainty, 151 (103 of which are from Nippur), or approximately 24 percent of the total number of sources (and approximately 23 percent of the Nippur sources), are Type I tablets. Type III sources are even more frequently attested, accounting for 455 of 634, or approximately 72 percent of the total number of sources (336 out of 454, or approximately 74 percent, of these sources are from Nippur). By contrast, prisms constitute only approximately 4 percent of the total number of sources (approximately 3 percent of which are from Nippur); and less than 1 percent of the extant sources for the Decad are Type II or Type IV tablets. While it is clear that Type I and Type III tablets were the primary formats used for exercises involving literary texts, less is known about the pedagogical function of these tablet types than for Type II and Type IV tablets, which were used during "Phase 1." One obstacle is that the function of these sources is not as immediately deducible from the physical aspects of these types of tablets as it is for Type II and Type IV sources. Unlike the latter, which generally contain a model extract from a text written in a better hand, and the pupil's copy of this extract. Type I

13. Tinney 1999. Since the appearance of Tinney's study, the Decad as a curricular grouping has been cited and discussed in numerous publications. For a recent detailed treatment of this topic, which includes references to earlier studies and a more extensive bibliography of editions and discussions of each of the individual texts in the Decad, see Delnero 2006: 22-147. 14. All of the numerical data that pertain to the distribution and function of different tablet types cited in this table and throughout this study are drawn from the registry of sources compiled for each of the texts in the Decad in Delnero 2006: 1857-63 (SA), 1909-16 (LÍA), 1961-68 (Al), 2020-28 (InB), 2108-14 (EnA), 2172-80 (KH), 2238-42 (ErH), 2290-97 (IEb), 2359-63 (Nu), and 2395-402 (GH). Unless otherwise indicated, all the sigla used to refer to specific sources in this study are the same as the sigla in these lists, which also include the tablet numbers of each source, together with references to previously published hand copies and photos. For a description and discussion of the use and meaning of these sigla, and an explanation of the abbreviations used to compile them, see Delnero 2006: 176-77.

60

PAUL DELNERO

TABLE l. DISTRIBUTION OF TABLET TYPES AMONG THE SOURCES FOR THE DECAD. Type I

Type III

Prisms

Type II

Type IV

Type Uncertain

SA

13(7)

42 (27)

4(2)

0

0

8(6)

LiA

21 (19)

36(18)

3(1)

1(1)

0

8(7)

Al

24 (16)

43 (37)

2(1)

0

0

18(16)

InB

19(14)

59 (41)

4(2)

0

0

16(15)

EnA

10(6)

47 (37)

3(3)

0

1(0)

11(9)

KH

19(12)

52 (39)

3(1)

0

0

10(9)

ErH

9(5)

40 (32)

1(0)

0

0

5(5)

leb

14(9)

46 (33)

2(2)

0

0

12(8)

Nu

7(5)

37 (32)

1(1)

0

0

4(4)

GH

15(10)

53 (40)

2(1)

0

1(0)

14(8)

Total

151 (103)

455 (336)

25 (14)

1(1)

2(0)

106 (87)

NOTE: The number to the left of the parentheses refers to the total number of sources of this type, and the number in parentheses refers to the number of these sources that are from Nippur.

and Type III sources were produced entirely by the pupil, and do not include model texts. It is thus not possible, on the basis of the formal features of these types of sources alone, to identify whether the model text was written on another tablet, which was then copied directly by the pupil, or whether the pupil produced the copy by other means, such as copying from dictation or memory. In order to determine the function of scribal extract tablets, it is therefore necessary to consider other factors.

The Function of Type III Extract Tablets The function of Type III sources was recently studied by Robson, who examined the function of extract tablets for copying mathematical texts (Robson 2002: 338-44). In addition to literary compositions. Type III tablets were also used to copy multiplication and division tables, which were probably learned during the second half of "Phase 1" of the scribal curriculum. By dividing a series of tables into four sections, each of which consists of ten reciprocal pairs, and examining the distribution of these sections across sources, Robson observed pafterns indicative of the function of individual tablet types. The multiplication series Robson examined is attested on three different types of sources: Type I tablets (including prisms). Type II tablets, and Type III tablets. All four sections are attested in more or less equal numbers on Type III sources and the obverses of Type II sources, which both contain shorter extracts than Type I sources and the reverses of Type II sources (Robson 2002: 339-440). The distribution of the number of sections across Type I sources and the reverses of Type II sources, on the other hand, is significantly different. Instead of equal numbers of all four sections, there are substantially more Type I sources and reverses of Type II sources from the first quarter of the series than from the remaining three (Robson 2002: 342-43). To account for this correlation, Robson concluded that Type III sources were used along with the obverses of Type II sources in the first stages of learning so that pupils could memorize the entire table by copying it in short sections (Robson 2002: 344). Once the shorter sections had been memorized, pupils would review the


61

text by copying longer sections from memory on Type I tablets and the reverses of Type II tablets, starting with the beginning of the series, but rarely reaching the end (Robson 2002: 344). Though it seems plausible that Type I and Type III tablets had the same function for learning literary compositions that they had for copying mathematical exercises, this assumption has yet to be tested. Moreover, the distribution of extracts on these two types of sources does not reveal as much about their function for literary texts as it does for metrological texts. Since literary compositions are generally shorter than a series of multiplication or reciprocal tables, they can be copied in their entirety on a single Type I tablet, and do not provide the same evidence for whether sections of a text were distributed evenly across a series of tablets or cluster toward the beginning of the text. There are, however, other patterns that more clearly reflect the use and function of Type I and III sources for copying literary compositions. One is the correlation between the percentage of a given composition that the preserved Type III tablets contain, and the relation between the average length of the extracts and the length of the composition. Another is the correlation between the number of Type I and Type III sources for each text. Since Type III sources typically contain extracts from a single composition, one means of determining the function of these tablets is to identify patterns among the types of extracts they contain. Examining the distribution of passages found on the Type III sources for the compositions in the Decad, it would appear that the selection of lines to include was influenced more by length than by content. With the exception of KH, which consists of distinct verses that end with the same refrain and rubric, the Type III sources for these texts do not contain extracts that reflect coherent narrative or thematic units, and the lines with which they begin and end generally do not bear any discernible relation to the literary structure of the text. By contrast, when the lengths of these sources and the percentage of the composition they contain are considered, more definite patterns emerge. To illustrate this, the average percentage ofa composition the Type III sources for the Decad contain was calculated by dividing the average number of lines per source with the total number of lines in the composition. To obtain the average number of lines per Type III source the total number of lines for all of the preserved Type III sources was counted and divided by the total number of Type III sources preserved for the text. To determine these totals only the number of lines in the sources for which the flrst and last lines are preserved, or for which the number of lines that are missing could be reasonably estimated on the basis of the amount of the physical tablet that is missing in relation to the distribution of lines on the existing portions of the obverse and reverse were included in the calculation. Since the line totals could be estimated for all but a relatively small number of sources, the exclusion of these sources is unlikely to have signiflcantly altered the how representative the flgures obtained are with respect to the entire corpus. The results of these calculations are presented in Table 2. In considering the number of lines on the Type III sources for the texts in the Decad, the average length of each extract for the entire group is 24 percent, or approximately one quarter the length of the composition it contains. This percentage can also be observed for many of the individual texts. The average length of the sources for flve of the ten compositions in the Decad—SA, Al, InB, EnA, and Nu—is within one or two percentage points of 25 percent, and if the smallest and largest percentages (14 percent for IEb and 36 percent for ErH) are dismissed as exceptions," then the remaining three compositions {LiA, KH, and CH) fall within 6 percent of this percentage.

15. One explanation for the relatively large percentage obtained for ErH is that this composition is relatively short (125 lines) and contains a large number of short lines with non-finite verbal forms and simple syntactic constructions. Since this composition was learned in the second half of the sequence of the Decad, and many of the same types of constructions would have already have been encountered in the texts in the first half of the sequence, it may have been easier to copy longer extracts of this composition than it would have for the other nine texts. That this may have been tbe case is further suggested by tbe preservation of at least four extracts from Nippur which contain half or slightly more than half of the text (Nmj, which contains lines 1-69; Nmj, which contains lines 1-65; N,|,j, which contains lines 1-60; and N,[,,g, which contains lines 60-125—the second half of the composition). Furthermore, when the relative shortness of the lines in ErH is taken into account, the impression that the extracts for this composition are significantly longer than the extracts for the other four texts in the second half of the sequence may be deceptive. Nu has approximately the same number of lines as ErH, but tbe lines in the text are much longer and typically contain twice as many signs. While the lengths of the extracts for Nu have significantly fewer lines (an average of 27 percent of the entire composition) they may contain approximately the same amount of text as the extracts for ErH.

62

PAUL DELNERO

TABLE 2. NUMBER OF LINES IN TYPE III SOURCES FOR THE DBCAD. Average # of Lines per Type III Source

Number of Lines in Composition

Average % of Text per Type III Source

§A

24

99

24%

LiA

33

108

31%

Al

26

109

24%

InB

35

154

23%

EnA

40

171

23%

KH

25

131-(-

19%

ErH

45

125

36%

IEb

25

181

14%

Nu

33

121

27%

GH

36

186-1-

19%

Averages

33

139

24 %

NOTE: The percentages are rounded to the nearest whole number.

If it is assumed that Type III sources were used in the process of learning shorter sections of a longer literary text, as they were for mathematical tables, it appears that in the initial stages of learning, there was a tendency to divide the numbers of lines to be learned into sections representing approximately one fourth of the composition being studied. Furthermore, if the distribution of lines that occur on the preserved Type III sources for the Decad is examined, the number of extant extract tablets is generally more or less equal for each section of the composition. For example, for the composition Al, there are five sources from House F at Nippur that contain lines that are all or almost all from the first quarter of the text (N,,,,,, N,,,2o + N,,m, N,,^^, N,,,23, and N,,,2,), as well as five sources with lines from the second quarter (N,,,25, N,,,^^, N,„„, N,,,^^, and N,,,^,), four sources with the third quarter (N,„3„, NiiBi. N,,,32, and N,,,33), and four sources with the final quarter (N,^^^^, N,¡¡¡¡, N„,3„ and N,„3,). Since all of these sources are from the same house, and there is nearly the same number of copies of each section, this would provide further evidence that extract tablets were used for copying entire texts with a series of extract tablets containing shorter sections. That compositions were copied on groups of Type III sources is also illustrated by extract tablets from Ur which begin and end with sequential lines and form a complete series containing an entire text, like the Type III sources UET 6/33,34, and 35, which contain lines 1-67,67-127, and 127-192 respectively of the 193-line composition The Debate between Grain and Sheep.'^

The reason for the shorter number of extracts for IEb is less certain, but may have been influenced by the presence of a twenty line section (in which Inana describes her plans to destroy Ebih), which is duplicated in 1.32-51 and 1.91-110. To avoid repetition, there may have been a tendency to copy the lines in this passage, which comprises approximately 11 percent of the text, only once, and to copy the remaining sections of the text before and after the first occurrence of these lines in shorter sections. Partial confirmation for this may come from the presence of significantly more extracts from Nippur with lines from the first occurrence of this passage (8) versus lines from the second occurrence (2). Furthermore, the exceptional series of five extract tablets from Ur (discussed in more detail below) with an average of only 30 lines (or approximately 17 percent of the text), may be contributing to the lower than average percentage of lines for the sources for this composition. 16. ETCSL 5.3.2. The line numbers given here follow the numbering of Alster and Vanstiphout (1987), who label these sources Up Uj, and U3 in their edition of this composition.


63

Assuming then, that there was a tendency to copy complete compositions on a series of approximately four extract tablets—as the figures in the Table 2 suggest—the approximate number of extant groups of Type III sources for each of the ten compositions in the Decad text can be calculated by dividing the total number of preserved Type III sources for each composition by four. This is of course not meant to imply that each group of four sources was written by the same scribe, or that every group of Type III sources for a composition consisted of exactly four tablets, but only to indicate a statistical probability based on a more general pattern. That the division of extracts on Type III tablets into fourths was not a rule, but a general tendency is suggested by the existence of extracts for all of the texts in the Decad that are either significantly shorter or longer than 25 percent of the entire composition. Extracts that contain as little as a sixth, or as much as half, of the text are attested for all ten compositions. The series of extracts for the composition The Debate between Grain and Sheep cited above, for example, show that in this instance a complete text was copied on a sequential series of three Type III sources; and a partial series of Type III sources for IEb with extracts of 25-35 lines (Ur^ Ur^, Ur;, and Ur^), two of which (Ur, and Urj) begin and end with the same line, shows that literary compositions could be copied on a series of as many as six extract tablets. When the total number of Type III sources for each composition in the Decad is divided by four to obtain the average number of preserved groups of sources containing the entire text, there appears to be a correlation between the number obtained and the number of Type I sources for each composition. These correlations are presented in the following table:

TABLE 3. NUMBER OF GROUPS OF TYPE III SOURCES VS. NUMBER OF TYPE I SOURCES. Total Number of Type I Sources

Total Number of Type III Sources

Number of Groups of Sources with Entire Text (Estimated)

SA

13(7)

42 (27)

11(7)

LiA

21 (19)

36(18)

9(5)

Al

24 (16)

43 (37)

11(9)

InB

19(14)

59 (41)

15(10)

EnA

10(6)

47 (37)

12(9)

KH

19(12)

52 (39)

13(10)

ErH

9(5)

40 (32)

10(8)

Nu

14(9)

46 (33)

12(8)

IEb

7(5)

37 (32)

9(8)

GH

15(10)

53 (40)

13(10)

Total

151 (103)

455 (336)

115(84)

NOTE: The number to the left of the parentheses refers to the total number of sources of this type, and the number in parentheses refers to the number of these sources that are from Nippur.

For many of the compositions in the Decad, the number of groups of four Type III sources corresponds directly to the number of Type I sources for this text. For the Nippur sources for SA and GH this number is identical; for six of the eight remaining texts (InB, EnA, KH, ErH, IEb, and Nu), it is nearly identical. Assuming that the same scribes who wrote the groups of Type III sources also wrote the Type I sources containing the same texts, this suggests that for each completed group of Type III tablets, scribes would produce one Type I tablet. This pattern is consistent

64

PAUL DELNERO

with Robson's assessment of the function of Type I and III sources for learning multiplication and division tables. In the initial stages of learning, the scribes practiced the entire text in short sections. After completing this process, they would then review the complete composition on a Type I tablet.

Two Type III Extract Tablets The initial stage in the process of learning literary texts is exemplified by two rather unique Type III sources for the composition LiA. Both of these tablets—NBG 1311 (Source X3) and NBG 1312 (Source XJ—are housed in the Yale Babylonian Gollection of Yale University in New Haven, and were published as hand-copies by Albert Glay in the second volume of the series Babylonian Inscriptions in the Gollection of James B. Nies in 1920. These tablets were purchased during the first decade of the twentieth century around the same time similar groups of Old Babylonian literary tablets now in the Louvre in Paris and the Hermitage in Saint Petersburg were acquired. While it is frequently assumed that many of these tablets are from Larsa, in the absence of more direct evidence, the provenience of the two sources remains unknown. What is significant about Sources X3 and X^ is that they both contain identical extracts (lines 1-31 oí LiA) with the same distribution of lines on the obverse and reverse (obv. lines 1-20; rev. lines 21-31), and have colophons that indicate they were copied on the same day by two different scribes. These colophons read: (Golophon X3) im-gidj-da il¡-su-i-di-su iti se-KIN-kUj u^-11-kam "Extract tablet" (imgida) of Ilsu-Iddissu. Month 12, day 11 (Golophon X4) im-gidj-da i-din-es^-tar^ iti se-KIN-ku; U4-ll-kam "Extract tablet" (imgida) of Iddin-Estar. Month 12, day 11 According to the colophons. Source X3 was copied on the eleventh day of se-KIN-kU; (Month 12) by a scribe identified as Ilsu-iddissu and Source X^ was copied on the same day and month by a scribe named Iddin-Estar. These colophons, together with the identical distribution of lines on the obverses and reverses of both tablets, provide direct evidence that these two scribes were being trained at the same time, and were given identical assignments. Furthermore, by analysing the variant writings that occur in the two sources, both with respect to the other preserved sources for these lines and with respect to one another, it can be demonstrated that these extracts were copied either from dictation, or more probably by memory, at an early stage of learning LiA. For the composition LiA a total of 71 sources have been identified. Over two-thirds of these sources (46) are from Nippur, but duplicates from Babylon, Isin, Kish, Uruk, and Ur, as well as other duplicates from unknown proveniences have also been discovered. For the lines copied in Sources X3 and X4 there are an average of fifteen to eighteen duplicates per line. When the contents of X3 and X^ are compared with how the lines are copied in the majority of sources, there is a substantial amount of overlap across the sources.'^ The preserved portions of the first ten lines in X3 and X^, which are broken in a few places, are identical to the majority of sources, as are all but nine of the remaining lines. There are, however, a significant number of instances in which one or both of these sources contain writings that differ from those in other sources. These instances are listed in Table 4. At least three of the variants that occur in Sources X3 and X^ are only attested in these two sources. Furthermore, the omission of the determinative na4 with za-gin3 in line 11 in X3 and X4 only occurs in one other source (N„),

17. The data pertaining to textual variation in this section, including the versions of the lines selected to represent the composite text of the lines cited below, were drawn from the score of LÍA in Delnero 2006: 1909-60. For a description of how these scores were compiled and a discussion of the criteria that were used to distinguish textual variants from "correct" writings in producing the composite text for this composition, see Delnero 2006: 148-82, esp. 175-76.

SUMERIAN EXTRACT TABLETS AND SCRIBAL EDUCATION TABLE 4. VARIANTS IN SOURCES NBC 1311 (X3) AND NBC 1312

65 (X4).

11: sUg "''•'za-ginj e3-a girj KESj.KESj-sa me-en X3 su, "za-gin, e,-âgis'(-)kes,? [...] X4 su, "za [...] a [...] a [...] me-en 13: uktim hus-hus-a hi-li dug-dug-a me-en Xj uktim(SIG7.ALAN) hus-hus-a hi-li du^-duj-O me-en X, uktim(SIG7.ALAN) hus-hus-a [...] du^-O me-en 15: aj il2-la su-si sa^-sa^ me-en X3 a^ ilj-la su-si sa,-â me-en X4 aj ilj-la su [...] sa^-â me-en 19: sag-kal kur-kur-ra kalam-ma il^-la me-en X3 sag-kal kur-kur-ra kalam-ma i\2-l and identifies this person with Nishur-Bel (4.3 no. 8).

5. Analysis First of all it should be noted that the evidence on which the conclusions of this paper are based is necessarily circumstantial. The reason for this is that, as already mentioned above, both logograms NIGIN and NIGIN^ may refer to both Akkadian verbs pahäru, "to gather, bring together" and sahäru, "to go around, turn, search." The oldest logographic writings oipahäru and sahäru as part of personal names date from the Middle Babylonian period. NIGIN is most frequently used, while NIGIN^ only occurs in four names. Both logograms occur in initial and in final position.

22:12, 26: rev. 11; TB 15:4. 106. SAA 14 317: rev. 5. 107. BIN 7 59:19 and its duplicate 60:22. 108. TR 2018:7. 109. H. W. F. Saggs, "The Tell Al Rimah Tablets, 1965," Iraq 30 (1968) 158. 110. J. N. Postgate, review of C. Saporetti, Onomástica medio-assira, OAR 13 (1974) 70. 111. KAJ 6:33. See C. Saporetti, Onomástica medio-assira I, 347-48. 112. E R. Kraus, Briefe aus dem Istanbuler Museum, AbB 5 (Leiden: Brill, 1972), 114, no. 220:20. 113. E. Grant, Cuneiform Documents in the Smith College Library, Biblical and Kindred Studies 1 (Haverford, PA: Haverford College, 1918), no. 263:8; YOS 13 252:6. 114. B£ 9 9:16,/MT 66:21 and 67:15. 115. ABL 281: rev. 19 (spelled "Nis-hur-'ÊN), 791: rev. 1,11 (spelled •"Ni-is-hur-''EN). 116. B£ 10 35:20. 117. ABI 1342:23. 118. L. Waterman, Royal Correspondence, IV 139. Collation by G. Frame. 119. J. M. C. T. De Vaan, Ich bin eine Schwertklinge des Königs: Die Sprache des Bél-ibni, AOAT 242 (Kevelaer: Butzon und Bercker, 1995), 321. The same reading is found in PNA 965.

JAN TAVERNIER

90

TABLE I . SYLLABIC READINGS OF PAHARU AND

SA.HAfiu ATTESTED IN PERSONAL NAMES.

(1) paharu Initial

Middle / Final

iphur

liphur

liphur

mupahhir

Old Babylonian mupahhir Middle Assyrian upahhir liphur

liphur

puhrat (only once)

mupahhir

Neo-Assyrian pahhir upahhir pahhir (only once) Late Babylonian upahhir

(2) saharu

Old Babylonian

Initial

Middle / Final

isahhur

issahram

nashiram nashur issahra

Middle Babylonian nashir

nashira

Middle Assyrian nashira Neo-Assyrian

nashir

nashira

nishur Late Babylonian nissahur (?)

Because of the occurrence of Nishur-Bêl (4.3 no. 8) in Neo-Assyrian sources one could argue that nishur is also attested in Neo-Assyrian names, but as Nishur-Bël is a Babylonian individual,'^" this idea should be rejected. When looking at this table a first conclusion becomes apparent. The examples oí paharu- and sahàru-names show that the element paharu mostly occurs in the final position of a name, while saharu is more frequently attested in initial position. There are also more names with syllabic spellings oí saharu in the beginning of the name than names with syllabic spellings oí paharu in the beginning of the name.

120. PNA 965.

AKKADIAN PERSONAL NAMES WITH PAHÄRU OR SAHÄRU AS INITIAL ELEMENT

91

Some of the elements remained in use for a long time: liphur (Old Babylonian to Neo-Assyrian; both initial and final), mupahhir (Old Babylonian to Neo-Assyrian; middle / final), nashir(am) (Old Babylonian to Neo-Assyrian; both initial and final), upahhir (Middle Assyrian to Late Babylonian; only in final position). Some are explicitly limited to one period: iphur (Old Babylonian; initial), issahur (Old Babylonian; initial), nashur (Old Babylonian; initial), nishur (Late Babylonian; initial) andpuhrat (Neo-Assyrian; initial). Old Babylonian names show a wide variety of forms of both verbs. The Neo-Assyrian names beginning with NIGIN or NIGIN2 have syllabic parallels, which indicates a reading using the element liphur, nashir (or nashur), and puhrat. As mentioned before, there is no full agreement among scholars, but nashir seems to be the preferred reading for NIGIN or NIGIN^. Late Babylonian names were apparently standardized, and only have the syllabic parallel ni-is-sa-hur I ni-ishur for elements in initial position. Nowadays most scholars believe that initial NIGIN / NIGINj should be read upahhir, partly because of the fact that final NIGIN / NIGINj is mostly read that way. Only two authors questioned this reading. Glay'^' realized that NIGIN / NIGINj may represent both pahäru and sahäru, and that both verbs occur in anthroponyms (his examples are Enlil-upahhir and Nis(sa)har-Enlil).'^2 He therefore concluded that it could not be determined whether NIGIN should be read nishur or upahhir. Krückmann'" went further by preferring a reading Nisshhar-Enlil for '"NIGIN-'*EN-LÍL (no. 2.2.4) and argued that NIGIN in the beginning of a name stands for nissahar, while at the end of a name it refers to upahhir. He compared this to KAR, which in the beginning of a name renders musëzib, while at the end of a name it stands for ëtir. The remarks of these two scholars were quickly forgotten. Nevertheless Krückmann seems to have been right. It has been made clear that not a single syllabic spelling of upahhir as the first part of an Akkadian anthroponym is attested in the aforementioned list of anthroponyms, while various examples occur of upahhir, spelled syllabically as the second part of a name (e.g., Bël-upahhir). This does not support a reading upahhir for NIGIN as the first part of a name. The second argument has already been mentioned, namely, that sahäru is attested more frequently in the beginning of a name, while pahäru in initial position is rare. A third argument against a reading upahhir is of a prosopographical nature. The names listed above contain three names occurring in the Murasû Archive: NIGIN-mas-tuk "'la-ba-si dumu lti.sid-''mas u lú.dub.sar ""•ag-sigj-iq a lti.su.ha kis^' iti.sig^ 'ûd''.7 .kam 15. mu.l2. ''gis.nUn-mu-gi.na '"'5 minas and 8 shekels of silver is the price of 154 sheep and goats belonging to Ninurta-usabsi under the responsibility of Bél-êtir. No other has the right of disposal over the sheep and goats until Ninurta-usabsi is paid his silver. ^'=The witnesses are: Etêru descendant of Dâbibi, Ina-têsî-étir descendant of Pahhàru (Potter), Suma-iddin descendant of Sîn-sadûnu, Bël-eriba descendant of Mastuku, Lâbâsi descendant of Sangû-Ninurta, and the scribe Nabû-mudammiq descendant of Ba'iru (Eisher). Kish. Simànu (III). Day 7, year 12, of Samas-suma-ukin.

Commentary 3. Ninurta-uáabsi may appear as witness with the family name Misiràya in BM 78167:18.

20. Fifty-five years separate Merodach-Baladans accession year (722) from 667, the year Ubäru sold a hanSú meadow to Bël-ëtir. Eliminating the possibility that the zu'uztu occurred during Merodach-Baladans second stint on the throne, the transferal of the land from Nabû-bêlsumäti and Nummuru to Ubaru could have occurred within a window of time from 55 to 43 years in length. 21. A. C. V. M. Bongenaar, The Neo-Babylonian Ebabbar Temple at Sippar: Its Administration and Its Prosopography, PIHANS 80 (Leiden: Nederlands Historisch-Archaeologisch Instituut te Istanbul, 1997), 34.

THREE EARLY NEO-BABYLONIAN TABLETS BELONGING TO BÈL-ÊTIR

103

5. Dr. Jonathan Taylor has collated the tablet and has confirmed that the sign is w. Negation is expected here, and it is assumed that the scribe mistakenly wrote ù instead of ul. 11. Suma-iddin appears as a witness in BM 78167:21 and BM 46799+:37. 12. Nabû-mudammiq appears as the scribe in BM 46799+:41. There are two good reasons to suppose that the man named Bël-ëtir in this tablet is identical to the man who purchased hansû land in the two tablets discussed above: first, this tablet and BM 78167 were both accessioned on the same day as part of a group of tablets acquired by the British Museum, suggesting that the two tablets may have been found together before making their way through the antiquities market to London; and second, there are men in the witness list of BM 78159 who appear in one or both of the other tablets. The fact that BM 78159 was dated in the same year as BM 46799+ and that all three tablets were dated at Kish and not Hursagkalama lends further support to the supposition that Bël-ëtir of the Misiràya kin group is featured in all three tablets.

BM 78159

According to BM 78159, 154 sheep and goats valued at 308 shekels of silver were placed at Bël-ëtir s disposal on III-7-656." The flock belonged to a man named Ninurta-uJabsi, and the agreement stated that no other man would have control of the flock until Ninurta-usabsi received his silver. As a private lease of animals, this document does not follow the model in which an investor leases animals with the agreement that he will then earn a share of the profits derived from using the animals." The tablet is formatted as a promissory note, and as such it would make sense if Bël-ëtir intended to profit from his management of the flock before returning it to Ninurta-usabsL However, the desired outcome stipulated in the contract is that Bël-ëtir should eventually pay the established value of the flock in silver to Ninurta-usabsi. We know that Bël-ëtir had sufficient funds to purchase two hansù fields seven months later but the cost of all the land that he purchased was significantly less than the value placed on Ninurta-usabsi's flock. It is possible that Bël-ëtir s control of the flock was part of his strategy to derive profit from the land. All three of the hansûs in Bëlr's possession are described as being located in tamirtus, which in some cases could be suitable for pasturing

22. Each animal cost 2 shekels of silver. 23. M. Jursa, Neo-Babylonian Legal and Administrative Documents, Typology, Contents and Archives, Guides to the Mesopotamian Textual Record 1 (Munster; Ugarit-Verlag, 2005), 39-40. 58.

104

JOHN R NIELSEN

a ffock." Had Bêl-êtir succeeded in his strategy and paid the 308 shekels to Ninurta-usabsi, the ffock would have become his property. If BM 78159 were part of Bêl-êtir's private archive, then the tablet's presence in the archive along with his titles to hansû land would be an indication that he was successful. The arrangement in BM 78159 may gain additional clarity if we consider the kin-group affiliations of Bél-êtir and Ninurta-usabsi, neither of who is identified by a family name in the text. A man named Ninurta-usabsi descendant of Misiràya was the first witness to Bêl-êtir's earlier purchase of hansû land in 667.^' If this is the same man as the owner of the ffock, then Bêl-êtir and Ninurta-usabsi were both members of the Misiràya kin group." One could propose a scenario in which Bêl-êtir was unable to buy the ffock outright but entered into a contractual arrangement with a fellow kin-group member that, if successful, would permit him to acquire the ffock. This reconstruction admittedly is speculative, but it highlights how underlying relationships based on presumed shared kin-group affiliation had the potential to shape transactions in ways that are not spelled out explicitly in the text of the tablet.

4. Conclusions Ultimately, the contents of these three tablets only accentuate what we do not know about Bêl-êtir's career, which at the very least spanned the years from 667 to 656. Absent are the promissory notes and contracts that could provide us with information about how Bêl-êtir managed the wealth he generated from his acquisitions and control of land and animals. Also absent are the family documents, administrative memoranda, and letters that characterize a private archive preserved in its complete form." What remains are documents that prove ownership of property (or at least usufruct in the case of the herd) and therefore possessed value. The outbreak of the Sama§-suma-ukîn revolt in 652, four years affer Bêl-êtir's last recorded act, may account for the eventual fate of these tablets and their owner in antiquity. Kish was only fifteen kilometers east of Babylon, and the presence of Ashurbanipal's army as it besieged Babylon from 650 to 648 certainly impacted Kish. It is unfortunate that the tablets were not scientifically excavated; their findspots could have provided useful clues about Bêlêtir's actions during the war. He may have kept these tablets separate from the less valuable tablets in his archive as circumstances became increasingly unstable. He may have hidden them, failing to retrieve them at a later date, or he may have opted to sell his possessions, in which case the tablets became retroacts in other men's archives. The fact that BM 46799+ became part of the Rassam collection, most of which come from Babylon and Dilbat, even raises the possibility that Bêl-êtir, or another holder of that tablet, made his way to Babylon.^' Whatever path Bêlêtir followed, he does not appear to have been successful. No tablets recording his activities date from after the war; and if he did survive, he may have lacked the capacity to continue his pre-war ventures.

24. CAD T, iamirtw, mng. 1 usg. 4' b lists usages of tamirtu with reference to grazing. 25. BM 78167:18. 26. Ninurta-usabSi would not be the only witness from BM 78167 to be named in BM 78159; Suma-iddin descendant of Sin-aadûnu acted as a witness in all three of the Bêl-êtir tablets (BM 78167:21, BM 78159:11, and BM 46799-1-46928:37). 27. M. Jursa, Neo-Babylonian Legal and Administrative Documents, 58. 28. The siege did prompt sales within Babylon of land located both at that city and elsewhere, including two sales of land located at Kish dated in 650 and 649 Babylon was besieged by the Assyrians by the first half of 650 and remained under siege until the city fell in the fifth month of 648. During this time, individuals sold land and other assets in order to pay for food. Grant Frame, Babylonia 689-627 B.C.: A Political History, PIHANS 69 (Leiden: Nederlands Historisch-Archaeologisch Instituut, 1992), 150-55. The sales tablets dated at Babylon are OECT 10 6 (dated XI-24-650) and 7 (dated XII-12-649).

SORCELLERIE HITTITE Alice Mouton (CNRS Strasbourg)

Dans le cadre de ma réédition du rituel d'Alli pour le « Digitale Publikation von Texten der Hethiter »' auquel je collabore modestement, je me suis intéressée à la sorcellerie. J'entends par « sorcellerie » les pratiques de magie nuisible (« harmful magic » dans les publications anglophones), celles qui sont utilisées pour faire du mal à autrui. C'est ce que les textes hittites désignent par le terme alwanzatar dont l'étymologie est, d'après la plupart des indoeuropéanistes, incertaine^. Certains africanistes anglophones font, depuis Edward E. Evans-Pritchard, la distinction entre « witchcraft »et « sorcery », le premier terme étant relatif à une qualité physique héréditaire que possèdent de rares individus, le second désignant l'usage de techniques magiques sans qu'une quelconque spécificité physique soit requise'. Cependant, le corpus hittite ne permet pas de faire cette distinction - qui n'existe d'ailleurs peut-être pas - , car le profil du sorcier reste presque totalement inconnu de nous. Nous ne pouvons, en effet, pas affirmer que les Hittites croyaient en l'existence de personnages possédant des dons héréditaires qui les prédisposaient à la sorcellerie. C'est pourquoi je ne ferai pas ici la distinction entre « witchcraft » et « sorcery » et j'utiliserai « sorcellerie » comme terme générique pour désigner la magie malveillante en général. Plusieurs travaux importants ont déjà étudié la question de la sorcellerie hittite. Citons notamment le travail de Maurice Vieyra (1966), d'Ahmet Ünal (1988, et plus particulièrement 64-65 pour la sorcellerie) ainsi que les différentes contributions de Volkert Haas (1977:173-75; 1994: 884-91), de Marie-Claude Trémouille (2004:162-64) et de Daniel Schwemer (2007a: 255-76). Lors de mon séjour à Mayence en Novembre 2007, j'ai eu l'occasion d'établir la liste de toutes les occurrences connues du terme alwanzatar, ainsi que d'autres expressions clairement associées à la pratique de la sorcellerie: « langue mauvaise », « yeux mauvais », « personne mauvaise », « malédiction », etc. Il en ressort un corpus très étendu, que je n'ai pas eu le temps d'étudier en entier. Par conséquent, le présent article ne prétend pas être exhaustif, il s'agit au contraire de sélectionner quelquesuns des témoignages qui me semblent les plus intéressants pour répondre à des questions précises. Ces questions sont: 1) Qui pratique la sorcellerie?; 2) Quelles sont les différentes manifestations de la sorcellerie, sous quelles formes apparait-elle ? ; 3) Quelles sont les techniques d'ensorcellement attestées par les textes; 4) Quels sont les effets

Les abréviations employées ici sont celles se trouvant dans; The Hittite Dictionary of the Oriental Institute of the University of Chicago (CHD), éd. H. G. Güterbock et H. A. Hoffner Jr. (Chicago, 1989-) et dans The Assyrian Dictionary of the Oriental Institute of the University of Chicago (CAD), éd. A. L. Oppenheim et al. (Chicago, 1964-). À ces abréviations, il faut ajouter Konkordanz = S. KoSak, Konkordanz der hethitischen Texte (online; http;//www.hethport.uni-wuerzburg.de/hetkonk/). 1. http;//www.hethiter.uni-mainz.de/ 2. HEDI, 46-47 et HEG I, 20-21. Voir toutefois la théorie d'E. Rieken et I. Yakubovich mentionnée à la fin de cet article. 3. Marwick, éd. (1970; 12-13). Abusch (1999; 112-13) utilise cette distinction pour le corpus mésopotamien; « I simply note that occasionally I find it useful to adopt the anthropological distinction between 'sorcery' and 'witchcraft' according to which 'sorcery' refers to the aggressive use of magical techniques for pernicious private ends, whereas 'witchcraft' refers to behavior of or acts by a human being believed to be endowed with a special mystical power and propensity for evil that results in harm to others. » Notons que cette distinction a tendance à être abandonnée dans les études récentes d'anthropologie, car elle est loin d'être universelle; voir, par exemple, Stewart et Strathern 2004; 1-2 et Favret-Saada 2004; 672. 105

ICS 62 (2010)

106

ALIGE MOUTON

causés par l'ensorcellement?; 5) Gomment neutraliser l'ensorcellement ? Je ne suis, bien entendu, pas la première à me poser ces questions, mais plusieurs aspects me semblent suffisamment intéressants pour être réétudiés. Notons que la sorcellerie a été beaucoup plus amplement étudiée dans le domaine mésopotamien, depuis les travaux de Gerhard Meier en 1937 jusqu'à ceux de Daniel Schwemer en 2007, en passant par ceux d'Erica Reiner, de MarieLouise Thomsen et de Tzvi Abusch. Je pense qu'il serait nécessaire de se pencher également plus en détails sur les pratiques hittites de sorcellerie, qui peuvent encore nous révéler bien des informations.

1. Les sources A. Le rituel de Mastigga de Kizzuwatna contre des querelles dans une maisonnée (CTH 404.1)'': ce texte a été réédité et étudié en détail par Jared Miller dans son ouvrage de 2004. Bien que l'incipit de cette composition ne le signale pas, nous avons bien affaire à un cas de sorcellerie. En effet, les « langues » sont mentionnées à plusieurs reprises dans le texte. Gette expression est une forme abrégée de « langues mauvaises », qui désigne probablement un ou des propos diffamant(s) provoquant un effet analogue à celui produit par une malédiction. Le terme hittite hurtai- « malédiction » est également utilisé en association avec la langue. Le fait que des querelles familiales soient le symptôme d'un ensorcellement est, d'ailleurs, également connu dans le monde de la Grèce ancienne (Bernand 1991: 328 et Graf 1994: 177-78). B. Un rituel contre l'ensorcellement du roi Tudhaliya (KUB 24.12+: CrH448.4.9): il s'agit d'un rituel d'exorcisme destiné à libérer le roi d'un ensorcellement pratiqué sur sa personne. La présence de la divinité hattie Zalipura tendrait à indiquer que ce rituel relève originellement de cette aire culturelle. Le texte a été édité par Piotr Taracha en 2000 (2000: 86-95). G. Le rituel contre Tensorcellement pratiqué par Ziplantawiya (CTH 443): ce rituel est destiné à pacifier la divinité solaire du sang et le dieu de l'orage après qu'une malédiction a été proférée par Ziplantawiya contre son frère Tudhaliya et sa famille, du moins du point de vue du roi^ Gabriela Szabó est responsable de Veditio princeps de 1971 et Alexei Kassian nous a offert une réédition en 2000. D. Le rituel contre l'ensorcellement de la Vieille Femme Hebattarakki de la ville de Zuhâruwa^ (CTH 397): plusieurs passages de ce rituel ont été édités par David Engelhard dans sa thèse de doctorat sur la magie hittite. Ge rituel témoigne de l'emploi de techniques de contre-magie tout à fait particulières. Le duplicat KBo 54.13 indique probablement que ce rituel provient de la région de Mukis.' E. Le rituel d'Alli d'Arzawa contre l'ensorcellement (CTH 402): ce rituel a été édité en 1972 par Liane JakobRost, mais plusieurs modifications peuvent être faites suite à la découverte de nouveaux duplicats et joints. Je viens de terminer une réédition de ce texte, réédition qui doit être mise en ligne sur le site internet de Mayence. Voici la liste des manuscrits après les modifications que je suggère: A: KBo 12.126 + KUB 24.9 + KBo 12.127 + }CS 24:37 B:

KUB 24.10

4. Pour la dimension politique éventuelle de cette composition, voir Hutter 1991: 38-39. 5. Concernant la possibilité pour que cet ensorcellement, comme beaucoup d'autres, n'ait jamais eu lieu, voir la discussion autour de l'article à paraitre de J. Miller, dans le paragraphe 2. 6. Lecture assurée par le duplicat KBo 54.13. 7. KBo 54.13 i 1: [... "™"sguG]i ^^^zuharuwa SA KUR "^"mu[kií]. Voir Schwemer (2007a: 267). Contra Hutter (2003: 252) qui le faisait venir du Kizzuwatna.

SQRCELLERIE HITTITE

107

C: KBo 57.28 (+) 911/u -i- KUB 24.11 -i- 987/v -h 1015/z (anciennement duplicat M)(-h?) 861/z (anciennement duplicat O) (-)-?) KBo 46.6 (anciennement duplicat K) (joints A. Mouton) D: ÍC50 11.12+ /Cßo 10.43 E: KBo 10.41 F: /J3or2.123 (+?) Bo 3582 (anciennement duplicat L) (joint A. Mouton) G: Bo 8752-I-KLÍB 41.2 (joint J.Miller) H: KUB 41.1 h KBo 21.8 J: Kßo 45.190 K: 452/u (anciennement appelé duplicat N) Le rituel d'Alli combine prescriptions rituelles et incantations. Parmi les gestes rituels marquants, on remarquera la manipulation de cinq figurines d'argile: detix masculines et trois féminines. Faut-il voir, dans la supériorité numérique de la femme sur l'homme, une association plus forte entre la sorcellerie et la femme ? Ce ne serait, en effet, pas la première fois que l'on remarque, dans une société traditionnelle, la suprématie de la femme sur l'homme dans le domaine de la sorcellerie^ À côté de ces rituels « spécialisés » dans les questions d'ensorcellement, de nombreux autres ont une fonction plus généraliste. Je pense plus particulièrement aux rituels de purification qui font souvent allusion au phénomène de l'ensorcellement comme source de souillure. Citons, à titre d'exemple, le célèbre rituel du neuve de Tunnawi(ya) édité dès 1938 par Albrecht Goetze. Lorsque l'origine d'une souillure n'est pas connue, il paraît raisonnable d'envisager, entre autres possibilités, un acte de sorcellerie. C'est ce que nous indique ce rituel qui fait tout d'abord allusion à « ceux » (au pluriel dans le texte: kuiës) qui pourraient être auteurs de l'impureté du patient, cette expression pouvant désigner aussi bien les divinités que les hommes. Puis on mentionne plus précisément des « mauvaises personnes », des personnes mal intentionnées qui auraient pu avoir recours à la sorcellerie (§ 15; Goetze 1938: 12-13). Le terme alwanzatar « ensorcellement » n'apparaît toutefois qu'un peu plus loin dans le texte (§ 19). Une autre composition bien connue des hittitologues, et qui fait elle aussi allusion au phénomène de l'ensorcellement, et plus précisément de la malédiction, est le rituel de purification d'une maison invoquant les divinités souterraines, CTH 446. Heinrich Otten a édité ce texte en 1961'. La malédiction y est mentionnée aux côtés d'autres maux, tels que le crime de sang, l'impureté et le parjure. Un troisième exemple de rituel « généraliste » faisant allusion au phénomène de la sorcellerie est le rituel d'Ammihatna de Kizzuwatna destiné à purifier une personne qui a été contaminée par une souillure (CTH 471). Lédition de Rita Strauss de 2006: 216-52 nous permet d'avoir l'ensemble du texte sous les yetix. Or, parmi les sources d'impureté évoquées, on remarquera des aliments ensorcelés (§ 2: NINDA-a« "^"I alwanzahhan). Parmi les autres rituels « généralistes », nous aurions également pu citer les rituels taknaz da-, étudiés, notamment, par Manfred Hutter (1988) and Piotr Taracha (1990, avec bibliographie). Mais pour revenir très rapidement sur la notion de « mauvaise personne », je voudrais souligner la présence, dans le rituel VS NF 12.26 appartenant au corpus de Lallupiya et Istanuwa sur lequel nous travaillons à l'heure actuelle Ian Rutherford et moi-même, de l'expression LÛ'^'^ârsanatallas LU""^^ harpana[ll] iés « les hommes envieux (et) les hommes hostiles » (Vo 22') dans un contexte lacunaire mentionnant une personne ensorcelée (alwanzahhanza: Vo 31'): La notion de jalousie, en particulier, est tout à fait caractéristique dans un contexte de sorcellerie. On pensera, par exemple, au phénomène du mauvais œil tel que le folklore méditerranéen le décrit.

8. Voir également les réflexions de Sefati et Klein (2002) et de Westbrook (2006: 50). L'africaniste Roger Brand a écrit à ce propos: « La femme est un lieu de rencontre de la vie: elle façonne ce que l'homme a déposé en elle. A ce titre, elle est soupçonnée de posséder un pouvoir d'agression et de sorcellerie. » (Brand 1986: 43). 9. Otten 1961. Voir également Collins 2003.

108

ALICE MOUTON

2. Qui pratique la sorcellerie? Les témoignages issus des Lois Hittites et de l'édit de Telepinu'" semblent indiquer que tout un chacun peut pratiquer la sorcellerie. Le fait que l'on puisse, par exemple, ensorceler quelqu'un simplement en tuant un serpent et en prononçant son nom en même temps (Hoffner 1997:136 § 170) indiquerait que certaines pratiques de sorcellerie sont accessibles à tous. Ce même passage des Lois hittites envisage en outre deux cas de figure: soit la personne qui a pratiqué cet acte de sorcellerie est une personne libre, soit elle est un esclave. Ainsi, comme le fait André Bernand au sujet du sorcier de la Grèce antique, nous pourrions dire que « le sorcier n'est pas dépositaire d'une science, mais dispose plutôt de recettes (Bernand (1991: 42). » Si cette interprétation se révélait correcte, cela signifierait que la sorcellerie est plus accessible que sa contrepartie la « magie blanche ». Une interprétation légèrement différente, mais qui n'est pas totalement en contradiction avec la précédente, consisterait à dire que la personne pratiquant l'acte de sorcellerie doit jouir d'un talent particulier pour cet art. Ainsi, la sorcellerie pourrait notamment être pratiquée par les mêmes experts que la magie bienveillante, comme semblent le penser Gabriella Erantz-Szabó" et Volkert Haas (Haas 2001: 55). Cette interprétation serait proche de ce que Ton observe dans l'antiquité gréco-romaine'l Un indice en faveur de cette vue est l'ambiguïté même qui caractérise certaines incantations de magie non agressive: les incantations de contre-magie d'une part, et celles accompagnant les serments d'autre part. Dans ces deux types d'incantations, on inclut une formule de malédiction à la récitation. La seule différence entre les deux contextes est que, dans le cas d'un serment, la malédiction est « conditionnelle », pour reprendre l'expression de Eritz Graf (1994:185): elle ne se produira qu'en cas de parjure. Concernant les serments hittites, je renvoie le lecteur aux travaux de Mauro Giorgieri (1995) et de Kirsten M. Reichardt(1998). Je tiens toutefois à citer un passage des serments militaires hittites édités par Norbert Oettinger, et qui se trouve clairement à l'intersection entre serments et magie malveillante {KBo 6.34+ iii 12-23): n[u=sm] as=kan AL[AM ... S]k=SUuidansu [INA Q]ATI=SUNUdä[i] nukisan tezzi [k]äs=wakuis Ü[L=wa] linkiskit [nu] DINGIR'^'^^-flS peran [link]atta namma=kan NlS DINGIR''[™ sar]radda n=an linkiantes épper n=as=san èk=§Usuttati nu=za sarhuwandan QATI=SUperan UGU-a karpan harzi nâsta kuiskûsNiS DINGIR™ iarri'zzz n=an kèNlS DINGIR'^'^^ appandu n=as=san SÀ=SLrsuttaru andurza=ma= [...] =kan INA èA=§UDUMU Îshära

[...-d]u n=an karipandu « II place dans leurs mains une figu[rine ...] ayant le cœur remph d'eau et il dit la chose suivante: 'Qui est-il ? N'a-t-il pas prêté serment ? Il a prêté serment en présence des dieux et ensuite il a brisé le serment des dieux. Les (dieux du) serment se sont emparés de lui. Il a eu le cœur rempli (d'eau). Sa main a porté son estomac vers le haut. Quiconque brise ces serments du dieu, que ces serments des dieux se saisissent de lui ! Que son cœur se remplisse (d'eau) ! Qu'à l'intérieur, dans ses organes, le(s) fils d'Ishara [...] ! Qu'ils le dévorent (Oettinger (1976: 12-13) !' » La manipulation, dans ce contexte, d'une figurine anthropomorphe associe cette malédiction « conditionnelle » aux pratiques d'envoûtement connues par les rituels de contre-magie. Enfin, un rituel prophylactique consiste peut-être à brûler la figurine en bois d'un ennemi du roi hittite". Cette pratique, que Daniel Schwemer identifie également dans le monde mésopotamien (2007), est bien à l'intersection entre magie bienveillante et malveillante.

10. Hoffmann 1984: 55 § 50: l'édit indique qu'un acte de sorcellerie pratiqué au sein d'une famille est possible. 11. Frantz-Szabó 1995: 2007: « On the whole, witchcraft, like magic, may perhaps be considered from the special viewpoint of an art practiced professionally by one with a talent for it, in order to help or harm fellow human beings or superior forces. » 12. Gaily 1978: 52: « Ce dernier terme [sorcière] renvoie en effet à une image créée par un mythe différent issu en majeure partie du Moyen-Âge. Une telle notion, engendrée principalement par l'idéologie chrétienne et sa façon de poser le problème du Bien et du Mal, n'existe pas dans l'Antiquité. La magie n'est pas alors décrite comme un phénomène collectif (sabbats - messes 'noires'...) mais reste une affaire de spécialistes qui opèrent seuls grâce à une science précise et à des pouvoirs particuliers qui en font les alliés des divinités plus que les victimes de forces mauvaises. Elle garde avant tout un caractère rituel et religieux. » 13. Voir en dernier lieu Schwemer 2007a: 271-72: il s'agit du rituel de Nikkaluzi KUB 7.61 sur lequel nous reviendrons ci-après.

SORCELLERIE HITTITE

109

Ces quelques exemples indiquent que la magie malveillante pouvait être pratiquée par les spécialistes de la magie « officielle » que nos textes mentionnent, la Vieille Femme étant le personnage prédominant". La magie hittite est donc ambivalente, tantôt bienveillante, tantôt malveillante et même parfois les deux à la fois'^. La preuve en est que quand, comme l'indiquent les Lois Hittites, un expert pratique un rituel de purification mais place les résidus du rituel dans le champ ou la maison de quelqu'un, cela est considéré comme un acte de sorcellerie (Hoffner (1997: 52-53 § 44b). Dans le même ordre d'idées, Tzvi Abusch a montré comment la sorcière kaüäptu du monde mésopotamien, bien que surtout mentionnée pour sa magie malveillante, n'était pas considérée comme méchante par nature et pouvait tout aussi bien pratiquer la magie bienveillante. Elle pouvait même intervenir en tant que guérisseuse (Abusch 2002: 8-10 et 84-85). La kassäptu mésopotamienne est donc en premier lieu ce que l'on appelle un « witch-doctor » dans la littérature anthropologique: magicienne et guérisseuse à la fois. L'ambivalence de la magie se retrouve dans cefle de l'un de ses principaux instruments: les plantes. En effet, le fait qu'une seule et même plante puisse tenir lieu tantôt de médecine, tantôt de poison est bien connu. Par ailleurs, la magie malveiflante ne relève pas des seules couches populaires, mais se retrouve aussi dans les élites sociales. Le rituel de Hebattarakki fait aUusion à la possibilité, pour la « mauvaise personne » d'avoir ses dieux, ses serviteurs, son char et ses chevaux, ce qui tend à prouver qu'une personne exerçant la sorcellerie peut provenir d'une classe sociale aisée. Les textes historiques se font en outre l'écho de plusieurs accusations de personnes proches de la sphère royale pour actes de sorcellerie (Arma-Tarhunta contre Hattusili III, Tawananna contre Mursili II, etc.), Raymond Westbrook les a réétudiés récemment"^. Cela pourrait indiquer que la sorcellerie n'était pas vue, dans le monde hittite, comme l'apanage des magiciens, mais que des personnes bénéficiant seulement de quelques rudiments de magie étaient vues comme capables de la pratiquer. Une autre interprétation est de considérer que chacun de ces actes de malveillance imputés à des membres de l'élite sont vus comme ayant été effectués par l'intermédiaire d'un spécialiste. Il me semble que les deux cas de figure doivent être envisagés, en fonction de l'acte de sorcellerie pratiqué. La malédiction, par exemple, est décrite dans les textes hittites comme pouvant être effectuée directement par le non spécialiste. En revanche, d'autres pratiques de magie malveillante, tel que l'envoûtement, par exemple, pouvaient être l'apanage exclusif des experts rituels. Cette dernière interprétation reste une simple supposition, les textes n'étant pas explicites sur ce point. Derrière ces accusations de sorcellerie, on cherche avant tout à donner un sens à un événement malencontreux. Comme l'écrit très bien Jean-Benoît Clerc au sujet de la magie de la Rome impériale, reprenant ainsi une idée d'Edward E. Evans Pritchard, « La magie et la sorcellerie apparaissent comme une mise en forme des malheurs ; elles introduisent un lien causal dans les événements malencontretxx de façon à leur donner un sens, une explication particulière qui déterminera une réaction et qui déclenchera un processus de résolution, d'annulation du malheur

14. Concernant la Vieille Femme, voir notamment Bin-Nun 1975:120 et suivantes. Notons toutefois que le passage plusieurs fois commenté du Testament de Hattuáili I ne lie pas la Vieille Femme et la sorcellerie de manière explicite. Le passage indique en effet (KUB 1.16+ iii 65-69: Sommer et Falkenstein 1938: 16-17 et Beckman 2003: 81): (version akkadienne) là inûma lugal kiäm iqab\bHum u\ dumu""* é.gal iqabbùium annû •"»"»•"«»äu.gi [t]astanal u lugal kiäm iqabbi adi inann[a] "'""'«m'isu.gi iStanal ul idê I (version hittite) ¡ë=man=Se [LUG]AL-«'a=s7 [ÎR"'=5=S] [/ANSE.KUR.RA"'^SG'SGIGIR'^'=S=5(;5^ IM=paf ienzi « On fait en argile une figurine (représentant) la mauvaise personne. On fait également ses dieux, ses serviteurs, ses chevaux et ses chars en argile. » Des parallèles mésopotamiens sont connus, notamment dans le recueil néo-assyrien Maqlû, dans lequel on lit: « J'ai confectionné la figurine de 'mon' sorcier et de 'ma' sorcière, de mon envoûteur et de mon envoûteuse, je vous ai mis sous serment, je (vous) ai attaqués en justice"*'. » A.2. Figurines non anthropomorphes: les figurines les plus fréquemment mentionnées dans les textes hittites sont celles qui sont en forme de langues. Cela est lié à la prédominance des « langues mauvaises » en tant que technique d'ensorcellement. Les figurines en forme de langues peuvent être en argile, en pâte ou en cire, principalement. Le rituel de Sehuzzi pour rétablir une personne ensorcelée {KUB 9.39 + KUB 43.59 i 1: man Uîi^-a[h] handan SlG^-ahmi) préconise la confection de figurines en forme de langues et en forme de bouche {KUB 9.39 + KUB 43.59 i 2-8: CHD S, 109): [n]u=za saluinaspurut lalauïsnaspurut wa[p]puwas IM-an NUMUN"^ hûmanda DIM4 BAPPIR n=at EGlR-paparasza mallanzi n=at IM-ni IGl-anda immianzi nu ALAM"^ KAxU EME"^ DÜ-m/ n=at KAxU-/s hassit sunnahhi n=at IM-nit istappi n=at siyami « De la terre de salwana-, de la terre de fourmis, de l'argile de la rive (d'un fleuve), toutes les graines, du malt et du pain à bière. On les moud à l'envers. On les mélange avec de l'argue. Je fais des figurines de bouche et de langues. Je les remplis, (elles et) la bouche, de cendre, je les recouvre d'argile et je les scelle. » Le rituel de Mastigga mentionne également des figurines en forme de mains dans un contexte faisant allusion à des malédictions. Volkert Haas indique très justement que figurines de mains et figurines de langues symbolisent les deux catégories d'actes magiques: manuels et oraux (Haas 1994: 884). Par ailleurs, le rituel d'AUaiturahhi de Mukis fait allusion à un chien et à un porc de l'ensorceleur". Peut-être s'agit-il de figurines représentant ces deux animaux, comme le proposent Volkert Haas et Use Wegner, les éditeurs de ce texte ? Le passage est trop lacunaire pour pouvoir l'affirmer.

B. Des malédictions contre une malédiction précédente Le rituel d'Alli contient une incantation assez inhabituelle (§ 6:36-39): [nu a]lwanzata dukanzi x[...] ésdu n=an hassannit däu ÍAH=ma=as=kan °'^GAG-az isparzasdu n=at EGIR-pfl BELI=SU paiddu « Que [ren]sorcellement soit... dukanzi ! Qu'elle le prenne par (sa) matrice ! Qu'il s'échappe du bâton du porc^^ ! Qu'il retourne auprès de sa propriétaire ! » Ce passage n'est pas sans difficulté. Outre dukanzi, dont le sens est peu clair dans le passage, mais que Craig Melchert suggère de traduire « cultivation, breeding » (Melchert 1999: 20-22), le lien logique entre les différentes propositions de cette incantation n'est pas évident à première vue. Seule l'interprétation proposée par Craig Melchert me paraît résoudre cette difficulté: selon lui, l'expression « bâton du porc » utilisée dans ce passage serait une façon de désigner le sexe de l'animal. On souhaiterait ainsi que l'ensorcellement soit renvoyé dans la ma-

42. 43. 44. 45.

Bawanypeck 2005: 276-77. Voir également Prechel 2002. Maqlû 115-17 (Meier 1937: 7): epuS nu "uSu-jfl u •""""'USU-JÎJ sa episya u mustepistiya aSkun ina saplikunuma adibbub dini. KUB 24.13 ii 29: Haas et Wegner 1988: 111. -as alterne avec -at dans les différents manuscrits, bien que -aS soit majoritaire. Voir Melchert 1999: 21 pour ce passage.

SORCELLERIE HITTITE

119

trice de l'ensorceleuse à la suite d'une relation sexuelle avec un porc. Nous aurions donc affaire une malédiction. Or, une telle malédiction n'est-elle pas un des moyens les plus logiques de lutter contre une malédiction préalable ? Un exemple de « contre-malédiction », si l'on peut dire, se retrouve dans le rituel KUB 17.27 d'Allaiturahhi (KUB 17.27 iii 12-15; 18): n=at=kan ANÊ-as sehurreskiddu"' [n=at]=kan G\J,-us kammarsieskiddu DUMU. LÚ.U,5.LU-MS=5a« [kuis] ser arha iyattari nu=(s)san ser [all]appahhiskiddu; iii 18: [...]x=war=an UH^-an UN-an L[I]M DINGIR'^'^^ hurzaskand[u] « Qu'un âne urine sur eux (= les mots de l'ensorcellement UH^-nas uddär) ! Qu'un bœuf défèque sur eux ! Qu'un homme qui passe (par ici) crache sur (eux) !(...) Que les mille dieux maudissent la personne ensorceleuse [...] (Haas et Wegner 1988:194; Haas 2007:18) ! » D'autres dieux devant maudire l'ensorceleur sont enumeres ensuite, dans un passage plus lacunaire. Qn pourrait également citer l'incantation prononcée lors du rituel de Pittei contre l'ensorcellement d'un nouveau-né, incantation assez difficile à comprendre mais dans laquelle interviennent plusieurs malédictions (Beckman 1983:178-79).

C. Ensevelissement Le rituel d'Alli mentionne la possibilité de renvoyer l'ensorcellement et les mauvais rêves dans la Terre Sombre et de les y fixer (§ 20:124-36): manninkuwahhi tékan paddäi nu=kan animas KIN'^'* anda däipùrut ser ishuwai nu istalakzi nu=za °'^GAG"'^ walahzi nu kisan tezzi kuis kùn alwanzahhiskit kinun=a apél alwanzata EGIR-pa dahhun n=at=kan taknai kattanda téhun n=at tarmänun nu alwanzata idalauës teshus tarmäntes asandu n=at=kan namma sarä lé u[i]zzi n=at dankuis daganzipas hardu « A proximité, elle creuse la terre et elle place les objets du rituel dans (le trou). Elle verse de l'argile par-dessus et elle aplanit (le tout). Elle enfonce les piquets et dit: 'Celui qui a ensorcelé celui-ci, maintenant, je lui ai retiré l'ensorcellement, je l'ai placé dans la terre et je l'y ai fixé. Que l'ensorcellement (et) les mauvais rêves (y) soient fixés ! Qu'ils ne remontent plus ! Que la Terre Sombre les (re)tienne !' » Une technique analogue est mentionnée dans le rituel kizzuwatnîen de Mastigga contre une querelle familiale: on y précise que les malédictions, sous la forme de la bouche et de la langue, doivent être fixées dans le sol et adressées à la divinité solaire (Miller 2004: 94-95 § 37).

D. Couper un lien Le rituel de Mastigga mentionne la technique consistant à « couper/séparer les langues » avec un couteau (§ 10"') nu SÍG mittan däi n=at=¡asma[s]=kan ISTU G Í R ser arha kuerzi nu kissan memai apedani=wa=smas=kan U4-Í1 kuit hassikkedumat kinun=a=wa=smas=kan kä[s]a apedas U^-as EME"'^ °Andäliyas I§[T]U GÍR karasta n=an=kan hassi [pe]ssiyazzi: « EUe prend de la laine rouge, la coupe au-dessus d'eux avec un couteau, et parle ainsi: 'Du fait que vous vous êtes quereflés en ce jour, maintenant, Andâliya vient de couper de vous les langues de ces jours.' Elle jette cela dans le foyer. » Le fait que l'ensorcellement soit vécu comme un lien qui unit magiquement l'ensorceleur et l'ensorcelé sera examiné ci-après.

E. Autres techniques Les textes attestent d'autres techniques de contre-magie. La façon dont la Vieille Femme nie le talent de l'ensorceleur dans un petit rituel attribué à Allaiturahhi est à cet égard intéressante: elle prétend en effet que l'ensorce-

46. D'après la photographie de la tablette mise en ligne sur le site de Konkordanz, il semble bien qu'un signe SE ait été inscrit sur un signe ES érasé, d'où la confusion de lecture possible pour ce terme entre eSharreSkiddu (HED K: 37) et Sehurreikiddu (Haas et Wegner 1988: 194). Par ailleurs, la lecture choisie par Haas et Wegner me parait plus satisfaisante dans le contexte. 47. Je suis la numérotation de Miller 2004.

120

ALICE MOUTON

leur n'a pas pratiqué son rite d'ensorcellement correctement"". Des techniques de purification connues par ailleurs (lustration, exorcisme voire substitution, etc.) peuvent également être utilisées dans le cadre d'un rituel de contremagie. D'autres techniques de contre-magie relèvent quant à elles, semble-t-il, de la thérapeutique. Ainsi, dans le rituel d'Alli, il semble que la Vieille Eemme fasse à deux reprises une sorte de cataplasme qu'elle enroule sur une partie du corps du patient tout en prononçant une incantation de magie analogique (§ 29 et 30). De même, dans le rituel de Mastigga, une « plante de la divinité solaire » est enroulée autour du corps des deux antagonistes en proférant une incantation demandant la suppression des « mauvais mots de la bouche et de la langue » (§ 43). Notons enfin l'utilisation, dans le rituel d'Hebatarraki {CTH 397), d'excrément de chien mélangé à de la farine d'orge et à différentes plantes qui, sous forme de pâte, est d'abord appliquée sur la victime (le texte utilise le verbe tamass- « to press ») puis apparemment brûlée en fumigation puisque la conjuration qui accompagne le rituel indique notamment {KUB 24.14 i 22-24): SA UR.Gl7=míJ=(í)ífl warsulan awan arhaparhun SA \JR.Gly=ma salpas {KUB 24.15 i 22: salpan) UZUUR.GI, "^"GÎR.PAD.DU UR.Gl7=>'a simisiyanun « J'ai chassé l'odeur du chien hors de toi, j'ai brûlé en fumigation les (ou le dans le duplicat) salpa- du chien, la chair du chien et les os du chien. »

Conclusion Lensorcellement peut consister à l'établissement d'un lien magique entre deux personnes: l'ensorceleur et l'ensorcelé. Le rituel mantalli(ya) est précisément destiné à dénouer ce lien"^ Dans ce contexte, l'expression mantalliés £]yI£HÁ.g| ¿JJJ5 jg rituel de Mastigga contre des querelles familiales est particulièrement intéressante, puisqu'elle atteste l'existence d'une relation entre la langue et le fait d'établir un lien. À côté des allusions faites au lien symbolique qui unit les antagonistes, quelques séquences rituelles matérialisent ce lien sous la forme d'une corde tressée qu'il s'agit de dénouer: c'est le cas dans le rituel d'Allaiturahhi KUB 17.27'°. Le rituel d'Uruwanda contre « les langues » CTH 411 représente lui aussi l'ensorcellement sous la forme d'éléments liés entre eux. Le texte indique en effet {KBo 11.11 i 3-5): n=asta "^"SA ANA GI anda huittahhari nu={s)san hamenkeskimi « Je tire le tendon dans un roseau. Je (les) lie (ensemble)^'. » Notons que la représentation de l'ensorcellement comme un lien, un nœud, se retrouve dans la magie de l'Egypte ancienne et de la Grèce antique. En Egypte ancienne, un article récent de Willemina Wendrich a montré à quel point les nœuds sont omniprésents dans les pratiques magiques (Wendrich 2006). Quant à la symbolique du lien dans la Grèce ancienne, elle a notamment été étudiée par André Bernand à travers l'analyse du terme grec katádesis « ligature » que le latin traduit par defixio (Bernand 1991: 75 et 162). Derrière cette image du lien magique se cache en outre un second concept: le lien entrave la personne ensorcelée, il la paralyse. Dans le même ordre d'idées, lier la figurine représentant la personne ensorceleuse comme on le fait dans le rituel d'Alli est une manière de l'immobiliser". Le rituel d'Allaiturahhi et ses nombreuses expressions utilisant le terme ishai- « lier » est une belle illustration de cette conception de l'ensorcellement. Si l'ensorceleur est parvenu à ses fins, c'est le plus souvent parce qu'il a réussi à éloigner les dieux de sa victime". Il a, pour ce faire, provoqué leur colère soit en accusant sa victime d'avoir commis une faute vis-à-vis d'eux (ce qui est l'un des principes de la malédiction mais semble également correspondre à certains emplois de l'expression « mauvais yeux »), soit en souillant sa victime, ce qui a comme effet d'éloigner les dieux. Le rituel d'Ammihatna fait notamment allusion à l'ensorcellement de la nourriture de la victime qui, une fois ingérée, provoque son impureté. Les dieux en colère considèrent alors la personne ensorcelée comme un parfait étranger. Dans un article d'Ilya

48. KUB 17.27 ii 28'-30'; Haas et Wegner 1988; 192. 49. Sur le champ sémantique du lien qu'on noue et qu'on dénoue, voir également Haas et Thiel 1978; 40-46. 50. KUB 17.27 ii 37'-41'; Haas et Wegner 1988; 193. 51. Pour d'autres citations mentionnant l'ensorcellement sous la forme d'un lien, voir CHD L-N: 2-3 sub la-. 52. Cette même image se retrouve notamment dans le rituel de guerre néo-assyrien récemment édité par Daniel Schwemer, où l'ennemi serait soupçonné de pouvoir « lier » (russû) les armes du roi. Pour ce sens de russû, voir Schwemer 2007a; 39. 53. Voir, sur ce thème, Abusch 1999.

SORGELLERIE HITTITE

121

Yakubovich5^ cet auteur, suivant une suggestion d'Elisabeth Rieken, indique que la racine hittite *alwanza- pourrait se retrouver dans l'adjectif louvite *aliwanna/i-, *alunna/i- qui signifierait « ennemi, étranger » à l'instar du latin aliénus. Gette théorie me paraît tout à fait séduisante à la lumière de ce qui a été dit ici: l'ensorcellement est avant tout ressenti comme une aliénation^'. Les techniques de contre-magie relèvent principalement de la logique suivante: il s'agit de renvoyer l'ensorcellement à l'ensorceleur, la malédiction à la personne qui l'a proférée. Pour ce faire, plusieurs techniques peuvent être utilisées voire de préférence combinées: 1) la personnification de l'ensorceleur sous la forme de figurines sur lesquelles on pourra agir plus aisément ; 2) maudire la personne ayant proféré une première malédiction. Gette technique est clairement mentionnée dans un passage du recueil néo-assyrien Maqlû, dans lequel il est dit: « Sa parole est retournée dans sa bouche, sa langue est hée^*. » Le même recueil décrit par ailleurs l'incantation suivante: « Je fais contre toi l'acte (de sorcellerie) que tu avais fait contre moi. Je t'ai fait te confronter à la machination à laquelle tu m'avais fait me confronter. Je t'ai retourné la faveur que tu m'avais faite" ! » Notons que des mesures préventives permettant d'éviter un ensorcellement devaient également exister, mais les textes ne les décrivent pas, du moins à ma connaissance. En revanche, un texte mésopotamien étudié par Walter Earber fait allusion à l'emploi d'une pierre magique et d'une incantation pour « empêcher la sorcellerie d'atteindre une femme enceinte et qu'elle ne perde son fœtus'* ». Pour neutraliser l'ensorcellement, il faut donc refaire les mêmes gestes à l'envers, pour défaire ce qui a été fait, comme l'indique très justement Maurice Vieyra (1966: 106). Gelui-ci écrit: « Défaire, d'une part, afin de remettre les êtres et les choses dans leur état primitif, renvoyer le mal là d'où il vient, c'est-à-dire, d'abord chez le sorcier: telle sera surtout la tâche du magicien. Mais aussi, il arrivera très fréquemment, plus fréquemment peut-être, que l'on ne se contentera pas de renvoyer le mal à celui qui l'a provoqué, mais bien dans 'la Terre sombre', aux Enfers, où - tirons la conclusion que les textes suggèrent - réside le principe du mal » (Vieyra 1966:113). Plus que défaire, c'est surtout dire qui semble rendre le rituel de contre-magie efficace. Le caractère performatif des incantations prononcées n'est plus à démontrer: en décrivant un phénomène, on souhaite l'ancrer dans la réalité autant que l'on suggère aux dieux d'agir d'une certaine façon. Après avoir défait ce qui a été fait, on éloigne à jamais la menace en clouant, en enterrant ou en brûlant l'élément symbolisant l'ensorcellement. Les deux premières techniques semblent témoigner de l'origine chthonienne de l'impureté ainsi provoquée. L'impureté étant une des principales conséquences de l'ensorcellement, son caractère chthonien est important à souligner: il témoigne du lien qui semble exister entre la sorcellerie et le monde des morts, lien qui est bien mieux connu dans d'autres sociétés méditerranéennes anciennes (Egypte, Grèce et monde romain). La defixio romaine consiste en effet à envoûter un adversaire en faisant appel à un défunt. À ma connaissance, une telle pratique n'est pas attestée dans les textes hittites, mais un lien magique relevant de la sorcellerie peut toutefois exister entre une personne vivante et un défunt, comme en témoigne le compte rendu oraculaire KBo 2.6 édité par Theo van den Hout (1998: 196-217). Ensorceler quelqu'un revient donc à l'attirer irrésistiblement dans le monde des morts. On se souviendra de l'expression taknaz da- « prendre/retirer de la terre » qui prend ici tout son sens. La personne ensorcelée entre malgré elle dans un état d'entre-deux, de liminarité comme diraient les anthropologues, qui la met en danger: elle a déjà un pied dans le monde des morts sans y être pourtant totalement intégrée. Le lien magique qui la paralyse et l'enchaine à l'ensorceleur doit impérativement être coupé pour éviter qu'elle ne soit définitivement happée par la Terre Sombre.

54. 55. 56. 57. 58.

Yakubovich 2008. Je remercie I. Yakubovich qui m'a aimablement fait parvenir le manuscrit de son article avant parution. Kitz 2007: 619 utilise, dans une logique analogue, le terme « segregation ». Maqlù I 28 (Meier 1937: 8): turrat inim=sa ana kaxu=ío eme=Sa kasrat. Maqlû VII 73-75 (Meier 1937: 49): ipSu tepusinni epuski mihir tuSamhirinni uSamhirki gimil tagmilinni utîragmilki. Färber 1989: ÎIO-U: ana munus eriti kiSpu nu te-e¡a-sa-¡á la iub-e.

122

ALICE MOUTON

References Ahusch, T. 1999

Witchcraft and the Anger of the Personal God. Pp. 83-121 in Mesopotamian Magic: Textual, Historical, and Interpretative Perspectives, ed. T. Abusch et K. van der Toorn, Ancient Magic and Divination 1, Groningen: Styx. 2002 Mesopotamian Witchcraft: Toward a History and Understanding of Babylonian Witchcraft Beliefs and Literature, Ancient Magic and Divination 5. Leyde: Brill. Bawanypeck, D. 2005 Die Rituale der Auguren, THeth 25. Heidelberg: Winter. Beal, R. H. 2001 Hittite Military Rituals. Pp. 63-76 in Ancient Magic and Ritual Power, 2nd ed., éd. M. Meyer et P. Mirecki. Leyde: Brill. Beckman, G. 1983 Hittite Birth Rituals. StBoT 29. Wiesbaden: Harrassowitz. 2003 Bilingual Edict of Hattusili 1. Pp. 79-81 in The Context of Scripture II. Monumental Inscriptions from the Biblical World, éd. W. W Hallo et K. L. Younger Jr., Leyde: Brill. Bernand, A. 1991 Sorciers grecs. Paris: Fayard. Bin-Nun, S. R. 1975 The Tawananna in the Hittite Kingdom. THeth 5, Heidelberg: Winter. Brand, R. 1986 Rites de naissance et réactualisation matérielle des signes de naissance à la mort chez les Wémémou (Bénin/Dahomey). Pp. 37-62 in Naître vivre et mourir. Actualité de Van Cennep, Musée d'ethnologie Neuchâtel (exposition du 28 juin au 31 décembre 1981), éd. J. Hainard et R. Kaehr, Neuchâtel: Musée d'ethnographie. Butler, S. A. L. 1998 'Mesopotamian Conceptions of Dreams and Dream Rituals. AOAT 258. Münster: Ugarit-Verlag. Gharpin, D. 1988 Les représentants de Mari à Ilân-§ûrâ. Pp. 31-137 in Archives épistolaires de Mari 1/2. Archives Royales de Mari XXVI, éd. D. Gharpin et ai, Paris: Editions Recherche sur les civilisations. Glerc, J. -B. 1995 Homines magici. Etude sur la sorcellerie et la magie dans la société romaine impériale. Publications Universitaires Européennes III/673. Bern: P. Lang. Gollins, B. J. 2003 Purifying a House: A Ritual for the Infernal Deities. Pp. 168-71 in The Context of Scripture I, 2nd ed., éd. W. W. Hallo et K. L. Younger. Leyde: Brill. Engelhard, D. H. 1970 Hittite Magical Practices: An Analysis, thèse de doctorat de l'Université de Brandeis. Farber, W. 1989 Schlaf Kindchen, Schlaf! Mesopotamische Baby-Beschwörungen und -Rituale. Mesopotamian Givilizations 2, Winona Lake, IN: Eisenbrauns. Favret-Saada, J. 2004 Sorcellerie. Pp. 670-73 in Dictionnaire de l'ethnologie et de l'anthropologie, 2nd edition, éd. P. Bonté et M. Izard. Paris: Presses universitaires de Erance. Erantz-Szabó, G. 1995 Hittite Witchcraft, Magic, and Divination. Pp. 2007-19 in Civilizations of the Ancient Near East 3, éd. J. M. Sasson, New York: Scrihner's Sons. Gaily M. 1978 Médée, ou le vertige d'un autre monde. Pp. 49-89 in La sorcellerie. Les Gahiers de Fontenay n° 11.12, Eontenay aux Roses: École Normale Supérieure. Gessel, B. van 1998 Onomasticon of the Hittite Pantheon I. HdO 1/33. Leyde: Brill. Giorgieri, M. 1992 Un rituale di scongiuro antico ittita per Laharna-Hattusili. SMEA 29: 47-98. 1995 / íesíi ittiti di giuramento. thèse de doctorat de l'Université de Elorence.

SORCELLERIE HITTITE Goetze, A. 1938 Graf, F. 1994 Haas, V. 1977 1988

123

The Hittite Ritual ofTunnawi. AOS 14. New Haven: American Oriental Society. La magie dans l'antiquité gréco-romaine. Paris: Les Belles Lettres.

Magie und Mythen im Reich der Hethiter. Hamburg: Merlin. Magie in hethitischen Gärten. Pp. 121-42 in Documentum Asiae Minoris Antiquae. Eestschriflfür Heinrich Otten zum 75. Geburtstag, éd. E. Neu et Ch. Rüster. Wiesbaden: Harrassowitz. 1994 Geschichte der hethitischen Religion. HdO 1/15. Leyde: Brill. 2001 Verfluchungen am hethitischen Hof und deren rituelle Beseitigung. Pp. 53-71 in Kult, Konflikt und Versöhnung. Beiträge zur kultischen Sühne in religiösen, sozialen und politischen Auseinandersetzungen des antiken Mittelmeerraumes, éd. R. Albertz. AOAT 285. Mütister: Ugarit-Verlag. 2003 Materia mágica et medica hethitica. Berlin: de Gruyter. 2007 Notizen zu den Ritualen der Frau Allaiturahi aus Mukis. AoE 34: 9-36. Haas, V et Thiel, H. J. 1978 Die Beschwörungsrituale der Allaiturah(h)i und verwandte Texte. Hurritologische Studien 2. AOAT 31. Neukirchen-Vluyn: Ugarit-Verlag. Haas, V. et Wêgner, I. 1988 Die Rituale der Beschwörerinnen ^î-SUGI. ChS 1/5. Rome: CNR. Hoffmann, I. 1984 Der Erlass Telipinus. THeth 11. Heidelberg: Winter. Hoffner, H. A. 1983 A Prayer of Mursili II about his Stepmother. lAOS 103: 187-92. 1997 The Laws ofthe Hittites. A Critical Edition. DMOA 23. Leyde: Brill. Hout, T. van den 1998 The Purity of Kingship. An Edition of CTH 569 and Related Hittite Oracle Inquiries ofTuthaliya IV. DMOA 25. Leyde: Brill. Hutter, M. 1988 Behexung, Entsöhnung und Heilung. OBO 82. Freiburg: Universitätsverlag Freiburg; Göttingen: Vandenhoeck & Ruprecht. 1991 Bemerkungen zur Verwendung magischer Rituale in mittelhethitischer Zeit. AoP 18: 32-43. 2003 Aspects of Luwian Religion. Pp. 211 -80 in 77!e Luwians, éd. H. G. Melchert. HdO 1/68, Leyde: Brill. Kassian, A. 2000 Two Middle Hittite Rituals Mentioning 'Ziplantawija, Sister ofthe Hittite King '"Tuthalija II/I. Moscou: Paleograph. Kitz, A. M. 2007 Gurses and Gursing in the Ancient Near East. Religion Compass 1/6: 615-27. Labat, R. 1951 Traité akkadien de diagnostics et pronostics médicaux 1. Travaux de l'académie internationale d'histoire des sciences 7, Paris: Académie internationale d'histoire des sciences; Leyde: Brill. Lebrun, R. 1976 Samuha. Foyer religieux de l'empire hittite. Louvain-la-Neuve: Université catholique de Louvain. Marwick, M. éd. 1970 Witchcrafl and Sorcery. Selected Readings. Harmondsworth: Penguin Books. Meier, G. 1937 Die assyrische Beschwörungssammlung Maqlû. AfO Beiheft 2. Berlin: Archiv für Orientforschung. Melchert, H. G. 1983 Pudenda hethitica. ICS 35: 137-45. 1999 Hittite tuk(kan)zi- 'Cultivation, Breeding'. Ktèma 24: 17-23. Miller, J. L. 2004 Studies in the Origins, Development and Interpretation ofthe Kizzuwatna Rituals, StBoT 46, Wiesbaden: Harrassowitz. Mouton, A. 2006 Quelques usages du feu dans les rituels hittites et mésopotamiens. RHR 223: 251-64. 2007 Rêves hittites. Contribution à une histoire et une anthropologie du rêve en Anatolie ancienne, GHANE 28, LeydeBrill.

124

2009

ALICE MOUTON

Le 'mauvais œil' d'après les textes cunéiformes hittites et mésopotamiens. Pp. 425-39 in Pensée grecque et sagesse d'Orient. Hommage à Michel Tardieu, éd. M.-A. Amir-Moezzi et al. Bibliothèque de l'École des Hautes Études Sciences Religieuses 139. Paris: Presses universitaires de France.

Oettinger, N. 1976 Die militärischen Eide der Hethiter. StBoT 22. Wiesbaden: Harrassowitz. Otten, H. 1961 Ein Beschwörung der Unterirdischen aus Bogazköy. ZA 54:114-57. Prechel, D. 2002 Betrachtungen zum Ritual der Pupuwanni. Pp. 277-88 in Silva Anatolica. Anatolian Studies Presented to Maciej Popko on the Occasion of His 65th Birthday, éd. P. Taracha, Varsovie: Agade. Reichardt, K. M. 1998 Linguistic Structures of Hittite and Luvian Curse Formulae. Thèse de doctorat de l'Université de Chapel Hill. Reiner, E. 1966 La magie babylonienne. Pp. 67-98 in Le monde du sorcier. Sources orientales 7. Paris: Éditions du Seuil. Schwemer, D. 2007a Witchcraft and War: The Ritual Fragment Ki 1904-10-9,18 (BM 98989). Iraq 69: 29-42. 2007b Abwehrzauber und Behexung. Studien zum Schadenzauberglauben im alten Mesopotamien. Wiesbaden: Harrassowitz. Sefati, Y. et Klein, J. 2002 The Role of Women in Mesopotamian Witchcraft. Pp. 569-87 in Sex and Gender in the Ancient Near East, 47' R.A.I., éd. S. Parpóla et R. M. Whiting. Helsinki: The Neo-Assyrian Text Corpus Project. Singer, I. 2002 Hittite Prayers. Vf AW 11. Atlanta: Society of Biblical Literature. Sommer, F. et Falkenstein, A. 1938 Die hethitisch-akkadische Bilingue des Hattusili L (Labarna IL), ABAW NF 16, Munich: Verlag der Bayerischen Akademie der Wissenschaften. Stewart, P J. et Strathern, A. 2004 Witchcraft, Sorcery, Rumors, and Gossip. Cambridge: Cambridge University Press. Strauss, R. 2006 Reinigungsrituale aus Kizzuwatna. Berlin: de Gruyter. Szabó, G. 1971 Ein hethitisches Entsühnungsritual für das Königspaar Tuthaliia und Nikalmati. THeth 1. Heidelberg: Winter. Taracha, P 1990 More about the Hittite taknaz da Rituals. Hethitica 10:171-84. 2000 Ersetzen und Entsühnen: Das mittelhethitische Ersatzritual für den Grosskönig Tuthalija (CTH *448.4) und verwandte Texte. CHANE 5, Leyde: Brill. Thomsen, M.-L. 1987 Zauberdiagnose und schwarze Magie in Mesopotamien. CNI Publikations 2. Copenhague: Museum Tusculanum. Torri, G. 2003 La similitudine nella magia analógica ittita. Studia asiana 2. Rome: Herder. 2004 A Hittite Magical Ritual to be Performed in an Emergency. JANER 4: 129-41. Trémouille, M.-Cl. 2004 I rituali magici ittiti. RANT 1:157-203. Unal, A. 1988 The Role of Magic in the Ancient Anatolian Religions according to the Cuneiform Texts from Bogazköy-Hattusa. BM£CC/3: 52-85. Vieyra, M. 1966 Le sorcier hittite. Pp. 101-25 in Le monde du sorcier. Sources orientales 7. Paris: Éditions du Seuil. Wendrich, W. 2006 Entangled, Connected or Protected ? The Power of Knots and Knotting in Ancient Egypt. Pp. 243-69 in Through a Glass Darkly: Magic, Dreams and Prophecy in Ancient Egypt, éd. K. Szpakowska. Swansea: The Classical Press of Wales. Werner, R. 1967 Hethitische Gerichtsprotokolle. StBoT 4. Wiesbaden: Harrassowitz.

SORCELLERIE HITTITE

125

Westbrook, R. 2006 Witchcraft and the Law in the Ancient Near East. Pp. 45-52 in Recht gestern und heute: Festschrift zum 85. Geburtstag von Richard Haase, éd. ]. Hengstl et U. Sick, Philippika. Marburger alterkumskundliche Abhandlungen 13, Wiesbaden: Harrassowitz. Yakubovich, I. 2008 The Luvian Enemy Kadmos 47: 1-19. Yoshida, D. 1991 Ein hethitisches Ritual gegen Behexung (KUB 24.12) und der Gott Zilipuri/Zalipura. BMECCJ 4: 45-61.

z u DEN AKKADISCHEN HEMEROLOGIEN AUS HATTUSA (CTH 546), TEIL II: DIE „OPFERBROT-HEMEROLOGIE" Jeanette C. Fincke (Leiden)

Emmanuel Laroche hat in seinem Catalogue des textes hittites (CTH) unter der Nummer 546 fünf akkadische Fragmente aus Hattusa (Bogazköy) der Textgruppe der Hemerologien - Texte, bei denen Ge- bzw. Verbote für bestimmte Handlungen an den einzelnen Tagen des Jahres festgehalten werden - zugeordnet, welche 1922 in den von Ernst E. Weidner angefertigten Keilschriftkopien als KUB 4,42-46 publiziert worden waren. Zwei weitere hemerologische Fragmente konnte Gernot Wilhelm bei seiner Durchsicht der unpublizierten akkadischen Texte aus Hattusa identifizieren und 1991 in KBo 36' als Nummern 54 und 55 in Kopie vorlegen. Eine Bearbeitung der Hemerologien aus der Hethiterhauptstadt ist bislang allerdings noch nicht erfolgt. Ein wesentlicher Grund hierfür ist die zur Zeit immer noch lückenhafte Bearbeitung der entsprechenden Niederschriften aus Mesopotamien, in deren Tradition die Textfragmente aus Hattusa stehen. Eine Edition dieser Textgruppe aus Babylonien und Assyrien wäre dementsprechend zum Verständnis und zur Einordnung der Fragmente aus Hattusa unentbehrlich. Diese Situation zu ändern hat sich Alasdair Livingstone zur Aufgabe gemacht, der seine Ergebnisse bereits in verschie-

Diese Studie ist im Rahmen des von Wilfred H. van Soldt, Leiden, geleiteten Projektes Transfer of Knowledge in a Cuneiform Culture entstanden, das von der Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO) finanziert und in Zusammenarbeit mit dem Leiden Institute for Area Studies (LIAS) organisiert wird. Nicole Pfeifer, Würzburg, sei an dieser Stelle für die Mühe gedankt, das Manuskript sorgfaltig durchzulesen und mit zahlreichen Anmerkungen zu versehen. Eür den ersten Teil dieser Studie, der eine Hemerologie über das „Rufen von Klagen" (sigü sasü) und das „Reinigen seines Gewandes" (subät-su ubbubu) behandelt, siehe ICS 61 (2009) 111-25. In diesem Artikel verwendete Abkürzungen, die AHw oder CAD nicht verzeichnen, sind folgende: BBVO = Berliner Beiträge zum Vorderen Orient (Berlin: Reimer, 1987ff.); CTH = Emmanuel R Laroche, Catalogue des textes hittites (Paris: Klincksieck, 1971); Emar VI.4 = D. Arnaud, Recherches au pays d'Astata. Emar VI/4. Textes de la bibliothèque : transcriptions et traductions. Mission Archéologique de Meskéné-Emar (Paris: Recherche sur les Civilisations, 1987); HZL = C. Rüster - E. Neu, Hethitisches Zeichenlexikon (Wiesbaden: Harrassowitz, 1989); KBo 34 = Heinrich Otten - Christel Rüster, Hethitische Texte vorwiegend von Büyükkale, Gebäude A, Keilschrifttexte aus Boghazköy Heft 34 (Berlin: Gebr. Mann, 1991); KBo 36 = Gernot Wilhelm, Literarische Texte in sumerischer und akkadischer Sprache, Keilschrifttexte aus Boghazköi Heft 36 (Berlin: Gebr. Mann, 1991); KUB 4 = Ernst F. Weidner, Akkadische Keilschrifttexte und Nachträge zu den Heften 1-3, Keilschrifturkunden aus Boghazköi Heft 4 (Berlin: Vorderasiatische Abteilung der Staatlichen Museen, 1922); SEL = Studi epigraftci e linguistici sul Vicino Oriente antico (Verona: Essedue, 1984-). 1. Das ebenfalls als hemerologisch klassifizierte Stück KBo 36,56 hingegen ist ein Omenfragment, wie sich aus den kopieren Zeichenspuren eindeutig ergibt: Hemerologische Texte nennen nur den Tag (UD.x.KAM, „x. Tag"), während die Formulierung ina UD.x.KAM, „am x. Tag," typisch für Omenprotasen ist. Zudem fehlt in KBo 36, 56 die für Hemerologien typische Einteilung der Angaben in Kolumnen. Folgende Spuren lassen sich auf dem Photo des Fragmentes (vgl. www.hethport.uni-wuerzburg.de/hetkonk/) erkennen: (V) [... ina UD.x]-i-13.K[AM ...] (2') [...] Una VD\17.KAM xlx...] (3')[...] îna UD"'.20.KAMx[x ...] (4')[...] ^>ia UD.21.KAMi x[x ...]. Die behandelte Sequenz von Tagen ([...], 17., 20., 21.,[...] Tag) deutet auf Mondfinsternisomina, wie sich aus dem Vergleich mit entsprechenden Texten aus Mesopotamien ergibt. Die einzige bislang aus Hattusa bekannte Tafel mit derselben Sequenz ist die Tafel mit hethitischen Mondfinsternisomina KBo 34,110 (+) 111 mit der Sequenz 14., 15., 16., 17., 20., 21. und 22. Tag. Ein möglicher Anschluß müßte an den Originalen überprüft werden. 127

JCS 62 (2010)

128

JEANETTE C. FINCKE

denen Artikeln vorgelegt hatl Weil er inzwischen zahlreiche weitere Textvertreter in verschiedenen Museen identifizieren konnte, wird die Gesamtbearbeitung dieser Textgruppe mit Ungeduld erwartet. Seit allerdings Maria C. Casaburi 2003 für eine der aus Assur bekannten Hemerologien die Bearbeitung der Textvertreter aus Assur (KAR 178,179 sowie sieben weitere Texte) und Ninive (K. 4131-1-, publiziert von R. Labat, RA 56 [1962] 1-8]) vorgelegt hat^ steht das Grundgerüst dieser Hemerologie" weiteren Forschungen zur Verfügung. Mit Hilfe ihrer in Partiturumschrift gebotenen Edition und des Wortindexes können nun fünf der sieben Fragmente aus Hattusa eindeutig dieser Hemerologie zugeordnet werden: KUB 4, 42-44 sowie KBo 36, 54 und 55. Die anderen beiden Fragmente lassen sich derzeit keiner bekannten Hemerologie zuordnen; für KUB 4,45 siehe unten S. 143-45, und für KUB 4, 46, das zusammen mit dem bislang als Omenfragment klassifizierten Fragment KUB 43, 1 zu einer Tafel gehört, siehe den ersten Teil dieser Studie zu den akkadischen Hemerologien aus Hattusa (vgl. Anm. 1). Mit der Bearbeitung der Hemerologie-Fragmente aus Hattusa, die hier vorgelegt wird, soll der Arbeit von Livingstone nicht vorgegriffen werden. Diese Studie versteht sich vielmehr als Ergänzung zur Arbeit von Casaburi unter Einbeziehung derjenigen Informationen, welche Livingstone in seiner Rezension zur genannten Edition in ZA 96 (2006) 138-41 vorgelegt hat.

Formale Besonderheiten der hemerologischen Texte der 2. Hälfte des 2. Jt.s v. Chr. Alle hemerologischen Fragmente aus Hattusa zeigen folgende Gemeinsamkeiten hinsichtlich der Verteilung des Textes auf der Tafel: 1. Die Schreiber bemühen sich nicht um einen sogenannten Blocksatz ihrer Zeilen, sondern lassen den nicht benötigten Raum am Ende der einzelnen Einträge frei. Dies haben diese Fragmente mit den hemerologischen Texten aus dem in Syrien gelegenen Emar' gemeinsam. Bei den etwa zeitgleichen Hemerologien aus Ugarit^ ebenfalls in Syrien gelegen, und aus Dür-kurigalzu' im Kernland Babyloniens - wie auch grundsätzlich bei den Hemerologien des 1. Jt.s v. Chr. - bemühen sich die Schreiber dagegen um einen Blocksatz, indem sie gegebenenfalls das letzte Zeichen eines Eintrages an den rechten Rand der Kolumne rücken und damit den Freiraum auf einen Bereich innerhalb der Zeile verschieben. 2. Diejenigen Einträge der Fragmente aus Hattusa, welche jeweils nur eine Zeile einnehmen, sind in zwei Kolumnen eingeteilt, so daß die Tagesangaben in der linken, die dazugehörigen Vorschriften und Deutungen in der rechten Kolumne stehen. Die etwa zeitgleichen hemerologischen Texte aus der Peripherie Mesopotamiens, sc. aus Ugarit und Emar, sowie aus Babylonien selbst, sc. aus Dür-kurigalzu, schreiben den Text dagegen durchgängig, ohne die Tage von den Vorschriften und Deutungen abzusetzend Gleiches gilt für die vergleichbaren neuassyrischen Tafeln. Letztere trennen jedoch die Einträge jedes einzelnen Tages von denjenigen des nachfolgenden Tages durch einen Paragraphenstrich ab. Bei den Exemplaren aus Hattusa dient der Paragraphenstrich lediglich zur Abgrenzung der einzelnen Monate zueinander. Hinsichtlich der Hemerologien aus Ugarit und Emar läßt sich

2. Vgl. z.B. „The Case of the Hemerologies: Official Cult, Learned Formulation and Popular Practice," Official Cult and Popular Religion in the Ancient Near East, Hg. B. E. Matsushima (Heidelberg: Winter, 1993), 97-113; „New Dimensions in the Study of Assyrian Religion," in Assyria 1995, Hgg. S. Parpóla - R. M. Whiting (Helsinki: The Neo-Assyrian Text Corpus Project, 1997), 165-77; „The Use of Magic in the Assyrian and Babylonian Hemerologies and Menologies," SEL 15 (1998) 59-67; „The Magic of Time," in Mesopotamian Magic: Textual, Historical, and Interpretative Perspectives, Hgg. T. Abusch - K. van der Toorn, Ancient Magic and Divination 1 (Groningen: Styx, 1999), 131-37; „Agriculture in Literary Calendar Texts," in Landwirtschaft im Alten Orient, Hgg. H. Kiengel - J. Renger, BBVO 18 (Berlin: Reimer, 1999), 375-79; „The Babylonian Almanac: A Text for Specialists?," in Die Welt der Götter, Hgg. B. Groneberg - H. Spieckermann, Beihefte zur Zeitschrift für die alttestamentliche Wissenschaft 376 (Berlin: de Gruyter, 2007), 85-101 (S. 85-86 verweist er darauf, daß sich sein Manuskript über die Neuedition aller hemerologischen Texte bereits weit fortgeschritten ist). 3. M. C. Casaburi, üme täbüti „I giorni favorevoli," History of the Ancient Near East vol. 8 (Padova: Sargon, 2003). 4. Vgl. hierfür die Rezension von A. Livingstone zu der von Casaburi vorgelegten Bearbeitung dieser Hemerologie, ZA 96 (2006) 138-41. 5. Vgl. D. Arnaud, Emar IV.4 Nr. 605-607. 6. Vgl. D. Arnaud, „Jours et mois d'Ougarit," SMEA 32 (1993) 124-29. 7. Vgl. R. Labat, „Un calendrier cassite de jours fastes et néfastes," Sumer 8 (1952) 17-36. 8. Die Hemerologien aus diesen drei Orten zeigen inhaltlich jedoch keine Übereinstimmungen mit denjenigen aus Hattusa.

z u DEN AKKADISCHEN HEMEROLIGIEN AUS H A T T U S A

129

diesbezüglich aufgrund ihres fragmentarischen Erhaltungszustandes keine Aussage treffen. Bei der mittelbabylonischen Hemerologie aus Dür-kurigalzu wird auf trennende Paragraphenstriche verzichtet, sondern stattdessen für jeden Monat eine neue Kolumne begonnen.

Formale und orthographische Merkmale der hemerologischen Texte aus Hattusa Abgesehen von den bereits genannten Übereinstimmungen lassen sich die sieben bislang identifizierten Fragmente aus Hattusa, deren Duktus sich aufgrund der geringen Anzahl typologisch relevanter Keilschriftzeichen nicht näher bestimmen läßt, zunächst anhand von vier formalen und orthographischen Merkmalen unterscheiden und in Gruppen einteilen: Formale Abweichungen finden sich z B. beim Verwenden einer sogenannten oberen und unteren Randleiste (1.), die bei Tafeln hethitischer Schreiber durchaus üblich ist', und bei der Anzahl der Kolumnen (2.), über welche hinweg eine Zeile geschrieben wird (für Hemerologien aus Hattusa sind zwei Kolumnen üblich, siehe oben). Was die Orthographie anbelangt, gibt es Unterschiede in der Schreibung der Zahl 9 (3.) und der Kennzeichnung der Ordinalzahlen (4.): Die Zahl 9 kann entweder auf die babylonische Art mit je drei vertikalen Keilen in drei übereinanderliegenden Reihen oder auf die assyrische Art mit nur drei übereinanderstehenden vertikalen Keilen geschrieben werden. Das Hethitische Zeichenlexikon {HZL) von Christel Rüster und Erich Neu, das sich zum Ziel gesetzt hat, sämtliche Zeichenformen der Bogazköy-Texte aufzunehmen, bucht nur die babylonische Schreibweise der Zahl 9, nicht aber die assyrische. Die babylonische Form findet sich auch auf den hemerologischen Tafeln aus Ugarit und Emar; die mittelbabylonische Hemerologie aus Dür-kurigalzu verwendet hingegen eine ausgefallene Zeichenform, welche in den oberen beiden Ebenen je vier Keile zeigen, wäbrend in der untersten nur ein Keil zu seben ist. Bei der Kennzeichnung der Ordinalzahlen weist das Zeichen KAM in den hethitischen Hemerologien entweder diejenige Zeichenform auf, welche einem HI-BAD gleicht, oder diejenige, welche fünf bis sieben schräge Keile zeigt und hier in Anlehnung an die Zeichenlisten von Rykle Borger'" mit KAM* gekennzeichnet wird. Die Verwendung des Zeichens KAM - die Variante mit den horizontalen Keilen - ist den Zeichenlisten zufolge" in dieser Funktion eine sekundäre Erscheinung und wird nur in späten Texten verwendet'^; eine erste Durchsicht mittelassyrischer Texte deutet darauf hin, daß sich KAM zur Kennzeichnung von Ordinalzahlen bereits unter Tukulti-Ninurta I. (1243-1207 v. Chr.) durchzusetzen begann" und Im späten 12. Jh. v. Chr. im ganzen Einzugsbereich von Assyrien als akzeptierte Schreibvariante galt, vgl. hierzu z. B. die Texte des mittelassyrischen Königs des Landes Märi, Aááur-ketti-léser, der ein Vasall Tiglat-pilesars 1. (1114-1076 v. Chr.) war'"*. Zur zeitlichen Verteilung

9. Einer untere Randleiste auf Vorder- und Rückseite sowie eine obere Randleiste auf der Rückseite ist bei hethitischen Tafeln durchaus nichts ungewöhnliches, wohl aber eine obere Randleiste einer Vorderseite, wie sie beim Fragment KUB 4, 44 zu sehen ist; freundliche Mitteilung von Willemijn Waal. Die Randleiste der hethitischen Tafeln ist Gegenstand der Untersuchung von Frau Waal über Form, Textgenre und Fundort der hethitischen Tontafeln, die sie zur Zeit in Leiden, Niederlande, im Rahmen ihrer Dissertation (The Source as Object: Studies in Hittite Diplomatics) durchführt. 10. Vgl. R. Borger, Assyrisch-babylonische Zeichenliste, AOAT 33/33A (Kevelaer: Butzon 8i Becker / Neukirchen-Vluyn: Neukirchener Verlag, 1978), 161 Nr. 406, und ders., Mesopotamisches Zeichenlexikon, AOAT 305 (Münster: Ugarit-Verlag, 2003), 389 Nr. 595 und 640. 11. So auch implizit R. Labat, Manuel d'épigraphie akkadienne (Paris: Geuthner, 1988), 101 Nr. 143, der in seiner Übersicht über die Paläographie bei den älteren Zeichenformen jedoch die Variante mit den schrägen Keilen anführt und damit zur Unklarheit hinsichtlich der Unterscheidung zwischen KAM und KAM beiträgt. W. von Soden - W. Röllig, Das akkadische Syllabar, AnOr 42 (Roma: Pontificium Institutum Biblicum, ''1991), 46 Nr. 235, halten das schräggestellte KAN hingegen für „eine jüngere graph. Variante von KAM und kein selbständiges Zeichen," womit sie Borger folgen. 12. Leider hat sich hinsichtlich dieser Unterscheidung der drei Zeichenformen in der Umschrift unter den Assyriologen noch keine Konvention durchgesetzt, so daß häufig die Variante mit den schrägen Keilen als KAM transliteriert wird. 13. Vgl. z. B. KA/228, 291, aber: KAJ 168, 301, 314, 316 mit KAM* (alle genannten Texte stammen aus dem Archiv M8 aus Assur), vgl. O. Pedersén, Archives and Libraries in the City of Assur, Part I (Uppsala: Almqvist & Wiksell International, 1985), 82-89. 14. Vgl. S. M. Maul, Die Inschriften aus Tall Täbän (Grabungskampagnen 1997-1999). Die Könige von Tàbêtu und das Land Märi in mittelassyrischer Zeit, Acta Sumerologica Suppl. Ser. 2 (Tokyo: Kokushikan University, 2005), Text Nr. 4 und 6.

130

JEANETTE C. FINCKE

der Zeichenformern KAM und KAM* läßt sich lediglich festhalten, daß KAM in mittelassyrischer und -babylonischer Zeit aufkommt, aber erst in den späten hethitischen Texten erscheint und dort KAM* teilweise ersetzt. Was die etwa zeitgleichen hemerologischen Texte aus anderen Orten betrifft, so verwenden die Texte aus Ugarit ebenfalls das Zeichen KAM*, während die Hemerologie aus Dür-kurigalzu KAM schreibt; auf den vergleichbaren Fragmenten aus Emar ist keines der entsprechenden Zeichen erhalten. Anhand der genannten Merkmale lassen sich die Fragmente aus Hattusa in folgende drei Gruppen einteilen: KUB 4, 42 und 43: Die Tafel weist eine tief eingedrückte sogenannte Randleiste am oberen und unteren Tafelrand auf. Die Zahl 9 wird auf die assyrische Art mit den drei übereinanderstehenden senkrechten Keilen geschrieben (KUB 4, 42). Das Zeichen für ICAM zeigt die Ausprägung mit den sieben schräg gestellten Keilen: KAM*. KBo 36,54-1-55'^: Der Schreiber dieser Tafel schreibt das Zeichen KAM in derjenigen Ausprägung, welche wie ein HI-BAD aussieht. Die Zahl 9 erscheint auf die assyrische Art mit drei übereinanderstehenden senkrechten Keilen. KUB 4,46: Bei diesem Text geht jede Zeile über vier Kolumnen: In der ersten steht i-na, in der zweiten der Monat, in der dritten der Tag, in der vierten folgen die Vorschrifien und Deutungen. Das Zeichen KAM zeigt die Ausprägung mit sieben schräggestellten Keilen und wird in der Umschrift demnach als KAM* transliteriert. Die Zahl 9 ist auf dem Fragment nicht erhalten. Der Text schreibt sigû mit §1, während alle anderen Fragmente hierfür das Zeichen SI verwenden. KUB 4, 45 schreibt ebenfalls KAM* mit den sieben schräggestellten Keilen; bei KUB 4, 44 ist keine Kennzeichnung der Ordinalzahl erhalten. Aufgrund der Aufteilung in jeweils zwei Kolumnen und der Schreibung von sigû mit si-gu-u könnten beide Fragmente zunächst dem Schreiber von KUB 4, 42 und 43 bzw. derjenigen Schule zuzuordnen sein, in welcher dieser ausgebildet wurde. Bei KUB 4,44 lassen sich zudem inhaltliche Übereinstimmungen mit derjenigen Hemerologie feststellen, welche von KUB 4,42 (-I-) 43 und von KBo 36 54-)-55 repräsentiert wird, während KUB 4,45 einem anderen Text zuzuordnen ist. Damit lassen sich die sieben hemerologischen Fragmente aus Hattusa aufgrund äußerer und inhaltlicher Kriterien drei verschiedenen Texten zuordnen, von denen zwei'^ im Folgenden vorgestellt werden sollen":

A. Vorläufer zur Opferbrot-^Hemerologie {KAR 178 und Duplikate) In seiner Rezension zu Casaburi, üme täbüti „I giorni favorevoli," hat Livingstone, ZA 96 (2006) 138, darauf hingewiesen, daß der Terminus üme täbüti in Mesopotamien nicht als Titel für die von Casaburi behandelte Hemerologie, sondern in Verbindung mit verschiedenen Hemerologien verwendet wird. Dieser Terminus ist demnach vielmehr als Bezeichnung für (eine Gruppe von)' Hemerologien zu betrachten. Aus diesem Grund soll die von Casaburi bearbeitete Hemerologie im Folgenden in Anlehnung an Livingstone als „Opferbrot-Hemerologie"'* bezeichnet werden. In seiner Rezension hat Livingstone ausdrücklich daraufhingewiesen (S. 139), daß die fragliche Hemerologie weder als „Assur-Hemerologie" (R. Labat: „Hemerologie dAssur,"" und U. Jeyes: „Assur hemerology"^") noch als neuassyrische Hemerologie (Casaburi, üme täbüti, 2) bezeichnet werden kann, weil

15. Die Zusammengehörigkeit beider Fragmente ist bereits von G. Wilhelm im Katalog von KBo 36 festgehalten worden; er konnte damals jedoch den direkten Anschluß, der sich aufgrund des ihm damals unbekannten Inhalts ergibt, nicht feststellen. 16. Für den dritten Text, KUB 4, 46, der mit KUB 43, 1 zu einer Tafel gehört, vgl. den ersten Teil dieser Studie über die hemerologischen Texte aus Hattusa (siehe oben Anm. 1). 17. Die Fragmente wurden mit Hilfe derjenigen Photos kollationiert, welche in der „Konkordanz der hethitischen Texte" im Internet einzusehen sind, vgl. www.hethport.uni-wuerzburg.de/hetkonk/. 18. So in: Matsushima, Official Cult and Popular Religion, 98 mit Anm. 9: „offering-bread hemerology" und ders., SEL 15 (1998) 60 sowie in: Groneberg - Spieckermann, Die Welt der Götter, 86: „Offering Bread Hemerology." 19. R. Labat, Hemerologies et menologies d'Assur (Paris: Libraire dAmérique et d'Orient Adrien-Maisonneuve, 1939). 20. U. Jeyes, JEOL 32 (1991-1992) 30.

z u DEN AKKADISCHEN HEMEROLIGIEN AUS HATTUSA

131

die Existenz des spätbabylonischen Textvertreters aus Uruk {LKU 52), den Casaburi bei ihrer Bearbeitung vernachlässigt, eindeutig die Zugehörigkeit dieser Komposition zum „stream of tradition" beweist. Diese Aussage kann mit Hilfe der Textzeugnisse aus Hattusa sogar noch weiter gefaßt werden, zeigt sich doch, daß die Angaben der fünf Fragmente zwar überwiegend mit denjenigen der Opferbrot-Hemerologie übereinstimmen, es für vereinzelte Abschnitte aber stärkere Übereinstimmungen mit dem sogenannten Babylonischen Almanach gibt (vgl. z. B. die Tage 18-20 des Monats du'üzu ['™gu.NUMUN.A], die in KUB 4,43 Vs. ii' l'-3' behandelt werden). Die in den Fragmenten aus Hattusa vorliegende Version dieser Hemerologie zeigt zudem noch nicht die für die Textvertreter des 1. Jt.s V. Chr. typische Fokussierung auf das Opferbrot (SUKU, kurummatu): der Hinweis auf das Opferbrot findet sich nur in dem Text KUB 4, 45, der aber inhaltlich stark von der Opferbrot-Hemerologie abweicht und deshalb gesondert behandelt wird (siehe unten B.). Die Feststellung Livingstones, SEL 15 (1998) 60, alle Hemerologien hätten sich aus dem sogenannten Babylonischen Almanach entwickelt, findet demnach eine weitere Bestätigung durch die Fragmente aus Hattusa. Weil diese Textzeugnisse jedoch wesentlich mehr Übereinstimmungen mit der Opferbrot-Hemerologie aufweisen als mit dem Babylonischen Almanach, sollten sie als Vorläufer dieser Hemerologie betrachtet werden. In die nachfolgende Partitur-Umschrift werden neben den Fragmenten aus Hattusa auch Textvertreter des 1. Jt.s V. Chr. einbezogen. Bei letzteren wird jedoch in der Regel nur derjenige Text zitiert, welcher die fragliche Parallele so vollständig wie möglich wiedergibt; es ist also keine Vollständigkeit bei den Textvertretern angestrebt worden. Die im Folgenden verwendeten Texte sollen zunächst kurz vorgestellt werden:

Eragmente aus Hattusa, etwa 13. Jh. v. Chr. Die fünf Fragmente dieser Hemerologie aus Hattusa lassen sich vier Tafeln zuordnen. Es handelt sich um KUB 4, 42, KUB 4, 43, KUB 4,44 und KBo 36, 54-1-55. KUB 4,42 (Bo 5408). Einseitig erhaltenes Fragment der rechten Kolumne einer Tafelvorderseite mit unterem Rand, an welchem eine sogenannte Randleiste sichtbar ist. Der Schreiber verwendet das Zeichen KAM*. Der Fundort ist unbekannt. Erhalten sind die Tage 2-9 eines zunächst nicht zu bestimmenden Monats. G. Wilhelm, KBo 36, VI, nahm an, daß dieses Fragment mit KUB 4, 43 zur selben Tafel gehört. Dies ist aufgrund der Aufteilung der Tafel KUB 4, 43 nicht möglich. Da KUB 4, 43 jedoch zu einer Serie von vier Tafeln gehört (siehe hierzu unten), ist es denkbar, daß KUB 4, 42 zu ebenfalls zu dieser Serie gehört, sofern alle diese Tafeln von demselben Schreiber geschrieben woirden. KUB 4,43 (Bo 5513). Beidseitig erhaltenes Fragment mit oberem Tafelrand (Vorder- und Rückseite der Kopie sind zu vertauschen). Die Tafel weist eine sogenannte Randleiste am oberen Rand der Vorderseite und am unteren Rand der Rückseite auf; der Schreiber verwendet das Zeichen KAM*. Der Fundort des Fragmentes ist unbekannt Auf der Vorderseite sind nur drei Zeilen zu erkennen, die wohl dem 4. Monat, du'úzu ('"'"SUNUMUN.A), zuzuordnen sind. Auf der Rückseite sind zwei Kolumnen mit Abschnitten des 5. Monats (Rs. iii'), abu (ITUNE. NE.GAR), und des 6. Monats (Rs. iv'), uMu (""KIN.'ÎNNIN), erhalten.; im Bereich der Kolumne Rs. iii ist KBo 36, 55 ii Duplikat. Bei der Tafel handelt es sich wohl um eine zweikolumnige Tafel, bei der die Monate III (Mitte) bis VI abgehandelt werden. Das bedeutet, daß dies die zweite Tafel aus einer Serie von vieren sein muß; eine andere Tafel dieser Serie könnte mit KUB 4,42 vorliegen (siehe dort). Die Zeilenzahl der Kolumnen läßt sich schwer abschätzen, weil die Vorhersagen für die einzelnen Tage dieser Monate in den Fassungen des 1. Jt. v. Chr. oftmals mehrere Zeilen umfassen, was in begrenztem Umfang auch für die Vorläufertexte gelten kann. Die Kolumnen müssen mindestens 20-22 Zeilen, wahrscheinlich aber ebenso wie KUB 4, 44 an die 40 Zeilen umfaßt haben. KUB 4,44 (Bo 5427). Beidseitig erhaltenes Fragment mit dem oberen und rechten Tafelrand. Die Tafel weist auf der Vorderseite am oberen Tafelrand eine tief eingedrückte Randleiste auf, während eine solche auf der Rückseite

132

JEANETTE G. FINGKE

am unteren Tafelrand nur mit zwei Linien angedeutet wird; der Ton war offensichtlich bereits sehr trocken, als der Schreiber am Ende des Textes angekommen war. Auf dem Fragment ist kein ÎCAM, KAM oder KAM* erhalten. Der Schreiber verwendet das in divinatorischen Texten aus Hattusa, Emar und Ugarit zu findende Zeichen ZA^' für amélu, „Mann," (Vs. 8), statt des geläufigen Logogramms LÚ. Es ist möglich, daß KUB 4,42 zur selben Tafelserie gehört (siehe dort). Der Fundort des Fragmentes ist unbekannt. Erhalten sind auf der Vorderseite Abschnitte des 2. Monats, ajjaru ('™GUD.SLSA), und auf der Rückseite Abschnitte des 4. Monats, du'üzu (''^"SU.NUMUN.NA). Das bedeutet, daß es sich bei dieser Tafel um eine dreikolumnige Tafel handeln muß, in der wohl nur die erste Jahreshälfte einschließlich des Schaltmonats nisannu-2 (•T"BÁR.ZAG.GAR.2.KAM) behandelt wird. In jeder Kolumne dürften etwa 36-38 Zeilen gestanden haben. Im Bereich der Rs. ist die Tafel Duplikat zu KBo 36,54 (-I-) 55 Vs. Kolumne i. Beide Texte wurden von derselben Vorlage abgeschrieben, wie sich an der Verwendung des Zeichens PAP „(Vorlage) zerstört" an derselben Stelle ablesen läßt, vgl. KUB 4, 44 Rs. 1' = KBo 36, 55 Vs. i 3'. KBo 36, 54 (1308/z). Beidseitig erhaltenes Fragment, das im großen Tempel (Tempel I), im Magazin 20 im Bereich der oberen Schufterde gefunden wurde^l Die Form des Zeichens KAM erscheint als Kombination von HI und BAD. An dieses Fragment schließt entgegen den Angaben von Wilhelm KBo 36, 55 unmittelbar an. Erhalten ist in der linken Kolumne der Vorderseite die Zeilenenden der Ge- und Verbote für den 4. Monat, du'üzu Cr^SU.NUMUN.NA), und in der rechten Kolumne das Ende des 5., abu (''^"NE.NE.GAR), sowie der Beginn des 6. Monats, ulülu ('™KIN.''INNIN). In der rechten Kolumne der Rückseite finden sich die Namen verschiedener Monate in ihrer korrekten Reihenfolge, was Rückschlüsse auf eine menologische Abhandlung zuläßt. Es handelt es sich demnach um eine mehrkolumnige Tafel, in der zunächst alle 12 Monate des Jahres hemerologisch behandelt werden, worauf - wohl abschließend - eine menologische Ausdeutung des Jahres folgt. Die einzelnen Kolumnen umfaßten ursprünglich mehr als 60 Zeilen. + KBo 36, 55 (721/z). Einseitig erhaltenes Fragment von der Vorderseite einer Tafel, das im großen Tempel (Tempel I) auf der Nordmauer des Magazins 46 gefunden wurde". Die Form des Zeichens KAM zeigt diejenige Ausprägung, welche wie ein HI-BAD aussieht. Dieses Fragment schließt unmittelbar an KBo 36, 54 an. Erhalten sind in der linken Kolumne Reste des 4. Monats, du'üzu (""SU.NUMUN.NA), und in der rechten Abschnitte des 6. Monats, ulülu ('•^"KIN-''INNIN), Tage 4-16. Im Bereich der linken Kolumne ist das Fragment Duplikat zu KUB 4,44 (siehe dort). Im Bereich der rechten Kolumne ist KUB 4, 43 Rs. iii Duplikat.

Texte aus Assur; neuassyrische Zeit KAR 178 (VAT 10564). Große Tafel mit sechs Kolumnen auf jeder Seite, deren Zerstörungen am oberen und unteren Rand sich unterschiedlich stark auf die einzelnen Kolumnen auswirken. Der Schreiber verwendet die Zeichenform KAM* mit sechs schrägen Keilen. Die Einträge zu den einzelnen Tagen sind durch Paragraphenstriche voneinander getrennt. Das Monatsende ist im Text in der Regel durch eine doppelte Linie gekennzeichnet^^ Der Text ist durchgängig geschrieben, ohne die Monate von den Vorschriften und Deutungen abzusetzen. Die Tafel behandelt jeden einzelnen Tag des Jahres. KAR 179 (VAT 10503). Große mehrkolumnige Tafel, von der auf der Vorderseite die Reste von vier und auf der Rückseite die Reste von drei Kolumnen erhalten sind. Auf dem Fragment ist kein Tafelrand zu erkennen. Der Schreiber verwendet das Zeichen KAM* und beendet die Einträge eines jeden Monats mit dem kryptischen Ko-

21. 22. 23. 24.

Vgl. hierfür ICS 61 (208) 115 mit Anmerkung 25. Vgl. www.hethport.uni-wuerzburg.de/hetkonk/. Vgl ibid. Siehe aber auch die nachfolgende Anmerkung.

z u DEN AKKADISCHEN HEMEROLIGIEN AUS H A T T U S A

133

lophon AL.TIL MAN MAN IGI.KÁR^'. Die Einträge für jeden einzelnen Tag sind durchgängig geschrieben und durch Paragraphenstriche voneinander abgetrennt. + MIO 5, PL 3-6 (VAT 10394). Fragment vom unteren Rand einer Tafel mit mindestens vier Kolumnen auf jeder Seite. Die Einträge für jeden einzelnen Tag sind von denen des nachfolgenden durch einen Paragraphenstrich abgetrennt. Der Text, der sich durch die Verwendung von he-pi „zerstört" als Abschrift erweist, ist durchgängig geschrieben. Der Schreiber verwendet das Zeichen KAM*.

Textvertreter des Babylonischen Almanachs aus Babylonien, neubabylonische Zeit VR 48-49 (76-11-17,2389). Große Bibliothekstafel mit 6 Kolumnen auf jeder Seite; in jeder Kolumne werden die Tage eines Monats behandelt. Statt jeden Tag mit einer Ordinalzahl zu bezeichnen, setzt der Schreiber die einfachen Zahlen ein; die Zahl 9 wird in Anlehnung an die neuassyrische Schreibweise durch drei in Reihe geschaltete schräggestellte Keile gekennzeichnet. Die Einträge für jeden einzelnen Tag sind von denen des nachfolgenden durch einen Paragraphenstrich abgetrennt.

Transliteration und Übersetzung: I. Ein nicht zu identifizierender Monat^* KUB 4, 42 Vs. ii' 1'

'"UD.2''.K[AM* 2. Tag [....].

4,42 Vs. ii'2'

UD.3.KAM* 3. Tag: [...].

4,42 Vs. ii'3'

UD.4.KAM* | ínflD[I Ê/NU.§E] 4. Tag: für einen Proz[eß ist es (un)günstig].

| [...]

KUB 4, 42 Vs. ii' 4'

UD.5.KAM* | NU.S[E ...] 5. Tag: ungün[stig].

KUB 4, 42 Vs. ii' 5'

UD.6.KAM* | GA.R[A§- NU GU^]" 6. Tag: Lauc[h darf er nicht essen].

KUB 4, 42 Vs. ii' 6'

UD.7.KAM* | SILA 7. Tag: auf eine Straße soll er ni[cht gehen].

25. Dieser kryptische Kolophon findet sich auch einmal auf der Tafel KAR 178 Vs. iv 40. AL.TIL bedeutet „beendet" und IGLKAR „kollationiert." Die Zeichen MAN MAN gehen auf den in der altbabylonischen Zeit ursprünglich verwendeten einfachen Winkelhaken zurück, der auf eine Doppellinie oder zwischen zwei Linien geschrieben das Ende einer Textsequenz auf einer Tafel markiert; vgl. z. B. VS II 1, 4, 26, 31, 49, und weitere Texte in diesem Band. 26. Diese Passage könnte den Monaten abu (""NE.NE.GAR) (V) oder kislimu ('™GAN.GAN.NA) (IX) zuzuordnen sein, denn für die entsprechenden Tage dieser Monate ist die neuassyrische Überlieferung lückenhaft. Mit anderen Monaten gibt es keine Übereinstimmungen über mehrere Tage hinweg. 27. Vgl. z. B. KUj ù GA.RAS NUGU, „Fisch und Lauch darf er nicht essen" (KAR 178 Vs. i 8 [se. Monat nisannu; '™BAR.ZAG.GAR; 1. Tag] ), '"GA\RAS^*'> NU Gü^ „Lauch darf er nicht essen" (KAR 176 Vs. i 11 ' [se. Monat nisannu, '™BÄR.ZAG.GAR; 2. Tag]), KUg GA.RAS SAR [NU].GU7 „Fisch und Lauf darf er [nicht] essen" (KAR 178 Rs. iii 55 [sc. Monat taSritu; '"DU^.KÜ; 7. Tag]), SUHUS GA.[RAS]S*'' N U . G U , „die Wurzel des Lauches darf er nicht essen" (KAR 178 Rs. iii 57 [sc. Monat tasritu; '™DUj.KÜ; 7. Tag]). 28. VgL iC[/ß 4, 44 Vs. 5.

134

JEANETTE C. FINCKE

KUB 4, 42 Vs. ii' T

UD.8.KAM* |SE[...] 8. Tag: günstig; [...].

KUB 4, 42 Vs. ii' 8'

UD.9.KAM* 9. Tag: guns [tig...].

KUB 4, 42 Vs. ii'

(Ende der Kolumne)

IL Monat ajjaru ("^GUD.SI.SA)

KUB 4,44 Vs. 1

[UD.20.KAM*

MU]S' (Text:]x-TI) ta-'GAZ'-ak UR.BI te-leq-

[20. Tag: eine Schlan]ge wirst du töten, (dann) wirst Du (alles) insgesamt nehmen (können). KUB 4, 44 Vs. 2

[UD.2I.KAM* I KI.MI]N- dá-sa-a-tú [21. Tag: Desgleich]en'. Respektlosigkeiten.

KUB 4, 44 Vs. 3

[UD.22.KAM* ] \'dá''-sa-a-tú'N\J.Ê [22. Tag]: Respektlosigkeiten; ungünstig.

KUB 4, 44 Vs. 4

[UD.23.KAM* ] I DU a-bu-ti §E [23. Tag]: Ausüben der Vaterschaft; günstig.

KUB 4, 44 Vs. 5

[UD.24.KAM* ] I SILA NU.DIB NU.Ê [t]UD.[24.KAM*...] [24. Tag]: auf eine Straße soll er nicht gehen; ungünstig.

KAR 179H-Vs. iii 1'

KUB 4, 44 Vs. 6 KAßl79+Vs.iii2'-3'

[UD.25.KAM* ] I DI SE KI.MIN DAM-5Ú ana É-sú KU^ É-sú DAGAL-is [1] UD.[25.KAM* (1-6 Zeichen)] §E / ^KI.MIN DAM''-s[u] ana É-s[u KU, É].BIDAGAL [25. Tag]: für ein Rechtsverfahren ist es günstig. Desgleichen: er soll seine Ehefrau in sein Haus eintreten lassen, (dann) wird sein Haus sich ausbreiten.

KUB 4, 44 Vs. 7 KAR 179-f Vs. iii 4'-5'

[UD.26.KAM* ] I NU.SE KI.MIN DAM TUK KI.MIN si-gu-u [1] UD.26.KAM* NU.SE KI.MIN DAM TUK / se-gu-u [26. Tag]: ungünstig. Desgleichen: er soll eine Ehefrau nehmen. Desgleichen: ein sigû-Gehet

KUB 4, 44 Vs. 8 KAR 179-1- Vs. iii 5'

[ \] ''NU"' i-sfl-5i ZA.BI SUMUN-fcflr NU DÉ NA.BI SUMUN soll er nicht,rufen,' (dann) wird der betreffende Mann alt werden.

KUB 4,44 Vs. 9

[UD.27.KAM*

I $]A'.HÜL'(Text: §Ú-§EN).LA KI.MIN HAL

29. Vgl. den babylonischen Almanach für diesen Tag {VR 48 Vs. ii 23-24): 20 MUS HÉ.EN.GAZ (24) SAG.KAL DU-aA:, „eine Schlange sollst Du töten, (dann) wirst Du (unter den Menschen) den ersten Platz einnehmen," vgl. R. Labat, Un almanach babylonien {VR 48-49) (Paris: Libraire dAmérique et d'Orient Adrien Maisonneuve, 1943), 13.

z u DEN AKKADISCHEN HEMEROLIGIEN AUS HATTUSA

135

KAR 179-t- Vs. iii 6'

UD.27.KAM* NU.SE KI.MIN SA.HÚL.LA [27. Tag]: [Fr]eude. Desgleichen: Geheimnis (KAR 179+: 27. Tag: ungünstig. Desgleichen: Freude).

KUB 4, 44 Vs. 10 KAR 179+ Vs. iii 7'-

[UD.28.KAM* I a-k]aV kar-si KI.MIN AS KI.3 ^si-gu'^-[u] UD.28.KAM* §E KI.MIN nak-Ha-a^-[t]e I KI.MIN si-gu-u [28. Tag: Ver]leumder. Desgleichen: es ist günstig. Zum Dritten: ein 5/^wGebet (KAR 179+: 28. Tag: günstig. Desgleichen: kunstvolle (Dinge). Desgleichen: ein sigû-Gebei) [

I NU] '"r-sa-5)' DU,, -"EGIR? ^-sú [0]

179-1-Vs. iii 8' soll er [nicht] ,rufen,' (sonst gibt es) Streit danach (KAR 179+: (sonst) wird Streit für ihn vorhanden sein). ilR178Vs.vi 28'

[UD.9.KAM* I ...i]-sa-ti [UD.9.KAM* I ...xxIZ]I H UD.9.KAM* ni-pi-ih IZI [9. Tag: Entzünden von F]euer (KAR 178: Entzünden von Feuer).

KLrß4,44Rs. 3' KBo 36, 55 Vs. i 5' KAR 178 Vs. vi 3O'-31'

[UD.IO.KAM* [

I ... xxxjx" [SET' ]. (KAR 178: [günstig']).

I ...]$E I ...xx$]E

[II UD.IO.BCAM* ... ina ] 'DV [§E] / [èVK-su ana \..] ""GAR-ma ma-hir''''

[10. Tag: ...] günstig (KAR 178: [für ein] Gerichtsverfahren [ist es günstig], [das Opferbrot für die Gottheit...] soll er hinstellen, (denn) es ist genehm). 4, 44Rs. 4' KBo 36, 55 Vs. i 6' KAR 178 Vs. vi 32'-33'

[UD.II.KAM*

KLfß4,44Rs. 5' Kßo 36, 55 Vs. i 7' K>1R 178 Vs. vi 34'

[UD.12.KAM*

I gf-li'-t]u^''n.Bl.ZA IGI™ >U
NU< lil-bi'-i[b]), wodurch sich dieser Text von den anderen hemerologischen Texten aus Hattusa unterscheidet. Auch die Schreibung der Prekativform von sasû mit dem Zeichen LfL findet sich in den entsprechenden anderen Texten nicht; der Schreiber von KL'^ß 4, 46 (+) KL^ß 43, 1 Vs. i verwendet das Zeichen LIL^". Nur für die Schreibung von sigû mit si-gu-u (Vs. 5', 8') finden sich Parallelen in KUB 4,44 (Vs. 7,10)^'; weil jedoch hier das Verbot (si-gu-u NU i-sa-si) steht, können keine weiteren Übereinstimmungen - z. B. die Verwendung von LIL oder LÍL - die Vermutung, es handele sich um denselben Schreiber oder dieselbe Schultradition, erhärten. Unter den Geboten dieser Hemerologie finden sich Anweisungen, bestimmten Göttern ihr Opferbrot (§UK. BI) hinzustellen (siehe Vs. 4', 11' und Rs. 4'). Damit ist die Verwandtschaft mit der sogenannten „Opferbrot-Hemerologie" und deren Vorläufer (siehe oben A.) offensichtlich. Der Textvertreter KUB 4, 45 scheint jedoch einer anderen Tradition anzugehören, da sich keine Parallelen mit der „Opferbrot-Hemerologie" des 1. Jt.s v. Chr. oder ihren Textvertretern aus Hattusa fmden lassen. Für einige Apodosen finden sich zudem keine Entsprechungen in anderen Hemerologien, wohl aber in Omentexten; ein Ergänzung der Vorhersagen bleibt damit zumeist unsicher. Trotz dieser Schwierigkeiten soll hier - soweit möglich - eine Bearbeitung geboten werden.

Transliteration: KUB 4, 45 Vs. 1' KUB 4, 45 Vs. 2' KUB 4,45 Vs. 3' KUB 4, 45 Vs. 4' KUB 4,45 Vs. 5' KUB 4, 45 Vs. 6' KUB 4, 45 Vs. 7' KUB 4,45 Vs. 8' KUB 4,45 Vs. 9' KUB 4, 45 Vs. 10' KLrß4,45Vs. 11' KL^ß 4,45 Vs. 12' KUB 4, 45 Vs.

[ \si-g]u-[u [ | DING]IR TNUÂS KI.MIN xix'"[UD.x.KA]M* | DINGIR ib-bal-^ma DINGIR-sú ¡-fe-xx xx xx^ [ [UD.x.KAM*] | SUKU.BI ana DINGIR'""-sú SUM A§ [ ] | KI.MIN si-gu-u lil-si-^TÚG-sú >NU< lil-bi'-i[b]" [ ] | \JD'i"'-sú '"GÍD"'.DA [O'] [UD.x.KAM* | D]AM NU.TUK gi-ih-lu-u GIG [ [ | GAR]-an AS KI.MIN si-gu-u lil-s[i [ | xx xx x]x^' HA MU-iii (i radiert') SIG5 [izakkar'] [ | AS'" KI.MI]N ina DI §E KI.MIN [ [ I XX XX XX x]xSUKU ana rDINGIR"'-[5w [ | xxxxxxxxx]xx[x (Rest der Kolumne ist weggebrochen)

50. Für diesen Text vgl. J. C. Fincke, JCS 61 (2009) 121-24. 51. Siehe oben Monat ajjaru (II), Tage 26 und 28. 52. Dieses Zeichen beginnt mit einem horizontalen Keil. 53. Der Keilschriftkopie zufolge steht hier úllil-na-[...]. Leider lassen sich die Zeilen anhand des in der „Konkordanz der hethitischen Texte" im Internet (www.hethport.uni-wuerzburg.de/hetkonk/) publizierten Photos nicht kollationieren, denn ein Photo des rechten Randes der Tafel liegt dort nicht vor. Dem Photo der Vorderseite ist jedoch zu entnehmen, daß der rechte Rand an dieser Stelle abgerieben ist. Es wäre also durchaus möglich, daß F. Köcher die Zeichenspuren falsch interpretiert und seine Interpretation - versehentlich - als vollständig vorhanden kopiert hat. 54. Dieses Zeichen endet auf einen vertikalen Keil. 55. Diese Ergänzung gründet sich ausschließlich auf den vorhandenen Raumverhältnissen vor dem Bruch.

144 KUB 4, 45Rs. V KUB 4, 45 Rs. 2'

JEANETTE C. FINCKE [ [

\ ...xxx]x[ | ... T]IR/ x]x-NIRpfl-'"«a='^"'-x[x

KUB 4, 45 Rs. 3'

[

| ...] '"w"' "120 SE KI.''MIN"' [

K[/ß4,45Rs. 4' KLiß4,45Rs. 5' KUB4,45Rs. 6' Kl7ß 4, 45 Rs. 7' KUB 4,45 Rs.

[ |...]flnfl""'-ííí7!-eaPGAR-''fln§I-xxM [ | ... xxx]x'"KP.MIN/ífl-W-û ana A.KÁR.[KÁRKU,...] [ | ... xxxxx]x'"/e-ê"'-eIZx[x [ | ... xxxxxxx]x (Spuren) (Rest der Kolumne ist weggebrochen)

Übersetzung: "'•'[... einsi]gû-[Gehet...]. ^'- ^ [... : der Got] t ist ungünstig (gestimmt) ; dito .[...]. ^' ' [Am x-te]n [Tag]: er soll den (persönlichen) Gott anflehen (sie!), dann wird sein Gott...[...]. ^"'[...:] er soll seinem Gott dessen Opferbrot geben, es ist günstig; ' [0] dito: er soll ein sigü ,rufen' (und) sein Gewand >nicht< rein[igen,] ** [0] (dann) werden seine Tage lang werden. ^'-'[...: eine Eh]efrau darf er nicht nehmen, (sonst gibt es) Trauer, Krankheit; [...] *[soll er hinstel]len, es ist günstig; dito: er soll ein sigü ,ru [fen'... ]. ^'^ '[...:...].., sein Name wird zum Guten [genannt werden]. ^'''°[... : es ist günstig; dit] o: für einen Prozeß ist es positiv; dito: [...]. ^ ' " [...: ...]. das Opferbrot für [seinen] Gott [soll er geben]. ^•"2 (Spuren) (Rest der Vs. ist weggebrochen)

"'^ [...:...] und Samas ist es günstig; dito [...]. '*^''[...:...]stellst du Venus hin;..[...]. •^"^ [...: ...].; dito; Verlust des Fischlai[ches . . . ] . "'•"^ [...: ...] das Nehmen von ..[...]. "' ^ (Spuren) (Rest der Rs. ist weggebrochen)

Kommentar: Vs. 3': Der hethitische Schreiber hat statt i-bal ib-bal geschrieben, was auf einen Hörfehler zurückzuführen ist. Vs. 7': gihlü, „Trauer (um einen Verstorbenen)," erscheint als Vorhersage in Omentexten, nicht aber in Hemerologien. Vs. 9': Wenn das Zeichen / nicht radiert ist, könnte es sich um eine halblogographische Schreibung für sum-su idammiq, „sein Name wird ,gut' werden," handeln. In Hemerologien findet sich jedoch nur der Ausdruck „sein Name wird zum Guten genannt werden" in der Schreibung MU-iti ana SIG5(-íe) MU.

56. Die Kopie zeigt ein deutliches UD, aber dem Photo zufolge steht hier NA. 57. Diese außergewöhnliche Schreibung für Venus konnte m. W. bislang nocb nicbt nachgewiesen werden; vgl. J. C. Fincke, N.A.B.U. 2008/1, 21-22 Nr. 13 zur Stelle.

z u DEN AKKADISCHEN HEMEROLIGIEN AUS HATTUÂ

145

Rs. 5': Entsprechende Vorhersagen mit Bezug auf den Fischbestand finden sich sonst nur in astrologischen Omentexten mit Bezug auf Mondfinsternisse, z. B. ACh Sin 25 Z. 63: DI§ ina '™NE U D . 2 I . K A M AN.GE^ GAR A.GAR.GAR KU^ ina A.AB.BA Z [ A H ] SU.GU^ [...], „Wenn im Monat abu am 21. Tag eine Mondfinsternis stattfindet, wird der Fischlaich im Meer vern[ichtet] werden - Hungersnot [...];" LKU119 Vs. 20: [DIS U ] D . 2 1 . K A M KI.MIN BA-w KI.MIN A.GAR.GAR KU^ ina .A[B.BA ...], „[Wenn] am 21. [Tjag dito: Minderung; dito; [wird] der Fischlaich im Me[er' ...];" ACh Sin 34 Z. 27: DIS UD.2LKAM a-gar-ga-ru-tu, hi-sib A.AB.BA Z A H , „Wenn am 21. Tag (dito), wird der Fischlaich, Produkt des Meeres, vernichtet werden" (EAE TF. 22 I § V.5, vgl. Rochberg-Halton, AfO Beih. 22,1988,257).

Copyright of Journal of Cuneiform Studies is the property of American Schools of Oriental Research and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.

Journal of Cuneiform Studies (JCS 62) 2010

Journal of Cuneiform Studies (JCS 60) 2008

Journal of Cuneiform Studies (JCS 61) 2009

Journal of Cuneiform Studies (JCS 55) 2003

Journal of Cuneiform Studies (JCS 56) 2004

Journal of Cuneiform Studies (JCS 58) 2006

Journal of Cuneiform Studies (JCS 57) 2005

Journal of Cuneiform Studies (JCS 59) 2007.pdf

Journal of Cardiology Cases (2010)

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

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

New Left Review 62 Mar-Apr 2010

The Journal of Antimicrobial Chemotherapy (Volume 62, Number 6, 2008)

European Coatings Journal 2010-11

International Journal of Diabetes Mellitus, December 2010

European Coatings Journal 2010, 06

Scientific American 03 2010 (journal magazine; March 2010)

Scientific American 10 2010 (journal magazine; October 2010)

Scientific American 08 2010 (journal magazine; August 2010)

Nature Biotechnology 02 2010 (magazine journal; February 2010)

Nature Biotechnology 10 2010 (magazine journal; October 2010)

Scientific American 06 2010 (journal magazine; June 2010)

Nature journal (2010-12-23 to 2010-12-30)

Nature Biotechnology 04 2010 (magazine journal; April 2010)

Nature Biotechnology 08 2010 (magazine journal; August 2010)

Scientific American 07 2010 (journal magazine; July 2010)

Nature Biotechnology 03 2010 (magazine journal; March 2010)

Scientific American 09 2010 (journal magazine; September 2010)

Nature Biotechnology 09 2010 (magazine journal; September 2010)

Nature Biotechnology 01 2010 (magazine journal; January 2010)

Nature Biotechnology 12 2010 (magazine journal; December 2010)

Journal of Cuneiform Studies (JCS 62) 2010

Journal of Cuneiform Studies (JCS 60) 2008

Journal of Cuneiform Studies (JCS 61) 2009

Journal of Cuneiform Studies (JCS 55) 2003

Journal of Cuneiform Studies (JCS 56) 2004

Journal of Cuneiform Studies (JCS 58) 2006

Journal of Cuneiform Studies (JCS 57) 2005

Journal of Cuneiform Studies (JCS 59) 2007.pdf

Journal of Cardiology Cases (2010)

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

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

New Left Review 62 Mar-Apr 2010

The Journal of Antimicrobial Chemotherapy (Volume 62, Number 6, 2008)

European Coatings Journal 2010-11

International Journal of Diabetes Mellitus, December 2010

European Coatings Journal 2010, 06

Scientific American 03 2010 (journal magazine; March 2010)

Scientific American 10 2010 (journal magazine; October 2010)

Scientific American 08 2010 (journal magazine; August 2010)

Nature Biotechnology 02 2010 (magazine journal; February 2010)

Nature Biotechnology 10 2010 (magazine journal; October 2010)

Scientific American 06 2010 (journal magazine; June 2010)

Nature journal (2010-12-23 to 2010-12-30)

Nature Biotechnology 04 2010 (magazine journal; April 2010)

Nature Biotechnology 08 2010 (magazine journal; August 2010)

Scientific American 07 2010 (journal magazine; July 2010)

Nature Biotechnology 03 2010 (magazine journal; March 2010)

Scientific American 09 2010 (journal magazine; September 2010)

Nature Biotechnology 09 2010 (magazine journal; September 2010)

Nature Biotechnology 01 2010 (magazine journal; January 2010)

Nature Biotechnology 12 2010 (magazine journal; December 2010)

Recommend Documents