On the Euphrates, ZA 95

On the Euphrates 0On 7 the Euphrates*0 by Christopher Woods – Chicago In seeking to recover the semantic relationships that led the Euphrates and Sippar1 to share the writing ud.kib.nun, this paper begins with an analysis of the earliest writings for river and city. It is found that the essential elements of this Diri compound, kib.nun, first designated the divine Euphrates; only as a secondary development did the city borrow the spelling from the river. Further, it is suggested that the writing ud.kib.nun belongs to the ud.gal.nun orthographic tradition. As for the city taking its spelling from the river, the explanation lies, on one hand, in the functional overlap of the Sun- and River-gods and, on the other, in the unique topography of the Sippar region. The aspect of divine judge defines both the Sun-god, the patron deity of Sippar, and the River-god, the divine Euphrates being a particular manifestation of d í d /dNaru – indeed, the textual and artistic records together describe a mythological and cosmographical conception that links these two deities. The Sippar region, it is argued, was an early cult center of the River-god and was regarded as a numen loci on account of the unique geomorphological conditions specific to the area. It is a curious fact of Mesopotamian toponymy that the Euphrates, B u r a n u n a /Purattum, the great artery of Mesopotamia, and Sippar, the renowned ancient cult center of the Sun-god, Šamaš, share a common logographic writing, ud.kib.nun, regarded by the native lexical tradition *0 I am grateful to the editors of ZA, Walther Sallaberger and Ursula Seidl, for their many corrections and suggestions from which this manuscript has greatly benefited. I also owe a debt of gratitude to Piotr Steinkeller, who read through several versions of this paper, making numerous suggestions and offering keen insights. Further, I would like to thank Robert Biggs, Miguel Civil, Robert Englund, Gene Gragg, Jennie Myers, Dennis Pardee, Seth Richardson, Martha Roth, and Gil Stein for their invaluable comments and assistance. Portions of this study were presented at the 214th meeting of the American Oriental Society (San Diego, CA March 13th , 2004). The abbreviations used are those of E. Reiner/M. T. Roth, The Assyrian Dictionary of the Oriental Institute of the University of Chicago, vol. R (Chicago 1999) ix–xxvii, and/or Å. W. Sjöberg, The Sumerian Dictionary of the University of Pennsylvania Museum, vol. A/3 (Philadelphia 1998) ix–xlii, with the following additions: ASJ = Acta Sumerologica. OBO = Orbis Biblicus et Orientalis. QdS = Quaderni di Semitistica. 1 Throughout this paper I maintain the traditional pronunciation /sippar/, although, at least for the OB period, sources indicate a pronunciation /sip(p)ir/ (see RGTC 3 s. v.). Zeitschr. f. Assyriologie Bd. 95, S. 7–45 © Walter de Gruyter 2005 ISSN 0084-5299 8 Christopher Woods as a Diri compound.2 It is a peculiarity of Sumerian writing that has largely escaped scholarly musings as to why this is so.3 Yet this orthographic identity must, no doubt, have had a meaningful semantic basis for those who laid the foundations of the writing system, although we have barely scratched the surface in terms of understanding the relationship between graph and meaning that must underpin such writings. Thus, at the root of our problem is the broader question of sign choice in Diri compounds, specifically, the rationale for grouping together certain graphs, the phonetic values of which have no obvious relationship to that of the composite. By the Old Babylonian period, when a systematized collection of these compounds was first integrated into the scribal repertoire, many were already very ancient writings and were most likely regarded as completely arbitrary spellings, the association of ideas that engendered each during the infancy of the writing system having long been forgotten. In this way the plight of the modern scholar is not so different from that of the Old Babylonian compilers of Proto-Diri. The problem at hand, in our case, is to recover the semantic associations that led to the Euphrates and Sippar sharing the spelling ud.kib.nun and to the choice of these particular signs to express these two toponyms. Sippar, of course, lay on a major branch of the Euphrates, but this claim could be made for any number of Mesopotamian cities whose orthographies have no relationship to the great river. Arguably, one might expect this shared orthography to be based on a particular relationship between Šamaš, the patron deity of Sippar, and the Euphrates. Indeed, an Eblaite incantation that invokes Šamaš also mentions the waters of the Euphrates,4 while a late text unequivocally attributes the Euphrates to the Sun-god: Idiglat sˇa ina maäar Enlil izziz(z)u … Puratti sˇa ina maäar Sˇamasˇ izziz(z)u “the Tigris which served Enlil … the Euphrates which served Šamaš.”5 However, this is the extent of the explicit association in the textual record, and in a much earlier Sîn-iddinam inscription it is the Tigris, 2 3 4 5 The following entries occur in Diri (cited according to the unpublished ms. of M. Civil; now MSL 15 [2004]), Zimbir: Diri I 142; Diri Nippur 328; Diri Ugarit I 121; Buranuna: Diri III 180, 197; Diri Nippur 347; Diri Sippar vi 5. As will be made clear below, the reading of the river is / b u r a n u n a / , not / b u r a n u n / . The only discussion of this orthographic phenomenon that I am aware of is the brief comment of R. McC. Adams, Heartland of Cities (Chicago 1981) 3, who is followed by F. Carroué, Études de Géographie et de Topographie sumeriennes II. À la Recherche de l’Euphrate au IIIe Millénaire, ASJ 13 (1991) 120–121. a bù-la-na-tim dùg “waters of the sweet Euphrates” (ARET 5, 3 rev. iv 1–4). RA 60 (1966) 73: 7´, 10´. On the Euphrates 9 in fact, that is conceded to Utu: Í d - I d i g n a í d h é - g á l - l a d U t u - k e 4 “the Tigris, the river of Utu’s abundance.”6 As to the question of orthographic primacy, here too the answer is more complicated than the evidence at first glance suggests. The Euphrates has been described as the “river Sippar” by Adams, Carroué, Edzard, and Nissen, among others.7 Behind this moniker is an understanding of Sippar as primary from the perspective of writing, the city lending its spelling to the river. Certainly, there are valid parallels in terms of the names of various third-millennium and OB waterways to justify this assumption. Writings of the type Í d - A d a b , (d) Í d - A k š a k ki , Í d - B à d ki , Í d - G í r - s u k i , G ú / Í d - E b - l a ki , Í d - Ì - s i - i n ki , d Í d - K i š ki , Í d - L a g a š ki , Í d - N i n a ki ( - š è ) - d u ( - a ) , Í d - Ú r i m ki - m a ,8 in which the names of canals and rivers derive from the settlements on their courses, would appear to serve as corroborating evidence. And confirmation would seem to come in the form of the common writing with the determinative ki, i. e., Í d - ud.kib.nunki , occurring as early as the Gudea corpus.9 Already at the end of the Early Dynastic period, as given by a Lugalzagesi inscription from Nippur,10 the writing for the city, ud.kib.nunki , must refer to the Euphrates based on context. Earlier at Abu Salabiä the Euphrates is written both ud.kib.nun and ud.kib.nunki as witnessed by variants to a Ninbilulu Z à - m í hymn,11 although elsewhere in this corpus ud.kib.nun, alone, stands for the Euphrates.12 By the Ur III period, the Euphrates was regularly construed with the ki determinative, i. e., Í d ud.kib.nunki , leaving little doubt that at the close of the third millennium, and perhaps earlier, scribes understood the Euphrates in these cases to be the “Sippar River” at least in terms of orthography. Later evidence is explicit on this point. The OB writings Í d -ud.kib.nunki-ri-tum and Í d -Si-ip-pi-ri-tum, i. e., the river Sippiritum, demonstrate conclusively 6 7 8 9 10 11 12 D. Frayne, RIME 4, 160: 39–40. It is perhaps the Larsa-centric perspective of this text, which opens with a mention of the building of the Ebabbar, that influenced the choice of this epithet. Adams, Heartland of Cities 3. 159; R. McC. Adams/H. J. Nissen, The Uruk Countryside (Chicago 1971) 44–45; F. Carroué, ASJ 13 (1991) 120–121; D. O. Edzard, Iturungal, RlA 5 (1976–1980) 223. RGTC 1–3, s. vv. Cyl. B x 20. xvii 10. BE 1/2 87 iii 2 f. (= Steible ABW, Luzag. 1 ii 7). OIP 99, p. 48: 61. 62. Only source E adds ki; in the other three witnesses (four for l. 62) Buranuna is written ud.kib.nun. E. g. OIP 99, 388 ii 4. 393 ii 4´. 10 Christopher Woods that in the second millennium the city could lend its name to the river at least in the vicinity of Sippar.13 This understanding is also reflected in the vast majority of the lexical attestations which include the ki determinative in the writing of the river.14 Yet as compelling as this argument may be, a review of all the evidence paints a very different picture – namely, that in origin the river represents the primary or unmarked writing and that it is the river which lent its spelling to the city. Indeed, contrary to the common assumption, Sippar is the “Euphrates city” from the perspective of orthography. In fact, this is what one would expect based on the cross-cultural observation that settlements tend to derive their names from topographical features rather than vice-versa. First, whereas the logographic writing for Sippar is stable in all periods,15 representing a fixed and, I shall argue, borrowed compound, the writing for the Euphrates shows considerable variation in third- and second-millennium texts, given that often the ud and occasionally the nun graphs could be omitted freely in the writing of the river. This demonstrates that these graphs are complements in the writing of Buranuna and that the essential element of the compound is kib.16 In second-millennium texts the writing kib.nun(.na) occurs sporadically in the south,17 but with greater frequency in the north, e. g., at Sippar,18 Emar,19 and particularly at Mari,20 prompting Charpin to state, “On sait qu’à Mari au IIe millénaire, le nom de l’Euphrate est généralement écrit kib.nun.na (et non ud.kib.nun.na).”21 In the Meturan exemplars of the Death of 13 14 15 16 17 18 19 20 21 See R. Harris, Ancient Sippar (Leiden 1975) 380–381. E. g., Í d -ud.kib.nunki = B u r a n u n a , Purattum (Diri III 180; Diri Sippar vi 5; Diri III 197; but note the earlier OB Í d -ud.kib.nun.na = Purattum [Diri Nippur 347], the significance of this plene spelling is discussed below); additionally, canonical Hh. XXII (and associated recensions and lists), as well as the OB forerunners, write Buranuna with the KI determinative. The one exception that I am aware of, and probably a scribal mistake, occurs in the spelling of the deity d L u g a l - Z i m b i r ki , written without nun, i. e., ud.kibki.e, in Or. SP 47/49 (1939) 369; see below for the interpretation of this deity as Lugal-Zimbir, rather than Lugal-Buranuna. For the use of this sign to write Buranuna, see C. Woods, The Paleography and Values of the Sign kib, in: Fs. R. Biggs (forthcoming). Lament for Sumer and Ur 25, 38 (texts DDa and DD from Ur). CT 4, 17c: 6 (cf. 1. 7 where the same PN is written with ud) and perhaps JCS 11 (1957) 29 no. 17 rev. 4. Aula Or Supp. 1 (1991) 50 no. 19: 4. ARM 1, 62: 17; ARM 2, 131: 11. 37; ARM 10, 155: 16; see ARM 15, p. 85. D. Charpin, Tablettes présargoniques de Mari, MARI 5 (1987) 72. On the Euphrates 11 Gilgamesh, kib.nun.na is written for Buranuna without exception.22 But rather than representing a corrupt, relatively late peripheral spelling, it is clear that this writing has deep roots in the third millennium. The writing Í d - nun.kibki at Umma23 shows that the ud element was omissible in the Ur III period; and earlier at the end of the Pre-Sargonic period there is the Nippur personal name U r - s a g -a.kib.nun(ki),24 a writing for the Euphrates that also occurs on the Barton cylinder.25 At Ebla we encounter the spelling kib.nun.a for the Euphrates, a spelling, as we shall see, that has antecedents at Fara and Mari.26 Finally, in another Ur III text from Umma Buranuna is written simply as kib,27 substantiating the claim that both the ud and the nun graphs are dispensable complements in the writing of the river. Curiously, there are no certain attestations of the Euphrates or any other watercourse in the Uruk texts.28 Perhaps this is simply a matter of isolating the graphs that conceal them – a task complicated by the fact that the determinative íd first appears in the Early Dynastic period. But references to the Euphrates are surprisingly rare in all periods, even in the vast corpus of Ur III administrative texts. The reason for this lies, at least in part, in the fact that B u r a n u n a /Purattum refers to the Euphrates river system in toto; not specifying any particular branch, the des22 23 24 25 26 27 28 A. Cavigneaux/F. N. H. Al-Rawi, Gilgameš et la mort. Textes de Tell Haddad VI. Cuneiform Monographs 19 (Groningen 2000) ll. M 241. M 242. M 247. M 249. MVN 16, 789: 9. Westenholz OSP 1, 31 i 3´ and TuM 5, 55 i 4 [= Westenholz Jena 56]). As in other third-millennium contexts, the graph a is presumably for i d 5; note the GN ( Í d - ) a - s u h u r (ki) (Steible ABW, Ean. 2 vi 19; ITT 6431: 6), written I d 5 (a)- a - s u h u r in Steible ABW, Urn. 26 iii 7 (see RGTC 1, 208); also i d 5 (a)- d u n “canal digger,” š u k u 6 - i d 5 (a)- d u n - a “fisher of the excavated canal” (see G. J. Selz, FAOS 15/2, p. 236, with references). Barton MBI 1 xiii 3. xvii 2 (according to the numbering of B. Alster/A. Westenholz, The Barton Cylinder, ASJ 16 [1994] 15–46). ARET 13, 15 iv 9. v 17 (reference courtesy of P. Steinkeller). g ú Í d - B u r a n u n a x (kib)- t a (Sigrist Princeton 347: 7). As pointed out by R. Englund (personal communication). It is doubtful that the graph kib occurs in the Uruk period contra M. W. Green/H. J. Nissen, ZATU 230 no. 290; thus the Euphrates must be sought under a different graphic composition than that of later periods (see Woods, Fs. R. Biggs [forthcoming]). The identification of idigna in the Uruk texts (ZATU 261) is similarly uncertain; further, the reading of the sign in question is complicated by the likelihood that in at least some cases it appears to designate a (water?-)bird, as the original pictograph suggests, e. g., idigna:ku6:a (MS 2862); cf. idignamušen (Deimel SF 58 x 20 [bird list]). R. Englund, however, understands the Tigris to be attested in the Uruk IV text W 9579, see Texts from the Late Uruk Period, in: P. Attinger/M. Wäfler (eds.), Mesopotamien: Späturuk-Zeit und Frühdynastische Zeit. OBO 160/1 (Freiburg 1998) 75 n. 148. 12 Christopher Woods ignation is largely limited to poetic uses. In administrative contexts especially, expediency required reference to the local name of the river, e. g., Í d - A d a b . Carroué has recently given evidence for the writing of the Euphrates in the third millennium, beginning with the Abu-Salabiä and Pre-Sargonic Nippur attestations. Overlooked in his study, but certainly to be included here, is the divine name d K i b - n u n , well-attested at Fara and the earliest identifiable writing for Buranuna. The designation refers to the divine Euphrates, a manifestation of the numinous quality of the river. The deity occurs most frequently in the personal name dkib.nun(.a)- u r s a g , a forerunner of the Pre-Sargonic PN U r - s a g - a.kib.nun(ki). The name appears no fewer than fourteen times in Fara administrative texts. In one case nun is omitted, i. e., U r - s a g - dkib – a writing that finds its counterpart in the Ur III spelling B u r a n u n a x (kib) (Sigrist Princeton 347: 7) discussed above and, possibly, in a Sargonic sealing which may read Í d - ud.kib.29 At first glance one might suspect that dkib.nun(.a) is a writing for Utu, given the Fara PN U t u - u r - s a g , but the co-occurrence of both PNs in at least eight administrative texts (in adjacent entries in one text)30 makes it most unlikely that these are variant orthographies of the same name. Decisive evidence that dkib.nun denotes the divine Euphrates comes from the Fara lexical text Deimel SF 72, which includes an enumeration of waterways. Included in this list are the consecutive entries [idig]na.min and dkib.nun.min (iv 13–14), which, incidentally, gives the regular ordering of this pairing, the Tigris followed by the Euphrates, commonly encountered in later texts.31 The god also appears in 29 30 31 See Pomponio Prosopografia 137 (nun omitted in Deimel WF 48 ii 7). Note the Fara PN U r - h u r - s a g without divine determinative but nevertheless reflecting the numinous character of the mountain ranges (Pomponio Prosopografia 252). As for the Sargonic seal, D. O. Edzard reads Í d - ud.kib ? (AfO 22 [1968–69] 15 no. 16: 9; see L. de Clercq/J. Menant, Collection de Clercq. Catalogue méthodique et raisonné 1 no. 44; RGTC 1, 209) – the sign is admittedly difficult. Deimel WF 9. 12. 13. 15. 53. 76. 78; Jestin Šuruppak 100. For this text, see M. Krebernik, Die Texte aus Fara und Tell Abu Salabikh, in: P. Attinger/M. Wäfler (eds.), Mesopotamien: Späturuk-Zeit und Frühdynastische Zeit. OBO 160/1 (Freiburg 1998) 316 and n. 761. Several other known waterways are recognizable in this list, suggesting, again, that this text is primarily a list of rivers and canals, e. g., dm[uš]!ir-hadin !.balag.nar.min (Deimel reads bu and geštin for muš and din respectively – see Krebernik Beschwörungen 298); G i b i l . min (cf. Í d - G i b i l [RGTC 1, 215]); éren.en.nun.min (cf. Enerennun [RGTC 1, 212]; this canal is located in the vicinity of Ur as recently demonstrated by P. Steinkeller, New Light on the Hydrology and Topography of Southern Babylonia, ZA 91 [2001] 44 n. 92). Krebernik understands min as an orthographic variant of a in this text, but, as will be discussed below, there is reason to believe that, in fact, the graph represents the dual in this text. On the Euphrates 13 an ud.gal.nun hymn to Nisaba, Deimel SF 56 vi 18, in connection with the Abzu – a fitting context for a riverain deity. Remarkably, the writing for the deified Euphrates is distinguished from the writing of Sippar at Fara, the latter occurring as ud.kib.nunki ;32 in the somewhat later Abu Salabiä corpus, however, no such distinction is maintained and the writings of both the city and the river include the ud graph. Outside of Fara, the deified Euphrates occurs, also without the ud element, at Pre-Sargonic Mari, where the god receives emmer offerings along with several other deities: 0.0.1 š e - z í z gi-ti-um dkib.nun.a/ dkib.nun.a “10 sila of emmer for the … of the (two) Euphrates-gods.” 33 Of particular interest here is the fact that the deified Euphrates is conceived as a dyad. The phenomenon, however, is not without parallel in the third millennium. At Ebla the deified Baliä – construed in the dual, dBa-li-äa(-a), 2-dBa-li-äa(-a) – was a full-fledged, if minor, deity of the local cult, as witnessed by offerings recorded in administrative texts, in one case specifying the cultic personnel for this god, as well as by the invocation of dBa-li-äa-a in several Eblaite incantations.34 The deified Baliä, again conceived as a dyad, dBa4-li-äa, also makes an appearance in the Abu Salabiä god list,35 while an Old Babylonian itinerary indicates that a conception of this River-god as a dual prevailed at least through the first 32 33 34 35 Jestin Šuruppak 881 vi 12´. Pomponio Prosopografia 111 s. v. gar.ab.si reads kib (as opposed to Jestin’s lam) after inspection of the photograph. MARI 5 (1987) 72 no. 7 ii 3–6. See F. Pomponio/P. Xella, Les dieux d’Ebla. AOAT 245 (Münster 1997) 78–79 for references and previous literature; also Krebernik Beschwörungen 130. 133–134. 316; W. G. Lambert, The God Aššur, Iraq 45 (1983) 84; P. Xella, ‘Le grand froid’ – Le dieu Bardu madu à Ebla, UF 18 (1986) 440. Note that the appearance of the deified Euphrates in Ebla administrative texts is debated and doubtful. G. Pettinato connected Baradu madu “le grand froid” ( dBa-ra-du ma-du / ma-ad), a minor deity who receives offerings at Ebla, with the deified Euphrates. Pomponio and Xella, however, have been adamant in their objections (see Pomponio/Xella, AOAT 245, 80–82, with previous literature, and Xella, UF 18 [1986] 440 n. 12) on the grounds that 1) madu is an adjective not expected with rivers; 2) the Euphrates is rarely deified – although they do not take into account the Fara and Mari evidence discussed here; 3) dmuš is associated with the Euphrates, but never with Baradu madu; 4) the Euphrates is written with its proper name, without divine determinative and without assimilation of the /n/, in ARET 5, 3 iv 1 f., i. e. bù-la-na-tim. Rather, Xella connects Baradu madu to the Biblical Barad (*BRD), a personification of hail (Xella, UF 18 [1986] 437– 444; see also P. Xella, Barad, in: K. van der Toorn et al. [eds.], Dictionary of Deities and Demons in the Bible [Leiden 19992] 160–161). Note that Pettinato’s interpretation is defended by P. Mander (MROA 2/1, 40–41) as an epithet or folk-etymology for the Euphrates. OIP 99, 83 x 10 (A. Alberti, A Reconstruction of the Abu Salabikh God-List, SEL 2 [1985] 14: 345). 14 Christopher Woods half of the second millennium.36 Similarly, if the Taban contains a dual suffix, i. e., Tab-an, then this river too was regarded as a dyad. Additionally, a dual form may lurk behind the OA personal name Sˇu-Äa-bu-ra and the toponym Äa-bu-ra ki37 – reminiscent of later Assyrian tradition that bears witness to the deified male-female pair Äabur and Äaburtu.38 Also to be noted here is the Sun-god tablet of Nabû-apla-iddina, whose relief – with its penchant for archaic iconography and its depiction of a twoheaded snake arising from the Apsû with the caption muš.igi.min – represents, quite possibly, an image that hearkens back to an old notion of the Araätum-Euphrates as a dyad.39 And perhaps of relevance are the mythical twin rivers of Eridu mentioned in connection with Šamaš and Dumuzi in the Kiškanû legend.40 Finally, it is plausibly in this light that we are to understand the min graph that follows each of the waterways in the Fara lexical list Deimel SF 72 discussed above.41 Remarkably, it is possible to isolate this conception of rivers as twin divinities in a seal motif that, like the textual references, is mainly attested outside of southern Babylonia, particularly at Mari.42 The first of these seals (fig. 1), from OB Mari, depicts an enthroned Enki holding an overflowing vase in the company of four water genies and his adjunct, the Laämu. The scene is played out on a two-headed water goddess. Given that this seal hails from Mari, there can be no doubt that the river in 36 37 38 39 40 41 42 W. W. Hallo, The Road to Emar, JCS 18 (1964) 59: 33. 77–78. RGTC 2, 72; RGTC 4, 44, 144. Frankena Takultu 124: 88. See C. Woods, The Sun-God Tablet of Nabû-apla-iddina Revisited, JCS 56 (forthcoming); cf. U. Seidl, Das Ringen um das richtige Bild des Šamaš von Sippar, ZA 91 (2001) 126 for a dissenting opinion. é - k u g - g a - a - n i - t a giš t i r g i s s u l á - e š a g 4 - b i l ú n u - m u - u n - d a - k u 4 - k u 4 - d è : ina biti elli sˇa kima qisˇti sillasˇu tarsu ana libbisˇu mamma la irrubu / š a g 4 d U t u [d] A m a - u š u m g a l - a n - n a - k e : ina qerebisˇu Sˇamasˇ Dumuzi / d a l - b a - a n - n a í d k a 4 2 - a - t a : ina birit pî narati kilallan “In the holy temple, where its shadow spreads like (that of) a forest and within which no one may enter, inside are Šamaš and Dumuzi, between the mouths of the two rivers.” (CT 16, 46–47: 193–198; after M. W. Green, Eridu in Sumerian Literature [Ph. D. diss. University of Chicago 1975] 188); see also M. J. Geller, Iraq 42 (1990) 23–51. This text may also include entities that, while not waterways, were likewise considered as inherently dual in character. Intriguingly, a sˇakkanakkum-period offering text from Tuttul (Tell Bi>a) lists two sheep for the temple of the River(-god), but only one for the temples of Dagan and Annunitum, suggesting, perhaps, in light of the above evidence, that the River-god is here conceived as a dual (note, however, that the form is singular, é I Na-<à-rí-im [WVDOG 100, 27: 1]). See P. Amiet, Notes sur le repertoire iconographique de Mari a l’époque du palais, Syria 37 (1960) 215–232. On the Euphrates Fig. 1 (Amiet, Syria 37 [1960] 215 fig. 1) Fig. 3 (Amiet, Syria 37 [1960] 217 fig. 3a) 15 Fig. 2 (Amiet, Syria 37 [1960] 216 fig. 2a) Fig. 4 (P. Beck, A Note on a Syrian Cylinder Seal, Tell Aviv 4 [1977] pl. 19 no. 3) question is the Euphrates. Contemporaneous motifs from Susa (fig. 2) and Syro-Anatolia (figs. 3 and 4) similarly depict a dyad river deity, but androgynous in nature, portrayed in the first as a female and in the others apparently as male; in a seal of uncertain provenance, but similar date, the River-god is likely again male and the central deity in question is probably the Sun-god supported by two human-faced bulls (fig. 5).43 A curious seal from Alishar Hüyük (fig. 6), of early OB, if not earlier date, displays strong Mesopotamian influences, being reminiscent of Sargonic 43 See D. Collon, Catalogue of the Western Asiatic Seals in the British Museum 3 (London 1986) 154: 378. 16 Christopher Woods Fig. 5 (Amiet, Syria 37 [1960] 216 fig. 2b) Fig. 6 (modified from von der Osten, OIP 29, 207 fig. 246: d 2199) seals in style and execution.44 Again the river terminates in two male protomes, here bearing a boat with an enthroned deity with bull ears, suggesting Nergal or a god connected with his circle. But the completion of the bovine imagery with the addition of two bison-men, deified and ithyphallic, and a divine human-faced bison serving as a footrest, speaks directly to the iconography of the Sun-god. What may be depicted here, 44 See H. H. von der Osten, The Alishar Hüyük Seasons of 1930–32, 2. OIP 29 (Chicago 1937) 205. 207 fig. 246: d 2199. The date is discussed by B. Buchanan, On the Seal Impressions on Some Old Babylonian Tablets, JCS 11 (1957) 50. As pointed out to me by U. Seidl, the photograph provided by H. Frankfort, Cylinder Seals (London 1939) pl. xxiv b and Amiet’s line drawing (Syria 37 [1960] 217 fig. 3b) are mirror images of the OIP 29 representation. J. Larson of the Oriental Institute has inspected the original negatives and informs me that the photograph in the original publication, OIP 29, is correct. The Frankfort image (and from there Amiet’s drawing) is, quite possibly, the result of printing the negative upside down. On the Euphrates 17 Fig. 7 (Hansen, in: D. Oates Festschrift 93 fig. 3) in light of this iconographic fusion, is a local rendition of Šamaš in his night or nether world aspect.45 While the exact date of this last seal is open to debate, the motif of the dual river god has clear third-millennium origins as shown by the remarkable limestone mold fragment recently published by D. P. Hansen (fig. 7) and attributed to Naram-Sîn. The king is depicted with Ištar atop a stepped temple; beneath, the temple is bordered by a female river goddess bearing offerings. While only the right half of the image is extant, clearly the watery body of the goddess terminated in two identical protomes in the manner of the second-millennium images.46 But the most 45 46 See already F. A. M. Wiggermann, Mischwesen. A, RlA 8 (1994) 235, who proposes that the principal deity in this seal is a local form of Šamaš; cf. Frankfort, Cylinder Seals 168. D. P. Hansen, Through the Love of Ishtar, in: L. al-Gailani Werr et al. (eds.), Of Pots and Plans. Papers on the Archaeology and History of Mesopotamia and Syria presented to David Oates in Honour of his 75th Birthday (London 2002) 102. I would like to thank P. Steinkeller for discussing with me the images in this paragraph and their relevance for the arguments made here. 18 Christopher Woods Fig. 8 (A. Parrot, Sumer [London, 1960] fig. 228) explicit visual evidence for the twin Euphrates gods comes from a unique (Pre-)Sargonic Mari seal (fig. 8).47 The image centers upon a mountain deity bearing a scepter; from the base of his mountain throne two juxtaposed bird heads emerge, each expelling a river from its mouth. These two rivers morph into twin goddesses adorned with vegetation, clearly embodiments of the divine Euphrates and the bounty that it makes possible. A fourth god completes the scene: striking a pose with one leg raised upon the flow of the river and bearing what may be punting pole, this god may be a more anthropomorphic form of the Boat-god who so often carries Šamaš in third-millennium glyptic (figs. 9 and 10).48 Similarly, a pair of water goddesses make their appearance in the so-called Investiture Fresco of Zimri-Lim (fig. 11), which was boldly displayed in the king’s Mari palace. In the top register Zimri-Lim, in a gesture of adoration, stands before Ištar in her martial aspect. Below, in the second 47 48 See Amiet, Syria 37 (1960) 219–220 fig. 5. Note, however, that the Boat-god often manipulates a two-pronged punting pole (in addition to figs. 9 and 10, see P. Amiet, La Glyptique mésopotamienne archaïque [Paris 1980] nos. 1438–1444. 1446). This seal motif is discussed by Frankfort, Cylinder Seals 67–70. 108–110. The suggestion that this theme in all cases involves the Moon-god traveling the night sky rather than the Sun-god (D. Collon, Mondgott, RlA 8 [1995] 372; eadem, Moon, Boats and Battle, in: I. J. Finkel/M. J. Geller [eds.], Sumerian Gods and their Representations [Groningen 1997] 11–12) does not adequately take into account the likely connection between the seal motif and the Šamaš literary text ARET 5, 6; OIP 99, 326+342 described below, or the apparent inclusion in one seal of the sˇasˇsˇaru (m)-saw (Boehmer, Die Entwicklung der Glyptik fig. 466). On the Euphrates 19 Fig. 9 (Boehmer, Die Entwicklung der Glyptik fig. 477; after Amiet, La Glyptique mésopotamienne archaı¨que no. 1505) Fig. 10 (Boehmer, Die Entwicklung der Glyptik fig. 478) register, each goddess bears an overflowing vase and, as in fig. 8, each wears a garment of parallel lines that reinforces the water imagery; fish swim up and down the streams and a sprig of vegetation, again symbolic of fertility, emerges from each vessel. Given that this is Mari, a location below the rain-fall line, where notions of water must necessarily evoke images of the Euphrates, these goddesses, appearing again as a dyad, would seem to be manifestations of the Euphrates, as in the glyptic. The prominence and, just as significantly, the placement of these River-goddesses in the bottom register of an image that conveys a message of no less import than the king’s legitimacy, his divinely sanctioned right to rule, speaks to the well-founded belief in the water as the basis of everything – it is an old visual motif in Mesopotamia, appearing already on the Uruk Vase. As is clear from the glyptic evidence, the sex of the River-deity in this motif varies, being portrayed alternatively bearded and with long flowing locks, thereby recalling the fact that in textual sources d Í d , although 20 Christopher Woods Fig. 11 (J. Oates, Babylon, rev. ed. [London 1986] 62 fig. 42) usually masculine, can also be construed as feminine.49 A case in point is the Incantation to the River, the numerous exemplars of which display both masculine and feminine agreement in free variation with this god.50 The sex of the River-god is further clouded by evidence that at least in some cases díd is to be read as dNaru which, naturally, displays feminine agreement in Akkadian and Semitic in general.51 While the an49 50 51 E. g., T. Jacobsen, The Harab Myth. SANE 2/3 (Malibu 1984) 6: 21. 30; Or. NS 39 (1970) 135: 21; see also the discussion in CAD I–J sub id. E. g., STC 1, 128. J. J. M. Roberts assumed that Íd, as a loanword in Akkadian, replaced the Semitic name for this god, presumably, Narum (The Earliest Semitic Pantheon [Baltimore 1972] 46). But syllabic evidence shows the existence of a dNaru (m) in northern Mesopotamia, e. g., OB Mari: ka-dNa-r [u ?] (ARM 7, p. 346 ad Naru, with additional PNs without the divine determinative), é Na-ri-im (ARM 7, 163: 5, Bottéro’s assumption On the Euphrates 21 drogyny of the River-god may in some cases be attributed to a confusion or syncretism between Semitic and Sumerian river gods, such intentional pairs as Assyrian Äabur-Äaburtum suggest an explanation beyond a mere blurring of genders. This supposition is strengthend by an intriguing parallel from Egypt. Here the deified Nile, Hap or Hapy (H<py), is hermaphroditic, depicted bearded with breasts, but without genitalia. Further, the god is often conceived as a dyad, depicted in pairs symbolizing the essential role of the Nile in unifying Upper and Lower Egypt.52 The duality that appears to be inherent to rivers may ultimately derive from the fact that a river is in essence defined by its two opposing banks, a notion that meshes well with the binary character of the male-female pair Äabur-Äaburtum and perhaps with the indeteminate sex of the river dyads in general. Indeed, the existence of an archaic deictic system in Sumerian built upon river banks, e. g., g ú - e ‘this bank’, g ú - r e ‘that bank’, expressing referential expressions as basic as ‘here’ and ‘there’, demonstrates the entrenchment of opposing river banks in the Mesopotamian conception.53 Thus, numinosity is not to be connected with the river’s distant terminal points, as many of our seals seem to suggest. Rather, this is an artistic device in which riverain duality finds its most expedient visual expression – an explanation which better accounts for those images in which duality is not tied to the river’s end points (figs. 8 and 11). Returning to our evidence for the deified Euphrates, the god, taken singularly, makes a brief appearance in the Ur III period in the form of 52 53 that this is a toponym and not a temple [ibid. 334, 343–344; RGTC 3, 43] is now made less likely by the existence of a é (I)Na-<à-rí-im, where offerings are received, at Tell Bi>a/Tuttul [WVDOG 100, 25: 7. 27: 1 – references courtesy of P. Steinkeller]); OB Sippar: Na-ru-um-dingir (CT 4, 50b: 8 [PN]); OA: ku-um-ri-im sˇa Na-ri-im “priest of the river(-god)” (AfO Beiheft 12, 82 ad 26; see H. Hirsch, AfO 22 [1968–1969] 38); NA: dNa-rum (K 4271: 2). This raises the possibility of a reading Naru (m) for íd in at least some northern, Semitic contexts such as the Pre-Sargonic Mari PN I-tì- díd (RA 31 [1934] 142; MAD 3, 191). The aforementioned OB and NA examples were collected by W. G. Lambert, Iraq 27 (1965) 11, who also cites the interchange of dÍd with Íd as suggestive of a reading Naru in CT 46, 45 col. iv. K. W. Butzer, Nile, in: D. B. Redford (ed.), The Oxford Encyclopedia of Ancient Egypt 2 (Oxford 2001) 550; also, W. Helck/E. Otto, Lexikon der Ägyptologie 4 (Wiesbaden 1982) 486–487. I would like to acknowledge M. Civil and a productive conversation we had at the 214th meeting of the American Oriental Society. For this riverain deictic system, see C. Woods, The Deictic Foundations of the Sumerian Language (Ph. D. diss. Harvard University 2001) 158–167; note the common designation of Egypt as jdbwy (Hr) “the two banks (of Horus)” (A. Erman/H. Grapow, Wörterbuch der aegyptischen Sprache [Leipzig 1926] 153). 22 Christopher Woods the personal name dud.nun.kib- t i - a .54 An echo of the divine river survives into Old Babylonian times as witnessed by the PNs Mar-Purattim “Son-of-the-Euphrates,” and Purattum-ummi “The-Euphrates-is-myMother,”55 to which we may add Marat-Araätim, Mar-Araätim, UmmiAraätim, and Ipiq-Araätim.56 And, in what may represent the latest reference to the Euphrates, a Kassite letter from Nippur invokes the “gods of the Euphrates,” a statement which, incidentally, may suggest that as late as the middle of the second millenium the Euphrates was conceived as a dual or plural divinity.57 At this point the evidence for the Euphrates-god in cuneiform sources comes to an end. Not until Classical times would a cult devoted to a deified Euphrates again flourish in Mesopotamia.58 The deification of rivers in Mesopotamia is subject not only to significant temporal parameters, as our evidence for the divine Euphrates suggests, but tends to be bounded geographically as well. It is primarily an early phenomenon and one with an unmistakably northern flavor. For instance, the deified Tigris occurs in the great OB god list as a member of Enki’s court,59 a reflex of which may be discerned in OB personal names, e. g., Idiqlat-ummi, Mar-Idiqlat, Ummi-Idiqlat. But as a theophoric element the river was particularly popular in Assyria during the second millennium, e. g., Urad-Idiqlat, Idiqlat-erisˇ, Idiqlat-remani, Kidin-Idiqlat, Siqe-idiqlat, Silli-Idiqlat, Sˇep-Idiqlat, Tasˇme-Idiqlat.60 And it is of temporal significance that the deified Baliä is attested at Ebla and in the Abu Salabiä god list, as well as in Old Akkadian personal names, but as a god 54 55 56 57 58 59 60 MVN 16, 908: 8. It is doubtful, however, that the god d L u g a l -ud.kib.nunki , who receives offerings among other gods at the Umma è š - è š festival, is to be taken as d L u g a l - B u r a n u n a . Rather, the name, which occurs without the íd determinative, is probably d L u g a l - S i p p a r ki ‘the Divine King of Sippar’, analogous to d L u g a l G ú - d u 8 a ki , d L u g a l - M a r a d - d a ki , and d L u g a l - uru×kárki also known from Ur III sources – that is, Šamaš of Sippar who is venerated at Umma. See Sallaberger Kalender 1, 248 (for attestations of this divinity in Ur III sources, see ibid. 2, Tables 90 and 99a; also J. A. Peat, An Offering-List from the Third Dynasty of Ur, RA 69 [1975] 19–22; N. Schneider, Die Götternamen von Ur III. AnOr 19 [Rome 1939] nos. 286. 287). RGTC 3, 305. RGTC 3, 274. dingir.meš sˇa ídPu-rat-t [i ] nap-sˇa-ti-ka li-is-su-ru “May the gods of the Euphrates protect your life” (BE 17/1, 87: 5–6; H. Waschow, MAOG 10/1 [1936] 15); cited in E. Ebeling, Flußgottheiten, RlA 3 (1957–1971) 93. F. Cumont, Études syriennes (Paris 1917) 247–256. RA 20 (1923) 100 ii 34. PNs from B. Alster, Tigris, in: K. van der Toorn et al. (eds.), Dictionary of Deities and Demons in the Bible (Leiden 19992) 870; C. Saporetti, Onomastica medio-assira 1. Studia Pohl 6 (Rome 1970) 310–311. On the Euphrates 23 the river barely survives into the post-OB periods as shown by the deity’s position at the very end of A n = Anum.61 Similarly, the Äabur occurs regularly with the divine determinative in Ur III texts from PuzrišDagan62 – but most often without it in the second millennium.63 And the appearance of the Taban as a theophoric element in personal names from Pre-Sargonic Dilbat (U r - d D a b 4 - a n ), as well as from Sargonic Ešnuna (ka-Ta-ba-an, {Ki }?-nam-Ta-ba-an) speaks to the deification of this Tigris tributary over an area stretching from western Akkad through the Diyala region in the third millennium.64 Further, the Diyala occurs in the Sargonic personal names Ì - m e - D u r - ù l and Sˇu-Dur-ùl/al,65 appearing once in the OB period written d D u r - ù l .66 Finally, the Pre-Sargonic and Sargonic periods bear witness to a number of deified lesser branches and canals, e. g., d Í d - A k š a k ki , d Í d - A m - s i - har, K a - d Í d - É n s i , d Í d - K i š ki , d Í d - M á - g u r ,67 but already by the Ur III period many of 8 these were stripped of their divine status, e. g., Í d - É n s i , Í d - M á - g u r 8 ( - r a ) .68 The promotion of rivers to gods is but one facet of the broader phenomenon of the numen loci that encompasses the deification of mountains and cities; it is the more tangible counterpart to the deification of heavenly bodies and natural phenomena, familiar aspects of Mesopotamian religious thought.69 As the geographical and temporal distribution 61 62 63 64 65 66 67 68 69 Lambert, Iraq 45 (1983) 85. RGTC 2, 266. RGTC 3, 277; RGTC 4, 144. Dilbat PN: OIP 104, p. 111 iii´ 5; Ešnuna PNs: MAD 1, 163 viii 40 and 72 rev. 5´ respectively. Ebla PNs attest the deification of a GN Ta-ba-an/nu (e. g., Isˇ-má-Ta-ba-an, en-ga-Ta-ba-an, Kùn-Ta-ba-an, Ti-Ta-ba-nu, kà-Ta-ba-an (cf. ka-Ta-ba-an at Ešnuna), but in light of the distances involved, it is uncertain, as pointed out by M. Krebernik (Die Personennamen der Ebla-Texte [Berlin 1988] 79 – reference courtesy of W. Sallaberger), that this designation refers to the Tigris tributary, as assumed by Lambert (MARI 6 [1990] 642). RGTC 1, 210. RGTC 3, 279. RGTC 1, s. vv. RGTC 2, s. vv. See already Roberts, The Earliest Semitic Pantheon 58. On the deification of rivers, mountains, and cities, see also J. Bottéro, Les divinités sémitiques anciennes en Mésopotamie, in: S. Moscati (ed.), Le antiche divinità semitiche. Studi Semitici 1 (Rome 1958) 43; Lambert, Iraq 45 (1983) 82–86; Lambert, MARI 6 (1990) 641–643; F. Stolz, River, in: K. van der Toorn et al. (eds.), Dictionary of Deities and Demons in the Bible (Leiden 19992) 707–709; P. Michalowski, The mountain and the stars, in: P. Marrassini (ed.), Semitic and Assyriological Studies Presented to Pelio Fronzaroli by Pupils and Colleagues (Wiesbaden 2003) 407–410. 24 Christopher Woods of our evidence suggests, the raising of topographical features of the Mesopotamian landscape to divine status may belong to the earliest discernable strata of the Semitic religious conception.70 In essence, topographical features that were considered to be of inherent significance, inspiring awe and reverence, were held to be imbued with a numinous quality and so were incorporated into the pantheon, if only in name. But the fact that the Euphrates and the Baliä both received offerings shows that at certain northern locations, such as Mari and Ebla, rivers could assume a personified form, as the evidence from the glyptic confirms, and thus stand on par with other gods. The Spelling UD . KIB . NUN The variations in the early writings of Buranuna demonstrate that the nun element, which may be omitted, was a phonetic complement and that the writing (d )kib.nun is to be interpreted as (d )B u r a n u n a x (kib)nun . At Fara the combination of the signs kib and nun was not infrequently rendered as a ligature, i. e., dkib+nun,71 a syllabically-glossed spelling that finds its parallel in writings of the type šeš+na for N a n n a x (šeš)na . In this connection, note that the spelling ending with na, i. e., ud.kib.nun.na, in the vast majority of cases, writes Buranuna and only exceptionally Sippar – a distinction that is maintained in Diri Nippur, i. e., ud.kib.nun.ki = Sí-[ip-pa/i ]-ra (328) and íd.ud.kib.nun.na = Pu-ra-at-tum ! (347) – thus confirming the phonetic character of this part of the logogram.72 The optional ud element, however, is a more complicated matter. Poebel, who sought phonetic solutions to so many problems of Sumerian 70 71 72 Roberts, The Earliest Semitic Pantheon 58. Although the dynamic landscape of Upper Mesopotamia no doubt influenced the more common occurrence of numen loci in the north, the deification of mountains being an obvious case, the phenomenon likely had a cultural component as well. This is borne out by the rivers, upon which no region depended more than the south, yet the deification and veneration of individual rivers, as we have seen, plays a lesser role in the Sumerian religious conception. E. g., Deimel WF 29 rev. iii 4. 76 rev. i 10´. See the comments of Å. W. Sjöberg/E. Bergmann, The Collection of the Sumerian Temple Hymns. TCS 3 (Locust Valley 1969) 141, who also note that the writing ud.kib.nun.naki for Sippar, rather than Buranuna, is rare. OB exceptions appear in the writing of the GNs Sippar-bàd(ud.kib.nun.na.bàd) and Sippar-edin.na (ud.kib.nun.na.edin.na) (RGTC 3, 209). Incidentally, the plene spelling, i. e., ud.kib.nun.na, shows that the name of the Euphrates is B u r a n u n a and not B u r a n u n ; an etymology based on an understanding of the toponym as a genitival construction, B u r a n u n a ( k ) , is suggested below. On the Euphrates 25 logography, interpreted the compound as a phonetic writing, í d . B í r - à l (or ù l ) - n u n u , which later changed to Buranunu.73 Lambert, on the other hand, interpreted kib as the basic logogram and both ud and n u n ( . n a ) as phonetic complements, i. e., bar6kibnun.na .74 The prima facie objection to understanding ud as a phonetic indicator is that a value /bur/ is not attested for the sign and a reading with the first syllable vocalized as /bar/ is known only from an OB Eme-sal text and is likely the result of a late vowel shift.75 But this obstacle is certainly not insurmountable, for the vocalic quality of CVC signs is not rigid in any period of cuneiform, particularly in the earliest periods when the writing system was ill-equipped to handle syllabic writings and phonetic approximations often sufficed, e. g., muš3+erin for šušin, where erin is a phonetic complement, known from later periods to have the value š e š 4 , but not the expected š u š x .76 More problematic, however, is the fact that the scribal convention handled the ud and nun elements differently. Whereas the ud graph is frequently omitted in the writing, attestations without nun are extremely rare, suggesting that the two belong to separate graphic classes. The solution to the problem of the ud element may ultimately rest in the fact that – as the earliest evidence of the writing of the river from 73 74 75 76 A. Poebel, Miscellaneous Studies. AS 14 (Chicago 1947) 11–12; also idem, Sumerische Untersuchungen II, ZA 37 (1927) 271. W. G. Lambert, Äalam, Il-Äalam and Aleppo, MARI 6 (1990) 642 n. 4. a - g i 6 B a - r a - n a - k [ a ] (Poebel, ZA 37 [1927] 162 iv 4). As noted by Poebel, /a/ is written for both /u/ and /i/ in this text (ibid. 270). The earliest syllabic writing for the river may come from Ebla. Krebernik suggests that the occurrence of b u r - n u n in an Eblaite incantation (Beschwörungen 180: xvi (d) 5. 182), mentioned in connection with Enki, may be a syllabic writing of Buranuna. Support for this interpretation may be sought in the divine pair d E n - k i - b u r - n u n (ITT 7567) and d E n - k i - g ú - Í d - i d i g n a (ITT 7310) (see Carroué, ASJ 13 [1991] 120; Schneider, AnOr 19, 21 nos. 110. 113), both of whom receive offerings in Ur III texts. However, the former may be related to the god d B u r - n u n - t a - s i / s á - a who bears the epithets “the one of wide understanding” and “native of Eridug” (CT 16, 45: 125–126; CT 17, 21: 112), and who is counted as one of the six sons of Enki in A n = Anum II 288. Note in this connection the Fara PN B u r - n u n - s i (Pomponio Prosopografia 62). This god is probably not to be connected with the Euphrates. Rather, this deity of Enki’s court – as his name suggests, “the one who fills the princely bowl” – was probably responsible for filling the overflowing vessel with which Enki is often depicted, a regular aspect of his iconography. Quite possibly, the two flowing streams that emanate from the bowl give visual representation to the conception of rivers as duals, if not symbolic of the Tigris and Euphrates themselves. See P. Steinkeller, Review of M. W. Green/H. J. Nissen, Zeichenliste der Archaischen Texte aus Uruk, BiOr 52 (1995) 695. 26 Christopher Woods Fara, as well as that from Pre-Sargonic Mari, makes clear – the Euphrates was considered a god and accordingly was written with the divine determinative, i. e., dkib.nun. In these cases ud is absent. Plausibly, the prefixed ud graph was employed as a substitute for dingir – that is, ud.kib.nun belongs to the ud.gal.nun (UGN) orthographic tradition, the allographic system of the Early Dynastic period. If this hypothesis is correct, then the writing for Buranuna and Sippar in later periods would represent a relic UGN spelling. As to why this particular spelling would have such perseverance when other UGN writings became obsolete, the answer may lie in the relationship between Šamaš(ud) and Sippar (ud.kib.nunki ), and in the distinct possibility that in later periods the orthography was reinterpreted based on the assumption, reasonably placed, that the inclusion of the god’s name was central to the writing of his cult center, as parallels from the south certainly bear out, e. g., Nippur, Larsa, Ur, etc. In support of this proposal is the Sargonic writing for Sippar, dud.kib.nunki , opposed to the contemporaneous writing of the Euphrates, ud.kib.nuníd , the former suggesting an association between the Sun-god and his northern cult center.77 But, as we have seen, an orthographic distinction between the river and the city was not a Sargonic innovation. Already Fara administrative texts differentiated the writing of the Euphrates, dkibnun(.a) from the writing of Sippar, ud.kib.nunki .78 It was only with the Abu Salabiä corpus that both river and city were regularly written with ud. More than a simple orthographic reform, this step gave primacy to the city and fostered an understanding of the Euphrates as the “river Sippar,”79 so well attested in later sources. Yet a vestige of the original writing of the river, and thus of the distinction between river and city, survived in the north, as witnessed by the frequent attestations of kib.nun(.na) for the Euphrates, while Sippar was written ud.kib.nun. These orthographic developments may be summarized as follows: 77 78 79 See B. Kienast, FAOS 8, 97–98. 116; RGTC 1, 144. Only at the close of the third millennium, in the Ur III period, do we have the composite spelling dud.nun.kib for the deified Euphrates, attested only once to my knowledge, in the PN B u r a n u n a - t i - a (MVN 16, 908:8, cited above). This writing may reflect a relatively late ignorance of the UGN origins of the compound or represent a corrupt transmission of the above Sargonic spellings. The compound ud.kib.nun also occurs in the Fara UGN text Deimel SF 55 ix 19–20 as well as in the lexical list Deimel SF 7, which is organized by sign form (viii 23–24, following Adab [viii 22]). As already suggested by the variant spelling ud.kib.nunki for the Euphrates in an Abu Salabiä Z à - m í hymn to Ninbilulu (OIP 99, p. 48: 61. 62, cited above). On the Euphrates Euphrates Sippar dkibnun(.a) Íd-kibnun(.na/a) 27 Íd-ud.kib.nun(.na)(ki) ud[=d]kib.nun (ki) Regarding the proposal that ud.kib.nun is UGN in origin, it must be pointed out that the writing of divine names in what may be described as a mixed orthography – part UGN, part regular orthography, insofar as UGN values differ from those of standard cuneiform – is quite common. The Abu Salabiä literary texts, for instance, are filled with writings of the type ud.M a r - t u , ud.mimušen , ud.N i n - u r t a, ud.Z a - b a 4 - b a 4 , etc.,80 while duplicates freely mix orthographies in identical passages, particularly with regard to the ud for an substitution, e. g., ud.gal.nun in OIP 99, 167 rev. xvi 7, versus dingir.gal.nun in OIP 99, 129 x 4´, and parallel passages within the same text show a free variation of orthographies, e. g., d gal - k i d N i n - k i (OIP 99, 114 i 3´) versus dgal.unug ud.N i n - k i (11´).81 As for a city deriving its spelling from the UGN tradition Michalowski has argued, quite convincingly, that the city written ub.pa+ru in the Abu Salabiä version of the City Names List82 and ub.pa.ru in the Fara exemplar83 corresponds to Nu-me-gi4ki in the later OB version from Ur.84 Since ub for me and pa+ru for gi/gi4 are regular UGN substitutions, the city had an UGN spelling at Fara and Abu Salabiä.85 As Krecher has shown, far from there being two distinct orthographies, UGN and the standard orthography are essentially two adjoining facets of the same writing system.86 Certain literary texts, for instance, while written in the regular orthography, employ signs more 80 81 82 83 84 85 86 W. G. Lambert, Review of R. D. Biggs, Inscriptions from Tell Abu Salabikh. OIP 99, BSOAS 39 (1976) 431. J. Krecher, ud.gal.nun Versus ‘Normal’ Sumerian: Two Literatures or One?, QdS 18 (1992) 296. OIP 99, 21 ii 5. Deimel SF 23 ii 4. UET 7, 80. P. Michalowski, On the Early Toponymy of Sumer: A Contribution to the Study of Early Mesopotamian Writing, in: A. F. Rainey (ed.), kinattutu sˇa darâti. Raphael Kutscher Memorial Volume (Tel Aviv 1993) 124. See p. 124 n. 17 on the omission of initial nu-. The relevant entry in the Uruk version reads [ub.pa].ru. If this restoration and Michalowski’s arguments hold true, then, significantly, UGN had its origins in the Uruk, rather than the ED period (ibid. 124). Note that Sippar does not occur in the Abu Salabiä, Fara, or Uruk versions of the City Names List. Krecher, QdS 18 (1992) 285–303. 28 Christopher Woods typical of UGN writing, but with uncertain values.87 And, as we have seen, scribes did not hesitate to mix orthographies in writing a single lexical item. Further, UGN has cropped up unexpectedly beyond the confines of religious and literary texts, being employed in several word lists.88 Temporally, it is of considerable significance that the one bi-orthographic text at our disposal (Westenholz Jena 173) is of Sargonic date, well after the floruit of UGN and thus refuting the assumption that UGN is a phenomenon confined to the Early Dynastic period. Certainly these are not isolated examples, and it would not be surprising eventually to find further Sargonic or even post-Sargonic UGN evidence, as well as values considered to be diagnostic of UGN in additional non-literary contexts. It is, therefore, not unreasonable to suggest that the mixed writing ud.kib.nun made its way into the “standard” orthographic tradition, if, indeed, we are justified in making this rigorous distinction. And it is quite possible that the writings of other cities may originate in the UGN tradition – although this hypothesis is yet to be proven.89 Whereas many of the major cult centers of Sumer have transparent orthographic origins, i. e., divine symbol + unug/ab, having their roots in symbolic and semiotic codes that pre-date the invention of writing, the logographic origins of others are far from obvious. Perhaps in some cases these early logograms represent archaic UGN writings.90 The proposal that ud.kib.nun represents an UGN substitution for dkib.nun requires further comment on the ud for an substitution, the hall-mark of ud.gal.nun orthography. As pointed out by Lambert, this replacement has yet to receive a satisfactory explanation,91 although more recently Krecher has suggested that the substitution is phonetically motivated, observing that phonemic changes between words in standard orthography and their UGN counterparts include vowels changing to /u/ and the omission of final consonants.92 The fact that ud can have the value u n x could be mustered in support of this suggestion.93 However, a 87 88 89 90 91 92 93 Krecher, QdS 18 (1992) 292. Krecher, QdS 18 (1992) 294 n. 30. Cf. Michalowski, Memorial Kutscher 123–124. See already Michalowski, Memorial Kutscher 129. W. G. Lambert, Studies in ud.gal.nun, OA 20 (1981) 92. Krecher, QdS 18 (1992) 299. See M. E. Cohen, JCS 28 (1976) 84 f., with references. K. Oberhuber proposed a phonetic solution for the ud for an substitution based on ud = /tam/ and a ProtoSumerian etymology of /*d/tem/g̃-r/ for d i g̃ i r, ES d i m e r (Linguistisch-philologische Prolegomena zur altorientalischen Religionsgeschichte [Innsbruck 1991] 14–18 – reference courtesy of W. Sallaberger). On the Euphrates 29 more compelling argument can be made for the ud for an substitution being semantically based.94 Obviously, the fact that both signs represent heavenly entities is the principal factor here, but the motivation may run deeper still, drawing parallels between the respective astral deities. As discussed by Myers,95 Šamaš and his northern cult center Sippar – celebrated for its great antiquity as u r u u l /al siati (m) “the eternal city” – played a crucial role in the northern politico-religious reality. While the Hammurapi stele provides the most striking evidence for the importance of Šamaš-Sippar for the first dynasty of Babylon, the phenomenon clearly has a third-millennium basis. The earliest known Semitic literary text, with ED manuscripts from Abu Salabiä and Ebla (ARET 5, 6; OIP 99, 326+342), is a mythic composition revolving around Šamaš and his cult seat, Sippar. More than a mere literary presence at Ebla, offering accounts from the royal archives demonstrate the existence of an active cult to Šamaš in northern Syria. That the Pre-Sargonic kings of Mari paid homage to Šamaš at his cult center is clearly shown by the discovery of a votive statue at Sippar dedicated to Šamaš by Ikun-Šamaš, king of Mari.96 Then there is the evidence from the glyptic. The most common mythological scene in Pre-Sargonic seals centers on the Sun-god, the wellknown “Sun-god in his boat” motif.97 The geographical distribution of these seals is revealing as well, being restricted to northern Mesopotamia, from Mari to the Diyala region. Taken as a whole, the evidence suggests that Šamaš was of great importance in the north and was acknowledged as such in a wide swath that stretched along the Euphrates from Ebla, to Mari, down to Sippar and Akkad, and into the Diyala region. And it is not without interest that later sources promoting the cultural and religious importance of Sippar – however suspect their claims – tend 94 95 96 97 The possibility of a semantic basis for the ud for an substitution has already been noted by Krebernik (OBO 160/1, 302). J. Myers, The Sippar Pantheon: A Diachronic Study (Ph. D. diss. Harvard University 2002); eadem, The Importance of Sippar as a Religious and Cultural Center for the First Dynasty of Babylon (paper delivered at the 213th Meeting of the American Oriental Society, Nashville, TN 2003); see also W. W. Hallo, Antediluvian Cities, JCS 23 (1970–71) 65. J. S. Cooper, Sumerian and Akkadian Royal Inscriptions 1 (New Haven 1986) 86–87 (Ma 2.1); J. E. Reade, Early Dynastic Statues in the British Museum, NABU 2000/82; C. B. F. Walker/D. Collon, Hormuzd Rassam’s Excavations for the British Museum at Sippar in 1881–1882, in: L. de Meyer (ed.), Tell ed-Der 3 (Leuven 1980) 96 no. 1; on the reading of the RN, see M. Krebernik, ZA 81 (1981) 139 (cf. I. J. Gelb/B. Kienast, FAOS 7, p. 9 [MP 8]). P. Steinkeller, Early Semitic Literature and Third-Millennium Seals with Mythological Motifs, QdS 18 (1992) 256. 30 Christopher Woods to do so by retrojecting the city’s preeminence into the remote past as witnessed by the Sumerian King List, where Sippar holds claim to being one of the five antediluvian cities and the only one located in Akkad, by Nabû-apla-iddina’s Sun-god tablet, where cultic claims find justification in ancient precedent,98 and by Nebuchadnezzar I’s self-legitimatizing claim of descent from Enmedurana, the antediluvian king of Sippar. Collectively, this evidence has led Myers to draw an analogy between Sippar-Šamaš in the north and Nippur-Enlil in the south – that Sippar essentially served as a northern religious counterpart to Nippur, both being prominent religious centers, hubs of scribal activity, and without significant political power throughout their respective histories.99 The ud for an substitution was a scribal jeu d’esprit that drew primarily upon the common astral qualities inherent to the common nouns a n ‘sky, heaven’ and u d ‘sun’. But a number of nouns share this quality and could conceivably substitute for an on this basis. What may be at work here in favor of ud is the Sun-god’s importance in the northern pantheon, a position which evoked similarities with An in the south, thus giving meaning to the primary ud for an replacement, namely, in the writing of the divine determinative. A semantic basis to this substitution, which relies in part upon the elevated status of the Sun-god in the north may go some way to explaining its early use in the spelling of the Sun-god’s northern cult center, particularly when the Euphrates was written dkib.nun. And as we have seen, for many northern scribes, as late as the Old Babylonian period, a spelling with ud was only appropriate for writing the city, kib.nun(.na) being the preferred spelling for the river. But far from this being an isolated geographical link, there is yet other evidence for connecting the UGN orthography to northern Babylonia. While the precise place of origin of UGN may never be known, it betrays a number of Semitic or northern characteristics – and it may not be too bold to suggest that the cities of Kiš and Sippar stand out as obvious candidates for centers of influence. As is well known, Semites account for half of all literary activity at Abu Salabiä and this city, lying north of Nippur, was also a major center for the production of UGN literature. Further, as Krecher has observed, certain signs that appear to be diagnostic of UGN orthography are also found in Semitic personal names in Pre-Sargonic texts notably from Kiš and Sippar,100 e. g., Il-gu-ru x(ku) for 98 99 100 Woods, JCS 56 (forthcoming). Myers, The Sippar Pantheon 4. Krecher, QdS 18 (1992) 300. On the Euphrates 31 Il-kurub (Kiš, ED II),101 or appear later in the writing of Semitic personal names in Sargonic texts from Nippur or further south, Umma, e. g., ku for urx , rux (ku 8-ku-ub-e-la-ak, /kurub-ilak/),102 ša for na5 (en-ša-il, /enna-il/).103 The substitution su for ni, occurring in writing of the possessive suffix - ( a ) n i,104 is very likely semantically motivated, inspired by the Semitic possessive -sˇu, which is written with the su graph in Old Akkadian and Eblaite. The substitution kiš for en is, no doubt, also based on meaning in deference to a hegemonistic northern Kiš state. Further, the content of the literature written in UGN orthography may hint at northern influences. As pointed out by Lambert,105 Enlil and Zababa figure prominently in UGN hymns and myths. Possibly the former,106 but certainly the latter, is a northern import, perhaps Semitic in origin. Zababa is, of course, closely linked to Kiš with only a minor southern presence. On the other hand, UGN texts regularly give supremacy within the pantheon to Enlil – a ud.gal.nun “father Enlil” – placing him above Enki and Nanna, while seeming to relegate An to a lesser position, in contrast to the big Fara god-list which begins with An.107 One of the two known UGN texts from Nippur, an enumeration of gods, gives the sequence ud, Nanna, Ningirsu, and Ašgi – where ud may represent Utu or An.108 101 102 103 104 105 106 107 108 B. Buchanan, Catalogue of Ancient Near Eastern Seals in the Ashmolean Museum 1 (Oxford, 1966) no. 137. I thank G. Marchesi for this reference. As suspected by Krecher, ku may have the UGN value urx /rux in the PN i -ku.gu-il appearing at Abu Salabiä (OIP 99, p. 34), Sippar(?) (DP 2 i 6´), and Dilbat (CT 37, 7 f. iii´ 4), but the form is admittedly difficult (QdS 18 [1992] 300 n. 49); perhaps the name is to be understood as I- guku-Il for Ikun-Il as suggested by I. J. Gelb/P. Steinkeller/R. M. Whiting, Earliest Land Tenure Systems in the Near East: Ancient Kudurrus. OIP 104 (Chicago 1991) 107 ad i 6´. Foster Umma 26: 10. Westenholz Jena 53 iii 2. These examples are taken from Krecher, QdS 18 (1992) 300, who further notes (ibid.) that nám interchanges with nam in a Pre-Sargonic literary text from Adab: n á m /n a m - m a - n i - r a (OIP 14, 53 vi 4–6); see also Krebernik, OBO 160/1, 299–302, for a list of known UGN substitutions; Lambert, BSOAS 39 (1976) 430–431. Westenholz Jena 173: 1. Lambert, OA 20 (1981) 92–93. For a contrasting view, see now D. O. Edzard, Enlil, Vater der Götter, in: P. Marrassini (ed.), Semitic and Assyriological Studies Presented to Pelio Fronzaroli by Pupils and Colleagues (Wiesbaden 2003) 173–184. Lambert, OA 20 (1981) 93, with references. Note, however, that the small god-list, SF 5–6, opens with Enlil followed by Enki (see M. Krebernik, ZA 76 [1986] 161–204). W. G. Lambert, Studies in UD.GAL.NUN: Addendum, OA 20 (1981) 305. 32 Christopher Woods The River-god, the Sun-god, and the Location of Sippar Nothing discussed thus far explains the relationship between the Euphrates and Sippar, or between the Euphrates and Šamaš, which must underlie the shared writing of Buranuna and Zimbir. The explanation, I suggest, is to be sought in the character of the River-god and the locations of the River-god cult. dBuranuna, as we have seen, is an old god attested at Fara and Pre-Sargonic Mari. But the deified Euphrates is merely one manifestation of díd, the primeval river lauded so often as banât kalama “creatrix of everything.” In the same way dBaliä and dIdiglat are manifestations of díd that embody the numinous powers inherent to these particular rivers. The divine river, ever cleansing in its constant flow, clearing the falsely accused, is first and foremost a god of justice – dini tenesˇeti tadinni atti “(River,) it is you who judges the cases of mankind.”109 As such, the River-god is best known in connection with the river ordeal – trial by river – appearing under the names d Í d - l ú - r u - g ú ‘River-who-confronts-the-man’110 and simply d Í d , or dNaru (m) in some Semitic contexts. It is this common judicial aspect that links the Sun- and River-gods – an association of sun, river, and justice made explicit in a hymn to Utu: d U t u z a - d a n u - è d i - k u d n u - k u d - d a k a - a š n u - b a r - r a / d U t u z a - d a n u - è d Í d - l ú - r u - g ú d i - k u d n u - k u d “Utu, if you do not come out, no judgment is given, no decision is decided / Utu, if you do not come out, the divine Ilurugu does not give judgment.”111 And to this we must add that in the Lugalbanda epic Utu is said to have a sevenmouthed subterranean river in whose waters the roots of Enki’s eagletree rest.112 But the association runs deeper still, as witnessed by the cosmic and mythological notion that claims of the underworld pair Lugalerra and Meslamtaea, m ì n - n a - n e - n e l u g a l Í d - d a - m e - e š d Í d - l ú - r u - g ú l ú z i d d a d a g - g a [ - à m ] “They are the two lords of the river, the River-god of the ordeal, which clears the true man.”113 The same Ibbi-Sîn hymn locates the primeval river, Í d - m a h ‘Great River’, 109 110 111 112 113 STC 1, 128: 8. There is little to distinguish d Í d from d Í d - l ú - r u - g ú . In the later tradition of the bilinguals the two are equated (KAV 218 ii 17. 20; 5R 13 ab), and in A n = Anum, d Í d - l ú - r u - g ú is simply another name for d Í d (An = Anum II 276–279). B. Alster, Incantation to Utu, ASJ 13 (1991) 44. Ll. 34–35. Or. NS 19–20 (1970–71) 143: 2–3. Note that in Ur III offering texts dIlurugu is often listed in conjunction with underworld gods (see T. Frymer-Kensky, The Judicial Ordeal in the Ancient Near East [Yale diss. 1977] 86–93). On the Euphrates 33 already encountered in the Barton Cylinder,114 at the place of the sun’s rising, í d - z u í d k a l a g - g a - à m Í d n a m - t a r - r a - à m , Í d - m a h k i u d - è i g i n u - b a r - r e - d a m “Your river is a mighty river, the River which determines fates, the Great River at the place where the sun rises, no one can look at it.”115 And this cosmographic conception also attributes to the place of the sun’s ascent the pronouncing of judgments, e. g., d i k u d - r u k i u d è “He (Utu) pronounces judgments at the place where the sun rises.”116 This nexus between the cosmic river, the underworld, and the east is captured in fig. 6, where the central figure is possibly Šamaš, in the predawn hours, accompanied by two bison-men and a human-faced bison. These creatures developed an association with the Sun-god by virtue of the bison’s home in the eastern hilly flanks117 and the Zagros as symbolic of the cosmic location of the sun’s daily ascent. This is an awesome place at the edge of the world, where the Sun-god rises, where the world of the dead meets the world of the living, where judgments are made, and where the primeval river runs. The River-god díd has a long history in Mesopotamia; already at Fara and Abu Salabiä the god occurs as a theophoric element in Semitic and Sumerian PNs.118 And, as with the deification of specific rivers, the appearance of díd in personal names is predominately an early phenomenon, being well-attested in the Pre-Sargonic and particularly the Sargonic periods, but relatively rare after the third millennium. Of significance for our argument is the telling geographical distribution of these names. Much of the earliest onomastic evidence for the River-god comes from Nippur, Mari, Sippar, and the Diyala, where the god is most often written with the divine determinative.119 This situation is in contrast to 114 115 116 117 118 119 Barton Cylinder MBI 1 ii 13 (according to ASJ 16 [1994] 15–46 numbering). Or. NS 19–20 (1970–71) 142: 19´–20´. TCS 3, 46: 489; for further references, see ibid. 89–90 ad 192; T. Frymer-Kensky, The Judicial Ordeal 611 n. 27. See F. A. M. Wiggermann, Mesopotamian Protective Spirits. Cuneiform Monographs 1 (Groningen 1992) 174. Fara: Í d - h i - l i - s ù , d Í d - i r - n u n (Pomponio Prosopografia 123); Abu Salabiä: U r d Í d ( - d a ) , I-ti - díd (OIP 99, pp. 34–35). Pre-Sargonic PNs include the aforementioned I-dì-díd (RA 31, 142; Parrot Documents 3: 1. 16: 1 [Mari]), I-dì- íd (RSO 32 [1957] 89 viii x+18 [Sippar]), d Í d - d è - d è (PBS 13, 27 rev. i 5 [possibly early Sargonic]), and Sˇu -íd (Iraq 7 [1940] 66 F. 1159 rev. 10 [Tell Brak]); from the Diyala, the following names occur during the Sargonic period: I-ti - díd, Pù-su- díd, puzur4-díd (MAD 1, p. 230); from Pre-Sargonic and Sargonic Nippur there is: d Í d - d è - d è d Í d - u r - s a g , d Í d - z a l a g - g a , d Í d - z i d , d Í d -dug×igi ?, d Í d - k i - g a l , ( D u m u - ) L u g a l - (d)Í d - m u , L u g a l - d Í d - s i , N i n - 34 Christopher Woods that encountered in the south, where the River, in names that suggest the deity, often lacks the divine determinative.120 Certainly, it is not coincidental that Nippur,121 Mari, and Sippar, in addition to Hit, were also closely associated with the river ordeal, although the only third-millennium evidence for this is from Nippur.122 But it is north of Nippur that veneration of díd is most pronounced. At Mari, for instance, there is evidence for a bit Narim ‘temple of the River(-god)’.123 As is well known, the River-god is well attested at Old Babylonian Mari in connection with the resolution of political matters and adjudication of legal cases.124 But the dedication of a stone vessel to díd and Ištarat on behalf of Ikun-Šamagan shows that already in the Early Dynastic period the River-god was the focus of royal patronage.125 To this we must add the contemporaneous appearance of the River-god as a 120 121 122 123 124 125 d {Í d - m u }, U r - d Í d - d a , U r - d Í d - m á - g u r (TuM 5, p. 27 and Westenholz OSP 1, 8 p. 108; see now M. Such-Gutiérrez, Beiträge zum Pantheon von Nippur im 3. Jahrtausend 1 [Rome 2003] 332). Late ED and Sargonic evidence includes the Lagaš names: { Í d } - k i - á g (Donbaz/ Foster Telloh 80 i 16´) díd-la-ba (Limet Documents 31: 6, 9; also MAD 3, 160 s. v. LB<x ), Í d - l ú , Í d - l ú - d a d a g (Struve Onomastikon 82), and the Ur PN Í d - d a - š a g 5 (Alberti/Pomponio UET 2 Suppl. 49: 5´). Also note the Ur III PNs from Lagaš and Umma: Í d - é - a , U r - Í d - d a , U r - Í d - e d e n - n a ( U r - Í d - d è - n a ) , U r - Í d N i n a ki , Í d - d è - b a - d u 7 (see Limet L’anthroponymie 254–255), Í d - a - b i - d ù g (MVN 6, 140 rev. 6. 512 rev. 6 – reference courtesy of W. Sallaberger). Although a great deal of the evidence for the river ordeal comes from Mari, it appears that the city itself was not a location for the ordeal (contra Frymer-Kensky, The Judicial Ordeal 175–180. 526–527); rather, subsequent review of the evidence first published by G. Dossin (Un cas d’ordalie par le dieu fleuve d’après une lettre de Mari, in: Fs. P. Koschaker [Leiden 1939] 112–118) reveals that the kings of Mari sent the accused to Hit (see J.-M. Durand, Archives épistolaires de Mari I/1. ARM 26/1 [Paris 1988] 521–523; idem, Les documents épistolaires du palais de Mari 1. LAPO 16 [Paris 1997] 397–398; J.-R. Kupper, Lettres royales du temps de Zimri-Lim. ARM 28 [Paris 1998] 27–28). See A. Falkenstein, Eine gesiegelte Tontafel der altsumerischen Zeit, AfO 24 (1941–44) 333–336, for a discussion of d Í d - d a in early river ordeal texts from Nippur. These texts are Edzard SR nos. 98 (= TuM 5, 49, with two river ordeal cases) and 99 (= TuM 5, 159, with seventeen cases); for discussion of these texts, see also Frymer-Kensky, The Judicial Ordeal 66–84. For the locations of the river ordeal, see ibid., pp. 93–96. 526–527; W. H. van Soldt, Ordal. A, RlA 10 (2003) 126, with previous references. ARM 7, 163: 5. See n. 51 for commentary on Bottéro’s suggestion that this is a toponym (ibid. 334. 343–344; RGTC 3, 43), which, even if correct, would likely derive from a temple name, as the author concedes. Frymer-Kensky, The Judicial Ordeal 162–175. Cooper, Sumerian and Akkadian Royal Inscriptions 87 (Ma 2.2), with previous literature. 35 On the Euphrates theophoric element in the aforementioned Mari PN I-dì- díd. For the sˇak kanakkum- period, devotion to díd is demonstrated by a votive inscription known from a NB copy of an original found, revealingly, at Sippar.126 The inscription opens with the claim that Itlal-Erra, king of Mari, the son of Puzur-Eštar, erected this statue before his lord díd, maš.tab.ba, and Ištaran. Clearly, a common judicial aspect unites these gods, but díd is given particular prominence not only by his appearing first among this triad, but by the fact that he is singled out as “his lord,” implying that the River-god enjoyed particular royal devotion at Mari – a supposition that is further strengthened in light of Ikun-Šamagan’s offering.127 It will of course be recalled that it is at third-millennium Mari where we find a cult to the Euphrates, as shown by the offerings made to dkib.nun.a/ dkib.nun.a as well as the glyptic evidence that presents an unparalleled rendering of this twin Euphrates-goddess (fig. 8). Royal veneration of díd, however, was not short-lived. Several centuries later, Zimri-Lim would write a letter to díd addressing him once again as (ana) beliya, soliciting his sign and protection.128 And it was this same Zimri-Lim who, in his “Investiture” Fresco, based his kingship, quite literally, on the fertile waters that derive from the Euphrates (fig. 11). Upriver of Mari at Tuttul (Tell Bi>a), administrative texts of the sˇakka nakkum-period specify the delivery of oxen and sheep offerings to the é (I)Na-<à-rí-im ‘temple of the River(-god)’, recalling the bit Narim attested at Mari. In one text, the temple of the River(-god) outstrips the temples of Dagan and Annunitum in importance, appearing first in the list and receiving two sheep to their one.129 Particular devotion to díd may be shown to extend downriver of Mari as well. The city Hit – the primary locale for the river ordeal130 with close links to Mari – reveals a tacit identity with the River-god in its very spelling, i. e., dídki . But more than a mere orthographic play, both the city, also written phonetically I-da ki and I-ta-i ki ,131 and the pitch for which it is famous, ittû ‘bitumen’, probably derive from Sumerian í d ‘river’, making sense of a late religious text which 126 127 128 129 130 131 Gelb/Kienast, FAOS 7, 366–367. The colophon reads, ki-ma pí-i na na.dù.a libir.ra sˇá ina Sip-pár ki I Re-mu-tum dub.sar bàndada dumu I Su -{x}-… ú-sab-bi-ma is-su-[uä] “According to the wording of an old stone stele which Remûtum, the junior scribe, the son of …, examined and copied in Sippar.” See already Frymer-Kensky, The Judicial Ordeal 177. Syria 19 (1938) 126; discussed by Frymer-Kensky, The Judicial Ordeal 177–178. MVDOG 100, 25. 27 (including the temples of Dagan and Annunitum). Heimpel, The River Ordeal at Hit, RA 90 (1996) 8; Durand, ARM 26/1, 521–523; idem, Les documents épistolaires du palais de Mari 3. LAPO 18 (Paris 2000) 151. For the writing of this GN, see RGTC 3, 104. 4 36 Christopher Woods associates d Í d with ittû ‘bitumen’.132 Hence the likelihood that Hit is in origin the ‘River(-god) city’. Parallels to cities taking their names from the rivers on which they lie may be sought in the GNs Äabura(tum) near the source of the Äabur, and Baliäu(m), perhaps identical to the OB toponym Apqu ša Baliäa,133 which, as the name itself bears witness, is at the headwaters of the (two) Baliä rivers. To these we may add, of course, Sippar, although in this case the identity with the river is only orthographic. Downriver of Hit we come finally to Sippar. It was here that the sˇakkanakkum-period king of Mari, Itlal-Erra, deposited a statue before íd, his lord, maš.tab.ba, and Ištaran. Similarly, Šulgi, claiming the particular patronage of the River-god with the opening phrase ana díd beliya – also used by Itlal-Erra and Zimri-Lim – left a foundation inscription dedicated to díd, written in Akkadian.134 This fragmentary inscription was found at Tell ed-Der, the Sippar which lay north of the river, and most likely detailed the building of a temple to díd in the immediate vicinity. The area maintained a special relationship with the river at least through the second millennium, for a site just upstream of Sippar, in a text to be dated not earlier than Nebuchadnezzar II’s reign, is specified as a location for the river ordeal.135 Clearly, in all these instances, from Tuttul to Sippar, díd refers to the deified Euphrates. Based on the substantial evidence from Mari, FrymerKensky has argued that the divine river was worshipped with special devotion along the stretch of the middle Euphrates between Mari and Hit.136 But the evidence from beyond Mari shows that this special devotion to the River-god in the third millennium extended upriver to Tuttul, and perhaps to Ebla in the form of the deified Baliä, but certainly and 132 133 134 135 136 PBS 10/4, 12 ii 18; this text and the relationship between d Í d , ittû, and Hit is discussed by Frymer-Kensky, The Judicial Ordeal 179–180. RGTC 3, 20. For discussion of the location and meaning of this toponym, see A. Goetze, An Old Babylonian Itinerary, JCS 7 (1953) 57. 61; Hallo, JCS 18 (1964) 77–78. D. Frayne, RIME 3/2, 137 (Šulgi E3/2.1.2.29). CT 46, 45 edited by Lambert, Iraq 27 (1965) 1–11: ma-äar-sˇú ib-bab-lu-nim-ma elisˇ Sippari ki kisˇad ídPuratti ma-äar dÉ-a sˇàr Ap-si-i ú-ma-<-ir-sˇú-nu-ti … “They were brought before him and he sent them upstream from Sippar to the bank of the Euphrates to the presence of Ea, king of the Apsû …” (iii 22–24); see also P.-A. Beaulieu, A Note on the River Ordeal in the Literary Text ‘Nebuchadnezzar King of Justice’, NABU 1992/77. W. Heimpel (RA 90 [1996] 9) claims that elisˇ Sippari ki is a reference to Hit, but the argument is circumstantial and no direct evidence is provided to support it. Note that the attribution of this text is debated – see H. Schaudig, AOAT 256, 579–580. Frymer-Kensky, The Judicial Ordeal 178–180. On the Euphrates 37 more importantly for our purposes, downriver past Hit and encompassing the area of Sippar in northern Akkad. In the wake of this evidence we are finally in a position to offer a hypothesis concerning the relationship between Sippar and the Euphrates, namely, that the location of Sippar was early on a holy site associated with the river. The basis of this association is simple geography. Sippar boasted a relationship with the river that no other Mesopotamian city could claim, for it is in the vicinity of Sippar that the Euphrates fans out, sending tendrils – the Zubi, Irnina, Araätum, and Abgal137 – down into the lower alluvium. This topographical reality is reflected in the GN Birit Narim situated in the Sippar region138 and captured in a fragmentary Old Babylonian map that depicts the city wedged between the Euphrates and one of its lesser branches or canals, ídtap-pí-isˇ-tum.139 As we have seen, 137 138 139 As reconstructed by Carroué, ASJ 13 (1991) 111–156, particularly figs. 2 and 5; cf. the earlier reconstruction of T. Jacobsen, The waters of Ur, Iraq 22 (1960) 176–178, followed by D. T. Potts, Mesopotamia Civilization: The Material Foundations (Ithaca 1997) 26. Jacobsen included the Iturungal, with its branches Í d - N i n a ki -du- a and N a n n a - g ú - g a l ; however, recently P. Steinkeller has persuasively argued that the Iturungal branched off from the Tigris, only joining the Euphrates below Uruk (ZA 91 [2001] 41–49). MHE 2/2, 2/4, 2/6 s. v. H. Gasche/L. de Meyer, Ébauches d’une géographie historique de la région Abu Habbah/Tell ed-Der, in: L. de Meyer (ed.), Tell ed-Der 3 (Leuven 1980) 6 fig. 3. While any suggested derivation of Buranuna is necessarily speculative, as with so many Mesopotamian toponyms, a possible etymology presents itself that, if correct, is revealing in terms of the inherent geophysical characteristics of the Sippar region. The final syllables suggest, plausibly, a Sumerian form ending in - n u n - a ( k ) ‘princely, of the prince (Enki-Ea)’. A convincing argument can be made by considering the first element as a loanword from a vocable related to Akkadian buru, burtu, meaning ‘well, cistern’, and in its broadest, and perhaps extended, sense ‘water source’, i. e., b u r a n u n - a ( k ) ‘preeminent water source’ (for attestations of the root b<r in the various Semitic languages, see D. Cohen, Dictionnaire des racines sémitiques [Paris 1976] s. v. B<R). Although the rationale for the association is unclear, the referent of this lexeme, bur, as argued by M. A. Powell, was considered a natural unit, hence, the area measure buru, and possibly the linear measure ber (Sumerian Area Measures and the Alleged Decimal Substratum, ZA 62 [1972] 209–210; idem, Maße und Gewichte, RlA 7 [1987–90] 480). One might surmise that originally the metrological unit referred to the average area or distance between or containing watering holes. That bur made its way into Sumerian as a loan, as our analysis requires, is shown by the Sumerian surface measure b u r, which, there can be little doubt, is related to its Akkadian counterpart bur. The loan, as convincingly argued by Powell, must be in the direction of Akkadian into Sumerian (ZA 62 [1972] 210). Thus, Sumerian b u r is one of a number of likely metrological loans from Akkadian including mana, lidga (Powell, RlA 7 [1987–90] 495), and possibly iku (the last also relating area to water sources, in this case, presumably, the area enclosed by a boundary canal or dike, cf. Sum. e g , 38 Christopher Woods Sippar’s relationship with the river manifested itself in the worship of the Euphrates, evidence of which can be found in the Itlal-Erra and Šulgi inscriptions, and in the very writing of Sippar with a graphic composition, kib.nun, borrowed from the spelling of the Euphrates.140 Finally, the bond between the city and the river may find expression in what appears 140 Akk. iku and ikû; see Powell, ZA 62 [1972] 205–206). Additional evidence for the loan into Sumerian presents itself if – given the semantic and phonological similarities – Sum. b ù r u can be convincingly derived from Akk. buru. Indeed, a relationship between the two is made explicit by the equation bu-ru : u = bu-rum (A II/4: 93 [MSL 14, p. 282]); cf. Powell, ZA 62 (1972) 210–211 n. 128. Originally, Buranuna may have referred to the area around Sippar, a manifestation of díd that was particular to the unique geomorphological conditions surrounding Sippar with its radiating river branches – the ‘preeminent water source’ (cf. the toponyms Äabura and particularly Apqu ša Baliäa discussed above). Since the designation was non-specific to any one branch, the region easily lent its name to the stretches south and north of Sippar, becoming a poetic designation for the entire river system. In this regard, it must be pointed out that a number of geographical names incorporate the lexeme bura – many of which are located in the vicinity of Sippar, e. g., Buratum (= NA Burati [RGTC 3, 46]), Bura-imdidi (Harris, Ancient Sippar 372), the waterway Buri (RGTC 3, 277), and in association with the last, the well-attested ugaru ‘watered field’ Bura located between the Euphrates and the Irninna (RGTC 3, 46; for the location of this ugarum, see M. Tanret, Le namkarum. Une étude de cas dans les textes et sur la carte, in: H. Gasche/M. Tanret [eds.], Changing Watercourses in Babylonia 1. MHEM 5/1 [Ghent 1998] 76). L. Dekiere has argued that the /a/ vowel is the adverbial -a found with measures that connotes a distributive sense (Quelques notes sur les noms d’ugaru, NAPR 10 [1996] 3; also M. A. Powell, The Adverbial Suffix -a and the Morphology of the Multiples of Ten in Akkadian, ZA 72 [1982] 89–105). The relationship between the area measure bur and the GN Bura is demonstrated by the writing of the latter logographically as 1.0.0 i k u . t a ( . à m ) (L. Dekiere, NAPR 10 [1996] 3). Of course, not to be overlooked in this discussion is the best known GN to claim a possible derivation from the Semitic root B’R, namely *Bi<rutu – Beirut – the logographic writing urutúläi.a demonstrating that at least for the Semites of the second millennium this was indeed the etymology (RGTC 12/2, 56–57; D. Pardee, Trois comptes ougaritiques, RS 15.062, RS 18.024, RIH 78/02, Syria 77 [2000] 60 n. 185). For additional GNs based on this root, see C. H. Gordon, Ugaritic Textbook, AnOr 38 (Rome 1998) 370: 437. In this connection, note the derivation of Tuttul discussed below. The etymology of Sippar is obscure, although Hallo draws a connection with siparru, when he writes “Sippar, the city of bronze” (JCS 23 [1970–71] 65), presumably the brilliant alloy symbolizing the sun. Support for this as a folk-etymology at least, if more than a mere sound play, may be sought in the variant spellings given in the previously discussed ARET 5, 6; OIP 99, 326+342, where zi !-bí-ra of the Ebla text corresponds to zabarx(ka+bar) of the Abu Salabiä copy in the passage érin+X du-sa aš murub4 zi !-bí-ra < šag4.gíd a.si.sá > <dutu> min murub4 < zi-bí-ra > šag4.{gíd} a.si[.sá] “érin+X du-sa in wisdom governs one side of Sippar, Šamaš in wisdom governs the other side of Sippar” (Krebernik, QdS 18 [1992] 81: C17, after the Ebla copy) – possibly a reference to Sippar-Amnanum and Sippar-Jaärurum of OB renown (ibid. 86 n. 10). On the Euphrates 39 to be a relatively high proportion of watercourse personal names at Sippar, particularly those referring to the Euphrates and the Araätum, e. g., Abdi- díd; díd-abi; díd-di.kud; díd-rabi; Ipiq-Idiglat; Ipqu-Araätum; MarAraätum; Marat-Araätum; Mar- ídPurattim; Narum-ili; Ummi-Araätum;141 idSilakum-ummi.142 Indeed, it is difficult to envision a location more at the mercy of the river. Traversed by a network of tributaries, shifting, branching, and rejoining – the off-shoots of a natural canal flowing towards the Tigris – the landscape is rutted by natural levees and basins. It is upon one such large levee that the settlements of Sippar and Tell ed-Der were founded, this high point providing the only protection from the devastating floods. Sand was heaped upon the levee so that the settlements were not surrounded by the city walls found elsewhere, but by dykes, which at Tell ed-Der reached more than 15 meters in height, perhaps the most telling fact of the relationship between the site and the river.143 The Tigris may also lay some claim to the inherent numinosity of the region. It has often been suggested that in prehistoric times, and perhaps as late as the proto-literate period, the Tigris and Euphrates joined in the area of the nearby Aqar Quf depression, creating a single great river in the vicinity of Sippar.144 Paepe has gone so far as to postulate that this link between the two rivers existed as late as the end of the fourth millennium: “A link was thus existing at Sippar between the two rivers, Euphrates and Tigris. It was probably that [sic ] after pushing the Tigris to a more eastern position, but still at the time contact between the two rivers existed, that civilization entered this part of the flood plain.”145 If this is so, it is all the more transparent why early settlers would have regarded the area as invested with a riverain numinosity. However, even if the rivers separated long before the region was inhabited, it is nevertheless a topographical fact that of the traditional major cult centers, only Sippar had the distinction of being situated near the point where the Tigris and Euphrates make their closest approach, lying on the very canal that joins the two. 141 142 143 144 145 From E. Woestenburg, OB Namenlijst (unpublished). MSS rev. ed. of previous work by G. Th. Ferwerda, Leiden. BM 79951; I thank S. Richardson for this reference (see also RGTC 3, 307). R. Paepe, Geological Approach of the Tell ed-Der Area, in: L. de Meyer (ed.), Tell edDer 1 (Leuven 1971) 21–23. Paepe, Tell ed-Der 1, 9–27; see also Adams, Heartland of Cities 16; McG. Gibson, The City and Area of Kish (Coconut Grove 1972) 22. Paepe, Tell ed-Der 1, 25. 40 Christopher Woods Geography also explains the orthographic identification of the city Hit with díd and this site as a riverain numen loci, for Hit marks the true beginning of the alluvium – at this point the Euphrates emerges from its deeply cut valley, thereby making gravity-flow irrigation, and thus life itself, possible in lower Mesopotamia. Wilkinson has pointed out that the region between Sippar and roughly Hit (specifically Falluyah), is a geomorphological unit, representing the boundaries of one of the principal “nodes of avulsion” in the alluvial lowlands, where the river decreases in slope and has a tendency to rise above the plain and break its banks, creating new channels.146 North of Hit the river is too deeply incised to permit gravity-flow irrigation. Thus, the long 250 km stretch of the Euphrates upriver of Hit is devoid of major cities until one comes to Mari, which is still well below the 250 mm isohyet, but where a number of topological factors conspire, including the confluence with the Äabur and the particular morphology of the valley and its terraces at this location, to allow sufficient irrigation to support large-scale settlement.147 Hence, the royal devotion enjoyed by díd at Mari appears to be based, again, on the geographical reality. The same, of course, holds true for Tuttul, which was uniquely situated at the confluence of the Baliä and the Euphrates. At all these sites – Tuttul, Mari, Hit, and Sippar – the particular veneration enjoyed by the River-god is a function of geography. And, as the river is literally identified with Hit, Tuttul, a toponym identifying not only Tell Bi>a but also Hit, is perhaps to be understood as a reduplicated form etymologically connected to Akk. tultu, Sum. t ú l , as suggested by the OB morphographemic spelling Tu-ul-tu-ul ki, i. e., ‘wells’ or perhaps better ‘water sources’, thus explaining the use of this place name for two locations, both defined by their geomorphological relationships to the river.148 146 147 148 T. J. Wilkinson (personal communication). On the geomorphology of this region, see S. W. Cole/H. Gasche, Second- and First-Millennium BC Rivers in Northern Babylonia, in: H. Gasche/M. Tanret (eds.), Changing Watercourses in Babylonia 1. MHEM 5/1 (Ghent 1998) 1–64; K. Verhoeven, Geomorphological Research in the Mesopotamian Flood Plain, ibid. 159–245. B. Geyer, Géomorphologie et occupation du sol de la moyenne vallée de l’Euphrate dans la région de Mari, MARI 4 (1985) 27–39; P. Sanlaville, L’espace géographique de Mari, MARI 4 (1985) 15–26; J.-C. Margueron, Mari, l’Euphrate, et le Khabur au milieu du IIIe millénaire, The Canadian Society for Mesopotamian Studies Bulletin 21 (1991) 79–100. With some skepticism, Frymer-Kensky allows for a possible topographical explanation for the presence of the river cult at Mari (The Judicial Ordeal 180). [Tu ]-ul-tu-ul ki = šu = uru I i-tú-it (MSL 11, 35: 23 [Hg.]). The various spellings of Tuttul are discussed by M. Krebernik, Ausgrabungen in Tall Bi>a/Tuttul – II. WVDOG 100 (Saarbrücken 2003) 3–4. In this connection note the above discussion of the toponyms Äabura and Apqu ša Baliäa and the suggested etymology for Buranuna (n. 139). On the Euphrates 41 Numerous parallels to this phenomenon may be found in upper Mesopotamia, where the phenomenon of the numen loci was an integral aspect of early Semitic religious thought. An instructive case is provided by the city of Aššur. If Lambert’s compelling hypothesis is correct, then the god Aššur derives from the numinous quality attributed to the site of Aššur. In prehistoric periods, he suggests, the location – a hill allowing control over the surrounding plain – was regarded as a holy spot, and the early inhabitants “exploited the holiness of their place by converting the ‘mountain’ into a city.”149 A similar development might be posited for Sippar. During the prehistoric periods the site of Sippar was regarded as a numen loci on account of its unique topographical relationship with the Euphrates. The locale was of great strategic importance for controlling not only traffic on several branches of the river radiating downstream from Sippar, but also the east-west trade routes between the Euphrates and the Diyala valley. It thus served as a crucial entrepôt between the alluvium and upper Mesopotamia. The importance of Sippar for trade is well documented for the Old Babylonian period150 and it is reasonable to assume that these same inherent characteristics of location that were so effectively exploited in the early second millennium played an important role in the occupation of the site during the late Uruk period. Indeed, already in the aforementioned Abu Salabiä literary composition, OIP 99, 326+342, Šamaš of Sippar is associated with mercantile activity. Surface surveys have allowed Adams and Gasche to date the onset of occupation at Sippar to at least the end of the Uruk period, c. 3300.151 Plausibly, it was at this time, during the so-called Uruk expansion, that the cult of the Sun-god was established at this strategic and holy site. For Adams and Carroué, as we have noted, orthographically speaking Sippar is primary and the river secondary, the latter borrowing its writing from the former. The significance of Sippar, the argument continues, lies in the fact that it is the first city encountered as the river debouches onto the alluvial plain.152 But, in actuality, the deictic perspective must be reversed. From the southern perspective Sippar marked the end of the easily habitable world and early settlers, dependent as they were on irri149 150 151 152 Lambert, Iraq 45 (1983) 85–86. W. F. Leemans, Foreign Trade in the Old Babylonian Period (Leiden 1960) 85–112. Note that in the OB period silver is advanced a-na kaskal ídburanuna “for the Euphrates trade” (VS 22, 35: 2. 39: 2. 40: 2. 44: 2. 49: 20); I thank P. Steinkeller for these references. R. McC. Adams, in: Gibson, The City and Area of Kish 192: 058; H. Gasche, Le système paléo-fluviatile au sud-ouest de Baghdad, BSA 4 (1988) 42. Adams, Heartland of Cities 3; Carroué, ASJ 13 (1991) 121. 42 Christopher Woods gation, no doubt recognized the alluvium as a distinct geographical region with specific boundaries.153 The establishment of a cult to the Sungod in a location revered for its relationship to the River-god would have been greatly facilitated by the common aspect of divine judge that defines both deities – the sun god’s ability to expose all to the light of day and the river’s cleansing power to clear the falsely accused. Further, Sippar, in the far reaches of the alluvium, on the very horizon of urban Babylonia, would have been a natural locale to establish a cult of the Sun-god, whose daily travels took him to the ends of the world, the god under whose aegis the fate of travelers rested: sˇa ruqqat kimtasˇu nesû alusˇu / [ina ] sˇurubat seri re<û imaääarka “He whose family is remote, whose city is distant / The shepherd amid the terror of the steppe confronts you (Šamaš).”154 Perhaps we are to attribute a synergistic result to the close association of the Sun- and River-gods in the vicinity of Sippar, one that served to enhance or magnify the judicial character of the former. While the aspect of divine judge was certainly inherent to the all-seeing, all-revealing Larsan Sun-god, there is some evidence to suggest that this characteristic may not have been as pronounced as with the Sipparian deity.155 In the temple hymns, for instance, it is said only of Šamaš of Sippar, d i k u d - r u k i u d è … “He (Utu) pronounces judgments at the place where the sun rises …” In the Larsa hymn, on the other hand, there is no hint of the god’s aspect of divine judge.156 And it will be recalled that at Umma, in the Ur III period, it is L u g a l - Z i m b i r k i of far-off Sippar and not Utu of Larsa who resides in the city, suggesting, perhaps, a meaningful distinction between the two. Even in the early Old Babylonian period, when any original difference would have been blurred, mention of Utu as divine judge is not as frequent as one might expect in the Larsa royal inscriptions.157 For the Sipparian deity, however, the earliest piece of Se- 153 154 155 156 157 See already the comments of Adams, Heartland of Cities 3. Šamaš Hymn 135–137, after Lambert BWL 135. See already Myers’ discussion of the differences between Utu of Larsa and Šamaš of Sippar (The Sippar Pantheon 4. 280). TCS 3, 27 (Larsa). 46: 489 (Sippar). I. e., d i n í g - g i - n a dU t u - t a “by the true judgment of Utu” (RIME 4, 149: 77 [Nur-Adad]); [… k a - a š - b ] a r a n k i … d [ i - k ] u d [ … ] - r a t i - l a u [ g 5 - g a è n t a r ] - b i - i m “[ jud]ge of heaven and earth … j[ud]ge … [who cares for] the living and the de[ad]” (ibid. 157: 7, 10–11 [Sîn-iddinam]); d i - k u d a n k i “judge of heaven and earth” (ibid. 163 [Sîn-iddinam]); d i - k u d s i g i g i - n i [ m - m a ] (ibid. 167: 4 [Sîn-iddinam]); d i - k u d - m a h a n k i “supreme judge of heaven and earth” (ibid. 221: 33 [Warad-Sîn]). On the Euphrates 43 mitic literature points to this aspect as a defining characteristic already for the mid-third millennium.158 Certainly, the character of the Sun-god in Akkad was, on some level, a composite, representing an early syncretization of Sumerian Utu with a Semitic Sun-god;159 but the judicial aspect of the latter appears to have been practically non-existent in the west, suggesting that the prominence of this characteristic of the Sipparian Sun-god was not merely the vestige of an early Semitic religious conception. The River-god, too, likely represents an early syncretism between Sumerian and Semitic gods. But in light of the evidence discussed here, which shows the veneration of the River-god in the north to be a function of topography, and the role of the numen loci in the early Semitic pantheon, it would appear that díd at Sippar – whom Šulgi invokes in Akkadian, rather than in Sumerian, and whom an OB personal name identifies as Narum160 – gravitates more toward the Semitic than to the Sumerian religious sphere. Certainly, dÍd is well represented at Nippur and to a lesser degree elsewhere in Sumer. But his presence in the south is almost entirely tied to the river ordeal and the River-god does not there appear to be topographically bound, nor does he enjoy the royal patronage or importance in offering texts known from northern sources.161 Of course, 158 159 160 161 <a5(ni).nun.gi di.kud nam.guruš i-ga-sar … i-ba-äar dutu íd !(a.lagab×an) nammu ù dištaran “The Anunna gods, the judges of the young men, he assembles … Šamaš, the River(-god), Nammu and Ištaran assemble” (Krebernik, QdS 18 [1992] 78: C11.3–4, after the Ebla copy) – note that a common judicial theme underscores the gathering of these gods – see Woods, JCS 56 (forthcoming). Although Sun-goddesses are well attested in the Semitic pantheon (note Ugaritic Šapš), there is no firm evidence for a female Sun-god in Mesopotamia (cf. Roberts, Earliest Semitic Pantheon 52). The OAkk. PNs Tamäur-Sˇamasˇ, Tulid-Sˇamasˇ, UmmiSˇamasˇ are ambiguous as to the gender of the deity. The possibility exists in the case of the first two that a preposition is omitted, or, more likely, that the gender agreement is with the name bearer and not with the deity, i. e., female PNs, although admittedly personal names with a masculine deity and female predicate are not otherwise known (see D. O. Edzard, mNingal-gamil, fIštar-damqat. Die Genuskongruenz im akkadischen theophoren Personennamen, ZA 55 [1962] 127); the third name may likewise be an abbreviation, feminine (cf. U t u - a m a - m u [HSS 3, 21 vi 22 (PN-fem.)] and U t u - a - m u [Edzard SR 60 iii 5]), or, perhaps, even metaphorical, e. g., a m a n u t u k u - m e a m a - m u z é - m e a n u - t u k u - m e a - m u z é - m e “I have no mother, you (Gatumdug) are my mother, I have no father, you (Gatumdug) are my father” (Gudea Cyl. A iii 6–7). Na-ru-um-dingir (CT 4, 50b: 8 [OB Sippar]), cited above. On the role of Ilurugu in Ur III administrative texts, see Frymer-Kensky, The Judicial Ordeal 86–93; in contrast to the Mari and Tuttul offering texts discussed above, Ilurugu receives fewer offerings than the other gods with whom he is listed. 44 Christopher Woods the ordeal is first attested in Sumerian texts and the River-god is known throughout Mesopotamia by the Sumerian designation (d) Í d ; however, it does not necessarily follow that the aspect of judge originates with the Sumerian god. Indeed, if the relatively late evidence from far off Ugarit may be brought to bear on this issue, then a judicial aspect was not at all foreign to the character of the Semitic River-god, as shown by the standard Yamm epithet, ©pt nhr, often translated as “Judge River.”162 Although the root ©pt, like its Akkadian counterpart, sˇapatu (m), is closer to the meaning ‘to issue decrees’ than ‘to judge’ in the sense of dianu (m),163 the epithet nevertheless meshes well with the role of the River-god in the ordeal, and in Akkadian, at least, the verbs were considered semantically close enough to be equated in a synonym list.164 Finally, the relationship between the River-god and the Sun-god at Sippar described in this paper may go some way to elucidating the major themes of the ‘Sun-god in his boat’ motif of third-millennium glyptic and its literary counterpart, the Early Dynastic Šamaš myth (ARET 5, 6; OIP 99, 326+342). The image invariably includes a plough, various vessels, and, of course, Šamaš and his boat, as well as, occasionally, a sprig of vegetation.165 Likewise, the text, centering on Šamaš of Sippar, refers to a boat, a plough, as well as mercantile activity involving a variety of foreign products, many of which, including aromatics, oils, and honey, can only be visually represented by the vessels that hold them. Both the image and text are of northern, Semitic origin, likely describing the same Semitic mythologem.166 It has even been suggested that the text was composed at Sippar, well-known as a scribal center.167 It is tempting, therefore, to see 162 163 164 165 166 167 C. H. Gordon, Ugaritic Textbook. AnOr 38 (Rome 1998) 506; W. G. E. Watson, Ugaritic “Judge River” and the River Ordeal, NABU 1993/95. See D. Pardee, in: W. W. Hallo (ed.), The Context of Scripture 1 (Leiden 1997) 245–246 n. 36. sˇa-pa-tú = da-a-nu (An VIII 187 = IX 1 [catchline]; see CAD Š/1 sub sˇapatu A). There is some evidence to suggest that the river ordeal was known at Ugarit and, moreover, that the god in question was referred to by the name Naru (m), not Id, i. e. tá-me-e a-na na-ri (Lambert BWL 116: 3; also Frymer-Kensky, The Judicial Ordeal 248–251; but see PRU 3, 316 for the interpretation of na-ri as a haplography for na-na-ri [van Soldt, RlA 10 (2003) 124]). Further arguments for the river ordeal at Ugarit are provided by Watson, NABU 1993/95. Much of my thinking connecting these images with the River- and Sun-gods at Sippar has been shaped by numerous discussions with P. Steinkeller. A brief comparison between the myth and the image is made by Steinkeller, QdS 18 (1991) 257–258. W. G. Lambert, Notes on a Work of the Most Ancient Semitic Literature, JCS 41 (1989) 25. On the Euphrates 45 this scene, on some level, as a projection of the Sippar reality onto the mythological plane, with its particular topology which allows the Euphrates to fulfill its fertility potential, with its bustling transit trade of oils, resins, aromatics, and wine,168 and, of course, with its River- and Sun-gods of justice. Many of the elements of the image and text find parallels elsewhere. If the central figure of figs. 5 and 6 is the Sun-god, as suggested above, then these scenes may represent relatively late and peripheral renditions of the ‘Sun-god in his boat’ motif. The Boat-god, whose stern often terminates in a snake’s head (figs. 9 and 10), is a symbolic play on the river itself, for the snake is a well-known motif in Mesopotamian metaphorical language, the god Iräan – dmuš – emblematic of the Purattum/Araätum, being the most obvious example.169 This is made explicit in figs. 1 and 4 where the River-god and the Boat-god are one and the same. Fertility, which plays a role in the myth and the image, both incorporating a plough, is underscored in the latter by the occasional inclusion of a vegetation goddess depicted in conjunction with the river (fig. 9). This same vegetation-bedecked goddess, appearing paired rather than singularly and representing the bounty of the Euphrates, also makes her appearance at Mari (fig. 8; see also fig. 11).170 And, as we have observed, the Sun-god and the River-god meet elsewhere on the mythological plain, namely, on the eastern horizon where Šamaš renders his judgments and where the Great River flows. 168 169 170 Based on Old Babylonian evidence, see Leemans, Foreign Trade 127, with references. For the metaphorical relationship between snakes and rivers, see Krebernik Beschwörungen 298–300; Woods, JCS 56 (forthcoming). Frankfort was inclined to see in this deal motif “the nightly journey by which he [the Sun-god] passes beneath the earth from west to east” (Cylinder Seals 68), passing, therefore, through Enki’s subterranean watery realm of the Apsû. Obviously, the riverain explanation offered here need not be at odds with Frankfort’s interpretation as the cosmographical tradition often associates, or even identifies, the cosmic river with the Apsû (see W. Horowitz, Mesopotamian Cosmic Geography [Winona Lake, IN 1998] 338–339). Note also the following astronomical omen: egir.meš sa5.me-ma ídburanunaki dukám nim.meš si.sá.meš “The rear stars are red: the Euphrates will flow and the early (crop) will thrive” (Koch-Westenholz, CNI 19, 194: 152–153). The natural association of the River-god with fertility finds expression on the theological plane as shown by the entourage of díd, which includes dK i - š a g 5 ‘Good-earth’ as his wife, and his son, dŠ a g - z ì g – a deliberate play upon ‘Rising-flood’ and ‘Erect-penis’, both being me4 taphors and instruments of fertility (A n = Anum II 280–281).