I offer here a few remarks to the article “More on the Archaic Writing of the Name of Enlil/Nippur” by P. Steinkeller (2010) that appeared in the Festschrift Owen, of which one might rather claim, “more is less, and less is more.” The question of the graphic realization of Enlil/Nippur in the period from Late Uruk to Old Babylonian (ca. 3500-1500 BC) is in fact, after extensive research, still quite fascinating, leading the reader along various paths that touch on early Mesopotamian totemic cult symbols; pantheistic syncretisms, often “marriages made in heaven” by court theologians of the Early Dynastic period; complex paleographic tables of the signs E₂ and KID; the history of Nippur; and the semantics of NUN and EN. Wang deals comprehensively with many of these issues in his recently published AOAT volume Metamorphosis of Enlil in Early Mesopotamia (2011), and this excellent study should be the basis of subsequent discussion. The following restricts itself to the method of presentation in the Steinkeller article, and on the reliability of his conclusions, with but passing notice given to the record of attestations of the sign combinations representing the Nippur pair Enlil and Ninlil in 3^rd millennium texts. The ultimate etymology of “Enlil” is left aside, although it will be obvious from the data presented that the Semitic origin *il-ilī (“god-of-gods”), long believed by Gelb, then popularized by Steinkeller (1999: 114, n. 36) and others (for instance, Michalowski 1998; wisely rejected by Jacobsen 1989), merely perpetuates an ancient folk eymology.

P. 239. “The logograms such as UNUG.UTU = Larsa, UNUG.INANA = Zabalam, UNUG×HA = Nina/Nimin, and ^kulUNUG = Kulab demonstrate the primacy of Uruk as a political and cultural center during the Uruk period.” UNUG is merely one of a string of complex signs and sign combinations based on the sign AB. Thus AB.MUŠ₃/INANNA = ZABALA(~a; of the 38 Late Uruk attestations, a singular instance [W 15771,w] is said to be written with the UNUG sign, but it derives from the reading of a poor photo of an inaccessible and heavily damaged fragment in the Iraq Museum); AB×KU₆ = NANŠE/NINA (12 Late Uruk attestations, one ED I-II, exclusively with AB); UD.AB = ARARMA₂/LARSA (12 Late Uruk attestations, three ED I-II, exclusively with AB). The source of the author's “^kulUNUG = Kullab” seems to be two instances of SI NUMUN UNUG in UET 2, 274 and 279, and thus not from the Uruk period, but also with no likelihood that NUMUN would signify anything other than “seed.” The author does not mention the combination URI₃.AB = URI₅ that, while attested only four times in the proto-cuneiform records, does lead both the lexical Archaic Cities List as well as the Jemdet Nasr Cities Seal that themselves also demonstrate only the presumed “household” referent AB in toponyms, but that offer the interesting graphic indication that the striations incised within the base of the AB sign to form UNUG might indicate the terraces of a stepped tower, or rather the niches of the Anu Ziggurat (cp. Mattthews 1993: 37-38). UNUG is, incidentally, in fourth position in the Cities List and the JN Cities Seal (disregarding Steinkeller 2002: 254-255) and therefore not likely, in the conceptual world of the ancients, to offer a good justification for an Uruk primacy exposed by proto-cuneiform orthography—if anything, one would have to argue that the replacement of AB with UNUG, inconsistently in the ED I-II period (for instance, Ur City Sealings nos. 4 and 9-11 in Matthews 1993: 61-62 and figs. 12-13) and more systematically in the ED IIIa period, signalled a shift of cultural focus to Uruk only after the close of the Late Uruk period ca. 3000 BC. The matter of EN KID_a//NUN_a found in the second (list) and third (seal) entries, respectively, and apparently representing Nibru, did not seem pertinent to the author in this connection (discussed in Matthews 1993: 34; but cp. Wang's argument [2011: 43-46] that the seal entry is to be read EN_x+KID_x^!?.NUN_x^?). Quite aside from any UD.GAL.NUN or rebus speculation, however, the pair is of some syntactical relevance, given the correspondences between Late Uruk Cities List l. 38 (ATU 3, pp. 35 and 147) and ED IIIa Cities witness OIP 99, 21, obv. iii 4 with E₂^?.NUN and KID.NUN, respectively (of the archaic witnesses, ATU 3, pl. 77 W 23998,1 appears to have E₂ NUN_a+E₂, while the other three are damaged, making the reading of the component “E₂” uncertain), on the one hand, and ED IIIa witness SF 23 obv. iii 1 with NIN.KID = NINLIL on the other, suggesting that the local cults of “EN.KID_a” and “E₂ (or KID_a).NUN_a” (Wang's comments [2011: 88] on ED IIIa NIN // NUN are unclear to me) were conjoined in the span between Late Uruk and ED IIIa times to form the union EN.E₂ and NIN.KID in Nippur (EN.E₂^ki). In the full archaic corpus, there are more than 100 instances of “EN NUN_a” and “EN+NUN_a” recorded, many reaching back to the Uruk IV period (cf. expecially the references in ATU 5; EN+NUN_a has, based on its appearance in the list entries Lu₂ A l. 92 and Officials l. 14, usually been considered a professional designation), and 25 of “E₂ NUN_a” (plus 14 in the ED I-II period). We should remember that NUN_a (NUN_b, with a single long horizontal stroke, is a glyph derived from the buckling [young male goat] sign MAŠ), as one of a series of signs representing early cult totems (Englund 1998: 102), is usually considered the standard of the Enki cult of Eridu to the far south, and therefore of Eridu itself—and that the classical Sumerian pantheon as it emerged in the ED IIIa god lists is nowhere to be found in proto-cuneiform texts, so that discussions of archaic cults remain highly speculative, but should also not leave the ground of the textual evidence.

p. 240. “Englund then chose to analyze the sign in question [namely, KID_a of ENLIL] as a variant of KID ...”. It remains a source of some frustration for those of us who have contributed to the Hans J. Nissen-led Uruk project that the basic precept of consistency in readings vs. preference in readings will not set in the minds of some casual users of our data. The proto-cuneiform sign list ATU 2 is most assuredly not “Green and Englund” as Steinkeller cites the volume in fnn. 7 and 9, but was created and published by Margaret W. Green and Nissen; it offers a variety of signs that the authors determined should be gathered under specific designations (and note that the lexical lists volume ATU 3 is authored by Englund and Nissen, whereby major contributions to that volume were made by Green and Nissen well before I became a project staff member in Berlin). As I and others have stated, in the course of transferring the proto-cuneiform transliterations to electronic form, we had to make decisions about a number of these graphic variants that were at odds with the leveling intent of ATU 2. One of many such instances is the case of KID that, in the Green/Nissen scheme, contained at least four relatively distinct graphic forms consisting of a rectangle with varying iterations and orientations of strokes incised within. Semantically neutral graphic variation will of course be greater in the earlier phases of writing systems, but we nevertheless assiduously assigned distinguishing tags to many of the Uruk IV period exemplars that in a more mature stage of decipherment can, where justified, be removed, so that KID signs were initially entered and may be referenced in our online files as KID_a, KID_b (Uruk III forms), KID_c and KID_d (Uruk IV forms; the Uruk III period KID_e was added later; all Late Uruk variants can be accessed via CDLI search). Of these entries, that labeled by us as KID_a was the grapheme which corresponded to the sign I inspected in the Baghdad text ATU 3, pl. 76, W 21126. Whether this KID_a or the graphically similar, but syntactically distinct sign KID_b (a tripartite KID frame but with oblique strokes top to bottom), is the true precursor of KID = reed mat is not clear; the pictographic referent of KID_a may have been something like three wooden slats woven together with reed or rope (see figure 1).

Pp. 240-241. Of the three ED I-II EN.E₂ references offered as corroborating evidence for the line of reasoning in this article, no. (1) is a seal rendering that could be E₂, could be KID_a—one needs merely move on from this seal to the second in Matthews' convenient compilation of Ur City Sealings (1993: 61 and fig. 12 nos. 1-2) to find the presumably intended rendering of KID_a in the sequence Ur – Larsa - Nippur; no. (2) (now CUSAS 17, 104), based on the much better images available at http://cdli.ucla.edu/P251646, could be E₂, could be KID_a; no. (3) may be KID_a, at least according to the hand drawing fig. 3 (the image p. 242 is illegible), but the discussion p. 241 of ED I and II has neither archaeological nor paleographic credibility. Read E₂ or KID_a as you will, all three corroborate nothing at all.

I offer in figure 1 above my own hypothetical paleography of Enlil (and Ninlil) that derives from evidence available through the CDLI. The Uruk IV forms are little more than idle speculation. In Uruk III, “Enlil”/”Nibru” is written EN.KID_a/NUN, whereby the stone text MSVO 4, 73, offers an excellent parallel to the form of EN (“EN_c”) found in the JN City Seal (and note the strong resemblance of this sign to the offering of stacked beveled-rim bowls carried by an officiant standing above the bearded ram in the uppermost frieze of the Uruk Vase), and “Ninlil” appears to be written E₂.NUN (quotation marks signal conventional pronunciations; incidentally, an image of the tablet MSVO 4, 73 = AO 19936 may be viewed at the commercial site and for those of a mind even purchased in limited license for about $100). Both the ED I-II tablets and the City Sealings excavated at Ur contain some evidence for the continued use of the Late Uruk sign combinations for Enlil and Ninlil, and we might posit that the transfer of the KID_a glyph to a seal surface effected its replacement by the more classical KID form found in the ED IIIa writing of Ninlil. Thus, during the initial phases of the Early Dynastic, the short vertical (using the conventional orientation) strokes impressed in the upper and lower registers of the left half of the sign, and the middle register of the right, are being simplified to full strokes top to bottom left and right, but retaining the bold stroke down the middle. The intrusion of the archaic E₂ form in the ED IIIa rending of Enlil cannot be explained on paleographic grounds, unless it was motivated by the syncretism of archaic EN.KID/Enlil and E₂.NUN/Ninlil. In this period, the sign E₂ represents both e₂, “household,” and lil₂ as a component of Enlil, but not of Ninlil (passim, for instance in the same text in ELTS 106, 122-123; MVN 10, 83; RTC 13; etc.—RIME 3/2.1.1.24, ex. 39c obv. 4 is a good example of an early Ur III royal text with exactly the same sign used for both e₂, and for lil₂ of Enlil). (It is unclear why KID in the early periods can also represent or qualify a very large number, for which compare the vocabulary ATU 3, pl. 81, W 20335,3 obv. i 3 [Uruk III]; TSŠ 190 obv. i 2 [ED IIIa]; MEE 3, 72 TM.75.G.2200 obv. 3 [“kid'u”!, for which cp. further the practice text OSP 1, 12 obv. ii 3'-4'], MEE 4, 78 TM.75.G.1678 rev. ii 2 (šar₂-KID = 10×60³?; both late ED IIIb]; and MAD 5, 112 rev. 1 [Old Akkadian surface measure text with fantastically large numbers, including šar₂-KID as a bundling step above šar'ugal, representing 216,000 bur₃ or ca. 5400 square miles—50 times the recorded arable land of neo-Sumerian Girsu; s. Friberg 2005, §4.7 with literature]). The stone tablet RIME 1.9.1.9 ex. 1 iv 3 appears to document use of ^den-KID in an Ur-Nanše inscription, but the line art publication of the Louvre text (CIRPL Urn 26 = AO 3177) requires collation, and the common genetive+ergative marker -ke₄ (=KID) found after ^den-lil₂ in royal inscriptions may have intruded in the orthography of the name itself. The KID form of lil₂ (reading “lil_x”) in Ninlil is replaced only in the Old Akkadian period (BiMes 3, 31 obv. ii 3 is a potential exception, but requires collation—and may read ^dnin-e₂-šum₂, as in SF 1 ii 17), and even the few examples in CDLI files of nin-E₂ are uncertain. We note that the Early Dynastic KID sign with reading ke₄ or ge₂ itself, as genetive+ergative and as an element of the “cult prostitute” kar-ke₄ (ke₄ replacing ED IIIa ke₃/AK in ED IIIb [VS 27, 33 obv. v 14]), held through the Old Akkadian period (but note one Old Akkadian instance of kar-ke₃ in MAD 1, 276 rev. 1), and was replaced by the E₂ form thereafter. Old Akkadian (for instance, Adab 632 obv. ii 2 [and cf. the less common older form in ASJ 14, 102 no. 6 obv. 4], with an often still prominent final vertical impression within the rectangular form) and Lagash II scribes used a form of E₂ (see figure 1, with wedge heads aligned diagonally within the rectangle, eventually dropping the lowest horizontal stroke; this graphic form is common, though not always displayed in published hand copies of 3^rd millennium documents—photographic images are currently available, for instance, for RIME 3/1.1.7.3, ex. 2 l. 8, and DCS 21 rev. 4, to cite both a royal and an administrative text) exclusively for e₂ “household,” in the Ur III period usually for lil₂ in Enlil, and seldom for lil₂ in Ninlil; Ninlil appears to retain the “new” kid (old e₂) form (this is best checked by searching CDLI for image-documented texts with the “dates referenced” entry for Šu-Suen 6, that is, ‘Shu-Suen.06,' which in its fuller form is mu ^dšu-^dsuen lugal uri₅^ki-ma-ke₄ na-ru₂-a-mah ^den-lil₂ ^dnin-lil₂-ra mu-ne-du₃, “year: ‘Šu-Suen, king of Ur, erected the Great-Stele for Enlil and Ninlil',” for instance in AAICAB 1/1, pl. 46, 1911-483, and CUSAS 3, 37). Old Babylonian scribes writing the name ^den-lil₂ may have merged the older form of E₂ (that is, new KID) for lil₂ still used for Ninlil in the Ur III period with that used for Enlil known from Old Akkadian, and Ur III royal inscriptions on stone. A clear and archaizing (Ur III) example of the distinction of the two signs is evident in the stone insciption of the Codex Hammurapi stele, comparing, for instance, i 3 ^den-lil₂ with ii 1 e₂ abzu.