The British Museum Nuzi tablets.

• Author: Maidman, M.P.

The very first Nuzi text to be published appeared 108 years ago. (1) That tablet had been acquired by the British Museum. By the time the volume under review appeared, about 160 British Museum Nuzi tablets had been published. (2) Gerfrid Muller's Londoner Nuzi-Texte (hereafter LNT) more than doubles that number. For this achievement and for the industry it represents, (3) we must be thankful to the author. Some thirty-eight additional British Museum texts now remain to be published.

LNT consists of an introduction, catalogue, copies, translations, and notes to 163 texts, assorted indices, and hand copies of the texts. Among the plates, and following the hand copies, there are also drawings of seal impressions. These should make the text publications and editions more useful still.

With such labor expended and broad coverage achieved, one would expect that LNT would quickly become an indispensable tool in the study of the important archives derived from illicit digs prior to 1925. (4) Unfortunately, it must be reported that LNT is a severely defective work and that Muller's copies and transliterations are frequently unreliable, sometimes painfully so.

The general level of the introduction and notes is also shallow and thus unhelpful. Comments are often jejune and betray less than adequate mastery of the literature. However, I shall not dwell on or specify those weaknesses (the heart of the book could, in theory, still pass muster despite those weaknesses). Rather, I shall focus on failures in the copies and transliterations; these are far more serious flaws. These failures are ubiquitous and particularly harmful to the scholar who wishes to employ this material for the reconstruction of ancient history or for other Assyriological aims. Though, ideally, the work should be redone and the results republished, I cannot see this happening within the foreseeable future. Therefore, the bulk of this review consists of corrections of major errors, this in order to render LNT at least minimally reliable. In this way, LNT might still achieve utility.

Corrections are based on my collations of the tablets. Actually, I have collated all the relevant material twice, once before the appearance of Muller's work and once with this volume in hand.

Apart from specific errors, enumerated below, there are general defects in the copies and transliterations which I usually do not note in detail. Sign fragments are frequently--and mistakenly--transliterated as if the signs were completely preserved. Conversely, fully-preserved signs are sometimes rendered as if they were partially broken. In the hand copies, anomalous sign forms are "normalized" as typical forms, yielding inaccurate "copies." (5) Often, cracked, damaged, or otherwise imperfect tablets are "copied" in an idealized form, failing to reflect idiosyncrasies and peculiarities of the object. (6)

There are occasional errors of interpretation which not even a novice to the Nuzi texts should make. Thus, in 22:3, Muller (hereafter, M) interprets the personal name Se-en-na-BE as Se-en-na-til, although this common name is correctly rendered elsewhere in the literature as "Sennape"; the final BE alternates with BI. Other transliterations also betray unfamiliarity with Nuzi scribal practice. (7) This is also troubling, given M's occasional unexplained restorations of lines. (8)

Less serious, but annoying, ends of lines running over from the obverse onto the reverse at times are not copied as ends of the obverse, but upside down as part of the reverse. (9)

Hundreds of minor errors or debatable interpretations have been noticed but not pointed out in this review. These are relatively harmless and probably do not greatly affect primary or secondary interpretation of texts. (10) However, when considered cumulatively, these defects undermine confidence in the text publication as a whole. Having stated the categories of error sometimes, but not generally, noted below, what kinds of mistakes are enumerated? First, weak or incorrect readings are corrected. Then, especially, are included major cases where the copy (hereafter "(c)") and the transliteration (hereafter "(tr)") agree and both are wrong; (11) where signs are omitted in (c) and (tr); (12) where entire lines are omitted in (c) and (tr) alike; (13) where (c) and (tr) disagree and (c) is correct while (tr) is wrong, indicating inaccurate transliteration from the copy; (14) where (c) and (tr) disagree and (tr) is correct while (c) is wrong. (15)

Having documented some of the defects of LNT, it must be said that M's readings are, in some cases, distinct improvements over my own, especially as documented in Maidman 1986. For these improvements, see below, in the list of texts, at nos. 1; 20:3; 66:6; and 98:7.

The remainder of this review is devoted to the individual texts.

No. 1. M's identification of the principal parties to this transaction is surely correct. Maidman 1986:264, is to be corrected accordingly. l. 5. First sign fragment: (c) right; (tr) wrong. l. 7. For "G[U.sub.4].MI," read "G[U.sub.4].AB!." l. 8. For the first six signs, (c) and (tr) read: "2 en-zu-u SU.NIGIN." This is impossible. The correct reading is [...] MIN x x 1 UDU sa. The two unidentified signs are peculiar but definitely not EN and ZU. l. 9. After "-se," it is possible to restore [u] at the damaged end of this line; the damage is not indicated in the copy. l. 13. For "lu-[??]u SIK?[??]", read lu-u [??]al?-pu?[??]. l. 15. After -la, read -ma. (c) and (tr) omit. l. 16. For "GIN," read u?. l. 25. After -a, read -u-ki. (c) and (tr) omit. l. 27. The third last sign is RA, not SA. l. 28. "[ -LU]GAL" consists of a single, trailing horizontal wedge. Above l. 30 there are two S.I.s, not one. After l. 30 and before l. "31" a fragmentary line intervenes: [...]-e / -[y]a.

No. 2. l. 6. (c) and (tr) read [??]u il[??]-ku sa A.SA. Both are incorrect regarding the first signs. The line reads [??]is[??]-pi-ku sa A.SA. l. 7. (c) and (tr) read na-su-u. Both are incorrect. The line reads u-ma-a[??]l-la[??].

No. 3. l. 5. After SE, read MES. (c) and (tr) omit. l. 7. (c) and (tr) read "3 BAN." Both are incorrect. Read 5 BAN.

No. 6. Subsequent to the publication, J. Fincke established a join between this text and no. 151 and reedited the fuller document. See Fincke 1999: 308-10. After l. 16 the upper edge contains a seal impression followed by: [??][.sup.N[A.sub.4]] KISIB [.sup.m]DINGIR-x[??]. (c) and (tr) omit.

No. 7. The following appears immediately below line 5': [...] DUMU A-ri-har-me. (c) and (tr) omit.

No. 9. The line transliterated after "(Lucke ...)" (presumably l. 19, though M's line numbering is confused at this juncture) seems actually to end line 7. In (c), these signs acceptably appear upside down on the reverse, after 1. 35. l. 25. The third sign probably represents r[a].

No. 10. (c) is very schematic. l. 20. After [.sup.m], read Ni-. (c) and (tr) omit. l. 21. [.sup.m] appears before DI. (c) is right; (tr) omits.

No. 11. After l. 16' six more lines appear. IGI begins five of them. Then there is a gap. The final line starts with the head of a horizontal wedge. (c) and (tr) omit.

No. 12. This text is to be compared to Jankowska 1961: no. 58, to which it is closely related. M (p. 53) also notes this text, as well as no. 36:11. The latter reference makes no sense to me. l. 8. The last word of the line is rendered differently in (c) (u-se-el-lu-u) and (tr) (u-se-el-lu-u). Both are wrong; read u-se-bi-lu-u. l. 10. (c) sa [.sup.m]Ke-el-sa!-ri; (tr) sa [.sup.m][??]Ke[??]-el-sa!-ri. I read sa [??]x[??]-el-hu-u.

No. 13. (c) is far from precise. M identifies the text as a letter, but the expected qibima is missing as is a seal impression. It is more like a memorandum. M's reference to "Gadd 76,1" with reference to Sehram-musni is wrong. He probably means Pfeiffer and Speiser 1936: no. 76:1 (= HSS XIV, 20:1; M does refer to this latter citation).

No. 14. l. 7. (c) and (tr) wrongly read, in effect A-WA. Read A-wi-lu. l. 11. (c) and (tr) read SI[G.sub.4]. The correct reading is SI[G.sub.4].MES. l. 12. (c) and (tr) incorrectly read [Ka.sub.4]-zi. The first sign may be [??]Ka[??] but cannot be QA. l. 14. (c) and (tr) read "[Qa.sub.4]-(Rasur)-zi." Both are wrong; read [Ka.sub.4]-ZI-ru. l. 15. (c) and (tr) read MES. Read rather MES! (= DIS). Lower edge. (c) and (tr) place two S.I.s and then 1. "25" on the upper edge. These appear, rather, on the lower edge. Reverse. Contrary to (c) and (tr), the upper one-third of the reverse is preserved on the left...