The Etymological Dictionary of Akkadian (EDA) provides the first fruits of a long-term project conducted since June 2013 at the Friedrich-Schiller-Universität Jena and the National Research University – Higher School of Economics, Moscow. Its usefulness is self-evident. Of the two major Akkadian dictionaries, only von Soden's Handwörterbuch, completed in 1984, includes consistent etymologies. Both new texts and the re-reading of old have produced a great deal of new material. The reasons for the lack of such studies are equally evident: the work demands a mass of philological minutiae and languages rarely of interest to most Semiticists, while Assyriology's long-term divorce from Philologia sacra has made the prerequisite knowledge of Semitic languages increasingly rare. It need hardly be mentioned that the authors are uniquely positioned to address these issues.
The volume begins with a general overview of Akkadian, its relationship to Semitic, and remarks on organization and use, including a methodological discussion of the thorny distinction between Fremdwörter and Lehnwörter (p. 21). The literature cited is minimal: newer references are generally missing, e.g. V. Meyer-Laurin, “Zur phonologischen Rekonstruktion von ‘Schin’ <Š>”, Altorientalische Forschungen 43, 2016, 77–146. The influence of A. Militarev and L. Kogan's earlier Semitic Etymological Dictionary (Münster: Ugarit-Verlag, 2005) can be seen throughout. The dictionary's most important feature is the use of “cover symbols” for the root consonants in the headwords: ↄ for ʾ, w, or y; H for ʾ or ḫ; K for k, g, or q; L for l or r; M for m or n; P for p or b; S for š, s, z, or ṣ; T for t, d, or ṭ; X for an unclear consonant; and ° for “a consonant whose presence is not certain” (p. 22). Suggested (phonemic) roots are separated by a dot: a relatively straightforward root such as balāṭum “to live” thus appears under the headword P-L-T.b-l-ṭ (P0596, pp. 336–8). The need for constant recourse to the indices to discover where a particular Akkadian word is sorted into this system makes the volume as a whole frustratingly cumbersome to use.
Each numbered headword (P0001–P1001) is followed by an indication of etymological background: 1. Inherited Semitic; 2. Internal development; 3. Foreign or loanword; 4. Unknown, all with various subgroups (p. 26). A list of derivatives, distribution, and references to the dictionary meanings in their respective sub-entries follow. The task of updating the lemmata was assigned to an unfortunately acronymed parallel project; see M.P. Streck, Supplement to the Akkadian Dictionaries, Vol. 1: B, P (Wiesbaden: Harrassowitz, 2018), 48. Each entry concludes with notes and an etymological discussion. Particularly noteworthy is the attempt to include Ethiopian and Modern South Arabian languages (pp. 621–35 in the index).
It is easy to quibble with many details. The distribution given for individual entries often requires revision, affecting any assessments of long-term lexical trends. Against the definition “V lex.” (first millennium lexical) assigned to napšurtu (P0828), see ina guḫaṣṣe ša lā napšurti “with the cord which cannot be released” in the bilingual van Dijk, Lugal, pl. 64 ff. iv 9′, discussed in S. Seminara, La versione accadica del Lugal-e (Rome: Dipartimento di Studi Orientali, 2001), 371. The bibliography is often spotty. Sabaic is usually referenced through A.F.L. Beeston et al., Sabaic Dictionary/Dictionnaire sabéen (Louvain: Peeters, 1982), with additions from A. Sima, Tiere, Pflanzen, Steine und Metalle (Wiesbaden: Harrassowitz, 2000). No reference is made to the updated material of the online Sabäisches Wörterbuch (http://sabaweb.uni-jena.de/).
The principle “separate until proven equal” leads to a separation of the entries bukru “son, child” (P0255) and the Early North Arabian loanword bakru “young camel” (P0256), both from the same root, even if they entered the lexicon through different paths. There is no reason why reference to David Cohen's entry in Dictionnaire des racines sémitiques 2 (Paris: Mouton, 1976), 64 should have been omitted; this work had already collected most of the material. Conversely, the entry on binu “son” and related lexemes (P0639: P-M.b-n.1) lists the root as either common Semitic or as a West Semitic loanword and suggests that its use may have been bolstered by “speakers’/writers’ knowledge of a highly prominent cognate”, since “the great majority of the attested examples are found in the first millennium literary compositions and in the Malku-type lexical works” (p. 357). But both binu and bintu are already found in earlier texts for which no West Semitic influence can be argued: see N. Wasserman, Akkadian Love Literature (Wiesbaden: Harrassowitz, 2016), 210 and 227 on KAR 158 rev. v 11′. Like bukru, binu belongs to the common literary language. The same cannot be said of bunu or bunatu: it may have been judicious to separate the entries more cleanly along the same lines.
The separate entries for būṣu “a bird” (P0744, P-S-ↄ.b-ṣ-ʾ.1) and būṣu “hyaena” (P0745, P-S-ↄ.b-ṣ-ʾ.2) highlight the authors’ erudition as a whole, connected with Modern South Arabian/Mehri bč̣aʿyōn “Tristram's grackle(?)” and Proto-West Semitic *ṣ̂abuʿ “hyena”, respectively, the latter by way of hypothetical *ṣ̂abuʿ- > *buʿaṣ- (or *buṣaʿ) > *buʿṣ- (or *buṣʿ-) > būṣ-. Such a metathesis is certainly possible – the doublet Akkadian laḫru and Hebrew Rāḥēl/Rachel “ewe” comes to mind. The suspicion remains, however, that the word is simply pūṣu and refers to both animals due to their white markings, as suggested by M. Stol in Bibliotheca Orientalis 77, 2020, 106.
Some entries explicitly reject the results of its sister Supplement. Based on the Neo-Assyrian writing BUR-DIŠ (Studien zu den Assur-Texten 2, 10: 4), M.P. Streck, Supplement 1, 48 had pointed to the need for revising the entry **burḫiš to burṭiš. The EDA's rejection of the reading based on a “rather uncertain” (p. 242) identification of the Neo-Assyrian attestation is needlessly contrarian: the context of the administrative text favours reading a designation of a natural animal, the burṭiš is also attested in other Middle Assyrian administrative texts: D. Kertai, “The creatures that protected the doors of Nineveh”, Mesopotamia 50, 2015, 151.
What picture of the lexicon emerges? D.O. Edzard (Zeitschrift für Assyriologie 60, 1970, 159) and M.P. Streck (Zeitschrift für Assyriologie 97, 2007, 149–52) had provided earlier statistics for CAD B and P, respectively. A simplified summary of the 1,001 headwords of the EDA yields 249 entries (or c. 25%) with no certain etymology, 426 entries either common Semitic (320) or internal developments (106) (together c. 42%), and 326 foreign or loanwords (c. 33%). The most common sources are West Semitic (114, or c. 11%), Hurrian (86, or c. 8.5%), and Sumerian (81, or c. 8%). Fewer derive from Indo-European (22), Elamite (10), Kassite (6), or Egyptian (8). Around 58 entries are hitherto attested only in Eblaite.
The first volume of any major dictionary usually serves as a test case to flush out any unforeseen gremlins in the system. As the volume demonstrates, these are mercifully few. Any disagreements on interpretation are compensated by the straightforward presentation of the evidence and ample indices. Nine additional volumes are planned. We can only wish the authors continued success and timely publication.
