Additions and Corrections to Wolf Leslau's Comparative Dictionary of Ge'ez (1987-2017).

• Author: Kogan, Leonid
• Position: Correction notice

In 2017 the scholarly world celebrated the thirtieth anniversary of Comparative Dictionary of Ge'ez (CDG) by Wolf Leslau. A major achievement of its time, CDG remains by far the most quoted tool of Semitic lexical comparison and, indeed, the only dictionary of a classical Semitic language which explicitly defines itself as "comparative"--in this context, practically tantamount to "etymological."

Due to the impressive development of several branches of Semitic lexicography in the past decades, upgrading Leslau's magnum opus inevitably suggests itself. The following pages of additions and corrections to CDG derive from many years of intensive use in my own scholarly work as well as in the classroom. Most of the additional material pertains to the following areas of Semitic linguistics and philology:

Assyriology. When CDG was in preparation, only a limited number of volumes of the Chicago Assyrian Dictionary were available, and even von Soden's Akkadisches Handworterbuch had only recently been completed (1981). Nowadays, when both dictionaries are fully accessible to Semitists, an important number of new and/or corrected lexical items from such a major Semitic idiom as Akkadian have become available for comparison with the relevant Ethiopian data.

Ugaritic studies. In Leslau's time, there was nothing remotely similar to Dictionary of the Ugaritic Language by G. del Olmo Lete and J. Sanmartin. Ugaritic lexical evidence--more than precious at times--had to be painstakingly gleaned from a huge number of text editions, translations, and specialized articles, from different periods and of varying quality. It is not surprising, therefore, that outdated or clearly mistaken interpretations of Ugaritic lexical elements are not rare in the pages of CDG. A systematic perusal of DUL carried out in the framework of the present study aims at improving this situation.

Modern South Arabian linguistics. By and large, the works of the Austrian South Arabian exedition must have been the main source of Leslau's lexical data for Mehri, Jibbali, and Soqotri. T. M. Johnstone's Jibbali Lexicon was published in 1981, when the bulk of the database behind CDG must have already been completed, (1) whereas the Mehri Lexicon by the same author, published posthumously in 1987, could not be taken into account at all. For Soqotri, Leslau's own masterpiece of 1938 was the only source to be consulted. Accordingly, additions and corrections pertaining to MSA on the pages of the present study are both frequent and substantial. A great deal of new and corrected evidence pertaining to Soqotri goes back to my own fieldwork research on this language, particularly on its exceedingly rich lexical treasures, from 2010 up to the present day.

It is my modest hope that this article will enable future readers to make an even more fruitful use of CDG as a classical tool of Semitic lexicography. (2) Last but not least, it is also a small token of admiration toward the memory of Wolf Leslau as a towering figure of the new age of comparative Semitics.

P. 2 (?ab 'father'). Add Soq. i?if- (LS 68), where -f- still awaits an explanation.

(?abd 'fool'). Note that in the fourteenth-century Arabic-Ethiopian glossary, where the etymological gutturals are usually preserved correctly, the relevant word is spelled with ? (Bulakh and Kogan 2016: 275), which, coupled with Tgr. ?abda 'to become crazy' (WTS 471) and Tna. ?obud 'crazy, mad' (TED 1879), makes less evident the traditional derivation of the Ethiopian root from *?bd.

P. 5 (?dber 'old woman'). Soq. ?abre means 'generation, people of the same age' (CSOLII 401) rather than specifically 'old age'. (3) Thus, in view of the unexplained difference of laryngeal, Leslau's comparison becomes rather unlikely. For the semantic connection between 'old age' and 'strength', thought to be unlikely by Leslau, v. Buck 1949: 276, Kogan 2015: 215-16 (under Hbr. ?abbir 'strong').

P. 10 (?afar 'soil'). It is hardly warranted to list Hbr. ?apar 'soil' (HALOT 861) and ?epar 'ashes' (HALOT 80) side by side as probable cognates to the Geez word. Indeed, the spelling with ? in the fourteenth-century Arabic-Ethiopic glossary (Bulakh and Kogan 2016: 152-53), as well as cognates with ? in Tigre and Argobba of Tollaha (ibid.), make it very likely that the Geez word (poorly attested, v. LLA 808) is an Amharism, with a non-etymological first guttural.

P. 12 (?ah exclamation of pain or grief). Note Soq. a?h with the same function (CSOL I 462; cf. LS 499).

P. 17 (?akaya 'to be bad, wicked'). Note Akk. aku A 'destitute, weak, powerless' and/or aku B 'crippled, deformed' (CAD A/I 283-84).

(?al element of negation). Akk. ul is unlikely to be related to PWS *?al, but rather goes back to ula

(?alle la- 'woe to'). Akk. allu with comparable meanings is not recorded in the standard dictionaries and is unlikely to exist.

(?ellu 'these'). Akk. ullu means 'that' (singular) and can hardly be directly equated with the WS forms meaning 'these' (plural); see further Kogan 2015: 68.

P. 18 (?allada 'to gather'). Syr. ?ulda 'acervi straminis, horrea' (LSyr. 21) can hardly be separated from from Akk. aldu 'store of barley' (CAD A/I 337, AHw 35), (4) itself borrowed from Sum. al-du (Lieberman 1977: 146). Its etymological relationship with the Geez verb is thus quite unlikely.

P. 21 (?ama 'when'). Soq. ?am with the same meaning (LS 312) is very unlikely to go back to PS *yawm- 'day' because of the intial ? and in view of the fact that a regular reflex of this term is attested in Soqotri as som (LS 419). Leslau's derivation of the Geez word from *yawm- is not very appealing either, particularly since *yawm- is the regular forerunner of yom 'today' (CDG 627).

P. 21 (?amd?ut 'intestine'). Add Soq. mi?ho with the same meaning (LS 248).

P. 26 (?amat 'cubit'). Soq. ?emeh, missing from LS, is unlikely to exist.

P. 27 (?enb- 'is not'). Ugr. yanu does not exist, the correct form being ?in [?ena] (DUL 74). As for Akk. yanu, it can scarcely be a direct cognate of *?ayn- because of its late attestation (v. Kogan 2015: 281).

P. 28 (?dnda?i 'perhaps'). A detailed etymological discussion of this and similar formations both within and outside Ethiopian can be found in Bulakh 2013: 7-9.

(?anf 'nose'). Probably related is Jib. ?enfi 'first, ancient' (JL 4, SED I No. 8).

P. 40 (?arwe 'animal'). For the semantic link between 'to live', 'animal', 'lion', and 'snake' in Semitic v. extensively Kogan 2006c: 294-95.

P. 42 (?askal 'cluster'). The correct Ugaritic cognate to this word is ?utkl (DUL 122).

P. 45 (?asar 'trace'). In view of the underlying *t in this root, Soq. ?eyhor 'to follow' (LS 54) is very unlikely to be related to it.

P. 56 (?adi 'still'). Arb. ?ad, pan-dialectal, but not codified by the classical lexicography, is worth mentioning here (Kogan 2015: 76-77).

P. 58 (?asm 'bone'). For the MSA cognates to this root and the underlying semantic shifts v. Kogan 2015: 537.

P. 59 (?aggala 'to revolt, to rebel'). (5) Note Soq. ?egol 'aller a la rencontre' (LS 297), with a rather feasible meaning shift.

P. 62 (?am 'year'). Add Soq. ?eno (LS 303), with an inexplicable alternation of sonorants (also in Jib. ?onut, JL 20).

P. 74 (?ask 'bough, branch'). Cf. Soq. ?eska 'branche, epi' (LS 330), with an unexpected s.

P. 79 (?ayg 'lake, pond'). Soq. ?ise 'lac d'ou les eaux ne s'ecoulent pas' (LS 307) has s

P. 82 (be?sa 'to be bad'). Ugr. b?s does not seem to exist; for the very uncertain bi?st v. DUL 202.

P. 83 (bd?sa 'to be strong, to grow mature'). Cf. probably Akk. bastu 'dignity, pride' (CAD B 142), 'Lebenskraft' (AHw 112), semantically and etymologically different from bustu 'embarrassment, distress' (CAD B 351), 'Scham, Scheu' (AHw 143). See Kogan 2003: 258.

P. 84 (be?la 'to be rich, wealthy'). Soq. ba?al 'master, lord' adduced by Leslau does not exist in this form. A comprehensive overview of the Soqotri reflexes of this root would include the following five positions (CSOLI 509, CSOL II 426-28): 1) ba?al 'to marry', (6) 2) meb?hel (fem. meb?elo) 'slave', 3) ba?- (pl. bi?hol'i-) 'owner of' (with pronominal suffixes only), 4) di-bahl'e (pl. il'-bahl'e) 'the owner of' (before nouns), 5) ba?l 'hi 'my lord' (referring to God). (7) Elsewhere in MSA, see Mhr. bal 'owner', ab[epsilon]li 'God' (ML 41), Jib. ba?al 'person owing', ?c?z 'God' (JL 22).

(mab?al 'iron tool'). Akk. belu 'etwa Waffe' (AHw 120) is now read as tillu (CAD B 199, T411).

P. 89 (behla 'to say'). Add Soq. bil'e 'thing', with a hypercorrect plural bissol (LS 83).

P. 91 (bahri, bahrey 'pearl'). Compare Arb. bahriyy- 'mother of pearl' (8) (Boson 1916-18: 417), Akk. bahru 'Koralle' (AHw 96, CAD B 29).

(?agzi?abaher 'God'). A curious structural parallel is found in Ebla: DINGIR KALAM-tim = BAD-lu ma-tim (VE 795b).

P. 95 (balha 'to be sharp; to be smart, clever'). For a tentative comparsion with Arb. bl[gamma] 'to reach the utmost point of something', bali[gamma]- 'eloquent' (Lane 250-52) v. Kogan 2005: 204.

P. 101 (bakalt 'date palm'). Add Soq. bekl'e 'sort (of date-palm)' (CSOL 1513, Naumkin et al. 2013: 69).

P. 105 (berk 'knee'). Add Soq. b[epsilon]rk (LS 96).

P. 108 (tabaraya 'to follow successively, to do by turns, to alternate with one another'). Compare Akk. bitru 'to be continuous' (CAD B 279, AHw 123), (9) with a plausible semantic shift.

P. 111 (bas?a 'to value, to evaluate'). Cf. perhaps Hbr. basa? 'gain' (HALOT 148), Arb. bida?at- 'merchandise' (Lane 215).

P. 113 (batul 'virgin'). Of considerable interest is Amh. battala 'to work...