Common Origin of Ethiopian Semitic: the Lexical Dimension*

Home , Amharic, Gurage languages, Languages of Ethiopia, Malik, Semitic languages, West Semitic languages

Leonid Kogan Russian State University for the Humanities Moscow COMMON ORIGIN OF ETHIOPIAN SEMITIC: THE LEXICAL DIMENSION*

Introduction Semitists have a tendency to be sceptical with regards to genetic classification and negative statements about «the questionable usefulness of classificatory schematizations» (Renfroe 1992:7) can easily be found on the pages of Semitological treatises. In spite of this trend, various aspects of genetic classification of Semitic have always occupied many of the best minds working in this branch of comparative linguistics. A proper evaluation of the subgrouping procedure as applied to Semitic languages has been admirably outlined in a recent survey by John Huehnergard: «Classification and subgrouping of language families are among the most important of the comparativist’s tasks, and this obtains in our field, too... Indeed, classification and subgouping should inform comparative work and historical reconstruction, for these activi- ties are inextricably interwined ... In other words, classification is not simply a mind game...» (Huehnergard 2002:130).1 A coherent classificatory pattern of Semitic has two facets: several major splits and unities are to be postulated and proved, such as East Semitic vs. West Semitic or Central Semitic vs. South Semitic; simultaneously, the common origin of each minor subdivision (Cannanite, Aramaic, ESA, Ethiopian,

* I am deeply grateful to Maria Bulakh and Alexander Militarev for their critical remarks on a preliminary draft of this article. The work on the present topic was carried out within the project «Studies in the Genetic Classification of Semitic» supported by the Center for Fundamental Research (project No. ÐÄ02-3.17-101) which deserves my most sincere gratitute for its assistance. My work on the present contribution was carried out in the framework of the projects 03-06-80435-a (ÐÔÔÈ) and 04-04-00324a (ÐÃÍÔ). I am grateful to both institutions for their help. 1 In reply to the following statement from ULLENDORFF 1961:30: «Classification is harmless, unobjectionable, and at times even useful if limited to describing present- day habitat and the prevailing geographical circumstances, but it becomes positively dangerous, i. e. obscuring rather than illuminating, if meant to explain genetic con- nections». [For the list of abbreviations used in this article see p. 392.] Ullendorff’s position is rightly qualified by Huehnergard as «without parallel in comparative work in other language families». For another highly positive assessment of linguistic classification v. GOLDENBERG 1998:461, quoting such outstanding figures of general and Indo-European comparative studies as Baudouin de Courtenay, Meillet and Greenberg.

Downloaded from Brill.com10/02/2021 09:12:28PM via free access 368 Scrinium I (2005). Varia Aethiopica

MSA etc.) is to be demonstrated as clearly as possible. Demonstration of the diachronic unity of Ethiopian Semitic (ES) is a model illustration of the second task: a compact, geographically and historically isolated group of languages whose similarity is intuitively realized by every student of Semitics should provide many specific features from which a reliable net of classificatory criteria could be worked out. According to a wide concensus, the principal method by which genetic classification of Semitic has to be guided is that of shared morphological innovations. This method, elaborated in a series of important contributions by R. Hetzron (e. g., 1972:13 and passim), was successfully applied by him and his successors to demonstrate the unity of West Semitic as opposed to East Semitic (Akkadian) and the unity of Central Semitic as opposed to other West Semitic languages. In spite of some difficulties connected with the formal shape of the corresponding classificatory features2 and the fact that con- flicting isoglosses have been adduced by some authors,3 the essence of Hetzron’s method proved to be valid and there is no doubt that it should be applied whenever possible — i. e., whenever important morphological innovations can be discovered for a given sub-branch of Semitic. Is this the case for Ethiopian Semitic? Regrettably, the answer seems to be overwhelmingly negative. As stated in Faber 1997:12, «although virtually all discussions of Semitic subgrouping assume a single Ethiopian Semitic branch which later split into North Ethiopic and South Ethiopic, there is virtually no linguistic evidence4 for such a Common Ethiopian stage». Indeed, even if Geez alone is confronted with Arabic, Sabaic or Mehri, reliable morphological innovations separating it from these languages are rather difficult to find, and the more so if such innovations are expected to be shared by, e. g., Tigre, Tigrinya and Amharic. Faber’s claim5 is perhaps exaggerated since

2 The origin of the second -a- in the Common WS New Perfect *katal-a³³ and that of -u in the Common CS New Imperfect *ya-ktul-u³³³ are still uncertain (HUEHNERGARD 2002:126 is well aware of this difficulty). 3 The principal one consists in the fact that Arabic shares several specific patterns of broken plural with Ethiopian and MSA (RATCLIFFE 1998:120). Cf. also ZABORSKI 1991, GOLDENBERG 1998:298ff. 4 Within Faber’s approach, «linguistic evidence» is clearly synonymous with «shared morphological innovations». This implicit equation — independent of its relevance for the subgrouping question — is terminologically inaccurate as it automatically excludes phonology, lexicon and syntax from linguistics. 5 No doubt, a fully justified reaction against many attempts to collect various allegedly specific Ethiopian features, all or most of which are actually absent from several languages of the group (notably, Geez), or turn out to be obvious retentions from PS, or can be qualified as typologically trivial developments possibly of areal nature. A typical example of such a list is LESLAU 1975 (as recognized already by Ullendorff in his comments on Leslau’s communication). HETZRON 1972:17–19 pro- vides a far more serious attempt but the final outcome is also somewhat disappoint-

Downloaded from Brill.com10/02/2021 09:12:28PM via free access L. Kogan 369 some morphological features which can be treated as Proto-Ethiopian can probably be detected. I mean first of all the gerund base *k³atîlÐ-, shared by Geez, Tigrinya, Amharic and Argobba (on the controversial evaluation of the Gurage evidence v. Hetzron 1972:101ff. as opposed to Goldenberg 1998:466ff.); the agent noun *k³atâl-îy- (partly replacing the PS active participle *k³³âtil-) and the infinitive in -ot can also be mentioned. However, none of these features appears to be of crucial value for determining the linguistic unity of ES (note that at least in one language — modern Harari — all of them are lacking) and, even if they do occur together (as in Geez), they hardly correspond to one’s expectations about a group of languages intuitively perceived as closely related.6 Does the lack of specific morphological innovations mean that the linguistic unity of Ethiopian Semitic cannot be demonstrated? Should we under- stand ES as a geographical and historical-cultural concept only? In my opinion, one important factor should prevent one from drawing such a conclusion, namely the fundamental unity of the basic vocabulary. A systematic demonstration of this unity and its key relevance for the question of the common origin of ES is the principal goal of the present article. Vocabulary is overtly or tacitly assumed to be persona non grata in most studies in the genetic classification of Semitic. According to Renfroe 1992:7, «that, under any circumstances, vocabulary is an unreliable measure by which to determine the relationship between any two given languages» is a «widely recognized fact». My comparativist’s intuition (no doubt shared by many col- leagues from various fields of historical linguistics) suggests that this — certainly unreflected — attempt to deny the classificatory value of the lexicon cannot be correct. The vocabulary can and should be taken into consideration ing. Thus, verbal expressions based on the verb ‘to see’ are scarcely attested in Geez but so common in Cushitic that their wide use in Modern Ethiopian may well be regarded as an areal phenomenon affecting each language independently. As for some syntactic phenomena discussed by Hetzron, the concluding statement of the corresponding section of his book is quite telling : «...the Cushitic evidence came later and was independent in the different branches of Ethiopian». In sum, it is very hard to agree with APPLEYARD 1996:207–208 who believes that Hetzron’s study «lays to rest the phantom of a dual or even multiple origin of Ethiopian Semitic». Personally, I am convinced that Hetzron’s attempt to defend the common origin of ES is among the weakest points of his otherwise brilliant monograph. 6 Prof. R.-M. Voigt kindly reminded me in personal communication about one Proto-Ethiopian innovation overlooked by Faber and myself, namely the front vowel after the first radical in the imperfect of B (Gez. y#-sebb# r, Tna. y# -s#bb#r, Har. yi-sîbri). I can only object to his observation that even this (no doubt, very important) innovative feature is absent from Tigre where the imperfect of B is identical to that of A (l# -sabb#r). Incidentally, Tigre lacks gerund and preserves *k³âtil- as the main pattern of the active participle (though k³âtlây and kat³ âl(i) are also in evidence).

Downloaded from Brill.com10/02/2021 09:12:28PM via free access 370 Scrinium I (2005). Varia Aethiopica in any subgouping procedure, especially if other criteria are scarce or vague (as in the case of ES). The main problem consists in elaborating a sound methodology for a proper assessment of its role. As far as the purpose of the present contribution is concerned, a simple and reliable method is at hand, in its essence almost symmetrically parallel to that used by Hetzron and his followers. Not only morphological, but also lexical units can be classified into retentions and innovations. While both morphological and lexical retentions are, in most cases, of no value for establishing the common origin for a group of languages, lexical innovations can be as telling in this respect as the morphological ones. The method of lexical innovations is not identical to the glottochronological method.7 The latter operates, in its statistical calculations, with both retentions and innovations indiscriminately, whereas the former is primarily interested only in innovations (though less trivial types of retentions may also be important, v. extensively below). The principal tasks of the two methods are also different: proving the common origin of ES need not be connected with establishing the time of its separation from Common West Semitic and its subsequent disintegration. However, since both are focused on lexical change, there may be some hope that the analysis proposed in the present article could provide arguments for a further refinement of the glottochronological procedure. R. Hetzron, a pioneer of the present-day Semitic and Ethiopian classification, was certainly aware of the important role of the lexicon in genetic classification: «Naturally, we do not mean that vocabulary must be discarded from among the criteria for linguistic classification. On the contrary, it is one of its most important bases» (Hetzron 1972:13). In spite of this, lexical evidence has been never used systematically in Hetzron’s studies on the topic.8

7 Which alone is critically mentioned (regrettably, without much discussion) by those who regard shared morphological innovations as the only tool of genetic subgrouping. Thus, APPLEYARD 1996:204 quickly moves from a critical evaluation of D. Cohen’s lexicostatistical study (COHEN 1970) to a general conclusion according to which «it is comparison of morphological forms and structures that must necessarily constitute the bedrock of any comparative work». Oddly enough, on p. 220 of his article the same author does not hesitate to suppose that the presence of the internal passive in MSA and Sabaic — in my view, one of the most important features linking these languages to WS as opposed to Akkadian — may be due to «diffusion, if not direct ‘borrowing’ from Arabic». If such an isogloss can be discarded as the result of borrowing and diffusion, I can only wonder what kind of morphological isoglosses still constitute «the bedrock of any comparative work». 8 Its most explicit manifestation is the term «Gunnän-Gurage» invented by Hetz- ron to denote West Gurage languages where terms for ‘head’ derive from *gunnän (as opposed to dum < *dVmâS- in East Gurage). HETZRON (1972:1) considers this to be «a very good lexical isogloss». Other scattered examples include the negative par-

Downloaded from Brill.com10/02/2021 09:12:28PM via free access L. Kogan 371

The special relevance of shared lexical innovations has also been recognized previously (see, e. g., Hackett 1980:122, relying on a personal communication by T. O. Lambdin; Tropper 1993:278ff. and especially Huehnergard 1998:275–276, with many penetrating remarks but a somewhat skeptical general evaluation).9 Nevertheless, to the best of my knowledge, this method has not yet been consistently applied to any sub-branch of Semitic.10 The present contribution, gratefully dedicated to the memory of Sevir Chernetsov — who some 13 years ago introduced those who are now the co-editors of the present volume into the field of Ethiopian studies — is intended to fill this gap.11

1. Swadesh wordlist: the evidence The Swadesh wordlist used in the standard glottochronological procedure will serve as a convenient starting point for the present investigation as it mostly consists of semantically unambiguous notions that are primitive enough to prevent — at least ideally — the corresponding terms from being borrowed from one language to another.12 ticles *Vy vs. *al (p. 28), the numeral ‘nine’ (*tis - vs. *zaht³ an³ -, p. 29), the verbs *hlf¼ vs. *hwr³ for ‘to go’ (p. 59) etc. It is also noteworthy that the only feature from among «the first independent innovations that separated it [Ethiopian] from South Arabian» (HETZRON 1972:18) mentioned by Hetzron on p. 129, namely the fact that «the same morphemic exponent is used adnominally in the meaning of ‘like’ and adverbially as a purposive ‘in order that’», is actually a lexical feature. It is attributed by Hetzron to an early Agaw influence but an identical use is observable for Akk. kîma at least in Old Assyrian: kaspam mehr¼ âtim irraminîkunu kîma tagammilâninni idâma ‘das Silber, die Gegenwerte, legt, um mir gefällig zu sein, von eurem ei- genen’ (HECKER 1968:255). That both Akk. kîma and Gez. kama are widely used to introduce dass-Sätze is commonplace. 9 For a somewhat contradictory presentation of the relevance of lexical isoglosses in the genetic subgrouping of Ethiopian v. APPLEYARD 1977:4–5. 10 A classical study dealing with shared lexical innovations in Indo-European is PORZIG 1954. 11 The method of shared lexical isoglosses has been applied by the present author to the problem of the genealogical setting of Ugaritic (to appear soon as a special study). Another contribution dealing with the lexical evidence for the continuity between Old and Middle Aramaic is KOGAN 2005. 12 In most cases, I will rely on the lists compiled for Geez, Amharic, Tigrinya, Soddo and Harari by Alexander Militarev (partly published in MILITAREV 2000 and MILITAREV 2004), a fruit of many years of thorough work with lexicographic tools and native speakers. My independent check of various positions of these lists has provi- ded abundant proof for the semantic accuracy of Militarev’s choice (in sharp contrast with RABIN 1975 or BENDER 1968, both full of quite arbitrary decisions), minor changes introduced by myself being thus quite insignificant. Needless to say, my etymological evaluation of many concrete cases is quite different from that proposed in Mili-

Downloaded from Brill.com10/02/2021 09:12:28PM via free access 372 Scrinium I (2005). Varia Aethiopica

My first step was to separate the stock of positions which can be more or less safely regarded as reflecting the hypothetic Proto-Ethiopian stage. In my opinion, a root reflected with the same basic meaning in Geez, Tigrinya or/ and Tigre, Amharic and/or Argobba, Harari and/or Eastern Gurage and one of the «Gunnän-Gurage» languages (for practical reasons, most often Soddo) can be attributed to Proto-Ethiopian without serious doubts. Such positions are represented by 41 examples. To these one can add, with a high degree of certainty, 16 roots which are present in Geez, Tigrinya or/and Tigre and at least two South Ethiopian subdivisions. A relatively small residual (11 examples) comprises various less reliable combinations (Gez., Tgr./Tna. and one S.-Eth. subdivision; all or most of Neo-Ethiopian but not Geez, etc.). In sum, no less than 68 positions can be qualified as Proto-Ethiopian, more than 80 % of them in a highly reliable way.13 By their diachronic nature, these Proto-Ethiopian roots can be subdivided into the following sections. 1. ´Trivialª retentions This group is constituted by terms which are exact phonological and semantic descendants of their Proto-Semitic ancestors. The latter, in their turn, are very likely to be the only (or at least the principal, or basic) terms for the respective notions in the proto-language.14 It includes the following positions: tarev’s studies so that all positive and negative solutions offered in the present article are my responsibility alone. For Tigre, a list compiled by Dr. M. Bulakh with the help of a native speaker was used. On the problem of borrowing in the Swadesh wordlists of ES languages v. Section 4 below. 13 This means that less reliable examples (11) can be easily eliminated from the discussion without prejudicing its basic conclusions. The same is true of the few cases which, in principle, could have been included in the Proto-Ethiopian stock but for various reasons were not (e. g., *tî³ s- or *tann- for ‘smoke’, *zV for ‘that’). 14 Such a conclusion is compelling if the respective term is reflected in all (or nearly all) the Semitic languages with the same basic meaning (positions 1, 11, 17, 21, 25, 32, 39, 41, 42, 44, 57, 80, 87, 89, 94, 95). It is almost certain in cases like 9, 37, 38, 40, 48 where the basic meaning is lost in one of the languages but the root in question is still present there and exhibits a transparent semantic (or even stylistic) shift (Akk. rçðu vs. k³akk³ ³adu head, idu vs. k³âtu hand and kabittu vs. amûtu liver are typical examples). As for the remaining cases, their attribution to trivial (rather than non-trivial) retentions certainly implies some degree of subjectivity (cf. the concluding remark of this section below). My decisions in such cases have been mostly guided by two criteria: the attestation, if not pan-Semitic, must be sufficiently wide and should affect geographically and historically unconnected areas; no alternative PS reconstruction with the same basic meaning should be available. Both criteria are well illustrated by PS *Vbn- which is reflected as the basic term for stone in Akka- dian, Hebrew, Ugaritic, Sabaic, Ethiopian and Soqotri, whereas no alternative basic synonym in PS seems to be known.

Downloaded from Brill.com10/02/2021 09:12:28PM via free access L. Kogan 373

1. ‘all’ — Gez. k«#ll-, Tna. k«#llu, Amh. hullu etc. (LGz. 281) < PS *kull-; 7. ‘to bite’ — Gez. nasaka, Tna. näkäsä, Amh. näkkäsä, Sod. näkkäsä, Har. näxäsa (LGz. 402) < PS *ntkÔ/nkt Ô(the metathetic variation, well attested for this root within and outside Ethiopian, is intriguing, but can hardly be regarded as an obstacle for postulating an eventual etymological identity of both variants); 9. ‘blood’ — Gez. dam, all Modern Ethiopian däm (LGz. 133) < PS *dam-; 10. ‘bone’ — Gez. as×m³ , Tna. asmi³ , Amh. at#³ nt, Sod. at#³ m, Har. ât³ (LGz. 58) < PS * at.mÔ-; 11. ‘claw, nail’ — Gez. s#³ fr, Tna. s³#fri, Amh. t#³ f#r, Sod. t#³ f#r, Har. tifir³ (LGz. 549) < PS *.tVprÔ -; 17. ‘to die’ — Gez. mota, Tna. motä etc. (LGz. 375) < PS *mwt; 19. ‘to drink’ — Gez. sätya, Tna. sätäyä, Arg. säèèa, Har. säèa (LGz. 518) < PS *šty; 21. ‘ear’ — Gez. #zn, Tna. #zni, Sod. #nz#n, Har. uzun (LGz. 52) < PS *udÔn-; 25. ‘eye’ — Gez. ayn, Tna. ayni, Amh. ayn, Sod. in, Har. în (LGz. 80) < PS * ayn-; 28. ‘fire’ — Gez. #sât, Tgr. #sat, Amh. #sat, Sod. äsat, Har. isât (LGz. 44) < PS *iš(-ât)-; 32. ‘full’ — Gez. m#lu, Amh. m#lu etc. (LGz. 342 < PS *ml); 37. ‘hand’ — Gez. #d, Tna. id, Amh. #¶¶, Sod.ä ¶, Har. i¶i (LGz. 7) < PS *yad-; 38. ‘head’ — Gez. r#s, Tna. r#si, Amh. ras, Har. urûs (LGz. 458) < PS *raš-; 39. ‘hear’ — Gez. sam a, Amh. sämma etc. (LGz. 501) < PS *šm ; 40. ‘heart’ — Gez. l#bb, Tna. l#bbi, Amh. l#bb, Sod. l#bb (LGz. 305) < PS *libb-; 41. ‘horn’ — Gez. k³arn, Tna. k³ärni, Amh. k³änd, Sod. k³är, Har. k³är (LGz. 442) < PS *k³arn-; 42. ‘I’ — Gez. anä, Tna. ane, Amh. #ne, Muh. anä, Har. ân (LGz. 26) < PS *anV; 44. ‘knee’ — Gez. b#rk, Tgr. b#r#k, Tna. b#rki, Sel. b#rk, Wol. b#rk (LGz. 105) < PS *bVrk-; 48. ‘liver’ — Gez. kabd, Tna. kabdi, Cha. xäpt, Har. kûd (LGur. 333) < PS *kabid-; 54. ‘moon’ — Gez. warh,¼ Tna. warhi³ , Har. wahri³ , Sel. wäri, Wol. wäri (LGz. 617) < PS *war(i)h¼-; 56. ‘mouth’ — Gez. af, Tna, af etc. (LGz. 8); as argued in SED I No. 223, the ES forms must be derived from PS *pay- ‘mouth’ (clearly the basic PS term with this meaning) but the a- extension is highly specific and almost certainly goes back to the Proto-Ethiopian period (to be explained by

Downloaded from Brill.com10/02/2021 09:12:28PM via free access 374 Scrinium I (2005). Varia Aethiopica the influence of similar Cushitic forms or/and contamination with PS *anp- ‘nose’);15 57. ‘name’ — Gez. s#m, Amh. s#m, Har. s#m etc. (LGz. 504) < PS *šim-; 59. ‘new’ — Gez. haddis³ , Tna. haddis³ , Amh. addis, Sod. a¶¶is, Har. ha³ ¶îs (LGz. 226) < PS *h³dtÔ; 60. ‘night’ — Gez. lelit, Tna. läyti, Amh. let, lelit, Gaf. litä (LGz. 314) < PS *lay(liy)-; 61. ‘nose’ — Gez. anf, Tgr. an#f, Tna. (Hamasen) anfi, Sod. afunna, Har. ûf (LGz. 28) < PS *anp-; 63. ‘one’ — Gez. ah³adu, Tna. hadä³ , Amh. and, Sod. att, Har. ahad³ (LGz. 12) < PS *ah³ad-; 68. ‘root’ — Gez. s×#rw, Tna. sur, Amh. s#r etc. (LGz. 535) < PS *Svr-; 73. ‘seed’ — Gez. zar, Tna. zari, Amh. zär, Sod. zär, Har. zär (LGz. 642) < PS *dÔar /-; 80. ‘star’ — Gez. kokab, Tna. kokob, Amh. kokäb, Sod. kokäb, Zwy. kokkäb (LGz. 280) < PS *kabkab-; 81. ‘stone’ — Gez. #bn, Tna. #mni, Sod. #mmayyä, Har. ûn (LGz.4) < PS *Vbn-; 87. ‘thou’ — Gez. anta, Tgr. #nta, Amh. antä, Gaf. ant, Sel. atä (LGur. 102) < PS *anta; 89. ‘tooth’ — Gez. s#nn, Tna. s#nni, Sod. s#nn, Har. s#n (LGz. 504) < PS *šinn-; 90. ‘tree’ — Gez. #s×,³ Tgr. #è³èät³ , Cha. äèä³ , Wol. #nèe³ (LGur. 12) < PS * is×-;³ mây, mîy 94. ‘water’_ — Gez. Tna. may, Har. , Sel. mäy, Wol. mäy (LGz. 376) < PS *may-õ; 95. ‘we’ — Gez. n#hna³ , Tna. n#hna³ , Amh. #ñña, Sod. #ñña, Har. #ñña (LGz. 395) < PS *nVhnV³ -; 96. ‘what’ — Gez. m#nt, Amh. m#n, Sod. m#n, Har. min (LGz. 352) < PS *mîn-; 97. ‘who’ — Gez. mannu, Tna. män, Amh. man, Cha. m«an, Har. mân (LGz. 348) < PS *man(n)-. It may be argued that in some of the above cases (Nos. 19, 28, 54, 59, 61, 63) the status of the respective PS root as the basic exponent of the notion in question is somewhat less certain than in the remaining ones. But even for these examples such a possibility is, in my opinion, so feasible that the preservation of the respective roots in Ethiopian can be regarded as highly non- specific.

15 In this sense, the present term can be treated not only as a trivial retention from PS but also as a rather specific formal innovation.

Downloaded from Brill.com10/02/2021 09:12:28PM via free access L. Kogan 375

2. ´Non-trivialª retentions For some notions included into the Swadesh wordlist the exponents are so diverse throughout Semitic that it has been impossible to reconstruct a single basic term for PS. Such cases may be explained in several alternative ways: either no basic term with this meaning altogether existed, so that all the corresponding words in the daughter languages are equally innovative; or one of the terms is the direct descendant of the basic PS term, which did not survive anywhere else with this meaning; or several synonyms without substantial semantic difference already coexisted in PS and have been subsequently ge- neralized in particular languages. In other words, the terms in question do have reliable cognates in other languages, but it cannot be demonstrated that the basic meaning like ‘bird’ or ‘black’ is a direct retention from PS and not an independent semantic evolution. Examples of this type include: 6. ‘bird’ — Gez. of, Tna. uf, Amh. wof, Sod. wof, Har. ûf (LGz. 78). # From PS * awp- ‘bird’, in its turn possibly going back to the verbal root wp ‘to fly’ (though a reverse development cannot be excluded). Ethiopian is the only branch of Semitic where * awp- became the basic term for bird, though this general meaning is also widely attested for Hbr. ôp (KB 801, mostly used as a collective; also applied to other flying beings), Jud. ôpâ (Ja. 1055) and Syr. awpâ (Brock. 517). 8. ‘black’ — Gez. sallim³ , Tna. sällim³ , Har. täy³ , Wol, tem³ , Sel. tçm³ (LGz. 556). # From *.tlmÔ , mostly attested with the meaning ‘(to be) dark’ rather than ‘(to be) black’ (v. Bulakh 2003:5–7). The latter is, however, also typical of Akkadian salmu³ (CAD S ³ 77). It is uncertain whether the meaning ‘to be black’ should be considered an independent innovation of Akk. and ES or a parallel retention of the original meaning (as Bulakh is apparently inclined to think). 77. ‘small’ — Gez. n#us, Tgr. n#uš, Tna. nuus, Amh. t#nn#š, Arg. mans, Gaf. #nsä, End. #ns (LGz. 382). # Probably a semantic evolution of PS *nš ‘to be weak’ represented by Akk. ençðu ‘to be weak, impoverished, shaky’ (CAD E 166), Hbr. nš ‘to be sickly’ (KB 73). The meaning ‘to be small’ is also present in Soq. énes (LS 68, without adjectival formations). An eventual etymological relationship between the two metathetic variants (N.-Eth. *ns and S.-Eth. *ns) can hardly be put to doubt. 16. ‘to come’ — Gez. mas³a, Tna. mäs³ e, Amh. mätt³ a³ , Sod. mätt³ a³ , Sel. mäta³ , Wol. mätä³ (LGz. 370). # From the PS verb of motion *mt.Ô (possibly ‘to reach, to arrive’): Akk. masû³ ‘to be equal to, to amount to’, D ‘to make reach to’ (CAD M1 344), Ugr. mt.Ô‘to meet, run into’ (DUL 608), Hbr. ms³ ‘to reach, meet, find’ (KB 619), Bib. mt³ ‘to reach, attain to; come over’ (ibid. 1914, with references to other Arm.), Sab. mt.Ô‘to reach, arrive, come to’ (Biella 273), ‘to go, proceed, march; to reach a place’ (SD 89), Qat. mt.Ô‘to enter, go through’ (Ricks

Downloaded from Brill.com10/02/2021 09:12:28PM via free access 376 Scrinium I (2005). Varia Aethiopica

96), Min. mt.Ô‘se trouver à un endroit’ (LM 64), Jib. míd. Ôí ‘to reach (to)’ (JJ 169), Soq. mty³ ‘venir, arriver, atteindre’ (LS 241). 26. ‘fat’ — Gez. s×#bh,³ Tna. s#bh³i, Amh. s#b, Cha. s#wä, Har. säbah (LGz. 525). # From *s×Vbh-³ , one of several Common Semitic terms for ‘fat, tallow’ (outside Ethiopian reliably attested in MSA, where it also became the main word for this notion, v. SED I No. 261). 31. ‘foot’ — Gez. #gr, Tna. #gri, Amh. #g#r, Sod. äg#r, Har. ingir (LGz. 11). # The origin of the ES terms for foot and their counterparts in some Arb. dialects (v. ibid.) is a matter of controversy: reflexes of an independent PS root (SED I No. 7) or an evolution of *rigl- ‘foot’ (so most recently Voigt 1998). 33. ‘to give’ — Gez. wahaba, Tna. habä, Arg. hawa, Sod. abä, Sel. wâbä, Wol. wabä (LGz. 609). # From *whb ‘to give’ (Arm., Arb., ESA, sparsely also Hbr.). Within the Common Aramaic suppletive paradigm of the verb ‘to give’ *whb became, side by side with *ntn, one of the two basic roots with this meaning. 43. ‘to kill’ — Gez. k³atala, Tna. k³ätälä, Wol. k³ätälä, Cha. k³ätärä³ (LGur. 508). # From *k³tl/k³tl³ ‘to kill’, mostly attested in Aramaic and Arabic where it also became the main verb with this meaning. 49. ‘louse’ — Gez. k³«#mâl, Tna. k³«#mal, Amh. k³#mal etc. (LGz. 432). # From *k³umâl- ‘kind of harmful insect, louse’: Old Arm. k³ml ‘louse’ (HJ 1013), Arb. k³aml- ‘poux’, k³ummal- ‘petites fourmis; petites sauterelles qui n’ont pas encore d’aile’ (BK 2 816), Sab. k³mlt ‘insect pests, locusts?’ (SD 105). For a metathetic variant *k³Vlm- v. Jud. k³alm#tâ ‘vermin’ (Ja. 1378), Syr. k³almâ ‘pediculus’ (Brock. 668), Sab. k³lm, k³lmt ‘insect pest, locusts?’ (SD 105), Qat. k³lm ‘Lausbefall’ (Sima 2000:131). Syr. k³almâ and Arb. k³ummal- also became the basic terms for louse in the respective languages. 65. ‘rain’ — Gez. z#nâm, Tna. z#nab, Amh. z#nab, Sod. z#nab, Har. z#nâb (LGz. 641). # From *dÔVnVn-/*dÔVnVm-, one of the common Semitic terms for rain (Akk. zunnu,Sab.dÔnm, v. ibid.). The last radical b in Neo-Ethiopian must be due to dissimilation. 69. ‘round’ — Gez. k#bub, kabib, Tna. käbbib, Amh. k#bb, Sod. k#bb, Wol. kub (LGz. 273). # From *kb (with various extensions), one of Common Semitic verbal roots for ‘to be curved, bent; to make a circular movement, to encircle’: Arb. kbb ‘pencher, incliner; pelotonner, rouler sur un peloton; faire des boulettes, des boules’ (BK 2 850), kbkb ‘renverser, culbuter’ (ibid. 855), Mhr. kbûb ‘to stoop’ (JM 201), Hrs. kbôb (JH 66), Jib. ekbéb id. (JJ 124), possibly also Hbr. pB. kabkâb (also kapkâp) ‘an arched round vessel’ (Ja. 608), Jud. kubbâ ‘vine cask; turret of a fort’ (Ja. 616), Akk. kabâbu ‘shield’ (CAD K 1). PS

Downloaded from Brill.com10/02/2021 09:12:28PM via free access L. Kogan 377

*kp with the same meaning represented by Akk. kapâpu ‘to bend, curve’ OB on (CAD K 175), kippatu ‘loop, tendril’ (ibid. 397), Hbr. kpp ‘to bend, bow down’ (KB 493), pB. kippâ ‘arch, doorway, bow; skull-cap’ (Ja. 635), Jud. kwp ‘to bend’ (ibid. 623), kiptâ ‘vault, arcade; bow, curve’ (ibid. 636), k#pap ‘to bend’ (ibid. 661), Syr. kâp ‘flexit’ (Brock. 323), kap ‘inclinavit, curbavit’ (ibid. 339), Mnd. kup, kpa, kpp ‘to bow, bend, curve’ (DM 208), Arb. kff X ‘entourer qch., faire un cercle autour; se rouler en spirale (se dit d’un ser- pent)’, kiffat- ‘tout object rond’ (BK 2 910) is almost certainly related as a variant root. 71. ‘to say’ — Gez. b#hla, Tna. bälä, Amh. alä, Sod. balä, Har. bâya (LGz. 89). # From *bhl, a verb of speaking (possibly with an original connotation of praying, imploring): Akk. baâlu ‘to pray, to beseech’ SB (CAD B 2), Arb. bhl ‘maudire’, VIII ‘implorer, invoquer, supplier’ (BK 1 173), Mhr. b#hlît ‘word’ (JM 45), Hrs. behelçt id. (JH 16), Jib. bµhlét id. (JJ 24), Soq. bíleh ‘chose’ (LS 83). 72. ‘to see’ — Gez. r##ya, Tna. räayä, Har. ria, Zwy. #rî (LGz. 459). # From *ry ‘to see’ whose attestation outside Ethiopian is mostly restricted to Hbr., Arb. and ESA (v. ibid.). It is the basic word for ‘to see’ also in Hebrew and Arabic. 79. ‘to stand’ — Gez. k³oma, Tna. k³omä etc. (LGz. 456). # From *k³wm ‘to stand, stay’, attested in Hbr., Arm., Arb. and ESA (v. ibid). It became the main term for ‘to stand’ also in Aramaic and Arabic. 3. Certain or likely innovations This group consists of those typically Ethiopian roots which can be more or less reliably traced back to PS terms with a different meaning, i. e., Ethiopian semantic innovations. The following positions can be classified in this way: 3. ‘bark’ — Gez. l#hs³ ,³ Tna. l#hs³ i³ , Amh. l#t,³ Muh. l#tä³ , Wol. l#è³è³aèe³ (LGz. 312, LGur. 383). # As suggested in LGz. 312, probably derived from PS *lhs¼ /³ hls¼ ³ ‘to draw off’: Arb. lhs¼ ³ II ‘épurer en séparant les parties moins propres; enlever, tirer, extraire la partie la plus pure et la meilleure’ (BK 2 980), Hbr. hls³ ³ ‘to draw off’ (KB 321), pB. Nip. ‘to be peeled off’, Syr. hls³ ³ pa. ‘rapuit, spoliavit’ (Brock. 237), see further KB 321–322. For a similar semantic development v. Latin cortex, according to WH I 279 < *(s)qer-t- ‘schneiden’ «als ‘abge- schälte, abgeschnittene Rinde’». 4. ‘belly’ — Gez. kabd, Tgr. käb#d, Tna. käbdi, Amh. hod (LGz. 273). # Generalization of PS *kabid- ‘liver’ (in Gez. and Tna. the meanings ‘belly’ and ‘liver’ are not distinguished at all whereas in Amh. *kabid- with the meaning ‘liver’ is replaced by gubbät). While the meaning shift ‘liver’ > ‘interior’ is also present in Arabic (‘cavité du ventre’, BK 2 852) and Ugarit- ic (‘innards, entrails’, DUL 424), the development of *kabid- into the main

Downloaded from Brill.com10/02/2021 09:12:28PM via free access 378 Scrinium I (2005). Varia Aethiopica term for external belly (abdomen) is specifically Ethiopian (the only — puz- zling — parallel is the expression ka-ba-tum-ma ‘on the belly’ in the Amarna Canaanite, cf. SED I No. 141). 22. ‘earth’ — Gez. m#dr, Tgr. m#d#r, Tna. m#dri, Amh. m#d#r, Sod. m#d#r (LGz. 330). # Generalization of a PS term for a type of soil, ground (probably ‘clod of earth’): Hbr. pB. mädär ‘ordure (material used for vessels)’ (Ja. 735), Syr. medrâ ‘gleba (terrae), terra, lutum, pulvis’ (Brock. 375), Arb. madar- ‘boue sèche et tenace, sans sable’ (BK 2 1078), Mhr. mdêr ‘Lehmziegel’ (Jahn 210). Replacing PS *ars×³- ‘earth’ (no trace in ES). A similar semantic evolution of *mVd(V)r- is apparently observable in Sabaic (mdr ‘territory, ground’, SD 83 and Biella 267) but rs× ³ clearly remained the main term for earth through- ×³ out ESA (note especially mr s1myn w rsn ‘lord of heaven and earth’ in the inscriptions from the monotheistic period, SD 7 and Biella 27). 23. ‘to eat’ — Gez. bal a, Tna. bäl e, Amh. bälla etc. (LGz. 95). # A semantic development from PS *bl ‘to swallow’ (Hbr., Arm., Arb., MSA, v. ibid. and KB 134). PS *kl is completely ousted as a verbal root throughout ES but a Proto-Ethiopian deverbal derivate *i/ukl- ‘corn, cereals’ is preserved in most languages (LGz. 15). 82. ‘sun’ — Gez. s×³ah³ây, Tna. säh³ ³ay, Amh. tay³ , EÅa èet³ (LGur. 190). # As argued in LGz. 149, probably derived from a verbal root *sh³ w³ /*s× h³ w³ ‘to shine, to be bright’ (v. ibid. 553 for an extensive list of cognates). An inherent connection with sun, sun-heat is possible as suggested by Hbr. h³ôm s³ah³ ‘glowing heat’ (Is 18.4), rûahõ ³ sah³ ³ ‘glowing wind’ (Je 4.11) and especially Akk. s³çtu ‘light, shining appearance of the sun, moon and stars, sultry weather, open air, open sun’ (CAD S³ 150), ‘Glut, heller Schein, Hitze’ (AHw. 1095). A similar derivation from this root is known from Arabic (d³uh³an ‘heure du jour où le soleil est déjà élevé sur l’horizon, matinée avancée; clarté, lu- cidité; soleil’, BK 2 12 and d³ihh³ ³- ‘soleil’, ibid. 10) but its evolution into the main term for sun (with a concomitant extinction of the main PS term for sun *SVmS-16 ) is spefically Ethiopian. 91. ‘two’ — Gez. k#le, Tna. k#l#tte, Amh. hulätt, Sod. kitt, Har. koot (LGz. 282). # A semantic development from PS *kil()- ‘both’: Akk. kilallân, Hbr. kilayim, Arb. kilâ, Mhr. k#lô (ibid.). The only remnant of PS *tin-âÔ ‘two’ in Ethiopian is Gez. sân#y ‘the next day’ (ibid. 509). 4. Proto-Ethiopian terms with uncertain status This section includes positions occupied by roots which are highly specific to Ethiopian but have no reliable etymology. It is, therefore, impossible to demonstrate whether they are inherited terms completely lost throughout the

16 Tgr. šäm#š is obviously an Arabism.

Downloaded from Brill.com10/02/2021 09:12:28PM via free access L. Kogan 379

Semitic languages (or, at least, with meager traces outside Ethiopian), or borrowed from an unknown source, or innovated through some unusual phonological and/or semantic shifts. The following examples should be attrributed to this category: 2. ‘ashes’ — Gez. hamad³ , Tgr. hamäd³ , Amh. amäd, Sod. amäd, Har. hamäd³ (LGz. 231).17 # None of the two alternative etymological approaches to this Proto-Ethi- opian root outlined by Leslau is fully convincing: — Arb. hmd³ ‘être intense (se dit de la chaleur)’, hamadat-³ ‘pétillement du feu qui brûle’ (BK 1 488); cf. also Hbr. pB. h³md ‘to produce shrivelling by heat’ (Ja. 475) and, possibly, Akk. h¼amadîru ‘shrivelled or withered’ (CAD H¼ 57); — Arb. h¼md ‘cesser de flamber (se dit du feu, quand la flamme s’éteint, quoqu’il y ait encore des tisons qui brûlent)’, h¼ammûd- ‘lieu où l’on couvre le feu, où on éteint les flammes, en conservant les tisons pour les rallumer’ (BK 1 630), Mhr. h#¼ mûd ‘to be extinguished, burnt out’ (JM 443). 46. ‘leaf’ — Gez. k³«asl³ , Tna. k³«äsli³ , Amh. k³#täl³ , Sod. k³#täl³ , Har. k³utt³ i³ (LGz. 450). # A number of terms with the root k³s³l and various botanical connotations are attested in Arabic: k³asl-³ ‘fleur de l’arbrisseau épineux salam-; rebut, parties que l’on jette en nettoyant le grain’, k³aslat-³ ‘tendre et flexible (arbre); gerbe (de céréales fauchées)’, k³as³îl- ‘fourrage vert coupé pour les chevaux’ (BK 2 755). One wonders whether these terms, together with the ES words for leaf, may go back to a verbal root *k³sl³ ‘to cut’ attested in Arb. (k³sl³ ‘couper; trancher, abattre’, XI prendre, saisir’, ibid.) and Soq. (k³ós³el ‘couper (les cheveux)’, LS 381). For possible examples of ‘leaf’ derived from ‘to peel, strip off’ v. Buck 525. 51. ‘man’ — Tna. säbay, Amh. säb, Sod. säb, Har. usu (LGz. 482). # Included in the present corpus of evidence because of its reliable attestation in Tna. and its wide spread throughout South Ethiopian. One cannot exclude that the collective meaning ‘people’ attested for Gez. sab (vs. b##si ‘man’) and Tgr. sab (vs. #nas) is an innovation of these languages but the reverse is also possible. The etymology of Proto-Ethiopian *sab- is unclear. It should probably be compared to Sab. s1b ‘carry out an undertaking (e. g., a military campaign)’, s1b t ‘expedition, undertaking, journey’ (SD 122), Qat. s1b ‘to set out, go’ (Ricks 157), supposing an original meaning ‘gang, mili- .18 tary or working commando’ A generalization of the ethnonym s1b ‘Saba,

17 The relationship between the present root and Tna. ham³ #kw#šti ‘ashes’ is uncertain (hamäd³ means ‘earth, soil, dirt’ in Tna.). 18 Cf. also Arb. subat- ‘long voyage’ (BK 1 1040), Mhr. h#bû (poor man) to go to people every day for food, hátbi ‘to struggle back home, make one’s way slowly to people to get help’ (JM 151); that Mhr. h is < *ð is suggested by Soq. miðteb ‘pauvre’ (LS 411; according to Leslau, < ðby).

Downloaded from Brill.com10/02/2021 09:12:28PM via free access 380 Scrinium I (2005). Varia Aethiopica

Sabaeans’ (cf. LLA 359 and Appleyard 1977:8) should, in my opinion, also be seriously considered. Comparison with Mhr. hâbû ‘people’ accepted in LGz. 482 is not convincing for various reasons (cf. the evidence collected in JM 2 which plainly suggests that hâ-/h³â- in this form belongs to the definite article rather than to the root whereas no expected š- is present in the cognate Jibbali form). 52. ‘many’ — Gez. b#zuh¼, Tna. b#zuh³, Amh. b#zu, Gog. b#Åä, Har. bä¶îh³ (LGz. 117). # Clearly related to Arb. bdÔh¼ ‘être grand, haut; s’élever à une grande hau- teur; e×tre fier, orgueilleux’ (BK 1 101) which is, however, hardly sufficient for postulating a reliable PS reconstruction. 66. ‘red’ — Gez. k³ayy#h,³ Tna. k³äyy#h³, Amh. k³äyy, Har. k³çh ³ (LGz. 456). # The origin of Proto-Ethiopian *k³yh³ ‘to be red’ is obscure. ESA k³yh³ ‘red’ adduced without reference by Leslau is Hapax Legomenon in the Hadramitic inscription RES 2693:2 where it is used as an attribute of dÔhb‘bronze’ (Sima 2000:320; cf. Pirenne 1990:75 for a different interpretation of dÔhb in this passage). The following possible cognates (none of them fully convincing) can be mentioned (some of them tentatively suggested by M. Bulakh in personal communication): — Arb. k³uhh³ ³- ‘pur, franc, sans mélange’ (BK 2 677), k³wh³ ‘balayer la maison’ (BK 2 831); — Sab. h-k³wh³ ‘complete satisfactory, finish off, level, plaster’, mk³h³ ‘suc- k³ ³ cess, triumph’ (SD 110), Qat. s1- h ‘to prepare, set in order’ (Ricks 143); — Arb. k³ayh-³ ‘pus (qui n’est pas mêlé de sang)’ (BK 2 844), k³wh³ ‘sup- purer (se dit d’une plaie)’ (ibid. 831); — Jud. k³#hah³ ³ ‘to be dull, faint’, k³#h³âh³â ‘faint-colored, gray?’ (Ja. 1345); — Akk. k³û ‘copper, bronze’ OB on (CAD Q 291). 93. ‘warm’ — Gez. m#ww#k³, Tna. mok³, Amh. muk³, Sod. muk³#nna, Har. muk³ (LGz. 375). # The origin of Proto-Ethiopian *mwk³ ‘to be warm, hot’ is unknown. 5. Certain or likely Cushitisms This group includes Ethiopian terms with no fully reliable Semitic etymology but widely attested in Cushitic. Some of them are obvious Cushitisms, the status of some other is disputed: 14. ‘cloud’ — Gez. dammanâ, Tna. dämmäna, Amh. dämmäna, Sod. dämmäna, Har. dân (LGz. 134–135). # Thought to be borrowed from Cushitic19 in Dolgopolsky 1973:51 and Appleyard 1977:36 (v. Militarev 2004:299–300 for a different opinion).

19 Which means, here and elsewhere below, from some particular Cushitic language(s) or from an undetermined stage of development of Common Central or East Cushitic.

Downloaded from Brill.com10/02/2021 09:12:28PM via free access L. Kogan 381

24. ‘egg’ — Gez. ank³ok³#ho³ , Tna. #nk³«ak³«#ho³ , Sod. ank³o, Har. ak³uh³ (LGz. 31). # Borrowed from Cushitic according to Dolgopolsky 1973:282, 319 but treated as cognate to the corresponding Cushitic forms in Militarev 2004:309 (cf. SED I No. 160). The situation is complicated by the fact that no basic PS term for egg can be safely reconstructed. 29. ‘fish’ — Gez. âs×â, Tna. asa, Amh. asa etc. (LGz. 73). # Clearly borrowed from Cushitic (Dolgopolsky 1973:293). 36. ‘hair’ — Tgr. è#³ gär, Tna. s³äg«ri, Amh. tägur³ , Sod. è#³ gär, Har. èigär³ (LGz. 550). # Included into the present corpus of evidence because of its wide attestation throughout Modern Ethiopian (the main Gez. term for hair seems to be s×# #rt though sag³ «r is also attested). Borrowed from Cushitic (Appleyard 1977:17). 53. ‘meat’ — Gez. s×#gâ, Tna. s#ga, Amh. s#ga (LGz. 526). # Clearly borrowed from Cushitic (Dolgopolsky 1973:99). This list is essentially in agreement with that proposed in Ehret 1988:649. Both include terms for cloud, fish, hair and egg. Ehret treats the term for meat as borrowed into Proto-North-Ethiopian (but observes: «also in Am- haric»). I have not found compelling reasons for treating *tîs-³ ‘smoke’, *brr ‘to fly’ and * wk³ ‘to know’ as the main Proto-Ethiopian terms for the respective notions though, obviously, such a possibility is not excluded for each of the three cases.

2. Swadesh wordlist: analysis and discussion As mentioned above, there are 68 probable Proto-Ethiopian roots and the majority can be reconstructed quite reliably. From the standpoint of the normal glottochronological procedure, this number is very high and unambiguously points to a rather close genetic relationship: compare, e. g., 56 positions in common between Harari and Geez as opposed to some 23 between Harari and Quranic Arabic or some 17 between Harari and Mehri. Within the present approach, however, bare numbers in themselves are not sufficient. The obviously heterogeneous mass of coincidences is to be carefully analysed in order to evaluate the classificatory relevance of each of its segments. Examples attributed to Group 1 («trivial retentions») represent a clear majority (37). In my opinion, these 37 positions do not prove the common origin of ES since the corresponding roots with the same basic meanings are well preserved in many other Semitic languages. A striking example is Heb- rew where all 37 roots are present in the respective positions, and Akkadian with its 33 examples is not too far behind. A more or less massive loss of these roots in a given Semitic language (cf. some 25 positions preserved in Arabic and some 20 in Mehri) may probably be interpreted as pointing to

Downloaded from Brill.com10/02/2021 09:12:28PM via free access 382 Scrinium I (2005). Varia Aethiopica their relatively prolonged separate status but certainly does not suggest that the remaining languages (such as Ethiopian, Hebrew and Akkadian) are closely related. In other words, if a text in a previously unknown Semitic language is discovered, the presence of these roots is likely to suggest that it is neither Arabic nor MSA, yet does not help to determine whether it is Ethiopian or Cannanite. In sum, trivial lexical retentions can hardly be helpful in the genetic subgrouping based on the lexical evidence, a symmetrical counterpart of the widely accepted opinion concerning the classificatory value of trivial morphological retentions: thus, the imperfect ya-k³tul-u is an important innovation proving the unity of Central Semitic but the preservation of *yV-k³attVl says nothing about the relationship between Akkadian, Ethiopian and MSA. 14 examples my be qualified as «non-trivial retentions» (Group 2). The classificatory value of each of them is conditioned by the degree of its spe- cificity, the evaluation of which is necessarily subjective. Roots specialized with the respective meaning in other West Semitic languages (*k³wm, *ry, *k³tl, *whb) are, in my opinion, the least significative in this respect (in some cases approaching those included in Group 1) whereas some others (*bhl or *ns) are so highly specific that practical difference between such terms and those included in the next section is sometimes elusive. The classificatory value of this segment as a whole is, therefore, not unimportant: a language exhibiting all 14 roots can be almost safely qualified as Ethiopian but even the combination of, say, * awp- for ‘bird’, *.tlmÔ for ‘to be black’, *whb for ‘to give’, *mt.Ô for ‘to come’ and *s×Vbh-³ for ‘fat’ is highly suggestive in this respect. Nevertheless, since all these terms are retentions (in each case inherited from a more or less restricted pool of options) rather than innovations, there is always the possibility (even if purely theoretical) for an identical or very similar set of options to be offered by another Se- mitic language. The six examples included in Group 3 («certain or likely innovations») are decisive for qualifying a Semitic language as Ethiopian and provide reliable evidence as to the common origin of ES. Admittedly, the relevance of particular cases is uneven. Thus, *kilay for ‘two’ is perhaps the most important one: highly specific from the semantic point of view (at least completely without precedent in Semitic), attested throughout Ethiopian and accompanied by an almost complete elimination of *tinÔ- (the only PS term for ‘two’).20 The least significant may be *kabd- ‘belly’ < ‘liver’, present only in North Ethiopian and Amharic and, possibly, with a precedent in early Canaanite. Nevertheless, even if one of the above roots is present in a Semitic text with the respective meaning, its Ethiopian background is very likely (and abso- lutely certain if there are two or three of such terms).

20 The high relevance of this innovation was fully realised by such a prominent authority of modern Semitics as J. CANTINEAU (1932:179).

Downloaded from Brill.com10/02/2021 09:12:28PM via free access L. Kogan 383

The diagnostic relevance of the six terms of Group 4 is high in spite of their uncertain status: as I have just shown, both groups to which they can be alternatively attributed (non-trivial retentions and likely innovations) are significant from the point of view of subrouping. As for the Cushitisms included in Group 5, their relevance depends on whether they are thought to be borrowed independently into various Ethiopian languages or into the hypothetic Proto-Ethiopian. No convincing solution to this interesting problem can at present be proposed (see further below).

3. Evidence from other lexical fields Since the method of lexical innovations is not bound to any fixed number of positions, one can operate freely with various basic notions not included in the Swadesh wordlist. Such an inquiry has yielded several additional examples of lexical innovations, probably going back to Proto-Ethiopian: * rk³ ‘to be naked’: Gez. ark³a, Tgr. ark³a, Tna. aräk³ä, Amh. arräk³ä (LGz. 71). It remains to be established whether the element #nè ³ attested with this meaning throughout West Gurage can be taken back to the present root with n < *r and è³ < *k³. # From PS * rk³ ‘to gnaw, strip away’: Arb. rk³ ‘dépouiller l’os de la chair’ (BK 2 228) and other cognates in LGz. 71; replacing PS * rw. *blh¼ ‘to be sharp’: Gez. balha¼ , Tgr. bälha³ , Tna. bälhe³ , Sod. bulä honä, Har. bäläha³ (LGz. 95). # As argued in Kogan forthcoming, probably from PS *blS ‘to be at the edge, to reach the point’ (Arb. blS ‘parvenir à un point; atteindre à..., arriver à...’, BK 1 161), cf. especially Arb. mablaS- ‘terme, point extrême où parvient une chose’ (BK 1 162) vs. Tgr. m#blah³ ‘point (of the horn)’ (LH 269). *batr- ‘branch, shoot, stick, rod’: Gez. batr, Tgr. bät#r, Tna. bätri, Amh. bätt#r, Sod. bätt#r, Har. bärti (LGz. 112). # Possibly derived from PS *btr ‘to cut off’ (cf. DRS 80): Hbr. btr ‘to cut in pieces’ (KB 167), Arb. btr ‘couper la queue à un animal’ (BK 1 82), Tgr. (t#)bättära ‘to become unarmed’, bätray ‘leafless (tree)’ (LH 287), Jib. ôt#r ‘to chop’ (JJ 30). For an exact semantic parallel v. Buck 523 in connection with Greek kládos. *grm ‘to be formidable, amazing, awesome’: Gez. garama, Tgr. gärrä- ma, Tna. gärämä, Amh. gärrämä, Sod. gärrämä, Sel. gärämä (LGz. 203). # Probably derived from PS *gVrm- ‘bone; body’ with a semantic evolution otherwise attested in Semitic (v. SED I No. 94). *hmm³ ‘to be sick, ill’: Gez. hamama³ , Tgr. hamma³ , Tna. hamämä³ , Amh. ammämä-w, t-ammämä (LGz. 233). # From PS *hmm³ ‘to be hot, feverish’ (v. LGz. 233 and SED I No. 120), replacing PS *mrs×.³

Downloaded from Brill.com10/02/2021 09:12:28PM via free access 384 Scrinium I (2005). Varia Aethiopica

*lk³h³ ‘lend’: Gez. lak³k³#ha³ , Tgr. aläk³ha³ , Tna. aläk³k³#he³ , Amh. täläk³k³a, Har. (a)lçk³äh³a (LGz. 317). # A causative formation from PS *lk³h³ ‘to take’. *nVfâs-‘wind’: Gez. nafâs, Tna. n#fas, näfas, Amh. n#fas, Sod. n#fas, Wol. n#fas (LGz. 389). # From PS *npš ‘to breathe’. *na a ‘come! up!’: Gez. na a, na â, n# â ‘come, come now’, Tgr. n# a, Tna. nä a, Amh. na, Sel. Zwy. na (LGur. 445). # This interjection (partly serving as a suppletive imperative for ‘to come’) goes back to a PS verb of movement *n with various extensions (v. KB 681 for references). The verbal origin of *na a is clear from its full conjugation in Gez. n# -û, n# -î, n# -â. *ma âr- ‘honey’: Gez. ma âr, Tgr. mä ar, Tna. mä ar, Amh. mar (LGz. 326). # From PS *wa r- ‘honeycomb’ (Hbr. ya ar, KB 423), a meaning still preserved in Gez. (note the meaning ‘wax’ for the reflexes of *ma âr- in Harari and Gurage, LGur. 386). Partly replacing PS *dibš- (preserved in Epi- graphic Geez, Harari, Gafat and Argobba, LGz. 122). See further Appleyard 1977:28–29. *amlâk- ‘god’: Gez. amlâk, Tgr. amlak, Tna. amlak, Amh. amlak (LGz. 344). # A broken plural from *malk- ‘king’ unattested in Ethiopian (Appleyard 1977:56), completely ousting PS *il-. *mar ât- ‘bride; daughter-in-law’ (and related forms connected with wed- ding, marriage): Gez. mar ât, Tna. mär at, Amh. m#rat, Sod. märat, Wol. märat (LGur. 424). # Derived from the verbal root *r w ‘to join’ present in Gez. ar ut, Arb. ur uwwat- ‘yoke’ (cf. Latin conjux, Russian ñóïðóã). Replacing PS *kallat-. *ngs×‘to rule, to be king’: Gez. nags×a, Tgr. nägsa, Tna. nägäsä, Amh. näggäsä, Sod. näggäsä, Har. nägäsa (LGz. 393). # From PS *ngs×‘to push, press, drive (to work)’, replacing PS *mlk, *mal(i)k-, supposed to be preserved in *amlâk- ‘god’ only (v. above). See further Appleyard 1977:51. *rkb ‘to find’: Gez. rakaba, Tgr. räkba, Tna. räkäbä, Amh. räkkäbä, Msq. räkkäbä, Sel. räkäbä (LGz. 469). # Almost certainly derived from PS *rkb ‘to ride; to join, connect’ even though exact details of the semantic evolution remain obscure (v. extensive discussion in LGz. 469). *rs ‘to forget’: Gez. ras a, Tgr. t#räss# a, Tna. räss# e, Amh. rässa, Sod. ärässa, Har. räsaa (LGz. 474). # Likely developed from PS *rš ‘to be wicked, ruthless’ via ‘to be negli- gent’ (v. especially Akk. ruššû ‘to behave thoughtlessly, imperiously, to act in contempt or disrespect of others’, CAD R 429). Almost completely ous-

Downloaded from Brill.com10/02/2021 09:12:28PM via free access L. Kogan 385 ting PS *nšy which is possibly preserved in Gez. nâhsaya³ (LGz. 395; the origin of -h-³ is unclear). *tbw³³ ‘to suck’: Gez. tabawa³³ , Tna. täbäwä³³ , Tgr. täba³³ , Amh. täbba³ , Sod. tobbä³ , Har. täba³³ (LGz. 587). # Derived from PS (and Proto-Ethiopian) *tVb³ - ‘teat, breast’ (SED I No. 247). *wald-, *lid-‘son’: Gez. wald, Tgr. wäd, Tna. wäddi, Amh. l#¶, Sod.wäld , Har. li¶i, waldi (LGz. 613). # PS *bin- is ousted by derivatives of *wld ‘to bear’ (preserved only in Gez. b#nta ayn ‘pupil of the eye’). The lack of distinction between ‘son’ and ‘boy’, ‘child’ observable throughout Ethiopian is atypical of Semitic and may be due to the substratum influence.21 While *wVld- for ‘son’ is attested in some Arabic dialects, its use there is restricted to filiations (DRS 546), *bin- clearly remaining the basic term for this notion. *zib- ‘hyena’: Gez. z#b, Tna. z#bi, Amh. #¶b, Arg. #¶b (LGz. 630). # A semantic evolution of PS *dÔbi - ‘wolf, jackal’, replacing PS *s×ab³ (u) - ‘hyena’ which is reflected in Gez. only (s×#³ b , LGz. 147; very scarcely attested). The notion ‘wolf’ is mostly expressed by Cushitic borrowings like Gez. tak« lâ. See further Appleyard 1977:44. Besides, a (by no means exhaustive) list of isoglosses with no reliable Semitic etymology, thus corresponding to group 4 in the above classification, can be adduced: * ilat-, *ma alt- ‘day’, *w l ‘to spend the day’: Gez. Tgr. mä al, m# #l, #lät, Tna. mä alti, #lät, Sod. may, Har. mäaltu, mâltu ‘day (in daytime)’, môy ‘day (24 hours)’ (LGz. 603). # Replacing PS *yawm- (preserved with the meaning ‘today’ in Gez., Tgr. and Tna., LGz. 627). *bk³« ‘to fit, to be sufficient’: Gez. bak³« a, Tgr. bäk³ a, Tna. bäk³« e, Amh. bäk³k³a, Sod. bäk³k³a, Har. bäk³a (LGz. 99). # Note Arb. bk³ ‘se contenter, être satisfait de qch.’ (BK 1 150). *ftn³ ‘to be fast’: Gez. fatana³ , Tgr. fätna³ , Tna. fätänä³ , Amh. fätt³ änä³ , Sod. fätt³ änä³ , Har. fätäna³ (LGz. 171). # Compared by Leslau to Arb. ftn³ ‘être intelligent, avoir de la sagacité’ (BK 2 613). MSA verbs with the meaning ‘to remember; to be clever’ (JM 108, JH 36, JJ 66) must be borrowed from Arabic. According to Biella 402, Sab. ftn³ ‘assign, ordain’ may also be related (supposing an original meaning ‘to proclaim’). *g«dl ‘to be missing’: Gez. g«adala, Tna. g«ädälä, Amh. g«äddälä, Sod. g«äddälä (LGz. 182).

21 V. such Cushitic examples as Saho âuÐk-â ‘kind, son, tochter, knabe, mädchen’ (RSa. 76) and Beja ôr ‘knabe, son’ (RBed.. 27).

Downloaded from Brill.com10/02/2021 09:12:28PM via free access 386 Scrinium I (2005). Varia Aethiopica

*gz ‘to dominate; to own, possess’: Gez. gaza, Tgr. gäza, Tna. gäze, Amh. gäzza, Sod. gäzza, Har. gäzaa (LGz. 210). *hwr³ ‘to go’: Gez. h³ora, Gaf. horä, End. wärää (juss. y#x«är), Har. h³âra (LGz. 249). # Compared by Leslau to Arb. rwh³ ‘aller’ (BK 1 945) and hwr³ ‘revenir, retourner’ (ibid. 509). That Sab. hwr³ ‘to settle in a town’ (SD 73) goes back to an original meaning ‘to go’ (as assumed by Leslau) is by no means certain. *hzl³ ‘to carry on the back’: Gez. hazala³ , Tna. hazälä³ , Amh. azzälä, Msq. ezzälä, Har. h³azäla (LGz. 253). # Note Arb. hud³ Ôal- ‘giron; ceinture dans le caleçon, cette partie par laquelle on le fixe sur le corps’ (BK 1 398). *h¼s×b³ ‘to wash’: Gez. has¼ ×aba³ , Tgr. has³ ba³ , Tna. has³ äbä³ , Amh. att³ äbä³ , Sod. att³ äbä³ , Har. ha³ täba³ (LGz. 259). # Compared by Leslau to Arb. hd¼ b³ ‘se teindre les mains, les ongles, la barbe ou les cheveaux’ (BK 1 584). Replacing PS *rhs³ × ³ which is preserved in Gaf. and East Gurage (in other languages with a meaning shift into ‘to sweat, perspire’ only). *k b ‘to do again, to double’: Gez. kâ #b, Tgr. kä abä, EÅa xabä, Wol. kabä, Ancient Har. ka ab (LGz. 271). # Note Arb. ka b- ‘jointure, articulation des os’ (BK 2 907). *k³«sl ‘to be wounded’: Gez. k³«asla, Tna. k³«äsälä, Amh. k³ossälä, Sod. k³«ässälä (LGz. 446). # Compared by Leslau to Arb. k³sl³ ‘couper’ (BK 2 755). *kyd ‘to tread, trample, walk’: Tgr. keda, Tna. kädä, Amh. hedä, EÅa hedä, Har. xe¶a (LGz. 301). # Compared by Leslau to Arb. kdd II ‘donner une chasse vigoureuse, re- pousser et poursuivre’, kadîd- ‘sol foulé par les bêtes à sabot’ (BK 2 872). Note Arb. kdkd ‘marcher lentement et lourdement’ (BK 2 875), Mhr. kdû ‘to make (a camel) trot’ (JM 203), Jib. kédé ‘(camel) to trot’ (JJ 126). *lmlm ‘to be green, verdant’: Gez. lamlama, Tgr. lämläma, Tna. läm- lämä, Amh. lämällämä, Sod. l#mällämä (LGz. 315). # Note Arb. lmm IV ‘arriver à son terme; avoir les dattes presque mûres’ (BK 2 1022) and Akk. lammu ‘almond tree; sapling’ (CAD L 68). *lsy³ ‘to shave’: Gez. lâsaya³ , Tna. lasäyä³ , Amh. laèè³ ä³ , Sod. läèè³ ä³ (LGz. 319). # Note Arb. lys³ ‘remuer et ôter une chose de sa place’ (BK 2 1048). *mhl³ ‘to swear’: Gez. mahala³ , Tgr. mähala³ , Tna. mäh³alä, Amh. malä, Sod. malä, Zwy. mâlä (LGz. 335). # Sab. mhly³ in Gl 1533.6 is tentatively interpreted as ‘oath’ in SD 84 and Biella 271. Note Hbr. pB. mâhal³ ‘to forgive, pardon’ (Ja. 761), Jud. id. (ibid.) compared by Leslau. It is unclear whether Arb. mhl³ III ‘agir avec astuce à l’égard de qn.’ (BK 2 1069) may also be somehow related. *makkân- ‘sterile, childless’: Gez. makkân, Tna. mäkan, Amh. mäkkan, mähan, Gog. mäkan, Zwy. mähan (LGz. 340).

Downloaded from Brill.com10/02/2021 09:12:28PM via free access L. Kogan 387

*mrh³ ‘to guide, lead’: Gez. marha³ , Tna. märh³e, Amh. märra, Sod. märra, Sel. mära (LGz. 358). # One wonders whether Syr. marrâh³â ‘audax’ (Brock. 404), Arb. mrh³ ‘marcher avec fierté, en levant la tête’ (BK 2 1087) may be related. *nbr ‘to sit; to live, stay’: Gez. nabara, Tgr. näbra, Tna. näbärä,Amh. näbbärä, norä, Sod. näbbärä, Har. näbära (LGz. 284). # At least partly replacing PS *wtbÔ‘to sit’ (usually thought to be preserved in Gez. awsaba ‘to take a wife’, LGz. 619). Etymology fully uncertain (note perhaps Arb. nabûr- ‘cul, derrière’, BK 2 1183). Nöldeke’s attempt to relate Gez. nabara with Arb. nbr ‘to be high, elevated’ (Nöldeke 1910:49) does not look attractive from the semantic point of view. *ndd ‘to burn’ (intr.): Gez. nadda, Tna. nädädä, Amh. näddädä, Sod. näddädä, Har. nädäda (LGz. 385). # Compared by Leslau to Arb. nd ‘faire un petit creux dans les cendres chaudes pour y mettre le pain etc., qu’on veut faire cuire’ (BK 2 1224). *ngd ‘to travel’: Gez. nagada, Tgr. nägda, Tna. nägädä, Amh. näggädä, Sod. näggädä, Har. nigdi âða (LGz. 391). # Compared to Syr. n#gad ‘traxit; se effudit’ (Brock. 413) in LLA 693 (with cognates in other Arm.: Ja. 871-2, DM 288). *s m ‘to kiss’: Tgr. sä ama, Tna. sä amä, Amh. samä, Sod. samä, Sel. sâmä (LGz. 481). # Compared by Leslau with Yemenite Arb. sa am ‘kiss’ (after W. W. Mül- ler; v. now Behnstedt 559) which, in view of its isolated position within Ara- bic, must be due to (Proto-)Ethiopian influence. *sbb³ ‘to be narrow’: Gez. sabba³ , Tgr. säbba³ , Tna. säbbä³ , Amh. täbbäbä³ , Sod. täbbäbä³ , Har. täbäba³ , Zwy. täbäbä³ (LGz. 545). # Compared by Leslau to Arb. d³bb ‘empoigner une chose, prendre avec toute la main’ (BK 2 2) and daff³ - ‘étroit’ (ibid. 30) and similar forms in -p in Hbr. and Arm. Likely replacing PS *s×y³ k³ (which may be preserved in Sel., Wol. tä³ k³äk³ä, Msq. tä³ k³k³äk³ä ‘to be narrow’, LGur. 628). *w y ‘to be hot, to burn’: Gez. w# ya, Tna. wä ayä, Enm. wiä, Har. w#y (LGz. 603). # Tentatively related by Leslau to Arb. w y ‘exciter la sédition, les tumul- tes (BK 2 1571)’. *wlt/³ lwt³ ‘to change’: Gez. wallata³ , Tna. läwwätä³ , Amh. läwwätä³³ , Sod. liwwätä³ (LGz. 614). # One wonders whether the Ethiopian roots can be related to PS *lwt³ ‘to twist, turn, do a second time’ (v. LS 230 for concrete forms). Akk. lîtu³ ‘hostage, pledge’ (CAD L 223) is rather close semantically to the meaning ‘to change’ but it is thought to be derived from lâtu³ ‘to confine, to keep in check’ (ibid. 113). To collect and analyse all non-trivial lexical retentions of ES is an important but difficult task clearly beyond the scope of the present article. Two examples of Proto-Afrosiatic roots widely attested in Ethiopian but with no

Downloaded from Brill.com10/02/2021 09:12:28PM via free access 388 Scrinium I (2005). Varia Aethiopica reliable parallel anywhere else in Semitic can show how instructive such an investigation could be: *maray-t- ‘earth, soil, dust’ (LGz. 361: Gez., Tgr., Tna., Amh.), clearly related to Berber *ta-mur-t- ‘earth’ (Militarev 2004:307); *hlw ‘to be’ (LGz. 218: passim), probably to be compared to Berber, Cus- hitic and Chadic terms with the same meaning (HSED No. 28). 4. Problem of borrowing One of the main reasons underlying Hetzron’s reluctance to deal with the vocabulary in his studies in Semitic classification was the possibility of lexical borrowing: «One must also be very careful with vocabulary in classification. ... Neighbour languages may use the same root and it is sometimes difficult to find out whether it is a recent borrowing from one of them or an old one going back to the proto-language. ... One must proceed with extreme prudence in dealing with the vocabulary of limitrophe languages» (Hetzron 1972:12–13). The first illustration adduced by Hetzron to justify his claim is rather unlucky: no one would group together Harari, Gurage and some Cush- itic languages because of the fact that they all have borrowed Arb. Sazw- ‘raid’ as gaz. His second example is of a different nature, however. Accor- ding to Hetzron, «the use of the Northern root *ngr for ‘to say’ in Amharic ... can be explained by the northern character of the Amhara civilization, and by the fact that (in historical times at least) Amharas had much more contact with the North than with other South Ethiopic speakers» (ibid.). Since ‘to say’ clearly belongs to basic rather than cultural notions, Hetzron’s argument can be easily expanded to a degree that all coincidences in the basic vocabulary of ES would be explained as interborrowings resulting from a series of «wave innovations» (Hetzron 1975:108), thus completely invalidating the results of the present investigation. In my opinion, there are several reasons to believe that Hetzron’s hypoth- esis, while not improbable in some particular cases (*ngr may well be one of them), is hardly suitable as a general explanation. First of all, in order to explain the pan-Ethiopian spread of such terms as *bl or *kilay one would have to postulate not just a single act of borrowing from language X into language Y but virtual chains of loanwords involving half a dozen languages at least. While quite conceivable for a term meaning ‘raid’, this spread would be a rarity for such highly conservative notions as ‘two’, ‘sun’ or ‘to eat’. Secondly, ES lexical innovations usually exhibit a deeply structured pattern common to all the languages affected by them. Thus, if PS *ars×- ‘earth’³ or *SVmS- ‘sun’ disappear from Ethiopian, they are eliminated completely from all the languages of the group (if one of them does not display the common Ethiopian innovation, in most cases it does not preserve the old root

Downloaded from Brill.com10/02/2021 09:12:28PM via free access L. Kogan 389 either but opts for another replacement, such as Har. îr for ‘sun’).22 But even if *kl ‘to eat’ does leaves a trace, it is — formally and semantically — the same trace in all Ethiopian Semitic languages, namely #kl ‘cereals, food’ (with fully regular correlates throughout ES). Not a shade of any other nom- inal or verbal formation from *kl is attested. Simultaneously, the original meaning of the newly established basic root undergoes a similar systematic shift: as soon as PS *bl becomes specialized with the neutral meaning ‘to eat’, its earlier meaning ‘to swallow’ disappears completely and is now expressed by a different root (*wh¼t).³ This picture is suggestive of a compact and well-shaped proto-language whose main lexical features were faithfully inherited by its daughter tongues rather than through a chain of borrowings that by their very nature could hardly be so systematic. Thirdly, cognate terms derived from Proto-Ethiopian roots undergo all regular phonological shifts typical of Modern Ethiopian languages: palatali- zation, loss of nasal sonorants, spirantization of velars and labials, loss of s,³ insertion of n, the shift *-rn- > -nd- etc. Since earlier stages of Modern Ethio- pian are poorly documented, the chronology of most of these processes cannot be established with certainty, but none of them is very recent (as one can learn from Podolsky 1991:22–23, 32, 35, 45, 48, 51, they are all more or less well represented in the earliest Amharic documents dating to the 14th–15th centuries). Accordingly, the hypothetic chains of borrowings, if they ever took place, must have occurred during a period whose cultural-historical circumstances (at least those which could justify a massive borrowing in the basic vocabulary) are completely out of our control. In sum, Hetzron’s caution is probably justified as far as some concrete examples are concerned but does not seem to be warranted for the bulk of lexical coincidences in the basic vocabulary of Ethiopian.23 This conclusion can be at least partly corroborated by an analysis of borrowings from ES into Central

22 This circumstance brings us to the interesting question of whether a system of shared losses in the lexicon is significant for genetic subgrouping (the essentially similar problem of morphological losses as classificatory criteria is well known to historical linguists). Even if the spread of *bl for ‘to eat’ is due to interborrowing, it is hard to suppose that its influence was strong enough to produce such a ubiquotous elimination of *kl. In my opinion, this elimination can only be explained as a Proto- Ethiopian fact. The same conclusion can probably be applied to more culture-bound notions: thus, the pan-Ethiopian spread of *amlâk- ‘god’ or *n#gûs×- ‘king’ is perhaps less impressive than a total absence of such virtually pan-Semitic terms as *il- and *malik-. 23 Even if some obvious examples of lexical borrowing and diffusion can be occasionally detected in Swadesh word lists of Ethiopian Semitic languages: note such telling cases as Amh. s³ähay ‘sun’ and Tna. hamlya³ ‘green’, clearly borrowed from Geez and Amharic respecively in view of their phonetic shape (suggested by M. Bulakh).

Downloaded from Brill.com10/02/2021 09:12:28PM via free access 390 Scrinium I (2005). Varia Aethiopica and East Cushitic and vice versa. As is well known, many years of mutual interaction between ES and Cushitic have produced a virtual linguistic sym- biosis whose impact on the lexicon (cultural above all, but basic as well) is hard to overestimate. In many lexical fields interborrowings between ES languages seem to be as frequent as those between them and a Cushitic language like Bilin. In this situation, it is important to check how deeply this interaction affected the 68 positions of proto-Ethiopian lexicon as separated above. Cushitic > Ethiopian Semitic While many Cushitic loanwords penetrated Swadesh lists of particular Ethiopian languages, as we have seen above, only five examples can be treated as pan-Ethiopian.24 Their extremely wide spread throughout ES makes the assumption of independent borrowings into particular ES languages from one or several Cushitic source(s)25 unlikely. Therefore a progressive inter- dialectal infiltration as envisaged by Hetzron is a real possibility in these cases. However, the number of such examples is so low in comparison with the one hundred positions under scrutiny (or even those sixty-eight which have been qualified as proto-Ethiopian) that the methodological relevance of this process is rather insignificant (though, of course, not to be discarded completely). On the other hand, one should not forget that another, much easier explanation for such cases is at hand: the respective terms may well have been borrowed into proto-Ethiopian and inherited by its daughter languages side by side with the native Semitic lexical items. This is especially likely in the case of * âs× â ‘fish’ since no PS term with this meaning can be reconstructed at all, this notion being expressed by different terms of uncertain origin throughout Semitic. Ethiopian Semitic > Cushitic In view of a total lack of etymological expertise with Cushitic, my evaluation of this problem heavily relies on Leslau’s comments «passed into Cus- hitic», «also in Cushitic» etc. scattered on the pages of LGz. and LGur.26 According to my initial impression, the degree of Semitic infiltration into the basic lexicon of several Cushitic languages (Agaw, especially Bilin; Beja;

24 I purposely omit a few cases of similar terms widely attested in both ES and Cushitic, but most likely dating back to a common Afroasiatic lexical stock such as *sim- ‘name’ (DOLGOPOLSKY 1999:44) or *SVr- ‘root’ (HSED No. 551). An interaction between such terms is certainly possible and even likely (see above in connection with Proto-Ethiopian *ap- ‘mouth’) but one is hardly entitled to qualify this interaction as borrowing. 25 With the possible exception of the term for ‘meat’ which is less widely attested (Gez. s×#gâ, Tgr. Tna. Amh. s#gâ). 26 I am therefore aware that additional examples not mentioned by Leslau may have escaped my attention but it seems that such cases can hardly be numerous.

Downloaded from Brill.com10/02/2021 09:12:28PM via free access L. Kogan 391

Saho-Afar) was remarkable enough to provide a solid argument for the possibility of a similar spread of genuine (i. e., non-Cushitic) lexical isoglosses within Ethiopian Semitic. However, this impression does not seem to be con- firmed by a closer examination of the evidence. In the majority of cases, one is faced with quite a traditional type of borrowing accompanied by semantic shifts and scarcely affecting the deep stucture of the native vocabulary.27 Thus, ES *slm³ ‘to be black’ is borrowed into Bilin as èalam³ but its meaning is ‘finster, dunkel werden’ (RBil. 171), the main term for ‘black’ being niðîr (ibid. 286). Bilin belâ is attested only in the derivatives with the meaning ‘die Speise’ etc. (RBil. 78), the normal word for ‘to eat’ being k³« i (ibid. 233). ES *bVrk- ‘knee’ was borrowed into Bilin as bäräk ‘auf die Knie fallen’ (RBil. 86), Saho barak ‘fallen’ (RSa. 87) and Oromo birki ‘joint of fin- ger’ (LGur. 153), the normal words for ‘knee’ being girib (RBil. 159), gulûb (RSa. 153) and jilba ‘knee’ (Gragg 236) respectively. Bilin bä¶aS and Quara bä¶âS mean ‘sich vermehren’ (RBil. 72, RQu. 41) while ‘viel’ is rendered as gär-î-uh ¼ and ayû (ayiSû) in RBil. 157 and RQu. 21 respectively. Beja behâl means ‘Wort’ and behâli ‘sprechen’ (RBed.³ 45) but ‘sagen’ is translated as an, di or yad ibid. 323; Beja sit means ‘Getränke’ (RBed³. 205), the normal word for ‘to drink’ being g«a (ibid. 86). Even if the borrowed term is attested with the basic meaning, it most often co-exists with inherited synonyms denoting the same notions: Bilin leheè³â ‘die Rinde; der Bast’ (RBil. 253) but also k³âf ‘die Rinde’ (RBil. 236); Saho bala ‘essen’ (RSa. 80) but also bay-t, bç-t ‘zu sich nehmen (Speise)’ (RSa. 93) and k³am ‘Tabak oder Speise in der Mund nehmen, kauen, essen’ (RSa. 233), the latter being accepted as the main term for ‘to eat’ in Black 295. The only examples of true replacements found by me were Bilin èiffer³ ‘Nagel, Klaue’ (RBil. 169), Bilin näkät- ‘beis- sen’ (ibid. 283) and Khamta bi¶eq ‘viel’ (RCha. 346), to be supplemented by the only example which can be qualified as pan-Cushitic: Beja dera (RBed.³ 70), Khamta zíra (Appleyard 1987:504), Saho zäri (RSa. 333) ‘seed’ (no doubt, culturally determined). Obviously, the degree of interaction between languages belonging to different families is not to be automatically projected on a group of closely related idioms. However, since geographical, historical and cultural circumstances accompanying this interaction are often similar in both cases, the results of these excursuses do provide some support for the main conclusions of the present investigation: — the degree of unity of the basic vocabulary of Ethiopian Semitic languages is comparatively high; — this unity is unlikely to be explained in terms of borrowing and wave spread but relects the lexical peculiarities of one hypothetic source-language;

27 Regrettably, I am unable to judge the degree of the phonological adaptation in such cases.

Downloaded from Brill.com10/02/2021 09:12:28PM via free access 392 Scrinium I (2005). Varia Aethiopica

— a deeply structured combination of innovations and speficic retentions makes the Proto-Ethiopian vocabulary markedly different from that of any other Semitic subgroup and unambiguously points to the separate status of this sub-branch. Abbreviations of Languages, Dialects and Linguistic Groups Akk. — Akkadian, Amh. — Amharic, Arb. — Arabic, Arm. — Aramaic, Arg. — Argobba, Bib. — Biblical Aramaic, Cha. — Chaha, CS — Central Semitic, End. — Endegeñ, Enm. — Ennemor, ES — Ethiopian Semitic, ESA — Epigraphic South Arabian, Gaf. — Gafat, Gez. — Geez, Gog. — Gogot, Har. — Harari, Hbr. (pB.) — Hebrew (post-Biblical), Hrs. — Harsu- si, Jib. — Jibbali, Jud. — Judaic Aramaic, Mhr. — Mehri, Min. — Minaic, Mnd. — Mandaic, MSA — Modern South Arabian, Msq. — Masqan, Muh. — Muher, N.-Eth. — North Ethiopian, OB — Old Babylonian, PS — Proto- Semiic, Qat. — Qatabanian, Sab. — Sabaic, SB — Standard Babylonian, Sel. — Selti, S.-Eth. — South Ethiopian, Sod. — Soddo, Soq. — Soqotri, Syr. — Syriac, Tgr. — Tigre, Tna. — Tigrinya, Ugr. — Ugaritic, Wol. — Wolane, WS — West Semitic, Zwy. — Zway Bibliographic Abbreviations

AHw. — W. von SODEN, 1965–1981. Akkadisches Handwörterbuch. Wies- baden. Behnstedt — P. BEHNSTEDT, 1992f. Die Nordjemenitischen Dialekte. T. 2: Glossar. Wiesbaden, BK — A. de BIBERSTEIN-KAZIMIRSKI, 1860. Dictionnaire arabe-français. Vol. 1–2. Paris. Biella — J. C. BIELLA, 1982. Dictionary of Old South Arabic. Sabaean Dialect. Chico, CA. Black — P. D. BLACK, Lowland East Cushitic: Subgrouping and Recon- struction. Unpublished Ph.D. dissertation (Yale University). Brock. — C. BROCKELMANN, 1928. Lexicon Syriacum. Halle. Buck — C. D. BUCK, 1949. A Dictionary of Selected Synonyms in the Principal Indo-European Languages. Chicago. CAD — OPPENHEIM, L., E. REINER, & M. T. ROTH (eds.), 1956ff. The As- syrian Dictionary of the Oriental Institute, the University of Chicago. Chicago. DM — E. S. DROWER, R. MACUCH, 1963. A Mandaic Dictionary. Oxford. DRS — D. COHEN, 1970ff. Dictionnaire des racines sémitiques ou at- testées dans les langues sémitiques. La Haye. DUL — G. DEL OLMO LETE, J. SANMARTÍN, 2003. A Dictionary of the Uga- ritic Language in the Alphabetic Tradition. Leiden — Boston. Gragg — G. B. GRAGG, 1982. Oromo Dictionary. East Lansing. HJ — J. HOFTIJZER, K. JONGELING, 1995. Dictionary of the North-West Se- mitic Inscriptions. Leiden — New York — Köln.

Downloaded from Brill.com10/02/2021 09:12:28PM via free access L. Kogan 393

HSED — V. E. OREL, O. V. STOLBOVA, 1995. Hamito-Semitic Etymologi- cal Dictionary. Materials for a Reconstruction. Leiden — New York — Köln. Ja. — M. JASTROW, 1996. A Dictionary of the Targumim, the Talmud Babli and Yerushalmi, and the Midrashic Literature. New York. Jahn — A. JAHN, 1902. Die Mehri-Sprache in Südarabien. Texte und Wör- terbuch. Wien. JH — T. M. JOHNSTONE, 1977. Harsù ûsi Lexicon. Oxford. JJ — T. M. JOHNSTONE, 1981. Jibbâli Lexicon. Oxford. JM — T. M. JOHNSTONE, 1987. Mehri Lexicon. London. KB — L. KOEHLER, W. BAUMGARTNER, 1994–2000. The Hebrew and Ara- maic Lexicon of the Old Testament. (revised by W. Baumgartner and J. J. Stamm). Leiden — New York — Köln. LGur. — W. LESLAU, 1979. Etymological Dictionary of Gurage (Ethio- pic). Vol. III. Wiesbaden. LGz. — W. LESLAU, 1987. Comparative Dictionary of Ge ez (Classical Ethiopic). Wiesbaden. LH — E. LITTMANN, M. HÖFNER, 1956. Wörterbuch der Tigrç-Sprache. Tigre-deutsch-englisch. Wiesbaden. LLA — A. DILLMANN, 1865. Lexicon linguae aethiopicae. Leipzig. LM — M. ARBACH, 1993. Le madâbien: Lexique, Onomastique et Gram- maire d’une langue de l’Arabie méridionale préislamique. T. I. Lexique mad- dâbien. Thèse de doctorat — Nouveau régime. Université de Provence Aix Marseille I. Centre d’Aix. Aix-en-Provence. LS — W. LESLAU, 1938. Lexique Soqotri³ (Sudarabique moderne) avec comparaisons et explications étymologiques. Paris. RBed.—³ L. REINISCH, 1895. Wörterbuch der Bedauye-Sprache³ . Wien. RBil. — L. REINISCH, 1887. Die Bilin-Sprache. Bd. 2. Wörterbuch der Bilin-Sprache. Wien. RCham. — L. REINISCH, 1884. Die Chamir-Sprache in Abessinien. Wien [Sitzungsberichte der philosophisch-historischen Klasse der Kaiserlichen Akademie der Wissenschaften zu Wien, 106/1]. Ricks — S. D. RICKS, 1989. Lexicon of Inscriptional Qatabanian. Roma. RQu. — L. REINISCH, 1885. Die Quara-Sprache in Abessinien. Wien [Sit- zungsberichte der philosophisch-historischen Klasse der Kaiserlichen Aka- demie der Wissenschaften zu Wien, 109]. RSa. — L. REINISCH, 1890. Wörterbuch der Saho-Sprache. Wien. SD — A. F. L. BEESTON, M. A. GHUL, W. W. MÜLLER, J. RYCKMANS, 1982. Sabaic Dictionary (English-French-Arabic). Louvain-la-Neuve. SED I — A. MILITAREV, L. KOGAN, 2000. Semitic Etymological Dictio- nary. Vo l . 1 : Anatomy of Man and Animals. Münster. WH — A. WALDE, J. B. HOFFMANN, 1938. Lateinisches etymologisches Wörterbuch. Vol. 1–2. Heidelberg.

Downloaded from Brill.com10/02/2021 09:12:28PM via free access 394 Scrinium I (2005). Varia Aethiopica

References

Appleyard 1987 — D. APPLEYARD. «A Grammatical Sketch of Khamtan- ga, II». BSOAS 50: 470–507. Appleyard 1977 — D. APPLEYARD. «A Comparative Approach to the Am- haric Lexicon». Afroasiatic Linguistics 5/2: 43–109. Appleyard 1996 — D. APPLEYARD. «Ethiopian Semitic and South Arabi- an. Towards a Re-examination of a Relationship». Israel Oriental Studies 16: 203–28. Bender 1968 — L. BENDER. «Remarks on Glottochronology of Northern Ethiopian Semitic Languages». JES 6/1: 1–11. Bulakh 2003 — M. BULAKH. «Etymological Notes on the Akkadian Co- lour Terms». In: L. Kogan (ed.). Studia Semitica (Festschrift for Alexander Militarev). Moscow, 3–17. Cantineau 1932 — J. CANTINEAU, «Accadien et Sud-arabique». BSLP (Bul- letin de la Societé de linguistique de Paris). 1932 [no number]: 175–204. Cohen 1970 — D. COHEN. «Le vocabulaire de base sémitique el le classe- ment des dialectes du sud». In: ID. Études de linguistique sémitique et arabe. La Haye—Paris, 7–30. Dolgopolsky 1973 — A. DOLGOPOLSKY. Comparative-Historical Phone- tics of Cushitic. Moscow (in Russian). Dolgopolsky 1999 — A. DOLGOPOLSKY. From Proto-Semitic to Hebrew. Phonology. Milan. Ehret 1988 — C. EHRET. «Social Transformation in the Early History of the Horn of Africa: Linguistic Clues to Developments of the Period 500 B. C. to A. D. 500». In: Taddesse Beyene (ed.). Proceedings of the Eighth Interna- tional Conference of Ethiopian Studies. Addis Ababa, vol. 1, 639–651. Faber 1997 — A. FABER. «Genetic Soubgrouping of the Semitic Langua- ges». In: R. Hetzron (ed.). The Semitic Languages. London, 3–15. Goldenberg 1998 — G. GOLDENBERG. «The Semitic Languages of Ethio- pia and Their Classification». In: ID. Studies in Semitic Linguistics: Selected Writingds. Jerusalem, 286–331. Hackett 1980 — J. A. HACKETT. The Balaam Text from Deir Allâ. Chico. Hecker 1968 — K. HECKER. Grammatik der Kültepe-Texte. Roma. Hetzron 1972 — R. HETZRON. Ethiopian Semitic. Studies in Classifica- tion. Manchester. Hetzron 1975 — R. HETZRON. «Genetic Classification and Ethiopian Se- mitic». In: J. Bynon, T. Bynon (eds.). Hamito-Semitica. The Hague, 103–121. Hetzron 1976 — R. HETZRON. «Two Principles of Genetic Reconstruc- tion». Lingua 38: 89–108. Huehnergard 1998 — J. HUEHNERGARD. «What is Aramaic?» ARAM 7: 261–282.

Downloaded from Brill.com10/02/2021 09:12:28PM via free access L. Kogan 395

Huehnergard 2002 — J. HUEHNERGARD. «Comparative Semitic Linguis- tics». Israel Oriental Studies 20: 119–50. Kogan 2005 — L. Kogan. Lexical Evidence for the Historical Unity of Aramaic. Babel und Bibel 2 (in print). Kogan, forthcoming — L. KOGAN.«S in Ethiopian». To appear in: Fest- schrift R.-M. Voigt. Leslau 1975 — W. LESLAU. «What Is a Semitic Ethiopian Language?» In: J. Bynon, T. Bynon (eds.). Hamito-Semitica. The Hague, 129–131. Militarev 2000 — A. MILITAREV. «Towards the Chronology of Afrasian (Afro- asiatic) and Its Daughter Families». In: C. Renfrew, A. McMahon, L. Trask (eds.). Time Depth in Historical Linguistics. Cambridge. Vol. 1: 267–307. Militarev 2004 — A. MILITAREV. «Another Step towards the Chronology of Afrasian (I)». Babel und Bibel 1: 283–342. Nöldeke 1910 — T. NÖLDEKE. Neue Beiträge zur semitischen Sprachwis- senschaft. Strassburg. Pirenne 1990 — J. PIRENNE. Les témoins écrits de la région de Shabwa et l’histoire. Paris. Podolsky 1991 — B. PODOLSKY. Historical Phonetics of Amharic. Tel-Aviv. Porzig 1954 — W. PORZIG. Die Gliederung des indogermanischen Sprach- gebiets. Heidelberg. Rabin 1975 — Ch. RABIN. «Lexicostatistics and the Internal Divisions of Semitic». In: J. Bynon, T. Bynon (eds.). Hamito-Semitica. The Hague, 85–102. Ratcliffe 1998 — R. RATCLIFFE. «Defining Morphological Isoglosses: the «Broken» Plural and Semitic Subclassification». Journal of Near Eeastern Studies 57: 81–123. Renfroe 1992 — F. RENFROE. Arabic-Ugaritic Lexical Studies. Münster. Sima 2000 — A. SIMA. Tiere, Pflanzen, Steine und Metalle in den altsüd- arabischen Inschriften. Wiesbaden. Tropper 1993 — J. TROPPER. Die Inschriften von Zincirli. Münster. Ullendorff 1961 — E. ULLENDORFF. «Comparative Semitics». In: G. Levi Della Vida, H. Cazelles (eds.). Linguistica Semitica: presente e futuro. Roma, 13–32. Voigt 1998 — R. M. VOIGT. «“Fuss” (und “Hand”) im Äthiopischen, Syro- arabischen und Hebräischen». Zeitschrift für Althebraistik 11: 191–199. Zaborski 1991 — A. ZABORSKI. «The Position of Arabic within the Semi- tic Dialect Continuum». In: K. Dévényi, T. Iványi (eds.). Proceedings of the Colloquium on Arabic Grammar. Budapest, 365–75.

Downloaded from Brill.com10/02/2021 09:12:28PM via free access 396 Scrinium I (2005). Varia Aethiopica

SUMMARY

The present contribution offers a comprehensive analysis of lexical isoglosses behind the linguistic unity traditionally defined as «Ethiopian Semitic». In spite of the relatively small number of pertinent phonological and morphological features common to all Semitic languages of Ethiopia and Eritrea, their common lexical heritage has never been systematically studied. As is revealed in the course of the present study, there are several important semantic innovations affecting a number of key positions of the basic lexicon of all or most Ethiopian Semitic languages. Such innovations (together with a considerable number of non-trivial lexical retentions) strongly suggest that all these languages developed from a common ancestor rather than from several independent sources.

Downloaded from Brill.com10/02/2021 09:12:28PM via free access