Notes on Slovene and Slavic Etymology The notes which follow find aspects of F. Bezlaj's EtimoloSki slovar slovenskega jezika to criticize. I hope this will not be taken adversely. On the other side of the balance we have an enormous debt of gratitude which we owe Bezlaj for the fact that he has given us the first major step in the production of a modern etymological inventory of the Slovene language and the scholarly exploration of the history of its lexicon. The critical nature of these notes is meant to point rather to the directions we may and must follow in the future, in my modest opinion; and which we may, thanks to the enormous aid that Bezlaj's work offers us, whereby we stand on his shoulders. Moreover, we must always remember that in the difficult days when Bez- laj was compiling much of his work there were many publications that were simply not available across borders in the world. The principal themes of the following notes are the first-order contribution of Slovene evidence to the comparison and history of the Slavic languages; their place in and contribution to the history of the Indo-European languages; the importance of Indo-European evidence to the solution of Slavic problems; the clarity and illumi- nation that comes from an insistence on the absolute regularity, in principle, of lin- guistic change; and the cumulative nature of scholarship. 1 kJma I , I I , and Krma 111 F. Bezlaj, EtimoloSki slovar slovenskega jezika 11 K-0, 1982, 95-6, agrees with received opinion in tracing kJma I 'pabulum' and its clear Slavic cognates to IE *(s)ker- 'rezati' with the suffixal formation *-mo-, -ma. This attribution must be roughly correct, but the specification offered for the aggregate derivation is impre- cise to a degree that today we must exert ourselves to avoid. Czech krmd is of course not precisely *-ma. Serbo-Croatian kr'me -ta 'svinja', collect. kr'mad is yet another formation. The semantics here also requires fuller spe- cification. One might think of pork as the preferred food or meat in Northern Eu- rope, as I have argued in the Festschrift for Marija Gimbutas (1987) and in Zeit- schrift fur Celtische Philologie 41, 1986, 257-8. But the prevalence of expressions with krmiti (also in Czech) for 'fatten' (esp. for slaughter) and for feeding pigs leads one to suspect that the formation of kr'me is based on feeding the pig, as in krmafa and in kfmek, rather than in the pig as potential food. On the other hand the IE comparisons are flawed. Lithuanian Sirti, Seriu etc. cannot be related to Lat. cena, kJma, and *(s)ker-; with its initial *k', however, it may be related to Albanian thier, thjerr(e) 'ervum lens', which I have analyzed (Bu- letin Shkencor, Seria e shkencave shoqerore IUniv. Luigj Gurakuqil 27 1471 111994, 89-90). Russ. Ukr. Br. korm, OPol. karm and Slovak krm < *kyma are not directly the expected *kor-mo-s, which is functionally equivalent to Lat. cena. Here * k r m must be contaminated with *kyma the well formed IE ancestor of kJma I, SCr. kJma, Bulg. karma, Mak. krma,' Ukr. korma, and Slovak krma. ' Mak. krna appears to be a misprint. E. P. Hamp, Notes on Slovene and Slavic Etymology 161 We recover here in kr'ma I an interesting addition to the examples I have ad- duced in KZ 96, 1982183, 171-7. kr'ma ZZ 'puppis' (Bezlaj 96) is accorded an account, with references to earlier literature. But despite the rich citation of Slovene derivatives no explication of the morphology is actually supplied; Etimologija ni razEiSCena. At least this morphology is discussed by this writer in KZ 96, 171-7, esp. 171 and 176. The toponym Krma 111 is referred to pre-Roman origins - often a counsel of despair - with the mention of Albanian karme 'cliff, wall'. I have, however, derived karme from *karp-na, an old collective. The Lautgesetz involved here is purely Al- banian, and imposes thereby a severe constraint on the presumed source. We have here, then, three different formations. The lexeme kfma II could have an origin in the base *krp-; cf. Latin corpus, Baltistica 15, 2 (1979) 145 (OPruss. kcrmens), and Etudes Celtiques 9 (1960) 139-40 (139, Welsh cryf < *kr(p)mo-). The Slovene evidence on this interesting etymon is very important, and the ac- count of F. Bezlaj ESSJ I1 (1982) 22-3 does not bring this out sufficiently. It is of course correct that the Proto-Slavic form must be reconstructed *kailjo- vel sim, as Bezlaj does. However the cited Church Slavonic kai'blb shows a written internal jer. It is therefore important to note the Slovene 18th century attestations which include kashl. Of course the question of precise (morpho)phonemic status enters here, as well as that of the phonetics of 16th century kashel. But the native testimony is it- self a valuable datum. The Indo-European account, and the setting within the IE dialects, for this item is deficient. For a fuller accounting see E. P. Hamp, in Ezikovedski prouc'vani- ja v c'est nu Akad. V . I. Georgiev, Sofia: BAN, 1980, 130-4; Zndo-Iranian Journal 30, 1987, 103-6; ~ r i u 39, 1988, 193-4; Scottish Gaelic Studies 16, 1990, 194. 3 bgenj As with kaielj the Slovene testimony for bgenj is rich and important (ESSJ 11, 1982, 243). We note especially the 17th century oghgn, a highly valuable form to be placed alongside the Church Slavonic ogrib. However the Indo-European account which follows the recital of Slavic cog- nates is not at all acceptable, and the appeal to linguistic tabu is simply a confession of failure to find any principled solution. On the whole question I would refer the interested reader to my discussion and analysis in T. F. Magner and W. R. Schmals- tieg eds., Baltic Linguistics, University Park: Pennsylvania State University Press, 1970, 75-9. On the exact background of vlgenj 'forge, hearth' I have now (Rocznik stawistyczny 49, 1, 13-14) revised my view and formulation as follows: Cz. vjheii, SCr v'iganj, Slov. vlgenj < *vjgnjb < (by Winter's lengthening) *ugnjo- < (by dis- similation or perceptual misassignment) *ungnjo- < *~gni-o-. The question of vatra has no part in this discussion, and I have dealt with that etymon elsewhere, esp. Opuscula slavica et linguistics: Festschrift fur Alexander Is- satchenko, Klagenfurt 1976, 201-10; Revue roumaine de linguistique 26, 1981, 315. 162 Slovenski iezik - Slovene Linguistic Studies 1 (1997) 4 (igIje As with the last item, the Slovene testimony for (igIje is highly valuable and rich in its variety and citation of variants (ibid. 243-4). The 18th century form ogl is particularly interesting, and the range of attested suffixation is instructive. But again the comparative Indo-European account is deficient. My 1970 essay mentioned under bgenj deals also with (iglje and kindred forms, and the reader is again referred to that. The etymon of (igel 'angulus' is also men- tioned there, but that is not part of the main focus there. However that discussion does deal with the Albanian loan from Slavic for which Bezlaj simply cites and re- produces Pisani's uninformed confusion. Prispelo maja 1995 Received May 1995 Eric P. Hamp University of Chicago UredniStvo je na sestanku 6. decembra At the meeting of the Editorial Board 1996 odlotilo, da se ob soglasju avtorja in on December 6, 1996, it was decided recenzentke objavi tudi recenzija tega pri- that with the agreement of the author spevka. and the reviewer, the review would be published accompanying the article.