Of the Satem groups, Slavic shows a
particularly large admixture of centum forms (larger than Baltic an far larger
than Indo-Iranian), e.g. *go~sI 'goose', *svekry 'mother-in-law', *korva
'cow'. At least those with *k cannot come from post-Grimm Germanic and
must be due to the influence of some extinct non-Satemic substrate of E/Central
Europe (Venetic again?). *kop- is certainly one of them. It is difficult to tell
at present whether we deal here with lexical or phonological diffusion (the
latter would be like "dispalatalisation" in Northern English dialects under Old
Norse influence, which is the reason why Modern English has "get, give,
carve" rather than "yet, yive, charve").
As for hippos, the geminated "p" is
explained as reflecting an IE consonant cluster (*k + *w), not a unitary
labiovelar *kW (there is also a Doric variant ikkos, cf. lakkos 'pond' <
*lakwos, pelekkao: 'cut with an axe' < *peleku-a-). Initially before a
consonant cluster Greek i in place of *e (already in Mycenaean) is at least
parallelled by i-prothesis in the same environment (as in ikhthys or iktis ~
ktis). The aspirate is a real mystery, although if I'm not mistaken it's
confined to Attic Greek.
Piotr
----- Original Message -----
Sent: Friday, October 27, 2000 4:05 PM
Subject: Re: [tied] names for HORSE
Joao wrote:
Ok, but remember the roots is *k'opH- (with
metathesis k'oHpo-, so Satem languages like Slavic must have *sop-. So,
kopyto must be a loan from some kentum language.
If caballus is Celtic and related to *k'opH-, I
think it's more plausible a compound *k'opHo-bolo- or alike.
It's not impossible to use the same root for a
castrated animal and a female. Remember English cow and Portuguese boi
(<Latin bo:s) "castrated ox".
kobyla < *kobhu:- ? or kobhaxw- ?
Talking abour horse names there's also the mystery
of name HIPPOS. The "normal" word must have been
*EPOS