On 2006-05-11 21:44, tgpedersen wrote:
> Like *gWow-s?
No. The original genitive of *gWo:u-s was probably *gWeu-s, as
reconstructed by Schindler. Skt. goh. isn't decisive (the /o/ may go
back to *eu or *ou, and the unpalatalised initial may well be due to
paradigmatic levelling), but the weak cases in <gav-> (instr. gava:,
dat. gave, etc.) point to *e, since *o would have been lengthened by
Brugmann's Law (it _is_ lengthened in nom.pl. ga:vah. < *gWówes).
Piotr