From: Joao S. Lopes
Message: 65992
Date: 2010-03-18
--- In cybalist@... s.com, "G&P" <G.and.P@... > wrote:
> [DGK]
> >The form *wenh1-us- avoids the difficulty that Old Latin *venos should have
> undergone /o/-umlaut of the /e/ in this >position, with a simple nasal
> between the vowels and /w/, /h/, or zero before the /e/. >We should expect
> Lat. *vonus, not <venus>, from an /es/-stem.
> [PG]
> Sihler discusses PIE *e > Latin *o before *w or l pinguis, and after
> consonant +w, then says, "Evidence for *we > Latin wo in other environments
> is scanty and suspect. ... A better case can be made for PIE *wemh1 > vomo
> ... Elsewhere *we- > vo only before l pinguis ..."
>
> So I'm not convinced we should expect *vonos from PIE *venos. As for
> *venh1-os, the slight evidence there is (PIE *vemh1o) suggests the h1 would
> not prevent this change (if there were such a change).
The /o/-umlaut which I have described is independent of the shifts before /w/ or /l/ pinguis, and after /sw/. That it is indeed umlaut is shown by <bene> from *dvene: (with iambic shortening) and <bellus> from *dvenelos, beside <bonus> from <dvenos>. Merely having a consonant (other than /s/) plus /w/ before the /e/ does not force the shift. I suspect that initial /y/ would also fit this scheme, but I have no examples of Old Latin *yeno- or *yemo-.
I leave <vomo:> sub judice, since I am doubtful that an umlauted root-vowel would be generalized from the 1sg. and 3pl. pres. ind. to the other forms.
DGK