Richard:
> Common sense says that the *-yo has to be nominative.
As I suggested, a locative still remains a better possibility.
There is no *-s and no reason why it should disappear, so
nominative it can't be without rewriting IE... which is
in effect what Jens is doing. You can shishkabob me if you
like but time will tell and I'm not scared.
> The only hope for a version of Glen's analysis is that our
> common sense is warped by the rarity, as opposed to
> non-existence of, of relative clauses expressing genitive
> relationships in which the possessum, not the possessor, is
> marked.
This is ONLY if you analyse the phrase one way. There's also
a possibility based on normal IE word order if we can tell
at all that a theoretical preform */wlkW&s ya hWa:kWs/
meant "wolf('s) with his eye" where *ya actually did refer
to "at his/hers/its/one's/someone's" and modified the possessum
like in the Hungarian example. That would be consistent with
a relative pronoun preposed to its clause. So your point about
the possessum/possessor order would be unwarranted in this
interpretation.
> Are double long vowels in Estonian a local development? My English
> idiolect seems to have three vowel lengths - the vowels in _beat_
> and _bid_ seem to be the same length! (These words are also
> distinguished by vowel quality and voicing of the final consonant.)
> Rare, but not impossible.
That doesn't mean that English has double-long vowels. I think
you misunderstand. In Estonian, there is a _contrastive_ difference
between short, long and double-long. It really does make a difference
in the meaning of the word you pronounce. This contrast doesn't
exist in English unless you can provide an example.
Regarding "bit" and "bid" (which Schwa Diffusion in mLIE is parallel
to, btw), the vowel length is phonetic, not phonemic, meaning that
the length is predictable by its environment and non-contrastive.
There's no real contrast of /bId/ and */bI.d/.
= gLeN