Re: w-glide

W dniu 2011-06-16 00:02, Sergejus Tarasovas pisze:

> Differently from *-va"/vo~s, I don't have a "gut feeling" of what this
> *h2(a)u might mean, especially in a combination with '1'. Besides 'on
> the other hand', dictionaries give meanings like 'away (from), back,
> again'. Do I get it right that in Greek aû is used in contexts like
> "secondly, thirdly" etc?

Yes, something of the sort. "Secondly again, thirdly again..."

> > Talking of
> > stress, wouldn't Nieminen's Law (retraction of stress from short *a in
> > the final syllable to a preceding long nucleus) be older than the
> > monophthongisation? Cf. die~vas < *deiwó-.
>
> My impression is that there's no mainstream view even on the exact
> formulation of this law, much less on its relative chronology. In view
> of forms like greitàsis 'fast (definite)', is it safe to suppose they
> were univerbated already in Early East Baltic? In any case, even in the
> pre-Nieminen state o-stems must have had enough barytone cases (in
> singular -- all but nominative?) to provide the necessary pressure for
> analogical leveling.

I know. Some of those forms can be explained differently anyway. For
example, there may have been a contrast between adj. *deiwó- 'divine'
and subst. *déiwos 'deity' already in PIE, even if Skt. doesn't show it.

> Slavic '1' is not without problems, too. If we want to start from an
> oxytone (*-nó-), we need a zero grade *(H)iH- to let Hirt's Law convert
> it into an (a)-word which Slavic *i"nU is (-eiH- would not attract
> stress). East autem Baltic forms require a full-grade *(h1)eiH-nó- or
> even better (to go together with Old Prussian) *(H)oiH-nó-. What to make
> of it?

If I knew, I would publish it immediately :). Most authors are
blissfully unaware of any problems with '1' and stick to traditional
*(H)oi-no-/*(H)oi-wo-/*(H)oi-ko-. To be sure, the second and the third
of these variants may well be correct, but the first just can't be. If
the first *o is apophonic, Saussure's Law should have removed the medial
laryngeal (so *(H)oi-no- would be justified for some branches), but
Slavic and East Baltic still require it to be there. Perhaps the true
morphological division is *(H)(o)i-Hn-ó-, which would probably allow us
to get round Saussurean effects.

Piotr