Thematic vowel etc

The Russian adj. inflection in m.nom. has ´-yj, but when stressed
it's -ój.
Now here's a hen-or-egg discussion:
1) did the /o/ attract the stress? or
2) did the stress change the /y/ to /o/?

The most reasonable seems of course to assume 2).
We know that o-grade of an ablaut vowel occurs when stress has been
shifted from somewhere else to that vowel. Supposedly, the first
step was that all stressed ablaut vowel got e-grade and the rest of
them zero grade. But the part about the zero grade can't be right,
since how can a vowel in zero grade, in other words, that isn't
there, develop into an /o/ when it's stressed, when there is nothing
there to be stressed? So I think Glen is right that in the first
step the unstressed ablaut vowels did not disappear (get zero
grade), but developed into some type of schwa, which then could
either get stress in the next step and become /o/, or not, in which
case it disappeared and truly became zero grade.

As for the verbal thematic vowel I can't say anything, but about the
one in nouns I can make this observation:
Of all the things that were inflected like nouns in the beginning,
the demonstratives were the only ones that were not syllabic. That
means they could not possibly have stress anywhere in the stem, but
must have it on the suffix. Therefore the demonstratives grew a
vowel and now look as if they have a thematic vowel.

Note this German example:
nom. das warme Wasser
dat. dem warmen Wasser "the warm water"
cf. without demonstrative
nom. warmes Wasser
dat. warmem Wasser "warm water"
Note the endings of the adj.!

Now why does German do that? I think because the
otherwise-inflected demonstrative acts as a left bracket to the
noun's right bracket; in other words, it delimits the extent of the
NP in the sentence. That device seems so useful that Germans don't
want to give it up even when there's no demonstrative present in the
NP, so they add the endings that the demonstrative would have had,
to the first member, namely an adjective, of the present NP.

Now, in PIE speakers felt the same way, they would in the NPs where
there were no demonstratives have transferred the ending of the
demonstrative (which had a stressed /o/ in it) to the first
adjective, which therefore ended up stressed on the suffix which
contained /o/. Thus was born the thematic inflection.

I like the idea that the thematic vowel is there, even where you
don't see it (because it's been elided by a zero-grade rule).
Perhaps even the athematic verbs contain a lost thematic vowel,
which then must have been the suffix that turned the verbal stem
into a gerund.


Torsten