Jens:
> But this is another matter, unless there is some relevance
> in the problem of getting the thematic vowel to alternate again
> so as to produce the input for *te-H2- > *ta-H2- and the like
> in pronouns where it plainly does that.
This has been shown to be a non-issue. It's the synthesis of
*t-e-x-, not *te-x-. We also see how *to- alternates with *te-
in the rest of the paradigm, which if analysed the same way
as other thematic stems, would mean that we should deconstruct
the stem as *t-[e/o]- for the purposes of declension even
though this is not the true etymological state of affairs.
This morphological analogy then is quite sufficient to produce
*sex and tex- long after the original event of Schwa Diffusion
that created the vowel alternation in the first place. Ergo, we
don't need a feminine gender in the oldest stage of IE which we
especially can't conclusively find in Anatolian anyway.
= gLeN