Differentiation

[Jens Elmegaard Rasmussen]

On Tue, 15 Jun 1999, Patrick C. Ryan wrote:

> [In reply to my - Jens' - rebelling against "differentiation" as an
> explanatory strategy,]

> Pat responds:

> Well, I think there is another clearer example of "differentiation": present
> secondary 1st sing. -m as opposed to 1st pl. -me.
[]

I fail to see the reason for taking them to have been ever identical. Why
can't the plural form have contained an additional morpheme (expressing
the plural) to which the *-e could be credited? It would have to be
something either developing into non-vanishing *-e or causing some other
material to take this shape. It appears to me that "differentiation" just
amounts to absence of an explanation.

As for the main problem of 3sg *bhe'r-e-t, the facts that (1) the second
*-e- is the "thematic vowel" which was a meaningful morpheme (marking i.a.
the subjunctive), and (2) that the "thematic vowel" is never lost by the
working of the accent, combine to make *bhe'r-e-t quite normal. Still, it
may demand an explanation that the "thematic vowel" does not vanish when
unaccented - what is so special about it? I can suggest two solutions,
both ad hoc: Either the position where it stands was prosodically such
that it was retained: the thematic vowel is the only kind of stem-final
vowels the language has as far as we know. Or it consisted of some
enigmatic element which immunized the vowel against the working of the
ablaut. That the "thematic vowel" is not just any old /e/ is seen from
the fact that it alternates in its own way: -e- before voiceless endings
(including zero), -o- before voice (incl. vowels).

Since the special rules applying to the "thematic vowel" make it
impossible to ascribe its existence to simple "differentiation", we may
also accept the endings *-me, *-te of the 1st and 2nd plural as problems
for which the material and the rules have yet to be found. I would suspect
that a comparison with Proto-Uralic *-k-me-k, *-k-te-k (or *-t-me-k,
*-t-te-k ?) holds some of the answer.

Jens