Re: [tied] Re: Decircumflexion, N-raising, H-raising: Slavic soundr

On Wed, 28 Apr 2004 14:26:46 +0200, Sergejus Tarasovas
<S.Tarasovas@...> wrote:

>> From: Miguel Carrasquer [mailto:mcv@...]
>
>> The rule as I gave it was:
>>...
>
>
>We seem to be starting a second round. During the first round, you
>formulated the rule, part of it being *-(i)ái > -ì (in nouns and adjectives
>except -ijo-contraction under stress and later analogy in nouns). I asked
>why we find *-ái > -ai~ (not -ì) in verbs then (sakai~, not +sakì); and why
>would unstressed contraction -- while yielding a *long* vowel (unstressed
>-ys still exists in dialects and is attested in Old Lithuanian) -- have
>yielded not circuflex, but acute (unlike *-ja:-stems). You ignored the first
>rebuttal

I forgot to address that. Sorry.

The verbal form comes from *-a:- + *-eí, which I'd expect to
give *-a:i~ (like the a:-stem. Dsg., although that's from
*-a:- + *-(e)i~), which I'd like to keep distinct from *-aí
(< *-oy) or *-ai~ (Slav. Lsg. *-oi, Lith. nom.pl. *-iyoy),
with short vowel.

>and reported the same rule one more time with one minor correction
>(probably to account for the second rebuttal): now didelì is accounted for
>not as "-soft B -ijaì > -jaì (heavy stressed root) *-ijo- : dìdelis ->
>didelì" , but as "-yoy > -iaí > -íe" (so it seems stress doesn't play the
>role and dìdelis is classified as an *-jo- -- rather than *-ijo-stem now).

No, that was simply a mistake. Read: z^ãlias, z^alì. It's
possible, however, that unstressed *-ijo- was reduced to
*-jo- early on, at least in the nom.pl. (not in the nom.sg.,
where there's a difference between -ias, -is and -ys). I'm
undecided really about whether didelì is the regular outcome
of unstressed *-iyoy, or whether it's analogical after gerì,
z^alì.

>If your answer to the first rebuttal is "ái *can* yield íe", that doesn't
>solve the problem since one feels the need of an exact rule: under what
>conditions "uncontracted" -ái yields -íe (pronouns, adjectives, nouns (as
>per your theory-- later analogically replaced)), and under what conditions
>-- -ai~ (eg., verbs).

Only short vowel -aí (not -á:i) was liable to yield -íe >
-ì.

=======================
Miguel Carrasquer Vidal
mcv@...