Re: [tied] 1sg. -o: [was Re: IE Thematic Vowel Rule]

On Tue, 23 Aug 2005 21:06:21 +0200, Miguel Carrasquer
<mcv@...> wrote:

>That leaves of course the question of where the -a:u in the
>perfect forms comes from.
>
>To come back to the point we were discussing earlier, the
>first person ending -o:, which phonetically can only come
>from *-oh3 (because of the Balto-Slavic acute), perhaps the
>answer is something along the following lines:
>
>In a cluster C+CW (consonant plus labialized consonant), the
>labialization of the second consonant is given up, cf.
>Armenian N.pl. *-esW, *-o:sW > -k` but A.pl. *-o:ns > -s.
>That would explain PIE *h2 for expected *h3 (< *-ku) in the
>1sg. perfect. After a vowel (in thematic verbs), the ending
>is maintained as *h3, so we have thematic 1s. subjunctive
>and present *-o-h3(u) > Toch. -eu, elsewhere -o:. There are
>no thematic perfects: the perfect stem always ends in a
>consonant. But perfects with roots ending in a laryngeal
>may have assimilated it to the laryngeal of the ending *-h3-
>(still labialized /xW/), so we get 1sg. *dhe(x)-xWe,
>*ste(x)-xWe, *de(x)-xWe. But I still don't understand why
>the final vowel dropped out.

Correction, these are perfects, I forgot the reduplication
and the o-grade:

*dhe-dho(h3)-h3o etc.

Melchert (Anat. Hist. Phon., p. 52) tries to make Jasanoff's
rule (-óHe > -óHu) phonetically more plausible: "Since at
least some cases of absolute final short -e probably reflect
earlier *-o (thematic vocative -e vs. regular -o-), the rule
becomes more plausible phonetically if we project it back to
pre-PIE as *-óHo# > *-óHu#", except that vocative -e does
not reflect "earlier -o-", and I cannot accept such an
unmotivated final -o.

On the other hand, -oH- by itself does seem to trigger a -w-
glide in cases like Slavic *stoh2-éie-tei > staviti,
*bhuoh2-éie-tei > baviti.


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