On Wed, 2 Oct 2002 01:59:23 +0200 (MET DST), Jens Elmegaard Rasmussen
<
jer@...> wrote:
>2. Final syllables are a nuisance. Being happy about the nice and tidy
>correspondences between Greek and Lithuanian in nom.pl. -oí : -ì/-íe- vs.
>loc.sg. -oi~/-ie~, one is just frustrated to the point of depair when the
>Greek optative 3sg lú:oi (not +lu~oi) is found to correspond to Lith.
>te-sukie~ (the socalled permissive, 'let him turn'), since here we know
>there was a laryngeal, PIE *-oyH1-t. To make it worse, Slavic has nom.pl.
>-i, loc.sg. -e^, ipv. (< opt.) -i. Do we have to posit an individual line
>of correspondences for each single item (when laryngeals are put in, the
>inputs *are* different)? Or could the Slavic imperative have been
>influenced by the form it would have after /y/, i.e. -i, just as the
>loc. of jo-stems ends in -i ??
Perhaps, but if we switch from soundlaw to analogy, why then only in the
Auslaut?
For my part, I would tend to consider the Lithuanian optative form as in need of
a special explanation (and Greek and Pre-Slavic -oí as expected).
I will assume the Lith. optative form continues a 3rd. person singular optative
(*-oih1t). Presumably, my explanation of the Armenian 3rd. sg. imperfect (<
optative) will have convinced nobody, but I'll sketch it out again:
1. *h1es-yeh1-m > ehye:n > eyi:~ > ei ( = eí)
2. *h1es-yeh1-s > ehye:s > eyi:z > eir ( = eír)
3. *h1es-yeh1-t > ehyeth > eyeD > eyr ( = é:r)
The point I'm getting at is the lack of syllable (< lack of length) in the 3rd.
person singular form, which in my opinion was caused by elimination of the
laryngeal through a merger with the following *t (*h1t > *th).
Perhaps the same thing happened in Lithuanian: *-oih1t > *-oith > *-oit > *-oi,
and therefore a "normal" diphthong, with default circumflex accentuation. In
other words: if we take the laryngeal *out*, the output may be different...
[In Slavic, the 3rd.sg. imper. form was of course at the mercy of the 2sg.
*-oih1s > *-oí > -i]
=======================
Miguel Carrasquer Vidal
mcv@...