Re: [tied] IE genitive

> From: Jens Elmegaard Rasmussen [mailto:jer@...]
> Sent: Tuesday, April 29, 2003 11:50 AM

> Not generally, but in a case like this one it certainly can.
> Skt. icchaïti has zero-grade as expected in an sk-present.
> That is certainly not the case with Lith. iïes^k-o- (older
> -a-), Slav. is^c^e- also with acute. The short -i- of icchati
> precludes a laryngeal, so the acute tone cannot be ascribed
> to a prestage with "*HeyHs-"; nor can it be Winter's Law
> since there is no voiced stop in it; it can however easily be
> *H2e:ys-, the expected lengthened grade of *H2eys-. The
> example is even so good as to be
> decisive: the Dehnstufe-to-circumflex theory is simply wrong.
>

I'm not so sure. First of all, what makes you reconstruct Proto-Slavic
*iskati rather than *jIskati (SCr.<ìskati>)? Or do you postulate an
alternation *jIskati ~ *is^c^e-? But does SCr. <îsc'e> point to an old
acute? As for the Baltic forms, taking into account extremely developed
metatony, how one can be sure we deal with something original here?
Lith. <ies^kóti> has the ictus retracted to the suffix by Saussure's
law, which would point to an original non-acute pitch (cf. <láidoti> ~
<léisti> etc., where the ictus is not retracted), and the accent of
<íes^ko-> well may have been influenced by forms such as <ýs^kus>,
<éis^kus> 'clear' (also by semantic contamination of 'seek' and
'explain, make clear'). One example is not enough to be decisive: IMHO,
it takes a representative sample to be guaranteed against metatonic
noise.

Sergei