31-01-04 21:56, Mate Kapovic wrote:
> I would write *s' (a *s with a single hook ยด) for BSl reflex of PIE *k'
> considering the reflexes.
Well, on the basis of Skt. /s'/ and Avestan /s/ you would probably feel
tempted to reconstruct the same for PIIr., and yet there's plenty of
evidence that the protophoneme was an affricate (we'd be able to prove
that even if the Nuristani languages had died out unrecorded).
Trubac^ev's argument that the Proto-BSl. was an affricate (retained as
*c in pre-Proto-Slavic down to a time when the contemporaneous ancestor
of East Baltic had already reduced it to a fricative) is based on the
existence of (pre-)Slavic loans in Lithuanian with /st/ for etymological
*k^. Trubac^ev interprets this /st/ as an attempted imitation of foreign
/c/.
> The type of sestra > se- > se-ka in Croatian is hardly IE. The hypocoristic
> shortening of the stem is hardly of an old origin.
Au contraire. Stem truncation in hypocoristic derivatives is and has
always been common cross-linguistically.
Piotr