On 2008-08-12 01:43, alexandru_mg3 wrote:
> 1.a *serp-en- > e/ya in closed syllable > PAlb *sjar-pe- so we can
> well have here a 'older' sh/zh from sj
>
> 1.b same if from *sek^s-ti- > e/ya in closed syllable > PAlb *sjak-
> sti-
>
> 1.c gjalp&r < *sjalp- < *selp- with e/ya again in closed syllable
>
> So *sup-no reconstruction for Albanian needs in this case to be taken
> with caution
But the last etymology happens to be very secure. Besides, *s- becomes
sh- also before consonants (e.g. *st > sht) and even in loans before a
variety of vowels (Lat. sagitta -> shëgjetë ~ shigjetë. The simplest
explanation is just retraction, not conditioned palatalisation (which
creates more problems than it solves). As for the fate of *sw- in
Albanian, opinions differ. In my opinion, the "regular" development is
*sw > *ð(w) > d(h) before a stressed vowel, and *sw > *þ(w) > þ
otherwise, but I'm the first to admit that the evidence is scarce. My
only decent example of the latter change is djath(të) 'right' < *deswo-
< *dek^swo-. <vjehërr> contains two aberrant developments, which makes
it likely that both are due to the same cause (metathesis). But we went
through that three years ago:
http://tech.groups.yahoo.com/group/cybalist/message/38674
Piotr