29-01-04 00:59, Abdullah Konushevci wrote:
> Because we are too late and I fill very tired, I like justo to know
> what about *s > h (*selk/solk > halq > heq 'to drag', *suil > hyll,
> *ke:sa > kohë, Slavic *c^asu) even you give *keh1k^-sah2) after back
> vowels and *s > sh (*su: > shi 'rain'; *pe:d-su > posh 'down'), etc.
If *pe:dsu (> *pe:su) > posh is correct, it only illustrates the
retention of voiceless *s in syllables that were unstressed in
Proto-Albanian. The story of <shi> must be more complex than Pokorny
makes it; there's no external support for PIE *su:- (*suh-) as a root
noun. I find it hard to accept either *s(w)olk-eje- or *h2wolk-eje-
(both have been proposed) as a valid etymology for <heq> (not that I can
confidently offer a better one). There is no PIE "*suil", and the
connection between *sh2wel- 'sun' and <(h)yll> 'star', once widely
accepted, is now largely abandoned (<diell> 'sun' is a far better
match). It's hard to be sure, but *h1uslo- 'glowing ember' (with
Germanic cognates, from *h1eus- 'burn') may be a good idea, especially
if *-sl-, like *-sr-, caused compensatory lengthening in Proto-Albanian.
Semantically, it parallels *h2(h1)s-tér-.
I forgot to mention the sporadic occurrence of <th-> /T/ as a Modern
Albanian reflex of initial *s-, cf. <thi> 'pig' (*su:s) and <tha(n)j>
'dry up' (*saus-n-jo:). Cimochowski suspects dissimilation here: *s...s
> *c...s, but if so, it must be older than the change of initial *s to
*z under stress and the dropping of inflectional -s.
Piotr