Re: [tied] Re: *gwistis

From: Piotr Gasiorowski
Message: 15602
Date: 2002-09-19

I didn't even check it in Pokorny, assuming that *gw- was self-evident (*gWi- would have had to yield Alb. zi-). To illustrate the (relatively late) loss of *w in this combination, cf. gjeth(e) 'leaf' (originally a collective of *gath < *gwosdos, meaning something like 'foliage, thicket', cf. Slavic *gvozdU 'young forest'). Albanian gisht- could reflect *gwist- or *gweist-. A strange omission on Pokorny's part. 
----- Original Message -----
From: Richard Wordingham
Sent: Thursday, September 19, 2002 2:26 PM
Subject: [tied] Re: *gwistis

--- In cybalist@......, Piotr Gasiorowski <piotr.gasiorowski@......> 
> As for  the absence of palatalisation, the reconstruction *gwist-
assumes an  initial cluster (*g + *w), NOT a labiovelar, and the
expected reflex  of *gwi- in Albanian is <gi->.

Actually I was being sloppy.  I hadn't realised that PIE had a
three-way _word-initial_ contrast gW v. gw v. g^w, spelt gw v. gv v.
g'v in .  Pokorny reconstructs gW, but
based only on Celtic and Germanic forms, which don't distinguish the
three initials.  If PIE *gwistis > Albanian gisht is correct, you have
refined the reconstruction to *gwistis (Cybalistic) / *gvistis