From: Piotr Gasiorowski
Message: 20973
Date: 2003-04-13
----- Original Message -----
From: "Abdullah Konushevci" <a_konushevci@...>
To: <cybalist@yahoogroups.com>
Sent: Sunday, April 13, 2003 5:11 AM
Subject: [tied] Re: albanian prefix "stër-"
> It will be easy to accept Your chain of changes (it sounds like a good cantilena) and to say that the question is solved or te accept that labiovelar *kw- is turned in Alb. s only before front vowels, but, still now, my undrunked mind, tell me that this is not always true.
If a drink or two would make it easier for you to accept what lots of linguists have already accepted, go ahead! :-))
> If any sane mind accept some of my etymologies sen(d)/cen < *kwoina 'price, value' (cf. Old Slavic cena)
An etymology is more than pairing similar with similar. It must involve a detailed derivation and its justification. "Any sane mind" would agree that <deus> and <theos> must be related, but linguistic analysis shows that "the obvious" isn't true in this case. Where's your demonstration? To the best of my knowledge *kWoinah2 would have become *<qerë> in Tosk, so I have to reject your comparison and challenge you to present serious arguments if you feel like defending it.
> or dhjes < kekwr.t/gen. *kekwnes (probably Pokorny and Demiraj may have other explanations or they didn't treated them at all),
This is unacceptable either. An etymology that raises more problems than it solves stands no chance against a straightforward one. <dhjes> corresponds exactly to Gk. kHezo: 'shit' (*g^Hedjo:) and is also related to Skt. hadati (*g^Hed-e-ti, same meaning). The <-s> is easy to account for as devoiced *-z < *-dj-.
> it seems possible also from a zero grade of *kw(e)tw(e)r to be derived the number stër 'four', preserved only as a prefix stër- or in compounds like stërgjysh, besides katragjysh, stërçel. In one reduced form we have also tërplote 'quarter': dy këmbë e një tërplote 'two feet and one
quarter'.
If not broken by an epenthetic vowel, *kWt (as well as *kt and *k^t-) > Alb. t, cf. *nokWt- > natë. It isn't a priori impossible to get <tër-> from unstressed *kWtwr.- > *tur- (which doesn't mean that other possibilities are ruled out; we haven't explored any yet), but even that would mean a _direct_ derivation (not via <stër->).
Piotr