Re: [tied] Re: Albanian names (2) -> Besa, Besiana ("Google 1720" t

04-11-03 20:16, alex wrote:

> If we accept that "s" in "sorë" ios because of a "ke, ki" then one
> should accept the "s" in besë" can be from an "ke, ki" as well":

No. In <sorë> the *s comes from *k^w-. There's no other way to get Alb.
s from *k^.

> since
> the last wovel now is an "ë", this one can be just form an "a" or an
> "e";

No. The usual source of Alb. -ë (word-finally) is *-a:. An *-e (or *-o,
for that matter) would have been deleted.

> in this way we exculde "i" but we remain with "e";(1)
> The question sould be which is the origin of the "e" here.
>
> One thought more about. Should be the word "bast" (Wette) related here?
> Semanticaly this should be too a kind of pactum.

I'll be hanged if I understand what you mean. Anyways, to get <besë>
from *pak^- (even ignoring the question of *p- > b-, which mustn't be
left unaccounted for) you'd need something like *pak^wih2 (so that *a
can become umlauted to give modern <e>, and *-k^w- can produce <s>)
transformed into an *-a: stem. But what the heck is *pak^wih2? An
unheard-of derivative (and the embarrassing question of *p- > b-
remains). Your proposal creates lost of problems and solves none.

Piotr