Re: Syncope

--- In cybalist@yahoogroups.com, "Abdullah Konushevci"
<a_konushevci@...> wrote:

> Yes, for you is probably too easy, but, if I have understood you
> well, Alb. relative pronoun <i cili/e cila> or <i cilli/ e
> cilla> 'which' is one of the darkest question in albanology.
> If we accept gen. té-s-yo it should derive tje-s-yo > se-syo, but

I

> am not aware what should yields -sy- or maybe we must assume a
> syncoped form of *tésyoi> tsyoi> ci- which, suffixed further in
> other pronominal stem -l-(cf. Lat. ollus, It. ille and Irish tall,
> anall, etc.) would yields Alb. <i cil(l)i/e cil(l)a>.
> According to H. Pedersen, Alb. relative pronoun <i cili> is a
> dialectal prefixed form of <i t-sili> and the root is <si> and,
> according to Brugmann, from PIE *kWi-.
> This view was backed up by Çabej, who treats it as a backformation
> of much older form <i sij>, looking as plural, so, due to this

fact,

> was reshaped latter the singular form <i silli>.

You have not understood me well, but that is probably my fault. I
was referring to the Albanian Gelenkartikel which offers a
typological parallel to the Iranian izafet construction. Only, in
Albanian the linking particle is really the definite article, while
in Persian it is the old relative pronoun. So on this point,
Albanian is not an exact parallel for the analysis of the IE
genitive in *-o-s-yo as a genitive in *-(o)s + a form of the
relative pronoun *yo-, but the Persian form is.

I would suppose the derivation of si- from *kWi- is correct,
including the further analysis of ci- as t-si-.

Jens