From: Piotr Gasiorowski
----- Original Message -----
From: "Abdullah Konushevci" <a_konushevci@...>
Sent: Saturday, April 12, 2003 5:05 PM
Subject: [tied] Re: Albanian-Romanian Concordances
> Alb. Sorrë is derived from PIE *kareka (about /k/ > /s/ in Albanian
we have discussed before)
There's no PIE *kareka to begin with (if you think otherwise, what forms in what languages point to such a reconstruction?). Rom. cioarã (as even Alex has observed) looks exactly like the reconstructed prototype of *sorrë, i.e. *c^orë. The emphatic rhotic /rr/ is usually taken to reflect a cluster such as *-rn-, *-rw- or *-rj-, but we should admit the possibility that in the name of a corvine bird it's an echoic "hardening".
Alb. s- < *c^- derives not only from palatalised *kW, but also from *k^w- (likewise, we have Alb. z- from *g^(H)w-). This allows us to pair <sorrë> with Balto-Slavic *C(w)árka: 'magpie' < *k^(w)orh2-k-ah2 (plus Skt. s'a:ri:, s'a:rika- ', and Gk. koraks 'raven'; a connection with *k(W)ers- 'black', as proposed by Demiraj, is far less compelling). I think a vr.ddhied formation *k^we:rh2-ah2 (or possibly *k^we:rh2-ih2, with PAlb. *-a: substituted for feminine *-i:, in which case it would correspond to Skt. s'a:r-i:) would work well for Albanian and is easier to justify than *k^we:rna: (Pokorny's choice). The lengthening would be as in Skt. na:ri: 'lady' or PGmc. *kWe:ni- (> Eng. queen).
To sum up:
*k^we:rh2(-ih2) > *cwa:r-a: > *c^orë > sorrë