On 2008-02-20 01:48, stlatos wrote:
> What would s- mean, is it supposed to be causative?
How about an 'in-law' morpheme? In Starostin's database, a *-st- is
proposed. I'm not in a position to evaluate the proposal with my limited
knowledge of Burushaski, but the parallel between <hir, gus> and <-skir,
-skus> can't be ignored.
> ... each sound must obey all the changes I've given. Is
> it really possibly that haGor shows both retention of h-, w>g, and
> kY>r before w/u and that each rule has at least 2, and sometimes more,
> examples?
Even so, it doesn't work without a metathesis, which greatly increases
the chances of getting a spurious match.
To make matters worse, your independent example of *w > g- has to be
disqualified: it is obviously Iranian with or without a Dardic
intermediary (gus'pur 'prince' <-- *wisah-puTra- 'son of the [royal]
clan', cf. MPer. wispuhr, Shina gus^pu:r, with *w- > g in the donor
language, whatever it was). The Iranian word was borrowed widely, e.g.
into Georgian.
Piotr