10-11-03 22:29, alex wrote:
> well, we discussed about Alb. "një" where I said so far I remember that
> it should be seen as Rom. "unã"(<*une). You shoued some dialectal
> examples from Albanian and you presented Hamp's ideas on this "një"
So what? Good linguistics and naive guessing are worlds apart. Eric Hamp
did the best thing that could be done: he took into account several
dialects of Geg as well as Tosk, treating them all as equally valid
evidence, rather than speculate on the basis of Standard Albanian alone.
That's why his analysis is so convincing: njân-/një < *nje:n <
*(V)ni-ain- < *eni + *oino-. Of course, there is a distant relationship
with Lat. unu- < *oino-, but Alb. një is inherited and independent of
the Romanian word (which is simply a reflex of the Latin numeral). For
details, see:
http://groups.yahoo.com/group/cybalist/message/23406
Piotr