Re: [tied] Question on Albanian sy

alexandru_mg3 wrote:

>>The development of the <ci:vit-> part was approximately like this:
>>*ki:wit- > *kiwët- > *kjut- > qyt-, with the /u/ fronted by
>>assimilation. This should not be directly compared with
>>developments in stressed syllables.
>
>
> I don't see you point here....is obvious that ju uj iwi (with or
> without long i) gave y ...I cannot see any distinction regarding
> stressed or unstresed syllables: iu,ui,iwi > y is available
> everywhere...maybe you have another examples to show the difference...

OK. We have <gjykoj> from *gjyik- < *jydik- < iu:dico:, but here the /y/
is due to the original length of /u:/, not to the initial /j/ or the
contraction with /i/. In <kujtoj> from *kuit- < *kugit- < co:gito: the
/i/ has become non-syllabic but hasn't influenced the /u/. In toponymic
<Rrjoll> from Lat. ri:vulus/ri:volus there's no effect of the /w/ on the
first vowel.

Piotr