Abdullah Konushevci wrote:
>> more simple for him to say " yes, shk= allowed and "shg"= not
> allowed",
>> at leas in the initial position.
>>
>> Alex
> ************
> Yes, in the beginnig of the words, at most, all voiced occlusives
> before intensive prefix sh- becames unvoiced (cf. <shkardhë> 'fence-
> gate', but <zgardhoj> 'to de-fence, to come lose (of barrel staves)'
>
> Konushevci
That is that. How expected, there are rulles they work very clear with
as well as no exception.And for that, for the phonetical interdictions
we have this make the basis to be very solide. This is why I say that an
explanation as " stimmhaftes k" , even if true, is weak in such
comparative cases.
I guess the presence of "shkardë" and "zgardã" with the same meaning in
Alb. And Rom. as well as presence of "gardh" and "gard" with the same
meaning, _exclude_ any Slavic loan here.
For the metathesised form in Rom. with "grã-", I won't be so sure. The
derivatives are all of "grãd-" and one will think imediately at the
Slavic metathesis. At least, at time I don't see another solution unless
there is a root in PIE as *ghre-, or *g^hre- with the meaning one need.
Alex