Piotr Gasiorowski wrote:
> 04-02-04 00:21, alexandru_mg3 wrote:
>
>> Hello Piotr,
>>
>> 1. How you justify here :
>>
>> Lat. di: > PAlb. *di
>>
>> and not
>>
>> Lat. di: > PAlb. *3i > *di (or Di later di)
>
> If it had been affricated, it would not have been lost in intervocalic
> positions (*3 < *g^(H) wasn't; nor was *3 < *dj). However, it _was_
> lost in words like mall(�)koj <-- maledi:co:, therefore it must have
> been a plain stop.
Under the marge of error which can appearsdue the accepted Latimons
(Latimons= Latin etynoms) as being the roots for the Albanian roots, there
is no /di:/ > PAlb "something". The reflexes of Latin "di:" are simply "0".
For "di" the actual lexical data is ambigous. There are words where "di" >
"0" but there are words where "di" > "dh"
This is a question which should be cleared.
Since Piotr argues there is of no importance the quality of "i", then I
would avance the ideea that the words where Alb. prezent an "dz" are not
loans from Latin but Loans from Romance where already the /di/ > /dz/(Alb
dh).
The disadvantage of this assumption is that Latin /di/ > /de/ in Romance and
just the diphtongation of /e/ could make Romance /de/ > /dz(e)/. But a
diphtongation does not appears from blue sky just because we want to have X
> Y. There are precise rules for this diphtongation and here seems there is
som more work to do for clearing the soup out.
> Piotr
Alex