The forms with s- are generally explained from the stem seen in Sanskrit
tásya:s, tásyai etc., meaning that they represent the encounter of *t- +
the reflex of word-internal *-sy-, cf. the normal change *ty > s. Once
this is accepted it can be stated without restriction that the Albanian
article has initial vowel where IE *so/to- had /s-/, and *reflex of *t"
where *so/to- had /t-/. This has been known for over a century.
I like Abdullah's interpretation of the first parts of k-y and a-í as
local adverbs 'to' and 'from' which connects them with well-known
prepositions.
Jens
On Tue, 7 Oct 2003, alex wrote:
[...]
> Assuming that (a)saj is the reflex of *so-, then here we have a "s"
> reflex of "so-" the Alb. (a)saj.
> Is there not a contradiction?
[...]