alexandru_mg3 wrote:
> And if you will try to explain this too 'by analogy'...next you
> will need 'to assert an analogical restauration', for vjedhull too
> where the intervocalic dh is well visible...and this 'analogical hocus-
> pocus' will never arrive to an end...
Of course it's highly likely that the <d> of <vjedull> owes its presence
there to restoration before a productive suffix. Cimochowski cites the
dialectal forms <vjedhullë> (the word was originally trisyllabic!) and
<vjetullë>, suggesting an original *vjed (> *vjedh) alternating with
*vjet (with normal final devoicing), possibly related to the verb
<vjedh> 'steal', and extended with <-ull(ë)> after the operation of
word-final phonetic rules. He also notes another dialectal variant,
<vjellë>, which may be the expected regular reflex of inherited
*vjed-ëla:, an elder cousin of the standard form.
Piotr