--- In
cybalist@yahoogroups.com, Rick McCallister <gabaroo6958@...>
wrote:
>
> Ivanov's remarks that Balkan vatra is from Iranian is
> interesting. I had read it was from
> pre-Albanian/pre-Romanian substrate. Any comments?
>
>
> --- Francesco Brighenti <frabrig@...> wrote:
>
>
> > see V.V. Ivanov's
> > discussion at
> >
> >
> http://www.humnet.ucla.edu/pies/pdfs/IESV/1/VVI_Horse.pdf
> > (on pp. 84-85 of the pdf document)
> >
>
I didn't see any agument there : maybe there are some but not in that
article
Vatra cannot be from Iranian because
1. Aromanian sg. vatra as a sg. variant 'veatra' (Papahagi)
2. Romanian pl. is 'vetre'
Both Romanian Dialects shows in addition the presence of an e
----------------------------------------------------------------
3. Albanian forms are vot&r ~ vat&r
So this can hardly has something to do with Iranian a:tra
In addition even the PIE a:tra doesn't contain any /e/
Piotr suspected that Romanian e 'in the pl. 'vetre' is a later
evolution in Romanian, than influenced also the Aromanian sg.
variant 'veatra' '
But such an assumption can hardly be taken seriously in account
having to many unproved assumptions in place to consider the direct
path : the presence of the /e/ in Common Romanian Time.
Next of course the origin of this e remains to be explained on his
side...that is not an easy task...
As on my side I saw a compound:
PIE *swai *a:tra 'his own fire' -> 'his sacred fire -> 'his home'
that fits perfectly the semantism of the word in Romanian and
Albanian and explained phonetically all the existing forms (with the
assumption that one of the Albanian dialects lost a length of the a:
*waia:tra > waiatra...all the other derivation Pre-Romanian *weatra,
Albannia *weotra/*weatra are regular)
This showing a PAlb/Dacian? form *waia:tra
If so, the Iranian form is a cognate.
I don't know any clear Iranian loan in Proto-Romanian and there are
very few candidates for PAlbanian
Marius