Re: [tied] i-verbs in Baltic and Slavic

On Sri, svibanj 24, 2006 1:01 am, Miguel Carrasquer reče:

> On Tue, 23 May 2006 22:32:23 +0200, Piotr Gasiorowski
> <gpiotr@...> wrote:
>
>>On 2006-05-23 16:14, Rick McCallister wrote:
>>
>>> Please tell me what the link is between Balto-Slavic, Albanian and the
>>> extinct Balkan IE languages (Dacian, Thracian, Illyrian). I've seen
>>> allusions to this. Is it just speculation or can it be backed up?
>>
>>Balto-Slavic and Albanian are both Satem languages, so _if_ the Satem
>>development of the *k^ series is a homologous feature derived from a
>>common ancestor, they are closer to each other phylogenetically than
>>either is to, say, Celtic, Tocharian or Greek.
>
> Are there any traces whatsoever of Winter's law in Albanian?
> The other day I discovered a major error in my thinking
> about the Balto-Slavic merger of /a/ and /o/, which I had
> assumed to be Proto-Balto-Slavic or even pre-Balto-Slavic,
> possibly Proto-Balto-Slavo-Albano-Indo-Irano-Germanic :-)
>
> However, Winter's law gives /ad/ > /a:d/ but /od/ > /o:d/
> (with some cases of /a:d/ as well), which implies that /a/
> and /o/ were still distinct when the law started to work.
>
> This can mean either that the merger of /a/ and /o/ is
> younger than I thought, or, less likely, that Winter's law
> is older than I thought, hence my question above.

The classical examples are Lith. núogas but obuoly~s.
I think that *o > *a must be very old (cf. also Germanic *o > *a), but
that Winter's Law is even older, thus very early. Since Winter's Law must
operated between *cH > *C as well, I would say that that makes it a pretty
old development. Of course, one cannot be sure.
However, I think Kortlandt believes *o > *a to be a late change, carried
out independently in Baltic and Slavic. I find that hard to believe.

Mate