Re: [tied] Re: Gypsies again

On Sun, 02 Oct 2005 17:18:15 -0500, Patrick Ryan
<proto-language@...> wrote:

>----- Original Message -----
>From: "Miguel Carrasquer" <mcv@...>
>> >> 3sg. caká:ra. The /a:/ must come from /o/.
>> >>
>> >
>> >***
>> >Patrick:
>> >
>> >And why is that?
>> >
>> >Old Indian <a:> can come from PIE *e:, *a:, or *o: according to the
>> >current
>> >theory.
>>
>> And *o in an open syllable (Brugmann's law).
>>
>> >What disproves that Old Indian simply used vRddhi where non-Old Indian
>> >used
>> >/o/?
>>
>> The paradigm goes:
>>
>> cakára
>> cakártha
>> caká:ra
>> cakr.má
>> etc.
>>
>> Notice that the singular has */o/:
>>
>> *k(W)e-k(W)ór-h2a => cakára
>> *k(W)e-k(W)ór-th2a => cakártha
>> *k(W)e-k(W)ór-e => caká:ra
>>
>> No other vowel will produce an /a:/ in the 3sg. only.
>>
>
>***
>Patrick:
>
>I am under the impression that the conversion upon which you are relying
>calls for PIE *o to become OI <a:>. Or will you say that it happens only
>with *o in PIE-open syllables?

Actually, *o in Indo-Iranian open syllables. I do not know
for sure if e.g. the accusative *podm. had an open or a
closed first syllable in PIE (I suspect there was a shift
from asyllabic --Polish-style-- pronunciation of resonants
to syllabic --Czech-style-- pronunciation, and that it may
happened before the breakup of PIE). In any case, Skt.
pa:dam has /o/ > /a:/ in open syllable.

=======================
Miguel Carrasquer Vidal
mcv@...