Re[4]: [tied] Religious terms , here saint

At 1:00:17 AM on Wednesday, July 2, 2003, alex wrote:

> Brian M. Scott wrote:
>> At 1:33:12 PM on Sunday, June 29, 2003, alex wrote:

>>> Miguel Carrasquer wrote:

>>>> Avestan sp&nta-, Lithuanian s^ventas, Slavic sveNtU point
>>>> to PIE *k^wentos and nothing else.

>>>> Avestan Lithuanian Slavic
>>>> PIE *sw- xv- sv- sv-
>>>> PIE *k^w- sp- s^v- sv-

>>> an *swentos shouldn't make any trouble in satem space ,
>>> would it?

>> Yes. It's right in front of you.

> Oh. I bet you mean that *k^v > sp in Avestan as _one of
> the changes_ which can happen;In the same manner *sw > xv
> in Avestan is too just _one of several_ possibilities who
> can occure.

Actually, I was also looking at the Lithuanian evidence.
The overall fit with *k^w- is obvious: when the usual
correspondences match the data, one needs a very good reason
to look further. Besides, I'm not aware of any other
Avestan reflexes except the one in '6', /xs^v/; any others
must be pretty rare.

Brian