Re: [tied] What is *IE for Soma?

14-12-03 06:02, wtsdv wrote:

> I don't think *po:tlo- is a proper reconstruction.
> Because although the verb *po:(i)-, and the suffix
> *-tro- are both reconstructed for P.I.E., I don't
> think that the complete *po:tro- is.

...

> Even if *po:tro- were a proper reconstruction to
> begin with, it still could never have meant anything
> like "brazier". The latter sense is applied only
> to the Japanese _compound_ "hi-bachi", not to "bachi"
> alone, not to the Chinese word from which "bachi"
> came, and not to the Sanskrit original. The original
> sense of "pa:tra-" was clearly "drinking vessel", from
> the Indo-Aryan reflex of *po:(i)- "to drink", with the
> Indo-Aryan reflex of the suffix *-tro, which forms the
> names of instruments. The entry you quoted had a link
> to another on *po:(i)-, appearing above as "p(i)-".
> Did you follow it?

*po:-tlom is OK. The instrumental suffix had four different variants:
*-dHlo-, *-tlo-, *-dHro- and *-tro-. Latin shows *-tlo- in <po:culum>
'drinking cup' < *po:-tlo-m, so that a PIE reconstruction like that is
justifiable (leaving aside the question of how the Brugmannian "long
diphthong base" *po:(i)-/*pi:- should be reconstructed properly).

Piotr