tgpedersen wrote:
> I like the proposed mechanism *Bw-i- > *B-i-, for B = labial stop.
> That might explain the variation *bay-/*baw- in the AfroAsiatic
> root.
Perhaps, provided that such a reconstruction is correct (few features of
Afroasiatic have been reconstructed with much confidence). You'd have to
ask a specialist in Afroasiatic.
> And BTW, I still think it would be nice to connect the *bh-w-
> (possibly *bh-y-) root with *bh-h2- "appear, come into this world".
The usual "master gloss" is 'shine', though of course 'appear' often
occurs as a related meaning. One can imagine all sorts of connections,
the problem is only that some of them may be hard to demonstrate.
> The semantic mesh nicely and it could be used to explain the
> otherwise unexplainable -ba- of Latin imperfectum.
I wouldn't say that it's otherwise unexplainable. Rix, for example,
explains it as extracted from the Proto-Italic reduplicated pluperfect
of *bHwah2- (his reconstruction of the full grade), i.e. *bHu-bHwah2- >
*fu-Ba:- 'had become (and continued to be...)', reanalysed as the
imperfect of *fu- (by then synonymous with *es-). One may disagree with
his explanation, but it is _an_ explanation, and one that has at least
some degree of plausibility. To come up with a better one you'd have to
explain just how a verb meaning 'appear' became compounded with other
verbs into a new grammatical form. How would you analyse a form like
<ama:bam>? 'I love-appear'?
Piotr