Re: IE *p and *b

From: etherman23
Message: 50777
Date: 2007-12-09

--- In cybalist@yahoogroups.com, "Patrick Ryan" <proto-language@...>
wrote:
>
> My own view is far simpler.
>
> I believe that stops had four manifestations in each articulatory
position (labial, dental, palatovelar): glottalized, glottalized
affricate, aspirated, aspirated affricate, with the following PIE
equivalents: t? > d; t?s > dz > dh; th > t; ts > th, merging with t.

I'm not sure how this is far simpler. We both posit 4 series for each
position, but two of yours merge in PIE. I posit the series of
voiceless aspirates to account for examples of correspondences that
are hard, or impossible, to explain with 3 series.

> One would expect p? to have simply become PIE b but instead, some b
> bh; and other b > w with a small number of b remaining b.

What are the conditions leading to this 3-way split?

> The evidence I have for b > w comes from cognates in what I consider
to be related languages: PIE *wi/i:, apart; Egyptian pj, *divide;
Sumerian be-4, divide, allot.

Is it fair to say that *b > *w before *j? Would we expect to see
variants like *wi v.s. *bei or *bHei?