On 2008-11-06 18:56, etherman23 wrote:
> The case for *h1 before labiovelars isn't looking too good either. I
> found two roots. *k'eh1kW, meaning hay. It occurs in Old Indian,
> Baltic, and Germanic. However, it is well represented in the Baltic
> and Germanic families. *yeh1gW meaning power. It occurs only in Greek
> and Baltic (Lithuanian and Lettish).
I was only looking at verb roots. Of course in derivatives, compounds
and inflected words all sorts of clusters may arise at morpheme
junctures. *-h2t- may be rare or nonexistent morpheme-finally, but of
course we have 3 sg. aor. *stáh2-t, 2sg. perf. *ste-stóh2-th2a, 3sg.
pres. *páh2-ti 'guards', *bHráh2-t(e)r- 'brother', etc. Likewise,
*-h3kW- (the zero grade of *h3okW-) is frequent in compounds like
*proti-h3kW-o- or *h2anti-h3kW-o-, though at the end of a morpheme it
would violate both constraints I have proposed (no *h3 before stops, no
labiovelars after laryngeals).
Piotr