>Not regularly, except of course when the rhotic is syllabic, as in
><mr.tá->. As for Indo-Iranian *mr- > Skt. br-, see Avestan mru:- / Skt.
>bru:- 'speak' (this, to be sure, is PIE *l, which hardly makes a
>difference). I think *mrég^H-mo:n is formally unassailable (or I wouldn't
>have suggested it); the question is if the semantics is plausible.
The change mr>br is attested in different languages and it appears to be a
very usual resolution of this group.
We could find a lot of examples for this change in different languages of
different families, and not only as a regular change but also as a sporadic
dissimilation in some words. I think for instance in sp.
hominem>homne>hombre (if that's correct). Well, anyway, we could find
clearer examples.
Like Piotr, I just think we should consider semantics, but it seems pretty
plausible having in mind the resolutions of this root in languages like
greek and latin.
ana
_________________________________________________________________
MSN - More useful every day
http://www.msn.de