On 2008-02-10 22:57, fournet.arnaud wrote:
>
> PIE *urughyo- is _not_ a PIE root!
> PR
> ============
> Pokorny : p. 1183
> It is a PIE Root *wrugh
> Variant form wrigh- (Thrakian)
*wrugHjo- is a ghost root. All that is warranted by Germanic and
Balto-Slavic is *rugHi- ~ *rugHjo-. Germanic, in particular, shows no
evidence of *wr-, which is a strong argument against an initial *w in
this root. Thrac. (?) briza is a poor match and one can't simply
_assume_ that it's a cognate of *rugHjo- just because it may mean 'rye'.
How can you rule out the possibility that the initial <b-> reflects *bH,
or that the <-z-> comes something entirely different from *-gHj-? Even a
connection with *bHr.h1g^-o- 'white' is formally more plausible.
Piotr