Re: [tied] Germanic KW

> One funny thing about PGmc. labiovelars is that they tend to lose
> their labiality next to root-internal apophonic *o, as in *gWolbHo- >
> *kalBa-, *dHogWHo- > *ðaGa- or *kWol(h1)so- > *xalsa-. That doesn't
> seem to work with *warma-, so perhaps it reflects *gWHermo- (as in
> Gk., Arm. and Alb.) with a coloured vowel.

There is something about the "neck" etymologies that worry me.

PGerm. *xalsa- "neck"
Lat. collus/collum "neck"
Lith. kãklas "neck"
(Gk. kúklos "wheel")
Estonian kael, -a "neck"

Why the o-grade?
Does the un-IE reduplication in the nominal stem *kWekWol- make *kWel-
non-IE?
How does the Estonian root fit in?


Torsten