--- Jens Elmegård Rasmussen <
elme@...> wrote:
>
> A very old theory put forward by Alois Walde has it
> that there is a
> Grassmann-type dissimilation in Latin if a liquid
> intervenes. That
> takes care of examples like barba, glaber, gradior
> and a few others I
> don't remember at the moment.
It doesn't seem necessary to me. Partially the
problem comes from initial *(ghR, g^hR, gWhR) behaving
differently from each other. Deaspiration of some Ch
next to a sonorant or glide is already needed for
Latin.
NgH>Ng etc.
ghR>gR initially
dhR>dR initially
dr>tr
C > -palatalized
aspirate > -voice
aspirate-voice > fricative
xW>f
x>h
h>0 before sonorant
So:
*g^hreu- > ruo: (and *en- > ingruo:)
*g^hreud- > ru:dus
*g^hroH3wo- > ra:vus
*gWhrendh- > frendo:
*ghladhro- > glaber
*ghredh- > gradior (a < ana. with *dhragh-)
*dhragh- > traho:
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com