28-06-03 12:25, tgpedersen wrote:
> We've seen some examples of these alternations in the past on
> cybalist. How frequent are they in IE?
What we find -- not very frequently but _regularly_ in several branches
-- is the realisation of the zero grade of roots beginning with *mer- or
*mel- as /br-/ and /bl-/ (examples of *ner- producing /dr-/ could also
be cited). It's a phonetic process, possibly (though not necessarily)
involving *mbr-, *mbl-, *ndr- as the intermediate stage (the insertion
of a stop is very common word-medially; cf. English thunder with -ndr- <
-nr-, or thimble with -mbl- < -ml-). Clusters consisting of two
sonorants are universally less acceptable than "obstruent plus sonorant"
ones, especially in onset positions.
Piotr