Piotr:
> Then why do we have the open-syllable (Brugmannian) lengthening of *o in
> acc. pa:dam < *podm., nom.pl. pa:das < *podes, and in nom.pl. a:pas <
> *h2opes ?
First off, I'm not talking about IE itself. I'm talking about eLIE, an
earlier form. I'm going to think some more on this but I'm thinking so
far that maybe this isn't so much about syllabics but rather vocalism
that's based on the syllabics of the _root_ of a given word.
So, for example, with *pedos, it may indeed be CV.CVC but the root here
is *ped-. This must have somehow affected how the pretonic *a in earlier
*pat:ása developped such that it became *e as if in a closed syllable
(because the _root_ of this word is a closed syllable *ped-). However,
in the reduplicated present *bHi-bHer-, the pretonic *a of an earlier
*bHa-bHer- does not lie in the "root syllable" (which would be *-bHer-)
and is therefore treated as an open syllable, becoming *i instead of *e.
Similarly, the open syllable at the end of thematic stems would yield
pretonic *i as well, given birth to the *i-stems.
In all of these examples, the rules are such that we would never ever
expect **pid- instead of *ped-, **bHebHer- instead of *bHibHer-, or
**e-stems instead of *i-stems.
= gLeN