On 2006-08-30 18:12, tgpedersen wrote:
> And therefore both vowels of Ce-CoC- *must* have survived the
> ablaut formation process? I'm a nice person, so I won't call
> other people's opinion 'preconceived'.
They _may_ have survived for that reason, but of course there's nothing
inevitable about it, as I have shown myself (in *g^í-g^n-e-ti the vowel
of the reduplication syllable has been reduced to *i and the root vowel
has been deleted). In the perfect, the *o survives under accent (i.e. in
the singular), otherwise it disappears. The vowel of the reduplication
syllable is *e in Greek and mostly a < *e in Sanskrit (the i- and u-
reduplications before roots containing the corresponding glides are
usually thought to be innovations). One possible explanation is that the
pattern established after the working of the PIE ablaut rules was
*gWHi-gWHón-h2a/*gWHé-gWHn.-me, and the variation of the reduplication
vowel was levelled out.
Piotr