On 2009-05-31 14:46, alexandru_mg3 wrote:
> ----------------------------------------------------------
> > There is no "root *menh2-/*mneh2-".
>
> Did you here about CERH-/CREH- roots?
The whole Schwebeablaut idea was demolished by Raimo Anttila in his 1969
book. A PIE root may have either form, but not both of them at one time.
We have *g^neh3- 'know' and *telh2- 'support', but no *g^enh3- or
*tleh2-. Even an "extended root" has only one vowel slot, not two. We
may have late analogical forms or vr.ddhied derivatives with *e infixed
in the "wrong" place (the *deiwo- type), but <moneo:> is neither of
these. It's generally regarded as the causative of *men-, and what could
be wrong with that? 'To make sb. think' --> 'to instruct sb.' is a
completely natural development.
Piotr