Balto-Slavic *intonations* don't, as far as
I'm aware, demonstrate laryngeals in the dual endings. If you believe otherwise,
show me the evidence. Long vowels do not always result from compensatory
lengthening. Why not contraction, e.g. *-o: < *-o-e?
In Greek duals there is so much analogy at
play (cf. N.du. no:e, spho:e with -e borrowed from consonantal stems) that it's
hard to believe that these particular forme should be very archaic while
everything else (logoin, etc., with no hiatus) is secondary.
The Kartvelian form doesn't prove much
about PIE. Anyway, if *ok^to: 'eight' is a dual, the singular should have been
**ok^tos 'four'. The assumption that the Proto-Kartvelians borrowed a dual form
with the meaning of the corresponding singular is odd and had
better be very carefully justified.
As for getting rid of *h3, I'll give you my
reasons; just allow me a little time.
Piotr
----- Original Message -----
Sent: Monday, November 13, 2000 3:02 AM
Subject: Re: [tied] Catching up again...
Kartvelian *os'txw- for one. All those long vowels. The
hiatus in
Greek no:ï(n), spho:ï(n). The Balto-Slavic
accent.
...
How would you reconstruct *deh3- "to give" then? And can you
expand
in general on the doubtfulness of *h3?