Re: [tied] Animate Dual in -h3 (was: IE Roots)

> >The retention of the sequence -u: #a- in the dual in Sanskrit, but not

in

> >other forms, where it goes to -va-, seems like evidence to me.
>
> I think we're talking about two different things. How does long /u:/

derive

> from *-o:u (or the unlikely Miguelian version **-oh3u)?

*-o:u is not established (as far as I know) for the dual, whereas the -H
ending has been suggested for quite a long time, and seems to have wider
acceptance. Szemerenyeni, who hates laryngeals, reconstructs a long vowel.
Beekes suggests -h1, and has been arguing for a laryngeal here at least
since 1969. The -u in Skt may be a feature that arose within that
language, as it did in other places where a laryngeal was present.

Peter