Re: Re[4]: [tied] Latin -idus as from dH- too

On Mon, 17 Mar 2008 15:00:27 -0400, "Brian M. Scott"
<BMScott@...> wrote:

>Oh? My understanding is that Hittite /h/ is far more likely
>to represent *h2 than *h3, so that /ah/ suggests *ah2 rather
>than *oh3. However, I freely admit that I have *not* read
>Melchert, so I'm going only by what I've picked here and
>from Beekes.

Melchert has *poh3s- > pass- "swallow" and *pah2s- >
pahs(s)- "protect" (p. 77). Hittite wouldn't show the
difference between /oh3/ and /ah2/, but *h3 has a tendency
to disappear more often than *h2, as in this case.


=======================
Miguel Carrasquer Vidal
miguelc@...