On Sun, Mar 23, 2008 at 06:32:38PM +0200, Yitzhak Sapir wrote:
> I've thought about it some more and it seems to me that the form with qamats
> is to be explained as an archaic form displaced by analogy. While analogy
> is
> a normal linguistic process which displaces older forms, the older forms
> sometime survive in restricted contexts. One would still have to figure out
> the exact analogy, however.
>
> Yitzhak Sapir