On Sat, 8 Mar 2008 16:43:51 -0800 (PST), Rick McCallister
<
gabaroo6958@...> wrote:
>I've seen that but wondered about it
>Was the *s- reanalyzed as s-mobile?
In Latin, *s is regularly lost before *m (and before most
voiced consonants in general).
I wonder if it's necessary to reconstruct a zero-grade in
*smih2 g^hslih2 > smi:ksli: > mi:lli(:) > mi:lle. Wouldn't
*smih2 g^heslih2 > smi:hesli:, with early loss of /h/ and
contraction if i:e(:) > i: also work?
=======================
Miguel Carrasquer Vidal
miguelc@...