16-07-03 20:25, Piotr Gasiorowski wrote:
> Since both Germanic and Sanskrit seem to have the "mainstream" plural as
> well, and there's no support for *-o:ses from other quarters, I think
> independent innovation is more likely than the other possibilities.
Oops, silly me. I forgot Av. -a: ~ -a:Nho: (as in <ahura:Nho:>) < *-a:s
~ *-a:sas, which pushes *-a:sas back to Proto-Indo-Iranian and makes a
connection with Germanic more believable (especially in view of the fact
that Germanic and Indo-Iranian are the chief witnesses for nom.pl.
*-o-es). If so, the best course is perhaps to assume for the ancestor of
Germanic the same set of o-stem plurals and pluralses: *-o-es in
competition with *-o-es-es.
Piotr