--- In
cybalist@yahoogroups.com, Mate Kapoviã <mkapovic@...> wrote:
> The other possibility could be that *-n drops and that *-m doesn't.
Yes.
> It couldn't. *-o:n (> *-a:n) would give -oN, not -y. Same as *-a:n
< *-a:m
> gives -oN in A. sg of eh2-stems
And yields circumflex in Baltic and (according to Dybo) in Slavic
(vs. a regular acute in Acc. pl?)? Looks rather like a laryngeal-
deletion (*-ah2m > *-am), whatever be its origin.
And I'm not sure of this *-o:n > *-a:n thing (at least on the
phonetical level), by the way. It could well stay [o:n]. As I've
already argued, Proto-Slavic */an/, */am/ seem to be [on], [om]
phonetically.
Sergei