Richard:
> I think Miguel's analysis would work here - an original contrast
> of /-s/ : /-tW/ : /-t/. Would the evidence for *Nt be satisfied by
> a rule converting final stops to nasals?
Assumption ignoring reality. There are no suffixes with labialized
phonemes at the end in IE. No **-hW but we have *-x as the collective
and *-ex as the later feminine, even instrumental *-eh with *h1. We
never see **-kW either. There's a pattern here and frankly it makes
sense since labialized phonemes are logically more difficult to
distinguish from plain ones in final position. So **-tW? I don't think
so. Not likely in phonetics.
= gLeN