-s -> -i

> 1) -s > -i as in Italian (doi, trei)
>
> =======================
> Miguel Carrasquer Vidal
> mcv@...

A nice chance to push my shibboleth theory
1) General confusion between nom. and acc. pl. in provincial dialects.
2) East Rome (incl. Italy) opts for generalizing the nom: -i, -e
3) West Rome therefore opts for generalizing the acc. -s
4) because of this -i/-s shibboleth East Rome further generalizes -i
for -s everywhere (eg 2 sg)

Or we might argue Greek and Germanic influence, respectively. In any
case a development -s -> -i is not phonetically plausible.

Torsten