--- In
cybalist@yahoogroups.com, Miguel Carrasquer <mcv@...> wrote:
> I have mentioned:
>
> Celtic *swi: 2pl.
> Greek spho: 2du.
> Hitt. sume:s 2pl.
> Alb u < *swey- (?) 2pl.
> Avest. x^sma 2pl.
>
>
> What more evidence do you want?
Something actually demanding *sw-. None of the clusters presented by
Greek or Avestan can reflect *sw-; we know what that yields and do
not find it. Since Celtic 1pl *sni: does not occasion you to posit
IE *sn- I cannot see how 2pl *swi: can be styled as solid evidence
for *sw-. And even if Hitt. sume:s does come from something
beginning with *swe- (which I am far from sure it does), that
preform does not have to be pushed back into the protolanguage.
Jens