Re: Hat

> > But PIE dental + dental > PGerm. -ss- too, without exception
>
> Not at all. I gave a long list of words with T+t > st after r()
> and other conditions fairly recently.

Where?

> It's also possible that *kaxdh+ >> *kx,dhtu+s > *xattu+z if there
> was no t>s after a fricative and if tt didn't become a interdental
> fric. gemin. later, depending on the rules. Otherwise an affix with
> a nasal (*-nu-?) with dhn > dn > tn > tt would be possible.

*kx,dhtu-s?? *-dhn > -tt- ??


What I think happened was
1) stop + stop > fricative + stop, *-tt- > *-þt-
2) þ > s, ð > z,
3) *-V´st-s > *-V´ss, *-VstV´- stays (cf. Lat os(s), Gk. osteon),
4) Languages with initial stress generalize *-V´ss-, the rest *-VstV´-

But *katt-/*kass- is substrate flotsam, and follows and breaks several
rules.


Torsten