Jens:
> A collective is obvious for 'bone', and in the case of Gk. kardÃa:
> it is apparently accepted by the language itself.
Ugh, I couldn't help but notice the flip-flop. You denied the
existence of an *i-collective before when we were discussing *megxis
(or your **mego:xs). At any rate, I suppose it matters little if I've
concluded that the Anatolian "mutant-i adjectives" :) are calandized
denominals of thematic nouns instead of involving collectives.
= gLeN