PIE Vowels Long By Nature

Piotr,

This is a very interesting topic to me.

Hittite confirmed the existence of a phone roughly corresponding to
Saussure's 'phonetic coefficient.

If we find an <h> in Hittite, I guess we generally can consider the question
of its presence in PIE in a given word pretty much closed. Perhaps not
exactly vice versa, though.

Given any word of the surface form *CV:(C), what criteria can we use to
determine whether ot stems from **CV:(C) or *CVH(C)?



Patrick