Exu:
> It may have been zero-graded later, but Anatolian evidence (Hittite,
> Luwian, &c) shows it to have been a vowel when the Anatolians broke off
> from the earliest IE band.
Of course it was a vowel. You misunderstood what I was saying. Yes, *-i
is a vowel for everyone to see. It was certainly phonetically [i].
However, by way of analysis of IE phonetics, *i is said to be technically
a consonant and not a true vowel. Only *e and *o (maybe *a too) are true
vowels in IE and are what define full-grade. When we find *i, it is
typically a zero-graded *y from a fuller form with one of the true
vowels accompanying that consonant. So *wid- is the zero-grade of *weid-
'to know', for example.
> I have not seen any evidence of any kind of sound law *-Vy -> -i when
> used as a present-tense marker.
No, it doesn't exist. It is always zero-grade in IE itself, just as the
related demonstrative stem *i- appears to be. When speaking of Pre-IE
however, there may have been full-grade preforms. In fact, I'm betting
on it.
= gLeN