>There is a contradiction of sorts if we stick to the claim that original
>*o is treated differently from *H3e. For these are the only examples I
>have seen - certainly this time around - of *H3e. If the words mentioned
>are disqualified as evidence this risks being a law formulated without any
>examples for it to work on.
It's not exactly a contradiction, but it is a problem. I hadn't actually
checked how many examples there are in Skt of PIE non-ablauting *-o- in open
syllables. But I'm prepared to believe that are not many! If you could
dispose of all of them, then Brugmann's Law falls. I suspect, however, that
we will be left with a handful where the -o- syllable is either undeniably
open, or not provably closed. Because the number is small, ad hoc
explanations may be possible - but that wouldn't make them right.
Have you also considered from *h3ekw, the Vedic an-ak blind?
Peter