Re: [tied] PIE Word Formation Q&A (1)

On 2006-03-31 22:50, Sean Whalen wrote:

> As I said, Sihler's explanation is different from
> mine, but there's nothing arbitrary about this
> derivation. It's to explain why a syllable most
> evidence shows to be * wlh2 shows up as if from * h2ul
> in Hittite (according to most; I think hu represents
> xW or GW here).

Most evidence is indecisive either way, since initial *h2w- and *w-
merge almost everywhere. Greek can't be used as evidence if it's the
Greek development we want to establish. There remains Anatolian, where
both <hulana-> and <hulija-> point to a root with initial *h2. If other,
independent data allow us to account for the absence of a prothetic
vowel in Greek, metathesis is simply unnecessary.

Piotr