From: Tavi
Message: 69395
Date: 2012-04-21
>Yesterday, I went to my local university library to consult one of the Adrados' volumes they keep there. He considers that some Greek prothetic vowels are reflexes of *h1,*h2,*h3, others are not, so he's also against the widespread pratice of reconstructing "laryngeals" from every Greek prothetic vowel. He has also developped an idiosyncratic system of "laryngeals" with apendixes which I don't need to explain here.
> > Do you have clear examples for which prothesis works better than a
> > laryngeal?
>
> To begin with, PIE "laryngeals" aren't real laryngeals, but rather
> *segments* corresponding to a consonant (not necessarily a laryngeal)
> plus an optional vowel support. For example, I myself discovered PIE
> *h3- corresponds to Hittite *s^a- in these words:
>
> Hittite *s^a:kuwa- 'eye' ~ PIE *h3ekW- 'eye'
> Hittite *s^ankuwai- 'nail; a unit of linear measure' ~ PIE *h3n(o)gh-
> 'nail'
>
> IMHO vowel support would be characteristic of *h2 and *h3 but not of
> *h1, so while *h2C- > Greek aC- and *h3C- > Greek *oC-, there's no point
> in positing *h1C-. This is why I think e- must be a genuine prothetic
> vowel.
>