Re: [tied] Pretonic laryngeals in roots

From: Piotr Gasiorowski
Message: 47882
Date: 2007-03-15

On 2007-03-15 21:07, tgpedersen wrote:

> I may be wrong, but it seems to me that the posited laryngeals in
> roots of the form *HCVC occurs in only two functions: as *He-, in
> which case they appear as the 'intended vowel' (which is the one found
> in Greek and Armenian), and *H-, in which case they disappear (in the
> other languages). Would it be possible to posit a vowel instead, with
> suitable rules to achieve the same result?
>
>
> Eg. *édons, *déntm etc

But they also do other things; most notably, they lengthen (and/or
otherwise modify) preceding syllabics in prefixations and compounds.
Thus, *h1 is required in *h1ei- and *h1es- because of forms like
3pl.ipf. *é h1j-ent > Ved. a:yan and *n.-h1s-n.t- > Ved. a:sat-, and *h3
is required in *h3kW-, the zero grade of *h3okW-, because of, among
others, *proti-h3kW-o-m > pre-Gk. *protjo:kWom > Gk. proso:pon 'face',
cf. Skt. práti:ka-.

Piotr