Re: [tied] Re: Question on PIE Root h2ep- 'to grab'

On 2006-08-18 00:43, alexandru_mg3 wrote:

> Viewing all these exceptions, could we imagine that finally this
> root was *h1eh2p- and 'a derived' form *h1h2ep- could be the source
> of Latin a in ap-i:scor, ap-tus as < PIE *h1h2p- (*h1h2p-tó- etc...
> (a nil-grade of *h1h2ep- )

How on earth can you get *h1h2ep- as a "derived" form of *h1eh2p-?
Laryngeals can't be paraded to and fro like this. Even those who believe
in "laryngeal metathesis" restrict its scope to the combinations *CHiC
and *CHuC > *CiHC, *CuHC, and only unidirectionally (NOT *CiHC > *CHiC).

Sequences like *HHC typically lost one of the larygeals (most likely the
middle one, it seems) already in PIE, hence *h2sté:r 'star' from
*h2ah-s- 'glow' and the loss of *h1 in the desiderative suffix *-h1s-
after some consonants (including laryngeals).

> Could also a PIE root *h1eh2p- to be valid for Hittite forms too as:
> epzi < *h1ep-ti

Why this and not *h1ah2p-ti (which would't end up as <epzi> in Hittite!)?

> and
> appanzi < *h1h2p-ónti

As far as I can see, *h1eh2p- solves no problems and creates new ones.
We would expect lots of *a:p- derivatives, for example, and I can see none.

Piotr