Re: Cognates & etymology of English net

From: Piotr Gasiorowski
Message: 50689
Date: 2007-12-04

On 2007-12-03 23:05, alexandru_mg3 wrote:

> Romanian/Albanian nëpërkë <
> *Dacian *neperka: < [k < tk; n < sn] <
> *Early-Dacian *(s)ne-*pertka: <
> PIE *(s)nh1-e *pert-(i)k-eh2
> with the original meaning 'horned viper' too...

I have a few objections to this:

(1) Morphological: *(s)nh1- is the reduced form of a _verb root_
(meaning approximately 'spin, twist'). No snake, anywhere, is called
*(s)neh1-; Latin, Celtic and Germanic all have a *-tor-/*tr-ih2 agent noun.

(2) Phonological: Why do you posit *e as a connecting vowel in a
compound? Where are such compounds attested? Why wasn't this *e
diphthongised to /ja/? And why is the *n asyllabic in this position?

(3) Semantic: In IE endocentric compounds the second, not the first
element is the head, i.e. an adder with a horn would have been called a
"horn-adder", not an "adder-horn".

Piotr