Re: [tied] Albanian N in CC-clusters (review)

From: Piotr Gasiorowski
Message: 42535
Date: 2005-12-16

alexandru_mg3 wrote:

> gn > nj (njoh < PAlb gna:-ska < PIE *g'neh1?-sko)

The laryngeal was *h3. Alb. /o/ is either (less likely) due to the
special development of *-n.H- > *-na:- in *g^n.h3-sk^é/ó- (but I know of
no other examples in Albanian) or (much more likely) reflects the
long-vowel grade *g^ne:h3-, with no laryngeal colouring, attested also
in Germanic (OE cna:wan), Hittite (ganess-) and Tocharian A (kña-). PIE
*g^noh3- may well have been a Narten root, hence the vocalism of Lat.
gno:sco:, OPer. xs^na:sa:tiy-, which point to an original
*g^nóh3-sk^e/o-, and the long *e: in pre-Albanian could have been
introduced analogically from other forms of the same root.

Piotr