Re: Hercynian (again)

2012/3/1, Tavi <oalexandre@...>:

> loans aren't the only trace of substrates, as there's also
> toponymy. And quite often toponymy roots become loans (e.g. Latin aqua
> 'water'). In a multi-layer model I've got no problem to consider items
> like this as IE, although certainly not "PIE" in the traditional sense.

———————————————————————————————————
OK, but You'll agree that PIE etymologies (in the sense of Celtic
evolution in situ of PIE names) are nevertheless ALWAYS possible in
full agreement with every phonological, lexical, and morphological
criterion
———————————————————————————————————

>
> But there're other toponymy roots which can't considered as IE at all,
> e.g the ones represented by Basque naba, (> Spanish nava) 'plain
> (between mountains)' and Gascon gava (> French gave) 'river', whose
> nearest cognates are respectively Kartvelian *neb- 'palm of the hand'
> and *qew- 'ravine'.

———————————————————————————————————
*naua < Celtic *nāuā (Hubschmid!) < PIE *nah2/4us
and
*gaua < Celtic *gāuā < PIE *g'hōu-ah2/4 (my ertymology)
are powerful explanations, I think.

Kartvelian OK, but PIE through Celtic still better!