--- In cybalist@yahoogroups.com, "Tavi" <oalexandre@...> wrote:
>
> > The Celtic Urheimat, strictly speaking, is where the phonetic and
> > other changes characterizing Proto-Celtic actually took place. Among
> > other things, this includes lenition of */p/ to */f/ (Matasovic')
>
> This fricative would be more likely a voiceless bilabial /p\/ (IPA
> "phi").
>
> > I think this took place in the Iberian peninsula, and the area very
> > likely included [Austria].
>
> So I must presume you can demonstrate those OEH river names you
> mentioned before underwent the same phonetic development, can't you?
> Relative chronology is extremely important, you know.
>
Of course, I was referring to the dephonologization of
*p (
*pH in the glottalic model), which we also find e.g. in Germanic and Armenian, although with divergent phonetic developments.
There's also evidence of dephonologization of voiceless stops (not only /
p/) in some Paleo-IE dialects.
For example, we've got the western
*pºrt-u- 'passage, way' (Latin
portus, Germanic
*furT-, Celtic
*p\ritu-), which corresponds to the eastern
*bred- 'to wade, to jump' (Balto-Slavic, Albanian) ~
*brod-o- (Slavic), whose stops correspond to series III (traditional "voiced aspirated") and has the IE ablaut. As external cognates we've got Kartvelian
*bo(r)d- 'to wander, to roam' and Berber
*barid- 'road' (Dolgopolsky's ND 241).
IE-ists usually link
*pºrt-u- to
*per- 'to pass, to go through', which has the Afrasian cognate
*?\a-bir- 'travelling (across a road), passing by, crossing (rivers)' (Militarev). If he's right in associating PAA to the
Natufian culture in the Levant (roughly 13,000-9,800 BP), this would put a terminus ante quem for the dephonologization of voiceless stops.