From: dgkilday57
Message: 65355
Date: 2009-11-03
>All right, that clears the air. Under no conceivable circumstances could a Proto-Samic adstrate have taken PIE *treh2-gH- and funneled it into Germanic and Slavic as *dra(:)g-. The initial cluster should have tipped me off.
> > Pokorny's derivation of Gmc. *drenk- by nasalization of PIE *dHreg^- is not bad. I see however that his *dHera:gH- (whence 'draw' etc.) is only represented in Gmc. and Slavic, and /g/ in the latter makes it hard to associate with *dHreg^- anyway. He brings in *tragH- (Lat. <traho:>, <tra:gula>, etc.; "dissimilation of spirants" in Proto-Latin from *dHra(:)gH- is ad hoc).
> >
> > If your water-transport-word theory indeed holds water, perhaps we are dealing with a PIE root *treh2- extended by *gH (thus no root-restriction problem); this *treh2gH-, *tra:gH- (in IE dialects losing aspiration, *tra:g-) might have been borrowed into Proto-Lappic as *Dra(:)G-, then back into both Gmc. and Slavic. Gutenbrenner suggested a similar mechanism for getting 'boar' into Gmc. beside 'farrow' etc., *pork- or *park- being borrowed into the adstrate language as *BarG-, then into Gmc. Since I know little about Uralic, this appeal to "Proto-Lappic" to get voiced fricatives may not itself hold water, however. It is really just a wild guess.
> >
> > DGK
>
> This indeed is not quite watertight - if you mean Proto-Samic, these particular words do not exist, and neither do initial consonant clusters nor *B *G exist either. Moreover, there is no motivation to turn onset tenuis stops to voiced spirants - no Uralic language does that (the Samic distinction between *t *D is inherited). You could just as well go with an entirely unrelated substrate.
> What exactly is the "water transport theory" anyway? Torsten's latest renames it "amber trail theory", but I thought this was supposed to cover all IE-Uralic interactions, not just those near the Baltic Sea.I should probably let Torsten explain it (or refer it to earlier messages). I was intrigued by the possibility of elucidating apparent voiced variants of PIE roots containing tenues, but the number of these is small anyway. Uhlenbeck, if memory serves, explained 'boar' as an epithet 'frightening' having good Indo-Iranian cognates, so it may have nothing to do with substrate or adstrate. Gutenbrenner thought Celtic *gabro- 'goat' and *ab- 'water' could also be explained as back-loans from a voicing adstrate, but there are other ways of handling these as well.