14-09-03 20:38, Sergejus Tarasovas wrote:
> But where's the evidence that interconsonantal *h2 yields Balto-
> Slavic *a rather than zero? Even such a laryngealistic purist as
> Beekes posits *nh2es- and *sh2el- (generalized accusative) here.
The 'nose' word is particularly problematic, since *h2 was basically an
obstruent and its inherent sonority was lower than that of a nasal. In
the sequence */nh2C/ I'd expect a syllabic nasal rather than a syllabic
laryngeal, i.e. *[n.h2s-] > BSl. *íns- or *úns-, yielding a Slavic nasal
vowel. But of course it's quite possible that morphological factors
(avoidance of samprasarana) overrode regular phonological development
and led to the appearance of *nas- as a secondary (non-inherited) weak
counterpart of *ná:s- < *nah2s-. At any rate, *nh2es- is an unacceptable
oddity, as far as I'm concerned.
Germanic has *nas- (as in OHG nasa, OE nasu) beside *nus- (as in OE nosu
[f.], presumably from *nuso:). Both forms are puzzling: the former may
perhaps be explained as in Slavic, whereas the latter looks either like
a metathetic transformation of an original nil grade (instead of
samprasarana *uns-?) or else an independent or contaminated formation
(cf. the verb root *neus-).
Piotr