28-04-2004 18:02, Miguel Carrasquer wrote:
> The genitive of -o:i stems is *-yos, not *-eis, and we would
> expect a Skt. paradigm *áva:, *áva:yam, *avyás (or *ávyur,
> like sákhyur), none of which are actually found in Skt. vé:s
> (vís), vím, vé:s, so I'm afraid such a reconstruction does
> not even yield a point of departure for analogy.
Right. What about a minimalist solution: pre-IIr. *h2wí-s, acc. *h2wí-m,
gen. *h2wéi-s, pl. *h2wéj-es, etc. declined like (and no stranger than)
*mn.tí-? For purely phonological reasons *h2wi- happens to be
monosyllabic (but cf. *trejes from virtual **tri-s). Of course it isn't
the kind of paradigm that could date back to the most archaic stages of
PIE, but it's real enough, and the only problem is the aberrant variant
of the nom.sg. As the word was a was monosyllabic and therefore unusual
i-stem, perhaps dat. *vay-e, voc. *ve, pl. vay-as were enough to
establish //vay-// as a pseudo-root noun, hence //váy-s// --> ves. The
paradigm has been preserved incompletely; perhaps there was a variant
gen.sg. *v(i)yas as well. Sorry, I can't think of anything better :-)
Skt. also has a neuter *es-stem, <vayas-> 'bird, (small) fowl'.
>>--> nom.pl. *h2aweyes > *aweyes > PToch. aw'&y& > pre-Toch.B ay&y& > awi
>
>
> The original nom.pl. form must have had /o/
> (*h2oweies/*h2owoies), so I think it's more likely that /a/
> comes from the (pl.) oblique, *h2aw-i-bhios etc.
The arrow "-->" was intended to suggest that pre-Toch. *h2aweyes had the
same analogical origin as *h2awis. The two just co-evolved together.
Piotr