Re: [tied] Re: Risoe fo the Feminine

29-04-2004 12:31, Miguel Carrasquer wrote:

>> ...If <suwais> is
>>related to the whole set, the relationship need not be straightforward.
>>It may be a different paradigm, possibly from an original collective
>>*(s)h2wó:is (hypothetically = 'fowl').
>
> But, but for the length mark, there you have Skt. nom. vé:s.

A coincidence, in my opinion. If you start with *h2wóih1-s, you can
account for <ves>, but what about the acc.sg. vim and the nom.sg. vis
(if it isn't the original form)? What about <víbHis> and <víbHyas>? What
are they all analogical to, if the weak stem was *h2wéih1-? (The gen.pl.
vi:na:m of course doesn't show a laryngeal in the stem, since -i:na:m is
found in all Skt. -i- nouns.) On the other hand, if you start with
pre-IIr. *h2wi-/*h2wei- (i.e., an ordinary *-i- stem), then <ves> is the
only difficult form. But at least there is a possible analogical basis
for it, though I agree that it isn't very solid.

>>Complicating the ablaut. Now you introduce another mobile vowel between
>>the *y and the final *t/*h1. How did it get there if the original form
>>was your **(s)xawá:yt?
>
>
> It may be a "vrddhi" vowel of the type *nu -> *n/e/w-os,
> *dy(e)w -> d/e/yw-os. Cf. also the vocalism of
> *ih2-derivatives (p(e)íhwr -> pih1w/e/r-ih2, etc.)

*pih-wer-ih2 has *pih-wo:n as its counterpart, so the *e occupies a
pre-existing vocalic slot. *h2woih1s/*h2weih1- ~ *h2weit- (not actually
attested) --> *h2wjetos? An acrobatic derivation (and why isn't the
syllabification resolved as *h2wijeto after an initial cluster?).

Piotr