Re: [tied] Re: Risoe fo the Feminine

On Thu, 29 Apr 2004 13:24:26 +0200, Piotr Gasiorowski
<piotr.gasiorowski@...> wrote:

>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.

I can't believe that.

>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.)

The plural is not really a problem: it is the same (-i-bhyas,
-i-bhis, -i-su) for _all_ i-stems, whether originally
*-ey-/*-oy- stems, whether making their genitive in *-eys or
*-yos. And note Avestan G.pl. vaya,m, with oblique stem
vay-.

The oblique forms of the singular are mostly consistent with
an oblique styem *h2wéih1- (G. vé:s (*h2wéih1-s), D. vaye:
(*h2weih1-ei > vayye: > vaye:), I. vyá: *(*h2weih-ét >
*h2wih1éh1 > v(i)yá:)).

Nom.sg. vís is analogical. Nom. vé:s (RV only) is the
original form (cf. also Middle Persian va:y "bird").

The acc.sg. _is_ a problem. I would expect relics of an
earlier *váyam or perhaps *ví:m (with reduction of *óih1C to
*íh1C). Instead, there is only vím.

>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.

But that is not (assuredly) the case in the other two
examples I gave. Moreover, *pih1werih2 (as shown by -r-, not
-n-), is derived from the neuter *<píh1wr.>.

>*h2woih1s/*h2weih1- ~ *h2weit- (not actually
>attested) --> *h2wjetos? An acrobatic derivation (and why isn't the
>syllabification resolved as *h2wijeto after an initial cluster?).

Good question. I'd have to think about that.

What about the Albanian forms? IEW gives vido, vito, vidheze
"Taube". EIEC gives vida "dove". I'm curious about the
dental. vito < *h2wéit-? (Leiden is down again, so I can't
check Demiraj).


=======================
Miguel Carrasquer Vidal
mcv@...