Re: [tied] Re: Nominative Loss. A strengthened theory?

From: Miguel Carrasquer
Message: 32133
Date: 2004-04-21

On Wed, 21 Apr 2004 16:16:19 +0200, Sergejus Tarasovas
<S.Tarasovas@...> wrote:

>> From: Miguel Carrasquer [mailto:mcv@...]
>
>> >(Sl.) *-ah2as > (-s analogicaly replaced by -x, as is often
>> the case in
>> >Sl.) *-a::x > ("raising of superlongs") *-o:x (/x#/ [h#]) > (Slavic
>> >raising before [-h#], otherwise *u < *o: would be expected)
>> *-u:(h) >
>> >*-y.
>> >(Lith.) *-ah2as > *-a::s > (no raising) -ãs (no * since that
>> /a:/ survives
>> >in the dialects) > (Standard Lith.) -õs
>>
>> Ingenious.
>
>Thanks.
>
>> >There seems to be no raising after *j in Sl. (*-ja::x >
>> *-ja:x [-jæ:h]
>> >> *-je^ ~ *-jeN, the nasalized variant being analogically introduced
>> >from
>> >accusative?)
>>
>> This would divorce the soft ending -jeN from the hard one
>> (-y), except for West Slavic -je^. But since the acc.pl.
>> shows the same distribution -je^/-jeN, and the nasal is
>> inescapable there, I still prefer the old fashioned
>> explanation that -a:(:)s mutated to *-a:ns generally in Slavic.
>
>... except for West *and East Slavic* -je^ (Standard ORuss -je^, and
>Krivichian even has that -e^ in the hard declension (<ruka> N.sg. ~ <ruke^>
>(sic!) G.sg.), so there's no -y at all there!), so it would probably be more
>appropriate to write "except for South Slavic -jeN". ;)
>Of course I realize that my analogical explanation is weak and unverifiable.
>
>> >Unfortunately, I've got no idea what to do with Sl.
>> *a:-stems Acc. pl.
>> >*-ah2ms > *-á:ms > ... > *y. Even if u:N(s) > *-y is
>> regular, a double
>> >raising "before nasal" and "before s" somehow doesn't make me happy,
>> >and again *j blocks the rasing (*-já:ms > *[-jæ:] ~ *[-jæN]
>> > *e^ ~ *eN
>> >).
>>
>> I don't know. I'm happy enough with the double raising. We
>> have something of the same thing in the dat.sg./ins.pl.,
>> where *-o:i(-) is first raised to *-u(:)i(-) (> *uo in
>> Slavic), and *uo is raised again to *u: before -h in the ins.pl.
>
>Yes, except that one wouldn't expect *s in *-á:ms to be replaced with *x (or
>do you postulate any *s# yields [-h#]?).

Let's see. *-eh2-ms > *-á:ns (acute, not circumflex). If
that first became *-á:Ns (nasal vowel), then *-s after vowel
can become *-h (*-á:Nh). Raising twice, we get -u: > -y.

Acc.sg. is then *-ah2m > -á:m (2)> -ó:m > -om > -oN.
Gen.pl. *-ah2om > -ã:m (1)> -õ:m (2)> -u~:m > -um > -U
Acc.pl. *-ah2ms > -á:ns (2)> -o:Nh (3)> -u:Nh > -u: > -y

(1) circumflex rising
(2) nasal rising
(3) aspirate rising

That works.

>And was it *-o:i(-), not
>*-o::i(-)? And if it was, why do we have circumflex in Lith. -ui (showing no
>Saussure-Leskien; BTW, the Z^emaitian dialects have -ô.u (broken tone),
>which points to *-o: (probably acuted, but I'm not sure), not *-o:i!) and
>-ai~s? Jens is wrong about V: > acute?

The dat.sg. and ins.pl. have circumflex (also in Greek, I
think), being contractions *-o-ei > *-o~i and *-oy-is(?) >
*-o~is. But I wouldn't expect the development of circumflex
õ (o::) before -i to be the same as that of plain -o::. I
have no problem with the Lith. developments -ui~ or -õ: (if
Z^emaitian points to -ó:, I would understand that less
well). Slavic probably also had *o:, which like *e:
subsequently developed into broken diphtongs *úo (ô) *íe
(ê).


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