Re: [tied] Slavic G. pl. (was: Re: Nominative Loss. A strengthened

From: Mate Kapovic
Message: 32271
Date: 2004-04-24

----- Original Message -----
From: "elmeras2000" <jer@...>
To: <cybalist@yahoogroups.com>
Sent: Saturday, April 24, 2004 2:45 AM
Subject: [tied] Re: Nominative Loss. A strengthened theory?


> --- In cybalist@yahoogroups.com, "Mate Kapovic" <mkapovic@...>
> wrote:
>
> > I can't see how could *-oom or *-o:m yield Slavic -7. And we know
> for sure
> > that *-om does.
>
> Well, this is supposed to be an empirical science. If there are no
> other cases of *-oom or *-o:m in Slavic, we really cannot know for
> sure what becomes of such a protoform if there was one.

Oh, we do. We have -oN in 1. sg present from PIE *-oh2-m > *-o:m.

>The fact is
> that there is no other evidence for a short vowel in the genitive
> plural in Indo-European.

Slavic (-7 is *very* short, it points to *short* origin, not long), Latin,
Old Irish, Hittite (-an in both A. sg and g. pl. in Old Hittite). Is that
enough? The reasonable thing to assume is that *-o-om > *-o:m originated in
o-stems and that other stems had *-om. Later, some languages generalized the
long variant, others the short one.

>Many languages have a shortening before
> final /-m/, so what is so terrible with a reflex of *-oom looking
> like that of *-om?

Because in Slavic *-om surely yields -7 and /7/ is a *reduced* vowel, it is
not very likely it stems from a long vowel. And in similar cases we have
nasal vowels from V: + -m# like -oN in 1sg present and -oN in Asg eh2-stems
< *-a:m.

>And is there absolutely no indication of a larger
> amount of compensatory lengthening in Slavic before the reduced
> gen.pl. ending than before that of the acc.sg.?

Yes, but it is because the situation Nsg *bra´´t7 - Asg *bra´´t7 - Gpl
*bra´´t7 was dealt differently. Asg had a tendency to take -a from Gsg and
gpl to change the intonation by imitating a. p. c Nsg vo^rt7, Gpl vo~rt7 <
*vort'7.
As for compensatory lengthening, it is not so compelling evidence. Firstly,
in Slovene and Kajkavian neocircumflex doesn't always become from comp.
lenthening. Secondly, Czech and Shtokavian don't have comp. lengthening
anywhere else so there is no need to claim that neocircumflex there must be
because of the old length in the next syllable. Thirdly, the idea of the
long jer is absurd. Jers are reduced vowels, they cannot be long. And if
they could be long, how come they were dropped anyway in the end? And no,
they were not reflected as -a: in Shtokavian. Written attestations clearly
show this -a: is an innovation.

Mate