Re: [tied] 1sg. -o: [was Re: IE Thematic Vowel Rule]

From: Miguel Carrasquer
Message: 39874
Date: 2005-09-03

On Thu, 01 Sep 2005 20:33:07 -0500, Patrick Ryan
<proto-language@...> wrote:

>From: "Miguel Carrasquer" <mcv@...>
>
>> The -m of the thematic neuter is simply the accusative
>> ending. Since thematic (i.e. definite) nouns and adjectives
>> are higher on the animacy scale than plain neuters, they
>> have a marked accusative, like the neuter pronouns, which
>> are even higher on the animacy scale. Neuters were not
>> allowed as subjects, so they lacked a nominative.
>
>
>***
>Patrick:
>
>So what, then, is the case of a neuter in a nominal sentence? e.g. 'the yoke
>is red'.


I should have said: "neuters were not allowed as subjects of
transitive sentences."

Still, that leaves the -om of neuter thematics in
intransitive sentences unexplained.

What can be said about the accusative morpheme *-m?

In the personal pronouns, it appears as a
genitive/accusative under the shape *-mé/*-wé:

1sg. Acc. *mé < *mmé (or *mwé), Gen. *méne < *méme
2sg. Acc. *twé, Gen. *téwe
3sg.refl. Acc. *swé, Gen. *séwe
1pl. obl. *n.smé
2pl. obl. *usmé
3pl.R. obl. *smé
1du. obl. *n.h3wé
2du. obl. *uh3wé
3du.R. obl. *sh3wé (*spHé)

This would point to pre-zero-grade PIE *-mWé in the personal
pronouns, versus unstressed *-mWe (< *-mua) elsewhere.

If we look at outside connections, we see an accusative *-m
in Uralic, and an accusative *-b/*-w (*-m after nasal) in
Altaic (Tunguz). The Altaic form confirms the pre-PIE
reconstruction *-mWe, with the same Altaic correspondence
*mW- => *b- (*m- before nasal) as in the 1sg. personal
pronoun.

Uralic and Altaic also contribute another important nuance:
the accusative is usually only marked for definite objects.
Indefinite objects do not take the accusative marker.

The *-m morpheme is not exclusively accusative: Eskimo-Aleut
has a genitive/ergative marker *-m, and Kartvelian an
ergative *-man (which is in fact the ergative of the
demonstrative pronoun and 3rd. person pronoun <is>, obl.
<ma->).

All of this seems to indicate that the accusative in *-m is
of pronominal origin, and based on the pronoun *ma (*mu-a in
IE and Altaic). This pronoun is usually an interrogative/
relative ("what?", "which"), but is occasionally also found
as a demonstrative (e.g. Turkic <bu> "this", Kartvelian
<ma->).

If we assume that this pronoun was added to nouns as a
definite marker already in Nostratic, it is not surprising
that we should find it as an accusative (IE, Uralic,
Altaic), a genitive (IE, EA) or a nominative (IE thematic
neuters, Semitic mimation(?)). In PIE, the result of
unstressed nominative [ergative?] *-mu(u) and
accusative/genitive *-mua would have been *-m in either
case.

Pre-PIE, however (perhaps as a result of the phonetic merger
between nominative and oblique forms), mostly abandoned the
use of the definite marker *-m in favour of the
demonstrative root *t-. Here the nominative / ergative *tu-
and the oblique *ta- remained phonetically distinct (*tu-V >
s(W)o(-), *ta-V > to-), also when appended to nouns/pronouns
as a definite marker. In final position, *-s(W) and *-t
gave *-z and *-d, the animate nominative marker and the
neuter pronominal NA ending, respectively.

The fact that neither in the pronouns (*-d), nor in the
definite (*-o-m), nor in the indefinite (*-0), nor in the
collective (*-(e-)h2) the neuters make any distinction
between nominative and accusative provides strong evidence
that the underlying original system of alignment cannot have
been exclusively accusativic.

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