Re: [tied] Re: 3rd. person *-s(V)

From: Miguel Carrasquer
Message: 31156
Date: 2004-02-17

On Tue, 17 Feb 2004 14:21:51 +0000, elmeras2000 <jer@...> wrote:

>--- In cybalist@yahoogroups.com, Miguel Carrasquer <mcv@...> wrote:
>> A brief review of the evidence:
>>
>> Afro-Asiatic
>> Hausa
>> 3masc. p.p. <shi:>, verbal prefix <shi>, <sa>, ...
>> 3pl. p.p. <su:>, verbal prefix <su>, <sun>, ...
>>
>> Berber
>> 3masc. suffix -s
>> 3pl. suffix -s&n
>>
>> Egyptian
>> 3masc. suffix -f < *-su, p.p. (old) swt, (new) nt-f
>> 3fem. suffix -s < *-si, p.p. (old) stt, (new) nt-s
>> 3pl. suffix -sn, p.p. nt-sn
>> 3su. suffix -snj, p.p. nt-snj
>>
>> Cushitic (Beja)
>> 3masc. suffix -s
>> 3pl. suffix -s-na
>>
>> Semitic
>> 3masc. p.p. Akk. s^u:, Arab. *s^uwa > huwa
>> 3fem. p.p. Akk s^i:, Arab. *s^iya > hiya
>> pl.masc. Akk. s^unu, Arab. *s^unu > hum(ma)
>> pl.fem. Akk. s^ina, Arab. *s^ina > hin
>
>In Afroasiatic the forms of the 3rd person do in fact appear to be
>fully parallel with those of the other persons. However, what the
>3rd person sibilant is parallel with is a 2nd person -k-. This does
>not look like a very close relative of IE, if it is a relative at
>all.

The pronominal system of AA is indeed different from that of IE, Uralic,
Altaic and Eskimo-Aleut. It is basically identical to that of Basque.
Kartvelian seems to occupy a middle position.

We have (simplifying):

First person *ni (Basque ni, PAA *ni) vs. *mi (Kartv *-m-, PIE *me, Ural
*mi-, Alt. *b- < *mw-?, EA du./pl. -u- < *-m-)

Second person *ki (Basque hi, -ga-, -k, PAA *ki, Kartv *-k-/*-g-) vs. *ti
(PIE *tu, Ural *ti-, Alt. *s- < *tw-?, EA *-n < *-t).

However, there are reasons to think that *ni in PAA goes back to *mi. The
plural in Hausa is /mu/, in a system ni *ki si / mu ku su, which can just
as easily be derived from *mi/*mu *ki/*ku *si/*su as from *ni/*nu, *ki/*ku,
*si/*su, as it usually is.

In the second person, I would tend to reconstruct PN *ki, rather than *ti.
But besides the pronoun, PN also had a prefix conjugation, with 2nd. person
morpheme *ta-. Perhaps that was a factor in the fronting of *ki to *ti in
the "northern" branches.

Of course the above is only an attempt to explain how the systems
*mi/*ti(/*si) and *ni/*ki(/*si) _could_ be connected, in case there is
sufficient evidence _elsewhere_ to think that that the pronominal and
grammatical systems _must_ be related. I believe this is so, but I realize
not everyone will agree with me.

>I should be allowed to overstate my case as much as you. For this
>once I will therefore suggest that it was the use of the sibilant
>morpheme for 3rd person, not the reflexive, in Semitic that caused
>Anatolian to change its use of the sibilant enclitic from reflexive
>to third person.

I'm not sure the Semitic influence on Anatolian was that significant, but I
would like to stress the fact, in general, that a change from reflexive to
3rd. person, or 3rd. person to reflexive, is rather trivial. That means
that neither of us is very likely to make a convincing case in the matter
at hand. My hunch is that *sV was originally a 3rd. person pronoun, which
trivially changed to a reflexive pronoun in PIE (perhaps only non-Anatolian
IE) and EA. Your hunch is that it was a reflexive pronoun (as in PIE and
PEA), which trivially changed to a third person (possessive?) in Uralic,
perhaps Altaic (Turkic).

Not to mention the possibility of confusion with a demonstrative root *s-,
which may very likely also have existed in Nostratic (however defined), and
your even more confusing proposal to throw the vowel demonstrative/anaphora
*a ~ *e ~ *o ~ *i into the mêlée.

>> Kartvelian
>> There is only the 3sg. (present/conjunctive) 3sg. marker -s
>(common to
>> Georgian/Zan and Svan).
>
>Vogt says and shows that the Georgian 1./2. persons have their own
>system, while the 3rd person has another. I can't make out any
>system in the verb.

The PK verb had S- and O-prefixes. The present, aorist and perfect behave
very differently syntactically (the present has AS nominative/ O dative
alignment, the aorist has A ergative / SO nominative, the perfect has A
dative / SO nominative) but there is no difference in the S- and O-prefixes
used. The system can perhaps be reconstructed as follows:

S O
1 *xw- *m-
2 *x- *k-
3 *0- *-s/*-a *0-, *x-

The O-set also has a 1pl. inclusive *gw-, but otherwise the plural
agreement prefixes seem to be identical to the singular ones (the
difference is made clear by plural suffixes).

I would consider the O set to reflect the Nostratic personal pronouns *mi
and *ki, third person is unclear (*x- seems to be indirect object only).

The S-prefixes can be related to the stative suffixes in other languages
(Semitic *-ku, *-ta/*-ka, *-0, IE *-h2, *-th2, *-0, Ural. *-k, *-n(k), *-0,
EA *-k, *-tk-?, *-0), so perhaps ultimately deriving from a periphrastic
construction "I am VERBING" xw-VERB, "you are VERBING" (t)x-VERB etc.

>> Altaic
>> The Turkic 3rd. person possessive is *-sI after vowels, *-I after
>> consonants. The third person imperative ending is *-sin.
>
>I fail to see the beautiful parallelism between the possessive
>endings
>
>1sg -im, 1pl -imiz
>2sg -in, 2pl -iniz and
>3sg -i, 3pl -leri

What beautiful parallelism do you want?

The possessive and verbal suffixes are:

present past imper poss. p.p.
1 -im -m -m, -im bän
2 -siN -N -0 -N, -iN sän
3 (-dir) -0 -sin -si, -i on-

1 -biz -k -miz, -imiz biz
2 -siNiz -Niz -iN(iz) -Niz, -iNiz siz
3 (-dir)-ler -ler -sin-ler -ler-i on-lar

The possessive suffixes are *-m, *-N and *-(s)i. Apart from the 2nd.
person *-ng, which has presumably spread from the past (= stative) paradigm
[cf. Mansi-Xanty and Permic in Uralic], the parallelism with Uralic *-m,
*-t, *-sa is as good as it gets. The plural forms are derived from the
singular (plural markers pronominal -iz and nominal -lar).

>We are getting too far outside for IE for my taste. If you take it
>to Nostratic-L I shall be happy to play along, to the extent I can
>find the resources.

I had already tried: one or two messages ago I cross-posted to Nostratic-L
and set Reply-To to Nostratic-L, but your answer came back in Cybalist
nevertheless.

I have done the same thing again.

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