KARTVELIAN

I had planned to stay for a while in the Caucasus, but I'm still
waiting for my copy of Kevin Tuite's "Kartvelian Morphosyntax", so
this visit will not be as in-depth as it should be. I will
concentrate on Georgian only, since I don't have some basic data on
the other languages that constitute the Kartvelian language family:
the closely related Mingrelian-Laz (Zan) and the more distantly
related Svan.


KARTVELIAN PERSONAL PRONOUNS

The forms given in Klimov's "Etymological Dictionary of the Kartvelian
Languages" are:

Georgian Mingrelian Laz Svan

1 me(n) ma ma(n) mi
poss. c^em- c^kim- c^kim- (mi-s^gwi- "me")
2 s^en si si(n) si
poss. s^en- skan- skan-, ckan- isgwi

1pl. c^ven c^ki c^kun gu-s^gwej (incl.)
ni-s^gwej (excl.)
poss. c^ven- c^kin- c^kun-
2pl. tkven tkva(n) tkva(n) sgäj
tkven- tkvan- tkvan- isgwe(j)


The reconstructed forms as given by Klimov are:

1. *men
*c^(w)em-
2. *sen
*s^(w)en-

1. *c^wen
*c^wen-
2. *tkwen
*tkwen-

Georgian 2sg. s^en is an intrusion from the possessive.

It is important to realize that Kartvelologists reconstruct three
different series of s(h)ibilants, the reflexes of which are the
following:

Proto-Kartvelian Georgian Zan/Svan
s z c c. dz s z c c. dz s z c c. dz
s' z' c' c.' dz' s z c c. dz s^ z^ c^ c.^ dz^
s^ z^ c^ c.^ dz^ s^ z^ c^ c.^ dz^ s^k -- c^k c^.k. dz^g
sk dzg

The reflexes of *s^w and *dz^w are Zan/Svan sk, dzg (Gamkrelidze's
rule).

This is why Geo. c^em- "my" corresponds to Zan c^kim- and Svan -s^gwi,
and why Geo. s^en- "thy" corresponds with Zan skan-/ckan- and Svan
(i)sgwi.

However, the traditional reconstruction is problematical. Why would
plain *s^ have given /s^k/ in three out of four languages? Why are
the s^ibilants (series III) so often found before *w?

A better solution seems to me to allow only plain sibilants
(traditional s, z, c, c., dz) and shibilants (traditional s', z',
c', c.', dz') in Proto-Kartvelian, and to explain the (less common)
third series as arising from earlier labialized *s^w/*sw, *z^w/*zw,
*c^w/*cw, *c^.w/*c^. and *dz^w/*dzw, with delabialization in Georgian,
and *w > /g/ in Zan and Svan (as in Armenian). While the exact
details still need some work, the reconstruction of the personal
pronouns should now be something like this:

1. *mi
*c^wem-
2. *ci
*cwen-

1. *c^wen
*c^wen-
2. *tkwen
*tkwen-

If we further assume that the *c^/*c^./*dz^ series arose out of
palatalization of earlier *k./*k./*g, while the *c/*c./*dz series
represents the palatalization product of *t/*t./*d, we can rewrite the
preceding as:

1. *mi
*kWim-
2. *ti
*t(k)Win-

1. *kWin
*kWin-
2. *tkwan
*tkwan-

First person *mi and second person *ti need no further comment.

The first person possessive is based on the stative *ku extended with
a (genitive?) *-i. The suffix *-m is likely to be the 1st person
marker *mi.

The second person possessive is built on the same model: stative *tku
extended with *-i and suffixed by a second person suffix *-n (which
we'll meet again in Uralic and perhaps PIE).

In the plural, we would have expected:

independent possessive
*ku-an > *kwan *ku-i-n- > *c^win-
*tku-an > *tkwan *tku-i-n- > *cwin- / *s^win-

Instead, in the 1pl. only the form *c^win- is attested, while in the
2nd. person only *tkwan(-) has survived (but note Svan sgäj < *s^win-
?)


GEORGIAN NOUN

The declension of the Georgian noun is the following:

pl.(I) pl.(II)
nom. -0, -i -eb-i -n-i
voc -o -eb-o -n-o

erg. -ma, -m -eb-ma -t
dat. -s -eb-s -t

gen. -is -eb-is -t

ins. -it -eb-it --
adv. -ad -eb-ad --

The genitive, instrumental and adverbial cases in the singular are
"weak" in the same sense as in PIE (the verbal root may be syncopated
before these endings). The nom., voc., erg. and dat. are "strong".

The regular plural is the one in *-eb-. The -n/-t plural is used only
in fixed phrases and "higher style".



GEORGIAN VERB

The Georgian verb is highly complex, and we'll focus here mainly on
the way subject, object and indirect object are expressed within the
verbal complex.

The Georgian verb has 3 series (present, aorist and perfect), which
are in turn divided into a number of "screeves" (mc.k.rivebi).

Present present ind. future ind.
imperfect ind. conditional ind.
present subj. future subj.
Aorist aorist ind.
optative (aorist subj.)
Perfect perfect ind.
pluperfect
perfect subj.

In the present series, the subject is in the moninative case, and the
direct object in the dative case (-s).
In the aorist series, the subject of a transitive verb is in the
ergative case, with the object (as well as the subject of an
intransitive verb) in the nominative case.
In the perfect series (as well as in all series of Class IV verbs --
verbs denoting feelings or states), the subject is in the dative, and
the object in the nominative.

First and second person markers are prefixed to verb, while third
person and plural markers are suffixed:

subject object/dative
1 v- (*xw-) m-
2. 0- (*x-) g-
3. -a/o/s 0/s/h-
1. v- -t gv-
2. 0- -t g- -t
3. -en/an/nen/es 0/s/h- -t

The 1st person subject prefix seems to be the possessive *ku- > *xW-,
while the (indirect) object uses the marker *m.

Likewise, the 2nd. person prefix can be derived from possessive
*(t)k(a)-. The (in)direct object prefix g- is mysterious. Perhaps it
contains 2nd. person *n-, combined with *(t)k-.

The 1st. person plural object prefix *gw- looks like the 1st.p.pl.
inclusive prefix gu- that we saw in the Svan personal pronoun.
Perhaps it can be derived from *m + 2nd. person poss. *(t)ku.
Alternatively, it could be 2nd. person *n + first person poss. *ku.

I can't say much about the 3rd. person markers at the moment.

The plural markers are *-t and *-n.


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