On Fri, 30 May 2003 18:46:30 +0400, Alexander Stolbov
<
astolbov@...> wrote:
>> > secondly, an innovation shared by all IE languages but Anatolian and
>> Tocharian.
>>
>> The "thorny" metathesis of *tk clusters, as in *dg^Ho:m and *h2r.tk^os .
>
>
>Piotr, could you please explain this in more details (examples are
>appreciated)?
The original order dental + velar is maintained in Hitt. te:kan,
tagnas, Toch A tkam. (Toch B kem.), while it is reversed in Greek
khtho:n, Skt. ks.am-. We don't have a Tocharian reflex of "bear", but
Hitt. hartaggas (/xartkas/) has dental+velar, as opposed to *h2r.k^tos
in the Rest (OIr. art, Lat. ursus, Grk. arktos, Arm. arj^, Skt.
r.ks.as)
>I can't estimate how serious are the following correspondences between
>Tocharian and Italic, however they have been reported:
>- Genitive singular -í
The Tocharian genitive is a thorny subject. The most common form is
TochA -s, TochB -ntse, which Adams ("Tocharian", p. 139) explains as
coming from the Proto-Tocharian definite declension (n-stem byforms of
other stems, with partially pronominal endings), gen.sg. *-nes(y)ó (>
*-n(ä)sé > -ntse), gen.pl. *-nesom > TochB -nts, TochA. -is. Besides
this, we find many other forms, some of them with transparent PIE
explanations (nt-stem -nte < *-ntos; u-stem TochA seyo < *soyous <
*su(:)y-ous "of the son"), others more mysterious (e.g. -epi/-(y)a:p).
One fairly common form is indeed -i. Adams derives it from the i-stem
genitive in *-eis, despite the fact that it is not found on any
original i-stem (ToB seyi "of the son" u-stem; ToB ma:tri "of teh
mother" r-stem).
The Latin o-stem genitive in -i: is also of disputed origin. It is
traditionally linked to the Celtic genitive in *-i:. Possibly both
the Latin and Celtic genitives are simply from *-osyo with early loss
of the *-s-: *-oyo > *-oy > *-i:. Celtiberian has -o (vs. C-stem -os,
so perhaps from -oso > -oo > -o), Old Lepontic -oiso (younger -i), Old
Latin and Faliscan stillhave -(i)osio, Osco-Umbrian -eis (< i-stems).
The Latin pronoun cuius is from *kWosyo > *kWoio + -s > *kWoios >
cuius.
>- Relative pronouns kwis
This is a shared feature of Anatolian, Tocharian, Italic, Celtic and
Germanic (probably a shared archaism).
>- Medium voice in -r (is this the mentioned "r-middle"?)
Again, this is found in Anatolian, Tocharian, Italic, Celtic, and is
likely to be a shared archaism.
>- Subjunctive mood with -á- / -é-
Subjunctives (futures) in -a:- and -e:- as well as preterites in -a:-
and -e:- can be found in most IE languages. Their origin is unclear.
I have suggested a perfect (stative) subjunctive in:
-a-h2a > -a:
-e-th2a > [-etha]
-e-e > -e:
This corresponds best with the Latin future of the i- and C-stems:
-a:-m > -am
-e:-s > -e:s
-e:-t > -et
-e:-mos > -e:mus
-e:-tes > -e:tis
-e:-nt > -ent
A perfect/stative subjunctive seems a good starting point for
subsequent specializations towards:
(perfect) preterit (Balto-Slavic, Latin, Tocharian etc.)
(stative) passive -e:- (Greek), middle -a:- (Armenian, Toch.)
(subjunctive) subjunctive/future (Latin, Celtic, Tocharian)
=======================
Miguel Carrasquer Vidal
mcv@...