On Sat, 26 Feb 2005 00:13:40 +0000, aquila_grande
<
aquila_grande@...> wrote:
>The perfect may be a development of the past forms of some derived
>verbs belonging to a "hi"-conjugation.
I was reading Jasanoff's chapter on the PIE s-aorist. There
is much there I can agree with, although I'm coming at it
from a different direction.
My take is that pre-PIE had a separate aorist category,
which was characterized by a slightly different set of
endings, in particular by a third person marker *-s (3sg.
*-s; 3pl. *-r.s, *-ér-s < *-én + -s). In the active aorist,
with endings *-m, *-s, *-s; ... *-érs, the irregular 3rd.
person sg. ending, inconveniently homophonic with the 2sg.,
was eventually replaced by *-t from the imperfect, but the
3pl. ending was more resistant, and it is still found (as
-ur) in the Vedic root aorist of certain verbs (those ending
in -a:, the injunctive kramur). Unlike Jasanoff, I would
suggest that the precatives (sigmatic optatives) made from
root aorists are not a Vedic innovation, extended from the
s-aorist, but represent an archaism with 3sg. *-s- preserved
(and extended), just like the optative in general preserves
and extends aoristic 3pl. *-rs.
Jasanoff proposes a stative aorist, from which the PIE
s-aorist developed. If I adapt that to my 3rd. person *-s
theory, we get:
*-h2e
*-th2e
*-s
...
*-r-s,
which is basically the Hittite hi-conjugation past tense
(and the Tocharian s-preterite). The development to
s-aorist involves replacement of the stative endings by the
active ones (which is the standard extra-Anatolian
procedure):
*-m
*-s
*-s-t
...
*-r.-s
Reinterpretation of 2sg. *-s as *-s-s leads to 1sg. *-s-m
and 3pl. *-s-r.s, etc.
Some problems remain:
Hittite has a/e-Ablaut (acrostatic *o ~ *e) in the
hi-conjugation, Vedic has /a:/ (Slavic /ê/) < *-e:- in the
s-aorist active indicative (*e elsewhere), which perhaps
points to generalization of e-grade, with
Szemerényi-lengthening in the 2/3sg. ending in *-s.
The sigmatic aorist subjunctive is a somewhat peculiar
formation. Its 2sg. form *-sesi was reduced to *-si which
became an imperative in Vedic, Old Irish, Tocharian. In
Tocharian, it also could become an s-present, which is also
slightly remarkable (normally subjunctives do not become
present indicatives in Tocharian). The Hittite
2pl.imperative nai-s-ten (and the middle imperatives
naishut, naisdumat) would seem to suggest that the category
had already been sigmatized in Anatolian, which is strange
given the absolute lack of *-s in Hittite in all these
"stative aorist" forms, except in the 3sg. and 3pl. where
the *-s properly belongs.
=======================
Miguel Carrasquer Vidal
mcv@...