Yet another areal/GENERALITIES

[Vidhyanath Rao]

I thank Professor Witzel for a nice summary of the received wisdom in IE
aspectology.  However I remain a skeptic and explaining why takes
much too long. In view of the new rules, we ought to move this off
Indology. I will summarize my objections below. If those interested will
drop me a note, I can start off a round table discussion with a more
detailed explanation. Alternatively, we can use the IE list
(indo-european at xkl.com, admin: ??; you can do it thru
http://www.liguist.org)

Firstly, there is the internal objection, as formulated by Szemerenyi:
The distinction in Slavic based on verbal prefixes is not inherited. The
imperfects of Latin, Armenian, OCS/Bulgarian, Lithuvanian  and Tokharian
are not morphologically similar to the Greek-I-Ir imperfect.
Furthermore, some forms cognate to the latter are found in Armenian and
Slavic >aorists<, as well as the Baltic preterite. The syntax of Greek
and I-Ir differ too much; in particular the pathway from the former to
the latter has not been explained in terms of diachronic syntax.

Secondly there is the external objection: The assumption that a stem
being usable in the present is equivalent to being atelic, common in IE
aspectology, does not apply to many (most?) languages. The presence of
such a feature must be demonstrated by syntactical study of complete
utterances or, for reconstructed languages, by diachronic syntax, before
we can start partitioning stems.

The same goes for the relation between completives and perfectives. My
interest in completives in Indian languages is that here we might be
able to see the evolution. Most of the languages with completives in
Bybee's sample do not have long known records.

-Nath