default aorist, middles

Mari Broman Olsen (
Fri, 1 Nov 1996 10:13:24 -0500 (EST)

Thanks, Rod, for analyzing Ken's specific example. I'm immersed in
something completely different right now (a paper on computerized
tagging of lexical aspect in English), so the warp speed on the switch
was potentially dangerous. I do want to point out, specifically to
Mark O'Brian, that I suggest (and I don't have Fanning in front of me)
that the aorist is only unmarked (or default) in tense (that is, time
reference), not in aspect. The fact that the aorist is perfective in
aspect accounts for how it conditions the ingressive shift of state
verbs (also happens in Russian), since the perfective focuses on the
end state of the event. The notion of perfective as end state (I call
it coda) focus also accounts for the strong sens of completion with
accomplishment verbs, since these (and the non-durative achievements)
are lexically provided with a coda, since they entail an inherent end.

For Don Wilkins (I feel like Casy Kasem) on the possibility of the
aorist being future-referring, whether one takes miscellaneous
examples as evidence against a particular account depends upon one's
view of science (and, to be sure, whether one considers onesself
engaged in scientific enterprise, which I do). Recall Popper:
theories can only be falsified, not proven. Recall Carson (I
mentioned in a previous note): 85% of aorists are past referring--15%
are not! Surely that is greater than the number of three- or five-
legged frogs, and perhaps more explainable (though one would surely
look for toxic waste before changing one's theory about the ontology
of frogs). More examples of future-referring aorists can be found in
Porter's big book (Verbal Aspect...): perhaps the little book (the
misnamed "Idioms" has them at well (but it's at home). Brooks and
Winbery (1979:94) also cite: Mark 11.24 'pisteuete hoti elabete',
claiming such uses "portray future events as if they were so certain
it's as if thay "had already happened". Again, it is not difficult to
imagine focus on the future completion of an event doing just that
(vs. the future-referring present forms, as in Mark 9.31 'ho huios tou
anthrwpou paradidotai eis xeiras anthrwpwn', where the focus is on the
future process)

And for all you working on middles and deponents out there, I'd love
to see someone investigate the Greek middle in the light of Suzanne
Kemmer's _The Middle Voice_ (1993, John Benjamins). Kemmer examines a
wide variety of languages with middle forms of various types
(including those that do and do not distinguish middles from
reflexives). She provides a semantic description of the middle (along
the lines of discourse representation theory and cognitive grammar)
that places it on a continuum with 'true' intransitive and true
ditransitive verbs at either end. According to her, the relevant
distinction is the relative distinguishability of participants in the
event. THe intransitive (John laughed) has one participant, whereas
middles and reflexives split apart the single participants into two
entities. Kemmer shows that this distinction allows for implicational
universals to be described, predicting a range of verbs that will have
middle marking, if a language has it. Good middle candidates are
verbs of grooming (shave onesself), nontranslational motion (wiggle),
change in body position (sit, stand) and verbs denoting mental
commitment (decide, promise vow).

This work has impact on (another line of) my current research, which
involves relating syntax and semantics: to what extent can semantics
(e.g. whether a verb has 'middle' or 'transitive' semantics) predicts
syntax (whether a verb has a second (object) argument). If anyone
DOES work on this in Greek, I'd be interested in hearing from him/her.

Back to the computer work...


Mari Broman Olsen
Research Associate

University of Maryland Institute for Advanced Computer Studies
3141 A.V. Williams Building
University of Maryland
College Park, MD 20742

(301) 405-6754 FAX: (301) 314-9658