Yet another areal feature in SA languages?

[birgit kellner]

To the extent that the specific example of Japanese "te shimau/shimatta"
constructions bears on more general issues of evaluating verbal
aspect/pragmatics, I'll briefly reply to Jonathan Silk's remarks -

jonathan silk wrote:

> Neither I nor Birgit (nor Michael Witzel, I may add) are native speakers of
> Japanese. But according to my wife, who is a native speaker, there
> certainly *is* a negative connotation in this expression; my wife says it
> may also mean, in addition to "oops... I really should not have drank
> this", "unwillingly or without intending to, I drank the sake." Therefore
> the English (a) is not correct.

With all due respect to your wife, whether or not one single native
speaker interprets an *isolated* utterance according to either of two
theoretically possible alternatives that are assumed to be valid
*within* a given context (dis)proves nothing at all. For the issue is
precisely that, when given *within* a certain context/situation, the
utterance in question could be understood in different ways. So in order
to test this utterance with native speakers, one would have to devise
contexts and ask whether or not these are viewed as natural or not (and
if not, this is of course not necessarily an indication of grammatical
mistakes, but may also be an issue of pragmatic inappropriateness).

On the issue of whether or not attitudes on the part of
speakers/listeners to actions can/should be analyzed under the heading
of verbal aspect:

> Perhaps this is so in general, but at least with this particular
> construction, aspect seems fine.  We need the comments of a Japanese
> linguist, I think...  But privately I think.

Why exactly would aspect seem fine in this case? This puzzles me.
Precisely such cases where specific verbal forms are *only* associated
with (un)desirability under specific pragmatic circumstances, but *not*
in isolation, seem to me utterly inappropriate for an analysis in terms
of verbal aspect. Aspect is concerned with whether or not an
action/state which is denoted by a certain verbal form is completed, in
the state of progression, and so on. In other words, it is related to
tense, and independent of speakers' evaluative attitudes to what was
done. Also, I am not aware of any contemporary linguistic literature
wherein factors such as the way in which a speaker feels or thinks about
an action or its result are considered to be a part of verbal aspect.

To the extent that (un)desirability of actions/states was at least
considered as a feature also of OIA languages, I believe that the issue
of how to deal with such factors within linguistic theory is also
pertinent to the classification of Indic languages (just to make it
clear that we're not just rambling away about Japanese :-)).

Regards,

--
birgit kellner
department for indian philosophy
hiroshima university