[b-greek] Re: Prototype theory

From: Kimmo Huovila (kimmo.huovila@helsinki.fi)
Date: Wed Mar 07 2001 - 14:54:54 EST

Randall already answered several points, but I add a little bit.

"Rodney J. Decker" wrote:

> I think my response, if I understand this prototype theory (PT) correctly,
> is to suggest that I am saying something quite similar though with
> different categories. (I suspect that is often the case when different
> linguistic systems are employed.) Your explanation does not use the
> terminology semantics vs. pragmatics. (Does PT include such?)

Yes, semantics and pragmatics are two different things, but their
distinction is not clear cut (or if it is made to be discrete, it is
arbitrary), though sometimes practical and real. Linguists differ in how
much they want to ascribe to semantics and how much to pragmatics. I
prefer putting more on the semantics side, because theories of
pragmatics do not really deal with conventionalized uses, but those are
easily dealt with in semantics. Too often some linguists just dump
everything they do not know how to deal with into a 'box', and label it
as a component of grammar (be it pragmatics or lexicon or whatever).

> I understand
> "semantics" to refer to those elements of meaning that *cannot* be
> cancelled, whereas *pragmatics* incorporates contextual factors/usage that
> *can* be cancelled. Thus I would argue that perfectivity (aspect) is the
> semantic value--in your terms, the more central meaning that cannot be
> canceled. The temporal reference of, say, the aorist indicative--which
> certainly *is* most commonly past time (I have no argument there)--is a
> pragmatic factor. Semantics and pragmatics often do overlap. The imperfect
> form is undoubtedly the best example where, at least in its temporal
> references, is nearly always past.

I just add slightly to Randall's explanation by illustrating the
difference between what I am saying and illustrating cancelability as I
am used to seeing the term used. I use English examples, since it does
not matter when I illustrate a methodological point. I assume that most
list members are more fluent in English than they are in Greek.
(1) Joe used to jog.
(2) Joe jogged.
Part of cancelable semantics of (1) is that he does not jog anymore.
Some would say it is a pragmatic implicature, but it is definitely part
of the meaning of 'used to', whether you label it semantic or pragmatic.
Yet it is cancelable, since it is OK to say 'Joe used to jog, and in
fact he still does.'
The situation is different with (2) and pastness. It is not cancelable.
It is not OK to say 'Joe jogged. That will happen tomorrow.'

While the Greek aorist pastness may be more like the semantics of 'used
to' in that it may be pragmatically canceled, the pastness of the
imperfect (whether English or Greek) cannot. The only way you can have a
non-past reference for it is that you use another focus of the prototype
category, for example of the type 'if Joe jogged'. But this is not
canceling the meaning pragmatically.

Another question is if there are limitations of when the pastness of the
aorist may be canceled, i.e. is it a purely pragmatically cancelable
category. I assume that perhaps with some speakers, but I am not sure
that it would hold true to all speakers. (Just assuming, I have not
studied it rigorously.) Yet if we probe the question further we are
bound to find clear patterns of use. It would be a good MA thesis topic
to find out what limitations there are on proleptic aorists. Perhaps I
just label it 'cancelable' because I am not aware of all the

One more comment, though I guess this should go to another thread.
Randall brought up the example of 'fox'. The metaphor can easily be
dealt with as a semantic phenomenon. Encyclopedic knowledge is part of
semantics according to prototype theory (though this is contrary to some
other schools of semantics). But of course pragmatics is also relevant
to the understanding of the passage. However, the classification is
really not so important, since we do not have any clear cut,
non-arbitrary distinctions between semantics and pragmatics, anyway.

Mark Wilson wrote:

> I rather like the contribution of Prototype Theory to Linguistics, what
> little I have read so far. I am far from one to disagree with Dr. Olsen, but
> I just can't seem to buy her idea that if in SOME instances a temporal
> element can be cancelled, then the temporal element of the tense(for her,
> Aorist and Present) is not part of the verb form SEMANTICALLY.

If she distinguished between semantics and pragmatics on the basis of
pragmatic cancelability, she seems to be right on the aorist (but see my
qualification of this above), which is primarily an aspect and only
secondarily a tense. But with the present I might disagree. It may be a
non-past tense, with a secondary focus of the prototype category as a
historic present serving a textual function. It cannot refer to a past
time event in any other context than as a historic present (a
prototypical case of non-past reference), can it? Then this would be
more parallel to imperfect. It does not change the situation that
historic presents can be very common - it is irrelevant to my basic

> Some suggest that 85 percent of Aorist's are past referring, but it is more
> than that if you include the temporal idea that the remaining 15 percent
> often INCLUDE the past ALSO (omnitemporal use, for example). This was my
> concern when I first questioned Decker on Olsen.

I think that the point is not to include the past, but to leave
omnitemporal use out of the count altogether, as it is a situation where
the temporal aspect of the verb fades out. But of course that depends on
what you want to do with the statistics.

> If the Aspect is more prominent in Greek, it is surprising to me that more
> Aorists do not appear in non-past referring situations. To state that the
> Perfective Aspect tends toward pastness seems correct, but I think that
> simply because the pastness can be cancelled in some instances does not
> render the temporal element of the Aorist non-existent.

True. Fanning argued that the future referring use is related to
conditionals and cases 'as if it were already done'. But is this latter
not better to be seen as a case where the aspectual use overcomes the
temporal reference? See for example Romans 8:30 where EDOXASEN (notice
the cohesive tie to DOXA in 18, where it is qualified by MELLOUSA) is
referentially future. Examples like this point out that aspect is
primary and tense secondary in the case of the aorist indicative (the
aspectual meaning is always perfective).

> How does Prototype Theory address this issue of cancelability as it relates
> to the semantics of a tense form?

Perhaps I already answered this one sufficiently. If not, try asking
again. Notice, however, that in your question 'on Cancelability', the
'If I ran' has not PRAGMATICALLY canceled pastness, but that canceling
is of a different nature as I explained above.

Okay, I think I may have already made enough controversial statements
and stepped on enough linguistic and exegetical toes to keep me busy the
rest of the week defending my views. Time to stop here.

Kimmo Huovila

B-Greek home page: http://metalab.unc.edu/bgreek
You are currently subscribed to b-greek as: [jwrobie@mindspring.com]
To unsubscribe, forward this message to leave-b-greek-327Q@franklin.oit.unc.edu
To subscribe, send a message to subscribe-b-greek@franklin.oit.unc.edu

This archive was generated by hypermail 2.1.4 : Sat Apr 20 2002 - 15:36:52 EDT