From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!cs.utexas.edu!usc!zaphod.mps.ohio-state.edu!mips!darwin.sura.net!jvnc.net!nuscc!hilbert!smoliar Wed Apr 22 12:03:51 EDT 1992
Article 5130 of comp.ai.philosophy:
Newsgroups: comp.ai.philosophy
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!cs.utexas.edu!usc!zaphod.mps.ohio-state.edu!mips!darwin.sura.net!jvnc.net!nuscc!hilbert!smoliar
>From: smoliar@hilbert.iss.nus.sg (stephen smoliar)
Subject: Re: Categories: bounded or graded?
Message-ID: <1992Apr17.022112.19379@nuscc.nus.sg>
Sender: usenet@nuscc.nus.sg
Reply-To: smoliar@iss.nus.sg (stephen smoliar)
Organization: Institute of Systems Science, NUS, Singapore
References: <1992Apr14.143822.10246@psych.toronto.edu> <1992Apr15.010721.17700@organpipe.uug.arizona.edu> <1992Apr15.150901.10959@news.media.mit.edu>
Date: Fri, 17 Apr 1992 02:21:12 GMT
Lines: 59

In article <1992Apr15.150901.10959@news.media.mit.edu> minsky@media.mit.edu
(Marvin Minsky) writes:
>In article <1992Apr15.010721.17700@organpipe.uug.arizona.edu>
>bill@NSMA.AriZonA.EdU (Bill Skaggs) writes:
>>In article <1992Apr14.143822.10246@psych.toronto.edu> 
>>christo@psych.toronto.edu (Christopher Green) writes:
>>>
>>>From: Stevan Harnad
>>>
>>>We disagree even more on categories. I think the Roschian view you
>>>describe is all wrong, and that the "classical" view -- that categories
>>>have invariant features that allow us to categorize in the all-or-none
>>>way we clearly do -- is completely correct.
>>
>>  With Wittgenstein in mind, I ask:  what are the invariant features
>>that characterize a "game"?
>>
>>  Wittgenstein, of course, argued that there are none.  His view,
>>and an increasing popular view among linguists (see, for example,
>>Lakoff's "Women, Fire, and Dangerous Things") is that natural
>>categories are not defined by invariant features, but rather
>>by radiation from prototypes.  (Thus a bird is anything that is
>>sufficiently similar to the prototypical bird, which is something
>>like a sparrow.)
>>
>   The problem both with Wittgenstein and Lakoff is the same; basing a
>definition on invariant features or on similarities to prototypes have
>the same bug when the categories depend on the intentions of the
>person who is trying to make the definition.  Consequently, what we
>usually need is something that sandwiches the definition between the
>two aspects: structural (that is, of the examples) and functional
>(that is, of the intended purpose or usage of the category).

I tried having a go-round with Harnad on Wittgenstein once;  but I was not, at
the time, able to wend my way through his dismissal.  However, I think I now
have a better idea of what he was trying to say.  I think the problem resides
in that word "invariant."  I think Harnad's position is that, in any given
context, there is some feature set which lies behind categorization;  but
the invariance of that feature set is only with respect to the context.
Thus, we always make an all-or-none decision;  but we do not always make
one which is consistent with the decision others might make.  This is why
Harnad likes to raise to issue of "MIScategorization."  The trouble is that
Harnad never talks explicitly about context, so I may be totally off the mark.
It just seems like the only position which makes sense of his argument.

What all goes into that context?  Here I certainly agree with Marvin that the
context has to include elements of the mental state of the agent making the
category decision (which would include that agent's intentions).  Actually,
this position is not that far from where Wittgenstein first started.  Back
when PHILOSOPHICAL INVESTIGATIONS was still a set of preliminary thoughts
in "The Blue Book," the USE of any category of symbol naming that category
was foremost among his concerns.  I'm not sure he lost that thread to the
extent that Marvin seems to be claiming, although it does seem as if more
than a few Wittgenstein scholars seem to have let it pass.
-- 
Stephen W. Smoliar; Institute of Systems Science
National University of Singapore; Heng Mui Keng Terrace
Kent Ridge, SINGAPORE 0511
Internet:  smoliar@iss.nus.sg


