From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!csd.unb.ca!morgan.ucs.mun.ca!nstn.ns.ca!aunro!ukma!wupost!think.com!ames!network.ucsd.edu!orion.oac.uci.edu!unogate!genisco!arcturus!berry Thu Feb 20 15:21:57 EST 1992
Article 3849 of comp.ai.philosophy:
Xref: newshub.ccs.yorku.ca comp.ai.philosophy:3849 sci.philosophy.tech:2161
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!csd.unb.ca!morgan.ucs.mun.ca!nstn.ns.ca!aunro!ukma!wupost!think.com!ames!network.ucsd.edu!orion.oac.uci.edu!unogate!genisco!arcturus!berry
>From: berry@arcturus.uucp (Berry;Craig D.)
Newsgroups: comp.ai.philosophy,sci.philosophy.tech
Subject: Re: Meaning by correspondence (was: <none>)
Keywords: consciousness,functionalism,meaning
Message-ID: <1992Feb18.211225.24162@arcturus.uucp>
Date: 18 Feb 92 21:12:25 GMT
References: <1992Feb7.232150.8611@husc3.harvard.edu> <1992Feb12.040025.14716@cs.yale.edu> <1992Feb13.125625.8790@husc3.harvard.edu> <6579@pkmab.se>
Organization: Rockwell International
Lines: 34

ske@pkmab.se (Kristoffer Eriksson) writes:

[Definition and discussion of statistical corelation deleted.]

>What may be a problem in using correlations as the basis for meaning, is
>rather the fact that the same variable (in this case, a symbol in the
>brain?) may correlate to many different other variables (outside the brain)
>at the same instant. Perhaps that can be solved by saying that the meaning
>then is the least common denominator between them. Some limitation on how
>contrived the real-world variables that are eligable for consideration are
>allowed to be may help too. Also, another problem is that correlation is
>not really a binary proposition, but rather a matter of degree.

>And lastly, even if one does subscribe to this view, one can still, out of
>curiousity, pose the question of *how* these correspondences are maintained,
>if indeed they do not occur by random chance alone. But that is another
>question.

To my way of thinking, the *real* problem is determining how the mind
selects *which* corelation to apply in a given context.  For example, if I
am cold, a tree might enter my consciousness primarily as a member of the
class "source of firewood"; if I am hot, "source of shade", and so forth.
So, if I am hot I am likely to judge parasols and trees as "conceptually
close", while if I am cold I would be less likely to see an immediate
relationship between them.  This gets back to the old point about perception
as an active process, depending as much on the state of the perciever as on
the nature of the thing percieved.  The corelations have to have enough
"contextual springiness" to handle this context-dependent perception.

-- 

- Craig Berry
"This sentence is true" : Epiminedes' Paradox -- NOT!


