Newsgroups: comp.ai.philosophy,comp.ai,comp.robotics,comp.cog-eng,sci.cognitive,sci.psychology
Path: cantaloupe.srv.cs.cmu.edu!europa.chnt.gtegsc.com!news.mathworks.com!news.duke.edu!godot.cc.duq.edu!hudson.lm.com!news.pop.psu.edu!psuvax1!uwm.edu!lll-winken.llnl.gov!ames!news.hawaii.edu!uhunix3.uhcc.Hawaii.Edu!roitblat
From: roitblat@uhunix3.uhcc.Hawaii.Edu (Herbert Roitblat)
Subject: Re: Grounding Representations: ("Grounding" is the wrong word)
X-Nntp-Posting-Host: uhunix3.uhcc.hawaii.edu
Message-ID: <D6FC96.Loq@news.hawaii.edu>
Followup-To: comp.ai.philosophy,comp.ai,comp.robotics,comp.cog-eng,sci.cognitive,sci.psychology
Sender: news@news.hawaii.edu
Organization: University of Hawaii
X-Newsreader: TIN [version 1.2 PL2]
References: <harnad-1503952148320001@sm1.psy.soton.ac.uk> <rwhite.795319853@superior> <departedD5xB4A.544@netcom.com> <3lkl8d$2gm@percy.cs.bham.ac.uk>
Date: Sun, 2 Apr 1995 20:15:06 GMT
Lines: 115
Xref: glinda.oz.cs.cmu.edu comp.ai.philosophy:26511 comp.ai:28711 comp.robotics:19475 comp.cog-eng:3077 sci.cognitive:7154 sci.psychology:39313

Aaron Sloman (A.Sloman@cs.bham.ac.uk) wrote:
> ...

: Note that any theory of meaning that requires there to be an
: existing object that is referred to (a `signified') is just WRONG as
: a theory of how human-like intelligence works. We frequently refer
: to things that don't exist (Mr Pickwick, Unicorns, that parrot
: sitting on your left shoulder, imaginary scapegoats, dieties,
: expected disasters that don't materialise, the largest prime number
: between 24 and 28, and many other things.) We also refer to things
: that are inaccessible in space and time and to things about whose
: existence we are unsure (Was he murdered, and if so by whom?).

: These are not just foolish quirks and foibles: the ability to create
: meaning without assuming a referent (discussed at length a century

> ...

: The phrase "symbol grounding problem" misleadingly suggests that
: every meaningful symbol has to be "grounded", and that leads to
: misguided theories requiring meaning to arise out of some sort of
: "contact" or fairly direct causal connection with reality, like the
: misguided "robot reply" to John Searle's chinese room argument.

Sloman is correct that meaning and reference are separable ideas.  I
disagree, however, in the assertion that symbol grounding necessarily
implies that reference is the mechanism for meaning.  As Harnad points
out, by themselves symbols are merely squiggles and squoggles to a
program.  The problem is how to get from a purely syntactic program to
an entity (even, I think, a program) that has some semantics.
Something has to constrain the interpretation given to the symbols.
Some of this constraint comes from the rules of the syntax, put some
of it has to come from something else.

Specifically, the syntax can transmit truth from a set of premises to
a set of conclusions.  No syntax, however, can prove that the premises
are valid (leave for now the difficult problem that validity entails
vis a vis the Golden Mountain or unicorns).  That is, no syntax can
specify what the symbols represent.  We need a separate theory of
representation that relates the symbols to their meanings, i.e., a
theory of semantics.  One theory that relates symbols to their meaning
is reference, that symbols acquire their meaning by allowing one to
pick out the thing they symbolize.  A quick examination of the
preceding sentence shows that not all meaningful words have referents.
Further, if the meaning of "apple" was the set of all apples, then
its meaning would change every time I ate an apple becuase the set of
all apples would be thereby changed.  I could go on.  

Symbol grounding highlights the need for such a theory, but does not
tie it necessarily to reference.  Harnad also raises the idea of a
transducer as a means for grounding symbols (I hope I am not doing too
much damage to his ideas by abbreviating them here).  The transducer
by virtue of its causal structure ties the symbols it generates to
physical quantities it measures.   Transduction thus constrains the
symbols it generates to be functionally related to physical
quantities.  This is not sufficient for a complete theory of
representation, but it contributes to such a theory.

Notice that I have been writing about constraints.  Symbols must allow
reference, even if they have "properties" that go beyond reference.
Consistency within the syntax and consistency among physical variables
that are transduced are part only of allow the system to "bootstrap"
from some straigtforward symbol-referent relations to more abstract
relations that transcend reference.  Consistency is not an
all-or-nothing property, but rather one of degree.  Few, if any, symbols
have precise specifiable unitary meanings, but rather are merely
generally consistent with the ways in which they are used.  Even
numbers, for example, have a graded typicality structure (4 is judged
more typical than 86).




I look forward to reading the papers that Sloman mentioned:

: `What enables a machine to understand?' in
:     Proceedings 9th International Joint Conference on AI,
:     pp 995-1001, Los Angeles, August 1985.

: `Reference without causal links' in
:     Proceedings 7th European Conference on Artificial Intelligence,
:     Brighton, July 1986. Re-printed in
:     J.B.H. du Boulay, D.Hogg, L.Steels (eds)
:     Advances in Artificial Intelligence - II
:     North Holland, 369-381, 1987.

: Both are available in the Cognition and Affect ftp directory:

:     ftp://ftp.cs.bham.ac.uk/pub/dist/cog_affect

: The files are
:     Sloman.ecai86.ps.Z
:     Sloman.ijcai85.ps.Z

: I make heavy use of Rudolf Carnap's concept of a "meaning
: postulate", explained in his book
: Carnap, R.,
:     Meaning and Necessity
:     Phoenix Books 1956.

: There's still much work to be done!

: Aaron
: ---
: --
: Aaron Sloman, ( http://www.cs.bham.ac.uk/~axs )
: School of Computer Science, The University of Birmingham, B15 2TT, England
: EMAIL   A.Sloman@cs.bham.ac.uk  OR A.Sloman@bham.ac.uk
: Phone: +44-(0)121-414-4775       Fax:   +44-(0)121-414-4281

--
Herbert Roitblat, Ph.D.             
Professor of Psychology             roitblat@uhunix.uhcc.hawaii.edu
University of Hawaii                (808) 956-6727  (808) 956-4700 fax
2430 Campus Road,                   Honolulu, HI 96822 USA
