COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	.require "let.pub" source
C00007 ENDMK
C⊗;
.require "let.pub" source
∂AIL Dr. Stevan Harnad↓%2The Behavioral and Brain Sciences%1
↓P.O. Box 777↓Princeton, N.J. 08540∞

Dear Dr. Harnad:

	I have read the paper %2Structural Theories in Psychology and
the Problem of Meaning%1 by Dennis Proffitt.
I found it stimulating but confused.  Here are some detailed comments:

1. Its criticisms of the inadequacy of defining meaning in terms of
the senses are well taken.

2. I must admit that
I evaluate Proffitt's ideas about meaning by comparing them with
some ideas I am developing, and I think his are muddled.  Meanings
(to me) are functions in a hierarchy of intensionality and go down
in the hierarchy.  Near the top of the hierarchy are
sentences and terms(in natural languages - noun phrases).
At the bottom are concrete objects like my dog Caspian and the number 4.
The middle includes concepts like that of %2McCarthy's dog%1 and %2its number
of legs%1.  There are many meaning functions, and their common feature
is that entities with certain meanings identical are intersubstitutable
in certain contexts.  The theory of meaning can be expressed mathematically
without any reference to thoughts, but a particular meaning function
may be important in discussing the thoughts of a particular person or
people in general.

3. I did not find the analogy with quantum mechanical complementarity helpful.
Complementarity in physics leave physicists and philosophers in
an intellectually difficult situation, but no-one knows how to
avoid it.  The uncertainty principle is a logical consequence of
a formulation of quantum mechanics that correctly predicts the results
of experiments, and no-one has found a theory without complementarity
that gives correct predictions.  There is no such excuse here, and I
think Proffitt is pursuing a profitless analogy.

4. Brentano and perhaps Chisholm are mistaken in supposing that
intentional concepts are useful only in discussing psychological
phenomena.  Terms like %2possibly%1, %2necessarily%1, %2caused%1, and %2can%1
are also intentional.

5. Mental concepts are also applicable to machines, and can be studied
more simply in relation to machines.

	I enclose two papers - %3which I don't want to submit for
publication at present%1 - that bear on meaning and on the mental
qualities of machines.
.NEXT PAGE
	As to whether you should publish Proffitt's paper, I don't know
the state of the psychological literature well enough to advise.  If
the views he criticize are widespread, then I suppose it should be
published.  If you do publish it, I will have comments to offer
along the above lines.
The paper is clearly written - given the unclarity of the ideas.

	I have no objection to your transmitting these remarks to the
author.

.sgn