COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	Comments on Dennett's %2True Believers: The Intentional Strategy and
C00008 ENDMK
C⊗;
Comments on Dennett's %2True Believers: The Intentional Strategy and
Why it Works%1.

1. It seems to me that belief and other intentional concepts are subject
to unjustified destruction by Socratic inquisition.  The Socratic
inquisition demands that they be made precise and equates this requirement
with defining how to tell whether a sentence involving the concept is
true in arbitrary states of the world.  However, intentional concepts are
meaningful only in approximate theories, and when one tries to extend them
beyond the limits of applicability of these theories, one is forced to
improvise, i.e. the common understanding of the concept does not define
the extension uniquely.  Worse yet, the concept may be unextendable, and
this leads some people to regard it as meaningless in its ordinary
application.

	Dennett's treatment will leave people with Socratic tendencies
to nibble at the fringes of concepts unsatisfied.

2. Maybe Dennett's justification for intentional concepts is overly
pragmatist.  More emphasis on the fact that other concepts we use are
similarly founded would help, but I suppose what is really required
is more analysis of how beliefs, etc. are represented physically, and
what can be said about this in our present ignorance of neurophysiology,
and why detailed knowledge of neurophysiology is unneeded for some
purposes.  In fact most of the analysis that needs doing would be
compatible with the ancient Greek belief that thinking took place in
the liver.

3. The statement that the stomach digests food can be regarded as
taking the design stance with evolution as the "designer".  Perhaps
one should regard the "intentional stance" as an example of the
"design stance".  Intentional systems have to be that way or they
wouldn't achieve goals with the information sources and computational
ability available to them.

4. The intentional stance and the design stance may be elaborated a
bit.  Namely, not all malfunctions force you to the physical stance.
You may observe that the alarm clock is running slow and predict when
it will ring, still not knowing whether it has gears or transistors.
Likewise human malfunctions can often be predicted as arising from
false beliefs or stubborn pursuit of subgoals that contradict main
goals.
In fact I am  not entirely convinced that the attribution of beliefs
always requires the attribution of goals.  My guess is that much more
partial intentional structures are reasonable.

5. Perhaps Dennett's requiring unnecessarily complete intentional
structures arises from the need to convince behaviorists, etc. that
the intentional strategy is worthwhile at all.  Therefore, he takes
an overly strong view of the criteria for intentional ascription.
Accepting these criteria will interfere with a proper theory of
intention, because it will be like starting the number system with
2 in order to convince skeptics that using numbers actually makes
sense.

6. In the thermostat example, I would mention the following:

	a. It is at the 1 level of intentionality or maybe at
the 2 level, i.e. calling it intentional is like admitting 1 or 2
to the number system.

	b. People are often in a state of information about a
thermostat or thermostatic system which can best be expressed
intentionally, even though the repairmen use a combination
of the design and physical stances.