Newsgroups: comp.ai.philosophy
Path: cantaloupe.srv.cs.cmu.edu!das-news.harvard.edu!news2.near.net!MathWorks.Com!europa.eng.gtefsd.com!howland.reston.ans.net!news.sprintlink.net!uunet!psinntp!scylla!daryl
From: daryl@oracorp.com (Daryl McCullough)
Subject: Re: Folk psychology (was: Is Common Sense Explicit or Implicit?)
Message-ID: <1994Oct4.125941.11683@oracorp.com>
Organization: Odyssey Research Associates, Inc.
Date: Tue, 4 Oct 1994 12:59:41 GMT
Lines: 180

>I do not dispute that we are able to predict the behavior of others.
>But I do dispute that our predictions are based on reasoning in terms
>of desires, beliefs, etc.

Then tell me what it is based on. Go through one of my "mind boggling"
examples, and explain how you would predict the behavior. How can you
predict that someone will go outside to roll up their car window
without invoking that person's belief that the window is down? You
have written endless pages of generalities about "empirical science"
and I've been asking you to show how you deal with concrete, specific
cases. You have made sweeping claims about how we make (or don't make)
predictions without even a single example of what you mean. I find this
entire discussion completely frustrating, because your claim to reject
folk psychology seems to have no content other than scientific posturing.

I don't care for the "more scientific than thou" attitude. If you have
a theory that explains and predicts human behavior as well as folk
psychology does, please share it.

>What I find mind boggling is the enormity of the set of unsupported
>assumptions one must make in order to argue for folk psychology.

Look, there are two basic observations that I make about behavior that
I consider intelligent. (1) Behavior seems to be influenced by
information about the state of the world. (2) Behavior seems to be
goal-directed.  Do you want to dispute that? Well folk psychology is
simply reasoning about behavior in those terms. For you to call folk
psychology "unsupported assumptions" makes me think that you aren't
living in the same world I am.

>>In the case of clouds, we can get good statistical data because
>>there is a very small number of parameters: the temperature, the wind
>>speed and direction, the color and size of the clouds, etc. In the
>>case of human beings, the number of parameters is enormous. There is
>>no way in our short life spans that we can come close to getting enough
>>statistical data to even start making statistical predictions. I don't
>>actually think that informal human reasoning ever involves statistics.
>>What it involves is model-building, and reasoning within models. And
>>our model of human beings (or at least mine) involves mental attributes
>>such as beliefs, purposes, etc.
>
>This is crazy. When it all boils down, all empirical evidence is
>statistical.  Yet you seem to deny the role of statistical evidence.

I am not denying the role of statistical evidence, I am denying that
statistical *reasoning* plays much of a role in human reasoning. You
need to understand the distinction between (1) evidence for a theory,
and (2) the theory itself. I don't believe that humans have any
particular abilities at statistical reasoning. What I think people do
is to take the evidence (which as you say is often statistical) and
formulate a model from which they can account for the evidence. And
then they reason from within their model.


>My main disagreement with this is that I expect a theory to have
>empirical support before it can count as empirical science.  For
>unsupported explanatory theories you should look to religion or
>astrology.

Why don't you try giving a coherent argument, instead of throwing
around what obviously you consider insults and put-downs? As I said
before, I am getting tired of this "more scientific than thou"
posturing.

I gave you a few simple examples in which reasoning about behavior of
a person in terms of goals and beliefs leads to testable predictions,
and you gave no coherent response other than "It boggles the mind".

>If folk psychology is to be an empirical scientific theory, then the
>beliefs, desires, etc, must play an actual role.  Now you can either
>take the beliefs and desires to be real entities, or you can take
>them to be theoretical entities hypothesized as part of the theory.
>If they are real entities, they should correspond to what people
>report when they discuss their beliefs and desires.  But there is
>considerable evidence to suggest that human behavior is not a
>rational product of belief and desire.

Whoa. There is a big gap between saying "Human behavior is not simply
a rational product of belief and desire" and denying that people ever
use reasoning about the world in order to accomplish a goal.

>The studies of Kahneman and
>Tversky come to mind.  Many people argue that humans behave
>irrationally, but that is just a way of saying that if beliefs and
>desires are taken to be real entities, then human behavior does not
>conform to the folk psychological model.

That's a non-sequiter. A folk psychological model does not predict
that people behave rationally, it predicts that people behave in a
way that is in keeping with their beliefs and goals. Those beliefs
themselves may be irrational.

>You can take beliefs, desires, etc to be theoretical entities.  That
>is essentially the intrumentalist
>position adopted by Dennett ("The
>Intentional Stance").  But if these theoretical entities are to play
>a role in the theory, they need to be properly defined, and reliably
>measurable, and the theory should be subject to empirical tests.  For
>example you could have three teams investigating some experimental
>subjects.  Team 1 would predict behavior, and write down their
>estimates of beliefs, desires and other intentional states that are
>relevant to the prediction.  Independently, team 2 would write down
>descriptions of the observed behavior, so that it could later be
>compared to predictions.  Team 3 would not observe the experimental
>subjects.  They would instead predict behavior purely on the basis of
>the intentional states recorded by team 1.  If folk psychology is
>correct, the predictions of team 3 should be close to the predictions
>of team 1, and should correspond to the behavior reported by team 2.
>
>My suspicion is that team 3 will do terribly.  They will quickly
>discover that they do not have enough information available to them
>to have any chance of making decent predictions.  However, if you
>know of documented empirical evidence actually supporting folk
>psychology, I would like to know of it.

Good idea. But you haven't told how team 1 is predicting the
behavior. How do you prevent *them* from using folk psychology?
Here's an offer: Team 1 will simply use statistics about past
behavior, while Team 3 will take into account the intentional
states. I am willing to bet that team 3 would do pretty well if the
type of behavior involved and the type of beliefs and goals involved
is simple enough.

But as I told Andrzej, I don't claim that folk psychology is a
scientific theory. It is more like a metatheory, a way of organizing
our information about human behavior.

>>>>                               If you had answered the question "Why
>>>>did that person go back to his car?" with an answer such as "Because
>>>>synapses fired, causing an electrochemical pulse to travel to his leg
>>>>muscles" people would have considered you insane. Not that the answer
>>>>is wrong, only that it is inappropriate, at the wrong level.
>>>
>>>I agree that, in most cases, this would be an inappropriate
>>>explanation. But it does not follow that the intentional explanation
>>>is correct.
>
>>Then give me an example of an explanation you would consider correct
>>and appropriate.
>
>An explanation has to be of the form A caused B.  But such an
>explanation is quite empty if there is no way to measure A.

I don't agree. The value of an explanation is in organizing
information to aid in reasoning and understanding. It is better
if the primitives are measurable, but not necessary.


>I can no more give you an explanation than you can measure the beliefs and
>desires that you claim are part of your explanation.  I have no
>explanation.

That's what I thought.

>You have no explanation. What you do have is a
>rationalizatn which you are presenting as an explanation.

I disagree. There is great organizing value of interpreting behavior
in terms of beliefs and purposes. And that's what an explanation is
*for*.

>I am not in the least denying that people have purposeful behavior,
>and that some of that behavior involves reasoning.

Okay. Well, that, in a nutshell is what I am calling "folk psychology".

>I am also
>interested in the computational aspects of producing purposeful
>behavior.  I am not even denying that purposes, desires, etc, play a
>role in behavior.  What I am arguing is that human behavior is far
>more complex than are the purposes, beliefs, desires which are
>purported to be the causes for that behavior.

Then why didn't you say that? I certainly agree with you there.

Daryl McCullough
ORA Corp.
Ithaca, NY


