Lecture 8

 

Dr. Ron Mallon                                                            Philosophy of Cognitive Science

 

1.  Midterm Quiz: In class, October 30.  Questions will be handed out next week.

 

2.  We’ve been considering reasons to endorse the language of thought.  I want to step back now, and consider what this way of viewing the mind is supposed to help us accomplish.  Remember that the point of thinking about ‘mental representations’ is to allow us to ‘naturalize’ key features of the mind.

 

3.  The story about mental representations does that because it allows us to see how thought processes involving internalized symbols could be a vehicle for rational though. The idea is that the internal causal relationships between symbols preserves or is ‘isomorphic to’ the logical, rational relationships among propositions.

 

4.  This way of thinking about the problem raises new questions - in particular, it raises very pressing questions about meaning.  For in ordinary psychological explanation, we think that mental states are causally efficacious in virtue of their meaning or ‘content’.  So, for example, we think that the reason Joe decided to go on a hike has to do with his beliefs about hikes and his desires that a hike could best fulfill. But, according to the mental representation story, the real cause of Joe’s behavior is the causal action of mental representations in his head.  So what’s the role of content?

      Some philosophers (e.g. Stich 1983) have taken such considerations to suggest that mental states like beliefs and desires are ultimately unnecessary to cognitive psychology.

 

5.  Here’s another problem.  Thinking about thought as being underwritten by mental representations raises an old problem about how we should think about knowledge.  The mental representations story encourages to think of a person’s beliefs as sentences in their ‘language of thought’.  But that raises a problem.  For suppose that my knowledge of how to do something (e.g. make a sandwich) is stored as a set of rules in my head.  Still, it would seem, I need to know how to interpret those rules.  Hence, I would need another set of rules that determines how to interpret the first set.  And then a need a set of rules for interpreting the second set.  It seems like we are trapped in a ‘Regress of Rules’!

     In fact, the comparison of a mind with a computer allows us to see the way out of this problem.  For computers exhibit rule governed behavior that is underwritten by symbol manipulations.  But at some point the computer no longer represents the rules by which it operates.  Rather, the computer is ‘hard-wired’ to act in a certain way, eliminating the need for a new set of rules.