What I'm reading 9/13/05 [Barsalou_99]

Barsalou (1999)

This paper is one of the most important papers in all the different areas of cognitive studies in at least the last 20 years.  It was successful in moving the field away from the computer metaphor, and the brain-in-the-vat mentality that had asphyxiated the field since Chomsky’s criticism of Skinner.  Since a summary can be found just about anywhere, I’ll spare the reader, and go directly to comments.

Comments:  I think what is most relevant to discussion is section 2.3.1 on Introspection.  Introspection is a very important component of Barsalou’s perceptual symbol systems, because many other aspects of his model (e.g., the knowledge of objects when they are not present, and abstract concepts) depend on it.  At least one aspect here is unclear to me.  For example, when a person considers the concept of TRUTH, Barsalou explains that an introspective, or interospective, process takes place where a comparison is made in order to assess the truth.  It seems to me that what the neural basis is for the comparison is still unclear, and it comes dangerously close to a homunculus problem.

Secondly, I’ve always noted the difference between Barsalou’s emphasis on symbols that are grounded in modality-specific areas of cortex, and Glenberg’s (1997) belief that action, and proprioception are the central component in grounding concepts and in how people understand concepts.  My question is are all modality-specific symbols separate but equal?  Or do certain modalities contribute more to our understanding of concepts than others, and in which situations?  I think that Conway and Christiansen (2004) provide a clue to this question.