Abstract

My Ph.D. dissertation is an argument that "connectionist" is the proper model of the mind because it can best model learning. Connectionism describes a family of computer approaches to modelling the mind based on the neural structure of the brain. Consequently, it is supported by those philosophers who look to neuroscience for explanations of human behavior. On the other hand, those philosophers who think beliefs and desires are appropriate for explaining behavior often reject connectionism as a suitable model. I think that in some ways both these camps are right. Belief-desire explanations of behavior are appropriate; however, this kind of explanation essentially involves learning. Thus the mind is best investigated according to a model which makes sense of learning. And therefore, I argue, connectionism is the proper model for our studies of the mind. ;My argument for this conclusion has four parts. I first examine the arguments given for the elimination of mental states from the ontology of theories purporting to explain behavior. I conclude that these arguments give us no reason to expect the elimination of mental states from our ontology. Then I turn to a more positive argument for retain mental states; I argue from the perspective of evolution for the presence of contentful, folk psychological states in our scientific ontology. Next I argue that behavior which is properly explained by folk psychological states is behavior of a learner. And finally I argue that this kind of learning is better modelled according to the connectionist, rather than the classical, paradigm.