ABSTRACT
INTRODUCTION This chapter continues the theme of the relationship between connectionism and logicist, or “classical”, cognitive science. In the last chapter, we examined the computational properties of symbolic systems, which are the most natural implementation of logic, and connectionist systems, which are loosely based on the structure of the brain. We argued that connectionist systems might be better at capturing important computational properties of the cognitive system —these were the “lures” of connectionism which have the potential to contribute to accounts of everyday, defeasible inference. These discussions focused on abstract computational issues. In this chapter, we turn to empirical evidence from the psychology of memory and inference, which, we argue, supports the conclusion that logicist cognitive science is not viable, and that connectionism holds the promise of providing a better alternative.
