What Have Computational Models Ever Done for Us?: A Case Study in Classical Conditioning

Computational Models As Models Of Learning

Computational models of learning have been considered as psychological models in themselves. This position, that constitutes a milestone in the annals of cognitive science and artificial intelligence, is in fact a misuse of the term. We are illustrating our contention by means of a paradigmatic example, the use of Artificial Neural Networks (ANNs) in conditioning theory. In what follows we discuss the inadequacy of such approach at different levels of analysis, namely, ontological, formal, representational, functional, and structural.