Propositional knowledge and mere responding

This commentary examines views recently offered by Furedy and Riley (1987) on the nature of classical conditioning. Their analysis identifies propositional learning as a component of conditioning separate from response acquisition. They reserve to this component the label cognitive on the epistemological grounds of falsifiability. We suggest that this analysis does not do justice to the concept of cognition in human conditioning. In particular it limits the scope of propositional knowledge to a formal role in the conditioning paradigm and ignores the recent development of propositional models in cognitive science. In seeking to integrate their propositional account with animal conditioning theory they usefully map their two processes on to the conditioned response/instrumental response distinction proposed by Gormezano and Kehoe (1975). However, their propositional contingency interpretation of the Rescorla and Wagner (1972) theory ignores the essentially associative mechanism used in the model to account for interevent relations.