Researchers once thought of conditioning as automatic and not involving much in the way of higher mental processes. However, now researchers believe that conditioning does involve some information processing.
The psychologist Robert Rescorla showed that in classical conditioning, pairing two stimuli doesn’t always produce the same level of conditioning. Conditioning works better if the conditioned stimulus acts as a reliable signal that predicts the appearance of the unconditioned stimulus.
Example: Consider the earlier example in which Adam’s professor, Professor Smith, pulled out a revolver in class and shot it into the air, causing Adam to cringe. If Adam heard a gunshot only when Professor Smith pulled out her revolver, he would be conditioned to cringe at the sight of the revolver. Now suppose Professor Smith sometimes took out the revolver as before and fired it. Other times, she played an audio recording of a gunshot without taking out the revolver. The revolver wouldn’t predict the gunshot sound as well now, since gunshots happen both with and without the revolver. In this case, Adam wouldn’t respond as strongly to the sight of the revolver.
The fact that classical conditioning depends on the predictive power of the conditioned stimulus, rather than just association of two stimuli, means that some information processing happens during classical conditioning. Cognitive processes are also involved in operant conditioning. A response doesn’t increase just because satisfying consequences follow the response. People usually think about whether the response caused the consequence. If the response did cause the consequence, then it makes sense to keep responding the same way. Otherwise, it doesn’t.