Context specificity of conditioning and latent inhibition: Evidence for a dissociation of latent inhibition and associative interference

Conditioning may generalize from one context to another when latent inhibition (the effect on subsequent conditioning of prior unreinforced exposure to the stimulus) does not. Experiment 1 studied conditioned approach and licking by rats to a stimulus paired with the delivery of water and confirmed that latent inhibition could be abolished by a change of context, while prior aversive conditioning to the stimulus, produced by pairing it with mild shock, interfered with the acquisition of conditioned approach and licking regardless of this change of context. Thus even when conditioning and latent inhibition were measured in the same way, the former generalized across contexts, and the latter did not. Experiment 2 showed that the effects not only of unreinforced exposure to a stimulus but also of presentation of the stimulus uncorrelated with the delivery of water were confined to the context in which they occurred. Thus the generalization of conditioning from one context to another and the failure of latent...