Context as an occasion setter following either CS acquisition and extinction or CS acquisition alone

Three experiments used conditioned lick suppression with rats in order to investigate the necessary conditions for obtaining occasion setting by context. Experiment 1 sought occasion setting by context as a product of conditioned-stimulus (CS) excitatory training only. Associative summation of the CS and context was observed, but no difference in responding inside versus outside the CS training context was found after devaluing direct context-unconditioned stimulus (US) associations and controlling for both familiarization and generalization decrement between training and nontraining contexts. In Experiment 2, two stimuli were conditioned in separate contexts, and subsequently each CS was or was not extinguished either in the same context in which it had been trained or in the training context of the other CS. This procedure equated the associative values of the two contexts during acquisition, extinction, and testing. No occasion setting by context was seen either with or without extinction of the CSs. In Experiment 3, a renewal paradigm was used in which a single CS was trained in context A and extinguished in Context B. Testing showed renewed responding to the CS in Context A. A context-CS summation test using an unambiguous CS trained in a third context found Contexts A and B to be equally excitatory. We hypothesize that the lack of occasion setting in Experiment 2, in contrast to that of Experiment 3, resulted either from extinction of the positive occasion-setting properties of the two contexts as a consequence of their both being sites of CS extinction or a failure of a contextual occasion setter to carry information about the particular stimulus qualities of a given CS with which it was conditioned (i.e., occasion setters act on US representations, not particular CS-US associations).