ABSTRACT
This chapter reviews some of the associative roles played by the training and test contexts at the time of testing and described three different associative roles of context during testing. The first is the well-established principle of direct context-US associations. The second is the recently observed potentiation of retrieval of associations between the nominal conditioned stimuli (CS) and unconditioned stimulus (US) when testing occurs in the conditioning context. And the third is the comparator hypothesis that views responding to the CS as a function of the discrepancy between the expectation of the US raised by the CS and that raised by the conditioning context regardless of where testing occurs. Direct context-reinforcer associations have frequently been proposed to account for the contextual cue effect, but only in a few instances has the context US strength been directly manipulated to support the validity of this position. The specific training and test conditions will determine the relative impact of each upon acquired behavior.
