Is autocontingency control established when a traditional contingency is simultaneously available?

Abstract

Using a conditioned suppression procedure, previous research has demonstrated that rats can use unsignaled shock occurrences to predict subsequent periods free from shock (Davis, Memmott, & Hurwitz, 1975). This stimulus arrangement, termed an autocontingency (if shock, then no shock), does not exert behavioral control if a traditional tone-shock contingency is simultaneously available. The present experiment examined whether anything is learned about a simultaneously presented autocontingency, even though behavioral control is not in evidence. A transfer test revealed that exposure to simultaneous and traditional autocontingencies in Phase 1 did not result in stronger or more rapid control by the autocontingency after the traditional contingency was removed in Phase 2. The apparent dominance of traditional over autocontingencies is discussed in terms of stimulus salience and information redundancy.