Autocontingencies: Suppressive and accelerative effects of pairs of shocks superimposed on a positively reinforced operant baseline.

Previous research has shown that unsignaled shock may accelerate positively reinforced operant responding if each shock signals a subsequent shock-free period. In order to explore the boundary conditions of this effect, two experiments were performed. In Experiment 1, pairs of unsignaled shocks separated by 15, 30, 60, or 120 seconds resulted in suppressed responding during the briefest intershock interval, and in accelerated responding during the longer intervals. When the second shock in each pair signaled a shock-free period of at least 3 minutes, accelerated responding also followed offset of the second shock in all but the 30-second condition. In Experiment 2, the addition of a conditioned stimulus prior to each pair of shocks restored baseline responding, and eliminated accelerative control following the second shock only under the briefest inter-shock interval. The results are discussed in terms of the similarity between autocontingencies (shock/no-shock relations; Davis, Memmott, & Hurwitz, 1975) and recent modifications of the feature-positive procedure (e.g., Reberg & Memmott, 1979), which stress stimulus control by shock/no-shock relationships.