Repeatability, recency, and response recovery: An experimental analysis of resurgence.

Four experiments were conducted to assess some of the variables controlling resurgence. The basic procedure involved three conditions: (1) reinforcement of key pecking, (2) reinforcement of treadle pressing, (3) the resurgence condition. In Experiment 1, the resurgence condition was conventional extinction. The extent to which resurgence is replicable was assessed by repeating the three conditions within subjects. Resurgence was demonstrated to be a repeatable effect. In Experiment 2, the effects of reinforcement recency were examined by manipulating the number of sessions (5 versus 30) of Condition 2, and the resurgence condition consisted of conventional extinction. Resurgence was not a function of recency. In Experiment 3, the resurgence condition was a VI schedule arranging a lower reinforcement rate than in Condition 2. Resurgence did not occur as reliably as with conventional extinction, although a correlation was found between key pecks and interval duration. In Experiment 4, a variable-time schedule was in effect for the resurgence condition. Resurgence was not produced by response-independent reinforcer delivery. The results are discussed in relation to the variables controlling resurgence, the relations between resurgence and other forms of response recovery, persistence, behavioral history and behavioral momentum, and potential applications of resurgence procedure