1) After a period of extinction, rate of response may be close to operant level. After some time, the organism is again placed in the setting and extinction is continued. Responding recovers, but over repeated sessions of extinction the amount of recovery decreases.

2) An increase in the magnitude of the CR after respondent extinction has occurred and after some time has passed.

3) A temporary recovery of the extinguished behavior during the first part of extinction sessions that follow the first session.

4) The reappearance of a target behavior following its extinction.

5) The reappearance of a response that had not occurred for an extended time interval during extinction conditions. Robert Epstein (personal communication) prefers the term resurgence because the phenomenon is not random but controlled by the state of the organism's conditioning history and current circumstances, as are all forms of operant responding.

6) A temporarily higher rate sometimes observed at the beginning of an experimental session, following a session in which the rate has declined (e.g in extinction). This traditional term suggests that the earlier rate has "recovered during the intervening time. A more plausible explanation is that stimuli closely associated with the beginning of the session control a higher rate because of earlier conditions of reinforcement and because there has not yet been an opportunity for this effect to be changed by the experimental changes made during the bulk of the preceding session.