Resistance to extinction versus extinction as discrimination

The hypothesis that response strength might be measured by persistence of responding in the face of extinction was discredited in the 1960s because experiments showed that responding persists longer following intermittent reinforcers than following continuous reinforcers. Instead, researchers proposed that the longer persistence following intermittent reinforcers arises because intermittent reinforcement more closely resembles extinction—a discrimination theory. Attention to resistance to extinction revived because one observation seemed to support the persistence hypothesis: Following training on a multiple schedule with unequal components, responding usually persisted longer in the formerly richer component than in the formerly lean component. This observation represents an anomaly, however, because results with single schedules and concurrent schedules contradict it. We suggest that the difference in results arises because the multiple-schedule procedure, while including extensive training on stimulus discrimination, includes no training on discrimination between food available and food unavailable, whereas comparable single- and concurrent-schedule procedures include such training with repeated extinction. In Experiment 1, we replicated the original result, and in Experiment 2 showed that when the multiple-schedule procedure includes training on food/no-food discrimination, extinction following multiple schedules contradicts behavioral momentum theory and agrees with the discrimination theory and research with single and concurrent schedules.     

Effect of effort and latent extinction on resistance to extinction

Sixty-four male Wistar rats were given acquisition training in an enclosed straight-alley runway which could be adjusted for angles of inclination. The 2×2×2 design involved two angles of inclination in acquisition (0° and 40°) and two angles of inclination in extinction (0° and 40°). Between acquisition and extinction, half the subjects were exposed to a latent extinction procedure and half served as controls. Number of responses in a 30-min extinction session was an inverse function of effort required in extinction. Additionally, latent extinction procedures resulted in reduced resistance to extinction, but only when the effort conditions of acquisition and extinction were constant. When the effort conditions of acquisition and extinction were dissimilar, latent extinction procedures resulted in increased resistance to extinction. The results raise questions about the nature of the learning which occurs during latent extinction training.     

Behavioral momentum and resistance to extinction across repeated extinction tests

The present experiments assessed whether resistance to extinction of pigeons' key pecking decreased across repeated extinction tests. An additional impetus for this research was to determine how the quantitative framework provided by behavioral momentum theory might be used to describe any such changes across tests. Pigeons pecked keys in two‐component multiple schedules (one component associated with a higher reinforcer rate and the other with a lower rate) in which baseline and extinction conditions alternated. In Experiment 1, baseline and extinction conditions alternated every session, and, in Experiment 2, these conditions lasted for 10 and 7 sessions, respectively. Resistance to extinction decreased across successive extinction conditions in both experiments. Fits of the behavioral‐momentum based model of extinction to the data returned uncertain results in Experiment 1 but implicated both generalization decrement and response–reinforcer contingency termination as the possible mechanisms responsible for behavior change in Experiment 2. Thus, these data suggest that experimental manipulations that affect discrimination of changes in reinforcement contingencies may influence resistance to extinction by modulating the disruptive impacts of removing reinforcers from the experimental context and of suspending response–reinforcer contingencies.     

Early extinction effects following intermittent reinforcement: Little evidence of extinction bursts

The occurrence of extinction bursts—transient increases in response rate in excess of those observed in baseline during the period immediately following discontinuation of reinforcement of a response—was examined. In Experiment 1, key pecking of pigeons was reinforced according to a multiple schedule in which a variable-ratio schedule alternated with an interval schedule in which the reinforcers were yoked to the preceding variable-ratio component. In Experiments 2 and 3, rats were screened such that the lever-press response rates of different rats maintained by variable-interval schedules were either relatively high or relatively low. Following these baseline conditions, in Experiments 1 and 2 responding was extinguished by eliminating the food reinforcer and in Experiment 3 by removing the response–reinforcer dependency. Responses immediately following extinction implementation were examined. Response increases relative to baseline during the first 20 min of a 324.75-min extinction session (Experiment 1) or during the first 30-min extinction session (Experiments 2 and 3) were rare and unsystematic. The results (a) reinforce earlier meta-analyses concluding that extinction bursts may be a less ubiquitous early effect of extinction than has been suggested and (b) invite further experimentation to establish their generality as a function of preceding reinforcement conditions.     

Response production during extinction training is not sufficient for extinction of evaluative conditioning

Two high-powered experiments examined the role of evaluative response production in the extinction of evaluative conditioning (EC) by positioning EC in the procedural and conceptual framework of classical conditioning (CC). According to Rescorla's response inhibition hypothesis, more frequent responding during extinction training results in larger extinction during testing. Experiment 1 used three extinction conditions following response acquisition in an EC procedure: evaluative responses were measured only after extinction; after acquisition and after extinction; or were continuously measured after acquisition, during extinction and after extinction. Based on Rescorla's response inhibition hypothesis, we predicted that extinction of EC would be the highest in the third condition. Experiment 2 was aimed at further facilitating extinction of EC by encouraging participants to experience that their evaluation may change over the course of the experiment. To this end, half of the participants completed pre- and post-acquisition ratings prior to practicing continuous response expression in the extinction phase. Contrary to our predictions, no extinction of EC was observed in either of these experiments. We conclude that Rescorla's inhibition response hypothesis may not apply to EC and discuss the theoretical implications of this finding.     

Avoidance extinction in goldfish

Independent groups of goldfish trained to avoid shock in a shuttlebox situation were presented with several extinction procedures in which the relationships between the conditioned stimulus and shock were altered and/or response contingencies removed. Random shock presentations, equivalent to the number of shocks received during avoidance acquisition, resulted in response decrements similar to those obtained when the conditioned stimulus was presented alone. Pairing the conditioned stimulus with shock on every trial, however, served to maintain response levels. When response-contingent punishment was superimposed upon these Pavlovian pairings, performance was facilitated slightly although punishment alone resulted in somewhat faster response reduction than that produced by exposure only to the conditioned stimulus. Extinction of avoidance responding produced by exposure to the conditioned stimulus alone was dependent on the total duration of exposure and independent of both number of stimulus onsets and response prevention. These experiments demonstrated that, in general, the procedures used to reduce avoidance responding in rats were equally effective for goldfish, with one exception: the introduction of a Pavlovian contingency following avoidance acquisition, making the previously avoidable shock unavoidable, maintained response probabilities near previously established levels.     

Cognitive processes in extinction

Human conditioning research shows that learning is closely related to consciously available contingency knowledge, requires attentional resources, and is influenced by language. This research suggests a cognitive model in which extinction consists of changes in contingency beliefs in long-term memory. Laboratory and clinical evidence on extinction is briefly reviewed, and it is concluded that the evidence supports the cognitive position. There is little evidence for a separate, noncognitive conditioning system. The primary implication for neural analysis is that learning and extinction are unlikely to be reducible to direct connections in which one stimulus simply activates or inhibits the memory representation of another. Rather, an adequate neural model will involve the integration of both low-level and high-level systems, including attention, representation of stimulus relations in long-term memory, and a dynamic performance mechanism based on anticipation, not just activation.