Relation Of Intermittent Reinforcementto Psychiatric Problems

execute a given activity, but whether the activity will be carried out at the required time. There is no question, for example, that the depressed patient who stays in bed all day knows how to dress or drive his car or carry out the activities associated with his work. The significant fact is that the performances do not occur. Nor is the ability of the mute schizophrenic to talk an issue. The required verbal responses are in the repertoire and potentially available. While a large number of different conditions are potentially capable of reducing the over-all frequency of an individual's behavior, as in clinical depression, intermittent reinforcement is one of the major factors that may be operative. The question is what variables are responsible for the low frequency. Unfortunately, there has been too little objective measurement of the patient's past behavior to know enough of the actual conditions of reinforcement that have occurred. A particular condition of intermittent reinforcement may weaken behavior profoundly enough so that it alone could be a significant factor in many types of behavioral pathology. We have already discussed animal experiments which show that requiring large amounts of behavior per unit of reinforcement, particularly where the amount of work required from time to time is relatively fixed, will decrease the frequency of responding by large magnitudes relative to other comparable schedules of reinforcement. A bird, for example, that has pecked a key several million times and produced hundreds of thousands of reinforcements, will suddenly stop responding and in some cases starve to death if the number of responses required per reinforcement is suddenly increased. A special history of reinforcement may be applied, however, to maintain the behavior under the low frequency of reinforcement, and normal performance may be reinstated if the amount of work required per reinforcement is reduced and then gradually increased. Since much of human behavior involves performances which have their effect on the environment only after a large and relatively fixed amount of work, there are many situations in which a repertoire may be potentially disrupted by sudden changes in environmental conditions comparable to the changes in schedules of reinforcement discussed above. If the performances which are weakened by the adverse schedules of reinforcement constitute a significant enough part of the individual's total repertoire, the result could certainly be a large reduction in the total behavior of the person, consisting of the magnitude of changes observed in clinical depression. As with the bird, the therapeutic procedure would involve, first, finding some conditions under which the schedules of reinforcement would be optimal enough to maintain behavior strongly before gradually exposing the person to environments in which less optimal reinforcement conditions would occur.

Just as the transition from adolescence to adulthood represents a shift in the kinds of stimuli maintaining the individual's behavior, it correspondingly represents a very radical change in the schedules of reinforcement. Much of the adolescent's behavior involves very short chains of responding and continuous reinforcement. The adolescent frequently has but to spend his allowance (continuous reinforcement). With adulthood, however, reinforcement becomes much more intermittent and rein-forcers which previously occurred after minimum amounts of behavior now occur with much less certainty and at the end of long chains. The youngster, for example, who previously had received an allowance which made possible some of his social behaviors, now works all day long for the money on which his social behavior depends. The student in high school who studied 30 to 60 minutes a day and could successfully pass all his courses now is required to read hundreds of pages, write long reports, and make many attempts at various possible solutions of problems in order to achieve the same net effect on his social environment. Even the schedules of reinforcement of social behavior become less optimal. The friends who were readily available for social interchange are now less available because of their new practices in the community. Dealing with the opposite sex introduces a new measure of uncertainty in social reinforcement. These large changes in schedules of reinforcement are potentially capable of radically reducing the size of the adolescent's behavioral repertoire.

Whether the severe changes in the disposition to behave that characterize many psychoses and neuroses are primarily results of over-all conditions of positive reinforcement, changes in schedules of reinforcement are undoubtedly substantial factors in many of these pathologic states.