Aversive Control And Positivereinforcement

The control of behavior by aversive stimuli is also closely related to the processes by which behavior is maintained through positive reinforcement. This is especially true in the case of punishment which is by definition the application of an aversive event to some behavior reinforced by another stimulus. One major function of aversive stimuli occurs in complex behavior where we are concerned with not only whether a given response occurs or does not occur but also the actual form of the response and whether it occurs on the exact occasion that is required. Laboratory experiments dealing with the frequency of occurrence of an arbitrary response have involved the intermittent reinforcement of a simple form of behavior, as, for example, a pigeon pecking at a single key. It is possible, however, to intermittently reinforce more complex and larger units of behavior. Such intermittent reinforcement is in fact the rule rather than the exception in human behavior and the particular schedule used has a large influence on how well the behavior is controlled. As an example of a more complex unit of behavior in a pigeon, we may reinforce a sequence of responses in which the pigeon looks at a sample and then pecks at one of two possible stimuli. A peck at the key corresponding with the sample is reinforced; a peck at the other is not. Under these conditions it has been discovered that the pigeon does not ordinarily attend to the stimuli unless errors are punished by interrupting the experiment for a brief period of time each time an error occurs. If the response unit, however, is reinforced on a fixed-ratio schedule - each correct match produces a conditioned reinforcement and every twentieth correct match is followed by food - then the bird pays attention to the stimuli. A similar process would go on when a machinist is turning a complex part in a lathe or milling machine. Each response in the sequence is carefully controlled by a particular stimulus and the successful completion of each operation is a conditioned stimulus because it permits the machinist to go on to the next step. Any mistakes simply postpone the completion of the chain as with the pigeon, and this factor alone brings the machinist's behavior under the control of the relevant stimuli. Even in a simple performance such as a child learning how to walk, this same process operates. A child walking from one place to another is engaging in a chain of responses which is essentially a fixed-ratio schedule requiring a certain number of steps. Walking is a complex performance, however, with the form of each movement depending critically on the position of the child. The movement which has just occurred previously, as with the machinist and the pigeon, determines which response is appropriate next. Simply the requirement of a fixed amount of behavior, all of which must be under the control of specific stimuli, leads to the elimination of incorrect responses. In many chains of responses an incorrect response, in addition to postponing the completion of the chain, requires that the chain be repeated from the beginning. Such would be the case when the machinist, turning a piece on the lathe, cuts in the wrong place. This kind of aversive control is often determined by the physical properties of the environment and is an important basis for maintaining many kinds of stimulus control.

16