The Environmental Control

of Behavior In general, the stimuli occurring when a response is reinforced come to control the future likelihood of emission of that response. This environmental control of behavior results from different stimuli being present when behavior is maintained by different conditions of reinforcement. Not all stimuli present when a response is reinforced come to control the future likelihood of emission of that response, however. The organism is exposed to a complex environment of which only segments are effective. However, some of the environment correlated with the reinforcement of a response may come to control that response without any explicit differential consequence. The bird that is reinforced for pecking at a red key, for example, will also peck at orange keys and yellow keys. In general, a generalization gradient can be specified: the frequency of pecking decreases continuously as the key color changes from the one present at reinforcement. It is a moot question as to which of the stimuli present will come to control the behavior and to what degree. The whole experimental situation is a complex set of stimuli, and we cannot be sure which elements will have an effect.

We may arrange that specific stimuli will control the bird's behavior, however, by differential reinforcement. The control that the key color has over the pecking response can be sharpened considerably if a differential contingency is provided: the color of the key is changed periodically and responses are reinforced when the key is green but go unreinforced when the key is red, orange, or yellow. The tendency to peck colors other than green falls continuously as a result of their nonreinforcement. Concurrently, the continued reinforcement of pecking in the green maintains a high disposition to continue pecking in this color.

Very small changes in stimulation, ordinarily ineffective, will come to control an individual's performance, depending upon what features of the environment are critically correlated with reinforcement. This accounts for the speaker's attentiveness to the eyes of the listener. The probability of reinforcement of broad classes of behavior depends very critically on the attention of the listener, and the position of the eyes, in general, determines whether or not or to whom the listener is attending. When we are in a group, a certain method for determining whether we have someone's attention is to ascertain whether his eyes are focused on us. If two potential speakers are close together, the difference in angular displacement of the eyes is of the order of a few degrees. The focus of the eyes determines where the listener's attention is. The contingencies of reinforcement in respect to eye focus are so important that most individuals in the community rapidly come under the control of these fine stimulus changes. The critical operation in the development of this stimulus control is that the behavior of the speaker goes unreinforced in the absence of the attention of the listener, and the attention of the listener is indicated by whether his eyes are focused on the face of the speaker. The exquisite control with which we attend to other aspects of facial expression depends upon the same kind of differential reinforcement for its development and maintenance. Changes in facial expressions have their effect not so much because they represent a change in mood but because a change in mood is, in itself, primarily a disposition to behave one way rather than another.

The way in which discriminative stimuli come to control behavior makes it more useful to say that a response is under the control of a stimulus than to say that we perceive the stimulus. What we see or notice depends on whether any behavior has been differentially reinforced, depending on the presence or absence of the stimulus. At any moment, a large complex of stimuli appear on the retina of the eye. Only a small portion of these, however, are effective stimuli in the sense of having any influence on behavior. The expert and the layman, for example, looking at the same scene are controlled by entirely different stimuli, and in many cases, we may be sure that the stimuli controlling the expert have virtually no effect on the layman. The behavioral nature of "seeing" is also illustrated in an episode where someone attempts to identify a person passing rapidly in an automobile. We are better able to describe a friend or acquaintance than a complete stranger, because we already have a history of differential reinforcement in respect to the familiar person which has strengthened particular verbal responses on the occasion of the specific stimuli that define that person. In the case of the friend, the stimuli strengthen (control) already existing behavior, while in the case of the unfamiliar person, the behavioral process is much more complex.

The developing child's performance also comes under the control of features of its environment as it acquires new performances which are correlated with reinforcement or nonreinforcement in various situations. The increase in environmental control occurs because many of the reinforcements maintaining the child's performances occur only in specific parts of its environment. At the simplest level, the behavior, or raising the arms and fingers in the line of vision, will produce its reinforcing effect on the child's environment only in daylight and, hence, there is a lowered disposition to manipulate the fingers in the line of vision in the dark. The relevant process responsible for the differential likelihood of responding in darkness versus light is the same as the one described in the pigeon demonstration. The differential reinforcement of the response in the two stimuli resulted in a difference in frequency of responding. The example developed earlier of a child reaching for a block also illustrates how stimulus control develops in the natural environment whenever a response is reinforced because it changes the environment in some special way. If there is a change in the environment (reinforcer) that will maintain an individual's behavior, the special circumstances in which reinforcement can or cannot occur will come to control the behavior. In an infant, the tactile and visual stimulation from handling a block brings the behavior under the control of the relevant stimuli. Initially, the child reaches for the block even when it is out of reach. On these occasions the behavior goes unreinforced by the tactile stimuli occurring when the block is touched. At closer distances, reaching for the block leads to touching the block. As a result, "distance" becomes a discriminative stimulus controlling the reaching behavior. At a distance, the appropriate response becomes "moving the whole body," which is reinforced by a change in position closer to the block. "Correct distance" is a reinforcer because it is an occasion on which further behavior can be reinforced. The analysis of this behavioral episode illustrates the close connection between stimulus control and shaping of new behaviors by special characteristics of the child's environment.

The parent provides many of the important events for the child, and, therefore, comes to function as a major controlling stimulus. For the very young infant, the sight of the parent, for example, is an occasion on which it may be cuddled and handled. Feeding and changing of diapers occur only in the presence of the parent, or at least a person. Even a newborn child is rapidly exposed to sufficient differential contingencies, so that the parent comes to function as a discriminative stimulus. The vocal behavior of the infant comes under very close control of these parts of the environment correlated with the reinforcement or nonreinforcement of crying. The parent who sleeps soundly in the morning, unaffected by the crying of the child, may effectively be producing a low likelihood of crying on that occasion by extinguishing the behavior. The same child in the evening hour, however, may show a very high disposition to cry when placed to sleep, if the now awake parent shows a high disposition to minister to or react with the child whenever it cries. Wherever the reinforcement of any of the child's performances occurs under special circumstances, these will control the child's performance. The time of day, extent of illumination, amount of activity in a particular part of the house, and presence or absence of the particular parent are examples of controlling stimuli potentially affecting even the youngest child.

In the example of the pigeon pecking the green color and not the red, because responses in only the one color are followed by food, the reinforcement conditions in the two stimuli are extremely different: reinforcement or extinction. More subtle difference in reinforcement contingencies will also come under the control of the particular stimuli correlated with them. Almost every nuance of behavior which can be produced by differential reinforcement can come under the control of a special stimulus. The special features of the various audiences to whom the speaker is exposed come very quickly to control the form of his verbal behavior by means of the differential reaction to his speech. In the presence of the infant and the child, very special forms have maximal effects, including baby talk, special intonations, and different vocabularies. It is no accident that persons who have experience with children talk baby talk to babies. These are the verbal forms which have the largest influence on the infant. More accurately, technically, the baby reinforces one manner of speech and extinguishes others. In professionals, such as nursery teachers, the reinforcement is so extensive that the special form of speech frequently carries over to other situations. The public speaker uses complete sentences, and formal speech, omitting colloquial expressions. Postural preparation for lifting a heavy object is another example in which more subtle stimuli control behavior because of differential reinforcement. The preliminary postures which the individual can assume may vary the reinforcement of the behavior of lifting the object depending upon the weight of the object in relation to the posture of the lifter. Most objects have identifiable properties correlated with their weight. In order to lift a sack of grain, a specific posture is required that is different from the posture required for lifting a large pillow. The differential consequences when the two kinds of objects are lifted gradually strengthen the appropriate posture on the occasions of lifting objects of varying weights. Most persons have had the experience of assuming the wrong posture for an object that was deceptively heavy.