Verbal Behavior

Verbal behavior differs from other kinds of performances because it is not reinforced directly. Rather, it provides a discriminative stimulus for another organism who provides, in turn, the reinforcing consequence. When verbal behavior has the form of a vocal response, the only consequence would be vibrations in the surrounding air unless an audience were attending; however, its significance lies in the social history of the organism, or in the interrelationship of the behavior of two individuals.

Skinner's analysis (1957) of a small verbal episode in which a child asks a parent for toast suggests some of the essential features of verbal behavior. Consider the behavior of the child as diagrammed in the top line of the diagram below.

In the presence of an adult, the child says, "Toast, please." This response is followed by toast from the parent, which is the occasion for the child's eating the toast. The delivery of toast reinforces the verbal request or mand, as Skinner defines this type of verbal response. The presence of the adult is a discriminative stimulus for the response, since the verbal response goes unreinforced in the absence of the parent.

Now consider the behavior of the adult, as it is outlined in the bottom line of the diagram. The sequence begins with the stimulus from the child's verbal response, "Toast, please." This response is the occasion on which the parent's behavior of giving toast has been reinforced by his pleasure in seeing the child eat the toast. Another possibility is that in the past the child has screamed aversively when he did not receive toast. Consequently, the adult gives toast in order to avoid the aversiveness of the possible tantrum.

This interaction (represented by the arrows crossing the line) between the two chains of responses occurring in the adult and child is social behavior. The behavior of asking for toast is under the control of the adult's presence because the behavior of speaking is unreinforced in the adult's absence. "Toast, please" is a response emitted by the child; for the adult, it is a stimulus in the presence of which the behavior of "giving toast" will be reinforced by his pleasure in seeing the child eat it. The child's eating the toast maintains both chains of responses. For the adult, seeing the child eat is a positive reinforcement because of a complex of reasons which need not be analyzed for the present purpose. The delivery of toast reinforces the child's verbal mand, "Toast, please." The main function of the child's verbal response was to provide a discriminative stimulus which specified a form of parental behavior that was reinforcing to the child. Even in the very young child, and perhaps especially in the very young child, a verbal response can achieve many important consequences in its environment, otherwise unavailable because of its limited physical and social development in dealing with the external environment.

In a psychologic analysis, the dimensions of a verbal behavior are the muscular activities of the voice apparatus. However, the physical description of this musculature during speech is an extremely difficult one, because of the large numbers of muscles, and their complex interactions, and the extremely fine differences in magnitude that are involved. The alternative to actually measuring the muscular changes of the voice apparatus is measuring the environmental consequences of the muscular activity. This alternative is taken in most psychologic experiments in which the behavior is defined by some simple effect on the environment, such as a switch which closes when a lever is depressed. Although the actual performance in such experiments may be extremely complex and may vary in subtle detail from occurrence to occurrence, its essential property may be recorded as the closure or nonclosure of a switch. Analogously, the objective measurement of speech is possible because most observers already possess a highly developed discriminative repertoire with which to record the speech they hear. Although there may be ample basis for disagreement as to causes of a verbal episode and the functional interrelations of its parts, there is seldom disagreement about what has been said when the observations are carried out with care. The ultimate criterion in objective recording, whether in physics, chemistry, or verbal behavior, is agreement among the observers.

Verbal behavior is therefore a natural activity of the organism, amenable to the same functional analysis of behavior as the individual's other muscular activities. The verbal response is maintained by its effect on the environment; it is shaped by differential contingencies; it comes under the control of relevant stimuli; it may be punished or suppressed by aversive stimuli; and it may avoid aversive consequences. For example, in the verbal episode in which the child asks for a piece of toast, the strength of the behavior is a function of the child's level of food deprivation. The actual form of the behavior is determined by its relation to the parental repertoire. Its reinforcement may depend upon its aversive effect on the listener. In this case, the form of the verbal response might be successively approximated in the intensity, repetition, and tonal quality that would have the maximum aversive effect on the listener. The child's disposition to ask for toast is a function of its relation to the reinforcement by the listener, and its frequency of occurrence depends upon the schedule of reinforcement by the listener and the particular occasions on which the listener is inclined to withhold or give toast.

In addition to the differential reinforcement, stimulus control, chaining, punishment, and multiple determination of individual verbal responses, the verbal repertoire as a whole is influenced by its general reinforcement conditions. The general disposition to speak is a function of the audience, for example. Audiences that have tended to reinforce a wide range of verbal responses will strengthen the verbal behavior of the individual, whereas audiences that have tended to extinguish and punish verbal behavior weaken this behavior. Ordinarily, the reinforcement conditions implied in an "audience" are sufficiently explicit so that the emitted verbal behavior corresponds very closely with the audience's requirements. We speak baby talk to the young child, but communicate in full sentences and more complicated grammatical constructions to the adult audience. Writing letters involves weak verbal behavior relative to those situations in which the verbal responses may be reinforced more immediately. Emotional variables also affect the frequency of verbal responses in terms of the general strength of the behavior and the variables maintaining it.