Private Events

A very difficult problem in verbal behavior arises when the stimulus controlling the speaker's behavior is accessible only to himself, as, for example, when he says, "My arm itches; I had a twinge in my muscle; I have a headache; I was about to leave the room." In these cases, the speaker is in unique contact with the critical events to which he is referring, and the question therefore arises as to how the speaker acquired verbal behavior about these stimuli when the verbal community responsible for the speaker's entire verbal repertoire has had no access to the crucial events. The stimuli being referred to in a toothache are the unique province of the person experiencing the toothache. These are private events occurring within the integument of the individual and, for the most part, inaccessible to the other members of the verbal community.

The report of a toothache is essentially a tact functionally equivalent to a response such as "There is a chair in the living room." We have already described the manner in which tacts are acquired and maintained. The problem of accounting for the verbal behavior about the toothache is a similar one, except that the community does not have access to the relevant stimulus as it does with the chair. How then is the verbal response acquired? The critical factor that allows verbal behavior to develop about private events is that many private events have public accompaniments. A person with a toothache grimaces, holds his jaw, cries in pain, and does not eat with the involved tooth. Most persons in the community have experience with this kind of symptom, which is closely related to tooth infections. The tact about the tooth inflammation and its resulting pain is therefore established through its public accompaniments. The child who exhibits all of the public accompaniments of tooth pain is told, "You have a toothache." On subsequent reports of tooth pain, he is subjected to a general test in which all possible public accompaniments of tooth pain are scrutinized, from asking, "Which tooth hurts?" to tapping the affected part and observing the child carefully over a longer period of time.

The Development Of Behavior Based On Private Events

The analysis of verbal behavior gives some lead as to how the community develops public behavior about these private events. The basic process by which verbal behavior about private events is developed is the same as with behavior under the control of exteroceptive stimuli, except that both the community and the individual have had access to the chair with the exteroceptive stimulus. On the other hand, with the private event, the speaker has access to the stimulus but the listener has access only to the public accompaniments of the stimulus. For example, in developing the verbal response, "chair," the community differentially reinforces or nonreinforces the verbal response of the individual, depending on the presence or absence of the chair; or its reinforcement is based on how closely the response conforms to the verbal practices of the community. For a toothache, the community reacts to signs of distress, as holding the mouth. When a man stubs his toe, he will hold it and rub it. A mosquito bite will usually be followed by scratching. The community infers the private event from these public accompaniments, which become the occasions on which responses such as "itch" or "hurts" are reinforced in the same way that the response, "chair," is reinforced only in the presence of a chair.

The environment provides many consequences which encourage an individual to emit behavior based on private events. When we see the child scratching, we ask, "Is something itching you?" We constantly inquire, "Where does it hurt?" The major difference between the way private and public events control verbal behavior is in the consistency of the community's reinforcing practices. With exteroceptive stimuli, the listener more narrowly restricts the kind of behavior which will be reinforced than is possible with private events. With private events, such narrow control is not possible because the listener is inferring from public accompaniments what stimuli are controlling the speaker. As a result of this inference, the reinforcing contingency is usually extremely variable; and these tacts do not have the same consistency as with exteroceptive stimuli, where the reinforcement conditions are more explicit.

When a patient describes a tension, sharp pain, stabbing sensation, or the pulling in a muscle, there is no way of knowing, from the verbal report, how similar the actual state of the muscle is to the state of other individuals who make similar reports. The physician has some check on the stimulus as he treats the patient, and he observes his condition in nonverbal ways. Eventually, he learns how much correspondence and variability there is between the report and the underlying condition responsible for the report. The meaning of the verbal report will vary from patient to patient and from culture to culture, depending on the verbal practices of the specific community supporting the patient's verbal behavior. Ultimately, however, the usefulness of the patient report depends on corollary information (public accompaniments) by which the physician can evaluate it. Alternatively, or probably concurrently, the physician establishes a new verbal repertoire by differentially reinforcing the patient in terms of the public accompaniments he observes directly or through medical technics. Often, for example, he specifies the major form of the patient's verbal report by questions such as, "Is it a sharp or dull pain?"

We cannot assume that each individual will have verbal behavior about each of the stimulus changes occurring within his integument, no more than we can assume that all of the exteroceptive stimuli falling on the sense organs control an individual's behavior. In the main, we are controlled by those stimuli which are correlated with differential consequences of our behavior in the environment, similarly with private events. Certain private events such as an itch come to control behavior readily, since scratching is reinforced by a change in the stimuli. Others control because the community constantly differentially reinforces in terms of public accompaniments, as, for example, in the case of pain. However, still others, such as muscular tension, have little effect on the social community. As a result, most individuals are unaware of the state of tension of most of their musculature. A trained observer can readily detect the public accompaniments of muscular tension, however, and training programs have been developed to create a verbal repertoire based on private events in the muscles.

MULTIPLE DETERMINATION OF BEHAVIOR BASED ON PRIVATE EVENTS The difficulty in dealing with behavior presumably based on private events is that the verbal report, or other public aspect of the private event, is an operant response the strength and form of which may be multiply determined by more than one variable simultaneously. We mistrust the person who makes excuses on the basis of a headache or the student who accounts for his error by saying, "That's what I meant to say." The general inability of the social community to reinforce consistently when dealing with private events makes it likely that other variables, particularly those related to current deprivational states of the speaker, will determine the form of the tact. The child's verbal response, "My tummy hurts," may be an avoidance response based on an aversive consequence in school, rather than a tact based on the state of his stomach. Possibly, the response might derive its strength from two variables simultaneously, either of which alone would not be sufficient to produce the verbal response. The proverbial "headache" may not be of a sufficient order of magnitude to bring about the emission of the relevant verbal responses. However, a difficult social situation, which can be avoided by saying, "I have a headache," will provide the necessary additional reinforcement for strengthening the response. Although it is generally impossible to check upon whether the individual, in fact, has a headache or a stomach-ache, particularly when the condition might be present to a limited degree, the continued use of this kind of behavioral control by a speaker will have a cumulatively unfavorable effect on the relevant audiences. The community generally suspects that a high frequency of avoidance responses, based on presumed private events, is evidence of a multiply determined response.

Verbal behavior about private events can also develop by metaphorical extension, related to the process of generalization discussed earlier, even when there is no public accompaniment. If an individual has developed the verbal response "sharp pain" in respect to being pricked by sharp objects, then private events, to the extent that they are similar to the stimulation from the sharp prick, may also elicit the response "sharp pain." Generalization in the private event differs from generalization about exteroceptive stimuli because the community has little control over the extent of the generalization. When we say, "He was as gentle as a lamb," we are making a generalization based on properties of a lamb. Although a great deal of latitude is allowed in such metaphorical extensions, the verbal community still has some control in how much generalization occurs.

Such limits are less likely to be placed on generalizations based on private events. Most of the verbal behavior about emotional states, pains, or muscular states are responses to private events established by generalization from verbal responses to exteroceptive stimuli, as well as the secondary public accompaniments of the private event. For this reason, these words do not have the precision of the terms involved in the description of objects fully available to the community for differential reinforcement. Terms such as happy, depressed, despondent, lighthearted, anxious, fearful, or jittery serve both as terms by which one individual describes another's public behavior or as descriptions of private events. In the former case, the terms are objective descriptions of the individual's public behavior. In the latter case, they are private events. The common language does not distinguish the extent to which the descriptive term is based on the private or public stimuli. In either case, the consistent use of the terms depends upon whether the community can have a consistent reinforcing practice, as it does, for example, with a tact. The difficulty, however, is that these verbal responses may be multiply determined by other variables completely unrelated to the community practice. For example, the psychotherapist who has a strong bias toward a particular kind of interpretation and inference of the private stimuli supporting a patient's behavior may differentially reinforce verbal forms independently of the state of the patient. The patient and the therapist comprise a small verbal community which may have special verbal practices unrelated in important ways to the verbal practices of the larger community. Although this multiple determination of verbal responses is true of any kind of verbal behavior, the problem is more serious with the private events, such as emotional states, because the general verbal community has no way to keep close control of the behavior. Patients' reports of private emo-tories of reinforcement incidental to the tact that is required.

Pain

The verbal repertoire about pain is of special relevance to medical practice and similar to that of other private events such as "felt" emotions. To the extent that the physician bases any diagnostic or therapeutic procedure on the patient's verbal report, he should know the nature of the event on which the verbal report is based. If a tact based on a private event were as pure as one based on a public event, medical diagnosis would be simplified. Unfortunately, verbal responses describing pain are subject to the same kind of environmental control described above for other private events. A behavioral process based on an aversive stimulus may have two different consequences which may be in large part independent of each other. The first and most basic function of an aversive stimulus is its potential for reinforcing a performance which terminates or avoids the stimulus. The second function is secondary effects and collateral behaviors such as verbal reports, physiologic changes, and incidental performance changes. From the point of view of a functional analysis of behavior, how painful a stimulus is would be measured by the amount of avoidance or escape behavior one could maintain with such a stimulus. The collateral effects of aversive control may occur because of secondary processes unrelated to the aversive stimulus. If, in fact, we could measure the painfulness (aversiveness) of a stimulus for a patient by recording a response which terminated or postponed the aversive stimulus, we would probably discover little correlation between the verbal or introspective evaluations of the aversive stimulus and how reinforcing was its termination. The relation between the patient's verbal behavior about the pain and the property of the pain in its ability to maintain avoidance and escape behavior is necessarily a loose one, because the verbal responses about the individual's pain may be under the control of entirely different variables from the particular inflammation or trauma responsible for the aversive state of affairs. Even in behavioral research with animals, the aversiveness of a stimulus is difficult to estimate by collateral operant or physiologic response. An electric shock which causes a vigorous reaction, including squealing, defecation, and urination, may turn out to be an ineffective aversive stimulus when its aversive properties are measured by the amount of avoidance or escape behavior that can be maintained.

SELF-DESCRIPTION The account of the development of behavior about private events already suggests that self-awareness, consciousness, self-description, and introspection involve specific behavioral processes above and beyond the organism's fundamental dispositions to action. We have, first, the behavior of the organism in its commerce with the environment under the control of the relevant variables. Given the behaving organism, however, there is still the further issue of verbal behavior about the individual's own behavior in the same sense as verbal behavior about any other event available to the organism. In experiments on concept formation, for example, in which a subject's response is differentially reinforced depending upon which of two classes of stimuli are presented, the subject's motor behavior comes under the control of the relevant stimuli long before he can verbalize the rule for reinforcement. The verbal responses are a second order process by which the subject describes his motor behavior. The effect of the main variables of the experiment occur independently of the verbal behavior. The final self-description is essentially a tact based on the subject's motor behavior, as with many variables that control an individual's behavior. The change in a person's behavior caused by a variable and the verbal behavior about the variable are two separate issues.

Although the emission of a verbal response often alters the effect of a potential controlling variable, the reverse is usually true. A variable changes an organism's performance, and the altered behavior becomes the stimulus for a tact based on the altered behavior. Self-description differs from other kinds of verbal behavior (tacts) because the relevant stimuli involve the individual's own behavior. Some of this behavior may be public, but its order of magnitude or form may be such that it is not easily available for public inspection. We develop this second order of behavior in the child by constantly demanding of him tacts that are based on his own behavior. We say: Where are you going? What are you going to do? What have you done? And we reinforce, extinguish, or punish depending upon whether the subsequent activity of the child conforms to his verbal statement. When the child is required to report on why he did something, the parent effectively makes a functional analysis of the child's behavior in order to determine whether the child's statements about his own behavior actually conform to the true probabilities of occurrence.

"Behavior based on an individual's own behavior" can be established, even in a pigeon, by technics which closely parallel those described for verbal responses. A pigeon can be trained to "pay attention" to his own behavior by reinforcing a response differentially, depending upon his own performance. For example, we could deliver food after 50 pecks, or after 5 minutes, but unpredictably. Under these conditions, we discover that the bird begins pecking immediately after reinforcement at the standard high rate of responding characteristic of a small fixed-ratio schedule. However, if the current schedule calls for the 5 minute fixed-interval schedule, the bird stops pecking soon after emitting 50 pecks; it pauses and then continues to perform appropriately to the time schedule. The stimulus controlling the fixed-interval performance is the bird's own behavior: the occurrence of 50 responses without reinforcement, which leads to the temporary cessation of responding because it is the occasion on which the schedule of reinforcement is a fixed interval of 5 minute duration. Although we often do not know how closely the bird normally attends or observes its own behavior, we clearly can make the bird attend for its own behavior by reinforcing some response differentially in respect to some performance just completed.

Like other problems in private events, the major difference between self-description and description of other parts of an individual's environment is that many aspects of an individual's own behavior are not readily available for public inspection.

Conversely, an individual's controlling environment may produce broad changes in his behavior about which he is completely "unaware" unless there is some verbal or even nonverbal audience that will reinforce some verbal response about his own behavior. In this sense, individuals differ enormously in the availability of a verbal repertoire with which they can comment on their own behavior. Even when the ability to react to one's own behavior is highly developed, large segments of the behavioral repertoire still remain about which one is completely unconscious, that is, has no verbal responses. The reaction of the individual hearing his recorded speech for the first time is one illustration of the small degree to which the individual normally observes his own behavior.

Much of psychiatric technic is concerned with providing differential contingencies which will generate behavior in the individual about the variables governing his behavior and about the various dispositions and strengths of the various portions of his repertoire. One of the functions of the psychotherapist, in a personal interview, is to provide the differential reinforcement for forcing the patient to attend to aspects of his own behavior about which he is unaware. Naturally, the process may be a very complex one because the absence of self-knowledge may be related to many processes of aversive control, as in repression, rather than simply representing an absence of a verbal repertoire. Even when aversive control is at issue, however, the release of behavior potentially available to the patient does not guarantee that the patient will be able to observe his newly available behavior.

The community develops self-description about various aspects of the behavioral repertoire.

Behavior about the Controlling Variables of which the Individual's Performance is a Function. A child's disposition to eat is a function of when he has eaten last, but there is no assurance that the child can report the condition responsible for his increased frequency of eating. Although a verbal repertoire about such variables may develop, it will be under the control of entirely different variables. In general, the community gives little instruction about the causes of behavior, and there are large variations in the degree to which this behavior is present from individual to individual. A potential reinforcer does exist for self-observation, however, when it becomes possible to manipulate a controlling variable which has been identified. For example, the individual who discovers that his tendency to drink alcohol at a cocktail party is increased by salted peanuts and water deprivation may control his drinking by having several glasses of water at the start of the party.

In many cases, the variables of which the performance is a function may be unavailable to the individual, as, for example, when repression, reaction formation, or denial are involved. The child who spills milk may do so as a result of his parent's counter-aversive control, but under these conditions, the child is not likely to be able to report the reason, particularly if the parental environment also included a history of severe punishment of aggressive behavior.

The child's verbal behavior about his own response tendencies will tend to be absent to the extent that there are no specific contingencies maintaining them. We constantly ask a young child: Why did you spill the milk? What are you doing? Did you put your toys away? Here, again, the verbal behavior about the child's own behavior occurs because of reinforcing contingencies supplied by the community. Because of the high probability of punishment when the child "explains" why he spilled the milk, variables other than the private event also control the behavior, and the child may give an incorrect account of his disposition to act. For example, in Freudian "Reaction Formation," the statement, "I love my brother," is not a result of a strong disposition to be reinforced by the brother's behavior or a way of providing reinforcing consequences for the brother. Aggressive acts toward the brother may have been punished severely, so that they generate conditioned aversive effects. As a result, any performance which is incompatible with aggressive action toward the sibling will be strengthened by reduction or removal of the conditioned aversive consequences resulting from the history of punishment of aggressive behavior toward the sibling. In this case, the verbal response, "I love my brother," is based on its effect in the verbal community of avoiding or escaping from aversive consequences rather than being a tact based on a repertoire in which the brother is a strong positive reinforcer. The lack of self-knowledge in this case occurs because the relevant verbal responses are suppressed due to the conditioned aversive stimuli they generate.

Many responses and even repertoires may occur without self-knowledge, even in the adult repertoire. In many cases, the individual may not even be able to identify the specific performances which comprise the repertoire, as, for example, in sports. Mathematicians commonly cannot describe the methods by which they develop theoretical proofs. The expert golfer usually cannot describe, the effective elements in his repertoire responsible for his success in drives of the ball.

Behavior about Past Performances. These are verbal responses which are based on past behavior. The discriminative stimulus is an item of the individual's past performance, and the verbal response is reinforced if it is of the same form as a response which would have been emitted by a direct observer of the behavior in question. The community encourages this behavior continuously when it asks questions such as "Where did you leave your toy? How did things go at work today? What was the movie about that you saw last night? The possibility of this kind of control depends upon an extensive self-descriptive repertoire. Only if the individual can describe his own behavior fairly continuously is he likely to be able to recount it. The process of recounting the day's activity is essentially repeating the verbal responses emitted earlier as the individual described his own performances as they occurred.

Responses to Potential Behavior. In many cases, the community is interested in the potential repertoire of an individual, as, for example, when it inquires, "Can you bake a cake? Do you know how to adjust a carburetor?" Here, again, a distinction must be made between the actual repertoire and the ability to emit a tact based on the potential ability to emit the behavior. The stimulus for which the response, "I can fix a carburetor," is a tact is a private event. In this case, however, the community has close control over the verbal behavior because it can observe the repertoire and can punish the individual if, in fact, the individual cannot fix the car.

Prediction of Future Behavior. Another type of private event involves the behavior of the individual which describes his own response tendencies. The community asks the individual to predict what he will do. "I am really going to dig in and master this thing. I will fix the faucet tomorrow." Both of these verbal episodes involve private events, and the development of the performance is again via the principles described above. These tacts of the individual's response tendencies are useful to the community, and, hence, they are systematically developed in the child by generalized reinforcement.