Chapter Three

Respondent Conditioning and Counterconditioning


Someone smiling at us produces a pleasant feeling. Pictures of good food may literally cause our mouths to water. In one type of fetishism a man is sexually aroused by the sight of a woman’s shoe. A woman with an automobile phobia may become anxious when she sees a car. Why should these stimuli (smiles, pictures of food, women’s shoes, automobiles) elicit these particular responses (a pleased feeling, salivation, sexual arousal, anxiety)? It is not instinctual that these stimuli elicit these responses; hence it probably is learned.


Perhaps one reason a smile now elicits a pleased feeling is that in a person’s learning history the stimulus of a smile was associated with other stimuli, such as affection, which produced a pleased feeling. The stimulus of the image of the food was associated with the stimulus of the taste of the food, with the taste eliciting salivation. Eventually the image of the food came to elicit salivation. Similarly, the sight of a woman’s shoe may have been paired with sexually arousing stimuli such as from masturbation. The image of an automobile may have been paired with an anxiety-producing stimulus such as seeing a close relative die in an automobile accident. The learned associations may have been gradually built up over time, as in the case of the smile and affection, or may have followed a single dramatic learning experience, as in the case of the automobile accident.


This type of learning is called respondent conditioning, the learning model in which one stimulus, as the result of being paired with a second stimulus, comes to elicit a response it did not elicit just previously. Usually this new response is similar to the response previously elicited only by the second stimulus. In this model the first stimulus is called the conditioned stimulus (CS) and the response it comes to elicit is called the conditioned response (CR), while the second stimulus is called the unconditioned stimulus (UCS) and the response it already elicited is called the unconditioned response (UCR). Figure 1 illustrates this for the case of the child who is gradually developing a dislike for school (CS) because the teacher emphasizes the use of corporal punishment (UCS), which makes the child anxious and fearful (UCR). I have had a couple of cases in which this was the first step for the children developing school phobias.



Numerous theories account for respondent conditioning (see Mikulas, 1974b, p. 98); but the following is satisfactory for our present purposes. Through association of the CS and UCS, the CS comes to provide information about the occurrence of the UCS. The more probable it is the UCS will follow the CS, the stronger the respondent conditioning and the more probable it is the CR will follow the CS. After the CR begins to occur, it may be rewarded or punished (Chapter 7), which affects its occurrence. In this sense the CR is often a response the person makes to prepare for the UCS. I was in a restaurant in which the lights would dim (CS) just before you would feel a blast from the air conditioner (UCS) because of the high energy requirements for the onset of the air conditioner. The scantily clad waitresses would shiver or cover their shoulders (CR) automatically when the lights dimmed. Respondent conditioning is often called classical conditioning and sometimes Pavlovian conditioning, although there is ambiguity in these terms (Hebb, 1956).


It is important to realize how prevalent respondent conditioning is in human behavior, particularly as it relates to emotional affect. Consider all the things that please or displease you and all the situations that elicit emotions such as affection, sexual arousal, anger, anxiety, or frustration. Consider how you differ in these areas from other people you know or people from different times or different cultures. Although some of these responses may be largely innate, such as some reactions to physical pain, most of these reactions are learned, primarily by respondent conditioning. Respondent conditioning in humans often is mediated by language. If a friend tells you that George is a “racist,” then perhaps some of your emotional affect to the word “racist” will become associated to your image or memory category of George. Respondent conditioning is often complex, involving more than associations between one particular CS and one particular UCS. Thus a person may have negative feelings around older, male authority figures based on experiences with his father, two elementary school teachers, and a local policeman. Or a person may have bad associations to bars in one part of town based on one personal experience, stories from friends, and newspaper accounts. A person’s present reactions to certain situations may be based on such a complex set of experiences that the person cannot readily remember them, and they may be extremely difficult to trace back historically. Fortunately, since behavior modification is ahistorical, this is not necessary. Rather, we would determine the person’s current reactions to specific situations and use our knowledge of respondent conditioning to change undesired reactions.


Let us consider a few more examples. In a classic study Watson and Rayner (1920) made a loud noise behind 11-month-old Albert whenever he reached for a white rat. This noise (UCS) was frightening (UCR) to Albert and resulted in a fear (CR) of rats (CS), as well as of other furry objects. A woman had two painful childbirths, so that she became anxious when she learned she was pregnant again. Some teenagers take up smoking tobacco even though the initial reaction to smoking may be aversive. Here the associations to smoking, through sources such as peers and advertisements, make smoking desirable. Chapter 7 deals with the effects of rewards (reinforcement) and punishments on behavior. In human behavior most of the things that are rewarding (e.g., attention, approval, money, good grades) or punishing (e.g., Ostracism, criticism) acquired their affect through respondent conditioning and are called conditioned reinforcement and conditioned punishment (see Mikulas, 1974b, p. 1 03).


Three important variables affecting respondent conditioning are (1) temporal order of the stimuli, (2) interstimulus interval, and (3) response dominance. Temporal order refers to the fact that you generally get the best Conditioning if the CS precedes the UCS (forward conditioning); while you generally get little or no conditioning if the UCS precedes the CS (backward Conditioning). Early attempts at treating alcoholism involved making the Person sick and then having him drink an alcoholic drink. This is backward Conditioning, which probably decreased the effectiveness of the program.


Backward conditioning, however, does work in some situations, such as those discussed below under response dominance; but the best approach is to establish forward conditioning. Also in some situations, the CS and UCS occur together, not one before the other. This may be inevitable, as when you cannot separate a person from one of his characteristics, and may readily produce respondent conditioning.


Assuming forward conditioning, the interstimulus interval, refers to the amount of time from the onset of the CS to the onset of UCS. Generally, with many exceptions, you usually get the best respondent conditioning with an interstimulus interval of about one-half second. Although conditioning may occur with much longer intervals (e.g., nausea from food poisoning may condition to the taste of the food that occurred many hours earlier), in most behavior modification programs, one-half second seems optimal.


Before conditioning, both the CS and the UCS elicit responses. How ever, people seldom list the initial response to the CS, for it is usually relatively minor in importance to the UCR and the later CR. Response dominance refers to the relative strengths of the responses eIicited by the CS and UCS before they are paired, the relative strengths of R1 and UCR in Figure 2. Now what happens when we pair the CS and the UCS? If R1 and UCR are compatible, both stimuli may come to elicit both responses. If the responses are incompatible, the conditioning will tend toward both stimuli eliciting the dominant response. Since the UCR is usually dominant to R1, R1 is usually not even mentioned and we have the result shown in Figure 1. However, if R1 were dominant to the UCR, the conditioning may go the other way, backward conditioning. If you have a fear of snakes and I throw a snake in your lap and tell you to relax, I will probably get anxiety associated with the word “relax” rather than relaxation associated with snakes. Response dominance is the critical component to counterconditioning, which is discussed later. Although the concept of response dominance is important in respondent conditioning, there is currently little research on it.



Other variables of respondent conditioning include that the more times the CS and UCS are associated (CS predicts UCS) the more the learning; within limits you often get better learning with a strong UCS than a weak UCS; and you get poorer learning if you have the learning trials (CS-UCS pairing is a trial) too close together.




Establishing a new response by respondent conditioning is often part of a behavior modification program. Consider a person who is not sexually aroused by what he considers desired heterosexual cues, but is only aroused by cues he considers undesirable, such as an unusual sexual fantasy or homosexual stimuli. If such a person is not offended by masturbation, we may gradually pair the desired heterosexual cues (e.g., imagined scenes, photos) with the sexual arousal associated with masturbation (e.g., Marquis, 1970). We would probably begin with the stimuli that already elicit sexual arousal and gradually change from these to the new desired stimuli. Then as the person later engages in sexual behavior in the presence of the desired stimuli, the natural forms of respondent association will take over. The respondent conditioning here is used to overcome an initial obstacle, get things started, and turn the process over to the regular course of events.


Enuresis (bedwetting) is a problem affecting many children at all ages, including about 10 percent of six-year-old children. In addition to being a problem by itself, it also leads to other problems such as anxiety and guilt. A possible component of some enuresis is that the child has a small bladder capacity and poor development of related muscles so that the child urinates more both during the day and at night (Yates, 1975, chap. 3). Treatment then often involves teaching the child to have a larger bladder capacity by rewarding him during the day for going for longer and longer times without urinating. In a few cases the enuresis is because of excessive anxiety in the child and reducing the anxiety eliminates the enuresis.


But the most common approach for dealing with enuresis is the bell and pad or urine-alarm procedure (Mowrer & Mowrer, 1938). The logic is that internal cues of the bladder and related muscles are too weak to awaken the child at night so that he may go to the bathroom before wetting the bed. Treatment uses a specially constructed bed pad that when moistened by urine sounds a bell or buzzer and wakes the child up. (Such devices are sold by Wards and Sears.) The child now also begins inhibiting more urination as the bell rings. Since increased muscle tension (CS) precedes urination and bell (UCS), which wakes up the child and inhibits more urination (UCR), then by respondent conditioning eventually the muscle tension alone will wake the child and inhibit bedwetting. Thus the child is taught to respond to the internal cues that most people use to control urination. This is a fast and relatively effective procedure by itself and couples well with training in increasing bladder capacity. It has also been successfully used with enuretic adults (Turner & Taylor, 1 974). There are other explanations for how the bell and pad procedure works (see Doleys, 1 977; Lovibond, 1 964), such as the Child learning to wake up to avoid the bell or buzzer.


Azrin, Sneed, and Foxx (1 974) improved on this urine-alarm method by adding training in inhibitory control, rewards for correct urination, training in rapid awakening, increased fluid intake to increase the response rate, self-correction of accidents, and practice in toileting. They reported significantly reducing bedwetting after one night of such intensive training.


A spectacular example of respondent conditioning is Efron’s (1957) report of treating an epileptic. Most epileptics can detect the onset of a seizure by a subjective aura that precedes the seizure. Efron found that one of his female clients could inhibit her seizures by inhaling the odor of jasmine during the early stages of the aura. Efron then respondently conditioned the smell to the sight of a bracelet. Then she could inhibit the seizure by staring at the bracelet. Eventually, just thinking about the bracelet could inhibit seizures. Interestingly, looking at or thinking about the bracelet also elicited a subjective sense of smelling jasmine. In terms of latency of effectiveness, the direct odor was faster than seeing the bracelet, which was faster than thinking about the bracelet. Eventually, the client would just have spontaneous experiences of the odor of jasmine; but by then she was no longer having any seizures. This suggests the whole process moved further back so that pre-seizure cues triggered off the sense of jasmine and inhibited seizures. (This is a common sequence in many self-control programs in which initially conscious components eventually slide out of consciousness.) Finally, the occasional smell of jasmine disappeared and there were still no more seizures. The generality of this case study to other epileptics needs considerable more research.


Although there are situations, such as those above, in which a new response needs to be established by respondent conditioning, more often in behavior modification it is a matter of changing or eliminating an undesired behavior. In respondent conditioning there are two ways of dealing with undesired behaviors: extinction and counterconditioning.




Respondent conditioning is accomplished by establishing a contingency (relationship) between the CS and the UCS: the CS predicts to a certain degree the onset of the UCS. If we terminate this contingency so that the CS is not associated with the UCS, eventually the CS will no longer elicit the CR. This process is called extinction. If a small child is scratched (UCS) by a cat (CS) and hurt (UCR), then the child may develop a fear (CR) of cats. If the child now encounters cats without anything bad happening, the fear may extinguish. Sometimes following extinction, the CR may gain in strength over time. This is spontaneous recovery. However, in practical situations, this is usually minimal; and with further extinction the CR will no longer reappear.


There are basically two ways of carrying out extinction: gradual and not gradual. The gradual approach consists of moving through a sequence of steps, called a hierarchy, toward the object or situation that elicits the strongest CR. The alternative is to bypass most of these intermediate steps and confront the final situation right away. (Actually these are not two different approaches, but two points on a continuum of how many steps there are until approaching the final situation.) For example, if a child had a fear of the water at the beach, a gradual approach would involve slowly approaching the water, perhaps first playing on the beach 20 feet away from the water, then playing 1 0 feet away, then at the edge of the water, then putting feet in the water, and so forth. The non-gradual alternative may be to put or carry the child into the water until the fear extinguishes.


Although extinction is applicable to any respondently conditioned response, it is most used with anxieties and fears. People are continually confronted with situations that elicit some anxiety, such as standing up to the boss, making a presentation before a class, or talking about something personal. If the person can approach and be in the anxiety situation without anything unpleasant happening, then some of the anxiety should extinguish. The following are some general guidelines for using a gradual approach to the extinction of anxiety and fears: First, it is necessary to establish a hierarchy of steps toward the feared object or situation. It is generally better to have too many steps than too few. Second, it is desirable to encourage, motivate, or reward the person for going through the hierarchy. However, the person should move through the hierarchy at a comfortable pace, extinguishing most of the anxiety at each step before moving on. Finally, it is often useful to provide the person with a way to reduce the anxiety while all this is taking place—perhaps by teaching the person how to relax, having the person imagine pleasant scenes, or having the person pretend to be someone who would not be anxious in this situation. These aids, plus any rewards the person receives, help to reduce anxiety, provide motivation, and produce some counterconditioning, as discussed later. A good variation of the above is to first have the person gradually go through the hierarchy of steps in his imagination and then in real life, the latter called in vivo.


The non-gradual approach to extinguishing fears involves immediately confronting the feared situation. If a child learning to ride a bicycle falls off and hurts himself, his parent may have him get right back up and try again. If a person feels anxious about dancing in front of others in a nightclub, he may force himself to get up and do it. This approach often works, but sometimes the resulting anxiety is too great and the person ends up more anxious rather than less anxious. Therefore, my bias is to generally favor the gradual approach, which although slower is also safer. A variation of the non-gradual approach involves bombarding the person with the anxiety-producing stimuli and/or keeping the person in the anxiety situation without escape. This approach is called flooding and will be discussed in Chapter 4.


As will be seen throughout this text, most behavior modification procedures can, to some degree, be carried out in the imagination (see Chapter 9). Extinction carried out in this way is called covert extinction (Cautela, 1971). There is currently little research in this area, but the following is an example: Götestam and Melin (1974) used covert extinction with four female amphetamine addicts (who were mainlining 100—200 mg., 3—5 times per day). They had the clients imagine situations in which they would inject themselves and had them imagine they felt no effect from the drug. Following one week of this treatment, about 100 trials, there was a decrease in the effect of the drug, even to the extent of the clients getting no effect when actually taking the drug. At a nine-month follow-up, three of the four women were not using amphetamines.


Respondent extinction can be seen to be a critical component of many therapeutic or change programs, although it is not conceptualized as extinction. Therapists, while maintaining a non-judgmental or permissive attitude, may encourage their clients to recall or discuss emotionally laden ideas or memories. Psychoanalysts, scientologists, and primal scream therapists may encourage their clients to recall and relive early painful experiences, perhaps real or unintentionally fabricated to suit the theory. People who feel uptight about some aspect of their body may attend a weekend marathon in which everyone goes around nude and each person tells sympathetic listeners about a problem that brings on uptightness. Peer-evaluation counseling may involve two non-professionals sitting down and gradually telling each other more and more personal-emotional things. As a spiritual tool, Ram Dass (Richard Alpert) may say to people, “anything you can think, feel, desire, fear, anything you can bring to your mind about any of these, that you have difficulty with, are embarrassed by, are made uncomfortable by sharing with me—share it with me.” Some forms of meditation help people free themselves from emotional attachments by letting their thoughts and behaviors run their course while holding in a calm conscious space. My bias is that all these examples contain respondent extinction as a component and realizing this can facilitate whatever is to be accomplished.




Counterconditioning is the reduction of undesired elicited responses by respondently conditioning incompatible responses to the eliciting situations. The first step is to determine the situations that elicit the undesired responses, as spiders may cause excessive anxiety in some people. The second step is to determine or establish ways to elicit a response incompatible with and dominant to the undesired response, such as some forms of relaxation may be to the spider anxiety. Finally, the incompatible response is respondently conditioned to the stimuli eliciting the undesired response, as stimuli producing relaxation may be paired with stimuli related to spiders. This counterconditioning is continued until the undesired response has been adequately reduced, usually until it no longer occurs.


Counterconditioning is often used to reduce unwanted emotional reactions such as anxiety, anger, or jealousy. Most clinical cases have an anxiety component that needs to be handled in some way. Desensitization, discussed in Chapter 5, is the counterconditioning of anxiety with relaxation. In other situations, the undesired response is a rewarding, approach response, as occurs in some aspects of alcoholism, drug-addiction, and over-eating. The sight of a bar may elicit a craving for a drink or the taste of one cigarette may lead to smoking another. In these cases, counterconditioning may involve conditioning in an unpleasant or aversive response to the stimulus situations eliciting the approach response. This is called aversive counterconditioning and will be discussed in Chapter 6.


It is important in counterconditioning that the incompatible response be dominant to the undesired response. Sometimes this is not a problem. For example, in aversive counterconditioning the aversiveness of electric shock or imagining unpleasant scenes may be dominant to the pleasing effects of having a second piece of pie. However, response dominance is often an issue. The way to insure the incompatible response is dominant is through the use of a hierarchy, similar to the gradual approach of respondent extinction. That is, rather than immediately starting with counterconditioning to the situation that most strongly elicits the undesired response, we begin with a situation which weakly elicits the undesired response. We then apply our counterconditioning to a sequence of situations, the hierarchy, that gradually approximates the situation which most strongly elicits the undesired response. In the case of a person with a fear of spiders, our counterconditioning using relaxation would begin with items low on the hierarchy (such as the word “spider”), work up the hierarchy through intermediate items (such as a picture of a spider), on to items at the top of the hierarchy (such as touching a live spider). The assumption is that the effects of the counterconditioning generalize (carry over to similar stimuli) up the hierarchy, thereby gradually reducing the strength of the undesired response elicited by the various situations. In the example of the spider anxiety, it may be that at the beginning of treatment the anxiety elicited by a picture of a spider or touching a live spider is dominant to any relaxation we can produce. But our relaxation is dominant to the anxiety elicited by the word “spider”; so we begin our counterconditioning there. Now as we countercondition out the anxiety to the word “spider” it is assumed the counterconditioning carries up the hierarchy and reduces somewhat the anxiety to the picture and the live spider. By the time we get to the picture, our relaxation is dominant to any remaining anxiety, which we can now countercondition out. And this counterconditioning generalizes up the remainder of the hierarchy. Thus if we choose a hierarchy of related items, have a sufficient number of items in our hierarchy, and do not move through the hierarchy too fast, we can insure that the incompatible response is dominant to the undesired response and Counterconditioning will move in the desired way. This approach will be seen in greater detail when discussing desensitization (Chapter 5), but it is important to remember it applies to all counterconditioning.


Counterconditioning is not simply replacing the undesired response with the incompatible response. Rather, it is a matter of moving along a continuum from the undesired response toward the incompatible response. Counterconditioning may be stopped anywhere along the continuum with the usual stopping point being some neutral middle point. Thus with the spider anxiety we would probably countercondition until the person felt neutral (not anxious, not relaxed) toward spiders, although we could countercondition less and leave some anxiety or countercondition more so we get relaxation. Aversive counterconditioning may be part of a program with a female homosexual to reduce sexual arousal to females. We would probably stop at the point the client felt neutral to females rather than continuing until she felt aversion. On the other hand, if we had a male client who would go to jail if he exposed himself at a playground again, it may be desirable to continue the counterconditioning until aversion is elicited in this situation.


Although the explanation of counterconditioning used in this discussion is practical for behavior modification, on a theoretical level it should be considered highly tentative until considerable more research is done related to the basic assumptions. Guthrie (1935) was one of the first major theorists to consider counterconditioning, and his conceptualization was similar to how counterconditioning is described above. Wolpe (1958) introduced the concept of reciprocal inhibition, borrowed from physiology, to account for counterconditioning, primarily desensitization. The assumption is that counterconditioning generally involves one part of the nervous system physiologically inhibiting another part, as desensitization may involve the parasympathetic nervous system inhibiting the sympathetic nervous system. Critics (e.g., Wilson & Davison, 1971) of this concept of reciprocal inhibition to account for counterconditioning suggest that the concept is an unnecessary addition to our explanation and not well supported by physiological research. However, “reciprocal inhibition” is a common expression in behavior modification related to counterconditioning.


In both respondent extinction and counterconditioning, the client is presented with the CS or a gradual sequence of approximations to the CS. Procedurally, the only difference between the approaches is that in counterconditioning we present stimuli or training that leads to an incompatible response to be associated with the CS. This similarity between the two approaches allows some theorists to assume they are the same. Thus it is possible to argue that counterconditioning is simply extinction. The purpose of the incompatible response is for motivation and/or to facilitate the extinction process. Desensitization, as the most common example, may be seen as extinction rather than counterconditioning (e.g., Wilson & Davison, 1971). The role of relaxation is a way to motivate the client to work through the hierarchy and/or reduce anxiety so the client is in a better state for extinction.


On the other hand, my bias and the position of theorists such as Guthrie is that respondent extinction is counterconditioning. The assumption is that extinction is not the passive weakening of a behavior, but the learning of new, perhaps incompatible, behaviors. If this position is correct, then in counterconditioning you are providing the client with an incompatible behavior, while in extinction you are relying on the incompatible behavior occurring some other way. Hence counterconditioning is potentially more effective.


But whatever the theoretical explanation for counterconditioning and how it differs from extinction, it is clear what to do procedurally in applied situations. Next I turn to relaxation, the most commonly used response for counterconditioning anxiety.




Teaching a client how to relax is often a powerful and needed therapeutic approach just in itself, for our culture provides more and more potential sources of stress and anxiety; and few people ever learn effective ways to relax. Thus many people report they are often anxious or uptight. This may be associated with specific fears, a racing mind, or inability to get to sleep easily. Also each year more research (e.g., Holmes & Masuda, 1972; Seligman, 1975) relates many physiological problems, such as colds, ulcers, and cancer, to the stress a person experiences and how the stress is handled. Thus relaxation training is a common part of many programs. It is often useful to introduce this training early in clinical sessions because it calms the client down, shows him you have some powerful tools at your disposal, and gives the client a sense that there are things he

can do about his own behavior.


Thus several programs have been geared toward teaching people how to relax and use this as a self-control skill (Goldfried & Trier, 1974; Mikulas, 1976a; Sherman & Plummer, 1973). The people learn how to sense when they begin to feel anxious and learn how to relax instead. They learn how to handle stressful situations, as well as how to reduce specific problems such as tension headaches, nervous muscle movements, and anxiety-produced distortions in thought and perception. Learning to relax is also an important component in treating insomnia (Borkovec et al., 1975; Nicassio & Bootzin, 1974). For these reasons, the best type of relaxation procedure is one in which the client learns a skill of relaxing rather than has something done to him to make him relax (e.g., drugs).


The most-used relaxation training in behavior modification is some shortened variation (see Bernstein & Borkovec, 1973) of the muscle relaxation procedure developed by Jacobson (1938). This involves the client alternately tensing and relaxing various muscles while focusing his attention on the different feelings. With instruction and practice, the client learns how to relax himself “at will”; identify the onset of stress or anxiety earlier, which facilitates self-control; and generally become more aware of the muscles in his body, which may reduce a variety of problems, such as unconscious chewing on the tongue, headaches resulting from muscle tension in the neck and head, excessive wrinkling of the face, and poor posture.


A variation on this relaxation training is cue-controlled relaxation (Russell & Sipich, 1973) in which the person, while relaxed via muscle relaxation and focusing on his breathing, associates a word such as “calm” or “control” with the relaxed state. The person can then use this word to help cue in the relaxed state.


There are many other ways of training and producing relaxation (White & Fadiman, 1976) that may be useful with different clients, problems, or situations. Hypnosis, in addition to facilitating relaxation, can also be used to increase the client’s motivation for parts of the program and may improve visualizing scenes in the imagination, if this is part of the treatment. On the other hand, hypnosis varies in effectiveness with clients and involves many dangers for the practitioner lacking substantial training. Autogenic training (Lindeman, 1973; Luthe, 1969) is a form of relaxation training similar to self-suggestion. It involves giving yourself such suggestions as “My right arm is heavy,” “My heart beats calm and regular,” and “My forehead is cool.” Autogenic training packages well with biofeedback. Biofeedback (Chapter 7), as it relates to relaxation, involves the use of machines to tell clients how anxious or aroused they are in terms of some physiological measure such as muscle tension, skin resistance, or skin temperature. This biofeedback information helps the clients learn to relax and is particularly useful when a set of muscles requires specialized attention. Meditation is being used more in behavior modification (e.g., Berwick & Oziel, 1973; Boudreau, 1972); and besides facilitating relaxation it is also useful when the client needs to learn to calm his mind (e.g., insomnia, racing mind) or just generally get a little more perspective on his life.


Finally, drugs such as tranquilizers or breathing a mixture of oxygen and carbon dioxide (Ley & Walker, 1973) may be used to produce relaxation. Ideally, they would only be used as a temporary adjunct to the treatment program and would gradually be phased out. But however relaxation is produced, once a person is relaxed or can relax on cue, we can use relaxation to countercondition anxiety.




Anxiety is the most common emotional reaction that is counterconditioned; relaxation is the most common incompatible response used to countercondition anxiety; and desensitization is the most common way to countercondition anxiety with relaxation. However, many other incompatible responses, elicited or facilitated by the practitioner, may be used to countercondition anxiety, such as laughter, assertive behavior, anger, music-elicited responses, eating, emotive imagery, and aversion relief. An incompatible response other than relaxation may be chosen because it is difficult to get the client to relax, the other response is already strong in the client’s repertoire, or the other response is one that independently needs to be strengthened.


Laughter is a good response for counterconditioning anxiety, if the person can learn to laugh at himself or the situation. At first, it may be necessary for the client to alter the situation in his imagination to facilitate making it seem humorous. Ventis (1 973) described a case of a coed who was anxious about attending a banquet at which she would encounter her ex-boyfriend. A humorous image involving the boyfriend was used for counterconditioning. After one session and a brief hierarchy she was able to attend the banquet that evening with little discomfort.


Assertive behavior is often used to overcome anxiety, for many people are anxious in situations in which they are unassertive. Learning to be appropriately assertive (see Chapter 8) may be an important part of social skill training with the client, as well as being a source of behaviors that will countercondition anxiety.


When using anger to countercondition anxiety we generally do not want to condition the person to feel anger in the anxiety situations. Rather we use the principle of counterconditioning that lets us stop at the intermediate neutral point. Goldstein and associates (1970) begin by having the client in the consulting room pair anger-arousing imagery, plus vocal and motor behavior, with imagined situations that elicit anxiety. Later, the client uses the anger-arousing imagined images for counterconditioning in vivo (real life) situations. The anger-arousing scenes can be used both for counterconditioning and as a self-control procedure for handling anxiety, these two functions often going together in counterconditioning. Butler (1975) used anger to countercondition a variety of fears in his client, including a fear of traveling more than about eight blocks from home (a form of agoraphobia, fear of open spaces). Treatment involved the client imagining an anxiety- provoking scene and responding with anger, angry verbalizations, plus vigorous muscular activity.


Lowe (1973) used the excitatory responses elicited by music to counter-condition anxiety in a client who had trouble learning to relax. The client was a former rock guitarist who would get excited by particular music and could augment the excitement by imagining such things as he was performing the music.


In a classic study, Jones (1924) used eating as one of the responses to countercondition a fear of rabbits in three-year-old Peter. The rabbit was introduced in a cage in the far part of the room where Peter played and ate. Each day the rabbit was brought closer (a hierarchy) until it was out of the cage and in Peter’s lap. This counterconditioning generalized to other furry objects that Peter had also been afraid of. (Jones was a student of Watson, whose conditioning of little Albert was mentioned at the beginning of the chapter.)


Emotive imagery is the counterconditioning of anxiety with images that arouse feelings such as pride, affection, self-assertion, or mirth. Therefore, it obviously overlaps with other counterconditioning approaches being discussed. So far it has primarily been used with children who are not easily trained to relax. With adults, emotive imagery may combine well with desensitization. The originators of the approach, Lazarus and Abramovitz (1962), treated a ten-year-old boy who feared the dark by using his passion for the radio series “Superman” and “Captain Silver.” Treatment involved the boy imagining a story involving himself, Superman, and Captain Silver. In the story he imagined himself in situations that gradually became darker. At the end of the third session he was able, without anxiety, to picture himself alone in his bathroom with all the lights turned off, awaiting a communication from Superman. This treatment eliminated the fear of the dark and was accompanied by an improvement in school work and a decrease in insecurity.


Aversion relief refers to the offset of an aversive event, such as the offset of electric shock. This offset, assumed to be pleasant, can be paired with words such as “calm” or used to countercondition anxiety by pairing the offset with an anxiety situation. This approach has not been used much, particularly with the number of preferable alternative approaches. It comes up the most in situations where the onset of an aversive event is being used for aversive counterconditioning (Chapter 6). A form of aversion relief is respiratory relief, which is based on the relief following holding your breath. Treatment of anxiety would involve having the person hold his breath and then begin breathing when presented with an anxiety situation (e.g., Orwin et al., 1975).


New approaches to countercondition anxiety continually come up in the literature, and creative practitioners will find many other ways that suit their particular cases. Kass, Rogers, and Feldman (1973) report several cases in which they counterconditioned anxiety with responses specific to the different individuals. One woman became distressed at being called names, which then led to fighting. Treatment involved her imagining herself laughing at the person calling her names. Another client was anxious about being alone, rejected, or in crowds. This was counterconditioned with a scene in which she was lying in bed looking at new drapes. And in another case, job anxiety was counterconditioned with sitting up exercises.




All of the logic and examples that apply to counterconditioning anxiety with relaxation also apply to the counterconditioning of other unwanted emotional responses such as anger, jealousy, or frustration. For example, Hearn and Evans (1972) used a desensitization-type approach in which relaxation was used to countercondition anger in student nurses. And Cotharin and Mikulas (1975) used a desensitization-type approach to reduce racially related emotional responses in high school students. Following this, white students could interact pleasantly with black students whom before they felt uncomfortable with and had to avoid.


Now relaxation is just one of many incompatible responses we can use to countercondition these unwanted emotional responses. We could use many of the incompatible responses mentioned above for counterconditioning anxiety. For example, Smith (1973) describes a case in which he used humor to countercondition anger. The client was a 22-year-old female who could not control her extreme anger responses with her husband and three-year-old son. The child’s misbehavior generally elicited extreme rage, including screaming, breaking things, and physically attacking the child. A hierarchy of situations was constructed, and the items of the hierarchy were greatly exaggerated to make them humorous. Counterconditioning with this hierarchy reduced the anger the situations elicited, gave the client control over her temper, and allowed her to view anger eliciting situations more objectively.


In general then, we can probably countercondition any unwanted, conditioned, emotional response (e.g., anxiety, anger, racially related emotional responses) with any incompatible response (e.g., relaxation, humor). Since we can stop our counterconditioning at a neutral point, it often may not matter whether the incompatible response is desirable or not. All this gives the practitioner enormous flexibility in tailoring a counterconditioning program to his client’s specific problems and skills. The choice of what incompatible responses to use for counterconditioning depends on which such responses the client already has in his repertoire or could readily learn and which responses would be profitable to learn for reasons more than counter- conditioning (e.g., relaxation, assertive behavior), My bias is that most, if not all, changes in affect that occur in any non-physiological treatment program, regardless of the theory behind the treatment, is because of respondent extinction and/or counterconditioning. Seeing it from this perspective may result in the practitioner being more aware of the need to strengthen or establish incompatible responses and perhaps use a hierarchical approach.


Masters and Johnson have determined more about the human sexual response and its dysfunctions than anyone else in history. Their program for the treatment of sexual dysfunctions (Belliveau & Richter, 1970; Masters & Johnson, 1970) is still the classic work in this field; and many other programs (e.g., Hartman & Fithian, 1972) have drawn on their work in varying degrees and added other treatment components. The Masters and Johnson treatment program has basically two overlapping parts, educative counseling and behavioral assignments. The educative counseling involves general sex education, discussion of motivation to change, discussion of basic problems, and instruction in the treatment approach. This is coupled with history taking, general assessment, and physical examinations. The behavioral assignments consist of programs, geared toward the specific type of dysfunction, that the couple carries out with each other in private.


Although Masters and Johnson conceptualize their treatment program as a form of psychotherapy, the behavioral assignments are much like behavior modification with counterconditioning playing a large role. A major part of most of the treatment consists in eliciting the sexual response in a non-coital situation, then gradually moving through a hierarchy of steps leading to coitus while maintaining the sexual response. The implicit assumption is that the sexual response will gradually countercondition the responses (e.g., anxiety) which are impairing the sexual response in coitus. Thus a man with impotence, often unable to achieve or maintain an erection sufficiently for coitus, will first be stimulated to erection in a non-coitus situation. Then while maintaining and restimulating the erection, the male is gradually moved through a sequence of steps leading to coitus. A similar approach is used with a female with orgasmic dysfunction, seldom or never experiencing an orgasm. First the woman’s sexual response is elicited in non-coitus situations. Then she moves through a sequence of stages toward coitus. Other things are involved in the treatment of these and other sexual dysfunctions, but counterconditioning seems to be a significant component.


Viewing the Masters and Johnson program from the behavior modification position suggests several ways the program can be improved (see Murphy & Mikulas, 1974). For example, in many cases anxiety is impairing the sexual response. Now Masters and Johnson are using the sexual response to countercondition the anxiety that is impairing the sexual response. This may often work, particularly with a good hierarchy. But sometimes the anxiety is too great. Many of the cases they report as failures are cases in which too much anxiety exists. This suggests the addition of something like desensitization to first reduce some of the anxiety. There is also a need for more individualizing of the hierarchies. For many people the first step in the behavioral assignment may be too anxiety producing. For others, the items in the hierarchy need to be geared more toward their specific problems. Masters and Johnson believe the sexual response will naturally occur if they can just remove the obstacles. A behavior modifier may go further and initiate or strengthen the sexual response to specific situations. This would probably involve respondent conditioning and perhaps the use of sexual fantasies, masturbation, vibrators, or pornography.


A case study by Davison (1968) will finish this section as it illustrates various aspects of counterconditioning. The client was a 21-year-old, unmarried male who was only sexually aroused by sadistic fantasies, involving torturing women, which he masturbated to about five times a week. He dated little and showed no interest in girls. The first step was for the therapist to argue against any disease interpretation of the client’s unusual behavior.


This is common as the behavior modifier often has to reduce his client’s fears about being inherently sick, abnormal, or evil. Davison’s next step was counterconditioning: The client was instructed to masturbate while looking at pictures of sexy, nude women, using his sadistic scenes to occasionally help initiate or maintain arousal and erection. From these pictures the client slowly moved along a sequence of pictures of women with more and more clothes on and finally to imagined situations from real life. Now Davison was helping and encouraging the client to begin asking out girls. Finally, aversive counterconditioning was used to reduce the sexual arousal to the sadistic fantasies. This involved associating in the imagination a sadistic fantasy with an unpleasant image, such as drinking from a bowl of urine and feces. All of this produced a decrease in sadistic fantasies and a positive sexual feeling toward girls. Six months after treatment ended, the client used what he had learned to return to sadistic fantasies for a while and then reverse it back. At this point he found he had no need for sadistic fantasies and was involved in a relatively vigorous program of dating.




Respondent conditioning is the learning model in which a stimulus situation comes to elicit a relatively new response or increase in response because of association with other stimulus situations. Formally, the conditioned stimulus (CS) comes to elicit the conditioned response (CR) because of the person learning that the CS is associated with (provides information about) the unconditioned stimulus (UCS), which elicits the unconditioned response (UCR). In most situations, respondent conditioning is best when the CS comes on about one-half second before the onset of the UCS and the UCR is dominant to the response originally elicited by the CS. Respondent conditioning is sometimes used in behavior modification to establish or strengthen a response, as in the treatment of enuresis or in building in sexual arousal to a situation. Undesired respondent behavior is changed by respondent extinction or counterconditioning, both of which may or may not be done gradually with a hierarchy of intermediate steps. Extinction consists in presenting the CS without its being paired with the UCS until the CR is suitably reduced. Counterconditioning consists in conditioning a desired response to the CS to gradually replace the undesired response, with a hierarchy often used to control response dominance. Practically, extinction and counterconditioning only differ in the degree to which the practitioner facilitates the occurrence of an incompatible response. Relaxation training, a useful procedure in itself, is also part of programs for the self-control and counterconditioning of responses such as anxiety. Anxiety can be counterconditioned with a range of incompatible responses, including relaxation, laughter, assertive behavior, anger, musicelicited responses, eating, emotive imagery, aversion relief, physical exercise, and sexual responses. Similar responses can be used for the counterconditioning of many other behaviors, including anger, jealousy, frustration, racially related emotional responses, and aspects of sexual dysfunction. Aversive counterconditioning is counterconditioning in which a response to an unpleasant stimulus, such as electric shock or imagining an aversive scene, is gradually conditioned to stimuli that elicit undesired, but pleasant approach behavior, as in parts of alcoholism or addiction to other drugs.




Define respondent conditioning, respondent extinction, and counterconditioning. What are the practical and theoretical differences among them? For each give a real life example you have observed.
Discuss the variables of respondent conditioning as they may relate to a case in which you are reducing anxiety with emotive imagery.
Describe at least five different ways to teach or produce relaxation in a client. For each give a situation in which this method would be the preferred approach.
Describe a hypothetical case in which you countercondition anxiety with a response other than one of those mentioned in the chapter. Why would you use this particular response for this case?
Describe a hypothetical case in which you use humor to countercondition jealousy.
Describe covert extinction using a hierarchy for the reduction of test anxiety.
Consider any of the things you value. To what extent may respondent conditioning have been involved in the learning of this value? Give possible examples. Does understanding part of the cause of the value affect the value itself? Why?




Bernstein, D. A. & Borkovec, T. D. Progressive relaxation training: A manual for the helping professions. Champaign, Ill.: Research Press, 1 973.

Eysenck, H. J. & Beech, R. Counterconditioning and related methods. In Bergin, A. E. & Garfield, S. L. (eds.), Handbook of psychotherapy and behavior change: An empirical analysis. New York: Wiley, 1971.

Salter, A. Conditioned reflex therapy. New York: Farrar, Straus & Giroux, 1 949. Capricorn paperback, 1961.