This article is for use only of students enrolled in GPS 520: Advanced Social Psychology for purposes associated with this course and may not be retained or further disseminated.

 Leary, M. R., & Miller, R. S. (1986).  Behavioral compliance and psychological change.  In M. R. Leary & R. S. Miller, Social Psychology and Dysfunctional Behavior: Origins, Diagnosis, and Treatment (pp. 166-178).  New York: Springer-Verlag Inc.

 Behavioral Compliance and Psychological Change

Underlying the social influence model discussed in the previous chapter is the assumption that, if people can be persuaded to change their beliefs or attitudes, changes in dysfunctional behavior should occur. Thus, the focus of the influence model is on factors that make the client more receptive to the counselor's interpretations, suggestions, and teachings.

 Although cognitive change can result in behavioral change, the opposite process also occurs: Inducing someone to change his or her behavior can, if certain conditions are met, result in cognitive change. As Madsen and Madsen (1972, p. 3 1) put it, "It is much easier to act your way into a new way of thinking than to think your way into a new way of acting." This simple fact suggests a useful mechanism for promoting psychological change. Simply stated, inducing a client to engage in new, more functional ways of acting can result in new, more functional ways of thinking and feeling. This process constitutes our focus in this chapter.

Attitude-Discrepant Behavior and Psychological Change

 Within experimental social psychology, the impact of behavior on cognitive change has often been studied within what is known as the "forced compliance paradigm." In hundreds of studies, subjects have been induced to say or do things that are inconsistent with their existing beliefs, and their resultant beliefs or attitudes have been assessed. Typically, people who engage in these counterattitudinal or attitude-discrepant actions subsequently change their attitudes so that they are more consistent with their prior behavior. (The "boring task" study, in which subjects were induced to lie to another subject [Festinger & Carlsmith, 1959], is probably familiar to most readers).

 Hundreds of studies over the past 25 years have demonstrated the forced compliance effect as researchers have attempted to identify the conditions under which the performance of attitude-discrepant behavior does and does not result in cognitive change. Although the phenomenon has proven to be quite complex, a few necessary conditions have been isolated. In particular, there is general agreement that in order for counterattitudinal behavior to result in attitude change the individual must feel responsible for his or her actions and their consequences. In other words, the person must have some degree of choice in performing the action and must have foreseen the consequences of the behavior (Wicklund & Brehm, 1976). As we will see, this caveat is important in applying counterattitudinal procedures to clinical practice.

 Lest the reader fail to see the relationship between the forced compliance research paradigm and the processes of counseling and psychotherapy, we should point out that many therapeutic techniques require clients to behave in ways in which they prefer not to behave. Clients may be urged to discuss distressing topics, engage in actions that produce anxiety or are inconsistent with their prevailing values, role-play positions with which they do not agree, abandon enjoyable vices, or otherwise engage in unpleasant, effortful, or attitude-discrepant tasks. For example, in marital therapy, feuding spouses may be told to convincingly argue one another's positions. In other cases, therapists may urge their clients to engage in atypical, counterattitudinal behavior by being assertive, more independent, or less authoritarian. Many clients seek professional help in order to relinquish dysfunctional but enjoyable habits such as smoking, gambling, or overeating. In each of these examples, clients are in a type of forced compliance situation-they are induced to behave in an attitude-discrepant fashion-and it is of considerable importance for clinicians and counselors to understand what is involved when they induce clients to do things they prefer not to do.

Theoretical Explanations of the Effect

 Before discussing the clinical implications of forced compliance research, let us briefly examine three theories that provide explanations of it. A full discussion of each of these approaches would go far beyond the confines of this chapter, so we will refer the interested reader to useful sources as we proceed.

 Cognitive dissonance theory. The earliest studies that examined the effects of counterattitudinal behavior on cognitive change were conducted as tests of cognitive dissonance theory (see Brehm & Cohen, 1962). According to this approach (Festinger, 1957), a state of dissonance is created when an individual has two cognitions (thoughts about oneself or the world) that imply the obverse of one another. For example, believing oneself to be highly intelligent is dissonant with scoring in the lowest quartile on an IQ test. Similarly, abhorring the Ku Klux Klan is dissonant with being best friends with the Grand Master of the local KKK. According to Festinger, the state of dissonance is aversive, thereby motivating the individual to avoid dissonance when possible and to reduce dissonance once it has been aroused.

 People may reduce dissonance by modifying their cognitions in ways that restore a state of consonance. For example, they can add new consonant elements or eliminate dissonant cognitions regarding the issue, such as when the friend of the KKK Grand Master justifies his friendship by concluding that what his friends do in their spare time is none of his business. Dissonance may also be reduced by changing dissonant cognitions into consonant ones. According to the theory, this is what happened in Festinger and Carlsmith's (1959) "boring task" study. To relieve dissonance, subjects who had told someone that the boring task was interesting changed their attitude so that their attitudes and behavior were no longer discrepant.

 According to the theory, people will not experience dissonance if they have no choice but to perform the behavior or are unable to foresee its consequences. From the perspective of the theory, it is simply not dissonance-arousing to do counterattitudinal things with the proverbial gun to one's head. For this reason, research has obtained less attitude change following counterattitudinal behavior the greater the pressures on the individual to perform the action. For example, subjects who are offered a large sum of money to behave in a counterattitudinal fashion typically experience less dissonance, and less subsequent cognitive change, than subjects who are paid less. With little justification for their actions, people experience greater dissonance and are more likely to change their cognitions to bring them in line with their behavior. The clinical implications of this fact will become clearer below.

 Self-perception theory. An alternative explanation of the effects of counterattitudinal behavior has been offered by Bem (1972). According to his self-perception theory, people sometimes come to know their personal attitudes and other internal states by examining their behavior and the conditions under which it occurs. When people do not have well-formed attitudes or beliefs in a particular domain, they infer their attitudes from their behavior, much in the same way that they infer other people's attitudes.

 According to self-perception theory, subjects in forced compliance studies seem to change their attitudes following counterattitudinal behavior not because they experience dissonance, but because they have inferred their attitudes from their behavior. Having told the waiting individual that the boring task was interesting, subjects in the Festinger and Carlsmith (1959) study inferred that they thought the task was interesting, for example.

 Like dissonance theory, self-perception theory emphasizes the importance of perceived responsibility. If people do not feel personally responsible for their behavior, they are unlikely to infer that they wanted to behave as they did or believed what they said. Thus, providing a large inducement to perform an action may lead to behavioral compliance but will not lead individuals to infer that they agree with their behavior.

 Self-presentation theory. In the quest for attitude-discrepant behavior, the typical forced compliance study induces subjects to cheat, lie, harm others, or otherwise appear immoral, unattractive, irrational, or incompetent to others in the experimental setting. In light of this, it is possible that the attitude change observed following counterattitudinal behavior reflects a self-presentational strategy designed to control how others view the subject (Tedeschi & Rosenfeld, 1981; Tedeschi, Schlenker, & Bonoma, 1971; chapter 5, this volume).

 When people think they are viewed in a socially undesirable fashion, they offer accounts in an attempt to lead others to interpret their behavior in more acceptable terms (Goffman, 1955; Schlenker, 1980). Imagine a subject who has lied in return for money, as in the Festinger and Carlsmith study. When subsequently asked his evaluation of the boring task, the subject has two options: He can either admit that the task was excruciatingly boring, in which case he clearly has lied to an unwitting person for profit, or he can maintain that the task was really not so boring after all, which portrays his actions in a more acceptable light. By espousing attitudes that are consistent with the behavior they have just performed, people may be viewed as consistent and honest rather than as inconsistent and deceitful. Thus, the self-presentation approach argues that what appears to be attitude change is really an interpersonal face-saving strategy.

 As with dissonance and self-perception theories, personal responsibility is central to the self-presentation formulation. People will perceive a need to save face in others' eyes only if they are seen as responsible for their undesirable behavior and its consequences. If one is forced to lie under penalty of death, there is no need to account for one's action by insisting that one believed the lie.

 Critique of the Three Approaches

Space does not permit a discussion of the enormous literatures that have been built around these three approaches, but a brief statement of their relative merits may be helpful. Although research exists that supports all three theories, support for the self-perception theory seems to be the weakest. The biggest problem with self-perception theory is that it applies only when the individual does not hold salient or important attitudes. As Taylor (1975, p. 13 1) observed, self-perception theory holds only under "conditions where one is asked one's attitude about an inconsequential issue."

 When attitudes or beliefs are strongly held, data tend to support predictions derived from dissonance theory (Fazio, Zanna, & Cooper, 1977; Green, 1974; Ross & Shulman, 1973; Woodward, 1972). However, this theory, too, has been criticized extensively and has undergone extensive revisions that have shifted its focus from cognitive inconsistency per se to a focus on behaviors that threaten one's self-concept. Nevertheless, dissonance theory has strong empirical support and provides a broad framework in which a great deal of behavior can be explained.

 Because the self-presentation approach is rather new, the verdict on its utility is still out. However, initial experimental results have been supportive. In a series of three studies, Schlenker, Forsyth, Leary, and Miller (1980) provided strong evidence that attitude change following counterattitudinal behavior serves self-presentational functions. For example, "dissonance-like" effects were obtained only when subjects expressed their attitudes to others who had observed their counterattitudinal, socially undesirable actions. When subjects expressed their attitudes to those who had not observed their actions, no effects were obtained, strongly suggesting that subjects strategically modified their attitude statements in order to "save face" (see Tedeschi & Rosenfeld, 1981).

 Clinical Applications

 We commend the reader's patience during this whirlwind tour of the social psychological literature dealing with the effects of counterattitudinal behavior. Armed with these theoretical perspectives, we can now turn our attention to the implications of this phenomenon for clinical and counseling practice.

 Role-Playing

 Many counseling and clinical psychologists use role-playing techniques to promote change in their clients. These techniques are commonly used in cases in which the client is in conflict with another individual (such as marriage counseling or organizational conflict resolution) and the therapist wants the client to understand better the perspective of the "opponent." Alternatively, the client may be asked to role play himself or herself in situations that are distressing or threatening, such as those requiring assertiveness. A considerable body of research has demonstrated the effectiveness of role playing in these contexts (Janis & Gilmore, 1965; Johnson, 1971; Muney & Deutsch, 1968). For example, in dyadic problem-solving and negotiation settings, people who role-play their opponent's position change their attitudes more than people who negotiate without role reversal or who only listen to their partner engage in role reversal (Johnson, 1967, 1971). Although it is clear that role playing promotes cognitive change, researchers and practitioners disagree regarding the best explanation of role-playing effects.

 Early role-playing studies showed that improvising speeches in support of positions with which one disagrees produces more attitude change than either listening to a persuasive speech or reading a prepared speech (Greenwald & Albert, 1968; Janis & Gilmore, 1965; Janis & King, 1954; King & Janis, 1956). The original explanation offered for these findings posited that attitude change occurs during active role playing because, in the process of improvising their speeches, people are brought into greater "cognitive contact" with opposing positions. The role-playing individual thinks up and uses those arguments that he or she finds most convincing, and thus generates persuasive arguments that are tailored to himself or herself. In essence, role playing helps convince the individual of the correctness of the other's position (Elms & Janis, 1965; Janis & Gilmore, 1965).

 One corollary of this explanation is that the greater the incentive offered to the individual to role-play, the better the arguments he or she should generate (~ la reinforcement theory) and the more attitude change should occur. However, as we have seen, greater inducements to perform attitude-discrepant actions typically result in less rather than more attitude change (see, however, Janis & Gilmore, 1965). According to cognitive dissonance theory, this is because the effects of role playing depend upon the generation of dissonance, which is minimized by external inducements that decrease personal responsibility. Within marriage therapy, for example, dissonance theory would advise that clients should perceive that they are role playing by choice and should be urged to improvise their own roles rather than rely on prompts from the therapist or spouse. The greater the perceived responsibility for one's behavior in a role-playing session, the greater the dissonance that is aroused; the realization that "I am freely arguing for positions with which I don't agree" produces maximum dissonance. And greater dissonance results in greater cognitive change toward the counterattitudinal position one has just argued.

 Self-perception and self-presentation theories make precisely the same recommendation, although for different reasons. To the degree that people infer their attitudes from their behavior only when their actions are not controlled by situational factors (Bem, 1972), self-perception theory suggests that the woman who role-plays her husband's position during marriage counseling will infer she endorses his position only to the extent that she argues it freely. Of course, in most cases in which role reversal is used in therapy, Bem's precondition that existing attitudes be weak or ambiguous seldom is met. Spouses who seek counseling, for example, usually enter treatment with strong, polarized positions. In such situations, self-perception theory has limited or no utility.

 According to the self-presentation approach, role reversal results in attitude change because the individual finds himself or herself in a self-presentational dilemma. Having convincingly argued his wife's side in a marital dispute, a husband will later have great difficulty trying to maintain that he does not understand her position or that it has absolutely no merit. Role playing may put people in the position of having to concede to others' views or risk appearing inconsistent and ridiculous.

 In brief, although these three theoretical approaches explain the effectiveness of role playing in different ways, they agree that external pressure on the individual to role-play effectively should be just sufficient to get him or her to participate in the exercise. Stronger pressure will, according to all three theories, lead to less cognitive change.

 Effort Expenditure by the Client

 All successful therapy requires clients to exert effort. Not only must the client go to the trouble to contact a psychologist, travel to his or her office, and give up time from other activities, but therapy itself exacts certain costs in terms of effort, distress, and, usually, money. In light of this fact (and, often, to keep the client from terminating treatment), therapists sometimes try to minimize the effort involved. They may rearrange schedules to suit the client, minimize the discomfort produced by the therapy session itself, maintain flexible payment systems, and so on. Interestingly, each of the theories discussed above argues against this approach, suggesting instead that effort expenditure by the client is therapeutically beneficial and that the amount of effort exerted by the client is positively related to psychological change.

 For the cognitive dissonance theorist, dissonance is created when people freely expend considerable effort on a task that they recognize may not be worth the cost. People may come to evaluate the task more positively as a way of justifying the effort and energy they have devoted to it, thereby reducing dissonance. As Festinger (1961, p. 11) noted, people "come to love things for which they have suffered. " In an early demonstration of this effort justification effect, Aronson and Mills (1959) showed that female subjects who underwent a severe "initiation" in order to participate in a boring group discussion rated the discussion and its participants more favorably than subjects who underwent a mild "initiation" or none at all (see Gerard & Mathewson, 1966; Zimbardo, 1965). Having freely exerted considerable effort to join the group, subjects would have experienced dissonance had they admitted to themselves that the experience was not worth the effort.

 Self-perception theory also predicts that increased effort enhances one's liking for an activity. According to this approach, people will be likely to infer that they like and are committed to activities for which they expend a great deal of effort.

 Self-presentation theory maintains that people become more attracted to effortful activities because they want to justify their effort to others, rather than to themselves. Having publicly expended a great deal of effort, energy, time, or money in a particular endeavor, people may be compelled to justify their actions to others, and may do so by enhancing their endorsement of the activity (Alexander & Sagatun, 1973; Schlenker, 1975b).

 Whatever the best theoretical explanation of the effect (and there is evidence that all three processes may operate), effort justification plays an important role in psychological change. In fact, Cooper and Axsom (1982) suggested that effort is a necessary, if not a primary ingredient of successful psychotherapy. In the course of exerting effort, the goals of therapy become more attractive, and the client becomes increasingly committed to the therapeutic objectives and involved in therapy itself. Cooper and Axsom reviewed earlier studies suggesting that, in some cases, the exertion of effort per se is sufficient to promote positive change, regardless of the specific activity on which effort is exerted. For example, subjects in control groups who engage in effortful but non-therapeutic tasks often improve as much as subjects who undergo standard psychotherapeutic techniques (e.g., Marcia, Rubin, & Efran, 1969; Sloane et al., 1975).

 Taking the effort justification idea a step further, Cooper and Axsom (1982) suggested that any activity may have therapeutic benefits if it requires sufficient effort. In a study designed to test this notion, Cooper (1980) exposed snake-phobic subjects to a 40-minute session involving either implosive therapy or a bogus "effort therapy." The latter required subjects to engage in 40 minutes of physical exercises, including jumping rope, running in place, and winding a stick that had a 5-pound weight attached. (Subjects were told that this activity reduced fear by increasing emotional sensitivity.) In addition, subjects either were or were not given the option of withdrawing from the study after learning that the procedures could be "effortful and anxiety-provoking. "

 Results showed that, when decision freedom was high and subjects had the choice of whether or not to continue, both implosive therapy and "effort therapy" worked, but when decision freedom was low neither approach produced effects. To the degree that both treatments were effortful, change occurred only when subjects felt responsible for their participation, as the theories predict.

 In a conceptual replication, Cooper (1980) gave subjects either a common assertive behavior rehearsal treatment or a meaningless "effort therapy" similar to that described above. In addition, subjects were again given low or high choice regarding their participation. Using a clever dependent variable in which subjects were deliberately underpaid by a receptionist, Cooper showed that subjects who were given a choice regarding participation were more assertive in demanding the correct payment than those who had no choice, regardless of the treatment they had received. We consider this finding quite remarkable: A treatment designed to teach assertiveness was no more effective than a nonsensical "effort therapy," as long as subjects exerted effort and perceived they had high choice.

 The importance of clients' perceptions that they have choice in their treatment was further highlighted by Mendonca and Brehm's (1983) study of weight reduction. In this study, some overweight children were given a choice regarding which weight-control program they would follow, whereas others were given no choice. In reality, children in both experimental groups were assigned to the same program, an effortful weight-loss regimen that required restriction of caloric intake, regular exercise, and detailed records of food consumption. The effects of the program clearly demonstrated the importance of perceived choice: Children who had the perception that they had chosen their program lost significantly more weight than those who had no choice, and this difference was maintained at a 4-week follow-up.

 In perhaps the most striking demonstration of the effort justification effect in counseling, Axsom and Cooper (1985) recruited overweight women for a weight-loss program. All of the women participated in four sessions of "therapy." In these sessions, the women engaged in perceptual and cognitive tasks that they were told would enhance their emotional sensitivity in a way that helps lead to weight reduction. In reality, these tasks had nothing to do with weight loss, involving activities such as matching the length of lines on a tachistoscope.

 The women were assigned to one of two groups that differed in the amount of cognitive effort required by these tasks; for half of the women, the tasks were relatively easy, whereas for the others they required sustained mental effort over a longer period of time. After only four sessions of these nonsensical activities, the women in the high-effort group had lost significantly more weight than those in either the low-effort group or a no-treatment control group (neither of which had lost any weight at all). A 6-month follow-up showed that the high-effort group continued to lose weight, weighing an average of 8 pounds less than they had at the start of the study, whereas the women in the low-effort and control groups lost no weight at all. These group differences were maintained up to a year after "treatment."

 Thus, using different target behaviors (phobias, assertiveness, weight control) and different procedures (physical versus cognitive "effort therapies"), these studies demonstrate that effort expenditure per se can have therapeutic effects in the absence of any intervention that could reasonably be expected to produce change. However, effortful therapies were only effective when subjects felt a sense of choice and responsibility regarding their participation in them, a finding that is consistent with all three theoretical explanations of the effect.

 Although Cooper (1980) interpreted the effort justification effects within a cognitive dissonance framework, they are as easily handled by the alternative approaches. According to self-perception theory, subjects who exert high effort under high-choice conditions are more likely to infer that they really wanted to change than those who engaged in low effort therapy or had no choice in the matter. According to self-presentation theory, subjects who engaged in an effortful "therapy" under high-choice conditions experienced pressure to show that the treatment was effective, either to publicly justify their involvement (so as not to look foolish for wasting their time in a worthless project) or to please the experimenter. Future research should test alternative explanations of effort justification effects using clinically relevant procedures, such as those employed by Cooper (1980) and Axsom and Cooper (1985).

One further therapeutic implication of the effort justification phenomenon is worthy of mention. Harari (1983) noted that clients are often required to complete extensive batteries of tests before undergoing treatment. Although many practitioners regard these "intake" procedures as detrimental to therapy, Harari showed just the opposite. He randomly assigned some clients in a mental health center to undergo an increased number of preliminary procedures. As laboratory research on effort justification predicts, clients who completed extensive intake procedures were more satisfied with therapy and their therapists, were less likely to miss their therapy sessions, and were less likely to terminate therapy prematurely.

 In brief, research suggests that effort expenditure facilitates psychological change, at least under certain circumstances. Of course, if the effort required of clients is too great, they are likely to withhold full effort or to terminate therapy. But until that point is reached, greater effort is associated with greater change.

 Paradoxical Therapy

 In paradoxical therapy, the therapist attempts to eliminate the client's dysfunctional behavior by instructing the client to engage willfully in the undesired behavior (Haley, 1963). The behaviors targeted for paradoxical injunctions usually involve those such as insomnia, smoking, procrastination, and overeating, that are resistant to traditional "talk therapies." For example, a student who is a habitual procrastinator might be directed to sit in front of her books but put off studying for some prespecified time each day (e.g., Lopez & Wambach, 1982; Wright & Strong, 1982). Or an insomniac who normally takes 4 hours to go to sleep each night might be instructed to lie awake for at least 6 hours. An adolescent who is in constant conflict with his parents might be instructed to purposefully argue with his parents at prespecified times each day.

 In a variation of this theme known as "reframing," the therapist interprets the client's difficulties in a positive fashion and encourages their continuance. For example, Palazzoli, Boscolo, Cecchin, and Prata (1978) found that interpreting pathological family interactions in a positive fashion (as demonstrating sensitivity and self-sacrifice, for example) reduced the dysfunctional interaction patterns among family members. Similarly, Beek and Strong (1981) demonstrated that reinterpreting clients' depressive symptoms in a positive fashion resulted in greater long-term reduction of symptoms than did a more traditional therapy.

 Paradoxically, then, paradoxical directives and reframing are often effective in reducing or eliminating undesired behavior (Jackson & Haley, 1968; Rohrbaugh, Tennen, Press, & White, 1981; Strong, 1982; Watzlawick, Beavin, & Jackson, 1967; Weeks & L'Abate, 1979). However, the psychological processes mediating these effects are not at all clear. Several explanations have been offered to explain the effectiveness of paradoxical therapy, but only some of them concern us here.

 Reactance theory. Perhaps the most widely accepted explanation of paradoxical effects is based on reactance theory, which deals with people's reactions to events that threaten their behavioral freedom (Brehm, 1966; chapter 3, this volume). Under certain circumstances, events that threaten individuals' behavioral freedom motivate them to restore their freedom, resulting in behavior contrary to the social pressure. In the case of paradoxical therapy, the therapist's insistence that the client perform the problem behaviors in particular ways at prespecified times may threaten the client's sense of freedom and autonomy and result in an oppositional attempt by the client to prevent the symptoms from occurring.

Cognitive dissonance theory. Alternatively, explanations based on dissonance theory emphasize the fact that, in most instances, clients assume that their "symptoms" are beyond their voluntary control. However, clients who follow a therapist's instructions to purposefully engage in the behavior soon learn that their behavior can, to some degree, be turned off and on at will. To the degree that this realization is inconsistent with the belief that one's behavior is out of one's control, the clients may experience dissonance, which can be resolved by concluding that their symptoms are controllable after all. Once this point is reached, the clients are able to exert conscious control over the problem behavior.

Self-perception and attributional explanations. Attributional explanations offer two other insights regarding paradoxical therapy. First, it has long been recognized that intrinsic interest in an activity is diminished when external rewards are offered for engaging in the activity (Deci, 1975; Lepper, Greene, & Nisbett, 1973). In one classic study, children who were rewarded for engaging in activities they previously enjoyed showed a subsequent decrease in interest in those activities (Lepper et al., 1973). According to self-perception theory, when people do something for no obvious external reason, they infer that they are doing so for internal reasons, and may conclude that they enjoy the activity (Bem, 1972). However, when external inducements are offered, people infer that they are performing the behavior to gain those external rewards, and their intrinsic motivation declines. In the same way, clients tend to make internal attributions for dysfunctional behaviors that occur for no obvious external reason (such as chronic insomnia or procrastination). Once the therapist orders that the behavior be performed, the internal reasons for doing so diminish in importance, and the dysfunction seems less dispositional.

 Further, clients who attribute their problems to themselves may conclude that their behavior is uncontrollable and experience learned helplessness in that regard (see chapter 3). However, when the behavior appears to occur voluntarily in compliance with the demands of the therapist, clients may conclude that the behavior is, in fact, controllable, either by the therapist or by themselves. Thus, paradoxical therapy may be effective because it leads clients to conclude that the problem behavior is controllable.

 In chapter 2, we discussed problems such as insomnia that are exacerbated when people make internal attributions for them (Storms & McCaul, 1976). Paradoxical therapy may have the additional effect of leading clients to attribute their difficulties externally (away from themselves), thereby reducing their distress over the problem and breaking the exacerbation cycle. For example, the insomniac who attempts to force himself or herself to stay awake has a plausible explanation for his or her sleeplessness, an explanation that reduces the degree to which he or she worries about going to sleep on any particular night. With this source of anxiety gone, the individual may fall asleep more quickly (e.g., Ascher & Efran, 1978).

 Self-presentation. A self-presentation approach to paradoxical therapy suggests two routes to behavior change. First, the "reactance" effects discussed above may reflect impression management strategies rather than intrapsychic attempts to restore behavioral freedom (Wright & Brehm, 1982). That is, clients may disobey the therapist's directives so as to be viewed as autonomous and self-directed. Clients may purposefully control their symptomatic behavior in order to show the therapist that they are not under his or her control.

 Additionally, as noted earlier, most clients maintain that their dysfunctional behavior is beyond their voluntary control. Once It is demonstrated that a certain degree of control is possible (e.g., I can fight with my parents 14 on demand"), the client faces a self-presentational dilemma. lf the client continues to engage in maladaptive behavior even in the face of public evidence that he or she has some control over it, negative evaluations and reactions from others will result. Thus, the client may begin to control the dysfunctional behavior as much as possible.

 Interactional approaches. Many writers have pointed out that dysfunctional behavior may serve interpersonal functions (Artiss, 1959; Carson, 1969). People sometimes engage in dysfunctional reactions as a way of controlling or punishing others. As noted in chapter 5, intimidating and supplicating strategies are quite effective in inducing others to respond in desired ways by instilling fear, guilt, pity, or some other reaction. The effectiveness of these strategies is due in part to their undesirability as people respond to the "disturbed" individual in a manner that will eliminate the undesirable behavior.

 Thus, reinterpreting problem behaviors in a positive fashion may deprive the client of the use of these sorts of strategies. Just as a child who throws a tantrum on the kitchen floor will stop once his parents calmly thank him for dusting the linoleum, clients may forgo their "symptoms" when they no longer offend, intimidate, or control others (see Beck & Strong, 198 1; Omer, 1981; Palazzoli et al., 1978; Weeks & L'Abate, 1982).

 Summary and conclusions. We have mentioned how reactance, dissonance, attribution, self-presentation, and interactional approaches can explain the effectiveness of paradoxical therapy in order to make two points. First, although paradoxical treatments are clearly effective in many instances, we have little empirical evidence that elucidates the processes involved. At this juncture, each of these explanations of the effect is plausible. This seems to be a very important topic for future research, for advances in the use of these techniques will depend on a more complete understanding of the factors underlying them.

 Second, as this chapter has amply demonstrated, more than one social psychological theory may be applied to many behavioral phenomena. Although this state of affairs is often regarded as frustrating and discouraging by professionals and students alike, this is not necessarily a bad state of affairs. On the one hand, there is no reason to assume a priori that any one of these theories can explain all instances in which paradoxical therapy or any other treatment is effective. To the degree that behavioral and emotional problems are multiply determined and various treatments have multiple effects on the client, it is possible that each of these theories, or at least some subset of them, are needed to understand every instance in which paradoxical treatment is effective. For example, paradoxical injunctions may serve an attributional function for an insomniac if she attributes her sleeplessness to her willful attempts to remain awake, rather than to her personal problems, and consequently stops worrying about it. For a schizophrenogenic mother, paradoxical techniques may serve to reframe her manipulative symptoms in more positive terms, thereby diminishing their effectiveness. The point is that paradoxical therapy may work for different disorders and for different people for different reasons. Only future research will uncover the relevant variables.

 Beyond that, having several theories that explain a single phenomenon encourages strong empirical tests of the various positions by the competing camps, thereby stimulating a great deal of creative and enlightening research (Elms, 1975). Feyerabend, a noted philosopher of science, argues that science is best served not by monotheoretical "normal science" (Kuhn, 1962) but by the presence of competing theories. Feyerabend noted that, "this plurality of theories must not be regarded as a preliminary stage of knowledge which will at some time in the future be replaced by the One True Theory" (Feyerabend, 1970, p. 321). Having just one accepted theory neither ensures nor indicates that the theory is the best one.

Conclusions

This chapter has examined three therapeutic strategies that share an emphasis on inducing clients to engage in behaviors that they would prefer not to perform. In role playing, clients act in ways that are atypical of their behavior or are counterattitudinal; in effort justification paradigms, clients engage in effortful treatment regimens; and in paradoxical therapy, clients are told to consciously create their dysfunctional behaviors in prescribed ways. Each of these procedures can, under appropriate conditions, create therapeutic change, but the mechanisms responsible for these changes are imperfectly understood, and they can be explained equally well by different theories. The picture is complicated further by the possibility that more than one process is involved in these effects. Improved use of these procedures in counseling settings awaits additional research, both basic and applied, that elucidates the processes involved.