Cognitive Dissonance: Reexamining a Pivotal Theory in Psychology - Eddie Harmon-Jones 2019

Dissonance, Hypocrisy, and the Self-Concept
The Role of The Self in Dissonance

Elliot Aronson

This chapter focuses primarily on the relationship between cognitive dissonance and the self-concept. At the outset, however, note that when Leon Festinger invented the theory of cognitive dissonance, he conceived of dissonance arousal and reduction as a much more universal phenomenon—not as tied to a person’s self-concept. Accordingly, before getting to the heart of this chapter, I trace the evolution of the theory from its exciting universalistic beginnings, in 1957, when it revolutionized the way social psychologists think about human behavior, through its “doldrums” period (roughly, 1975—1990), when it was largely ignored by most researchers, to its reemergence in the 1990s as a powerful means of predicting and changing human behavior in a variety of areas, including those that have abiding societal importance (such as condom use and water conservation).

Because dissonance theory arrived on the social psychological scene at virtually the same moment I started graduate school, my involvement with the theory is a personal one as well as an intellectual one. Accordingly, it might be useful to tell the story of the dissonance “revolution” through my own rather fortuitous experience with the theory—almost from its inception.

As a 1st-year graduate student at Stanford, in 1956, I had very little interest in social psychology, and what little I knew about that discipline seemed both boring and pedestrian. Central to social psychology was the issue of social influence, which is certainly an important topic, but in the mid-1950s, the existing knowledge of social influence seemed fairly cut and dried and rather obvious. What did social psychologists know for sure about social influence at that time?

1. If you want people to go along with your position, offer tangible rewards for compliance and clear punishments for noncompliance.

2. Present an audience with a reasonable communication, attributing it to a highly credible communicator.

3. Present the individual with the illusion that everyone else in sight agrees with one another.

4. If a member of your discussion group disagrees with you, you will send him more messages (in attempt to get him to see the light) than if he agrees with you. If he persists in being stubborn, you will try to eject him from the group.

In those days, the overwhelming trend in all of American empirical psychology was “Let’s find the reinforcer.” If a person (or a rat) does something, there must be a reason, and that reason has to be the gaining of an identifiable reward, such as food, money, or praise, or the removing of a noxious state of affairs, such as pain, fear, or anxiety. If food will induce a hungry rat to press the lever of a Skinner box or turn left in a Y maze, surely conceptually similar rewards can induce a person to adopt a given opinion. The classic experiment that seemed to epitomize experimental social psychology in the mid-1950s was the still-classic Asch (1951) experiment, in which a unanimous majority apparently disagreed with an individual on a simple, unambiguous perceptual judgment. Why do most people conform to this kind of group pressure? Perhaps it makes them anxious to be alone against a unanimous majority; they fear being considered crazy, being held in low esteem, and so on. It’s comforting to be in agreement with others. That’s the reward for conformity.

Or take the equally classic experiment done by Hovland and Weiss (1951). Why do people tend to believe a statement attributed to a credible source (such as Oppenheimer) rather than a noncredible one (such as Pravda)? Perhaps it increases the probability of being right, and being right reduces anxiety and makes one feel good, smart, and esteemed. That is the reward for changing one’s belief.

These data are true enough, but hardly worth getting excited about. My old bobbeh (grandmother), a fountainhead of folk wisdom, could have told me those things without having done an elaborate experiment to demonstrate the obvious. Then, in 1957, Leon Festinger invented the theory of cognitive dissonance, deftly combining cognition and motivation, and produced a revolution that revitalized social psychology and changed it forever. I first read Festinger’s book in the form of a prepublication carbon copy that he thrust into my hands (rather disdainfully!) after I told him I was trying to decide whether to enroll in his graduate seminar. Reading that manuscript was something of an epiphany for me. It was (and still is) the single most exciting book I have ever read in all of psychology.

The core proposition of the theory is a very simple one: If a person were to hold two cognitions that were psychologically inconsistent, she or he would experience dissonance. Because dissonance is an unpleasant drive state (like hunger, thirst, or pain), the person will attempt to reduce it—much like she or he would try to reduce hunger, thirst, or pain. Viewed more broadly, cognitive dissonance theory is essentially a theory about sense making: how people try to make sense out of their environment and their behavior and, thus, try to lead lives that are (at least in their own mind) sensible and meaningful.

As I implied above, one of the theory’s most important aspects was in the challenge it presented to the long-standing dominance of reinforcement theory as an all-purpose explanation for social psychological phenomena. To illustrate this challenge (as well as its importance), I put forth the following scenario: A young man performs a monotonous, tedious task as part of an industrial relations experiment. After completing it, he is informed that his participation as a participant is over. The experimenter then appeals to him for help. He states that his research assistant was unable to be there and asks the participant if he would help run the experiment. The experimenter explains that he is investigating the effect of people’s preconceptions about their performance of a task; specifically, he wants to see if a person’s performance is influenced by whether he or she is told either positive things about the task (in advance), negative things about the task (in advance), or nothing at all about the task. The next participant, who is about to arrive, is assigned to be in the favorable-information condition. The experimenter asks the participant if he would tell the incoming participant that he had just completed the task (which is true) and that he found it to be an exceedingly enjoyable one (which is not true, according to the participant’s own experience). The participant is offered either $1 or $20 for telling this lie and for remaining on call in case the regular assistant cannot show up in the future.

The astute reader will recognize this as the scenario of the classic experiment by Festinger and Carlsmith (1959). I regard this experiment, because of the enormous impact it had on the field, as the single most important study ever done in social psychology. The results were striking. The participants who said that they found the task enjoyable to earn the paltry payment of $1 came to believe that it actually was enjoyable to a far greater extent than those who said it to earn the princely payment of $20. The experiment was a direct derivation from the theory of cognitive dissonance. Needless to say, reinforcement theory would suggest that, if you reward individuals for saying something, they might become infatuated with that statement (through secondary reinforcement). But dissonance theory makes precisely the opposite prediction. If I were a participant in the Festinger—Carlsmith (1959) experiment, my cognition that the task I performed was boring would be dissonant with the fact that I informed another person that it was enjoyable. If I were paid $20 for making that statement, this cognition would provide ample external justification for my action, thus reducing the dissonance. However, if I were paid only $1, I would lack sufficient external justification for having made the statement (I would be experiencing the discomfort of dissonance) and would be motivated to reduce it. In this situation, the most convenient way to reduce dissonance would be for me to try to convince myself that the task was somewhat more interesting than it seemed at first. In effect, in the process of persuading myself that the task was actually interesting, I would convince myself that my statement to the other student was not a great lie.

Similarly, in another early experiment aimed at testing dissonance theory, Aronson and Mills (1959) demonstrated that people who go through a severe initiation, to gain admission to a group, come to like that group better than people who go through a mild initiation to get into the same group. Reinforcement theory would suggest that we like people and groups that are associated with reward; dissonance theory led Aronson and Mills to the prediction that we come to like things for which we suffer. All cognitions having to do with the negative aspects of the group are dissonant with the cognition that we suffered to be admitted to the group; therefore, they get distorted in a positive direction, effectively reducing the dissonance.

Even in the early years, it was crystal clear that dissonance-generated attitude change was not limited to such trivial judgments as the dullness of a boring task or the attractiveness of a discussion group. The early researchers extended the theory to much more important opinions and attitudes, such as a striking reassessment of the dangers of smoking marijuana among students at the University of Texas (Nel, Helmreich, & Aronson, 1969), and the softening of Yale students’ negative attitudes toward the alleged antistudent brutality of the New Haven police (Cohen, 1962).

IMPACT ON THE FIELD

It is hard to convey the impact these early experiments had on the social psychological community at the time of their publication. The findings startled a great many social psychologists largely because they challenged the general orientation accepted either tacitly or explicitly by the field. These results also generated enthusiasm among most social psychologists because, at the time, they represented a striking and convincing act of liberation from the dominance of a general reward-reinforcement theory. The findings of these early experiments demonstrated dramatically that at least under certain conditions, reward theory was inadequate. In doing so, dissonance research sounded a clarion call to cognitively oriented social psychologists, proclaiming in the most striking manner that human beings think, they do not always behave in a mechanistic manner. It demonstrated that human beings engage in all kinds of cognitive gymnastics aimed at justifying their own behavior.

Perhaps most important, dissonance theory inspired an enormous number and variety of hypotheses that were specific to the theory and could be tested in the laboratory. The wide array of research that dissonance theory has produced is truly astonishing. Dissonance research runs the gamut from decision making to color preferences, from the socialization of children to curing people’s snake phobias, from interpersonal attraction to antecedents of hunger and thirst, from the proselytizing behavior of religious zealots to the behavior of gamblers at racetracks, from inducing people to conserve water by taking showers to selective informational exposure, from helping people curb their temptation to cheat at a game of cards to inducing people to practice safer sex.

The impact of dissonance theory went even beyond the generation of new and exciting knowledge. Given the nature of the hypotheses we were testing, dissonance researchers were forced to develop a new experimental methodology—a powerful, high-impact set of procedures that allowed us to ask truly important questions in a very precise manner. As we all know, the laboratory tends to be an artificial environment. But dissonance research made it necessary to overcome that artificiality by developing a methodology that would enmesh participants in a set of events—a drama, if you will—which made it impossible for them to avoid taking these events seriously.

In my writing on research methods (Aronson & Carlsmith, 1968; Aronson, Ellsworth, Carlsmith, & Gonzales, 1990) I have referred to this strategy as establishing experimental reality where, within the admittedly phony confines of the lab, the experimenter makes certain that real things are happening to real people. Because of the nature of our hypotheses, we could not afford the luxury, so common in contemporary research, of having participants passively look at a videotape of events happening to someone else and then make judgments about them. Rather, our research questions required the construction of an elaborate scenario in which participants became immersed. Thus, what dissonance research brought into focus more clearly than any other body of work is the fact that the social psychological laboratory, with all of its contrivances and complex scenarios, can produce clear, powerful effects that are conceptually replicable in both the laboratory and the real world.

DISSONANCE AND THE SELF: THE POWER OF SELF-PERSUASION

Research has shown that the persuasive effects in the experiments discussed previously are more powerful and more persistent than those resulting from persuasion techniques based on rewards, punishments, or source credibility (e.g., Freedman, 1965). The major reason is that the arousal of dissonance always entails relatively high levels of personal involvement and, therefore, the reduction of dissonance requires some form of self-justification. From the very outset, some of us who were working closely with the theory felt that at its core it led to clear and unambiguous predictions, but around the edges it was a little too vague. Several situations arose in which it was not entirely clear what dissonance theory would predict or, indeed, whether dissonance theory even made a prediction. Around 1958, the standing joke among Festinger’s research assistants was, “If you really want to be sure whether A is dissonant with B, ask Leon!” Although this was said with our tongues firmly planted in our cheeks, it reflected the fact that we argued a lot about whether dissonance theory applied in a wide variety of situations.

What comes to mind most specifically are two strenuous running arguments that Festinger and I had about two of his classic examples. The first involved a person stepping outside in a rainstorm and not getting wet. Festinger was convinced that this would arouse a great deal of dissonance, whereas I had considerable difficulty seeing it. My disagreement went something like this: “What’s that got to do with him? It’s a strange phenomenon, all right, but unless he feared he was losing his mind, I don’t see the dissonance.”

The second was Festinger’s classic example of a situation in which dissonance theory did not apply. This was the case of a man driving, late at night, on a lonely country road and getting a flat tire (Festinger, 1957, pp. 277—278). Lo and behold, when he opened the trunk of his car, he discovered he did not have a jack. Leon maintained that, although the person would experience frustration, disappointment, and perhaps even fear, there were no dissonant cognitions in that situation. My argument was succinct: “Of course there is dissonance! What kind of idiot would go driving late at night on a lonely country road without a jack in his car?” “But,” Leon countered, “where are the dissonant cognitions?”

It took me a couple of years, but it gradually dawned on me that what was at the heart of my argument in both of those situations was the self-concept. That is, when I said above that dissonance theory made clear predictions at its core, what I implicitly meant by at its core were situations in which the person’s self-concept was at issue. Thus, in the raindrop situation, as far as I could judge, the self was not involved. In the flat tire situation, the self-concept was involved; what was dissonant was (a) the driver’s cognition about his idiotic behavior with (b) his self-concept of being a reasonably smart guy. Accordingly, I wrote a monograph (Aronson, 1960), in which I argued that dissonance theory makes its strongest predictions when an important element of the self-concept is threatened, typically when a person performs a behavior that is inconsistent with his or her sense of self. Initially, I intended this not to be a major modification of the theory, but only an attempt to tighten the predictions a bit. That is, in my opinion, this tightening retained the core notion of inconsistency but shifted the emphasis to the self-concept—thus, clarifying more precisely when the theory did or did not apply. I believe that this apparently minor modification of dissonance theory turned out to have important ramifications inasmuch as it increased the predictive power of the theory without seriously limiting its scope.

In addition, this modification uncovered a hidden assumption contained in the original theory. Festinger’s original statement and all of the early experiments rested on the implicit assumption that people have a reasonably positive self-concept. But if a person considered himself or herself to be a “schnook,” he or she might expect to do schnooky things—like go through a severe initiation to get into a group or say things that he or she didn’t quite believe. For such people, dissonance would not be aroused under the same conditions as for people with a favorable view of themselves. Rather, dissonance would occur when negative self-expectancies were violated, that is, when the person with a poor self-concept engaged in a behavior that reflected positively on the self.

To test this assumption, Merrill Carlsmith and I conducted a simple little experiment that demonstrated that under certain conditions, college students would be made uncomfortable with success; that they would prefer to be accurate in predicting their own behavior, even if it meant setting themselves up for failure. Specifically, we found that students who had developed negative self-expectancies regarding their performance on a task showed evidence of dissonance arousal when faced with success on that task. That is, after repeated failure at the task, participants who later achieved a successful performance actually changed their responses from accurate to inaccurate ones, to preserve a consistent, though negative, self-concept (Aronson & Carlsmith, 1962). (In recent years, Swann and his students have confirmed this basic finding in a number of experiments and quasi-experiments; Swann, 1984, 1990, 1996; Swann & Pelham, 1988; Swann & Read, 1981).

A few years later, I carried this reasoning a step further (Aronson, 1968; Aronson, Chase, Helmreich, & Ruhnke, 1974), elaborating on the centrality of the self-concept in dissonance processes and suggesting that in this regard, people generally strive to maintain a sense of self that is both consistent and positive. That is, because most people have relatively favorable views of themselves, they want to see themselves as (a) competent, (b) moral, and (c) able to predict their own behavior.

In summary, efforts to reduce dissonance involve a process of self-justification because, in most instances, people experience dissonance after engaging in an action that leaves them feeling stupid, immoral, or confused (see Aronson et al., 1974). Moreover, the greater the personal commitment or self-involvement implied by the action and the smaller the external justification for that action, the greater the dissonance and, therefore, the more powerful the need for self-justification. Thus, in the Festinger—Carlsmith experiment, the act of deceiving another person would make one feel immoral or guilty. To reduce that dissonance, one must convince oneself that little or no deception was involved, in other words, that the task was, in fact, a rather interesting activity. By justifying one’s actions in this fashion, one is able to restore a sense of self as morally good. In the Aronson and Mills’ (1959) experiment, going through hell and high water to gain admission to a boring discussion group was dissonant with one’s self-concept as a smart and reasonable person, who makes smart and reasonable decisions.

SCOPE VERSUS TIGHTNESS IN THEORY BUILDING

All theories are lies. That is, all theories are only approximations of the empirical domain they are trying to describe. Accordingly, it is inevitable that theories will evolve and change to accommodate new data that are being generated. Indeed, it is the duty of theorists to modify their theory in the face of new data and new ideas. Festinger understood this better than most theorists. At the same time, understandably, he was deeply enamored of both the elegant simplicity and the breadth of his original theoretical statement. When I first came out with the self-concept notion of dissonance, Festinger was not pleased. He felt that although my revision had led to some interesting research, conceptually, I was limiting the scope of the theory far too much. I agreed that the scope was a bit smaller, but I believed that the increased accuracy of prediction (the added tightness) was worth the slightly more limited scope. In 1987, while serving as a discussant at the American Psychological Association symposium on cognitive dissonance, Festinger acknowledged the dilemma of the theorist who has a hard time seeing his theory change yet knows that change it must. (One might say that this is a situation bound to produce considerable dissonance in the theorist!). In typical fashion, Leon poked fun at himself for trying to cling to the original conceptualization even though he knew better (see Appendix B, this volume, for the complete transcript of this talk):

No theory is going to be inviolate. Let me put it clearly. The only kind of theory that can be proposed and ever will be proposed that absolutely will remain inviolate for decades, certainly centuries, is a theory that is not testable. If a theory is at all testable, it will not remain unchanged. It has to change. All theories are wrong. One doesn’t ask about theories, can I show that they are wrong or can I show that they are right, but rather one asks, how much of the empirical realm can it handle and how must it be modified and changed as it matures?

As a lot of people know, I ended up leaving social psychology, meaning dissonance theory, and I want to clarify that. Lack of activity is not the same as lack of interest. Lack of activity is not desertion. I left and stopped doing research on the theory of dissonance because I was in a total rut. The only thing I could think about was how correct the original statement had been . . . how every word in that book was perfect. So to me, I did a good thing for cognitive dissonance by leaving it. I think if I had stayed in it, I might have retarded progress for cognitive dissonance for at least a decade. (pp. 382—383)

In theory building, there is always a tension between scope and precision; generally speaking, one usually gains precision at the price of scope. The self-concept notion strikes a pretty good balance between scope and precision. My guess is that sooner or later, someone will come along with a richer conceptualization that will strike a better balance. When that happens, I hope I will have the good grace to applaud.

THE SELF-CONCEPT AND THE INDUCTION OF HYPOCRISY

In recent years, the self-concept notion of dissonance has led us into areas of investigation that would not have been feasible under the rubric of Festinger’s initial formulation. One of these involves the induction of feelings of hypocrisy. This discovery came about quite by accident. At the time, I was not even thinking about theory development but was struggling to find an effective way to convince sexually active college students to use condoms in this, the era of AIDS. The problem is not an easy one to solve, because it transcends the simple conveying of information to rational people. College students already have the requisite information, that is, virtually all sexually active college students know that condoms are an effective way to prevent AIDS. The problem is that the vast majority are not using condoms, because they consider them to be a nuisance, unromantic, and unspontaneous. In my research, I had run into a stone wall; I had tried several of the traditional, direct persuasive techniques (powerful videos, aimed at arousing fear or at eroticizing the condom) with very limited success. Whatever impact my videos did have was of very short duration; our participants would try condoms once or twice and then stop using them.

Eventually, I thought about using the counterattitudinal-attitude paradigm. That is, why not try to get people to argue against their own attitudes, as in the Festinger—Carlsmith (1959) experiment? On the surface, it seemed like a great idea. After all, we had found that this strategy was powerful and, when judiciously applied, had long-term effects on attitudes and behavior—precisely what was needed in this societal situation. But wait a minute: In the condom use situation, there were no counterattitudinal attitudes to address. That is, our surveys and interviews had demonstrated that sexually active young adults already were in favor of people using condoms to prevent AIDS. They simply weren’t using them. They seemed to be in a state of denial: denying that the dangers of unprotected sex applied to them in the same way they applied to everyone else. How could we invoke the counterattitudinal-attitude paradigm if there was no counterattitude to invoke?

It occurred to me that the solution had to come from the self-concept, because being in denial is not an attractive thing to be doing. The challenge was to find a way to place the person in a situation where the act of denial would be infeasible, because it would conflict, in some way, with his or her positive image of their selves. And then it struck me. Suppose you are a sexually active college student and, like most, (a) you do not use condoms regularly, and (b) being in denial, you have managed to ignore the dangers inherent in having unprotected sex. Suppose, upon going home for Christmas vacation, you find that Charlie, your 16-year-old brother has just discovered sex and is in the process of boasting to you about his many and varied sexual encounters. What do you say to him? Chances are, as a caring, responsible older sibling, you will dampen his enthusiasm a bit by warning him about the dangers of AIDS and other sexually transmitted diseases and urge him to, at least, take proper precautions by using condoms.

Suppose that I am a friend of the family who was invited to dinner and who happened to overhear this exchange between you and your brother. What if I were to pull you aside and say, “That was very good advice you gave Charlie. I am very proud of you for being so responsible. By the way, how frequently do you use condoms?” In other words, by getting you to think about that, I am confronting you with your own hypocrisy. According to the self-concept version of the theory, this would produce dissonance because you are not practicing what you are preaching. That is, for most people, their self-concept does not include behaving like a hypocrite.

My students and I then proceeded to design and conduct a simple little experiment following the scenario outlined above (Aronson, Fried, & Stone, 1991). In a 2 × 2 factorial design, in one condition, college students were induced to make a videotape in which they urged their audience to use condoms; they were told that the video would be shown to high school students. In the other major condition, the college students simply rehearsed the arguments without making the video. Cutting across these conditions was the “mindfulness” manipulation: In one set of conditions, our participants were made mindful of the fact that they themselves were not practicing what they were preaching. To accomplish this, we asked them to think about all those situations where they found it particularly difficult or impossible to use condoms in the recent past. In the other set of conditions, we did nothing to make the students mindful of their past failures to use condoms.

The one cell we expected to produce dissonance was the one high in hypocrisy, that is, where participants made the video and were given the opportunity to dredge up memories of situations where they failed to use condoms. Again, how did we expect them to reduce dissonance? By increasing the strength of their intention to use condoms in the future. And that is precisely what we got. Those participants who were in the high-dissonance (hypocrisy) condition showed the greatest intention to increase their use of condoms. Moreover, 2 months later, there was a tendency for the participants in the high-dissonance cell to report using condoms a higher percentage of the time than in any of the other three cells.

In a follow-up experiment (Stone, Aronson, Crain, Winslow, & Fried, 1994), we strengthened the manipulations of the initial experiment and used a “behavioroid” measure of the dependent variable. Specifically, in each of the conditions described above, participants were subsequently provided with an opportunity to purchase condoms at a very substantial discount (10¢ each). The results were unequivocally as predicted. Fully 83% of the participants in the hypocrisy condition purchased condoms; this was a significantly greater percentage than in each of the other three conditions, none of which were reliably different from each other. The effect was a powerful and long-lasting one: Three months after the induction of hypocrisy, a telephone survey indicated that 92% of the participants in the hypocrisy condition were still using condoms regularly, a figure that was significantly different from the control conditions.

Subsequently, we increased our confidence in the efficacy of the induction-of-hypocrisy paradigm by testing the paradigm in a different situation, one in which we could get a direct behavioral measure of the dependent variable. We found one in the shower room of our campus field house. As you may know, central California has a chronic water shortage. On our campus, the administration is constantly trying to find ways to induce students to conserve water. So we decided to test our hypothesis by using dissonance theory and the induction of hypocrisy to convince students to take shorter showers. What we discovered was that although it is impossible, within the bounds of propriety, to follow people into their bedrooms and observe their condom-using behavior, it was easily possible to follow people into the shower rooms and observe their shower-taking behavior.

In this experiment (Dickerson, Thibodeau, Aronson, & Miller, 1992), we went to the university field house and intercepted college women who had just finished swimming in a highly chlorinated pool and were on their way to take a shower. Just like in the condom experiment, it was a 2 × 2 design, in which we varied commitment and mindfulness. In the commitment conditions, each student was asked if she would be willing to sign a flyer encouraging people to conserve water at the field house. The students were told that the flyers would be displayed on posters, and each was shown a sample poster: a large, colorful, very public display. The flyer read: “Take shorter showers. Turn off water while soaping up. If I can do it, so can you!” After the student signed the flyer, we thanked her for her time, and she proceeded to the shower room, where our undergraduate research assistant (unaware of condition) was unobtrusively waiting (with hidden waterproof stopwatch) to time the student’s shower. In the mindful conditions, we also asked the students to respond to a water conservation survey, which consisted of items designed to make them aware of their proconservation attitudes and the fact that their typical showering behavior was sometimes wasteful.

The results are consistent with those in the condom experiment: We found dissonance effects only in the cell where the participants were preaching what they were not always practicing. That is, in the condition where the students were induced to advocate short showers and were made mindful of their own past behavior, they took very short showers. To be specific, in the high-dissonance cell, the length of the average shower (which, because of the chlorine in the swimming pool, included a shampoo and cream rinse) averaged just over 3 min 30 sec (that’s short!) and was significantly shorter than in the control condition.

How can we be certain that dissonance is involved in these experiments? Although the data are consistent with the self-concept formulation of dissonance theory, there is another plausible interpretation. It is conceivable that the effects of the hypocrisy manipulation may have been due to the effects of priming. The combination of proattitudinal advocacy and the salience of past behavior may have served, in an additive fashion, to make participants’ positive attitudes toward condom use or water conservation highly accessible, thus fostering a stronger correspondence between their attitudes and behavior (e.g., Fazio, 1989). What is needed to pin it down is evidence that the hypocrisy effect involves physiological arousal, thereby indicating the presence of dissonance rather than the mere influence of attitude salience.

An experiment by Fried and Aronson (1995) provides exactly this sort of evidence. Within the context of the hypocrisy paradigm, this experiment used a misattribution-of-arousal manipulation, a strategy brilliantly developed in earlier research to document the existence of dissonance as an uncomfortable state of arousal by Zanna and Cooper (1974). Zanna and Cooper found that when participants were given an opportunity to misattribute their arousal to a source other than their dissonance-arousing behavior—for example, to an overheated room, a placebo, or glaring florescent lights—the attitude change typically associated with dissonance reduction no longer occurs.

Using a modified version of the earlier condom experiments, Fried and Aronson’s (1995) study required participants to compose and deliver proattitudinal, videotaped speeches advocating the importance of recycling. These speeches were ostensibly to be shown to various groups as part of a campaign to increase participation in recycling programs on campus and in the larger community. Hypocrisy was induced in half the participants by asking them to list recent examples of times when they had failed to recycle; the other half simply wrote and delivered the speech, without being reminded of their wasteful behavior. In addition, half the participants in each condition were given an opportunity to misattribute arousal to various environmental factors within the laboratory setting. Specifically, participants were asked to answer questions regarding the room’s fighting, temperature, and noise level, including how these ambient factors might have affected them. (This was accomplished under the guise of asking participants to rate the room’s suitability for use as laboratory space—a request that was made to appear unrelated to the activities in which participants were participating.) To summarize, proattitudinal advocacy was held constant, and the salience of past behavior and the opportunity for misattribution were manipulated, yielding a 2 × 2 factorial design with the following conditions: (a) hypocrisy (high salience), (b) hypocrisy with misattribution, (c) no hypocrisy (low salience), and (d) no hypocrisy with misattribution. Dissonance reduction was measured by asking participants to volunteer to help a local recycling organization by making phone calls soliciting support for recycling.

The results of this experiment revealed that arousal was indeed present within the hypocrisy conditions. Hypocrisy participants who were not afforded the opportunity to misattribute the source of their arousal volunteered significantly more often, and for longer blocks of time, than participants in the other experimental conditions. Moreover, volunteer behavior for hypocrisy participants who were allowed to misattribute their arousal was no greater than for participants who were not exposed to the hypocrisy manipulation.

More recently, the hypocrisy paradigm has been tested and extended to a great many areas, including health, the environment, and politics by Jeff Stone and his students (e.g., Focella & Stone, 2013; Focella, Stone, Fernandez, Cooper, & Hogg, 2016; Stone & Fernandez, 2011; Stone & Focella, 2011).

HYPOCRISY: DISSONANCE IN THE ABSENCE OF AVERSIVE CONSEQUENCES

As I stated above, the initial reason for the development of the hypocrisy paradigm was couched in my attempt to apply dissonance theory to the solution of a societal problem. The hidden bonus was that it also shed some light on an interesting theoretical controversy among dissonance theorists. I should say at the outset that dissonance research tends to be a family business; thus, the controversies are invariably friendly arguments, around the dinner table as it were. One such controversy involves the new look theory, developed several years ago by Cooper and Fazio (1984). In examining the early forced-compliance experiments, such as the Festinger—Carlsmith (1959) experiment and the Nel et al. (1969) experiment, Cooper and Fazio made an interesting discovery: In these experiments, participants not only experienced cognitive dissonance but also inflicted aversive consequences on the recipient of their communication; that is, lying to another person is presumed to have aversive consequences to that person. Their next step was a bold one: Cooper and Fazio asserted that dissonance was not due to inconsistent cognitions at all, but, rather, was aroused only when a person felt personally responsible for bringing about an aversive or unwanted event. Or, to put it in my terms, dissonance was caused solely by the person’s doing harm to another person, which was a threat to the person’s self-concept as a morally good human being.

Although I always appreciated the boldness implicit in Cooper and Fazio’s (1984) theorizing, I could never bring myself to buy into the notion that aversive consequences are essential for the existence of dissonance. Moreover, in terms of my earlier discussion of scope versus tightness, it seems to me that Cooper and Fazio’s conception is limiting the scope of the theory enormously while gaining nothing in tightness that wasn’t already present in the self-concept notion.

How would one test this difference empirically? Several years ago, I was at a loss as to how to produce inconsistency in the Festinger and Carlsmith (1959) type of experiment without also producing aversive consequences for the recipient of one’s message. That is, if you are misleading another person, by telling her or him something you believe is false, then you are always bringing about aversive consequences. But without quite realizing it, my students and I seem to have stumbled onto the solution with the hypocrisy experiments. In this procedure, the participants are preaching what they are not practicing (and are, therefore, experiencing dissonance), but where are the aversive consequences for the audience in the condom experiment? There are none. Indeed, to the extent that the “hypocrites” succeed in being persuasive, far from producing aversive consequences for the recipients, they may well be saving their lives. And still, it is clear from the data that our participants were experiencing dissonance. For a fuller discussion of this theoretical controversy, see Thibodeau and Aronson (1992).

SELF-JUSTIFICATION AND SELF-ESTEEM—SOME UNSUBSTANTIATED SPECULATIONS

From the very beginning, I found dissonance theory to be a powerful explanation for a wide swath of human behavior. The scientist in me was always delighted by the exciting, nonobvious predictions generated by the theory, as well as the creative experimentation used to test these predictions. But at the same time, the humanist in me was always a bit troubled by the rather bleak, rather unappetizing picture the theory painted of the human condition—forever striving to justify our actions after the fact. Over the past few years, my reasoning about dissonance and the self-concept has led me to speculate about how a person’s self-esteem might interact with the experiencing and reduction of dissonance. These speculations might suggest a more complete picture of human nature. Note two intriguing experiments in the dissonance literature. First, consider an experiment I did a great many years ago with David Mettee (Aronson & Mettee, 1968), in which we demonstrated that if we temporarily raised a person’s self-esteem, it would serve to insulate him or her from performing an immoral act such as cheating. We found that the higher self-esteem served to make the anticipation of doing something immoral more dissonant than it would have been otherwise. Thus, when our participants were put in a situation in which they had an opportunity to win money by unobtrusively cheating at a game of cards, they were able to say to themselves, in effect, “Wonderful people like me don’t cheat!” And they succeeded in resisting the temptation to cheat to a greater extent than those in the control condition.

Now consider an experiment performed by Glass (1964). In this study, people were put in a situation where they were induced to deliver a series of electric shocks to other people. They then had an opportunity to evaluate their victims. Dissonance theory predicts that if individuals are feeling awful about having hurt someone, one way to reduce the dissonance is to convince themselves that their victim is a dreadful person who deserved to suffer the pain of electric shock. What Glass found was that it was precisely those individuals who had the highest self-esteem who derogated their victims the most. Consider the irony: It is precisely because I think I am such a nice person that if I do something that causes you pain, I must convince myself that you are a rat. In other words, because nice guys like me don’t go around hurting innocent people, you must have deserved every nasty thing I did to you. On the other hand, if I already consider myself to be something of a scoundrel, then causing others to suffer does not introduce as much dissonance; therefore, I have less of a need to convince myself that you deserved your fate. The ultimate tragedy, of course, is that once I succeed in convincing myself that you are a dreadful person, it lowers my inhibitions against doing you further damage.

Aronson and Mettee (1968) showed that high self-esteem can serve as a buffer against immoral behavior, whereas Glass (1964) showed that once a person commits an immoral action, high self-esteem leads him or her into a situation where he or she might commit further mischief. In pondering these two experiments, I have come to the conclusion that it is far too simplistic to think of self-esteem as a one-dimensional phenomenon—either high or low. My reasoning here is similar to that of Baumeister (1998); Kernis, Cornell, Sun, Berry, and Harlow (1993); Rohan (1996); and Waschull and Kernis (1996). My notion is that self-esteem can be high or low and either fragile or well-grounded. Well-grounded in this context means that a positive self-image has been developed and held during a great deal of past behavior, whereas fragile suggests that a positive self-image has never been securely developed. People with high and well-grounded self-esteem need not be concerned with developing or verifying their self-image and can enter situations with the confident knowledge that they are competent, moral people. On the other hand, people with high and fragile self-esteem, because of their lack of a secure self-image, are overly concerned with trying to preserve images of themselves as being competent and moral at all costs.

Typically, people with high and fragile self-esteem, in their zeal to maintain a belief in their own competence and virtue, often boast about their achievements—trying desperately to convince themselves and others that they are terrific. But their boasting behavior, their misjudgments, their errors, and the wrong turns they make because they are thinking more about themselves than the situation they are in all tend to shatter the fragile image they are trying to defend. As a result, they frequently feel like impostors and are forever trying to prove that they are not. Thus, they are trying to win every possible argument, pushing themselves to believe they are always right, justifying their behavior to themselves at every turn, and explaining away failures and mistakes instead of attending to them long enough to learn from them.

In contrast, people with high and well-grounded self-esteem are not invested in winning arguments for winning’s sake, do not need to believe they are always right, do not need to explain away failures and mistakes, and do not need to engage in the almost frantic self-justification in which high and fragile self-esteem people constantly engage. Instead, when they fail or make mistakes, people with high and well-grounded self-esteem can look at their failures and mistakes and learn from them. For example, a person with high, well-grounded self-esteem can look at his or her errors and say, in effect, “I screwed up. I did a stupid (or hurtful or immoral) thing. But just because I did a stupid (or hurtful or immoral) thing this time, this doesn’t make me a stupid (or hurtful or immoral) person. Let me look at it. How did it come about? How can I make it better? What can I learn from this situation, so that I might decrease the possibility that I’ll screw up in a similar way again?”

At this point, I have no idea whether the fragile and well-grounded dimension of self-esteem is normally distributed. If I were to hazard a guess, I would speculate that high, well-grounded self-esteem is not a common thing; accordingly, my guess is that the majority of people who score high on general measures of self-esteem would cluster near the fragile end of the continuum. If this were true, it would certainly account for the behavior of Glass’s (1964) high-self-esteem participants who did not hesitate to derogate their victim. In contrast, people with high, well-grounded self-esteem would not use derogation of the victim as a way to reduce their dissonance; rather, they would be more likely to take responsibility for their actions and try, in some way, to make amends for their cruel behavior.

This strikes me as an important area of inquiry. Again, these are mere speculations; I have no data to confirm them. Somehow, it seems reasonable to end this chapter with unsubstantiated speculation because, for me, it has always been the interesting loose ends that make science such an exciting enterprise.

REFERENCES

Aronson, E. (1960). The cognitive and behavioral consequences of the confirmation and disconfirmation of expectancies [Grant proposal]. Harvard University.

Aronson, E. (1968). Dissonance theory: Progress and problems. In R. P. Abelson, E. Aronson, W. J. McGuire, T. M. Newcomb, M. J. Rosenberg, & P. H. Tannenbaum (Eds.), Theories of cognitive consistency: A sourcebook. Chicago, IL: Rand McNally.

Aronson, E., & Carlsmith, J. M. (1962). Performance expectancy as a determinant of actual performance. Journal of Abnormal and Social Psychology, 65, 178—182. http://dx.doi.org/10.1037/h0042291

Aronson, E., & Carlsmith, J. M. (1968). Experimentation in social psychology. In G. Lindzey & E. Aronson (Eds.), The handbook of social psychology (2nd ed., Vol. 2, pp. 1—79). Reading, MA: Addison-Wesley.

Aronson, E., Chase, T., Helmreich, R., & Ruhnke, R. (1974). A two-factor theory of dissonance reduction: The effect of feeling stupid or feeling awful on opinion change. International Journal for Research and Communication, 3, 59—74.

Aronson, E., Ellsworth, P., Carlsmith, J. M., & Gonzales, M. H. (1990). Methods of research in social psychology. New York, NY: McGraw Hill.

Aronson, E., Fried, C., & Stone, J. (1991). Overcoming denial and increasing the intention to use condoms through the induction of hypocrisy. American Journal of Public Health, 81, 1636—1638. http://dx.doi.org/10.2105/AJPH.81.12.1636

Aronson, E., & Mettee, D. R. (1968). Dishonest behavior as a function of differential levels of induced self-esteem. Journal of Personality and Social Psychology, 9, 121—127. http://dx.doi.org/10.1037/h0025853

Aronson, E., & Mills, J. (1959). The effect of severity of initiation on liking for a group. Journal of Abnormal and Social Psychology, 59, 177—181. http://dx.doi.org/10.1037/h0047195

Asch, S. E. (1951). Effects of group pressure upon the modification and distortion of judgments. In H. Guetzkow (Ed.), Groups, leadership and men (pp. 177—190). Pittsburgh, PA: Carnegie Press.

Baumeister, R. E. (1998). The self. In D. Gilbert, S. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (4th ed., pp. 680—740). Boston, MA: McGraw-Hill.

Cohen, A. R. (1962). An experiment on small rewards for discrepant compliance and attitude change. In J. W. Brehm & A. R. Cohen (Eds.), Explorations in cognitive dissonance (pp. 73—78). New York, NY: Wiley.

Cooper, J., & Fazio, R. H. (1984). A new look at dissonance theory. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 17, pp. 229—266). Orlando, FL: Academic Press.

Dickerson, C., Thibodeau, R., Aronson, E., & Miller, D. (1992). Using cognitive dissonance to encourage water conservation. Journal of Applied Social Psychology, 22, 841—854. http://dx.doi.org/10.1111/j.1559-1816.1992.tb00928.x

Fazio, R. H. (1989). On the power and functionality of attitudes: The role of attitude accessibility. In A. R. Pratkanis, S. J. Breckler, & A. G. Greenwald (Eds.), Attitude structure and function (pp. 153—179). Hillsdale, NJ: Erlbaum.

Festinger, L. (1957). A theory of cognitive dissonance. Evanston, IL: Row, Peterson.

Festinger, L., & Carlsmith, J. M. (1959). Cognitive consequences of forced compliance. Journal of Abnormal and Social Psychology, 58, 203—210. http://dx.doi.org/10.1037/h0041593

Focella, E., & Stone, J. (2013). The use of hypocrisy for promoting environmentally sustainable behaviors. In H. van Tripp (Ed.), Encouraging Sustainable Behavior (pp. 203—218). New York, NY: Psychology Press.

Focella, E., Stone, J., Fernandez, N. C., Cooper, J., & Hogg, M. (2016). Vicarious hypocrisy: Bolstering attitudes and taking action after exposure to a hypocritical ingroup member. Journal of Experimental Social Psychology, 62, 89—102. http://dx.doi.org/10.1016/j.jesp.2015.09.014

Freedman, J. (1965). Long-term behavioral effects of cognitive dissonance. Journal of Experimental Social Psychology, 1, 145—155. http://dx.doi.org/10.1016/0022-1031(65)90042-9

Fried, C. B., & Aronson, E. (1995). Hypocrisy, misattribution, and dissonance reduction. Personality and Social Psychology Bulletin, 21, 925—933. http://dx.doi.org/10.1177/0146167295219007

Glass, D. C. (1964). Changes in liking as a means of reducing cognitive discrepancies between self-esteem and aggression. Journal of Personality, 32, 531—549. http://dx.doi.org/10.1111/j.1467-6494.1964.tb01357.x

Hovland, C. I., & Weiss, W. (1951). The influence of source credibility on communication effectiveness. Public Opinion Quarterly, 15, 635—650. http://dx.doi.org/10.1086/266350

Kernis, M. H., Cornell, D. P., Sun, C.-R., Berry, A., & Harlow, T. (1993). There’s more to self-esteem than whether it is high or low: The importance of stability of self-esteem. Journal of Personality and Social Psychology, 65, 1190—1204. http://dx.doi.org/10.1037/0022-3514.65.6.1190

Nel, E., Helmreich, R., & Aronson, E. (1969). Opinion change in the advocate as a function of the persuasibility of his audience: A clarification of the meaning of dissonance. Journal of Personality and Social Psychology, 12, 117—124. http://dx.doi.org/10.1037/h0027566

Rohan, M. J. (1996). The performance-integrity framework: A new solution to an old problem (Unpublished doctoral dissertation). University of Waterloo, Waterloo, Ontario, Canada.

Stone, J., Aronson, E., Crain, A. L., Winslow, M. P., & Fried, C. B. (1994). Inducing hypocrisy as a means of encouraging young adults to use condoms. Personality and Social Psychology Bulletin, 20, 116—128. http://dx.doi.org/10.1177/0146167294201012

Stone, J., & Fernandez, N. C. (2011). When thinking about less failure causes more dissonance: The effect of elaboration and recall on behavior change following hypocrisy. Social Influence, 6, 199—211. http://dx.doi.org/10.1080/15534510.2011.618368

Stone, J., & Focella, E. (2011). Hypocrisy, dissonance and the self-regulation processes that improve health. Self and Identity, 10, 295—303. http://dx.doi.org/10.1080/15298868.2010.538550

Swann, W. B., Jr. (1984). Quest for accuracy in person perception: A matter of pragmatics. Psychological Review, 91, 457—477. http://dx.doi.org/10.1037/0033-295X.91.4.457

Swann, W. B., Jr. (1990). To be adored or to be known? The interplay of self-enhancement and self-verification. In R. M. Sorrentino & E. T. Higgins (Eds.), Handbook of motivation and cognition (pp. 408—448). New York, NY: Guilford Press.

Swann, W. B., Jr. (1996). Self-traps: The elusive quest for higher self-esteem. New York, NY: Freeman.

Swann, W. B., Jr., & Pelham, B. W. (1988). The social construction of identity: Self-verification through friend and intimate selection (Unpublished manuscript). University of Texas at Austin.

Swann, W. B., Jr., & Read, S. J. (1981). Acquiring self-knowledge: The search for feedback that fits. Journal of Personality and Social Psychology, 41, 1119—1128. http://dx.doi.org/10.1037/0022-3514.41.6.1119

Thibodeau, R., & Aronson, E. (1992). Taking a closer look: Reasserting the role of the self-concept in dissonance theory. Personality and Social Psychology Bulletin, 18, 591—602. http://dx.doi.org/10.1177/0146167292185010

Waschull, S., & Kernis, M. (1996). Level and stability of self-esteem as predictors of children’s intrinsic motivation and reasons for anger. Personality and Social Psychology Bulletin, 22, 4—13. http://dx.doi.org/10.1177/0146167296221001

Zanna, M. P., & Cooper, J. (1974). Dissonance and the pill: An attribution approach to studying the arousal properties of dissonance. Journal of Personality and Social Psychology, 29, 703—709. http://dx.doi.org/10.1037/h0036651