Moral Rationalization and the Integration of Situational ... · PDF fileMoral Rationalization...

Moral Rationalization and the Integration of Situational Factors andPsychological Processes in Immoral Behavior

Jo-Ann TsangSouthern Methodist University

Moral rationalization is an individual’s ability to reinterpret his or her immoral actionsas, in fact, moral. It arises out of a conflict of motivations and a need to see the self asmoral. This article presents a model of evil behavior demonstrating how situationalfactors that obscure moral relevance can interact with moral rationalization and lead toa violation of moral principles. Concepts such as cognitive dissonance and self-affirmation are used to explain the processes underlying moral rationalization, anddifferent possible methods of moral rationalization are described. Also, research onmoral rationalization and its prevention is reviewed.

Religious scholars, philosophers, and lay-people alike have been puzzled for centuriesover the problem of evil. When horrendousatrocities such as the Holocaust occur, peoplescramble for explanations, but they seem toraise more questions than answers. How could agroup like the Nazis get away with such ex-treme immorality? Why did entire societiesseem to close their eyes to the evil around them?How can we act to prevent such moral mon-strosities in the future?

To answer these questions, people often turnto dispositional explanations to describe the or-igin of evil behavior (see Darley, 1992, for anextended discussion of naive theories of evil).Laypeople often posit that pathology or charac-ter flaws cause evil individuals to disregardmoral standards and commit evil. Psychologistshave long stated that moral standards are im-portant in bringing control and order into ourrelations with others and to society as a whole(e.g., Freud, 1930). These standards are learnedearly in life so that by the time most individualsbecome adults, they have internalized whatBandura (1990, 1991) termed “moral self-sanc-tions,” what Freud labeled the “superego,” andwhat is called in everyday language our“conscience.”

Whatever the name given to these internal-ized moral principles, it is clear that morality isso ingrained in people that they are not easilyignored (Bandura, 1991). But how does oneexplain, then, the prevalence of crime through-out all societies? A dispositional approachmight focus on the idea that because of certainpsychopathologies, some people have failed tointernalize the moral standards of society andthus perform evil deeds with impunity. Or at thevery least, immoral individuals have certainpersonality characteristics that predispose themto psychological processes that make it easier toact immorally (e.g., Post, 1990). In other words,evil actions come from evil people.

In contrast, much current research has begunto move away from an exclusively dispositionalmodel of the pathological evil actor to moreinteractional explanations of evil. Rather thanoriginating from a few evil people, evil arisesfrom a combination of situational (e.g., Darley,1992; Zimbardo, 1995) and psychological fac-tors present in the majority of individuals (Ban-dura, 1999). For example, certain situationaland psychological factors can prevent peoplefrom realizing how their behavior violates theirmoral principles. Evil in this case arises fromour failure to activate our moral standards. Thissocial psychological explanation claims that allof us have the potential to commit evil actions,given the right circumstances. Once individualsrealize the moral ramifications of their actions,they can work actively to convince themselvesand others that their evil behavior falls withinmoral standards.

I would like to thank C. Daniel Batson for his manyhelpful reviews of this article.

Correspondence concerning this article should be ad-dressed to Jo-Ann Tsang, Department of Psychology,Southern Methodist University, Dallas, Texas 75275. E-mail: [email protected]

Review of General Psychology Copyright 2002 by the Educational Publishing Foundation2002, Vol. 6, No. 1, 25–50 1089-2680/02/$5.00 DOI: 10.1037//1089-2680.6.1.25

25

I call this psychological phenomenon moralrationalization: the cognitive process that indi-viduals use to convince themselves that theirbehavior does not violate their moral standards.The occurrence of moral rationalization is anormal and prevalent psychological phenome-non and can be involved in small unethical acts,such as cheating on taxes, as well as largeatrocities such as the Holocaust. An explanationof evil that incorporates moral rationalizationposits that people can violate their moral stan-dards because they have convinced themselvesthat their behavior is not immoral at all. Whenmoral rationalization interacts with certain situ-ational variables, extreme evil can result.

In this article, I attempt to show how bothsituational and rationalization factors are impor-tant concepts in understanding evil. I present amodel that illustrates how these two factorscontribute to immoral behavior. I also use psy-chological theories such as moral disengage-ment, cognitive dissonance, and self-affirmationto explain why individuals might feel the needto rationalize immoral behavior. Once research-ers are familiar with the different methods ofmoral rationalization and the processes behind

them, they can work to short-circuit the cycle ofevil, leading the actor to behave morally insteadof immorally. To illustrate these points, I relyheavily on examples from the Holocaust ofWorld War II, one of the epitomes of extremeevil in contemporary society.

Precursors to Moral Rationalization:Values, Perceptions, and Motivations

Figure 1 illustrates a model of moral ratio-nalization and evil behavior. (The terms evil,immoral, and unethical are used interchange-ably in this article to refer to any behavior thatobjectively violates the actor’s moral princi-ples.) In this model, the individual is faced withthe choice of either upholding moral principlesor engaging in moral rationalization to justify apotentially immoral behavior, allowing him orher to violate moral principles but still preservethe semblance of being moral. The first assump-tion of the model is that morality is valued bythe individual. Second, the individual needs toperceive that moral principles are relevant to theparticular situation. Certain structures present inthe situation, such as an obedience-induced

Figure 1. Model of moral rationalization and evil behavior.

26 TSANG

agentic state, or routinization, can prevent mo-rality from becoming salient. Third, once mo-rality is salient, the individual weighs the costsand benefits of acting morally versus choosing apotentially immoral behavior. Acting to upholdmoral principles can be a benefit, in that theindividual can receive praise from the self andperhaps others. However, if that person hasother motivations that compete with morality,and if the goals of these motivations are mutu-ally exclusive with moral goals, then behavingmorally will carry the cost of forfeiting thegoals of the competing motivations.

Acting on competing motives carries its owncost, that of self-condemnation for engaging inimmoral behavior. If the ratio of costs to bene-fits of acting morally is high, then instead ofchoosing to uphold moral principles, the indi-vidual may instead engage in moral rationaliza-tion and reconstrue potentially immoral behav-ior as being moral or at least irrelevant to mo-rality. This lowers the cost of acting immorallyby reconstruing the act as being in fact moral. Ifmoral rationalization and evil action occurs, thissets the stage for escalation: the increased like-lihood and intensity of further evil behavior.These steps leading to moral rationalization andevil action are affected by different factors. Ilook at the factors that affect an individual’sdecision at each point in this moral decision-making process.

The Universal Value of Morality

Ironically, rationalization of immoral behav-ior begins with the high value that individualsplace on moral standards. If people did not careabout upholding moral standards, there wouldbe no need to rationalize immoral behavior intobeing moral. Morality itself emerges out of thesocial nature of human beings. Society developsin part because an individual who works inconcert with others can reach goals that cannotbe reached alone. Yet, along with these socialbenefits come conflicts arising from people’scompeting goals. Because of this, in every gath-ering of individuals there arise moral principlesthat govern people’s interactions with one an-other. Thus, some form of morality is present inand valued by all societies (Nowell-Smith,1967). Within each society, individuals areraised with a specific set of moral codes that aregiven almost sacred importance. Well-social-

ized people from any culture should hencestrive to be moral or at least make efforts torefrain from being seen as immoral.

Because the majority of individuals placesuch a high value on upholding moral standards,they cannot simply violate their moral princi-ples when they wish to reach goals in conflictwith morality. Instead, individuals rationalize tothemselves that their actions still fall withinmoral standards. But before this rationalizationoccurs, the individual has to perceive that moralprinciples are relevant to the situation at hand.These perceptions of the situation are the sec-ond step on the road to moral rationalization.

Perceptions of the Situation

Certain elements present in the situation canobscure the moral relevance of one’s evil ac-tions. These situational factors represent thefirst potential cause of evil behavior. At thispoint, the individual is not aware that his or herbehavior violates moral principles and thereforehas yet to engage in moral rationalization.These situational factors that can conceal moralrelevance include obedience, roles, deindividu-ation, routinization, norms, and the inaction ofothers. In addition, the specific context of evilorganizations brings together a number of thesesituational factors, increasing the potential forimmoral behavior.

Obedience to authority and the agentic state.Milgram’s (1974) classic obedience studies areprime examples of the strong effect situationalfactors can have on moral behavior. In thesestudies, a naive participant was ordered by anexperimenter to shock a confederate. WhenMilgram originally asked other psychologists topredict the outcome of this experiment, the vastmajority of them hypothesized that only apathological few would obey the experiment-er’s commands to incrementally increase theshock to dangerous levels. The prediction ofknowledgeable researchers was, in fact, a dis-positional one: The obedient person who obeysevil commands is sadistic and ill. In running theexperiment, Milgram quickly discovered thestrength of the experimental situation over dis-position, as about 65% of participants in theoriginal studies obeyed orders to shock a con-federate the maximum 450 volts. In the inter-views following the experiment, many partici-pants claimed that, though they themselves dis-

27MORAL RATIONALIZATION

liked shocking the other person, they feltobliged by the experimenter’s orders to con-tinue. Rather than perceiving their actions asimmoral, these participants pointed to the situ-ation as the cause of their harmful behavior.

Milgram (1974) explained the participants’obedient behavior in terms of an “agentic state.”He believed that the individual in an obediencesituation moves from an autonomous state ofacting on his or her own initiative to an agenticstate that he described as “ the condition a per-son is in when he sees himself as an agent forcarrying out another person’s wishes” (Mil-gram, 1974, p. 133). Similarly, Bandura (1999)pointed out that many individuals in an obedi-ence situation have a shift in attention from theirresponsibility as moral agents to their duty asobedient subordinates. Rather than being com-pletely agentless, “good” obedient individualsfeel a strong duty to carry out their orders well.Because they are focusing on the importance ofobedience, these individuals might not realizetheir own part in evil behavior.

So circumstances requiring obedience cancause an individual to perceive a situation asone in which moral questions do not need to beasked, at least not by the actor. Rather thanfocusing on the moral ramification of actions,the individual focuses on the duty to obey andinterprets the situation in terms of how well heor she is following orders. If the obedient indi-vidual does not perceive his or her actions interms of moral principles, there is no need forrationalization; it is instead the responsibility ofthe person in authority to sort through the moralissues.

This shift of attention from moral values tothe duty to obey can be seen in some of thequalifications given by Eichmann of his guilt inthe Holocaust. He claimed that “what he haddone was a crime only in retrospect, and he hadalways been a law-abiding citizen, because Hit-ler’s orders, which he had certainly executed tothe best of his ability, had possessed the ‘ forceof the law’ in the Third Reich” (Arendt, 1963, p.24). Eichmann asserted that, at the time, his rolein the Holocaust did not seem to contradictmoral principles. Rather, the situation called forhim to be an obedient soldier, which he was toan extreme.

The effect of roles. One of the reasons whyindividuals in an authority situation may notrealize the moral relevance of their actions is

that they focus on their role as an obedientsubordinate (Kelman & Hamilton, 1989). Blu-menthal (1999) theorized that a focus on roles isan important contributor to immoral behavior. Ifa given role includes aggressive or unethicalbehavior, individuals may engage in that behav-ior without realizing that it violates their moralprinciples. Ordinary people who are given rolesof oppressive power over others will begin todisplay pathologically aggressive behavioragainst the powerless (Staub, 1989).

Haney, Banks, and Zimbardo (1973) demon-strated the powerful effects of roles in theirStanford prison experiment. They randomly as-signed normal, well-adjusted male college stu-dents to be either prisoners or guards in a sim-ulated prison. They found marked differences inattitudes and behaviors between the two groups.“Guards” became increasingly aggressive to-ward the “prisoners” as the study progressed.The majority of their interactions with the pris-oners took the form of commands, threats, andpunishments. Those guards who were not ag-gressive did nothing to speak out against theaggression of their peers. In contrast, the pris-oners became increasingly passive with time. Afew prisoners, though judged as emotionallystable before the study began, displayed severeemotional breakdowns and had to be releasedearly from the study. Prisoners fell into theirrole from the beginning, obeying intrusive anddegrading commands from the guards, whereasthe guards quickly became facile in their role ofcontrolling the prisoners. Although there were afew weak effects of personality, the authorsconcluded that the biggest influence on partici-pants’ behavior was the role assignment. TheStanford prison experiment is a stunning exam-ple of how roles can influence behavior andalter moral perceptions in a situation.

Kelman and Hamilton (1989) argued that aninteraction between personality characteristicsand roles contributes to immoral behavior. Spe-cifically, they believed that individual differ-ences in political orientation could predict peo-ple’s behavior in obedience situations. An indi-vidual with a rule orientation complies withsociety’s laws and customs to garner reward andavoid punishment. People with a rule orienta-tion are obedient to authority only when theauthority has direct influence over their rein-forcement. On the other extreme, an individualwith a values orientation has internalized the

28 TSANG

values and moral principles of society. The obe-dience of those with a values orientation isconditional on whether orders from authorityare consistent with their internalized values. Anindividual can also have a role orientation,identifying with a particular role within society,such as the role of a good citizen. Those with arole orientation are the most obedient to author-ity: They do not need direct rewards or punish-ments to comply with orders, and they are lesslikely to question the orders’ morality. Individ-uals with strong role orientations might not evenperceive the immorality of an evil order fromauthority; their focus would instead be on theirmoral roles as obedient subjects. According toKelman and Hamilton (1989), these individualdifferences in political orientation can affectpeople’s sensitivity to roles in a given situation,leading to more or less obedience to unethicalorders.

Roles may have partially influenced Naziconcentration camp guards. Although manyguards may have been self-selected (Staub,1989), it is also possible that merely beingplaced in a position of power caused some non-pathological individuals to display the aggres-sive behaviors that came with their role as aguard. Many Nazis may have had a role orien-tation, causing them to pay less attention to themoral ramifications of their actions and moreattention to their role as good Nazi party mem-bers in the killing of Jewish people. In this way,individuals’ roles in a situation can reduce thesalience of relevant moral principles and aid inimmoral behavior.

Deindividuation. The concept of deindi-viduation is another example of how situationalfactors can obscure moral relevance. Certainaspects of a situation can cause people to losetheir sense of individual identity, allowing themto engage in behavior that is discrepant withtheir internal standards. For instance, individu-als in a crowd of onlookers obscured by dark-ness have been known to encourage suicidalpeople to jump to their death (Mann, 1981).However, with smaller crowds in daylight, jeer-ing is less likely to occur. In the Stanford prisonexperiment, the guards were given identical uni-forms with reflecting sunglasses to promotedeindividuation (Haney et al., 1973). Zimbardo(1995) explained deindividuation by stating thatan increase in anonymity increases immoral be-havior by reducing cues for social evaluation

and self-evaluation. However, when the situa-tion changes to make individuals more self-aware, they are less likely to engage in the samebehavior (Diener, 1980). In this way, the dein-dividuating structure of the situation can serveto disguise divergences between behavior andmoral standards or, instead, make discrepanciessalient.

Routinization. Circumstances that are taskoriented and routine can also obscure the rele-vance of moral principles. Kelman and Hamil-ton (1989) described how tasks become routin-ized when actions are so structured that there isno chance to question the morality of the situ-ation. People see themselves as small cogs in ahuge machine. Routinization causes individualsto avoid the implications of what they do byencouraging them to focus on the details of thejob rather than the meaning of the tasks. Thesituation is then transformed from one of moralright and wrong to a focus on the process andperformance of a specific assignment.

Examples of routinization can be found inEichmann’s description of being sent to Bra-tislava to discuss mass evacuations of Jewishpeople (Arendt, 1963). Rather than focusing onhis memory of the arrangements that were madebetween him and his superiors there, Eichmanninstead recalled such details as bowling withSano Mach and hearing of the assassination ofHeydrich. Not until he was further probed didEichmann even remember that he had been sentto Bratislava to organize evacuations. Arendt(1963) speculated as to why Eichmann focusedon what he did: “To evacuate and deport Jewshad become routine business; what stuck in hismind was bowling” (p. 82). For Eichmann andothers, making arrangements to eliminate theJews was just a routine task.

Todorov (1996), in his discussion of moralityin the concentration camps, commented onEichmann’s routinization of his duties: “Eich-mann focuses his attention exclusively on themethods of execution, never on what is at stakein the action itself” (p. 174). Although the rou-tinization inherent in Eichmann’s tasks mayhave prevented him from realizing the immo-rality of his actions, it is likely that after hiscapture, when he was being charged with warcrimes, Eichmann also strategically portrayedhis actions as routinized to rationalize his be-havior to himself and to the people putting himon trial. Routinization can thus be used as a


rationalization for purposefully immoral behav-ior, but it first serves to hinder perceptions ofmoral relevance.

Euphemisms—sanitized words or phrases—can contribute to routinization (Kelman &Hamilton, 1989). If wrongful behavior is calledsomething less thought provoking, then actorscan view their tasks as just part of their jobs.The labeling of the Holocaust as “ the final so-lution to the Jewish problem” allowed manyNazis to routinize their actions by focusing onsolving the “problem,” taking the focus off ofthe moral implications of their actions.

Depersonalization of the actor. Other as-pects of the situation can cause individuals tofeel depersonalized and take away their feelingsof moral agency. Todorov (1996) combined theideas of obedience, routinization, and deindi-viduation and applied them to concentrationcamp guards in his description of depersonal-ization of perpetrators. He stated that actorsbecame depersonalized when they were con-verted into a means to another end. Similar toMilgram (1974), Todorov claimed that the veryact of obedience or focus on a particular taskwithout consideration of its implications leadsto depersonalization: “Someone who only fol-lows orders is no longer a person” (1996, p.166). He then pointed out that the totalitariangovernments under which concentration campsfunctioned held up submission and obedience asvirtues. Because of this emphasis on obedience,concentration camp guards were reduced to ameans to another end, and the question of moralchoice no longer existed for them. Todorovstated that when depersonalization occurs, per-petrators of evil no longer see themselves asthinking, breathing human beings with worth oftheir own. Instead, as with routinization, actorssee themselves as cogs in the bigger machine ofthe political movement. For example, becausemany members of the Nazi party felt as if theywere tools for a greater cause, they felt strippedof volition to choose between good and evil andno longer saw themselves as responsible fortheir actions, moral or otherwise.

Depersonalization is also related to deindi-viduation, in that factors such as the arousal thatcomes with wartime, the feelings of transcen-dence and anonymity from being a member ofan organization such as the Nazis, and the dif-fusion of responsibility inherent in the concen-tration camp system could have deindividuated

people by obscuring cognitions of discrepanciesand responsibility, making it harder for individ-uals to interpret the situation as morally rele-vant. However, although deindividuation issomething imposed on the individual throughthe situation, Todorov (1996) characterized de-personalization as something individuals di-rectly seek out. In other words, people have anactive role in their own depersonalization. Forinstance, because submission was highly valuedin the Nazi party, Nazis would strive to fulfilltheir duties of obedience, actively seeking tobecome depersonalized. In this way, individualscan interact with the situation to obscure themoral relevance of evil action.

The inaction of others. The reactions of vic-tims, bystanders, and other perpetrators in agiven situation can also affect one’s perceptionof moral relevance. For example, if victims donot protest harm done to them, the actor mightassume that the victims are willing to be harmedor that his or her behavior is not even harmful.In contrast, cries of injustice or protest fromvictims can make moral principles salient to theperpetrators, curbing their immoral actions(Staub, 1989). Bystander inaction can also workto keep moral principles from being salient.Latane and Darley’s (1970) concept of plural-istic ignorance asserts that the inaction of by-standers can prevent others from perceiving anemergency. This inaction could also work toprevent people from perceiving moral relevancein a situation. In turn, bystander protest in theface of immoral action can serve to bring moralprinciples to the fore (Staub, 1989). Finally,when perpetrators are seen to commit crimeswithout apparent remorse, they serve as models,teaching people that these acts are acceptable.Other potential actors then accept the moralityof the perpetrator’s action without question. Incontrast, if a perpetrator is seen as remorseful oras suffering punishment as a result of the im-moral action, others might be more likely torealize the relevance of morality in the situation.In this way, the actions of other people in thesituation can affect the salience of moralprinciples.

Evil organizations. One avenue throughwhich these situational factors become imple-mented is “evil organizations.” Darley (1992,1996) described how structures imposed on asituation by an institution such as a corruptedcorporation or military can propel individuals

30 TSANG

unknowingly into a spiral of immoral actions.These institutions have situational factors suchas obedience and routinization already in place,obscuring the moral ramifications of unethicalactions. It is not until some disaster occurs, andthe individual is made aware of the immoralityof his or her actions, that the need for rational-ization arises. Darley stated that at this point ofinitial awareness, if the individual chooses thepath of rationalization and cover-up, he or she istransformed into an evildoer. The Nazi govern-ment in Germany can be seen in terms of an evilorganization. It constructed euphemisms andelaborate systems of routinization and obedi-ence to disguise the nature of the “fi nal solu-tion.” Yet, although some members of the Naziparty may have started out oblivious to therelevance of morality in the Holocaust, becauseof the extremely immoral nature of this event itis probable that almost all perpetrators eventu-ally became aware of the violation of moralprinciples. This set the stage for rationalizationamong the Nazis, transforming many normal,law-abiding citizens into agents of evil.

Despite factors that can work to obscure therelevance of moral principles, oftentimes mo-rality does become salient. What affects what anindividual will do once he or she reaches thisawareness of moral relevance? Once moralprinciples are salient, the presence of other mo-tivations will be influential in determiningwhether one chooses to uphold morality or torationalize away moral principles.

Motivations

The conflict between morality and other,equally strong motivations creates the need formoral rationalization. If an individual has nostrong motivations that come into conflict withmoral principles, then moral action is lesscostly, and she or he will be more likely to actmorally. Yet, there are often motivationspresent that compete with moral motivations.Selfishness or even sympathy might then leadan individual to do something that violatesmoral principles (Batson, Klein, Highberger, &Shaw, 1995; Batson, Kobrynowicz, Dinner-stein, Kampf, & Wilson, 1997). In this section,I first discuss the nature of moral motivation andthen describe other motivations that can conflictwith morality.

Components of morality. Certain character-istics of an individual’s moral motivation mightmake that person more prone to rationalizeaway morality in favor of another motivation.An individual’s moral motivations can differalong four possible dimensions: (a) a focus onconsequences versus intentions, (b) whether themotivation toward morality is instrumental toanother goal or an ultimate goal in itself, (c)whether the moral motivation is conceived as an“ought” or a “want,” and (d) the direction of themoral motivation (a person might be motivatedto be moral or to not be immoral).

Philosophers make a distinction between mo-rality based on consequences (utilitarianism)and morality based on intentions (deontologicalethics). Utilitarianism, in its most basic form,posits that morality depends on the nature of theoutcome: The outcome that benefits the mostpeople is the most moral alternative (Smart,1967). Robin Hood, who stole from a few richpeople to redistribute wealth across many poorpeople, would be considered moral by utilitar-ian standards. In contrast, according to a deon-tological ethic, moral rules are independent ofoutcomes (Olson, 1972). Rather than focusingon consequences, deontological ethics focus onthe intention to uphold universal laws of moral-ity. Deontological ethics would posit that RobinHood acted immorally, in that the act of stealingis morally wrong and cannot be willed a uni-versal law.

The Western conception of morality has beeninfluenced by both deontological (Kant, 1788/1898) and utilitarian (Bentham, 1829/1983;Mill, 1861/1957) ethics. The balance betweenthe two can be seen in instances in which con-sequences are discrepant with intentions. Forinstance, the United States still carries a penaltyfor manslaughter; the consequence, the death ofothers, is serious, and even though the guiltyindividual did not intend for people to die, he orshe still has to be punished for their deaths.However, manslaughter, with its lack of inten-tion to kill, also carries a lesser punishment thanpremeditated murder, wherein the individualboth intends to kill and brings forth the conse-quence of another’s death.

It is useful to know the extent to which aparticular individual’s conception of morality isbased on intentions or consequences, becausean action leading to a breach of moral principlesbased on utilitarianism could still uphold deon-


tological moral principles, and vice versa.Knowing the components of an individual’s orculture’s morality would allow us to know whena person might perceive a violation of moralstandards.

Whether morality is an instrumental or ulti-mate goal also has important ramifications formoral action and rationalization. Someonemight be motivated to uphold moral principlesas a means to another end, for example, toreceive social and self-rewards or avoid punish-ments (Bandura, 1986; Nowell-Smith, 1967). Incontrast, another person might consider moral-ity an ultimate goal, upholding moral principlesas an end in itself (Nowell-Smith, 1967). Thisdistinction is potentially useful because it couldhelp predict whether or not an individual willuphold moral principles in varying situations.Specifically, in the case of a person motivatedtoward morality as a means to another end, ifthere arises another, less costly opportunity toreach the goal of appearing moral, that alterna-tive will be taken in place of upholding moralprinciples (Batson et al., 1997). However, if theindividual holds a strong moral motivation as anultimate goal, an opportunity to appear moralwithout upholding moral principles will not betaken.

Heider (1958) made a distinction between“ought” and “want,” and this distinction can beapplied specifically to moral motivation. Ac-cording to Heider, when a person wants some-thing, this desire is centered around the self.However, when the individual believes that heor she ought to do something, this motivation isindependent of the individual’s personal de-sires. In addition, although a person might wantone thing at one time and a different thing atanother, what someone ought to do is invariantacross time in similar situations. Finally, al-though “wants” can be particular to specificindividuals, “oughts” are universal and apply toall people (Heider, 1958).

Although moral motivations are by definition“oughts,” it is also possible that an individual’spersonal preference is for morality: The “ought”is also a “want.” The combination of “ought”and “want” would help an individual both uni-versally apply moral principles and be moti-vated to personally uphold them. Therefore,knowledge of Heider’s concepts of “ought” and“want” can be useful in predicting moral ratio-nalization: When only feelings of “ought” are

present, the motivation to act morally may beweak and moral rationalization might be morelikely to occur, whereas a combination of“ought” and “want” might lead an individualtoward moral action.

In a similar manner, an individual’s moralmotivations might belong to differing self-do-mains (Higgins, 1987). The motivation to be amoral person might be a part of one’s ideal self,which includes one’s hopes and aspirations, orit might be a part of one’s ought self, whichincludes one’s duties and obligations. Whereasmoral motivations might normally be catego-rized with one’s ought self, there may be timeswhen the individual views morality as part ofhis or her ideal self. In general, characteristicsassociated with the ideal self are related to thepresence or absence of positive outcomes,whereas characteristics associated with theought self are related to the presence or absenceof negative outcomes (Roney, Higgins, & Shah,1995). Roney et al. (1995) found that the type offeedback that individuals were given during atask affected their motivation and subsequentperformance: Participants who were given feed-back about negative outcomes (“You didn’ tmiss that one” ), which is related to the oughtself, were less motivated and performed worseon an anagram task than participants who weregiven feedback about positive outcomes (“Yougot that right” ), which is related to the ideal self.In the same way, individuals with a moral oughtself might focus on negative outcomes and beless motivated to act morally than individualswith a moral ideal self, and thus they would bemore likely to engage in moral rationalization.

Individuals may also possess different typesof moral goals. Diversity in moral goals can berelated to Higgins’s (1997) theory of regulatoryfocus. This theory distinguishes between a pro-motion focus, in which the individual strivestoward positive outcomes and is focused onaccomplishments and aspirations, and a preven-tion focus, in which the individual avoids neg-ative outcomes and is focused on safety, duty,and responsibility. Self-regulation related to theideal self is tied to a promotion focus, whereasself-regulation related to the ought self is tied toa prevention focus. These differences in regu-latory focus are related to the differential use ofgoal-promoting strategies: Individuals with apromotion focus tend to approach matches todesired end states, whereas individuals with a

32 TSANG

prevention focus tend to avoid mismatches todesired end states.

In terms of moral motivation, individualswith a promotion focus, who view morality asan aspiration, would be motivated to approachmatches to morality. Their goal is to be moral.This motivation might manifest itself in actionssuch as giving all of one’s belongings to thepoor or dedicating one’s life to civil rights ac-tivism. This promotion-focused moral motiva-tion may apply to only a select few, such asMahatma Gandhi or Mother Teresa. On theother hand, individuals with a prevention focus,who view morality as an obligation or respon-sibility, would be motivated to avoid mis-matches to immorality. Their goal is to not beimmoral. All the prevention-focused person hasto do is refrain from breaking certain prohibitivestandards, for example, not stealing, not mur-dering, and not breaking the law. Because awider number of behaviors fit into the categoryof not being immoral than fit into the categoryof being moral, it is much easier to not beimmoral than it is to be moral. It is thereforemost likely the former that individuals wouldhold as a goal. Because individuals have morelatitude in which to move when their moral mo-tivation is preventative, rationalization would bemade easier.

Bandura (1999) drew a parallel distinctionbetween inhibitive and proactive moral agency.Inhibitive moral agency occurs when an indi-vidual refrains from immoral action. In contrast,proactive moral agency, although including ab-stention from immorality, also consists of ac-tively engaging in humane behavior, even atgreat personal costs. The type of moral agencythat an individual possesses may make it moreor less likely that he or she will use moralrationalization.

Some evidence for the importance of the mo-tivation to “not be immoral” can be found inBaron’s (1995) studies on the motive to “do noharm.” When acting on a “do no harm” princi-ple, individuals become concerned with not tak-ing any action that might end up hurting some-one who would not have been harmed if theaction had not been taken, even if, overall, theaction benefits more people than it harms. Inother words, crimes of commission are seen asmore detrimental than crimes of omission.Thus, in certain circumstances, individuals be-lieve that they are acting morally as long as they

do not actively cause harm, whereas failing toact morally, an act of omission, is more tolera-ble and many times not even seen as immoral.

Alternative motivations. These different char-acteristics of moral motivation are brought intothe moral decision-making process and make itmore or less difficult to rationalize around moralprinciples. In addition, many variables can af-fect the motivations that are competing withmorality. One powerful factor that can influencemotivation is the presence of difficult life con-ditions. Staub (1989) proposed that difficult lifeconditions affect people’s hierarchy of motives,causing them to focus on their own psycholog-ical needs and making a society more conduciveto genocide. He believed that conditions such aswar, crime, political violence, and even eco-nomic insecurity can lead to an intense needwithin a population for a coherent, positive self-concept. This would be especially likely in so-cieties with cultural self-concepts that contain asense of superiority or collective self-doubt.

The needs triggered by harsh circumstanceslead to self-protective motives that find theiroutlet in aggression against others for the pur-poses of building a positive social identity andproviding a scapegoat for the difficult life con-ditions. Although the forming of groups to sat-isfy psychological needs and goals stemmingfrom difficult life conditions does not necessi-tate the additional formation of hostility towardanother group, certain factors facilitate inter-group aggression: differentiation of the in-group and out-group along with a devaluation ofthe out-group, respect for authority, a mono-lithic society, and a history of aggression. Cer-tain ideologies can shift a population’s tenden-cies and make it more prone to these processes(Staub, 1993).

Staub (1989) applied this model of genocideto the period before the Holocaust. Germany’sdifficult life conditions took the form of theTreaty of Versailles, imposed after the countrylost the war in 1918. The strict stipulations inthe treaty, as well as the loss of the war, wereexperienced as great humiliations to Germansand were discrepant with their cultural self-concept of superiority. Economic difficultiesstemming from the treaty and from a worldwidedepression also added to difficult life conditionsin Germany. These combined factors led to ad-ditional political instability and violence. All ofthese difficult life conditions made the German


people particularly receptive to Hitler, whopromised to restore economic stability and apowerful Germany and provided a scapegoat inthe Jewish people.

According to Staub’s personal goal theory,motivations competing with morality take theform of extreme needs arising out of extraordi-nary conditions or rampant nationalism. Histheory does not take into account the presenceof everyday motivations and how they mightcompete with morality. What role do normal,mundane motivations such as egoism play inmoral conflict and rationalization?

Bandura’s (1990, 1991) theory of moral dis-engagement can be applied to both everydayand extraordinary motivations that conflict withmorality. Bandura stated that when competingmotivations—be they simple self-interest ormore complex needs linked to difficult life con-ditions—become sufficiently strong, individu-als seek to violate their moral principles andfulfill these other goals. However, because peo-ple in a society have been raised with certainmoral self-sanctions that keep them in adher-ence to moral principles in the absence of ex-ternal reinforcements, they cannot simply be-have immorally. Instead, they need to disengagetheir moral self-sanctions to engage in non-moral behaviors without the self-condemnationthat the sanctions would normally bring. There-fore, according to the theory of moral disen-gagement, (a) people can be motivated to en-gage in actions that violate moral principles,and (b) moral self-sanctions need to be short-circuited to enable individuals to act immorally.Bandura labeled this short-circuiting of self-sanctions “moral disengagement,” and it isakin to what I have been calling “moralrationalization.”

Along with difficult life conditions and ev-eryday self-interest, obedience is another moti-vation that might come into conflict with moralprinciples. Milgram (1974) explicitly character-ized obedience as a motive that competes withmoral standards in an authority situation. Forsome individuals, a shift to the agentic state iscomplete, and other moral principles do notcome into play. Yet, for many others, the agen-tic state is only partially achieved. These indi-viduals may then experience acute conflict be-tween the duty to obey and competing moralmotives such as the motivation not to hurt aninnocent person. Individuals can engage in dif-

ferent techniques to release the strain associatedwith this conflict of motivations, the most dif-ficult of which is to disobey the authority, ef-fectively choosing moral motivations over obe-dience motivations.

These are only a few examples of motivationsthat can potentially compete with moral moti-vations. Any motivation that would lead to theviolation of an individual’s valued moral prin-ciples can conflict with moral motivations andlead to moral rationalization.

Dynamics of Moral Rationalization

In summary, there are a number of stepsleading to the decision to rationalize evil. Ini-tially, the individual internalizes the society’smoral standards. Certain aspects of the situationmay obscure the relevance of morality andcause people to engage in evil behavior withoutrationalization. Yet, at some point many indi-viduals do realize that their behavior violatestheir moral principles. When the moral rele-vance of their behavior becomes salient, indi-viduals are faced with motivational conflict.One can choose to behave in line with one’smoral principles, but often moral action iscostly and one’s competing motivations may bestrong. However, violating moral principles isalso costly. It is at this point that the individualcan reduce the cost of acting immorally byengaging in moral rationalization. In this sec-tion, I examine how motivated reasoning under-lies the process of moral rationalization. I alsoexplore why individuals would have the needfor moral rationalization and outline the differ-ent methods people can use to rationalize theirimmoral behavior.

Motivated Reasoning: The Mechanics ofMoral Rationalization

Moral rationalization allows people to con-vince themselves that their preferred unethicalchoice is consistent with moral standards. Kun-da’s (1990) theory of motivated reasoning canbe used to explain the cognitive component ofmoral rationalization. According to this theory,when individuals approach a situation with apreference toward a particular solution, thispreference distorts their cognitions in the direc-tion of the desired decision or interpretation. As

34 TSANG

Kunda (1990) put it, “people motivated to arriveat a particular conclusion attempt to be rationaland to construct a justification of their desiredconclusion that would persuade a dispassionateobserver” (p. 482). Individuals are motivatednot only to justify themselves to others but toconvince themselves of the rationality of theirconclusion. Because motivated reasoning willnot work if the individual blatantly distorts in-formation, the motivation toward a predeter-mined outcome and its effects on the decision-making process are largely unconscious (Kunda,1990). Baumeister and Newman (1994) charac-terized individuals using this preference-drivenprocess as “ intuitive lawyers,” in contrast to themore objective, accuracy-driven approach of“ intuitive scientists.”

Applying the theory of motivated reasoningto morality, people about to engage in moralrationalization have a preference for seeing theirdesired immoral behavior as still consistentwith moral principles. If individuals can thenproceed as “ intuitive lawyers” and conclude thata certain behavior does not in fact make themimmoral, then they can have their proverbialcake and eat it too. They approach the decisionwith a preference toward any plausible interpre-tation of events that would allow them to appearmoral and still choose the immoral action.

But why engage in motivated reasoning tobegin with? What drives individuals to perceivethemselves as moral? Although the theory ofmotivated reasoning can describe how moralrationalization works, it does not explain themotivation behind rationalization.

Why Rationalize?

A major assumption of moral rationalizationis that people need to perceive that their actionsare consistent with their valued moral standards.Theories such as moral disengagement, cogni-tive dissonance, and self-affirmation addressdifferent aspects of the motivation to avoid theappearance of immorality.

Moral disengagement. According to Ban-dura’s (1991, 1999) theory of moral disengage-ment, an individual, through social learning,internalizes moral principles as self-sanctionsthat bring self-worth when they are upheld andself-condemnation when they are violated. Be-fore a person can engage in behaviors that vio-late moral principles, he or she has to disengage

moral self-sanctions to avoid self-condemna-tion. From this perspective, moral rationaliza-tion is motivated by guilt avoidance.

Bandura and his colleagues have found lim-ited support for moral disengagement. Bandura,Barbaranelli, Caprara, and Pastorelli (1996) hadjuvenile delinquents complete a pencil-and-pa-per measure of moral disengagement askingthem whether they agreed or disagreed withstatements about their use of different mecha-nisms of disengagement such as dehumaniza-tion of the victim and displacement of respon-sibility. They then correlated these scales withteacher observations of delinquent behavior inparticipants. They found that participants usedvarious methods of moral disengagement butrelied more heavily on moral justification anddehumanization of victims. In addition, pathanalyses showed that methods of moral dis-engagement facilitated delinquent behaviorthrough reduction of prosocial motivation andanticipated guilt while promoting aggression-related cognitions and affect.

Although Bandura et al.’s (1996) study dem-onstrated that individuals have many differentmethods of moral disengagement at their dis-posal, it did not rule out the presence of othermotivations in addition to guilt avoidance. Be-cause self-report measures were used for suchvariables as moral disengagement and anticipa-tion of guilt, this study gives us the knowledgeof individuals’ stated motivations for their ac-tions but still leaves open the possibility of othermotivations that may remain unconscious toparticipants. For instance, individuals engagingin moral decision making might also experiencea drive toward consistency between the cogni-tion of a moral self and their behaviors. In thiscase, the methods of moral disengagement out-lined by Bandura (1991, 1999) could serve torationalize these cognitions into being consis-tent while also warding off guilt.

Cognitive dissonance. The need for consis-tency between a valuing of morality and themoral perceptions of one’s behavior can be ex-plained through the theory of cognitive disso-nance. Festinger (1957) stated that two cogni-tions are in a dissonant relationship when theopposite of one cognition arises from anothercognition. A cognitive dissonance perspectivewould be that dissonance arises in moral situa-tions because of socialization: Individuals havelearned the value of moral principles and thus


possess the need to see themselves as good andmoral. Whenever something occurs that is in-consistent with an individual’s moral self-con-cept or violates valued moral principles, theindividual experiences dissonance. Moral ratio-nalization would then be the individual’s at-tempt to eliminate this dissonance by addingcognitions and reconstruing the immoral act as,in fact, moral.

The fact that morality is universally valuedmakes the theory of cognitive dissonance par-ticularly relevant. E. Aronson (1969) stated thatat the center of dissonance theory were “situa-tions in which the self-concept or other firmexpectancies are involved—and in which mostpeople share the same self-concepts or otherfirm expectancies” (p. 29). Because most peoplein a given society share similar moral self-concepts but experience many other motiva-tions that compete with moral ones, there isgreat potential for individuals to feel dissonancein the realm of morality.

Dissonance may first arise when an individ-ual is contemplating engaging in a behavior thatmight violate his or her moral principles. Thiswould take the form of what Festinger (1964)called anticipated dissonance: If the individualchooses to be moral, then there will be disso-nance between wanting to do something im-moral and choosing not to; if the individualchooses the immoral option, then there will bedissonance in seeing the self as moral but actingimmorally. Although Festinger (1957) firststated that it is the postdecision period that ischaracterized by dissonance, whereas the pre-decision period is characterized by conflict, helater allowed the possibility of anticipated dis-sonance (Festinger, 1964). He outlined two pos-sible ways of reducing anticipated dissonance:The individual can decide that the decision is oflittle importance, or he or she can avoid makingthe decision altogether. An experiment con-ducted by Braden and Walster (1964) showedthat individuals will, in some cases, attempt toreduce anticipated dissonance by avoiding thedecision.

A tendency to avoid moral action was presentin many bystanders during the Holocaust. Agreat number of people living in Nazi-occupiedcountries in Europe felt morally opposed tohanding over their Jewish citizens to be killed.Yet, if they had openly protested or even co-vertly hindered the Nazis, their lives would

have been in great danger. On the other hand,failure to protest was dissonant with the cogni-tion that one thought the Nazis’ treatment of theJews was wrong. It is possible that individualswho wanted to remain safe and avoid protestingagainst the Nazis anticipated dissonance be-cause of their opposition to killing Jews. Insteadof facing this anticipated dissonance directly,many people did nothing; they neither openlysupported the Nazis nor protested the deporta-tion of the Jews. Similarly, an individual in thissituation might have also resolved anticipateddissonance by reducing the importance of thedecision, pointing to diffusion of responsibility:“ I am only one person, I cannot make a differ-ence, so it doesn’ t matter what I decide.” Butare avoiding the decision and reducing its im-portance the only two options available in thecognitive dissonance paradigm?

Festinger (1964) described the predecisionperiod as a time of objective weighing of twoequally attractive options. This might be thereason why he predicted only two ways to re-solve anticipated dissonance. In his character-ization of the predecision period, Festinger de-termined that “ the information gathering andevaluative activities with respect to the alterna-tives are impartial and objective” (p. 8), that theindividual is not biased in favor of one alterna-tive or the other. If the individual is in favor ofa particular alternative, he or she can be said tohave already made the decision. Thus, Festingermade a clear distinction between the predeci-sion period and the postdecision period; it isonly after the decision is made that a differenceof preference for one alternative over the othershould be seen. This assumes that the individ-ual approaches the situation as an “ intuitivescientist,” with the motivation to make theobjectively best choice possible between twoalternatives.

But if we reframe the situation in terms ofBaumeister and Newman’s (1994) “ intuitivelawyer,” there emerges the possibility of theindividual being predisposed toward a particu-lar outcome before he or she has consciouslymade a decision as to whether to engage in abehavior. If this specific behavior toward whichthe individual is motivated happens to violatemoral principles, this may cause dissonance be-tween the individual’s cognition of a moral selfand the fact that an immoral behavior is insteaddesired. According to Baumeister and Newman,

36 TSANG

the individual will not blatantly distort the de-cision process to pursue an immoral action;instead, he or she will resolve the dissonance byengaging in subtle moral rationalizations so asto behave immorally but still appear moral. Inthis way, a dissonance-driven desire to restoreconsistency by appearing moral can act as amotivator in the motivated reasoning of moralrationalization.

Thus, cognitive dissonance can instigatemoral rationalization in the period before anyimmoral action has been taken. In addition,consistency motivations might also be relevantin circumstances wherein individuals realizethat behaviors in which they are currently en-gaged violate moral standards. In this scenario,situational factors have prevented moral princi-ples from being salient, and the individual actsimmorally without realizing that he or she isviolating those principles. However, at a certainpoint the individual realizes the inappropriate-ness of the action, and cognitive dissonance istriggered.

Cognitive dissonance is classically discussedas a phenomenon that occurs after an action hasbeen taken. When an individual has violatedmoral principles, the fact that he or she actedimmorally is dissonant with his or her self-concept of being a good and moral person (E.Aronson, 1969). Alternatively, the individualcan experience dissonance from the fact that heor she values particular moral principles but isnot acting consistently with those principles.Faced with these inconsistent cognitions, theindividual then has to find a way to reduce thisdissonance. In the traditional dissonance para-digm (Festinger, 1957), the individual canchange the action, change the situation, or addcognitive elements that are consonant with themost resistant cognition.

In the moral domain, changing the actionmay manifest itself as a choice to act morally.Although the individual may not be able to undothe past immoral behavior, he or she can decideto no longer engage in that behavior. For exam-ple, Nazis stationed in Denmark had the moralramifications of their actions made salient tothem by the opposition of Danish citizens. Theythen potentially resolved their dissonance bydisobeying orders to deport Jews under theirjurisdiction (Arendt, 1963). Although changingone’s action may be the most moral option, it isnot always the easiest one. The immoral behav-

ior may be one that is rewarding for the indi-vidual in some way, or, conversely, changingthe action might carry with it grave conse-quences. For instance, many individuals whohelped hide Jews were caught and killed fortheir actions. It was widely known among theGerman ranks that “anyone who had seriouslyprotested or done anything against the killingunit would have been arrested within twenty-four hours and would have disappeared”(Arendt, 1963, p. 232). In cases such as these,people might choose other options for reducingdissonance, such as changing their cognitions.When individuals change their cognitions toresolve dissonance in a moral context, they areessentially engaging in moral rationalization.

One of the first dissonance-related experi-ments was indirectly associated with moral ra-tionalization. Festinger and Carlsmith (1959)found that when participants were given $1 tolie about how interesting a boring task was, theychanged their attitudes toward the task andfound it more interesting than participants whowere paid $20 to lie or participants who did notlie at all. Participants who were paid only $1 tolie about the task experienced cognitive disso-nance between the cognition that the task wasactually boring and a second cognition that theytold someone else that the task was interesting.Whereas those who were paid $20 for lying hadsufficient justification for lying, individuals whowere paid only $1 did not have a good reasonfor lying and had to resolve their dissonance bychanging their attitudes toward the task. Interms of moral rationalization, participants inthe $1 condition rationalized their unethical be-havior—lying to another participant withoutsufficient justification—by telling themselvesthat the task was not really that boring, andhence they were not really lying. In this way,Festinger and Carlsmith’s study can be inter-preted as supporting the role of cognitive dis-sonance in moral rationalization.

Self-affirmation. The theory of self-affirma-tion is another possible explanation for moralrationalization. Steele (1988) presented self-af-firmation as an alternative explanation to cog-nitive dissonance. He explained that when indi-viduals are faced with dissonant cognitions, ten-sion is produced not because of the lack ofconsistency between the cognitions but becausepeople’s intelligence and goodness are calledinto question (e.g., “ I just said that a boring task


was interesting, I must be a liar” or “ I’m smok-ing and I know it’s bad for me, I must be stupidor lack self-control” ). Individuals are then mo-tivated to reaffirm themselves as good. There-fore, rather than positing a consistency motive,self-affirmation posits that, within the disso-nance paradigm, inconsistency serves as a threatto self-integrity. Whereas cognitive dissonancerelates the moral motivation to a drive evokedby inconsistencies between cognitions, self-af-firmation presupposes the goal of maintaining apositive self-concept.

Unlike E. Aronson’s (1969) proposition thatthe clearest predictions in dissonance can bemade in relation to behaviors that are inconsis-tent with an individual’s self-concept, Steele’sconception of self-affirmation does not dealwith inconsistency at all. Instead, self-affirma-tion is a motivation to globally view the self asgood. It follows that when addressing a threat tothe self, an individual can reaffirm the self inways that would not resolve the apparent “ in-consistency.” For example, the smoker mightsay “ I may be smoking, but I’m an excellentworker” ; this has nothing to do with the incon-sistency between health and smoking, but itdoes counteract the feelings of badness that maystem from doing something that is unhealthy.

When individuals discover that their behav-iors have violated their moral principles, self-affirmation would predict that this realizationwould prompt them to reaffirm their own good-ness. This might be done in a manner consistentwith the predictions of cognitive dissonance:The immoral behavior can be stopped or thesituation redefined as amoral, for example.However, according to self-affirmation, becausepeople are motivated by self-integrity ratherthan self-consistency, they have more optionsthan these; they can even affirm parts of theirselves that are irrelevant to the immoral behav-ior. For instance, individuals who discover thatthey have done something immoral might asserttheir intelligence to make up for their lack ofmorality. A self-affirmation explanation forwhy we rationalize would then be broader thana dissonance explanation of moral rationaliza-tion. Moral rationalizers motivated by cognitivedissonance would have to rationalize immoralbehavior into being moral, whereas rationalizersmotivated by self-affirmation could admit thewrongness of their behavior if they affirmedanother equally valuable part of the self.

Several studies have demonstrated the effec-tiveness of self-affirmation opportunities in at-tenuating dissonance-reduction behaviors. Forexample, Steele and Liu (1983) placed individ-uals in a “ forced-compliance” situation inwhich they freely chose to write a counteratti-tudinal essay. In traditional dissonance studies,researchers have found that participants whofreely choose to write an essay that is inconsis-tent with their own personal opinions will re-solve their dissonance by later changing theiropinions to match the essay. Steele and Liufound that when participants were allowed to fillout a self-relevant value scale after writing acounterattitudinal essay, they showed less opin-ion change than both participants who did notfill out the scale and those for whom the scalewas not self-relevant. Self-affirmation, in theform of filling out a self-relevant value scale,decreased attitude change, suggesting that re-solving dissonance is less related to consistencymotivations and more related to the need to seethe self as good and moral.

Dissonance versus self-affirmation. Apply-ing the theory of self-affirmation to moral ra-tionalization underscores the importance ofmoral self-concept in the moral rationalizationprocess. However, because the theory of self-affirmation cannot completely account for allcognitive dissonance phenomena, self-affirma-tion is not the best explanation for moral ratio-nalization. It is beyond the scope of this articleto outline all of the evidence in favor of con-sistency motivations, but I present a few exam-ples of research showing that cognitive disso-nance does exist independently of self-affirma-tion. For instance, Simon, Greenberg, andBrehm (1995) demonstrated that the mecha-nisms behind cognitive dissonance are indeedconsistency related. In their experiments, par-ticipants were exposed to issues that were notimportant to them but were important in gen-eral. Simon et al. found that this served toreduce dissonance and thus attenuate attitudechange after participants wrote a counterattitu-dinal essay. Their studies suggest that self-af-firmation may actually be a method of disso-nance reduction that allows an individual toreduce the importance of the inconsistent rela-tionship by making other, more important val-ues salient, a method they labeled “ trivializa-tion.” Simon et al. (1995) concluded that self-

38 TSANG

affirmation results are in fact related tocognitive consistency.

Whereas Steele (1988) emphasized how fo-cusing on self-relevant values can work to af-firm the self and remove threats presented byinconsistency, J. Aronson, Blanton, and Cooper(1995) posited that self-focus can both reducethreat and increase it. They pointed out that, inself-affirmation studies such as that of Steeleand Liu (1983), participants were allowed toself-affirm in a realm unrelated to their incon-sistent behavior. They predicted that if partici-pants were made to focus instead on valuesclosely related to their inconsistent action, thiswould serve to increase threat rather than de-crease it, functioning as a self-discrepancyrather than self-affirmation.

Their studies showed that, when given achoice, participants who engaged in counterat-titudinal behaviors did not affirm aspects of theself that were related to their inconsistent be-havior. Instead, they affirmed aspects of the selfthat were unrelated to the inconsistency, buteven this affirmation of unrelated traits did notkeep participants from changing their attitude inthe typical counterattitudinal direction. In otherwords, participants were addressing their cog-nitive inconsistencies, even when given oppor-tunities to self-affirm. Because participantschose to focus on traits unrelated to their incon-sistent behaviors and reduce the importance oftraits related to the inconsistency, Aronson etal.’s (1995) results are similar to Simon et al.’s(1995) findings on trivialization. Therefore,rather than self-affirmation, the primary moti-vation in counterattitudinal situations seems tobe that of resolving inconsistencies.

Likewise, the motivation behind moral ratio-nalization is unlikely to be primarily self-affir-mation. However, exploring self-affirmationpredictions regarding rationalization helps tounderscore the fact that inconsistencies thatmight cause individuals to rationalize are inti-mately related to the self-concept.

The most viable candidates for the motiva-tion behind moral rationalization are moral self-sanctions and cognitive dissonance. Most likelya combination of these motivations is actingwithin the individual when he or she is makinga moral decision. Individuals may be simulta-neously motivated to avoid self-condemnationand garner praise, as well as perceive them-selves as acting consistently with valued moral

principles. Ultimately, the question of whichtheory best explains moral rationalization isopen to empirical investigation.

How Do We Rationalize?Methods of Moral Rationalization

What, then, are the specific methods of moralrationalization that people can use to reduce thecosts of immoral actions? In what ways canindividuals’ motivations cause them to recruitcognitions that make them appear moral? Ban-dura’s (1991, 1999) theory of moral disengage-ment identifies four different categories of ra-tionalization that can lead to moral disengage-ment: reconstruing conduct, obscuring personalagency, disregarding negative consequences,and blaming and dehumanizing victims. In ad-dition to these mechanisms, individuals can ra-tionalize by focusing on the actions of othersand by fragmenting their identity. These meth-ods of moral rationalization function to changeindividuals’ perceptions of the behavior as fall-ing within moral standards.

Reconstruing Conduct

Moral justification. Individuals can ratio-nalize by directly addressing the nature of theirimmoral actions, reconstruing their conduct tomake potentially immoral behavior appear pos-itive (Bandura, 1991, 1999). This can be accom-plished in a number of ways. The method ofmoral justification reconstrues conduct by por-traying it as being in the service of a valuedsocial or moral purpose. For example, the Nazigovernment of Germany made use of moraljustification in representing the genocide ofJewish people not as murder but as “ the ‘holiesthuman right and . . . obligation,’ which is ‘ to seeto it that the blood is preserved pure and, bypreserving the best humanity, to create the pos-sibility of a nobler development of these be-ings’” (Lifton, 1986, p. 431). For the Nazis, themurder of innocent people was transformed intheir minds into an almost sacred duty.

Euphemistic language. Immoral actionscan also be rationalized to be morally accept-able through the use of carefully selected wordsand phrases that depict immoral actions asharmless and individuals as less causally linkedto their actions (Bandura, 1991, 1999). The Na-


zis had knowledge of the power of euphemisms,and Arendt (1963) described them as construct-ing elaborate “ language rules” that dictated theavoidance of words such as extermination, liq-uidation, and killing in reference to the Holo-caust; instead, the terms final solution, evacua-tion, and special treatment were to be used. Theterm concentration camp itself was a euphe-mism; it referred to the camp’s function in con-centrating the Jewish people in one area butmade no mention of the horrors that occurredthere. Yet, apparently this euphemism was notgood enough for the Nazis, because concentra-tion camps such as Auschwitz were labeled“charitable foundations for institutional care”(Arendt, 1963, p. 109). These language ruleswere used not to keep people ignorant of themassacres but to “prevent them from equating[the massacres] with their old, ‘normal’ knowl-edge of murder and lies” (p. 86); if the deedswere given a cleaner name, individuals couldrationalize the immorality out of their actions.

Euphemistic language might already be inplace when an individual enters a particularsituation. In that case, euphemisms would workto affect the individual’s perceptions of moralrelevance as well as serve as tools for rational-ization. However, the original instigation of eu-phemistic language will have been for the pur-poses of active rationalization.

Advantageous comparison. Individuals canalso rationalize their evil behavior by compar-ing their own immoral actions with somethingworse (Bandura, 1991, 1999). Nazi doctors whogave lethal injections to Jewish prisoners inconcentration camps could have used advanta-geous comparison by comparing their methodof killing with the more painful method of ex-ecution by shooting. In this way, they couldhave reinterpreted their actions as “mercy kill-ings” rather than cold-blooded killings.

Obscuring Personal Causal Agency

Not only can immoral behavior be rational-ized by reconstruing it into something positive,but the individual can also preserve his or hermoral self-concept by separating the self fromthe immoral action (Bandura, 1991, 1999). If theindividual does not feel personally responsible forevil actions, then he or she need not experiencediscomfort about the violation of moral principlesthat those actions might entail. People can obscure

personal causal agency through displacement anddiffusion of responsibility.

Displacement of responsibility. The obscur-ing of causal agency can occur in displacementof responsibility, when an individual is orderedor feels compelled to perform a particular ac-tivity. Rather than blaming the self, the individ-ual blames those who gave the orders. Mil-gram’s (1974) obedience studies, used earlier todemonstrate how obedience situations can con-ceal the moral relevance of a situation, alsoshow how people might experience moral con-flict in a situation and displace responsibility soas to rationalize their immoral behavior. Forinstance, Milgram described how many of hisparticipants would show obvious signs of psy-chological distress during the experiment, suchas sweating, shaking, and even laughing hyster-ically. From these behaviors, one can infer thatthese particular participants clearly believedthat what they were doing was wrong; yet, whenasked why they continued to shock the confed-erate, many of these same participants wouldrationalize their behavior by pointing out thatthey were not responsible for harming the con-federate: It was the experimenter, the person inauthority, who held the responsibility.

Because obedience was highly valued withinNazi ideology, displacement of responsibilitybecame an effective means of rationalization forperpetrators in the Holocaust. In his own de-fense, Eichmann stated that he was simplyobeying Hitler’s orders, which had the force ofthe law (Arendt, 1963). His lawyer claimed thathe was only carrying out “‘ acts of the state,’what had happened to him might happen in thefuture to anyone” (p. 247). Eichmann may havebeen strategically using displacement of respon-sibility to rationalize his behavior to both him-self and those judging him.

Diffusion of responsibility. Individuals canalso deny their part of the responsibility for anegative consequence that resulted from com-bined action. Whereas displacement of respon-sibility is related to obedience to authority, dif-fusion of responsibility occurs when there is adivision of labor, with different members per-forming separate parts of a task (Darley, 1996).Diffusion of responsibility can also occur in theabsence of a division of labor, with the merepresence of other actors enough to deflect re-sponsibility onto others and away from the self.For example, group decision making can lead to

40 TSANG

diffusion of responsibility, each single individ-ual believing that the group, not himself orherself, is responsible for a poor decision. Inaddition, diffusion of responsibility can be fa-cilitated by group action. When harm is done bythe group, immorality can be attributed to oth-ers, or to the group, and not the self. As withdisplacement of responsibility, diffusion of re-sponsibility can serve to keep people from per-ceiving moral relevance as well as function asan active rationalization for immoral conduct.An individual, motivated to rationalize an im-moral action, may point out that he or she wasonly one of many perpetrators and therefore isnot the one responsible.

The Holocaust was set up as a group action,allowing diffusion of responsibility to serve as arationalization for individual perpetrators. Thus,Eichmann was able to state:

With the killing of Jews I had nothing to do. I neverkilled a Jew, or a non-Jew, for that matter—I neverkilled any human being. I never gave an order to killeither a Jew or a non-Jew; I just did not do it. (Arendt,1963, p. 22)

The nature of the Nazi killing machine enabledEichmann to make this claim; the labor was infact divided up, and he never really did have tokill anyone with his own hands or give directorders to kill. Instead, he gave orders for ship-ments of Jews to be moved to concentrationcamps, where, although death was certain, hewas not the one who directly caused the death.In fact, the work was divided up such that theindividuals directly involved in killing wereJewish prisoners themselves. Bandura (1999)pointed out that those in the middle level of ahierarchy of authority, who neither make deci-sions nor directly witness the consequences ofevil behavior, have the easiest time rationalizingtheir role in evil. In this way, people such asEichmann can convince themselves that theyare not responsible for actions that violate moralprinciples.

Eichmann also pointed to diffusion of re-sponsibility from group decision making in de-fending his actions. This was apparent when hedescribed being at the Wannsee conference,where he first got word of the “fi nal solution” ofkilling the Jews. Because no one spoke outagainst this violent action, Eichmann felt as ifthe blame for these murders would no longer behis: “Here now, during this conference, the mostprominent people had spoken, the Popes of the

Third Reich. . . . At that moment, I sensed akind of Pontius Pilate feeling, for I felt free ofguilt. Who was he to judge?” (Arendt, 1963, p.114).

Disregarding or Distorting theConsequences of an Action

Individuals can also disregard or distort theconsequences of immoral actions to preservetheir moral self-concept. Bandura (1991, 1999)delineated two ways in which consequences canbe disregarded or distorted. First, an individualcan use selective inattention and cognitiveavoidance of the effects, basically not facing theharm caused or actively trying to minimize thatharm. This was apparent in the previously men-tioned structure of the concentration camps,with Jewish prisoners the ones who were forcedto place the gas pellets into the gas chambers.Then Nazi soldiers did not have to actually seepeople dying or hear their screams. Anotherway consequences can be distorted is for indi-viduals to engage in active efforts to discreditevidence of harm. Even today, many neo-Nazisclaim that the statistics of people who died inthe Holocaust are greatly inflated or that theHolocaust did not actually happen at all.

Blaming and Dehumanizing Victims

Dehumanization. Another way to preserveone’s sense of morality while still behavingimmorally is by dehumanizing victims (Ban-dura, 1991, 1999). When individuals dehuman-ize victims, they take away the victims’ humanqualities or give them animal-like ones. Thevictims are then seen as subhuman, without thesame feelings or hopes as the perpetrators, andthus one can rationalize that normal moral prin-ciples do not apply. During World War II inGermany, the Nazi government enacted theNuremberg laws to dehumanize Jews before theHolocaust. These laws took away politicalrights from Jewish people in Germany; for in-stance, they were no longer considered citizens,but they were still members of the Germanstate. Sexual intercourse between Jews and Ger-mans was also prohibited, as were mixed mar-riages, and no German woman under the ageof 45 years was to be employed in a Jewishhome. The Nuremberg laws functioned to make


official the separation and dehumanization thatJews suffered in Germany preceding the Holo-caust and allowed other Germans to see Jews assubhuman.

Consistent with a theory of moral rationaliza-tion, Todorov (1996) characterized dehuman-ization of the victim as a technique that Nazisoldiers used to cope with the fact that theywere committing such evil acts. Todorov illus-trated the imposed transformation that the Nazisinflicted on the Jewish prisoners: how theyforced them to live in their own filth, becomescavengers because of starvation, and strip na-ked before being killed. These things made theprisoners seem more like animals than humanbeings, making it easier for guards to then killthem; yet, it was the guards themselves whoexacted these alterations in the first place.

Large numbers contributed to the dehuman-ization of the Jews and the rationalization oftheir massacre. Because of the huge amount ofpeople being killed, Nazi soldiers saw their pris-oners as masses of flesh, no longer individualpeople. Therefore, even though the large-scalemurder in the Holocaust made it all the moreatrocious, the fact that a huge amount of peoplewere killed actually made it easier to carry out.

Todorov explained how the Nazis also dehu-manized Jews by incorporating them into ab-stract categories. Hitler, by classifying Jewishpeople as dangerous “enemies of the people,”was better able to elicit cooperation from Ger-man citizens in mistreating Jews. The victimswere no longer human beings but members ofan abstract entity of “enemy.”

The minimal face-to-face interaction betweenguards and victims in the concentration campsalso served to dehumanize the Jews. Whenguards did not have much contact with theirprisoners, there was less chance for the human-ness of the prisoners to bring the guards tocompassion.

Euphemistic language can be important in theprocess of dehumanization. Todorov (1996)gave examples of how the Nazis referred to theJews as “pieces” or “ items” but never “people”(p. 160). These innocuous words made it easierfor the Nazis to think of the Jews as subhuman,placing them outside the realm of normal hu-man morality.

As was the case for euphemistic language,dehumanization can be present both beforemoral relevance is perceived and afterward. An

individual might enter a situation in which vic-tims are already dehumanized, and this couldwork to keep him or her from perceiving therelevance of moral principles in the situation.However, early use of dehumanization in a sit-uation will take the form of active rationaliza-tions for potentially immoral behavior. De-humanization is used by the first actors in asituation as a rationalization, affecting theperceptions of morality of later actors.

Attributions of blame. Perpetrators of evilcan also rationalize their actions by shiftingattributions of blame (Bandura, 1991, 1999).For instance, perpetrators can shift the blamefrom themselves to their victims (Lerner, 1980),as when the Nazis claimed that Jewish peopledeserved to be killed. Alternately, perpetratorscan blame the circumstances around them fortheir crimes. Actors might feel controlled byother people or by the situation and feel thatthey are faultless, helpless victims. For instance,many Nazi doctors who participated in forcedexperiments and killings in concentrationcamps felt morally opposed to these actions.Yet, they continued to engage in these actsbecause they believed that if they disobeyedorders, they would be killed (Lifton, 1986).Therefore, they were able to justify their actionsby assigning responsibility to their circum-stances and not to themselves.

Moral exclusion. Dehumanizing and blam-ing victims can aid moral rationalization byexcluding victims from the moral field. Moralexclusion occurs when individuals define a par-ticular group as outside the realm of morality(Opotow, 1990; Staub, 1989). Deutsch (1990)captured the rationalizing nature of moral ex-clusion: “Excluding the ‘ they’ from one’s moralcommunity permits one to consider oneself as amoral person even while one engages in whatwould normally be considered depraved ac-tions” (p. 24). Deutsch posited that moral ex-clusion was a psychological need that arose incertain individuals. In normal development, anindividual learns to integrate good and bad as-pects of the self and other people.

However, factors such as harsh circum-stances, an authoritarian family, or an ethnocen-tric culture may cause certain people to continueto separate good and bad. These individualswould be more prone to moral exclusion be-cause of the tendency to project bad aspects ofthe self onto out-groups. Along with disposi-

42 TSANG

tional differences in moral exclusion, Deutschdescribed different social conditions that mightfacilitate moral exclusion: difficult life condi-tions, an unstable government, authoritarian so-cial institutions, claims of superiority (racial,religious, and so forth), culturally sanctionedand salient violence, remoteness of victims, andno objecting observers of the violence. Thesedifferent social factors may make it easier forindividuals to use moral exclusion as a rational-ization for violence.

The Actions of Others

The actions of victims, bystanders, and otherperpetrators can all be used as rationalizationsfor immoral deeds. For example, perpetratorsmight point to lack of protest from victims as away to rationalize immoral behavior into some-thing benign. However, simply because the vic-tim does not protest does not mean that he orshe desires to be victimized. During the Holo-caust, many Jewish organizations encouragedJews to be passive, believing that Jewish coop-eration would prevent Nazi actions from wors-ening. Thus, there was a lack of protest from theJewish community, but by no means did theJews consent to what was being done to them.Instead, they saw passivity as their only optionin a situation in which the Nazi government hadsupport from much of the German population(Staub, 1989).

Just as actions of bystanders affect individu-als’ perceptions of moral relevance, they can beused to rationalize behavior. If bystanders donot react to the actor’s behavior, or if they infact show their support for the actor, then theactor can rationalize that his or her harmfulbehavior is moral. Arendt (1963) illustratedhow the inaction of bystanders allowed Eich-mann to rationalize his duties: “As Eichmanntold it, the most potent factor in the soothing ofhis own conscience was the simple fact that hecould see no one, no one at all, who actuallywas against the Final Solution” (p. 116). Be-cause there is no outright protest, the actor canassume that bystanders agree and assuage theguilt he or she may be feeling over immoralbehaviors. On the other hand, when there isbystander action, this often causes perpetratorsto change their behavior: Their avenues of ra-tionalization severely restricted, perpetrators arefaced with the fact that their actions are im-

moral, leading to a possible change in behavior.For instance, when Nazi soldiers stationed inDenmark during World War II were faced withstrong Danish opposition to the mistreatment ofJewish people, they themselves began to dis-obey orders from Germany and resisted thedeportation of the Jews under their jurisdiction(Arendt, 1963).

Finally, reactions of other perpetrators canalso be used to rationalize immorality. A per-petrator of evil actions can point to the actionsof fellow wrongdoers, adding consonant cogni-tions to resolve dissonance over his or her ownimmoral behavior (a la Festinger, Riecken, &Schachter, 1956). Use of the inaction of bothbystanders and perpetrators as a rationalizationpoints to the importance of social support inrationalization; it may be easier to convinceoneself that one’s potentially immoral actionsare in fact moral if there are others presentwho are also trying to rationalize the action’smorality.

Fragmentation

Not only can people look to others to ratio-nalize their own immoral behavior, but they canfragment their own identity to preserve a moralself-concept. Todorov (1996) characterized frag-mentation as a way that a perpetrator of evil canseparate the bad identity—the part of the indi-vidual that is committing evil acts—from thegood identity. Fragmentation can manifest itselfin a number of ways. Todorov illustrated howmany of the concentration camp guards showeddiscontinuity, a form of fragmentation charac-terized by unpredictable behavior (e.g., beingcompassionate one minute and evil the next). Inaddition, the guards exhibited discontinuities interms of their interests, showing themselves tobe cultured in literature or music, tastes thatseemed to clash with their sadistic behaviorstoward the prisoners.

Perpetrators of the Holocaust were also frag-mented in terms of their private and publiclives. Todorov (1996) described citizens of atotalitarian government: “By deciding to submit‘only’ in their outward behavior, in their publicwords and gestures, they would console them-selves with the thought that they remain mastersof their consciences, faithful to themselves intheir private lives” (p. 129). Therefore, individ-uals are able to rationalize acceptance or perpe-


tration of immoral behaviors as long as they donot privately agree with those actions, the“comforting illusion that in their innermostselves they remain honorable and pure”(Todorov, 1996, p. 129). In compartmentalizingtheir public and private lives, individuals canthen claim that, even though they are publiclybehaving in evil ways, their real, private iden-tities are as loving spouses or caring parents.Thus, concentration camp guards might writeheart-felt letters home, inquiring about theirchildren and wives. In doing this, they hopedthat “ their admirable private lives might serveas a counterweight, at least in their own minds,for the things that may have troubled themabout their professional activities” (Todorov,1996, p. 147).

Fragmentation as a method of moral rational-ization can be better understood in light of self-affirmation theory. In this view, participating inthe immorality of the concentration campswould have threatened the soldiers’ view ofthemselves as intelligent and good. By thenstating that, for instance, they were good fa-thers, they were able to restore a global sense ofself-worth. Examples from Todorov’s own de-scriptions provide support for a self-affirmationinterpretation. Todorov cited Lifton’s (1986) in-terview with Eduard Wirths, the head doctor atAuschwitz:

In Wirths’s mind, his love for his family seemed tomake up for the drawbacks of his professional life. . . .The faster the selections succeeded one another, themore he filled his letters with questions about thechildren’s first teeth and with commentaries on thephotographs she had sent him. (Lifton, 1986, as cited inTodorov, 1996, p. 146)

One of the jobs of the Nazi doctors was to makethe selections, meaning they had to look overthe Jewish prisoners directly after they steppedoff of the trains, choose those who were fit towork, and send the others off to be executed. Asthe selections elevated, Wirths’s self-concept asa good and moral person was increasinglythreatened. Here we see a relationship betweenan increase in selections and Wirths’s affirma-tion of himself as a good and loving husbandand father. Although Todorov described frag-mentation of one’s identity as having “ two non-communicating compartments,” a self-affirma-tion perspective would view fragmentation notin terms of different “parts” but in terms of theneed to affirm one’s global self as good.

Wirths’s letters home were an attempt to affirmthe morality of his overall character. Moral ra-tionalization is reached not through separatinginto a “good” and “bad” identity but by affirm-ing the entire self as good and moral. The theoryof self-affirmation is thus especially relevantto fragmentation as a form of moralrationalization.

To summarize, while making the decision ofwhether to engage in moral versus immoralaction, individuals may use specific methods ofmoral rationalization. These rationalizations aidin the process of motivated reasoning, allowingpeople to conclude that the desired immoralbehavior is actually moral. They are then able toviolate their moral principles while still main-taining a moral self-concept. Moral rationaliza-tion allows normal individuals to engage insometimes horrific behaviors. Yet, the processdoes not end here. Once an individual usesmoral rationalization to make evil behaviorseem acceptable, a process of escalation canbegin wherein the individual is led into furtherviolations of moral standards so that small un-ethical acts incrementally increase into enor-mous acts of evil.

Escalation

Moral rationalization not only allows individ-uals to preserve their moral self-concept whencommitting immoral behavior; it also has thegrave consequence of allowing evil to continue.If people do not realize the evil in their actions,they can continue their behavior or even in-crease it while posing little threat to their moralself-concept. In addition, the commission ofimmoral behavior makes it more costly to actmorally in the future, increasing the likelihoodof further evil. Many theories that stress thecontribution of situational factors to evil behav-ior emphasize this step-by-step escalation intoatrocity. Individuals begin with small, unethicalacts, but their behavior can quickly increase toalarmingly immoral actions.

For instance, Staub (1989) explained thateven if all of the conditions are present forgenocide or mass killings, these events do notoccur instantly but are eased into through a“continuum of destruction” (p. 17). When smallacts of harm are committed, perpetrators dehu-manize their victims to rationalize their harmfulbehavior. This dehumanization further obscures

44 TSANG

the relevance of morality, paving the way formore intense harm and, finally, genocide. Smallunethical acts might also feel rewarding; forexample, stealing without getting caught bringspsychological reinforcement of euphoria and asense of power, along with material reinforce-ment of acquiring stolen property. The individ-ual feels positively reinforced for immoralityand begins to seek further rewards in moreextreme behavior.

Blumenthal (1999) described an incrementalpraxis of evil in which situational factors suchas deindividuation and psychological factorssuch as rationalization allow individuals to en-gage in small unethical actions. When perpetra-tors realize they can engage in unethical behav-ior without punishment, they incrementally es-calate their immoral behavior. Blumenthal usedthe Holocaust to illustrate the idea of incremen-tal praxis, noting that the Nazis began withdiscrimination against the Jews and moved topersecution, imprisonment, and, finally, system-atic extermination. With each incremental in-crease, people were desensitized to the levelof immoral action, until the idea of concentra-tion camps seemed reasonable and morallytolerable.

Milgram’s (1974) studies on obedience pro-vide good illustrations of escalation. Probably,few of Milgram’s participants would have con-sented at the onset of the study to shockingsomeone at 450 volts. Instead, the experimenterinstructed them to begin at the lowest shocklevel, 15 volts, and move up at 15-volt incre-ments for each incorrect answer the victimgave. Milgram termed this the “sequential na-ture of the action.” He believed that the sequen-tial set-up of the shocks facilitated increases inevil behavior by making it difficult for the par-ticipant to disobey the experimenter. This oc-curred because, if participants chose to stop theshocks, they had to admit that many of theshocks that led up to their disobedience werealso wrong and that they had therefore actedimmorally. As a result of this, Milgram believedthat many participants might have continuedwith the shocks as a way to rationalize to them-selves that what they had done in the past wasright because they continued to do it. He statedsuccinctly: “Earlier actions give rise to discom-forts, which are neutralized by later ones. Andthe subject is implicated into the destructivebehavior in a piecemeal fashion” (Milgram,

1974, p. 149). Milgram also mentioned thatonce the sequence of obedient events has beeninitiated, the individual can passively continueto obey immoral orders, whereas disobedienceentails a much more difficult active break withauthority.

Cognitive dissonance can help explain theeffects of escalation on immoral behavior. Oncean individual has engaged in a particular behav-ior, he or she is committed to it: Admitting thatthe behavior is wrong and needs to be changedis dissonant with the fact that one has engagedin the behavior in the past and views oneself asa moral person. To resolve this dissonance, anindividual may continue to engage in evil be-havior while rationalizing that the behavior ismoral.

The Nazis used escalation during World WarII to gain national acceptance of their plans forthe annihilation of the Jews. Eichmann de-scribed how a proposed plan to ship Jewishpeople to Madagascar was not serious butserved to familiarize

all concerned with the preliminary notion that nothingless than complete evacuation from Europe would do. . . when, a year later, the Madagascar project wasdeclared to have become “obsolete,” everybody waspsychologically, or rather, logically, prepared for thenext step: since there existed no territory to which onecould “evacuate,” the only “solution” was extermina-tion. (Arendt, 1963, p. 77)

The proposal to deport Jewish people to Mada-gascar thus functioned to desensitize people toviolations in moral standards, preparing themfor the next step in the escalation of evil.

Darley (1992, 1996) stressed the step-by-stepnature of evil behavior in organizations. Hedescribed the process of escalation:

People move toward these disasters in small steps, andthe later steps seem no different from the earlier ones.When one faces the realization that something ethicallyuntoward is going on, one seems somehow committedto allowing and assisting in its continuance, because ofcommitments generated by previous actions. (Darley,1996, p. 33)

Darley believed not only that rationalization ledto escalation of immoral actions but that the actof rationalizing evil behavior transformed theperson into an evil individual who would bemore likely to commit evil actions in the future.

Thus, escalation can lead to increased im-moral behavior through the various processes ofoperant conditioning, desensitization, and cog-


nitive dissonance. Although these factors canlead to escalation of immoral behavior, moralbehavior might also be facilitated through esca-lation. Milgram (1974) described the steps thatmany participants in his studies took towarddisobeying the experimenter. He noted that thevoicing of dissent often served as a steppingstone toward other behaviors that gradually ledto a break with authority. As he stated, “Theinitial expression of disagreement, however ten-tatively phrased, provides a higher plateau fromwhich to launch the next point of disagreement”(Milgram, 1974, p. 163). Not only did dissentfacilitate disobedience within the participantsthemselves, but dissent by peers greatly assisteddisobedience among participants. Thus, moralas well as immoral behavior is subject toescalation.

Empirical Evidence forMoral Rationalization

Research relevant to moral rationalizationcan be categorized into three groups: researchon situational factors in harmful or immoralbehavior, research looking specifically at differ-ent methods of moral rationalization, and re-search on the prevention of moral rationaliza-tion and evil. Taken together, this research dem-onstrates the importance of both situational andeveryday psychological factors in the commis-sion and persistence of evil behavior.

Situational Factors

Much research has been done looking at howsituational factors affect immoral behavior. Al-though it is beyond the breadth of this article tolist all relevant research, I provide a few exam-ples of studies that illustrate how situationalfactors may be involved in setting the stage formoral rationalization. For instance, Milgram’s(1974) study of obedience and the Stanfordprison experiment (Haney et al., 1973) demon-strated how a particular situation can elicit im-moral behavior from participants. The situa-tional factors in these studies not only preventedmoral standards from becoming salient to par-ticipants but provided moral rationalizations forindividuals to use once they realized they wereviolating their moral principles: “ I was onlyfollowing orders” and “ It was my job as a prisonguard to keep the prisoners in line.”

Other studies have used experimental manip-ulations of a situation to alter people’s moralbehavior. For example, Kilham and Mann(1974) varied Milgram’s (1974) procedure toinclude a transmitter role in addition to theteacher (executant) role. As in Milgram’s orig-inal experiment, the executant shocked thelearner. However, it was the transmitter whomonitored the answers of the learner and gavethe executant the order to shock. The transmitterwas therefore not directly responsible for harm-ing the learner but played an essential role in theprocess. Kilham and Mann found that partici-pants were significantly more obedient in thetransmitter role than in the executant role,showing the importance of diffusion of respon-sibility in obedience. Individuals in the trans-mitter role could claim that they were not vio-lating their moral principles because they werenot directly harming the learner.

Bandura, Underwood, and Fromson (1975)found that when situational factors that affecteddiffusion of responsibility and dehumanizationwere in place, participants behaved more puni-tively toward a victim than when responsibilitywas personalized and the victim was human-ized. In addition, participants faced with a de-humanized victim were more likely to exoner-ate themselves from their aggressive behavior, arough indication of moral rationalization.

Moral Rationalization

Whereas the studies just mentioned did notdirectly posit mechanisms of moral rationaliza-tion in the relationship between a given situa-tion and immoral behavior, other studies haveexamined the role of specific rationalizations inimmoral actions. Similar to Bandura et al.(1975), Diener, Dineen, Endresen, Beaman, andFraser (1975) found that altered responsibilityincreased aggressive behavior in the laboratory,as did witnessing an aggressive model. Relevantto moral rationalization, Diener et al. also foundthat a manipulation of cognitive set increasedaggression: Those participants who were toldthat the activity they were about to engage inwas “ like a game” acted more aggressively thanparticipants who were told that the activity wasabout aggression. Possibly, the participants whowere told that the activity was a game used thiseuphemism to mask the aggressive nature oftheir behavior to themselves.

46 TSANG

Bandura et al. (1996) measured the effect ofdifferent methods of moral disengagement ondelinquent behavior. They found that frequencyof endorsement of moral disengagement influ-enced delinquent behavior through mediatorssuch as lowered prosocialness, lessened an-ticipatory guilt, and higher proneness toaggression.

These studies demonstrate how situationalfactors, along with individual differences inmoral rationalization, can affect moral behav-ior. Not only can Nazis commit evil, but normalpeople in everyday situations have the potentialto violate their moral standards.

Prevention of Moral Rationalization

The study of moral rationalization and evilcan seem pessimistic and bleak, but there existsa positive counterpoint: In understanding thevariables that affect rationalization, researchersare also uncovering factors that can encouragemoral behavior. Although the research just de-scribed demonstrates how certain aspects of theindividual and the situation can make moralbehavior less likely, it also suggests that othersituational and psychological factors mightmake moral rationalization more difficult andmoral behavior more prevalent.

Some studies relevant to the prevention ofmoral rationalization include situational manip-ulations that make it more difficult to use certainforms of moral rationalizations. Although moralrationalization was not measured in these stud-ies, it is possible that the changes in moralactions found were related to lessened rational-ization. Milgram (1974) conducted a number ofvariations of his obedience study and foundseveral factors that lessened obedience in hisparticipants. For example, obedience signifi-cantly decreased as a function of the feedbackreceived from the learner. Physical absence ofthe experimenter also affected obedience, withfewer participants obeying the experimenterwhen he gave his commands over the telephonerather than in person. In addition, participantswere less likely to follow commands given byanother participant rather than the experimenter,especially if the experimenter was the one beingshocked. Moreover, they were less likely toobey orders when confederate peers also defiedthe experimenter’s orders.

These variables of increased feedback fromthe victim, characteristics of the authority, andthe presence of defiant others were thus impor-tant factors that could prevent immoral behaviorthat originated from obedience. These situa-tional factors may have affected the availabilityof different moral rationalizations such as denialof consequences, displacement of responsibil-ity, and the actions of others. With increasingsalience of the victim, it would have been moredifficult for participants to deny the harmfulconsequences of the shocks they were adminis-tering. With the experimenter physically absentor replaced by another participant, it becomesharder to shift responsibility to the absent or lesscredible authority. And when peers defy theexperimenter, it again becomes harder to dis-place responsibility to others or point to others’harmful actions as a justification for what one isdoing. In these ways, the different situationalfactors that were varied in Milgram’s studiesmay have changed the effectiveness of differentmoral rationalizations, making it harder to harmthe victim without perceived violations of moralstandards and more likely that participantswould defy authority.

Tilker (1970) used a variant of Milgram’sprocedure, with the participant acting as an ob-server of the teacher and learner interactions.He found that when participants were made tofeel more responsible for the welfare of thelearner and when they had maximum feedbackregarding the harm that was being done, theywere more likely to stop the experiment and tostop it earlier. This study illustrates that certainaspects of the situation that make it more diffi-cult to displace responsibility or deny conse-quences will encourage moral rather than im-moral behavior.

Staub (e.g., 1998, 2000) stressed the theoret-ical importance of bystander action in the halt-ing of genocidal violence. In terms of moralrationalization, bystanders can work to makemoral principles more salient or to refute per-petrators’ rationalizations. Staub also suggestedaddressing the difficult life conditions that in-stigated genocide, which would alleviate theneed to act immorally before rationalizationprocesses began.

Additional research has attempted to measuremore directly the mechanisms of moral ratio-nalization and their effects. McAlister, Ama,Barroso, Peters, and Kelder (2000) examined an


intervention program that promoted racial tol-erance and reduction of moral rationalization.They used a method called “behavioral journal-ism,” which presented high school students witharticles written by their peers about positiveexperiences with interracial friendships and dat-ing and refuted ideas about violence that couldbe used as moral justification for intergroupconflict. After the behavioral journalism inter-vention, students had more positive attitudestoward interracial friendships and dating, higherperceived value similarity, and less perceivedracial superiority than they had before the in-tervention and also in comparison with studentsfrom a nearby school who did not experience anintervention. In addition, these students pro-vided less endorsement for violent moral justi-fications. Students also reported fewer experi-ences of verbal racial aggression after the inter-vention. This study showed that it is possible todesign a successful intervention that inhibits atleast some forms of moral rationalization forintergroup conflict.

Bersoff (1999) studied the occurrence of un-ethical behavior in the laboratory, examininghow situational changes affected the ease withwhich individuals could rationalize their im-moral behavior. His dependent variable waswhether participants spontaneously mentionedthat they were overpaid for an experiment. Inthe first condition, the person reimbursingthe participant directly asked the participantwhether the payment was correct, making itharder for participants to use rationalizations ofdenial of responsibility when taking an over-payment. In another condition, the experimentwas headed by a graduate student rather than afaceless organization. This condition varied therationalizations of dehumanization as well asthe distortion of consequences (“ It wouldn’ thurt a big corporation if they paid me a fewextra dollars” ). In further conditions, Bersofftested a method called “ inoculation” by havingparticipants criticize a subject in a scenario forcommitting a similar unethical action. Theseconditions made overall moral rationalizationmore difficult for participants.

The results of this experiment showed thateach of these “anti-neutralization” (or anti-ra-tionalization) techniques increased the amountof ethical behavior exhibited by participants. Inaddition, Bersoff found that multiple anti-ratio-nalizations had additive effects. This study

demonstrated not only that individuals use dif-ferent methods of moral rationalization but thataspects of the situation can be changed to makerationalization more difficult, stymieing uneth-ical behavior.

The aforementioned study conducted by Ban-dura et al. (1975) can also be cited as an exam-ple of situational factors working to inhibitmoral rationalization. When individuals findthemselves in a situation in which they areclearly responsible for their actions, and victimsare humanized, they will be less likely to en-gage in aggressive behavior that violates moralstandards.

Although situational factors might play animportant part in preventing moral rationaliza-tion, it is also fruitful to examine the existenceof individual differences in the propensity formoral rationalization. Bandura, Caprara, Bar-baranelli, Pastorelli, and Regalia (2001) foundthat perceived social and regulatory self-effi-cacy in adolescents was related to a decrease intransgressive behavior through lowering moralrationalization and increasing prosocialness.Children who felt positively about their abilityto maintain interpersonal relationships and re-sist peer pressure were less likely to endorsedifferent methods of moral rationalization andmore likely to value helping others; also, theyreported less antisocial activity both initiallyand 2 years later. These results raise the possi-bility that interventions that increase children’sself-efficacy might help curb mechanisms ofmoral rationalization that contribute to delin-quent behavior.

Conclusion

Hence, the explanation of evil resides notsimply in evil situations or evil people but in acomplex interplay between situational factorsand normal psychological processes. Moral ra-tionalization plays an important role in allowingan individual to autonomously engage in im-moral behavior while still seeing the self asmoral. The individual then engages in more andmore extreme behavior, until small unethicalacts escalate into large atrocities. Because ofmoral rationalization, large-scale evil and smallbreaches of morality are, on a certain level,qualitatively similar.

I have endeavored in this article to present amodel of evil behavior that integrates different

48 TSANG

situational and intrapsychic factors, placing aspecial emphasis on the role of moral rational-ization in evil. Knowledge of the processes be-hind moral rationalization and evil behaviorwill not only help researchers explain the oc-currence of such events as the Holocaust butalso will allow us to find variables that canattenuate rationalization and halt evil behavior.An understanding of moral rationalizationmight help us know not only when and whypeople act immorally but also when they willuphold moral principles.

References

Arendt, H. (1963). Eichmann in Jerusalem: A reporton the banality of evil. New York: Penguin Books.

Aronson, E. (1969). The theory of cognitive disso-nance: A current perspective. In L. Berkowitz(Ed.), Advances in experimental social psychology(Vol. 4, pp. 1–34). New York: Academic Press.

Aronson, J., Blanton, H., & Cooper, J. (1995). Fromdissonance to disidentification: Selectivity in theself-affirmation process. Journal of Personalityand Social Psychology, 68, 986–996.

Bandura, A. (1986). Social foundations of thoughtand action: A social cognitive theory. EnglewoodCliffs, NJ: Prentice Hall.

Bandura, A. (1990). Selective activation and disen-gagement of moral control. Journal of Social Is-sues, 46, 27–46.

Bandura, A. (1991). Social cognitive theory of moralthought and action. In W. M. Kurtines & J. L.Gewirtz (Eds.), Handbook of moral behavior anddevelopment: Vol. 1. Theory (pp. 45–104). Hills-dale, NJ: Erlbaum.

Bandura, A. (1999). Moral disengagement in theperpetration of inhumanities. Personality and So-cial Psychology Review, 3, 193–209.

Bandura, A., Barbaranelli, C., Caprara, G. V., &Pastorelli, C. (1996). Mechanisms of moral disen-gagement in the exercise of moral agency. Journalof Personality and Social Psychology, 71, 364–374.

Bandura, A., Caprara, G. V., Barbaranelli, C., Pas-torelli, C., & Regalia, C. (2001). Sociocognitiveself-regulatory mechanisms governing transgres-sive behavior. Journal of Personality and SocialPsychology, 80, 125–135.

Bandura, A., Underwood, B., & Fromson, M. E.(1975). Disinhibition of aggression through diffu-sion of responsibility and dehumanization of vic-tims. Journal of Research in Personality, 9, 253–269.

Baron, J. (1995). Blind justice: Fairness to groupsand the do-no-harm principle. Journal of Behav-ioral Decision Making, 8, 71–83.

Batson, C. D., Klein, T. R., Highberger, L., & Shaw,L. (1995). Immorality from empathy-induced al-truism: When compassion and justice conflict.Journal of Personality and Social Psychology, 68,1042–1054.

Batson, C. D., Kobrynowicz, D., Dinnerstein, J. L.,Kampf, H. C., & Wilson, A. D. (1997). In a verydifferent voice: Unmasking moral hypocrisy. Jour-nal of Personality and Social Psychology, 72,1335–1348.

Baumeister, R. F., & Newman, L. S. (1994). Self-regulation of cognitive inference and decision pro-cesses. Personality and Social Psychology Bulle-tin, 20, 3–19.

Bentham, J. (1983). Deontology; together with atable of the springs of action; and the article onutilitarianism. New York: Oxford UniversityPress. (Original work published 1829)

Bersoff, D. M. (1999). Why good people sometimesdo bad things: Motivated reasoning and unethicalbehavior. Personality and Social Psychology Bul-letin, 25, 28–39.

Blumenthal, D. R. (1999). The banality of good andevil: Moral lessons from the Shoah and Jewishtradition. Washington, DC: Georgetown Univer-sity Press.

Braden, M., & Walster, E. (1964). The effect ofanticipated dissonance on pre-decision behavior.In L. Festinger (Ed.), Conflict, decision, and dis-sonance (pp. 145–151). Stanford, CA: StanfordUniversity Press.

Darley, J. M. (1992). Social organization for theproduction of evil. Psychological Inquiry, 3, 199–218.

Darley, J. M. (1996). How organizations socializeindividuals into evildoing. In D. M. Messick &A. E. Tenbrunsel (Eds.), Codes of conduct: Behav-ioral research into business ethics (pp. 13–43).New York: Russell Sage Foundation.

Deutsch, M. (1990). Psychological roots of moralexclusion. Journal of Social Issues, 46, 21–25.

Diener, E. (1980). Deindividuation: The absence ofself-awareness and self-regulation in group mem-bers. In P. B. Paulus (Ed.), Psychology of groupinfluence (pp. 209–242). Hillsdale, NJ: Erlbaum.

Diener, E., Dineen, J., Endresen, K., Beaman, A. L.,& Fraser, S. C. (1975). Effects of altered respon-sibility, cognitive set, and modeling on physicalaggression and deindividuation. Journal of Per-sonality and Social Psychology, 31, 328–337.

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

Festinger, L. (1964). Conflict, decision, and disso-nance. Stanford, CA: Stanford University Press.


Festinger, L., & Carlsmith, J. M. (1959). Cognitiveconsequences of forced compliance. Journal ofAbnormal and Social Psychology, 58, 203–210.

Festinger, L., Riecken, H. W., & Schachter, S.(1956). When prophecy fails. Minneapolis: Uni-versity of Minnesota Press.

Freud, S. (1930). Civilization and its discontents.New York: Anglobooks.

Haney, C., Banks, C., & Zimbardo, P. (1973). Inter-personal dynamics in a simulated prison. Interna-tional Journal of Criminology and Penology, 1,69–97.

Heider, F. (1958). The psychology of interpersonalrelations. Hillsdale, NJ: Erlbaum.

Higgins, E. T. (1987). Self-discrepancy: A theoryrelating self and affect. Psychological Review, 94,319–340.

Higgins, E. T. (1997). Beyond pleasure and pain.American Psychologist, 52, 1280–1300.

Kant, I. (1898). Kant’s critique of practical reasonand other works on the theory of ethics (T. K.Abbott, Trans.). New York: Longman, Green.(Original work published 1788)

Kelman, H. C., & Hamilton, V. L. (1989). Crimes ofobedience: Toward a social psychology of author-ity and responsibility. New Haven, CT: Yale Uni-versity Press.

Kilham, W., & Mann, L. (1974). Level of destructiveobedience as a function of transmitter and execu-tant roles in the Milgram obedience paradigm.Journal of Personality and Social Psychology, 29,696–702.

Kunda, Z. (1990). The case for motivated reasoning.Psychological Bulletin, 108, 480–498.

Latane, B., & Darley, J. M. (1970). The unresponsivebystander: Why doesn’t he help? New York:Appleton-Century-Crofts.

Lerner, M. J. (1980). The belief in a just world: Afundamental delusion. New York: Plenum.

Lifton, R. J. (1986). The Nazi doctors: Medical kill-ing and the psychology of genocide. New York:Basic Books.

Mann, L. (1981). The baiting crowd in episodes ofthreatened suicide. Journal of Personality and So-cial Psychology, 41, 703–709.

McAlister, A. L., Ama, E., Barroso, C., Peters, R. J.,& Kelder, S. (2000). Promoting tolerance andmoral engagement through peer modeling. Cul-tural Diversity and Ethnic Minority Psychology, 6,363–373.

Milgram, S. (1974). Obedience to authority. NewYork: Harper & Row.

Mill, J. S. (1957). Utilitarianism. Indianapolis, IN:Bobbs-Merrill. (Original work published 1861)

Nowell-Smith, P. H. (1967). Religion and morality.In P. Edwards (Ed.), The encyclopedia of philoso-phy (Vol. 7, pp. 150–158). New York: Macmillan.

Olson, R. G. (1972). Deontological ethics. In P. Ed-wards (Ed.), The encyclopedia of philosophy(Vol. 2, p. 343). New York: Macmillan.

Opotow, S. (1990). Moral exclusion and injustice:An introduction. Journal of Social Issues, 46,1–20.

Post, J. M. (1990). Terrorist psycho-logic: Terroristbehavior as a product of psychological forces. InW. Reich (Ed.), Origins of terrorism: Psycholo-gies, ideologies, theologies, states of mind (pp.25–40). Cambridge, England: Cambridge Univer-sity Press.

Roney, C. J. R., Higgins, E. T., & Shah, J. (1995).Goals and framing: How outcome focus influencesmotivation and emotion. Personality and SocialPsychology Bulletin, 21, 1151–1160.

Simon, L., Greenberg, J., & Brehm, J. (1995). Trivi-alization: The forgotten mode of dissonance reduc-tion. Journal of Personality and Social Psychol-ogy, 68, 247–260.

Smart, J. J. C. (1967). Utilitarianism. In P. Edwards(Ed.), The encyclopedia of philosophy (Vol. 8, pp.206–212). New York: Collier Macmillan.

Staub, E. (1989). The roots of evil: The origins ofgenocide and other group violence. New York:Cambridge University Press.

Staub, E. (1993). The psychology of bystanders, per-petrators, and heroic helpers. International Journalof Intercultural Relations, 17, 315–341.

Staub, E. (1998). Early intervention: Prediction andaction. In H. J. Langholtz (Ed.), The psychology ofpeacekeeping (pp. 31–39). Westport, CT: Praeger.

Staub, E. (2000). Genocide and mass killing: Origins,prevention, healing and reconciliation. PoliticalPsychology, 21, 367–382.

Steele, C. M. (1988). The psychology of self-affir-mation: Sustaining the integrity of the self. In L.Berkowitz (Ed.), Advances in experimental socialpsychology (Vol. 21, pp. 261–302). New York:Academic Press.

Steele, C. M., & Liu, T. J. (1983). Dissonance pro-cesses as self-affirmation. Journal of Personalityand Social Psychology, 45, 5–19.

Tilker, H. A. (1970). Socially responsible behavior asa function of observer responsibility and victimfeedback. Journal of Personality and Social Psy-chology, 14, 95–100.

Todorov, T. (1996). Facing the extreme: Moral life inthe concentration camps. New York: Henry Holt.

Zimbardo, P. G. (1995). The psychology of evil: Asituationist perspective on recruiting good peopleto engage in anti-social acts. Japanese Journal ofSocial Psychology, 11, 125–133.

Received November 30, 2000Revision received May 28, 2001

Accepted June 4, 2001 �

50 TSANG

Moral Rationalization and the Integration of Situational ... · PDF fileMoral Rationalization...

Documents

Transcript of Moral Rationalization and the Integration of Situational ... · PDF fileMoral Rationalization...