7400655 - Psychological Inquiry 1992, Vol. 3, No. 2,...

Info iconThis preview shows pages 1–21. Sign up to view the full content.

View Full Document Right Arrow Icon
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 2
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 4
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 6
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 8
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 10
Background image of page 11

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 12
Background image of page 13

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 14
Background image of page 15

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 16
Background image of page 17

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 18
Background image of page 19

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 20
Background image of page 21
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Psychological Inquiry 1992, Vol. 3, No. 2, 199—218 Copyright 1992 by Lawrence Erlbaum Associates, Inc. BOOK REVIEW ESSAY Social Organization for the Production of Evil John M. Darley Princeton University There doesn’t seem to be a sustained tradition within psy- chology of confronting questions about evil, its origins, and its effects. Or perhaps this comment should be given a nar- rower scope: There does not seem to be that confrontation going on in the academic traditions of psychology within which most of us work. In this essay, I attempt to present a social psychological perspective on evil, and contrast that perspective with two others: (a) the views of evil that we all hold at an unexamined level of everyday thought, and (b) the View of evil that I think might be drawn from the clinical perspective. Finally, I suggest some conclusions we might draw about evil from the clash of these different per— spectives. I was born in 1938, into a cohort that was ordained to think about evil—indeed to be haunted by it, given the events of the Nazi era. But that social psychology could come to grips with what this evil might mean I did not have the Vision, or perhaps the courage, to see. Even now, having thought about the topic under the stimulus of these books, I would not want to say that my thinking has gone very far. This essay is a reconnaissance, not anything I would regard as a finished work. Ifl can invite others to the debate, then perhaps our conceptualizations of evil will be advanced. The authors of the books considered here have had the Vision and courage to analyze evil. In the past few years, we have seen the publication of Lifton’s report of his interviews with physicians who participated in the Nazi death camps, the long-awaited book by Herbert Kelman and Lee Hamilton containing their observations on the interpretations of their survey data on My Lai and the related trial of Lt. Calley, and Ervin Staub’s comparative study of several genocidal events. Reflecting on these books, I have been led back to Stanley Milgram’s (1974) book detailing his obedience studies and his final interpretations of them, and to Hannah Arendt’s (1963) seminal and controversial book Eichmann in Jerusa- lem: Report on the Banality of Evil. Each author has been willing to consider the events or people we label as evil. (Milgram does so somewhat implicitly, the others quite ex— plicitly.) Each author deserves considerable credit for this; evil is not a topic easily accessible within the confines of Crimes of Obedience: Towards a Social Psychology of Authority and Responsibility, by Herbert C. Kelman and V. Lee Hamilton, 1989, New Haven, CT: Yale University Press. The Nazi Doctors: Medical Killing and the Psychology of Genocide, by Robert Jay Lifton, 1986, New York: Basic. The Roots of Evil: The Origins of Genocide and Other Group Vio- lence, by Ervin Staub, 1989, New York: Cambridge University Press. those psychological movements that are the inheritors of our earlier operationalist, positivist traditions. These people are working outside the mainstream of at least the sorts of psy- chological content customary within our academic enclaves. They are not, however, always working outside the meth- odological traditions of modern social science. Milgram, famously and controversially, carried out social psychologi— cal experiments on harmdoing. Kelman and Hamilton (al- though their book is by no means confined to it) report survey results about the “Lt. Calley trial,” the military court marshal of a platoon commander who presided over the My Lai massacre. Lifton conducted interviews with doctors who had served in Nazi concentration camps. Only Staub does not present some form of new data, choosing instead to construct an account ranging across different fields of social science and different sources of social—science thinking. Each book contributes to what becomes, at a general level, a consistent picture of the origins of many of the evil acts in the world. Reading and rereading them, I have come to see that they sharply contradict our ordinary ways of thinking about the origins of evil, and that this is one of the major messages to be extracted from them. The message they jointly present is a disconcerting one: Our everyday under- standings of evil (to be sketched in my next section) are frequently incorrect or perhaps, more accurately, irrelevant to most acts of evil and therefore deeply misleading. Instead, the authors provide the material on which to form an alternate View, one that I have called the “social psychological per— spective.” Because that View clashes with views all of us hold, their message is an uncomfortable one, whose implica- tions are not easily grasped. Like most ordinary people, we psychologists are in the grip of another View. That other view exists at the level of our day-to—day, understood rather than examined, beliefs—the beliefs we hold as persons rather than as psychologists. To understand the social psychological view, it is useful to ex— amine common—sensical views first, the ways that our culture currently thinks about the related notions ‘of “moral wrong- doing” and “evil.” Everyday Thinking About “Wrongdoing” and “Evil” Moral wrongdoing has proved to be a troubling concept to define. Uneasily recalling our exposure to philosophy courses, in which we discovered the difficulties in defining and defending our everyday concepts, we “ordinary per— sons” recognize at some level that we may not be able to 200 DARLEY clearly define and defend our particular concepts of “good” or “wrongdoing.”l Unlike some other cultures, or our own culture in the past, we do not have an agreed-on, author— itative list of actions that constitute wrongdoing. We used to have lists of sins, but because we no longer put much stock, as a culture, in the religious principles that generated these lists, we cannot much rely on them as defining the set of moral wrongs.2 Thus, we have to generate such a list from some underlying philosophical principle. Principle-Based Definitions of Wrongdoing When we cast around for that agreed-on principle that might give us a tenable conceptualization of wrongdoing, we are not greatly helped. One prevailing theory of human moti- vation, the theory of rational choice, does not easily justify the moral preferencing of certain desires over others, and thus does not give us any easy mechanism for designating certain actions as morally wrong. Contributing to the diffi- culties with the definition of “wrong,” is our learned recog- nition, within our present society, that it is difficult to state our grounds for disagreeing with another individual ’s person— al preference structure. One of us likes the paintings of Tit— ‘ ian, another paintings of Elvis on black velvet; to each his own.3 As this hints, all these notions lead us to retreat to the culturally familiar stance of a last—resort utilitarianism. With— in that system, we can all agree that causing harm or pain to others is the essence of wrongdoing. The definition of im- mortality that arises from utilitarian considerations centers around a notion of actions that inflict pain on others. As the old characterization went, “my freedom of action ends just a centimeter away from hitting your nose.” We might say that, within a utilitarian perspective, my possibility of doing wrong begins about at your nose. Wrong actions, more for— mally, are those that impinge on the other, and cause that other pain or harm. This definition needs qualification, but the moves to qualify it are well understood. Briefly, we need to rule out certain harmful actions, such as unforeseeable accidents, as necessarily morally wrong. So we add to the definition a 11f this is taken to describe the present state of philosophy, then it is a rather poor description. Within philosophy, there are several clear-headed and cogent analyses of evil that are internally coherent and well reasoned. But it is not the current state of philosophy that I am addressing. Instead, I am—I think accurately—trying to sketch the ways in which most of us ordinary people, perhaps dimly remembering a college philosophy course or two, would enter into discourse about evil. 2These remarks raise rather acutely the question of who the “ordinary person” I am talking about is taken to be, and just how many courses in moral philosophy I am assuming he or she has taken. One group that initially appears to be an exception includes those who hold fundamentalist Christian beliefs, and who do believe in sin. But I suggest that they are also prone to the utilitarian groundings of morality that seems to be the bedrock for the rest of us. As a test, ask a fundamentalist why some particular action is wrong. The initial answer may be some variant of “Because the Bible pronounced it a sin.” However, if one continues to ask why God, Christ, or the Bible pronounced it a sin, answers will be forthcoming, and in my experience are based on causing harm to others or to self. 3Not irrelevant here is the recent “deconstructionist” movement in liter— ary analysis, which argues that those works of art we have characteristically regarded as masterpieces, and thus exemplars of the good, are in fact catego- rized as such because certain privileged elites have foisted this perspective on the rest of us. Thus, they too have been read as arguing for a relativistic world. qualification: a notion of a wrongdoing actor as somehow intentionally, knowingly, or recklessly harming or causing pain to another individual. A few more qualifications bring us to a preliminary defini— tion. We recognize that in certain times (such as wartime) or in certain places (such as prison death rows) actions that bring harm or even death to other individuals, are regarded by many not only as not evil, but as morally required actions, so we add a reference to “unjustified harmdoing” as the morally wrong sort of harrndoing. We also add qualifications pertaining to excused harrndoing incidents, or mitigated ones (Austin, 1956). What should count as justifications, mitiga- tions, and excuses requires some elucidation, but we have generally agreed on examples to guide our judgments (Darley & Shultz, 1990; Darley & Zanna, 1982). Naturally, some tricky questions remain; we can be made uneasy by certain borderline cases.4 Nonetheless, I want to suggest that this is how ordinary Americans resolve questions when they are pressed concerning their definitions of “morally wrong actions.” Staying at a self-conscious level of discourse, having struggled to arrive at this definition of “wrong,” we seem to have very little left to say about evil. Evil is an even more difficult concept to define than “wrong.” Uncharac- teristically, Webster’s flounders: “not good morally; causing or tending to cause harm; the antithesis of good; something that is injurious to moral or physical happiness or welfare” (Webster’s New Third International Dictionary, 1963). Here evil is simply equated with moral wrongness. One gets the feeling that evil is a word falling out of use; it seems redun- dant with the notion of moral wrongness, and bringing archa— ic baggage such as the notion of sin along with it. Actually, I think we have difficulties with the notion of evil only when we take a certain analytic perspective. The per— spective, or level of thought at which we have been analyzing wrongdoing thus far, is one that we might call considered thought that one is prepared to defend against challenge. The challenge might arise from an acquaintance with differ— ent values, a public opinion pollster, or a philosophy pro— fessor. We are all capable of functioning on that considered level of thought, and do so when we are facing definitional questions. At this level of thought, the notion of evil becomes difficult of definition or redundant. However, it is not always at this level of thoughttthat ordinary people think about evil; in everyday thought, including that of psychologists, the notion of evil is alive and well. It may not be easy to define, but none of us have much trouble recognizing it. Everyday Definitions of Evil At the level of our everyday functioning, we operate on the basis of “naive psychology,” which comprises the under- standings we all carry around with us to analyze the events and actions of our everyday life. These are the sorts of under- standings Heider (1958) most successfully called to our at— tention: those often unexamined understandings, frequently built into everyday vocabularies, that categorize our every- day world and enable us to function in it. At this level, we have both a conceptualization of evil acts that is generally 4For instance, even if somebody consented to being tortured, would not it still be evil to torture him? BOOK REVIEW ESSAY 201 shared and coherent and a theory of what causes these evil actions. First, how do we conceive of evil acts? As follows: Evil actions are a subset of bad actions; all evil actions are bad, but not all bad actions are evil. A sports fan, provoked in a barroom by another who vilifies the losing performance of the fan’s much-loved sports team, punches his tormenter. A morally bad action, but not one we would be prone to call evil. This casting about does not yet give us a definition of evil actions, however. To be labeled as evil, the wrongdoing act often has to have a quality of egregious excess, such as a murder gratuitously committed in the course of a crime. A bank robber gratuitously shoots some elderly and disarmed bank guard as he exits the bank; the evil lies in the sense— lessness of the act. However, even if the elderly and disarmed guard was killed to demonstrate to the other victims the seriousness of the robber’s threats to kill those who might pursue, we would still consider that act evil. The actor is seen to put such a low value on human life as to provide the moral outrage that triggers the label of evil. At other times, the evil actor shows an equal disregard for humanity, but the evil act is not so much described as egregious excess as depraved excess. Those individuals who derive pleasure from the torture of children display this as- pect of evil. One is struck by two things in this case; (a) the deviant and inhuman nature of the perverse impulse, and (b) the disproportion of the act in terms of the pleasure gained versus the pain inflicted. To inflict vast pain on the innocent to derive a fleeting and perverse pleasure is breathtakingly horrible. The evil we have been describing contains elements of intention. The criminal intended to kill his hostage; the tor- turer intentionally chooses his victim. However, there is felt to be something bizarre about the intention. Sometimes the bizarre quality resides in the disparity between the intention and the grounds giving rise to that intention. A parent repeat— edly batters an infant child because the child cries. We see from where the impetus to batter arises, but we simply cannot grant any sympathetic validity to the act arising out of the impetus. Again, a wildly disproportionate disregard for the humanity of the harmed individual seems indicated.5 At other times , we simply cannot fathom the sort of person who would intend to commit such terrible actions. We posit evildoing when we see that an individual is following our general societal template for the commission of an inten- tional action, and that the action is not only wrong but horri- bly wrong. The contrast between the apparent rationality of the sequence of behaviors that lead up to the action and the irrational character of the act is one cue we use to assign evil. To buy the instruments that will be used to torture another, or to dig the grave in which the kidnap victim will be buried alive, are evil acts. To plan to kidnap an individual for the purposes of deriving pleasure from torturing that individual 5An individual who tortures animals is generally seen as evil. Thus, the inference of evil arises from crimes other than those directed against people. In fact, many feel mistreatment of animals is somehow particularly evil. I suspect that this is exactly because to hurt a creature who cannot have any reasoned role in the causal circumstances surrounding the incident seems particularly egregious. Still, this means that it is not the infliction of suffer- ing on a human being that is necessary for evil, but the infliction of suffering on an organism capable of sufi‘ering—experiencing pain and perhaps antic- ipatory dread about that pain. is an act from which, among other things, an inference of evil is likely to arise.6 The buildup of the impulse in a serial killer to kill again, which I imagine to be experienced by the killer as an outside force, does not fit our usual definition of inten— tional action. Perhaps the imputation of evil arises because the killer deliberately organizes the acts to fulfill his “irre— sistible impulse.” In essence, then, an evil action occurs when an individual inflicts a highly negative state on another, without this nega- tive state being balanced in any way in the perpetrator. I kill you to avoid being tortured. Terrible, but not evil. I kill you or torture you because it gives me some brief pleasure or avoids a slight annoyance to me. This is evil. Further, the evildoer knowingly violates society’s norms. Thus, first, the actor puts his or her needs above the needs of those victimized, and, second, the actor puts his or her own judgments above others’. The imbalance here shows a chilling disregard for the humanity of the victim or the community; the other is given so little standing, as compared to the actor’s pleasures and needs, as to be denied human existence. An approach from another direction is possible. So far I have attempted to define evil as we use the concept in our everyday life. N 0w let me suggest a marker reaction that tells us when evil is present. Moral and jurisprudential philosophy typically identifies five reasons for incarcerating or other— wise punishing an individual who has committed wrongdo- ing, including (a) providing an opportunity for the correction or rehabilitation of the individual, (b) deterring that indi— vidual from future wrong acts, (c) deterring other individuals who witness the punishment, and (d) incapacitating the indi- vidual for a period to prevent the commission of other wrong actions, just as one would cage a dangerous animal. The last reason (e), which seems to me to exist on a somewhat differ- ent plane, is called “just deserts,” or closer to the bone, “retribution.” The notion here is one of [ex talonis~society must punish justly an individual who commits certain sorts of wrong actions. Evil actions invariably are seen by people as requiring this last sort of punishment; mere wrongdoing does not always elicit this requirement. These are the ways that I think ordinary people in our culture come to identify acts as evil. I should admit that, armed with these definitional remarks, I do not think that it is possible to definitely, unequivocally categorize certain acts as evil, and certain others as merely “bad.” That limitation is not fatal to the kind of psychological enterprise I am attempt— ing. Instead I claim three things. First, a person who identi— fies an act as evil, when operating at the level of day—to—day understandings, does so with considerable personal certain- ty. Second, and again at this level, there is a considerable consensus among people about acts they will classify as evil. (Not surprisingly, this consensus contributes to the sense of certainty of classification that each individual feels.) Third, principles ordinary people say lie behind their intuitions about evil are the ones I have spelled out.7 Of these three 6Among the other things that arise is an inference of craziness. The two do not preclude each other. 7The reader is invited to explore this. Imagine a series of acts, some of which seem bad and others evil, and examine the difi‘erences. Ask others about ese cases. If your experience matches mine, the others will agree with yd}: classifications of most of the acts. Some may be classified differ- ently; t is is because the individuals doing the classifying are emphasizing different aspects of the ways of thinking about evil that I have elucidated. Alternatively, they may be bringing to bear genuinely different perspectives. 202 DARLEY claims, the first—that people, in their day-to-day lives, make untroubled judgments of what actions are evil——is cen— tral to my argument. Evil Actors and the Kernel of Evil If we have an intuitive sense of an evil act, how do we pass from that to a recognition of the evilness of a person? Ob- viously, an evil person is one who has committed an evil action, and we are even more certain of our attribution of evilness if that individual has committed many evil actions, particularly if they all seem to point to some consistent origin of that evil, that is, to locate the evildoer in some particular corner of the linked set of domains that represents our con- ceptualization of evil. Saddam Hussein is a current candi- date, as was Ted Bundy, the psychopathic serial killer of young women.8 Second, I suggest that we intuitively require that the evil- doer will himself or herself be found to contain an element of evil, something with an almost physical characteristic (al— though, as in the suspense thriller, this evil may be hidden from outside scrutiny behind a mask of ordinariness and require unmasking). The evil person is expected to possess a “quality of evilness” having properties like the ones Allport (1937) attributed to central and cardinal traits. And intu— itively, this internal quality of evil is matched in magnitude to the quantity of evil that we assess as having resided in his or her evil actions. As good crime novelists recognize, we re— quire our evildoers to be major figures, with something of the demonic about them, rather than pathetic figures in the grip of impulse. Putting this in a related but somewhat different way, it is as if there is a naive assumption of an enduring kernel of evil which, once detected in the act, must be present in the actor. Behind evil actions must lie evil individuals. At some day-to—day, “gut” level, we conceptualize evil actions as springing from the depraved minds of evil persons who will be found to contain a core quality of evil. I do not have what psychologists usually count as evidence for this assertion, but I offer two reasons for you to take it seriously. First, I think it probably accords with your intuitions, and remember that it is these intuitions, shared by members of our culture, thatl am attempting to articulate. Second, take seriously the evidence presented by the existence of the sus— pense novels to which I have just alluded. Consider the enor- mous popularity of a rapidly growing genre of suspense novel, the “serial killer” novel. In it, a series of serial mur- ders are committed, often in a bizarre and horrible fashion. At some point, the reader is led into the mind of the serial 8Because Saddam Hussein has some possibility of playing the role for this generation that Hitler has played for earlier generations, a few remarks about him may be appropriate. It needs to be remembered that these remarks are made soon after the conclusion of the ground war in Iraq. Briefly, I have no difficulty asserting that he is evil, and that the commander of the Allied forces (I suppose ultimately, George Bush) is not. This is so even though we assume for a moment that the Allied forces killed more civilians than did the Iraqi forces. The discussion to back that up would involve reference to Hussein’s indiscriminate gassing of fleeing Iraqi Kurdish tribespeople, quasi-random launching of Scud missiles, and so on. Certainly Hussein will be enshrined within the institutional memories of those determining what sorts of conflicts American military must be prepared to fight. We will be, and probably should be, prepared to fight evil nations for years to come. It is also likely, however, that Hussein will not be as great a figure in the pantheon of demons as Hitler because he fought so ineptly. Perhaps complete evil needs to almost prevail. killer, and the pattern generating the sequence of killings is made intelligible, although no less horrible, to the reader. The detective’s task is to intuit imaginatively the patterning in the sequence, predict the next crimes, and confront the evil killer in the act. The suspense is generated by the reader’s knowledge of the true patterning, and the reader’s neces- sarily passive watching, as the detective struggles toward understanding the pattern. Watching the killer rationally and logically set about preparations for the next killing act shows the discrepancy between the rationality of the plans and the horror of the contemplated act, and illustrates my earlier remarks about a depraved intention. For those unacquainted with the terrain of the suspense novel, some examples may be useful. Thomas Harris seems to have an excellent sense of the genre. His first novel intro- duced a psychopath, a psychiatrist, which created the in- teresting possibility of the psychopath bringing to bear some very sophisticated interpretations of his own pathology. In two later novels, The Red Dragon (1981) and The Silence of the Lambs (1989), he has created other plausible serial killers and an interesting system of cross-referencing by having his original killer, now in solitary confinement, consulted by the detectives trying to enter the mind of their current quarry. Harris’s books succeed, others fail. One way in which suspense books fail is that they trivialize the evildoer, and the failure is on terms revealed by our analysis. He is portrayed by the author as pathetic rather than demonic; the killer, once discovered, does not have the chilling quality of evil that his actions signaled and that is required by our everyday concep— tualizations of evil.9 (His is appropriate here. Both in sus- pense novels and real criminal statistics, the serial murderer is almost always a man.) The “quantum of evil” that we require to be preserved between the act and the actor is not present. To give an example of this in an otherwise quite well—conceived serial-murder novel, one might read Patricia Cornwell’s (1990) Postmortem. Psychological Functions of the Everyday Conceptualization of Evil Sometimes, like the poor suspense novel, the real world fails us. That is, we cannot find the requisite quantum of evil when we examine the perpetrator of some particularly evil set of crimes. When this is so, I suggest, we fall back on some alternate models of explaining evil actions. A fellow named Whitman, who one day went up to the top of the University of Texas bell tower and shot and killed a number of his fellow students, had a rather wholesome, Boy-Scoutish, back— ground. Rather fruitlessly, for several days after the incident, the newspapers scrabbled for “the story”—on my terms scrabbled for some prior evidence that Whitman was evil behind his amiable mask, and some further story that would 9A comment about mental illness, and the relations between concep- tualizations of individuals as crazy and as evil: Probably the most famous example of the serial-killer plot in movie form is Alfred Hitchcock’s Psycho. Certainly, one would not deny that Norman Bates (played by Anthony Per- kins) is crazy. This is brilliantly suggested by his conversations with his mother, whom we discover is mummified. However (and this is why the movie works), we retain a perception of Bates as containing that quantum of evil that I suggest is necessary to fulfill our stereotyped vision and for the psychological drama to grip us. In the current classic of the genre, Silence of the Lambs, Anthony Hopkins does an equally good job of showing us that it is possible to be both crazy and evil. BOOK REVIEW ESSAY 203 account for the origins of that evil. None was found, but rumors of a brain tumor detected on autopsy provided an alternate, physical, source of causality for the acts that other- wise would require an evil individual to produce them. My point is this: These alternate explanations for evil actions have a particular function; they are the “licensed exceptions” to the quantum-of—evil view and thus protect the application of that View to other unexamined cases. For most of us, cerebral dysfunction whether by tumors, or involuntarily in— gested mind—altering drugs, would explain acts we would otherwise ascribe to an evil actor. For some, other more psychological disturbances , such as posttraumatic stress syn— drome or schizophrenia, also provide an alternative explana- tion for acts that would otherwise be taken as revealing the evil nature of the actor. Notice that these alternate explana— tions work best when the entities being postulated, such as tumors , are imagined to have a thinglike character, much like the quantum of evil I have suggested people imagine. In common-sense psychology, we conceive of the brain as oc- casionally intruding into the mind, so the more physical and palpable one can imagine the suggested intruder, the more comfortably it fits this model.10 Here is where I think that clinical psychology and psychia— try fit in. I want to tread carefully here. More accurately, here is where I think that ordinary people’s conceptualizations of psychology and psychiatry fit in. Although psychologists and psychiatrists sustain complex perspectives on these is— sues, I think that a frequent cultural use of the everyday equivalents of concepts in those fields is to provide alternate accounts of evil actions that attribute the act to other than an evil person. For instance, a culturally accepted form of the psychodynamic perspective, popularly thought to trace adult thought disorder to experiences inflicted on the person as a child, and inaccessible to current conscious control, removes the onus of evil from the individual who commits evil actions. Sometimes , as I have already noted, the brain does intrude into the mind, and a tumor or hormonal dysfunction causes deviant and sometimes harmful behavior. As competent de- fense lawyers have long realized, sometimes a diagnostic label can be made to serve the same apparent explanatory function, although sometimes spuriously so to my mind. Much of the debate around certain diagnostic categories, such as posttraumatic stress disorder, stems from the fact that when they are used as legal defenses, lawyers can lead jurors 10Readers will recognize this as a special version of Goifman’s general account of retrospective biographical reinterpretation. In asylums, he point- ed out (Goffman, 1961), the process of interviewing and testing the mental patient is the process of constructing a biography leading up to the mental breakdown or criminal episode that caused the inmate to be incarcerated in the asylum. This “leading up to” is accomplished by discovering patholog- ical meanings in events that might otherwise have been regarded as neutral, and emphasizing the frequency of occurrence of negative events in the individual’s biography. Rosenhan (1973), in his famous article on the mental hospital careers of the normal individuals that he signed in to those hospitals, gave a particularly poignant example of this. He reported what one of the “pseudopatients” (I suspect himself) told the interviewer about his child— hood, and then gave the interviewer’s write-up of the case. The description is normal, the case write—up reeks of pathology. Kai Erikson (1989) gener- alized the notion of reconstruction further. He argued that, although history is lived‘from the past to the future, whether we are constructing the history of a person or a social movement, we do so from the present into the past, emphasizing those elements of the past that can be seen as leading into the present state of things. To us, for example, the South of the 1850s will always be the “pre-Civi1 War South.” to accept the “medical” metatheory underlying the particular diagnosis in question. What are the psychological functions served by the view of evil that these exceptions serve to defend, the notion that behind evil actions lie evildoers who can be identified as possessing an inherent and inward evilness? At first glance it seems to add an unnecessary component of terror to a world that is not short on other experiences of terror. Why do we imagine a world inhabited by evildoers who could, for ra— tionally unfathomable reasons, inflict horrors on ourselves and those we love, essentially at random? Although this will be initially counterintuitive, I want to suggest that it preserves our belief in a just and ordered world. Just as the brain tumor functioned as the licensed exception to the requirement of an evil personality to lie behind the evil action, the notion of the evil individual is a licensed exception to, and thus protects our notion of , a just and ordered society. The problem is this; in this era, we cannot sustain a belief in a world in which only good things happen to good people. Television and the newspapers all too often remind us that joggers are beaten and raped by “wild- ing” teenagers, innocent passengers are blown up in air— planes or machine—gunned in airports, or good Samaritans are murdered in the course of their helping activities. So, to some degree, it is not a just and ordered world. But we do not want to relinquish the notion of the just world; it gives us the courage to go out into the world and to send our children out into the world. How can we maintain the notion of the just and ordered world, and yet recognize the undeniable occur— rence of unjust actions?11 We recognize the unjust action but provide ourselves with a rule that at least partially restores order and justice and gives us some predictive power concerning those times when the order—and—justice rules are not in effect. They are not in effect when evil persons are around. And evil persons are generally recognizable; they contain the quantum or essence of evil that I have described. Of course, we cannot always perfectly identify evil people and we are sometimes taken in because their evilness is hidden behind a mask of normality. Thus , terrifyingly, evil actions happen to people because they did not discern the evil character of the perpetrator. However, and here wereturn to the suspense novels, the evil is recog- nizable in principle once we learn to “see it.”12 So, if the world is not a completely just and orderly place, we can know when it is not——when the general rules are in tempo- rary abeyance. They are in abeyance when evil individuals enter the picture. 13 This, I submit, is a more tolerable excep— tion to the principle of the just world than would be many 11Humanity has a long history of struggling with the problem posed by the existence of evil. For instance, it is obviously in conflict with the concept of an all-powerful God, and thus has generated various views that the early Christian church labeled heretical. 12It is the task of the suspense novel’s protagonist to discover which individual is evil by conceptually grasping the pattern of the serial killings. Here we use the notion of “protagonist” in the technically incorrect but usual sense of hero-protagonist—the detective who is tracking the killer. In the better—done suspense books, it is possible to make the case that the role of protagonist is shared equally between the villain-seeker and the villain. 13It is as if the world fits a generic plot for science-fiction novels. Some- how living among us is a race of aliens, creatures from another planet, who commit evil. Under certain conditions, we can recognize the aliens for whom they are, and thus guard ourselves against their evil actions. However, terrifyingly, sometimes these aliens appear in the guise of humans, making it difficult to recognize them and guard against their actions. 204 DARLEY other realizations about evil. Specifically, it is a more com- fortable and containable exception to the principles of order and justice than are the views on the origins of evil jointly contained in a recently published and deeply disturbing set of books by social psychologists. It is to a consideration of those books that I now turn. The Social Psychological Conceptualization of Evil Consider a hypothetical experiment. By sampling news- paper accounts and other sources, we identify a large number of evil actions, and we set out to interview the actors who committed them, looking for this quantum or kernel of evil- ness. Next consider a possible but disturbing outcome: When one probes behind evil actions, one normally finds, not an evil individual viciously forwarding diabolical schemes, but instead ordinary individuals who have done acts of evil be— cause they were caught up in complex social forces. The quantum of evil that we look for in the individual cannot be found. Instead we encounter again what Hannah Arendt found so striking about the Nazi mass murderer, Adolf Eichmann: the banality and ordinariness of an individual whom we expected to be demonic. Surely, though, we can discover some such evil indi- viduals and would expect to find them among the group of people Lifton studied. He, you will remember, studied the participation of medical doctors in horrendous acts of torture and murder in Nazi death camps. Surely, it is among Lifton’s respondents, the medically trained upper—middle-class indi— viduals who apparently chose to participate in horrible ac- tivities, that we would most expect to confront evil face to face. It is among them that we would expect to see the evil motives and evil hearts of the evildoing actors. Sensing this agenda, this search, Lifton early on warns us that our search will fail. “The disturbing psychological truth [is] that participation in mass murder need not require emo— tions as extreme or demonic as would seem appropriate for such a malignant project. Or to put the matter another way, ordinary people can commit demonic acts” (p. 5). Staub remarks that “I believe that tragically human beings have the capacity to come to experience killing other people as noth- ing extraordinary” (p. 13). These, I want to assert, are examples of the major message arising from the books. Its validity is strengthened by the independent convergence of these books on this conclusion. The books examine a variety of evil acts. Lifton, using a close lens, examines the involvement of doctors in the Nazi death camps. Staub, using a longer lens, examines many episodes of genocide, mass killing, and torture. Milgram, as is well known, examined the behavior of individuals in a “psychological experiment” in which they were ordered to give what seemed to be painful and harmful electric shocks to others, thus giving us what may be regarded as the closest possible experimental analogue to evil. Kelman and Hamil- ton tell us a good deal, indeed sometimes as much as we can bear to read, about the circumstances leading US. Army units to the massacre of Vietnamese women and children at My Lai. As we will see, the proposition that arises from all these books is that many evil actions are not the volitional products of individual evildoers. Instead, they are in some sense societal products, in which a complex series of social forces interact to cause individuals to commit multiple acts of stunning evil. 14 In that process, the individuals committing the evil are themselves changed. They become evil although they still do not show the demonic properties suggested by our conventional views of evil. The social psychological perspective suggests that gener— ally organizations are required to produce sustained evil ac- tions. The specific social forces that alter individuals are those produced in organizations. One needs a Nazi dic- tatorship, a Vietnam war, a Stalinesque gulag, or an Argenti- nean military dictatorship to train, reinforce, and sustain killing activities (although as shown later, it is not only these sorts of organizations that socialize their members into evil— doing). This realization leads to several questions. What forces create these organizations and put their evil activities in motion? How do they alter the character of those indi— viduals caught up in their activities? How do these organiza- tions grow and change? Of these three questions, the second, involving the alteration of the individual by organizational and small-group processes, is the one most congenial to social psychologists, and I consider it first. But we need to consider the other two questions as well. Organizations Socialize Individuals Into Evildoing How does an organization enlist individuals in harrndoing, V and how are they altered by their involvement? This question recognizes that the output of these organizations is twofold; first and horribly, corpses, and second, and less commonly recognized, individuals who have been fundamentally al- tered by their participation in the harrndoing activities of the organization. Killing organizations produce those who are killed, and those who kill. Lifton, Milgram, Staub, and Kel— man and Hamilton tell us how this is so; how organizations produce killers. The Doubled Personality As those of us who have read him over the years are well aware, Lifton’s continuing concern has been with the darker issues of human existence: the meanings we attach to life when impersonal forces frequently threaten life, and the meanings we attach to deaths that we inflict on others. He has examined the collective forces that lead to these deaths, and the ways those collective forces act and interact with the perceptions and constructions of the individual to produce the actions that the individuals take. The wars of this century have furnished him with a rich set of materials, and brought out in him a correspondingly richly nuanced and intertwined analysis. In his recent book, Nazi Doctors: Medical Killing and the Psychology of Genocide, he continues this examina- tion. Drawing on his work, we see how human beings adapt- ed to participate in evil and were altered by it. Lifton interviewed German doctors stationed in Ausch— witz during World War II. He also interviewed prisoners who 1“Of course, the authors would not want to deny the existence of evil done by evil individuals, who do meet the specifications of our intuitive require- ments for the evil actor. However, if we can extend their argument, surely they are showing us that the huge predominance of evil actions committed in the world are in keeping with their model rather than the intuitive individual- origins model. And surely they are correct here; surely this is one thing that the fate of the 6 million means. BOOK REVIEW ESSAY 205 had been in the concentration camps, particularly those who had been both prisoners and somehow involved in the medi- cal system set up within the camps—frequently those who were themselves doctors. My reconstruction of Lifton’s anal- ysis takes this path. When transferred to duty at Auschwitz, the doctor was confronted with the discovery that the ma— chinery of state had certain aspects that were, on the face of them, morally terrible. The ideology of Nazism, and the resurgence of the German state, apparently required the in— carceration and eventual killing of the “lesser” races. Still, as it always is, it was possible to see these terrible actions as somehow required to achieve generally good actions. The doctors perhaps had a feeling akin to what is called “dramat— ic inevitability” in the theater. Some entirely unexpected and perhaps violent outcome occurs, and yet the witnesses to it saw why that outcome was inevitably contained within the seeds of what had gone before. At Auschwitz, the Nazi doc— tors saw the inevitable unfolding of the meaning of the oath they took when they pledged allegiance to the state. At the moment of confronting the horrible reality of Auschwitz, although the doctors’ thinking was likely to be confused, nightmarish, and self-contradictory, certain bed- rock truths would be confronted if the doctor reasoned far enough. First, the terrible machine would go on, whether or not the doctor participated in it. Auschwitz was, among other things, a vast complex of buildings and trains and medical wards and persons and schedules and procedures, an enter- prise that would continue working regardless of the doctor’s degree of participation. Second, although apparently the doctor could have declined to participate, that choice led at a minimum, to the dangers of the Russian Front. Doctors were in the military and were assigned to Auschwitz; that there was a choice about their participation was not necessarily apparent to them. (Of course, it was apparent to some, be- cause they chose not to participate, an act of considerable courage in the circumstances)” Meanwhile, old hands were available to socialize them, to help them in the process of conversion from outsider to insid— er. Let us examine that process. Apparently “selection” was one of the most frequent and taxing ordeals faced by the doctors. It was their task to select those who would be al— lowed to live and those who would immediately be sent to the gas chambers. The determination was supposed to be on the basis of those who remained fit to work, as against those who were exhausted by near starvation, were ill, or simply looked “unfit.” Given the Nazi ideology, this killing of the lesser races was conceived of as a public health decision, somehow continuous with their previous program of euthanasia for the mentally inferior; thus, selection was fitted into a version of the medical ethic under the heading of the ruthless extirpa- tion of germs, of loathsome diseases. Selection was a medi- cal decision made by the SS doctors. And night and day they made it. When the prisoners left their boxcars to enter the camp, an SS doctor stationed on the arrival ramp selected 15Lifton wrote me about this (personal communication, 1991). Probably the most accurate description would be this: A doctor was often aware that it might have been possible to escape ramp duty, but to refuse it would have meant risking front-line duty, perhaps on the dreaded Eastern Front, with the risk of death that that entailed. This realization would have led to a desire to put the possibility of choice out of mind. The doctor’s perceptions of choice and no—choice, therefore, were likely to rotate and revolve. The reader really must look at Lifton’s very careful discussion of these matters. those who would live and those who would die. When the prisoners went out to work for the day and when they came back at night, they passed a selection doctor—or failed to pass. Selection was incessant. In the medical wards, those too sick were “selected” for death; when new arrivals reached the camp, many of those already in the camp were selected to make room for the new arrivals. Selection was by no means the most unambiguously mor— ally wrong thing the doctors did. However, Lifton’s inter- views reveal that the doctors found it extremely stressful. They often did it drunk, or got drunk after it. Apparently, it caused them to face the moral implications of what they were doing, and did so in a particularly pointed way for the new doctors soon after their arrival in the camp. The task for the new doctor was to fit into this machine. Other doctors provided whatever rationalizations were nec- essary to promote this “adaptation.” This being the task, then it was in some sense better to still one’s doubts about what one was doing, to develop a network of beliefs that stilled the moral doubts. Doctors became preoccupied with adapting them— selves to that reality, and moral revulsion could be converted into feelings of discomfort, unhappiness, anxiety, and despair. Subjective struggles could re- place moral questions. They became concerned not with the evil of the environment but with how to come to some terms with the place. (Lifton, 1986, p. 199) What Lifton is suggesting is that human beings have the capacity to adapt to moral wrongdoing taking place within organizational settings and, although at some psychic cost, to blank out the implications of those actions and function as a cog within the terrible machine. He quotes one doctor as remarking “after a few weeks in that milieu, one thinks, ‘Yes’ ” (p. 199). Lifton’s doctors found several ways of playing their roles. It is important to realize that the machine can tolerate differ— ent levels of commitment and even actions from its partici— pants. An inmate, writing retrospectively, thought that the doctors seemed to fall into three categories: . Zealots who participated eagerly in the extermination process and even did ‘extra work’ on behalf of killing; those who went about the process more or less me- thodically and did no more and no less that they felt they had to do, and those who participated in the exter— mination process only reluctantly. (Langbein, para— phrased in Lifton, p. 194) But, looking at it from an outside perspective, any level of participation was suflicient to keep the machine in motion. Organizations that kill do not need all individuals to partici- pate in the most direct acts of killing; many individuals are needed to fill subsidiary and support roles. Initially one suspects, the doctors stumbled through their dreadful activities, largely perceiving themselves as follow— ing orders, the implications of which they did not completely understand. Descriptively, however, their participation next began to be more voluntary, less well conceptualized as fol- lowing orders, now functioning more independently and au- tonomously, and drawing on their skills and knowledge to increase the efiectiveness of what they did within the camp. Famously, Lifton suggests that they adapted by the act of 206 DARLEY “doubling.” Doubling is “the division of the self into two functioning wholes, so that a part-self acts as an entire self” (p. 418). Doubling created a self that would function within the Auschwitz walls, that still remained in contact with, and drew on the knowledge—~and strength—of the pre- Auschwitz self. Doubling takes place largely outside of con— sciousness, and promotes the avoidance of guilt because it is the doubled self that commits actions, and the doubled self is the one that renders coherent the entire Auschwitz environ— ment. The Auschwitz self avoids guilt because it is upholding the moral principles of the Auschwitz surroundings, promot- ing the values of the state, achieving racial purity, staying loyal to one’s oath of obedience, and so on. Lifton is offering us two propositions, or at least I have abstracted two from him. The first one is shocking, but I think in keeping with what many social psychologists would want to say. Situations can be created in which it is possible to enlist the ordinary participant in the commission of evil, and in the process the participant is transformed into a crea— ture capable of autonomously and knowledgeany commit- ting evil actions. Importantly, this conversion is, as most conversions are, a process. Lifton’s second proposition is that this conversion process produces a doubled individual, [Putting this another way, that it is useful to have the concept of doubling because all of the processes that work on the inductee to the machine of terror converge to produce a personality that can only be described as doubled] Putting this yet another way, those processes work in unison to pro— duce a person whose personality is split in a particular way. A personality is formed that is designed to cope with the ex- igencies of the killing situation, but one which can and does access the skills and knowledge of the prior personality. Milgram’s Contribution: The Agentic State The problem to assess now is whether we ought to postu- late that those who pass through experiences such as the Nazi doctors had, are altered in a way that requires concepts such as “the doubled individual.” I waver about the necessity for the concept of doubling. In its favor, it makes clear that the person in question has been altered in permanent ways. [B utI am not sure that the evidence points toward a unified concep- tualization of the altered individual, and if it does, that “dou— bled” is the concept of choice for the resulting product] Seeking guidance on this, I returned to the Milgram experi- ments, and looked again at Milgram’s (1974) Obedience to Authority: An Experimental View. Recall the Milgram paradigm. A subject comes to partici- pate in an experiment on “teaching and learning” and is randomly assigned to give electric shocks to the learner when the learner makes mistakes in identifying the correct associ- ate of a stimulus word—this is an experiment about the effects of punishment. Instructed to do so, and with those instructions reiterated by an experimenter who is present throughout the process, the subject administers increasing levels of shock to the learner, even when the marking on the shock apparatus reveal that these shock levels are dangerous and the learner calls out protests. The copyright date, 1974, reminded me that the book had been published more than a decade after Milgram began his series of experiments, published by his own report, after he had spent some years wrestling with what he wanted to say about the meaning of his own work. Because the book came such a long time after what the psychological community regarded as the completion of the experimental program, I think that it has not played the role it should in shaping our interpretations of the Milgram findings. In fact, I suspect that it is not much read. If so, this is a pity because it contains reports of many experimental variations in the research para- digm that are reported nowhere else, as well as Milgram’s own interpretations of his findings. These interpretations I often find deeply insightful and, occasionally, deeply bizarre. Milgram certainly agrees with Lifton that a concept like doubling is required. He postulates the existence of an “agentic state” into which his subjects pass to administer shocks to the other individual. The assertion of the agentic state is one that I find startling and bizarre, but we ought to mark that two social scientists who have spent many years examining individuals involved in the commission of evil have both come to the conclusion that one commits evil in an altered state.16 What is the “agentic state” according to Milgram? He characterizes it in several ways, from which an image of it gradually emerges. The first characterization concerns its evolutionary nature. Human beings must often function within organizations. Thus, evolutionarily, according to Milgram, they have developed the potential for obedience. That is, the standard workings of evolutionary selection pres- sures have brought about an inherited propensity to obey. From an evolutionary standpoint each autonomously functioning element must be regulated against the un‘ restrained pursuit of appetites, of which the individual element is the chief beneficiary. The superego, con— science, or some similar mechanism that pits moral ideals against the uncontrolled expression of impulses fulfills this function. However, in the organizational mode, it is crucial for the operation of the system that these inhibitory mechanisms do not significantly con— flict with directions from higher—level components. Therefore when the individual is working on his own, conscience is brought into play. But when he functions in an organizational model, directions that come from the higher level component are not assessed against the internal standards of moral judgment. Only im— pulses generated within the individual, in the autono- mous mode, are so checked and regulated. (pp. 128— 129) Thus, we see the agentic state is one within which one is not governed by the operations of one’s own conscience; instead, the conscience has been switched off in the individual. Milgram’s views about the physiological substratum of the agentic state, and the events that “trigger” an individual into that state, also require examination. Where in a human being shall we find the switch that controls the transition from an autonomous to a sys~ l6Staub reminded me (personal communication, 1991) that he holds that it is not necessary to postulate an altered state to account for evildoing. Instead, his central explanatory mechanism is the motivated profound de- valuation and dehumanization of a target group or groups. These groups may be ethnically or religiously defined ( “Armenians,” “slopes,” “Jews,” “in- fidels”), but they can also be more abstractly defined (“Comrnunists,” “liberals,” “landowners”). The point is that a certain group of people are defined as those toward which the ordinary rules of morality no longer apply. BOOK REVIEW ESSAY 207 temic mode? No less than in the case of automata, there is certainly an alteration in the internal opera- tions of the person, and these, no doubt, reduce to shifts in patterns of neural functioning. Chemical in— hibitors and disinhibitors alter the probability of cer— tain neural pathways and sequences being used. But it is totally beyond our technical skill to specify this event at the chemoneurological level. (p. 133) Milgram also made clear how seriously he takes his con- cept. “The agentic state is the master attitude from which the observed behavior flows. The state of agency is more than a terminological burden imposed on the reader; it is the key— stone of our analysis . . . ” (p. 133). And further, Since the agentic state is largely a state of mind, some will say that this shift in attitude is not a real alteration in the state of the person. I would argue, however, that these shifts in individuals are precisely equivalent to those major alterations in the logic system of the auto- mata considered earlier. Of course, we do not have toggle switches emerging from our bodies, and the shifts are synaptically effected, but this makes them no less real. (p. 134) The first time I read this, I was startled and appalled by what I took to be the odd and pseudoscientific/ pseudophysiological concept of the agentic state, by the no- tion of the “trigger” that switches an individual between normal and agentic functioning, and by the dichotomous and all-or-nothing character of being in one state or another. And I continue to be. On rereading Milgram’s work, however, I also see that he gives an inherently more social and less dichotomous account of the agentic state. Milgram wrote: From a subjective standpoint, a person is in a state of agency when he defines himself in a social situation in a manner that renders him open to regulation by a person of higher status. In this condition the individual no longer views himself as responsible for his own actions but defines himself as an instrument for carry- ing out the wishes of others . . . An element of free choice determines whether the person defines himself in this way or not, but given the presence of certain critical releasers , the propensity to do so is exceedingly strong, and the shift is not freely reversible. (p. 134)17 Still I find this construction overly dichotomous, and by its reference to releasers, a term borrowed from a now some- what outmoded notion of physiological reflexes, overly pseudobiological. However, this more phenomenological construction of the process, coupled with Milgram’s later discussion of the situational events that enable an individual 17Lifton (personal communication, 1991), I think, has the most revealing construction of what I am trying to say about Milgram’s theory. Recall that Milgram’s subjects completed their task of administering what they regarded as high—level shocks within 1 experimental hour. The acuteness of this shift, or rather Milgram’s perceptions that this required an acutely abrupt shift, “made him think in terms of a sudden psychophysiological shift into an ‘agentic state.’ ” Lifton gently continued: “In contrast, what I observed was a more gradual process, though it had an acute or transition element—the anxiety Nazi doctors felt during their first weeks in Auschwitz until they made their adaptation, which in my view took the form of developing an Auschwitz self.” to construe himself or herself into obeying authorities, seems to me to contain many insights.18 For instance, he pointed out the importance of the sub- ject’s perception that he has willingly entered into a transac— tion governed by an authority that is legitimate and has the scope to command the particular actions in question. Sec- ond, once the interaction starts, other forces bind the subject into the situation. The cues that somebody is possibly being harmed occur only later, after a “momentum” has been built up around the legitimate definition of the punishing actions, and the shock—giving participant has incurred all the obliga- tions (that Goffman, 1961, has so convincingly pointed out) to continue an ongoing social activity and the definition of that activity. From these materials an account could be cre- ated of why the modal subject in many of the Milgram condi— tions gave the maximal level of shock. Albert Bandura, in his 1986 book, Social Foundations of Thought and Action, gave us an account that addresses what is accomplished within the individual by the conversion pro— cess. He suggested that normal socialization processes pro— duce what he called a “self—regulatory system,” which func- tions to regulate and control the actions of the normally functioning individual. 19 However, these control mecha— nisms do not operate in invariant ways. “Development of self—regulatory capabilities does not create an invariant con— trol mechanism, as implied by theories of internalization that incorporate entities such as conscience and superego as con- tinuous internal overseers of conduct” (p. 375). He went on to remark that “self-evaluative influences do not operate unless activated” (p. 375), and it is also the case that they can be selectively disengaged as well. This might work as fol- lows. Normally, people do not indulge in censurable behav— iors because these will produce self-devaluative conse— quences as a result of the working of the self—regulatory system. However, “what is culpable can be made honorable through cognitive restructuring” (p. 376). One can morally justify harmdoing, find euphemistic labels for the action, minimize the harmful consequences, and dehumanize and blame the victim. There are also the usual possibilities of displacing responsibility for the detrimental actions else— where in the system. These are the family of processes that seem to me to be involved in creating individuals who willingly do evil. That they often occur together, and that many of them are frequently caused by the same circum- stances, is undeniable. It is the fact that many of these pro- cesses are involved in the conversion process, and that the individual is in some unstable and dynamic state—now rely- ing on one justification, now on another—that seems to me to becomeobscured when an altered state is postulated. The Conversion Process This returns us to the issue of the conversion process. Psychologists are, understandably, reluctant to write the 18Kelman caused me to see this first. In his discussion of a talk I gave on the Milgram work, he pointed out that, physiological readiness aside, a good many elements in modern culture generated what is clearly a habitually assumed agentic role. When in that role, the role incumbent accepts the authority’s judgments of the right and wrong things to do. 1‘I’Bandura distinguished his account from one that would be given by what he called a “radical behaviorist” by giving causal weight to cognitive factors in governing behavior. However, the life events and external influ- ences shaping these cognitions would be similar to those that behaviorists would point to, including evaluative feedback, instruction, and modeling. 208 DARLEY “production of torturers” handbook, but social psychologists certainly have the knowledge to do so. The insights for it are there in such articles as Zimbardo (1969) on deindividuation, Gibson and Haritos—Fatouros (1986) on the recruitment of the Greek torturers during the reign of the colonels, and others. The essence of the process involves causing individuals, under pressure, to take small steps along a continuum that ends with evildoing. Each step is so small as to be essentially continuous with previous ones; after each step, the individual is positioned to take the next one. The individual’s morality follows rather than leads. Morality is retrospectively fitted to previous acts by rationalizations involving “higher goods,” “regrettable necessities,” and other rationalizations men— tioned by Bandura and others. Other books under consideration here give alternate ac- counts of Lifton’s conversion process, but do not contradict it. They draw on Milgram’s experimental work to illuminate their accounts. Without pausing here to sketch the account each gives, although that is a worthy task, I comment on what other elements of the process they call to our attention. Staub reminds us of the causal role of the bystander in the process—the interpreted meaning of the actions of those who stood by, not protesting, as harm was done to persecuted minorities. Their apparent indifference was certainly taken as tacit approval by others who also watched, and who might have been otherwise moved to protest. In a dynamic that Latané and I (1970) have described, this leads to a “con- tagion of inaction” among all bystanders. Thus, bystanders who fail to intervene, perhaps because they are stunned into passivity, are read by both perpetrator and Victim as condon- ing the acts of the perpetrator and approving the victimiza- tion of the harmed. Kristallnacht was a signal to the Jews of what the Nazi regime would do. The general lack of protest by other Germans, whatever the reasons, was a signal to the Jews that they would not be protected by non—Nazi Germans. Staub has thought deeply about the role of “bystanders” in the social processes we are describing, and he calls our atten- tion to the occasionally absolutely critical importance of their actions—0r more tragically their inactions. Often forces we are in the process of conceptualizing lead the “perpetrator” group to be constrained to continue along what Staub calls the “continuum of destructiveness.” J ustifying and ra- tionalizing what they have done, they are led to do more and do worse. Given this, it frequently will be the responses of the other elements in the society that determine whether this appalling progress of the perpetrator group continues and enrolls other elements of society, or is checked. For instance, an independent judiciary, with powers enough to stand against the usurping actions of the executive branch, can block those actions. Watergate. The Kelman and Hamilton book is the one I do the least justice to as I try to construct an account of the processes that socialize individuals toward evil, because they focused on how persons make moral judgments about crimes of obe- dience. (They did bring their thinking to bear on explaining how people pressed to commit those crimes chose or de- clined to participate.) In partial amends, later I sketch their general line of thinking, as they developed it in their study of people’s reactions to the trial of Lt. Calley. Here I take up another of their points, which begins with Kelman’s famous distinction of three modes of social influence: compliance, identification, and internalization. Each of those orienta— tions, they suggested, can link individuals to the society in which they find themselves. Compliance considerations pro- duce a rule orientation, in which a person is integrated into society via considerations of the rewards and punishments society delivers to those who follow or break its rules, and the social approval and disapproval that signal those rewards and punishments. Identification implies a commitment to a particular role within society as a part of the individual’s self— definition. Generally, through the processes of socialization, a person comes to accept the values of a society, and thus can be said to internalize them. The society’s values become his or her own, and naturally the person will act to further those values in the future. Given this, when a person violates some element of what society requires of him or her, he or she feels a mixture of fear of sanctions, distress for role failure, and regret for not living up to espoused values. All these are powerful enforcers of the person’s tendency to do what soci- ety asks of him or her. If the society is asking that individuals obey orders resulting in evil outcomes, still these enforcing forces move the person toward obedience. To the degree to which the nation-state, or any other orga- nization, is viewed by the individuals within it as legitimate, it has these powers to induce obedience with its demands. Organizations are perceived by their members as legitimate insofar as they engage their sentiments, such as loyalties, and fulfill their instrumental needs, desires, and interests (Kel— man & Hamilton, 1989, chap. 5, particularly pp. 112—119). We are analyzing how it is that people are socialized through an organization to commit evil. Two relevant im- plications flow from Kelman and Hamilton’s analysis of forces linking an individual to an organization. First (and the conclusion that we need here), within organizations that members perceive as legitimate, the forces leading to obe— dience are multiple, mutually reinforcing, and very strong. Obedience rather than disobedience to authority can be un— derstood as the expected outcome. Second, as shown later, the different conditions binding an individual to an organiza- tion may produce different behavioral outcomes under cer— tain kinds of pressures to obey. The Product of the Process: The Evil Individual Two intertwined issues are found in discussions of the individual who is the product of the perverse socialization process just described. First, how is the individual altered by the process, and second, are those alterations so great as to require the postulation of an discontinuous, dissociated state, such as an “agentic state” or “doubled personality”? Al- though we need to come to our own conclusions about those questions, let us first look at Milgram and Lifton’s con- clusions. As we have seen, Milgram, like Lifton, concluded that the evidence requires the postulation of a different and basically discontinuous state created in the harrndoing individual. But the two sharply differ on how that state is created. Lifton made a comment on Arendt’s thesis about the banality of evil, which can also be read as a comment on Milgram’s concept of the agentic state: What I have noted about the ordinariness of Nazi doc- tors as men would seem to be further evidence of her thesis. [Recall that the thesis involved the banality of evil—in this context the ordinariness in the present BOOK REVIEW ESSAY 209 of the individuals who had committed horrible acts of evil in the past] But not quite. Nazi doctors were banal, but what they did was not. Repeatedly in this study, I describe banal men performing demonic acts. In doing so—or in order to do so—the men changed; and in carrying out their actions, they themselves were no longer banal. (p. 5) Here Lifton reiterated his first point, that it is the indi- Vidual’s encounter with the killing machine that results in a conversion process altering that individual and creating a doubled personality. Although I have indicated my reserva- tions about the concept of the doubled personality, I com— pletely agree with Lifton that the encounter begins a process that morally alters the person who participates in that pro- cess. Although I am not sure that Lifton would agree with me, let me draw a line somewhere along the continuum of participation in the encounter, and suggest that people who go beyond that point are evil———more precisely, have become evil. The person who is induced into participation, and who goes far enough in the conversion process so that he or she autonomously and intelligently initiates evil actions, is an individual who has become evil. Examples may make this clearer. The soldiers who, yelled at by Lt. Calley, with tears in their eyes, fired into crowds of innocents were not evil. Those soldiers who coldly and knowingly killed innocents while operating independently, were evil. Staub reported a case of two young, rosy-checked Hitlerjugend whose fre— quent habit it was to “hunt” in the Warsaw ghetto. They simply wandered into the ghetto and shot whoever it captured their fancy to shoot.20 As the reader will be aware, many have debated whether humankind is inherently good, inherently evil, or any of several other possibilities. Psychologists working within the academic and experimental traditions of modern psychology do not enter this debate willingly. Nonetheless, the previous analysis suggests an answer. The possibility of being evil is latent in all of us, and can be made actual and active, among other ways, by the conversion process. The person who goes a certain distance in the process has been fundamentally changed, and is now capable of doing harm in an autono- mous way. He or she has “changed, changed utterly,” has become evil. But that is not at all what Milgram said. He argued that, as an inevitable condition of life, people come “prewired” as it were, with two possible states of functioning. Any person, therefore, could be put into the agentic state by the right combination of authoritative pressures. Milgram did not sug- gest that having once been thrown into that state, a person is fundamentally altered or that once a person has been through this process, it will be efiected more easily in the future. Not surprisingly, Milgram did not come to these conclu— sions. Were Milgram to have accepted Lifton’s construction, that it was the obedience to the initial commands of authority that began the conversion-to—evil process, he would have found himself wearing a very uncomfortable shoe indeed. It would mean Milgram had begun the process of converting his innocent subjects to “doubled” individuals, capable, if they went further down that path, of independently acting to inflict harm on other individuals in the name of science, as 20This description is quoted in Staub (1989, p. 139). It is taken from Karski (1944). did the Nazi doctors in the name of the state. To use Milgram’s vocabulary, he took himself as showing that the agentic state already existed in his subjects, and his experi— menter could rather easily “flip on” that state. But an un- comfortable alternate reading is possible: that Milgram had set up a terrible machine, and had begun to create (to again use the Lifton vocabulary) this doubled state in those unlucky individuals who were fed into the machine.21 Why do I suggest this? For both theoretical and empirical reasons. Many of us still find considerable explanatory power in dissonance theory, particularly in the attitude- changing effects produced by the forced-compliance para— digm, of which Milgram’s is one version. Thus the attitudes of the subjects about what they did can be predicted to have changed in directions favorable to repeating the actions. (Whether or not this is altered by the debriefing is unclear.) Nor is it only dissonance theory that would lead us to this conclusion. Any theory recognizing that the production of morally ambiguous actions can be rationalized after the fact by the actor would lead to these conclusions. Specifically, I would expect that Milgram’s subjects, who were implicitly preselected to put a high value on “science,” might have increased the value they placed on science as an important source of discoveries that would help humankind, and per- haps also derogated the intelligence of the individual who received the shocks. In both these ways, inflicting pain on the learner became justified.22 There is empirical evidence for the occurrence of these sorts of changes following the commission of a morally am- biguous act. In a set of studies involving an experimental cover story very similar to Milgrarn’s, and in which subjects were led to give electric shocks to others, Brock and Buss (1962) demonstrated that the subjects’ perceptions of the individual to whom they gave shocks altered in ways that 21Does this mean I am accusing Milgram of behaving unethically in inflicting his paradigm on subjects? The answer to this is complex, but has more elements of “no” than it does of “yes.” We now have a much clearer picture of the dynamics of the Milgram experiment than was possible at its inception. 22Perhaps this is a useful place to discuss the importance of “individual differences” in the account of harmdoing that I am shaping here. In it, given my background as a social psychologist, I take what might be called the “strong” situationalist perspective—that the social forces acting on the individual are sulficient to convert any individual, regardless of the strength of his or her personality and character, to being a cog in a killing machine. I should acknowledge two difficulties with this view: First, in this extreme form, it is likely to be wrong———a fact I would argue is relatively unimportant to the argument; second, my unfolding of the process does not emphasize one more historical truth that contributes to the development of the killing machine. Individuals with certain personality characteristics are likely to be recruited into the harmdoing organization and to contribute to its progress. The group of malcontents that originally formed around Hitler was certainly violence prone. Certainly, as Staub reminds us, destructive organizations, particularly in the early stages of their development, have the chance to recruit individuals who are predisposed to fit in with their destructive pur- poses, and the timely recruitment of these cadres may contribute greatly to their later domination of the societies in which they exist. Returning to the point about heroic individuals, in truth there are probably individuals who have the strength to “stand against” the forces that the present authors and I are conceptualizing. But my task in this essay is to elucidate the strengths of the forces that surround individuals caught up in these organizations, the forces that lead individuals along the continuum of destruction. 1 agree that there are some individuals who can stand against these pressures, but I do not accept the implication that is sometimes drawn from that—that the “real task” in resisting these atrocities is to produce more such heroes or martyrs. I see the central task of preventing these atrocities as existing at a different level, at' the level of societal institutions. 210 DARLEY justified their morally ambiguous actions; they dero gated the victim, implying that the Victim somehow deserved the punishment. I have asserted that being “processed through a killing machine” can create an “evil individual.” What exactly do I want to say about how that individual has been changed? And therefore, what do I want to assert that individual will think and, more to the point, do? What is the cash value of calling an individual “evil”? There are, I think, two answers to this, the first a partial perspective on the matter, the second a deeper perspective. First, in ways that the sociologist Weber initially conceptualized, the processes of doing evil have become routinized or alternately the person doing evil has become “a bureaucrat.” Actions that initially were shocking have become routinized, habitual, and at the end of the day in the concentration camps, the executioner can go home and read his children bedtime stories. This, I think, best fits the case of the individual who has some fragmented role in pro— ducing the evil action. When death, like cars or chairs, is produced on assembly lines, each individual eventually con— centrates on the microrequirements of his or her part in the process; the eventual outcome is rarely thought of. A group of police in a city round up the Jews and take them to a stadium. Later an army contingent takes them to the boxcars. A railroad worker throws the switches that bring the train to one or another subdestination on the way to the concentration camp. The fact of the eventual deaths is so remote that no participant finds it salient. Each person doing a subtask does so in a routinized way; it is only the final assembly of those subtasks that is horrible, and no individual “sees” that final solution. This explanation is good as far as it goes, but to my mind doesn’t go far enough. As described earlier, many people participated in the direct acts of killing, and many others knew where the boxcars were going. More needs to be said about the mental alterations that took place in those indi— viduals; Weber’s “routinization of bureaucratic subroutines” is not enough. They normally have been permanently moral— ly altered in ways that change their thinking and behavior. The continuing mark of their past experiences with the kill— ing machine is mental, and consists of the structures of moral thought that they were led to use to rationalize their actions in the first place. Unless they have had some sort of moral epiphany, they continue to believe what the killing machine taught them: “regrettable necessities,” “for the good of the state,” “the alien communist ideology threatened the Argen— tine way of life.” (This was the point of the banal conversa— tions that Arendt had with Eichmann, in which he went on and on about “necessities of state.” These statements re— flected the moral rationalizations that he had formed to justi— fy his conduct.) Two consequences flow from these mental adjustments. Both these consequences go some way toward answering the question: In what ways should we conceptualize the altera— tions in the evildoing individual? First, contained within these adjustments is a definition of the target groups toward whom harmdoing is deemed appropriate. In some ways this limits the operation of the forces that have been created. It is appropriate to execute the defined groups, but certainly not the groups on whose behalf one strives. One kills lesser races on behalf of the higher races, or one kills Communists to preserve the purity of Argentinean life. Of course, the target groups tend to grow larger. From our own history, not only “Communists” but also “fellow travelers” were seen as a threat to the “American way of life.” And soon, “unknow- ing fellow travelers” or “Communist dupes” also needed to be ferreted out. To cite a particularly horrible example, some of the Argentinean military were unable to conceive chil— ‘ dren, and so the practice grew of identifying young women who had just had a child as “Communists” so they could be killed. The children were then given to the military families. The groups that this sort of evildoer feels justified in attack— ing are contained, but only partially. The second consequence of the mental adjustments caused by participating in harmdoing organizations is straightfor— ward: an increased readiness to participate in harmdoing activities again if any of a number of social conditions are recreated. The normal outcome of the kind of socialization process I have described is a permanent one. This is not to deny the possibilities of a moral reorganization taking place, in which the individual turns away from his or her previous actions, and painfully reconstitutes a morality in which those previous activities are seen as morally wrong. However, the guilt produced by this, certainly in the moral sense and possi- bly in the legal sense, is going to be high. Lifton’s doctors had lived in a culture in which the wrongs of the Nazi era were about as thoroughly acknowledged as we can ever real— istically expect, and I think that Lifton does not see many signs of a moral reorganization on their part. My sense is that the negative moral change caused by the perverse socializa- tion that I describe generally persists, making the individual so socialized permanently susceptible to being caught up in harmdoing institutions in the future.23 One way of characterizing these mental adjustments is as a neutralization or even a positive valuation of actions that are generally regarded as morally reprehensible. This tells us how the person will continue to act vis-a—vis those actions. That individual will autonomously and independently con— tinue to harm others, but if and only if the harmdoing actions are rule governed by the rationalizations of previous harms, and if the social conditions are generally supportive of harmdoing in the present. Those who have become evil in this fashion require cultural or small-group support for the rationalizations that supported their doings. (And in this fashion they are quite different from those, such as serial murderers, who fit the common—sense prototype of evil- doers. They either are able to independently support what- ever conceptual system justifies their acts or have no need for such a system.) If the Nazis had successfully invaded Britain, then Nazi doctors would have helped design and build con— centration camps in Britain, brilliantly and logically adapting what they knew of selection ramps to the local conditions of 231 am wrong here. Interestingly, Lifton, whose concept of doubling I have read as creating a permanently doubled individual who is therefore available in the future to participate in evildoing activities, would want to say something different and more hopeful. Doubling can be both “prolonged and temporary, not necessarily permanen ” (Lifton, personal communica— tion, 1991). Although his doctors were for “the most part no longer actively doubling, though remnants of the ‘Auschwitz self’ could at certain moments be evoked,” I regard them as essentially and unequivocally “doubled” and thus recruitable to staff the death ramps if the bugle were to blow again. Lifton has a more nuanced and charitable View about this. He, we should remember, has firsthand knowledge of the individuals he interviewed; I did not. “Upon interviewing Nazi doctors decades later, . . . they could not then be called ‘evil men.’ So one can cease being an evil person when one’s life returns to ordinary pursuits, even without confronting the evil one had been part of” (Lifton, personal communication, 1991). BOOK REVIEW ESSAY 211 Manchester or Liverpool. History turned out otherwise, and many of them returned to the conventional practice of medi- cine. Those who tortured in Argentina or Greece can now be encountered on the streets, going about mundane activities. All would commit evil again if social conditions altered. The Creation of Killing Organizations My argument thus far is simple. Most evildoers are pro- duced by a process of socialization into doing evil, a process that makes them capable of doing evil autonomously and independently in the future.24 If this argument is correct, then we come facesto-face with a question that we now rec— ognize as urgent. How are the organizations that socialize an individual into doing harm created and sustained? If they are an important source of harmdoers, then how do they them— selves come about? I do not think we have a complete ac- count of how this happens, and what I have to suggest is tentative and incomplete. But I am sufficiently convinced of the importance of the task to make those suggestions, to advance the debate. First, let us look at the cultural and social conditions conducive to the development of such organiza— tions, and next at the specific events that are the origins of such organizations. The Cultural Preconditions for Destructive Organizations Staub sets himself the task of describing the cultural condi- tions that lead to genocide. Specifically, he has analyzed the Nazi Holocaust, the killings of the Armenians in Turkey from 1915 to 1916, the Cambodian massacre of their own coun— trymen during the 1970s, and the mass killings of “leftists” under the Argentinean dictators during the 19705. His analy- sis deserves our attention for two reasons. First, genocidal movements are certainly evil ones, and therefore of direct interest. Second, as I have come to see, we can extract some generalizations from his conclusions, which can be applied to explain the origins of other destructive social organiza— tions, in the interest of understanding these origins and the hope of preventing them. Staub’s story begins when a society or powerful groups within that society are subjected to difficult life conditions. The possible sources of these difficult life conditions are numerous and various; they can include economic hardship, political conflict between groups within a society (with the associated feelings of loss of control), perceived threats from criminal violence, as we currently experience in this country, and so on. As a psychologist is likely to do, Staub includes such things as threats to a sense of security, well-being, and even self—esteem as conditions that can be experienced as difficult life experiences. “The threat may be to life, to se— curity, to well-being, to self-concept, or to world view” (p. 14). These difficult life circumstances bring about physical 2"’Again Staub, who, most among the authors considered, has taken on the task of describing the many ways of coming to evildoing that a historical examination reveals, reminds us of an alternate path. The present account encases the individual in a social organization and shows how organizational pressure operates on that individual to turn him or her into an independent originator of evil actions. Staub points out that, in certain instances, an individual alone can move along this path. The husband who first strikes his wife, later with increasing frequency batters her to the point of injury, and finally kills her is one such example. and psychological needs in people that are sometimes filled in positive ways, in ways we would regard as effective and morally appropriate. At other times though, the circum— stances give rise to feelings of hostility directed at whomever can be made to seem responsible for the problems. Staub puts his case this way: Blaming others, scapegoating, diminishes our own responsibility. By pointing to a cause of the problems, it ofiers understanding which, although false, has great psychological usefulness. It promises a solution to problems by action against the scapegoat. And it allows people to feel connected as they join to scape— goat others. Devaluation of a subgroup helps to raise low self—esteem. Adopting an ideology provides a new world View and a vision of a better society that gives hope. Joining a group enables people to give up a burdensome self, adopt a new social identity, and gain a connection to other people. This requires action, but it is frequently not constructive action. Often all these tendencies work together. The groups that are attractive in hard times often provide an ideological blueprint for a better world and an en- emy who must be destroyed to fulfill the ideology. Sometimes having a scapegoat is the glue in the forma- tion of the group. But even if the ideology does not begin by identifying an enemy, one is likely to appear when fulfillment of the ideological program proves difficult. Thus these psychological tendencies have violent potentials. (p. 17) Certain cultural tendencies can make the forces unleashed during difficult times lead to scapegoating. What motives arise and how they are fulfilled depend on the characteristics of the culture and society. For example, a society that has long devalued a group and discriminated against its members, has strong respect for authority and has an overly superior and/ or vul- nerable self—concept is more likely to turn against a subgroup. (p. 4) Staub goes on to make another point, to which I have alluded earlier in this discussion. The scapegoating group is not ca— pable of leaping immediately to genocide, to killing the members of the scapegoated group just because of their group membership. There is the familiar progression of acts. Open criticism of the scapegoated groups produces deroga- tion, which licenses brutality; brutality is justified, and leads to further derogation and the discovery that the scapegoated group is somehow not included in humanity. Finally, kill- ings, and then systematic killings. Genocide. Initially, I found Staub’s essential conceptualization of “difficult life conditions” too broad. It is a notion that, if defined narrowly, we can all understand and know when to apply. For example, the rampant economic inflation of Weimar Germany led many previously well—to—do people into poverty, and created the difficult life conditions he al— ludes to. However. Staub’s notion of difficult life conditions is broader and less concrete. Essentially, he psychologizes it. Experiencing difficult life conditions is after all a psycholog— ical state, and may be caused not by obvious economic hard— ships or other deprivations such as those caused by famines, but by more symbolic and subjective disruptions. Threats to self-esteem, for instance, can cause the perception of diffi— 212 DARLEY cult life circumstances.25 By psychologizing the concept, Staub has made it potentially much more arguable in its application. It is relatively easy to determine when the mate— rial and economic conditions of an individual’s life have declined. It is going to be much harder to determine when, historically, it makes sense to say that individuals were experiencing psychological feelings of deprivation. Some peasants in Cambodia were experiencing painful economic hardship but conditions for some were improving. Some Argentineans were doing more poorly, some were not. Staub makes the argument, particularly in the case of Argentina, that the feelings of decline were, at least considerably, psy- chological in nature, stemming from a perception that Argen- tina’s hope of becoming a world power were fading. Al— though I do not think Staub stretches the application of his concept, I think it could be stretched by those who might use it in the future. (In fact, I am about to do so.) The core of my problem is this: I would be hard pressed to think of a time when one could not make the case that any population had reasons to feel psychological difficulties, either by compar- ing their status with that of past generations, individuals in other cultures, or some glorified notions of what they were entitled to; or alternatively, that some subgroups within a nation or culture had reasons to feel this way, given similar comparisons. Thus it seemed that Staub’s “difi‘icult life con— ditions” precondition for genocide could be found to exist in, if not all places at all times, then at least most places at most times.26 The part of Staub’s analysis that I have presented gives us the preconditions for a culture becoming “genocidal,” more specifically, for the national leadership turning to killing members of an out-group existing within or at the margins of that nation. But the analysis seems equally applicable at the organizational level. The cultural forces that he identifies must be the background conditions for the formation of orga- nizations that actually carry out the genocide. With this as background, we can now (at last) turn to questions concem— ing the formation of those organizations. If it is a central claim of the social psychological analysis that evildoing is frequently a product of organizational processes, then it is necessary to delineate how the organizations that come to produce evil outcomes and evildoing individuals come into being and reproduce themselves. 25There is an interesting parallel between Staub’s notion of life difficul- ties, and the hypothesis about the revolutionary potential generated in popu— lations experiencing rising expectations. That hypothesis suggests that revo— lutions occur when the gap between life conditions the masses are experiencing and the life conditions they can imagine experiencing grow larger. We could unify that hypothesis with Staub’s hypothesis by noting that both terms in the equation, the assessment of the current state and the assessment of what is possible, have psychological components (the expec- tations component, in fact, seems almost completely psychological in nature). Let us add the notion that people are aware of the source of the discrepancy between what is and what ought to be; that, for instance, they sense when their own conditions have worsened or those of other groups have improved, and they have or can be given some sense of the cause of this gap. This information may guide their choice of remedy, which can be either removing the government (revolution) if they identify it as the source of the gap, or eliminating subgroups within the population, if those subgroups can be scapegoated as the source of the gap. 26Obviously modem polling techniques create the possibility of determin— ing with some precision who it is that is experiencing significant feelings of deprivation. It is difficult to do this retrospectively, for populations of an- other culture, in which these sorts of surveys are not taken. The Origins of Destructive Organizations It seems, though, that I am avoiding my own central ques- tion. How do killing organizations come into being in the first place? Is it not the case that I have just spent a great deal of time simply making minor shifts in the origins of evil? Are the people who put together the first killing machines not evil in the demonic sense that our common—sense analysis sug- gests? The answer, of course, is frequently “yes.” Certainly Hitler was evil. Certainly too, the commanders of the Argen- tine forces who ordered the torture and killing of large seg— ments of Argentinean society completely intended the kill— ings they caused, and are evil.27 However——and this is perhaps the most disturbing element of my case—I believe that organizations can lurch toward evil, in ways not intended by any of the participants in the organization. Organizations of social control. First, however, con- sider the “yes” part of the story. Notice the organizations that we recognize as having a propensity for harm. As Kelman and Hamilton remark, The most obvious sources of crimes of obedience are military, paramilitary and social—control hierarchies, in which soldiers, security agents, and police take on role obligations that explicitly include the use of forces. These hierarchies are the classic ones from which the term chain of command is borrowed; author— ity is bureaucratically stringent. The goals of these bureaucracies and the role definitions of actors within them in fact require harm to certain categories of oth— ers (such as an enemy or subversive). The sole ques- tion concerns the scope and definition of the target of harm rather than the existence of such a target. (p. 3 l 4) In these organizations, coercive pressures are high. One “obeys orders,” and often one’s own life is in danger. Those to be controlled are the enemy and are often dehumanized. Criminals are called “scum”; Vietnamese are called “ gooks.” What one can do to the enemy or who counts as the enemy is rigidly rule-bound, but the reader will be well aware of the pressures to bend those rules and replace formal rules with informal rules that prescribe different and more lax standards. These informal rules come to govern behavior, and are well understood at the level of the police officer on the beat or the soldier in the field. We do not have much difficulty seeing how those organizations can shift toward becoming illicitly destructive machines; they are destructive machines to begin with. Of course, this does not imply that those who command such organizations are always evil, or that those organiza- tions must necessarily shift in this fashion. The Argentine generals and Hitler were evil; I, at least, do not think that the Army high command in the Vietnam conflict was. Yet large segments of the American Army did massacre civilians. So we face the fact that organizations can somehow be subverted or otherwise altered to turn persons within those organizations into evildoers, even when the apex of the orga- nization does not direct this. Again, we do not have too much 27They are also free. As of the end of 1990, they were pardoned by the President of Argentina, Carlos Menem. BOOK REVIEW ESSAY 213 trouble understanding how this happens in an organization set up for purposes of social control. We will need to say something specific about this, and will do so shortly. Before exploring this specifically, let us consider a more disturbing possibility. The analysis we have constructed so far can be read in the following way. Organizations of a particular type, roughly those concerned with social control, have the unique capacity to turn their members into evildoers. We have a reasonably clear notion of which organizations those are. Our task, therefore, is to be particularly vigilant in monitor— ing them, so that they do not consciously stray or uncon- sciously slip into creating evil actions, and in the process, create evildoers. Success in this task, admittedly difficult to achieve, will protect us from this problem. This analysis seems accurate as far as it goes, but it has one unfortunate implication. The truth is bleaker. Although I agree that organizations of social control are particularly vulnerable to this process, I disagree with the implicit con— tention that other kinds of organizations are not subject to similar problems. Normal organizations also bring their members to harm others. Normal organizations’ propensities for harm. Many organizations exist that would not be conventionally re— garded as organizations of social control: schools and univer- sities, manufacturing firms, research organizations. Do they need to figure in this discussion in any way? What is their potential for socializing individuals into the harmdoing pro— cess? My answer is that their potential for the incubation of harm is high, and in many cases that we can cite, that poten- tial has been actualized. Only one of the books we are exam- ining answers this question directly. Kelman and Hamilton invite us to consider several examples in which corporations or other organizations have gone far down this road. Recall the design of the Ford Pinto, sold for years by a company in which many executives were aware that it had a gas tank likely to rupture in low-speed rear-end crashes, and thus incinerate its passengers. Recall Watergate or the Iran-Con- tra affair. Consider the silence of Morton Thiokol executives who were aware of the dangers to the space shuttle O—rings of a low—temperature launch. We could add to Kelman and Hamilton’s list. Think about executives who continued to have shipyard employees work with asbestos long after its carcinogenic properties were known to them, or government bureaucrats who kept uranium miners at work long after the dangers of that occupation were known to them. Consider any number of defense contractors who have delivered mili— tary weapons systems to the defense department with faked safety and effectiveness tests and substandard internal elec- tronic components. A complicated set of issues are raised here. First, we have the case of an organization whose activities bring a great deal of harm to individuals, but in which it is hard to fix the responsibility for that harm within the organization. Second, we have the case of the organization in which evil individuals are produced but produced in a more complex way than they are produced in the concentration camp. Third, I argue that the division of organizations into those engaged in social control and related activities versus those engaged in, for instance, production or other purposes, is less useful for identifying organizations that may engage in harrndoin g than we might think. Let’s look first at the case in which harm results from the unfortunate assembly of a set of innocent actions. When an organization does harm, that action can be the result of the interaction of many other actions, each of which is, on the face of it, innocent. Assume an organization has produced and marketed a drug that is later found to have terrible side— efi’ects—thalidomide, or diethylstilbestrol (DES), for in— stance. One corporate unit can develop a drug, and assume it will be tested for side-effects. Another unit can arrange for it to be marketed, assuming those safety checks have been completed. Those who actually carried out the drug tests may be aware that their tests were not the sort of tests that can determine side-effects with any sort of precision. (For in— stance, consider DES, a drug given to pregnant women to prevent miscarriages and reduce nausea during pregnancy. Only many years later could it be discovered that it produced various effects, including increased likelihood of cancer, among the young women who had been in utero when their mothers were taking the drug.) No individual intentionally brought about the horrid side-efl’ects produced.28 More to the point, it is difficult to identify exactly who within the organi- zation was negligent in allowing the mistake to happen. This becomes apparent in those few cases in which members of organizations are put on trial for the consequences of their actions. Rarely are any specific individuals found criminally liable. As Kelman and Hamilton point out, this conclusion can be generalized to other cases in which, on first glance, there seem to be obvious wrongdoers to hold accountable. They report what they found in their survey of Americans’ reac— tions to “the Lt. Calley trial.” Recall the circumstances. During the Vietnam War, at the hamlet of My Lai, U.S. soldiers knowingly shot, clubbed, and bayoneted Viet- namese woman, children, old people, and babies. Once the cover—up was exposed, an investigation revealed that superi— or officers had ordered a “search and destroy” mission into a hamlet that intelligence indicated would be empty of civil- ians, and occupied only by enemy. Certainly nothing was said about the care to be taken if civilians were encountered in the hamlet. The company commander interpreted these orders to involve leveling the hamlet, and transmitted them to his subordinates, including Lt. Calley, who heard those or— ders as including the killing of the inhabitants. Calley’s platoon did so. They rounded up the inhabitants; saw that they were old people, women, and children; and massacred many of them. Calley both gave the orders to do so and shot many of them with his automatic weapon. En— listed men also shot civilians. Who was responsible for this? As Kelman and Hamilton analyze the situation, the answer depends on what kinds of 28A science fiction reader reminded me of Asimov’s famous laws govem- ing the conduct of robots. Obviously, in a thoughtful society, one would build a prohibition against killing a human being into the governing program of a robot. In one disconcerting short story, I recall that the author figured out a way that robots could bring about the death of a human being. The pro- grammer simply created a series of apparently innocent steps that brought about a human’s death. For instance, one robot puts rat poison in a container in the tool shed, another robot is told to move the container from the shed to a kitchen shelf, the third robot is told to put a spoon of “sugar” from the container into the victim’s coffee. To return from the fanciful, the general point is that a series of subroutines, each not morally outrageous, can sum to an outrageous act. 214 DARLEY responsibility we are considering. Clearly, Calley and the enlisted men had direct causal responsibility; they did it. They killed people and they did it knowingly. Yet we also hold the notion of role responsibility; in the military this is discussed under the notion of “command responsibility.” Officers are responsible for the conduct of their subordinates whether they ordered the actions of those subordinates or only allowed them to happen. From that point of View, the higher—ups were responsible for the massacre. They certainly gave no orders concerning the protection of innocent life; some listeners read them as “saying” that they wanted the inhabitants killed.29 Calley, the only individual who had both role and direct causal responsibility, was the only individual tried and con— victed for his actions. This makes intuitive sense. The en— listed men were seen as the physical cause of the killings, but it was recognized that they had a role responsibility to obey orders in general, which made them unattractive candidates for punishment. But looking at higher authorities, many felt that they had a plausible claim to deny that murder of inno— cents was what they meant. True, they in some sense allowed the murders to happen, because the murders did happen; but they were not present, perhaps couldn’t conceive that anyone would murder innocents, and so on. Only Calley, having both kinds of responsibility, was convicted. Kelman and Hamilton went on to discover another in— teresting fact. Their survey found that citizens had quite different reactions to Calley’s conviction, and those reactions could be related to which View of responsibility they held. Some respondents held to the View that the individual actor is responsible for his actions, and that this responsibility is not canceled if those actions are committed under orders. A majority of them approved of the Calley trial. Interestingly, they were also likely to say that the higher-ups should be tried. Others believed that in the massacre situation they themselves would have obeyed orders, most people would have obeyed orders, and it was unfair to try Calley for “doing so.” As people in this latter group also tended to disapprove of trying higher-up officers, they remind us of the way in which responsibility can be extraordinarily diminished when an organization commits an evil action. As Kelman and Hamilton remark, individual responsibility and commanded responsibil— ity together can add up to 200 percent responsibility, as they did for many (who asserted personal responsi— bility); or to 0 percent responsibility, as they did for many (who denied personal responsibility); or (pre— sumably) to anything in between. With disturbingly high frequency, joint responsibility for the My Lai massacre added up to zero in the public ’s eye. (p. 223) 29When uttered in different contexts, sentences sometimes “mean more than they say.” This is the psycholinguistic notion of pragmatic im- plicatures. For instance, I am teaching a seminar. John the student comes late to the seminar and leaves the door to the hall open as he slinks to a seat near the door. If I say, “John, it is noisy out in the hall,” I report a fact about noise, but surely mean that John should reach out and shut the door. It may go a step further; I may be telling John that his lateness is noticed and marked up against his name. This provides an interesting retrospective illumination of Milgram’s experiments. At some points the experimenter said to the shock giver, “Please go on; the experiment requires that you go on.” Because the shock giver knew that the experimenter had heard the victim’s protests, he read the meaning of that sentence to be, roughly, “I am in charge here and I am sure that the learner is not getting harmed. He may be getting a little hurt, but he agreed to that in the beginning." In sum then, sometimes in organizations an act harming others is innocently or unknowingly “assembled” from the actions of many individuals who are not aware that the conse— quence of the act to which their actions contribute will be destructive. Thus sometimes organizations perpetrate major evils, with no single individual having evil intentions, or being guilty of more than, arguably, negligence in not fore- seeing the harmful consequences. Even when the organiza- tion acts through an individual who knowingly commits evil actions, if the individual’s phenomenology is such that the individual may have regarded himself as acting at the behest or command of the organization, many people decrease the responsibility assigned to that person. Consider now a second path by which organizations that are not military or social-control organizations can bring about harmdoing actions, which may in turn change the ac— tors into autonomous sources of harm in the future. In many organizational settings in which an action that is going to result in harm to others is taken, at least initially, there is no overt target for the actions committed, no salient other human who is seen to be a victim of the actions. The indi— vidual who decides to let the assembly line use substandard cord in the fabrication of radial tires is not thinking of the accidents that decision could cause; he or she is simply keep— ing the assembly lines moving. Because a good many of the forces that cause people to avoid doing harm to others rely on the salient presence of specific or specifically imagined Vic— tims, if they are not present, then restraining forces are con— siderably weakened. “These opposing forces rest ultimately on the actor’s awareness that he or she is connected to a Victim” (p. 313), as Kelman and Hamilton remark. What this creates is the possibility that individuals within organizations can lose sight of the fact that individuals may be harmed in the course of fulfilling the other goals of the corporation or bureaucracy. Then, let us assume that, sud- denly and dramatically, it is discovered that the actions of the corporation have already harmed large classes of others. It is now realized that certain actions of the organization inevita- bly led to harmful outcomes. Pintos are actually rear-ended, gas tanks actually catch on fire, and actual passengers are actually, horribly killed. Memos are found to exist within the corporation in which design engineers warn about exactly these possibilities. To an outside observer, it seems apparent that those in the organization must have been aware of the harms risked, and thus, somewhere there must be evil indi— viduals who have knowingly brought about those evils.30 Inside the organization, however, the phenomenology is very different. It is possible that the negative outcomes simply could not have been anticipated by any individual— an unan- ticipated drug side-effect might be an example of this. It is more likely, however, that some evidence existed calling attention to the negative outcomes, but that the evidence was not given sufficient attention or weight within the organiza— tion. The people within the organization were focusing on other organizational goals and missed the meanings of the danger signals because they were “negligent, hurried, slop— 30And sometimes this is true. There are such evil individuals. One thinks of the bosses of the company that recovered precious metals using processes that involved terribly dangerous chemicals, took almost no safety precau— tions to shield the workers from the effects of these chemicals, and hired illegal immigrants who spoke little English so they would not understand what was happening or be able to reveal it to others. BOOK REVIEW ESSAY 215 py, or overworked” (Kelman & Hamilton, p. 3 12). To this list could be added interpersonal processes involving breakdown of communication and diffusion of responsibility. However it came about, harmful actions have been com- mitted (in this scenario), and now the individuals who had some responsibility for those actions have become aware of those consequences. To my mind, this is a critical point at which those individuals can become evil actors. Notice that the question is not whether or not to commit an immoral act. It is what to do when such an act has been committed and is now recognized. To those “organizationally responsible” for the harrndoing act, there are several choices, none of which is comfortable. These choices all seem to me to take the system in the direction of rationalization and cover-up, rather than toward acknowledgement and amends. Again, the mor- al essence of the situation is this: An organization has un— foreseeably, carelessly, or in some sense willfully harmed others. In the clear light of hindsight, to the organizational higher—ups, it must seem, as it seems to the potential outside observers, that the negative outcomes were at least foresee- able and perhaps, in the complex sense that an organization can be said to intend something, “intended.” There may be internal evidence that all the information was available with— in the organization to know that the effects would be harmful.31 Thus, were they to publicly or privately admit to the existence of the outcomes, or their role in producing them, they would be publicly convicted of harrndoing and internally faced with feeling shame and guilt. These are negative outcomes, which do not fit in with the people’s dim memories of the paths that led them to the present predica- ment. At this point, it must be extraordinme tempting to “cover up” the evidence if it is possible to do so. Several mechanisms are available for doing so, depending on who has become aware of what. If the negative consequences are known only within the organization, then their existence can be minimized or denied. Apparently executives in cigarette companies to this day deny that cigarettes cause cancer. Thus the person denies the negativity of the consequences, or the responsibility for those consequences, and in so doing denies guilt, both to himself and to others. The second part of the motive—to avoid appearing immoral in the eyes of others— leads to concealment of the harms from the outside world. Individuals in corporations, when they discovered, for exam- ple, that asbestos used by workers was leading to high rates of lung cancer, sometimes chose to conceal that fact, perhaps because they were concerned with the liabilities they would incur if they revealed that information. Concealment has a price. Covering up past evidence is also likely to lead to maintaining the current practices that bring about the harms; it is at this point that I think such an organizational actor becomes evil, becomes an independent perpetrator of further negative acts now knowingly done. Historically, it is clear that shipyard managers or nuclear— plant managers concealed increasing amounts of evidence that made clear to them that working with asbestos or mining “In the case of the Ford Motor Company, Kelrnan and Hamilton tell us that there were internal memos acknowledging the flaw, the danger it repre- sented, and the cost of a redesign fix. The fix was about $11 per car, which was then cost—benefit compared to the costs of the estimated payouts to persons killed in rear—end accidents, estimated at about $200,000 per death (p. 31 1). It would have been rather difficult to deny corporate foreseeability or intentionalin here. uranium (or even living downwind from nuclear plants) caused cancer. But often the failure to acknowledge past harms is to continue to commit those harms in the present. On one hand, this may be exactly what those who conceal intend to accomplish; they can continue practices that they now know are unsafe. But the psychological dynamics can be more complex. Consider the plight of the manager who, although he wants to conceal evidence of past harms, also wants to change current practices. In may not be possible. How can a shipyard worker who has worked with asbestos for 20 years interpret a sudden request to put on a filter mask? Thus it is often the case that, driven by a desire to hide past inadvertent harms, managers continue to have their workers operate in what they now know are dangerous settings, or otherwise engage in dangerous activities. They now do in— tentionally what they had previously done unknowingly. Concealment of harm within organizations is not easy. To maintain it, further concealments are likely to be necessary, even though these were perhaps not contemplated by the organizational actor at the moment of choice between ac— knowledging and denying the harms done. The evidence of previous harms had better disappear. Those in the organiza- tion who might discover the previous harms had better be hindered or muzzled. Computer memories had better be wiped clean. A chain of repugnant and evil actions are found to be required following the initial decision to conceal the initial harm. People who don’t think of themselves as corrupt find themselves burning incriminating documents and pay- ing out bribes to potential informers. (Needless to say, one thinks of Watergate here, and the Iran—Contra scandal.) At some point the “face” or honor of the organization becomes committed to the concealment and the processes of denying that real harms were done or real wrongs perpe- trated. For those versed in history the Dreyfus affair comes to mind. An individual was falsely convicted of treason on rather flimsy evidence. This became apparent and when con— siderable new evidence pointing to Dreyfus’s innocence had accumulated, a new trial was finally ordered. The original conviction was affirmed. A more likely candidate for the treasonous act was later tried, and although the evidence was better that he had committed the crime, he was acquitted. Why? Because those doing the retrospective reviewing fear- ed that to reverse the verdict would be to dishonor the mili— tary or admit to national disgrace. And they were right. To do it would have dishonored the military justice system and the French government. So they chose a path that further dishon- ored and discredited the system.32 What we have discovered here, I claim, is a second way that an individual can be caught up in and altered by a harrndoing process. Whatever else might be said about the Nazi doctor who stood on the selection ramp, designating those who would live and those who would die, he knew what he was doing. But often an individual within an organi— 32Lest anyone think this is an isolated incident, let me remind them that the British justice system is currently seriously dishonored by exactly the same pattern of events in the case of the “Birmingham Six,” a case where the initial conviction and subsequent denial of appeals of individuals accused of terrorist bombings were upheld. Initially, the system brought in a flawed verdict based on faked evidence. It dishonored itself by willfully blinding itself to this during the appeals process. More recently, the Los Angeles police department brutally beat a Black Speeder they stopped, then filed papers to cover this up. An amateur’s videotape of the beating derailed that cover—up. 216 DARLEY zation carries out what seem to be routine actions, and then discovers those actions had negative consequences that now seem to have been anticipatable. When he or she denies or conceals those consequences, and becomes enmeshed in a widening circle of actions necessary to maintain this denial and concealment, the person has moved to become an inde- pendent and autonomous perpetuator of the harms done. He or she has become evil. But the process here is an after—the— fact one, in which the person faces not the prospective choice to do harm, but the retrospective choice to acknowledge that his or her actions have already done harm. The more the person now sees that those harms should have been forseen, the more guilt, shame, and blame is acquired if he or she chooses to acknowledge them. “Normal” organizations sometimes intend harm. I suggest there is a third way that many organizations cause harm. Bluntly put, they set out to do it. That is, their corpo- rate ideologies make it appropriate to harm others, and the conditions of life include elements Staub identified as impor— tant in facilitating the development of genocidal practices. Reading Staub, while reading several other similar books at the same time, caused, as the saying goes, the penny to drop. That is, it occurred to me that many elements of Staub’s - analysis could be used to analyze some otherwise puzzling things that occur within organizations, which may explain how those organizations come to initiate harrndoing acts and convert their members into evil individuals. In many organi- zations, the members are in fear of losing their positions; thus, even if they are well—to—do, they are experiencing diffi- cult life conditions. Only if they perform up to a certain level will they keep their jobs. Or perhaps more interestingly, only if they perform up to very high levels will they advance within a corporation. Echoing Staub, to be deprived of op- portunities is to be in danger in a competitive corporate en— vironment. Second, within a good many corporate struc- tures, there exist certain well—identified groups whose interests are in some zero—sum relationship with the interests of the corporate group. When one is in a zero—sum rela- tionship with another group, it is easy to depersonalize mem- bers of that group and rationalize harming them. A few ex— amples may clarify this point. Union management relations often take on this perspective. Corporations competing for the same markets tend to regard the others as the enemy and act accordingly. Political parties competing for votes cer- tainly are in this relationship with one another, and we have recently seen cases that make the point. For instance, to justify to oneself launching the infamous Willie Horton cam— paign ads, one must have been convinced that the “other side” consisted of people who would so disastrously govern America that any means of stopping them was warranted. At the end of that path lies Richard Nixon, Watergate burglaries , and beating up demonstrators who might cause the electorate to vote against incumbents. In a similar vein, it is interesting to speculate about what certain corporations think of their customers—cigarette companies, for instance. Michael Lewis (1989), who wrote Liar’s Poker, provides a richly detailed case of how those joining a stock brokerage firm were socialized into regarding their customers as sheep to be fleeced. A good many customers’ lives were destroyed in the process, as the book reveals. At one point the narrator, Michael Lewis, describes selling a bond to a customer that somebody within his brokerage house had advised him was a good bond to sell. It fell, taking the customer down with it, and Lewis discovered that it was indeed “a good bond to sell.” It was one that the brokerage house held a large in— ventory on and had inside information of its impending fall. Thus, the brokers moved them out of inventory onto custom- ers, letting the customers take the upcorrring loss. Considerably hilarity ensues within the brokerage house. Perhaps bent, although not morally broken by a similar set of experiences, Lewis leaves the firm,33 but the socialization process that he describes corrupts many of the participants. Of course, it was intended to. That is, his book makes clear the willing participation of the firm’s managers in the corruption, and their calculated efforts to corrupt lower—level stalf. As an example, higher commissions were paid for moving poor-quality bonds off on unwary customers. One is reminded of the people who worked for the now—defunct Lincoln Savings and Loan Company, which sold a good many nongovemment-insured investments to elderly cus— tomers, allowing them to believe they were insured. Many lost their life savings. Thus, in several complex ways, organizations that are not social—control organizations can still corrupt their members. In fact, they can do this in several complex ways and in one simple way. In the simple way, that is what they intend to do, to corrupt their members into dealing unethically with people that they regard as the enemy or their appropriately de- humanized “marks to be fleeced.” The Reproduction of Destructive Organizations We now consider another aspect of the usually unthinka- ble. If one were to take on the task of duplicating a killing organization, how would one do so? The question of organi- zational reproduction is quite easy to answer. Implicit in our previous analysis is an account of how such organizations reproduce themselves and grow. Organizations such as those involved in the Nazi death camps have not one but two out- puts. They produce not only death, but individuals who be- come autonomously capable of and committed to producing other deaths. They produce evil individuals who become available for the reproduction of the evil organization. Con— cretely, SS officers and soldiers who first murdered civilians on the eastern front could then be used to stalf the concentra— tion camps and initiate and socialize other individuals into the new organizations. Older soldiers in the U.S. Army in Vietnam made clear to the new inductees how the war was really to be fought. Using a single evil organization “intel— ligently,” it can be made to produce a surplus of individuals who can be used to replicate the organization in other set— tings. Given that the individuals who have been “processed” by the evil organization have been brought to a point where they use their intelligence in the service of their evil actions, the replicated organizations can be counted on to transcend whatever local obstacles stand in the way of reproducing the results of the original organization. The staff of various con- centration camps made numerous grisly procedural refine— ments that increased the efficiency of their activities. The realization that evildoing organizations have the ca— pacity for self-replication provides part of the explanation for 33He was, after all, a Princeton man. BOOK REVIEW ESSAY 217 one of the facts that so bewilders us about, for instance, the Nazi death camps: Why were so many individuals willing to participate in their immoral activities? One answer is that different individuals were “trained” (a horrible word, used in this context) at different times, and they in turn trained others, and so the camps were staffed. Conclusions The argument has been a long one and I am not so con— vinced of its validity that I will try to summarize it. Instead, to draw others into the discussion, it seems useful for me to suggest some of its implications. We now have in modern culture a well-developed psychology concerned with the ori- gins of antisocial acts in the personality structures of those who originate those acts. The clinical investigation of “psy- chopaths,” “sociopaths,” “antisocial personalities,” and other diagnostic categories into which we encode those who best fit our intuitive definition of the evil individual, While by no means concluded, has much to say about the origins of those individuals.34 This analysis is the sort we are all natu- rally drawn to because it fits with our everyday concep- tualizations of evildoing, in which the person who does the evil contains the appropriate quantum of evil, which is ulti- mately recognizable even if it might not be apparent from casual scrutiny of the individual. But, the present argument goes, that individual—level psychology is largely irrelevant to the occurrence of a much more common source of evil ac- tions—~produced by what I call “organizational pathology.” We now need to create—and the authors mentioned in this essay are creating—a psychology and sociology of how human institutions can purposiver move or accidentally lurch toward causing these actions, somehow neutralizing, suspending, overriding, or replacing the moral scruples of their members. That psychology will inevitably be a social and organizational one, rather than one centered on the indi- vidual acting alone, although, as all of them show, it will draw on the conceptualization of an individual—level psy- chology, particularly to explain how the individual partici- pates in his or her training in the social movement and con— tinues to access his or her own particular skills in the service of the pathological group projects. _A polemical message lies behind this scientific one, or at least I have extracted one. It is too easy to defuse conceptually the chilling implications of evil actions by psychologically distancing them. Reading about an evil action, we assume that it was committed by an evildoer, a person who, because of a psyche twisted by genetic mischance or developmental trauma, is abnormal and evil. We assume evil actions nor— mally flow from the actions of individuals who contain the quantum of evil I described earlier. By doing this, I would argue, we preserve our belief in the essential justness of the world, by having a limited and contained generalization about when that justness will not prevail. We are assisted in this process in that we are led to commit what social psychol- ogists call the fundamental attribution error, to attribute 3“This is not to say that the current trend in clinical psychology is to label these individuals as evil. Indeed, it seems to be the opposite. As the pathology of each individual is examined, the individual’s evil character first shimmers, then disappears. Genetic influences hint at a deterministic story that makes the individual a victim rather than a cause of his impulses. Early childhood trauma has the same cause-removing effect. behavior to the internal dispositions of an individual rather than to recognize that it stems from situational pressure. Thinking about evil actions, we call to mind typical or modal representations (“prototypes” or exemplars) of such actions, and in examining those representations, we find that they include images of evil individuals. In our minds evil acts are committed by evil individuals.35 The point that the social psychologist wants to stress is this: Evildoing is not confined to individuals who are evil at the time of committing the act. Each of us has the capacity to do evil actions if our surroundings press us to do so. The wonderfulness of our upbringings and the goodness of our personalities do not protect us from doing so. What social psychologists generally call the forces of the situation, but what I have argued here is the recruitment into a killing organization, “socialize” us into committing evil actions?6 Many social psychologists, myself included, have made a good deal of intellectual and career yardage by demonstrat- ing the complex and compelling nature of the forces of the contextual pressures on people, and the definitions of the situation that they engender in those people. We thus show how actions that seem from the outside to be apathetic, or inhuman, are actually very human responses to flawed social situations. Frequently, having shown this, we end our lecture with a second message, which goes something like this: “Students. I have now made you aware of the ways in which social forces bring about the social constructions of the actors in the situation, and lead them to participate in doing evil. By making you aware of those social forces, I intend to enable you to resist the apparent imperatives of the situations in which you find yourselves, so you can avoid doing evil or step forward to do good.” It is a rare lecture on bystander responses to emergencies that I haven’t ended with that ring- ing affirmation of individual powers. I continue to believe in that message, but I find I cannot apply it with such conviction as I used to when I think about the cases discussed here. To resist the psychological forces characteristic of the organizations discussed herein, pres— sures which strike me as often highly coercive and reinforced by real physical threats, requires a rare degree of individual strength indeed. This plea for resistance at the individual 35An example of this may make it concrete. Stanley Milgram made a film of his experiments (Milgram, 1965), in which he included several sequences of shock-giving individuals who “go all the way.” That is, a person in the role of teacher was to give an ascending series of painful shocks to the learner if the learner made mistakes on an associative learning task. Several individuals shown in the film, under what appeared to be mild prodding from the experimenter, escalated to the maximum levels of shock, even in the face of the protests of the individual receiving the shocks. When this film is shown to introductory students, they invariably attribute a sadistic person- ality to the “teacher.” The experienced lecturer sometimes allows this per- ception to be created, then dismantles it by showing that the degree of situational control of this behavior was so high as to preclude this explanatoa ry possibility. V 36One needs to put this cautiously. SeVeral social psychology studies including Zimbardo’s (1969) on deindividuation, Latané’s and my own on responding to emergencies, and Milgram’s, have demonstrated a high de- gree of situational control over actions usually thought to be largely under dispositional control. Certainly, the Milgram (1974) findings of a high per— centage of subjects behaving obediently defeats the attribution of sadistic personalities unless We want to make the rather improbable claim that a majority of New Haven dwellers are sadistic. However, as Funder and Ozer (1991) pointed out, there is still plenty of variance in those studies that potentially could be explained by individual—difierence variables. True. Still, it may not be useful to conceive of these individual differences as trait— described dispositions acquired during childhood socialization. 218 DARLEY level to the contrary, the real action in evil—prevention may lie elsewhere. Where does it lie? What I suggest is that the prototypes we carry around about the sources of evil actions, which assign those actions to individuals who are themselves in some way intrinsically evil, causes us to ignore the more likely source of harmdoing actions, which is organizational in nature. If harmdoing actions are in the main committed by individuals caught up in organizations and their pathology, then preven- tion or amelioration of evil may be best done at the organiza— tional level. How could we do this? We have a standard set of interventions designed to prevent the development of pathology in military and social—control organizations. What can we learn from these that could be applied in the context of conventional organizations? Given my comments about the somewhat different ways conventional organizations slip into wrongdoing, ways of preventing evildoing in those orga— nizations may need to be tailored to their special charac- teristics. How, for instance, could we halt or limit the tenden— cy of organizations to cover up past harms , and in the process inadvertantly commit themselves to perpetrate future ones? Other writers who have contributed to what I have called the social psychological perspective on evil clearly direct us to these questions. One last word about the banality of evil: One way of wording the insight that arises from considerations such as have been examined here, is that it is generally only possible for a person to do evil when that evil has been “banalized”—— rendered routine and morally neutral. To analyze these pro- cesses to better understand them, we give phenomenological accounts of how ordinary people neutralize evil as they are caught up by forces urging them to commit it. By doing this, do we not banalize evil at second hand, as we render it understandable, and make its commission easier?37 I began with other discomforts; I end with that one. Notes I am grateful to Dan Batson, Kai Erikson, Dan Gilbert, Genevieve Pere, and Nelson Polsby for their comments on this essay and for, more generally, their perceptions on the question of evil. As I am acquainted with several of the authors of the books under review, I should also make that clear to the reader. Kelman is a much—valued graduate school professor of mine; Hamilton, a later student of his and there— 37As the philosopher Lichtenberg remarked in another context, “warm, slightly burned, thanks” to Dan Robinson for forcefully reminding me of this point during a talk I gave onprocesses socializing individuals into evildoing. fore a “younger sib” of mine, is a friendly acquaintance; Milgram was a generous, slightly older colleague and a friend; Staub is a friendly acquaintance within the small circles of experimental social psychology; and Lifton, the only one of the authors I have not met, had commanded my admiration for a number of years. John M. Darley, Department of Psychology, Princeton University, Princeton, NJ 08544. References Allport, G. W. (1937). Personality: A psychological interpretation. New York: Holt. Arendt, H. (1963). Eichmann in Jerusalem-A report on the banality of evil (rev. & enlrg. ed.). New York: Viking. Austin, J. L. (1956). A plea for excuses. Proceedings of the Aristotelian Society, 57, 1—30. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice—Hall. Brock, T., & Buss, A. (1962). Dissonance, aggression, and the evaluation of pain. Journal of Abnormal and Social Psychology, 65, 197—202. Cornwell, P. (1990). Postmortem. New York: Macmillan. Darley, J. M., & Shultz, T. R. (1990). Moral judgments: T heir content and acquisition. Annual Review of Psychology, 41, 525—556. Darley, J. M., & Zanna, M. P. (1982). Making moral judgments. Ameri— can Scientist, 70, 512—521. Erikson, K. (1989). Big Daddy Lipscomb: A parable. Unpublished manu— script, Yale University, New Haven, CT. Funder, D. C., & Ozer, D. J. (1991). Behavior as a function of the situa- tion. Journal of Personality and Social Psychology, 44, 1074112. Gibson, J. T., & Haritos-Fatouros, M. (1986). The education of a torturer. Psychology Today, 20, 50—58. Goffman, E. (1961). Asylums: Essays on the social situation of mental patients. Garden City, NY: Doubleday/Anchor. Harris, T. (1981). Red dragon. New York: Dell. Harris, T. (1989). The silence of the lambs. New York: St. Martin’s. Heider, F. (1958). The psychology of interpersonal relations. New York: Wiley. . . Karski, J. (1944). Story of a secret state. Boston: Houghton Mifflin. Latané, B., & Darley, J. M. (1970). The unresponsive bystander: Why doesn’t he help? New York: Appleton—Century. Lewis, M. (1989). Liar’s poker. New York: Norton. Milgram, S. (1965). Obedience [A filmed experiment]. New York: New York University Film Library. Milgram, S. (1974). Obedience to authority: An experimental view. New York: Harper & Row. Rosenhan, D. (1973). On being sane in insane places. Science, 179, 250— 258. Ross, L. (1977). The intuitive psychologist and his shortcomings: Distor— tions in the attribution process. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 10, pp. 174—221). New York: Academic. Webster’s New Third International Dictionary. (1963). Springfield, MA: Merriam. Zimbardo, P. G. (1969). The human choice: Individuation, reason, and order versus deindividuation, impulse, and chaos. In W. J. Arnold & D. Levine (Eds), Nebraska Symposium on Motivation. Lincoln: Univer— sity of Nebraska Press. Copyright© 2002 EBSCO Publishing ...
View Full Document

This note was uploaded on 03/14/2011 for the course ENG 102 taught by Professor Emelbekişoğlu during the Fall '11 term at Bilkent University.

Page1 / 21

7400655 - Psychological Inquiry 1992, Vol. 3, No. 2,...

This preview shows document pages 1 - 21. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online