Reflect upon the Merits and Ethical Dilemmas of Three Classic Psychology Experiments

British Journal of Social Psychology (2015), 54, 55–83 ©2014 The British Psychological Society www.wileyonlinelibrary.com ‘Happy to have been of service’: The Yale archive as a window into the engaged followership of participants in Milgram’s ‘obedience’ experiments S. Alexander Haslam 1*, Stephen D. Reicher 2, Kathryn Millard 3and Rachel McDonald 4 1School of Psychology, University of Queensland, Brisbane, Queensland, Australia2School of Psychology, University of St. Andrews, UK3Department of Media, Music, Communication and Cultural Studies, Macquarie University, Sydney, New South Wales, Australia 4Department of Psychology, University of Kansas, Lawrence, Kansas, USA This study examines the reactions of participants in Milgram’s ‘Obedience to Authority’ studies to reorient both theoretical and ethical debate. Previous discussion of these reactions has focused on whether or not participants were distressed. We provide evidence that the most salient feature of participants’ responses–and the feature most needing explanation–is not their lack of distress but their happiness at having participated. Drawing on material in Box 44 of Yale’s Milgram archive we argue that this was a product of the experimenter’s ability to convince participants that they were contributing to a progressive enterprise. Such evidence accords with an engaged followership model in which (1) willingness to perform unpleasant tasks is contingent upon identi cation with collective goals and (2) leaders cultivate identi cation with those goals by making them seem virtuous rather than vicious and thereby ameliorating the stress that achieving them entails. This analysis is inconsistent with Milgram’s own agentic state model. Moreover, it suggests that the major ethical problem with his studies lies less in the stress that they generated for participants than in the ideologies that were promoted to ameliorate stress and justify harming others. As we write, it is exactly 50 years since Stanley Milgram’s studies on obedience to authority were rst published. The details of these are so well known that they only need the briefest sketch here (for details, see Milgram, 1963, 1965, 1974). Motivated by a desire to understand the psychological processes that might allow ordinary citizens to perpetrate atrocities of the form witnessed in the Holocaust, his research sought to investigate when and why people obey the destructive instructions of those in authority.

The paradigm that he devised for this purpose was carefully honed (Russell, 2011) and centred on an elaborate setup in which participants (mostly men) were led to believe that they were acting as ‘Teachers’ in a scienti c experiment devised to study the impact of punishment on ‘Learners’ memory performance. The logic of this meant that Teachers came to the Yale Psychology Department to take part in a study in which they needed to *Correspondence should be addressed to S. Alexander Haslam, School of Psychology, University of Queensland, Brisbane, Qld 4072, Australia (email: [email protected]).

DOI:10.1111/bjso.12074 55 administer shocks of increasing magnitude to a (male) Learner every time he made an error in a word recall task. They did this by depressing 1 of 30 switches arranged in ascending order on an imposing shock generator–starting with a mild 15 V and progressing to a deadly 450 V. Unbeknown to participants, the shocks were not real and the Learner was a confederate of Milgram’s, as was the lab-coated Experimenter who oversaw the entire procedure. Furthermore, Milgram had no interest in memory; instead he wanted to know how far the participants would go in carrying out the Experimenters’ lethal instructions.

The answer to this question not only shocked Milgram, but went on to shock the successive generations of students and readers who have been exposed to his studies’ ndings in both psychology textbooks and popular commentaries (Blass, 2004). For rather than abandon their task once it posed a threat to the Learner’s health and well- being, a great many participants proved willing to continue administering shocks right up to the 450 V mark. Indeed, in an early pilot study, where the Learner never voiced any objections, all participants went this far; and in the variant that has become known as the ‘baseline’ condition (Russell, 2011), all participants went to 300 V and 65% continued to the very end. It is true that in a range of other variants far fewer participants proved willing to go so far (Milgram, 1974; Reicher, Haslam, & Smith, 2012) and that Milgram himself made much of the dif culty participants experienced in progressing–noting the high levels of con ict and nervous tension that the paradigm produced. Nevertheless, the studies are routinely understood as demonstrating people’s alarming propensity to ‘blindly’ comply with the orders of those in authority (e.g., see Zimbardo & Gerrig, 1999)– even when those orders violate deeply ingrained notions of morality and civility.

Conventional ethical analysis of the Milgram studies When Milgram’s ndings were rst published they sparked one of the most famous exchanges in the history of our discipline. This exchange was about ethics. In the rst volley, Baumrind accused Milgram of manipulating, embarrassing, and discomforting his participants and of showing a singular detachment from their plight (Baumrind, 1964, p.

422). These claims were based on Milgram’s own descriptions of his participants’ experiences. For instance, in his initial report on the studies Milgram had claimed:

‘I observed a mature and initially poised businessman enter the laboratory smiling and con dent. Within 20 min he was reduced to a twitching, stuttering wreck, who was rapidly approaching a point of nervous collapse’ (Milgram, 1963, p. 377, cited in Baumrind, 1964, p. 422). On the basis of such evidence, Baumrind dismissed Milgram’s suggestion that these tensions dissipated before participants left the laboratory as ‘casual’ and ‘unconvincing’ (Baumrind, 1964, p. 422; for a recent and trenchant critique along similar lines, see also Nicholson, 2011).

Milgram retorted by accusing Baumrind of being ‘de cient in information that could have been obtained easily’ (Milgram, 1964, p. 848). To back this up, he reported follow-up data showing that 84% of respondents to a post-experimental survey were ‘glad’ or ‘very glad’ to have taken part in the experiment, whereas only 1% were ‘sorry’ or ‘very sorry’ (p. 849). He further cited evidence from a psychiatrist who had examined 40 participants and concluded that ‘none was found by this interviewer to show signs of having been harmed by his experience’ (p. 850). Milgram observed: ‘such evidence ought to be weighed before judging the experiment’ (p. 850).

This exchange–or at least the way it has been remembered–set the terms for a controversy that has hardly diminished over the ensuing half century. On the one hand, 56S. Alexander Haslam et al. Milgram’s and Baumrind’s positions have continued to be taken as polar opposites. But it is perfectly plausible to take them both seriously, and to accept that people were both extremely distressed during the study but subsequently genuinely glad to have participated. That raises the intriguing question of how exactly this dramatic turnaround was brought about.

This, however, is a question that most commentaries overlook. Indeed, for most readers of Milgram’s work, the question ofwhyparticipants were (or were not) distressed has been overshadowed by the question onwhetherthey were distressed. As a consequence, the ethical controversy surrounding the Milgram studies has become separated from the theoretical debate about why his participants behaved as they did (Blass, 2004). For, treated as a phenomenon, distress can be viewed as something very different to obedience. However, when it comes to how they are produced, it is reasonable to conjecture that distress and obedience may have much in common. That is, the processes which govern whether people felt it was acceptable to deliver shocks in the rst place may well give us some insight into the processes that reconciled them to their actions subsequently (and vice versa). Indeed, this was what Milgram himself proposed when he remarked to Baumrind: ‘The same mechanisms that allow the subject to perform the act, to obey rather than to defy the experimenter, transcend the moment of performance and continue to justify his behavior for him’ (Milgram, 1964, pp. 849–850).

Milgram was never explicit as to what precisely these mechanisms might be and he never provided any evidence that allows us to understand how participants justi ed their behaviour or why they supported his research. However, he does root his claim in a more general point about the behaviour of his participants, namely, that they are active adults and that they exercise choice. Indeed, one of his arguments in defence of the studies is that the experience may help people exercise that choice more effectively when confronted by authority in the future. ‘If there is a moral to be learned from the obedience study’, Milgram argues, ‘it is that every man must be responsible for his own actions’ (p. 852).

In his experimental notes, located in Box 46 of the Yale archive, Milgram takes the point still further, to the extent that he questions the very terminology of his studies.

Speci cally, he speculates as to whether the act of shocking constitutes cooperation with authority rather than obedience to authority. He asks, rhetorically, what is the difference between the two, and he answers: Cooperation implies a certain willingness to perform the action or help out, a certain internal desire to assist, while obedience implies an action that is totally in response to a command, with no motivational support from inner sources. (Milgram, Box 46, Yale archive) This observation should lead us to use and interpret the word ‘obedience’, now so indelibly associated with this work, with a large degree of caution.

Over time, though, Milgram himself abandoned such caution, as he came to describe the behaviour of his participants as unquali ed obedience, and as a matter of submission rather than of choice (Blass, 2004). Moreover, he never returned to the relationship between what people did in his studies and how they felt about that research. In this study we wish to revisit this un nished business. Our interest lies precisely in why people who had experienced such acute tensions came to feel so happy about their participation and so good about the research itself. Here, we take seriously Milgram’s comments about the relationship between mechanisms that lead people to perform an act and which lead them to justify their acts. That is, our ambition is to use material that ostensibly relates to the ethics of the study to shed light on the psychology of ‘obedience’. Reconceptualizing the Milgram studies57 To do this, it is necessary to go beyond the raw data published by Milgram in his 1964 study and elsewhere (notably the appendix to his 1974 bookObedience to Authority). We need to access the understandings which lie behind the numbers. Fortunately, such information is available. For, as already indicated, after his studies were completed, Milgram sent a post-experimental debrie ng report to his participants along with a questionnaire that included both closed questions and an invitation to provide elaborated responses. This was an invitation that a majority of his participants accepted, and their comments were subsequently transcribed onto cards. Today, these cards can be found in Box 44 within the Milgram archive at Yale. Before we get to these, however, it is necessary to consider Milgram’s own theoretical explanation of his results, and to consider alternative accounts that have been proposed more recently. For these will provide a conceptual framework within which to examine participants’ responses. Conventional theoretical analysis of the Milgram studies: The inattentive bureaucrat We have already noted that Milgram’s own explanation of his ndings changed markedly over time. In particular, his initial emphasis on the active choices of his participants gave way to his better-known claim that obedience results from the relinquishing of agency. For instance, in his rst published account of the research, Milgram (1963) lists 13 considerations which bear on participants’ behaviour. These have to do with the nature of the participants’ obligation to both the ‘Experimenter’ and the ‘Learner’, and the dilemma of choosing between them. The second consideration is particularly interesting in this regard. For Milgram notes that the experiment is ostensibly designed with a worthy purpose, and that ‘obedience occurs not as an end in itself, but as an instrumental element in a situation that the subject construes as signi cant and meaningful’ (1963, p. 377). This point is spelled out in further detail in Milgram’s experimental notes, located in Box 46 of the Yale archive. There he observes: Even in this experiment we must disguise the character of obedience so that it appears to serve a productive end. Therefore we are not dealing with ‘blind obedience’.... For every command is justi ed as serving some productive end. The important thing is, in exacting obedience, that end must appear rationally correlated with the role of the authority. (Milgram, Box 46, Yale archive) Traces of these explanations endure and many are reproduced in Milgram’s, (1974) book.

But by and large he dropped the idea that people are concerned with their obligations to the Experimenter and the Learner, as well as the idea that behaviour depends upon active construals of context. Instead, he came to treat obedience as an end in itself. More speci cally, he argued that participants obey because they disengage from the task and focus simply on doing the bidding of the Experimenter as well as they possibly can (Milgram, 1974; for a discussion see Blass, 1999).

In this, Milgram’s thinking was heavily in uenced by exposure to the writings of Hannah Arendt in which she theorized about the processes that had allowed Adolf Eichmann to play his part in the Nazi Holocaust 1. Eichmann was the bureaucrat chie y 1Indeed, it is interesting to observe that Milgram’s extensive experimental notes (contained in Box 46 of the Yale archive) make no reference to the Holocaust, suggesting that these were connections that became more salient only once he had been exposed to Arendt’s writing.

58S. Alexander Haslam et al. responsible for devising and enacting ‘The nal solution to the Jewish problem’ which saw millions of Jews (and other targets of Nazi hatred) deported to death camps across German-occupied territories as part of an elaborate but highly ef cient policy of extermination (Cesarani, 2004). Arendt’s views about Eichmann were shaped largely through her attendance at his trial in Jerusalem in August 1961 (the same month that Milgram’s own research programme commenced). There she had expected to encounter a monster, but was instead taken aback by the sheer ordinariness of the man she encountered in the dock–an ordinariness famously captured in the subtitle of her 1961 book on the subject:The Banality of Evil(Arendt, 1963).

Although it is much more rich and nuanced than commonly supposed (Jetten & Mols, 2014; Newman, 2001), Arendt’s portrayal of Eichmann as a man who was focused narrowly on the bureaucratic challenges that he faced without concern for the broader implications of his actions proved to be enormously in uential (Miller, 1995, 2004). Most particularly, it provided a sharp focus for Milgram’s evolving understanding of his participants’ behaviour, such that in the Introduction to his 1974 book he observed: After witnessing hundreds of ordinary people submit to the authority in our own experiments, I must conclude that Arendt’s conception of the banality of evil comes closer to the truth than one might dare imagine. The ordinary person who shocked the victim did so out of a sense of obligation–a conception of his duties as a subject–and not from any peculiarly aggressive tendencies. (Milgram, 1974, p. 6) More formally, Milgram conceptualized the process through which participants became immersed in their task as one that involved entry into an ‘agentic state’ (1974, pp. 132– 134). The key idea here is that, in the presence of a powerful authority, individuals come to focus their attention on the challenge of enacting the authority’s wishes rather than on the question of whether those actions are right or wrong. Thus, just as Eichmann was seen by Arendt ‘to never realize what he was doing’ (1974, p. 287), so Milgram’s participants were seen to have failed to grasp the lethal signi cance of the punishments they were required to mete out in furtherance of the Experimenter’s goals. They abstained from active and re ective thought. They succumbed to the power of the situation. They became passive and uncritical agents of authority.

Over the ve decades since his publication on the topic, Milgram’s agentic state account has remained the dominant framework for understanding his ndings (Blass, 1999, 2004). Moreover, partly through its close links to the banality of evil thesis (see Haslam & Reicher, 2007, 2012b), it has proved to be a key analytic resource not only for social psychologists (e.g., Zimbardo, 2007) but also for researchers in a range of other academic disciplines (e.g., Akerlof, 1991; Helm & Morelli, 1979; Overy, 2014), as well as for the public more broadly (Novick, 2000). The extent of this in uence is surprising for a number of reasons. Principal among these is the fact that there is a general consensus among those who have looked closely at Milgram’s research that the agentic state account fails to provide a convincing explanation of his ndings. In particular, this is because it does not explain why levels of obedience differed across the many variants of the paradigm, why participants were clearly tormented by the tasks they had to perform, or why (to varying degrees) they were in uenced by the protestations of the Learner (Blass, 2004; Mantell & Panzarella, 1976; Reicheret al., 2012; Rochat & Modigliani, 1997).

A second key problem is that Milgram himself presents relatively little data to support the analysis. The data he does present are taken from participants’ own accounts of their experiences, but it is unclear how representative these are of those accounts or what Reconceptualizing the Milgram studies59 precise status they have. For example, he notes that, after the experiment, several of his participants expressed a belief that Yale was ultimately responsible for participant welfare (Milgram, 1974), but it remains unclear whether such views were widespread or how they accord with other features of participants’ accounts. It is, for example, perfectly plausible to invoke the accountability of the institution without counter-posing it to one’s own. To get a clearer picture of participants’ re ections on their experiences, these therefore need to be examined in the round rather than by looking at selected statements in isolation.

An alternative theoretical analysis of the Milgram studies: The engaged follower For all its shortcomings, one key factor that has contributed to the durability of the agentic state model is the lack of an alternative theory that might provide a more plausible explanation of Milgram’s ndings. Nevertheless, in recent years researchers have begun to challenge the claim that destructive acts are the product of any inherent inclination to conform passively to the wishes of those in authority. Thus, among other things, it has been suggested that the behaviour Milgram observed was a product of the structure of his paradigm (Darley, 1992)–speci cally, the way that it created a sense of obligation to, and identi cation with, the Experimenter (Gilbert, 1981; Rochat & Modigliani, 1997), and also relied upon, and communicated, particular behavioural norms (Navarick, 2009; Nissani, 1990). In other contexts, researchers have also observed that it is those individuals who identify with and glorify the in-group who are most likely to countenance cruel treatment of others (Castano, 2008; Reicher, Haslam, & Rath, 2008; Roccas, Klar, & Liviatan, 2006).

Such research has also contributed to the development of a radically different model of the processes at work in Milgram’s studies. This asserts that rather than representing passive ‘obedience’ (Lutsky, 1995), behaviour in the paradigm is better understood as a form ofengaged followership(Haslam & Reicher, 2012a,b; Reicher & Haslam, 2011; Reicheret al., 2012). This alternative analysis is derived from a social identity perspective on social in uence and leadership which asserts that people’s willingness to accede to the requests of others is predicated upon social identi cation with them, and an associated sense that they are legitimate representatives of shared group goals, values, and aspirations (Haslam, Reicher, & Platow, 2011; Turner, 1991; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987; Tyler, 1990, 1998). In these terms, Milgram’s participants did not administer shocks because they were oblivious as to their purpose, but rather did so because they were identi ed with the scienti c goals that underpinned the experiment and that the Experimenter appeared to embody. In other words, and in line with Milgram’s own assertions in his 1963 study, those who administered shocks were aware of the consequences of their actions. Indeed, they continued shocking precisely because these consequences were ones they supported and identi ed with, and because their actions were construed to be contributing to a moral, worthy, and progressive cause.

In this regard, it is interesting to refer again to Milgram’s experimental notes and observe that his early deliberations focused extensively on the importance of identi ca- tion with the experimenter as a factor responsible for his ndings. In particular, he observed: The subjects have come to the laboratory to form a relationship with the experimenter, a speci cally submissive relationship in the interest of advancing science. They have not come to form a relationship with the subject, and it is this lack of relationship in the one direction and the real relationship in the other that produces the results.... Only a genuine relationship 60S. Alexander Haslam et al. between the Victim and the Subject, based on identi cation, or marriage, etc. could reverse the results. (Milgram, Box 46, Yale archive) Moreover, in the context of this observation, he also goes on to outline plans to explore this hypothesis experimentally: How could this be fostered experimentally?

Perhaps, the Negro–White variations.

It would be possible to vary the characteristics of experimenter, subjects and victim in such a way as to study the importance of identi cation in the obedience act. (Milgram, Box 46, Yale archive) Milgram’s notes then specify a series of theoretically interesting variants of an experimental design in which the identity of the ‘Subject’, ‘Experimenter’, and ‘Victim’ (as ‘Negro’ or ‘White’) is varied systematically.

Evidence that supports a reconceptualization in which such identi cation assumes a central role comes from two sources. First, recent work by historians has served to question suggestions that the atrocities orchestrated by Eichmann and his ilk resulted from an inability to recognize the signi cance of their actions or from blind obedience (Cesarani, 2004; Lozowick, 2002; Vetlesen, 2005). This was certainly an argument used after the war, but critics have argued that it is better viewed as an excuse than as a plausible account of the relevant facts (Mandel, 1998). Indeed, evidence that has recently emerged from secret tape-recordings of senior German of cers in British Prisoner of War camps suggests that they actively conspired to ensure that everyone would con rm this stance so as to lend it credibility (Neitzel, 2007).

By contrast, the wartime behaviour of Nazis such as Eichmann stood out by virtue of the energy and enthusiasm with which it was performed (Cesarani, 2004; Sofsky, 1993).

Moreover, rather than slavishly following orders or resembling dull automatons, it is clear that Nazi bureaucrats brought considerable creativity to bear on their work, and were proud of their ability to surmount the many challenges they confronted (Haslam & Reicher, 2007). A key point here too is that in most instances there were no explicit orders for bureaucrats to follow; instead, they had to ‘work towards’ their leaders based on an understanding of, and sympathy for, those leaders’ goals (Kershaw, 1993; Reicheret al., 2012).

Alongside this historical reanalysis, a second body of evidence that supports the engaged followership model has resulted from a recent renewal of interest in Milgram’s work among social psychologists. Here, there are at least four aspects of this work that support this model. First, several researchers have remarked upon the lengths that Milgram went to to ensure that participants construed his project as a worthy scienti c enterprise and hence became bound to its ostensible goals. This was something that Milgram worked at meticulously and which shaped the careful operationalization of his paradigm. Examples include the scrupulous attention paid to the design of the shock machine and to the process of recruiting and greeting participants (Russell, 2010, 2011; Russell & Gregory, 2011) as well as major departures from the experimental script that positioned Teachers as in-group collaborators rather than detached actors (Gibson, 2011).

Second, a meta-analysis by Packer (2008) has shown that the two key points where participants break off from administering shocks are 150 and 315 V. The signi cance of these is that they are the points at which the Learner rst asks to be released and then voices his most categorical objections, and hence they are the junctures at which Reconceptualizing the Milgram studies61 identi cation with the Experimenter is challenged by, and needs to be reconciled with, an alternative source of identi cation.

Third, a limited replication of Milgram’s baseline study by Burger (2009; Burger, Girgis, & Manning, 2011) has shown that when they are given a prod which orders them to continue (speci cally, being told ‘you have no other choice you must continue’), participants are strongly inclined todisobey the instruction. This observation has recently been con rmed in an experimental analogue of Milgram’s paradigm which disentangles issues of causation (Haslam, Reicher, & Birney, 2014). This suggests that orders provoke reactance not obedience (Brehm, 1966) and one potential reason for this is that they undermine participants’ sense that the authority is acting in terms of an identity that they share (Haslamet al., 2011).

Fourth and nally, recent studies by Reicheret al.(2012) have shown that across the many variants of Milgram’s paradigm reported inObedience to Authority, the willingness of participants to administer the maximum level of shock can be predicted with a high degree of accuracy by observers’ estimates of the degree to which a given variant encourages participants to identify with the Experimenter and the science that he represents rather than with the Learner and the general community that he represents. For example, identi cation and obedience both drop dramatically when two Experimenters contradict each other, when the Experimenter is absent from the laboratory, or when the Experimenter’s role is enacted by someone who appears to be another na€ ve participant.

The importance of this alternative analysis, then, is that it is the rst account which systematically explains the variation in obedience across the various Milgram studies. That is, it explains why people do not obey as well as why they do (Jetten & Hornsey, 2011; Jetten & Mols, 2014).

The present research: Two key questions about the experience of participants Despite the accumulation of evidence that has challenged the agentic state account, it remains the case that this was the model that Milgram himself derived on the basis of participants’ accounts of their behaviour, and this proximity to the data has carried considerable weight (Blass, 2004). Yet, as noted above, Milgram’s analysis of this data was far from systematic, and, at the time, he had no elaborated alternative hypothesis against which to assay the validity of his interpretation.

For this reason there would appear to be considerable value in a more structured analysis of the experiences reported by Milgram’s participants’ which seeks to establish the relative merits of agentic state and engaged followership explanations. It is here that the Milgram archive at Yale proves invaluable. Indeed, increased scrutiny of the archive in recent years has been one of the key factors in a resurgence of interest in Milgram’s work (see Blass, 2004; Gibson, 2011, 2014; Millard, 2011, 2014; Perry, 2011; Russell, 2010, 2011).

Our particular interest in this resource relates to material that addresses how people felt about both the science of the study and about their own role in contributing to this. To be more speci c, through a systematic analysis of this evidence we seek to address two research questions: RQ1. Was participants’ general orientation to the science of the study one of disengagement or engagement? 62S. Alexander Haslam et al. RQ2. Did the lack of stress reported by Milgram’s participants (as discussed by Milgram, 1964) re ect a state of disengagement or a state of engagement? In regard to these questions, the key point to note is that Milgram’s (1974) agentic state explanation and our engaged followership explanation predict different answers. On the one hand, an agentic state explanation implies (1) that participants were generally disengaged from the task of administering shocks and (2) that this disengagement protected them from the stressful implications of their actions. On the other hand, an engaged followership explanation implies (1) that participants were generally engaged with the task of administering shocks and (2) that it was engagement with this project that protected participants from stress because it allowed them to construe their actions as virtuous rather than vicious (Haslam & Reicher, 2012b).

To determine which approach better accounts for the evidence, we will rst consider responses to Milgram’s post-experimental questionnaire, some of which we have already referred to in describing his response to Baumrind (1964). We will then explore these questions more closely, drawing upon both qualitative and a quantitative analysis of the elaborated comments that are recorded on the cards in Box 44.

Quantitative examination of responses to Milgram’s post-experimental questionnaire In his 1964 study, Milgram describes the steps he took to reassure his participants. In addition to an immediate debrief (or ‘dehoax’ as he terms it), this included a detailed ve- page report that was sent out when the experimental series was completed. In this Milgram explained the nature of the deception in his studies and why it was necessary. He then described some of the key ndings. Signi cantly, where participants had previously signed up to take part in a study examining the effect of punishment on learning that was important because ‘almost no truly scienti c studies have been made of it in human beings’ (Milgram, 1963, p. 373), the report used very similar language to inform them that ‘The problem of obedience to authority may well be the crucial issue of our time. The experiments you took part in represent the rst efforts to understand the phenomenon in an objective, scienti c manner’ (Milgram, no date b, p. 5).

Participants (‘by this time numbering 800’; Milgram, no date a, p. 1) were able to react to the report through a follow-up questionnaire. A rst point to note–which itself is suggestive of high levels of engagement with Milgram’s project–is that completed questionnaires were received back from at least 659 participants (this being the maximum number of participants who responded to any single question), a response rate of approximately 82%. This is far higher than the response rate that one would expect under the circumstances. Indeed, Fox, Crask, and Kim’s (1988) meta-analysis would lead one to expect that with pre-noti cation, university sponsorship, return postage paid, and a postcard reminder, only 32.7% would reply.

The questionnaire consisted of 10 items that invited participants to re ect on their experiences before, during, and after the study and to respond to statements on 3- or 5- point scales. Milgram (1964) himself presents only percentage data pertaining to Question 8: concerning whether people were glad or sorry to have participated. As noted above, most said that they were either very glad (44%) or glad (40%) to have done so. Two other questions also relate to the issue of how people felt about their participation. Here, 64% of participants indicated that, after it was over, the experiment had not bothered them at all (q6) and most agreed (43%) or de nitely agreed (31%) that they had learned something Reconceptualizing the Milgram studies63 important through their participation (q10). These gures are lower, but they still indicate that a clear majority of participants had reasons to feel positive about having taken part in the study and lacked reasons to feel negative. Accordingly,t-tests con rm that on all three measures, responses were signi cantly above the scale midpoint–q6 not bothered(1–3) :

M=2.57,t(653)=23.27,p<.001; q8 glad to participate(1–5) :M=4.26,t(655)=41.72, p<.001; q10 learned(1–5) :M=3.94,t(642)=25.11,p<.001.

In addition, the questionnaire contained two items relating to the institution to which the Experimenter was af liated and to the science of the experiment itself. Here, most respondents indicated that participating had improved (10%) or not changed (89%) their views about Yale (q7) and agreed (39%) or de nitely agreed (41%) that further studies of this form should be carried out (q9). Once again we see a paucity of negative feelings and a preponderance of positive feelings, such that responses were signi cantly above the scale midpoint–q7 unchanged views of Yale(1–3) :M=2.12, t(656)=2.65,p=.008; q9 support more studies(1–5) :M=4.18,t(651)=35.52,p<.001.

On the whole, then, participants were as positive about the studies as they were about their own participation within them. Unfortunately, however, the way that Milgram recorded his data makes it impossible to look at the relationships between these various responses. That is, he analysed answers to different questions separately and the Yale archive provides no information that would make it possible to relate a person’s response on any one question to their response on any other. This again re ects the fact that Milgram was more concerned to clarifythatparticipants were not harmed than to explore whythis was the case. We cannot, therefore, establish whether higher levels of support for the science are associated with more positive feelings about having participated.

However, we can at least note that people are generally very engaged with the science.

This speaks to RQ1 and is clearly consonant with the engaged followership approach. To address RQ2 we need to look at the more detailed comments of participants that are archived in Box 44. Qualitative examination of engagement and positivity in post- experimental responses After participants had responded to the 10 questionnaire items that we have just been considering, they were told: ‘Any additional comments you would care to make would be most helpful; space is provided on the accompanying sheet of paper. Thank you very much for your cooperation’ (Milgram, no date a, p. 3). Of the 659 participants who returned the questionnaire, 431 provided some form of comment (i.e., 65% of respondents, approximately 54% of all participants).

The comments were typed out onto 1,057 6 inch94 inch cards which, together, constitute Box 44 of the Milgram archive in the Sterling Library at Yale. A sample card is presented in Figure 1. On each card the author of the comments was identi ed by a four- digit code number in the top right-hand corner. The rst two digits indicate the experimental variant in which the participant took part (29 different variants are identi ed; 11 more than in Milgram, 1974); the second two digits indicate the participant number. Rather than being organized by participant or by study, cards were sorted into 1 of 17 numbered sections, with the section number marked in the top left-hand corner of the card (noting that there is no Section 15 and that the rst section is labelled ‘T.R.’).

There is no index that provides an explicit rationale for the sections. However, it appears that cards are organized around different themes. To retrospectively identify 64S. Alexander Haslam et al. these themes, the rst and fourth author read the contents of all the cards and made independent notes about the similarities within each category as well as the differences between them (thereby clarifying the content of themes following the principle of meta- contrast; Haslam & McGarty, 2014; after Turner, 1985). Discussion between coders then sought to resolve differences in interpretation (which were minimal). The resultant themes are presented in Table 1.

From this table it can be seen that there are a number of sections of Box 44 that contain material pertinent to RQ1 and RQ2. Nevertheless, the section that is most directly relevant to the present analysis is Section 13–whose content was parsed as ‘Thoughts about the value (or otherwise) of (having participated in) the research’–as this contains comments that relate directly to participants’ sense of whether or not the research process as a whole was worthwhile and valuable. We therefore focus our analysis on the cards in this section. 2Narrowing our focus in this way also reduces the risk of oversampling that would otherwise arise from the fact that comments in other sections often relate very closely to the comments found in Section 13.

As noted in Table 1, Section 13 contains 140 cards (131 containing comments and 9 cross-referencing comments in other sections). As the illustrative examples in Table 2 suggest, the majority of these comments were extremely positive. To examine the structure of this positivity more closely, the rst, second, and fourth authors read the cards in Section 13 closely with a view to identifying key themes in participants’ responses. As before, this activity was guided by the principle of meta-contrast (Haslam & McGarty, 2014; see also Braun & Clarke, 2006) in such a way as to maximize (1) similarity among the data that are considered illustrative of the same theme and (2) difference between data that are considered illustrative of different themes. As Haslam and McGarty (2014) argue, where either of these principles cannot be satis ed this suggests either that one needs to create additional themes (because intra-theme similarity is too low) or that some themes need to be merged (because inter-theme difference is too low).

On the basis of this process, and following discussion, our coders agreed that three distinct themes could be identi ed (while noting that the inter-thematic component of Figure 1.A sample card from Box 44.

2Box 44 contains more than one card per respondent and so the cards in Section 13 contain only those comments that Milgram (or his research assistants) saw as tting this section’s theme. Having looked across the material in all sections, it would appear that these cards capture most (if not all) of the comments that pertain to this theme. Nevertheless, it is worth emphasizing the point that the cards that are the focus of our analysis contain feedback from only some of Milgram’s participants and only some of the feedback from those who responded to his questionnaire. Reconceptualizing the Milgram studies65 Table 1.Contents of Box 44: Sections, content, and representative examples SectionNcards (blank a) Theme Summary of content Illustrative examples b T.R.

c 14 (6) Troubled responses Ps who had particularly strong and lengthy reactions to the study.

All are particularly engaged with its meaning and implicationsAfter leaving, while driving home, I came to the conclusion that, for purposeful reasons, I had been hoaxed. Interesting concern enveloped me as to why. Later at the Yale Med School I found out why. Partial satisfaction. With your report, almost complete satisfaction.|...Brie y, I was happy to have been of service....

Continue your experiments by all means as long as good can come of them. In this crazy mixed up world of ours, every bit of goodness is needed. [2205] 1 41 (2) Reactions to receiving reportMost Ps express gratitude for receipt of report. Many express reliefIt was a very interesting experiment and I should like to thank you for sending me the report. [0223] I think this report should have been sent out shortly after the experiment as I would have had more peace of mind. After the test I was very surprised that Yale would have such a cruel experiment. My feelings have now changed [0614] 2 19 (0) Reactions to study’s main ndingsVaried reactions. Some Ps express surprise, some do notThe thing that amazes me is the fact that 60% obeyed to the end.

I had sincerely thought that about 2% would complete the shock waves. This is really a shock. [0616] 3 31 (0) Thoughts about administering shocksVaried reactions. Several Ps want to know what the study says about them as individuals.

Several state their belief that Yale was responsibleDuring the administration of the shocks I kept thinking that if I later had to change places with the learner, I was going to give back the $4.50 and leave. [0517] 4 23 (1) Thoughts about experimental designMost Ps comment on the ingenuity of the study. Many wish Milgram well with his researchI am very delighted to be a part of this project. I have often thought I was the subject but I could not be any happier. I’ve been waiting very anxiously for this report to really put my mind at rest and curiosity satis ed. Believe it or not Dr. this was a well planned project.... We wish you continue success in the future.

Last but not least I sure hope my efforts, and cooperation have been somewhat useful for your project. [1817] Continued 66S. Alexander Haslam et al. Table 1.(Continued) SectionNcards (blank a) Theme Summary of content Illustrative examples b 5 37 (1) Thoughts and possible suspicions about the actorsMost Ps report being completely fooled by the actors. A few say they had doubtsI was taken in completely by your ‘actors.’ I hate to admit this; I gured I was a pretty world-wise guy. I keep a big safety-pin on my wallet pocket now. [2324B] 6 131 (3) Feelings and suspicions during studyMost Ps indicate they had doubts that shocks were real. However, many indicate that this only occurred to them afterwards.

Some disturbed by lack of debrie ng at end of sessionI felt somewhat guilty and nervous but as soon as I left the building I began to think and realized something was wrong and he could have been acting. [0201] After leaving the experiment, I gave it a great deal of thought for several hours. I came to the conclusion that I was the one being studied. Several months later, quite by accident, I found out the truth, and my suspicions were con rmed. [1013] 7 42 (4) Reasons for (not) continuing to administer shocksVaried responses, but many Ps mention prior experiences –particularly in the military, at work, and in their upbringingI am one who went all the way on the experiment. The reason: I don’t know. Perhaps it was because of military (service) training or perhaps it was that I have absolute faith in my leaders or superiors. [0233] I agree with your comments that it was an inner struggle to obey your instructions and at the same time hurt another person, defying my parents’ teaching. [0436] 8 103 (7) Reasons for (not) ceasing to administer shocksThose who continued refer to con dence in E and Yale. Those who stopped typically refer to (1) sympathy for the L and his heart condition (2) con ict with other roles (e.g., as health professional, lawyer), and (3) lack of good motiveI kept thinking, why didn’t I refuse to give pain to my fellow man, instead of going through as directed to the end. [0116] I did not like the idea of giving the shocks, but had complete con dence in the instructor and the nature of the experiment. [0922] Because as a lawyer, I felt there would be personal liability on my part and that of my colleagues (actors). Our learner had claimed a previous heart condition. As he continued to object and yell, I saw the possibility of my being held liable increasing. Knowing that as a lawyer I would be held to a higher degree of care, knowledge etc.

I decided to break off...[1320] Continued Reconceptualizing the Milgram studies67 Table 1.(Continued) SectionNcards (blank a) Theme Summary of content Illustrative examples b 9 202 (13) Signs of (lack of) stress during or since the experimentMost Ps indicate that they were (extremely) stressed during the study at the process of having to in ict pain, and afterwards when re ecting on what they had done (especially prior to debrie ng).

Many express relief after receiving reportYes, this experiment had bothered me because I could not see how Yale would allow such experiments to go on. I even went to a priest at the university where I work to ask about the morality of such an experiment.| The experi. left such an effect on me that I spent the night in a cold sweat and nightmares because of the fear that I had that I might have killed the man in the chair. This fear was aroused from the fact that I had to sign papers that I would bring no charges against Yale. [0711] I was upset to think that I was hurting someone even in the name of science. [1127] It was a decision, after having made the decision I was relatively calm.|...The fact that I followed the experimenter’s instructions to the end was based upon a decision, not upon the following of orders.

[2323B] 10 74 (7) Thoughts about experiment’s broader relevanceMany Ps comment on relevance of the study to WWII and the Holocaust. Most take conclusions at face value, but many also question them (on a range of grounds)I nd it a very depressing thought to consider that if I were recruiting Sonder Kommandos for Auschwitz I would nd 6 out of every 10 men in the army–maybe even more–would be suitably submissive to authority to commit atrocities under the rationale of obedience.

[0837] I really hesitate to draw any quantitative conclusions from the experiment were I the experimenter, because the experiment is not really conducted ‘in situ’.... As a teacher, I would not have gone on in a real life situation, but went on because I felt (1) Some good would come of it–(You needed the data, etc.), (2) You must have taken precautions to protect the learner from any harm–etc.

etc.... [1409] Continued 68S. Alexander Haslam et al. Table 1.(Continued) SectionNcards (blank a) Theme Summary of content Illustrative examples b 11 24 (3) Thoughts about obedience and complianceA range of relatively idiosyncratic re ections on the nature of obedience in societyThe problem of obedience is indeed one of the most crucial issue of our time (as you so aptly put it). The problem for each individual is to decide whether he should obey every order given to him, even if this violates his conscience.... [1706] 12 22 (1) Thoughts about deceptionMost Ps still aggrieved at deception and delay in receiving feedback. Many Ps also urge caution in selection of participants...| Upon re ection, I seriously question the wisdom and ethics of not completely dehoaxing each subject immediately after the session.

The standard ‘decompression’ treatment I received was not successful in reducing my anger and concern below the boiling point.... [0623] Since I became so upset during the experiment, I’m not sure that you were entirely responsible in picking your subjects. Suppose I’d had a heart condition? [2032] 13 140 (9) Thoughts about the value (or otherwise) of (having participated in) the researchMost Ps report feeling happy and privileged to have played a part in contributing to scienti c progress. Many make reference to the report. Many offer to participate in future researchAfter reading the results of this experiment, which I took part of, I will always be glad to help Yale in experiments. I rmly believe in experiments that will help to understand people. I did not realize scientists enth [went] to that extent to learn more of human behavior. [0125] Some of the self-distaste Some of the self-distaste is somewhat expunged now; at least I was making a contribution to an important search for truth. [0222] ...The above answers are my personal belief and I hope the may be of some small help, If in the future Yale does have any experiments I would gladly cooperate but I think that offering a person a sum of money for his help is a sence a insult, I believe a person should go through the whole thing willingly or not at all.

[2314B] Continued Reconceptualizing the Milgram studies69 Table 1.(Continued) SectionNcards (blank a) Theme Summary of content Illustrative examples b 14 130 (6) Insights gained through participationMany Ps report nding out that they are easily manipulated, others that they can resist pressure and think for themselves. Many report they have learned not to trust psychologistsIn psychological experiments they don’t always tell the truth as to what they are going to do. [#] I learned that I could resist group pressure and go according to my own feelings as to what I thought was right. [1121] 16 8 (0) Thoughts about E Most Ps express anger with E and/or MilgramI remember blaming the leader of the experiment more than myself, the executioner. Approx. 2:1 ratio. [0240] 17 18 (5) Thoughts about L Ps express mixture of sympathy and frustration with LI was angry at the learner for being so slow and forcing me to shock him harder. I was upset at having to punish someone. [0226] Total 1,057 (69) Note. aSome cards were blank because they referred to the primary section in which the relevant comment was led (e.g., ‘see card Category 9’).

bAll quotes verbatim (including typos, spelling errors, etc.).

cIt is unclear what T.R. stands for (we interpret it as something akin to ‘Troubled Responses’). 70S. Alexander Haslam et al. meta-contrast was relatively low for the rst two). These were labelled (1) support for Milgram’s project, (2) support for behavioural investigation, and (3) support for Yale. This thematic structure was then validated by the third author on the basis of an independent reading of the data. To be clear, the purpose of this exercise was not to interrogate the ways in which people make, justify, or warrant particular claims about their participation (e.g., by means of discourse analysis; see Gibson, 2011, 2014). It was to identify broad themes in the ways that people framed their feelings about having participated. However, having identi ed themes in this way, it is instructive, next, to re ect more closely on their speci c content.

Theme 1: Support for Milgram’s project The dominant theme that characterizes the comments in Section 13 relates to the sense of ful lment that participants experienced from participating in a scienti c study that they considered to be extremely important. These comments mainly elaborate on responses to Question 8 in the post-experimental survey (being glad or sorry to have participated) and re ect participants’ support for the speci c research programme in which they had taken part.

This occurs at various levels, depending upon what is pre-supposed in the response. In some cases participants pre-supposed that they have contributed to the research and that the research has contributed to society. Then pleasure and pride is expressed simply at having been part of the programme. Indeed, participants articulate a sense of being ‘special’ simply as a result of having been part of such an important project, as indicated by the following response: [1618] Felt pride at being chosen to participate. In other cases, it is pre-supposed that the project is of bene t and satisfaction is expressed at having been of use in the research: [no number] The experiment was a unique experience for me, and I believe everyone gets satisfaction from helping in a group endeavor.

[1817] I am very delighted to be a part of this project. I have often thought perhaps I was the subject, but I could not be any happier. I’ve been waiting very anxiously for this report to really put my mind at ease and curiosity satis ed. Again I will say I am happy to be a part of this project....Last but not least I sure hope my efforts, and cooperation, have been of somewhat useful for your project. In yet other cases, the broader bene t of the project is explicitly referenced, and the satisfaction relates directly to its societal impact: [0202] It seems to me that every experiment of this type cannot but help, even if only in a small measure, to teach us something.

[0603] To be a part of such an important experiment can only make one feel good, especially because for the reasons stated in your report.

[1716] It was my privilege to be of some help in what I think is a worthwhile experiment.

When you gentlemen have fully completed your research, and it is possible to do so, I would like to receive your nal report on this experiment, and the good that it might achieve for mankind.Reconceptualizing the Milgram studies71 Finally, elaborating on their responses to Question 9, there were cases where the notion of scienti c bene t led participants to argue for the need for more obedience studies, to advocate that others should be involved in them, and also to offer to participate in them themselves: [0319] If it [is] your belief and conviction that these studies will bene t mankind then I say we should have more of them...I found the experiment very interesting and wouldn’t hesitate to participate in future ones.

[0436] As you mentioned in your report, something has de nitely been accomplished to aid science. Why stop now? I am extremely grateful for being allowed to be a small part of this experiment.

[1205] I feel I have given something for the experiment. I think these experiments should go on, because it will make better citizen of the younger people, through better understanding. Theme 2: Support for behavioural investigation Closely related to Theme 1, many participants took the opportunity to express their enthusiasm for the systematic investigation of human behaviour. As the following Table 2.Levels of participant engagement with Milgram’s scienti c project based on analysis of responses in Section 13 of Box 44 Engagement with Milgram’s scienti c project–and illustrative commentN% 7 Very highly engaged I feel I have contributed in some small way toward the development of man and his attitudes towards others. I would be glad to participate in other studies. | I thoroughly enjoyed participating in the program and hope I will be called on again. [0203]33 23.6 6 Highly engaged The experiment was very interesting and worthwhile. | I think that studies of this kind are very helpful and should continue [0320]27 19.3 5 Moderately engaged Any study with an aim, if properly conducted, can do no harm and might be of some value [0436]34 24.3 4 Neither engaged nor disengagedIt is good to know that you would not permit me to give the learner the actual shocks under the condition of this experiment [0537]33 23.6 3 Moderately disengagedIt was only after speaking to you on the phone that I concluded the experiment had been prearranged and in all truthfulness somewhat silly. I would suggest that more experiments are conducted but that they be conducted on the more serious side.... [1922]8 5.7 2 Highly disengaged You might be interested to know that my opinion of Yale is quite low because of this experiment. | Kindly furnish me with the name & address so that I can satisfy my own thought about this experiment. [0820]5 3.6 1 Very highly disengaged00 Total 140 72S. Alexander Haslam et al. examples illustrate, these references have two distinct foci. First, participants stressed the importance of understanding behaviour: [0406] Where would we be without constant studies of man’s behavior–there is still an abundance of information to be learned.

[0841] Any deeper understanding of human behavior can only strengthen our countries ‘needs’ when applied by them who are ‘in support’ of our welfare. Second, they stressed the importance of studies that are scienti c and rigorous: [0208] I am in favor of all types of scienti c research. I am particularly in favor of research that is of a pure scienti c nature as compared to practical nature.

[0133] I could only say I am a rm believer in research and experiments which I believe will make life more comfortable as they are being done. I will always take part so long as I feel I am not doing harm to myself physically.

[0608] I am a rm believer in experimentations, any and all types. This may be construed as either quest for knowledge of plain inquisitiveness. If at any time I can be of assistance to you or others, please feel free to contact me.

[0415] Any study with an aim, if properly conducted, can do no harm and might be of some value. As with Theme 1, participants did not only endorse the work that had already been done, they also called for more research into human behaviour (and not simply into human obedience). Moreover, as is already clear from the words of Participant 0841 above, such work was often argued not only to be desirable but also to be necessary for human well- being and progress: [0518] Any experiment which might, by deep probing and intensive observation of human behavior, give us a clue to that which makes man ‘click’ is both worthwhile and necessary.

[0904] In this ever changing World, more should be learned about man in what he does and why. When this is learned we may have a better world to live in.

[2211] I consider research essential to progress and think everyone should encourage and aid when possible. Theme 3: Support for Yale A third theme had to do with support for, and con dence in, the institution where Milgram’s studies were conducted–Yale University: [0621] Being a citizen of New Haven I have always felt that Yale was good for New Haven. My feelings are now, that Yale is very good for New Haven.

[0236] I have always thought of Yale University as one of this country’s best institutions of learning. I think you are really learning what makes people ticks, so as to speak.

[1503] I have always had great con dence in any method developed by Yale. My admiration is unchanged. It was an experience I was happy to have added to my ‘memories’. As the second of these extracts intimates, it is not just that participants have con dence in Yale, it is that the association with the University is seen by them to confer scienti c Reconceptualizing the Milgram studies73 legitimacy on the studies and also to guarantee their value to humanity. This becomes clearer in the following extracts: [1706] I have the greatest respect for Yale, and what this great institution has done for mankind. I consider it a privilege that I could have participated in your experiment.

[1121] I have always had a high regard for Yale as a university and a research center. Although at times during the experiment I wished I wasn’t there, I believe I would do it over again.

I think all studies on human behavior are important. I wish to thank you for sending me the report. If I can ever again aid you in your studies or research projects please feel free to call on me. There are two further elements of interest here (speci cally in the latter extract), both of which we have encountered before. The rst is that the Yale connection, and the guarantee of scienti c credibility does not simply lead to positive evaluations, it resolves any ambivalence participants may have had by actively dispelling negative feelings. The second is that the resultant commitment is prospective as well as retrospective: Not only were participants happy to have participated but, having done so, they now actively seek out further involvement. As the next (and nal) extract indicates, participation in such prestigious research within such a prestigious institution is therefore understood to be a special privilege: [1204] My only desire is that some day I will be able to afford to go and learn more about the human race. And I hope I will be able to attend the greatest school in this country Yale of course. I wish to thank Prof. Milgram for allowing me to take part in such a dramatic experiment. Please feel free to call on me whenever you could possibly use me. Overall incidence and valence of themes It is important to stress that we are not suggesting that all the comments about Milgram’s work, about behavioural science in general, or about Yale were positive. Indeed, some of these more negative reactions have been emphasized in other researcher’s examination of material in the Yale archive (in particular, see Nicholson, 2011; Perry, 2011). Thus, to give a number of counter-examples to the above, some people felt that Milgram’s studies were of questionable value: [0909] I have no way of knowing if the expperi. was of value or whether further experi. are necessary to prove anything.

[2040] I feel it was a waste of time and money on behalf of the department. Some felt that psychology in general is of questionable value: [1905] To me, psychology remains a pseudo-science. These types of studies are little value in the study of human behavior since the variables are too many. And, nally, some felt negatively about the role of Yale in the research: [0820] You might be interested to know that my opinion of Yale is quite low because of this experiment. 74S. Alexander Haslam et al. Instead, our argument is, rst, that participants’ comments were predominantly positive, and, second, that being positive about the science was associated with coming to feel more positive about having participated in the studies even to the extent of being willing to undergo (and for others to undergo) similarly stressful experiences in the future. We will deal with the former point here and the latter point in the next section.

To provide a systematic summary of the incidence of the different themes that we have identi ed, and the proportion of positive, neutral, and negative comments relating to each, the rst and fourth author read and coded each card for both theme and valence.

Given the overlap in references to the value of Milgram’s science and the value of behavioural science more generally, we collapsed the rst two themes for this purpose.

The raters agreed that 137 of the 140 cards in Section 13 (98%) referred to the scienti c value of the research. Of these, 84 (61%) were positive, 40 (29%) were neutral, and 13 (9%) were negative. The raters agreed that 26 of the 142 cards (18%) referred to Yale University.

Of these 16 (61%) were positive, 7 (27%) were neutral, and 3 (12%) were negative.

Two broad points can therefore be drawn from this analysis of the material in Section 13 of Box 44. The rst is to con rm that the vast majority of the comments relating to participants’ experience of the studies are indeed positive and have to do either with the bene ts that scienti c study will bring to humankind or else with the worthiness of the institution that guarantees the scienti c value of the research. Following on from this, the second point is that qualitative analysis suggests a relation between individuals feeling good about their participation in the studies and their feeling positive about the scienti c value of the studies. That is, participants feel goodbecausethey feel that they have been part of something good, something progressive, something that will bene t humanity.

Several of them feel privileged to have been allowed to take part. Many of them want to do more and canvass the possibility of future participation. This suggests that being part of important science was something they actively wanted–and something they continued to want.

Quantitative examination of engagement and stress in post-experimental responses To summarize thus far, qualitative examination of Box 44 addresses RQ1 by suggesting that Milgram’s participants were closely engaged in his scienti c project and also provides some insight into RQ2 in so far as this engagement seems generally to have been associated with low levels of post-experimental stress. This is consonant with the ‘engaged followership’ explanation but not with the agentic state model (which would suggest that reduced stress is associated withdisengagement).

To provide more formal assessment of these patterns, we undertook a quantitative analysis of the material in Box 44. The rst and fourth authors made independent ratings of the degree to which the comments on each card in Section 13 were indicative of high or low engagement with the science of the study. This was done on a 7-point scale, where 1=very low engagement,7=very high engagement. There was a high degree of consensus in the two raters’ assessment of participants’ engagement (r=.91), and in no case did their ratings differ by more than one scale point. Accordingly, the two ratings were averaged to create a single engagement score. The distribution of engagement scores is presented in Table 2 (for this purpose, engagement ratings were rounded up when averaging had produced non-whole values). Reconceptualizing the Milgram studies75 The rst thing to note from this distribution is that, across the board, participants were generally highly engaged in the science. Consistent with this, a one-samplet-test indicated that participants displayed levels of engagement that were signi cantly above the scale midpoint–M=5.06,t(139)=9.14,p<.001. In relation to RQ2, then, these data suggest that the lack of stress experienced by Milgram’s participants after the study was associated with engagement in his scienti c enterprise rather than with disengagement. The same is not true of their stress during the study. In other words, a belief in the science does not remove the dilemma that participants face. It does not make them unaware that they are in icting pain. It does not make performing the task any easier (as seen by the comments in Section 9 of Box 44). But it does help them resolve the key dilemma they confront, it does help them live with themselves, and (as we saw from the qualitative material) it does makes them prepared to support more of the same in the future. Again, all this is consistent with the engaged follower explanation but not with the agentic state account. Integrative summary of evidence from Box 44 The rst point to emerge from the above analysis–the rst systematic, detailed examination of the views of participants in one of the most famous and controversial psychological research programmes ever conducted–is that it con rms Milgram’s claim that, by and large, those who took part in his studies were not distressed by their participation. On the contrary, they felt pleased, privileged, and were eager for more. This also con rms our contention that the controversy between Baumrind and Milgram may have been focused (and led subsequent debate to be focused) on the wrong issues. This is because the key question that needs addressing is not whether or not participants were distressed but ratherwhy they were so happy.

Related to this, a second point that our analysis draws attention to is the degree to which Milgram’s participants were engaged with the science of his study. By and large they saw science–especially science associated with a prestigious institution like Yale–as a social good and as an important driver of human progress. Being associated with this was something they took pride in and that they felt good about. Accordingly, in Section 13 we see very little of the ‘inattentive bureaucrat’ who is so focused on the minutiae of his performance that he loses sight of the bigger picture. Here, the bigger picture is very much in focus. Indeed, it dominates the attentional eld.

It needs to be borne in mind, of course, that participants wrote these comments some time after their participation in the studies, when the stress they had endured was in the past and they had just been reminded, through Milgram’s report, of the scienti c goals of his research. This is an obvious and important proviso to bear in mind in re ecting on our analysis, especially as we move from looking at constructs in isolation to the more critical issues concerning the relationships between them.

We can start to do this by considering the relationship between participants’ feelings about their participation in the study–in particular, the level of post-experimental stress that they experienced–and their engagement with the science of the study. Here, three features of the data seem signi cant. First, our qualitative analysis suggests that it is the science that gives meaning to participants’ experience in the study and that makes the tensions they confronted seem worthwhile. It ennobles an otherwise unpleasant experience and transforms it into something to embrace–and, if possible, to repeat. In short, participants feel good about themselves because they have been part of, and assisted in, scienti c progress.

76S. Alexander Haslam et al. Yet second, signi cantly, we see that it is only after the study that an awareness of scienti c value creates a positive feeling about participation and stress is ameliorated. As Participant 1817 put it ‘I’ve been waiting very anxiously for this report to really put my mind at ease and curiosity satis ed’.More generally, then, there is evidence that participants were decidedly uncomfortable about what they had done up to the point that Milgram’s report provided them with the scienti c rationale for their participation (see Nicholson, 2011, for detailed evidence on this point). Once that was delivered, they, like Participant 1817, could af rm that they were ‘delighted to be part of the project’ and ‘could not be any happier’.

The role of Milgram’s words in shaping the feelings and action of participants also alerts us to a critical problem with the agentic state account (and with explanations of destructive behaviour more generally; Reicheret al., 2012). These generally draw our attention only to the behaviour and psychology of participants. They thereby neglect the active role that Milgram (and his associates) played in actively cultivating forms of engagement that encouraged particular acts of followership (Gibson, 2011; Haslamet al., 2011).

We therefore suggest that Milgram’s own leadership constitutes a critical third factor that contributed to how participants felt and acted both during and after the studies. For not only did Milgram work hard at both junctures to encourage forms of engagement and identi cation that would encourage compliance (Rochat & Modigliani, 1997), but so too, after the experiment, his debrie ng promoted forms of engagement and identi cation that would mitigate against stress and facilitate post-traumatic growth (Joneset al., 2011).

Indeed, from this perspective, the skill of his report (which is replicated, we suggest, in many experimental debrie ngs) was precisely to provide participants with an explan- atory narrative that made them feel that they were part of a collective enterprise and had contributed to a collective good–speci cally the advancement of progressive science.

Once more, the content of participants’ comments suggests that in this he was generally very successful. Conclusion: Reintegrating the empirical and ethical dimensions of Milgram’s studies One of the consequences of separating out what happened during the obedience studies from what happened afterwards–in part occasioned by Milgram’s abandonment of the idea that both were rooted in the active choices of the participants–is that, over time, the scienti c and ethical dimensions of Milgram’s work have become largely divorced. On the whole, the question of why people shocked has been pursued in isolation from the question of whether it was acceptable to subject them to a shocking situation.

The analysis we have pursued in this study brings the two back together, suggesting that cognate processes govern what people did in the studies and how they felt about them subsequently. In both cases, we suggest that this hinged on participants’ engagement with science as a source of human progress and welfare. Accordingly, it was as engaged followers of Milgram’s leadership and of his scienti c mission that participants responded to the Experimenter’s instructions during the study and then, after it was over, re ected positively on their involvement.

Part of the power of thisengaged followeranalysis, then, has to do with its ability to account parsimoniously for the complex patterns of data observed both within the studies and–as we have just seen–after they were completed. Indeed, despite clear limits to our Reconceptualizing the Milgram studies77 ability to make inferences about the patterns we have discussed, this analysis is certainly far more consistent with those patterns than Milgram’s own agentic state account.

At the same time, a further part of the power of this fresh account lies in its resonance with recent historical analyses that have called into question the banality of evil thesis that underpinned Milgram’s own emergent theorizing. Indeed, if the emblematic event for Milgram’s ‘agentic state’ approach is Eichmann in Jerusalem, or, more accurately, Arendt’s (1963) portrait of Eichmann as a seemingly non-descript bureaucrat, so the emblematic event for the engaged followership analysis would be Himmler’s visit to Poznan in October 1943. There he addressed SS Of cers responsible for conducting operations to exterminate Jews in occupied Poland, and in one notorious passage urged them on by reminding them of the nobility of their cause: I am now referring to the evacuation of the Jews, the extermination of the Jewish people. It is one of those things that is easily said: ‘The Jewish people is being exterminated’, every party member will tell you...[But] none of them has seen it happen, not one has had to go through with it. Most of you men know what it is like to see 100 corpses lie side by side, or 500 or 1,000.

To have stood fast through all this and...at the same time to have remained a decent person...

has made us hard. This is an unwritten and never-to-be-written page of glory in our history....

All in all, however, we can say that we have carried out this most dif cult of tasks in a spirit of love for our people. (Grobmes & Landen, 1983, pp. 454–455) What we see here is Himmler facing up to the realities of mass murder. He does not hide the fact that it is deeply unpleasant. On the contrary, he claims that it is precisely because it is unpleasant that it confers merit on those who kill in a ‘noble’ cause. For it is only the best who will subject themselves to such a foul activity for the collective good.

The importance of this speech is that it encapsulates a toxic logic that, according to recent scholarship, pervaded the Holocaust (e.g., see Cesarani, 2004; Herf, 2008; Lozowick, 2002; Vetlesen, 2005). The perpetrators of this were no inattentive bureaucrats. They knew full well what they were doing. They acted strenuously and creatively to nd ways of deporting and killing Jewish people and others besides. And they did so because they identi ed with the Nazi cause and because they believed the Nazi claim that Jews threatened the glorious German Reich. In the simplest terms, they did it because they believed that what they were doing was right.

It is worth adding that the toxic logic of killing in the name of racial progress was not limited to the Nazis. Duggan (2013) shows how Mussolini’s exterminationist war against Ethiopia (then, Abyssinia), in which up to 275,000 people were killed mainly by chemical warfare, was justi ed mainly in terms of the superiority of Italian civilization and the bene ts that would accrue to the indigenous people from its introduction. He quotes from the diaries of many of those involved in the war including one soldier, Manlio la Sorsa: There were no discernible chinks in Manlio’s moral armour. He was able to witness thousands of Ethiopian corpses strewn across a battle eld almost with dispassion: they were a regrettable but necessary price for making ‘our dear fatherland greater, stronger and more respected’ and bringing civilization to ‘this dark and shadowy land’. (Duggan, 2013, p. 270) One must always be careful in drawing parallels between Milgram’s studies and Fascism, Nazism, and–particularly–the Holocaust, not least because of the dangers of losing its uniqueness and trivializing the slaughter of so many people. Certainly, whatever we think of what happened in Milgram’s studies, it bears no direct comparison to the death camps.

78S. Alexander Haslam et al. Nonetheless, Milgram himself sought to draw parallels at the level of the underlying processes which allow people to harm others–even if the form and level of harm is qualitatively different (Mastroianni, 2002; Miller, 2004). We do likewise, but we differ from Milgram as to the nature of the process (just as contemporary historians have differed from Arendt). For us, it is this selfsame identi cation with a ‘noble’ cause that led his participants to prove willing to administer what they thought were lethal shocks to a helpless stranger and then, ultimately, to feel happy about what they had done. (Indeed, although we have not discussed it here, very similar forms of identi cation can be seen to have underpinned the willingness of Milgram’s research assistants to play their important role in this process; see Nicholson, 2011; Russell, 2014). Moreover, as we noted towards the start of this study, examination of Milgram’s experimental notes suggests that, at the time that he was conducting his research, this was a point of which he was very much aware.

There is one more point we wish to make. As may be obvious by now, our analysis does not simply reconnect the science and the ethics of Milgram’s research, it also reframes the nature of the ethical issues that the studies raise. Tellingly, Baumrind (1964) concludes her critique of Milgram by writing: ‘I would not like to see experiments such as Milgram’s proceed unless the subjects were fully informed of the dangers of serious aftereffects and his correctives were clearly shown to be effective in restoring their state of well-being’ (1964, p. 423; see also Nicholson, 2011). The implication, of course, is that if there were informed consent and if participants could be made happy about their involvement, then the ethical doubts about Milgram’s research would be dispelled. In other words, the ethical issues associated with the obedience studies are seen to revolve entirely around participants’ (lack of) distress.

Our analysis supports Milgram’s claims that most of his participants were not distressed. Indeed, as we have seen, the majority professed that they were very happy about their participation. But, in contrast to Milgram himself, we do not see this as dispelling the ethical problems. Indeed, we see this as raising more problems as it resolves.

For Milgram restored his participants’ well-being by skilfully engaging them with the notion that science is such an overwhelming good for humanity that any incidental ‘collateral damage’ along the way is entirely unproblematic. That is, participants were made comfortable at the cost of accepting an ideology with considerable potential for social harm–a potential vividly realized through Milgram’s own demonstration that this was a cause for which participants were prepared to kill an innocent person. This, we suggest, is a very signi cant ethical dimension of the obedience studies, but it is one that, hitherto, has been largely overlooked.

Final comment Our analysis supports the view that people are able to in ict harm on others not because they are unaware that they are doing wrong, but rather because–as engaged followers– they know full well what they are doing andbelieve it to be right(Haslam & Reicher, 2012a,b; Reicher & Haslam, 2012).

Thankfully, notions of racial superiority and racial hygiene–and hence the ability to annihilate people in the name of racial progress–no longer have the purchase that they once did. But that does not mean that there are not other notions of progress, in the name of which various forms of viciousness can still be justi ed. Our close investigation of what Milgram’s participants had to say about their experiences in his studies points to the way in which ‘science’ itself has the potential to be invoked as a ‘warrant for abuse’. This in Reconceptualizing the Milgram studies79 turn suggests that the horizons of scienti c ethics (and the practices of ethics committees) need to extend beyond a concern to ensure that participants are content, to also re ect on what they are encouraged to be content about. Certainly, all the evidence suggests that Milgram’s participants were ‘happy to have been of service’. As scientists, we need to ask whether this is the kind of service with which we want people to be quite so happy.

Acknowledgements We are grateful to two anonymous reviewers for their comments on an earlier draft of this study. Work on this study was supported by grants from the Australian Research Council (FL110100199; DP1301108) and the Economic and Social Research Council (ES/L003104/1).

References Akerlof, G. A. (1991). Procrastination and obedience.American Economic Review,81,1–19.

Arendt, H. (1963).Eichmann in Jerusalem: A report on the banality of evil. New York, NY:

Penguin.

Baumrind, D. (1964). Some thoughts on ethics of research: After reading Milgram’s “behavioral study of obedience”.American Psychologist,19, 421–423. doi:10.1037/h0040128 Blass, T. (1999). The Milgram paradigm after 35 years: Some things we now know about obedience to authority.Journal of Applied Social Psychology,29, 955–978. doi:10.1111/j.1559-1816.

1999.tb00134.x Blass, T. (2004).The man who shocked the world: The life and legacy of Stanley Milgram. New York, NY: Basic Books.

Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology.Qualitative Research in Psychology,3,77–101. doi:10.1191/1478088706qp063oa Brehm, J. W. (1966).A theory of psychological reactance. New York, NY: Academic Press.

Burger, J. (2009). Replicating Milgram: Would people still obey today?American Psychologist,64, 1–11. doi:10.1037/a0010932 Burger, J. M., Girgis, Z. M., & Manning, C. M. (2011). In their own words: Explaining obedience to authority through an examination of participants’ comments.Social Psychological and Personality Science,2, 460–466. doi:10.1177/1948550610397632 Castano, E. (2008). On the perils of glorifying the ingroup: Intergroup violence, in-group glori cation and moral disengagement.Social and Personality Psychology Compass,2, 154– 170. doi:10.1111/j.1751-9004.2007.00040.x Cesarani, D. (2004).Eichmann: His life and crimes. London, UK: Heinemann.

Darley, J. M. (1992). Social organization and the production of evil.Psychological Enquiry,2, 199– 218. doi:10.1027/s15327965pli0302_28 Duggan, C. (2013).Fascist voices: An intimate history of Mussolini’s Italy. London, UK: Vintage Books.

Fox, R., Crask, M. R., & Kim, J. (1988). Mail survey response rates: A meta-analysis of selected techniques for inducing response.Public Opinion Quarterly,52, 467–491. doi:10.1086/ 269125 Gibson, S. (2011). Milgram’s obedience experiments: A rhetorical analysis.British Journal of Social Psychology,2, 290–309. doi:10.1111/j.2044-8309.2011.02070.x Gibson, S. (2014). Discourse, de ance and rationality: ‘Knowledge work’ in the ‘obedience’ experiments. Journal of Social Issues,70, 424–438. doi:10.1111/josi.12069 Gilbert, S. J. (1981). Another look at the Milgram obedience studies: The role of a graduated series of shocks.Personality and Social Psychology Bulletin,7, 690–695. doi:10.1177/01461672 8174028 80S. Alexander Haslam et al. Grobmes, A., & Landen, D. (1983).Genocide: Critical issues of the Holocaust. West Orange, NJ:

Berman House.

Haslam, S. A., & McGarty, C. (2014).Research methods and statistics in psychology(2nd ed.).

London, UK: Sage.

Haslam, S. A., & Reicher, S. D. (2007). Beyond the banality of evil: Three dynamics of an interactionist social psychology of tyranny.Personality and Social Psychology Bulletin,33, 615–622. doi:10.

1177/0146167206298570 Haslam, S. A., & Reicher, S. D. (2012a). When prisoners take over the prison: A social psychology of resistance.Personality and Social Psychology Review,16, 154–179. doi:10.1177/10888683 11419864 Haslam, S. A., & Reicher, S. D. (2012b). Contesting the “nature” of conformity: What Milgram and Zimbardo’s studies really show.PLoS Biology,10, e1001426. doi:10.1371/journal.pbio.

1001426 Haslam, S. A., Reicher, S. D., & Birney, M. (2014). Nothing by mere authority: Evidence that in an experimental analogue of the Milgram paradigm participants are motivated not by orders but by appeals to science.Journal of Social Issues,70(3), 473–488. doi:10.1111/josi.12072 Haslam, S. A., Reicher, S. D., & Platow, M. J. (2011).The new psychology of leadership: Identity, in uence and power. New York, NY and Hove, UK: Psychology Press.

Helm, C., & Morelli, M. (1979). Stanley Milgram and the obedience experiment: Authority, legitimacy, and human action.Political Theory,7, 321–346. doi:10.1177/009059177900700303 Herf, J. (2008). Post-totalitarian narratives in Germany: Re ections on two dictatorships after 1945 and 1989.Totalitarian Movements and Political Religions,9, 161–186. doi:10.1080/ 14690760802094792 Jetten, J., & Hornsey, M. J. (Eds.) (2011).Rebels in groups: Dissent, deviance, difference and de ance. Chichester, UK: Wiley-Blackwell.

Jetten, J., & Mols, F. (2014). 50-50 hindsight: Appreciating anew the contributions of Milgram’s obedience experiments.Journal of Social Issues,70(3), 587–602. doi:10.1111/josi.12080 Jones, J. M., Haslam, S. A., Jetten, J., Williams, W. H., Morris, R., & Saroyan, S. (2011). What doesn’t kill you can make you stronger (and more satis ed with life): The impact of identity on well-being after acquired brain injury.Psychology and Health,26, 353–369. doi:10.1080/08870440 903440699 Kershaw, I. (1993). Working towards the F€ uhrer: Re ections on the nature of the Hitler dictatorship.

Contemporary European History,2, 103–108. doi:10.1017/S0960777300000382 Lozowick, Y. (2002).Hitler’s bureaucrats: The Nazi security police and the banality of evil(H.

Watzman trans.). London, UK: Continuum.

Lutsky, N. (1995). When is “obedience” obedience? Conceptual and historical commentary.Journal of Social Issues ,51,55–65. doi:10.1111/j.1540-4560.1995.tb01334.x Mandel, D. R. (1998). The obedience alibi: Milgram’s account of the Holocaust reconsidered.

Analyse und Kritik,20,74–94.

Mantell, D. M., & Panzarella, R. (1976). Obedience and responsibility.British Journal of Social and Clinical Psychology,15, 239–245. doi:10.1111/j.2044-8260.1976.tb00030.x Mastroianni, G. R. (2002). Milgram and the Holocaust: A reexamination.Journal of Theoretical and Philosophical Psychology,22, 158–173. doi:10.1037/h0091220 Milgram, S. (1963). Behavioral study of obedience.Journal of Abnormal and Social Psychology,67, 371–378. doi:10.1037/h0040525 Milgram, S. (1964). Issues in the study of obedience: A reply to Baumrind.American Psychologist, 19, 848–852. doi:10.1037/h0044954 Milgram, S. (1965). Some conditions of obedience and disobedience to authority.Human Relations, 18,57–76. doi:10.1177/001872676501800105 Milgram, S. (1974).Obedience to authority: An experimental view. New York, NY: Harper & Row.

Milgram, S. (no date a).Questionaire[sic]for subjects in memory and learning project. Milgram Archive, Yale University Library, Box 45, Folder 158.Reconceptualizing the Milgram studies81 Milgram, S. (no date b).Report to memory project subjects. Milgram Archive, Yale University Library, Box 45, Folder 158.

Millard, K. (2011). The window in the laboratory: Stanley Milgram as lmmaker.The Psychologist, 24, 658–660.

Millard, K. (2014). Revisioning obedience: Exploring the role of Milgram’s skills as a lm-maker in bringing his shocking narrative to life.Journal of Social Issues,70(3), 439–455. doi:10.1111/ josi.12070 Miller, A. (1995). Constructions of the obedience experiments: A focus upon domains of relevance.

Journal of Social Issues,51,33–53. doi:10.1111/j.1540-4560.1995.tb01333.x Miller, A. G. (2004). What can the Milgram obedience experiments tell us about the Holocaust?

Generalizing from the social psychology laboratory. In A. Miller (Ed.),The social psychology of good and evil(pp. 193–239). New York, NY: Guilford.

Navarick, D. J. (2009). Reviving the Milgram obedience paradigm in the era of informed consent.The Psychological Record,59, 155–170.

Neitzel, S. (2007).Tapping Hitler’s Generals. Barnsley, UK: Frontline Books.

Newman, L. S. (2001).The banality of secondary sources: Why social psychologists have misinterpreted Arendt’s thesis. Unpublished manuscript. Syracuse, NY: Syracuse University.

Nicholson, I. (2011). “Torture at Yale”: Experimental subjects, laboratory torment and the “rehabilitation” of Milgram’s “Obedience to Authority”.Theory & Psychology,21, 737–761.

doi:10.1177/0959354311420199 Nissani, M. (1990). A cognitive reinterpretation of Stanley Milgram’s observations on Obedience to Authority.American Psychologist,45, 1384–1385. doi:10.1037/0003-066X.45.12.1384 Novick, P. (2000).The Holocaust and collective memory. London, UK: Bloomsbury.

Overy, R. (2014). Ordinary men, extraordinary circumstances: Historians, social psychology, and the Holocaust.Journal of Social Issues,70(3), 515–530. doi:10.1111/josi.12075 Packer, D. J. (2008). Identifying systematic disobedience in Milgram’s obedience experiments:

A meta-analytic review.Perspectives on Psychological Science,3, 301–304. doi:10.1111/j.1745- 6924.2008.00080.x Perry, G. (2011).Behind the shock machine: The untold story of the notorious Milgram psychology experiments. Brunswick, Vic.: Scribe.

Reicher, S. D., & Haslam, S. A. (2011). After shock? Towards a social identity explanation of the Milgram ‘obedience’ studies.British Journal of Social Psychology,50, 163–169. doi:10.1111/j.

2044-8309.2010.02015.x Reicher, S. D., & Haslam, S. A. (2012). Obedience: Revisiting Milgram’s obedience studies. In J. R.

Smith & S. A. Haslam (Eds.), Social psychology: Revisiting the classic studies(pp. 106–125).

London, UK and Thousand Oaks, CA: Sage.

Reicher, S. D., Haslam, S. A., & Rath, R. (2008). Making a virtue of evil: A ve-step social identity model of the development of collective hate.Social and Personality Psychology Compass,2, 1313–1344. doi:10.1111/j.1751-9004.2008.00113.x Reicher, S. D., Haslam, S. A., & Smith, J. R. (2012). Working toward the experimenter:

Reconceptualizing obedience within the milgram paradigm as identi cation-based followership.

Perspectives on Psychological Science,7,315–324. doi:10.1177/1745691612448482 Roccas, S., Klar, Y., & Liviatan, I. (2006). The paradox of group-based guilt: Modes of national identi cation, con ict vehemence, and reactions to the ingroup’s moral violations.Journal of Personality and Social Psychology,91, 698–711. doi:10.1037/0022-3514.91.4.698 Rochat, F., & Modigliani, A. (1997). Authority: Obedience, de ance, and identi cation in experimental and historical contexts. In M. Gold (Ed.),A new outline of social psychology (pp. 235–246). Washington, DC: American Psychological Association.

Russell, N. J. C. (2010). The making of an (in)famous experiment.The Psychologist,23, 780–783.

Russell, N. J. C. (2011). Milgram’s obedience to authority experiments: Origins and early evolution.

British Journal of Social Psychology,50, 146–162. doi:10.1348/014466610x492205 Russell, N. J. C. (2014). The emergence of Milgram’s bureaucratic machine.Journal of Social Issues, 70(3), 409–423. doi:10.1111/josi.12068 82S. Alexander Haslam et al. Russell, N. J. C., & Gregory, R. J. (2011). Spinning an organizational “web of obligation”? Moral choice in Stanley Milgram’s “Obedience to Authority” experiments.The American Review of Public Administration,41, 495–518.

Sofsky, W. (1993).The order of terror: The concentration camp. Princeton, NJ: Princeton University Press.

Turner, J. C. (1985). Social categorization and the self-concept: A social cognitive theory of group behaviour. In E. J. Lawler (Ed.),Advances in group processes(Vol. 2, pp. 77–122). Greenwich, CT: JAI Press.

Turner, J. C. (1991).Social in uence. Milton Keynes, UK: Open University Press.

Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987).Rediscovering the social group: A self-categorization theory. Oxford, UK: Blackwell.

Tyler, T. R. (1990).Why people obey the law. New Haven, UK: Yale University Press.

Tyler, T. R. (1998). The psychology of authority relations: A relational perspective on in uence and power in groups. In R. M. Kramer & M. A. Neale (Eds.),Power and in uence in organizations (pp. 251–260). Thousand Oaks, CA: Sage.

Vetlesen, A. J. (2005).Evil and human agency: Understanding collective evildoing. Cambridge, UK: Cambridge University Press.

Zimbardo, P. (2007).The Lucifer Effect: How good people turn evil. London, UK: Random House.

Zimbardo, P., & Gerrig, R. (1999).Psychology and life(15th ed.). New York, NY: Longman.

Received 29 May 2013; revised version received 2 March 2014Reconceptualizing the Milgram studies83 Copyright ofBritish Journal ofSocial Psychology isthe property ofWiley- Blackwell andits content maynotbecopied oremailed tomultiple sitesorposted toalistserv without the copyright holder'sexpresswrittenpermission. However,usersmayprint, download, oremail articles forindividual use.