Obedience lecture PDF

Preview

Summary

Lecturer Stefanie Scmeer...

Description

Obedience In the early 1960s Stanley Milgram conducted a classical series of experiments on obedience. His original motivation was to find an answer to the question of why during the Nazi regime ordinary human beings obeyed orders resulting in the killings of large numbers of innocent people. His experiments took place in the psychology laboratory at Yale University. Participants were usually males between 20 and 50 years old, recruited by a newspaper advertisement that asked for volunteers to participate in a study of memory and learning. Participants had a wide range of occupations and educational backgrounds. They were paid $4.50 for turning up. For each session two people arrived at the laboratory. Only one of them was a real participant, however, the other person was a confederate of the experimenter. The experimenter explained that the experiment was designed to investigate the effect of punishment on learning. For this purpose one person would have to be the teacher and the other the learner. It would be the task of the teacher to read a series of word pairs to the learner and then test the learner’s ability to recall the pairs by reading the first word of each pair along with four other words. The learner was to indicate which of the four words had originally been paired with the other word. Wrong answers would be punished by an electric shock. Both the participant and the confederate were then asked to draw lots to determine who would be the teacher and who the learner. The real participant always got the teacher role, because both slips of paper contained the word "teacher". Teacher and learner were then taken to an adjacent room where an electrode was attached to the learner’s wrist. Electrode paste was applied "to avoid blisters and burns". The electrode was allegedly attached to a shock generator in an adjacent room (but the confederate did not actually receive any shocks). The participant was taken to this room and the shock generator was explained to him. The shock generator consisted of 30 lever switches in a horizontal line, each clearly marked with a voltage value that ranged from 15 to 450 volts and groups of four switches with a verbal label ranging from ‘slight shock’ to ‘XXX’. Each participant was given a sample shock of 45 volts (from a battery) to further convince him of the authenticity of the machine. Participants were told to give a shock to the learner each time he gave a wrong response and increase the shock level by one each time the learner gave a wrong answer. Since the confederate seemingly made many mistakes, the shocks the participants were supposed to apply soon reached a high level. In the first experiment nothing was heard from the confederate until a shock of 300 volts, when he seemed to bang on the wall. After the next shock of 315 volts no more answers were given. The participant was instructed to treat no answers as wrong answers and to continue to deliver the shocks. If the participant refused, the experimenter used a series of prods to bring him into line, such as "Please continue", "The experiment requires that you continue", "It is absolutely essential that you continue", “You have no choice but to continue”, always delivered in this order. The experimenter’s tone of voice was firm, but not impolite. If the participant still refused to obey, the experiment was terminated. The dependent variable was the maximum shock administered before the participant refused to continue. Various people including psychologists, psychiatrists, and lay people were asked to make predictions of the participants’ behaviour in this situation. The expectation was that few, if any, people would go beyond ‘very strong shock’. Nobody expected maximum obedience, i.e., an administration of a 450-volt shock three times. The expected mean maximum shock was 135 volts. These expectations, however, proved to be incorrect. In the first experiment no participant stopped before 300 volts (when the confederate banged on the wall and no longer gave answers), 65% of the participants showed maximum obedience and the mean maximum shock administered was 405 volts. These results attracted considerable attention and prompted a series of replications and variations, both by Milgram himself and by other people.

1. Closeness of the victim a) remote see above b) voice feedback The victim was in an adjacent room, but tape-recorded verbal protests could be heard clearly. 62.5% of the participants obeyed, the mean maximum shock given was 367.5 volts. c) proximity The victim sat in the same room, a few feet away from the participant and therefore was visible as well as audible. 40% of the participants obeyed, the mean maximum shock given was 312 volts. d) touch proximity The victim received the shock only if his hand rested on a plate. He refused to do so after 150 volts and the experimenter ordered the subject to force the victim’s hand onto the plate. 30% of the participants obeyed, the mean maximum shock given was 268 volts. Obedience declined as the proximity of the victim to the participant increased, because it became increasingly difficult for the participant to ignore the victim’s signs of suffering and deny that their own actions would have caused these consequences. 2. Authority of the experimenter a) absent When the experimenter left the laboratory and gave his orders over the phone 20.5% of the participants obeyed, the mean maximum shock given was 272 volts. Some participants administered lower shocks than they were supposed to and lied to the experimenter when he asked whether they had followed the instructions correctly. In order for high obedience levels to be obtained, the experimenter has to be present. b) experiment in an office building The experiment did not take part in the Yale laboratories but in a private office building, allegedly carried out on behalf of a private research institute. 48% of the participants obeyed, the mean maximum shock given was 315 volts. The fact that the experiment was no longer carried out at Yale University with its high reputation reduced the experimenter’s legitimate power. c) an ordinary person gives the orders The experimenter was called away without having given special shock instructions yet but asked that the experiment should be completed in his absence. A confederate of the experimenter whose task it had been to record reaction times (while the participant administered the shocks) announced that he had thought of a good system and insisted that the procedure of increasing the shock level for each incorrect answer was followed. 20% of the participants obeyed, the mean maximum shock given was 243 volts. The ordinary person was not perceived as an expert and therefore could not elicit high levels of obedience. However, when a participant refused to continue and the confederate administered the shocks while the participant just watched, 69% of the participants did not stop the confederate from administering the maximum shock, the mean maximum shock the confederate was allowed to give was 375 volts. Participants did not feel responsible for questionable actions they saw someone else performing.

d) two authorities, contradictory commands Two experimenters with equal status were in charge of the experiment. At the 150 voltlevel one wanted to stop the experiment, the other wanted to continue. All participants broke off. There must be a consensus between authorities in order for people to obey. 3. Group influence a) another person administers the shocks The experiment involved two teachers, one of them was a confederate of the experimenter who administered the shocks. The participant performed some subsidiary task. 92.5% of the participants did not stop the confederate from administering the shocks, the mean maximum shock the confederate was allowed to give was 429 volts. b) two other people rebel The experiment involved three teachers, two of them were confederates of the experimenter. One read the word pairs, one told the victim whether his answers were correct, the participant’s task was to administer the shock. At 150 volts the first confederate refused to continue, at 210 volts the other. 10% of the participants obeyed, the mean maximum shock given was 246 volts. The example of disobedience provided by others helps the participants to rebel against the orders of the authority themselves. Several explanations for why obedience occurs have been put forward: Firstly, some people suggested that the errors of the learner or the situation more generally frustrated the teacher and he therefore became aggressive. This explanation is not valid, however, because when participants were allowed to choose the shock levels themselves, almost all chose the lowest possible level. The mean maximum shock given was 54 volts. Also, in the absence of the experimenter obedience levels dropped and participants started to cheat. If the teacher was aggressive towards the learner, the absence of the experimenter should not have made any difference. 1. Personality characteristics? Some differences were observed between obedient and defiant participants, however, these are unlikely to be the sole explanation for the behaviour of the participants: • Obedient participants showed significantly higher authoritarian values (F-Scale, Adorno et al., 1950) • Obedient participants had served longer in the army (exception: officers were more often defiant) • Obedient participants had a lower level of education • Obedient participants had technical professions more often than social professions In a study by Kilham and Mann (1974), men were found to be more obedient than women. In this experiment participants had to either order someone else to inflict pain on another person or to carry out an order to inflict pain on another person. Generally, obedience levels were higher in the first condition. Women were less obedient than men, especially when it came to carrying out an order to inflict pain and when the pain was to be inflicted upon another female. However, the experimenter was always male and one therefore has to be careful with generalising the results. In other studies, either no difference was found between men or women (Milgram, 1974) or women were found to be more obedient (Sheridan & King, 1972). Sheridan and King's victim was a puppy, their participants were college students. The puppy received real electric shocks, which were not as strong as the labels indicated but nevertheless strong enough to make the puppy jump and yelp. Contrary to the authors’ expectations, all women (n=13, which is very low) applied the maximum shock, whereas only 54% of the men showed maximum obedience.

2. Milgram’s explanation Milgram assumes that we are born with a potential for obedience. This has been strengthened because of the survival value of hierarchy, i.e., obedience allows people to cope better with threats from the environment and gives stability and harmony to relations within a group. Also, people have learned that obedience is usually rewarded. In the particular context of Milgram’s experiments the scientific ideology that gives a unique power to experimenters in scientific experiments was also crucial. Participants were led to believe that it was necessary to do what the experimenter told them to in order to help science. Milgram proposes that his participants moved from a self-directed autonomous state to a so-called agentic state. Milgram states that “the person entering an authority system no longer views himself as acting out of his own purposes but comes to see himself as an agent for executing the wishes of another person” (1974, p. 133). This agentic state, Milgram suggested, has a number of psychological effects: • Participants wanted to perform competently and make a good impression. Therefore they attended to the features of the situation that would suggest to them what was required for a competent performance, i.e., they ‘tuned in’ to the signals of the authority (the experimenter) while the signals of the learner were muted. • Participants accepted the definition of their actions as provided by the experimenter who was perceived to be a legitimate authority and redefined the meaning of the situation as one where they helped the experimenter and not one where they were applying deadly electric shocks to an innocent victim. • There was a loss of perceived personal responsibility; participants felt responsible to the authority, but not for the content of any actions that the authority prescribed. • Their self-ideal, which might have otherwise kept them from behaving in the way they did, was inhibited and self-evaluation was largely absent, because their actions were not self-motivated. According to Milgram certain factors put strain on the participants and encouraged them to leave the agentic state, like the cries of pain from the learner, the contradictory demands made upon them by the experimenter and the learner, the conflict between the norms to be obedient and to not hurt other people, the fear of retaliatory actions from the learner, and the incompatibility of their actions with their self-image. These factors were counteracted by binding factors, which kept participants in the agentic state, such as the sequential nature of the action. Breaking off would have meant admitting that everything done to this point had been wrong. The importance of this argument was discussed by Gilbert (1981). If the shock box had only two switches, one for 15 volts and one for 450 volts, and the teacher would be asked to administer a 450 volt shock after the learner had failed a multiple choice test on the word list, obedience would be negligible. This suggests that participants in Milgram's experiments were unable to break away from their prior obedience, which was obtained before they fully realised the implications of their actions. A compliance technique which makes use of this effect is the foot-in-thedoor technique, where making participants agree to a small request makes them more likely to comply with a larger related one compared to participants who are asked to agree to the larger request only. Other binding factors may have included the situational obligation caused by the agreement to take part in the experiment, and anxiety caused by anticipating the conflict that would arise when refusing to continue with the experiment. Buffers, i.e., features of the situation that increase the distance between actions and their consequences, for example, the physical distance between the participant and the victim and the shock generator that impersonalised cruel actions, also made it easier for participants to remain in the agentic state. Nevertheless, the longer the experiment went on for, the more difficult it became for participants to obey the experimenter’s orders. They looked for various ways to resolve the strain, such as avoidance and denial, showing minimal obedience, searching for loopholes, shifting responsibility to the victim, stating dissent and, finally, as an ultimate means, showing

disobedience. Interestingly, in the latter case people often offered to return the money they had received in order to free themselves from the commitment they had made. Milgram’s experiments were widely debated and strongly criticised by some authors. One major criticism concerned their internal validity, i.e., doubts about whether the experiments actually measured obedience. Orne and Holland (1972) proposed that most participants would not have believed that the electric shocks were real, because they knew that the experimenter would not risk health problems, let alone the death of a person. Also, participants might have thought that participants in psychological experiments are often not told the truth. However, during the experiment Milgram’s participants displayed a number of stress symptoms, which would have been difficult to fake. Also, in post-experimental interviews the participants typically indicated that they had believed that the shocks had been extremely painful for the victim. Another issue that may create doubt as to whether it was obedience that was measured is that the experimenter gave the participants inconsistent information on a number of occasions when he said that high-level shocks would not cause tissue damage although the respective switch for administering the shock was labelled as dangerous, hence creating demand characteristics and leading participants into a particular course of action. Some researchers have also pointed out that Milgram's sample of participants was not random, as he recruited his participants via a newspaper advertisement and paid them. Reicher and Haslam (2011) also feel that Milgram’s experiments don’t really demonstrate obedience, because • If the participants were obedient, they should have been more obedient the more the prods they received sounded like an order, but the opposite was the case; e.g., in Burger’s (2009) replication of the Milgram experiment all participants stopped after the 4th prod (“you have no choice but to continue”), which sounded most like an order (Burger et al., 2011) • When one examines the available evidence including Milgram’s (1974) analysis of ‘responsibility clock’ data, there is, according to Reicher & Haslam, little evidence to support the assumptions about the agentic state (see also Mandel, 1998) and it may be considered a rather reductionist construct, as it only focuses on the relationship between the participant and the experimenter, and not also between the participant and the victim. Instead, Reicher and Haslam (2011) propose a social identity explanation. Behaviour in Milgram’s experiments is assumed to depend on whether the participant perceives a shared social identity with the experimenter or the victim. Obedience levels fall as soon as participants are encouraged to stop categorising themselves with the experimenter, e.g., because of • spatial arrangements (e.g., the experimenter is absent) • the experimenter being less prototypical of the ‘science’ category (e.g., when the experiment is carried out in an office building rather than at Yale) • the experimenter imposing himself over the participant (e.g., with the 4th prod) • the participant becoming more aware of the victim (e.g. the victim’s first dissent at 150 volts or in the proximity variations) It was also doubted that Milgram’s results would be obtained in different contexts, i.e., there were doubts about the external validity of the research. Milgram, however, had already varied the context considerably and still found high levels of obedience. Also, field studies like Höfling et al.’s (1966) in a psychiatric hospital setting showed obedience in applied contexts rather than the laboratory. Furthermore, similar results were obtained not only in the US but also in other countries (see Blass, 1999, for a review). A recent partial replication by Burger (2009) that measured participants’

willingness to go beyond a shock level of 150V when the learner first protested found that 70% of participants were willing to continue as compared to 82.5% in Milgram’s original experiment (this difference was not statistically significant). Cultural norms and values have occasionally been found to influence obedience levels to some extent. In Kilham & Mann’s (1974) Australian study, the level of obedience was found to be considerably lower than in other countries. The participants were students and it was speculated that the student movement as well as the public discussion about the American massacres in Vietnam might have had an effect on obedience levels at the time. However, interestingly, Blass (1999) found no significant relationship between when replications have been carried out and the extent of obedience observed, i.e., there is no evidence that obedience rates have become lower over time. Mandel (1998) points out a number of limitations of Milgram’s obedience account when relating it to the Holocaust. In particular, Mandel criticises that Milgram considered obedience to be a sufficient cause of Holocaust crimes and did not take into account other motives such as the opportunity to achieve higher social status and career advancement. Also, Mandel points out the circularity of the obedience and the agentic state concept. Milgram uses the terms obedience and agentic state...