It Is Time to Reevaluate the Experiment of Milgram

Akihiro Eguchi


During the trial of Adolf Eichmann for war crimes he committed in the Holocaust, his defense was consistently that “I was just following orders” (Weiten 658). Stanley Milgram was a psychologist who constructed an experiment to find what causes the obedience that made Eichmann and people around him do the immoral tasks. The result of the experiment amazed people because the human’s tendency for obedience under authority was much more than experimenter predicted, and this posed a question about the nature of obedience which people had never thought of before. However, according to an article “Reflections on ‘Replicating Milgram’” by Arthur G. Miller which reviews Jerry Burger’s 2009 experiment, “the obedience research had stimulated […] ethical and methodological controversy” (26), and uncountable attempts to replicate the experiment have been rejected because of the controversy. He talks about the importance of continuing the research Milgram started.

On the other hand, in her article “Psychology in Action,” Diana Baumrind has doubts about the validity of the experiment and argues that the experiment was unethical because it tried to force the subjects to follow absolute directions in order to estimate the degree of obedience. However, despite the ethical problems discussed by Baumrind, more experiments should be conducted because many problems caused by obedience, such as terrorism, could be avoided by seeking knowledge through more studies. Although ethics in research is very important, ethical concerns should be flexible in order to solve many problems for our common good. Milgram’s assessment is still correct in terms of validity, effects, and ethics.

In his article “The Perils of Obedience,” Milgram describes the experiment conducted at Yale University to identify the factors which cause ordinary citizens to follow orders to hurt another person even though it goes against their ethics. In the experiment, the two participants of the study were told that it would measure effects of punishment on learning, and each of them was designated either “teacher” or “learner” by the experimenter. Actually, the person who became the learner was an accomplice of the experimenter, and the participant always became the teacher. The teacher watched as the learner was tied up to an electric chair and was told that he would receive electric shocks of increasing intensity if he made an error on memory questions.

The teacher was taken to a different room and ordered to ask memory questions to the learner through a microphone and shocked the learner according to the answer he or she received, making the voltage higher and higher up to 450 volts. The higher the voltage got, the stronger the protests the teacher heard from the learner in the next room. When the voltage was increased to around 300 volts, the teacher heard no further responses from the learner. Nevertheless, the teacher was ordered to continue shocking regardless of what happened to the learner in order to see if the teacher follows the orders. As a result, although many people showed internal conflicts in proceeding with the experiment, around 60 % of the subjects obeyed the orders of the experimenter to the end, apparently harming the learner until they reached the highest shock on the board. This experiment also observed people’s peculiar behaviors such as nervous laughter under extreme conditions. Milgram described the experiment using the phrase "banality of evil" (Milgram 324) which expresses the fact that even ordinary people would act unbelievably sadistically in certain conditions even though they thought they would not.

After Milgram’s report of the experiment, people started arguing about the validity of the experiment. While some people like Miller praise his work, some of the other people like Baumrind doubted the validity of the experiment. She thinks “his experimental situations are not sufficiently accurate models of real-life experience” (332). First of all, she points out the problems of the anxiety and situation. She believes that “the interpersonal relationship between experimenter and subject additionally has unique features which are likely to provoke initial anxiety in the subject” (330). In addition, she assumes that the situation, which is in a laboratory, would make people behave passively. She also mentions that in an experimental situation, everyone usually behaves in obedience; therefore, she thinks the Milgram’s experiment is not valid.

On the other hand, Miller thinks it was a valid experiment because Baumrind’s criticisms proved the purpose of Milgram’s experiment. Miller introduces the term “agentic shift” which is the tendency of people to lose their personal responsibility in certain situations. Therefore, he writes there is no problem with the experiment since it includes “a fear of defying authority, a hesitancy to ‘make a scene’ and embarrass oneself or ruin the project” (25). He also mentions that “obedience could simply be a reasonable reaction to a seemingly legitimate source of influence in a highly ambiguous setting” (25); therefore, a place like the laboratory is very suitable for the experiment. Thus, although Baumrind has doubts about the validity, pointing out the same problems, Miller believes these qualities are crucial for the experiment. It seems like Baumrind worries about the anxiety caused by the experiment too much thus ignoring the main purpose which was to observe the obedience. Although her argument seems to be logical, it should not replace the original purpose of the experiment; otherwise the outcome would be spoiled.

Moreover, in her article, Baumrind sees the effects that participants would experience as another problem for the experiment. She indicates the possibility that the unstable tempers observed in the experiment could cause the loss of participants' self-esteems and faith in authorities (332), and she doubts if Milgram really took care of the subjects after the experiment. However, after Baumrind wrote the article in 1963, Miller notes in 2009 that there has been no report of negative effects of the participants in the experiment (23). Therefore, there seem to have been ultimately no serious effects on the participants of the experiment.

Miller also shows the different experiment conducted by Burger in 2009 according to Baumrind’s recommendations. Burger is one of few experimenters who tried to replicate Milgram’s experiment by making two significant changes from original experiment “to minimize the likelihood of unacceptable levels of stress, tension, or harmful after effects in participants” (22). By reducing the maximum shock to 150 volts and using multifaceted screening of participants, Burger obtained data that looks similar to Milgram's experiment. Miller praised the experiment attempting to replicate the experiment; conversely, he points out that this outcome does not show some important aspects of Milgram’s experiment such as nervous laugher because it does not test the obedience under extreme conditions. He states that "only further research can provide the convincing means to sharpen theoretical understanding of destructive obedience” (21). Therefore, this conflict shows that Milgram could have followed a more humane way like Burger by accepting many ethical modifications; however, the important discovery of how people respond to extreme stress and under stronger orders would not have been gained.

Finally, Baumrind insists on the importance of the ethics of the research, and Miller’s article shows that her hope has come true because most of the replications of Milgram’s experiment have been rejected since 1963. Baumrind argues that psychologists must conduct research very carefully by giving priority to ethical concerns before pursuing the outcome (334). Also, she shows her strong unwillingness to see the recurrence of Milgram’s error to disrespect subjects’ rights. Because many people have similar opinions to hers, the American Psychological Association has laid down that researchers have to clearly announce the subjects' right to quit the study before the experiment begins (Miller 20).

In contrast, however, Miller shows his discomfort about the situation: “because of the need to modify the paradigm for IRB (Institutional Review Board) approval, it is ‘good’ that intense displays of stress and emotion were not observed in this study” (24). Therefore, he laments that any further study of obedience since Milgram’s experiment was implicitly stopped. The strong ethical concerns make people hesitant to take further study in this field. This struggle shows the very important balance between ethics and science: we cannot go so far with ethics that no new information is found, and at the same time, we cannot ignore ethics to the point that people are harmed.

Miller brings up the danger of “the rise of terrorism and the use of torture” (21) which is becoming one of the biggest issues of society. Reviewing the recent trial of modified paradigms by Burger, Miller insists on the importance for studying about obedience again. Although the study about obedience has been avoided because it reminds researchers of the ethical problems caused by Milgram's experiment, the most important thing for researchers is to seek the best way for solving problems in the world by continueing to do what they can, but accepting changes to satisfy ethics (Miller 23).

There is no rigid answer to which is more important: protecting ethics or seeking science. In 1963, Baumrind pointed out problems about the validity and ethics of the study of obedience conducted by Milgram, but almost 50 years has passed since then. Miller suggests reassessing Milgram's achievement by reviewing the new experiment conducted by Burger in 2009. He talks about the difficulty of replicating Milgram’s experiment because of ethics, but Miller praises Burger’s attempt to replicate the experiment by keeping a balance between ethics and the study even though it lacked some features of original experiment. In the present age surrounded by many threats, the new discoveries about obedience may lead to solutions. It seems to be time for re-evaluating the obedience experiment and finding a new approach which implements both ethics and a significant outcome.

Work Cited