The Shocking Truth of the Notorious Milgram Obedience Experiments

By Guest Blogger | October 2, 2013 9:22 am

By Gina Perry

The original Milgram “shock box,” on display at the Ontario Science Centre. Image by Isabelle Adam via Flickr

It’s one of the most well-known psychology experiments in history – the 1961 tests in which social psychologist Stanley Milgram invited volunteers to take part in a study about memory and learning. Its actual aim, though, was to investigate obedience to authority – and Milgram reported that fully 65 percent of volunteers had repeatedly administered increasing electric shocks to a man they believed to be in severe pain.

In the decades since, the results have been held up as proof of the depths of ordinary people’s depravity in service to an authority figure. At the time, this had deep and resonant connections to the Holocaust and Nazi Germany – so resonant, in fact, that they might have led Milgram to dramatically misrepresent his hallmark findings.

Nazi Overtones

Stanley Milgram framed his research from the get-go as both inspired by and an explanation of Nazi behavior. He mentioned the gas chambers in the opening paragraph of his first published article; he strengthened the link and made it more explicit twelve years later in his book, Obedience to Authority.

At the time Milgram’s research was first published, the trial of high profile Nazi Adolph Eichmann was still fresh in the public mind. Eichmann had been captured in Buenos Aires and smuggled out of the country to stand trial in Israel. The trial was the first of its kind to be televised.

Suddenly the Holocaust was in American living rooms.  A procession of witnesses, most of them survivors, gave testimony while Eichmann, impassive and ordinary looking, looked on from inside a glass bullet proof box. Hannah Arendt, covering the trial, dwelled on his terrifying ordinariness, and famously coined the phrase the “banality of evil” to describe it.

Milgram stressed the connection between Nazi functionaries like Eichmann and the subjects in his lab. His findings appeared to demonstrate that ordinary people would inflict pain on someone else simply because someone in authority told them to.

In his 1974 book Milgram described how a person surrenders his will with that of an authority, entering an “agentic state.” For subordinates of Hitler in Germany and Stalin in Russia, this state was a “profound slumber” compared to the “light doze” of subjects in his lab, but the process was the same. Once a person merges with an authority who gives the orders, and enters the twilight zone of the agentic state, even though he might be doing inhumane things that are “alien to his nature” he feels “virtually guiltless.”

The trouble was that this zombie-like, slavish obedience that Milgram described wasn’t what he’d observed.

Truth in Numbers

The statistical story of the obedience experiments is not nearly as straightforward as you’d think. The 65% headline figure, of people who followed the experimenters’ orders and went to the maximum voltage on the shock machine, implies that there was a single experiment. In fact there were 24 different variations, or mini dramas, each with a different script, actors and experimental set up.

It’s surprising how often Milgram’s 24 different variations are wrongly conflated into this single statistic. The 65% result was made famous because it was the first variation that Milgram reported in his first journal article, yet few noted that it was an experiment that involved just 40 subjects.

By examining records of the experiment held at Yale, I found that in over half of the 24 variations, 60% of people disobeyed the instructions of the authority and refused to continue.

Then there are the methodological problems with the experiment. The highly controlled laboratory study that Milgram described actually involved a large degree of improvisation and variation not just between conditions but from one subject to another. You’d expect this to happen in the pilot phase of a study when the protocol is still being refined, but not once a study has begun.

Straying From Script

In listening to the original recordings of the experiments, it’s clear that Milgram’s experimenter John Williams deviated significantly from the script in his interactions with subjects. Williams – with Milgram’s approval – improvised in all manner of ways to exert pressure on subjects to keep administering shocks.

He left the lab to “check” on the learner, returning to reassure the teacher that the learner was OK. Instead of sticking to the standard four verbal commands described in accounts of the experimental protocol, Williams often abandoned the script and commanded some subjects 25 times and more to keep going. Teachers were blocked in their efforts to swap places with the learner or to check on him themselves.

The slavish obedience to authority we have come to associate with Milgram’s experiments comes to sound much more like bullying and coercion when you listen to these recordings.

Subjects’ Suspicions

Milgram went to great lengths with the stagecraft of his experiment, and was dismissive of subjects’ claims that they had seen through the hoax. Yet unpublished papers at Yale show that suspicion was alive and well among many of Milgram’s subjects (which is not surprising, given that Candid Camera was the most popular TV show in the United States at that time).

Subjects wrote to Milgram or called him afterwards to describe what had made them suspicious. Some commented on how the learner’s cries seemed to be coming from a speaker in the corner of the room, suggesting it was a tape recording. Others noticed the check given to the learner looked dog-eared and worn, an indication that it had been handed over many times before. The experimenter’s lack of concern for the learner and failure to respond to the learner’s complaints suggested there was nothing to worry about. Some subjects described how they had surreptitiously pressed switches of lower voltage but still the learner’s cries intensified.

But this skepticism of the subjects – whose belief in the experimental set up is pivotal to the validity of the experiment – has consistently been downplayed in discussions of the relevance and meaning of the results.

Poet Scientist

In the fifty years since publication of Milgram’s first journal article the obedience research continues to be cited as evidence of an enduring psychological truth: inside all of us is a Nazi concentration camp guard waiting to be called into service. Yet my archival research and examination of primary sources and that of other scholars contradicts this claim.

Milgram himself was privately aware of the methodological weakness of his research and struggled with many of the issues about the validity of experiments and their generalisability beyond the lab. Privately Milgram reflected that his work was more art than science, and described himself as a “hopeful poet.”

Poet or scientist, his determination to make a contribution to an understanding of one of the pressing issues of his generation led him to frame, shape and edit the story of his research for maximum impact. And while Milgram may have not measured obedience to authority in his lab his findings do offer us a powerful lesson: to question the authority of science and to be more critical of the stories we’ve been told.

 

Gina Perry is a psychologist and author of the book Behind the Shock Machine: The Untold Story of the Notorious Milgram Psychology Experiments, published in September 2013.

 

CATEGORIZED UNDER: Mind & Brain, Top Posts
MORE ABOUT: psychology
  • JonFrum

    Untold story? I seem to recall a debunking of the Milgram study done years ago. It was always obvious that this study was more advocacy than research. The lurid language used to discuss the results should be all the red flags you need.

  • One Smart Gal

    how about new studies using brain scanning tech on the train crash type study–redirect thr train or not, kill one or kill more by doing nothing? If this can show solid pathways for decision making Milgrim can be relegated to the bin of ancient bad science history where he belongs.

  • Buddy199

    Another urban legend.

    • APK247

      If you
      think Denise`s story is inconceivable,, 5 weaks-ago my friend’s cousin
      basically also got paid $5562 working 40 hours a month from there house and
      their buddy’s sister`s neighbour has been doing this for three months and
      easily made over $5562 in there spare time on there labtop. apply the
      instructions from this site…,www.Pro67.ℂom

  • John McIntire

    Great article, thanks. This leaves a bit of a glaring gap, it seems, in that psychology then still has no good idea or explanation as to how/why the Nazis were able to carry out what they did on such a monumental scale with so little apparent resistance from within.

  • Colton Johnson

    The article provided some interesting insights, but failed to account for several key factors in the test. Of note, Milgram tested for proximity of the shocker to the learner. Some of the levels include, “in the same room,” “in another room but with audio connected,” “in another room and no audio.” He found that the farther the shocker was removed from the learner, the more likely people were to give lethal shocks.

  • Dimitri Ledkovsky

    So are we to believe that the reported horrors of Guantanamo are also staged exaggeration?

  • johnwerneken

    Some people emphasize self and/or the now, others community and/or the future; most in between somewhere. There are choices and a range of behaviors.

  • http://neuroautomaton.com/ Zachary Stansfield

    Is this an accurate description of what went on in those experiments? Because it rather grossly deviates from Milgram’s own report (which made no mention of any “off-script” activities) and clearly implies that the entire setup was a fraud. It would be nice if Discover would point its readers to a digital copy of the original recordings of these experiments, as alluded to in this excerpt.

    Having read the published paper multiple times, I would note that the main issues, with the actual report, were more mundane: like the fact that the confederate “getting shocked” didn’t complain at all until the shock intensity read “very severe” (the first shock level at which Milgram reported any discontinuation), or the clear fact that the researchers employed strong coercion within the official protocol.

    Additionally, it’s not clear to me why laboratory experiments should be expected to replicate Nazi Germany. What a gross over-estimation of the powers of psychologists–a point of hubris on Milgram’s part, to be sure.

    • Moonshadow74

      There’s a link in the article to wikipedia that gives footnotes for references.

  • Cole Davis

    Milgram-type experiments have been replicated over different eras and in different countries.

    • http://blogs.discovermagazine.com Ron Hansing

      MORE RECENTLY THE STANFORD PRISON EXPERIMENTS… NOT TO MENTION, INTERVIEWS WITH PRISON IMATES AND GUARDS.
      THEY HAD TO STOP THE EXPERIEMNT BECAUSE SOME THE “PRISIONERS” WENT BONKERS.
      Also, anadotal information such as the wounded dog syndrome…and Schadenfreude. How many of us had glee in our hearts when tony Weiner, the NYC mayor candidate was crucified. and not to mention, in the Bible… “those without sin throw the stones.” The brain scan train test could just as well as comparing apples and oranges. Also, we are herd animals… we have a genetic urge to follow… instructions, just like cow, led to the slaughter house.Or the Jews to the ovens.
      Lastly, the conclusions is not black or white… but a strong suggestion. hence, a theory… and theories need to be continually refined, supported, sometimes rejected… (scientific expertise) vs scientific authority… (the case is closed, no mor ediscussion needed. ie, global warming.
      So lets just leave it at this point… Can we ever reach an absolute when dealing with biological varibility?
      This article is probably based on a recent book that the author had an agenda… The reality is that the facts speaks for themself.

      • Cole Davis

        Let’s just leave it at this point? But you’ve got the wrong series of experiments. The Stanford experiments were by Zimbardo. We are talking about Milgram. What is the point of pontificating if you have only the faintest clue of what you are talking about?

        • setarcos89

          I believe this individual was attempting to add other studies whose conclusions parallel those of Milgram.

          • Cole Davis

            Well then perhaps that individual might like to know that the Milgram experiments have been replicated over subsequent decades and in various countries and the results are pretty much the same as those from Milgram’s studies. Re Zimbardo: Yes, Stanford experiment also showed how ordinary people could become influenced by their environment and by being on one side of an authority divide (in this case, ‘guards’ and ‘prisoners’ allocated roles from a random sample of students). The fact that Zimbardo’s experiment got out of hand does not detract from what it revealed and it certainly does nothing to detract from Milgram’s epic work.

            I know that the writer of the article has a book out. It should be noted that her book is not reputed to be particularly objective. Having found some methodological holes in his work – which she was right to report about – she has tried to monster Milgram altogether and to detract from the importance and the validity of his findings.

  • JoeM

    So what if the experiments weren’t the greatest by today’s standards. Do you have trouble believing that many people in society obey orders because they get a sense that it’s ok if they see everyone else doing it….and/or a superior told them some reason as to why it’s the right thing to do. I would argue most of us need only reflect on our own lives at different times to see how we are influenced. I believe the majority of people act like sheep, waiting for some herder to follow as to right or wrong…(as opposed to actually thinking independently). I see it all the time!

  • http://blogs.discovermagazine.com Michael Rodriguez

    The point that many posters here are missing is that the conclusions of a procedurally flawed experiment simply must be dismissed – especially one as deeply flawed as the Milgram experiments apparently were. This experiment absolutely reeks of bias. Milgram clearly twisted the experiment to produce the results he wanted and failed to publish either an accurate description of his procedures or many of the observations that were contrary to desired result. For those of you who think Milgram is being unfairly judged, would you take a medicine that was tested using such methodologically flawed experimental procedures – for instance if the receivers of the placebo were informed when they reported improvement of their condition?

    • Cole Davis

      This is not in fact true. You are taking the blogger’s word for it.

  • P.I. Staker

    If Milgram wanted to test the ‘Germans are Different’ hypothesis, then why did he conduct the study in New Haven, USA and not in Germany because if he did carry out the experiment in Germany, then he will have a really good idea into obedience levels in Germany. You cannot conduct the study in America and then use this data to compare levels of obedience in a country that is thousands of miles away never mind the rest of the world.

    • SouthernGal

      … because he wasn’t trying to test a “Germans are Different” hypothesis? So he didn’t need to do it in Germany?

      Is that really non-obvious?

  • Esdavis

    This article is earnest, but I am concerned that it falls into the vast category of simple rationalizations to explain away things we don’t like about human behavior. Life is full of bullying! Bosses have power over their subordinates, sometimes the power to destroy their career and decimate their family’s income. Police have power over citizens. Department chairs have power over researchers. To assert that improvised bullying outside of the script destroys the validity of this scientific experiment strains my ability to ignore the writer’s biases.

    There are some valid criticisms of the scientific processes used in Milgram’s studies, but non that I can see as challenging the validity of the outcomes. While acknowledging that the studies would not be able to be performed today under our current research ethics standards, this does not undermine the research, in my opinion.

    One would expect the research subjects to dislike being characterized as weak-willed. One would expect them to rationalize their terribly unattractive and humiliating behavior in every conceivable way. And I can understand this as normal, given that their consent to be filmed in the research would not stand muster today.

    So, does the research, itself, offend the sensibility of a modern researcher trained in our current iteration of the ethical treatment of the subjects of scientific research? Of course! But, that does not make Milgram’s research any less revealing of human nature, of our ability to do terrible things under various forms of peer or authority-driven social pressure.

    Milgram lived in an era in which many Nazis defended their horrible actions with, “I was just following orders.” He offered important evidence to shed light on how atrocities could be perpetrated by anyone, from any nationality. He showed that Germans were not uniquely prone to atrocities. Americans have participated in our share.

    So, my request to the writer is to step back from the reductionist realm in which Milgram violated our contemporary standards of research ethics and look at his experiments and what they reveal. There is much of vital import here, and it should not be suppressed by a modern day witch hunt. Trashing the important work of Milgram may yield a book contract, but that is an economic matter, not a scientific matter.

  • GABA

    hi, very interesting, loving the blog dudes!!

  • Bryony Bratchell

    I think something people forget is that Milgram wasn’t just trying to prove or disprove that people would blindly follow orders from an authority figure, but that he was also trying to determine what characteristics and factors were essential for the subject to obey. So the 65% obedience rate in the simplest form of his experiment is in fact the case, but the much lower levels of obedience in the other conditions move towards identifying specific factors at play, rather than supporting/not supporting the claim of the effect of authority. He looked at factors such as proximity, physical presence of the authority, double authority figures, changed status of authority etc. So no, the obedience figures are not replicated through all his studies, but that’s kind of the point! He was systematically going through possible additional factors which effect obedience, to discount them. I’ve read his book and this is clear, and although I am not saying that his experiments were perfect, I just think some people have not completely grasped the significance of the lower obedience levels in some conditions. Also, I can’t seem to find any references to support the claims in this? Where is the information about subjects’ suspicion coming from?

NEW ON DISCOVER
OPEN
CITIZEN SCIENCE
ADVERTISEMENT

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

The Crux

A collection of bright and big ideas about timely and important science from a community of experts.
ADVERTISEMENT

See More

ADVERTISEMENT
Collapse bottom bar
+

Login to your Account

X
E-mail address:
Password:
Remember me
Forgot your password?
No problem. Click here to have it e-mailed to you.

Not Registered Yet?

Register now for FREE. Registration only takes a few minutes to complete. Register now »