Revisiting Milgram's obedience experiment: what did he actually prove?


Well, shucks. Milgram is taught to every Psych student in the world, including this superannuated one.

I recall that his results were replicated a few times, including by a very unethical researcher in (I think) Egypt, who used children. I wonder if these studies also exhibit the same flaws?

1 Like

Well, color me shocked.


So…humans not as totally scummy as we thought?


At least the Stanford Prison Experiment is still real, right?

1 Like

Even if the experiment is discredited or disproved. It did get people talking and has been used as a inspiration for other experiments. Experiments don’t really succeed or fail, they just test assumptions, theories, and hypotheses.


I was a subject in this experiment performed at the University of Missouri-Columbia in 1975. Subjects were offered extra credit in the intro Psychology class. I am proud to say that I was disobedient. Once I was shocked once, I didn’t want to do it to the other person. To me that makes it more of an experiment in empathy than obedience.


No matter the validity of Milgram’s experiment, I’d still do it.

1 Like

Caveat: I haven’t read the book Cory’s referencing. However, the allegations he mentions are not new, and when I studied the matter extensively years ago I found them very unconvincing, to say the least.

I will give you Milgram’s own response: “Orne’s suggestion that the subjects only feigned sweating, trembling, and stuttering to please the experimenter is pathetically detached from reality, equivalent to the statement that hemophiliacs bleed to keep their physicians busy.” (1972)

Milgram’s experiment is repeatable and has been repeated.


Radiolab devoted an hour to Milgram last year. I haven’t read the book discussed here, but this synopsis makes it seem rather narrow-minded and dumbed-down compared to the Radiolab episode. It has been awhile since I listened to it, but as I recall they said it is Milgram’s first set of experiments that get all the attention despite the fact that he did many, many of them over a very long period of time. Each time – and again, I’m going on memory here – he would change some of the parameters, and very frequently a seemingly small change would make a big difference in the experiment’s results. For example, people behaved quite differently if the person telling them to administer a shock was in the room with them, as opposed to being a disembodied voice over a speaker.

I’m probably not doing the Radiolab episode any justice. It should be available as a free podcast on their website, if anybody is interested…


Incrementally. Small increments.

It significantly decreases compliance if you do anything that establishes a connection between the shocker and the shockee, or do anything that gives the shocker a model for non-compliance. You could almost certainly decrease compliance by something as simple as putting up a poster of Ghandi in the room (although I don’t remember anybody trying that particular variation).

Milgram showed that most people do what they are told, even if it’s wrong. However, he also showed that people are just as easily encouraged to do the right thing, by simple methods like the one you mentioned, or just by having accomplices refuse to comply. This tells us that we have an obligation to stand up when nobody else is doing so; even if you have to eat the first bullet, you can still inspire others who would otherwise support evil.

1 Like

Isn’t the sample group being self-selected from a fairly narrow demographic a flaw also?

In all the time I’ve heard Milgram brought up, I’ve never heard of, nor considered the idea, that some of the participants saw through the ruse.

If you are in on the “joke,” you would pull the “X X X - 460 VOLTS” switch every time, just to see what the actor would come up with.


As I’ve mentioned before, I was one of the student archivers of the Milgram tapes at Yale, working under the actual librarians in charge of the project. There were a number of conditions that Milgram divided the experimental protocol into, and not all of them had the strict script of four prompts. Not having read Perry’s book, I can’t say that this is what she is talking about. But merely having the existence of these multiple researcher scripts on tape doesn’t mean Milgram was faking it. Regarding the other point–of people knowing that it was a set-up–this too is pretty obvious from the tapes. There was nervous, inappropriate laughter, but there were also some people who laughed and didn’t sound nervous–it sounded like they knew that they weren’t shocking someone.

But I’m not sure how you would separate the believers from the non-believers, given that once you asked someone who crossed the line and shocked another person into unconsciousness “Did you think it was real?” there are going to be people who claim they knew it was a ruse to cover up their behavior.


I hope it’s true, and not an urban legend, but I’ve heard somebody did an inverse experiment, using simulated pleasure instead of pain… and the subjects balked much, much faster. But according to Perry, Milgram’s own subjects did too, so we’re back to square one.

Time to reproduce both experiments. Do I hear any volunteers?

1 Like

He proved nothing, literally.

If anything, he proved buttons are attractive, and we’ve known that for centuries…Next. :wink:


Really? Rubber, anyone?

To bounce.


Every time I go to the DMV I find confirmation that humans will inflict misery if they think everyone else is doing it. I don’t want to consider the possibility that they are all just toying with me thinking it is a big joke.