Who’s Fooling Who?
Rod Dreher refers to the famous Milgram experiment in which subjects were led into a mock experimental laboratory and told to press a button that supposedly delivered increasingly severe electrical shocks to other “subjects” (who were actually other researchers) at the instruction of the scientists running the experiment. It was all a set-up: no real shocks were delivered, but the subjects believed that they were delivering up to lethal voltages to other humans. Milgram was shocked (ha ha) to discover that a majority of subjects were willing to deliver all the shocks.
His interpretation of the experimental results was that Americans would have willingly participated as employees in Nazi death camps. Rod cites Ron Bailey as saying that our institutions deserve the credit for preventing Nazism or it equivalent form reaching our shores; Rod makes the point that Anglo-Protestant culture has something to do with it as well. Both are correct, in my view.
But what’s always struck me as ironic about the typical interpretations of the Milgram experiment is that the subjects did not actually deliver painful shocks. Now, you will say, yes, but they thought that they did, which indicates that they would be willing to do so. But I’m not so sure about that.
After all, they were making a (likely unarticulated) judgment about whether an action was “right” in the context of what presumably seemed like legitimately professional scientists at a major research university in the middle of a law-bound, peaceful republic. Operating in the way that most decision-making under pressure works, which is not necessarily linear processing of logical statements, they likely thought something that could be crudely represented as “Well, they wouldn’t be allowed to do something that was really heinous, and would destroy their careers if they did. I don’t understand exactly what’s going on, but all my assumptions about how the world works would be violated if Yale University could really run a torture chamber operated by random people picked off the street. It would just be too crazy.” This muddled-headed, superficial thinking turns out to have been a correct judgment.
If you took these exact same people, formed by the institutions and culture of contemporary America, and had them live in an America that within a few years had been taken over by a dictatorial regime premised on ethnically cleansing a good fraction of the population, and then had them go into a chamber with a plaque saying “work makes you free” populated by scary guys in military uniforms, they might or might not go along with electro-shocking people strapped to chairs; but if they did, it would most likely be out of a fear for personal safety.
milgram understood that objection and so he replicated the experiment in an ordinary rented office space and did not identify himself as associated with Yale (or anything comparably prestigious). the acquiescence rate went down, but not by that much.
— Gabriel Rossman · Jan 13, 10:47 PM · #
Thanks for the informed comment.
I’d say that’s actually consistent with my interpretation, since one factor that would lead this to be in violation or normal assumptions was dropped, but not many of the others, you would expect a reduction in compliance commensurate with that.
What if you did a series of experiments that started with an extremely unlikely environement for torture (I was going to say, doing this in the White House, but maybe not so much) and proceeded through environments of increasing likelihood of being places and situations in which torturing people would happen – ending maybe with a set up of mock gangsters interogating a “rival”. My prediction would be lower and lower rates of compliance. Unfortunately, I assume no IRB would ever approve the Milgram experiments today, but who knows?
— Jim Manzi · Jan 13, 11:11 PM · #
He also had the experiment repeated in several different countries and in widely varying conditions. And the results were never reassuring. But you could still argue that in each of these cases what was functioning in the experimental subjects was a background belief that “things like that don’t happen here.” Of course, that may not be reassuring either.
— Alan Jacobs · Jan 13, 11:11 PM · #
Alan:
Yes, I’ve read (casually and non-profesionally) about the replications in other societies. A couple of quick points. One, is that (from my limited reading) these have happened predominantly in other similar societies, rather than, say, 1970s Chile or contemporary Iran. Second, It may also be that people generally, including those currently living in the US, would become habituated to this (or maybe, their children would after growing up in a dictatorial regime), but that if you took the actual people now living in the US and after a short transition to dictatorship or quasi-dictatorship (as per the thought experiment), they might or might not be willing to go along for reasons other than fear. The reason I think this second point is significant if it is correct is that this attitude forms another bulwark aginst a transition to tyranny.
— Jim Manzi · Jan 13, 11:17 PM · #
Jim, I’m afraid you’ve thrown out this objection to the experiment that’s been addressed a lot in discussions of it. Like Gabriel said, the experiment has been replicated in less prestigious circumstances. Moreover, the experimental subjects showed extreme distress in the original experiment, often while continuing to deliver the shocks, which strongly suggests their actions felt real to themselves.
Lastly, as the wikipedia page mentions, there was a replication done with a puppy actually being objected to shocks in front of subjects.
— Justin · Jan 13, 11:19 PM · #
That’s always a huge problem with doing experimental studies with students along the lines of psychology, economics, etc. – the students try and guess the game they are being asked to play. If you play the confess/don’t-confess game with two students, many will already know quite a bit about what you are testing for.
I don’t think that’s relevant here though – it’s not clear why proximity to the research “lead” would have explanatory power if the testee was just following social cues, and assuming medicalized research knew what it was doing. Or that they thought that the whole thing was suspicious on its face.
— rortybomb · Jan 13, 11:37 PM · #
“His interpretation of the experimental results was that Americans would have willingly participated as employees in Nazi death camps. Rod cites Ron Bailey as saying that our institutions deserve the credit for preventing Nazism or it equivalent form reaching our shores; Rod makes the point that Anglo-Protestant culture has something to do with it as well. Both are correct, in my view.”
Just as an aside: It is my understanding that german and “anglo-protestant” culture are very similar and share the same roots.
— cw · Jan 14, 12:14 AM · #
Mr. Manzi-
You left out the funnest part of the experiment, namely, that the subjects giving the shocks were not simply told that they were giving increasingly-intense shocks to the subjects, they were given a loudspeaker so they could hear the “subject” in intense pain and beg for his life, and still the subjects gave the shocks.
“If you took these exact same people, formed by the institutions and culture of contemporary America, […] but if they did, it would most likely be out of a fear for personal safety.”
Such a motive is still sufficient, as a sonderkommando would tell you. And cultural-authoritative “this must be the best way because the boss says so” reasons were quite sufficient for people like Rudolf Kastner. These people weren’t parachuted into the situation; they’d been living in and around the Nazi regime for a decade, and the party and its international affiliates had been a reality in central Europe since the end of World War 1. And, as I think Arendt demonstrated, most of these movements’ despicable attitudes towards the other can be attributed to the way colonial powers had historically treated their subject populations, which had been happening for at least a century previous, using most of the same racial-theoretic justifications the Nazis did.
I guess my point is that the attitude that “it can’t happen here” is very easy for people to find when they need it, because in practice it was very difficult for even well-informed people to see how the whole culture of Europe had been ever so slightly bent sinister over the course of decades. For example, I don’t think people are being mass-murdered at Guantanamo, but I can see that if you were to take the justifications and reasoning behind Guantanamo and apply them to the body politic in general, we could have problems.
— Jamie · Jan 14, 01:11 AM · #
Justin:
I’m not saying that the experience didn’t “feel real” to the subjects, and was explicitly not saying that they rationally deduced that this must be a set-up, but instead that using a manner of thinking that is not linear processing they “weighed the odds” (speaking metaphorically) that this was exactly what the experimenters wanted them to think it was. This would be consistent with a high degree of stress and subsequent distress.
I think that the experimenters implicitly had an overly simplistic view of human decision-making that led the researchers to an over-interpretation of the finding.
— Jim Manzi · Jan 14, 02:24 AM · #
You can always find people willing to hurt other people. When I lived in the Chicago area in the 1980s, the police routinely tortured confessions out of murder suspects.
— Steve Sailer · Jan 14, 04:48 AM · #
Jim,
There’s a deep logical difficulty in your position. Germans during WW2 were broadly behind Hitler. There was a war on, their boys were fighting, etc. Those Germans who were in the Army and civil service – most of whom resisted Nazification, and whose leadership predated Nazi rule – enthusiastically backed the Nazi leadership and that includes tripping over themselves to help carry out the Final Solution.
That civil service, and much of the bureaucracy, and certainly the business class stayed intact after the war. Somehow we were allies with West Germany and quickly came to see them as a peace-loving enlightened western democracy.
A huge percentage of white America is of (at least partial) German descent. Let’s say it’s Germanness that led people to let down their defenses and back Hitler – how do you reconcile our acceptance of 60’s or 70’s or 80’s West Germany as an ally? What about modern Germany? And unless you’re a pure behaviorist, much of white America is prone to these “German” moral failures.
No, the real answer is we can’t know for sure.
But I can’t imagine how you can watch Torturing Democracy and believe that given the right systemic rot regular people wouldn’t adapt and happily, say, send Japanese people to interment camps, or tell themselves that wiping out the Indians isn’t so bad, or enslaving black people is ok.
We can all sit here in a secure peaceful society pursuing happiness and say we wouldn’t do these things. But after reading about the types of people who went along with Nazi policies (pretty much everyone ruled by the Nazis except the Danes), seeing interviews with the American soldiers in Torturing Democracy – these are all good people in peacetime. I can’t say for certain that I wouldn’t be vulnerable too, given the right conditions.
What we can say from these case studies is that the leadership at the very top is what creates the systemic conditions for either brutality or resistance to it. We don’t have the legal concepts for dealing with a “systemic” crime – but it’s clear that though individuals who committed actual harm should be prosecuted, the real harm is the creation of an environment where “normal” moral judgments are turned on their heads.
— Steve C · Jan 14, 07:04 AM · #
One question I have about the methodology of the experiments is how to control for the charisma and persuasiveness of the person ordering the subjects to torture the victims. Leadership charisma varies gigantically among individuals. Con men, for instance, can talk people into hurting giving away money, which is at least as hard to explain as hurting strangers.
— Steve Sailer · Jan 14, 09:09 AM · #
To riff off cw’s aside: I’m no historian, but I am pretty sure it is wrong to say that Anglo-Protestant and German Protestant cultures are that similar, at least as far as the last 1000 years are concerned. Politically, it’s the difference between the Magna Carta and the Kaisar. Philosophically, it’s the difference between the determinism of Baron d’Holbach and other continentals, and the liberalism of Locke and Mill. Religiously, the history of German Protestantism and its English counterpart are very different, and even today Anglicanism has more in common philosophically and ecclesiastically with Catholicism than most other Protestant churches.
— Blar · Jan 14, 04:33 PM · #
Jim, interpretations abound, and yours is certainly one way of looking at it.
Here’s the frame: A person submits into an authority/subject dynamic, as subject, with no threat of material loss if he fails to finish the activity. What kind of tasks will the subject be willing to perform?
At the first level, we know from everyday experience that a subject will perform simple activities — menial tasks which have no moral salience — without much if any reflective delay, tasks for which they have no motive but for obedience to authority.
The question is, in what circumstances will the subject’s “conscience” — however defined — inhibit the performance of an authority sanctioned/mandated task?
Several variables can be tuned, one being the perceived qualities of the authority (i.e., your point). Another, which Jamie noted, is the kind of feedback you get when performing the task. The most obvious is the type of task involved, and that is the meat of Milgram’s experiment.
Milgram chose a task which ceteris paribus a person’s moral machinery would inhibit him from performing on innocent others. And sure enough, during the experiment the subject would usually pause and express doubt or unwillingness to continue. To eliminate effects of charisma as much as possible, the experimenter was given a set of four standardized “prods” to get the subject to continue, always delivered in order.
Jim, where your interpretation falters is here, where you write:
That is muddle-headed thinking, but that wasn’t the subjects’ judgment. The subjects were given a legitimate pretext about negative reinforcement and learning, were told that no burning or blistering would result, were told and indicated belief that the shocks would be extremely painful but “cause no permanent tissue damage” — and they bought it. Utmost care was taken that the subject believed in the authenticity of the shock generator. Everything looked serious and expensive. During the experiment subjects were continually reminded of the increasing pain they were administering. In a post-experiment interview, subjects were asked to rate the painfulness of the last few shocks they administered. On a 14 point scale which ended in “extremely painful”, the modal response was 14 with the mean at 13.42.
Even more telling is this little bit from the study:
And yet, no subject stopped prior to administering Shock Level 20 (the border between intense and extremely intense shock, at which point the victim kicks on the walls and no longer supplies answers to questions). 26 of the 40 went to Shock 450, two steps beyond the designation DANGER: SEVERE SHOCK. (Before the experiment, an informal poll of psychology seniors predicted at worst that only 3% would go to the end, with the class mean at 1.2%, so the results were indeed shocking.)
The important point to take away is this: the State is more convincing and compelling an authority than a University (however prestigious the latter is), and native culture causes a significant difference in results.
— JA · Jan 14, 05:51 PM · #
I’m still not reassured that personal charisma doesn’t play a sizable role in this. Consider two movie stars — Liam Neeson and Steve Buscemi. What if you did the experiment twice, one time with an authority figure in the experiment radiating the calm benevolent assurance of Liam Neeson doing the voice of the lion in the “Narnia,” the other time with the authority figuring being a sweaty little weasel like Steve Buscemi in “Reservoir Dogs”?
— Steve Sailer · Jan 14, 07:42 PM · #
The other question about the methodology is self-selection. How did they get a representative sample of the population? All the statistics presented above are just for people who sat down to shock some poor bastard of a stranger. How many people didn’t show up for the experiment or bowed out at an earlier point?
— Steve Sailer · Jan 14, 07:51 PM · #
Anyway, the point is that, however credible the statistics generated by the Milgram experiment, a government or organized crime outfits can always find enough men to torture people. Police work seems to attract a certain number of guys who would be willing to do what it takes. I doubt if you could easily persuade the same percentage of firemen to do it, though. Policemen and firemen are pretty similar on a lot of dimensions, but not on measures like wife-beating, which I suspect has predictive value for the question of whether they’d torture people.
Conscripted soldiers (i.e., average men) can be trained to do very bad things, but it helps to put them through a huge amount of training and then have severe punitive measures (e.g., firing squads and blocking battalions) in place to keep them from shirking.
In particular, war helps justify atrocities. The Holocaust would probably have been unfeasible outside of wartime.
— Steve Sailer · Jan 14, 08:01 PM · #
Returning to Hitler, offstage, Hitler was a sort of twitchy little Steve Buscemi character, although a little more effeminate than most roles that Buscemi plays. Nazi Foreign Ministry officials often apologized to visiting dignitaries in advance for Hitler being so unprepossessing in person. They would reassure the foreign visitors that when he got up to speak on stage, he turned himself into a hugely charismatic authority figure.
— Steve Sailer · Jan 14, 08:12 PM · #
Steve Sailer, on population selection, there are strict statistical and selection criteria to try and minimize selection-bias. That is where the cross-cultural and cross-class replication studies become important. Given that this study is replicated multiple times a year with many, many variations and varieties of complexity, with similar results, it carries heavy weight.
Similarly, with regards to your “charisma” question, you can actually view film of the original Milgram sessions. One of the most frightening things is that very little “persuasion” is actually needed: the “authority figure” simply repeats the instruction in a calm tone of voice. Many, many subjects comply, even after hearing screams of pain and protest from the person being “shocked.” Indeed, some continue to deliver shocks even after the “shocked” individual ceases to respond after letting out a shriek.
Profesor Bob Altemeyer’s book, “The Authoritarians,” offers further thoughts on this.
— James F. Elliott · Jan 14, 09:46 PM · #