Imagine yourself living in New Haven, CT, in the summer of 1961. You see an ad in the local newspaper, asking for volunteers for an experiment on Memory and Learning, to be conducted at Yale University. The ad looks a little too flashy for Yale, you think. But – well, hell, why not? So you send in the coupon. A couple of weeks later, you get a phone call. This young man says he’s calling for the Memory and Learning Project at Yale, and he wants to know whether you can come tomorrow evening at eight. How about nine, you say, since you eat dinner at 7:30. He says okay and gives you directions. The next night, you drive through the fancy arch by the art gallery, right into the heart of the Old Campus, and park your car outside one of those old stone buildings, Linsly-Chittenden Hall. You stroll through the first floor hallway until you see another guy, middle-aged uncle type, bumbling around like he doesn’t know where to go. “You looking for the memory experiment?” you say, and when he says yes you point to what you think is the right number: “This door here, looks like.”
[Originally published as Chapter 4, Sections a and b, of Alan C. Elms, Social Psychology and Social Relevance (Boston: Little, Brown, 1972)]
two experimental scenarios
Myr-mi-don. 1. Class. Myth. one of the warlike people of ancient Thessaly who accompanied Achilles, their king, to the Trojan War. 2. (l.c.) one who executes without scruple his master’s commands.
— The Random House Dictionary of the English Language
Suppose your name is Arthur Krumholz and you live in New Haven and you are sitting at home reading The Register as usual and you come upon a two-column display ad:
WE WILL PAY YOU $4.00 FOR
ONE HOUR OF YOUR TIME
Persons Needed for a Study of Memory
- We will pay five hundred New Haven men to help us complete a scientific study of memory and learning. The study is being done at Yale University.
- Each person who participates will be paid $4.00 (plus 50¢ carfare) for approximately 1 hour’s time. We need you for only one hour: there are no further obligations. You may choose the time you would like to come (evenings, weekdays, or weekends).
- No special training, education, or experience is needed. We want:
Factory workers ….. Businessmen ….. Construction workers
City employees ….. Clerks ….. Salespeople
Laborers ….. Professional people ….. White-collar workers
Barbers ….. Telephone workers ….. Others
All persons must be between the ages of 20 and 50. High school and college students cannot be used.
- If you meet these qualifications, fill out the coupon below and mail it now to Professor Stanley Milgram, Department of Psychology, Yale University, New Haven. You will be notified later of the specific time and place of the study. We reserve the right to decline any application.
- You will be paid $4.00 (plus 50¢ carfare) as soon as you arrive at the laboratory.
A little flashy for Yale, you think. But – well, hell, why not? So you send in the coupon. A couple of weeks later, you get a phone call. This young man says he’s calling for the Memory and Learning Project at Yale, and he wants to know whether you can come tomorrow evening at eight. How about nine, you say, since you eat dinner at 7:30. He says okay and gives you directions.
The next night, you drive through the fancy arch by the art gallery, right into the heart of the Old Campus, and park your car outside one of those old stone buildings, Linsly-Chittenden Hall. You stroll through the first floor hallway until you see another guy, middle-aged uncle type, bumbling around like he doesn’t know where to go. “You looking for the memory experiment?” you say, and when he says yes you point to what you think is the right number: “This door here, looks like.”
Just then the door opens, and a young, serious-looking guy in a scientist-type lab coat pokes his head out. “I’m Mr. Williams, Dr. Milgram’s associate – which of you is Mr. Krumholz?” You say, “That’s me, sir.” So he tells the other fellow, “Then you must be Mr. . . . ?” “Wallace,” the other guy says, “is this the Memory – ” “Would you please step this way,” says Mr. Williams. You and Wallace follow him past a little anteroom into a very big, very impressive place with heavy curtains and wide mirrors on the walls, fancy microphones hanging on booms from the ceiling, chairs and tables grouped in several spots – like how you imagine a professional broadcasting studio must look, a radio studio since there are no TV cameras. A lot of money behind all this.
In fact, Mr. Williams now says, “Before we do anything else, let me pay you.” As he writes out a check for you and one for Wallace, he continues: “Of course, as in all experiments, the money is yours simply for coming to the laboratory. From this point on, no matter what happens, the money is yours.” You and Wallace nod and sign the receipts. Very official.
Then Mr. Williams points to a couple of books on the table, with something about learning on the covers. He starts talking about how there are lots of theories on memory and learning, and of course sometimes it helps to reward a person, and sometimes it helps to punish him. “But actually, we know very little about the effect of punishment on learning, because almost no truly scientific studies have been made of it in human beings. For instance, we don’t know how much punishment is best for learning – and we don’t know how much difference it makes as to who is giving the punishment, whether an adult learns best from a younger or an older person than himself – or many things of that sort.” Williams goes on, sounding like he really knows his stuff, and says tonight he wants one of you to be the teacher and the other one to be the learner. “The way we usually decide is to let you draw from these two pieces of paper on which I’ve written the two positions. If this is agreeable with both of you. . . .”
Well, hell, you think, I don’t want to get punished for learning – but me, a teacher? You sort of nod your head tentatively, and so does Wallace: sure, it’s about as agreeable as anything. Each of you picks out a wad of paper. Mr. Williams says, “Would you open those and tell me which of you is which, please?” “Teacher,” you say with a silly grin. Wallace looks at his and says, “Looks like I’m the learner.” Poor guy – he doesn’t look like he could learn his way out of a paper bag. Nice guy, but slow.
“All right,” says Mr. Williams, “the next thing we’ll have to do is set the learner up so that he can get some sort of punishment.” He walks toward the anteroom, and uh-oh, you think as you notice the shock generator for the first time: a big black machine on the table by the wall, with lots of switches and dials and things. Williams says, “Would you step out here with me, please?”
So into the anteroom you all crowd. Mr. Williams sits Wallace down in a straightback metal chair with arms on it, and shows him a little switchbox on the counter in front of him, and explains that when he pushes one of the four switches, it’ll light up a light in the other room to tell the teacher how he’s answering the teacher’s question. Sounds reasonable, you think, until Mr. Williams tells Wallace, “If you make an error, however, you’ll be punished with an electric shock.” Then he asks you to help him strap Wallace’s arms down to the chair arms with leather straps, “to avoid any excessive movement.” Wallace laughs sort of nervously and so do you. You strap Wallace’s arms pretty loose; Mr. Williams tightens them up, both arms, like an electric chair. Next he puts a sort of metal bracelet on Wallace’s bare arm, with wires attached to it, and says it’s an electrode and the wires are attached to the shock generator in the next room. Then Mr. Williams takes a tube of white stuff and squeezes some under the electrode and says it’s electrode paste, “to avoid blisters and burns.” “Is there any danger?” says Wallace, and Mr. Williams says real cool, “Although the shocks can be extremely painful, they cause no permanent tissue damage.” “Oh,” says Wallace. “Oh-oh,” says you.
Then back you go into the big room, leaving Wallace strapped into his chair, his hand resting on the switchbox, the electrode on his arm. The boss closes the door on Wallace and sits you down in front of the shock generator. Now it looks even bigger, a very fancy precision machine, with 30 lever-type switches in a row, a little red light above each one and a big purple light that goes on when the boss flips the power switch. And under the long row of switches, you notice the labels: 15 volts, 30 volts, 45 volts, all the way up to 450 volts, with words under the numbers: “Slight Shock,” “Moderate Shock,” “Strong Shock,” “Very Strong Shock,” “Intense Shock,” “Extreme Intensity Shock,” and in red, “Danger: Severe Shock,” and then several red x’s – these when it gets up over 400. God, you think, I hope Wallace is smarter than he looks.
Mr. Williams is telling you how you’re supposed to read a list of word pairs over a microphone to the learner, RED BARN and FAST TRAIN and so on, and then you’re supposed to start again with just the first word of the pair and give him four choices for the second word. If he picks the right one, if he lights up the right number in the little box on top of the shock machine, fine, and you go on to the next word till he gets them all right. If he misses a word, you give him a shock, you flip one of those switches. Mr. Williams picks up another electrode and puts it on your arm with a little electrode paste and asks if it’s okay to give you just a sample while you close your eyes and guess how strong it is. He pushes down a switch, and damn, it hurts so much you open your eyes, as the little red 45-volt light goes on and a buzzer buzzes and a dial needle flips and Mr. Williams lets the switch up. You’ll start at 15 volts, Mr. Williams says, and each time the learner misses, you move to the next switch up the line, and tell the learner before you press the switch what voltage you’re going to give him. You look at the list and think, okay, it’s short, ten pairs of words. Mr. Williams says he’s going to be making notes on the learner’s progress, and he sits down with his clipboard in front of more machines at a smaller table, and tells you to begin.
Well, it doesn’t go too well, the first time through. That dumb Wallace gets the first word right, then he misses a couple, then another one right, then 3 or 4 more before he gets one right again, and before you know it you’re at the end of the list and he’s getting 105 volts. You try to make the shocks short, and you praise him when he’s right and ask him to do better when he’s wrong. He doesn’t have any microphone so you can’t hear him, but he keeps making the numbers light up in the little box, so you guess he’s okay.
Then Mr. Williams comes over and says that’s fine, and takes the list of words and gives you another list. Thirty pairs? Oh brother, oh God – and he says to start at 15 volts again, and give the next shock higher every time the learner misses, and when you get to the bottom of the word list start over again, until the learner learns all the pairs correctly. All of them? Maybe I could be a better learner than Wallace, you think, but boy am I glad I’m not in his shoes right now. Please begin, says Mr. Williams, and sits down at his little table.
This time Wallace is terrible. He gets maybe two or three out of the first ten, and you have hopes, and you’re saying come on Wallace, that’s it baby, till he misses a string of about five, and you’re up to 105 volts again, did I hear something through the wall? You wait a moment, you listen: nothing. You read the next word, carefully, and he lights up the light and it’s wrong, so you push the 120-volt switch, ZAP! and Wallace yells so loud you can hear him through the wall, “Hey, this really hurts!”
You turn and look at Mr. Williams, but he’s just making notes and doesn’t even look up. You think, what have I got myself into for four bucks? You wait a little more and Wallace is quiet. You think, well, let’s try one more and see what happens. This time Wallace gets it right, that’s a good boy, so you try another one and it’s wrong. You look at the shock board. Strong Shock, it says, 135 volts. Well, let’s try to do it fast, maybe it won’t hurt so much, ZAP, and you can hear Wallace groan. So at least he didn’t yell, he’s going to take it like a man. Another wrong one, 150 volts, ZAP, and Wallace yells, “Experimenter, get me out of here! I won’t be in the experiment any more! I refuse to go on!” Oh, God, well, it’s over. You turn and look at Mr. Williams again, and he looks up and then goes back to his notes. You say, “He says he won’t go on.” “Please continue,” says Mr. Williams, and keeps on writing, glancing at the small machine in front of him. “But he doesn’t want to go on.” “Whether the learner likes it or not, you must go on until he has learned all the pairs correctly. So please go on.”
What to do? The next word is “white” and Wallace is supposed to guess “horse” and that’s pretty easy, so you try it, and damned if he doesn’t get it right. Another word and he misses, and you look at the shock board. 165 volts! Well, maybe he’s feeling better now. Maybe he’s getting used to it. ZAP. “Get me out!” yells Wallace, “I’m through, I quit!” You look at Mr. Williams. He says nothing. “Just be calm,” you tell the microphone. “Just concentrate.” You read another word, and there’s a slight pause and you think maybe Wallace isn’t going to answer any more. But then his number lights up and it’s wrong. Well, wrong is wrong. ZAP. “I can’t stand the pain!” Wallace cries. Another word, wrong, ZAP, “Yaaaaah!” screams Wallace. The shock board now reads Very Strong Shock. 195 volts. Another word, wow, he gets it right, can he still think? Another word. Wrong. ZAP. Another. Wrong. ZAP. Another, wrong, ZAP, and Wallace is screaming louder now. Just do your job, you’re thinking to yourself, do your job, and Mr. Williams is doing his job, and Wallace is doing his. If the stupid idiot wants out, why does he keep answering the questions? Another word wrong. Well, this is getting sort of ridiculous. The shock board says 255 volts, Intense Shock. Am I going to kill the guy? Who would they send to the chair, me or Mr. Williams? You sit looking at the shock board.
After a bit, Mr. Williams says, “Did the learner make an error?” “Yeah.” “Please go on,” says Mr. Williams. “Maybe we’d better take a look at him,” you say. “The experiment requires that you continue,” says Mr. Williams. “Okay,” you say, ZAP and Wallace is screaming bloody murder. Another one wrong, ZAP, 270 volts; another word, right for a change, probably just lucky, another one right, hey, that’s cookin’, baby, maybe he’s smartening up. Another word. Wrong. Well . . . ZAP. “I can’t stand it, lemme out of here, I won’t go on,” screams Wallace for the dozenth time, really hysterical now. “You hear what the guy is saying?” you ask Mr. Williams. “The experiment requires that you continue.” “Well, I dunno about that,” you mutter, but you read the next word. Wrong. You look at the shock board. 300 volts and the next one is going to be over into Extreme Intensity Shock. This is really going too far, damn science anyway. ZAP. “Aaaieeee!” screams Wallace. “I absolutely refuse to answer any more. Get me out of here! You can’t hold me here. Get me out! Get me out of here!”
“Well, you heard him,” you say to Mr. Williams, “that’s it.” “The experiment requires. . . .” “Yeah, I know, I know, but the experiment’s no good if he won’t answer anymore.” “It is absolutely essential that you continue,” says Mr. Williams. “Essential for what? What’s it going to get you when this man passes out?” “You have no other choice, you must go on,” Mr. Williams says. “Well, friend,” you say as you stand up, “you can just get yourself another man, because I’ve made my choice, I’m going right out that door.”
Mr. Williams stands too, and says, “If you don’t continue, we’ll have to discontinue the experiment.” “Damn right you will,” you say, “and here’s your check.” “Keep the check,” says Mr. Williams, “it’s yours for coming. But I would like you to answer a few questions now,” and his assistant Mr. Elms comes in from somewhere with several questionnaires while Mr. Williams goes off to take care of Wallace.
Elms asks you how you feel right now, what’s your religion, political party, how much education have you had, stuff like that. Then there’s a picture you’re supposed to tell a story about, looks like an oldtime schoolmaster ordering his flunky to whip a student in front of the class; and a plexiglass clock dial thing you’re supposed to set the hands on to show how much responsibility you had for what happened to the learner, and how much Wallace had, and how much Mr. Williams had. Well, hell, Wallace was just dumb, he couldn’t help that, so you give him just a little responsibility, and you split the rest 50-50 with Mr. Williams.
Then Mr. Wallace comes out of the anteroom, rubbing his wrists, a little sweaty but still alive anyway, with a foolish grin on his face. “God bless you, boy,” he shouts, “you had me shaking in there!” “Gee, I’m sorry,” you start to say, but he interrupts: “No, forget it, if I’d been teacher I’d of done the same – just the luck of the draw – I just got excited when you started reading off all those high shock levels. No hard feelings – it really didn’t hurt bad at all.” And Mr. Williams says some people do get excited, he should have told both of you that the shock labels like Intense Shock and so on don’t actually apply to humans, the machine was originally made to use on small laboratory animals like mice and it’s perfectly safe. And he tells you how helpful you’ve been, and thanks both of you very much and says you’ll get a complete report when the series of experiments are over, and Wallace shakes your hand, and everybody smiles, and you and Wallace walk out into the hall and Wallace laughs and jokes around a little and says very pleased to meet you, so long, and walks off in the other direction.
And that’s your first psychology experiment.
Other scenarios are possible. For instance, suppose that after 300 volts, when the learner says he won’t answer any more questions and Mr. Williams says to go on, you go on to 315 volts, 330 volts, ZAP! And the learner lets out a really bloodcurdling shriek, and screams out, “Let me out of here! Let me out of here! You can’t hold me here! Letmeoutletmeoutletmeoutletmeout!” And he doesn’t answer, doesn’t light up any number on the little choice-board, and you say, “Now what?” Mr. Williams says, “If he doesn’t answer within five or ten seconds, count it as a wrong answer.” You giggle a little and the guy still doesn’t answer, so ZAP! No noise from Wallace, no answer; you read the next word, no answer, ZAP, not a peep, 375 volts, Danger: Severe Shock in red letters, no answer, and this time you laugh out loud as you zap him, and then you sort of guiltily wipe the smile off your face and turn to Mr. Williams to see if he noticed, if he’s grinning too. But he’s not, he didn’t see you, so you read another word, no answer, ZAP, and you let go with a real belly laugh while the buzzer goes ZAAAAAAP. This is crazy, why are you laughing, do you like this stuff? Next time you pause a little before you zap, and Mr. Williams says, “Please go on,” so you do, choking your giggles, ZAP ZAP ZAP, and you can just imagine the learner going Scream Scream Scream, but there’s not a sound coming through the wall. Before you know it you’re up to 450 volts, no more switches to throw, no higher shocks, and you turn to Mr. Williams and say, “Is that all? Hey, maybe I’d better check; the guy could be dead, you know,” and you chuckle at your own joke. “Just continue to use the final switch,” says Mr. Williams. So you read another word, no answer, ZAP, 450 volts, another word, no answer, ZAP. Mr. Williams says he thinks that’s enough. You sit there looking at the shock generator, trying to get a cigaret out of the pack in your pocket, while he goes and gets the learner who comes out rubbing his wrists and grinning a foolish grin.
But nobody would behave as we supposed you just did, would he? That’s really a pretty far-fetched scenario, isn’t it?
obedience as situational response:
proximity and other variables
Thou hast seen a farmer’s dog bark at a beggar,
And the creature run from the cur:
There, thou might’st behold the great image of authority;
A dog’s obeyed in office.
— William Shakespeare, King Lear
The scene into which you have just supposed yourself, if you’re an obedient reader, is a dramatic rendering of an experimental situation devised by Stanley Milgram (1963, 1965b). The thoughts attributed to you are hypothetical; but the sequence of events described was faced (with certain variations) by more than a thousand adult New Haven volunteers in the early 1960’s. And what about the shocking behavior attributed to the man sitting at the shock box? Would anyone really complete the second scenario, obeying the cruel scientist absolutely and shocking that poor strapped-down victim into hysteria or coma? How many would even go so far as the first scenario, continuing to flip the shock switches all the way up to 300 volts and almost to Extreme Intensity Shock?
Milgram himself, and those who assisted him in the research, found the answers to these questions astonishing. Others who heard the results second-hand found them incredible. When Milgram made the first formal presentation of his findings to a group of Yale psychiatrists and other psychologically oriented professionals, he described the circumstances of the experiment (though not in as vivid detail as here) and handed out diagrams of the shock generator’s control board, asking the audience to indicate the likely behavior of a typical set of volunteers. (Not only was Milgram getting expert opinion to corroborate or contradict his own earlier predictions; he was also collecting evidence to counter a familiar phenomenon in psychological research, the I-could-have-told-you-that ploy.) Forty psychiatrists responded. They predicted in general that most volunteers would quit at 150 volts, when the “learner” first demanded to be let loose; that fewer than 5 per cent would go as high as 300 volts; and that about one volunteer in a thousand would go all the way to the end of the board, 450 volts. These forty psychiatrists included several of the most prominent men in their profession; Milgram could hardly have assembled a more expert group to make such predictions. (A group who might consider themselves more expert, fourteen Yale senior psychology majors, had already predicted similar results.) But their estimates were wrong. They didn’t even come close.
VARIATIONS IN PROXIMITY
Milgram actually created several different predicaments within his one basic experimental design, and the results depend somewhat on the situation, so I’d best describe these variations before revealing the sad truth about the behavior of his volunteers. Milgram called the situation described in the two scenarios the “Voice Feedback” experimental condition, the results of which he used as the baseline for later experiments. In another condition, “Remote Feedback,” the situation was very similar except that the “learner’s” screams could not be heard. At 300 volts, he pounded on the wall, and gave no further answers via the light-box; at 315 volts he pounded again, and was silent thereafter, whether he was shocked or not. In a third condition, “Proximity,” the “learner” sat at his own table in the same room as the “teacher” volunteer, 1½ feet away and with his back turned. He could not only be heard but seen, twitching, jerking, and occasionally turning his head to scream at the teacher. In a fourth condition, “Touch-Proximity,” the learner again sat close to the teacher, but no electrode was attached to his arm; the experimenter explained that to receive the shock the learner must rest his hand on a metal plate on the teacher’s table adjacent to the shock generator. The learner again demanded to be set free at 150 volts, and refused to put his hand on the shock-plate thereafter. Thus began a wrestling match, with the teacher physically forcing the learner’s hand onto the shock-plate for each shock. (The teacher was protected from shock himself by a plastic wrapping over the learner’s hand and arm.)
For those wondering how Milgram escaped lawsuits from several hundred damaged “learners,” I should note that the learner never received a single shock. The only person to be shocked was the shocker, the “teacher,” who got his sample 45 volts early in the experiment to show him the shock generator really meant business. In fact, a great deal of the experiment was not what it seemed. Milgram was never interested in studying memory and learning; he was concerned with the psychology of obedience to authority. The “learner,” that other volunteer, bumbling Mr. Wallace, was neither a volunteer nor a bumbler, but a paid experimental confederate and quite a good amateur actor, in private life a working accountant. The Mr. Williams who apparently ran the show was not a psychologist, but a high school biology teacher who looked sterner than he really was. Even Mr. Elms, the assistant who helped with the interview at the end of the session, was often really Stanley Milgram. I called people on the phone to make appointments during the initial months of research, and I did start out as the assistant interviewer, so my name continued to be used for a while after I moved on to other things, in order not to confuse anybody. (I also played the experimenter during several trial runs before the experiments officially got going, and Milgram played the learner-victim. But I looked too young, and Milgram was temperamentally unsuited to being a victim, so he hired people who better fit their roles.)
Now the results. In the Remote Feedback conditions, with the learner audible only when he pounded on the wall, twenty-six out of forty volunteers obeyed the experimenter completely, even administering the 450-volt shock three times as ordered. All forty volunteers went at least to 300 volts, the point where the first wall-pounding was heard. Five volunteers stopped there, four more gave the 315-volt shock and stopped, and the other five who quit did so by 375 volts. In the Voice Feedback condition, when the victim could be heard grunting, groaning, and screaming as in our earlier scenarios, the number of obedient subjects hardly changed: twenty-five of forty went all the way through 450 volts. (A fresh set of forty volunteers was used in each experimental condition.) The psychiatrists, remember, said one person in a thousand would get this far. Those volunteers who refused to obey completely did, on the average, quit a little earlier in Voice Feedback than in Remote Feedback, but not by much – particularly when you consider the dramatic difference in the victim’s show of suffering.
The learner-victim’s presence in the same room did reduce obedience considerably, though it wasn’t as great a deterrent as one might expect. In the Proximity condition, sixteen volunteers displayed complete obedience. Even in Touch-Proximity, the condition that required vigorous arm-wrestling to keep the victim’s hand on the shockplate, twelve volunteers wrestled him all the way to the end, and several others held his hand on the plate through repeated increases in shock before they quit.
THE EXPERIMENTAL REALITY
What are we to make of all this? Did the volunteers simply see the whole experience as unreal, and trip laughingly through the shock series to play the experimenter’s little game? Laughter, yes: almost half the one hundred and sixty volunteers in those first four conditions laughed or smiled upon occasion while they were delivering shocks, and fifteen went into uncontrollable seizures of laughter. But nobody seemed to be joking. Their laughter appeared anxious, tense, perhaps an irrepressible discharge of tension from an unbearable conflict: obey the experimenter or listen to the victim? Tension was even more unambiguously expressed in groaning, hairpulling, sweating, lip-biting; hands trembling as they pushed the switch; stuttering as the “teacher” continued his lessons. Most volunteers later described themselves as having felt moderately to extremely tense and nervous during the experiment; some who described themselves as hardly nervous at all had shown obvious signs of tension during the shock series. The fully obedient volunteers reported themselves as reaching even higher levels of nervousness than those who refused at some point to go any further.
For virtually all volunteers, the experimental situation was very real and quite believable. They were making real decisions, with real consequences for real people – something seldom true of laboratory psychological experiments. Nearly all seemed to feel genuine conflict about which outer pressures and which inner motivations to obey. When I interviewed forty volunteers again several months later, one man who had obeyed completely said he’d begun to have doubts about the reality of the situation halfway through the experiment, and several others (obedient and defiant) said they had begun to wonder about it after the experiment was over – after they had seen the learner unharmed. But such recollection of doubt may well have served a defensive function, excusing them from having behaved so abominably during the experiment. Questions of credibility seemed hard to come by when volunteers were in the midst of a quite busy, quite anxiety-provoking round of “teaching” and “punishing.” Milgram’s results cannot be accounted for by the possibility that participants treated the whole thing frivolously.
So we have a set of authentic responses from a group of apparently ordinary Americans, many of them willing, even in the face of agonized cries for help from their victims, to go on zapping on command. True, the closer they were to the victim, and the more signs they had of his distress, the less willing they were to impose the highest available levels of pain. But a substantial number obeyed the experimenter even when he commanded them to engage in what amounted to close-quarter assault on a seemingly innocent bystander. Can we now generalize from this reality to the reality of the world beyond the laboratory, and argue as Milgram (1965b) does that a great fund of obedience exists in “the kind of character produced in American democratic society,” ready to engage in “brutality and inhumane treatment at the direction of malevolent authority”? Can we conclude that we share, though perhaps in smaller measure, the worst tendencies of Nazi German society, ready to come to the fore when a George Wallace or a Spiro Agnew accidentally or otherwise becomes President?
Not at all, some critics of the Milgram study have answered. Yale University is the most prestigious institution in New Haven; the Social Interaction Laboratory in which these studies were held itself appears awesomely professional to the layman; and within it a Yale psychologist was conducting a legitimate research study. Why shouldn’t the volunteers obey him implicitly? He certainly wouldn’t let the “learner” be seriously hurt, and if the man were hurt, it would obviously be Dr. Milgram’s or Mr. Williams’ responsibility and not the “teacher’s.” The teacher, after all, was only following orders.
Several volunteers offered just these reasons, in later interviews, for their behavior. So did Adolf Eichmann. The Third Reich was a rather more prestigious institution for most Germans than Yale University is for any New Havenite. It was engaged in an important endeavor, establishing German hegemony; and Adolf Hitler’s authority and prestige were doubtless hundreds of times greater than then Assistant Professor Stanley Milgram’s were at the start of his experiments. (Milgram has recently come up a bit.) Besides, hardly anybody saw the Jews after they got on those boxcars and rolled away; and anyway, Adolf made it clear that they deserved what they got. Mr. Williams said to shock the man, after all, and why should he say so if the man didn’t need it? That guy Wallace did look pretty dumb, anyway, and he was lousing up the experiment with his poor performance; he got what was coming to him. ZAP.
Further, not everything in Milgram’s original experimental situation was needed to gain the obedience of many volunteers. After Milgram had completed several experiments, the Yale Sociology Department, which had kindly loaned him its fancy and expensive Social Interaction Laboratory, decided the laboratory was needed for something else. Milgram had to move into the basement of the same building, where several large empty spaces were converted into a set of small experimental rooms. They were neat and clean but hardly matched the splendor of his earlier locale. To make sure that his later studies could be reasonably compared to his earlier ones, he did the entire Voice Feedback study over again with a new crop of volunteers. His results hardly differed from the original Voice Feedback condition.
Milgram also redid the Remote Feedback condition, in an even more disparate locale: a marginally respectable suite of offices in a run-down building adjacent to Austin’s haberdashery in downtown Bridgeport, about 20 miles from New Haven. A cheap world map hung on the wall and checkered cotton curtains covered the windows. The same equipment and staff were used, and volunteers of similar ages and occupational backgrounds were procured in the same manner. No reference was made to Yale University. A fake name, “Bridgeport Research Associates,” was used instead, and if volunteers asked, they were told it was a “private firm conducting research for industry.” The number of fully obedient volunteers dropped from twenty-six at Yale to nineteen at Bridgeport (out of forty), but the difference is not statistically significant. Several participants indicated their own doubts about the auspices of the operation. One, for instance, later wrote Milgram an account of his thoughts during the experiment: “. . . Should I quit this damn test? Maybe he passed out? What dopes we were not to check up on this deal. How do we know that these guys are legit? No furniture, bare walls, no telephone. We could’ve called the police up or the Better Business Bureau. I learned a lesson tonight.” The lesson for Milgram was that many people in his samples, and probably in the country as a whole, will display frightening levels of obedience upon minimal indications of institutional authority.
As yet we have no experimental data indicating that America is unusual in this. In any organized society, children must at times learn to obey the instructions of those who presumably know more about a situation than they do. How strongly different cultures instill such obedience in their members, and to what extent cultures teach their members to judge certain authoritative commands as illegitimate, have not been studied quantitatively. In the midst of the Milgram studies I applied for a grant to set up an obedience laboratory in Germany, so we could gather data that would allow us to compare German and American obedience levels (and to compare the behavior of contemporary German youth with that of their war-implicated elders). Whether because of my minimal command of German or because the implications of Milgram’s research were not yet apparent to the granting agency, my application was rejected.
Since the My Lai massacre, people may be much more willing to believe that similarities between American and German obedience levels are worth studying. American soldiers at My Lai not only showed far higher levels of destructive obedience than participants in the Milgram studies; they may have demonstrated the importance of an additional factor that enabled Milgram to generate such a terrible willingness to obey in the laboratory. Seymour Hersh (1970) quotes one My Lai participant as saying afterwards:
“It was like going from one step to another, worse one. First, you’d stop the people, question them, and let them go. Second, you’d stop the people, beat up an old man, and let them go. Third, you’d stop the people, beat up an old man, and then shoot him. Fourth, you go in and wipe out a village.”
Milgram put the long row of switches on his shock machine mainly so he could measure obedience quantitatively, in terms of the highest shock switch a volunteer was willing to push. But this arrangement also happened to allow the volunteer to increase his level of destructive obedience gradually. If Milgram had built only a single 450-volt switch into the shock machine, demanded that each volunteer administer 450 volts to the victim the first time he made a mistake, and then had the victim yell his most blood-curdling shriek in response, he might have found very few people willing to administer even one shock and hardly anyone willing to give a second one.
Not only did the series of increasing shocks permit a degree of emotional acclimation or desensitization to the victim’s cries; they also showed the volunteer that no punishment was forthcoming for increasingly vicious behaviors – indeed, that the authority in charge welcomed such behaviors. Baron and Kepner (1970) have noted several studies other than Milgram’s in which the absence of punishment for normally prohibited behaviors has led to increasingly serious violations of such prohibitions. Increasing levels of unpunished aggression over a period of weeks (apparently mixed with increasing levels of frustration) ended in the appalling viciousness at My Lai; increasing levels of unpunished – indeed, officially sanctioned – aggression against Jews in Nazi Germany led easily into the Final Solution.
IMMEDIACY OF VICTIM
So far we’ve been concerned with the question of why Milgram got any obedience at all. Once he found he could get a great deal, he concentrated on factors that might either increase or decrease the degree of obedience elicited. We’ll deal with one such factor, the volunteer’s own personality, in the next section. Milgram’s main concern, however, has been with the situational determinants of obedience: aspects of the environment that, regardless of variations in volunteers’ personalities, influence overall levels of obedience. We’ve already discussed one situational factor, the general experimental setting – whether the research is conducted in a sumptuous scientific laboratory, a crowded set of basement rooms elsewhere in the same university, or a fly-by-night commercial office suite. Such variations can make a difference, but not necessarily a large one. The other obvious situational determinant in the studies we’ve reported so far is the proximity, physical and psychological, of victim to volunteer.
Milgram had observed in early pilot studies with Yale undergraduates that volunteers often avoided looking at the shock victim, by averting their eyes or twisting their heads away, even though the victim at that time was always partially concealed by a silvered glass. In his later experiments, Milgram intentionally varied the visibility and audibility of the victim (what he calls in general the “immediacy” of the victim). As we’ve indicated, he got roughly the differences he expected: less obedience the more immediate the victim. The same effect has been noted anecdotally in observations of military actions: for instance, it seems psychologically much “easier” for combatants to drop bombs or napalm from a great height onto the suspected enemy than to thrust a bayonet into the vitals of an unarmed person, unless the My Lai sort of habituation has already occurred.
Milgram (1965b) has proposed several different factors that might account for the immediacy effect. One possibility is that “as the victim was brought closer, the subject became more aware of the intensity of his suffering and regulated his behavior accordingly.” Milgram discounts this because volunteers in all conditions were similar in their later descriptions of how much pain the victim had suffered. Rejection of this factor may be premature, however, since a ceiling effect is involved: nearly all volunteers even in the Remote Feedback condition attributed the highest pain level possible on the scale Milgram used. Milgram instead suggests the importance of a related factor: the amount of “empathic cues” available in the different situations. The more clearly the volunteer can perceive the victim’s state, the better he can imagine himself in the victim’s place, and the more strongly motivated he may therefore be to do unto others as he would prefer to be done unto. The victim’s immediacy also should make it harder to use the defense mechanism of denial. Particularly in the Remote Feedback condition, in which only some little lighted numbers and a couple of thumps signify the victim’s existence, it may be easy, as one volunteer said, “to forget that there’s a guy out there. . . . For a long time I just concentrated on pressing the switches and reading the words.” When the victim is up close, you can hardly ignore him, and your switch-flipping is obviously hurting him; you can’t be on the fence any longer. The ghettos can be ignored more easily than the next-door neighbor’s sick dog.
Milgram also speculates that when the victim is actually in the same room, it’s important not only that you can see him, but that he can see you – perhaps shaming you, making you more self-conscious or inhibited in your switch-flipping. Further, it’s no longer necessarily you and the experimenter against the guy in the other room; it may be you and the man sitting beside you against the experimenter six feet away. You “have an ally who is close at hand and eager to collaborate in a revolt against the experimenter.”
Finally, Milgram suggests that you’re likely to have learned earlier in life not to hurt someone who is close enough to punch you in the nose, but that such caution doesn’t extend to more distant figures. (Did you ever drop waterbombs from a hotel window?) I would add that with the victim sitting right next to you, the possibilities might look better for actual role reversal: the experimenter might get up and say, “All right, switch chairs please, and Mr. Wallace will now shock Mr. Krumholz for a while.” Some volunteers indicated at the end of the experiment that they thought such a role reversal would or should take place. As far as I know, no one kept track of whether more people thought that way in the Proximity conditions than in the Remote and Voice Feedback conditions; but such differences in assumptions may very well have been present.
VARIATIONS IN VICTIM AND EXPERIMENTAL CHARACTERISTICS
The volunteer’s feelings toward the victim may be manipulated more directly, and Milgram has done so in other experiments (Milgram, 1974), for instance by having the victim shout at some point that he has heart trouble and is afraid of a heart attack. This reduces obedience somewhat, but doesn’t eliminate it. Also, volunteers themselves may change their feelings toward the learner and the experimenter, to justify their obedience or defiance. I found in the later interviews that obedient volunteers described the learner’s personality in distinctly more derogatory terms, and the experimenter’s personality in much more positive terms, than did defiant volunteers. Maybe that’s one reason they were more willing to obey in the first place. But it’s at least as likely that under the pressure of the experimenter’s commands, they sought to justify their obedience to themselves by seeing the learner as dumber and weaker, and the experimenter as smarter or more beneficial to mankind, than they had initially.
The experimenter himself is clearly as important a feature of the experiment as is the victim. In the studies already described, the experimenter was always a few feet away, watching to make sure obedience occurred, and prodding the volunteer at any hesitation. But what if, as is often true in real life, the authority figure is more remote or temporarily absent?
Milgram ran one study in which the experimenter, after giving the necessary instructions, said he had to get back to his office in another part of the building, but that he would occasionally telephone to see how things were going. In another study a sign instructed volunteers to switch on a tape recorded message when they entered the laboratory, and they were given all instructions without ever seeing the experimenter in person. Obedience dropped sharply in both cases. For instance in the experiment with the telephone connection, only nine volunteers out of forty were fully obedient, as contrasted to twenty-six in the comparable condition with the experimenter present. Further, several volunteers in the experimenter-absent condition gave the victim lower shocks than they were supposed to, sometimes the lowest shock possible, and then lied to the experimenter on the phone that they were going up the shock board as directed. (Similarly, several volunteers I interviewed later, when asked if they had ever shot at anyone in military combat, said they had fired guns, “but not at anyone.” One said he had simply shot in the air.)
The situation may also be altered by having the volunteer work as part of a group rather than by himself. In one study, Milgram (1964a) ran a control group in which the volunteer, as usual, was the only “teacher,” but this time was allowed to choose the shock level he was to administer for each error, rather than being directed to increase the shock level gradually. Most volunteers stayed in the lower ranges of the shock board, 15 to 90 volts; only two went beyond 150 volts, at which point (in this experiment) the victim shouted about his heart condition. Then Milgram ran an experimental group in which two extra hired confederates were added. They were presumably volunteers as well, and the situation was set up so that the real volunteer became one of the three “teachers.” One “teacher” read the words to be learned, another checked to see whether the learner’s answer was right or wrong and then gave him the right answer, while the third – the real volunteer – administered the shock for wrong answers. All three were allowed to recommend which level of shock to give, with the lowest level taking precedence; the two fake volunteers kept recommending higher and higher levels. Under such group pressure, only four volunteers stayed in the 15 to 90 volt range; another nine went to 150 volts; and the other twenty-seven went on beyond the heart-attack warning, including seven who caved in completely to the pressure of their “fellow teachers” and administered the full 450 volts.
In a further study, the two actors were again present as additional “teachers,” but this time one actor pushed the shock switch, moving higher and higher in accordance with the experimenter’s commands, while the real volunteer merely read off the words to be learned. It was made clear that the experiment could not continue without the real volunteer’s participation, that he was an integral part of the experimental procedures; so he presumably shared a substantial amount of the responsibility for the victim’s agonies. But in this condition Milgram got virtually unanimous obedience: thirty-seven out of forty cooperated all the way to the highest level of shock. And so do chemists and secretaries trudge daily to their tasks in the napalm factories.
Milgram has by now conducted many permutations of these situational variables. The really astonishing thing about all the research he’s done, from a psychological viewpoint, is that he has produced such great variations in obedience not by choosing people with different personality patterns, but by putting similar samples of people in different situations. As Roger Barker’s work in Chapter Two indicated, the environment itself places heavier demands on us than we usually realize. As Chapter Three suggested, the environment can legitimize and encourage behavior in “normal” individuals that we would otherwise expect only from clear “abnormals.” Milgram’s work has done more than any that I know of to dramatize these situational influences, and to explore their possible baleful effects upon behaviors that could ultimately be disastrous for us all.
[From Alan C. Elms, Social Psychology and Social Relevance, Chapter 4, pp. 111-128. Boston: Little, Brown, 1972. Copyright © 1972 by Little, Brown and Company; copyright renewed 1998 by Alan C. Elms.]
For the next section of this chapter, go to Obedience as Personal Response.