Alan C. Elms
Virtual Library
  Alan C. Elms Home Page

 Alan C. Elms Home Page   Virtual Library   Books Online

Personality in Politics [pdf]
Uncovering Lives [pdf]

The Sin of Conformity

Alan C. Elms

the sin of conformity:
laboratory studies of group influence

The land of the free! This the land of the free! . . . Why, I have never been in any country where the individual has such an abject fear of his fellow countrymen.

-- D. H. Lawrence, Studies in Classic American Literature

Stanley Milgram's obedience studies were not the product of an isolated genius creating a radically different body of research. They came out of a heavily worked vineyard within social psychology, the field of experimental conformity research. Milgram's virtue is that he has drawn new wine from old vines. But conformity research itself has not been unproductive. It has even had its vintage years.

A few years ago, the word "conformity" was bandied about so often and in so many contexts that it was left with hardly any meaning beyond a vague connotation of opprobrium. Even current dictionary definitions are looser than they might be: "being in harmony or agreement" (Webster's New World Dictionary, 1962), "compliance or acquiescence" (Random House Dictionary of the English Language, 1967), "action or behavior in correspondence with current customs, rules, or styles" (American Heritage Dictionary, 1969). Your agreement with me that 2 + 2 = 4 wouldn't be evidence of conformity, the way psychologists see it; the bohemian's rejection of common "customs, rules, or styles" wouldn't necessarily be evidence against his conformity; and acquiescence as a result of physical force had better be called compliance rather than conformity. Agreement with another person as the result of logical arguments or elaborate emotional appeals has also seemed to psychologists to fall into a somewhat different psychological category. Psychologists usually define conformity as the act of behaving similarly to a group's majority, as a result of simple expressions of opinion by group members. It's not that the group is doing something you'd do anyway, even if there were no group. It's not that they're berating you or threatening you or delivering oratorical declamations at you. They're just stating a judgment and you're going along with it because that's what the group says.


Several early studies of group influence didn't really make such a clear distinction between conformity and persuasion, freely mixing simple judgmental statements with more complicated attempts at influence. The first important research on what we now would call conformity behavior, though it was then referred to as "formation of social norms," was done by Muzafer Sherif (1936). He seated volunteers, singly or in small groups, at one end of a darkened room, with a stationary point of light at the other end. Under such circumstances the light will soon appear to start moving. This is known as the "autokinetic effect" ("self-moving"), since the light really remains stationary and appears to move only because of the volunteer's own perceptual processes. Left to himself (and not told the light isn't moving), a person will establish his own norm of apparent movement: he'll see the light moving only within a rather narrow range of distances and directions. Sherif found that when several people are placed together and are asked to announce aloud how much the light is moving, most will come to agree on the amount of movement over time. Establishing such a group norm is actually a case of "mutual conformity": group members start off with no standards or with individual standards of movement, and are influenced by each other's statements until they agree closely on the light's movement.

Sherif's technique of reducing the group's operations to a bare minimum, and of using a task for which no norms exist outside the experimental laboratory, was much more susceptible to experimental manipulation than previous methods of studying group influence. But his results were not really surprising, either to the participants or to other psychologists. The autokinetic situation offers no real basis for judging what scale to measure the perceived movement against anyway, and other people are as likely to be "good" estimators of movement as oneself. So because their judgment is the only external reference point in the situation, why not use it to decide on your own scale? Conformity it may be, but of a relatively innocuous kind. Further, the sort of conformity involved – a mutual influence that ultimately establishes a group norm where none had previously existed – is rather different from most of our daily experiences with group influences, where long traditions are often involved and where even the new groups we form are likely to be so similar to previous groups that we can carry over old norms bodily.


Fifteen years later, Solomon Asch (1951, 1956) began to publish the results of research on another conformity situation, results that were not only surprising but that to a greater extent duplicated our experience with the groups we join – perhaps our experiences with society itself. Again, a situation ingenious in its simplicity: the volunteer finds himself sitting with seven or eight fellow students, all ready to participate in a visual discrimination study. The experimenter holds up two cards, one card with one vertical line on it and one card with three, and asks the students to say which of the three lines on the one card is the same length as the one line on the other card. Each student announces his choice in turn, our volunteer being next to last. Because the differences in the lengths of the three lines are quite noticeable, everybody agrees on the answer. The experimenter holds up successive pairs of cards, with still other lengths of lines, but the pattern is always the same: everybody chooses the one line out of three which is the same length as the line on the other card.

Soon, however, the first student to respond makes a mistake, picks what is obviously the wrong line. The other members of the group, for some strange reason, make the same mistake. What is our hero, alas, to do – tell it as he sees it or go along with the group, dissent or conform? Whichever he does, the experimenter moves along without pause to other cards, other lines; and occasionally the group members agree with our hero on the right line. But most of the time, over and over again, they choose a line that is definitely wrong, sometimes the wrongest line possible. They aren't demonstrative about it, they aren't disturbed at dissent; they just continue to answer dispassionately, whatever the volunteer does. And each time, the volunteer faces the choice: are his eyes right, or are the seven or eight pairs of eyes that belong to his fellow volunteers?

Knowing already about the occasional deviousness of social psychologists, you may immediately suspect (quite rightly) that those seven or eight other volunteers are paid shills of the experimenter, and that they've agreed in advance to agree on the wrong answers, to see whether the one real volunteer will go along or not. Most real volunteers didn't suspect; they believed the situation was as it appeared to be, and they became obviously upset at the conflict between what their own vision told them and what other group members reported. (If this conformity research, in particular, resembles Milgram's obedience studies, it should; Milgram worked as Asch's assistant for a year at Princeton's Institute for Advanced Studies.)

Asch's findings were as startling upon first publication as Milgram's obedience results were later. Three-fourths of the volunteers apparently ignored their own eyes and went along with the group at least part of the time; 27 per cent agreed with the group most of the time (on eight or more out of the twelve wrong group judgments). Only 24 per cent stuck with their own perceptions. In a control study, less than 5 per cent of volunteers tested alone chose anything other than the right line, and then only once or twice at most.

"Horrors!" a good many of Asch's readers seemed to think, "here we have a breed so conformist it will look at black and see white, if only the group says so! Conformity at its worst – blind, slavish conformity! Thank God, at least, for that virtuous 24 per cent who retained their own independence!"


But as it turned out, the nonconformists were not completely without guilt, and the conformists still saw black when they looked at black. They just said it was white. The contradiction between group judgment and personal judgment seemed highly stressful for practically everyone who participated, whether they remained confident that their own judgment was correct, or gave in to the group every time. Very few really came to think they were seeing things the way the majority was seeing things. A few did get so confused that they pretty much ignored the visual cues of the lines themselves, and simply took the majority version as accurate. Still others continued to see the lines as they really were, but went along verbally with the majority, for various reasons. Some independents never had major doubts about their own judgments, though they found their announced disagreements distressing. They speculated that the group might be subject to an optical illusion that didn't affect them, or that the rest of the group had misunderstood instructions. Other independents thought perhaps they themselves had misunderstood the instructions – that maybe, for instance, they were supposed to be judging which lines were similar in width, not length (though all lines were of the same width). But they weren't sure enough to change, and blundered ahead giving the right answers. Still others were pretty sure the group knew what it was doing, but just couldn't see any reason to change their reports of what they themselves saw, because they'd never had cause to distrust their vision before. As one volunteer told the group, "You're probably right, but you may be wrong!"

The behavior of the nonconformists in this situation is readily understandable. The experimenter had asked them to pick the right answers, so in order to help science, show their own intelligence, defer to the presumably powerful experimenter, or fulfill a commitment that they had voluntarily undertaken, they picked the right answers. The behavior of the conformists seems less immediately explainable. Asch lists several reasons given later by the conformists themselves: belief in the legitimacy of majority rule in a democracy, for instance, or a kind of "logical" calculation (if they're right and I disagree, I'm wrong; if they're wrong and I agree, everybody's wrong, so what?). One might even speculate that some volunteers were responding less to the current situation and the apparent demands of the current group than to much earlier conflicts, "transferred" to the present situation a la Freud – though Asch's post-experiment interviews were not designed to probe so deeply.

But what appear to be the main reasons for conformity, not necessarily identified as such by the conformists, work down to a small group of motives. Different psychologists have called the motives by different names, but they fall essentially into two categories. In one, the motives for giving answers acceptable to the experimenter are similar to those of the nonconformists; but rather than depending solely on their own senses to determine the correct answers, the conformists defer in part or whole to the group's judgment as well. We're very much accustomed to using the judgments of others to augment our own, when we don't have first-hand information or when a stimulus is ambiguous. In the Asch situation, the habit seems to carry over for a portion of the volunteers, even though the stimuli are unambiguous. Two reliable sources of information are, after all, in direct conflict, so why reject one source – the social one – entirely?

So some defer to group judgments to help them judge objective stimuli. For others, the group may be an even more important source of information in judging themselves. Both our self-esteem and our self-identity come in large part from the voiced or unspoken reactions of others. Social relationships play a major role in human life for many reasons, and the only reliable way to judge your own efficacy in such relationships is to see how others respond to you. In the Asch situation, volunteers were often fearful of seeming strange or silly to the rest of the group. Some even reasoned that if they behaved so stupidly in this temporary group, they would behave similarly and suffer similar consequences outside the group. "I hope you didn't think I was different," volunteers said after the experiment. "I like to be one of the boys, so to speak." "For a while it [disagreeing with the group] made me feel funny; it seemed as though I was a fool." "They probably think I'm crazy or something." These statements came despite the fact that the group never indicated any thought that the real volunteer was strange or lunatic. In this respect the group was quite artificial. The response such nonconforming behavior would likely get in a "real" group is illustrated by another Asch experiment, in which sixteen genuine volunteers and one paid stooge participated together. The real volunteers, when asked to judge the length of the lines, of course agreed consistently on the right answers; the lines were, after all, pretty easy to match. But at a certain point the stooge began giving wrong answers, minority-of-one answers, just as in the other experiments. The real group just laughed at him, laughed repeatedly and with great glee. Those minority-of-one volunteers in the usual Asch groups not only accurately perceived the lines on the cards; they appear also to have accurately perceived the likely group response to any dissenting judgments.


As with obedience, the particular situation at hand can make a big difference in how many people conform and how much. For one thing, the issue on which one is being pushed to conform can be crucial. The judgments demanded in conformity studies have ranged from which two pictures have the same number of paratroopers dotting the sky, to which of several arrangements of thumbtacks look more "friendly," to whether volunteers believe they could be happy without any friends at all. (Richard Crutchfield [1955] and others have built machines which can present many different kinds of visual stimuli or opinion items in rapid sequence, so that a battery of conformity situations may be used in one session.) In each case, the volunteer has no information from the group except its majority vote; he decides to dissent or conform on that basis. The general finding about such variations is that the more ambiguous the stimulus, the more conformity occurs and the easier the conformity is on the volunteers (the less conflict or anxiety.) The Sherif autokinetic situation is about as ambiguous as possible, and the Asch line-judging situation is about as unambiguous. A related factor is how far out the group's judgment is. If the stimulus is rather unambiguous and if the group's judgment is way off, volunteers will likely show less conformity than to a more reasonable but still wrong judgment. But through most of the range of possible judgments, at least some volunteers seem willing to tag along. In a study by Read Tuddenham (1958), for instance, volunteers working by themselves never estimated that people over 65 constituted 60 per cent or more of the United States population; but over half the volunteers in a group situation agreed with the group that this was the correct percentage. In the same study, only 2 per cent of the non-group volunteers agreed with the statement, "I could be perfectly happy without a single friend," but 40 per cent of a volunteer sample in a group-influence setting agreed with the statement after the rest of the group did.

Characteristics of the group make a difference too. A group whose members are seen as more attractive, more competent with regard to the issue at hand, or higher in status generally, is more likely to induce conformity. The group's size is important only up to a point. Asch eventually found that he didn't need to pay eight or nine assistants every time he tested one real volunteer; three assistants could elicit just as much conformity. (Two assistants got noticeably less and one assistant even less than that.) Perhaps this leveling off of conformity, once you go beyond three group members influencing one, happens partly because the unanimous agreement of larger numbers begins to look a little suspicious. Or perhaps it’s partly because (as some volunteers have indicated) the response pattern of a larger group looks too much like one instigator leading his flock of sheep into error, thus reminding the real volunteer of his own sheepish tendencies.

All these conclusions are predicated on a unanimous group influencing the lone volunteer. Give him a single ally, in a large group or small, and his likelihood of conformity drops drastically. Give him an ally who abandons him midway, and his likelihood of conforming jumps back toward normal. But if the ally can give a convincingly urgent excuse for leaving the scene, other than growing dislike for the volunteer or withdrawal from group pressure, the volunteer is likely to retain a good bit of the backbone with which his now departed buddy invested him.


Certain personality and demographic factors also influence conformity, but just how is not always clear. In various studies using a single type of conformity measure, a considerable assortment of personality variables has been associated with conformity. But such associations have then usually turned out to be absent or contradicted in other studies, although since different experimenters have seldom used the same personality measures, it's a bit hard to tell whether even the contradictions are contradicting whatever was originally studied (see Mann, 1959). Among the individual-difference factors most often found in these one-conformity-measure studies are that women are more conforming than men; that older people are more set in their ways and less conforming to the temporary group than the young; that the stupid are more conforming than the smart; and that authoritarians are more conforming than nonauthoritarians. But each of these differences has been found not to apply in at least some conformity situations.

An Australian researcher, Graham Vaughan (1964), tried several rather different ways of eliciting conformity in the same set of volunteers. He found that about 80 per cent conformed in some situations but not others, so that they couldn't be described as having any general conformity trait that could then be correlated with other traits. Vaughan feels, probably rightly, that most conformity-and-personality studies have exaggerated the relationships involved by locating personality characteristics that are important in determining one particular kind of conformity behavior, and then generalizing to conformity in all situations. Vaughan found only about 10 per cent of his volunteers conforming in all four of the conformity situations he used, and about 10 per cent consistently remaining independent of conformity pressures. Several factors differentiated these two groups: the independents were higher in intelligence and in assertiveness (on measures separate from the conformity situation); conformists were higher on measures of authoritarianism and dogmatism. However, even Vaughan's results are not without their problems. For instance, one of his conformity measures was the student volunteer's response to a request to turn in an extra class assignment by a deadline – behavior that would appear to partake more of obedience to authority than of social conformity, and that might be contaminated by other little-related factors such as procrastinatory tendencies. So Vaughan could be underestimating both the amount of general conformity and the number of personality factors related to conformity.


Attempts have also been made to relate conformity to certain national or cultural differences. Herbert Barry (Barry et al., 1959) has speculated that in hunting and fishing subsistence-level cultures, individualism and assertiveness would be most likely to promote survival, while in pastoral or agricultural subsistence-level cultures, compliance and conservatism would best guarantee survival. To test this hypothesis, J. W. Berry (1967) administered a simple modification of the Asch line-judging technique to adults of the Temne culture of Sierra Leone, whose food supply is mainly dependent on a yearly rice crop, and to adult Eskimos of Baffin Island, who are hunters and fishers entirely. A sample of Scotsmen was used as a sort of Western-culture control group. As expected, the Temne sample was highly conforming, the Eskimo almost completely nonconforming. The Scots were rather closer to the Eskimos than to the Temnes, as any Englishman might have predicted.

Stanley Milgram (1961), prior to his obedience research, did a cross-cultural study of conformity in Norway and France. He carefully selected volunteers in both countries so that such background factors as age, educational level, and social class were fairly equal. The volunteers were asked to participate in another modification of the Asch situation, in which they judged the relative length of musical tones heard on earphones. The instructions were delivered in the appropriate language by a native speaker; five other volunteers were presumably judging the same tones over a microphone system. Actually these other judges' responses were taped and only the one volunteer and the experimenter were present in the laboratory. Quite consistently, the Norwegians conformed more to the "group" judgments than the French – 62 per cent of the time versus 50 per cent in the basic experiment. When subjects in a separate experiment were told the tone-judging results would be used in designing accurate safety signals, conformity was slightly lower but the Norwegians were still ahead, 56 per cent to 48 per cent. When the situation was changed so that the real volunteer didn't have to announce his judgment aloud, French conformity dropped sharply to 34 per cent; Norwegian conformity less so, to 50 per cent. Milgram was able to increase conformity, on the other hand, by tape-recording snickers and belittling comments ("Are you trying to show off?"), supposedly originating from the group, which were played from a second tape recorder whenever the volunteer made an independent judgment. Norwegian conformity jumped to 75 per cent, French to 59 per cent. The Norwegians just took the criticism, but in France, over half the volunteers talked back when criticized, and two "became so enraged they directed a stream of abusive language at their taunters." Overall, 12 per cent of Norwegian volunteers conformed completely to the group, but only 1 per cent of French volunteers; 25 per cent of the Norwegians showed "strong independence," compared with 41 per cent of the French. Milgram suggests a relationship between these results and the "highly cohesive" Norwegian society, oriented toward social responsibility and group identification, versus the politically unstable, low-consensus French society, with a "tradition of dissent and critical argument."


Hollander and Willis (1967) have suggested that Milgram's results need not indicate any virtuous independence from group pressure on the part of the French, but that a good many Frenchmen might instead be what social psychologists call "anticonformists" or "counterformists." Whereas nonconformity or independence, in or out of the conformity laboratory, refers to doing your own thing regardless of group pressure, counterformity refers to doing the opposite of whatever the group is doing, just because the group is doing it. For the true independent, social pressure is either irrelevant or (more often) is not the critical factor in his behavior. But the counterformist is as dependent on the decisions of the group as the conformist is. He has to know what the group feels, so he can do something different – and he will do it even if the group is right.

Many people suspect their children, or youth in general, of being counterformists (so a song in The Fantasticks warns you not to caution your children against putting beans in their ears). There are good theoretical reasons for postulating the existence of counterformists in conformity situations, but experimenters have seldom run across consistently anticonforming individuals. Robert Frager (1970) has observed occasional anticonformity behavior in an Asch-type situation among Japanese college students, and finds it correlated with feelings of social alienation. But no one to my knowledge has found enough American counterformists to see what their problems might be. It probably wouldn't be hard to set up a laboratory situation that would temporarily generate a lot of counterformists: make the volunteer fighting mad at the group just before the line-judging starts, or present the group as John Birchers and use only Maoists as real volunteers, or vice versa. But I don't see what that would prove, other than that social psychologists can waste a lot of time being inventive.

The search for counterformists has come, I suspect, from more than abstract theoretical considerations alone. Psychologists seem to grow uncomfortable if they go on condemning one category of person too long; and a good bit of the conformity research has implied, if not stated, condemnation of conformists. On the other hand, we all like to think of ourselves as rather independent-minded. (Even the conformists themselves, in the studies that took the time to ask them, seldom praised their own conformity, and conveniently forgot or overlooked most of it.) So who else can we condemn to make up for our rough treatment of the conformists? Well, those hypothetical counterformists are definitely baddies – nonthinking troublemakers who don't even have the graciousness to go along with the group, which after all is right part of the time.


Conformity has been a popular research topic among social psychologists for several of the same reasons as rightist extremism – e.g., they don't like it and would like to undercut its rational bases or find ways to discourage it. But the psychologists have usually been careful (perhaps as a result of Asch's own thoughtful analyses) to grant the validity of conformity in certain situations. Obviously someone who allows all his behaviors to be determined by the group is a spineless ninny. But partial conformity to the group's standards can save us a lot of trouble, and permits enough efficiency in unimportant things to leave us time to make our own important decisions. A little less time spent on developing one's own unique artistic style, I daresay, would give us a rather larger quantity of substantial art. If you don't worry too much about whether your clothes are sufficiently avant-garde, you might be able to put a little more time into serious reflection on your own place in the universe, or whatever else you care to reflect seriously about. It is the content rather than the form of art or thought or life that can be seriously damaged by conformity.

Social conformity can on certain occasions even lead to a positive good. Such occasions have not often been studied; but Stanley Milgram (1965a) modified his obedience situation to provide us with one example. In his Voice Feedback condition, remember, twenty-six out of forty volunteers were completely obedient to the experimenter's commands to shock the protesting learner. Milgram ran another forty volunteers through this same basic condition, but added two more "volunteers," really paid confederates, who divided up the job of "teaching" with the real volunteer. When the "learner" yelled at 150 volts that he wanted out, one fake teacher refused to go any further, and walked away from the shock board. At 210 volts, the other fake teacher did the same, saying, "I'm willing to answer any of your questions, but I'm not willing to shock that man against his will; I'll have no part of it." So the real volunteer was left at the shock board to complete the experiment, and the experimenter told him it was essential that he do so.

Here, experimenter pressure for obedience conflicts with group pressure for disobedience. The volunteer could be a nonconformist, and shock the victim for the sake of science; a few people did. (One obedient volunteer sounded like the more self-righteous independents of the Asch studies: "Momentarily I was ready to go along with them. Then suddenly I felt that they were just being ridiculous. What was I doing following the crowd? . . . They certainly had a right to stop, but I felt they lost all control of themselves.") But thirty-six out of the forty volunteers in this experiment conformed to the behavior of their peers, and defied the demands of the experimenter to continue – about as high a level of defiance as Milgram ever got. Other factors obviously intrude: for instance, the volunteer sees the fake teachers quitting and getting away with it. But the important feature here is that an effective situation for positive social conformity was created, one which, in Milgram's term, "liberated" the volunteer's more humanitarian impulses. With a little more of that kind of social pressure and that kind of conformity in the world, it might be harder to fight a war than to stop one.

[From Alan C. Elms, Social Psychology and Social Relevance, Chapter 4, pp. 136-146. Boston: Little, Brown, 1972. Copyright © 1972 by Little, Brown and Company; copyright renewed 1998 by Alan C. Elms.]

For the final section of this chapter, go to Experimental Ethics. For previous sections of this chapter, go to Acts of Submission.

Alan C. Elms Home Page   Virtual Library   Books Online Text Version | *Standard Version

Alan C. Elms
Alan C. Elms Home Page

All text, layout and images on the domain are Copyright © 1999-2007 by Alan C. Elms. All rights reserved.