These studies are relevant to examples of destructive obedience in modern history. It is proximity to suffering, rather than the magnitude of destruction that determines how much horror we feel, and this affects levels of obedience. These studies predict that it is easier to drop an atomic bomb from a great height, than to attack someone close up when ordered.
The Status of the source of authority also affects the levels of obedience. Thus when the experimenter left the lab, and was replaced by an ‘ordinary man’ rather than the professor, maximal levels of obedience fell to 20%. Another variation hinged on the status of Yale as an institution. When the study was moved to a slightly run down office block in Bridgeport, obedience again fell, this time to 48%. These variations indicate how authority must be legitimate or appear so, otherwise levels of obedience fall. In an effort to isolate the personality of the experimenter, Milgram switched the experimenter and learners roles, obedience fell marginally to 50%. In this case the switched experimenter still maintained a level of legitimacy.
Milgram’s studies investigated the effect of authority on single participant targets, suggesting that this might explain the high levels of obedience. In a study by Gammon et al (1981; cited in Brown, 1986) known as the MHRC study, groups of nine participants were required to give false evidence to a court in support of an oil companies litigation against one of it’s employees. Although many other factors distinguish this from Milgram’s earlier studies, the fact that participants were in groups rather than as single targets, seems to account for the fact that of 33 groups, successful rebellion against the experimental authority occurred in all but four. Rebellion is the result when an immoral, or illegitimate authority is dissipated amongst a group.
Thus number of targets/ participants is a crucial factor in predicting levels of obedience. This explains why in both studies, even though the external force to obey was in conflict with privately held attitudes and beliefs, such different levels of obedience occurred. Moscovici et al, (1969) illustrated how minorities were able to exert social influence. Their findings suggest that if rebellion and disobedience is to occur, it will be a slow process. The minority must be consistent, unified and undogmatic. Influence will begin with private acceptance of minority views, and is greater if minority views are in sympathy with current social trends. Importantly, the influence is reduced when faced with a majority that has strong prior convictions.
The source of interest in obedience studies came from seeking an explanation of how Nazi atrocities might have occurred. The high levels of obedience in Milgram’s studies were determined by proximity, status and number (often known as Social Impact Theory), but it is doubtful whether these factors alone explain the blind obedience of war criminals, or that of Milgram’s participants.
Some aspects of the experimental situation may have resulted in artificially high levels of obedience. One factor must be the incremental nature of the experiment, where the shocks were administered gradually. Another factor is that of cognitive narrowness: a preoccupation with the role and duties of being a participant, rather than the suffering of the victim, and the ensuing moral issues. To disobey would have caused conflict that would have been at odds with the participants intended role.
This implies a shift in responsibility for the action from individual to experimental organisation. Many participants thought the experiment was more important than the individual, and were told at the start that the experimenter and not they would be responsible for the victims well being. Also, if we accept that the participants accepted the legitimacy of the experimenter’s authority, they had nothing to lose by obeying: the experiment gave them anononymity. This reduces the costliness of compliance, in terms of social disapproval directed at the individual. Known as deindividuation, research has shown that a loss of personal identity occurs when an individual adopts a role which is sanctioned by legitimate authority, and this disinhibits aggression and violence. For example, Zimbardo’s famous prison study (1969; cited in Aaronson, 1992).
Conformity, another example of social influence should be distinguished from obedience, where changes in behaviour are due to authority. Conformity, is a pressure to behave in the same manner as other people in a group, and is caused by the influence of peers. It is an implicit, rather than explicit influence.
The extent of this implicit influence on others was studied by Sherif (1936; cited in Berkowitz, 1988) using an autokinetic perception task. The effect is subjective, and thus he created an ambiguous situation. Where participants made judgements alone, they quickly established personal norms that were not maintained during a group situation. The participants conformed to others judgements. Whereas a second group making judgements together quickly established a group norm that persisted when judgements were made alone.
This is an example of informational influence. We use other people’s information to inform our own judgements, especially in ambiguous or difficult situations (see Crutchfield, 1955; cited in Hewstone et al, 1996). Information exerts influence because we have a desire to be right and we rely on other people’s actions and opinions define social reality. In Sherif’s study, agreement among participants affected each individuals understanding, a principle known as conversion.
Asch (1951, 1956) set up an unambiguous situation to study whether individuals are influenced by a majority that is obviously wrong. In a control situation, when judging the relative length of lines, 99% accuracy was achieved. Surrounded by six confederates, who made wrong judgements 12 out 18 times, the participants mean error rate went up to 37%. Importantly, 76% of participants conformed at least once, and 27% made between 8 and 12 errors.
This high level of conformity can be explained in terms of normative influence. Conformity leads to approval and acceptance from those around us, consequently we succumb to normative social influence to maximise our popularity. We alter our behaviour to meet others expectations. This kind of compliance, which is public, should be distinguished from conversion, since it does not imply that Asch’s participants believed the judgements of others: they knew the majority were wrong but they still conformed. When participants were allowed to respond privately, thus reducing the effect of normative influence, conformity fell, mean error rates were now 12.5%.
Cohesiveness is the degree of attraction to a group. A study by Crandall(1988; cited in Baron et al, 1994) found that as cohesiveness developed in two groups of college students, individuals became more like each other in terms of their binge eating habits. The more we like others, and seek their approval, the more we are influenced by them.
In a variation, Asch planted a social supporter into the previously unanimous group who responded correctly. Error rates fell to 5.5%. The supporter provides an independent assessment of reality, which is enough to outweigh the potential informational value of the majority. When a confederate was planted, who not only dissented with the participant but also with the rest of the group, conformity was still only15%, even though the dissenter was more inaccurate in his judgements than the rest of the group. This illustrated the point that a unanimous majority is an important factor in conformity. Later research suggests that where the situation is ambiguous, and people are relying on informational influence, only social support will lead to reduced conformity (see Allen and Levine, 1968, 1969; cited in Hewstone et al, 1996).
Although these studies of conformity and obedience provide evidence of social influence on individual behaviour, there are many variations that must be considered. 24% of participants in Asch’s study did not conform at all, and over all, most resisted group pressure on most occasions. There was no one situation that produced 100% conformity.
Personality has an effect, although it is a less important factor in conformity than those mentioned so far. A study by Burger and Cooper (1979; cited in Baron et al, 1994) found that individuals rated by questionnaire as high on desirability for self control, conformed less than those low on this dimension, when judging how funny cartoons were. Low self confidence has been shown to increase levels of conformity.
An individuals status in the group is yet another factor. Thus Dittes and Kelley (1956; cited in Hewstone et al,1996) found that conformity was highest in medium status individuals: high status individuals can afford to deviate, and low status individuals have nothing to lose.
Although earlier studies found that women tended to conform more, Sistrunk and McDavid (1971; cited in Byrne et al, 1994) found no significant difference. They found that familiarity with the task had confounded earlier results. Males were more likely to yield to females when the tasks were those more familiar to females and vice versa. Thus informational influence, and not gender has caused conformity.
There have been many attempts to replicate Asch’s original study, but these have failed. Perrin and Spencer (1983) succeeded, at a time when the Falklands war was being fought. In war time a moral consensus develops which favours conformity over individualism. This illustrates that conformity is not a universal or cross cultural psychological phenomena. It is both historically and culturally determined. Although situation has an impact on conformity and obedience, they are transient aspects of human behaviour, and are subject to prevailing social values.
Aronson, E. (1992). The Social Animal. (6th ed.). W.H.Freeman and Co.
Baron, R. A. and Byrne, D. (1994). Social Psychology; Understanding human interaction. (7th ed.). Allyn and Bacon.
Berkowitz, .L (1988). A survey of social psychology. (3rd ed.). CBS publishing.
Brown, R. (1986). Social Psychology. (2nd ed.). Macmillan.
Hewstone, M. Stroebe, W. and Stephenson, G.M. (Eds). (1996). Introduction to Social Psychology (2nd ed.). Blackwell.
Milgram, S. (1963). Behavioural study of Obedience. Journal of abnormal and social psychology, 67, 371-78.
Moscovici, S. Lage, E. and Naffrechoux, M. (1969). Influence of a consistent minority on the responses of a majority in a colour perception task. Sociometry, 32, 365-380.
