400 Social Influence and Group Dynamics
complicate matters even further, research began to find evi-
dence of movement in bothdirections after a group discus-
sion (Doise, 1969; Moscovici & Zavalloni, 1969), suggesting
that both risky and cautious shifts were different manifesta-
tions of a more basic phenomenon. Based on a review of this
research, Myers and Lamm (1976) identified what they felt
was the underlying process. According to their group-
polarization hypothesis, the “average postgroup response
will tend to be more extreme in the same direction as the
average of the pregroup responses” (p. 603). Imagine two
groups, each consisting of four individuals whose opinions
vary in their respective preferences for risk. The average
choice of members is closer to the risky end of the caution-
risk dimension in one group, but closer to the cautious end of
this dimension in the other group. The group-polarization
effect predicts that the first group should become riskier as a
result of group discussion (i.e., a risky shift), but that the sec-
ond group should become more cautious during its delibera-
tions (i.e., a cautious shift). The evidence cited by Myers and
Lamm (1976) is consistent with this prediction and is widely
accepted today as a valid empirical generalization regarding
group dynamics.
This straightforward generalization proved to be resistant
to a simple theoretical account. Most theorists eventually en-
dorsed the value account (e.g., Myers & Lamm, 1976; Pruitt,
1971; Vinokur, 1971), although it didn’t take long for differ-
ent variations on this general theme to emerge. Of these, two
have stood the test of time (thus far). Social comparison the-
ory holds that people attempt to accomplish two goals during
group discussion: evaluating the accuracy of their position by
comparing it with the positions of other group members, and
creating a favorable impression of themselves within the
group. The confluence of these two motives results in a ten-
dency to describe one’s own position in somewhat more ex-
treme terms (e.g., Goethals & Zanna, 1979; Myers & Lamm,
1976). Persuasive-arguments theory, meanwhile, stresses the
importance of the information obtained during group discus-
sion. Whether there is a shift toward risk or toward caution
depends on the relative persuasiveness of the arguments
favoring each position (e.g., Burnstein & Vinokur, 1977;
Vinokur & Burnstein, 1974). The distinction between these
two accounts corresponds to the distinction introduced earlier
between normative and informational influence. Social com-
parison theory, with its emphasis on self-presentation at-
tempts to match the perceived group norm, can be understood
in terms of normative influence. The persuasive-arguments
perspective, meanwhile, is practically synonymous with
the rationale of informational influence. As noted in our ear-
lier discussion, these two forms of influence often co-occur,
so it should come as no surprise that social comparison and
persuasive arguments often work together to promote polar-
ization in groups (cf. Forsyth, 1990).Minority InfluenceIn the film Twelve Angry Men,the character played by Henry
Fonda turned his one-man minority into a unanimous major-
ity during jury deliberations so that an innocent man could go
free. In the face of virulent opposition, Galileo struggled for
acceptance of his proof of Copernican theory that the planets
revolve around the sun. This acceptance did not come during
his lifetime, but his influence lived on and eventually turned
the intellectual tide for subsequent generations. Martin
Luther King Jr. and Mahatma Gandhi both defied the prevail-
ing norms of their respective cultures and brought about sig-
nificant social and political change. And in everyday life,
people with opinions or lifestyles out of step with those of the
majority often manage to preserve their personal perspective,
sometimes even overcoming the majority’s disapproval and
winning acceptance. If conformity were the only dynamic at
work in social groups, these examples could be dismissed as
aberrations with no implications for our understanding of so-
cial influence processes. One can envision groupthink and
group polarization carried to the extreme, with the complete
suppression of minority opinion and a resultant interpersonal
homogeneity.
Far from representing aberrations, these examples suggest
that there is more to social life than accommodation by the
minority to majority influence. Even in small social groups, it
is possible for a lone dissenter to be heard and to convert oth-
ers to his or her point of view. At a societal level, minority in-
terests and opinions manage to survive in the face of majority
disapproval and hostility, and can sometimes manage to be-
come dominant forces in the culture. In recognition of these
facts of social life, minority influencehas emerged as an im-
portant topic in social psychology (cf. Moscovici, 1976).
Much of this research attempts to identify factors that enable
minority opinions to persist in groups. Experiments in the
Asch tradition, for example, have found that both group size
and unanimity of the majority have important effects on con-
formity. The relation between group size and conformity
appears to be logarithmic, such that conformity increases
with increasing group size up to a point, after which the addi-
tion of more group members has diminishing impact (Latané,
1981). Asch’s own research showed that conformity is re-
duced if the group opposing the subject is not unanimous.
Even one dissenter among the confederates emboldens the
naive subject to resist group pressure and express his or her