Page 116 - The Social Animal

98 The Social Animal

wrong or based on misinformation. The greater the disagreement,
the greater our discomfort. How can we reduce this discomfort?
Simply by changing our opinions or actions. The greater the dis-
agreement, the greater our opinion change will be. This line of rea-
soning, then, would suggest that the communicator should argue for
the daily program of rigorous exercise; the greater the discrepancy,
the more the opinion change. Indeed, several investigators have
found that this linear relation holds true. A good example of this re-
54
lation was provided by an experiment by Philip Zimbardo. Each of
the college women recruited as participants for the experiment was
asked to bring a close friend with her to the laboratory. Each pair of
friends was presented with a case study of juvenile delinquency, and
then each of the participants was asked, separately and in private, to
indicate her recommendations on the matter. Each participant was
led to believe her close friend disagreed with her—either by a small
margin or by an extremely large margin. Zimbardo found that the
greater the apparent discrepancy, the more the participants changed
their opinions toward what they supposed were the opinions of their
friends.
However, a careful look at the research literature also turns up
several experiments disconfirming the line of reasoning presented
55
above. For example, James Whittaker found a curvilinear relation
between discrepancy and opinion change. By curvilinear, I mean
that, as a small discrepancy increased somewhat, so did the degree of
opinion change; but as the discrepancy continued to increase, opin-
ion change began to slacken; and finally, as the discrepancy became
large, the amount of opinion change became very small. When the
discrepancy was very large, almost no opinion change was observed.
Building on Whittaker’s finding, Carl Hovland, O. J. Harvey,
56
and Muzafer Sherif argued that, if a particular communication dif-
fers considerably from a person’s own position, it is, in effect, outside
of one’s latitude of acceptance, and the individual will not be much in-
fluenced by it. In the experiment by Hovland and his colleagues, the
communication was based on a red-hot issue—one the participants
felt strongly about: whether their state should remain “dry” or “go
wet”—that is, whether to change the law prohibiting the distribution
and sale of alcoholic beverages. The voters of the state were virtually
equally divided on this issue, and the participants were a representa-
tive sample: Some of the participants felt strongly that the state