Professional Documents
Culture Documents
Latané y Wolf 1981
Latané y Wolf 1981
Previous theorizing about social influence processes has led to the emergence of
two research traditions, each focusing on only a subset of influence situations.
Research on conformity looks at the influence of the majority on a passive mi-
nority, whereas research on innovation considers the influence of active minorities
on a silent majority. In the present article, we review these two lines of research,
as well as some recent evidence, from the perspective of a new theory of social
impact. This theory views social influence as resulting from forces operating in
a social force field and proposes that influence by either a majority or a minority
will be a multiplicative function of the strength, immediacy, and number of its
members. Social impact theory offers a general model of social influence processes
that integrates previous theoretical formulations and empirical findings and ac-
counts for the reciprocal influence of majorities and minorities. By viewing social
influence as a unitary concept, social impact theory permits comparisons between
conformity and innovation and predicts the relative magnitude of their effects.
Traditional approaches to social influence 1980; Moscovici, 1976; Moscovici & Ne-
have concentrated almost exclusively on sit- meth, 1974) suggests that individuals and
uations in which the majority serves as the active minorities can serve as sources of in-
source of influence pressure (Allen, 1965; fluence pressure. When the minority is viewed
Darley & Darley, 1976; Kiesler & Kiesler, as a source, rather than as a target, it is clear
1969). Considering dependence to be the that traditional models of social influence
process by which influence operates, re- must be modified. A minority, by definition,
searchers have sought to identify those vari- is disadvantaged in terms of those charac-
ables that increase the dependence of indi- teristics—size, status, and power—that
viduals and minorities upon the majority. would make a majority dependent upon it.
Thus, the majority's size, status, and power Turning away from dependence, Moscovici
have been systematically investigated. In the has argued that behavioral style is the mech-
context of research on conformity, the mi- anism by which influence operates, partic-
nority has been viewed as the passive recip- ularly when the influence source is a rela-
ient of forces emanating from the majority. tively powerless minority. In the context of
Recent research on innovation (Levine, research on innovation, the majority has
been viewed as the passive recipient offerees
Preparation of this article was supported by National emanating from the minority.
Science Foundation Grant BNS-7619629 to the first Moscovici's challenge to traditional con-
author. Portions of this article were presented at the ceptions has led to a rift in our theorizing
International Symposium on Social Influence Processes,
Laboratoire Europfeen de Psychologie Sociale, Barce- about social influence processes. Two dis-
lona, Spain, September 1980; at the annual meeting of tinct research traditions have emerged, each
the Kentucky Academy of Science, Lexington, Novem- offering a different view as to what consti-
ber 1980; and at the annual convention of the Mid- tutes a crucial independent variable, an ap-
western Psychological Association, Detroit, Michigan,
April 1981. Vernon Allen, Jack Brehm, Sharon Brehm,
propriate dependent variable (Moscovici,
Timothy Brock, John Darley, James Davis, John Harvey, 1980), and indeed a valid understanding of
John Herstein, Jeffrey Jackson, Norbert Kerr, Serge the process by which social influence oper-
Moscovici, Charlan Nemeth, Anthony Pratkanis, John ates. Although Moscovici (1976, p. 68) has
Pryor, Richard Sorrentino, and Maryla Zaleska pro- suggested that both the majority and the
vided valuable comments on an earlier draft.
Requests for reprints should be sent to Bibb Latane, minority can be simultaneously the source
Behavioral Sciences Laboratory, 404B W. 17th Avenue, and target of social influence, this possibility
Columbus, Ohio 43210. has not been investigated. Our purpose in
438
MAJORITIES AND MINORITIES 439
the present article will be to explore its im- minority. The greater the number of indi-
plications. viduals who espouse a position, the better
By considering the simultaneous and mu- basis they provide for establishing social
tual influence of majorities and minorities, reality (Festinger, 1954). Further, the greater
social impact theory (Latane, 1981) would the number of people advocating a position,
seem to provide a means of bridging the the- the greater are their resources for rewarding
oretical gap that has been created and in- those who conform to that position and pun-
tegrating the findings of both research tra- ishing those who deviate. Supporting the in-
ditions. This theory proposes that social creased effectiveness of larger numbers is the
influence can be understood as resulting finding that conformity to the majority po-
from social forces operating in a social force sition increases with increases in majority
field. The amount of influence produced by size, at least up to three, where conformity
either a majority or a minority will be a appears to reach an asymptote (Asch, 1951).
multiplicative function of the strength, im- Since majorities are seldom in a position
mediacy, and number of its members. of dependence upon the minority, it is not
Our plan in the present article will be to surprising that research in this tradition has
review briefly a traditional dependence model considered the majority as the source and
of majority influence and Moscovici's model the minority as the target of influence pres-
of minority influence. Then the principles of sure. In fact, most of this research has em-
social impact theory will be presented and ployed trained or simulated confederates as
their application to an understanding of so- the majority and naive subjects as the mi-
cial influence will be specified. Findings of nority. Minority influence has not been a
previous research and of some recent re- potential consequence of the influence pro-
search will be evaluated from the perspective cess in this research. From the perspective
of this theory. Finally, we will argue for the of dependence models, it would be difficult
theoretical advantages of viewing social in- to account for influence produced by a nu-
fluence as a unitary concept. merically disadvantaged minority.
quences of consistent behavior mediate in- difficult to account for majority influence
fluence. and particularly for increases in influence
The effectiveness of consistency as a source with increases in the number of influence
of influence pressure is not determined by sources.
dependence relations, and thus it is a par-
ticularly potent source of influence when the Summary
influence agent is a minority. Moscovici and
Faucheux (1972, p. 158) claim that if de- The apparent conclusion of research in
pendence relations are not salient, the num- these two traditions is that majority and
ber of individuals involved in an interaction minority influence require different explan-
should not directly affect the amount of in- atory frameworks. Minorities are inherently
fluence exerted. According to this model, the disadvantaged with respect to the variables
minority's numerical disadvantage does not hypothesized to mediate majority influ-
preclude it from exercising influence. In fact, ence—size, status, and power. Majorities are
the minority's smaller size may indirectly at a disadvantage with respect to the process
increase the influence it exerts by fostering hypothesized to mediate minority influ-
attributions of confidence and commitment. ence—perceptions of confidence and com-
Moscovici and Faucheux (1972, pp. 179- mitment created by a consistent behavioral
180) have argued that the "testimony of a style. Two models of social influence have
single subject who is perforce more consis- been proposed, neither of which can account
tent with himself is more influential than the easily for influence by both factions.
testimony of a theoretically less consistent There would appear to be theoretical ad-
subgroup." Increasing the size of a minority vantages of viewing social influence as a uni-
beyond one almost inevitably implies a re- tary concept. A general model of social in-
duction in consistency and consequently in fluence processes, by viewing majority and
influence. Likewise, Nemeth, Wachtler, and minority influence as potential outcomes of
Endicott (1977) suggest that a minority has a single process mediated by a common set
two stylistic advantages inversely related to of variables, would offer a more parsimo-
its size. By standing out against the crowd, nious account of influence data covering a
the minority gains visibility and becomes the greater variety of influence situations (cf.
focus of attention in the group. As minority Doms, Note 1). Social impact theory (La-
size increases, the majority's attention be- tane, 1981) attempts to provide such a per-
comes divided, and individual minority spective.
members become less salient. Second, by
advocating its position consistently in the Social Impact Theory
face of possible sanctions, the minority forces Social impact may be defined as any of
the attribution that it is confident and com- the great variety of changes that occur in an
mitted to its position. As minority size in- individual as a result of the real, implied, or
creases, the potential for ostracism and re- imagined presence of other individuals. So-
jection is reduced, resulting in a decrease in cial impact theory describes these effects in
perceived confidence and commitment and terms of social force fields analogous in some
consequently in influence. respects to the physical force fields that gov-
In light of the hypothesized advantage of ern the transmission of light, sound, gravity,
smaller numbers, it is not surprising that and so forth.
research in this tradition has considered the
minority as the source and the majority as Social Forces: I = f(SIN)
the target of influence pressure. In fact, this
research has typically employed trained or As an example of a social force field, La-
simulated confederates as the minority and tane (1981) describes Figure 1 in terms of
naive subjects as the majority. Majority in- the impact of several sources on an individ-
fluence has not been a potential outcome of ual target. He identifies three factors that
the influence process in this research. From determine the combined effect of the sources
the perspective of Moscovici's model, it is on the target: the strength (S) or intensity
MAJORITIES AND MINORITIES 441
(status, power, ability, etc.) of the source person in the specific situation and t is an
persons (represented by the areas of the cir- exponent with a value of less than one. That
cles), their immediacy (/) or proximity in the value of the exponent must be less than
space or time to the target, and the number one implies that impact will grow as some
(N) of source persons present. root of the number of source persons.
Thus, the social impact experienced by a Latane (1981) describes 10 diverse areas
target person should be a multiplicative of application of social impact theory in-
function of these three factors, with / (im- cluding the interest value of news events,
pact) =f(SIN), suggesting that the effect of social inhibition of response to emergencies
any one variable will be greater the greater and requests for help, tipping in restaurants,
the value of the other variables. As any one inquiring for Christ, group productivity, and
of these variables increases, there should be crowding in rats. With its few basic propo-
a corresponding increase in impact. sitions, it is clearly a broad theory applicable
to a wide range of social phenomena, which
Marginal Impact: 1 = sN' integrates and formalizes ideas developed by
psychologists working in such disparate areas
As in the case of physical stimuli, the psy- as Kurt Lewin in group dynamics and S. S.
chological effect of other people may not be Stevens in psychophysics. It does, however,
a simple linear function of their number or lead to highly specific and verifiable predic-
strength. Rather, social impact may obey tions. Here we illustrate its application to
psychosocial laws similar to the psycho- social influence phenomena.
physical laws that govern the subjective im-
pact of physical stimuli such as light and
sound intensity (Stevens, 1975). Social im- The Social Influence Situation
pact theory proposes that the effect on a tar-
get person of an increase in the number of Majority Influence
sources will be a power function, with each When influence pressure is generated by
additional source producing less impact than a unanimous majority, all of the social forces
the previous source. This principle of mar- acting on an individual target will pull him
ginally decreasing impact can be expressed in the same direction. Social impact theory
by the equation / = sN1, where s is a scaling proposes that conformity pressure will in-
constant reflecting the impact of a single crease with increases in the strength, im-
mediacy, and number of individuals advo-
cating the majority position. Further, as the
number of majority members increases, their
impact on the target person should grow as
some root of N, with the largest differences
in conformity being associated with the first
few increments in N.
Asch and the magic number three. This
situation is precisely that of the subject in
a typical Asch (1951, 1952, 1956) experi-
ment. Asch brought together groups of stu-
dents for the ostensible purpose of making
a series of perceptual judgments. Their task
was relatively easy—to choose from among
three lines the line that was equal in length
to a standard. Students did this alone, or
after 1, 2, 3,4,8, or 16 other people (actually
Figure 1. Multiplication of impact: I =f(SIN). (Figure experimental confederates) had first re-
from "Psychology of Social Impact" by Bibb Latane,
American Psychologist, 1981, 36, 343-356. Copyright sponded. When alone, the students were vir-
1981 by the American Psychological Association. Re- tually errorless in their judgments. When
printed by permission.) judging in groups, however, they were faced
442 BIBB LATANE AND SHARON WOLF
with a difficult conflict: On one third of the three individuals are necessary for the ma-
trials, the other judges unanimously reported jority to be perceived as consensual.
an obviously incorrect response. The stu- Social impact theory predicts a clearly
dents were thus forced to choose between different relationship between majority size
stating a judgment that contradicted the and conformity from the one suggested by
physical evidence or one that contradicted Asch's interpretation of his data. The failure
the reported evidence of the other judges. to find increases in conformity when major-
Asch set up a conflict between the physical ity increases in size beyond three poses only
impact of the stimulus situation and the so- a small problem for the theory, The principle
cial impact of from 1 to 16 other people. of marginally decreasing impact implies that
Overall, Asch found that the students con- the conformity curve should level off, al-
formed to the erroneous judgments of the though perhaps not quite so completely.
majority on about one third of the critical More problematic for social impact theory
trials. The amount of conformity, further, is that majorities of one and two had so little
depended upon the size of the incorrect ma- effect on conformity. One should always ex-
jority (Figure 2, a). When the majority con- pect the first person in a social force field to
sisted of but one or two individuals, there have the greatest impact. It may be that
was very little conformity. With the addition Asch's college students were sufficiently in-
of a third majority member, conformity in- dependent, at least with respect to the evi-
creased dramatically. Increases in majority dence of their own senses, to require a sub-
size beyond three, however, did not result in stantial amount of social pressure just to
increasing amounts of conformity. bring them up to a yielding threshold. The
Asch's explanation of his data focused on impact of the first two confederates, then,
the Gestalt notion of group consensus: Con- might have been in reducing restraints
formity results from the perception of agree- against making incorrect judgments.
ment among group members and not from The missing magic number. Asch's
the mere addition of their individual influ- (1951, 1952, 1956) research has led to a
ences. Once this perception has been created, number of attempted replications, the results
further increases in majority size should of which have tended to confuse rather than
have little effect. It appears that at least clarify the relationship between majority
<D
E 50
•§ 50
o
M—
§ 40 §40
u
n
'C 30 . <? ---- o ------ IT
30
"5
:E 20 8
L-
••C 20
u
I=14N"
*. 10 0,0 r's.80
0)
1234 8 16 1 2 3 4 5 6 7 8
a. b.
Figure 2. Conformity as a function of number (N) of majority members: (a) data from Asch (1951);
(b) data from Gerard, Wilhelmy, & Conolley (1968). (/ = impact.)
MAJORITIES AND MINORITIES 443
were not even present at the time of influ- majority member producing a smaller incre-
ence. On the other hand, the issues involved ment in conformity than the previous mem-
value judgments rather than matters of ber. Taken together, they provide strong sup-
physical reality, and thus the respondents' port for a social impact theory analysis of
opinions might have been more susceptible the majority influence process.
to change. We might also note the existence
of consistent sex differences in conformity Minority Influence
(Wolf & Latan6,1981). Persons with female
names were more influenced (42%) than per- In the situations just described, the social
sons with male names (30%). There were forces acting on the target were unidirec-
also sex differences with regard to specific tional, each influence source contributing to
issues—both sexes had similar views with the total pressure on the target to adopt the
respect to the adequacy of foreign and na- majority position. In other social situations,
tional, and state and local, news coverage; when influence sources are not unanimous
but females were much more satisfied with with respect to the positions they advocate,
the number of columns and features and, not the forces impinging on the target may pull
too surprisingly, with the adequacy of sports him in different directions. Consider, for ex-
coverage. However, these differences in orig- ample, the situation of an individual who is
inal attitude did not affect conformity, and the target of influence by others who are
sex did not interact with the majority size divided on the issue in question. The force
effect. exerted by the larger faction will pull the
The results of this study, coupled with target toward the majority position, whereas
those from Gerard et al. (1968), challenge the force of the smaller faction will pull the
Asch's conclusion that conformity is essen- target toward the minority. According to
tially unrelated to majority size both by their social impact theory, the magnitude of each
demonstration of systematic increases in force will be a multiplicative function of
conformity with increases in majority size the strength, immediacy, and number of
beyond three and by the significant amounts subgroup members; and the resultant force
of conformity obtained with majorities of on the target will be a simple function of the
one and two. More precisely, they suggest difference in impact imparted by each. All
that conformity is related to majority size else being equal, conformity to the majority
by a power function, with each additional position should increase as a power function
60
t
>- 40 o •<5'
11 9X I=24N- 38
Z J2 r'=.88
o o
20
P
a. .
0 1 2 3 6 12
of majority size and decrease as a power gether with one, two, three, or four confed-
function of minority size. erates to make a series of perceptual judg-
Asch and the magic number one. In ad- ments. Their task was to indicate the color
dition to his research with unanimous ma- of slides, all of which were objectively blue.
jorities, Asch (1952) conducted several stud- The confederates unanimously labeled each
ies of minority dissent. In the first study, he slide as "blue-green." The results showed
simply introduced a "partner" who re- that the number of "blue-green" responses
sponded correctly on the critical trials. The given by the naive participants was signifi-
presence of the partner reduced conformity cantly higher in all four experimental con-
from 33% to 13%. Asch suggested two fac- ditions than in a control condition composed
tors that might operate when a dissenter is entirely of naive people. A minority of three
present: (a) presence of social support or was found to be significantly more influen-
(b) lack of consensus. If social support were tial than minorities of one or two, but a
the critical factor, the dissenter would have minority of four did not differ significantly
to agree with the subject, but if breaking from the others. Although Nemeth et al.
consensus were the important factor, the report that the only significant component
dissenter could be even more incorrect than of the minority size effect was linear, our
the majority. To help choose between these reanalysis of their data shows that a power
factors, Asch introduced dissenters who ei- function with an exponent of .5 achieves a
ther answered between the majority estimate slightly better fit.
and the correct response or who were even Unfortunately, a methodological feature
more extremely incorrect than the majority. of this experiment makes it impossible to
In both cases, conformity was greatly re- obtain a precise estimate of the minority size
duced, although not eliminated. effect. All of the responses were made pub-
From these results, Asch concluded that licly, and each time a naive participant con-
the breaking of consensus was the major fac- formed to the minority position, the majority
tor responsible for conformity reduction in to minority size ratio changed. In a group of
the presence of a dissenter. Asch essentially six naive participants and four confederates,
views the group as a dynamic whole, the for example, the first person to agree with
majority and the minority being part of the the minority position obscures the majority-
total configuration. Consensus is a property minority distinction. The value of their con-
of the group as an entity, and once it is bro- clusions with respect to minority size, there-
ken, additional minority members should fore, is limited.
cause no further decrease in conformity. Minorities make a minor impression. A
This general conclusion has received only more precise investigation of the effect of
partial support from the study by Allen and minority size was conducted by Davis and
Levine (1969) and from the studies reviewed Latane (Note 3) in the context of a study
by Allen (1975). on social influence and person perception.
Rather than viewing the group as a whole, Employing a procedure used in the study of
social impact theory views the majority and impression formation, they asked partici-
the minority as separate sources of influence. pants to integrate information from several
Holding majority size constant, increases in sources in order to form a coherent impres-
the size of an opposing minority should result sion of another person. Each of up to 24
in decreasing amounts of conformity to the different persons who knew the target person
majority position, with the largest decre- gave a one-word trait description, either pos-
ments in conformity being associated with itive or negative, and participants were asked
the introduction of the first few minority to report how much they would expect to
members. like the person described on a scale from
The missing magic number. Only one -100 to +100. Each respondent rated 216
study in the minority influence tradition has different target persons described by from
investigated parametric variations in minor- 0 to 24 different people, with a combination
ity size. Nemeth, Wachtler, and Endicott of 0 to 12 extremely positive and 0 to 12
(1977) brought six naive individuals to- extremely negative adjectives listed on a sin-
446 BIBB LATANE AND SHARON WOLF
multiple positions on an issue are plausible. affecting the strength, immediacy, and num-
In the case of group members discussing a ber of sources and targets of influence, and
relevant issue, in fact, the number of units does not itself predict the effect of variables
in the force field will correspond precisely involving the judgmental issue (e.g., stimu-
to the number of discrete positions advocated lus ambiguity, task difficulty, opinion rele-
by the group members. vance, etc.) nor those concerning the per-
It may be noted parenthetically that Fig- suasive message (e.g., opinion discrepancy,
ure 7 depicts the social force field operating primacy and recency effects, one-sided vs.
in an experiment by Moscovici, Lage, and two-sided arguments, fear appeals, etc.).
Naffrechoux (1969). In that study, each na- Kelman (1958) has suggested that influence
ive subject was brought together with five can best be understood in terms of three
other subjects and two confederates to make distinct psychological processes—compli-
judgments concerning the color of each of ance, identification, and internalization—
a series of blue slides. There are two ways which differ with respect to their antecedent
to view this situation from the perspective conditions and behavioral consequences.
of the subject, depending upon the confi- Similarly, Moscovici (Moscovici, 1980;
dence with which he holds the majority po- Moscovici & Lage, 1976; Moscovici & Per-
sition and his perceived relationship to the sonnaz, 1980) has distinguished manifest
other naive subjects. In one case, the subject from latent (i.e., a change in the cognitive-
may view himself as independent and as the perceptual code underlying an overt re-
lone target of forces generated by the other sponse) influence and has proposed that
group members (he would be represented in while majorities produce the former, minor-
Figure 7 by the single I). From this per- ities produce the latter. This proposition,
spective, the other naive subjects who cor- however, has not been supported by recent
rectly label the slides as "blue" constitute evidence (Doms & Van Avermaet, 1980;
a majority, whereas the confederates who Sorrentino, King, & Leo, 1980). Social im-
label the slides as "green" constitute a mi- pact theory does not make differential pre-
nority. The total impact experienced by the dictions for these mechanisms or processes,
subject in this case should be a function of nor does it specify when the public or overt
the majority's impact minus that of the mi- expression of an opinion will become private
nority. In the second case, the subject may or internalized.
view himself as a member of the majority
and as the target of forces coming from the Conclusions
minority (he would be represented in Figure
7 by a sixth M and the force attributable to Social Influence as a Unitary Concept
the subgroup of independents would be
zero). In this case, the total influence pres- The present analysis suggests that minor-
sure experienced by the subject should be a ity influence is governed by the same prin-
function of the minority's impact divided by ciples and mediated by the same variables
that of the majority. as majority influence, the difference between
Scope of the model. While social impact the two sources of influence being purely
theory offers a broad and descriptive, yet quantitative.
precise and testable, model of social influ- The theoretical advantage of viewing ma-
ence, the model does not specify when or jority and minority influence as instances of
where social influence will occur, nor does a general influence process is both parsimony
it detail the exact processes by which social and the ability to account for a wide range
influence is transmitted (Latane, 1981). of influence phenomena. More importantly,
Rather, it provides general rules that deter- however, this formulation considers the in-
mine the magnitude of influence when it does dividual and the group as adaptive social
occur and is most useful when combined with agents. An individual may change his mind,
specific theories relevant to each area of ap- move from one position on an issue to an-
plication. other, without having to engage in a new
Social impact theory focuses on variables psychological process and without disrupting
450 BIBB LATANE AND SHARON WOLF
the ongoing influence process in the group. upon each of them. Thus, we agree with
Further, neither the majority nor the mi- Moscovici (1976) that single individuals and
nority is viewed as the passive recipient of relatively powerless minorities must inevi-
social forces from the other faction. What- tably have some influence.
ever position an individual assumes with re- Variables that have traditionally been the
spect to a given issue, he will be an active focus of dependence models—such as ma-
participant in the influence process. From jority si/e, status, and power—however, are
this perspective, the influence of majorities at the very heart of social impact theory.
and minorities is simultaneous and re- Attractiveness, expertise, control over re-
ciprocal. sources, and so on are strength variables in
Finally, the present formulation views so- the present formulation; and variations in
cial influence as only one example of the their magnitude should lead to correspond-
wide variety of effects that the presence or ing changes in impact. Further, since strength
actions of other people have on a target per- is multiplicatively related to the number of
son. By analyzing the influence situation in influence sources and majorities are by def-
terms of a social force field, social impact inition more numerous, these changes should
theory helps to integrate research on social be more pronounced when the source of in-
influence with the vast body of research on fluence is a majority than when it is a mi-
other social phenomena such as bystander nority. It is not surprising, therefore, that
intervention, responses to crowding, produc- research on majority influence has empha-
tivity in groups, and so forth. sized the role of these variables.
Some properties of the group as an entity
Relationship of Social Impact Theory to may also be viewed as strength variables.
Previous Models Consider, for example, the role that group
cohesiveness might play in the generalized
Previous models of social influence have social force field described earlier. As the
been concerned primarily with conformity cohesiveness of the group as a whole in-
or innovation, but not both. Two models of creases, the ability of any subgroup to influ-
social influence have been reviewed, each of ence members of another subgroup should
which accounts for a portion of influence increase. Similarly, as the cohesiveness of a
phenomena. Our purpose in the present ar- given subgroup increases, the ability of that
ticle has not been to suggest that either subgroup to influence its own members, or
model is inaccurate with respect to the type to resist influence pressure from other
of influence it purports to explain. Rather, subgroups, should increase as well (Back,
we have proposed a general framework by 1951; Festinger, Schachter, & Back, 1950).
which to understand the influence of both Social influence is only one of a number
majorities and minorities, which integrates of possible reactions to disagreement of
previous theoretical formulations and em- opinion in groups; others include commu-
pirical findings. Thus, while there are some nication, changes in affect, overt hostility,
points of disagreement, social impact theory and rejection (Levine, 1980). In fact, studies
has much in common with each of these ear- by Schachter (1951) and Emerson (1954)
lier models. have demonstrated a tendency for groups to
Dependence models. In contrast to pre- reject a consistent deviate, especially under
vious conceptions of majority influence, so- conditions of high group cohesiveness. In the
cial impact theory suggests that influence present formulation, rejection of the minor-
may occur in the absence of an explicit de- ity may be viewed as an active attempt to
pendence relationship between the target remove the minority from the social force
and source. The simple presence of an in- field and thus as a potential source of influ-
dividual in a social force field insures that ence (cf. Wolf, 1979). Issues such as what
he will be affected, at least to some extent, constitutes the psychological boundary of a
by every other individual in that field and group (Festinger, 1950) and how others are
that he will simultaneously have some effect chosen for the purposes of social comparison
MAJORITIES AND MINORITIES 451
(Festinger, 1954) appear to relate to the are necessary components of successful in-
more general problem of how the area of a novation by a numerically disadvantaged
social force field is determined. minority. Behavioral style may in fact be the
Finally, the present analysis suggests that only means by which a powerless minority
whenever a majority and a minority are si- can increase its strength sufficiently to have
multaneously present and of comparable a noticeable impact. It is not surprising,
strength (i.e., the minority does not possess therefore, that research on minority influ-
special resources or power), the majority will ence has emphasized the role of this variable.
have greater impact than the minority. In
the absence of constraints from external But Minorities Have Changed the World
reality, a group consensus that results from
interaction between the majority and mi- While history is replete with examples of
nority will remain on the majority side of elite minorities who have commanded and
the issue, although it should be less extreme manipulated the masses through the use of
than it would be if there were no minority political, economic, and military might, the
position advocated. present formulation suggests minorities
Moscovici's model. Consistent with lacking such resources will have a modest
Moscovici's analysis of minority influence, impact at best. When the majority and the
social impact theory views individuals and minority are of comparable strength and im-
minorities as potential sources, as well as mediacy, minority impact will be manifested
recipients, of influence pressure. In contrast primarily as a reduction in conformity to the
to Moscovici, however, the present formu- majority. From this perspective, it would be
lation suggests that their impact will be pro- difficult to account for conversion to the
portional to their strength, immediacy, and minority position or for fundamental changes
number. It will be neither excessive nor rel- in the status quo. As Moscovici has clearly
atively greater than that of a majority. pointed out, however, there are many ex-
It is possible that a single individual may amples, past and present, of individuals and
appear to exert disproportionate influence as small minorities who, although lacking pres-
a result of the power function relating influ- tige and power, have seemingly changed the
ence to the number of influence sources. The world by convincing large majorities of their
first person in a social force field should al- views. Innovation and social change are facts
ways have the greatest impact. As minority of history, and it would seem that they must
size increases, additional minority members represent an exception to the law of majority
should have a marginally decreasing effect. rule and thus to the predictions of social im-
It is also possible, although not well docu- pact theory.
mented, that a minority may derive some We would argue that the apparent ability
strength from its deviant status. Nemeth et of a minority to produce conversion, or a
al. (1977) have suggested that minority size complete abdication of the majority position,
may be inversely related to attributions of under certain circumstances actually derives
confidence and commitment that, as signs from the pervasive influence of the majority.
of strength, may increase influence. This Often social influence processes are suffi-
analysis suggests a trade-off as minority size ciently powerful that members of the ma-
decreases, the minority gaining in strength jority keep each other from noticing fun-
what it loses in numbers. Because of the damental changes in the underlying structure
multiplicative relationship between number of affairs. In the face of changing times and
and strength, however, the attributional ad- conditions, majorities continue to convince
vantages would have to be substantial to each other of truths of the past, whereas
offset a decrease in minority size. minority positions may be fresher, more el-
Much research has focused on the behav- egant or realistic than those of the outmoded
ioral style of the minority. Consistent with majority. The insistent behavior of the mi-
this line of research, social impact theory nority thus forces the majority to attend to
suggests that greater strength or immediacy the discrepancy between its position and the
452 BIBB LATANE AND SHARON WOLF
current state of the world, creating the con- (Eds.), Contemporary topics in social psychology.
ditions for a profound and often rapid Morristown, N.J.: General Learning Press, 1976.
Deutsch, M., & Gerard, H. B. A study of normative
change in the status quo. Thus, the minority and informational social influences upon individual
may serve mainly as a trigger to create an judgment. Journal of Abnormal and Social Psy-
explosive release from the powerful confor- chology, 1955, 5/, 629-636.
mity pressures exerted by the majority, Doms, M., & Van Avermaet, E. Majority influence,
minority influence and conversion behavior: A repli-
rather than as a direct source of influence cation. Journal of Experimental Social Psychology,
(cf. Wheeler's 1966 discussion of behavioral 1980, 16, 283-292.
contagion as resulting from disinhibition). Emerson, R. M. Deviation and rejection: An experi-
We suggest that opinion conversion takes mental replication. American Sociological Review,
place only when the minority position is in- 1954, 19, 688-694.
Festinger, L. Informal social communication. Psycho-
herently and demonstrably more correct logical Review, 1950, 57, 271-282.
than that of the majority. Like the child who Festinger, L. A theory of social comparison processes.
first remarked on the Emperor's lack of Human Relations, 1954, 7, 117-140.
clothes, non-elite minorities may be effective Festinger, L., Schachter, S., & Back, K. Social pres-
mainly when majorities have blinded them- sures in informal groups: A study of human factors
in housing. New York: Harper, 1950.
selves to naked reality. If the Emperor were Gerard, H. B., Wilhelmy, R. A., & Conolley, E. S.
in fact dressed, the child would, of course, Conformity and group size. Journal of Personality
be ignored. and Social Psychology, 1968, 8, 79-82.
Goldberg, S. E. Three situational determinants of con-
formity to social norms. Journal of Abnormal and
Reference Notes Social Psychology, 1954, 49, 325-329.
1. Doms, M. Moscovici's innovation effect: Towards Jones, E. E., & Gerard, H. B. Foundations of social
an integration with the conformity effect. Paper pre- psychology. New York: Wiley, 1967.
sented at the International Congress of Psychology, Kelman, H. C. Compliance, identification, and inter-
Leipzig, East Germany, 1980. nalization: Three processes of attitude change. Jour-
2. Latane, B., & Davis, D. Social impact and the effect nal of Conflict Resolution, 1958, 2, 51-60.
of majority influence on attitudes toward news me- Kidd, J. S. Social influence phenomena in a task-ori-
dia. Unpublished manuscript, 1974. ented group situation. Journal of Abnormal and So-
3. Davis, D., & Latane, B. Minority vs. majority in- cial Psychology, 1958, 56, 13-17.
fluence and impression formation. Unpublished Kiesler, C. A., & Kiesler, S. B. Conformity. Reading,
manuscript, 1974. Mass.: Addison-Wesley, 1969.
Latane, B. Psychology of social impact. American Psy-
chologist, 1981, 36, 343-356.
References Latane, B., & Nida, S. Social impact theory and group
Allen, V. L. Situational factors in conformity. In L. influence: A social engineering perspective. In P. B.
Berkowitz (Ed.), Advances in experimental social Paulus (Ed.), Psychology of group influence. Hills-
psychology (Vol. 2). New York: Academic Press, dale, N.J.: Erlbaum, 1980.
1965. Levine, J. M. Reaction to opinion deviance in small
Allen, V. L. Social support for nonconformity. In L. groups. In P. B. Paulus (Ed.), Psychology of group
Berkowitz (Ed.), Advances in experimental social influence. Hillsdale, N.J.: Erlbaum, 1980.
psychology (Vol. 8). New York: Academic Press, Moscovici, S. Social influence and social change. Lon-
1975. don: Academic Press, 1976.
Allen, V. L., & Levine, J. M. Consensus and conformity. Moscovici, S. Toward a theory of conversion behavior.
Journal of Experimental Social Psychology, 1969, In L. Berkowitz (Ed.), Advances in experimental so-
5, 389-399. cial psychology (Vol. 13). New York: Academic
Asch, S. E. Effects of group pressure upon the modi- Press, 1980.
fication and distortion of judgments. In H. Guetzkow Moscovici, S., & Faucheux, C. Social influence, con-
(Ed.), Groups, leadership and men. Pittsburgh, Pa.: formity bias, and the study of active minorities. In L.
Carnegie Press, 1951. Berkowitz (Ed.), Advances in experimental social
Asch, S. E. Social psychology. Englewood Cliffs, N.J.: psychology (Vol. 6). New York: Academic Press,
Prentice-Hall, 1952. 1972.
Asch, S. E. Studies of independence and conformity: A Moscovici, S., & Lage, E. Studies in social influence
minority of one against a unanimous majority. Psy- III: Majority versus minority influence in a group.
chological Monographs, 1956, 70(9, Whole No. 416). European Journal of Social Psychology, 1976, 6,
Back, K. Influence through social communication. Jour- 149-174.
nal of Abnormal and Social Psychology, 1951, 46, Moscovici, S., Lage, E., & Naffrechoux, M. Influence
9-23. of a consistent minority on the responses of a majority
Darley, J. M., & Darley, S. A. Conformity and devia- in a color perception task. Sociometry, 1969, 32, 365-
tion. In J. W. Thibaut, J. T. Spence, & R. C. Carson 379.
MAJORITIES AND MINORITIES 453
Moscovici, S., & Nemeth, C. Social influence II: Mi- Journal of Abnormal and Social Psychology, 1951,
nority influence. In C. Nemeth (Ed.), Social psy- 46, 190-207.
chology: Classic and contemporary integrations. Chi- Sorrentino, R. M., King, G., & Leo, G. The influence
cago: Rand McNally, 1974. of the minority on perception: A note on a possible
Moscovici, S., & Personnaz, B. Studies in social influ- alternative explanation. Journal of Experimental So-
ence V: Minority influence and conversion behavior cial Psychology, 1980, 16, 293-301.
in a perceptual task. Journal of Experimental Social Stevens, S. S. Psychophysics: Introduction to its per-
Psychology, 1980, 16, 270-282. ceptual and social prospects. New York: Wiley, 1975.
Nemeth, C., Swedlund, M., & Kanki, B. Patterning of Wheeler, L. Toward a theory of behavioral contagion.
the minority's responses and their influence on the Psychological Review, 1966, 73, 179-192.
majority. European Journal of Social Psychology, Wolf, S. Behavioral style and group cohesiveness as
1974,4, 53-65. sources of minority influence. European Journal of
Social Psychology, 1979, 9, 381-395.
Nemeth, C., Wachtler, J., & Endicott, J. Increasing the Wolf, S., & Latane, B. If laboratory research doesn't
size of the minority: Some gains and some losses. square with you, then Qube it: The potential of in-
European Journal of Social Psychology, 1977, 7,15- teractive TV for social psychological research. Per-
27. sonality and Social Psychology Bulletin, 1981, 7,
Rosenberg, L. A. Group size, prior experience, and con- 344-352.
formity. Journal of Abnormal and Social Psychol-
ogy, 1961,65,436-437. Received August 19, 1980
Schachter, S. Deviation, rejection, and communication. Revision received April 22, 1981 •