Download PDF
Intergroup Conflict and Cooperation: The Robbers Cave
Muzafer Sherif, O. J. Harvey, B. Jack White, William R. Hood, Carolyn W.
Sherif (1954/1961)
Classics in the History of Psychology
An internet resource developed by
Christopher D. Green
York University, Toronto, Ontario
(Return to Classics index )
(Classics Editor's Note: Many thanks to Joseph Trimble of Western Washington University for
bringing this work to my attention, and making it available to the "Classics" project.)
Intergroup Conflict and Cooperation: The Robbers Cave Experiment
Muzafer Sherif, O. J. Harvey, B. Jack White, William R. Hood, Carolyn W. Sherif
Preface (1954)
Preface (1961)
Chapter 1. Integrating Field Work and Laboratory in Small Group Research
Chapter 2. Approach, Hypotheses and General Design of the Study
Chapter 3. Role of Staff, Subject Selection, Experimental Site
Chapter 4. Experimental Formation of In-Groups
Chapter 5. Intergroup Relations: Production of Negative Attitudes Toward the
Chapter 6. Intergroup Relations: Assessment of In-Group Functioning and
Negative Attitudes Toward the Out-Group
Chapter 7. Intergroup Relations: Reducing Friction (Stage 3)
Chapter 8. Summary and Conclusions
Preface to Original Release -- 1954
Livros Grátis
Milhares de livros grátis para download.
In the chapters to follow the main points of a large-scale experiment on intergroup relations are
reported. It was carried out as a part of the research program of the Intergroup Relations
Project at the University of Oklahoma. In this first presentation, sufficient time and facilities were
not available to make use of data contained in recorded tapes and half a dozen short moving
picture reels. Nor was it found feasible to include introductory chapters surveying major
theories on intergroup relations and elaborating on theoretical outlines of the present approach,
which determined the formulation of the hypotheses advanced and the design of the study in
successive stages. These are presented more fully in our Groups in Harmony and Tension
(Harper, 1953), which constituted the initial work unit in the present intergroup relations project.
Therefore, a brief statement of the cardinal considerations that shaped the conception of this
approach to the study of intergroup relations is in order. It is not unfair to say that the major
existing theories fall within two broad categories in terms of the emphasis placed in formulation
of the problem and methods involved.
In one broad category of theories, the problems are expressed in terms of actualities of events
in group relations as they exist in everyday life. On the whole, theories advanced by many
social scientists fall in this broad category. In this concern over actualities the problem is
frequently not stated and discussion not developed in a way that can be tested rigorously. In
the second broad category of theories, problems are stated and analysis carried out in terms of
more rigorous-appearing concepts and units of analysis. Theories coming from psychologists
and social scientists heavily influenced by them fall within this broad category. In this line of
approach, theories are advanced without due regard to actualities, and consequently they are
plagued with serious questions of validity.
The present approach starts with a serious concern over the rise and functioning of actual small
groups in social life. The hypotheses advanced are formulated on the basis of recurrent events
reported in sociological accounts of small groups. Testing these hypotheses under conditions
that appear natural to the subjects has been a theoretical and methodological consideration of
prime importance. Therefore, a great point was made of carrying on observations without the
awareness of subjects that they were being observed and of giving priority to the uninterrupted
and uncluttered flow of interaction under experimentally introduced stimulus conditions. The
techniques of data collection were adapted to the flow of interaction, rather than cluttering or
chopping off interaction for the convenience of the experimenter. This imposed the task of
securing an experimental site which is isolated from outside influences so that results could not
be accounted for primarily in terms of influences other than the experimentally introduced ones
and the interaction on that basis.
In such a natural, life-like interaction situation, there are so many items that can be observed at
a given time that it becomes impossible to observe and report all behavioral events. Therefore,
there is the possibility of being selective in the choice of events to be observed. In testing vital
hypotheses related to intergroup relations, restricting the number of subjects to just a few is not
the proper remedy. Circumscribing the number of reactions of the subjects is no remedy.
Asking the subjects to remain within optimal distance of a microphone and asking them please
to speak one at a time will destroy the very properties of the interaction process in which we are
interested. The dining hall adjacent to the kitchen is not the place conducive to getting the
subjects to cooperate in preparing a meal of their own accord. By trying to eliminate selectivity
through such resorts we would have eliminated at the same time the essential properties of the
very things we set out to study.
(1) One remedy lies in unmistakable recurrences of behavioral trends so that the observer
cannot help observing them even if he tried to ignore them. If these trends are independently
reported by the observers of two different groups, then they serve as a check against each
other. We have secured such checks time and again in this study.
(2) The danger of selectivity can be avoided (without disrupting the flow of interaction) by
having outside observers in crucial problem situations and by having them make, for example,
their own independent status ratings in terms of effective initiative in getting things started and
(3) The most effective way of checking selectivity is the use of a combination of techniques.
This consists in introducing at a few choice points laboratory-type experiments and sociometric
questions. If the trends obtained through laboratory-type and sociometric checks are in line with
trends obtained through observations, then selectivity of observation need not worry us as far
as the relevant hypotheses and generalizations are concerned. The actual use of observational,
experimental and sociometric techniques in a combined way, whenever feasible without
cluttering the main flow of interaction, has been a major point of emphasis in our study. In our
previous work, the feasibility of using judgmental indices to tap norm formation and intra- and
intergroup attitudes was established in various studies. This series of experiments, whose logic
and techniques were made part-and-parcel of this large-scale experiment, are summarized in a
paper "Toward integrating field work and laboratory in small group research" (to appear in
Small Group Research Issue, American Sociological Review, December, 1954).
The present study has for its background the invaluable experience of the 1949 and 1953
experiments, both carried out under my direction. In 1949 the design (in three stages) went as
far as the end of Stage 2 of this 1954 study, namely in-groups were formed and intergroup
friction was produced experimentally. The 1949 study was jointly sponsored by the Attitude
Change Project of Yale University and the Department of Scientific Research of the American
Jewish Committee, to both of whom grateful acknowledgment is extended. Without the effective
help of Professor Carl I. Hovland this start could not have materialized. The second study was
attempted in 1953 in four successive stages. We succeeded in completing only two stages in
this attempt, which covered the experimental formation of in-groups. The experiment reported
here, as well as other units during the last two years, were carried out with a grant from the
Rockefeller Foundation to the University of Oklahoma, for which we are grateful.
It is a pleasure to note here the active participation of O. J. Harvey during the last four years in
the development of this program of research. Especially his doctoral thesis, entitled, "An
Experimental Investigation of Negative and Positive Relationships between Small Informal
Groups Through Judgmental Indices," constitutes a distinct contribution in demonstrating the
feasibility of using laboratory-type judgmental indices in the study of intergroup attitudes.
Without the untiring and selfless participation of O. J. Harvey, Jack White, William R. Hood, and
Carolyn Sherif the realization of this experiment and the writing of this report would have been
This program of research in group relations owes a special debt to the dedication of the
University of Oklahoma and its administrative agencies to making development of social
science one of its distinctive features. The close interest of President George L. Cross in social
science has been a constant source of encouragement and effective support. Professor Lloyd
E. Swearingen, Director of the Research Institute, has cleared our way for smooth sailing
whenever occasion arose. We have turned again and again to the encouragement and unfailing
support of Professor Laurence H. Snyder, Dean of the Graduate College.
chapter 1
Preface -- 1961
The report of this large-scale experiment dealing with factors conducive to conflict and
cooperation between groups was first released in August, 1954 and was sent in multilithed form
to colleagues active in small group research. Since then, it has appeared in condensed form in
books and journals and has been presented in lecture form at various universities and
professional associations.
In view of numerous requests from colleagues engaged in small group research, instructors in
institutions of higher learning, and the interest expressed by colleagues in political science,
economics and social work in the applicability of the concept of superordinate goals to
intergroup problems in their own areas, the original report is being released now with very
minor editorial changes.
Two new chapters have been added in the present volume. Chapter 1 presents a theoretical
background related to small group research and to leads derived from the psychological
laboratory. It was written originally at the request of Professor Fred Strodtbeck of the University
of Chicago, editor for the special issue on small group research of the American Sociological
Review (December, 1954). This chapter summarizes our research program since the mid-
thirties, which was initiated in an attempt to integrate field and laboratory approaches to the
study of social interaction. Chapter 8 was written especially for this release to serve as a
convenient summary of the theoretical and methodological orientation, the plan and procedures
of the experiment, and the main findings, with special emphasis on the reduction of intergroup
conflict through the introduction of a series of superordinate goals.
We are especially indebted to Mrs. Betty Frensley for her alert help in typing and other tasks
connected with the preparation of this volume. Thanks are due Nicholas Pollis and John Reich
for proofreading several chapters.
The experiment could not have been realized without the utmost dedication and concentrated
efforts, beyond the call of duty, of my associates whose names appear with mine on the title
page. However, as the person responsible for the proposal prepared for the Rockefeller
Foundation in 1951 and with final responsibility in the actual conduct of the experiment and
material included in the report, I absolve them from any blame for omissions or commissions in
this presentation.
On this occasion it is a pleasure to acknowledge the understanding support and
encouragement extended by the Social Science Division of the Rockefeller Foundation to this
project on intergroup relations, a research area notably lacking in systematic experimental
studies in spite of its overriding import in the present scheme of human relations.
This preface is being written with a heavy heart. The research program of which this experiment
was an important part lost a great friend by the death of Carl I. Hovland of Yale University in
April, 1961. It was Carl Hovland who, from the very inception of the research project on
intergroup relations in 1947, gave an understanding and insightful ear and an effective hand to
its implementation. The give-and-take with his searching questions, wise counsel and steadfast
friendship through thick and thin will be sorely missed in the continuation of our research
Muzafer Sherif
Institute of Group Relations
The University of Oklahoma
Norman, Oklahoma
June 5, 1961
chapter 1
Integrating Field Work and Laboratory in Small Group Research[1]
The study of small groups has become one of the most flourishing areas of research, involving
men in various social sciences and psychology. The influences responsible for the increased
preoccupation with small groups spring both from developments within various academic
disciplines and from agencies instituted for devising practical solutions for immediate
application. Brief mention of influences contributing to the flourishing state of affairs in small
group research will be helpful as orientation:
1. Theoretically and empirically, works of sociologists have historical priority in showing
persistent concern with the topic of small groups (Faris, 1953). Since the early 1920's a definite
research development in sociology related to small groups has been carried on, as represented
by the works of men like Thrasher, Anderson, Clifford Shaw, Zorbaugh, Hiller, and Whyte. In
the recurrent findings reported in this line of research, which was carried out over a period of a
good many years, one cannot help finding crucial leads for a realistic approach to
experimentation in this area.
2. Another of the major instigators of the extraordinary volume of small group research stems
from the practical concern of business and military agencies. A series of studies initiated by
Elton Mayo and his associates at the Harvard Business School in the late 1920's has
proliferated in various institutions, both academic and technological. Another impetus along this
line came from the concern of military agencies for establishing effective techniques for the
assessment of leaders.
3. Another major influence in the development of small [p. 2] group studies comes from
psychological research. Regardless of the theoretical treatment, the results of psychological
experiments almost always showed differential effects on behavior when individuals undertook
an activity in relation to other individuals or even in their presence, as can be ascertained
readily by a glance at Murphy, Murphy, and Newcomb's Experimental Social Psychology. F. H.
Allport's experiments which started around 1915 are illustrative of this point. In the l930's, it
became increasingly evident that social behavior (cooperation - competition, ascendance -
submission, etc.) could not be properly studied when the individual is considered in isolation.
Psychological "trait" theories or personality typologies fell far short in explaining social relations.
Therefore, when Moreno's work appeared in this country in the mid-thirties presenting his
sociometric technique for the study of interpersonal choices and reciprocities among individuals
(i. e., role relations), it quickly found wide application. A few years later Kurt Lewin and his
associates demonstrated the weighty determination of individual behavior by the properties of
group atmosphere. This line of experimentation was the basis of other subsequent studies
coming from the proponents of the Group Dynamics school. Some other major influences
coming from psychology will be mentioned later.
Interdisciplinary Cooperation and the Concept of "Levels"
It becomes apparent even from a brief mention of the background that men from various
disciplines contributed to make the study of small groups the going concern that it is today. As
a consequence there is diversity of emphasis in formulating problems and hypotheses, and
diversity in concepts used. This state of affairs has brought about considerable elbow-rubbing
and interdisciplinary bickering among sociologists, psychologists, and anthropologists. In this
process and through critical appraisal of each others' approaches, the interdisciplinary
approach has become a necessity for achieving a rounded picture.
Faced with the task of dealing with both psychological and sociocultural factors in human
relations problems, psychologists have too often yielded to the temptation of improvising their
own "sociologies" in terms of their preferred concepts. Sociologists, on the other hand, have
sometimes engaged in [p. 3] psychological improvisations. While sociological or psychological
improvisation at times proves necessary on the frontiers of a discipline, it is difficult to justify on
topics for which a substantial body of research exists in sociology or in psychology, as the case
may be.
On the whole, interdisciplinary cooperation has usually turned out to mean rallying
psychologists, sociologists, anthropologists, and other social scientists to toss their theories and
concepts into the ring. But, mere juxtaposition of utterances made by psychologists,
sociologists, etc., in the same room or between the covers of the same book does not bring
interdisciplinary cooperation. Nor is interdisciplinary integration possible by laying down
segments from each discipline along the same line -- one yard from psychology, one yard from
sociology, then a foot each from history and economics.
The outlines of an interdisciplinary approach appear more clearly with the realization that
"psychological" and "sociological" signify different levels of analysis. Men studying human
relations are approaching related, similar, or even the same problems at different levels of
analysis, necessitating units and concepts appropriate for dealing with events on that level. If
we are working on the psychological level, our unit of analysis is the individual; hence our
treatment must be in terms of his psychological functioning -- in concepts such as motives,
judging, perceiving, learning, remembering, imagining, etc. If we are working on a sociological
or cultural level, our concepts are in terms of social organization, institutions, value systems,
language, kinship systems, art forms, technology, etc. (Note 1).
The concept of levels holds a fairly obvious but invaluable check on the validity of research
findings. If it is valid, a generalization reached on a topic at one level of analysis is not
contradicted and, in fact, gains support from valid generalizations reached at another level. For
example, the psychologist's findings of differential behavior of an individual when participating
in the activities of his group should be (and are) substantiated by findings on the sociological
level, namely that collective action in a group has properties peculiar to the group. Checking
and cross-checking findings obtained at one level against those obtained at another level on
the same topic will make interdisciplinary cooperation the integrative meeting ground that it
should [p. 4] be.
During the last century in the social sciences and more recently in psychology, the dependence
of sub-units upon the setting or superordinate system of which they are parts has gained
increased attention, especially in view of unrewarding attempts to account for the functioning
system in an additive way. Understanding part processes is possible only through analysis of
their relations within the functioning system, as well as by analysis of unique properties of the
part process itself. Unless knowledge of the superordinate or larger functioning system is
gained first, before tackling the part processes, there is the likelihood of unwarranted
generalizations concerning the parts, and misinterpretation of the true functional significance of
the processes observed.
In this connection, an illustration from Malinowski (1922) is instructive. Malinowski describes the
complex exchange system of the Argonauts of the Western Pacific called the Kula. The
Argonauts themselves "have no knowledge of the total outline of any of their social
structure...Not even the most intelligent native has any clear idea of the Kula as a big,
organized social construction, still less of its sociological functions and implications. If you were
to ask him what the Kula is, he would answer by giving a few details, most likely by giving his
personal experiences and subjective views on the Kula...Not even a partial coherent account
could be obtained. For the integral picture does not exist in his mind; he is in it, and cannot see
the whole from the outside."
This point can be illustrated in relation to small group studies. Since Lewin's experiments in the
1940's comparing lecture and group discussion methods in changing attitudes, various studies
have shown that in the American setting skillfully conducted group discussion in which
members participate is more effective than lecture presentation of the same material. On the
basis of results obtained in the American setting, it would seem that the superiority of group
discussion methods might be universal. That this is not the case is indicated by one of the
studies in the UNESCO project in India (Murphy, 1953). In an attempt to modify caste attitudes
among college students in India using various methods, the greatest changes arose as a result
of a lecture method using emotional appeals. The [p. 5] experimenter wrote: "Contrary to our
original expectation and hypothesis, these young boys do not seem to be in a position to exploit
fully the discussion technique, in bettering their social relationships. Does it indicate that our
boys have got to be used to the democratic ways of discussion and at present prefer to be told
what are the right attitudes rather than to be allowed to talk them out?" Within a social
organization whose values clearly encourage dependence on authority and effectively
discourage settling issues on a give-and-take basis in small sub-units, particular dependencies
may become so much a part of the individual's ego system that group discussion techniques
would be less effective than methods more in harmony with the social organization in which
they take place.
Such comparative results illustrate the value of starting first with due consideration of the
sociocultural setting with its organization and values before generalizations are made about
small groups functioning as parts of that setting (cf. Whyte, 1951; Arensberg, 1951). For small
groups are not closed systems, especially in highly complex and differentiated societies such
as the United States.
Facts obtained concerning the group setting are in terms of concepts and units at the social or
cultural level of analysis. They will not give the step-by-step analysis of the particular interaction
process; they will not be adequate for the task of dealing with interpersonal relations or the
behavior of particular individual members. At this point, psychological concepts are needed for
a detailed analysis of reciprocal relations, for handling motives, perceptions, judgments, etc.
Experimental Steps toward Integration
The rest of the Chapter will be devoted to a summary statement of the prior attempts on our
part toward pulling together some relevant findings in sociology and in psychology in the study
of small groups. In these attempts the guiding considerations have been the following:
1. To extract some minimum generalizations from the sociological findings on small groups on
the one hand; on the [p. 6] other, to extract relevant principles from the work coming from the
psychological laboratory.
2. To formulate problems and hypotheses relating to one another the indications of the two sets
of relevant findings, that is, from sociological and psychological research.
3. To test hypotheses thus derived with methods and techniques which are appropriate for the
particular problem -- experimental, observational, sociometric, questionnaire, or combinations
thereof, as the case may be.
Let us start with the term "small group" itself. The term "small group" is coming to mean all
things to all people. If the concept of small groups is considered at the outset, research on
small groups will gain a great deal in the way of selection of focal problems for investigation,
and hence effective concentration of efforts.
"Small group" may mean simply small numbers of individuals. If this is the criterion, any small
number of individuals in a togetherness situation would be considered a small group. But a
conception of small groups in terms of numbers alone ignores the properties of actual small
groups which have made their study such a going concern today.
One of the objectives of concentrating on small group research should be attainment of valid
generalizations which can be applied, at least in their essentials, to any group and to the
behavior of individual members. Accordingly, one of our first tasks was that of extracting some
minimum essential features of actual small groups from sociological work. In this task there is a
methodological advantage in concentrating on informally organized groups, rather than formally
organized groups in which the leader or head and other positions with their respective
responsibilities are appointed by a higher authority, such as a commanding officer or board. In
informally organized groups, group products and the particular individuals who occupy the
various positions are determined to a much greater extent by the actual interaction of
individuals. If care is taken at the beginning to refer to the general setting in which small groups
form and function, their products and structure can be traced through longitudinal observation
of the interaction process.
[p. 7] On the basis of an extensive survey of sociological findings, the following minimum
features in the rise and functioning of small groups were abstracted:
(1) There are one or more motives shared by individuals and conducive to their interacting with
one another.
(2) Differential effects on individual behavior are produced by the interaction process, that is,
each individual's experience and behavior is affected in varying ways and degrees by the
interaction process in the group (Note 2).
(3) If interaction continues, a group structure consisting of hierarchical status and role
relationships is stabilized, and is clearly delineated as an in-group from other group structures.
(4) A set of norms regulating relations and activities within the group and with non-members
and out-groups is standardized (Note 3).
Interaction is not made a separate item in these minimum features because interaction is the
sine qua non of any kind of social relationships, whether interpersonal or group. Since human
interaction takes place largely on a symbolic level, communication is here considered part and
parcel of the interaction process.
When group structure is analyzed in terms of hierarchical status positions, the topic of power
necessarily becomes an integral dimension of the hierarchy. Power relations are brought in as
an afterthought only if this essential feature of group hierarchy is not made part of the
conception of group. Of course, power does in many cases stem from outside of the group, and
in these cases the nature of established functional relations between groups in the larger
structure has to be included in the picture.
Our fourth feature relates to the standardization of a set of norms. The term "social norm" is a
sociological designation referring generically to all products of group interaction which regulate
members' behavior in terms of the expected or even the ideal behavior. Therefore, norm does
not denote average behavior (Note 4). The existence of norms, noted by sociologists, has been
experimentally tested by psychologists in terms of [p. 8] convergence of judgments of different
individuals (Sherif, 1936), and in terms of reactions to deviation (Schachter, 1952). A norm
denotes not only expected behavior but a range of acceptable behavior, the limits of which
define deviate acts. The extent of the range of acceptable behavior varies inversely with the
significance or consequence of the norm for the identity, integrity, and major goals of the group.
With these minimum essential features of small informally organized groups in mind, a group is
defined as a social unit which consists of a number of individuals who, at a given time, stand in
more or less definite interdependent status and role relationships with one another, and which
explicitly or implicitly possesses a set of norms or values regulating the behavior of
the individual members, at least in matters of consequence to the group.
Common group attitudes or sentiments are not included in this definition because social
attitudes are formed by individuals in relation to group norms as they become functioning parts
in the group structure. At the psychological level, then, the individual becomes a group member
to the extent that he internalizes the major norms of the group, carries on the responsibilities,
meets expectations for the position he occupies. As pointed out by various authors, his very
identity and self conception, his sense of security become closely tied to his status and role in
the group through the formation of attitudes relating to his membership and position. These
attitudes may be termed "ego-attitudes" which function as constituent parts of his ego system.
On the basis of findings at a sociological level, hypotheses concerning the formation of small in-
groups and relations between them were derived and tested in our 1949 camp experiment
(Sherif and Sherif, 1953). One of the major concerns of that study was the feasibility of
experimental production of in-groups among individuals with no previous role and status
relations through controlling the conditions of their interaction.
The hypotheses tested were:
(1) When individuals having no established relationships