Theories of Social Conflict

Author(s): Anthony Oberschall

Source: Annual Review of Sociology , 1978, Vol. 4 (1978), pp. 291-315
Published by: Annual Reviews

Stable URL: https://www.jstor.org/stable/2945973

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Annual Reviews is collaborating with JSTOR to digitize, preserve and extend access to Annual
Review of Sociology

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
Ann Rev. Sociol. 1978. 4.291-315
Copyright (3 1978 by Annual Reviews Inc. All rights reserved

THEORIES OF SOCIAL *:.10557

CONFLICT

Anthony Oberschall

Department of Sociology & Anthropology, Vanderbilt University,

Nashville, Tennessee 37235

INTRODUCTION

Conflict results from purposeful interaction among two or mor

a competitive setting. It refers to overt behavior rather than to p
action and to subjective states. According to Deutsch (1973:10), "c
tion implies an opposition in the goals of ... interdependent parties such
that the probability of goal attainment for one decreases as the probability
for the other increases." Whereas a competitive situation might exist with-
out any awareness of it by the parties concerned, according to Boulding
(1963:5) conflict "is a situation of competition in which the parties are aware
of the incompatibility of potential future positions and in which each party
wishes to occupy a position that is incompatible with the wishes of the
other."
"Social" conflict refers to conflict in which the parties are an aggregate
of individuals, such as groups, organizations, communities, and crowds,
rather than single individuals, as in role conflict. Group conflict is used as
a synonym of social conflict in this essay. Finally, social conflict refers in
common usage to interaction in which the means chosen by the parties in
pursuit of their goals are likely to inflict damage, harm or injury, but not
necessarily in every case. With this small proviso, Coser's definition of social
conflict conveys its meaning very well (1967:232): "social conflict [is] a
struggle over values or claims to status, power, and scarce resources, in
which the aims of the conflict groups are not only to gain the desired values,
but also to neutralize, injure, or eliminate rivals." Social conflict encom-
passes a broad range of social phenomena: class, racial, religious, and
communal conflicts; riots, rebellions, revolutions; strikes and civil disorders;
marches, demonstrations, protest gatherings, and the like.
291
0360-0572/78/0815-0291 $01.00

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
292 OBERSCHALL

THE SCOPE OF CONFLICT THEORIES

Just what do theories of social conflict seek to explain? Any comprehensive

theory of social conflict should encompass the following topics:

1. The structural sources of social conflict, in particular structures of domi-

nation that make struggles over values and scarce resources likely. At
this stage, a theory of social conflict will rely heavily on stratification,
social change, and macrosociological theories. These theories will iden-
tify the most important explanatory variables in conflict theories.
2. Conflict-group formation and the mobilization for collective action of
challenging groups and their targets. For this topic, theories of collective
action, recruitment, participation, commitment, and internal structure
will be especially useful.
3. The dynamics of conflict: processes of interaction between conflict
groups; the forms of conflict; its magnitude scope, and duration; escala-
tion and de-escalation; conflict regulation and resolution; the conse-
quences of conflict outcomes for the contending groups and the larger
society. These are the most important dependent variables of social
conflict theories.

This review is mainly concerned with describing and evaluating the two
principal approaches to conflict theory: the breakdown-deprivation ap-
proach and the solidarity-mobilization approach. The older classics and the
well-known theories based on them like Dahrendorf (1959) and Coser
(1956) are not reviewed here, because their contributions have become part
of the shared fund of sociological knowledge. However, I indicate in a short
example how the classics might be more tightly incorporated into current
conflict theories.1
This review also omits evaluations of the conflict perspective in sociologi-
cal thinking, because a perspective or orientation is not a theory of conflict
defined by the three broad topics. For instance, Randall Collins' Conflict
Sociology, Towards an Explanatory Science (1975), does not contain the
statement of a theory of social conflict. It is rather a work of general
sociology written from a conflict perspective. Collins writes little about
overt conflict, such as strikes, rebellion, and collective violence. He does not
discuss the forms of conflict, the formation of conflict groups, mobilization,
recruitment, and leadership, nor outcomes and conflict resolution. And he
deals only briefly with major and common group conflicts such as ethnic,
racial, and communal conflicts. He seeks instead to explain how structures
of domination based on social class, gender, and age differences have come

'A comprehensive review of the literature on social conflict is provided by Kriesberg (1973).
This essay is primarily concerned with modes of theorizing and theory construction.

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
 SOCIAL CONFLICT 293

about and how they are maintained. These structures of domination have
a potential for conflict, but the link between these structures and overt
conflict is not explicitly spelled out.
However, I am concerned with two issues of theory construction that are
particularly salient in theorizing about social conflict. The first is levels of
analysis. Some theorists treat a conflict group as a single collective actor.
Their theories consist of statements about groups and the relations between
groups. Other theorists start with individual behavior as the unit of theoriz-
ing, and derive group behavior from a summation of individual tendencies
and behaviors. Still other theorists handle the problem of levels of analysis
and of moving from the individual to the group level in more complicated
ways. Does the mode of theorizing lead to different insights and different
kinds of conflict theory?
The second issue is the use of different assumptions about human behav-
ior. Some theorists utilize the rational-choice model derived from economic
analysis; others incorporate complex psychological and social psychological
assumptions into models. Are these approaches mutually exclusive or com-
plementary? Do their conclusions differ?

Theory, Method, and the Classics: A Short Example

Both of these issues in theory construction, as well as the incorporation of
knowledge derived from the classics into current theory, are best discussed
in the context of a short example.
Two frequently noted empirical generalizations analyzed at length by
Coser concern the internal structure of conflict groups: "conflict with an-
other group leads to ... increased cohesion of the group," and "close-knit
groups in which there exists a high frequency of interaction and high
personality involvement of members have a tendency to suppress conflict"
(Coser 1956:95, 152). Granted that conflict with outgroups creates cohesion
and internal conformity within a group, and granted also that these are
common consequences of conflict, how can such knowledge about internal
group structure contribute to explanations about the process of conflict and
the forms it is likely to take?
One important set of consequences that appears to follow from Coser's
propositions was described in Janis' (1972) recent work on "groupthink".
It refers to a concurrence-seeking tendency that results in a deterioration
of mental efficiency, reality testing, and moral judgement. Janis' idea can
be summarized by his proposition that "the more amiability and esprit de
corps among members of a policy making group, the greater is the danger
that independent critical thinking will be replaced by groupthink, which is
likely to result in irrational and dehumanizing actions against outgroups"
(1972:13). Groupthink fosters overoptimism, lack of vigilance, and slogan-
istic thinking about the immorality of outgroups.

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
294 OBERSCHALL

If conflict leads to cohesion and concurrence-seeking tendencies resulting

in groupthink within the leadership circle of conflict groups, the leadership
will underestimate the cost of further conflict and overestimate the chances
of success (overoptimism, risk-taking), and will also be more ready to resort
to coercive means (belief in the group's superior morality, taking dehuman-
izing actions against outgroups, ignoring the moral consequences of ac-
tions). If leadership in both conflict groups undertakes such actions,
escalation of the conflict will result. Thus, supplementing Coser's proposi-
tion with Janis' theory allows one to link changes in the internal structure
of conflict groups with the mode of conflict they are likely to choose and
the form that conflict is likely to take. Cohesion, under some circumstances,
lowers the quality of leadership decision making in such a way that destruc-
tive strategies will be chosen by both sides, which in turn will escalate the
conflict.
The foregoing example merits two methodological comments. The first
concerns the inclusion of social psychological processes into rational-choice
models in conflict studies. Janis' groupthink proposition specifies the condi-
tions under which rational decision makers collectively arrive at "bad,"
"irrational" decisions as a result of distortions in their perceptions and their
critical faculties. Rational decision makers weigh the benefits and costs of
alternative courses of action, and choose the alternative that maximizes
expected net benefits, i.e. the excess of benefit over cost. Janis' proposition
suggests that groupthink results in underestimation of the costs of coercive
action and in overoptimism about its chances of success. Thus expected
benefits of coercive actions are overestimated.
Rather than sacrifice the rational-choice model of decision making in
conflict situations, Janis' work suggests the manner in which psychological
and social psychological variables might improve it. The framework of the
rational-choice model can be kept. It is the actors' calculation of costs,
benefits, and chances of success and failure that are changed under various
circumstances. Decision makers who make "bad" decisions seek to maxi-
mize expected net benefits, as others do. It is the information and percep-
tions utilized in the calculation of costs and benefits that become faulty
under groupthink conditions and lead to "unrealistic" decisions.
The second lesson of the groupthink example for conflict analysis con-
cerns the relationship between the individual and the aggregate level of
analysis. For Janis, each individual participant in the policy making leader-
ship group is a rational, moral individual. As individuals they are "the best
and the brightest," to use Halberstam's felicitous term. Yet the output of
the group is not simply an average of these individual tendencies. Groups
have emergent properties, here the concurrence-seeking tendencies of the
cohesive group, that fundamentally alter the group's collective output.

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
 SOCIAL CONFLICT 295

Theories of group conflict will have to pay special attention to group struc-
ture and the emergent properties of group. Group behavior should not
simply be assumed to be analogous to individual behavior. Social psycholog-
ical knowledge will help locate emergent group properties for conflict theo-
rists.

Theory and Method: Does Mobilization Theory

Make a Difference?

A further issue in conflict theory is whether it is possible to account for

forms of conflict directly from an analysis of structures of domination,
without recourse to mobilization and collective action theories. Rather than
discussing the issue in the abstract, I evaluate Agrarian Revolution (Paige
1975) from this perspective, because it is an ambitious and systematic effort
to link structures of domination in agriculture export sectors with forms of
conflict. The variables and hypotheses are clearly stated. An empirical test
of the theory based on 135 agriculture export factors in 70 underdeveloped
countries and colonies in the period 1948-1970 is provided and is supple-
mented with three case studies. For Paige, conflict is the product of interac-
tion: It is the interaction of two groups, and their characteristics, that
determines the forms of conflict. Mobilization and collective action are not
treated by Paige as distinct processes analyzed in their own right. Rather,
they deterministically follow from the structures of domination. In his
approach to theorizing, Paige employs a series of dichotomies. Group be-
havior and the behavior of the typical member of the group are essentially
similar. This mode of theorizing allows for great clarity of exposition at the
price of overlooking the diversity of rural economies, social classes, and
ethnic groups.
The most fully developed part of Paige's theory is on structures of domi-
nation. The structure of domination in agriculture export economies is
determined, according to Paige, by the conjunction of the principal sources
of income of ordinary cultivators and of rural elites. The commercial haci-
enda economy is characterized by cultivators and elites both deriving their
incomes from land. Because of low technology and productivity, and be-
cause of a competitive disadvantage in world markets, more land must be
acquired by either group for increased incomes. Thus the hacienda economy
tends to provoke bitter conflicts over land. The form such conflict usually
takes is the agrarian revolt. When cultivators derive their income from
wages and rural elites from ownership of land, which is typical of migratory
labor estates or of a sharecropping economy, the power of rural elites rests
on political control that enables them to maintain a hold over land, capital
and labor. Conflict will then ultimately center on control of the political

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
296 OBERSCHALL

system and will take on revolutionary forms, of either the socialist or

nationalist variety.
When rural elites derive their income primarily from the control of
capital (i.e. exports, marketing, machinery, storage, and transportation),
and cultivators derive their income from land, a situation characteristic of
small-holdings export agriculture, Paige hypothesizes that commodity
movements for limiting the power of middlemen will take place. The planta-
tion economy, the remaining possibility, consists of elites deriving income
from capital, and cultivators earning wages. It gives rise to conflicts over
wage levels within the context of labor movements. In sum then, when rural
elites earn their living from control of land, conflict is more intense and aims
at the overthrow of the structure of domination. When rural elites derive
their living from capital, conflict is over a greater share of the economic pie
and has a reformist, rather than a revolutionary, character.
Paige's theory further specifies the variables and processes that lead from
the structure of domination (income sources) to overt conflict by way of
economic behavior and political behavior (1975:21 ff.). For instance, cul-
tivators who derive their income from wages, he argues, will tend to develop
cooperative relations among themselves, will be independent of elites, and
will accept a risk strategy (e.g. willingness to strike), which in turn will lead
to radical, solidary, collective political action. Cultivators dependent on
land for their income, on the other hand, will tend to be conservative and
low on solidarity, and will engage in individual acts of opposition. Paige
short-circuits a theory of mobilization and collective action when he derives
political behavior (ideology, solidarity, collective action) deterministically
from the structure of domination and the economic behaviors and outlooks
associated with it.
By doing so, Paige's theory creates internal contradictions and inconsis-
tencies, which can be overcome only by an ad hoc appeal to variables that
are exogenous to his theory. It follows from Paige's own theory that in the
commercial hacienda elites will have strong political control, and that the
cultivator's opposition will be conservative, individualist, and weak on
solidarity. Though the theory predicts that "there will be little or no peasant
political activity," yet "periodic uprisings have been a constant part of
manorial economies" (Paige 1975:42). To explain the inconsistency between
theory and data, Paige notes that "peasant rebellions in commercial haci-
enda systems depend on the weakening of the repressive power of the landed
aristocracy, the introduction of organizational strength from outside the
peasant community, or both" (1975:42). These usually result from internal
divisions within the landed elite, loss of military power due to a major war,
and organizing drives by urban militants. It is precisely these kinds of
variables that theories of mobilization and of collective action incorporate
more systematically into an explanatory scheme.

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
 SOCIAL CONFLICT 297

One also suspects that in the revolutionary movements based on socialist

and communist ideologies, which Paige finds characteristic of sharecrop-
ping systems, mobilizing groups from outside the agriculture export sector
typically play an important part, and that the same is true for nationalist
opposition movements in colonial landed estates based on migratory labor.
Indeed, Paige recognizes this when he writes that "the explanation of these
revolts [in migratory labor systems], like the explanation of revolts in
commercial hacienda systems, must depend on the introduction of political
organizations from outside the workers' community" (1975:68).
Structures of domination in the agriculture export sector in many parts
of the world existed for a long time prior to the post-World War II period
Paige studied. One seriously doubts, however, that for the 1920s and 1930s
he would have found, as he did for the 1950s and 1960s, that "most of the
events in the study population were revolutionary, socialist, or nationalist,
which altogether accounted for 70% of the total" (1975: 101). The loosening
social control in overseas empires after World War II and the much greater
likelihood of outside support, both domestic and international, for rural
movements, must have been decisive.
In conclusion, what can be learned from an analysis that relies on struc-
tures of domination as the cornerstone of conflict theory? First, it shows the
usefulness of macrosociological views of stratification and of social change
for understanding the potential for conflict in a social structure, who the
likely antagonists will be, and their relative power. Second, it signals to
conflict theorists that to pass directly from structures of domination to
group conflict, overlooking mobilization and collective action, makes the
explanation of many of the dimensions of conflict (e.g. magnitude, duration,
timing, forms, outcomes) difficult.

Breakdown and Solidarity

With these preliminary remarks on theory construction completed, a com-

parison and evaluation of the two principal approaches to conflict theory
can be undertaken. What accounts for the production of conflict and the
formation of conflict groups? The most common view in sociology has been
the "breakdown" theory. Its social psychological foundation rests on the
grievance-frustration notion.2
Briefly stated, breakdown theory points to the dissolution of traditional
social formations and communal solidarities as a result of rapid social
change. Social disorganization, demographic pressures, and ecological im-

2The social change and modernization literature that incorporates breakdown ideas is im-
mense. Two sophisticated exponents of breakdown views are Huntington (1968) and Smelser
(1968).

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
298 OBERSCHALL

balance lead to the accumulation of strains, frustrations, insecurity, and

grievances, and the resulting pressure cooker has a tendency to explode in
collective violence and civil disorders. After a time, processes of integration
take the upper hand. Individuals become incorporated into new social
formations and associations. Strain decreases, and grievances are pursued
through regular institutional channels.
For breakdown theorists a sharp discontinuity exists between collective
violence and more institutionalized forms of social conflict. The two forms
of conflict require different conceptualization and theory. Breakdown theo-
rists stress the similarity between the roots of collective violence and other
forms of deviant and anomic behavior such as crime, mental illness, and
suicide. They emphasize the marginality of participants in collective vio-
lence. They expect conflict to locate in growing industrial centers where
anomie prevails, or else in weakened, disorganized rural communities. Fre-
quently, they see collective violence as irrational tension release rather than
as purposeful collective action to defend or obtain collective goods (Tilly
et al 1975:4 ff.).
In recent years, breakdown theories have been increasingly criticized on
theoretical as well as empirical grounds, and a rival solidarity theory has
taken shape (Tilly et at 1975). Solidarity theorists maintain that uprooted
masses do not account for most collective protest; stress that conditions that
lead to violent protest are essentially the same as those that produce other
forms of collective action; view all forms of collective action, including
violent ones, as essentially purposeful, rational pursuits or defenses of col-
lective interests; and note that violence is most frequently initiated and
perpetrated by the agents of social control. Solidarity theorists do not
minimize the effects of large-scale social changes upon the incidence and
forms of social conflict, but they maintain that their impact on social
conflict does not derive primarily from the production of tensions and
grievances. Grievances and disaffection are a fairly permanent and recur-
ring feature of the historical landscape. Social, economic, and political
change act indirectly upon incidence and forms of conflict by changing the
mobilization potential of various social formations, by changing the social
milieu and ecological locus of conflict, and by changing the social control
capabilities of the authorities.
Still another line of criticism of breakdown theories challenges its central
assertion that rapid social change destroys traditional social formations and
communal solidarities. In a recent book entitled Internal Colonialism: The
Celtic Fringe in British National Development, 1536-1966 Hechter (1975)
deals with the persistence of ethnic domination and stratification in the
British Isles over four centuries despite the alleged leveling and integrating
effects of the spread of education and industry, administrative centraliza-

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
 SOCIAL CONFLICT 299

tion, increased communications, labor mobility, the expansion of citizen-

ship rights, and policies of linguistic and religious assimilation. In contrast
to diffusionist and evolutionary views, which maintain that these trends
undermine the cultural, political, and economic foundations of ethnic
solidarities and identifications, Hechter puts forward the notion of "internal
colonialism." In this view, the superordinate core group institutionalizes its
original advantages over the peripheral ethnic minority in a stratification
system based on a cultural division of labor in which the minority's access
to certain roles are denied and their share of societal resources limited. The
leveling effects expected by diffusion theory are neutralized, because the
economic development of the periphery proceeds in a dependent mode, and
because control of political power and of the administrative apparatus of
the state remains in core hands. In a stratification system based on the
observable cultural differences of religion, language, and life styles, the
disadvantaged group keeps reasserting its own culture and ethnic worth
despite considerable costs and pressures for assimilation. Thus a system of
ethnic stratification in the internal colonialist setting shows remarkable
staying power.
For the specific case of the British Isles, Hechter shows that in compari-
sons of English, Scottish, Welsh, and Northern Irish counties, economic
indices show only slight convergence and indicate that long-term regional
economic inequalities persisted. Even though the use of the Celtic language
has declined largely as a result of the spread of compulsory public education
in the English language, differences in religious affiliation between core and
periphery have persisted and served to maintain ethnic solidarity and
boundaries. Hechter's analysis of election returns also shows that the insti-
tutionalization of class-based politics and voting patterns in the English
core took hold only partially in the Celtic periphery where status-group
voting persisted and Scottish, Welsh, and Irish separatism and nationalism
periodically kept erupting.
The persistence and resurgence of ethnic and nationality conflicts in old
states and developed industrial societies are troublesome to explain from the
vantage point of theories that associate major social conflicts with the
dissolution of traditional social bonds.

RELATIVE DEPRIVATION AND BREAKDOWN AS A THEORY OF CON-

FLICT Because it is a well-known, recent statement of the breakdown
theory of conflict, I turn to a critical examination of Gurr's causal model
of civil strife (1968, 1970). Gurr's central hypothesis is that relative depriva-
tion is the basic precondition of civil strife, and that the greater the depriva-
tion, the greater the magnitude of strife. Relative deprivation in turn is
produced by a discrepancy between what people think they are entitled to

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
300 OBERSCHALL

and what they are actually getting. As deprivation increases, frustration and
anger will ensue. These psychological states will produce aggression. At the
level of aggregates, many aggressive acts and tendencies will produce civil
strife.
The relationship between discontent and strife is mediated, however, by
a number of intervening conditions either facilitating or inhibiting overt
conflict. Just as punishment, actual or threatened, will inhibit individual
aggression, increased size and use of social control agents against regime
opponents can be expected to have a deterrent and depressing effect on
group conflict. Gurr refers to such inhibiting conditions simply as the
"coercive potential" of the regime.
A second inhibiting condition, called "institutionalization" refers to the
existence of stable, enduring and strong associations and solidarities beyond
the primary group level (Gurr 1968:1105). For Gurr, such groups provide
members with an opportunity to obtain what they think they are entitled
to. Their presence will thus lower deprivation, or else provide nonviolent
means of voicing discontent. In either case, institutionalization will inhibit
civil strife.
Other facilitating conditions identified by Gurr are inaccessibility of
regime opponents due to poor transportation in rugged terrain, the strength
of subversive groups such as the Communist party, external support for the
initiators of civil strife, and past levels of strife, which create a climate of
opinion tolerant of violent responses to discontent (Gurr 1968:1106,1114,
1115). Finally, legitimacy of the regime is assumed to have an inhibiting
effect on civil strife.
Many problems of operationalizing and measuring these variables and
processes are admitted by Gurr. There are also difficulties of interpreting
the result of multivariate analyses based on cross-national, static compari-
sons. Here I comment only on the theoretical issues raised by Gurr's model
as an example of conflict theory. The most completely spelled out parts of
the model are those concerned with relative deprivation and institutionali-
zation. Both together make Gurr's model a variant of the breakdown-
frustration explanation of social conflict. The other inhibiting and
facilitating variables in the model come from empirical generalizations in
past studies of civil strife, or from widely accepted notions about conflict,
rather than from theory.
As used by Gurr, the relative deprivation notion lacks explanatory
power. Relative deprivation is the discrepancy between what people think
they are entitled to, and what they actually get. Relative deprivation, by
definition, involves social comparison. One develops notions about what one
is entitled to relative to what some other group is getting, or relative to some
norm of equity. Gurr's theory is silent on the choice of the comparison

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
 SOCIAL CONFLICT 301

group and equity norms. Robbed of this specificity, the r

notion reduces to little more than that hardship produ
grievances (Salert 1976:Ch. 3). That is plausible enough
common sense, but is not very useful as a theory of the so
Did blacks in the US in the late 1950s feel deprived relativ
being made by Africans toward independence, relative to t
immediately after the war, or relative to the growing afflu
Did blacks outside the south in the mid 1960s become r
as a result of the progress being made by southern bla
expectations raised and disappointed by the promises of th
Gurr's theory does not provide a means of answering t
Gurr's theory is basically psychological and individualist
ior and tendencies are the result of the sum of individ
tendencies. The deprivation level of a collectivity is th
deprivation in the group. This conceptualization treats ind
pendent, isolated behavioral units. But is this realistic?
that if a frustrated and angry individual (say, as a result
of force) is released in his immediate social milieu, oth
become angered, not because they themselves have experie
frustrating experience, but because they sympathize and
victim. Social interaction may well have such "multipli
or collective behavior cannot be deduced from individ
sumptions alone. It must incorporate effects resulting fr
individuals in the group, i.e. the effects of group structu
This observation and criticism is not just a sterile co
theorists who have different styles of theorizing. Import
for theory do follow from it. This can be clarified in an
been observed that though considerable progress was bein
in the 1960s in the area of civil rights, the magnitude and
riots increased, at least until 1969. An explanation base
psychological assumptions would be that racial progr
expectations about further progress even faster. Thus rel
had increased, and hence also the magnitude of collect
But another interpretation based on social interaction an
ture is also possible. The civil rights movement increased
communication, and shared symbols among blacks. The
solidary group. Though the incidence of grievance-produc
blacks did not substantially change or might even hav
events that did occur became more visible to them an
intense collective reaction, becuase group cohesion had inc
concerns of group members for each other. I know of no
to decide between these two views, and I certainly do not s

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
302 OBERSCHALL

view is by itself a satisfactory explanation of the growing

What the illustration indicates is that neglect of group struc
theory will lead to quite a different perspective on conflict
explanations of it.
Even some of the policy implications differ. The rising expectation view
leads to despair with ever satisfying minorities' aspirations. It might also
lead to the view that in order to reduce expectations, racial progress will
have to be slowed, which will ultimately reduce expectations more rapidly
than accomplishments, and thus also lower relative deprivation. The other
model views increased conflict as a likely by-product of the greater
solidarity, and hence also mobilizing capacity, of the challenging group, and
thus as a "normal cost" to pay for racial progress. It does not make sense
to slow racial progress in order to decrease overt conflict, nor even to
weaken group cohesion, which would increase overt conflict in the short run
almost certainly. Raising the costs of rioting by taking firm action against
rioters is the only viable policy prescription from this viewpoint, since lower
mobilization costs resulting from increased solidarity will have to be coun-
teracted by increasing collective action costs.

CRITICS OF THE BREAKDOWN-FRUSTRATION APPROACH Writing

from different disciplinary backgrounds and viewpoints, several rece
ics of the breakdown-frustration theories of social conflict have shown
surprising agreement on what other approach might prove useful. I review
three important works in this context.
In The Rebellious Century, 1830-1930, the Tillys squarely confront
breakdown theories with their findings on civil strife in France, Italy, and
Germany (Tilly et al 1975). In their view, local, regional, and national
struggles for power, and especially state making, account for a high propor-
tion of violent events, rather than immediate responses to misery and eco-
nomic deprivation. There is no tendency for collective violence to
concentrate during or after surges of urban growth. Despite major changes
in French social structure, France remained as violent in the twentieth
century as it was in the nineteenth. Group violence ordinarily grows out of
collective actions that are not necessarily violent, such as festivals, meetings,
strikes, and demonstrations. Collective action is not anomic tension release,
but purposive and political. Periods of strong repression and firm central
government control experience little or no collective violence, because all
collective action diminishes and thus also the occasion for collective action
becoming violent. Contrary to the breakdown theory, in the short run, rapid
social change depresses the level of social conflict, because it weakens many
groups' means of mobilization for collective action faster than it creates
other groups with a high mobilizing capability. A case in point is that of

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
 SOCIAL CONFLICT 303

recent urban migrants. They are members of solidary groups in villages and
small towns, but remain unorganized for a time in the city milieu.
Urbanization, industrialization, and centralization of government stimu-
late political conflict, because the authorities make claims on the resources
of groups that retain a viable social organization and hence are capable of
defensive mobilization. Typical collective actions of the earlier 19th century
in which participants resisted the claims of national structures were the food
riot, the tax rebellion, and resistance to conscription. The sites of collective
action were the natural gathering places of rural folk: markets, church
services, festivals. As traditional groups weakened and new solidarities,
associations, and trade unions came into being, collective action shifted
from the defense of existing rights and resources to a claim for more rights
and a greater share of societal resources. The demonstration, strike, and the
deliberate assembly in an urban setting became the typical form of collective
action.
Other critics of breakdown-frustration theory find it deficient in explain-
ing internal war and rebellion in the contemporary world. Laqueur (1976)
is profoundly skeptical of social science theories and generalizations about
internal war and collective violence. He is even skeptical of the comparative
approach employed in civil strife studies: "The guerrilla phenomenon
presents endless variety ... a comparison between China and Vietnam or
between Angola and Mozambique, or even between the IRA and Basques
may be of value and of interest. Moving further afield in time or space,
generalizations can be made only with the greatest of caution" (1976:386)
Laqueur's Guerrilla, a Historical and Critical Study (1976) is an exhaus-
tive examination of secondary sources on partisans, guerrillas, and wars of
national liberation, ranging widely in geography and time. Rich in detail,
the book stresses the great diversity of such movements. Laqueur debunks
the notion that guerrilla war is a recent historical phenomenon, a rare event,
and an effective weapon. He holds that recent doctrines of revolutionary
warfare are neither original nor able to account for success and failure.
According to Laqueur, nationalism, patriotism, and separatist ethnic senti-
ment have been the most important driving force of guerrilla wars. He
doubts the usefulness of the breakdown-frustration explanation of popular
uprising in traditional societies undergoing rapid change. He questions the
assumptions that insurgency is deviant behavior, "as if acceptance of for-
eign occupations and domestic tyrants is the norm, and the decision to
oppose them a deviation" (Laqueur 1976:386). He notes that a consistent
bias of theorists of internal war has been to analyze "the forces which propel
societies towards violence rather than those which inhibit it" (1976:389). He
suggests that more attention be paid to incumbents, the targets of the
insurgents.

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
304 OBERSCHALL

Based on the numerous cases he examines, Laqueur is willing to make

some low-key empirical generalizations. The most useful ones for conflict
theory are the following:
1. Guerillas tend to locate in inaccessible regions (mountains, jungles,
forests, swamps) where they are difficult to find and where their opponent
cannot deploy full strength. Moreover guerrilla wars often occur in areas
where they have previously occurred, i.e. peripheral regions where the hold
of governments has been weak. Successful guerrillas operate, however, near
some centers of population where they are not entirely cut off from the
people. These generalizations suggest that the cost of social control for the
authorities is an important variable for explaining the likelihood of success.
2. Propaganda and terror have always been essential parts of guerrilla
warfare. Though a noncombatant fringe has been very helpful to such
movements in providing them with money, intelligence, and supplies, the
majority of people take a neutral, passive attitude in the struggle between
insurgents and incumbents. Moreover, terror is used to frighten government
collaborators. Local people are more often the casualties of guerrillas than
foreigners and social control agents. In summary, "no guerrilla movement
can ... survive ... against an overwhelmingly hostile population. But in the
light of historical experience the measure of active popular support required
by guerrilla movements need not be exaggerated" (Laqueur 1976:393-409).
These observations suggest that theories of insurgency inspired by the
imagery of an uprising of large, disaffected masses, as some breakdown-
frustration theories are, provide a mistaken view of participation in and
support for insurgents. Such also is the opinion of Leites & Wolf (1970) in
Rebellion and Authority where they contrast the "hearts and minds" ap-
proach with a systems approach to insurgency.
The hearts-and-minds approach is a variant of the breakdown-frustration
theories of social conflict. In that view deprivation causes rebellion, and
poverty and inequality account for deprivation. Popular attitudes, sympa-
thies, and support for the rebels then play a decisive role in ensuring their
success or failure. Whichever side will gain the hearts and minds of the
population will win. The hearts-and-minds approach represents a demand-
pull view of rebellion: the demand for change by the people is the driving
element of rebellion.
Leites & Wolf level several criticisms at the hearts-and-minds approach.
Like Laqueur, they do not find evidence in the contemporary record that
rebellion thrives on poverty and inequality of wealth, that a high proportion
of the population is required to sympathize with and support the rebels, and
that the scope of support must be broad and come from within the rebel
region. In contrast, they maintain as Laqueur did, that only a small propor-
tion of the population need be active supporters. Nondenunciation to the

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
 SOCIAL CONFLICT 305

authorities is the only critical action required by the majori

rebel success. Rebels can create support and nondenunciation
cion and the fear of retaliation instead of conquering peo
minds. The individual caught between authorities and rebels
short-term, damage-limiting behavior. His expectations abou
side will largely determine the side he will collaborate with.
Leites & Wolf state an alternative systems approach to insu
like the hearts-and-minds approach, which focuses on th
change on the part of a population, they emphasize the supp
rebellion's growth, i.e. the opportunities and costs that determ
of rebels and the amount of opposition activity. Rebellion
action. Resource inputs are used to provide output activit
a conversion mechanism, the insurgents' organization. Inputs
information, shelter, and food, usually secured from within t
and weapons, financing, and publicity, frequently obtained f
Local inputs are obtained through persuasion and coercion
conflict, the population calculates the costs and benefit of d
and chooses a course of action that maximizes short-term ga
(Leites & Wolf 1970:Ch. 3). For decreasing rebellion, the d
insurgency prescribes costly, major reforms that are difficul
in a conflict. The authors' cost-push version emphasizes f
crease the costs of rebel activity: raising rebels' costs of obtai
impeding the conversion of inputs into outputs by reducing
of rebel organizations, destroying rebel outputs, and bluntin
outputs by increasing the capacity to absorb rebel activities.
Leites & Wolf s systems approach and their conceptualiz
lion as a production process is a promising innovation for co
Both challenger and target have a resource base from wh
inputs. Assembling resources for inputs in the conflict syste
to as mobilization in the literature. The output produced by
mechanism is collective action. The mechanism of conversion
tion apparatus of the insurgents, is the organizational st
challengers. The conversion process itself will use up reso
tion maintenance is costly. Collective action that results in co
be destructive of resources. Outcome of the conflict will dep
on each party's ability to deny or destroy the other's resour
increase the other's cost of mobilization, organization m
collective action.
The central concern of conflict theory becomes mobilizatio
tion, and collective action. Grievances, discontent, and societ
are not ignored insofar as they will affect the resource base
parties and their mobilization, organization and collectiv

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
306 OBERSCHALL

People may harbor intense grievances; yet they may possess few resources
and may be vulnerable so that they can provide but negligible inputs to
challengers. Others with but mild grievances, yet plentiful resources, might
make substantial contributions even if they provide but a fraction of their
resources to challengers. Societal breakdown is, to be sure, associated with
the generation of discontent, but its main impact on conflict is by way of
the costs of mobilization and collective action. Mobilizing the resources of
disorganized collectivities is far costlier than doing it within solidary
groups, because the latter already possess well-defined leadership and orga-
nization that can be enlisted on behalf of the challenger.

Mobilization and Solidarity

What are the salient features of a theory of conflict that thrusts processes
of conflict group formation into center stage? As other theories do as well,
such a theory starts with an existing structure of domination that identifies
the number and type of major collective actors in the system, their collective
interests, and the major resources at their command. Mobilization refers to
the processes through which individual group members' resources are sur-
rendered, assembled, and committed for obtaining common goals and for
defending group interests. Because mobilization is facilitated or impeded by
the internal organization and structure of the collectivity, group structure
is a major variable in the analysis. The extent and forms of collective action
taken in pursuit of collective goals depend on levels of mobilization and on
repertoires of collective action.
A widely applicable general model of structures of domination has been
put forward by Tilly (1975, 1978). It consists of a government that controls
the principal means of coercion in a population, and contending groups that
apply their resources for influencing the actions and outputs of government
on their behalf. Contenders are of two kinds. Members of the polity, favored
in the structure of domination, have routine, low cost access to the govern-
ment. Other contenders, called challengers, seek to obtain greater influence
and more government outputs, though the costs of challenge are consider-
able for them. Backed by the government, members of the polity resist the
demands of challengers for a greater share of societal resources. A success-
ful challenger modifies the initial structure of domination. When he
becomes a member of the polity, the challenger has his interests routinely
recognized and shifts some of the costs of obtaining resources onto institu-
tionalized politics. For instance, the civil rights movement in the US was
eventually able to shift the costs of desegregating and integrating public
facilities to the Justice Department and other government agencies instead
of bearing the cost of direct action itself, as it did in the earlier challenge
period.

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
 SOCIAL CONFLICT 307

A major issue in mobilization theory has been identified in Olson's theory

of collective action (Olson 1968). Challengers frequently seek to obtain a
collective good. A collective good is a good that once supplied to one
member of the group cannot be witheld from any members of the group,
even those who did not contribute to the cost of obtaining it. According to
Olson, all members of the collectivity have a common interest in obtaining
the collective good, yet they have no common interest in paying for the cost
of providing it. This gives rise to the so-called free-rider problem in large
groups. Each individual knows that his own efforts will have no noticeable
effect on the chances of obtaining the collective good; he also knows that
he can enjoy the benefits of others' collective efforts even when he has
contributed nothing to these efforts himself. The rational individual will
thus have a tendency to be a free rider. But if all or most members of the
group are free riders, no joint action will be mounted to obtain collective
goods in the first place.
Olson suggests that in order to circumvent this dilemma groups seeking
collective goods mete out negative sanctions to noncontributors, or else
provide selective incentives to contributing members. Selective incentives
are individual benefits obtained exclusively by contributing group members.
Olson's analysis accords well with the observations of Laqueur and of Leites
& Wolf on insurgencies and guerrilla movements: only a fraction of the
population that might benefit from success of the insurgents' goals will
voluntarily support the insurgency; the insurgents typically rely on negative
sanctions to secure collaboration of the people. As for the active insurgents
themselves, positive selective incentives, e.g. future leadership and opportu-
nity for advancement should they be successful, enter their calculation. The
notion of selective incentives as a motivating force for leaders and political
entrepreneurs who initiate mobilization has been systematically developed
within the framework of a rational-choice model of collective action (Froh-
lich et al 1971). Moreover, insofar as resources are provided to mobilizing
groups from without, which frequently occurs, the free-rider hurdle is more
likely to be overcome (Oberschall 1973:Ch. 4).
Nevertheless a mobilization theory based on the utilitarian rational-
choice model leaves one somewhat unsatisfied, because strong passions,
group consciousness, ideological appeal, and appeals to solidarity in group
conflict are relegated to secondary place, if they are taken into account at
all (Gamson & Fireman 1978). Psychological gratification from participat-
ing in a collective effort, or from personal commitment to a cause, must
surely have some bearing on recruitment and participation. And one would
like to know why certain causes rather than others enlist greater enthusiasm
and loyalty, and why some movements are able to provide more satisfaction
to joiners than others.

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
308 OBERSCHALL

Group structure and solidarity do enter directly into mobiliz

Mobilization, it will be recalled, refers to the transfer of indiv
to agents or organizations that commit them to group goals. Groups may
already be organized in such a way that substantial amounts of individual
resources are routinely allocated through existing associations and leader-
ship for group ends (Oberschall 1973:Ch. 4). As the need arises, existing
leadership and organizations can then rapidly commit mobilized resources
to new group goals, and can expand the reach of their mobilizing effort at
low costs by making use of existing networks among group members. For
nonsolidarity groups, mobilization will cost much more: a mobilizing
agency will have to be created; its link to the population will have to be
forged; and since organization is not a routine by-product of existing group
structure, organization maintenance costs will be considerable.
Tilly (1978:Ch. 3) summarized the basic relationship between group
structure and mobilization in a particularly useful fashion. Solidary orga-
nization is a product of catness and netness, called catnet for short. Catness
refers to the strength of a shared identity in a group and to the sharpness
of social boundaries that comprise all those who share a common character-
istic. Netness refers to the density of networks among group members that
link them to each other by means of interpersonal bonds. Solidarity in-
creased with catness. Mobilization in turn can be measured by the amount
and kinds of resources in a group multiplied by the probability that these
will be delivered for the pursuit of group goals, when needed. The greater
the solidarity of the collectivity, the higher the probability of delivery.
Collectivities poor in resources may in this view compensate for it by greater
catness and solidarity, and may exceed in resources delivered for group
goals the mobilized resources of more richly endowed, low-solidarity
groups.
The notion that solidarity is positively associated with the probability of
delivery of individual resources to group ends allows one to enter psycho-
logical gratification from group membership into a theory of mobilization
based on the rational-choice model. In solidary groups, people derive many
material and psychological benefits from membership, which they cannot
provide for themselves individually. Social fellowship, social support, sense
of belonging, and other benefits are possible only through the preservation
of the solidary group itself. Because there is a visible relationship between
individual welfare and group solidarity, people in a high-solidarity group
come to value the preservation of their group and of its qualities more than
people in low-solidarity group. Consequently, members of high-solidarity
groups can be expected to allocate more resources for collective ends,
including the preservation of its solidary qualities, than members of low-
solidarity groups. This would especially be true in situations of conflict
where the preservation of the group itself is at stake. And one should not

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
 SOCIAL CONFLICT 309

forget that groups high on solidarity, and therefore also on catnet, have a
better chance of identifying free riders and pressuring them to conform to
group norms. For both these sets of reasons then, the mobilization capabil-
ity of high-solidarity groups are greater than that of groups low on
solidarity.
Though mobilization theory is relatively well developed, much work
remains to be done to increase its explanatory power. In particular, it
appears useful to differentiate several categories of participants in conflict
groups, because their recruitment into a conflict group and their commit-
ment to a collective movement will be generated and sustained by a different
mix of selective incentives, psychological gratifications, and ideological ap-
peals. Useful work along these lines has been done by McCarthy & Zald
(1973) who distinguish constituents (those who provide resources to a
conflict group), adherents (those who value the collective good), bystander
publics, and opponents. Constituents may further be divided into a leader-
ship cadre or full-time activists, and transitory teams of part-timers. Finally,
all these categories may differ according to the direct benefits they expect
to derive from the realization of the conflict group's goals. These distinc-
tions have allowed McCarthy & Zald to identify the important categories
of conscience adherents and conscience constituents, i.e. participants and
supporters who are not direct beneficiaries of the collective good. Strategies
of mobilization can then be analyzed from the point of view of the different
resources each of these categories contributes to the struggle and the differ-
ing selective incentives and appeals necessary to ensure resource delivery.
Internal splits and factionalism within conflict groups may also be usefully
understood by looking at the internal heterogeneity of conflict groups.

The Forms and Dynamics of Conflict

Mobilization provides a potential for collective action. Conflict however
results from collective action, and in particular from the interaction of
challenger's and target's collective actions. One party's initial advantage
may spur the other party to greater mobilizing efforts and to different
strategies and forms of collective action, which in turn may induce changes
in the first party's mobilizing efforts, strategies, and collective action. Unfor-
tunately no theory of the dynamics of collective action has yet been formu-
lated in sociological theories of conflict. Recent work by Gamson (1975),
however, showed that a static analysis linking mobilization variables and
forms of collective action to outcome variables produces useful empirical
generalizations. Much interesting work has also been done on an empirical
documentation of the forms and magnitudes of violent conflicts in the
contemporary world, on types of conflict, and on linking these to structures
of domination and to mobilization variables (Gurr 1969). At a microlevel,
Tilly's (1978) work on the repertoires of collective action is especially

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
310 OBERSCHALL

promising for uncovering relationships between mobilizati

collective action. Finally there exist fragments of theory o
dynamics (Heirich 1971; Coleman 1957; Deutsch 1973; Oberschall 1978b).
The empirical generalizations in Gamson's The Strategy of Social Protest
(1975) are based on a sample of 53 challenging groups drawn from the
American experience between 1800 and 1945. The central dependent vari-
able is the degree of success achieved by challengers. Success is conceptual-
ized along two dimensions: gaining full acceptance, i.e. recognition as a
member of the polity, and gaining new advantages, i.e. achieving the collec-
tive goods, objectives, and other goals of the challenge. Among other
findings, Gamson shows that challenger's size is positively associated with
acceptance but only slightly so with new advantages. Challengers who seek
to overthrow a target group have a much lower success rate than challeng-
ers who limit themselves to influencing a target. On the other hand, chal-
lengers who resort to violence or the threat of it, or to coercive tactics, have
somewhat higher success rates than those who use noncoercive means
alone. Moreover, there exists a fairly high positive association between
success and challenger's use of selective incentives, as opposed to providing
members with psychological gratification alone. Finally, variables measur-
ing the internal organization of the challenger are associated in complex,
yet systematic ways, with success. Gamson demonstrated that for conflict
theory some of the variables identified as important in mobilization theory
do indeed have significant observed associations with a central dependent
variable of conflict theory, success of the challenge.
Much empirical work on collective action is essentially descriptive, classi-
ficatory, and cross-national. The cross-national literature on collective vio-
lence has grown immense in recent years. Here I only selectively discuss
some findings from a representative of this school (Gurr 1969). For a set
of countries, over a limited span of years, specific strife events are coded
from news sources for number and social categories of participants, their
goals and their targets, the human costs of strife, as well as many other
variables. Strife events are categorized in various ways along the dimensions
of internal organization, popular participation, and objectives. This proce-
dure allows one to distinguish turmoil, conspiracy, and internal war, or
some equivalent categories at the aggregate level. Finally, at the level of a
country, indices of pervasiveness, duration and intensity of civil strife are
calculated, and a summary index of the magnitude of civil strife is com-
puted. Comparisons are made of these indices and types of conflict with
countries grouped by world regions, political regimes, level of economic
development and other variables. Such studies often compare and rank
nations by magnitude and types of conflict as well. Findings are also used
to test models of civil strife.

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
 SOCIAL CONFLICT 311

Aside from differences in quality of data across a variety of countries, a

further problem is the particular time period investigated, because many
countries experience sharp spurts and declines of collective violence. A
study based on the 1950s and early 1960s would certainly draw different
conclusions about levels of civil strife in the US than one that included the
late 1960s. Whatever time frame one chooses, some countries will have quite
different levels of strife in a somewhat different time span.
Nevertheless, cross-national studies provide many insights into conflict.
One particularly useful conclusion is expressed in Gurr's statement that
"among nations generally, political groups most often mobilize people for
strife ... strife in the more developed and democratic nations is often more
organized by political groups than in other nations ... the implications are
that strife is a recurrent facet of the political process and that the effect of
economic development and of political democratization is to channel it into
the political process rather than to insulate politics from violence" (Gurr
1969:558). These and similar findings confirm empirically the view of
solidarity and mobilization theorists that conflict, whether violent or nonvi-
olent, is firmly embedded in the larger political process.
Specific empirical generalizations from cross-national studies (e.g. the
relationship between regime coercive force size and magnitude of civil
strife) are sometimes difficult to utilize in conflict theory. The direction of
causation may be uncertain. Nations differ on so many variables that even
if the direction of causation is obvious, the observed relationship may be
spurious. For conflict theory, a study of the magnitude and forms of collec-
tive action in relation to the size and utilization of coercive forces within
the same polity over an extended period of time would probably yield more
interesting information about interaction between challengers and author-
ities. Indeed, it is the study of collective violence in France over somewhat
more than a century that allowed Tilly to uncover the relationships between
broad social trends, mobilization variables, social control, and collective
action that I described above (Tilly et al 1975:Ch. 2).
In accounting for the forms of collective action, Tilly (1978) introduced
the useful concept of repertoires of collective action. Given the range of
possible collective action, the observed forms of it are surprisingly limited.
For instance,

Most "twentieth century Americans" know how to demonstrate. They know that a
group that has a claim to make assembles at a public place, identifies itself and its
demands or complaints in a visible way, orients its common action to the persons,
property, or symbols of some other group that is seeking to influence. Within those
general rules, most Americans know how to carry on several different forms of demon-
stration: the mass march, the assembly with speechmaking, the temporary occupation
of premises (Tilly 1978:Ch. 5).

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
312 OBERSCHALL

Such a repertoire is learned directly by participating in such actions or

witnessing them in the mass media. Other forms of collective action that
were once part of the repertoire of some American groups, such as lynching,
have dropped out of it.
Tilly argues that the innovation, diffusion, and incorporation of certain
forms of collective action into a group's collective action repertoire can be
accounted for by the daily routines of the population, its internal organiza-
tion, its past experience with prior collective action, patterns of repression,
and prevailing standards of right and justice. He shows how the growing
legality of electoral assemblies and associations in the modern era promoted
the demonstration as a form of collective action because the protection
enjoyed by electoral assemblies spilled over to other collective actions that
were marginally related to elections. Such protection then lowered the
mobilization and collective action costs of demonstrations compared to
other forms and led to its increased adoption (Tilly 1978:Ch. 5).
One of the most promising yet little-used ways of studying the dynamics
of conflict is the investigation of the diffusion patterns of collective action
in physical and social space and over time, for it allows the formulation of
fairly precise models subject to empirical testing. A decade of research on
urban riots has shown that an atomistic analysis of riots that centers on the
characteristics of participants and the characteristics of cities does not have
much explanatory power (Spilerman 1970). The surge and decline of collec-
tive action occurs within a system that has dynamic properties of its own.
Contagion effects, bandwagon effects, strategic interaction among partici-
pants, focal points, and pacesetter-follower relationships are important
properties of such systems, which merit further study in their own right
(Oberschall 1978a,b).
Another area worth more study is escalation and de-escalation. Much
collective action is essentially legal and nonviolent, though often threaten-
ing, to the target. Why is it that in some conflicts there is a convergence
on destructive means, whereas in others conciliation is successfully em-
ployed? A useful approach to escalation and de-escalation is suggested by
Deutsch's social psychological analysis of destructive conflicts (Deutsch
1973).
Destructive conflict becomes independent of its initiating causes and is
likely to continue after these have become irrelevant. Expansion of the
conflict occurs along the various dimensions of conflict: the number of
participants drawn into the conflict; the number of issues at stake; the
intensity of hostility towards the other party; the increased reliance on
threats and coercion. The tendency to escalate results from the conjunction
of three interrelated processes: increased competition, biased perception,

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC12:34:56 UTC
All use subject to https://about.jstor.org/terms
 SOCIAL CONFLICT 313

and heightened commitment (Deutsch 1973:52). Communication between

the conflicting parties becomes unreliable and impoverished. The competi-
tors become more sensitive to differences and threats than to similarities and
conciliatory gestures. Perceptual distortions arise from pressures for self-
consistency and social conformity. Heightened group identity and solidarity
increase hostility against the competing outgroup. Actions acquire moral
connotations such that errors of judgments and mistakes based on lack of
information are interpreted as intentional measures designed to harm and
as indications of bad faith. As a result the usually accepted norms of
conduct and morality that govern one's behavior become suspended. De-
spite mounting costs, the parties persist in a spiraling effort to destroy each
other.
In Coleman's view (1957:14), an escalating conflict obeys "Gresham's
Law of Conflict: the harmful and dangerous elements drive out those which
keep the conflict within bounds." This process results from both attitudinal
and social polarization, "the division of the community into two socially
and attitudinally separate camps, each convinced it is absolutely right"
(Coleman 1957:13). Polarization grows with the formation of new partisan
organizations and the emergence of new extremist leaders, while existing
leaders and organizations are cross-pressured into inaction. The issues un-
der dispute have a tendency to change from specific to general, new issues
are added to the original ones, a wider set of people become drawn into the
controversies, and disagreements give way to bitter antagonisms. Eventu-
ally the conflict may become independent of the initial disagreements and
is sustained by the antagonists' goal of winning and of harming each other.
Many of these same processes are noted and analyzed in Heirich's (1971)
elaborate effort to provide a theory of escalating conflict based on a detailed
study of the Free Speech Movement at Berkeley. Although full of insight
and very suggestive, the conceptual and theoretical apparatus he develops
lacks simplicity and will be difficult to apply.
Can escalation and de-escalation be analyzed from the perspective of
rational-choice models? A start in this direction has been made by Ober-
schall (1978b). During confrontation, challenger and target both have three
options: abandon the conflict and submit to the other side, make a concilia-
tory move, or make a coercive move. They choose the alternative that
maximizes expected net benefit, i.e. benefits minus costs. As in mobilization
theory benefits can be collective goods and selective incentives. Costs are
opportunity costs and the costs of collective action itself. This model shows
that, under a variety of circumstances, conflict groups persist in continuing
highly destructive conflict even in the face of low success chances. Because
of the certainty of high penalties, abandoning the conflict can be costly. This

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
314 OBERSCHALL

conclusion applies to targets as well as to challengers. It suggests that a

relatively low-cost exit from destructive conflict has to be provided to the
protagonist for conflict termination. Frequent demands for amnesty and
safe conduct in highly destructive conflicts suggests that the cost-benefit
approach to escalation is worth pursuing.
Only a beginning has been made in modeling conflict as a dynamic
system. Much work needs to be done to combine the social psychological
and rational choice approaches. This task is perhaps even more necessary
for a satisfactory theory of conflict dynamics than it is for a theory of
mobilization. This difficulty plagues the application of game theory to stra-
tegic analysis in realistic settings. Although yielding rich insights, the intro-
duction of psychological concepts, such as trust, commitment, intentions,
expectations, and promises into game theory leads to a proliferation of
variables that hinder precise deductions from formal models, which is their
principal advantage (Schelling 1963).

CONCLUSION

The topic of conflict cuts across a range of disciplines, from eco

biology by way of history, political science, sociology, anthropol
psychology. Conflict has been studied by a number of methodolo
techniques, ranging from detailed case studies to abstract mathematical
models. Theories dealing with particular classes of conflict, such as interna-
tional conflict, strikes, race riots, guerrilla war, and ecological competition
have received far more attention than a general theory of conflict that might
be applicable to all manner of conflict.
Sociological theories of conflict at the present time are a mixture of
insights, hypotheses, conceptual analyses, typologies, observed empirical
regularities, and some modest attempts at systematic modeling. Social con-
flict analysis overlaps with existing macrosociological theories about stratifi-
cation, social change, group formation, and collective action. Insofar as
there are clearly formulated and tested theories of the relationship between
structures of domination and the origin of conflict, and of the links between
group structure, mobilization, and collective action, this knowledge can be
and has been readily incorporated into conflict theory.
I have shown above that two competing clusters of sociological thinking
and theorizing, referred to as the breakdown-frustration and the mobiliza-
tion-solidarity approaches, have provided the sociological underpinning for
explaining and understanding the forms, incidence, and outcomes of social
conflict. Whatever the relative merits of the two approaches, it should be
said that neither is yet close to providing a dynamic analysis of conflict as
a process. I have also shown in the discussion and in several illustrations

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms
 SOCIAL CONFLICT 315

that the choice of levels of analysis (group vs individual) and of fundamental

behavior assumptions (rational choice vs a more complicated psychology of
the actor) does make a difference for the interpretation of conflict. At the
present time, neither the substantive theoretical controversy nor the two
methodological issues are close to being resolved. And that is all to the
good, for better theory will probably result from the existence of contro-
versy over conflict theory.

Literature Cited

Boulding, K. 1963. Conflict and Defense. Janis, I. 1972. Victims of Groupthink. Bos-
New York: Harper Torchbooks. 349 pp. ton: Houghton Mifflin. 277 pp.
Coleman, J. S. 1957. Community Conflict. Kriesberg, L. 1973. The Sociology of Social
New York: Free Press. 28 pp. Conflicts. Englewood Cliffs, NJ: Pren-
Collins, R. 1975. Conflict Sociology, Towards tice-Hall. 300 pp.
an Explanatory Science. New York: Laqueur, W. 1976. Guerrilla, A Historical
Academic 584 pp. and Critical Study. Boston: Little,
Coser, L. 1956. The Functions of Social Con- Brown. 462 pp.
flict. New York: Free Press. 188 pp. Leites, N., Wolf, C. J. 1970 Rebellion and
Coser, L. 1967. Continuities in the Study of Authority. Chicago: Markham. 174 pp.
Social Conflict. New York: Free Press. McCarthy, J., Zald, M. 1973. The Trendof
272 pp. Social Movements in America. Morris-
Dahrendorf, R. 1959. Class and Class Con- town, NJ: General Learning. 30 pp.
flict in Industrial Society. Stanford Oberschall, A. 1973. Social Conflict and So-
Univ. Press, Calif. 336 pp. cial Movements. Englewood Cliffs, NJ:
Deutsch, M. 1973. The Resolution of Conflict. Prentice-Hall. 371 pp.
New Haven, CT: Yale Univ. Press. Oberschall, A. 1978a. The decline of the
420 pp. 1960s social movements. In Research in
Frohlich, N., Oppenheimer, J., Young, O. Social Movements, Conflict, and
1971. Political Leadership and Collec- Change, ed. L. Kriesberg. Johnson As-
tive Goods. Princeton Univ. Press, NJ. soc. In press
161 pp. Oberschall, A. 1978b. Protracted conflict. See
Gamson, W. 1975. The Strategy ofSocial Pro- Gamson & Fireman 1978
test. Homewood, Ill.: Dorsey. 217 pp. Olson, M. Jr. 1968. The Logic of Collective
Gamson, W., Fireman, B. 1978. Utilitarian Action. New York: Schocken. 176 pp.
logic in the resource mobilization per- Paige, J. 1975. Agrarian Revolution. New
spective. In The Dynamics of Social York: Free Press. 435 pp.
Movements: Resource Mobilization, Salert, B. Revolutions and Revolutionaries.
Tactics, and Social Control, ed. M. New York: Elsevier. 161 pp.
Zald, J. McCarthy. Beverley Hills, CA: Schelling, T. 1963. The Strategy of Conflict.
Sage. In press New York: Oxford Univ. Press. 309 pp.
Gurr, T. R. 1968. A causal model of civil Smelser, N. 1968. Toward a theory of mod-
strife. Am. Polit. Sci. Rev. 62,4:1104-24 ernization. In Essays in Sociological Ex-
Gurr, T. 1969. A comparative study of civil planation, ed. N. Smelser, pp. 125-46.
strife. Violence in America, ed. H. Gra- Englewood Cliffs, NJ: Prentice-Hall
ham, T. Gurr, pp. 544-605. New York: Spilerman, S. 1970. The causes of racial dis-
New American Library turbance: A comparison of alternate ex-
Gurr, T. 1970. Why Men Rebel. Princeton planations. Am. Sociol. Rev. 35:627-49
Univ. Press, NJ: 421 pp. Tilly, C. 1978. From Mobilization to Revolu-
Hechter, M. Internal Colonialism: The Celtic tion. Reeding, Mass.: Addison-Wesley.
Fringe in British National Development, In press
1536-1966. Berkeley: Univ. Calif. Tilly, C., Tilly, L., Tilly, R. 1975. The Rebel-
Press. 361 pp lious Century, 1830 to 1930. Cam-
Heirich, M. 1971. The Spiral of Conflict: bridge, Mass.: Harvard Univ. Press.
Berkeley 1964. Columbia Univ. Press, 354 pp.
NY. 462 pp. Tilly, C. 1975. Revolutions and collective vi-
Huntington, J. 1968. Political Order in olence. In Handbook of Political
Changing Society. New Haven, CT: Science, ed. F. Greenstein, N. Polsby,
Yale Univ. Press. 488 pp. Vol. 3. New York: Addison Wesley

This content downloaded from

5.156.6.228 on Tue, 06 Sep 2022 05:13:59 UTC
All use subject to https://about.jstor.org/terms