Escalation, Stalemate,

and Settlement

Dean G. Pruitt

State University of New York at Buffalo

Jeffrey Z. Rubin

Tufts University



Processes That Produce


Chapter 5 described the transformations that occur during escalation and proposed sets of conditions under which conflict is most and least likely to escalate. In this and the next chapter, we look at the processes of escalation-the chains of events that produce and maintain these transformations when conflict occurs under conditions favorable to escalation. To put it another way, when the situation is unstable and the parties fall into conflict, what actually happens that pushes them in the direction of escalation?

The present chapter focuses mainly on the production of escalation: the processes that encourage the use of heavier tactics, cause issues to proliferate, and produce increasing absorption in the struggle. Chapter 7 examines the persistence of escalation once it has occurred: the processes that encourage the continued use of heavy tactics and a consistently deep absorption in the struggle.

We now introduce a second example of escalating conflict to supplement the case study of a campus crisis presented in the last chapter. This is the sequence of events that took place during the development of the Cold War between the United States and the Soviet Union. The fact that both of these cases involve conflict between collectives (groups, organizations, or nations) does not imply that escalation is found only in such relationships. Much of our theory also applies to the escalation of interpersonal conflict, such as conflict between spouses, neighbors, business associates, and the like.



The development of the Cold War between the United States and the Soviet Union immediately after 1945 is a prime example of conflict escalation. These two major nations were allies during the Second World War, which ended with high hopes for continued cooperation. But the Soviets emerged from the war with deep suspicions of the West. This led them to adopt a goal of controlling the nations adjoining their territory, making it difficult to maintain EastWest cooperation. They built a communist satellite system in Eastern Europe, supported communist guerrillas in Greece, and put political pressure on Tur key. In 1947 the United States responded to these actions in three ways: lt gave military aid to Greece and Turkey. It created the Marshall Plan, which was designed to revitalize the economy of Western Europe and weaken communist parties in Western European countries. And (in conjunction with Britain and later with France) it began the slow process of unifying West Germany and rebuilding its economy, as a further bulwark against Soviet expansion.

The latter move was viewed with considerable alarm by the Soviet Union, which had been at war twice with Germany in the prior 30 years. The Soviets responded at first with protests. Then, in 1948, they tried sporadically interrupting communications between Berlin (which was under joint control but was an enclave surrounded by the Russian-controlled portion of Germany) and .West Germany. Finally, after the West introduced a currency reform in West Germany, they installed a full blockade of Berlin, claiming that they were repairing the routes to the city. The United States and its allies responded by launching a successful airlift between Berlin and West Germany and by beginning negotiations that led to the formation of the North Atlantic Treaty Organization (NATO), a military alliance involving the United States and most of the Western European nations. This latter development led eventually to the rearmament of West Germany, which caused considerable further alarm in the Soviet Union.

The story of this conflict continues to the present day, but we stop it at this point because we have said enough to give a massive and extremely significant example of conflict escalation. This escalation illustrates most of the transformations described in the last chapter: tactics went from light (protests) to heavy (blockading a city, forming a military alliance); issues proliferated; the parties became increasingly absorbed in the struggle; more and more elements of the relationship between the superpowers were affected; and goals changed from self-advancement to subverting the adversary.

We look now at the processes that encourage such transformations, especially those that foster the use of increasingly heavy tactics, a key transformation in most escalative episodes.


Most theories of escalation can be classified under one of three broad conflict models (Pruitt & Gahagan, 1974): the aggressor-defender model, the conflict spiral model, and the structural change model. Though all three models have some value (they all account well for some episodes of escalation), the first model has generally been overemphasized and the last underemphasized.

The Aggressor-Defender Model

The aggressor-defender model draws a distinction between the two parties.One party, the "aggressor," is viewed as having a goal or set of goals that places it in conflict with the other, the "defender. " The aggressor ordinarily

starts with mild contentious tactics because of the costs involved in escalation. But if these do not work, he or she moves on to heavier tactics, continuing to escalate until the goals are attained or a point is reached at which the value of goal attainment is outweighed by the anticipated cost of continued escalation. The defender merely reacts, escalating his or her efforts in response to the aggressor's escalation. Escalation persists until the aggressor either wins or gives up trying.

The terms "aggressor" and "defender" in this model are not intended to be evaluative. In other words, they do not imply that one side is wrong and the other right in the controversy. The aggressor is simply a party who sees an opportunity to change things in the direction of his or her interests; the defender, a party who tries to resist this change.

The aggressor-defender model helps to explain one of the stages in the development of the Cold War. This is the point at which the Soviet Union adopted the goal of blocking the unification of West Germany. At first the Soviets employed the mild tactic of protest. When this did not work, they moved to a heavier tactic of sporadically interrupting communications between Berlin and West Germany. When this was unsuccessful, and the West introduced a currency reform that contributed further to German unification, they employed an extremely heavy tactic, a full blockade of the city. This explanation is cogent, but despite its popularity, the aggressor-defender model is incapable of interpreting many other stages in the development of the Cold War-and indeed in most conflict escalation. In other words, this model provides a useful but incomplete account of the processes underlying escalation.

The Conflict Spiral Model

The conflict spiral model of escalation is found in the writings of many theorists (North, Brody & Holsti, 1964; Osgood, 1962, 1966; Richardson, 1967). This model holds that escalation results from a vicious circle of action and reaction. Party's contentious tactics encourage a contentious response from Other, which contributes to further contentious behavior from Party, completing the circle and starting it on its next iteration.

Two broad classes of conflict spirals can be distinguished. In a retaliatory spiral, each party punishes the other for actions that it finds aversive. An example is an argument followed by a shouting match followed by a fist fight. In a defensive spiral, each party reacts so as to protect itself from a threat it finds in the other's self-protective actions. An example is an arms race. In a defensive spiral, each party can be thought of as alternately the aggressor and the defender.

Conflict spirals produce escalation of tactics when, as is often the case, each reaction is ' more severe and intense than the action it follows. They also contribute to the perpetuation of tactical escalation-that is, to the fact that heavy tactics continue to be used on both sides once they are first employed. If I hit you, you will often hit me back, which leads me to hit you once again, and so on.

In addition to explaining the escalation of tactics, the conflict spiral model helps us understand the growing size of an escalating Conflicts the proliferation of issues, and increased attention to the conflict. The point is that each retaliatory or defensive action in the spiral provides a new issue-a new grievance for the target of this action, producing a growing sense of crisis in the mind of this party.

The conflict spiral model provides further insight into the dynamics of Cold War escalation. In response to Soviet moves in Eastern Europe and in Greece and Turkey, the United States and its allies began to establish a West German state. In response to this action, the Soviet Union instituted a blockade of Berlin. In response to this blockade and all that had come before, the United States and its allies formed NATO and began to arm West Germany. And so on. The sequence of actions in the UB campus crisis, which started with stones thrown at the president's window and ended with an ugly confrontation between a mob of students and a number of city police officers, also illustrates such a spiral.

The aggressor-defender and conflict spiral models are compared diagramatically in Figure 6. 1. In the aggressor-defender analysis, causation flows in only one direction; the aggressor acts and the defender reacts. In the conflict spiral analysis, causation flows in both directions; each party reacts to the other party's actions. The conflict spiral diagram involves an oversimplification (which is necessary to help us distinguish the forest from the trees) in that it pictures each party's action as a response to the other's immediately preceding action. In reality, each action is a "result of the cumulative impression from all the previous actions by the other side" (White, 1984, p. 95), though more recent actions are usually given greater weight than earlier actions.

The conflict spiral model.should not be seen as an improved version of, or

replacement for, the aggressor-defender model of escalation. The latter is

useful whenever one party develops a goal that places it at odds with another party and pursues this goal through an escalating sequence of actions. Many cases of escalation exhibit this form. However, even in controversies where an aggressor-defender analysis is useful, the conflict spiral model frequently provides additional insights. Quite often the goal that impels the aggressor is, at least in part, a reaction to the defender's prior actions. This point is often missed by participants in conflict, who attribute the cause of the conflict exclusively to their adversary'.s aggression. It is also often missed by involved observers, who assign the cause of the confict to actions ofthe side with which they have weaker relations or the side that has employed the heavier, less defensible tactics. But a careful analysis usually reveals that causation has flowed in both directions.

A case in point is the Soviet effort to prevent the unification and ascendancy of West Germany, which took the form of an escalating series of protest actions that were progressively resisted by the West. Even though these actions are properly labeled "aggression" (in the nonevaluative sense of the term used here), they can also be seen as a reaction to Western efforts to strengthen Germany. Hence they are also part of a larger conflict spiral. Likewise, German efforts to conquer Europe in the 1940s, which were surely aggression by any definition of the term, can also be seen in part as reactions to the humiliation of Germany after the First World War-and thus as part of a conflict spiral lasting many years.

The Structural Change Model

Our picture of the forces producing escalation is rounded out by a third model, which is implied by the writings of Burton (1962), Coleman (1957), and Schumpeter (1955, first published in 1919), among others. This structural change model argues that conflict, and the tactics used to pursue it, produce residues in the form of changes in the parties and the communities to which the parties belong. These residues then encourage further contentious behavior, at an equal or still more escalated level, and diminish efforts at conflict resolutions Thus escalated conflict is both antecedent and consequent of structural changes.

Three kinds of structural changes can be distinguished: psychological changes, changes in groups and other collectives, and changes in the community surrounding the parties.

Psychological changes are many and diverse. As conflict escalates, negative attitudes and negative perceptions of the adversary typically develop. The adversary is blamed for the growing controversy and comes to be distrusted in the sense of being seen as indifferent or even hostile to our welfare. Negative traits are attributed to the adversary, such as being self-centered, morally unfit, or (in extreme cases) a diabolical enemy. The adversary is dehumanized and deindividuated. Anger, fear, and wounded pride become the dominant emotions. Zero-sum thinking develops-it's either victory for them or victory for us. New goals come to the fore: to look better than, punish, discredit, defeat, or even destroy the adversary. The capacity for empathy with the adversary is eroded. There are also changes in the approach taken to joint decision making: Positions become rigid, there is little room for compromise, and there is a dearth of imagination and creativity. Emphasis is placed on proving how tough and unyielding one is, so as to persuade the adversary that one cannot be pushed around. Coupled with this is an exaggerated fear of small losses, which are seen as diminishing one's status vis-a-vis the other and possibly encouraging the other to take advantage of one. There is also a tendency to break contact with the adversary-to be unwilling to communicate with him or her. All of these changes typically occur on both sides of the dispute.

These psychological changes occur in all escalated conflicts, whether the actors are individuals or collectives. When collectives (groups, organizations, or nations) are involved in a controversy, structural changes also occur at the collective level. The psychological reactions just described are accentuated by collective discussion and tend to become collective norms. Collective goals of defeating the enemy tend to develop, and subgroups are set up to implement these goals. Increased cohesiveness, resulting from having an outside enemy, contributes to the force of these norms and to the dedication of individuals in the collective to the newly found goals and the means of implementing them. New, more militant leadership often emerges, contributing further to the collective orientation toward struggle. In other words, doves are replaced by hawks. If one of the parties is an unorganized set of individuals, conflict sometimes encourages the development of a struggle group-precipitated out of the mix of strong individual emotions-which then takes up the cudgel against the adversary.

Structural changes may also take the form of polarization in the broader community of which the antagonists are a part. Third parties join one or the other antagonist, forsaking the constructive neutral role they might otherwise play.

The important point about all of these changes is that they contribute to the cycle of escalation: They result from the use of escalated tactics and encourage further escalation. This contribution takes three forms: One is that psychological and collective changes are the mechanisms at work in conflict spirals. As can be seen in the top part of Figure 6.2 (model 1), heavy tactics used by Party tend to produce structural changes in Other, which motivate similar heavy tactics from Other, provoking structural changes in Party, which start the spiral around once more. For example, Party's yelling at Other (a contentious tactic) causes Other to think of Party as an unpleasant person (structural change), making it easier for Other to yell back (more contentious tactics), encouraging Party to develop the goal of punishing Other (another structural change), motivating Party to make a fist (still more contentious tactics), and so on. Any or all of the changes described earlier, except community polarization, can be written into the boxes marked "structural changes" in this figure.

The second way in which structural changes contribute to the cycle of escalation is illustrated in the bottom part of Figure 6.2 (model 11). These changes often result from Party's own use of heavy tactics and contribute to further use of these tactics. For example, by a dissonance theory (Festinger, 1957) analysis of attitude change, the very fact of defending one's nation against another nation is likely to create more negative attitudes toward the adversary, encouraging further defensive efforts.

The third impact of structural changes is to erode some of the safeguards against escalating conflict that were describe in the last chapter: positive attitudes, respect, friendship, perceived similarity, common group membership, and future dependence. Crosscutting relationships also tend to disappear as communities polarize, and the effectiveness of conflict-limiting institutions in the broader community (such as legislatures and courts) may be adversely affected. The result is not only to encourage further escalation in the current situation but also to make escalation more likely in future conflicts involving the same parties. Thus escalated conflict often weakens a community's capacity to deal effectively with further conflict.

In addition to reinforcing escalation, those changes involving the development of hostile goals actually increase the divergence of interest between the parties. New issues come to the fore, resulting from the desire each party has

to defeat the other. Such developments confirm each party's distrust of the other. Distrust of this kind involves no misunderstanding; as a result of the development of hostile goals, each party has become an actual eneiny of the other.

The short, diagonal arrows at the top left of the diagrams in Figure 6.2 are meant to suggest that the initial impulse for these circular processes has its impact on behavior rather than on structural change. We believe that this is the way such circles ordinarily arise. In response to a perceived divergence of interest (or in response to some stimulus entirely outside the relationship between the two parties), one of the parties engages in contentious behavior. This produces structural changes tha't start the system on the road to escalation. However, it is possible for escalation to start with a structural change. A conversation with a third party may turn me against my neighbor and kick off an escalative process. A radical leader may urge a crowd of students to stage a sit-in, producing a cohesive struggle group that is capable of mounting a concerted campaign against the university administration. (This was the sequence of events in the 1968 Columbia University student crisis.)

Structural changes help to account for the escalation that led to the Cold War and for the flinty persistence of this escalation. Such psychological changes as anger, hostility, profound distrust, blackened images, and an inability to empathize took root in the United States during the early period of the Cold War and persist today. Zero-sum thinking ("What is good for them is bad for us, and vice versa") set in and is still noteworthy in the modern era. Considerations of national pride and face-saving gripped every American president from Truman onward-none wanted to be in office when another country fell to communism. Most Americans became unable to empathize with the genuine Soviet security needs that underlay a large proportion of their actions. For a period of time in the late 1940s and early 1950s, most communication with the Soviet Union was broken off, and today it is at an uncomfortably low level again.

Collective changes also took place. At the depth of the Cold War, hostile norms were so strong that people who had a good word for the Soviet Union were made to feel uncomfortable and were sometimes hauled up before congressional committees. The country even flirted for a time in the 1950s with highly militant leadership, in the person of Senator Joseph McCarthy, a virulent anti-communist with a large political following. Fortunately, some of these collective excesses were overcome by the 1960s. But the establishment of a semi-autonomous military machine, with political support from technological sectors dependent on it, provides constant input into. the still-festering controversy.

We have described these changes as we have seen them in the United States, because we know our own society better than that of the Soviet Union. However, there is good reason to believe that comparable changes have taken place and have been perpetuated in the Soviet Union (see White, 1984).

In addition to these psychological and collective changes, there has also been some polarization in'the international 'community, with many other nations choosing up sides between the two giants. This trend has abated somewhat in recent years.

In summary, we have described three models of conflict escalation that are commonly found in the literature: the aggressor-defender, conflict spiral, and structural change models. These models are not mutually exclusive. Rather, each helps us understand certain aspects of escalation. The most complicated model, by far, is the structural change model, because there are so many possible residues of escalation that keep escalation going. We shall devote most of the rest of the chapter to a further elaboration of this model, examining in more detail selected psychological, collective, and community changes. The concluding comment looks at the impact on escalation of the models that the parties themselves use in analyzing their conflict.


This section deals only with psychological changes that have been subject to careful research: the desire to punish (aggress against) the other, negative attitudes and perceptions, and deindividuation.

The Desire to Punish (Aggress Against) the Other

A great deal of research has been conducted on the sources of "aggression" (see Baron, 1977; Berkowitz, 1962; Zillmann, 1979). Because aggression is defined in this research as intentionally hurting another person, this research sheds light on the antecedents of the desire to punish the other party.

Research on this topic indicates that aggression arises mainly from aversive (unpleasant) experiences: deprivation, failure to achieve aspirations, inequitable treatment, pain and suffering, and the like. Such experiences are, of course, quite frequent when Other is engaged in contentious behavior. Hence the desire'to punish Other can be seen as one link between Other's contentious behavior and Party's subsequent contentious behavior.

Blaming Other. Aggression is more likely when the perceived source of an aversive experience can be blamed for his or her actions-that is, when these actions seem to be Other's fault. Blame has a number of sources. For example, Other is more likely to be blamed for actions that seem voluntary than those that seem involuntary (Schneider, Hastorf & Ellsworth, 1979). Actions that seem freely taken are more likely to evoke blame than those apparently resulting from heavy environmental pressures, unless Other is clearly responsible for resisting such pressures. Actions that appear contrary to the norms of society are also especially enraging (Mallick & McCandless, 1966).

The implication of these points is that conflict is especially likely to escalate when the parties see each other's contentious behavior as arbitrary and not attributable to extenuating circumstances, because under these conditions the parties are especially likely to develop a desire to punish one another.

To forgive Other because of extenuating circumstances requires a high level of cognitive activity. Party must analyze Other's circumstances and motivation and hold them up to standards of reasonable conduct. It is clear that this is socially learned behavior. Small children do not behave this way, being much more likely than adults to hold people responsible for their actions regardless of why these actions were taken (Shaw & Sulzer, 1964). Furthermore, Party pays less attention to extenuating circumstances when he or she is autonomically aroused, as might be expected with any complex cognitive behavior (Zillmann, Bryant, Cantor & Day, 1975).

The Role of Anger. Theorists differ on the role of anger in aggression, but it seems reasonable to assume that anger is often at work when aversive experiences lead to aggression.

Like other emotions, anger can be interpreted as resulting from cognitive labeling of an undifferentiated state of autonomic arousal. According to Scbachter (1964), the first stage in emotional experience is arousal-activation of the autonomic nervous system. The second stage is interpretation of this arousal. To make this interpretation, people employ whatever information is available. Not all interpretations lead to emotion; for example, if they have recently taken a new medicine or exercised, people may assume that this is the cause. But there are certain standard interpretations that produce emotion. If people see danger, they are likely to interpret their arousal as fear and feel fear; if they are aware of a recent aversive experience, their reaction is likely to be anger. When such interpretations are made, greater arousal leads to stronger emotion and more extreme emotional behavior.'

It follows that people become particularly emotional when they are aroused by other stimuli in addition to those that produce the emotional interpretation. Such an effect has been found in the case of anger leading to aggression. For example, Zillmann (1979) has shown in a controlled experiment that people who exercise and then have an aversive experience aggress more vigorously against the source of the aversive experience than they do if they have not exercised. This result is particularly apparent when the exercise comes somewhat earlier than the aversive experience so that people lose track of why they are aroused and attribute it entirely to the aversive experience. In addition, it has been shown that emotion can be destroyed by encouraging people to attribute their arousal to a neutral experience such as taking a pill. This misattribution" effect is also found with anger (Loftis, 1974).

The ideas just presented imply that heavily contentious behavior and hence escalation are most likely to occur when people approach conflict in an aroused state. The arousal can be due to any source, including physical exercise (folklore to the contrary, a fast game of basketball is more likely to exacerbate controversies than to cure them) and listening to loud or complex music

'Schachter's theory has recently come under attack because some of his earliest research could not be replicated (Marshall & Zimbardo, 1979; Maslach, 1979). However' it still provides the best available theoretical account of the body of research results cited in this section.

98 Social Conflict

(Konecni, 1975). The chances of escalation are diminished when people's emotions are subdued-for example, as a result of listening to soft, sweet music (Konecni, 1975)-or when they believe their state of arousal originates outside the controversy, in concern about health, tensions at home, or the like.

Inhibition of Aggression. The desire to punish another person is by no means always expressed in actual behavior. Indeed, aggressive impulses are usually inhibited. Inhibition may be due to a number of sources, including social condemnation of aggression, conscience, ability to empathize with the target's potential suffering, fear of punishment by the target, and ties with the target.

If social condemnation of aggression contributes to its inhibition, social endorsement should be disinhibiting. Thus, if Party is angry against Other and a third party acts aggressively, this should make it easier for Party to actually aggress. There is ample research support for this conclusion. The model for Party's aggression can be a fellow group mem ' ber (Wheeler & Caggiula, 1966), a character in a movie (Berkowitz & Geen, 1966), or even a football player in a game Party is watching (Goldstein & Arms, 1971). It follows that, to help avoid escalation in a difficult controversy, the disputants should shun aggressively tinged entertainment and that people who are in contact with them should try to behave peacefully.

People who have difficulty projecting ahead are especially likely to act aggressively when they are tempted to do so. This is because they are not deterred by fear of punishment from the target or third parties. For this reason, people who are under the influence of high emotion (Zillmann, 1979) or alcohol (Mulvihill & Tumin, 1969) are particularly likely to get involved in escalating controversies.

Displacement. The preferred target for the expression of aggression is the party who is blamed for the aversive experience. However, it is not always possible to indulge this preference. The source of annoyance may be well protected, or there may be extenuating circumstances that reduce his or her culpability, or it may be impossible to -identify the source, Under these circumstances, the desire to punish is sometimes displaced onto another target. If one cannot hit an offending boss, one can yell at one's spouse or kick the cat.

Evidence of displacement can be seen in a historical study by Hovland and Sears (1940), who found an inverse correlation between the price of cotton in the South and the number of blacks lynched over a 49-year period. The lower the price of cotton, the more lynchings. What presumably happened is that white farmers were frustrated by the decline in the cotton market but could not legitimately aggress against the cotton merchants who were paying less. Hence, they took it out on a handy displacement object, the black man. More recently, Berkowitz, Cochran, and Embree (1981) has shown that subjects who are forced to hold their hand in very cold water (another aversive experience) are more aggressive toward fellow subjects than are those who hold their hand in moderately cold water, another demonstration of displacement.

The phenomenon of displacement suggests that people who have had earlier aversive experiences in settings that do not permit aggressive behavior are especially likely to choose escalated tactics in the current conflict.

Negative Attitudes and Perceptions

An attitude is a positive or negative feeling toward some person or object. A perception is a belief about, or way of viewing, that person or object. Like all structural changes, negative attitudes and perceptions are both cause and effect of the use of contentious tactics. Accordingly, they are way stations in the escalation of conflict.

Attitudes and perceptions tend to be consistent in valence in the sense that, if I have negative (positive) feelings toward somebody, I tend also to have predominantly negative (positive) perceptions of that person. However, they are not 100 percent consistent. I may dislike a man and generally think ill of him but nevertheless trust him, because I have had a positive experience whenever I have relied on him. Hence, we must, to some extent, deal separately with these two psychological elements.

The following kinds of perceptions are particularly characteristic of escalated conflict. Adversaries tend to be seen as deficient in moral virtue-as dishonest, unfriendly, or warlike a They also tend to be distrusted; we believe them to be hostile to our welfare. In addition, they are sometimes seen as lacking in ability or achievement (Blake & Mouton, 1962), though this kind of perceptual distortion is less likely because of the greater availability of sound evidence about these characteristics (Brewer, 1979). By contrast, one's own side is seen often as more moral and sometimes as more able (White, 1984).

When groups are in conflict, a variant of these perceptions is sometimes found, which White (1984) has called the "evil-ruler enemy image." This is the perception that ordinary members of the other group feel neutral of even positive toward us but that their leaders are hideous monsters. In this view, aggressive actions taken by ordinary members of the other group are attributed to the fact that they are misled by their leaders, rather than to their moral degeneracy. Such images can be seen in American views of the Soviet Union and in Soviet views of America. They were also found in. the UB campus crisis described in the last chapter. In the midst of the crisis, the university president went on the air to denounce a small group of "vicious vandals" who were misleading the larger body of normally reasonable students. The evil-ruler enemy image appears to permit a decidedly negative view of the opponent while realistically acknowledging that not all members of any group can be evil.

Attitudes and the perceptions that accompany them tend to be similar on both sides of a controversy. This is the so-called "mirror image" hypothesis (Bronfenbrenner, 1961; Frank, 1982). For example, the profound distrust felt by most Americans toward the Soviet Union is mirrored in Soviet attitudes toward the United States.

Unfortunately, the existence of a mirror image is often not recognized by

parties involved in a conflict, who tend to distrust the adversary without realizing that the adversary also distrusts them. This lack of insight can contribute to the conflict spiral in the following way: If Other is behaving in a contentious fashion and we do not recognize that Other fears us, we assume that Other's behavior is aggressively motivated and therefore feel the need to escalate our response further.

Effects on Behavior. Negative attitudes and perceptions encourage escalation and discourage the settlement of conflict in at least seven ways.

The first is by encouraging a tendency to blame the object of the attitude for one's unpleasant experiences. People who have aversive experiences commonly seek a culprit to blame. However, the evidence about who is to blame is often ambiguous. Even when another party is clearly the source of the aversive experience, there may be extenuating circumstances. When evidence is ambiguous, attitudes tend to structure perceptions. This means that disliked parties are blamed and that liked parties are given the benefit of the doubt.

A finding by Blumenthal, et al. (1972) illustrates the impact of at ' titudes and perceptions on blame. During a period of political turmoil in the United States, in the summer of 1969, people were found to blame the conflict on groups whose views they did not like. Liberals blamed the police, whereas conservatives blamed the demonstrators. Both tended to use the term cc violence" to describe the behavior of groups they disliked and the term "justified force" to describe the behavior of groups whose views they favored. People were also more sympathetic to the use of force against the groups they blamed.

A second, and related, mechanism leading to escalation is that parties who are distrusted tend to be seen as threatening when their actions are ambiguous (Pruitt, 1965). They are given little benefit of the doubt or credit for good intentions. This produces an escalation of defensive and deterrent moves, and it often creates new issues for conflict.

Examples of this process are seen regularly in relations between the United States and the Soviet Union. Take, for example, American interpretations of the Soviet invasion of Afghanistan. This invasion is commonly seen as a sign of expansionist intent-as evidence that the Soviet Union is ready to invade other Middle Eastern countries such as Saudi Arabia. The United States has invested billions of dollars in additional armaments as a result of this perceived threat. Yet the evidence of this threat is ambiguous in the sense that there are other highly plausible interpretations of the Soviet invasion. Most notable is the view that the Soviet Union was simply trying to protect a beleaguered communist government from disintegrating and being reabsorbed into the Western community of nations. Under the latter interpretation, the Soviet invasion probably would still be labeled "aggression. " But it would not be evidence of a major threat to the West, because Soviet intentions would be seen as strictly limited (White, 1984).

The point of all this is that American distrust of the Soviet Union has led many Americans to choose, from the various possible interpretations of the Soviet invasion, the one that forbodes the greatest future danger. In a strictly parallel fashion, recent increases in the Ai-nerican arms budget have apparently been misunderstood in the Soviet Union as evidence of increased military threat. As in the case of the United States, distrust serves to shape Soviet perceptions of this inherently ambiguous evidence.

A third way in which negative attitudes and perceptions encourage escalation is by diminishing inhibitions against aggression among people who have been provoked. People are reluctant to aggress against parties they like and respect, even when these parties can clearly be blamed for unpleasant experiences, but they are quite willing to aggress against parties they do not like or respect. The finding that southern white students (many of whom can be assumed to have been prejudiced) retaliated more vigorously when insulted by a black than by a white (Rogers & Prentice-Dunn, 1981) supports these generalizations.

When negative attitudes lead to name-calling, inhibitions against aggression are particularly likely to fall away. Name-calling strengthens the impression that the other is morally inadequate and dissimilar to oneself Some names (such as the epithet "pig," which was hurled at policemen by student demonstrators in the 1960s) even make the other seem subhuman. Hence they tend to erode normal inhibitions against aggression.

A fourth way in which negative attitudes and perceptions encourage escalation is by interfering with communication. People tend to avoid those toward whom they are hostile. The point is well put by Coleman (1957): "As controversy develops, associations . . . wither between persons on opposing sides" (p. 11). This contributes to misunderstandings and hence to the proliferation of conflict issues. It also makes it difficult to reach a peaceful settlement of the controversy.

The reasons for this common development are not altogether clear. Why stop meeting and talking when one becomes hostile toward another party? Conceivably, people view association as implying friendship and endorsement. If so, contact with an opponent might send out false signals that one accepts the opponent's position. This is part of the reason for mutual nonrecognition between Israel and the Palestine Liberation Organization. Or, conceivably, the phenomenon has deeper emotional roots. According to balance theory (Heider, 1958). negative attitudes toward any object psychologically imply a negative relationship with that object-and a desire to put psychological distance between oneself and that object.

A fifth mechanism is that negative attitudes and perceptions tend to make it difficult to empathize with the adversary (White, 1984). Adversaries seem so different from us that it is hard to put ourselves in their shoes. Furthermore, there is an easy explanation for all of their actions which makes empathy seem unnecessary: these actions stem from evil motives. Absence of empathy is like absence of communication in that it fosters misunderstandings. It also encourages escalation by blocking insight into the conflict spiral. Awareness that the adversary's hostile behavior is a reaction to our own hostile behavior causes us to limit our escalation as a matter of self-protection.

102 Social Conflict

But if we lack empathy into'the adversary's motives, we are unaware of our own role in encouraging the other to aggress and are likely to escalate unthinkingly.

A sixth point is that negative attitudes and perceptions foster "zero-sum thinking"--the belief that what is good for you is bad for me, and vice versa. This happens partly because of distrust. If I believe you are hostile toward my interests, I will tend to doubt that you can endorse any alternative that helps me. Zero-sum thinking also arises from negative feelings toward the other party. Why should I help you if I don't like or respect you? An example of this phenomenon can be seen in the results of a study by Sillars (1981), in which it was found that students who blamed their roommates for past conflict tended not to take a problem-solving approach in dealing with their roommates. Syna (1984) has found a similar effect with married couples.

Zero-sum thinking is the antithesis of perceived integrative potential (PIP), a concept introduced in Chapter 2. Hence it tends to diminish perceived common ground (PCG), which makes problem solving an unlikely alternative. To resolve the controversy, Party must either yield or contend, and the latter is the more common approach because there so often are constraints against yielding. Thus zero-sum thinking often leads to conflict escalation.

The seventh and final point is that, when negative attitudes and perceptions grow severe, the adversary comes to be viewed as a "diabolical enemy" (White, 1984) and the conflict is seen as a war between light (our side) and darkness (their side). We are the chosen people; they are the "evil empire" (to quote President Reagan's statement about the Soviet Union). In such circumstances, we are ready to blame them for all that goes wrong, communication often takes a nose dive, empathy is especially weak, and problem solving is extraordinarily hard to sustain. Heavy tactics tend to become the rule and new controversies develop regularly, confirming our view of them and theirs of us.

Sources of Negative Attitudes and Perceptions. Like all structural changes, negative attitudes and perceptions result from escalation as well as contributing further to it. More precisely, they are affected both by Other's escalated actions and by Party's own escalated actions.

Aversive behavior that is blamed on.Other tends to produce an initial angry reaction characterized by a state of autonomic arousal. This is often followed by a cooler, longer-lasting residue, which serves to goad people toward aggression and hence toward escalation. The latter state is sometimes an active goal to punish Other, leading directl' to escalation. But more often it is simply a negative attitude toward Other, which acts in the seven ways we have described.

An alternative reaction to Other's contentious behavior is sometimes found. Instead of blaming Other, one might instead blame oneself as the source of behavior to which Other is reacting or blame both parties equally on the assumption that a conflict spiral is at work. However, this is less likely to happen, especially if there are multiple conflicts between oneself and Other (Syna, 1984). There are two reasons for the tendency to blame Other for controversy. One is ego-defensive: Finding fault with oneself is,more painful than finding fault with another person. The other is perceptual: We are much more aware of other peoples' contributions to controversy than of our own, because their contribution is more apparent to the senses. In Heider's (1958) terminology, their contribution "engulfs the field." For these reasons we tend to see Other as more causally central to the controversy.

The perceptions of other people or groups that have the greatest effect on our conflict behavior are the inferences we draw about their stable dispositions-their basic traits and motives and their attitudes toward us. If these dispositions seem negative, we are more likely to escalate our tactics; if they seem positive, we are less likely to do so. To understand the origins of such perceptions, we turn to two principles of attribution theory: the discounting principle and the augmentation principle (Kelley, 1973; Schneider, Hastorf & Ellsworth, 1979).

According to the discounting principle, Other's actions are more likely to be attributed to an underlying disposition when alternative causes can be confidently ruled out. Thus role-related behavior is less likely to be seen as due to underlying attitudes or personality traits than is behavior that departs from Other's role (Jones & Davis, 1965). For example, a bank employee who forecloses on our loan is less likely to be viewed as morally destitute than is a friend who takes the same action, because the bank employee can be seen as simply doing his or her job.

According to the augmentation principle, Other's actions are viewed as expressing an underlying disposition to the extent that he or she incurs risks or costs in enacting them. This implies that a person who jeopardizes his or her job to harm our interests is especially likely to be seen as having negative dispositions. If a porter should refuse to carry our bags for no apparent reason,, we are likely to view him or her in particularly negative terms, because his or her behavior can only be explained by assuming hostility toward us or toward people in our category.

It is no surprise to learn that Other's behavior affects Party's attitudes toward Other. But many people are astonished to learn that there is evidence (Bem, 1972; Festinger, 1957) that Party's own behavior toward Other also affects these attitudes and perceptions. Through a process of rationalization, people's feelings and beliefs tend to become consistent with their behavior.

There is no inconsistency between regarding behavior as a source of attitudes and regarding attitudes as a source of behavior. When people can choose among several courses of action, attitudes often influence their choice. But when environmental pressures push them in a particular direction, attitudes tend to fall in line.

We can conclude from these generalizations that Party's hostile behavior toward Other causes Party to dislike Other and see him or her in a negative light, whereas friendly behavior tends to erode such attitudes and images. (Evidence favoring the latter point can be seen in the experimental finding that people tend to trust others to whom they have directed promises of cooperation [Loomis, 1959].) This implies, in turn, another mechanism that can generate conflict escalation: Party takes contentious action toward Other, which encourages negative feelings and perceptions about Other, motivating heavier contentious tactics in the next time period.

Behavior affects attitudes only when the actor believes that it was freely taken and was not due to external pressures. If I call a policeman a pig and believe I did so of my own free will, I shall probably develop a negative attitude toward that policeman or all policemen. But if I believe that I was coerced into taking this action or (and this is rare) acknowledge that I was trying to impress somebody, my attitude will not be affected. The issue is not whether the behavior is actually freely taken (the authors of this volume take no stand on the question of whether people ultimately have free will) but whether the actor believes his or her behavior was freely taken. Because people usually believe that their actions are freely taken, behavior usually has a considerable impact on attitudes.


Another person is deindividuated when he or she is perceived as a member of a category or group rather than as an individual. This perception has no valence and hence cannot be viewed as a negative attitude or perception. But it nevertheless encourages contentious behavior, apparently by eroding in'hibitions against acting aggressively. What may happen is that people who are deindividuated seem less human than those who are individuated, and hence seem less protected by social norms against aggression.

Deindividuation was probably at work in an experiment by Milgram (1974) in which subjects in the role of "teacher" gave especially severe shocks to others in the role of "learner" when the latter were at a distance or out of sight. Deindividuation of the enemy may be what makes it easier for fliers to drop bombs on people they cannot see than for foot soldiers to shoot those they can see. Deindividuation is countered by receipt of information about others that makes them seem unique. Hence, guards in Nazi prison camps are said to have treated prisoners more leniently when they knew their names (Zimbardo, 1970).

Another way to discover that outgroup members are individuals is to have friendly relations with them over a period of time. It follows that residential settings that foster interracial friendships should lead to a reduction of white prejudice against blacks, an effect that has been demonstrated in two survey studies (Deutsch & Collins, 1951; Hamilton & Bishop, 1976). It is said that, before he led a protest demonstration, Gandhi would ask for hospitality from the local English governor and thus make friends with him. This was presumably a way of individuating himself-and his movement by association with him-in the eyes of the authorities, thereby reducing the aggressiveness of tactics used against his followers.

Similar reasoning suggests that aggressive or discriminatory impulses should lead to deindividuation of the prospective target, In this way, people

prediction emerges from a study by Worchel and Andreoli (1978). It was found that subjects who were angry with, or were expected to shock, another person were especially likely to forget individuating information about that person (such as his or her name) and to remember deindividuating information (such as his or her race). Name-calling during conflict may in part reflect this effect. An epithet such as "nigger or pig" reduces the individuality of the other, making it easier to aggress.

This finding implies that deindividuation is another way station in the circle of conflict escalation. Each participant deindividuates the other in order to rationalize his or her initial contentious moves. This state of mind then makes it easier to take more severe measures against the other.

In addition to viewing others as deindividuated, it is possible to see oneself in this way-in other words, to lose awareness of one's own distinct identity. This also facilitates aggression. Among the sources of self-deindividuation are acting in concert with others, wearing nondistinctive clothing, emotional arousal, and lack of sleep. In a study of the effect of clothing on aggression, Zimbardo (1970) found that college women playing the role of punitive teachers were especially likely to give shocks when they were wearing a hood. Such apparel reduces one's distinctiveness and, presumably, one's inhibitions. Military and police uniforms probably have a similar effect.


When groups, organizations, or nations become involved in contentious conflict, these collectives tend to change in at least six ways that contribute to escalation of the conflict. In describing these changes, we shall speak only of groups, because groups have been mainly studied in this context. But there is good reason to believe that these changes are found in all collectives.

First, group discussions often cause individual group members to become more extreme in their hostile attitudes and perceptions. This is due to the group polarization mechanism (Moscovici & Zavalloni, 1969). When group members share any view and discuss it with one another, this view tends to become stronger. Two main mechanisms apparently account for this phenomenon (Lamm & Myers, 1978). One is that the group members hear one another's views and the arguments underlying them. Finding that others agree with them, they feel their views are validated and also learn new arguments favoring them. The other is that a sort of competition develops among the group members, in which each strives to hold an opinion that is at least as extreme in the direction favored by the group as that advocated by the average group member. As a result, at least half of the members shift in the direction favored by the group.

In the context of social conflict, this means that psychological changes such as hostility and distrust are magnified when groups are involved. We see this phenomenon most clearly in mob action; a group of people who are upset about some incident gather and, through discussion, strengthen one another's sentiments to the point of angry action. But the phenomenon is not limited to


mobs. It can occur in perfectly stable and respected groups and organizations, including the United States Senate and the Soviet Politburo.

A second kind of change is the development of runaway nornm supporting a contentious approach to the controversy (Raven & Rubin, 1983). A norm is any attitude,' perception, goal, or behavior pattern that is shared by the dominant segment of a collective. Norms come to be seen as "right thinking" by most members of the group and are taught to new members and imposed on old members who appear to question them. Most of the psychological changes mentioned earlier in this chapter-including negative attitudes, distrust, zero sum thinking, and a reluctance to communicate with the other party ans become the subject of norms. When this happens, they gain more strength and stability than they would have if they belonged to a single individual or to a set of individuals who did not share a common group membership. Hence, escalation becomes more likely.

The development of contentious group goals is a third common outcome of conflict. Such goals include defeating or even destroying the adversary. Goals such as these arise from the conflict experience and fuel it. In addition, groups are capable of acting on their goals in ways that are not available to individuals. The activities of a number of individuals can be coordinated. Furthermore, a division of labor is possible, permitting highly complicated contentious routines, such as the recruitment and outfitting of an army. Hence groups are particularly efficient at conflict escalation if their members are so inclined.

A fourth kind of change that can contribute to escalation is the development of group cohesiveness or "solidarity," as it is commonly called. Groups are cohesive to the extent that their members find them attractive. Cohesiveness affects group behavior in three important ways. It encourages conformity to group norms (Festinger, Schachter & Back, 1950). This conformity is due in part to enhanced communication within the group (Back, 1951), in part to member fear of being ostracized (Festinger, 1950), and in part to social pressure, which is especially strong in cohesive groups (Schachter, 1951). Cohesive groups are also capable of especially vigorous action in pursuit of their goals. And there is reason to believe that members of cohesive groups are particularly convinced of the rightness of their cause and the effectiveness of their intended actions (Janis, 1972; Kriesberg, 1982).

For all of these reasons, we can expect group cohesiveness to augment or multiply the effect of the psychological states discussed earlier in this chapter. If the attitudes toward an outgroup are generally negative, they should be particularly strong in a cohesive group. If the other group is distrusted or seen as a threat, cohesiveness should strengthen these perceptions. If a goal of defeating the adversary is adopted and contentious tactics for achieving this goal are developed, a cohesive group mounts a particularly vigorous campaign against the adversary. In making these points, we do not intend to say that cohesiveness per se encourages antagonism or escalation. Research evidence (Dion, 1973) does not support such a position. The point is simply that cohesive groups are likely to be particularly militant in contentious conflict.

Contentious conflict has been repeatedly shown to enhance group cohe-

siveness (Dion, 1979; Harvey, 1956; Ryen & Kahn, 1975; Worchell & Norvell, 1980).' It follows that enhanced cohesiveness is still another mechanism that results from and encourages conflict escalation.

The fifth type of change that often occurs in groups engaged in heavy conflict is that they take on militant leadership (Sherif, et al., 1961). Every group has its leaders. Some are formally designated as such; others can be called leaders because group members are influenced by what they say. Groups ordinarily choose as their leaders people who resonate with the dominant sentiments of the members and are good at the activities to which the group is dedicated (Hollander, 1978). This is as true of groups in conflict as of groups engaged in any other kind of activity. If conflict involves negotiation, people with bargaining skills are likely to gain status. If it involves heavy contentious activity, leadership is more likely to fall into the hands of militants, who can mirror the anger of the membership and build a fighting force. Such leaders have particularly strong negative attitudes and perceptions of the adversary and are especially rigid in the demands they make. Accordingly, they tend to reinforce and augment the group's commitment to extreme tactics.

Leadership changes of this kind occurred on both sides in the UB campus crisis. The heaviest part of the controversy began when the campus police clubbed several demonstrators in the Student Union. At first, officers of the Student Government tried to exercise leadership over the campus, promising to negotiate with the university administration. But the students were so angry at the administration that they shunted these officers aside in favor of a group of radicals who had not previously exerted much influence. Similar changes occurred in the university administration. A vice-president who wanted to mediate the controversy was excluded from decision making, while other officers who advocated sterner measures came to the fore.

In addition to devising tactics for dealing with the opponent, leaders of groups that are in conflict often try to strengthen their members' dedication to the struggle, for example, by blackening the image of the adversary (Bowers & Ochs, 1971).

The sixth and last type of collective change that often occurs in escalating conflict is the development of new and more militant groupings. Sometimes this involves the organization of a new subgroup in a well established collective a committee or department to deal with the emerging struggle. At other times, it involves the development of an entirely new struggle group. The latter phenomenon will be discussed in some detail.

When a party in a conflict is an unorganized collection of people, a struggle group is sometimes born, which takes responsibility for defending that party's interests. The development of such a group is particularly likely to encourage further escalation, because, in addition to having all the other attributes of a group in conflict, a struggle group exists for the primary purpose of prevailing over the adversary. Such a development is often a turning point in the conflict.

Many community conflicts have developed in this way (Coleman, 1957), including the UB campus controversy, in which student radicals organized a Strike Committee that kept the campus in turmoil for the next six weeks. What happens is that people who have not previously communicated with each other gradually become aware of their common interests and collective identity. A sense of ingroup and outgroup begins to develop, often in conjunction with growing pride about (Apfelbaum, 1979), and favoritism toward, the ingroup. Radical spokespeople now emerge, whose pronouncements help crystalize the developing consciousness. An organized group is then formed, at first in miniature but often growing to a sizable membership with a large following of sympathetic nonmembers. This group has the dual function of defeating the opponent and fostering its own further growth.

At the end of the controversy, most struggle groups simply wither away. A few, however, go on to assume a permanent, legitimate place in the community as advocates of the interests of the people who gave them their origin. Several American social movements have followed such a line of development, most notably the labor movement.


When two groups come into heavy conflict with each other, it is often hard for other community members to remain neutral. One reason is that the participants in the controversy frequently seek support from others and demand that they decide whether they are "with us or agin' us." Another is that the use of escalated tactics is often annoying or frightening to the broader community. It is hard to remain indifferent when people are yelling at each other, hurting each other, or damaging each other's property. There is a tendency to cast blame in such circumstances, causing many third parties to join the side of the party to which they were initially closer or which seems to have escalated less and hence to be more properly considered the defender. This is the process of community polarization.

When communities polarize, their conflicts tend to escalate further. This is because of the destruction of crosscutting group memberships and the disappearance of neutral third parties who would otherwise urge moderation and mediate the controversy.



The three conflict models (aggressor-defender, conflict spiral, and structural change) were presented earlier as aids to a scholarly analysis of conflict. But they can also be seen as models of participant thought--oncepts that describe

the way the parties in an escalating conflict understand what is happening to them. Each model has implications with respect to action, so we can sometimes make predictions about the direction taken by a conflict if we know what models the participants in the conflict subscribe to.

A firm belief in the aggressor-defender interpretation often serves to exacerbate the conflict spiral. This is because parties who interpret their conflict in this way usually see themselves as the defender and the adversary as the aggressor. If surrender is out of the question, they must redouble their efforts at deterrence and defense to prove to their adversary that aggression does not pay. Seeing this, the other is likely to redouble its efforts as well, spawning@ a new round of contentious activity. The result is a conflict spiral. Thus the pace of the current arms race between the United States and the Soviet Union is bound to be accelerated by President Reagan's belief that "There is no arms race; they are racing and we are just trying to catch up."

On the other hand, belief that one is in a conflict spiral can serve to dampen this spiral. Parties who make such an analysis of their conflict are likely to avoid overly contentious actions in order not to antagonize the other, and to be conciliatory on the grounds that the other party will probably reciprocate (Tetlock, 1983). These are the "doves," by contrast to the "hawks" who make an aggressor-defender analysis. If the doves are right about the nature of the conflict (as they often are), this stance can contribute to de-escalation. For example, in 1977 President Sadat of Egypt, concluding that his country was involved in a conflict spiral with Israel, made a gesture of good will in the form of a personal journey to Jerusalem. This started a de-escalatory spiral in relations between these countries that resulted in the eventual resumption of diplomatic relations. If the doves are wrong (as they sometimes are), a soft conciliatory stance may simply encourage the adversary to redouble his or her efforts to force them to yield. For instance, after surrounding Indian outposts in 1961, Chinese forces withdrew in an effort to signal a desire to be conciliatory. Unfortunately, Indian leaders "interpreted the Chinese withdrawal as a sign of timidity [and] became even bolder in their efforts to occupy as much of the disputed territory, east and west, as was possible" (Lebow, Jervis & Stein, 1984).

A structural change analysis of the conflict one is experiencing implies a number of possible tactics. 'Some of these are similar to the implications of conflict spiral analysis. For instance, people are likely to mitigate harsh words and heavy tactics when they believe such actions will encourage negative perceptions and the emergence of hawkish leaders on the other side. Other tactics are peculiar to structural change analysis. For example, one may try to avoid structural changes in one's own party that will contribute to further escalation of a controversy. Thus a leader who suspects that a permanent defense establishment will become a strong advocate for hawkish policies may insist that only temporary agencies be formed to meet a current threat.

In addition, structural change analysis implies the importance of timing in reversing any actions one has taken that are resented by the other party (Pruitt & Gahagan, 1974). For example, it seems reasonable to assume that

110 Social Conflict

the UB campus crisis would have dissipated quickly if the administration had publicly apologized for the initial violence by the campus police, made restitution to the students who were assaulted, and arranged to drop the charges against those initially arrested. Such actions would probably have prevented the formation of the Strike.Committee. But once this committee had developed and numerous students had taken important positions in it, the campus was consigned to an extended period of heavy conflict.

The next chapter continues our discussion of escalation, focusing mainly on the question of how a conflict stays escalated once it has moved along this path.