Adaptation and Learning in Conflict Management- Yaacov Bar-Simon-Tov

ADAPTATION AND LEARNING IN CONFLICT MANAGEMENT, REDUCTION, AND RESOLUTION

Yaacov Bar-Siman-Tov

The question of how self-interested actors in a protracted conflict can cooperate so as to manage, reduce, or even resolve their conflict is one of the most interesting ones in the literature of conflict and resolution. This literature essentially suggests situational and ripeness conditions that reflect developments in the real conflict environment and changes in the parties' perceptions in the wake of crucial developments.

The situational and ripeness conditions are what determines the appropriate time for conflict reduction and resolution. Specifically, the following relevant conditions come into play: (1) the conflict environment (the relationship between the parties, the intensity of the conflict, and the balance of power); (2) the external environment (regional and global developments, the relationships between the adversaries, and external powers); and (3) internal environment (internal political and economic developments on each side, such as new elections, new leadership, economic crisis or difficulties). Generally, the views with respect to propitious situation fall into two further categories. First are those asserting that such conditions arise only at a certain stage of the conflict, such as at its inception or after its escalation (into a severe crisis or war). Second are those focusing on an appropriate mix of external or internal conditions that may develop at any stage of a conflict, and that arise from changing relationships between the adversaries and external sides, from developments in the global or regional system, or from internal developments involving the sides to the conflict (Zartman, 1985, 1999, 2000; Haass, 1990; Kriesberg, 1991; Rubin, 1991; Stedman, 1991; Kleiboer, 1994; Mitchell, 1995).

The ripeness condition is a necessary though not sufficient for conflict reduction and resolution. The rival parties can consider resolving their conflict only when the appropriate time or, more accurately, the appropriate set of conditions arrives. At the ripe moment, however, the adversaries must jointly perceive themselves to be in a costly hurting stalemate, and that unilateral military means are not only ineffective to accomplish their goals but also costly and risky. This may happen after an indecisive war or a series of such wars. The appropriate conditions are defined in terms of the conflict costs that are experienced by the rivals and their failure to gain any significant military or political advantages from existing strategies. In addition, the adversaries have some sense that there is a way out of the conflict via a negotiated solution (Zartman, 1985, 2000; Rubin, 1991; Kleiboer, 1994; Mitchell, 1995).

Both situational and ripeness conditions induce the parties to reconsider and modify the conflict. But because decisionmakers are the first to be influenced by changing conditions in the conflict and are responsible for reevaluating policies and adopting new ones, "it is the interpretation of these conditions by those leaders that determines whether the time, is indeed ripe" (Mitchell, 1995: 10). Explanations dealing with ripeness and initial processes of conflict reduction or resolution should focus, therefore, on the decisionmakers themselves. The question, then, is what leads decisionmakers to undertake a major change of policy in a conflict? The ripeness theory stresses the role of cost-benefit analysis, based on the assumption that the extended cost and pain entailed by a hurting-stalemate situation is the most important factor inducing a reevaluation by decisionmakers. This implies that leaders learn and change their minds mainly through experiencing the pain of damage and loss (Mitchell, 1995; Zartman, 2000). However, it is not clear why certain crises or wars trigger this kind of required learning whereas others do not, and how much pain over damage and loss is necessary for the learning that leads to conflict reduction or resolution. For example, we are not sure why four wars in the Arab-Israeli conflict (1948-1949, 1956, 1967, 1969-1970) failed to induce such learning and only the 1973 war did so. Was it because of the extent of the cost and pain? Was it because both sides' leaders at that time were more sensitive to the cost and pain than their predecessors? Was it because of the indecisive outcomes of the war? Or was it because of an energetic mediator (Kissinger)? We can assume that decisionmakers' learning is a major factor that mediates between hurting stalemate and ripeness for a change in a conflict. Since decisionmakers experience and react differently to extended pain because of their different beliefs and personalities or their different perceptions and interpretations of the conflict conditions, their learning processes are a main key to understanding how conflicts deescalate.

This study utilizes of theories of adaptation and learning, which recent studies of foreign policy have applied from social psychology in order to understand foreign policy shifts, and specifically the role of cognitive processes in major turning points in international conflict (Nye, 1987; Haas, 1990, 1991; Breslauer, 1991; Tetlock, 1991; Levy 1994a, 1994b; Stein, 1994). Adaptation and learning may mediate between structural and ripeness conditions, and are processes of evaluating conflict developments by decisionmakers that may lead them to change their attitudes, beliefs, and even behavior in the direction of conflict reduction and resolution.

Adaptation and Conflict Management

According to Haas, an adaptation process takes place when an actor changes its behavior in response to new events but without questioning its beliefs about basic causation or underlying values. Adaptation involves a realization that a previous set of measures or strategies are no longer effective in attaining one's aims. Although the ultimate goal remains the same, new paths of action are sought (Haas, 1990, 1991).

Adaptation, then, does not require a basic change of outlook. So long as a decisionmaker is reasonably successful in carrying out his policies through an adaptation process, there is little incentive to learn (i.e., change attitudes and beliefs). But even when the new means still prove ineffective in terms of the original aims, a further adaptation may occur: "If the decision makers then conclude that an alternative set of ends ought to be considered, without at the same time questioning the underlying cognitive schemata that establish a belief in cause-and-effect relations, we are still encountering adaptation" (Haas, 1991: 73). An adaptation process in a conflict may explain the shift from an unmanageable stage (zero-sum, or military and political decision) to a more manageable one (mixed-motive, or limited war or violence), or even to different manageable stages (shifting from one type of limited war to another), in which the parties use war and violence as a political means and a bargaining tool rather than as way of eliminating each other from the conflict.

An adaptation process is a change in the sides' behavior so as to meet the new military and political challenges following an unsuccessful war or a crisis, without seriously redefining their basic objectives. New political and military objectives are adopted without concern for their coherence with the original ones.

Adaptation is a rational adjustment of political and military means and objectives because of changes in the conflict environment; it stems from a perceived need to match means and ends more effectively. This requires an improved understanding of the nature of the conflict, and of how one can achieve one's objectives, leading to the selection of new military means and strategies. Adaptation is a process of changing one's problem-solving behavior that avoids a thorough reevaluation of one's attitudes and beliefs in the conflict. The parties, however, still believe that ultimately the only way to accomplish their respective, incompatible goals is through unilateral and militarily means, and they still await the right moment to do so.

Adaptation may result from repeated failures to accomplish one's objectives, or from new external or internal constraints that limit the rivals' freedom of action and make the accumulation of costs and risks undeniable. Repeated disappointment may be a necessary, though not sufficient, condition for adaptation. A failure in a war may constitute an incentive or a constraint toward changing one's behavior so as to function more effectively in the conflict or meet new challenges. However, one or both parties may ignore failures or constraints and persevere in old conflict patterns. On the other hand, repeated failures after a process of adaptation may lead to a new and more ambitious form of adaptation, where, for example, a side adopts a new military strategy, but still without questioning the underlying cognitive schemata. Thus, a side might shift from static war to offensive war.

Adaptation may also lead the parties to perceive the conflict more realistically and to limit their political and military objectives. The sides may realize that in the given circumstances, especially taking into account external or internal constraints or limited capabilities, it is not possible to achieve conclusive results but only partial ones. War should be only a political means to promote limited objectives. Sometimes the rivals to a conflict may enter the adaptation stage for asymmetrical reasons. Whereas one side may do so after a failure in a war and because of limited capabilities, the other may do so after winning the war and because of internal or external constraints on fully exploiting its capabilities in the conflict.

Although this change in behavior entails a shift to a limited war or low-intensity conflict, there is still no change in the belief that war and violence are the only effective means to advance one's political goals. When adaptation takes place, the aim of conflict management is not to prevent violence but to ensure that it will remain limited (Bar-Siman-Tov, 1994). The main characteristic of conflict management at this stage is regulation (Mitchell, 1981), which refers to developing internal or external mechanisms to keep war limited. The most important factors likely to influence the limitation of the conflict are the following: the balance of power between the parties, the balance of means of punishment in case of violations, the balance of advantages of keeping the war limited, the degree of autonomy of the parties, and third-party intervention.

The balance of power and the balance of punishment are probably the most important factors in the entire adaptation process. If the parties are relatively equal, they will be more interested in regulation of the conflict because the prospects that unilateral attempts will attain the sides' objectives are minimal, if not nonexistent, and entail great risk. As for equalization of advantages, it depends on the assumption that the nonuse or control of certain military means will not favor one party over the other in terms of increasing the probability of achieving its objectives in a less costly way. The degree to which the limitations are maintained depends mainly on their success in equalizing the cost-benefit tradeoff of violating versus upholding the limitations for both parties (George, 1958). The parties' degree of autonomy and the factor of third-party intervention are linked. When the sides fail to observe limitations and are dependent on external actors for expanding or limiting the conflict, then the role of a third party (generally a power or superpower) in the adaptation process becomes crucial.

Given the realities of protracted conflicts, such as the Arab-Israeli conflict, it is reasonable to assume that adaptation is not possible without external intervention or support. Because of difficulties in keeping the conflict limited, the adversaries need help from an external party. The third party's role is mainly to influence the rivals to reduce the violent intensity of the conflict. To encourage an adaptation process in the conflict, the third party can use various approaches. It may act directly or indirectly (via another power) to persuade one or both sides of the limitations of using military means to resolve or achieve substantial unilateral gains. The third party may also stress the risks involved in resorting to violence, which may include not only the dangers of limited local war but also its possible expansion into a less limited war, including military intervention by an external power. The third party's main task at this stage is to make clear to both sides that they will not be permitted to attain a total victory in the war (Bar-Siman-Tov, 1980, 1987).

In acting as a patron, the third party may resort to threats or inducements to get its client to keep the conflict limited. On the one hand, the third party may threaten to suspend military aid or deprive the client of other vital resources, and to disassociate itself from the client in case of noncompliance. On the other hand, it may promise military supply and economic aid, and even a military umbrella, to reward the client's self-restraint. The third party may act to strengthen the client's deterrent capacity against its rival in order to avoid it from preemption or escalation.

Even if adaptation is limited to regulation of the conflict (i.e., containing conflict behavior rather than preventing it), adaptation is an important technique of management that is based on some degree of mutual interest, understanding, and cooperation. Adaptation can succeed not only in keeping violence limited but also in making the parties realize the low effectiveness of military options for overcoming their incompatible interests. The limited outcomes of limited war, which reflect mutual concern about escalation or about constraints imposed by third parties, encourage the parties to seek a shift in the conflict.

The question is to what extent, and how, adaptation can modify a protracted conflict. The common assumption is that successful regulation transforms the means by which incompatible goals are pursued, rather than preventing all attempts to attain them (Mitchell, 1981). In this case, the modification of a conflict involves means rather than substance; the fundamental differences between the parties may remain unaltered. It may also be possible, however, to affect the substance of a conflict by gradually modifying the conflict discourse and transforming some of the underlying attitudes. A modification of this sort can sometimes be achieved through learning.

Learning, Conflict Reduction, and Conflict Resolution

For the purposes of this study, which maintains that learning is necessary for conflict reduction and resolution, I find useful Levy's definition of learning as "a change of beliefs (or the degree of confidence in one's beliefs) or the development of new beliefs, skills, or procedures as a result of the observation and interpretation of experience" (Levy, 1994a: 283). Use of this definition means that learning is restricted mainly to cognitive change (beliefs) on the individual level (decisionmaker) resulting in a different understanding of a certain reality, though not necessarily in changes in policy or behavior. Learning generally involves a basic transformation in a mode of thinking, including a thorough reassessment of fundamental beliefs and values (Tetlock, 1991: 45).

Learning may occur for such reasons as the following: (1) a negative experience involving repeated failures, unexpected failures, disappointments, or an outright disaster; (2) the failure of adaptation to solve a crucial problem; (3) a new understanding of how to solve a crucial problem; (4) the attainment of new information that may call into question current beliefs and policies; or (5) past policy successes (Haas, 1991; Nye, 1987; Tetlock, 1991; Levy, 1994a). Such developments may lead to two kinds of learning: causal or diagnostic. In causal learning, people change their beliefs about "cause and effect, the consequences of actions, and the optimal strategies under various conditions"; in diagnostic learning, they change their beliefs about "the definition of the situation or the preferences, intentions, or relative capabilities of others" (Levy, 1994a: 285).

The reevaluation of a situation, or a change in attitudes and beliefs, has great potential to lead to a behavior or policy change. However, personal, institutional, political, and economic obstacles may prevent the translation of learning into such a change. Learning may also strengthen decisionmakers' current attitudes and beliefs and actually discourage a policy change (Levy, 1994a).

Learning may lead to a policy change in a four-stage, causal process: (1) the observation and interpretation of experience lead to a change in decisionmakers' attitudes and beliefs; (2) attitude and belief change may lead to consideration of a policy change when decisionmakers acknowledge that this is necessary to reach their objectives; (3) adopting and implementing a policy change then depend not only on decisionmakers' willingness to do so, but also on their effective coping with potential obstacles to the change, as well as on building a wide consensus for it; and (4) institutional and political processes must mobilize the political support for translating decisionmakers' learning into a policy change (Levy, 1994a: 291).1 Since policy changes do not necessarily result because of belief changes, a claim that this has occurred needs to be demonstrated.

Learning is not necessarily linked to effectiveness or to positive connotations. The effectiveness of cognitive change sometimes becomes evident only much later, or sometimes not at all; alternatively, learning may bring liked and disliked changes. The question of whether decisionmakers learn rightly or wrongly, positively or negatively, from a particular experience is one of subjective judgment (Nye, 1987: 379-380).

Learning is necessary (though not sufficient) for a change in a conflict, but only when it is translated into a policy or behavior change; otherwise, learning remains only a potential factor. Learning in a conflict involves three types of belief change: about oneself in the conflict, about the other side, and about the conflict itself. Only mutual learning makes possible a change in a conflict. In this regard, the relevant questions are: (1) What changes in a conflict require changes in attitudes and beliefs? (2) What attitudes and beliefs need to be changed? (3) When are these cognitive changes translated into policy changes that are necessary for an overall change in the conflict?

Learning takes place at different levels and at different stages in the conflict process. Simple or tactical learning may bring about a shift from conflict management to conflict reduction and institutionalization; for a shift to conflict resolution, complex or strategic learning may be required (Nye, 1987: 380).2

Simple or Tactical Learning and Conflict Resolution

Simple or tactical learning is a low-magnitude change in the attitudes and beliefs of the sides in a conflict that amounts to a high-magnitude change in their behavior. Although the parties have not yet redefined their fundamental attitudes and beliefs and are not yet ready to resolve the conflict by peaceful means, their attitudes toward conflict management have seen a substantial transformation. Simple or tactical learning occurs when the parties begin to realize that war is no longer an effective means for achieving military and political objectives.3

This may occur not only when the parties internalize the potential costs of a new war, but also because they conclude unilateral gains are not feasible, and only by some cooperation in conflict reduction is there any chance not only of avoiding mutual damage but of achieving some of their incompatible goals. Such understanding is likely to develop after a crisis or an indecisive war that fails to achieve its minimum objectives or whose costs cannot be justified, with the sides finding themselves in a hurting stalemate. When the rivals are hurting, they may realize that while their military means can thwart the other side, making victory impossible for it, they are not effective for achieving victory for their own side.

When, however, a mutually hurting stalemate continues over time, both parties may realize that maintaining the conflict in its nature and intensity will make them worse off both in absolute and relative terms, so the most rational alternative is to find a way out of the conflict. Mutual simple or tactical learning may lead the sides to ripeness for conflict reduction. On the other hand, sometimes the very painfulness of the stalemate may stymie the development toward ripeness, because it intensifies the parties' mutual distrust, creates a sense that there is no prospect of deescalation, and hence may impel them to consider escalating again.

Because the adversaries expect the conflict will continue for a long time, they assume only conflict reduction and institutionalization can secure some degree of stability in their strategic relations. Their concordance is limited to security issues in a way that will not necessarily require a major shift in attitudes and beliefs, while at the same time assuring each side that its concessions are being reciprocated.

Institutionalization refers to formal or informal attempts to put conflict relations between the parties on "a more stable basis and predictable footing in order to reduce the magnitude, scope, and possibility of armed confrontation" (Hampson and Mandell, 1990: 194). The functions of institutionalization are the following: (1) to prevent crises; (2) to remove or reduce incentives for escalation; (3) to promote and facilitate deescalatory measures; (4) to establish new patterns of behavior leading to the development of more durable norms of conflict management; (5) to establish confidence-building measures or security regimes; and (6) to encourage expectations about resolution of the conflict (Hampson and Mandell, 1990: 196).

Institutionalization, therefore, has the potential not only to foster a better stabilization of the conflict but also to prepare the ground for its resolution. To institutionalize the conflict, the sides must accept the restructuring of their security relations, which requires some tacit or even explicit cooperation (Mandell, 1990). In order to manage their security relations, the sides need to create new norms and mechanisms. The new norms will define the limits of the sides' behaviors in their new conflict system, and the mechanisms will provide the means to enforce those limits (Mandell, 1990; Mandell and Tomlin, 1991).

The most important factors that can influence the degree of institutionalization are, the balance of power (current and future) between the parties, the degree of shared interests, the issues at stake in the conflict, the degree of autonomy of the parties, and third-party intervention. The balance of power between the parties may influence institutionalization in different ways. If the parties are relatively equal, they will be more interested in institutionalization because the prospects for unilaterally attaining their objectives are more doubtful and risky. In cases of power inequality, the dominant party may try to impose its preferred mode of institutionalization, whereas the weaker side may bargain to avoid this or seek to manipulate its alliance or its patron-client relationship so as to achieve a more favorable mode of institutionalization. The prospective power balance may also influence the degree of institutionalization. The party that expects to increase its power in the future tends to resist an unfavorable mode of institutionalization in the present; whereas the side that believes the balance will change to its disfavor will seek a high degree of institutionalization in the present (Kriesberg, 1973, 1991).

The degree of shared interests also plays a role. When the parties share only a concern to avoid undesirable outcomes, such as a crisis or war, institutionalization will be limited to minimal military arrangements for monitoring behavior and reducing uncertainties. The sides choose to cooperate not because it offers substantial gains but because it prevents losses. However, when the parties have mutual interests in extending their cooperation beyond military issues to the political sphere, this will augment institutionalization (Stein, 1992).

The types of issues involved in the conflict also affect institutionalization. When parties are dealing with military and security issues that they believe require immediate and clear-cut solutions, this poses an acute security dilemma. Therefore, institutionalization is likely to be more formalized so as to minimize problems of defection, violation of the understandings, and uncertainty. When cooperation is limited to security issues, while political issues remain untouched, cooperation is easier.

In terms of autonomy, parties that belong to a coalition or depend on external powers for arms supply, guarantees, or economic assistance will prefer less autonomous forms of institutionalization than parties that are more independent. Even in the latter case, however, external patronage may encourage a more stable institutionalization.

Third-party intervention can be important in determining whether simple or tactical learning will occur and in influencing the degree of institutionalization. In situations where the parties fail to begin the learning process by themselves, the role of the third party may be particularly important. The third party's role depends on the sides' relationships with each other and with the third party, the sides' willingness to cooperate with the third party, and the third party's willingness and ability to help the parties modify their relationship.

The third party can employ three strategies to change a conflict in the direction of reduction: pressing, integration, and compensation. Pressing refers to attempting to create a perception by the sides that the moment is ripe for changing the conflict by reducing the set of nonagreement alternatives available to them, while emphasizing the costs and risks of continuing the conflict. Integration is an effort to identify a solution in the context of common ground between the parties, while stressing the benefits of an agreement. Compensation means trying to induce the parties to make the necessary concessions by offering them guarantees and tangible aid so as to reduce the uncertainty associated with security cooperation and to reward them for their sacrifices (Carnevale, 1986; Mandell and Tomlin, 1991).

The third party can help the sides to create new norms and mechanisms for managing their security relationship so as to prevent undesired escalation and reduce the conflict. The third party may help the sides reach an agreement that is limited to security issues and will not threaten basic interests and core values on the one hand, and not necessarily require a major shift in basic attitudes on the other. The third party may also play an important role in convincing each side that its concessions are reciprocated by the other side, or that there is not necessarily a symmetrical reciprocity.

Security regimes are the most important outcome of simple or tactical learning. Once are they formally institutionalized, they may not only prevent war but also reduce the conflict. Effective security regimes intensify the learning process, helping each side to change its mode of thinking.

Complex or Strategic Learning and Conflict Resolution

When actors question original, implicit, and explicit attitudes and beliefs about a conflict, they may enter a process of complex and strategic learning. This may lead to a change in their beliefs, or the development of new beliefs about the conflict. In particular, the actors thoroughly reevaluate their beliefs about the basic causation of the conflict and/or diagnostically examine the conditions under which causal generalizations about the conflict apply (Haas, 1991; Larson, 1994). Complex or strategic learning also involves change in the parties' images of the conflict environment, which prompts new thinking about the conflict. This type of learning fosters changes in leaders; schemata that shape, in turn, a new policy direction for the conflict. When such learning occurs, a new understanding of the conflict issues emerges, new solutions are identified, and ultimately the goals in the conflict are redefined (Campbell, 1969; Hedberg, 1981; Sitkin, 1992; Levy, 1994a; Stein, 1994). The parties become ready to give up some of their goals so others can be achieved. National interests are redefined so that higher-order national values can be attained. For negotiations to succeed in resolving a conflict, complex or strategic learning is a necessary, though not always sufficient condition.

It has been suggested that complex or strategic learning results from failure, especially unexpected policy failures, crisis, or past policy successes. However, a notable success in conflict reduction as a result of simple or tactical learning may foster a process of further learning to the point that the nature of the conflict is affected and possibly even transformed. Effective security regimes may convince parties with incompatible goals to find some peaceful solution to their conflict.

On the other hand, sometimes effective conflict reduction can form a serious obstacle to conflict resolution. Because the need for alternative political outcomes is not urgent enough and the costs of resolving the conflict may appear higher than those of continuing it under controlled conditions, as in a security regime, the incentives to attempt conflict resolution may disappear. The costs of conflict resolution are usually not only territorial, political, or economic but also ideological and emotional.

The need to change attitudes, beliefs, and values often creates a situation of cognitive inconsistency, which causes distress to decisionmakers because it involves inconsistency in their value systems. People usually want their beliefs and values to be interconnected and mutually coherent. In peacemaking, a certain threshold of inconsistency is often crossed.

The need to change attitudes and beliefs explains why complex or strategic learning, and the shift toward conflict resolution, are so psychologically difficult for decisionmakers. Rather than moving toward conflict resolution, the sides often prefer to stay at the conflict reduction stage. In this case the situation that emerges is what Galtung (1967) calls "negative peace," (i.e., the relations between the parties are limited to conditions of maintaining and balancing power, freezing the status quo, and preserving security rather than seeking conflict resolution).

Certain conditions are, however, conducive to the emergence of complex or strategic learning: (1) The parties realize conflict reduction, even if it stabilizes the security and strategic relationships between them and minimizes the risks of war, lacks the potential to secure even some of the goals in the conflict, and these can be achieved only by conflict resolution. The parties, in other words, must reach a stage where conflict resolution seems to offer a better alternative than a continuation of conflict reduction. In this situation the parties are motivated not only by loss avoidance, as is the case with security regimes, but also by expectation of relative gains (Stein, 1992). (2) The parties realize, because of the success of the security regime, there are no immediate or even long-term opportunities for unilateral gain by war, or more limited violence. (3) The parties are aware that because of the success of the security regime it will be very difficult, if not impossible, to gain domestic and/or external legitimacy for the resort to military means. (4) There is an apprehension that without progress toward resolution, the conflict may revert to its violent stage. (5) There is a change in governmental leadership. New leaders are relatively free to undertake new initiatives, both because there is some domestic expectation that they will do so and because they are less committed to their predecessors' beliefs and policies. Furthermore, a change in leadership may well signal a change in orientation to adversaries (Breslauer, 1991; Kriesberg, 1991; Stein, 1994). (6) A significant change occurs in the international or regional system (such as the resolution of another major conflict, a change in the structure of the system, the withdrawal of a major ally or patron, or a dramatic change in its international orientation) that forces the sides to reconsider the cost/benefit of effective conflict reduction versus that of conflict resolution. Such external events usually help to legitimize the elite in power and to reinforce a prevailing set of new ideas (Stein, 1994). (7) An energetic and powerful third party induces or coerces the parties to settle the conflict.

In essence, the third party's role in the complex or strategic learning process is to create the perception among the parties that the moment is, indeed, ripe for conflict resolution and, at the same time to convince them that certain possible solutions are preferable to the existing situation. In other words, the third party has to generate complex or strategic learning by altering the preference structure of the parties (Mandell and Tomlin, 1991). The latter need to realize that the stabilization of their security relationships is not enough to attain even part of their political goals in the conflict, whereas conflict resolution has the potential to do so; moreover, without progress toward resolution the conflict may revert to its violent stage.

The third party also may alter the nature of compensation to make conflict resolution more attractive, such as by substantial financial and military aid. By committing itself to guarantee compliance with all agreements reached between the sides and by assenting to be a signatory to agreements, the third party makes the shift toward resolution less costly (Mandell and Tomlin, 1991).

From Adaptation to Learning

For effective deescalation and eventual resolution of a protracted conflict, the sides must pass from the unmanageable stage of the conflict to adaptation, simple or tactical learning, and then to complex or strategic learning. The three processes may occur sequentially or out of sequence; nor do the parties necessarily pass through all of them (though, as observed below, simple or tactical learning may be a necessary condition for complex or strategic learning). There is also, in a protracted conflict, a high potential for regression to former stages.

Such factors as changes in the balance of power, destabilizing domestic forces, or changes in the perception of interests may cause regression to a less advanced stage. Any regression may result in the preference of conflict reduction to resolution, the preference of war to reduction, or a failure to prevent a war. Generally when a regression occurs, the parties fail to shift to institutionalization and find themselves again in a regulation situation that precedes adaptation. A failure in complex or strategic learning does not necessarily mean a reversion to simple or tactical learning or even to adaptation; the reversion may be more drastic. Concern about regression may restrain the sides from undertaking initiatives that will endanger the positive shifts in a conflict.

Whereas the emergence of adaptation and simple or tactical learning may result from repeated failure in war or an indecisive war, the emergence of complex or strategic learning may result mainly from past success. Effective simple or tactical learning may encourage the emergence of complex or strategic learning. Adaptation and simple or tactical learning aim to avoid unilateral or mutual damage; complex or strategic learning also aims to achieve mutual gains.

Since complex or strategic learning demands far-reaching change in decisionmakers' attitudes, beliefs, and goals, the common assumption is that it is a rare phenomenon, whereas adaptation is more common. Although actors tend to adjust their policies in response to repeated failure so as to avoid damage or to match means and goals more effectively, far-reaching redefinition of the basic nature of a conflict, or of the national interest, is indeed extremely rare. Haas (1991), for example, maintains that learning occurs only when conditions of perceived urgency, feasibility, and desirability of the necessary change converge. Similarly, Weber (1991) argues that a "critical learning period" is required for a significant change to take place.
The shift of a conflict toward resolution indeed requires a perception of urgency, feasibility, and desirability by the sides. The parties should perceive resolution as crucial to the national interest and its feasibility as very high. Simple or tactical learning may, via institutionalization of the conflict, provide a critical learning period.

Complex or strategic learning as an outcome of a linear development in a conflict is assumed to secure the shift toward resolution. But learning is also important in internalizing the norm that peace is better than war, not only because it provides some goods that war cannot provide and because war is too risky, but because war and violence are not appropriate means for resolving conflicts. The internalization of that norm is crucial for stabilizing peace agreements. Although ideally complex or strategic learning will ensure peaceful change, it need not subordinate all differences in values, culture, and identity. Compatibility of interests, of attitudes and beliefs, does not entail compatibility of political orientation and political institutions.

The Institutionalization of Learning

Given the realities of a protracted conflict, it is reasonable to assume that complex or strategic learning (defined as changes in core attitudes, beliefs, and goals) is a necessary condition for conflict resolution. Nevertheless, when we use learning as the main independent variable to explain significant change in a protracted conflict, there is a danger, as Breslauer (1991: 843-845) notes, of inadequate appreciation of the political bases of policymaking in a conflict, and overestimation of the importance of learning in how a conflict develops, as opposed to explanations that stress the origins of a conflict in deeper, objective conditions that constrain or impel actors to undertake different policies.

Although objective conditions, both external and internal, must not be ignored, learning by individuals and by the leadership in general plays an important role in understanding those conditions and translating them into concrete policies that effectuate change in a conflict. Indeed, greater understanding of the conflict environment (learning what) is distinct from greater willingness to take steps toward conflict resolution (learning how). Even though personal, economic, or political constraints may prevent implementation of the preferred policies based on what leaders have learned, it is difficult to suppose that without such learning (i.e., change in beliefs) leaders can undertake dramatic changes. Learning is, therefore, a necessary condition for contemplating a change in a conflict, but insufficient for implementing a policy of change.

The question, therefore, is how an individual's learning is transformed into foreign policy initiatives that dramatically change a conflict. As Stein (1994: 180) points out, "institutional and political processes must intervene to build the political support to transform individual learning into changes in foreign policy behavior.… At a minimum, learning must be institutionalized in the central political agencies, a dominant political coalition must be committed to the new representations of problems, and new policies must be created". Institutionalization of learning, in turn, requires systematic analysis of such political and institutional variables as type of regime, type of leadership, leadership skills, coalition building, legitimacy for peacemaking, and so on.

However, for transformation of learning into effective change in a protracted conflict, institutionalization of learning in only one party is insufficient. Learning in one side may trigger learning in the other side; in any case, only mutual learning has the potential to change the conflict. Learning in one side that is not recognized as such by the other side may even prevent learning by that side. Therefore, each side needs to signal to the other via positive declarations or actions that its learning is sincere and not a manipulative act to gain domestic or external support. A third party may also be helpful in convincing both sides that learning is taking place.

The Rate and Scope of Learning in Different Dyadic Conflicts

Learning in each dyadic conflict may develop separately, independently, and differently from other dyadic conflicts in the same conflict system, such as the Arab-Israeli framework. The Egyptian-Israeli conflict, for example, was resolved long before the beginning of the resolution process of the other dyadic conflicts in the system. The assumption is that different rates and scopes of learning in each dyadic conflict influence the shifts in each conflict toward resolution (the rate of learning refers to its extension and scope refers to its depth). What determines the rate and scope of learning in each dyadic conflict?

The following variables may be important: type of conflict; issues in the conflict; history of violent interaction; mutual trust or mistrust; the sides' interests in the conflict; hurting stalemate; balance of power; type of leadership; domestic environment; external environment; and existence of a third party.

The type of conflict relates to the kinds of actors who participate in the conflict. In a conflict in which the actors are states, the rate and scope of learning may be faster and deeper than in conflicts where the actors are not states, or where one is a state and one is not. The issues at stake in the conflict refer to its substance together with its territorial, security, economic, political, and ideological dimensions. The rate and scope of learning in a territorial or security conflict are usually slower than in an economic or political conflict, because of the difficulty of changing attitudes and beliefs on issues perceived as the most momentous by the parties. Moreover, questions of deception and uncertainty are more critical in this kind of conflict.

The violent history of a conflict may induce bitterness and diminish the rate and scope of learning. Conflict reduction that eliminates or decreases violence may be required to enhance confidence and encourage greater learning. Mutual mistrust is also an obstacle to learning; in such cases, trust between the sides must be strengthened if the learning process is to be improved. The more vital the sides' interests in the conflict, the greater their difficulty in making painful concessions; so the rate and scope of learning tend to be low.

A hurting stalemate may induce faster and deeper learning so as to ease the parties' difficult situation. However, this kind of situation may impel the sides only toward simple or tactical learning, which is sufficient only for conflict reduction. In terms of balance of power, when the parties are relatively equal they will be more interested in speeding up the learning process, mainly because unilateral attempts to attain their objectives are less likely to succeed. Both rivals realize that not only this kind of balance of power does not enable them to wipe out the other side by force but any employment of force will be costly and risky. Moreover, a stable balance of power may serve better security interests.

The type of leadership most likely to accelerate and deepen the rate and scope of learning is a leadership that, for reasons of accountability, has the utmost interest in reducing or resolving the conflict. Accountable leaders are those who believe that they are obligated to their people and to history to reduce or resolve a conflict. Such leaders will try to institutionalize their learning, so that ruling and competitive elites, pressure groups, and the public itself will adopt their learning and become oriented to conflict resolution.

The domestic environment also affects the rate and scope of learning. Domestic factors are broadly defined to include the basic political conditions, such as the political system and processes. The political system includes both the type of regime and the political actors who play a role in formulating and implementing foreign policy. When domestic actors are concerned about a conflict's negative effects on the state, the society, and their own interests, the rate and scope of learning will be greater than in a case where they perceive the conflict as serving their interests. A strong, motivated leadership may manipulate a hurting situation to deepen the learning among domestic actors.

The external environment may indirectly influence the rate and scope of learning. A movement toward peace in the regional or global system, or successful cases of conflict resolution in different areas, may encourage further progress toward learning. An energetic third party that is willing and able to help change the conflict may encourage a greater rate and scope of learning by the parties.

Complex or strategic learning in one of the dyadic conflicts in a multiparty conflict system may develop independently and separately from the other conflicts, because of different influences of the variables surveyed. Nevertheless, such a learning process, especially if its outcomes are successful, may spill over to the other conflicts in the system. But because of different influences of the relevant variables, the parties in the other conflicts may face different difficulties in advancing their learning.

Conclusions

This study maintains that shifting from war to peace can be explained by processes of adaptation and learning. The latter are cognitive processes whereby decisionmakers perceive their conflict environments differently and decide to change their policies toward conflict. The use of adaptation and learning as explanatory variables does not ignore the "objective" conditions of the conflict, which may also bring about changes; the study has focused on the perceptual consideration of these conditions by the decisionmakers. The assumption is that leaders are primarily responsible for shifts in a conflict, and that shifts are made only following a cognitive process. Adaptation is a change in behavior that results from observation and interpretation of experience in a conflict. Adaptation stems from failure to achieve one's aims in the conflict, which mainly means failure in a war. Adaptation does not involve change of attitudes, beliefs, and values; it is primarily an attempt to employ means and strategies in the conflict more effectively, in the hope that better understanding of constraints may lead to better results.

Adaptation may lead to a more realistic understanding of the conflict in general, and this may lead to different management of the conflict. Consideration of the limitations of military means, because of limited capabilities or domestic and external constraints, may lead the parties to use violence as a more restricted, political means. Limited war and its regulation may then be the net result of the adaptation process.

Ineffective limited war may lead the sides to a new observation and interpretation of experience, which may lead in turn to simple or tactical learning; the latter does entail a change of attitudes. Simple or tactical learning may lead to a dramatic change in the conflict. Although the sides are not yet ripe for conflict resolution, they realize that war, because of its limited benefits and high costs, cannot accomplish their objectives; hence their immediate aim is to prevent war which means they must cooperate militarily.

Simple or tactical learning may lead to the institutionalization of the conflict, which takes the form of security regimes that may restructure the security relationship between the sides. Simple or tactical learning may be a necessary condition for the emergence of complex or strategic learning.

Complex or strategic learning entails a change of beliefs and is a necessary, though not sufficient, condition for shifting the conflict from reduction to resolution. Individual learning seems to be necessary but not sufficient for conflict resolution, especially in democratic states where leaders need to convince others in the political echelon and in the society at large before they can translate this learning into operational terms. In other words, learning must be institutionalized.

The institutionalization of learning is necessary for a transformation of national interests and values that will enable peacemaking. The rate and scope of learning in a conflict may explain why different conflicts move differently toward conflict resolution.

Notes

1. I have extended Levy's schema from three to four stages.
2. I have borrowed the terms simple and complex learning from Nye (1987, p. 380), while elaborating them differently.
3. This definition of simple or tactical learning differs from that suggested by Nye (1987, p. 130), who defines it as adapting the means "without altering deeper goals in the end-means chain".

References

Bar-Siman-Tov, Yaacov. 1994. "The Arab-Israeli Conflict: Learning Conflict Resolution." Journal of Peace Research, Vol. 31, No.1, February, pp. 75-92.

Bar-Siman-Tov, Yaacov .1987. Israel, the Superpowers and the War in the Middle East. New York: Praeger.

Bar-Siman-Tov, Yaacov. 1980. The Israeli-Egyptian War of Attrition 1969-1970: A Case-Study of Limited Local War. New York: Columbia University Press.

Breslauer, George B. 1991. "What Have We Learned about Learning?" In George B. Breslauer and Philip E. Tetlock, eds., Learning in U.S. and Soviet Foreign Policy.
Boulder, Col.: Westview Press.

Campbell, Donald T. 1969. "Reform as Experiments." American Psychologist, Vol. 24, No. 1, April, pp. 409-429.

Carnevale, Peter J. 1986. "Strategic Choice in Mediation." Negotiation Journal, Vol. 2, No ,1 January, pp. 41-56.

Galtung, John . 1967. "Peace Research: Science or Politics in Disguise?" International Spectator, Vol. 21, No. 19, pp. 1573-1603.

George, Alexander L. 1958. The Quid Pro Quo Approach to the Study of Limitation. Santa Monica, Cal.: Rand.

Haas, Ernst B.1991. "Collective Learning: Some Theoretical Speculations." In George B. Breslauer and Philip E. Tetlock, eds., Learning in U.S. and Soviet Foreign Policy.
Boulder, Col.: Westview Press.

Haas, Ernst B. 1990. When Knowledge Is Power. Berkeley, Cal.: Berkeley University Press.

Haass, Richard N.1990. Conflicts Unending. New Haven: Yale University Press.

Hampson, Osler F. and Brian S. Mandell. 1990. "Managing Regional Conflict: Security Co-operation and Third Party Mediators." International Journal, Vol. 45, No. 2, Spring, pp. 191-201.

Hedberg, B. 1981. "How Organizations Learn and Unlearn." In Paul C. Nystrom & William H. Starbuck, eds., Handbook of Organizational Design, Vol. 1. New York: Oxford University Press, pp. 3-27.

Kleiboer, Marieke. 1994. "Ripeness of Conflict: A Fruitful Notion?" Journal of Peace Research, Vol. 31, No.1, February, pp. 109-116.

Kriesberg, Louis. 1991. "Timing Conditions, Strategies, and Errors." In Louis Kriesberg & Stuart J. Thorson, eds., Timing the De-Escalation of International Conflicts.
Syracuse: Syracuse University Press.

Kriesberg, Louis. 1973. The Sociology of Social Conflicts. Englewood Cliffs, N. J.: Prentice-Hall, 1973.

Larson, Deborah W. 1994. "The Role of Belief Systems and Schemas in Foreign Policy Decision-Making." Political Psychology, Vol. 15, No. 1, March, pp. 17-33.

Levy, Jack S. 1994a. "Learning and Foreign Policy: Sweeping a Conceptual Minefield." International Organization, Vol. 48, No, 2, Spring, pp. 279-312.

Levy, Jack S. 1994b. "Learning from Experience in U.S. and Soviet Foreign Policy." In Manus I. Midlarsky, John A. Vasquez, and Peter V. Gladkov, eds., From Rivalry to
Cooperation: Russian and American Perspectives in the Post-Cold War Era. New York: HarperCollins.

Mandell, Brian S. 1990. "Anatomy of a Confidence-Building Regime: Egyptian-Israeli Security Co-operation, 1973-1979." International Journal, Vol. 45, No. 2, Spring, pp.
202-223.

Mandell, Brian S and Brian M. Tomlin. 1991. "Mediation in the Development of Norms to Manage Conflict: Kissinger in the Middle East." Journal of Peace Research, Vol.
28, No. 1, March, pp. 43-55.

Mitchell, Christopher R . 1995. "Cutting Losses: Reflections on Appropriate Timing." Working Paper, No. 9, Winter. Fairfax, Va.: Institute for Conflict Analysis and
Resolution.

Mitchell, Christopher R. 1981. The Structure of International Conflict. New York: St. Martin's.

Nye, Joseph S. 1987. "Nuclear Learning." International Organization, Vol. 41, No. 3, Summer, pp. 371-402.

Rubin, Jeffrey Z.1991. "The Timing of Ripeness and the Ripeness of Timing." In Louis Kriesberg and Stuart J. Thorson, eds., Timing the De-Escalation of International
Conflicts. Syracuse: Syracuse University Press.

Sitkin, Sim B. 1992. "Learning through Failure: The Strategy of Small Losses." In Larry L. Cumming & Barry H. Straw, eds., Research in Organizational Behavior. New
York: JAI Press, 1992, Vol. 14, pp. 231-266.

Stedman, Stephen J. 1991. Peacemaking in Civil War: International Mediation in Zimbabwe, 1974-1980. Boulder, Col.: Lynne Rienner.

Stein, Janice G. 1994. "Political Learning by Doing: Gorbachev as Uncommitted Thinker and Motivated Leader." International Organization, Vol. 48, No. 2, Summer, pp.
155-183.

Stein, Janice G. 1992. "International Co-operation and Loss Avoidance: Framing the Problem." International Journal, Vol. 47, No. 2, Spring, pp. 202-234.

Tetlock, Philip. 1991. "Learning in U.S. and Soviet Foreign Policy: In Search of an Elusive Concept." In George B. Breslauer and Philip E. Tetlock, Learning in U.S. and
Soviet Foreign Policy. Boulder, Col.: Westview Press. pp. 20-61.

Weber, Steven. 1991. "Interactive Learning in U.S.-Soviet Arms Control." In George B. Breslauer and Philip E. Tetlock, Learning in U.S. and Soviet Foreign Policy. Boulder, Col.: Westview Press.

Zartman, William I. 2000. "Ripeness: The Hurting Stalemate and Beyond." In Paul C. Stern and Daniel Druckman, eds., International Conflict Resolution after the Cold War. Washington: National Academy Press.

Zartman, William I.1999. "Negotiations as a Mechanism for Resolution in the Arab-Israeli Conflict." Davis Occasional Papers, No. 72, August, Jerusalem: Leonard Davis Institute for International Relations.

Zartman, William I. 1985. "Conflict Reduction: Prevention, Management, and Resolution." In Francis M. Deng and I. William Zartman, eds., Conflict Resolution in Africa. Washington: Brookings Institution.