A Cognitive Approach to the Study of Foreign Policy

[
A Cognitive Approach to the Study of Foreign Policy
Jerel A. Rosati, UNIVERSITY OF SOUTH CAROLINA
Editors' Introduction
In this chapter Jerel Rosati introduces the role of cognitive studies in foreign policy analysis. As Rosati illustrates, cognitive analyses consider the role of policymakers' beliefs and images and are pre-sent in both the first and second generations of foreign policy analysts. Indeed. much of what this chapter discusses can be found in literature focusing on the "individual" level of foreign policy analysis. Rosati makes explicit the changes occurring in cognitive studies throughout the evolution of foreign policy analysis and emphasizes the contributions of recent , second-generation. cognitive studies. This chapter complements the following chapter by Keith Shimko, which looks specifically at the metaphors (a cognitive concept) leaders use in developing foreign policy.
In considering the role of cognitive sources of foreign policy, can you think of any modem examples of foreign policy makers whose beliefs and images influenced foreign policy? For example, would Président Clinton's belief System have led him to a different policy in the Gulf War than the one pursued by President Bush? What type of belief system might have been behind Iraqi leader Saddam Hussein's decision to invade Kuwait? Consider your own belief system. How would il influence your foreign policy actions if you were a policymaker? How important are cognitive factors as compared to other sources of foreign policy (e.g., bureaucratic. domestic political systemic)? ■
How do policymakers view the world? What affects the beliefs and images of policy-makers over time? What impact do the beliefs of policymakers have on foreign policy making and the practice of foreign policy? These are the kinds of questions about the role of cognition that interest many students of foreign policy and are addressed in their scholarship, hence contributing to knowledge and understanding about the for¬mulation and conduct of foreign policy. Although the value of a cognitive approach should be obvious, it has not always received the attention it deserves in the study of foreign policy.1

50 Second-Generation Foreign Policy Analysis
Traditionally, foreign policy has been explained from a rational actor perspec¬tive common to the realist and power politics tradition. The assumption has been that governments, and their political leaders, think and act in a rational manner in their quest for power and order. Such rationality assumes that individuals perceive the world accurately and arrive at decisions through an open intellectual process: goals are ordered, a search is made for relevant information, a wide range of alterna¬tives is considered, and the option that maximizes the benefits while minimizing the costs is selected. Since policymakers act rationally in pursuit of power and order, there is no need to delve into their psychological predispositions or closely examine the governmental policymaking process. Instead, one should focus on how the inter¬national System constrains foreign policy action, treat the government as a rational actor, and speak in terms of an overriding shared national interest in the making of foreign policy.
A cognitive approach challenges much of Western thought and practice premised on the assumption of individual rationality (see Allison 1971, 10-38; Simon 1957b; Snyder, Bruck, and Sapin 1962; Steinbruner 1974, 25-46). Where the rational actor perspective assumes individual open-mindedness and adaptability to changes in the environment, a cognitive approach posits that individuals tend to be much more closed-minded due to their beliefs and the way they process informa¬tion—thus, they tend to resist adapting to changes in the environment. A cognitive perspective emphasizes the importance of examining the individuals involved in the policymaking process, for they are likely to view their environment differently. This approach has grown in visibility, prominence, and sophistication since the 1950s as social scientists have attempted to be more systematic in identifying and explaining major patterns of foreign policy.
This chapter is organized along three lines so the reader can arrive at a better understanding of the development of a cognitive approach and its relevance for contributing to an understanding of the theory and practice of foreign policy. First, a brief history of the evolution of political psychology and its impact on a cognitive approach to the study of foreign policy is provided. Second, a number of significant research programs and individual studies are highlighted to illustrate different cognitive approaches to the study of foreign policy. Finally, an assessment is made of what has been learned by a cognitive approach to foreign policy over the past few decades and what issues require future attention. Together, this should provide the basis for clarifying how a cognitive approach can contribute to a better understanding ot the dynamics of foreign policy.
■ The Political Psychological Study of Beliefs and Foreign Policy
Scholars and individuals have been interested in examining the beliefs of individuals and the workings of the human mind throughout the ages. Yet it was not until the 1930s, with the development of psychology as a discipline of study, that systematic efforts to apply psychological approaches to the study of politics and international relations really began. The evolution of political psychology and its relevance to inter¬national relations and foreign policy is marked by three distinct periods. With each new period the political psychological study of beliefs has grown tremendously in

A Cognitive Approach to the Study Of Foreign Policy 51
sophistication and has contributed to a powerful understanding of the theory and practice of foreign policy.2
Early Efforts
Early efforts were made beginning in the 1930s and continuing through the early 1950s to apply explicitly psychological concepts to the study of politics, especially the nature of war and peace. Most of this research focused on national stereotypes, atti¬tudes toward war, and public opinion on foreign policy issues (Kelman 1965b; Klineberg 1950; Pear 1950). Unfortunately, while most of these early efforts were made by scholars and individuals who were well versed in psychology, they usually lacked a strong foundation in the study of international relations and world politics. This was best exemplified by the studies on individual irregularities and pathologies that were directly projected onto the nation-state, the "war begins in the minds of men" approach, and national character studies of the war proneness of different societies.
This well-intentioned effort by many psychologists to apply different psycho¬logical concepts and knowledge about the individual "directly" to the complex arena of the state and world politics was not well received by students of international rela¬tions (see Holsti 1976, 16-20; Jervis 1976, 3-10). Most international relations scholars found these studies unrealistic and believed they were not relevant to the study of foreign policy and world politics (see, e.g., Waltz 1959). The problem was clearly stated by Herbert Kelman in International Behavior: A Social-Psychological Analysis: "Only if we know where and how these individuals fit into the larger process, and under what circumstances they operate, are we able to offer a relevant psychological analysis" (1965b, 6). Clearly, until political psychology was well grounded in politics and international relations such studies would likely remain naive and simplistic.
Although the initial effort to incorporate psychological approaches into the analysis of international behavior failed to influence the field, some scholars began to question the lack of psychological input into the study of international relations and foreign policy. Quincy Wright, in his magnum opus, The Study of International Relations, proposed that psychology belongs at the "core" of the discipline: "International relations cannot, therefore, be confined to intergovernmental relations and conclusions based on the assumption that they [i.e., psychological studies] tail to provide an adequate foundation for prediction and control. The minds of individuals who constitute the world's population, the influences that affect them, and the influ¬ences they exert, both domestic and foreign, must be taken into account by examining their minds" ( 1955, 433).
The First Generation of Scholarship
Beginning in the mid- 1950s, the contribution of psychological approaches to the study of international relations grew in importance due to the interaction of the "peace research" movement and the "behavioral" revolution in the social sciences (see Kelman 1965b; Kelman and Bloom 1973). A number of psychologists, sociologists, economists, anthropologists, and other scientists became interested in applying

52 Second-Generation Foreign Policy Analysis
the knowledge and techniques of their disciplines to the problems of war and peace. At the same time, many international relations scholars became interested in making their field more empirical and scientific. Nowhere was this interaction greater than among those who advocated a decision-making approach to the study of foreign pol¬icy (see Snyder, Bruck, and Sapin 1962).
In comparison to the early efforts by psychologists, the late 1950s and 1960s represented the beginning of a qualitative leap forward for psychological approaches and their application to the study of international relations. Psychological approaches increasingly were applied from an international relations and political perspective. Two groups of specialists "emerged and interacted closely with one another: students of international relations, with a political science background, who are thoroughly grounded in social-psychological concepts and methods; and social psychologists (as well as students of other disciplines outside of political science) who have systematically educated themselves in the field of international relations" (Kelman and Bloom 1973, 263).
These developments meant that the systematic study of the beliefs and images of foreign policy makers grew in popularity and significance during the 1960s and 1970s. The working assumption was that the ideas and thoughts about the environment held by policymakers affect the foreign policy-making process. Policymaker images "mav be partial or general. They may be subconscious or may be consciously stated. They may be based on carefully thought-out assumptions about the world or they may flow from instinctive perceptions and judgements. In any event ail decision-makers may be said to possess a set of images and to be conditioned by them in their behavior on for¬eign policy" (Brecher, Steinbcrg, and Stein 1969, 86-87).
Much of the psychologically oriented research on foreign policy was influenced by the study of attitudes and attitudinal change in psychology that began as early as the 1930s (see Calder and Ross 1973; Oskamp 1977). Most of the work on attitudes and attitudinal change in psychology during this time was based upon theories of cognitive consistency, including cognitive dissonance, congruity, and balance theory (see Abelson et al. 1968; Festinger 1957; McGuire 1969; Oskamp 1977). The assump¬tion behind cognitive consistency is that individuals make sense of the world by relying on key beliefs and strive to maintain consistency between their beliefs. Under cognitive consistency, individuals maintain coherent belief Systems and attempt to avoid acquiring information that is inconsistent or incompatible with their beliefs, especially their most central beliefs. In other words, "individuals do not merely sub-scribe to random collections of beliefs but rather they maintain coherent Systems of beliefs which arc internally consistent" (Bem 1970, 13).
The study of the impact of propaganda and communications on individuals reinforced research on cognitive consistency (McGuire 1969; Oskamp 1977; Sears and Whitney 1973). The literature on persuasive communications at the time indicated that most individuals are indifferent to persuasive appeals, especially political propaganda, and when they are attuned they tend to be surrounded by people and communications with which they sympathize. In other words, incoming information typically gets interpreted in accordance with an individual's existing central beliefs and predispositions.

A Cognitive Approach to the Study of Foreign Policy 53
The prevalence of cognitive consistency, and the tendency of most communications to reinforce an individual's belief System, found in the first generation of political psychological research provided the foundation for many studies of foreign policy decision making since the 1960s. This psychological literature and its relevance for the study of foreign policy were brought together by Robert Jervis ( 1976) in Perception and Misperception in International Politics.* jervis provided a significant service to the advancement of a cognitive approach to foreign policy through his rich survey of the processes of perception for foreign policymakers, including a discussion of how cognitive consistency affects decision makers, how decision makers learn from history, how attitudes change, and an analysis of common patterns of misper¬ception among policymakers. Not only did he illustrate the relevance of a cognitive approach for foreign policy and international relations, he also used such a perspec¬tive to critique what he saw as simplistic assumptions of both deterrence theory and the spiral model of state interaction in world politics. As Jervis concluded, "lt is often impossible to explain crucial decisions and policies without reference to the decision-makers' beliefs about the world and their images of others" ( 1976, 28).
Second-Generation Scholarship
Beginning in the 1970s, psychology underwent what has been referred to as a "cogni-tive revolution" in the study of attitudes and how individuals process information. The revolution involved a different conception of the individual and his or her inter¬action with the environment: away from a "passive agent who merely responds to environmental stimuli" to a concept of the individual as more likely to selectively respond to and actively shape his or her environment; away from an individual who "strives for consistency" toward an individual who is more likely to act as a "problem solver" in order to make sense of a complex environment involving great uncertainty (George 1980b, 56; see also Fiske and Taylor 1991, 9-14; Lau and Sears 1986c). Responding to developments in cognitive theory, psychological research increasingly moved beyond the study of cognitive consistency to a more sophisticated study of attitudes and information processing often referred to as social cognition theory and schema theory.4
Social cognition theory and schema theory assume that individuals are "cognitive misers" who tend to rely on existing beliefs and schema- that is, mental constructs that represent different clumps of knowledge about various facets of the environ¬ment—for interpreting information. Although schema necessarily simplify and structure the external environment, they are the basis from which individuals are able to organize new information, use their memory, and ineligibly make sense of the world around them. The more complex and uncertain the environment, the more likely individuals will rely on simple schema and cognitive heuristics—shortcuts in information processing to make sense of the world and the situation at hand (Fiske and Taylor 1991; Lau and Sears 1986b; Milburn 1991 ).5
The social cognition theory and schema theory approach to attitudes and beliefs builds upon and extends the previous work on cognitive consistency but is based on a more complex and sophisticated understanding of the nature of atti-tudes and how information is processed by the mind. Social cognition theory and schema theory emphasize the dominant role of preexisting beliefs in interpreting

54 Second-Generation Foreign Policy Analysis
new information, much like cognitive consistency theory. Yet, where the theory of cognitive consistency assumes the existence of a belief system with a high degree of coherence and interdependence between beliefs that are extremely resistant to change, a social cognition perspective depicts individual belief systems as much more fragmented internally, with different beliefs or schema being invoked under different situations for making sense of the environment. This suggests a greater likelihood that some beliefs may change over time. From this perspective, although the beliefs held by an individual may appear incompatible and contradictory to an outside observer, the overall belief System is likely to make sense to the individual of concern, suggesting a rather complex cognitive process. Although more sophisticated than cognitive consistency theory about the workings of the human mind, cognitive psychology and social cognition nevertheless continue to represent an alternative to the rational actor perspective.
Two major works published near the beginning of the so-called cognitive revolution anticipated many of these developments in the area of social cognition theory and schema theory, integrating them into the study of foreign policy. One, already discussed, was Perception and Misperception in International Politics by Robert Jervis ( 1976i. The other was another landmark in cognitive approaches to the study of foreign policy, entitled The Cybernetic Theory of Decision by john Steinbruner (1974). Actually, Steinbruner's title was somewhat of a misnomer; his book really developed a "cognitive theory of decision" that made three significant contributions. First, Steinbruner demonstrated that the rational actor model (what he called the analytic paradigm) has great difficulty in explaining governmental decision making and per-formance, especially under real-world conditions of complexity and uncertainty. Second, he synthesized a large body of knowledge from the fields of cybernetics and cognitive theory in developing a cognitive process model as an alternative to the rational actor model. He argued that relatively simple decisions could be explained b\' a simple cybernetic process of "satisficing" and "incrementalism" commonly found in bureaucracy. However, to explain more complex decisions one had to turn to a cognitive process model (or cognitive paradigm; in order to understand the regularities of how the mind deals with uncertainty. According to this model, the mind "constantly struggles to impose clear, coherent meaning on events" (Steinbruner 1974, 112), relying on "cognitive structures" involving memory, centrality, consistency, simplicity, and stability. Finally, Steinbruner not only demonstrated how individuals arrive at decisions under the cognitive process model but discussed how the model applies within a larger collective decision-making context as occurs within the government: "In essence, it is cognitive operations of the human mind working in interaction with the organizational structure of the government which set work-able limits on highly diffuse decision problems" (Steinbruner 1974, 14). Ultimately, what Steinbruner did was present a coherent and powerful theoretical foundation for the cognitive study of foreign policy.
■ Major Foreign Policy Studies
The evolution of the psychological study of attitudes provided the foundation for a number of research programs and studies employing a cognitive approach to the

A Cognitive Approach to the Study of Foreign Policy 55
study of foreign policy, especially since the first generation of scholarship. The most significant approaches and studies over time will be highlighted here; these both reflect the evolution of the cognitive study of foreign policy and have contributed to its development.6 These works vary in terms of conceptual orientation, substantive focus, and research strategy, as discussed subsequently.
Images of the Enemy and Mirror Images
The earliest and most popular studies employing a cognitive approach in a system-atic fashion have focused on general images held by political elites, especially images of the "other" and self-images. Such a cognitive approach to foreign policy has been a particularly powerful source for better understanding international conflict and war.
One of the most powerful studies from the first generation of scholarship has been Ole Holsti's (1962, 1967; Finlay, Holsti, and Fagen 1967, 25-96) work on the imagf of the enemy.' Holsti believed that the concept of the enemy helped to explain and sustain international conflict over time. In order to better understand U.S.-Soviet hostility and the evolution of the cold war, he employed a case study to examine the image of the Soviet Union held by former Secretary of State John Foster Dulles. Holsti (1967, 17) assumed that hostile images of the enemy, represented by Dulles's image of the Soviet Union, tended to be self-perpetuating, which could be explained by the psychological literature on cognitive dynamics.
Dulles's image of the Soviet Union was derived from a content analysis of his public statements "supplemented by contemporary newspapers, secondary sources, questionnaires sent to a number of Dulles's closest associates, and memoirs written by those who worked closely with him" (Holsti 1967, 18). Holsti found that Dulles's extremely hostile image of the Soviet Union was very rigid and resistant to change regardless of changes in Soviet behavior. This was consistent with the psychological literature on cognitive dynamics based on the nature of attitudes, cognitive consis-tency, and persuasive communications.8 Dulles resisted new information inconsistent with his image of the Soviet Union by engaging in a variety of psychological processes: discrediting the information, searehing for other consistent information, reinterpret-ing the information, differentiating between different aspects of the information, engaging in wishful thinking, and refusing to think about it. According to Holsti, the self-perpetuation and resistance to change found in Dulles's image of the Soviet Union was based on an "inherent bad faith" model of the enemy: "As long as the Soviet Union remained a closed society ruled by Communists, it represented the antithesis of values at the core of Dulles' belief System. Furthermore, information that might challenge the inherent-bad-faith model of the Soviet Union generally came from the Soviets themselves a low-credibility source and was often ambiguous enough to accommodate more than one interpretation" ( 1967, 24).
The study of mirror images is consistent with the study of the image of the enemy but involves the images held by both parties in a hostile relationship. The con¬cept of mirror images refers to the fact that each party holds an image that is diametrically "opposite" the other: each party has a positive and benevolent self-image while holding a negative and malevolent image of the enemy. Ralph White (1966, 1968) popularized the notion of mirror images in Nobody Wanted War: Misperception in Vietnam and Other Wars.9 Analyzing the two World Wars and focusing on the

56 Second-Generation Foreign Policy Analysis
Vietnam War, White discussed how each party in a conflict situation holds a "diaboli-cal enemy-image" and a "virile and moral self-image" that become the source of selective inattention, absence of empathy (for the other), and military overconfidence. Although the resuit of diverse social and psychological sources, such "black-and-white" thinking leads to escalation and war. As White concluded, "The ethnocentric black-and-white picture is a transcultural, almost universal phenomenon, the details of which vary greatly from nation to nation, but the essence of which remains much the same. ... In view of this it is not surprising that there is a 'mirror image' quality in the reality-worlds of combatants" (1968, 275).
The Operational Code
One of the most popular approaches for describing a political leader’s belief system has been the development of what has been called the operational code, based upon Nathan Leites's (1951, 1953) study of communism in the Soviet Union. However, it was not until Alexander George (1969) refined the concept that the operational code became popular as a research program in the area of foreign policy involving numerous scholars. The operational code is based on the concept of cognitive consistency, for it assumes an overall coherent and interconnected set of beliefs about the nature of political life.1"
The operational code approach consists of two fundamental types of beliefs— philosophical and instrumental beliefs organized around ten questions. Philosophical beliefs refer to assumptions and premises regarding the fundamental nature of politics, political conflict and one's opponents, the future, and historical development. Instrumental beliefs refer to beliefs about strategy and tactics concerning political action, risk taking, timing, and means for advancing one's interests. The operational code serves as a guide to political decision making: the individual's philosophical beliefs help diagnose the definition of the situation, while the instrumental beliefs affect the likely choice of action (see George 1979b).
One of the more interesting uses of the operational code has been Stephen Walker's ( 1977) study of former National Security Adviser and Secretary of State Henry Kissinger.11 Walker examined the interface between Henry Kissinger's opera-tional code and his bargaining behavior during the Vietnam conflict. The study was particularly informative in depicting how Kissinger saw the world, including the role of great powers, and his understanding of the use of force and negotiations in inter-acting with an adversary. Walker also found a close relationship between Kissinger's operational code (through his academic writings) and his conduct of the Vietnam negotiations (through governmental and secondary sources). While the operational code has been a powerful cognitive approach for determining the "content" of politi¬cal leaders' foreign policy beliefs, Walker's study represents an early landmark in actually examining the relationship between foreign policy beliefs and behavior.
Cognitive Mapping
In the early 1970s an attempt was made to examine the beliefs of policymakers in a more specific and rigorous way through the use of a cognitive map. A cognitive map refers to a specific set of beliefs, and their interconnectedness, as they exist in the






A Cognitive Approach to the Study of Foreign Policy 57
mind of a decision maker concerning some aspect of the environment. It is a "specific way of representing a person's assertions about some limited domain, such as a policy problem" (Axelrod 1976, 55). The foundation of cognitive mapping is based upon the early formulations of Robert Axelrod (1972, 1973) and Michael Shapiro and G. Matthew Bonham (1973), derived from the literature on cognitive consistency and schema theory.
From the work of Shapiro and Bonham (see, e.g„ Bonham 1976), a decision maker goes through five basic cognitive steps when he or she perceives a new interna¬tional situation that may require a policy response: initial amplification (of relevant beliefs), search for antecedents (explaining the events), search for consequences {of likely future behavior of actors involved), search for policy alternatives, and policy choice. Given the focus on specifie mental constructs and causal beliefs over different information-processing stages, the cognitive mapping approach allows for a mathematical modeling and computer simulation of the mind. Such an approach has been particularly useful for examining specific decisions and analyzing the cognitive complexity of decision makers in a variety of contexts, as demonstrated by the collection of studies in Structure of Decision: The Cognitive Maps of Political Elites, edited by Robert Axelrod (1976b).12
Attribution Theory and International Conflict
Since the 1950s psychologists have been interested in understanding how people per-ceive the causes of human behavior, resulting in the generation of a body of knowledge known as attribution theory. It was not until the late 1960s and 1970s, how-ever, that attribution theory grew in prominence and began to influence the study of political psychology. Attribution theorists have attempted to identify the various types of attribution errors and biases that people commonly make. For example, the "fundamental attribution error" is the tendency to attribute (explain) our own behavior in terms of external or situational causes. This would be akin to someone saying, "The situation forced me to do what I did." However, this error also entails attributing the behavior of others to internal causes. In this case the same person might say, "The other person acted as he or she did because he or she is mean-spirited." This is further affected by the "self-serving bias," which is the tendency to take credit for good or pos-itive behaviors but to deny responsibility for bad or negative behaviors, and the "halo effect," which is the tendency to see friends in a positive light and enemies in a negative light (Baron and Byrne 1981; Freedman, Carlsmith, and Sears 1970; Jones 1972; Nisbett and Ross 1980).
Attribution theory has not received as much attention from scholars who take a cognitive approach to the study of foreign policy as have other psychological con-cepts discussed previously. Nevertheless, an important study employing this perspective, along with consistency theory, was conducted by Daniel Heradstveit (1979) in The Arab-Israeli Conflict: Psychological Obstacles to Peace.13 Based on inter¬views of political activists conducted in Egypt, Israel, Jordan, Lebanon, and Syria, Heradstveit found strong support for the existence of the fundamental attribution error in explaining the behaviors of participants in the Arab-Israeli conflict. In fact, he found that the inference of internal causes of the opponent's behavior is enhanced when the observer dislikes the actor who performs the blameworthy act, which is

58 Second-Generation Foreign Policy Analysis
consistent with the self-serving bias and the halo effect. As stated by Heradstveit, "If I have a devil-image of the opponent and the opponent behaves in an indisputably friendly way, I can still maintain my beliefs about the opponent by explaining his friendly behavior as caused by environmental influences and constraints. His disposi-tion to act in an unfriendly way remains the same, but certain characteristics of the situation have forced him to be temporarily friendly. In other words, the opponent is not given responsibility (credit) for what he is doing" ( 1979, 74). Such research based on attribution theory contributes to a further understanding of the rise and rigidity of the image of the enemy and mirror images during times of conflict.
More Recent Studies
Beginning in the 1980s a number of individuals built upon earlier work in an effort to further extend knowledge and understanding of the role of cognitive approaches to the study of foreign policy. Much of this work emphasized the need to synthesize available knowledge and to be more eclectic in the study of foreign policy, reflecting the growing diversity and interdisciplinary nature of the study of international rela¬tions. Four studies that reflect such concerns and have contributed to them are highlighted here.14
One important study has been Origins of Containment: A Psychological Explanation, by Deborah Larson (1985), which employs cognitive social psychology to explain the origins of the cold war. Larson specifically attempted to explain the changes that transpired in the beliefs of key U.S. policymakers W. Averell Harriman, Harry S. Truman, James F. Byrnes, and Dean Acheson toward the Soviet Union from 1944 to 1947 in terms of five rival theories of attitude change: the Hovland attitude change approach (involving persuasive communications), cognitive dissonance theory (the classic cognitive consistency approach), attribution theory, self-perception theory (which makes the unique argument that individual beliefs are derived from observing one's own behavior), and schema theory. Through an analy-sis of the archival evidence, Larson concluded that Harriman, Truman, and Byrnes, unlike Acheson, did not have coherent, structured belief Systems; instead, their inconsistent and contradictory beliefs resulted in each eventually developing an enemy image of the Soviet Union in his own way. In tact, since "U.S. policymakers used different types of cognitive processes to interpret information about Soviet behavior...No one theory of attitude change explains the origins of American leaders’ Cold War belief System" ( Larson 1985, 342). Instead, self-perception, schema, and attribution theories were all useful in explaining the individual policymakers' beliefs about the Soviet Union over time. Accordingly, the Larson study is particularly powerful in demonstrating the complexity and messiness of individual belief Systems and cognition, especially during times of considerable uncertainty.
Another important cognitive study is Perceptions and Behavior in Soviet Foreign Policy, by Richard Herrmann (1985; see also 1984, 1988). Building on the work of Richard Cottam ( 1977), this study addressed "the problem of inferring the perceptions and motives of a nation's leaders" in order to better understand Soviet foreign policy (Herrmann 1985, xix). Drawing on attribution and cognitive consistency theory, Herrmann developed a framework that presents three common types of perceptions policymakers have of rival countries and their likely implications for

A Cognitive Approach to the Study of Foreign Policy 59
foreign policy behavior. the "enemy" stereotype that produces a defensive foreign policy pattern; the "degenerate" stereotype that produces an expansionist foreign policy pattern; the "child" stereotype that produces an imperialist foreign policy pattern. Based on an examination of the stereotypical justifications invoked in Soviet speeches, public statements, and foreign policy behavior, Herrmann found that the Soviet image of the United States resembled the classic enemy stereotype in 1967, evolved into a more complex and less stereotypical enemy image in 1972, and then reverted to a more stereotypical enemy image in 1979, although more differen-tiated and less intense than in 1967. Herrmann's cognitive approach and empirical findings contribute to an understanding of the motives underlying foreign policy behavior in general and Soviet foreign policy in particular, especiaily in its implica¬tions for U.S. foreign policy.
Another significant cognitive study of foreign policy produced during the eighties is The Carter Administrations Quest for Global Community: Beliefs and Their Impact on Behavior, by Jerel Rosati (1987; see also 1984, 1990). Given the reports of policy disputes between high-level officials, especiaily Secretary of State Cyrus Vance and National Securitv Adviser Zbigniew Brzezinski, and conflicting interpretations of President Jimmy Carter's foreign policy views during the late seventies, Rosati saw the Carter administration as an excellent case study for applying some of the insights of social and cognitive psychology to advance an understanding of U.S. foreign pol-icy. Employing a content analysis of public statements made by Carter, Vance, and Brzezinski throughout their four years in office, Rosati found that in 1977 the Carter administration initially had an optimistic worldview that was shared by the principal policymakers, that individual differences emerged in 1978 when Brzezinski's image of the Soviet Union hardened. that the administration's image fragmented in 1979 as Carter began to waver between Vance's optimism and Brzezinski's increasing pes-simism, and that by the beginning of 1980 a new consensus emerged based on a pessimistic image of an increasingly unstable world in which Soviet power had to be contained. The stability of Vance's image and the change in Brzezinski's and Carter's images were explained through use of the literature on cognitive consistency, persua¬sive communications, individual personality, the impact of external events, and the role of domestic politics. Ultimately, Rosati found that the greater the consensus in beliefs among individual policymakers, the greater the likelihood that the Carter administration's foreign policy behavior was congruent with those beliefs. Mot only does the study shed light on the nature and significance of U.S. foreign policy during the Carter years, but it also demonstrates the power of a broad cognitive approach for contributing to an understanding of continuity and change in policymakers beliefs over time and their impact on foreign policy behavior.
Finaliy, there is the important study by Yuen Foong Khong( 1992 ) Analogies at War: Korea, Munich, Dien Bien Phu, and the Vietnam Decisions of 1965, which focuses on "how and why policymakers use historical analogies in their foreign policy decision-making" (Khong 1992, . It specifically addresses the question of whether the use of "lessons of history" invoked by policymakers actually shapes policy or pri-marily serves as its justification. Relying on the public record, interviews, and archivai documents, Khong demonstrates that the lessons policymakers drew from Munich, Dien Bien Phu, and, most importantly, the Korean War had a powerful influence on

60 Second-Generation foreign Policy Analysis
the decision-making process relative to Vietnam because they predisposed them toward military intervention. In other words, policymakers use analogies not merely to justify policies but also to perform specific cognitive tasks essential to political decision making: to help define the nature of the situation facing the policymaker, to help assess the stakes, and to provide policy prescriptions. The Khong study con-tributes to an improved understanding of why humans analogies and why policymakers often use analogies poorly due to a common cognitive process.
■ What Do We Know? Where Should We Go?
In surveying the development of political psychology and cognitive approaches to foreign policy, it should be relatively clear that much knowledge has accumulated concerning the role of policymakers' beliefs and images on foreign policy. In this sec¬tion Ibriefly review what has been learned about the cognitive study of foreign policy and discuss some questions that remain unanswered or need to be addressed in the future,
Content of Beliefs
One feature ail cognitive approaches share is that they provide an excellent perspec¬tive for determining the actual contents of the beliefs and images held by individual policymakers. A myriad of cognitive approaches can be drawn upon. For instance, the operational code emphasizes a policymaker's broad beliefs about politics; the image of the "other" focuses on the policymaker's perceptions and beliefs concerning a particular actor in world politics; cognitive mapping tends to examine a policy-maker's foreign policy beliefs toward very specific issues. Whether a broad or narrow perspective is taken, a cognitive approach allows one to examine any aspect of a poli-cymaker's foreign policy beliefs to contribute to a better understanding of foreign policy. This is particularly important because historians and observers of contemporary politics often make judgments and disagree about the contents of policymakers' beliefs. In this respect, a cognitive perspective allows for a more sophisticated foundation to examine the contents of the beliefs of policymakers. In the study of U.S. foreign policy, for example, cognitive approaches have contributed to a better under¬standing of the origins of the cold war (Larson 1985), American perceptions of Soviet communism during the "high" cold war era (Holsti 1967), the American decision to intervene in Vietnam (Khong 1992), the years of détente as represented by the beliefs of Henry Kissinger (Walker 1977), and the rise and décline of a world order approach during the Carter administration (Rosati 1987).
In order to describe foreign policy beliefs and images, it is important to determine the relevant policymakers involved in the formulation of policy. Most scholars who have systematically studied the foreign policy perceptions and beliefs of politi¬cal leaders from a cognitive perspective hâve tended to focus on one key leader, such as Holsti's (1967) study of John Foster Dulles or Walker's (1977) study of Henry Kissinger. Also, a government and sometimes even a Society has been treated as a single actor—that is, the equivalent of a single individual—such as in White's (1968) general discussion of mirror images or Heradstveit's (1979) analysis of the




A Cognitive Approach to the Study of Foreign Policy 61
Arab-Israeli conflict. Although analyzing an individual leader minimizes the empiri-cal problems of identifying the content of beliefs, it considerably simplifies the "collective" nature of most decision making and the differences in beliefs that one might find across individuals (Steinbruner 1974). Some of the more recent work in this area has attempted to be more sophisticated in the cognitive analysis of foreign policy, as demonstrated by Larson's (1985) study of Harriman, Truman, Byrnes, and Acheson within the Truman administration; Rosati's (1987) study of Carter, Brzezinski, and Vance (and Muskie) within the Carter administration; and Khong's (1992) study of Lyndon fohnson and his advisers during the Vietnam policymaking process. More studies need to identify the critical individuals within the policvmak-ing process and examine the contents of their beliefs and the decision-making dynamics between them so as to better understand the formulation and conduct of foreign policy.
Also, most of the work on the contents of beliefs has centered on the policy-makers of the great powers, especially the United States, and has focused on the "high policy" issues of national security and international conflict, a possible legacy of the initial interest by political psychologists in addressing the paramount ques-tions of war and peace.15 Clearly, cognitive studies of images of the enemy and mirror images; crises (see, e.g., Holsti 1990; Lebow 1981 ); deterrenee theory (see, e.g., Jervis, Lebow, and Stein 1985; Snyder 1978); and intelligence warning and sur¬prise attack (see, e.g., Betts 1978; Wohlstetter 1962) have contributed to a better understanding of the dynamics of international conflict and war. At the same time, much more needs to be done in systematically examining the contents of beliefs about "low policy" issues such as international economies and the environment and in studying images of other actors from other parts of the world using a cognitive perspective.16
The Structure of Beliefs
The development of political psychology and the cognitive study of foreign policy has contributed to a better understanding of the structure of beliefs in the minds of individuals in a political setting. Both cognitive consistency theory and schema theory agree that central beliefs are consequential, although they differ about the level of coherence and interconnectedness between beliefs. On the one hand, the literature on cognitive consistency theory emphasizes that individuals acquire beliefs and images that are interconnected and form coherent belief Systems. This was the basis, for example, of some of the earlier research approaches, such as the study of the image of the enemy as represented by Holsti's (1967) examination of lohn Foster Dulles or the study of an individual's general orientation toward political life found in the operational code approach. On the other hand, the literature on social cogni-tion theory and schema theory describes cognitive structures within the minds of individuals as quite complex and messy. From this perspective, people's beliefs tend to be much less coherent, less interconnected, and more contradictory than originally conceived by cognitive consistency theory.
As the more recent studies demonstrate, both cognitive patterns are likely. Larson's (1985) study of the origins of the cold war found that only Acheson had a coherent belief System about world politics. Harriman, Byrnes, and especially

62 Second-Generation Foreign Policy Analysis
Truman had much less coherent beliefs and were more prone to make sense of world developments and Soviet behavior in ways that were consistent with schema, attribu-tion, and self-perception theory. It was only in 1947 that a more coherent belief System crystallized for these policymakers around an enemy image of the Soviet Union and a policy of containment. Likewise, Rosati (1987) found both sets of cog-nitive patterns operating in his study of the Carter administration. Whereas Vance and Brzezinski maintained relatively coherent belief Systems throughout, Carter s beliefs became increasingly inconsistent during 1978 to the point that, by 1979, he was constantly wavering between Vance's more optimistic image and Brzezinski's increasingly pessimistic one, until eventually siding with Brzezinski's interpretation of world politics in 1980.
What can possibly account for these contrasting cognitive structures suggested by cognitive consistentcy and schema theory? It appears that much depends on the individual, on his or her background, personal development, and role. According to studies on the differences between "experts and novices," coherent images of phenom-ena are more likely to be formed if individuals have developed considerable expertise and experience relative to those phenomena. Fragmented, inconsistent, and contra -dictory beliefs are more likely to exist for phenomena about which individuals have given little thought or have acquired little knowledge and experience (see, e.g., Lau and Erber 1985; Lau and Sears 1986c; Milburn 1991), From this perspective it is not surprising that Acheson, Brzezinski, and Vance each an expert in foreign policy with considerable experience developed relatively coherent belief Systems, while Harriman (to a Lesser extent), Byrnes, Truman, and Carter as politicians and nonex-perts were more prone to uncommitted thinking {see Larson 1985, 346).
Such an interpretation is consistent with the emphasis placed bv cognitive theo-rists over the years on the importance of the situation for perception and cognition ( Fiske and Taylor 199] ; Milburn 1991 ; Rokeach 1968). Different situations are likely to trigger different schemas and sets of beliefs.17 This is one of the major conclusions drawn by Herrmann ( 1985) in his work on the role of the enemy, degenerate, and child stereotypes in Soviet foreign policy. And such differences in images are likely to vary among individuals. Although individuals are likely to have little difficulty making sense of situations that are relatively stable and familiar in terms of their beliefs, situa¬tions of great complexity and uncertainty may result in significant cognitive inconsistency and confusion, especially if individuals do not have considerable knowl-edge and well-developed belief Systems for imposing some degree of understanding on the environment. This may help account for why Truman administration policy-makers were vulnerable to such inconsistency in their beliefs and dependent on the mechanisms of schema, attribution, and self-perception theory (Larson 1985)— clearly, the immediate post-World War II years were a time of great uncertainty and flux in global developments, similar to the current collapse of communism and the rise of the so-called post-cold war era. This may also explain why policymakers are prone to rely on historical analogies, as Khong ( 1992) found in his study of Vietnam decision making. Such cognitive responses by policymakers in differing times and set-tings needs to be explored further.18



A Cognitive Approach to the Study of Foreign Policy 63
Continuity and Change in Beliefs
Acognitive approach to foreign policy also contributes to an understanding about the extent to which policymakers' beliefs are likely to change over time. Cognitive theory, as informed by the literature on both cognitive consisteney and schemas, is based on the premise that central beliefs are the most consequential in understanding the process of perception and cognition, As described by Milton Rokeach, "First, not all beliefs are equally important to the individual; beliefs vary along a central-peripheral dimension. Second, the more central a belief, the more it will resist change. Third, the more central the belief changed, the more widespread the repercussions in the rest of the belief System" ( 1968, 3; see also Bem 1970).
Beyond this common core, cognitive consisteney and schema theory differ as to the likelihood and nature of attitudinal change. Cognitive consisteney theory tends to emphasize the overall rigidity of belief Systems due to their interconnectedness and, should change occur, the abrupt and all-encompassing nature of belief System change. As explained by Jervis (1976, 170), "If a person's attitude structure is to be consistent, then incremental changes among interconnected elements cannot be made. Change will be inhibited, but once it occurs, it will corne in large batches. Several elements will change almost simultaneously." A number of studies from different cognitive research programs support such an explanation. Hotsti (1967) found that Dulles's enemy image of the Soviet Union was fundamentally constant over time. Heradstveit (1979) found stability in Arab and israeli images throughout the 1970s, while Bonham, Shapiro, and Trumble (1979), employing cognitive mapping, found no difference in the belief Systems of American policymakers before and after the 1973 Yom Kippur War. Harvey Starr (1984) reviewed the literature on Henry Kissinger's foreign policy beliefs and found considerable stability in his preoffice, officiai, and post-office beliefs. Finally, Ben-Zvi (1978) found that the images of Japan held by American leaders were dramatically altered following the attack on Pearl Harbor and the fighting of the war in the Pacific, as predicted by cognitive consistency theory.
Schema theory, on the other hand, stipulates that beliefs are much more isolated and inconsistent with each other, and hence are less resistant and more open to piecemeal, incremental changes over time. This is what Larson (1985) found in her study of the origins of the cold war: the years 1944 to 1947 represented constant fluc-tuation in the beliefs of Harriman, Byrnes, and Truman, who moved sporadically but inevitably toward an enemy image of the Soviet Union. Likewise Rosati (1987) found that not only did Carter experience changes in his beliefs and increasingly waver between Vance's and Brzezinski's opposing views over a period of two years, Brzezinski also experienced considerable change in his worldview over time: relatively optimistic during 1977, skeptical of Soviet behavior and its implications for world order after the Ethiopian-Somali war in early 1978, and increasingly pessimistic about Soviet expansionism by 1979.
What the various studies suggest is that patterns involving both belief stability and change may in fact occur. According to cognitive consistency theory, the more coherent and interconnected the belief system, the more resistant to change and, should change occur, the more likely that it will be abrupt and profound. At the same time, from the perspective of schema theory, the fragmented and inconsistent nature


64 Second-Generation Foreign Policy Analysis
of most people's beliefs indicates that incremental changes may also occur with some frequency. Both patterns were observed by Larson (1985) and Rosati (1987) in their studies of Truman and Carter administration policymakers over time. In this respect it is important to remember that while consistency theory and schema theory have different implications for belief system change, both emphasize the Level of centrality of beliefs: the more central the belief, the more stable and resistant it is to change.
These two patterns of stability and change are consistent with the earlier dis-cussion of the structure of beliefs, and probably for the very same reasons: much depends on individual background and personal development, the role occupied, and the situation at hand. According to Rosati (1987, 30), "While individuals strive to maintain consistency and the communication of information acts to reinforce one's belief system, images may still undergo change under certain conditions." He specifically highlighted the role of individual personality, external events, and domestic forces to explain the continuity of Vance’s image and the changes in Brzezinski's and Carter's images. As summarized by Rosati (1987, 102-3):
While Carter's personality was open to new information, Brzezinski's insecurity and rigid personality prompted his return to a previously held image. Major events played a critical role in shaping their alternative views: Soviet and Cuban interventionism in Africa was particularly crucial for Brzezinski; the Iranian hostage crisis and the Soviet invasion of Afghanistan for Carter. And finally, the development of a more pessimistic image for Brzezinski and particularly for Carter was reinforced by the changing mood of the country in a more nationalistic and conservative direction.
Larson (1985) also offered a rich explanation, although less systematic in conceptualization, in examining continuity and change in the beliefs of Truman policymakers. Clearly, further study of the patterns of continuity and change in the beliefs of policymakers is warranted.
Beliefs and Their Impact on Behavior
A considerable amount of research has been conducted in analyzing the significance of individual attitudes and beliefs: describing their content, structure, and relative degree of stability over time. All of this work has been based on one fundamental assumption: that beliefs are major sources of behavior and, therefore, explain and predict human action. Unfortunately, this assumption is so embedded in the attitudinal research that the relationship between beliefs and behavior has rarely been tested and demonstrated. The following comment by Fishbein and Aizen (1975, 355) pertains to the present day: "Despite the commitment of the social sciences to the study of human behavior, relatively little research in the attitude area has investigated overt behavior as such." The results of the few studies conducted in social and political psychology have been inconsistent and indicate that the link between individual beliefs and behavior is quite complex and varied (see Barner-Barry and Rosenwein 1985; Calder and Ross 19/3; Deutscher 1973; Oskamp 1977; Wicker 1969).
A similar problem exists with a cognitive approach to foreign policy. Holsti (1976, 52), in his review of the literature, found that the linkage between foreign pol-icy beliefs and behavior has rarely been tested. "It is not uncommon to find in the

A Cognitive Approach to the Study of Foreign Policy 65
Conclusion a statement to the effect that, 'the preceding analysis of X's belief system established its utility for understanding X's political behavior.1 Less often do we find AN explicit and compelling demonstration of why this is the case." George (1979b): Has outlined two basic techniques for determining the impact of policymakers' Beliefs on behavior: the process-tracing procedure and the congruence procedure. Each has been applied, although only in very select cases, to better understand this relationship between beliefs and behavior.
The process tracing procedure "seeks to establish the ways in which the actor's Beliefs influenced his receptivity to and assessment of incoming information about die situation, his definition of the situation, his identification and evaluation of options, as well as, finally, his choice of a course of action" (George 1979b, 113). By Closely examining the process of decision making, process tracing is able to establish Causality. However, it tends to be oriented around a specific case study and requires large amounts of detailed information about the decision-making process not readily available. The major study operating within a process-tracing perspective is Larson's , 1985) study of the Truman administration.19 Relying on archival data, Larson found that although individual beliefs did influence individual behavior, what was more surprising was that policymakers' beliefs about the Soviet Union often flowed from their decisions, as suggested by self-perception theory. These contradictory patterns indicate the complexity of the belief-behavior relationship, which, as suggested earlier, may have been accentuated by the fluidity and uncertainty of the immediate post-World War II era.
The congruence. Procedure examines the level of consistency between the con¬tent of the beliefs and the content of the decision outcome. Since the congruence procedure focuses on the level of association between beliefs and behavior, it cannot determine causality although it requires much less specific information about the decision-making process and allows for broad generalizations over time. The few studies conducted also point to the existence of contradictory patterns. Starr (1984) found that Kissinger's images of the Soviet Union and China were not consistent with American behavior toward the Soviet Union and China. However, when Walker 11975, 1977) examined the relationship between Kissinger's beliefs and his bargaining behavior toward the Soviet Union, the Arab Israeli conflict, and the Vietnam War, he found consistency. Loch Johnson (1977, 85) studied the operational code of Senator Frank Church and found a strong correlation between Church's "beliefs and subsequent voting behavior which reinforces my impression (as participant observer) of his belief-behavior consistency." In Rosati's (1987) comparison of the beliefs of Carter, Brzezinski, and Vance with the administration^ major foreign policy behavior over four years, he found both sets of belief-behavior patterns. He found that the Carter administration’s foreign policy behavior was extremely consistent with its beliefs during 1977 and 1980, in contrast to 1978 and 1979, when the consis¬tency between behavior and beliefs declined considerably. The major factor accounting for these differing patterns was the level of consensus in beliefs among Carter's policymakers—a high level of shared beliefs promoted congruent foreign policy behavior, but little congruency existed when consensus in beliefs broke down.
One of the impressive aspects of the study of Vietnam policymaking by Khong (1992) is that he relies on both procedures—process tracing and congruence to

66 Second-Generation Foreign Policy Analysis
demonstrate how historical analogies affect decisions. Despite such efforts to explain decision-making and foreign policy behavior as a function of the beliefs held by policymakers, this is a topic that needs considerably more exploration and study. Ultimately, as described by George, beliefs "serve as a prism or filter that influences the actor's perception and diagnosis of political situations and that provides norms and standards to guide and channel his choices of action in specific situations. Neither his diagnosis of situations nor his choice of action for dealing with them is rigidly prescribed and determined by these beliefs" (1980b, 45). Although beliefs often impact behavior, many other causal factors also play a role in influencing for¬eign policy behavior.