The ORE of Politics will be available for subscription in late September. Speak to your Oxford representative or contact us to find out more.

Show Summary Details

Page of

PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA, POLITICS ( (c) Oxford University Press USA, 2016. All Rights Reserved. Personal use only; commercial use is strictly prohibited. Please see applicable Privacy Policy and Legal Notice (for details see Privacy Policy).

date: 16 August 2017

Symbolic Politics as International Relations Theory

Summary and Keywords

The symbolic politics theory of ethnic war starts from the insight that most political behavior is not rational but intuitive, driven by “symbolic predispositions” such as ideological beliefs, normative values, and prejudice. The way leaders lead is by using rhetoric not to appeal rationally to followers’ interests but to appeal emotionally to their symbolic predispositions.

According to symbolic politics theory, the path to ethnic conflict begins with group narratives that are hostile to another group. These narratives help to generate hostile and prejudiced symbolic predispositions. If group members perceive a social threat, such as to their group identity or status, they become more likely to join mass movements agitating for a politics of redistribution—discriminating in favor of their own group at the expense of rival groups. If people feel physically threatened, they become more likely to support a politics of protection leading to violent ethnic conflict. These popular attitudes and moods are turned into social movements or military mobilization if aggressive leaders emerge, framing political issues in terms of these threats, and if those leaders are both credible and supported by effective organizations. A series of case studies has demonstrated that this process—from narratives to prejudice and threat perceptions, harnessed by leadership and organization—is what occurred in ten ethnic civil wars, including the former Yugoslavia, Rwanda, Israel-Palestine, and the Philippines. The theory also explains less violent cases such as Gandhi’s nationalist movement in India.

This theory is hypothesized to apply to international war, as the politics of national identity is similar to the politics of ethnic identity. The theory also suggests a way of reconciling realist, liberal, and constructivist accounts of international relations through political psychology and a scientific realist ontology.

Keywords: symbolic politics, political psychology, ethnicity, ethnic conflict, nationalism, war, civil war, scientific realism, empirical international relations theory


Common sense, as well as a raft of studies in social and political psychology, tells us that people are not the value-maximizing rational actors assumed by most theories of international relations. Voters are notoriously uninformed about even basic facts, while elite decision making is driven by the beliefs, ideologies, and values of the decision maker. Yet, because the rational actor assumption is convenient, its use in international relations theory is predominant.

The value of symbolic politics theory is that it is based on an alternative, and more empirically grounded, model of human decision making. In the past, the most prominent symbolic politics analyses concerned American domestic politics, as in the work of Murray Edelman and David O. Sears. Starting in 2001, however, symbolic politics theory has been a superior alternative for explaining ethnic conflict, most notably ethnic civil war.

This topic has a twofold relevance to international relations theory. First, civil wars outnumber international wars by a large and increasing margin, and most civil wars are partially, if not primarily, ethnic in nature, so the symbolic politics theory of ethnic war has the status of a middle-range theory of a key international relations topic. Second, attempts have already been made to expand symbolic politics into a theory of international relations. This nascent theory is a rare attempt to turn political psychology into the basis of a general theory of international relations, instead of merely serving as an adjunct to other theories. It is also epistemologically and ontologically distinctive, relying on a scientific realist rather than a positivist or poststructural philosophy of science (see Jackson, 2010; Wight, 2006). Symbolic politics therefore deserves attention as a potentially fruitful new departure that already has a solid track record in its first area of application.

Contemporary symbolic politics theory (Kaufman, 2015) is based on a fusion of findings from political psychology and social mobilization theory. Based on findings in psychology, it explains preferences as being the result of the interaction between symbolic predispositions (essentially, emotionally charged attitudes) and emotions themselves, especially feelings of threat. Based on findings in sociology, the theory also recognizes the necessity of social organization and leadership for translating attitudes into social action. Ethnic civil war is most likely to occur when at least one group is prejudiced against and feels threatened by another, and when both groups can generate the organization and leadership needed for them to mobilize and fight. The direct application to international relations begins with the insight that the same dynamics apply to nationalism, and can therefore explain the decision of a nation-state to mobilize and fight.

Origins of Symbolic Politics Theory

The first seminal statement of symbolic politics theory was Murray Edelman’s (1971) Politics as Symbolic Action. Edelman (1971, p. 4) pithily states his main thesis early on: “The significant ‘outputs’ of political activities are not particular public policies labeled as political goals, but rather the creation of political followings and supports: i.e., the evocation of arousal or quiescence in mass publics.” Myths and symbols play a critical role in such politics. Edelman (1971, pp. 14–15) explains: “The word ‘myth’ signifies a belief held in common by a large group of people that gives events and actions a particular meaning.” Such myths usually define an enemy who is plotting and must be opposed, or a “savior-hero … who is to be followed and obeyed.”

An example of such politics is a grandstanding congressional hearing, the main purpose of which is to allow committee members to denounce some proclaimed villain and thereby to cast themselves as heroes in a drama. The more epic or mythical the scale of the drama, the more political support ensues. Legislative action to fix the problem is neither the purpose nor, in most cases, the result.

The idea of “meaning” is central to Edelman’s (1971, p. 31) logic: “Meaning is basically different from information and incompatible with it.” Indeed, “information destroys meaning”; it is either redundant or contradictory to the myth, which works by supplying a frame for an issue. The Vietnam War, for example, was framed variously as being “about” either communist aggression or napalming children (Edelman, 1971, p. 69). Edelman went on to fuse these anthropological insights with psychological ones, generating what would later be called a “psychocultural” approach (Ross, 2007). For example, he notes, “the personification of threat is psychologically useful to provide a target of action” (Edelman, 1971, pp. 30, 62).

This thinking naturally lends itself to ethnic conflict if one applies what Anthony Smith later called the “ethnosymbolist” understanding of ethnicity (Smith, 2009). Smith asserts that any ethnic identity is defined by a “myth-symbol complex,” a set of mythified (i.e., meaning-laden) narratives and symbols that define who is in the group and what group membership means, identifies villains, heroes, and enemies, and so on. In this view, ethnic politics is about manipulating myths and symbols of ethnicity to enable political leaders to gain support from members of their ethnic group.

The second wave of symbolic politics theorizing and analysis pushed aside Edelman’s anthropological concern with myth and meaning but kept and deepened his emphasis on the priority of feelings over information. Epitomized in the seminal Something for Nothing (Sears & Citrin, 1982), this school of thought particularly aimed to trace the impact of racial prejudice in America’s post–civil rights era.

As the political psychologist David Sears (2001, p. 16) later explained, his symbolic politics theory begins with the presumption that “people acquire stable affective responses to particular symbols [i.e., symbolic predispositions] through a process of classical conditioning.” He showed that it was these symbolic predispositions, not material self-interest, that determined public opinion on issues and political candidates. The most important of these symbolic predispositions are party identification, ideological beliefs, and racial prejudice. One early study (Sears et al., 1980, p. 678), for example, found that symbolic attitudes were five times more powerful than measures of self-interest in explaining issue voting. Symbolic predispositions even influence people’s factual beliefs, with partisans tending to get even basic economic facts wrong if those facts reflect unfavorably on their preferred party (Bartels, 2002).

In Something for Nothing, Sears and Citrin (1982) explored the political psychology of California’s 1978 tax revolt—the prelude to the nationwide “Reagan revolution” two years later. They found that two ideologically based symbolic predispositions (feelings about government and about “welfare’), along with a measure of prejudice, were the most powerful factors influencing opinions about whether to cut property taxes. Strikingly, measures of self-interest mattered much less: being a homeowner (and thus a property taxpayer) had little effect, and having a child in school (which was mostly funded by property taxes) made no difference at all. Instead, the logic behind the movement was racial resentment: those who suspected that their tax money was being wasted on undeserving black welfare recipients tended to support the tax cut. Compatibly with Edelman’s notion of meaning crowding out information, the facts did not matter: property taxes went primarily to schools, not “welfare”; “welfare” was paid for by state taxes, not local property taxes; most welfare recipients were not black. But none of it mattered. Voters chose to “teach the government a lesson” and express their racial resentment.

Since the line between racial difference and ethnic difference is thin, Sears’s concept of symbolic politics is also a theory of ethnic politics, demonstrating that what drives ethnic politics are symbolic predispositions, especially prejudice. It is not, however, a complete explanation of the range of possible outcomes. Going from these insights to an explanation of events like the Rwanda genocide obviously requires more theoretical work.

The Symbolic Politics of Ethnic Conflict

The first statement of the symbolic politics theory of ethnic conflict was in Kaufman’s (1996) account of ethnic conflict in 1990s Moldova. The main theoretical roots of the argument are in social mobilization theory, Horowitz’s (1985) hypothesis about the role of fears of ethnic extinction and the spiral model of international conflict (Kaufman, 1996, pp. 111–112). In this account, escalation to ethnic war starts with an atmosphere of ethnic hostility and fear of group extinction, which gives political leaders an incentive to engage in “ethnic outbidding.” Such outbidding generates threats to the other group, resulting in a security dilemma spiral.

These factors are summed up in three processes that lead to ethnic war—mass hostility, elite outbidding, and a security dilemma—and necessary preconditions for each are identified. Most importantly, “Mass hostility requires reasons for both rational dissatisfaction and emotional heat,” as of course does fear (Kaufman, 1996, pp. 113–114).

There is a distinction between mass-led and elite-led conflict processes (Kaufman, 1996, pp. 115–116). Mass-led conflicts begin with grassroots mobilization motivated by ethnic hostility and fear, which lead to the rise of extremist leaders and a security dilemma. In elite-led processes, the order is reversed: incumbent leaders use state resources to whip up mass hostility and fear to create a security dilemma. The concept is thus that of a recursive spiral process in which hostility and fear, outbidding and mutual insecurity, all mutually reinforce each other. The order in which these processes occur is less important than the positive feedback interaction they generate.

The symbolist research program is characterized by its fierce opposition to the theoretical assumption of individual rational action:

Contrary to the claims of rational choice theorists, rational grievances alone … are not sufficient to motivate ethnic war … Donald Horowitz has shown that the [necessary] emotional heat comes from a fear of group extinction which is usually “anxiety-laden” or exaggerated by emotion. This fear turns stereotypes into antipathies, making people feel hostile enough to risk their skins in battle, wreck their economic infrastructure, and put aside basic humanitarian values.

(Kaufman, 1996, p. 112)

By the time Modern Hatreds (2001) appeared, Kaufman had discovered Edelman’s work and found in it an overarching logic that could draw together most of these ideas. Combining Edelman’s ideas with later social psychological work, Modern Hatreds asserts: “The core assumption of symbolic choice theory is therefore: people choose by responding to the most emotionally potent symbol evoked. Those symbols, in turn, “get their meaning from emotionally laden myths” (Edelman, cited in Kaufman, 2001, p. 28, emphasis in original).

In the new version of the theory, the necessary conditions have been reduced to three: myths justifying hostility, ethnic fears, and opportunity to mobilize and fight. The processes that then lead to ethnic war are essentially the same as those identified previously: mass hostility, chauvinist mobilization (including elite outbidding and organization), and a security dilemma. These processes are still conceptualized as mutually reinforcing in a positive feedback loop. The theoretical development concludes by identifying relevant conflict resolution tools: peacebuilding efforts that replace hostile myths with benign ones; elite pacts to forestall chauvinist outbidding; and reassurance policies to ameliorate security dilemmas.

Reflecting its debt to Edelman’s anthropological approach, the resulting logic is essentially constructivist. Psychological concepts such as prejudice or stereotypes fade into the background—mostly because of lack of evidence—so the focus is on discourse: preexisting hostile myths provide the symbols that chauvinist elites can use to provoke hostility.

The most recent statement of the theory, presented in Nationalist Passions (Kaufman, 2015), brings psychology back in, based on a tacit agreement with Shannon and Kowert (2012) that political psychology and constructivism are “ideational allies.” The emphasis is on Sears’s concept of symbolic predispositions, understood as measurable attitudinal variables, thus returning to the 1996 conception of a causal chain from hostile myths to negative stereotypes and then to mass hostility. Drawing on Duckitt’s (2003) definition of prejudice as including negative stereotypes, negative feelings, and a predisposition to act in a negative way, this new version shows how the logic of the 1996 article dovetails with Sears’s thesis that symbolic predispositions result from classical conditioning. In the case of prejudice, the conditioning is accomplished by the repetition of well-known ethnic narratives. This is the theory’s first hypothesis: “The more hostile a group’s narratives are toward another group, the more prejudiced members of the first group are likely to be” (Kaufman, 2015, p. 62).

The discussion of “narratives” reflects a terminological shift away from the previous references to “myths.” “Myth” is actually the more technically precise term, referring to a specific kind of meaning-making narrative, which in this case defines an ethnic identity. However, it is widely misunderstood to imply that the narrative in question is false, and it can therefore be offensive to those whose narratives are under discussion. Therefore, following the lead of Ross (2007), later articulations of the theory discuss ethnic “narratives” instead of “myths.”

Nationalist Passions begins with findings in experimental psychology and neuroscience showing that the theory of symbolic choice is a more accurate account of how the human brain actually works than is the theory of rational choice. According to the “dual systems” model of mental processing, the human brain uses at least two different decision-making processes based on different neural systems: the intuitive system and the deliberative system (Evans & Frankish, 2009, p. 15), with the intuitive system being much more powerful (Haidt, 2012, p. 46). Furthermore, what drives intuitive decision making on a routine basis are, essentially, symbolic predispositions such as prejudice and ideological beliefs. The most likely cause of a shift out of routine decision making is a perceived threat, which typically just shifts behavioral tendencies toward a different set of symbolic predispositions (e.g., fight-or-flight responses).

National Passions then identifies four starting propositions based on well-established findings in social and behavioral sciences, which generate six hypotheses. First, again, is the existence of symbolic predispositions. Prejudice, the most important symbolic predisposition for ethnic conflict, is taken to be a variable that varies in intensity and targeted outgroup, both across time and across populations. The first hypothesis, restating Sears’s account of the origins of symbolic predispositions, asserts: “The more hostile a group’s narratives are toward another group, the more prejudiced members of the first group are likely to be (Kaufman, 2015, p. 62). Duckitt’s definition of prejudice then yields the theory’s third hypothesis: “The more prejudiced group members are, the more they will support hostile and discriminatory actions toward the disliked group” (2003. p. 559).

Prejudice theory generates the theory’s second hypothesis: “the more prejudiced group members are, the more likely they are to perceive a threat from the disliked group.” Furthermore, if people face a “perceived social threat—whether a threat to power, wealth, values, or status” (p. 45)—this feeling of threat “increases hostility and prejudice” (Kaufman, 2015, p. 52, emphasis added). Yet another set of findings comes from studies in “terror management theory,” which show that priming people to think about their own mortality induces them to become more nationalistic, intolerant, and punitive toward outgroups (Kaufman, 2015, p. 46; cf. Greenberg et al., 1990; Cuillier et al., 2010). These findings add up to the fourth hypothesis: “Stronger threat perceptions lead to more support for hostile action” (Kaufman, 2015, p. 62). Subhypotheses specify that (nonoverwhelming) physical threats spur support for violent and provocative action, while social threats spur support for discriminatory action.

The third and fourth propositions are that the transition from individual inclination to sustained social action requires social organization and leadership—both necessary conditions for ethnic mobilization. Returning to the core notion of symbolic politics as involving elite appeals to emotive symbols, Nationalist Passions draws on studies of framing to generate hypotheses about the circumstances under which different kinds of symbolic appeals are likely to work. “Briefly, leaders lead by framing, and followers follow if the frame resonates with them” (2015, p. 49).

Summarizing one strand of framing research, the book states: “the resonance of a frame depends on the credibility of the leader and of the frame; on the frame’s ‘salience’ …; and on its ‘narrative fidelity’,” or consonance with the audience’s symbolic predispositions” (Kaufman, 2015, pp. 49–50; cf. Benford & Snow, 2000). In ethnic conflicts, the credibility and salience of the frame depend primarily on threat, suggesting an addendum to the fourth hypothesis: stronger threat perceptions lead to more resonance for more assertive frames. The next (fifth) hypothesis focuses on frame effects, stating: “the use of assertive frames by credible leaders increases the likelihood of ethnic mobilization.” Furthermore, if these frames refer to physical threat and are resonant, the likelihood and probable scale of ethnic violence increase.

The overall message of these propositions is that narratives, symbolic predispositions, frames, and threat perceptions all interact, primarily through emotions. In the long run, the reiteration of hostile narratives generates hostile, emotionally charged symbolic predispositions—that is, predispositions to feel hostility when presented with anything that symbolizes the outgroup. In the short run, however, the causality runs the other way: preexisting symbolic predispositions determine which frames (that is, which narratives leaders deploy) will resonate with the audience. Hostile narratives and symbolic predispositions make it more likely that ambiguous action by the other group will be seen as a threat, causing threat frames to resonate. If, however, the outgroup poses an obvious threat, for example, by launching an attack, the resulting feelings of threat are likely to generate hostile narratives and symbolic predispositions.

The final proposition of the theory is that organization is also necessary for sustained social action. The resulting hypothesis is: “the stronger the organizational network supporting an assertive leader, the greater the degree of social mobilization (violent or nonviolent) that leader can achieve” (Kaufman, 2015, p. 63).

The tenor of ethnic politics in a multiethnic society depends on the interaction of these four factors— symbolic predispositions, threat, leadership, and organization—with changes in perceived threat working as the key switching mechanism among them in the short term (Kaufman, 2015, p. 19 and passim). Briefly, an overwhelming physical threat generates a politics of submission and a repressive ethnic peace. A manageable physical threat empowers aggressive leaders by making threat frames resonant and so can result in a politics of protection that may lead to ethnic violence or war. A perceived social threat makes “injustice” frames resonant, likely leading to contentious mobilization and a politics of redistribution. Finally, if ethnic threats are minimal or absent, threat frames will not resonate, so the likely result is an inclusive ethnic peace featuring a low-conflict politics of distribution.

As the other hypotheses suggest, however, threat alone does not determine the outcome. Any degree of mobilization requires adequate organization and a credible leadership. High levels of prejudice may make relatively mild threats seem larger, increasing the degree of conflict, but only if credible leaders adopt the more aggressive frame. Leaders can also attempt to swim against an aggressive tide, appealing for calm instead of encouraging feelings of threat, but they are vulnerable to being outflanked by fear-mongering chauvinist rivals. The outcome of such a contest turns in part on the credibility of the rival leaders. Finally, low levels of prejudice will tend to stabilize a low-conflict politics of distribution, but such a politics is likely to be fragile if prejudice is high.

Symbolic Politics as International Relations TheoryClick to view larger

Figure 1. Reprinted from Nationalist Passions, by Stuart J. Kaufman. Copyright © 2015 by Cornell University. Used by permission of the publisher, Cornell University Press.

The path to conflict (or not) is thus conceived as a series of steps, as illustrated in the accompanying figure. The thick, downward-pointing arrows in the figure indicate the main process leading to ethnic mobilization and conflict; the thinner, upward-pointing arrows identify long-term feedback effects. In the first step (hypothesis 1), narratives shape symbolic predispositions in the population. In the second step, those symbolic predispositions interact with other groups’ behavior and with the frames generated by leaders to shape perceptions of threat (hypothesis 2). Symbolic predispositions, threats, and frames then interact in the third step to determine the degree of public support for mobilization (hypotheses 3 and 4). The final step is the jump from attitudes to action: leaders’ frames, the degree of public support, and the level of organization interact to determine the level of ethnic mobilization (hypotheses 5 and 6).

As a brute theory of ethnic war, the theory is fairly simple. Hostile narratives, prejudices, and other symbolic predispositions s, perceived threats, hostile framing by credible leaders, and strong ethnic organizations are all necessary conditions for war, needing to be present on at least one side if war is to occur. The main process that generates war is symbolic politics, with leaders using threat frames to appeal to hostile symbolic predispositions to justify the fighting. The intensity of the conflict typically reflects the intensity of the hostile symbolic predispositions and feelings of threat.

Evidence and Criticism

The evidence that has been marshaled so far in support of symbolic politics theory consists mostly of a dozen case studies of ethnic conflicts in Asia, Africa, and southeastern Europe (Kaufman, 2001, 2009, 2015). These include ten cases of civil war (including one study of a civil war settlement) and two cases of relative ethnic peace; there are also four briefer looks at low-violence cases. Unsurprisingly, the author claims victory: all of the cases of war show evidence of the processes hypothesized to lead to war, and nonwar cases all seem to look, as hypothesized, like explicable cases of the politics either of submission, distribution, or redistribution.

A meta-analysis of the 16 cases and mini-cases shows that the first hypothesis, on the role of hostile narratives, performs almost perfectly. In virtually every case of ethnic war, mainstream historical narratives on at least one side identified the group on the other side as an enemy. The main Armenian narrative, for example, identified Azerbaijanis with Turks and then identified Turks as having been a genocidal enemy of Armenians for centuries (Kaufman, 2001). Similarly, the dominant Rwandan narrative identified the minority Tutsis as descendants of foreign conquerors who wanted only to renew the subjugation of the majority Hutus (Kaufman, 2015); and the Palestinian nationalist narrative is defined in terms of opposition to Zionism (Kaufman, 2009). In cases of ethnic peace, in contrast, national narratives were either aimed against external enemies (among Kazakhs, Tanzanians, and Macedonians) or were inchoate (Moldova’s Gagauz). The only partial exception is South Ossetia, where the hostile narratives all seem to have been published after the conflict; it is not clear how widespread they were before.

The Tanzanian case also provides process-tracing evidence for how inclusive nationalist narratives can promote tolerance and ethnic peace. Tanzania’s founding President Julius Nyerere promoted a nationalist narrative promoting pan-ethnic unity against colonialist “exploiters,” energetically spreading Swahili-language education and successfully turning Swahili into a symbol of national unity. The result was impressive: by the early 21st century, 83% of poll respondents in Tanzania claimed to “trust other tribes,” and 69% said they felt “only Tanzanian,” rather than identifying with their ethnic group (2015, pp. 216–221).

In other cases, evidence on prejudice is uneven, though the performance of the imperfect measures that are used is fairly good. In most violent cases, there is suggestive anecdotal evidence. Armenians admitted to widespread loathing of Turks; rioting Azerbaijanis endorsed genocide; Abkhazians called Georgian a “dog’s language”; Sudanese northerners referred to southerners as “dogs,” “slaves,” and “monkeys”; Rwandan Hutus openly joked to Tutsis about killing them all; and even Mahatma Gandhi endorsed bigoted stereotypes about Muslims, writing: “the Mussulman is as a rule a bully” (2015, p. 155). Some survey data support the hypothesis: in violent cases, Filipino Christians were found to have pervasively negative attitudes toward Muslims, Georgians reported negative attitudes toward minority Abkhazians and Ossetians, and Palestinians are similarly negative toward Israelis. On the other side, black prejudice against whites in 1990s South Africa was in single-digit percentages, which helps explain why the peace held after the end of apartheid.

As might be expected, moderate levels of prejudice have an indeterminate effect. For example, 35% of Kazakhs were found to oppose mixed marriages with Russians—a common measure of prejudice—but that case remained peaceful owing to the ethnically inclusive leadership of Nursultan Nazarbaev (Kaufman, 2001, p. 80). In contrast, a similar proportion of Croats reported feeling disadvantaged as a group (a measure of ethnic resentment)—meaning that nearly two-thirds did not feel that way—but that case resulted in war due to the aggressive leadership of Slobodan Milosevic in Serbia and Franjo Tudjman in Croatia (Kaufman, 2001, p. 183).

While all of this evidence is imperfect, it seems fairly reliable. For example, anecdotes about Sudanese northerners’ racism toward southerners are common enough to make it hard to believe the attitude was uncommon. Similarly, the poll of Filipino Christians showing anti-Muslim sentiment was performed in 1975, after the war began, so it probably showed levels of hostility greater than before; but the findings were so overwhelming that it is implausible to suggest that the feelings were entirely a result of the war, and anthropological studies show evidence that it was widespread before as well.

Regarding feelings of threat, the case studies find evidence for such feelings in every case of ethnic violence, and a startling lack of such concern in some of the peaceful cases (Tanzanians, Azerbaijanis in Georgia). The evidence is mostly discursive: in the violent cases, nationalist leaders framed the situations in terms of threats to the group, and crowds responded (survey data on ethnic fear remains scanty). This evidence supports a constructivist interpretation of symbolic politics theory, focusing on the interaction of discourses, but it is indicative at best in support of the hypothesized social psychological mechanism.

Again, however, this “indicative” evidence is highly plausible. Given the pervasiveness of genocide discourse among Armenians, it would be absurd to suggest that Armenians did not feel threatened; and of course the same goes for Israelis. Similarly, given their circumstances and histories, it would be absurd to deny the significance of feelings of threat among Rwandans, Sudanese southerners, or South African blacks under apartheid. The strongest evidence probably comes in the Rwandan case, but even there the survey data on Hutu fears (McDoom, 2009, 2012) is retrospective, and so is vulnerable to the problem of inaccurate recall. Thus, none of these data fully meet the usual positivist evidentiary standards; but they do meet Bayesian standards: it is more probable than not that feelings of threat in these groups were widespread and consequential.

Narratives, prejudices, and threats are, however, contextual variables in the symbolist story; most of the action is in elite rhetoric—the way political leaders rally support by appealing to emotion-laden symbols. This part of the story is well documented in the cases. In Sudan, for example, Jaafar al-Numayri initially gained power as a late-1960s Arab socialist but then slowly evolved into an early-1980s Islamist. Thus, the formerly hard-drinking army officer successively established a Holy Quran Festival, began publicly performing Friday prayers, cracked down on drinking and gambling, and began dressing as an imam (Kaufman, 2015, p. 105), all in an effort to appeal to popular Islamist sentiment. The acts that triggered the north–south civil war in 1983—especially the dissolution of the south’s autonomy and the imposition of sharia law nationwide—were part of this same campaign. Thus, it is clear that ethnic civil war stemmed directly from symbolic ethnic politics.

The other violent cases involve similar stories. The secessionism of the Armenians of Mountainous Karabagh was powered by pervasive reference to the 1915 genocide that is at the heart of the Armenian myth–symbol complex. The Abkhazian separatist campaign, responding to the majority’s “Georgia for the Georgians” rhetoric, featured commemoration of the “Mohajirstvo,” the nineteenth-century ethnic cleansing suffered by Abkhazians and neighboring peoples (2001, p. 102). Slobodan Milosevic rocketed to prominence among Serbs by visiting Kosovo Field, the most revered spot in Serbian nationalist mythology. And of course Rwanda’s Hutu extremists tried to maintain their power by appealing to and encouraging anti-Tutsi racial attitudes, most famously through their sponsorship of the “hate radio” station RTLMC (Kaufman, 2015).

The strongest empirical criticism of one of the cases is Gagnon’s (2004) account of the Yugoslav conflict, which demonstrates that ethnic prejudice, hostility, and threat perceptions were relatively low in Yugoslavia before 1990. Thus, he notes survey data showing that most people in Bosnia trusted other groups in Bosnia; that most rated interethnic relations at the community level as good in both Bosnia and Croatia; and that 80% or more of Serbs and Croats did not perceive threats to “national rights.” Furthermore, the majority of Serbian draftees in 1991 refused to serve in the army, and pre-1991 efforts at chauvinist outbidding failed. Even in the 1990 elections, the parties with the more extreme ethnic platforms were defeated; leaders like Milosevic promised moderation in their campaigns and then shifted their stances after the elections (Gagnon, 2004, pp. 2, 36–50).

This evidence demonstrates that Yugoslavia was not a case of mass-led ethnic mobilization, but it is compatible with the symbolist argument (Kaufman, 2001) for an elite-led process of symbolic politics. With hostile narratives prominent but prejudice rare, Milosevic fed threat perceptions by sponsoring the actions of violent extremists. The context of rising fear in turn made threat frames more widely resonant. Gagnon’s (2004, pp. 7, 25) argument is that “violence constructs political space” by raising the salience of ethnic identity and works to “demobilize [less nationalist] opponents.” The symbolist account does not deny this argument, however. It explains how the process worked, agreeing that Milosevic “was not carrying out a freely developed consensus of opinion,” but rather “created and fed the fear of ethnic extinction” (Kaufman, 2001, p. 180).

The fact of widespread symbolic politics is therefore well established. Process-tracing also demonstrates the pivotal role of such political appeals as a causal factor in promoting ethnic civil war. It was, again, the nationalist and Islamist appeals of the Armenian National Movement and Numayri, respectively, that were at the core both of their strategies for gaining and holding power and of the processes that led to ethnic civil war. This process of ethnic outbidding is widely recognized in the 1990s literature on ethnic conflict (e.g., Lake & Rothchild, 1996).

There is also variation on the dependent variable: several cases demonstrate how and why symbolic ethnic politics can fail to gain traction. In the Philippines, an early effort to mobilize Muslims against the government failed, even in the aftermath of a widely publicized massacre, because the movement’s leader lacked credibility on the issue (Kaufman, 2015, pp. 79–80). In South Africa, anti-white racial appeals by the African National Congress’s more radical rivals fell flat due to the low levels of anti-white prejudice among blacks. Finally, in Tanzania, a strong national identity, coupled with low levels of ethnic prejudice, explains the complete failure of efforts to mobilize Tanzanians along ethnic lines (2015, p. 224).

As for the final variable, organization, the studies amply demonstrate its importance in launching and sustaining any serious ethnic challenge to the state; but they equally demonstrate the great range of different kinds of organizations or social ties that can be harnessed. In elite-led conflicts like Moldova and the former Yugoslavia, the rebellions were led by local and regional governments and Communist Party organizations (Kaufman, 2001). In Sudan and Rwanda, the rebel core consisted of army units. Mobilization in South Africa was led by the United Democratic Front, which in turn was an umbrella group composed of student, labor union, and civic (local protest) groups. Gandhi combined his vast personal charisma, which inspired imitators and followers across India, with shrewd networking skills in coopting the previously existing Indian National Congress and Home Rule League. In the Philippines, rebel organizing strategies varied across ethnic groups, with Maguindanaoan datus (traditional local headmen) serving as brokers for rebel organization, while Tausug organizers evaded the datus, relying on school and kinship ties (Kaufman, 2015).

This diversity validates the theoretical decision to emphasize the importance of organization but not to assume the necessity of any particular type. Once organizational ties exist, they can be repurposed for protest or for war if the other conditions are also present, a fact well understood in the social mobilization literature (Tilly, 2001).

In sum, the internal validity of the symbolic politics account of ethnic war is firm. It is hardly disputable that in these cases of ethnic civil war, hostile narratives and prejudice exist, and that against that background, a combination of perceived threats, ethnic outbidding by elites (using symbolic appeals), and organizational capacity are instrumental in causing escalation to war. Symbolic politics theory is sometimes criticized as primoridalist: Hale (2008, p. 18), for example, classifies it as assuming that ethnic identity is “inherently conflictual,” and Laitin (2007, p. 2) has it as assuming “nationalism is dangerous.” This, however, is an overstatement. The argument is that while ethnic and national identities often conflict, their likelihood of doing so varies with the degree to which group narratives recount conflict and the degree of prejudice. If prejudice is absent and national myths are not hostile toward the other group, conflict is unlikely unless some external factor causes threat perceptions to rise.

The larger question about symbolist theory is external validity: do the 16 cases so far considered represent what happens in most or all ethnic conflicts, and are there alternative accounts or omitted variables that better explain the phenomenon? Nationalist Passions (2015) argues for external validity on the basis of a most-different-systems logic: if societies as different as Rwanda and the Philippines (and Armenia, Serbia, and Israel) all display a similar pattern of politics, the case for external validity is highly plausible.

Other approaches, however, are more popular. In the early 2000s, symbolic politics logic was starkly contradicted by a series of quantitative studies (e.g., Collier & Hoeffler, 2004; Fearon & Laitin, 2003), suggesting that ethnicity is irrelevant for explaining civil war. Working with global datasets, these studies found that proxies for ethnic diversity—religious and ethnolinguistic fractionalization—were insignificant as causal factors in civil war. Rather, Fearon and Laitin (2003) argued that the pivotal cause of civil war is opportunity: weak states and rough terrain create circumstances that are essentially a civil war waiting to happen. Collier and Hoeffler (2004) argued that the main motivation is greed: the existence of “lootable resources” such as Africa’s famous conflict diamonds is what prompts rebellion. Thus, in this view, civil war is simply organized crime on an especially large scale.

These conclusions, however, were artifacts of flaws in the data and have since been abandoned by most scholars. Reynol-Querol (2002; Montalvo & Reynol-Querol, 2005) constructed a measure of “polarization” that reaches its maximum in a situation in which two ethnic groups each comprise half of a population. Reynol-Querol showed that measures of both linguistic polarization and religious polarization were strongly correlated with civil war outbreak, even if “fractionalization” was not. This finding broke the back of the “ethnicity doesn’t matter” school, and indeed Collier and Hoeffler abandoned it as well (Collier, Hoeffler, & Rohner, 2009).

Later quantitative studies moved the literature further in the direction of symbolist logic by recognizing that ethnic issues, and not only ethnic identities, play a role in causing civil war. “Horizontal” inequality—that is, inequality across ethnic groups—was shown to be associated with civil war, explaining why economic issues come to be defined in ethnic terms (Cederman et al., 2011; Ostby, 2008). Bhavnani (2009) proposes a formal model in which the probability of civil war varies with the salience of ethnic identity.

With these studies, the mainstream quantitative literature—even in economics journals—stepped out of a purely economic ontology and into one compatible with the symbolist paradigm. Most notably, Esteban et al. (2012), writing in American Economic Review, concede the potential role of “hatreds and resentments” in motivating civil war, and also recognize the importance of symbolic issues such as “ideological or religious supremacy” among the public goods at stake. Indeed, their central argument is that ethnic polarization is most important when public (including symbolic) goods are at stake and when ethnic group cohesion (measured in terms of popular values) is high. Little of this diverges from the symbolic politics account.

The main theoretical challenge to symbolic politics theory comes from advocates of rational choice. Again leading the charge is Laitin (1998, 2001), building in particular on Fearon’s (1998) suggestion that ethnic violence be understood as resulting from problems of credible commitment. As Laitin (1998: 853) argues, “the key to secessionist war … is whether leading elites in the national homeland … make credible commitments [of support] to the minority population.” While this is often true—and compatible with symbolist logic—the evidence Laitin uses to attack symbolic politics theory is weak. The most relevant of his evidence is his comparison of scores for “cosmopolitanism” among different ethnic groups in six post-Soviet states, which measures the inverse of prejudice. Ironically, the lowest score on cosmopolitanism among “titular” groups is that of Azerbaijanis toward Armenians (Laitin, 1998, p. 848)—that is, the only case among his six that involved mass-led violence. Thus, Laitin’s key measure actually supports the symbolic politics hypothesis.1

Other applications of rational choice theory to ethnic conflict are common. Valentino (2004, p. 186), for example, attributes the Rwanda genocide to “the cold logic that led Colonel Bagosora to … [conclude], ‘The only plausible solution … would be the elimination of the Tutsis’.” But the evidence does not support such claims: by distracting the army from its battle with the Tutsi-led rebels while alienating all potential foreign supporters, the genocide ensured the defeat of the genocidaires. Similarly, Posner (2005, p. 97) asserts that African voters support candidates from their own ethnic group in the belief that leaders divert economic goods to co-ethnics. However, his own evidence shows that most members of the favored group in his case did not believe this, and it may well not have been true. Insistent on denying the relevance of symbolic goods, these theorists rely on assumptions that are dubious at best.

Arguably the most sophisticated alternative to the symbolist theory of ethnic war is Wimmer’s (2013) “power configuration” approach. Fusing an updated version of mobilization theory with quantitative data analysis and formal modeling, Wimmer proposes an impressive argument centered on the proposition that the key driver of ethnic violence is ethnic exclusion: the larger the group excluded from political participation and the receipt of public goods, the greater the likelihood of civil war. The motive for rebellion, he states, is “fear of political domination by ethnic others” (2013, p. 29). The likelihood of war rises further if the state is weak, poor, or lacks legitimacy. Furthermore, Wimmer suggests that different kinds of ethnic wars—secessions, rebellions seeking power in the center, and infighting among elite groups—should display different causal patterns. His data analysis supports these hypotheses.

Since this power configuration approach shares roots in mobilization theory with the symbolist approach, these arguments and conclusions are generally compatible with symbolic politics theory. From a symbolist point of view, legitimacy is a symbolic predisposition that typically grows out of nationalist narratives, and “fear of domination” means threat perceptions, which are a result of ethnic exclusion. Wimmer also agrees on the importance of social organization; indeed, he presents a highly nuanced and detailed understanding of how variations in social organization matter, something absent in symbolic politics analyses.

Another first-rate work in this vein is McDoom’s (2009) study of the Rwandan genocide. In McDoom’s account, the three key factors that led to genocide were the security threat posed by the ongoing civil war, the opportunity presented by the death of the president, and the organizational strength of the Rwandan state. He also mentions the roles of the genocidal leadership and of the myths of national identity that contributed to violent mobilization. Thus, his only disagreement with symbolic politics theory is over how significant a role prejudice and racial hatred played in the genocide. But the effects of cultural and attitudinal variables are often overshadowed in single-case studies such as this one; it is in cross-case comparisons that variation in narratives and prejudice have the most clearly identifiable effects.

Wimmer’s approach is also generally compatible with symbolic politics theory. The main difference between the two is in Wimmer’s purely material account of legitimacy and—despite his references to “symbolic power”—his rejection of the importance of symbolic goods such as group status. From a symbolic politics perspective, however, the two approaches are actually complementary, with symbolic politics theory providing the superior account of legitimacy, symbolic power, and the psychological microfoundations of political behavior; and the power configuration approach providing the superior account of the role of institutions and social ties.

Broader Applications and Allied Approaches

As mentioned earlier, a distinctive feature of symbolic politics theory is that it is based on a philosophy of science that is neither positivist nor poststructural, but rather scientific realist. Growing out of the philosophy of Roy Bhaskar, as adapted for international relations theory by Colin Wight (2006) and others, the central objective of scientific realism is “causal analysis focused on mechanisms” (Joseph, 2014). In this perspective, statistical correlations cannot suffice to demonstrate causality; rather, proving causality requires hypothesizing a specific causal mechanism and then demonstrating that this mechanism actually occurs and results in the outcome of interest. Furthermore, these hypothesized mechanisms are real; they cannot be mere theoretical constructs. An analogy might be to organic chemistry, where an explanation of how reagents combine to form a new compound must specify the order in which reactions occur and then demonstrate that the intermediate products hypothesized at each stage actually do appear during the reaction.

This kind of thinking represents a significant trend in international relations theorizing, as illustrated in George and Bennett’s (2005) Case Studies and Theory Development in the Social Sciences, which spells out a methodology that fits with scientific realism. The future prospects of symbolic politics theory depends on the degree to which dissatisfaction with neopositivist theorizing in international relations leads to increased acceptance of the scientific realist ontology and a willingness to permit the questioning and replacement of the rational actor assumption.

Why should international relations theory move in this direction? Nationalist Passions suggests that symbolic politics theory can be generalized into the basis for a theory of international relations, an idea elaborated in a yet-unpublished paper (Kaufman, forthcoming). The starting point is the suggestion, “the same set of arguments that explains ethnic war and peace can also explain international war and peace” (2015, p. 263). The result is an innenpolitik argument that since most powerful states are nation-states, and nationalist politics looks very much like ethnic politics, it follows that the same political processes that lead to ethnic wars also lead to international wars.

The American path to the war in Iraq in 2003 illustrates the idea. The starting point was a widespread narrative that Saddam Hussein was an enemy, combined with strong symbolic predispositions against his regime stemming from the previous conflict in 1990–1991. In the post–9/11 context, when threat perceptions were high, the “terror management” effect interacted with these narratives and symbolic predispositions to provide a veneer of plausibility to the Bush administration’s use of threat frames to conflate the threats from al-Qaeda and Iraq to justify war against Iraq. The institutional power of the American state, coupled with the arousal of American nationalist passions, did the rest.

The clearest advantage of the symbolist argument is that it suggests a way to explain civil wars—a large majority of 20th- and 21st-century wars—as well as international wars using the same logic. It does so by starting with the assumption that the relevant actors are organized groups, not only states (cf. Goddard & Nexon, 2016). Nation-states combining the symbolic power of nationalism with the institutional power of the state would obviously be most powerful, but nonstate nations (a.k.a. ethnic groups) can also be important (Kaufman, 2015, p. 264). Non–nation-states are expected to be relatively weak, and therefore vulnerable to civil war, because their institutional strength is not supported by the symbolic resources of a national identity.

At the same time, a diverse set of existing findings in conflict research can all be understood in symbolic politics terms (Kaufman, forthcoming). Owen’s (1994) account of the democratic peace, for example, looks very much like symbolic politics. He argues that liberal values (symbolic predispositions) encourage reduced threat perceptions among liberal states, while democratic institutions (organizations) permit popular pressure to influence government decisions, but they have a pacifying effect only if leaders frame rivals as fellow liberal states worthy of trust. Symbolist theory also explains why enduring rivalries make war more likely (by generating hostile narratives and symbolic predispositions). It explains why (Huth, 1988, p. 75) bullying tactics in one crisis makes war more likely in the next crisis between the same states: because the loss of status for the bullied state generates hostile narratives and symbolic predispositions as well as feelings of resentment. It also explains Mueller’s “movement of ideas” thesis for the decline of international war (owing to the spread of antimilitarist narratives and symbolic predispositions). In sum, symbolic politics theory can serve as a general theory of conflict in world politics that provides an overarching explanation for much of what we already know, tying many existing findings together with a single logic.

The focus on symbolic predispositions also allows a synthesis of the insights of offensive realism, defensive realism, and liberal institutionalism. They differ primarily in their assumptions about state goals—respectively, the maximization of power, security, or broad material interests. The symbolist hypothesis would be that state goals should vary with the symbolic predispositions of those who make state policy. In general, states led by liberals should be expected to behave as liberal theory predicts, and states led by realists should be expected to behave as realist theory predicts. The security dilemma is not a constant but a variable (2015, p. 262)—as constructivism (Wendt, 1999) and defensive realism (Glaser, 1997) also contend. The mechanism by which the security dilemma works is the terror management effect that induces people feeling threatened to become more aggressive. However, its strength depends not only on structural factors but also on national narratives and symbolic predispositions: liberals will be slow to perceive threats from those they trust, for example, but quick to perceive threats from traditional rivals.

The “neo-neo synthesis” thus emerges as a fusion into symbolic politics theory—a fusion in which constructivism also has a place. This is especially true of constructivist approaches (e.g., Katzenstein, 1996; Berger, 1998; Hopf, 2002) that focus on cultural narratives, values, and discourses in explaining political outcomes, such as the relative pacifism of post–World War II Japan and Germany. Symbolic politics also provides a way to conceptualize the status-conscious policies of leaders such as Vladimir Putin and Xi Jinping (see Larson & Shevchenko, 2010). Lind’s (2008) discussion of how Japanese leaders’ apologies for Japan’s World War II behavior have led to backlash at home and “jangled nerves” in neighboring countries is another illustration of how more-or-less constructivist analyses often shade into symbolist discussions of the relationship between discourse and emotion. Also in this category is Gries et al.’s (2012) experimental exploration of the importance of feelings of symbolic loss and gain in influencing popular attitudes toward U.S.-China relations in both countries.

Recent works specifically focused on emotion also fit into the symbolic politics family. Mercer’s (2010) discussion of “emotional beliefs,” for example, is about symbolic predispositions by another name. Sasley (2010) has shown that such “affective attachments” do influence elite foreign policy behavior, just as symbolic politics logic expects. Crawford (2014) also makes a symbolic politics argument in discussing how emotions can be institutionalized in perceptions and political practice.

To take a more specific example, Kim (2012) explicitly applies symbolic politics theory to explain the “history problem” dividing Japan and South Korea. Kim demonstrates that efforts by the Japanese and South Korean governments to cooperate against common security threats, especially from North Korea, have repeatedly foundered owing to disputes over Japanese history textbooks, the “comfort women” issue, and high-level Japanese visits to the Yasukuni shrine. These symbolic issues outweigh security concerns, Kim shows, because of symbolic politics, with clashing historical narratives fueling mass-led political mobilization in South Korea against offensive Japanese actions, and elite-led pushback in Japan.

The focus of symbolic politics on organized groups as the unit of analysis in world politics also opens the door to a systematic account of the roles of the whole array of nonstate actors considered in liberal institutionalist theory. For example, environmental and human rights nongovernmental organizations (NGOs) can be understood as “pure players in symbolic politics,” appealing to humanitarian symbolic predispositions among their supporters to elicit resources and generate influence (Kaufman, 2015, p. 264). Multinational corporations rely primarily on organizational power, buttressed by a one-dimensional symbolic appeal (an appeal to the symbolic predisposition that might be labeled “greed”). Organized crime groups generate organizational and material power by appealing to a mixture of greed and fear. Thus, symbolic politics theory suggests a single, simple, analytical framework for explaining international conflict and cooperation that accounts for the diversity of actors, motivations, and logics of behavior that characterize world politics.

A final point is that the scientific realist foundation of symbolic politics theory suggests a richer account of international structure than the common neorealist one. Briefly, if the units in the international system are organized groups, then the structure of the international system consists of the social ties or relationships among people with social roles in those groups, as well as the pattern of symbolic predispositions underlying them.2 From this perspective, the structure of the early-21st-century international system is characterized by, inter alia, an especially powerful set of nation-states that are great and medium powers; a military alliance system linking most of them to the United States; an international economic system that guarantees poverty for those excluded from it; social media that empower nonstate groups of all kinds, from human rights NGOs to transnational terrorist organizations; and great diversity in symbolic predispositions from nearly pacifist western Europeans to the apocalyptically violent inclinations of Islamic State supporters.


As a middle-range theory of ethnic conflict and ethnic war, symbolic politics theory has proven itself. There is variation in groups’ narratives about each other, and more hostile narratives reliably play a key role in causing ethnic violence. There is also cross-national variation in prejudice, and good evidence has been presented that higher levels of prejudice in a society increase the likelihood of ethnic violence. As has long been known, feelings of threat also play a critical role in creating ethnic conflict—but low levels of prejudice tend to reduce threat perceptions.

Furthermore, the processes of symbolic politics are what lead to ethnic conflict. Ethnic leaders lead in large part by using symbols based in group narratives to appeal to popular symbolic predispositions and thereby stir emotions. Such appeals are not empty rhetoric but a fundamental tool of politics. Threat frames are what mobilize people for violence. Injustice frames can generate a contentious politics of redistribution. Low levels of prejudice and threat are conducive to a low-conflict politics of distribution. Finally, political mobilization at any level is also heavily reliant on the nature and quality of political and social organizations available to ethnic leaders.

The case study evidence for these propositions constitutes a strongly supportive track record for the theory, though certainly not proof of universal validity. The stark difference between Rwanda’s hate-filled ethnic narratives and next-door Tanzania’s narrative of national unity is a salient example of the role of narratives. The variation in prejudice is demonstrated inter alia by the data showing very high levels of prejudice against Muslims in the Philippines and very low levels of ethnic prejudice in Tanzania. The importance of threat perceptions, threat frames, leadership, and organization are less controversial but also supported by the evidence available. Real verification of the theory, however, awaits more large-scale systematic efforts to measure both narratives and symbolic predispositions cross-nationally, to demonstrate their relationship to patterns of ethnic politics.

As international relations theory, symbolic politics is still in an embryonic stage. Its basic logic certainly suggests a useful theory of foreign policy, but it has not yet been fully specified. Its basis in scientific realism suggests an understanding of international system structure that may be very promising, but as yet it has only been hinted at. The promise of symbolic politics theory is that it can generate explanations that track closely with real politics on the ground and with common-sense intuitions about it. Leaders make so many speeches for excellent reasons; the speeches are neither cheap talk nor empty rhetoric but the essential stuff of politics. Symbolic politics theory offers the most useful way of understanding what sort of rhetoric is likely to be politically successful in what circumstances, while taking into account the centrality of social organization for translating ideas into collective action.


Bartels, L. M. (2002). Beyond the running tally: Partisan bias in political perceptions. Political Behavior, 24(2), 117–150.Find this resource:

Benford, R. D., & Snow, D. A. (2000). Framing processes and social movements: An overview and assessment. Annual Review of Sociology, 26, 611–639.Find this resource:

Berger, T. (1998). Cultures of anti-militarism: National security in Germany and Japan Baltimore: Johns Hopkins University Press.Find this resource:

Bhavnani, R (2009). Ethnic Polarization, Ethnic Salience, and Civil War. Journal of Conflict Resolution, 53(1), 30–49.Find this resource:

Cederman, L-E., Weidmann, N. B., & Gledtitsch, K. S. (2011). Horizontal inequalities and ethnonationalist civil war: A global comparison. American Political Science Review, 105(3), 478–495.Find this resource:

Collier, P., & Hoeffler, A. (2004). Greed and grievance in civil war. Oxford Economic Papers, 56, 563–595.Find this resource:

Collier, P., Hoeffler, A., & Rohner, D. (2009). Beyond greed and grievance: Feasibility and civil war. Oxford Economic Papers, 61(1), 1–27.Find this resource:

Crawford, N. (2014). Institutionalizing passion in world politics: Fear and empathy. International Theory, 6(3), 535–557.Find this resource:

Cuillier, D., Duell, B., & Joireman, J. (2010). The mortality muzzle: The effect of death thoughts on attitudes toward national security and a watchdog press. Journalism, 11(2), 185–202.Find this resource:

Duckitt, J. (2003). Prejudice and intergroup hostility. In D. Sears, L. Huddy, & R. Jervis (Eds.), Oxford handbook of political psychology (pp. 559–600). Oxford: Oxford University Press.Find this resource:

Edelman, M. (1971). Politics as symbolic action: Mass arousal and quiescence. New York: Academic Press.Find this resource:

Esteban, J., Mayoral, L., & Ray, D. (2012). Ethnicity and conflict: An empirical study. American Economic Review, 104(4), 1310–1342.Find this resource:

Evans, J. St. B. T., & Frankish, K. (2009). The duality of mind: An historical perspective. In J. St. B. T. Evans & K. Frankish (Eds.), In two minds: Dual processes and beyond (pp. 1–29). New York: Oxford University Press.Find this resource:

Fearon, J. D. (1998). Commitment Problems and the Spread of Ethnic Conflict. In D. A. Lake and D. Rothchild (Eds.), The International Spread of Ethnic Conflict: Fear, Diffusion, and Escalation (pp. 107–126). Princeton: Princeton University Press.Find this resource:

Fearon, J. D., & Laitin, D. D. (2003). Ethnicity, insurgency, and civil war. American Political Science Review, 97(1). 75–90.Find this resource:

Gagnon, V. P., Jr. (2004). The myth of ethnic war: Serbia and Croatia in the 1990s. Ithaca, NY: Cornell University Press.Find this resource:

George, A., & Bennett, A. (2005). Case studies and theory development in the social sciences. Cambridge, MA: MIT Press.Find this resource:

Glaser, C. (1997). The Security Dilemma Revisited. World Politics, 50(1), 171–201.Find this resource:

Goddard, S., & Nexon, D. (2016). The dynamics of global power politics: A framework for analysis. Journal of Global Security Studies, 1(1), 4–18.Find this resource:

Greenberg, J., Pyszczynski, T., Solomon, S., Rosenblatt, A., Veeder, M., Kirkland, S., & Lyon, D. (1990). Evidence for terror management theory II: The effects of mortality salience on reactions to those who threaten or bolster the cultural worldview. Journal of Personality and Social Psychology, 58(2), 308–318.Find this resource:

Gries, Peter Hays, Peng, Kaiping, & Crowson, H. Michael. (2012). Determinants of Security and insecurity in international relations: A cross-national experimental analysis of symbolic and material gains and losses. In V. P. Shannon & P. A. Kowert (Eds.), Psychology and constructivism in international relations (pp. 170–196). Ann Arbor: University of Michigan Press.Find this resource:

Haidt, Jonathan. (2012). The righteous mind: Why good people are divided by politics and religion. New York: Pantheon.Find this resource:

Hale, Henry E. (2008). The foundations of ethnic politics: Separatism of states and nations in Eurasia and the World. Cambridge: Cambridge University Press.Find this resource:

Hopf, T. (2002). Social construction of international politics: Identities and foreign policies, Moscow, 1955 and 1999. Ithaca, NY: Cornell University Press.Find this resource:

Horowitz, D. (1985). Ethnic groups in conflict. Berkeley: University of California Press.Find this resource:

Huth, Paul K. (1988). Extended deterrence and the prevention of war. New Haven, CT: Yale University Press.Find this resource:

Jackson, P. T. (2010). The conduct of inquiry in international relations. New York: Routledge.Find this resource:

Joseph, Jonathan. (2014). The impact of Roy Bhaskar and critical realism on international relations. E-International Relations . Retrieved from this resource:

Katzenstein, P. (Ed.). (1996). The culture of national security: Norms and identity in world politics. New York: Columbia University Press.Find this resource:

Kaufman, S. J. (1996). Spiraling to ethnic war: Elites, masses and Moscow in Moldova’s civil war. International Security, 21(2), 108–138.Find this resource:

Kaufman, S. J. (2001). Modern hatreds: The symbolic politics of ethnic war. Ithaca, NY: Cornell University Press.Find this resource:

Kaufman, S. J. (2009). Narratives and symbols in violent mobilization: The Palestinian-Israeli case. Security Studies, 18(3), 400–434.Find this resource:

Kaufman, S. J. (2015). Nationalist passions. Ithaca, NY: Cornell University Press.Find this resource:

Kaufman, S. J. (forthcoming). The symbolic theory of world politics. Presented at the International Studies Association Conference, Atlanta, GA.Find this resource:

Kim, Ji-Young. (2012). Symbolic politics, history problems, and South-Korea-Japan Security relations during the post-Cold War period. PhD diss., University of Delaware.Find this resource:

Laitin, D. D. (1998). Identity in formation: The Russian-speaking populations in the Near Abroad. Ithaca, NY: Cornell University Press.Find this resource:

Laitin, D. D. (2001). Secessionist rebellion in the former Soviet Union. Comparative Political Studies, 34(8), 839–861.Find this resource:

Laitin, D. D. (2007). Nations, states and violence. Oxford: Oxford University Press.Find this resource:

Lake, D. A., & Rothchild, D. (1996). Containing fear: The origins and management of ethnic conflict. International Security, 21(2), 41–75.Find this resource:

Larson, D., & Shevchenko, A. (2010). Status seekers: Chinese and Russian responses to U.S. primacy. International Security, 34(4), 63–95.Find this resource:

Lind, J. (2008). Sorry states: Apologies in international politics. Ithaca, NY: Cornell University Press.Find this resource:

McDoom, O. S. (2009). The micro-politics of mass violence: Authority, security and opportunity in Rwanda’s genocide. PhD Dissertation, London School of Economics and Political Science.Find this resource:

McDoom, O. S. (2012). The psychology of threat in intergroup conflict: Emotions, rationality, and opportunity in the Rwandan genocide. International Security, 37(2), 119–155.Find this resource:

Mead, W. R. (2002). Special providence: American foreign policy and how it changed the world. London: Routledge.Find this resource:

Mercer, J. (2010). Emotional beliefs. International Organization, 64(1), 1–31.Find this resource:

Montalvo, J., & Reynol-Querol, M. (2005). Ethnic polarization, potential conflict, and Civil Wars. American Economic Review, 95(3), 796–816.Find this resource:

Ostby, G. (2008), Polarization, horizontal inequalities and violent civil conflict. Journal of Peace Research, 45(2), 3–162. Special Issue on Polarization and Conflict.Find this resource:

Owen, J. M. (1994). How liberalism produces democratic peace. International Security, 19(2), 87–125.Find this resource:

Posner, D. N. (2005). Institutions and ethnic politics in Africa. New York: Cambridge University Press.Find this resource:

Rathbun, B. (2011). Trust in international cooperation: International security institutions, domestic politics and American multilateralism. Cambridge: Cambridge University Press.Find this resource:

Reynal-Querol, M. (2002). Ethnicity, political systems, and civil wars. Journal of Conflict Resolution, 46(1), 29–54.Find this resource:

Ross, M. H. (2007). Cultural contestation in ethnic conflict. Cambridge: Cambridge University Press.Find this resource:

Sasley, B. E. (2010). Affective attachments and foreign policy: Israel and the 1993 Oslo Accords. European Journal of International Relations, 16(4), 687–709.Find this resource:

Sears, D. O. (2001). The role of affect in symbolic politics. In James H. Kuklinski (Ed.), Citizens and politics: Perspectives from political psychology. New York: Cambridge University Press.Find this resource:

Sears, D. O., & Citrin, J. (1982). Tax revolt: Something for nothing in California. Cambridge, MA: Harvard University Press.Find this resource:

Sears, D. O., Lau, R. R., Tyler, T. R., & Allen, H. M., Jr. (1980). Self-interest vs. symbolic politics in policy attitudes and presidential voting. American Political Science Review, 74(3).Find this resource:

Shannon, V. P., & Kowert, P. A. (Eds.). (2012). Psychology and constructivism in international relations: An ideational alliance. Ann Arbor: University of Michigan Press.Find this resource:

Smith, A. D. (2009). Ethno-symbolism and nationalism: A cultural approach. New York: Routledge.Find this resource:

Tilly, C. (2001). Dynamics of contention. Cambridge: Cambridge University Press.Find this resource:

Valentino, B. A. (2004). Final solutions: Mass killing and genocide in the twentieth century. Ithaca, NY: Cornell University Press.Find this resource:

Wendt, A. (1999). Social theory of international relations. Cambridge: Cambridge University Press.Find this resource:

Wight, C. (2006). Agents, structures and international relations: Politics as ontology. Cambridge: Cambridge University Press.Find this resource:

Wimmer, A. (2013). Waves of war: Nationalism, state formation, and ethnic exclusion in the modern world. Cambridge: Cambridge University Press.Find this resource:


(1.) Laitin (1998, p. 850) attributes this finding to measurement error.

(2.) This is based on Wight’s (2006) preferred definition of structure as “systems of human relationships among social positions.”