No sympathy for the devil: Emotions and the social construction of the democratic peace

Abstract

Constructivists claim that the democratic peace is socially constructed via mutual recognition between liberal subjects. Mutual recognition is rooted in shared moral attitudes and cognitive perceptions, thereby creating liberal intersubjectivity. What is largely missing from these accounts is the fact that shared meanings and identities are not solely rooted in cognitive perceptions and moral attitudes but significantly depend upon shared emotions that underpin and reproduce intersubjectivity. Building on interdisciplinary insights from social constructivist emotion theories, it is argued here that collectively shared emotions provide a way by which liberal subjects choose particular meaning frames and interpretations, which help align and sustain mutual attitudes and perceptions in constructing categories of ‘us’ and ‘them.’ Accordingly, the theoretical question concerning how liberal democracies recognize each other as friends can be more fully answered by the high degree of emotional convergence among them. Moreover, I suggest that it is precisely this emotional convergence that underpins the way liberal selves construct non-liberal others as their enemies.

Keywords

Constructivism democratic peace emotions

Introduction

This article investigates the emotional underpinnings of the democratic peace proposition. The latter’s central empirical claim states that democracies have been reluctant to engage in armed conflict with other democracies, even though they have fought numerous wars with non-democratic states. Various theoretical accounts have embarked upon explaining this puzzle, including utilitarian accounts focusing on rational choice, institutional accounts emphasizing the pacifying role of domestic and international institutions, and, more recently, constructivist accounts privileging liberal norms and identities.¹

In this study, I will engage with constructivist explanations of the democratic peace. Constructivists claim that the democratic peace is socially constructed via mutual recognition between liberal subjects. Mutual recognition is based on shared moral attitudes and cognitive perceptions that construct categories of ‘us’ and ‘them.’² Oren’s (1995, 2003) seminal constructivist statement on the democratic peace as a ‘peace of our kind’ suggests that this peace is the outcome of the dialectical logic of self–other relationships, and thus at least implicitly points to the socio-psychological link between cognitive and emotional categories that constitutes the basis for constructing liberal intersubjectivity. This speaks directly to an ongoing debate pertaining to the underlying mechanisms that produce the democratic peace. Scholars are increasingly pointing to significant gaps in our understanding of these mechanisms (Dixon and Senese, 2002; Hayes, 2011). Particularly in conventional constructivist studies that center around notions of shared liberal perceptions and attitudes, the question looms: how are these shared attitudes and perceptions created and sustained? The prevailing assumption among conventional constructivists in International Relations (IR) is that cognition is ‘a property of intentional actors that generate motivational and behavioural dispositions’ (Wendt, 1999: 224). However, as many emotion scholars in IR have demonstrated, the claim that ideas, beliefs, and mental processes of acquiring knowledge have motivational force of their own is at least questionable.³ Knowing about ‘self’ and ‘other’ certainly influences thought processes and mental imagery, but it does not necessarily motivate to act in a certain way. It is when one gets angry at the other that one feels inclined to seek revenge and embark on retaliatory acts. Conversely, if one sympathizes with members of a group, one will likely be more trustful and behave in a conciliatory way toward members of that group. Cognition that lacks emotional input fails to produce a sense of obligation or loyalty necessary for collective identification. Conversely, emotion that lacks cognition has no object, so there is nothing to get angry or sympathetic about.

This article ties together research on the social construction of emotions with scholarship on the social construction of the democratic peace in a novel way. It is argued that the social construction of the democratic peace is not solely rooted in shared cognitive perceptions and moral attitudes but significantly depends upon collectively shared emotions that underpin liberal intersubjectivity. More specifically, I suggest that such emotions form part of the sociocultural structure by which liberal agents choose meaning frames and interpretations, which help align and sustain their cognitive perceptions and moral attitudes. In other words, the democratic peace is not solely constructed via cognitive knowledge about ‘us’ and ‘them’ but is, at the same time, underpinned by corresponding emotions of sympathy for ‘us’ and anger toward ‘them.’ This summarizes the main argument put forward here that the theoretical question of how liberals recognize each other as friends can be more fully answered by the high degree of emotional convergence among them. Moreover, I suggest that it is precisely this emotional convergence that can help explain why liberal selves construct non-democratic others as enemies.

This article limits its intervention to the theoretical level and will not present new empirical data. It explores the role of emotions within the constructivist research framework of the democratic peace, thus examining the role of emotions as a theoretical problem inside a theoretical framework. To be clear, and as I will discuss in more detail, constructivists who study the democratic peace certainly do not turn a blind eye to emotions. But while constructivist scholarship on the democratic peace presents convincing accounts of the socially constructed nature of liberal intersubjectivity in world politics, it nevertheless centers on cognitive constructions of liberal intersubjectivity and arguably takes for granted its emotional underpinnings. It is argued here that our theoretical understanding of the social construction of the democratic peace will advance significantly if we appreciate more fully and systematically integrate emotions into the study of liberal norms and communities—indeed in any norms and communities—and that a fuller engagement with the ways this reveals subtle yet powerful processes of aligning liberal subjects provides valuable insights into how collectively shared emotions underpin the social construction of liberal intersubjectivity.

Constructivism itself is not a uniform research program but rather a loose paradigm of related interpretations (usually divided into conventional, critical, and post-structural strands). A key difference that is significant to the argument put forward here is that conventional constructivist (as opposed to critical constructivist and post-structuralist) approaches do not dismiss rationalist-liberal conceptions outright but reconceive them in terms of the logic of appropriateness. This notion presents the specific puzzle that particularly highlights the absence of emotion and the need for its addition in the social construction of the democratic peace. Brent Steele (2007: 32), for example, follows Wendt in arguing that such rationalist-liberal conceptions de-socialize or have the capacity to de-socialize agents so that friends and enemies may not emerge as easily as we might think. Put differently, by sticking to a rationalist-liberal understanding of the democratic peace, conventional constructivists have a hard time explaining how individually internalized liberal norms and identities provide the social glue for the emergence of liberal-democratic communities. I suggest that this is precisely why we need to incorporate emotions to explain the presence and persistence of such communities. Collectively shared emotions arguably provide the glue that helps liberal subjects ‘stick’ together, creating the liberal intersubjectivity on which liberal communities are built, thus bridging the tension between individual liberal subjects and the political communities that make liberal subjectivity possible.

To develop this argument, the article is divided into three parts. In the first part, I briefly reflect on the relationship between constructivism and liberalism and review social constructivist accounts of the democratic peace that explain liberal intersubjectivity via processes of mutual recognition. In the second part, I move beyond cognitive explanations of the social construction of the democratic peace by examining how intersubjectivity hinges on socially constructed feeling rules. In the third part, I employ these insights to develop a framework for the role emotions play in underpinning liberal intersubjectivity.

Liberalism, constructivism, and the democratic peace

Liberalism constitutes a fuzzy and inherently contested concept. I treat liberalism here not as a deterministic variable endowed with certain ‘natural’ characteristics but as a historically specific and temporary narrative that builds allegiance by constituting the ways that subjects are connected to each other. Rather than answering ‘what’ liberalism is by listing its attributes and properties, I ask ‘when’ and ‘how’ it comes into being (Adler, 1997: 347). Emotions are understood here as evaluative moral judgments that are not simply personal or universal but socially constructed (Averill, 1980; Harré, 1986). Hence, when I say ‘liberal emotions’, I do not mean that emotions are either ‘liberal’ or ‘non-liberal’ in and of itself but rather how they are being framed as belonging to members of liberal or non-liberal communities, respectively.

The basis of that framing is rooted in a radical departure from the overwhelming dominance of rationalism in much of conventional constructivist research, which has stood in the way of taking emotions seriously in the social construction of the democratic peace. As pointed out above, one of the shortcomings of conventional constructivist approaches to the democratic peace lies in its ontological adherence to a rationalist-liberal understanding that provides a very thin social basis for the emergence of liberal intersubjectivity. As a result, the liberal narrative underlying the social construction of the democratic peace has been a very rationalist narrative that hardly accounts for emotions. If anything, emotions are regarded as confused, irrational (often violent) bodily motions that prevent any self-reflection about the conduct of liberal subjects. By contrast, Martha Nussbaum (2001: 658) argues that the expression of emotions is a moral act that underpins ‘the model of rational judgment required by a democratic nation.’ In IR, Renée Jeffery (2014: 7) makes the important argument that emotions, along with reason, form an indispensable element in motivating ethical behavior and constraining unacceptable actions in democratic deliberations. Khaled Fattah and Karin Fierke (2009: 70) extend this view by adding that the appraisal or value judgments discussed by Nussbaum are not purely individual but draw on collectively shared cultural knowledge and social institutions. Finally, Reinhard Wolf (2011: 118) explores the emotional impact of respectful behavior across group boundaries, employing a similar relational ontology. These scholars explicitly challenge the rationalist-liberal assumption that reason ought to be liberated from the negative impact of emotions and instead confirm the view that emotions are not irrational forces but have a social pattern to them that underpins liberal (inter)subjectivity in world politics.

This conception is consistent with the constructivist underpinnings of the argument put forward here. It builds on the more critical constructivist scholarship on the democratic peace. This body of literature has pointed to some important dimensions of this phenomenon that mainstream research essentially misses (Hobson, 2011; Ish-Shalom, 2006; Oren, 1995, 2003; Steele, 2007). Particularly relevant to the argument at hand, this literature not only demonstrates how constructivism can and needs to be distinguished from liberalism but also specifies which liberal strain of the democratic peace it is going after.

Again, Brent Steele’s (2007: 33, 45) conception of liberalism is particularly helpful here. He distinguishes between so-called rationalist ‘liberal-idealism’, which views democracy as a universal outcome and promotes a value-free understanding of science, and ‘reflexive liberalism’, which emphasizes self-awareness, both of theory and of the social world it studies. From the latter point of view, liberalism—and in a similar way the democratic peace—is not some abstract form of ‘absolute Kantian principles’ but a social construction that looks at a much wider and more reflexive terrain: a theoretical concept that affects—even changes—social reality and, simultaneously, reflects the social context in which it is produced. Such a reflexive understanding of liberalism can then be understood as the social construction of liberal spaces, subjects, and institutions to which the democratic peace can be linked and from which it emerges.

Emotions and the problem of recognition

A fundamental critique of the democratic peace is the problem of recognition. For the democratic peace to work, liberal selves have to somehow know that the other is a democracy in order to recognize it as such. It is precisely this knowledge about the democratic character of likeminded others that, according to social constructivism, reassures democratic states not to fight wars against each other, providing the epistemic foundation of the democratic peace.

In what has been labeled the third wave of democratic peace scholarship (Hayes, 2011), constructivist research has centered around shared notions of perceptions and attitudes, both at the level of individual policymakers (Farnham, 2003; Hermann and Kegley, 1995; Widmaier, 2005) as well as at the state level and policy elites (Hayes, 2012; Kahl, 1999; Owen, 1994; Peceny, 1997; Risse-Kappen, 1995; Williams, 2001). Williams (2001: 529), in particular, raises the central question: ‘what, precisely, do … decision-makers look at in deciding that they are part of the shared democratic “us” and who gets to decide? What is focused on as “like”, and what is ignored?’ Williams uncovers how the liberal self is produced in the first place and views mutual recognition between such selves as a social construction involving common identity, norms, and practices. In making his argument, Williams (2001: 534) emphasizes shared moral attitudes and cognitive perception—‘an evaluation of their moral character … emerging from an evaluation of their moral attitudes’ (emphasis added)—in recognizing others as members of a liberal community.

This emphasis, in more or less similar ways, features prominently in several other constructivist studies that have sought to trace the cognitive process of how liberal democracies consider each other to be liberal democracies. Risse-Kappen (1995: 509) suggests the following:

Democracies do not fight each other because they perceive each other as predisposed toward peacefulness and then act on this assumption. … This perception then creates a security dilemma leading to behavioural patterns that confirm the presumption of enmity [toward non-democracies].

MacMillan (2004: 180) proposes that liberal ideas, a set of shared beliefs, are doing the work in the democratic peace as they give rise to ‘perceived legitimacy … in terms of liberal principles and values’ (emphasis added). In a similar way, Owen (1994: 92) claims that liberal ideas produce the moral order of the democratic peace, ‘but only when the actors’ perceptions are taken into account.’ Oren (1995: 152) postulates that mutual recognition is based on shifting perceptions of domestic and international political and historical processes ‘that made political enemies appear subjectively further and friends subjectively closer.’ Peceny (1997) argues that democratic identity is the result of an intersubjective consensus on perceptions, a finding that echoes William’s notion of mutual recognition. Adler (1997: 347) claims that ‘the democratic peace is about the historical development and spread over part of the world of an “intersubjective liberal identity” that, cutting across national borders, becomes an identity marker and indicator of reciprocal peaceful intentions.’

All of these studies share the basic notion of emphasizing cognitive explanations for the social construction of liberal intersubjectivity—centering on mutual attitudes and perceptions—as the basis for distinguishing self and other. While emotions do feature in some of these accounts, their prescriptive and purposive function in the social construction of liberal intersubjectivity remains somewhat ambiguous. Risse-Kappen (1995: 508), for example, argues that ‘the absence of publicity in autocratic systems increases the feeling of uncertainty by liberal states and might lead to increased suspicions’, and, as a result, ‘democracies will feel threatened and act accordingly.’ Williams (2001: 534, 536) claims that ‘respect is something which is due to others – which …, we naturally feel towards others – only to the extent that they live up to the moral worth … intrinsic to their humanity’ as opposed to ‘the lack of trust which liberals feel toward non-liberal states.’ As these examples illustrate, emotions seem to matter in the social construction of the democratic peace, but this hardly explains why or how. While these studies highlight the emotional connections that link liberal subjects, they do not explicitly theorize its underlying mechanisms and how this impacts on mutual attitudes and perceptions.

Williams and Risse-Kappen are both correct to call attention to the emotional footings of mutual recognition in constructing the democratic peace. But this focus needs to be deepened by an enquiry into the actual emotional constitution of liberal intersubjectivity. The current neglect of emotions means scholars miss critical facets to the creation of liberal intersubjectivity. Emotions play a decisive role in the social construction of the democratic peace, but its importance is not widely appreciated. Unpacking the ways emotions underpin liberal intersubjectivity acknowledges the political role of emotions in securing cohesion and allegiance as well as in negotiating power relations among liberals and non-liberals.

For this argument to hold this means that if there is something like a ‘liberal self’, and if the democratic peace is constructed via mutual recognition between these liberal selves, then there must be a distinct way in which emotions underpin liberal intersubjectivity. And it is here that interdisciplinary insights offered by social constructivist emotion theories turn out to be particularly instructive. In what follows, I suggest that a focus on the social nature of emotions offers a fruitful way to enter into dialogue with social constructivist theories on the democratic peace: how socially appropriate emotions function in the collective representation of shared meaning and knowledge and, ultimately, the social construction of liberal intersubjectivity.

The conceptual foundations of emotional intersubjectivity

Emotions may be a pervasive element in the social construction of the democratic peace, but, as argued above, precisely how they play a part in constituting the social and institutional structures and processes that bind liberals together is an area that conventional approaches appear to lack tools to systematically examine. Social constructivist approaches to understanding emotions have been among some of the first to take the social nature of emotions seriously. A social constructivist perspective argues that emotions are cultural products that owe their meaning and purpose to learned social rules. A social constructivist approach to emotion does not constitute a single, coherent theory of emotions but rather combines a set of generalized assumptions about the nature and function of emotions that have emerged across various disciplines, ranging from philosophy (Coulter, 1979), psychology (Averill, 1980; Harré, 1986), sociology (Gordon, 1986; Hochschild, 1979), feminist theory (Ahmed, 2003), and anthropology (Abu-Lughod, 1986; Lutz, 1988). What unites constructivist approaches is the notion that emotions are socially constructed in the sense that ‘what people feel is conditioned by socialization into culture and by participation in social structures’ (Turner and Stets, 2006: 2). Below, I summarize three core assumptions of social constructivist emotion theories to carve out the conceptual link between emotion and intersubjectivity. The purpose is to build a conceptual foundation for the emotional underpinnings of liberal intersubjectivity in the social construction of the democratic peace.

Beliefs and identity

Social constructivists reject the view of emotions as solely physiological states or natural objects. As Averill (1980: 309) underlines, ‘emotions are not just remnants of our physiological past, nor can they be explained in strictly physiological terms. Rather, they are social constructions, and they can be fully understood only on a social level of analysis’ (emphasis added). A social constructivist viewpoint holds that emotions are characterized by beliefs and social appraisals about how the world should be. Through social appraisal, emotions acquire meaning in the sense that they define individuals with unique mental and psychological configurations (Lazarus, 1991). The content of such beliefs is not a natural thing but depends upon an agent’s ability to appraise a social situation as warranting a particular emotion. While representing very different views on emotions, many scholars (Crawford, 2014; Hutchison, 2016; Mercer, 2010; Sasley, 2011) generally agree that individual emotions can be collectivized among members of a group through the participation in social discourses and cultural practices. In this view, ‘emotion finds expression only in a language and a culture, which is linked to a moral order and moral appraisal’ (Fattah and Fierke, 2009: 70). While differing about the scope of emotions as being socially constituted and how they can be studied, constructivist emotion scholars agree that emotional meanings are not confined solely to individual subjects but created in and through social interaction with others. Whenever individual subjects identify with a particular social group, they also assume its collective emotions as part of their self. Whenever individual subjects identify with a particular social group, they also assume its collective emotions as part of their self. As Tajfel (1981: 255) underlines, social identity is ‘that part of the individual’s self concept which derives from his knowledge of his membership of a social group (or groups) together with the value and emotional significance attached to that membership.’ The way we make sense of and attach meaning to our emotions depends, to a significant degree, on how we interpret and make sense of the world around us. For constructivists, emotions are inextricably linked to and reproduce sociocultural structures. This point is perhaps best stated by Lutz (1986: 5), who argues that emotional experience ‘is not precultural but preeminently cultural’ (emphasis in the original).

Knowledge and learning

The constructivist claim that emotions underpin collective beliefs and identities rooted in sociocultural structures implies that emotions are learnt as part of an agent’s socialization that reflects the moral values and beliefs of particular communities. As much as socialization involves processes of learning cognitive facts, socialization is simultaneously driven by emotional attachment to significant others based on ‘emotionally charged processes of identity formation’ (Berger and Luckmann, 1966: 131, 178). By the socialization of emotion, I mean the process of ‘how people come to feel as they do as a result of their relationship over time with others’ (Saarni, 1993: 435). Members of a particular community synchronize and converge their emotions in appropriate ways to be in tune with the collective identities and beliefs of their respective community or culture. For example, members of a church congregation collectively express grief at a funeral as a community-building experience. College students cheer for their sports team to experience a sense of ‘we-feeling.’ This notion of emotional intersubjectivity presupposes a pool of affective memories and emotional knowledge that is paradigmatically linked to social identities. Emotional knowledge can be defined as an agent’s ability to morally categorize emotional expressions and to emotionally connect these categories to others’ identities based on experience over time. In other words, subjects have to be able to know what it means to be angry, ashamed, or happy in order to understand its social implications and evoke appropriate emotional reactions toward others within a particular social context and based on a moral history. For example, anger can be interpreted as destructive to close relationships because one may have experienced the destructive nature of anger in previous relationships (being angry means I’m careless). Conversely, anger may be perceived as displaying the closeness of a relationship based on a very different emotional experience (being angry means I care).

Emotional knowledge is based on intersubjective learning; that is, the habituated establishment and recurring exchange of emotions that shape the identities of social actors. One member communicates emotions to other members who then give emotional feedback and, in turn, receive emotional feedback on their part, and so on. Crucially, this involves power and status differentiation. Agents enter preconfigured social structures within which they encounter significant others of a higher rank who educate newcomers about the meaning of their emotional expressions, thereby acting as ‘schools of emotions’ (Berger and Luckmann, 1966: 131). Through this process of emotional socialization, members of a community can enter a stage of mutual understanding by building a common emotional history together, which contributes to the establishment of shared meanings. Converging emotional expressions attune and align through ‘webs of associative knowledge’ (Kirkpatrick and White, 1985: 19). In other words, knowledge is not only shaped by emotions through the cognitive representation of a concrete object or event (such as the rattling snake as dangerous) that enables the subject to acquire information about the world and construe their reality in terms of a basic understanding of the causes of things. Knowledge is also about the emotions themselves by providing meaningful categories for affectively understanding the self in terms of others, to organize the differences in feeling among members of a group, to establish the limits of meaningful emotional expression, and to delineate legitimate power and epistemic authority (Rosaldo, 1980: 85). To sum up, socialization and learning emotional meanings is about orientation as part of the epistemic landscape of actors in world politics. It is the accumulation of memories, founding myths, identities, and symbolic patterns that enables agents to make sense of the world around them within an emotionally shared reality.

Rules and norms

Finally, constructivists assume emotions to have significant social functions that emphasize their prescriptive and purposive character. Emotions can serve both regulatory and constitutive functions in the maintenance and cultivation of shared meanings. First, in terms of regulating social behavior, emotions underpin the moral hierarchy of values and beliefs within a group by assigning emotional meaning to norms which members of the group care about, on the one hand, and by restraining undesirable attitudes and behavior, on the other hand (Hall, 2012). Emotions not only designate significance to certain norms, but, in case of a deviation from the norm, they also serve as an indicator of how bad the violation actually is. As Mercer (1996: 23) argues: ‘One way to test for the presence of norms is to look for emotion.’ Emotions can give reassurance that norms have been internalized by members of the group, which serves as a commitment to the community’s shared values and beliefs. As Bleiker and Hutchison (2014: 501) note, ‘emotions become intersubjective when they relate to something social that people care about.’ Second, and in addition to these regulatory functions, emotions serve important constitutive roles. Emotional expressions can and often do vary significantly across groups and these variations are indicative of culturally local conceptions of emotions, including their meaningful expression (Abu-Lughod, 1986; Lutz, 1986). Emotions are culturally constructed and institutionalized in the sense that communities provide a ‘structure of feeling’ through social conventions that constrains and compels members’ affective experience in order to facilitate group cohesion (Williams, 1961: 47). Many communities are composed of emotion norms or feeling rules that set the frame for appropriate interpretations and meanings of emotional performance among members of a particular group and thus incorporate sociocultural standards into the emotional lives of agents. Feeling rules are collective standards about what to feel and how to express emotions in appropriate ways that are internalized through their invocation and active cultivation (Gordon, 1988; Hochschild, 1979). They provide intersubjective patterns of standardized emotional expressions that constitute particular communities or cultures by setting them apart from others.

To sum up, social constructivist emotion theories highlight the particular sociocultural spaces in which emotions are contextualized, the institutional and discursive mechanisms through which they are expressed, and the prescriptive and purposive social functions they serve. From this viewpoint, the concept of intersubjectivity hinges on collective internalization and shared practice of socially constructed feeling rules: rules that govern how agents should feel in appropriate ways. As I hope to show in the next section, such an understanding provides a conceptual key to unlock the emotional underpinnings of liberal intersubjectivity in the social construction of the democratic peace.

The emotional underpinnings of liberal intersubjectivity

Whereas biological and cognitivist theories of emotions stick to a subjective ontology of emotion, social constructivist emotion theories shift the analytical focus from their internal phenomenological perception and psychological appraisal by individuals to their representative articulation and communication within social spheres. Social constructivists are less interested in the inner feelings and thoughts of individuals but instead focus on the socially shared emotional patterns between individual actors or groups. In other words, constructivist emotion scholars employ a social ontology of emotions that is less concerned with investigating the subjectivity but rather the intersubjectivity of emotions. From this social ontology of emotions follows a social epistemology. A social epistemology of emotions states that emotions can only be fully grasped within the collectively shared meaning systems and social worlds in which they are represented and known. Emotions are not natural objects of inquiry but unfold in a world already occupied by systems of meanings, discourses, and beliefs while simultaneously impacting on and shaping these meanings, discourses, and beliefs. Emotions are the product of sociocultural knowledge, for it is through systems of meanings that emotions can become ‘what they are.’ At the same time, emotions produce knowledge about the world by contributing to the construction of collective meanings, status hierarchies, and the exercise of power at the international level.

Building on these insights, I suggest that, rather than being solely grounded in cognitive facts, liberal intersubjectivity is simultaneously rooted in the affective relationship between liberal selves based on collectively shared feeling rules: the evaluative moral content and the dynamic intersubjective bonds of emotional exchange and reciprocity through which mutual recognition among liberals takes place. It is, to a large part, through the appraisal and expression of socially appropriate emotions that liberal agents come to recognize who are the ‘liberals’ and who are the ‘non-liberals.’ In other words, the self-conception and collective identity of liberal communities made up of liberal selves is both cognitively and emotionally constructed.

In this final section, I develop a tentative framework for the role that emotions play in underpinning liberal intersubjectivity based on the conceptual groundwork outlined above. The assumptions raised here are neither meant to be exclusive nor conclusive but should rather be viewed as theoretical building blocks that are open to further empirical exploration. As previously pointed out, my goal is to expand and elaborate on the constructivist research framework for studying the democratic peace to more systematically include the role of emotions and to conceptualize its underpinning mechanisms. In particular, my framework addresses two central constructivist assumptions concerning the democratic peace. First, it suggests that part of what, according to constructivism, constitutes liberal intersubjectivity, namely mutual recognition among liberal subjects, is underpinned by the collective framing and sharing of socially constructed ‘liberal’ emotions, based on the emotion norm of amity. Second, liberal insiders cultivate different emotions in relation to non-liberal outsiders, based on the emotion norm of enmity. Such a framework for liberal intersubjectivity significantly broadens and extends the constructivist argument pertaining to the democratic peace as one that is not solely based on liberal intersubjectivity but one that is, at the same time, deeply rooted in liberal emotional intersubjectivity.

The liberal emotion norm of amity

Constructivist assumptions on the pacifying role of liberalism posit that democracies exercise peaceful restraint toward other democracies based on mutual recognition, thereby establishing a separate peace among them. The emotional underpinnings of this assumption rest on two interrelated claims, derived from the constructivist argument cited above that mutual recognition is rooted in shared moral attitudes and cognitive perceptions.

First, emotions underpin liberal intersubjectivity based on what I call the attitude claim: the claim that socially constructed ‘liberal’ emotions help liberal agents identify with each other as one liberal family. Collective emotional meanings are not value-free as people enter relationships with a considerable package of worldviews and biases. Liberal subjects do not simply express and share emotions freely, but they express and share the right emotions as a group. As Levy (1984: 254) has argued, certain emotions are ‘the objects of considerable attention and knowledge … [and are] richly expressed’, while other emotions, by contrast, receive little attention or are excluded. Put differently, liberals not only attach meaning to the emotions they share, but they link emotions with moral meaning in order to orient fellow liberals toward the values and beliefs that matter for them. To the extent that there is some pattern to intragroup emotional attachment to key values and beliefs driven by shared relational concerns, this emotional pattern will come to be seen as normative, thereby reinforcing the community. In other words, for abstract liberal norms to become meaningful, liberal agents arguably need to affectively invest in these norms. As Alexis de Tocqueville’s ([1840] 2002: 613) notes in his chapter on The Influence of Democracy on the Feelings of the Americans: ‘The first and most intense passion which is engendered by the equality of conditions is, I need hardly say, the love of that same equality.’ I am not suggesting here to equate democracy with American democracy nor do I wish to imply that the same emotional expressions can necessarily be found in all democracies around the world. The example is simply meant to illustrate a theoretical point: socially constructed ‘liberal’ emotions hold the power to align the attitudes of liberal selves with the liberal norm of equality, thereby facilitating collective identification. Emotions are neither inherently moral nor beyond the scope of moral consideration, but they are deeply rooted in the collective values and beliefs of liberal communities. By tracing such emotions, we can learn a lot about the values and moral meanings of these communities.

Second, emotions play an important role in the social construction of liberal intersubjectivity based on what I refer to as the perceptual claim: the claim that ‘liberal’ emotions are perceived as an appropriate mode of expression through which liberal selves recognize each other as liberals. If collective identification, and the norms and institutions that are part of the liberal self, lie at the heart of how liberal subjects recognize each other, then emotions clearly underpin and shape these processes. This is because liberal intersubjectivity, as pointed out above, includes a shared commitment to the ‘liberal’ emotion norm of amity: the expression of appropriate emotions in a given situation that reinforces moral meaning, and ultimately, solidarity and trust. Without the ability to emotionally connect with each other, mutual understanding as the basis of trust among liberal subjects would remain sketchy at best (Lazarus, 1991: 287). To give an example, when the US faced a collective trauma on September 11, 2001 North Atlantic Treaty Organization (NATO) members collectively expressed sympathy, thereby creating a ‘we-feeling.’ Some 200,000 people took the streets of Berlin to show their solidarity with the US. In Britain, the US National Anthem was played during the change of guard in front of Buckingham Palace. In such a liberal security community, members will be inclined to share their emotions with each other simply because they trust each other (Adler and Barnett, 1998). Hence, in the event of an outside attack against a community member, all members expect each other to react with the appropriate emotional expressions (sympathy with the ‘attacked’/anger at the ‘attacker’).

Michel’s (2013: 886) conception of trust is helpful here. He distinguishes between trust as an emotive disposition, which precedes cooperative behavior, on the one hand, and strategic trust, on the other hand (which he calls ‘reliance’), which follows from cooperative behavior. A similar account is presented by Booth and Wheeler (2008: 229) who distinguish between functional ‘trust-as-predictability’ and emotional ‘trust-as-bond.’ The latter conception of trust represents an emotive and moralistic disposition: ‘Trust emerges here as a moralistic disposition which guides and influences behaviour by structuring our engagement with the world’ (Michel, 2013: 886). This conception of trust-as-bond mirrors the main argument about the binding role of emotions inside liberal communities put forward here. Since trust-as-bond is based on normative rather than strategic cooperation, harming strategic trust (reliance) will result in mere disappointment without questioning the meaning of action. The loss of emotive trust, on the contrary, will generate feelings of betrayal, which shake the foundations of liberal communities. A perceived betrayal by members of liberal communities would thus result in much deeper and intense emotions (for example, anger or revenge) than feelings of disagreement or disappointment. In other words, the link between trust and the idea of an emotional connection among liberal agents underpins mutual perceptions of moral integrity. This mutual perception is expressed and reinforced through the empathetic sharing of the emotion norm of amity.

To sum up, liberals cultivate socially constructed ‘liberal’ emotions that emphasize cosmopolitan equality with other liberals, thereby creating a shared sentiment of liberal equals, pacified and united in an emotional community of liberals. Such an understanding of liberal intersubjectivity embodies collective emotional attachment to liberal norms (attitude claim). Moreover, it allows liberals to recognize each other through the expression of such emotions, emphasizing the emotion norm of amity (perceptual claim).

The emotion norm of amity serves two main functions. First, it strengthens a moral hierarchy among liberals by assigning emotional meaning to the liberal norms that are important and whose violation can have grave consequences. Moreover, emotions do not only designate significance to certain liberal norms but, in case of norm breaking, they also provide an indication of the degree of norm violation. A fundamental violation of liberal norms will then trigger a much more intense emotional reaction than a minor gaffe.

Second, the emotion norm of amity gives reassurance that liberal norms have been internalized by all members of the liberal community, which underpins the commitment and confirmation of perceived ‘likeness’, ‘as a result of which they do not feel potentially threatened’ (Risse-Kappen, 1995: 509). Studies in social psychology support this assumption by showing that sympathy and empathy occur more likely when they correlate with real or perceived similarities between groups (Tajfel, 1981; Wispé, 1991). By repeatedly expressing appropriate emotions of amity, liberal actors reassure each other that their liberal norms and way of life are unquestioned, that the world is as it should be. Conversely, the expression of inappropriate emotions, such as morally ungrounded hostility directed toward liberal insiders, alert members that their shared world is out of bounce, which should trigger an emotional response to correct the imbalance.

I should like to add two caveats, however. First, there is a significant element of power and hierarchy involved. Inside liberal communities, members are not always treated as approximate equals but are woven together in asymmetrical power relationships. The self-image of liberal communities is often formed based on the minority of its ‘best’ members (for example, the US within NATO). Members can then only participate in the liberal community by internalizing and complying with certain emotional patterns of affect control. Members who fail to comply, for example by showing sympathy toward non-liberal outsiders, will risk losing their status rank within the liberal community.

Second, conflicts between liberals certainly do arise (Widmaier, 2005). What remains important is that, in resolving their conflicts, members follow the script of previously agreed feeling rules; for example, when and how anger may be an acceptable form of emotional expression, which should allow them to resolve disagreements in a peaceful manner. Inside liberal communities, emotions are managed by its members in a way that enables them to share and align their emotional expressions in appropriate ways. For example, sympathy operates to increase the level of intersubjectivity while the control of anger inhibits aggression toward likeminded others.

The liberal emotion norm of enmity

The second constructivist assumption on the democratic peace has been characterized by a well-known moral ambiguity: even though liberal norms seem to promote peace among democratic states, these same norms may also generate violent conflicts between democratic and non-democratic states: ‘Democratic peace – warlike democracies’ (Risse-Kappen, 1995; see also: Müller et al., 2013; Reiter and Stam, 1998).

To conceptualize the emotional underpinnings of how liberals recognize non-liberals, and thereby construct their non-liberal other, I will proceed again with the ‘attitude claim’ and the ‘perceptual claim’, respectively. To begin with the attitude claim, the defense and promotion of liberal norms against socially constructed ‘aggressive non-democratic states’ implies a strong emotional element of contempt and hostility in terms of othering (Wendt, 1999: 226–227). Analyzing the way emotional expressions such as anger refer to the other in discourse is to search for the construction of chains of connotations between words or pairs of concepts and their emotional meaning related to identity and moral attitudes, which are often reproduced via polarizing speech acts. As Campbell (1992: 89) suggests,

the ‘barbarian’ invoked connotations … energized by moral concerns …; these moral concerns naturalize the self … by estranging the other …. Each has its own emotional valence … the combined valuations of which constitute a position of being occupied by any one of a number of identities.

This notion suggests that the emotional attachment to liberal norms explained above appears to be accompanied by a corresponding emotional detachment from non-liberal norms. In other words, the more liberal subjects attach emotional significance to socially constructed liberal norms as part of their ‘way of life’ (and thus align their moral attitudes), the more they emotionally reject the norms of their non-liberal other because the latter’s ‘evil (and) hostile attitudes’ (Kant, [1795] 1949: 453) threaten their liberal way of life: sympathy with the ‘attacked’/anger at the ‘attacker.’

Second, in terms of perception, there appears to exist an underlying fear among liberal subjects of what non-liberals might do to them, as well as prevalent anger against what liberals perceive authoritarian states are doing to others as ‘public violators of human rights’ and ‘enemies of free enterprise’ (Doyle, 1983: 320, 322). In the eyes of the liberal self, the non-liberal other breaches the lower boundary between humanity and animality by degrading itself given how it treats others. This reinforces a point raised by Risse-Kappen (1995: 509) that ‘this (liberal) perception then creates a security dilemma leading to behavioural patterns that confirm the presumption of enmity.’ Such a perception, projected onto a non-liberal other, arguably evokes a concerted emotional response on the outside that, in turn, further stabilizes the democratic peace on the inside (Müller, 2014). Emotion thus constitutes a vehicle for representing specific perceptions that liberals hold toward non-liberals:

Understanding of difference is not just a cognitive process, particularly when it is built on a foundation of threat and challenge – which in turn are built on emotional scaffolds, such as fear. States can thus only identify themselves as themselves when they feel that other states are different. (Sasley, 2011: 466, emphasis in the original)

In short, the emotion norm of enmity underpins and sustains the hostile attitudes of liberals toward non-liberals as well as liberal perceptions of intentional and unjustifiable moral transgressions by non-liberals, which both allow liberals to recognize and construct their non-liberal other.

Again, this involves power dynamics and international hierarchies. The emotion norm of enmity forms part of the asymmetries of power and status between liberals and non-liberals, in which liberals are willing to secure their way of life vis-a-vis non-liberal others, if necessary through force. Emotional meanings are used to justify the exclusion of non-liberals from positions of power and responsibility and to legitimize their disadvantaged moral position. Accordingly, liberals maintain and reproduce a particular self-image of moral superiority vis-a-vis non-liberal states, which is undergirded by emotional meanings.

To sum up, liberal subjects seem to inhabit two emotional worlds: one in which sympathy is cultivated to stabilize liberal communities on the inside (amity), and one in which the social construction of ‘non-liberal’ others simultaneously nurtures antipathy on the outside (enmity). Of course, none of this is to say that antipathy does not exist in and among liberal societies. It clearly does (Dean, 2002; Young, 2003). What is important to note, however, is that, at least according to Democratic Peace Theory (DPT), it should not exist. From a constructivist theoretical perspective, emotions of enmity are supposed to be devalued and discouraged among liberals in favor of the cultivation of the socially constructed ‘liberal’ emotion norm of amity.

Conclusion

This article explored the emotional underpinnings of liberal intersubjectivity in the social construction of the democratic peace. It was argued that the way liberal democracies recognize each other has to do with the high degree of emotional convergence among them. Liberal intersubjectivity is constructed not solely via cognitive perceptions and moral attitudes, but also via the construction of socially appropriate collective emotions. The emotion norm of amity assures that liberal selves gravitate toward one and feel comfortable and reassured in one’s presence as morally likeminded selves: an intersubjective self-validation. Conversely, the emotion norm of enmity facilitates the construction of non-liberal others, generating hostility and antipathy between liberals and non-liberals. In sum, liberal intersubjectivity can be said to be intimately linked to the affective quality of its relationships, based on the social construction of collectively standardized emotional meanings and expressions. Emotions help liberals to make sense of their self and how they are situated in relation to others.

These findings contribute to several theoretical debates in the field of IR. First, there is an ongoing theoretical debate about the nature, role, and significance of emotions in world politics. Whereas macro approaches emphasize the need for abstraction and generalizable theories of how particular emotions and politics intersect, micro approaches highlight the need for exploring how particular emotions acquire meaning and function within specific cultural and political contexts (Bleiker and Hutchison, 2014: 497). While these approaches may appear incompatible, I argue that combining them offers ideal opportunities to more fully grasp the crucial link between individual and collective emotional meanings in world politics, thereby bridging the aforementioned liberal tension between individual liberal subjects and the political communities that make their liberal subjectivity possible. As Bleiker and Hutchison (2014: 493) attest:

We see these poles between macro and micro approaches as neither fixed nor mutually exclusive. Indeed, a combination of them – through a focus on the links between individual and collective emotions – offers great opportunities to bring out the best from both traditions and to carve out a promising way forward.

This also contributes to a burgeoning literature that theorizes the link between power and emotions in IR (Hutchison, 2016; Koschut, 2017; Solomon, 2014, 2015). As the present study underscores, emotions and their appropriate expression function to position subjects within relations of power and are mobilized by social actors to include or exclude subjects from entering the boundaries of liberal communities.

Second, the article speaks to recent theoretical debates on the state and nature of constructivism itself. One of the discontents with constructivism concerns its relationship with rationalism. As pointed out above, conventional accounts of the social construction of the democratic peace have pursued a research agenda that is complimentary to rationalism and thus remains skeptical to fully incorporate emotions into its research frameworks. As a result, this body of constructivism has figured less in constructivist debates about emotions than critical and post-structural variations that take emotions seriously. By showing how the social construction of the democratic peace is undergirded by emotions, this article points to the need to rethink conventional constructivist conceptions of liberal rationality.

In the end, this suggests a rich agenda for empirical research. For example, collective sympathy potentially undergirds a democratic peace in liberal security communities such as NATO and the European Union. Conversely, collective anger toward autocratic rule may have facilitated military interventions by NATO states in Libya and Afghanistan. Crucially too, I suggest that the theoretical findings shed light on the role of emotions as conduits of political change, which, given the recent rise of populist and chauvinistic social movements and political parties in Western liberal democracies and elsewhere, begs for more attention.

Footnotes

Funding

This research has been funded by the German Research Foundation (DFG) as part of the research network Constructivist Emotion Research (KO 4078/3-1).

Notes

Author biography

Simon Koschut is Visiting Professor in International Relations at the Otto Suhr Institute at the Free University Berlin. Previously, he held positions at the Weatherhead Center for International Affairs at Harvard University and the University of Erlangen-Nuremberg. His work appeared in journals such as Review of International Studies, International Studies Review, Millennium, and Cooperation and Conflict. He recently edited a forum on ‘Discourse and Emotions in International Relations’ in the International Studies Review (2017).

References

Abu-Lughod

(1986) Veiled Sentiments: Honor and Poetry in a Bedouin Society. Berkeley, CA: University of California Press.

Adler

(1997) Seizing the middle ground: Constructivism in world politics. European Journal of International Relations 3(3): 319–363.

Adler

Barnett

(1998) Security Communities. Cambridge: Cambridge University Press.

Ahmed

(2003) The Cultural Politics of Emotion. Edinburgh: Edinburgh University Press.

Averill

(1980) A constructivist view of emotion. In: Plutchik

Kellerman

(eds) Emotion: Theory, Research and Experience, vol. 1. New York: Academic Press, pp.305–339.

Berger

Luckmann

(1966) The Social Construction of Reality: A Treatise in the Sociology of Knowledge. New York: Doubleday.

Bially Mattern

(2014) On being convinced: An emotional epistemology of international relations. International Theory 6(3): 589–594.

Bleiker

Hutchison

(2014) Theorizing emotions in world politics. International Theory 6(3): 491–514.

Booth

Wheeler

(2008) The Security Dilemma: Fear, Cooperation and Trust in World Politics. Houndmills: Palgrave Macmillan.

10.

Campbell

(1992) Writing Security: United States Foreign Policy and the Politics of Identity. Manchester: Manchester University Press.

11.

Coulter

(1979) The Social Construction of Mind. London: Macmillan.

12.

Crawford

(2014) Institutionalizing passion in world politics: Fear and empathy. International Theory 6(3): 535–557.

13.

De Tocqueville

([1840] 2002) Democracy in America, vol. 2. Chicago: University of Chicago Press.

14.

Dean

(2002) Liberal government and authoritarianism. Economy and Society 31(1): 37–61.

15.

Dixon

Senese

(2002) Democracy, disputes, and negotiated settlements. Journal of Conflict Resolution 46(4): 547–571.

16.

Doyle

(1983) Kant, liberal legacies, and foreign affairs. Part 2. Philosophy & Public Affairs 12(4): 323–353.

17.

Farnham

(2003) The theory of democratic peace and threat perception. International Studies Quarterly 47(3): 395–415.

18.

Fattah

Fierke

(2009) A clash of emotions: The politics of humiliation and political violence in the Middle East. European Journal of International Relations 15(1): 67–93.

19.

Gordon

(1988) Changing emotion norms in marriage: Love and anger in US women’s magazines since 1900. Gender & Society 2(3): 308–342.

20.

Hall

(2012) Sympathetic states: Explaining the Russian and Chinese responses to September 11. Political Science Quarterly 127(3): 369–400.

21.

Harré

(ed) (1986) The Social Construction of Emotions. Oxford: Blackwell.

22.

Hayes

(2011) The democratic peace and the new evolution of an old idea. European Journal of International Relations 18(4): 767–791.

23.

Hayes

(2012) Securitization, social identity, and democratic security: Nixon, India, and the ties that bind. International Organization 66(1): 63–93.

24.

Hermann

Kegley

(1995) Rethinking democracy and international peace: Perspectives from political psychology. International Studies Quarterly 39(4): 511–533.

25.

Hobson

(2011) Towards a critical theory of democratic peace. Review of International Studies 37(4): 1903–1922.

26.

Hochschild

(1979) Emotion work, feeling rules and social structure. American Journal of Sociology 85(3): 551–575.

27.

Hutchison

(2016) Affective Communities in World Politics. Cambridge: Cambridge University Press.

28.

Ish-Shalom

(2006) Theory as a hermeneutical mechanism: The democratic-peace thesis and the politics of democratization. European Journal of International Relations 12(4): 565–598.

29.

Jeffery

(2014) Reason and Emotion in International Ethics. Cambridge: Cambridge University Press.

30.

Kahl

(1999) Constructing a separate peace: Constructivism, collective liberal identity, and democratic peace. Security Studies 8(2): 94–144.

31.

Kant

([1795] 1949) Perpetual peace. In: Friedrich

(ed) The Philosophy of Kant. New York: Modern Library, pp.449–459.

32.

Kirkpatrick

White

(1985) Exploring ethnopsycholgies. In: White

Kirkpatrick

(eds) Person, Self, and Experience: Exploring Pacific Ethnopsychologies. Berkeley, CA: University of California Press, pp.3–34.

33.

Koschut

(2017) The power of (emotion) words: On the importance of emotions for Discourse Analysis in IR. Journal of International Relations and Development (Published online ahead of print).

34.

Lazarus

(1991) Emotion and Adaptation. Oxford: Oxford University Press.

35.

Lebow

(2005) Reason, emotion, and cooperation. International Politics 42(3): 283–313.

36.

Levy

(1984) Emotion, knowing, and culture. In: Shweder

LeVine

(eds) Culture Theory: Essays on Mind, Self, and Emotion. Cambridge: Cambridge University Press, pp.214–237.

37.

Lutz

(1988) Unnatural Emotions: Everyday Sentiments on a Micronesian Atoll and Their Challenge to Western Theory. Chicago, IL: University of Chicago Press.

38.

MacMillan

(2004) Liberalism and the democratic peace. Review of International Studies 30(2): 179–200.

39.

Mercer

(1996) Approaching emotion in international politics. Paper Presented at the International Studies Association conference, San Diego, CA, 25 April 1996.

40.

Mercer

(2010) Emotional beliefs. International Organization 64(1): 1–31.

41.

Michel

(2013) Time to get emotional: Phronetic reflections on the concept of trust in international relations. European Journal of International Relations 19(4): 869–890.

42.

Müller

(2014) Evilization in liberal discourse: From Kant’s ‘Unjust Enemy’ to today’s ‘Rogue State’. International Politics 51(4): 475–491.

43.

Müller

Geis

Schörnig

(eds) (2013) The Militant Face of Democracy. Cambridge: Cambridge University Press.

44.

Nussbaum

(2001) Upheavals of Thought. Cambridge: Cambridge University Press.

45.

Oren

(1995) The subjectivity of the ‘Democratic’ peace: Changing US perceptions of imperial Germany. International Security 20(2): 147–184.

46.

Oren

(2003) Our Enemies and US. Ithaca, NY: Cornell University Press.

47.

Owen

(1994) How liberalism produces democratic peace. International Security 19(2): 87–125.

48.

Peceny

(1997) A constructivist interpretation of the liberal peace: The ambiguous case of the Spanish–American War. Journal of Peace Research 34(4): 415–430.

49.

Reiter

Stam

(1998) Democracy, war initiation and victory. American Political Science Review 92(2): 377–389.

50.

Risse-Kappen

(1995) Democratic peace – Warlike democracies? A social constructivist interpretation of the liberal argument. European Journal of International Relations 1(4): 491–517.

51.

Rosaldo

(1980) Knowledge and Passion. Cambridge: Cambridge University Press.

52.

Ross

(2006) Coming in from the cold: Constructivism and emotions. European Journal of International Relations 12(2): 205–206.

53.

Ross

(2014) Mixed Emotions: Beyond Fear and Hatred in International Conflict. Chicago, IL: Chicago University Press.

54.

Russett

Oneal

(2001) Triangulating Peace: Democracy, Interdependence, and International Organizations. New York: W.W. Norton & Company.

55.

Saarni

(1993) Socialization of emotion. In: Lewis

Haviland

(eds) Handbook of Emotions. New York: Guilford Press, pp.435–446.

56.

Sasley

(2011) Theorizing states’ emotions. International Studies Review 13(3): 452–476.

57.

Solomon

(2014) The affective underpinnings of soft power. European Journal of International Relations 20(3): 720–741.

58.

Solomon

(2015) The Politics of Subjectivity in American Foreign Policy Discourses. Ann Arbor, MI: University of Michigan Press.

59.

Steele

(2007) Liberal-idealism: A constructivist critique. International Studies Review 9(1): 23–52.

60.

Tajfel

(1981) Human Groups and Social Categories. Cambridge: Cambridge University Press.

61.

Turner

Stets

(2005) The Sociology of Emotions. Cambridge: Cambridge University Press.

62.

Weart

(1998) Never at War. Why Democracies Will Not Fight Each Other. New Haven, CT: Yale University Press.

63.

Wendt

(1999) Social Theory of International Politics. Cambridge: Cambridge University Press.

64.

Widmaier

(2005) The democratic peace is what states make of it: A constructivist analysis of the US–Indian ‘near-miss’ in the 1971 South Asian crisis. European Journal of International Relations 11(3): 431–455.

65.

Williams

(2001) The discipline of the democratic peace: Kant, liberalism and the social construction of security communities. European Journal of International Relations 7(4): 525–553.

66.

Williams

(1961) The Long Revolution. Westport, CT: Greenwood.

67.

Wispé

(1991) The Psychology of Sympathy. New York: Plenum.

68.

Wolf

(2011) Respect and disrespect in international politics: The significance of status recognition. International Theory 3(1): 105–142.

69.

Young

(2003) The logic of masculinist protection: Reflections on the current security state. Journal of Women in Culture and Society 29(1): 1–25.