The Dark Matter of World Politics: System Trust, Summits, and State Personhood

Abstract International relations theory has had a trust revival, with scholars focusing on how trust can enhance interpersonal cooperation attempts between leaders. We propose there is another type of trust at play in world politics. International system trust is a feeling of confidence in the international social order, which is indexed especially by trust in its central unit, state persons. System trust anchors ontological security, and its presence is an unstated assumption of the international relations trust scholarship. In this paper we conceptualize system trust. We illuminate its presence by flagging the production of state personhood in a familiar case in international relations trust scholarship, the 1985 Geneva Summit between Reagan and Gorbachev. Interpersonal and system trust perspectives highlight different aspects of the same summit. The juxtaposition suggests new lines of research into the production of state persons in diplomacy, the relationship between interpersonal and system trust, and the impact of the rise of personalistic/patrimonial leadership on diplomacy and international order.

But interpersonal trust is just the tip of the trust iceberg. IR trust scholarship treats trust among leaders as a psychological, interpersonal event that takes place (or not) in so-called realist anarchy, a harsh environment where the lack of enforceable supranational rules or authority makes trust scarce, fragile, and a heroic accomplishment.
We propose that there is another form of trust, which sits in the background of world politics, unseen but invisibly conditioning the meaning and effects of interpersonal trust. Following Anthony Giddens, Niklas Luhmann, and others, we call it system trust, trust in the continuity of the overarching, shared sociopolitical order. 4 In world politics, system trust is invested in the existence of an international system, an anarchy of states that seem to act like persons. 5 Such trust is a sociological phenomenon that is generally plentiful and durable in well-established systems and contributes to the production of order.
To get at the phenomenon of international system trust, consider some assumptions anchoring expectations at a presidential summit. Summit participants know that the United States, North Korea, Russia, and so on, are "states," which are simultaneously structures of political authority, charged with protecting citizens, and actors or "persons," capable of intentional action and human-like feelings and relationships, including trust. Summit staff know what to do, from managing entrances and exits to choreographing photo ops and signing ceremonies. Leaders know not to bring weapons or engage in violence. All know that the personal signatures of leaders commit their state.
Shifting from the horizontal perspective of participants to the vertical perspective of observers, as the summits' many audiences watch leaders laughing together and shaking hands, they see, simultaneously and without cognitive dissonance, two persons and two states. Audiences know that agreements reached-or not-are between two states. This fusion 6 of state and human person allows audiences to feel collective relief at the handshake and expect that summitry might de-escalate tension. Regardless of whether either leader personally feels trust, a summit, like other types of diplomacy, relies on those and other background interpretive certainties. Summits may result in interpersonal trust, or they may not. Either way, even at a summit, interstate trust is not merely an interpersonal thing.
In this paper, we argue that the salutary effects of interpersonal trust in world politics rely on the ongoing production of international system trust, especially trust in the existence of the basic unit of the international system, the state person. Sociologists of trust propose that trust in the unit indexes holistic system trust, often using the example of paper currency in a monetary system. Members of the system, from those in authoritative roles to consumers, must treat certain paper as currency, bestowing on it day to day, without question or reflection, the power to represent value. 7 Trust in state personhood is similar. Through anthropomorphizing the state, daily and without thinking, the multiple audiences of diplomacy simplify unimaginably complex relations into a familiar vernacular of interpersonal relations, making the world and their place in it comprehensible. This largely backgrounded "feeling of knowing" that the state is a person in a system of state persons is part of the usually stable backdrop of choices and relations, fostering individual ontological security.
Indeed, system trust operates much like "dark matter" in the universe, unobservable but invisibly conditioning patterns of action, even holding entire galaxies together. Like dark matter, if system trust is removed from a system, energy that is crucial to system continuity is lost. But unlike dark matter, which is physical, the largely social dark matter of system trust requires ongoing human participation to keep working. International system trust is (re)produced through the ongoing interplay of horizontal dynamics, as those in authoritative positions competently, recognizably perform their roles, and vertical dynamics, as performances and symbols register, consciously or subconsciously, with relevant audiences.
International system trust is reproduced through practices at many levels, from professionals and practitioners to lay publics. Pulling it from the background has two implications for the study of trust in world politics. First, system trust is causally relevant even if it, like other structural causes, does not yield point predictions. 8 That is, explanations of how interpersonal trust matters in world politics depend on contemporary anarchy's continual reproduction in ways extant IR trust scholarship cannot capture because it treats anarchy as an analytic prior rather than as a sociological production.
Second, little is known about the dynamics of mere interpersonal trust in world politics, that is, interpersonal trust under conditions of unstable or withdrawn system trust. This is important because many scholars have charted widespread declines in public trust in state institutions and bureaucracies, often linking them to political trends such as leaning on charismatic leaders, preferences for patrimonial rule, and populist politics. 9 Such trends in domestic rule delegitimate and/or thin out state institutions, possibly including those reproducing the state person, while potentially reverberating as personalistic diplomacy. We suggest that mere trust among leaders is not necessarily a mechanism for international order and may be destabilizing.
System trust is continuously reproduced in multiple locales and through a range of horizontal and vertical dynamics, some backgrounded and others salient. Presidential summitry is one such locale. This claim may be surprising, since summits are locales where interpersonal trust seems to shine, and decision makers often rationalize summits for this reason. But we propose that system trust in state personhood is central to summitry. To illuminate the dark matter of system trust, we revisit a 8. Waltz 1996. 9. Homolar 2019. Weeks 2012 isolates the dangers of personalistic rule; Hanson and Kopstein 2022 relate the loss of trust in institutions to a "patrimonial wave." On public trust erosion in the US and OECD, see Drezner 2021;Pew Research Center 2019;Smith and Son 2013. familiar case in IR trust scholarship, the 1985 Reagan-Gorbachev summit in Geneva. Nicholas Wheeler and Marcus Holmes, separately and together, argue that interpersonal trust between these two men, forged through summitry, helped end the Cold War. Our approach contextualizes these arguments, bringing the state person into view and showing how the state system was flagged and reproduced in the summit process. We infer from its presence the maintenance of system trust that contributed to public confidence in the shifting interstate relationship. A counterfactual implication is that had Geneva been solely a meeting of minds, its international political effects would have been at most transient and at worst destabilizing.
Because it produces the political meaning and effects of leaders' interpersonal trust, and because it is potentially under threat, international system trust warrants analytic attention. With the goal of complementing IR trust scholarship, the paper is structured as follows. After reviewing IR scholarship on trust, we introduce the concept of system trust, specifically in state personhood. We then place the two interpretations of the Geneva summit side by side. In conclusion, we reflect on the implications of personalistic diplomacy and person-centered research programs on trust in world politics.

Interpersonal Trust in International Relations
The study of trust at any level begins with the intuition that positive expectations about others' actions and intentions are the basis of social life. Acting with others requires simultaneously managing awareness of the uncertainty of the outcome of these actions and our vulnerability to others' whims and desires. Uncertainty is managed via trust, understood as a leap-of faith, cognition, or some combination -that enables a feeling of knowing others' future behavior. 10 The idea is that, paraphrasing Luhmann, because we cannot erase our vulnerability as we contemplate social action, we replace unattainable external certainty with attainable inner trust. 11 Trust makes action, and social life, possible.
IR scholarship on trust has developed in two streams focusing on two different subjects, or "trusters": persons (leaders, negotiators, elites) and states treated like persons, that is, as if they were intentional actors, capable of feelings and relationships. Within each stream, scholarship varies along several dimensions, and we cannot do justice to the vibrancy of this research program here. 12 What is important for our argument is that it begins from the same model of trust and its challenges in anarchy. Trust is treated as a micro-level, psychological phenomenon operating interpersonally and usually consciously, a mental state or personality trait that can facilitate cooperation. Such trust can be said to operate horizontally, reducing (the perception of) risk 10. All definitions of trust include the idea of action under uncertainty, along with the perception of vulnerability.
11. Luhmann 1979, 26-27. 12. Reviews of IR trust literature include Rathbun 2018;Ruzicka and Keating 2015. between decision makers in an interaction. Parties are aware of the difficulties of forging trust in anarchy and their role in growing it (or not). From this shared model, scholars unpack the trust mechanism in various ways. Andrew Kydd's influential rationalist account treats the state as a person-like unit of analysis and defines trust cognitively, as the belief that another state will reciprocate a cooperative action, which would make cooperation a rational choice. 13 The problem of trust is an information problem rooted in the problem of other minds: information about intentions (and thus the game one is playing) brings trust. That information can be acquired through costly signaling. The choice of whether to trust results from an assessment of the extent to which others' interests align with or "encapsulate" the chooser's interests. 14 That is, trust emerges from calculation as to whether it is wise to trust the other party, given uncertainty.
Security community scholars also treat interstate trust on an interpersonal and cognitive model. Deutsch developed security community formation as a process of growing trust, understood as learning to have benign expectations. 15 Adler and Barnett's agenda-setting conceptualization defines trust as cognitive certainty about intentions/actions, treating it as an indicator of collective identity in a pluralistic security community. 16 Not all trust scholars agree that states can be treated like persons or that trust is primarily cognitive. Many treat the individual as unit of analysis and take a psychological approach, bringing in emotions. For example, Deborah Larson analyzes Cold War dynamics, drawing out the emotions of regret and moral outrage associated with misplaced trust. 17 Jonathan Mercer characterizes trust as an "emotional belief" in the sincerity of others that makes the truster feel less vulnerable. 18 Brian Rathbun treats trust in dispositional terms. 19 Individuals might be generalized or particular trusters. Prosocial types are more likely to be susceptible to trust mechanisms. Some scholars conceptualize trust as a relationship rather than an interaction, insofar as it implies an obligation for the trustee to follow through. Aaron Hoffman conceptualizes trust as a relationship in which the truster confidently expects the trustee will do what is right. 20 Wheeler argues that trust is a binding relationship, as in his argument with Jan Ruzicka that the Nuclear Non-Proliferation Treaty persists because of the commitment to fellow signatories, which has become through habituation a thick social bond. 21 A focus on relationships turns attention from individual dispositions to trust-formation processes. This focus characterizes the arguments of, for example, Holmes, Wheeler, and others who argue 13. Kydd 2007. 14. Hardin 2006. 15. Deutsch 1958; see also Ruzicka and Keating 2015, 13-14. 16. Adler and Barnett 1998. 17. Larson 1997. 18. Mercer 2010compare Michel 2013, 813. 19. Rathbun 2014. 20. Hoffman 2002. 21. Ruzicka and Wheeler 2010 that speaking face to face enables parties to assess sincerity, connect empathically, and potentially develop lasting bonds; 22 or Marina Henke's argument that transactional modes of negotiation can undermine trust processes. 23 Despite the change in unit, this work retains the premise of state-centric accounts, that attempts at international trust take place between people in the uniquely dangerous environment of realist anarchy. Thus, having come down to the individual, dyadic, or small-group level to "find" the trust mechanism, this scholarship faces practical problems that seem to undermine the promise of trust as a stabilizing force in world politics. First, if international trust relies on dispositions, then trust in world politics is idiosyncratic: change the negotiator and the outcome changes. Successful cases of trust are nothing more than rare, even heroic interpersonal success stories. 24 Second, if trust relationships develop among leaders and negotiators as human persons, there is the problem of aggregation. As Rathbun points out, scaling up from leader to country is a tall order analytically and politically: "Is it possible to say that a country trusts another country or only that a leader of a particular country trusts another leader? If a leader comes to trust another leader, how could that relationship be more broadly institutionalized in a government or in a society?" 25 As Wheeler stresses, the challenges of "state trust" are of "embedding" elite trust in society and growing cultures of intersocietal trust. 26 In other words, for these scholars the fact that trust works through individuals (it is "ontologically individualist" as a phenomenon) translates to an "explanatory individualism," reducing the institutions these trusting humans are embedded in to ingredients to be added onto foundationally human processes, with anarchy as the institution-free environment. 27

System Trust in World Politics
Ironically, IR's "problem of trust in anarchy" relies on deeply held "trust in anarchy." This latter kind of trust is what we call international system trust. Sociologists have long posited that a distinct form of trust is necessary for sociopolitical systems to function. 28 System trust is trust in the "whole," the feeling of knowing that the set of basic institutions that shape everyday life, including the political, social, and moral systems, are as they are and will continue in their present form. That is, system trust is trust in an abstraction, not in a human person. 29 Lewis and Weigert 1985;Luhmann 1979. 29. Among IR scholars, Mercer 2014 and Michel 2013 highlight the distinctiveness of trust in abstractions, although neither treats the state system as a trust object. world politics, system trust is so heavily invested in the existence of a world of states that anarchy can function as an analytic prior in IR theory and as the unquestioned background of political practice. But international system trust cannot be taken for granted, and its decline or withdrawal affects the system, altering the dynamics and effects of interpersonal trust.
System trust differs from interpersonal trust in that it operates between members of a society/political system and the system itself. If interpersonal trust operates horizontally, between people, system trust can be said to operate "vertically," between persons and the systems they are embedded in. Verticality creates a particular vulnerability. The acceptance of vulnerability is at the heart of all trust relationships, but with interpersonal trust that vulnerability is symmetric: each party chooses to trust (or not). With system trust, the vulnerability is asymmetric, since the institutional order simply exists, at least in the medium run.
The verticality and asymmetry of system trust make it similar to another phenomenon analyzed in IR scholarship, institutional trust. Scholars of institutional trust, such as Lisa Maria Dellmuth and Jonas Tallberg, have examined the sources of public trust in international institutions such as the World Trade Organization or the European Union, to determine whether it is rooted in, for example, dispositions, culture, or the institution's performance. 30 The phenomenon they measure is similar to system trust, but there is an important difference. For Dellmuth and Tallberg, the object of trust-the institution-is presumed to continue to exist. The question is, what accounts for public trust in the institution, and to what extent is that trust dependent on the institution's performance?
System trust, in contrast, refers to trust in the existence of the institutional order, not the performance of a given institution within it: the existence of a market, not the performance of a particular firm or of the market overall; the existence of an anarchic system, not the performance of a given state or international organization. Performance matters: those in institutional roles must competently enact those roles and institutions must perform as they were set up to do. Without competent performances, system trust is vulnerable to decline. But analytically, the existence of a social and political system and the performance of institutions within it can be treated as distinct.
It is helpful to refer to Keynes, for whom system trust or confidence means that people in business roles are "oriented to the same object," that is, the economy as a whole. 31 The common assumption all rely on, that the economy is an object, is the product of system trust even though it feels objective. An "economy" seems to exist outside of their imagination, but without their practices, it would not. By trusting that institutional objects exist, members instantiate them as existing. If members stop 30. Dellmuth and Tallberg 2020. See also Harteveld, van der Meer, and De Vries 2013;Sattler and Urpelainen 2012;Voeten 2013;compare Alter and Meunier 2009. 31. Discussed in Barbalet 1996. backgrounding the social order, and instead constantly question and/or stop participating in its institutions, those institutions ultimately will break down. Sometimes, as with Luhmann, system trust is referred to relatively interchangeably with the term confidence and distinguished from the interpersonal trust necessary in modern risk society. Confidence is a feeling that things are what they seem. Luhmann treats confidence as a glue of premodern societies, which were characterized by deep interpersonal bonds within familiar systems of rules and sanctions. Such societies did not require what we think of as trust, in that they could function without any conscious leap of faith or cognition among members. Stability was assured through socialization, and alternative societal organizations did not cross members' minds. 32 In contrast, trust is the emotional orientation necessary for a society of strangers, where confidence cannot be relied on, as in industrialized and especially modern risk society. It is not realistic to expect interpersonal trust alone to carry the burden of social integration among strangers. That is accomplished largely through their relations to abstract systems like money, political power, and law. 33 Trust in the ongoing existence of those systems, as much as trust in one another, is crucial for maintaining social and political cohesion. 34 System trust is the confidence-like phenomenon for modern society, in all its complexity and populated by strangers. 35 Trust reduces complexity and, in Jack Barbalet's words, "overcomes the problem of an unknown future by emotionally committing persons to the risk of cooperative action." 36 In modern society, some actions and relations entail specific choices to trust, that is, to bet on the future contingent actions of others, trusting others to follow through on their commitments. In other cases, trust is the norm, such as when we rely on technological expertise. Despite the objective risk, we get on an airplane. 37 Most of the time, for most of us, the risks of flying do not cross our minds, although if pressed we could articulate them. System trust or confidence is similar. However, unlike the aviation system, which ultimately depends on machines and technology, the sociopolitical order is constituted and held in place by our ongoing participation in it even as it feels presocial, like the ground beneath our feet. 38 The naturalization of the order is important psychologically, as a platform for agency, and socially, for maintaining order.

A Taken-for-Granted Background
System trust is produced largely in the background of everyday life, not as the product of conscious choice. As Barbalet puts it, system trust operates "often below the threshold of awareness and therefore consciousness," as a "calm, unobtrusive" emotion lacking any particular behavioral, intellectual, or reflective commitment. 39 People do not "choose" to hold in mind that the international system exists or that states are actors. Rather, they (we) approach the international institutional order with a positive emotional orientation, expecting it to continue.
Other kinds of trust and trust-like phenomena also can operate in the background. Taking a psychological approach, many scholars of interpersonal trust propose that trusters' dispositions condition choices from the background. Generalized trust, or the propensity to trust other people, implies that the truster brings a tacit confidence to all interpersonal relations. Particularized trust, too, might work in the background. In a given encounter, trusters might not recognize the extent to which a category identity informs their feelings about a trustee's future action. 40 Many IR socialization arguments adopt a similar model, focusing on belief and norm internalization. Here, getting to trust is a process that begins with a self-conscious choice to trust, which over time becomes a backgrounded habit. 41 Two things stand out. First, these psychological phenomena-dispositions, beliefs, norms-occur in a world whose existence is treated as analytically prior to and independent of them. None of this work questions the objects constituting the shared social order, much less links the existence of those objects to beliefs and dispositions affecting interpersonal trust. Second, these phenomena are treated as affecting a choice to trust. With interpersonal trust, the confident expectation a trustee will follow through encourages cooperation where it otherwise might not happen. 42 Backgrounded phenomena feed into that choice, making it more or less likely.
With system trust, in contrast, what's backgrounded is not beliefs about others but the knowledge that the shared social order is a construct, a product of its members' social action. In this sense, system trust can be understood as a practice reflecting and reproducing tacit knowledge about the social world. 43 Paraphrasing Vincent Pouliot, we do not think about system trust, we think from it. This links system trust to Berger and Luckmann's understanding of socialization as "induction" into the social order, or the immersion in its social practices. 44 The resulting "knowhow" enables competent action, even as it operates as tacit, even inarticulable 39. Barbalet 1996Barbalet , 91, 2011. In IR, see Larson 2018. 40. Barbalet 2011, 39;Uslaner 2002. 41. Checkel 2005Finnemore and Sikkink 1998. 42. Although the relationship between trust and cooperation is not direct. 43. Adler and Pouliot 2011;Adler-Nissen 2014;Bicchi andBremberg 2016, 392-95. 44. Berger andLuckmann 1966, 135. In IR, see Adler-Nissen 2014;Epstein 2012. knowledge. Interpersonal trust is a conspicuous accomplishment; system trust is the inconspicuous status quo, at its most visible when it breaks down.
Pouliot conceptualizes the practice of interstate trust along these lines, as "knowhow" among security community practitioners. If practitioners did not trust in the continuity of the security community, they would find it difficult to build trust interpersonally. They are not thinking about trust. Trust is "forgotten as knowledge" and embedded in diplomatic practice. 45 Pouliot focuses solely on the horizontal dimension among practitioners. We extrapolate the idea of "trust as practice" to lay audiences. Embedded in laypersons' orientations to action is the practical sense that there is an international system of states. Such backgrounded "knowing" that the states system exists, and that states are person-like intentional and feeling actors, is part of being fluent in the practices of contemporary life.
There are two further senses in which we depart from Pouliot's approach. First, for Pouliot, practices, including trust, are ultimately cognitive. 46 In contrast, we follow IR scholars such as Janice Bially Mattern, who stresses the inseparability of the cognitive and the emotional in our practices, and Torsten Michel, who treats trust as an "emotive disposition" that "structures our inarticulate horizon of expectation." 47 To be ready for action is an emotional orientation, which all of the knowledge in the world cannot sum up to. 48 Second, as a feeling of knowing that the basic institutions that order our lives will endure, system trust is linked to the production of subjectivity. 49 It is therefore linked to ontological security. Being able to take the system for granted buffers the awareness not just of day-to-day risks and vulnerabilities, but also of the existential uncertainty lurking underneath. Thus, while the opposite of interpersonal trust is mistrust or distrust of an Other, the opposite of system trust is existential anxiety or dread, a "particular unease associated with the awareness of human mortality and the ultimate meaninglessness of human life." 50 Combining ontological security and trust-as-practice language, we could say that trust know-how is an emotional orientation of calm confidence toward the world, which is part of how we know who we are and how to "be" ourselves. 51 Roland Bleiker and Emma Hutchison suggest a similar link between institutions and the sense of self in their discussion of trust building for the post-conflict space. 52 John Cash argues that institutional cultures, by legitimating certain practices and the mentalities that sustain and support them, produce those very mentalities. 53 These links from institution to subjectivity connect with insights such as Neta 45. Pouliot 2008, 258, 278, 271. 46. Pouliot 2008. See also Koschut 2014, 544. On trust as an emotional practice, see Bleiker 2014. 47. Bially Mattern 2011;Michel 2013, 880. 48. Barbalet 1996, 77-78. 49. Duvall and Choudhury 2011. 50. Kinnvall and Mitzen 2020; see also Gustafsson and  Crawford's on the institutionalization of emotions in world politics, 54 suggesting such institutionalization is linked to a tacit abiding by of what Arlie Hochschild would call the "feeling rules" sustaining the social and institutional order. 55 The states system organizes the globe. The production of trust in its continuity entails emotional governance, a disciplining of individual dispositions through a set of feeling rules, rituals, and practices that maintain the calm confidence that the states system is the natural order of things. 56 This emotional governance organizes subjectivity at a macro level, although it remains separate from our individual sense of self. It also reproduces our communities or groups as constant social environments for action and sometimes as collective actors in their own right.

A Macro-Level Phenomenon
System trust is a macro-phenomenon not reducible to individual psychology or relationships. It is not independent of individual trust, insofar as it is a feeling of knowing, and feelings and motive states can only be properties of human persons. 57 From a psychological perspective, system trust refers to an internalized trust in the takenfor-grantedness of the social order that factors into one's "basic trust," the "cognitive cocoon" of confidence that contributes to ontological security and conditions intentional action and choice. This micro or "bottom-up" approach to system trust, treating it as a personality variable, is consistent with much of the IR scholarship on trust. 58 But while rooted in individuals, the sociological phenomenon of system trust is not an aggregation of individual psychological states or dispositions.
Closer to our understanding is Wheeler's argument that trust is relational, an emotional bond between truster and trustee irreducible to individual traits or dispositions. 59 Bonding trust is an "emergent property of face-to-face interpersonal dyadic interaction." 60 But while Wheeler treats the relational whole as separate from the sum of its psychological parts, like other psychological approaches he treats trust in particularistic terms, as a conspicuous phenomenon between specific people. Trust is analytically separable from trusters, but any trust relationship depends on those specific persons. If one truster changes their mind, the emergent effect disappears. System trust, in contrast, is nonspecific and durable. Not every member of a social system, and not any particular person(s) in it, must feel trust in order for system trust to obtain. We propose that system trust is best understood in constitutive terms, as a relation between a social whole and its parts. 61 Mercer draws on the concept of emergence, 54. Crawford 2014; see also Berezin 2002. 55. Hochschild 1979. 56. Berezin 2002Koschut 2019;compare Wiley 1983, 45. 57. Barbalet 2019. Many sociologists begin here, for example, Barbalet 1996;Sztompka 1999. 59. Wheeler 2018. 60. Holmes and Wheeler 2020. Lewis and Weigert 1985, 966. taking it beyond Wheeler's dyadic case, to derive state trust as based on but not analytically reducible to the individual trust states of its population. 62 Andrew R. Hom and Brent Steele interpret international anarchy along these lines, treating it as effectively an emergent effect of national narratives. 63 To the extent that each state's autobiography positions it in an environment of like actors, states instantiate the macro-structure of anarchy. The resulting international system is not a unitary or even harmonious vision, since state autobiographies are anything but coordinated. But with emergent structures, different constellations of preferences or desires at the micro level are consistent with a given macro-level state. 64 Alexander Wendt proposes that macro-level phenomena "supervene" over a range of motive or feeling states at the micro level, and that macro-phenomena are multiply realizable at the level of individual motive or belief. 65 Applied to international system trust, we could say that while system trust is rooted in individuals' "trust systems," not every individual needs to "trust" in a particular way or even to trust at all, in order for international system to be characterized by system trust. 66 To get at the macro dimension of system trust, it is helpful to turn again to a Keynesian analogy, this time to the phenomenon of a bank run, which manifests a withdrawal of system trust. The cause of a bank run is not necessarily that each individual has lost trust in the banking system-indeed it is unlikely that everyone would lose trust at the exact same time. But it is possible to say that the run was due to withdrawn system trust. The bottom-up and top-down levels of system trust are analytically distinguishable even if in practice they are entangled. Both are required to keep sociopolitical order realized as real for its members. In sum, system trust, like all social phenomena, is rooted in individuals and can be found at the micro level as a feeling of confidence in the continuity of daily life and the naturalness of the institutional order. But the explanatory power of international system trust is as a macro-level phenomenon. Even in an environment where interpersonal distrust can be rampant, such as anarchy, day-to-day social interaction relies on system trust in the basic ingredients of the shared social order. In world politics, trust is not just an interpersonal thing.

Indexed by the Central Unit: State Person
The idea that trust in a system is indexed by or realized through trust in its central unit appears repeatedly in sociological discussions of trust, especially drawing on the example of the monetary system, in which each person, each day, reproduces the shared idea that pieces of paper called "currency" represent value. Another example is that for a democratic system to function, each citizen, each day, must 62. Mercer 2014; see also Hall and Ross 2015. 63. Hom and Steele 2020. 64. Ejdus and Recevic 2021Ross 2015. 65. Wendt 1999, chapter 4. 66. Barbalet 1996, 77-78 reproduce the shared idea that pieces of paper called "ballots" represent votes that determine political representation. If members of society lose confidence that their vote does what they all assume it does, the system can keep running for a while. But under some conditions the result can be a run on the statehouse. Losing confidence in the constitutive relationship of the unit to the system calls into question the entire system.
We propose that the state person is an analogous unit for the international system, in that system continuity requires that on a global scale, laypersons and statespersons, each day, reproduce the shared idea that the state is an intentional actor on the world stage. This makes trust in state personhood an unrecognized source of order. Indeed, the state's legal personhood arose historically in part to buffer the system against the idiosyncrasies of leader personality and mood, in the service of systemic stability. 67 Centering state personhood means pulling it from the background of both theory and practice. Mainstream IR systemic theory routinely treats states as unitary corporate actors capable of acting with intentionality toward fellow states. The common sense that states are persons similarly anchors the English School concept of international society, in that its primary institutions, mutual recognition of sovereignty, international law, diplomacy, great power management, war, and the balance of power assume and/or tacitly depend on it. 68 Constructivist scholars have developed sovereignty as a dynamic institution constituting state actorhood; 69 and separately, scholars have debated whether the state is "really" a person. 70 But few scholars treat the personhood of sovereign agency as distinctly important, much less as a source of order. 71 In social/political practice, state personhood is similarly prevalent and similarly unnoticed. That states are people-like is widely taken for granted. In media imagery and reporting, states have feelings, and relationships, and hold grudges, and act with intention. But on reflection, the knee-jerk attribution of personhood to states is a huge accomplishment, a feat of imagination, that we all-from leaders, to diplomats, to lay audiences-engage in each day, regularly and without thinking. 72 Alisher Faizullaev describes this feat as a "phenomenological transformation": a mental leap from epistemically accessible material-our perceptions, experiences, and practices-to something invisible and epistemically inaccessible, namely, the state as a unitary and purposive person-like entity. 73 What people see and participate in as members of states are sets of practices and behaviors, complex and mundane, across many contexts. Each time we attribute such practices to the state we are 67. Holm andSending 2018, 832-36. 68. Buzan 2004 introduces the distinction between primary and secondary institutions, which grounds recent English School scholarship (e.g., Friedner Parrat 2017). Personhood as intentionality is not discussed.
69. Lyons andMastanduno 1995. 70. Jackson 2004. 71. But see Holm andSending 2018. 72. Faizullaev 2007;compare Wendt 2004, 289. 73. Faizullaev 2007 using our imagination to bridge a vast gap and consolidate these as the intentional agency of a unitary entity that appears diplomatically and in international law as a corporate person, "the state." 74 In that feat, we are aided by the quotidian symbols and routines of collective life, as well as precise protocols and ritualized performances on many scales, including the international. Each acceptance of the state person, whether consciously or subconsciously, by not questioning it, manifests trust in the state's continued personhood. It is a huge leap, but it is not felt that way. We just do it: we fuse disparate signs and traces into a state person. Faizullaev's conceptualization links up with experimental work in social psychology on state "entitativity," the apparent thing-ness of a social process or idea, supporting the link between state personification and trust in the continuity of the social world. 75 Scholars have found that when people see the state as a real entity, they are more likely to attribute intentionality to it and to perceive it as an actor on the world stage. 76 This, in turn, enhances confidence in the state as protector, that is, its ability to represent and defend its inhabitants. 77 The process is not a conscious one. Entitativity works through forgetting the fact that it is a construct. In other words, reifying the state in this way has psychological advantages, facilitating trust in it as a source of identification and locus of action, and confidence in the continuation of social and political life. The fact that states are not just "entities" but a specific kind of entity we are especially familiar with-a person-allows each of us to make inferences about state action and to understand the international political threat environment in ways we otherwise would not be able to. A world of vulnerabilities that is difficult to articulate or comprehend gets translated into a world managed by human persons through frameworks we already understand from interpersonal life, like friendship and enmity.
By stabilizing and making comprehensible the complex, dangerous realm of world politics, entitativity contributes to ontological security. 78 For the average person, international politics is far away, while the institutions in which people actively invest (dis)trust-Congress, Parliament, the presidency-are local. State personhood enables people to bridge the gap between the everyday and the international. The fact that the state "is" a person-like actor capable of feelings and relationships provides a means of identification; and it contributes to a sense of belonging and trust in the state as problem solver and protector. Also, the fact that the state is corporate and not human allows those qualities to transcend any individual leader and continue in time.
From here, the state's entitativity as a person could be said to supervene over the social system. How we each personify the state may be idiosyncratic. But those micro-level imaginative leaps and inferences occur in the context of authoritative 74. Sacchi, Castano, and Brauer 2009. 75. Ibid. compare Mercer 2014. 76. McGraw and Dolan 2007. Sacchi, Castano, andBrauer 2009. 78. Faizullaev 2007, 535. This claim echoes Luhmann 1988, 95: "We can live within a familiar world because we can, using symbols, reintroduce the unfamiliar into the familiar." system-level narratives, routines, rituals, and symbols reproducing and giving life to a world of state persons on an anarchical world stage. That is, while the state person coheres through myriad individual processes, it does so disciplined by the supportive discourses and cultural norms of a shared symbolic order, often flagged by background symbols and practices. 79 The continuing existence of state persons thus operates as a taken-for-granted assumption of daily life.
Erik Ringmar uses dramaturgical theory to make a complementary argument, focusing on the impact of diplomatic performances on lay audiences. The mechanisms of identification found in the theater are at work in diplomacy. As he puts it, a theatrical production "conjures up a certain atmosphere," to which audience members attune themselves. Actors' performances activate audience's own experiences of the expressed emotional states, so that audience members understand the action even before cognitively interpreting it. 80 The preconscious connection aids the cognitive interpretation of what's going on, helping audiences imagine, reflect on, and understand their place in the world. 81 Similar to entitativity scholarship, Ringmar's point is that diplomacy's many audiences participate in world politics through such identifications, which enable people to process information about world politics the way they process it about one another. This vastly simplifies the individual quest for protection and security, making it possible to feel more at home in the world.
In sum, we propose that international system trust exists and is maintained through an ongoing combination of the widely shared, taken-for-granted background expectations, first, that the order will continue, especially manifest in quotidian practices that reproduce it, and second, that those in institutional and symbolic roles, such as statespersons and diplomats, will competently perform and follow through on obligations. The resulting picture of international system trust is analogous to that painted by Bernard Barber at the domestic level, where the background and performance aspects of system trust dynamically interact, amid external stressors and the imperfect performances of those in institutional roles, to produce order. 82

International System Trust as Dark Matter
We have proposed that in world politics, the dynamics of interpersonal trust unfold in the context of taken-for-granted trust in the international system, especially the institution of state personhood. Interpersonal trust formation among leaders/negotiators is therefore less a micro-level story than a micro-instantiation of macro-processes. 83 This makes international system trust the dark matter of contemporary world politics, 79. McNamara 2015. 80. Ringmar 2016, 114. 81. Berezin 2002Ringmar 2016, 115. 82. Barber 1983. 83. Jepperson and Meyer 2011 a background condition so unobtrusive that it seems not to exist, but whose decline or loss would have profound implications.
The distinction between outcomes that system trust versus interpersonal trust can account for is similar to Waltz's distinction between theories of foreign policy versus international politics. 84 Forces operating at the level of interaction, from leaders' dispositions to domestic political dynamics, and so on, account for outcomes in decision situations. Applied to IR trust scholarship, if the question is why a summit takes place, or why a particular relationship changes from mistrust to trust, then for any given relationship, answers will be found at the level of interaction. Structural forces, on the other hand, shape and shove social action, setting the parameters of likely outcomes. They offer probabilistic assessments of what can happen. As a structural phenomenon, system trust is not a competing cause of outcomes that IR trust scholars examine. Rather, its existence raises new questions, such as the extent to which interpersonal trust is stabilizing or destabilizing, or the meaning of interpersonal trust in a given situation, and/or what would constitute transgressive trust behavior.
As for how these two forms of trust relate, Barber is once again helpful. He discusses the relationship between interpersonal trust and trust in the institutional order at the domestic level. 85 As he notes, a posture of active distrust toward others-while stressful-implies an underlying trust in the social system/order. Indeed, revealing an individual to be untrustworthy is possible only given shared understandings and a shared social system where the idea of "untrustworthy" behavior makes sense. Democratic systems work best when citizens trust the system but distrust power, that is, when they maintain a presumption of distrust of those in symbolic, authoritative roles. Too much trust in a given powerholder can undermine democratic mechanisms of accountability, and thus trust in the system. System trust also buffers individuals from the effects of misplaced interpersonal trust. A trusted institutional order can withstand untrustworthy individuals because institutional systems embed individual choices to (dis)trust in shared roles, rules, expectations, and accountability mechanisms. Accountability might be formal, as it is in a state; or it could be informal, as in dynamics of social shame and betrayal, which many scholars have shown to be operative in anarchy. 86 System trust makes trusting a less costly choice than it otherwise would be.
Applied to IR scholarship, our claim is that international system trust is necessary for interpersonal trust dynamics to have lasting salutary effects. System trust makes possible the assumption that states are already-existing actors. It provides the unstated reason that the particular individuals analyzed by IR trust scholars are talking to one another in the first place, as well as the institutional backdrop these scholars treat as trust's "aggregation problem." Similarly, anarchy is a trust system. The central actors 84. Waltz 1996. 85. Barber 1983, 164-66. 86. Ilgit andPrakash 2019;Terman and Voeten 2018;Zarakol 2011. each trust that others are states and that there is no world government. Citizens trust that state persons act in their name, committing the state to a course of action (or not). The distrust that IR scholars focus on therefore analytically presupposes the ongoing sociopolitical production of this deeper trust.
Consider the counterfactual, that is, a world in which the continued existence of state persons cannot be relied on, and interpersonal agreements among leaders are no more than leaders "doing favors" for one another based on a perceived meeting of the minds. Interpersonal trust "horizontally" among leaders still would be possible. But it is reasonable to expect that such trust would have a different normative valence and empirical effects on world politics. Where system trust is low, the choice to trust entails a greater leap of faith than in a system with mutually understood institutionalized mechanisms of accountability. The choice might therefore be based on different characteristics in a trust partner-for example, membership in an in-group rather than shared interest in a desired outcome-than in situations of high system trust. Lewis and Weigert warn that declines in system trust, which signify higher uncertainty, prompt a greater need for interpersonal trust, but where characteristic interactions are among strangers and embedded in complex systems, interpersonal trust is not a stable basis for social order. 87 We propose that this is true internationally: remove trust in state personhood and the king has only one outward-facing body. 88 Without system trust in state personhood, trust may well still exist among individual leaders. But the political implications would be different. For example, outcomes based on the trust relationship would be less durable because they would not reliably obligate parties other than those leaders. Trust would not necessarily be a reliable mechanism for conflict resolution or order in world politics. 89 A lack of international system trust also matters "vertically"; that is, it could produce effects at the lay level. For individuals, trusting in the institutions of the sociopolitical order solves at least some ontological-security problems, providing behavioral expectations and links from cause to effect in a wide range of spheres of action. Take that away and the world feels as if is it changing underfoot. It becomes difficult to feel confident that tomorrow will be like today, and therefore difficult to realize a sense of agency and identity. 90 Harold Garfinkel's classic ethnomethodological work on breaching lends some empirical support to this claim. 91 He shows empirically, through experiments, the anxiety that erupts when trust in background conditions is exposed and put to the test. The breaching of linguistic or other expectations can spur a moment of estrangement as the normative backdrop of social life is revealed by the dissonance between what we did not realize we 87. Lewis and Weigert 1985, 974. 88. Holm and Sending 2018, 834. 89. Jepperson and Meyer 2011 suggest something similar in their claim that outcomes without institutionalized support will be less stable.
90 . Giddens 1990;Mitzen 2006. 91. Garfinkel 1967 expected and what is said. Breaching is anxiety provoking, not just-and not necessarily-to the actor engaged in breaching, but to those around them. A breach of system trust exposes the constructedness and ultimately fragility of the social order, laying bare the fictions we rely on to get by, which can have profound effects on our sense of self or identity. Thus, a lack of international system trust can have collective and political effects. Giddens's diagnosis of late modernity as a period of heightened ontological insecurity is rooted in this kind of reasoning, as is Keynesian thinking on economic panics. 92 Catarina Kinnvall has linked collective ontological insecurity to dynamics of securitizing subjectivity, the reducing of self-other relations into simplified relations of othering, which provides some cognitive and emotional stability by clearly identifying friend and foe, but also makes violence more likely. 93 More generally, in situations where system trust is low, that is, where people feel they cannot trust in the continuity of the sociopolitical order, they can become more vulnerable to a politics of fear and willing to trade off political freedom for stability and certainty provided by charismatic and personalistic leadership. 94 The withdrawal of system trust makes anarchy a more dangerous space.
These counterfactual implications suggest the importance of being attentive to variation in international system trust. On the one hand, as a macro-level phenomenon, international system trust is relatively durable. Not everyone needs to trust equally or at all for trust in the system to continue. Not every breach of state personhood signifies a decline in system trust. Systems can sustain breaching, norm/rule violations, and delegitimation without system trust being destroyed. Indeed, the calling out bad behavior is part of the process of reproducing system trust.
On the other hand, over time, or given some conditions, trust in the system can be undermined. Especially in periods of high uncertainty, such as contemporary world politics, it is important to pay attention to how international social reality is being reproduced: when symbols' usage declines; when competing symbols emerge; when performances are not competent, or not recognized as such. 95 It also is important to pay attention to performances of symbolic roles. Audiences of diplomacy expect competent, socially legible performances of state personhood. Incompetent or transgressive performances, that is, performances that subvert the identity they are expected to adopt, effectively "breach" the social order and can have political effects. 96 Repeated breachings, violations that go unremarked, the rise of competing institutions, and delegitimizing discourses all can undermine system trust. In sum, system trust is a missing factor in IR analyses of inter-"personal" trust, whether between individual or state persons. Examining trust in exclusively interpersonal terms overlooks the crucial stabilizing role of the naturalized attitude toward the 92. Giddens 1990. 93. Kinnvall 2004. 94. Kinnvall and Mitzen 2020. 95. Jepperson and Meyer 2011, 63. 96. Duvall and Chowdhury 2011 international system and the seeds of instability and system change we might find in the subversive or transgressive attempts at trust.

Seeing as a State Person
We propose that system trust in state personhood is the dark matter of IR interpersonal-trust scholarship. That scholarship steers analytic attention to trust-formation processes between individual or state persons. It does not consider the phenomenological transformation through which leaders' human relationships become the states' relationships. Without that transformation, where trust is seen by audiences to be merely among persons, that trust can be experienced as a breach in order rather than a contribution to it. State personhood is reproduced in world politics in many locales and practices, from the routine flagging of state intentionality ("the US closes the border") and the use of feelings language ("China risks embarrassment") to describe interstate relations; to the know-how among diplomats of practices of "estrangement" from idiosyncratic interests as they embody the state; 97 to staged meetings between and among states, from the UN General Assembly to the G8; to summitry, in which the state person is embodied and the relationships performed for all to see.
It is beyond our scope to map the multitude of symbols, practices, and ceremonies that reproduce state personhood and contribute to international system trust. Our purpose is to show that IR interpersonal trust arguments take system trust for granted. To do so we revisit a familiar case in IR trust scholarship, the 1985 Reagan-Gorbachev Geneva summit. We place the dominant, interpersonal-trust interpretations of Holmes and of Wheeler beside our own, system-trust interpretation and highlight how the different perspectives lead us to "see" different things in the summit. The difference between the two interpretations is akin to Wittgenstein's "duck-rabbit" image illustrating what philosophers call aspect perception. 98 IR trust scholars argue that any trust forged (or not) at summits is between individuals representing their respective states. In our view, their analyses take summit trust as between human persons. We interpret summit trust as between state persons. That is, we argue that summitry is a staged performance of interstate rapport, whose success is a function of both performance and reception. Audiences expect competent performances of state personhood. Interpersonal trust forged in the context of the performance is trust between states.

The 1985 Reagan-Gorbachev Meeting: An Interpersonal-Trust Perspective
The Geneva summit was the first of a series of summits between Reagan and Gorbachev, culminating in the 1987 Intermediate-range Nuclear Forces Treaty. These summits, aimed at de-escalating bilateral tensions surrounding nuclear 97. Sharp 2009. Wille 2019 develops state personhood as emergent from diplomatic practices. 98. Baz 2020. proliferation, became possible after months of careful planning by both leaderships. For Reagan, who started his second term in 1985, improving US-Soviet relations was a priority. Gorbachev prioritized domestic reforms but treated the transformation in Soviet relations with the West as vital to their success. The conventional wisdom is that the two leaders' personal charisma and intimate communication were central to the summits' success.
For Holmes and for Wheeler, this summitry is a case of reassurance, rapport, and ultimately social bonding, made possible by face-to-face diplomacy. Both suggest that but for face-to-face diplomacy, interpersonal trust or intuition reading would not have been possible, and there would not have been a breakthrough in US-Soviet relations. For Holmes, face-to-face interactions allowed the two leaders to understand each other's intentions through a mechanism of mirroring in our brains whereby an individual can know the intentions of another by simulating how they might behave in the other's position. 99 For Wheeler, the face-to-face intention signaling enabled bonding trust to form between the two leaders, which allowed them to understand their own and their counterpart's identity and interests. 100 Neither treats summits as intrinsically important. International trust is an interpersonal story, and summits are a good locale for drawing out the mechanism because they highlight leaders' interpersonal relations. Whether the leaders met secretly in a private bunker or on stage in Geneva is not relevant. The relationship is forged between the two men first, then institutionalized (or not). Distinct attributes of summitry are not treated as causally relevant. Indeed, summit protocols and playbooks can get in the way of the mechanisms Holmes is after. 101 Thus, the evidence each relies on is primarily reports of the feelings and beliefs of participants about the personal connection between the two men to explain how interpersonal chemistry between them helped mitigate the mutual distrust between Washington and Moscow. Both Holmes and Wheeler highlight remarks made by the leaders and their inner circle that underline the importance of face-to-face meetings in building interpersonal trust. Both quote Gorbachev recalling that there was a "spark of electric mutual trust which ignited between us, like a voltaic arc between two electric poles." 102 Similarly, Reagan recalled that there was a "chemistry" between him and Gorbachev that was close to a "friendship." 103 These quoted recollections support the claim that face-to-face interactions nurtured trust between the two leaders.
They acknowledge that individual dispositions, structural conditions, and the institutional environment matter, but treat them as dimensions that get factored in separately, either as independent effects on the outcome or as relevant to the problem of aggregating from individual minds to the state level. Holmes asserts that causality 99. Holmes 2018, 5-6. 100. Wheeler 2018, 1. 101. Holmes 2018, 174-200. 102. Holmes 2018Wheeler 2018, 156. 103. Holmes 2018Wheeler 2018, 161. at the biological level "interacts" with other mechanisms at other levels to result in a political outcome, suggesting that the mental states of individuals have causal priority. 104 For Wheeler, interpersonal bonding between individual leaders changes the individuals' perceptions of the other, affecting state behavior at the international level. 105 The 1985 US-Soviet Summit: A System-Trust Perspective The (re)production of international system trust at a summit depends on the phenomenological transformation from leaders as human persons to leaders as state persons. That requires a backgrounded confidence in the state person, including through the reliable reproduction of the symbols and ceremonies of state summitry and competent performance of the leaders in their roles. In Geneva, judging from the memoirs and recollections of those involved, rapport clearly developed between the two leaders. But the summitry also was characterized by ongoing, multilevel, and public reproduction of the international system as a system of state persons. This is bracketed in those accounts.
Our narrative therefore begins from a different starting point: summitry is a distinctive practice. As David Dunn points out, summitry is not so much about the issues discussed as it is about the form of communication and level at which the meetings are taking place. 106 As public meetings among heads of state, summits are formal diplomatic events. In a sense, the summit itself is the goal of the summit: summits perform interstate rapport, drawing on mechanisms of interpersonal connection, as a means of constituting that rapport, for each other and for broader audiences. 107 They also perform the state person. Carl Death draws an analogy to the Balinese court analyzed by Geertz: summits similarly are "an illustration of the power of grandeur to organise the world." 108 Summits are constitutively public. Summit performances connect the audiences to the dynamics of world politics, communicating their place in the world, 109 especially through media representations. Just as stage actors alone do not determine the quality of their performance, a summit's participants alone do not determine its effects. Summits are media events for domestic and international audiences, who often are drawn to the spectacles-from rituals and ceremonies down to sartorial elements. 110 A foreign leader can leave an impact on a domestic audience by visiting or through media exposure. 111 Audience members thus connect and identify with states in a way that enables them to comprehend state action and envision world politics. 104. Holmes 2018, 13-16, 255-60. 105. Wheeler 2018, 20. 106. Dunn 1996, 18. 107. Death 2011. 108. Geertz quoted in Death 2011, 7. 109. Constantinou 1994. 110. Cohen 1987. 111. Balmas 2018Goldsmith, Horiuchi, and Matush 2021. Understood in this way, state personhood is crucial for and reproduced in summit performances. 112 Staging a summit requires that state personhood is-and is seen to be-diplomatically maintained. That is, summits require the ongoing "production of estrangement," the removal of individual summit participants from their local and idiosyncratic selves and placement as a member of the international society of states. 113 This is accomplished by drawing on repertoires, forms of action and self-presentation that amount to the "accumulated experience of social actors in a context, the knowledge of what can be done." 114 Summit repertoires include bureaucratic and diplomatic procedures, protocol, the rules on precedent, recognition, and the symbolic production of the states system through, for example, ceremonies and pageantry, flag placement, gift giving, and dinner menus. 115 By pulling from this repertoire of recognizable practices to perform their stateness, the trust that ensues (or not) is between state, not human, persons.
A summit is thus a complex performance of the international system, especially the state person. Elites, media, and the viewing audiences draw on state-as-person repertoires, consciously and unconsciously, as they plan, participate, and reflect on a summit. Each summit performance is different; leaders display "communicative virtuosity," 116 which may well include expressions of personal chemistry. But insofar as each performance taps sufficiently into the recognized repertoire, the state is rarely far behind. 117 Of course, not all leaders are virtuosos. Some go self-consciously off script, or subvert the rules of decorum, or refuse to maintain the requisite estrangement. Such incompetent and transgressive performances potentially expose the fiction of the state person, which could prompt anxiety or ontological insecurity in audiences. These effects can be transitory, as in Garfinkel's experiments. But in conditions of declining international system trust, transgressive behavior and acting out by those in symbolic roles offer sites to examine for seeds of systemic change. 118 Holmes and Wheeler, like most IR trust scholars, home in on specific leaders' virtuosity. Our goal is to show that such virtuosity is firmly within the boundaries of their roles. Reagan's and Gorbachev's performances reflect thinking from system trust and, as such, reproduce the state person. We look beyond the transcripts to show the possibility of inferring successful "phenomenological transformation." The Geneva Summit: From Interpersonal to System Trust Three aspects of the Geneva summit illuminate the dark matter of the trusted international system: the boathouse meeting, the joint communiqué, and the public 112. Neumann and Sending 2020. 113. Adler-Nissen 2016;Sharp 2009. 114. Ringmar 2016. 115. Cohen 1987. See also Danielson and Hedling 2022. 116. Bauman 2004, 9. 117. Faizullaev 2007Neumann andSending 2021. 118. Duvall andChowdhury 2011. response. Each aspect highlights the constitutive relationship between summit performance and system trust. Staging involves actors other than the leaders themselves-the officials and even the journalists are all part of the performance.
The boathouse meeting. As both leaders negotiated for their second plenary meeting on the first day, tensions rose in their discussions of the Strategic Defense Initiative (SDI). Gorbachev was fixated on the US's intentions behind SDI; Reagan countered that it was a defensive shield and not an offensive weapon. At this moment, Reagan suggested that they take a walk together outside, and he led them to a boathouse. In this quiet locale, accompanied only by their interpreters, Reagan and Gorbachev spoke frankly about the source of tension: the purpose of SDI and the mutual distrust about each other's intentions to strike first. Creating a private space to meet, outside the glare of public view, kept the summit on track. Wheeler attributes the interpersonal bonding process to this boathouse chat, where the leaders "shared some very human moments." 119 Without minimizing the "communicative virtuosity" of either leader, we propose that the institutional structure was the precondition for trust formation. This private chat, with its seemingly affable atmosphere, was enfolded into a role and institutional structure of state personhood, which was reliably reproduced. "Going to the boathouse" was not specifically on the summit agenda, 120 but it was in the playbook. William J. Henkel, head of the advance team to Geneva, paid attention to every detail of imagery that would be captured by the press and masterminded the idea of having a fire in the fireplace in the boathouse. 121 The Soviet side was following the US's lead on protocols for state production. 122 By the end of the summit Gorbachev, who looked to Reagan for cues during photo ops and press meetings, learned from Reagan to sit casually with one leg crossed over the other. 123 The visual "symmetry" of these two leaders sitting by the fire, smiling, with both their shoe soles showing became the iconic photograph symbolizing the thaw in US-Soviet relations.
The joint communiqué. At the end of all the meetings on the second day, the two leaders agreed to produce a document describing their accomplishments, one of which was that they "agreed that a nuclear war cannot be won and must never be fought." 124 Wheeler and Holmes attribute this statement to the relationship between the two leaders. Holmes points to this communiqué as indicating growing intuitions between the leaders about each other's intentions. 125 Wheeler similarly situates the communiqué as a product of interpersonal trust: the two leaders disagreed on the US's intentions behind SDI, yet they had a "breakthrough moment" because 119. Wheeler 2018, 155. 120. Marshall 2020. 121. Matlock 2004, 152. 122. Palazchenko 1997, 42. 123. Taubman 2017. "Joint Soviet-United States Statement on the Summit Meeting in Geneva," 21 November 1985. 125. Holmes 2018 their face-to-face interactions during the first day of meetings and their dinner together humanized US-Soviet distrust. 126 Shifting the lens to illuminate the background, the joint communiqué was made possible by, and maintained, the international system of state persons in two ways. First, a communiqué, while not a treaty, is a formal diplomatic document declaring states' commitments. Such public joint commitments buffer interstate relations from the idiosyncrasies of interpersonal dynamics. 127 Here, the communiqué statement that "a nuclear war cannot be won and must never be fought," 128 which was used repeatedly by the leaders in subsequent summits, and the statement that there would be future reciprocal summits in the two capitals similarly mitigated the future uncertainty of human factors in US-Soviet relations. 129 The Geneva meeting and the joint document thus "set a mechanism of bilateral relations in motion." 130 Importantly, while the leaders' interpersonal chemistry may have inspired the timing and enthusiasm, proposing these reciprocal visits was part of the US's playbook well before the boathouse. 131 Second, the agreement was bookended by ceremonies. As the New York Times reported after the first day of talks, the "US and Moscow" had "agree[d] on future talks." 132 The interstate communiqué was ritualized through a choreographed signing ceremony. Both foreign ministers, Shultz and Shevardnadze, signed the agreement, surrounded by the two leaders and other aides, on a stage with the American and Soviet flags as the conspicuous backdrop, in an auditorium filled with journalists. 133 Public response. Holmes and Wheeler overlook the constitutively public nature of summitry. The implication of their arguments is that whether public or private, diplomacy can have a bonding effect among the humans participating in it that can "matter" politically. In our view, the bonding or fusion in summitry is as much between leaders and lay audiences as it is between the leaders themselves. Performing interstate trust and rapport reminds and reassures citizens of their state's place in the world and its role as caretaker and mediator of "the international," even if subconsciously, by not upending it.
It is difficult to empirically ascertain the success of this "fusion" at reproducing system trust except in a "dog that didn't bark" manner. System trust entails the reproduction of the "normal." In the Geneva case, if system trust did not obtain, we might expect to find expressions of dissonance between the leaders' performance and people's expectations about interstate relations. One indirect measure of such 126. Wheeler 2018, 155. 127. Mitzen 2013. "Joint Soviet-United States Statement." 129. Shultz 1993, 605. 130. White House, "Memorandum of Conversation: Reagan-Gorbachev Meetings in Reykjavik, October, 1986. First Meeting," 11 October, 1986. 131. Matlock 2004compare Holmes 2018, 97. 132. Weinraub 1985a. 133. Gorbachev 1995 dissonance is polling. 134 If the Reagan-Gorbachev summitry increased interstate trust, we might expect a reduction in their populations' nuclear anxiety. The Geneva summit was more than a media event for the American public, which was reckoning with nuclear anxiety at the news of the summit. 135 Although expectations were low that it would improve US-Soviet ties, 136 opinion polls showed a visible decline in the American public's expectation of a nuclear war as the summits progressed. 137 The New York Times featured on its front page a photograph of Reagan shaking hands with Shevardnadze, and Gorbachev shaking hands with Shultz, at the closing ceremony. The sub-headline said "no breakthroughs," 138 but the picture communicated otherwise, a clear symbol of thaw between the two adversaries. 139 Gorbachev later described this moment as "a sight the whole world had hoped for" after years of ideological war. 140 Gorbachev also scored foreign policy success among Soviet audiences for his performance at Geneva, as the summit "went down well with the people at home." 141 An interpersonal-trust perspective on summits measures success and failure in terms of relationships among individuals and/or interstate agreements. From a system-trust perspective, the constitutive publicness of summitry matters, making the audience a part of the event and any effects it might have. The maintenance of system trust, especially through competent performances of state personhood, transformed the meeting into a feeling about US-Soviet relations in a way that an interpersonal perspective cannot capture.

Conclusion
The international institutional order currently faces many pressures, empirical and normative, suggesting that the state may not be the best vehicle for meeting basic human needs.
But as criticism of the state mounts, it can be easy to forget that the orderliness that the states system provides, and the ability to trust in the system's continuity, is a source of ontological security. State personhood enables lay audiences to trust in the system, an abstract, far-away structure that nonetheless affects-and is felt to affect-day-to-day security. It also enables interpersonal trust among leaders to have the salutary effects on international order that IR trust scholarship lauds. IR's 134. Polling is an imperfect measure of public experiences. On using surveys to capture trust, see Brehm and Savel 2019. 135. Trafford 1987. 136. Shipler 1985. 137. Smith 1988, 559. 138. Taubman 2017, 283. 139. Apple 1985Weinraub 1985b. 140. Gorbachev 1995, 410. 141. Palazchenko 1997 research program on trust assumes that it can be understood on the interpersonal model and as taking place in so-called realist anarchy. The effect of this focus is to foreground idiosyncratic interpersonal relations at the expense of constitutive international structures, while obscuring how interpersonal virtuosity may indicate transgressive international behavior-a possibility that is of both analytical and political significance. Because knowledge of social life readily becomes knowledge in social life, in closing we want to highlight the synergy between the critique motivating our argument, that IR trust scholarship overly stresses interpersonal dynamics, and the political costs we identify of a decline in international system trust, the leaning on charismatic leaders and populist politics. 142 Both reflect valorization of interpersonal dynamics and de-emphasize structure.
IR's scholarly debates and teaching contribute to the reality of international society. 143 As many scholars have argued, an unquestioned scholarly reliance on anarchy contributes to methodological nationalism, 144 that is, the unthinking reproduction of the states system by assuming it as the unit of analysis. We propose, in addition, that assuming realist anarchy while foregrounding interpersonal dynamics in IR trust research can have the unintended effect of elevating the personal, as if individual persons heroically overcome the constraints-and complexities-of the world around them to produce international order. That is the promise of the charismatic leader; and the instability it portends is one of the dangers that legal state personhood was introduced in order to curb. Our methodological choice to present our argument as an exercise in aspect perception, by asking readers to hold both interpretations in mind, offers a way to navigate these two poles.
The moment of the individual in IR trust research does not solve or address the problems of the states system. Scholars and practitioners may advocate the state's demise for other reasons. But in so doing it is important to attend to the anxieties the states system has held at bay and/or devise political responses to a potential "bank run" on state trust. We have sought, therefore, to see and name system trust in state personhood, and show how it produces order, as a step toward imagining, organizing, and harnessing the energies of alternatives. 145