Introduction
We often speak of international organizations (IOs) as though they possess independent agency, shaping the world through their own deliberate actions.Footnote 1 Public discourse abounds with statements such as: ‘The UN brokered a peace agreement,’ ‘NATO expanded its military presence,’ or ‘The WHO issued new health guidelines.’ International Relations (IR) scholars often echo these linguistic patterns. Some, however, go a step further, seeking to ground them theoretically and arguing that such phrases, rather than serving as mere metaphorical shorthand, reflect a deeper reality about the actorness of IOs. Choosing as their criterion the ability of IOs to influence world affairs independently of member states’ interests and preferences, rational institutionalists and constructivists typically locate IO actorness in delegated roles, specifically within bureaucracies and diplomatic staff.Footnote 2 More recently, Thomas Gehring and Kevin Urbanski have extended this argument beyond the internal apparatus of IOs, suggesting that even member-driven bodies such as the United Nations Security Council can exhibit a distinct form of actorness.Footnote 3 They too, however, subscribe to the idea that such actorness must be rooted in independence from individual state interests and preferences. In their view, this occurs when IOs acquire action capacity by pooling governance resources and achieve autonomy through decisions grounded in distinct organizational rationales.
Yet what existing scholarship often overlooks, and thus fails to account for theoretically, is how closely the agency of IOs is tied to their personification. When IOs appear to influence world politics as independent actors, they do so not as abstract forces but by exhibiting properties typically associated with human beings, such as intentions, beliefs, and even emotions. Crucially, this personification is not merely reflected in lay and academic discourse but constitutes an integral component of decision-making in many IOs, including those dominated by member states. For instance, the Security Council – used in this article as an illustrative case – has required its members, since 1945, to frame decisions as resolutions in which the Council itself is the subject performing a wide array of human-like actions. These actions include making authoritative demands when it ‘requests’ or ‘approves’ certain measures; expressing opinions when it ‘determines’ facts and ‘concludes’ from them; passing judgments when it ‘praises’ or ‘condemns’ the actions of other actors in world politics; and expressing emotions when it conveys ‘shock’ and ‘outrage’ at mass atrocities such as genocide or ethnic cleansing. Against this background, the aim of this paper is to account for the personification of IOs by member states while simultaneously advancing a novel perspective on IO actorness, thus treating the two as fundamentally interconnected.
To develop this perspective, the paper turns to Margaret Gilbert’s work on plural subjects. In contrast to other philosophers, like Philip Pettit, who have extensively explored the problem of collective intentionality, Gilbert’s contributions remain largely overlooked in IR. Yet her contributions are particularly significant as she is among the few philosophers pursuing this line of inquiry to also give sustained attention to the questions of collective agency, subjectivity, and traits of personhood. According to Gilbert, plural subjects – a term she uses to denote collective actors – are formed when individuals jointly commit to act as a single body. Footnote 4 Here, ‘act’ is understood broadly; not only as doing something but also as intending, believing, or feeling. Therefore, plural subjects can, in Gilbert’s view, perform most of the functions individuals can; not only ordinary actions but also complex cognitive and emotional activities such as feeling. This, in turn, grants them traits of personhood, although Gilbert does not claim a perfect parallel between plural subjects and individual persons.
What, on the other hand, renders plural subjects actors in their own right is that their actions, intentions, beliefs, and emotions cannot be reduced to the individual actions, intentions, beliefs, or emotions of their members. When, for example, John and Alice decide to go shopping together, this intention is not merely the sum of John’s personal intention to go shopping and Alice’s personal intention to go shopping, but the product of their joint commitment to do so as a unified entity or, in Gilbert’s language, a body. Another important aspect of Gilbert’s approach to plural subjects is that, while she differentiates between joint commitment and individual interests, intentions, preferences, and commitments, she does not posit a strict relationship between the two. The two may converge or diverge – i.e., John and Alice might personally feel like they want to or do not want to go shopping – but what will constitute the plural subject is not what they personally want but whether they have jointly committed to doing so. In Gilbert’s words, only the joint commitment constitutes the ‘relevant mental state’Footnote 5 necessary for the formation of plural subjects and the myriad of actions it can undertake.
By treating joint commitment and plural subjects as foundational to social ontology and thus applicable across contexts – from formal to informal settings, and from basic to derived commitments – Gilbert places no conceptual limits on using these ideas to theorize IO actorness and personhood. In the case of the Security Council, decision-making through draft and adopted resolutions thus displays clear features of a Gilbertian plural subject. When states draft a resolution, they already frame the Council as the agent of proposed actions. Rather than negotiating state-by-state preferences, they implicitly ask each other: ‘Are you ready to jointly commit to espousing these intentions, actions, beliefs, and emotions as the Security Council?’ A resolution is finalized when a majority of members – excluding any permanent-member veto – are prepared to do so. This act of joint commitment affirms the Council’s actorness, as its expressed positions cannot be reduced to those of individual states. Moreover, since plural subjects are not inherently limited in the kinds of traits they can embody, member states routinely attribute to the Council person-like qualities such as authority, judgment, opinion, and emotion.
The paper contributes to the debate on collective agency in IR, and more specifically on IOs, in four key ways. First, following Gehring and Urbanski, it argues that member-dominated IOs can also be considered actors in their own right. Second, it expands this inquiry by observing that IO actorness is not an abstract quality but a quality inseparable from the personhood traits ascribed to IOs by member states. Actorness and personification are two sides of the same coin, and both require accounting for. Third, the paper does so by drawing on a little-used concept from analytical philosophy and, in doing so, identifies a micro-foundational social process by which IOs’ actorness and personhood traits come to life – a process that is grounded in broader social ontology rather than organizational specificities such as resource pooling. In this way, the paper seeks to address the criticism often leveled in IR that the actorness of collective entities such as IOs and states is more often asserted than demonstrated in terms of the process by which it arises.Footnote 6 Fourth, by drawing on Gilbert’s differentiation between joint and individual commitments, the paper rejects a longstanding criterion in IR: that IOs can only be considered actors in their own right if they can be shown to act independently from state interests and preferences. In so doing, the paper contends that IO actorness is not a transient or situational phenomenon, but a stable and enduring feature of member-dominated IOs.
The paper unfolds in five steps. It begins by closely examining Security Council resolutions – the Council’s main decision-making tool – to show how member states personify IOs, thereby highlighting the close connection between personification and actorness. Next, it offers an in-depth review of the existing literature on IO actorness, alongside related debates on the state-as-a-person problem. It then introduces Gilbert’s notion of plural subjects, focusing on her concepts of action, the body, and joint commitment. These theoretical insights are subsequently applied to the Security Council as the paper’s illustrative case. Finally, the discussion outlines the broader theoretical implications and identifies avenues for future research, particularly in relation to two key themes: institutional moral agency and IO authority.
Beyond actorness: the personification of international organizations.
Security Council resolutions have been a centerpiece of world politics for nearly eight decades. Since the establishment of the UN in 1945, state representatives have adopted 2779 resolutions, addressing and affecting almost every aspect of international relations, from Iran’s nuclear program to the impact of armed conflict on women and girls.Footnote 7 Given their prevalence, it might seem that little remains unknown about these resolutions. However, as is often the case in international relations – and perhaps in social science more broadly – certain phenomena and practices become so routine that they are perceived as a given and thus fail to attract much scrutiny. In the case of Security Council resolutions, one such taken-for-granted aspect is their format and the practice that shapes it.
From the outset, state representatives in the Security Council have consistently followed the same template when drafting resolutions. As a result, every resolution is composed of a series of paragraphs – some no longer than a single sentence – each beginning with a capitalized verb in either the present participle or the present tense. At the top of these resolutions, preceding the paragraphs, the Security Council is listed as the subject of the actions, states, and behaviors to which these verbs refer. For resolutions to take this form, state representatives must engage in a common but seldom explicitly recognized practice: they must craft a series of action sentences that personify the Security Council, attributing human-like capabilities and characteristics to it. As with state personification, some might dismiss this characterization, or even the practice itself, as merely an inconsequential play with metaphors. However, it is important to recognize that we are not discussing a scholarly practice of personification but rather a practice of personification routinely exercised by representatives of member states during their decision-making. The following point cannot be emphasized enough: No formal decision in the history of the Security Council could be adopted without first being translated into language that personifies the Security Council.
To gain deeper insight into this phenomenon, I built a text corpus consisting of 2431 Security Council resolutions adopted between 1949 and 2018. Using methods of automated text analysis, I extracted verbs located at the beginning of the resolutions’ paragraphs – verbs that directly attribute humanlike actions, states, and behaviors to the Security Council. This exercise yielded 146 unique verbs, which, upon closer examination, reveal that, in its nearly eight decades of existence, the Security Council has been ascribed a surprisingly robust ‘personality.’Footnote 8
First, in line with its mandate to ensure international peace and security, it has been constructed as a highly authoritative ‘person.’ Accordingly, verbs like decide, direct, declare, request, authorize, and approve are among the most frequently used in resolutions. Closely trailing these are verbs that attribute to the Security Council the capacity to establish facts and express opinions. As a result, the resolutions are replete with verbs such as determine, confirm, affirm, maintain, observe, note, believe, realize, and conclude.
Second, the representatives of member states attribute to the Security Council the ability to pass judgment on international actors and events. Verbs that stand out in this regard include praise, applaud, salute, warn, denounce, welcome, congratulate, and condemn. Accompanying these are evaluative verbs through which the Security Council appears to assess the significance of various events and decisions, such as stressing, emphasizing, underlining, underscoring, reminding, re-emphasizing, and highlighting. There are also verbs by which the Security Council appears to address other actors during crises. It most commonly calls upon them to take certain actions, but also frequently urges, encourages, supports, and exhorts them.
Third, and perhaps most surprisingly, the representatives of the member states regularly attribute emotions and consciousness to the Security Council. These are arguably the most human-like qualities assigned to it. The Security Council is thus often said to be concerned, mindful, shocked, dismayed, outraged, distressed, disturbed, regretful, and grieving. In several resolutions, it is also presented as being aware and conscious.
Despite consistently using personifying language in its decisions, the Security Council may still appear to be an unlikely case of personification as it lacks many of the qualities we typically associate with a unified actor.Footnote 9 Its membership is deeply divided, especially among the five permanent members, who often pursue conflicting national interests and can unilaterally block collective action through the veto. As a security-focused institution, the Council responds to urgent crises shaped by divergent threat perceptions, strategic calculations, and geopolitical rivalries, with little shared vision or stable normative foundation. Moreover, unlike more technocratic or bureaucratic IOs, it lacks a consistent administrative core. Instead, it functions as a political arena in which power politics dominate. These conditions make it difficult to imagine the Council speaking as one. And, yet, through its resolutions, it is persistently described in precisely those terms: as having intentions, beliefs, and emotions.
The Security Council, however, is not alone in its use of personifying language. States in other international security organizations, such as the OSCE and NATO, regularly employ similar rhetoric. For example, official documents describe the OSCE as a ‘promoter of the notion of comprehensive and indivisible security’ and assert that NATO ‘will deter and defend against any threat of aggression.’Footnote 10 This tendency is also evident among IOs operating in diverse policy areas, from health and food to labor rights. In a resolution outlining its response to the COVID-19 pandemic, the World Health Assembly of the World Health Organization (WHO) is described as ‘deeply concerned’ about the ongoing crisis, while simultaneously ‘expressing solidarity’ and ‘optimism.’ The resolution further states that the Assembly ‘calls’ on the Member States and ‘requests’ the Director-General to take appropriate public health measures.Footnote 11 Similarly, in its 2019 centenary declaration, the International Labour Conference of the International Labour Organization (ILO) ‘recalls’ and ‘reaffirms’ its founding aims and principles, while ‘declaring’ its future intentions.Footnote 12 The Conference of the UN Food and Agriculture Organization (FAO) is likewise attributed the ability to ‘note,’ ‘emphasize,’ and ‘decide.’Footnote 13 In short, by crafting sentences of this kind, state representatives ascribe the actions, intentions, and judgments on which they have collectively agreed to singular organizational entities. The result of this practice is a universe of internally personified IOs – bodies that appear to feel, evaluate, reason, decide, and act.
The widespread nature of self-personification among IOs suggests that this practice is not primarily determined by institutional characteristics such as shared norms and values, issue area, level of authority, or the binding nature of their decisions. Instead, it likely rests on a deeper underlying mechanism. Against this background, this paper seeks to address several key questions: Why is the practice of IO personification an integral part of member states’ decision-making within IOs? How is it connected to the broader issue of treating IOs as actors in their own right? And what theoretical and practical consequences does it produce?
Before proceeding, a caveat is in order. In the subsequent sections of this article, I draw on the Security Council as my central example. Given the depth of my engagement with it, the Security Council represents the case of IO personification with which I am most familiar. While IO personification is a widespread phenomenon – and Margaret Gilbert’s notion of plural subjects is a statement on a general social ontology – I consider the arguments developed here to be broadly generalizable. Nevertheless, variations across cases should be expected, and I hope these will serve as a basis for refining and expanding the claims advanced in this article.
State of the art
International organizations as actors in their own right
The practice and significance of IO personification by member state representatives during decision-making have largely escaped the attention of IR scholars. Instead, they have focused on a related yet distinct question: Do IOs qualify as actors in their own right within world politics? This line of inquiry gained prominence during the so-called neo–neo debate of the 1980s and early 1990s, in which neorealists and neoliberal institutionalists examined the relationship between states – conceived as rational utility-maximizers – and international institutions, including IOs.Footnote 14 Neither camp, however, acknowledged IOs as autonomous actors. Neorealists dismissed them as mere arenas in which states enact power relations and strategically maneuver in their pursuit of relative gains, maintaining that IOs cannot produce effects beyond state interests and preferences.Footnote 15 In contrast, neoliberal institutionalists – most notably Robert Keohane – conceived of IOs as information-rich social structures that not only constrain the behavior of member states but also shape their expectations by, among other things, lowering transaction costs and reducing uncertainty.Footnote 16 Yet, despite not viewing IOs as standalone actors, neorealists and neoliberal institutionalists have established largely uncontested criteria for recognizing IOs unique actorness: scholars who adopt this position must demonstrate that these organizations operate autonomously, are dissociated from, or significantly diverge from state interests and preferences, and, as such, can exert independent effects in world politics.
Rational institutionalists and constructivists were among the first to identify scenarios in which these criteria for IO actorness could be met. In both cases, the strategy has involved shifting attention away from IO member states and toward actors operating within these organizations. Drawing on principal–agent theory, rational institutionalists have thus conceptualized states as principals who delegate specific tasks to intra-organizational agents – such as diplomatic staff and bureaucrats – authorizing them to act on their behalf.Footnote 17 While principals typically seek to maintain tight control over their agents, delegation always introduces the possibility of agency slack – a deviation in agents’ behavior from the principals’ interests, preferences, stated missions, or delegated tasks. Such slack may arise when agents choose to pursue their own agendas, but it can also result from structural complexities such as collective principals (i.e., multiple member states) issuing contradictory mandates or extended chains of delegation distorting their original demands.Footnote 18 And it is these deviations – intentional or incidental – that rational institutionalists cite as evidence that IOs can, at times, exert independent influence in world politics. In this view, IO actorness is thus not fixed but transient: organizations move in and out of it depending on whether principals retain control or agents assert autonomy through agency slack.
In contrast, constructivists view the actorness of IOs as a stable organizational attribute. This position arises from their exclusive focus on IO bureaucracy, which they, drawing on Max Weber, depict as an inherently independent entity with a distinct identity and unique culture. Leading proponents of this view, Martha Finnemore and Michael Barnett, assert that IOs are, at their core, bureaucracies.Footnote 19 While they acknowledge that bureaucracy depends on member states for survival, they maintain that a distinct bureaucratic culture predisposes it to address world political issues in specific ways, granting it considerable autonomy from the declared IO missions and member states’ interests and preferences.Footnote 20 Barnett and Finnemore further contend that this culture and autonomy can manifest so strongly that it sometimes leads to dysfunctional and pathological behavior within the bureaucracy, causing related effects in world politics.Footnote 21 Therefore, from the constructivist standpoint, the inherent characteristics of the bureaucracy, along with behavioral deviations they may lead to, conclusively establish IOs as independent actors on the global stage.
But not everyone is satisfied with exploring IO actorness by focusing solely on intra-organizational staff and bureaucracy. In their recent contribution, Thomas Gehring and Kevin Urbanski – alongside Gehring’s related work – take a step toward arguing that even state-dominated IOs can be considered standalone actors in world politics.Footnote 22 However, they too remain bound by the criteria established during the neo–neo debate, namely, that such actorness exists only if there is clear evidence that decisions and actions – even those taken by the states themselves – in some way depart from state interests and preferences and, as such, produce independent effects in global governance. To meet these criteria, Gehring and Urbanski argue that member-dominated IOs attain actorness whenever they acquiesce to sovereignty loss by developing action capability and autonomy for an IO. Drawing on James Coleman’s work, they argue that action capability is established when member states pool their resources and empower an IO to use them for the organization’s collective objectives rather than for the individual goals of the states.Footnote 23 IO autonomy, on the other hand, is evident when the organization’s policies cannot be explained as a straightforward compromise among the interests and preferences of its member states.Footnote 24 Gehring and Urbanski argue that this occurs whenever ‘distinct organizational rationales’ are at work, such as when established organizational norms prioritize claims and proposals aligning with the organization’s broader objectives rather than the preferences of individual states, when previous decisions shape future ones, or when expert perspectives gain greater prominence and influence.Footnote 25 Citing Christian List and Philip Pettit’s classical work on group agency, they also argue that the dissociation of organizational policies from state preferences can occur when certain decision-making criteria prevent the simple aggregation of member states’ preferences.Footnote 26
In this respect, their argument aligns with that of Toni Erskine, who examined IO actorness in relation to institutional moral agency and learning.Footnote 27 Also drawing on List and Pettit, Erskine contends in her later work that majority voting and consensus decision-making that is strongly influenced by organizational culture can commit group members to policies or actions that diverge from their individual preferences, thereby helping transform the group into a purposive actor.Footnote 28 Conversely, Erskine maintains that decision-making models based on unanimity fail to produce outcomes reflecting the independent will of an IO. Since they are indistinguishable from the individual preferences of member states, these outcomes, therefore, do not contribute to establishing IO actorness.Footnote 29
Although Gehring, Urbanski, and Erskine’s exploration broadens and deepens the discussion of IO actorness by extending it beyond intra-organizational actors and tying it to the problem of preference aggregation, they, similar to neoliberal institutionalists, advance a transient notion of IO actorness. Erskine makes this point explicitly. In her analysis of the Security Council, she argues that when one of the five permanent members exercises its veto, it effectively replaces the majority voting procedure with a demand for unanimity. This veto power, therefore, leads her to classify the Security Council as an institution that transitions between possessing and losing actorness.Footnote 30 While Gehring and Urbanski do not explicitly acknowledge this transience, they note that their two dimensions of IO actorness – action capability and autonomy – are subject to empirical investigation and vary within and across different IOs, suggesting that IO actorness can, at the very least, differ in degree.Footnote 31
Neither Gehring and Urbanski nor Erskine, therefore, adopts a strong ontological stance on group actorness. To satisfy the criteria for IO actorness established during the neo–neo debate, they draw on non-reductionist accounts of collective agency, as developed by scholars such as James Coleman, Peter French, Christian List, and Philip Pettit. However, they stop short of fully endorsing non-reductionism. Instead, they contend that the simple aggregation of preferences and intentions is both possible and common within IOs, rendering IO actorness a transient phenomenon. In other words, they do not treat reductionism and non-reductionism as ontologically incompatible, but rather as empirical states between which IOs may shift. This perspective leads them to formulate criteria for identifying when member-dominated IOs operate in non-reductive states (signifying the presence of IO actorness) and when they revert to reductive states (where IO actorness is absent). In contrast to Gehring, Urbanski, and Erskine, and drawing on Gilbert, this paper adopts a non-reductionist ontological perspective on group actorness to argue not only that member-dominated IOs are actors, but that this actorness is stable and imbued with personhood. However, to firmly establish this argument within the IR literature, it is essential first to examine the IR scholarship that has addressed corporate personhood, not just corporate actorness, and delve deeper into the questions of reductionism and non-reductionism.
State-as-a-person: proxy literature
The literature that addresses the attribution of person-like characteristics to corporate agents while also providing deeper insights and a clearer understanding of the broader debate on reductionism versus non-reductionism is the state-as-person literature. The puzzle that has animated this body of work is why scholars, practitioners, and laypersons persistently attribute person-like qualities to states. Are such ascriptions merely metaphorical (‘as if’), or do they correspond to some deeper reality?Footnote 32
Advocates of the metaphorical interpretation – often categorized as reductionists and methodological individualists – adopt the position that only individual humans possess the capacity for action, and they do so based on subjective personal motivations.Footnote 33 For them, the notion of state personhood serves merely as a pragmatic tool to organize experience and build theory, but it does not imply any real ontological presence.Footnote 34 Or, as Sean Fleming puts it, those advocating this position view action sentences about states as merely figurative shorthand for action sentences about individuals.Footnote 35 A prominent advocate of this perspective is Robert Gilpin. He dismisses the metaphorical use of personified language about states, considering it merely a tool of convenience employed widely for the sake of brevity.Footnote 36 In a similar vein, Peter Lomas views it as a harmful convention, stating that ‘[s]tates do nothing’ and that ‘[we] have no need for a verb for them.’Footnote 37
Non-reductionists, most prominently Alexander Wendt, reject this position.Footnote 38 They assert that if the language of state personification were merely useful fiction, it would be possible to dispense with it and discuss the real thing or switch to a different metaphor.Footnote 39 Wendt also contends that seeing state personhood as merely metaphorical would require accepting what he calls a ‘miracle argument’: it would be nothing short of extraordinary for theories predicated on state personhood to work so well if they did not correspond to some underlying reality.Footnote 40 As an alternative, he offers a rare defense of states as purposive actors with a sense of self, controversially declaring that ‘states are people too.’Footnote 41
Wendt builds his argument by first outlining three general criteria of personhood: intentionality, being an organism, and having consciousness (understood as a first-person subjective experience).Footnote 42 However, he quickly offers a disclaimer, clarifying that he will primarily defend state personhood by arguing that states possess intentionality. He grounds this defense in non-reductive physicalism, whose proponents in philosophy have already established arguments for the reality of group and corporate intentions.Footnote 43 Arguing that states are organisms, rather than just super-organisms like beehives, and that they possess consciousness would require Wendt to go beyond physicalism into panpsychism, which views the mind as ontologically fundamental as matter or, potentially, as an intrinsic aspect of matter itself. Although Wendt is sympathetic to such approaches, their speculative nature leads him not to pursue them fully, offering instead only tentative arguments for how they may be used for the defense of state personhood.Footnote 44
Accordingly, Wendt establishes three criteria for state intentionality: (1) constitutive members possessing a collective ‘idea’ of the state, (2) an institutionalized decision-making structure, and (3) centralized authority capable of imposing binding decisions on its members.Footnote 45 He supports these criteria with the concepts of supervenience and emergence – philosophical notions that adhere to the idea of a stratified reality where higher levels (e.g., group intentions) cannot be reduced to the sum of their lower-level components (e.g., individual intentions).Footnote 46 Supervenience and emergence, therefore, counter reductionism, which asserts that We-intentions are simply the interlocking I-intentions toward a group.Footnote 47 Citing evidence proposed by non-reductionist philosophers to dispute reductionism – some of which, as previously mentioned, Gehring, Urbanski, and Erskine repurposed as criteria of IO actorness – Wendt argues that supervenience and emergence are supported by observations showing that groups can possess intentions that none of their members individually hold. They can also persist despite complete member turnover and undertake actions that individuals alone cannot, such as imposing sanctions on misbehaving states.Footnote 48
The concept of supervenience helps explain these ideas through the notion of multiple realizability. It means that if the lower-level components (like people) are exactly the same, the higher-level structure (like a state) will also be the same. But different people or arrangements can still produce the same kind of state.Footnote 49 So, while a state’s intentions depend on its members, they do not depend on any one person’s intentions in particular.Footnote 50 This is what makes multiple realizability a form of non-reductive physicalism. By contrast, Wendt’s interpretation of emergence rejects the idea that individual intentions come before group intentions. Instead, he sees both as shaped by shared social meanings, making the relationship between individuals and the group two-way, not one-directional.Footnote 51 His view draws on structuration theory and loosens the physicalist assumption that only physical elements – like individual people or brains – can define what a group is. Wendt accepts both supervenience and emergence in his account of state personhood, though his non-physicalist leanings, however speculative, lead him to argue that emergence offers stronger support, as it grants greater ontological autonomy to group intentions. Supervenience, however, remains the minimal condition for justifying this personhood.Footnote 52
Despite its theoretical sophistication, Wendt’s idea of the state as a person is subject to two criticisms relevant to the present study. The first concerns Wendt’s reliance on intentionality as the primary quality and a minimal criterion for defining state personhood. While acknowledging that attributing person-like traits to states extends beyond intentionality to include aspects like emotions, he maintains these traits depend on arguing that states are organisms that possess consciousness. Since such arguments remain speculative, he leaves these traits largely unaccounted for, or at best, only tentatively addressed. Furthermore, referencing actor–network theory and its notion of actants, Colin Wight argues that intentionality may actually be a key characteristic of agency rather than personhood.Footnote 53 According to this theory, anything can have agency – i.e., be an actant – if it can influence other entities within a network. Actants, such as technologies or institutions, are designed with specific purposes in mind and influence the network to realize this purpose, thereby demonstrating a form of intentionality without personhood. Wendt’s theory, therefore, may show that states have agency, but not personhood. As a result, it falls short when applied to the Security Council, since it cannot fully account for the person-like traits discussed earlier.
The second issue with Wendt’s understanding of state personhood concerns the ambiguous link between his criteria for state intentionality – namely, the existence of a collective ‘idea’ of the state, institutionalized decision-making structures, and centralized authority – and the concepts of supervenience and emergence. Specifically, once these criteria are in place, it is unclear which process produces supervening or emerging state intentionality and, by extension, state personhood. In his discussions on supervenience and emergence, Wendt is more engaged with the philosophical debates about these concepts – e.g., their relations to physicalism, panpsychism, and reductionism, and questions about the existence of group minds and the nature of collective and distributed cognition – than he is in detailing how these ideas connect to his framework of corporate intentionality or foster state personhood.Footnote 54 But he does provide some clues that can facilitate further theorization. In his discussion of supervenience, Wendt highlights Gilbert’s concept of plural subjects – leveraged in this paper to theorize both IO agency and personhood – as among the most significant recent contributions to the study of group intentionality.Footnote 55
But before moving to the next section, it is crucial to note that beyond metaphorical and realist interpretations of state personhood, there exists a third approach – though less coherent – that focuses specifically on the linguistic aspect of this personification. Given that the problem addressed in this paper, along with the proposed solution, involves linguistic elements, these perspectives are noteworthy. Accordingly, Erik Ringmar argues that the subjectivity of both individuals and states is formed through narratives – stories that unite their diverse experiences into a coherent self.Footnote 56 In this ‘narrative conception of self,’Footnote 57 one exists only under a description; there is no ‘real’ entity to which the description refers. Patrick Jackson adopts a practice-based approach, contending that states do not exist independently of practices, including language practices. Accordingly, while states are constantly personified, he maintains that they never become persons with a stable ontological essence, rendering the search for such an essence futile.Footnote 58 Iver Neumann asserts that language is inherently metaphorical. Therefore, the important question is not whether a particular phenomenon is metaphorical but which metaphors are used to describe it, what effects they produce, and what the associated costs are.Footnote 59 Finally, adopting a language-before-ontology approach and drawing on Hobbes, Sean Fleming argues that states are fictional artificial persons – they act vicariously through an agent authorized by someone else.Footnote 60
Margaret Gilbert’s notion of plural subjects
In contemporary philosophy, the question of group agency and personhood is part of the broader debate on collective intentionality.Footnote 61 As in the IR literature it has influenced, this debate is divided between reductionists such as Michael Bratman, who view collective intentionality as interlocking personal intentions, and non-reductionists, who argue that ‘we-intentions’ cannot be reduced to ‘I-intentions.’Footnote 62 Yet not all non-reductionists address the agency or personhood of the third-person ‘we.’ Only a few explore this dimension in depth. Besides Philip Pettit and his collaborators – who have influenced scholars like Erskine, Gehring, and Urbanski to argue that IO agency emerges, among other factors, from premise-based majoritarian voting – Raimo Tuomela and Margaret Gilbert stand out as key figures in this area.Footnote 63 This paper draws primarily on the work of Margaret Gilbert. As a non-reductive physicalist, she offers a supervenient account of group agency and its associated traits of personhood. Unlike the IR theories discussed earlier, her approach does not require a strict division between group and individual intentions. And unlike Wendt, she provides both a concrete mechanism for the emergence of group agency and a detailed account of how groups can embody cognitive and psychological features without invoking metaphysical notions like a ‘social spirit’ or ‘group mind.’Footnote 64 The next section introduces Gilbert’s concept of plural subjects and its core components: joint commitment, doing, and a body. It then applies this framework to the actorness and personification of the UN Security Council and outlines its broader implications for understanding IOs.
Joint commitment, doing, and a body
To understand Gilbert’s concept of plural subjects, it is helpful to begin with one of the examples she commonly uses.Footnote 65 Suppose John and Alice have agreed to go shopping this afternoon. When Scott, unaware of their arrangement, asks John about his plans, John gestures toward Alice and replies, ‘We intend to go shopping.’ Scott responds, ‘I see, you both intend to go shopping.’ Somewhat irritated, John clarifies, ‘No, no, we intend to go shopping together.’ Based on this example, Gilbert concludes that (at least in English) the sentence ‘We intend to do A’ carries a meaning that is not equivalent to ‘Each of us intends to do A.’Footnote 66 She contends that the former signifies the formation of a plural subject, whereas the latter does not. This distinction also suggests that the intentions of plural subjects are more than a simple sum of individual intentions. If summing individual intentions were sufficient – or even possible – the two sentences would be equivalent. John’s and Alice’s separate intentions to go shopping would make ‘We intend to go shopping’ true, but that is clearly not the case. Instead, Gilbert concludes, other conditions must be met, and, as will be discussed below, personal intentions are not even among them.
‘People form a plural subject,’ Gilbert asserts, ‘when they are jointly committed to doing something as a body, in a broad sense of “do.”’Footnote 67 Or, in the language of analytical philosophy: ‘When persons P1 and P2 are jointly committed to X-ing as a body, they constitute … a plural subject.’Footnote 68 Both definitions capture the essence of the earlier example: plural subjects are groups whose members are jointly committed to acting together. Given how frequently joint commitments occur in everyday life, Gilbert views plural subjects as paradigmatic social phenomena.Footnote 69 Her claim is ontological: the joint commitments that give rise to plural subjects, she argues, form the very foundation of human sociability – they constitute ‘the structure of the social atom.’Footnote 70 As such, there is no principled obstacle to extending the plural subject approach beyond informal groupings like John and Alice to IOs and other more or less complex collectives. What matters is not the form a group takes – whether formal or informal, small-scale or large – but whether it is grounded in a joint commitment to act as a body, the defining feature of a plural subject. While joint commitment is constitutive of plural subjects, my discussion will begin by examining Gilbert’s concepts of a body and of doing – or X-ing – in order to clarify the kinds of entities and actions such commitment brings into being, before turning more directly to the nature of joint commitment itself.
Gilbert’s notion of a body
By using the word ‘body’ in the phrase ‘doing as a body,’ Gilbert conveys that a joint commitment among group participants is a commitment to act in unity. She acknowledges that this body is not equivalent to a human body and always remains ‘in some sense plural.’Footnote 71 Yet, this plurality is still capable of ‘emulat[ing] as far as possible a single body’Footnote 72 and becoming ‘unified in such a way as to count as the subject of a single intention – the intention that is ours.’Footnote 73 Importantly, doing something as a body is not about each individual doing the same thing but rather about everyone acting in a way that unifies a body to perform the action.Footnote 74 Gilbert also emphasizes that she is not wedded to the phrasing ‘doing as a body’ – she merely prefers it over other possible ways of expressing the idea of unified plurality. Alternatives, she notes, may include ‘doing as a unit,’ ‘doing as one,’ or, particularly relevant for our discussion, doing as a ‘person.’Footnote 75
That Gilbert allows plural subjects to be represented as persons is neither coincidental nor metaphorical. While she clearly distinguishes between plural and individual persons, she permits a degree of conceptual overlap between the two. This is because, as we will see next, plural subjects can engage in multiple actions typically associated with personhood, such as forming intentions, holding beliefs, and expressing emotions. More fundamentally, plural subjects do not engage in forms of action that lie outside the domain of human experience. On the contrary, and it is crucial to underscore this, everything they do remains firmly within the realm of human action. There is nothing in the language or logic of joint commitment that introduces an alien or supra-human mode of agency. This is precisely why the personification of plural subjects (including in the case of IOs) is not merely a rhetorical device, but a reflection of the kinds of actions, responsibilities, and normative expectations such collectives can undertake. To echo Wendt, were personification merely rhetorical, it would be possible to dispense with it or replace it with another metaphor.
Gilbert’s notion of doing (X-ing)
Against this background, Gilbert adopts a broad understanding of doing, an approach she formalizes as ‘X-ing as a body.’ She thus argues that people can jointly commit not only to ordinary and externally observable actions but also to complex cognitive and psychological states, such as intentions, beliefs, and emotions.Footnote 76 To illustrate how these distinctly human characteristics operate at the group level, Gilbert examines how plural subjects can express doubt, guilt, and remorse.Footnote 77 Contra Wendt, she maintains that such emotions do not require phenomenological manifestations and therefore do not strictly depend on consciousness.Footnote 78 Instead, they function as cognitive and normative categories, embedded and expressed in and made meaningful by specific social contexts. As a result, a group need not possess a collective mind to experience doubt, guilt, or remorse – nor must any or all of its members personally experience these emotions. Gilbert’s only requirement is that the members jointly commit to expressing the emotion as a body or, indeed, a person. Compared to Wendt’s higher-order markers of personhood, which rely on the notion of emergence, Gilbert’s approach is more modest and supervenient in nature, yet it still expands our understanding of what groups can do. This allows her to avoid the criticism Wight levels at Wendt – that he captures group agency but not necessarily the traits of group personhood.
A further implication of Gilbert’s broad understanding of doing is that a joint commitment to intend something is sufficient to constitute a plural subject; such a commitment need not culminate in joint action. In other words, a plural subject does not require the immediate – or even practical – possibility of acting on its intention. That said, not acting on a jointly intended goal may eventually lead to the reconsideration or dissolution of the plural subject, but actual execution is not a precondition for its existence. What defines the plural subject is, therefore, not the capacity for action (or even action capability) but the capacity to form a joint commitment. This commitment constitutes an action in its own right: an act of intending as a body. Alice and Scott can thus jointly commit to go shopping – and thereby form a plural subject – even if they are on an island without access to shops. The agency at stake is not reducible to physical action; it lies instead in their shared capacity to commit to doing something together, to recognize that commitment, and to hold one another accountable within it, whether in the present or at a later time.Footnote 79 For Gilbert, this capacity is foundational to collective agency and underpins the person-like features ascribed not only to informal groupings but also to institutional actors such as IOs – even when they do not, or cannot, act on their commitments immediately or, indeed, ever.
Gilbert’s notion of joint commitment
Now, let us turn to joint commitment, the process that brings supervening plural subjects – i.e., their body and doing – to life.Footnote 80 Gilbert has arrived at joint commitment as the foundation of plural subjects by insisting that aggregating individual preferences, intentions, and commitments is untenable when individuals do things together. As previously discussed, John and Alice did not decide to go shopping together merely because John expressed his personal preference, intention, and commitment to shop, followed by Alice doing the same. Nor did their decision result from both of them expressing these intentions and commitments simultaneously.Footnote 81 Instead, Gilbert contends, their intention to act together came about through their expression of a ‘readiness to be jointly committed to espouse the relevant goal as a body.’Footnote 82 Accordingly, for John and Alice to go shopping together, John had to express to Alice his readiness to do so jointly, and Alice had to express the same readiness to John.
Therefore, contrary to reductionists such as Bratman, who argue that group intentions are merely complexes of interlocking personal commitments,Footnote 83 Gilbert views them as joint commitments of the involved parties to act as the components of a unified entity with a particular goal. Importantly, Gilbert insists that joint commitments are not, strictly speaking, discrete components of a plural subject.Footnote 84 A joint commitment is only actualized when all parties have expressed their readiness to be jointly committed to the relevant goal as a unified body. Only then can we speak of a plural subject represented by the sentence ‘We intend to go shopping’ rather than ‘We both intend to go shopping.’ Gilbert further argues that once a joint commitment of this kind is established, the parties become subjects to a commitment that depends on them but of which they are not the sole authors.Footnote 85 Joint commitment and the resulting plural subject, therefore, supervene on their participants, thereby also exerting a form of downward causation on them. As Gilbert puts it, they become a ‘command center,’ instructing the parties to act accordingly – even if this ultimately proves unfeasible and becomes a reason to rescind the joint commitment, thereby dissolving the corresponding plural subject.Footnote 86
Joint commitment and its relation to individual interests, intentions, preferences, and commitments
For our analysis of IOs, it is crucial to underscore that Gilbert does not posit a necessary link between joint commitment – the basis of plural subjects – and the individual interests, intentions, preferences, or commitments of the actors involved.Footnote 87 This contrasts sharply with dominant IR theories, which, as discussed earlier, typically assess IO agency based on its deviation from state preferences. Gilbert’s framework resists such logic. For her, what matters is not the nature of what is jointly committed to but the act of joint commitment itself.Footnote 88 The content of a joint commitment (i.e., the x-ing) may align with individual preferences, or it may not.Footnote 89 For instance, John and Alice might jointly commit to go shopping because they both want to. But they could just as easily commit to it without either of them intending to go. As she turns away, Alice might say, ‘I’m heading back. I never intended to go to the mall,’ to which John could respond, ‘Same here.’ Despite lacking corresponding individual intentions to go shopping, their earlier joint commitment still formed a plural subject. Furthermore, for Gilbert, the motivations for or sources of joint commitment, whether personal, social, cultural, or institutional, also do not determine the existence of a plural subject. These factors may explain why actors made a commitment or what they committed to, but they do not determine the formation of group agency. What matters is solely the act of jointly committing to do something as a body. This logic also extends to complex emotional or psychological states. A group may thus jointly commit to express a collective emotion, such as guilt, by saying, ‘We feel guilty for the harm we caused,’ regardless of whether any individual member truly feels guilty or whether the commitment arose from social pressure or cultural and institutional norms.Footnote 90 It is always the joint commitment itself, and nothing else, that constitutes the group as a collective agent.
Rescinding of joint commitment
Gilbert’s non-summative conception of plural subjects carries an important implication: members of a plural subject cannot unilaterally rescind a joint commitment.Footnote 91 This contrasts with personal commitments, such as John’s decision to go shopping alone, where no such constraint exists. As the sole author of his commitment, John can change his mind at any moment.Footnote 92 Joint commitments, however, work differently. Imagine John and Alice on their way to the mall when Alice suddenly stops, says nothing, and turns back. Surprised, John objects, ‘What are you doing? We planned to go shopping. You cannot just turn back!’ Most of us would find John’s reaction understandable, and Alice might even acknowledge his right to respond this way. John thus seems to hold a certain standing due to the joint commitment he and Alice made. Since Alice is acting within the bounds of this commitment, she would need John’s agreement to change their plans. Participants in a joint commitment, therefore, understand that they cannot dissolve this commitment unilaterally and at will; they can only violate it with consequences.Footnote 93 Gilbert explains this using the language of rights and obligations: each participant, she maintains, has a right to the relevant action and performance by the others.Footnote 94 They owe each other behavioral conformity and are obliged to pursue the shared intention as best they can.Footnote 95 And, as noted earlier, an inability to do so may provide grounds for dissolving the plural subject.
In summary, Gilbert’s ontological conception of group agency as plural subjects departs significantly from earlier accounts. It challenges the predominant view in IR that IOs qualify as agents in their own right only when their actions diverge from member state interests, as well as giving intention priority over action. In contrast to Wendt, it offers a clear mechanism – i.e., joint commitment – for how supervenient group agents take shape and, by privileging cognition over phenomenology, shows that joint commitment can endow groups with psychological traits such as emotions. This non-reductionist perspective also provides a foundation for challenging the transient conceptions of IO agency proposed by Erskine and, to a degree, by Gehring and Urbanski. The following section applies these insights to develop a new account of IO agency and its associated personification.
International organizations as Gilbertian plural subjects
As shown previously, for eighty years, the language of Security Council resolutions has consistently constructed the Council as a unified, person-like actor, seemingly capable of action, intention, belief, and even emotion. This reflects a deeply embedded diplomatic practice in which states, even when primarily motivated by the pursuit of national interests, must advance their positions by translating them into the language of a collective body. Gilbert’s notion of plural subjects allows us to move beyond viewing this personification as a mere rhetorical convention. Instead, it invites us to understand it as a constitutive practice, one that transforms the Security Council into a supervenient group agent, brought into being through the joint commitments of its member states.
While there are no definitive rules governing the drafting of a Security Council resolution, established practices – some more formal than others – demonstrate that the drafting process typically unfolds through five stages, followed by voting in the Council’s chamber.Footnote 96 At the outset, one state or a group of states takes the initiative to prepare the draft resolution and maintains control over the subsequent drafting process. In the second stage, these states engage with other missions that are particularly interested in the issue, even if they are not members of the Security Council, to discuss the draft. In the third stage, the draft is shared with other Security Council members, either bilaterally or multilaterally. The fourth stage involves detailed, paragraph-by-paragraph discussions by all Council members during informal consultations or within unofficial groups. Finally, in the fifth stage, the text of the resolution is circulated as an official Security Council document.
These five stages highlight several important aspects of the resolution-drafting process. From the very beginning, states structure their negotiations in the Security Council around draft resolutions, effectively positioning the Council as the main vehicle for any proposed action. Whatever proposals they put forward or agreements they reach must, therefore, be framed in a way that enables the Security Council to serve as the implementing agent. Therefore, the negotiations, both in practice and outcome, cannot be described as a simple aggregation of power-weighted state interests. States, in other words, do not approach one another by asking, ‘My state plans to do such and such; what does your state plan to do?’ and then base the agreement on how closely their positions align. Instead, through negotiations, they seek a point at which the majority of involved states is willing to jointly commit to a course of action as a body, the Security Council. When presenting a draft resolution, the resolution sponsors therefore implicitly pose a Gilbertian question: ‘Our state(s) are ready to make a commitment to espouse these actions, intentions, beliefs, and emotion as a body called the Security Council. Is your state equally ready to commit?’ The final draft of the resolution is, therefore, the one in which a substantial number of member states express such readiness – that is, reach a joint commitment. Resolution drafting is thus not merely a procedural exercise but also a constitutive act: a practice through which the Security Council is formed and sustained as a plural subject or agent in its own right. At the same time, given that every joint commitment is a commitment to something, the practice of resolution drafting inevitably imbues the Council with person-like characteristics, such as authoritativeness (e.g., decide, direct, request), the expression of opinion and establishment of facts (e.g., determine, confirm, maintain), the capacity to pass judgment (e.g., praise, salute, warn), and the expression of emotion (e.g., concerned, outraged, grieving), as discussed earlier.
One important implication of Gilbert’s theory is that all joint commitments – whether to a visible action, an intention, a belief, or even an emotion – carry equal constitutive weight in forming and sustaining the Security Council as a plural subject. Therefore, while such a scenario may seem unlikely, Gilbert’s framework allows, in principle, for a Security Council whose members are solely committed to jointly expressing emotions to qualify as an agent in its own right. This is because the defining feature of a plural subject is not its capacity for action or operational capability but its capacity to form joint commitments. Put differently, what makes the Security Council an agent is not its ability to act on intentions but the collective willingness of its members to forge and sustain a joint commitment – to adopt a common stance, recognize its binding nature, and hold one another accountable to it. Even if the Council is, for various reasons, unable to act on its commitments – as is often the case – this inability does not immediately negate its status as a plural subject. Over time, however, a persistent failure to fulfill its joint commitments may serve as grounds for reconsidering or even dissolving the Council.
Negotiations on Security Council draft resolutions – both regarding the issues addressed and the motives driving states to express or withhold their readiness to commit – can and do hinge on a wide range of considerations. Normative, cultural, institutional, and geopolitical logic can influence which issues make it onto the agenda and shape the final form of the resolution. Also, in jointly committing to these resolutions, states may be guided by national interests and preferences, as well as by political pressure, power asymmetries, issue linkages, and other strategic or contextual factors. Yet, regardless of these factors, and following Gilbert, what ultimately gives rise to the Security Council as a plural subject with agency and traits of personhood is not why states commit to a resolution, what they committed to, or how closely it aligns with state interests and preferences, but rather the very act of jointly committing to do something as a body called the Security Council. Some states may support a resolution because it reflects their interests; others may agree despite reservations, driven by pressure or the need for compromise. Neither of these motives, however, undermines the validity of their joint commitment to constituting the Security Council as a plural subject.Footnote 97 Gilbert’s interpretation of collective agency thus marks a significant departure from dominant thinking in IR, which, as indicated throughout the paper, typically defines the agency of IOs in terms of acts and decisions that diverge from state interests, whether through bureaucratic independence or institutional logics overriding individual state preferences.
This point can be further reinforced by applying Gilbert’s perspective to the Security Council’s voting procedure – the moment when draft resolutions either become adopted resolutions or fail to do so, and, by extension, when certain intentions, actions, beliefs, and emotions are either attributed to the Security Council or not. Voting, therefore, represents a final expression of joint commitment. A resolution is adopted when it receives nine votes in favor, including the cases when a permanent member chooses to abstain. As with acquiescing to or rejecting the draft version of a resolution or its parts, a vote on the final draft need not reflect state interests and preferences. Instead, it is merely an institutional mechanism by which states jointly commit to espousing certain intentions, actions, beliefs, and emotions as a body called the Security Council. But, voting also has another important peculiarity. While the voting rules are designed to determine joint commitment, they are themselves the product of a basic joint commitment made by states when the Security Council was formed: the commitment to accept as the Council’s intentions, actions, beliefs, or emotions those that emerge through this voting procedure. This means that once a majority is achieved, even states that voted against the resolution or abstained are jointly committed to it by virtue of their accepting the voting process. In other words, when a majority is achieved, both positive and negative votes represent an expression of joint commitment.
Once a resolution is adopted, another characteristic of plural subjects comes into play: member states cannot unilaterally rescind the commitments they made or undo the plural subjects that resulted from them. By jointly committing to espousing specific action as a body called the Security Council, they have co-created a complex of rights and obligations that they can only violate but not dissolve individually. An example could illustrate this point. When the United States, along with its Coalition of the Willing, invaded Iraq, it did not claim the right to do so when accused of violating Resolution 1441, nor did it remain silent on the matter. The U.S. leadership at the time understood that it was entangled in a web of international rights and obligations, extending beyond Resolution 1441 to international law more broadly, none of which could be unilaterally rescinded. Consequently, the U.S. administration provided a detailed justification, asserting that its actions were, in fact, consistent with Resolution 1441 and that, through an intricate system of references between resolutions, they could also justify the invasion of Iraq based on the earlier Resolution 678. The key point here is that, regardless of its intentions, the United States acknowledged its entanglement in the web of rights and obligations established through previous joint commitments. This situation is analogous to our example in which Alice was expected to justify herself to John before turning and leaving.
Given these considerations, it is crucial to underscore two further dimensions of the Security Council’s status as a plural subject. First, since Gilbert’s conception of plural subjects is ontological, the Council should not be viewed as a special instance of collective agency. Rather, it represents a particular instantiation of a plural subject. For this reason, the example of John and Alice is not merely a didactic simplification but an ontological equivalent to the Security Council, albeit one that is vastly less complex.Footnote 98 On this basis, the notion of joint commitment and plural subjects should apply not only to other IOs, but also, importantly, to other forms of state cooperation, such as bilateral and multilateral treaties and conventions. At the same time, caution is warranted: the concept of plural subjects should not be extended to encompass structures such as international norms, customary or codified law, or institutional regimes. Nonetheless, it can be reasonably argued that many of these structures – though not all – are the product of actions and joint commitments undertaken by plural subjects.
What differentiates an IO from other forms of state cooperation is thus not the possession of collective agency but the particular institutional form it assumes. For instance, beyond differences in organizational complexity, such as the presence of formal bodies and decision-making procedures, a key distinction between an IO and a bilateral treaty lies in the singularized institutional name. In an IO, all joint commitments are attributed to the named entity – in our case, the Security Council – rather than to a collective ‘we.’Footnote 99 This practice reinforces the organization’s continuity and contributes to the formation of its distinct social identity. Gilbert’s notion of the plural subject thus invites a clear distinction between the process that constitutes the agency of an IO – i.e., joint commitment – and the institutional form that such commitment takes, and through which further acts of joint commitment are enabled. This distinction is often overlooked by scholars of international relations, who, as previously noted, tend to treat organizational features such as voting procedures and resource pooling as themselves constitutive of IO actorness, rather than as one of many social forms through which joint commitment can be realized.
Second, notwithstanding this distinction, the institutional complexity within which the Security Council and other IOs are embedded should not be glossed over. The United Nations, of which the Council is a part, comprises multiple distinct bodies and agencies, each performing specific functions and pursuing its own agenda. The UN, in other words, is a site of complex interplay among various forms of collective agency: authorized components such as the bureaucracy, which manage administrative and policy responsibilities, and member states, which engage in deliberative and executive roles. Gilbert addresses this complexity by positing that social units founded on joint commitments can exist within other such units. Consequently, plural subjects can be nested within one another, forming an intricate network of what Gilbert terms basic and derived joint commitments.Footnote 100 A basic joint commitment thus gives rise to a derived one when it leads to the appointment of a new authority responsible for advancing goals that, while different, remain closely tied to the original agreement.Footnote 101
The UN Charter can be understood as a collection of basic joint commitments – an expression of the participating states’ 1945 agreement to establish a body named the United Nations, tasked with maintaining international peace and security and promoting social progress. Among these commitments are the creation of the Security Council and the specification of its composition and mode of operation. The joint commitments exercised within the Council are thus derived from this basic commitment, requiring member states to continually espouse actions, intentions, beliefs, and emotions collectively, as a body named the Security Council, in pursuit of its mandate. This may appear to introduce a tension in the Council’s ontological status. On the one hand, the Security Council is constituted on an ongoing basis through member states’ joint commitments, formalized in its resolutions. On the other hand, it derives from a basic joint commitment established in the UN Charter. Yet these claims are not contradictory. Rather, the Charter functions as a basic joint commitment that creates the institutional space – the named body – within which further, derived commitments are made. Each resolution sustains the Council as a plural subject, while the Charter remains the foundational source of its authority and identity. In this way, the Council’s agency is both historically rooted and actively maintained.
Theoretical and empirical implications of treating international organizations as Gilbertian plural subjects
While framing the Security Council as a Gilbertian plural subject – an autonomous agent possessing attributes of personhood – is conceptually meaningful in itself, its full significance emerges when we consider its theoretical and empirical implications. These are particularly salient in two domains central to the study of IOs: institutional moral agency and authority.
Erskine has already argued that the Security Council can be considered an institutional moral agent, subject to praise or blame for what it does or fails to do.Footnote 102 However, as indicated earlier, she treats this status as transient rather than permanent. It depends on several criteria for institutional actorness, most importantly, the Council possessing a distinct identity and the capacity to act independently of its member states.Footnote 103 For Erskine, decisions reached through majoritarian procedures demonstrate such independence and, by extension, support the claim of institutional moral agency. However, when a permanent member exercises its veto, she contends that the Council’s capacity for moral deliberation – and thus its independent agency – is compromised. In such cases, the Council can no longer be said to act as a moral agent.Footnote 104
In contrast, viewing the Security Council as a plural subject allows us to move beyond the idea of institutional moral agency as transient, and instead understand it as an enduring feature. Gilbert’s distinction between basic and derived joint commitments is key here. It shows that resolutions are not a precondition for the Council’s existence, but rather context-specific expressions that sustain, in specific contexts, a plural subject already established through the basic joint commitment embodied in the UN Charter. Its moral agency, therefore, does not depend on constant re-articulation through resolutions. When a resolution is rejected or blocked, this does not dissolve the underlying basic commitment; it simply reflects a failure to act on it and to sustain the plural subject in a given context. In Gilbert’s terms, the basic joint commitment is violated when it fails to materialize into a derived one, yet the rights and obligations it establishes – and which confer its moral weight – remain intact. Inaction resulting from a permanent member’s use of the veto is thus a morally significant event, rather than, as Erskine suggests, a moment in which the Security Council’s agency dissolves. Such inaction may render the Council itself – not just the states exercising the veto – subject to moral blame. The Security Council’s moral agency also persists regardless of whether resolutions are adopted by majority or unanimous vote. This is because a basic joint commitment always underlies its existence, and because, in Gilbert’s framework, voting procedures are not a precondition for agency, but merely one of the many ways joint commitment can be expressed.
With the moral agency of the Security Council understood as an enduring feature of its status as a plural subject, we can shift the focus from questioning whether it possesses this agency to analyzing how it exercises it, thereby opening new avenues for both theoretical and empirical research. For instance, the nature of the Security Council’s personification can be explored by examining the language through which it conveys moral judgment, and vice versa. My analysis, for example, shows that verbs personifying the Council are often accompanied by modifiers that intensify the effect of moral judgment. The phrase ‘is concerned’ is frequently paired with modifiers such as deeply, gravely, or strongly. Similarly, when the Council ‘calls,’ it often does so solemnly, and when it ‘requests,’ it does so urgently. These linguistic patterns may offer further insight into institutional learning and change. According to Gilbert, blameworthiness, praiseworthiness, and moral judgments – both internal and external – can prompt reflexivity, encouraging the plural subject to realign its values and actions. Shifts in the verbs used in resolutions reflect this dynamic. For example, the Council became ‘concerned’ in the 1950s but was most ‘concerned’ in the mid-1990s, during a surge in ethnic conflicts. ‘Requests’ followed a similar trend, while ‘calls’ have steadily increased since the mid-1980s. These developments may help us trace how the Security Council’s moral judgments shift with political context and evolve in response to the prevailing values of their time.
The authority of IOs – another institutional feature for which the plural subject approach has important implications – is often defined by examining their role as governors of specific social domains or issue areas. Michael Zürn and colleagues thus define IO authority as the ability to perform regulatory functions, consisting of rule-making, enforcement, and compliance monitoring.Footnote 105 Others, such as Cooper and colleagues, are more explicit about what this authority means for states. They argue that IOs possess political authority when states acknowledge these organizations’ ability to make binding decisions on matters within states’ domestic jurisdictions.Footnote 106 In this interpretation, IO authority represents a force that contradicts state sovereignty. It can be argued that viewing IOs from a Gilbertian plural subject perspective introduces an additional layer, or at the very least, a new perspective on IO authority. When member states jointly commit to Security Council resolutions, encompassing shared intentions, actions, beliefs, and emotions, they are expected to align their behavior with these commitments and uphold the collective goal to the best of their ability. The authority of the plural subject arises from this binding commitment, compelling members to act in accordance with the group’s objectives, thus effectively granting the collective entity’s influence over individual behavior. Members cannot unilaterally withdraw from these commitments and are mutually accountable for fulfilling their roles, further reinforcing the group’s authority through shared responsibility.
Accordingly, as noted by Gilbert, members of a plural subject operate within a framework of rights and obligations, where authority functions more like a social norm, shaping behavior through accepted principles rather than through formal governance or binding decisions on domestic matters. The example of the United States and its Coalition of the Willing attempting to justify the invasion of Iraq through a complex web of mutually referencing resolutions illustrates the powerful influence that rights and obligations established through joint commitments can exert, including on the most dominant states. Future research could, therefore, examine how IOs, as plural subjects, shape member states’ behavior even when they do not encroach on state sovereignty or engage in norm creation, compliance monitoring, or enforcement.
Conclusion
The independent actorness of IOs, though not universally accepted, has garnered substantial theoretical support. Rational institutionalists and constructivists explain it by focusing on inter-organizational staff and the various rational and cultural factors that influence them. Gehring and Urbanski, by contrast, identify this actorness in member-dominated IOs, attributing it to the action capability derived from pooled resources and autonomous decision-making by member states. However, these accounts are neither definitive nor comprehensive, as they have yet to fully address the empirical reality surrounding the actorness of IOs. Notably, they fail to recognize that this actorness is not exercised in abstraction, but rather through the assignment of traits of personhood to IOs. By also focusing on member-dominated IOs – specifically the Security Council – this paper has sought to introduce a novel perspective on IO actorness and personhood by applying Margaret Gilbert’s notion of plural subjects. It has been argued that, as plural subjects in everyday life, the Security Council’s actorness and traits of personhood emerge through member states jointly committing, via resolutions, to espouse certain actions, intentions, beliefs, and emotions as a single body known as the Security Council. In this process, as demonstrated by the analysis of the resolutions’ action sentences, the Security Council assumes a robust ‘personhood,’ endowed with authority characteristics, the ability to express opinions and establish facts, pass judgments, and convey emotions.
Gilbert’s perspective on collective actorness, when applied to IOs, challenges several long-standing assumptions in IR. Most notably, it rejects the view that individuals – and by extension, member states – can act collectively in a purely summative way, where personal interests and preferences are simply aggregated. Instead, Gilbert argues that whenever individuals act together toward a shared goal, they form a plural subject, an entity she understands in strong ontological terms as foundational to human sociability. From this standpoint, the claim that IOs qualify as actors only when their behavior can be clearly separated from the interests of member states becomes unnecessary. For Gilbert, such separation is not required: plural subjects – i.e., collective agency and personhood traits – emerge through joint commitment, regardless of whether those commitments align with the individual state preferences or not. This non-summative view of agency, coupled with her distinction between basic and derived joint commitments, also implies that IOs do not fluctuate between being and not being agents; rather, agency is a stable feature of their existence. Finally, by grounding collective actorness in broader patterns of human sociability, Gilbert’s framework also enables a conceptual distinction between the social processes that constitute IO agency and traits of personhood, and those that form their formal institutional structure.
Against this background, the Gilbertian approach to IO actorness and personhood traits has the potential to spark new debates, generate fresh theoretical insights, and encourage innovative empirical investigation in the field of global governance. First, it might prompt IO scholars to pay more attention to ontological and meta-theoretical questions, much like the ‘state-as-a-person’ debate, where issues of supervenience, emergence, and the nature of language have yielded many productive insights and disciplinary self-reflection. Second, it could offer new perspectives on theoretical, normative, and empirical questions related to IOs, as demonstrated in discussions of their institutional moral agency and authority. Finally, Gilbert’s approach may shed light on the broader question of actorness in IR, not only by extending to other actors, such as NGOs, multinational corporations, or coalitions of the willing, but also by encouraging the examination of the processes through which collective agency emerges.
Acknowledgments
I would like to express my sincere gratitude to Yushu Soon for assisting with the literature review, to Strahinja Mišić for his help with the analysis of Security Council resolutions, and to my colleagues who provided insightful comments during the WZB Global Governance colloquium.
Competing interests
The author declares none.
Open access
