Change is a constant in international politics and takes many forms. One type of change that has lurked at the edges of scholarly discussions of international politics—often assumed, invoked, and alluded to, but rarely interrogated—is learning. Footnote 1 Learning entails a very particular type of change. It is deliberate, internal, transformative, and peaceful (in the sense of being uncoerced). Moreover, and crucially, the potential for this type of change attaches only to a specific type of entity. Systems and structures—whether the so-called liberal international order or the market—cannot learn, at least not in a way that is anything other than merely metaphorical. (They can adapt and evolve.) Norms are incapable of learning. (The actors that variously champion, flout, and are constrained by them can, however, learn.) And it makes no sense to describe informal associations such as social movements, transnational advocacy networks, and noninstitutionalized intergovernmental groupings like the G-20 or G-7 as learning. (Their discrete constituents may be capable of learning in ways that affect how they interact and collectively exert influence, but these associations themselves are not potential learners.) Learning is necessarily the purview of entities that are purposive agents.
We most often consider learning in the context of individual human beings. Yet, here, I want to ask whether the class of entity capable of learning also includes the corporate agents that are central to international politics. Can formal organizations, or what we might call “structured institutions,”Footnote 2 learn in a way that is not reducible to their individual members? Asking this question matters for (at least) two reasons.
First, there are intriguing assumptions made across a range of theoretical approaches in the discipline of international relations (IR) that corporate entities—particularly states, but, according to some, also intergovernmental organizations (IGOs)—are agents imbued with interests, aims, and impressive aptitudes for reasoning and acting in ways analogous to individual human actors. In short, they possess capacities that would seem to make some conception of learning a logical and unavoidable corollary. I have argued elsewhere that formal organizations in international politics—including (most) states, transnational corporations, and, at least transiently, IGOs—can possess the sophisticated, integrated capacities for deliberation and action necessary to qualify as moral agents, or bodies to which we can coherently assign moral responsibilities and apportion blame.Footnote 3 The proposal that formal organizations can learn follows, and extends, this work on what I call “institutional moral agency.” The capacities for self-reflection and understanding the consequences of one's acts and omissions that help define moral agency are exactly what is required for learning. Like investigating the moral agency of formal organizations in international politics, exploring the viability of “institutional learning” takes seriously, and lends substance to, common IR conceptions of corporate entities as purposive agents.
Second, outside scholarly debates, bodies like credit-rating agencies in the context of the global financial crisis, BP after the Deepwater Horizon oil spill, the United Nations following tragedies in Rwanda and Srebrenica, and particular states in relation to the COVID-19 pandemic are commonly described as appropriate objects of blame in practical discourses about international politics. When an organization's acts or omissions are deemed to have led to harm, we often direct charges of moral failure at the organization itself. We identify and lament its missteps that led to wrongdoing or a serious abrogation of responsibilities. Moreover, we seem to assume that such censure can prompt the sort of self-reflection necessary for the organization to consider the consequences of what it did—or failed to do—and chart a different path in the future. Asking whether it is possible for such bodies to learn aims to shed light on the appropriateness of such responses. In other words, if corporate bodies like the UN are not the sorts of entities that can engage in critical self-reflection, consider the consequences of their past acts and omissions, and endeavor to behave differently as a result, then our cries of condemnation and demands that they do better are likely misdirected.
Can formal organizations in international politics learn in a way that is comparable to the paradigmatic learning of individual human beings? What would such institutional learning look like and what sets it apart from other types of change? In the context of the present roundtable on international institutions and peaceful change, I will focus on structured international institutions, or IGOs—bodies such as the UN, the North Atlantic Treaty Organization (NATO), the Arab League, and the World Trade Organization (WTO)—in exploring these questions. This category of actor introduces an additional element of complexity. By suggesting that IGOs might learn, we are talking about a process of deliberate, internal change that is not at the familiar level of the individual human actor, or even at the level of a formal organization such as the state, but, rather, at the level of a corporate entity that is itself constituted by multiple groups, integrated, and formally organized. In other words, IGOs add another level of analysis to the already provocative proposition that formal organizations in international politics are capable of learning.
My starting point in questioning the possibility of institutional learning in formal organizations generally, and IGOs specifically, is an understanding of learning as a process of reflecting on past experiences—and the consequences of previous acts and omissions—in a way that leads to enduring change to subsequent behavior. This, to me, encompasses the central components of what we understand as learning in individual human agents. In addressing the puzzle of how—and indeed whether—this concept can be meaningfully applied to IGOs, I will take three steps. First, I will explore references to corporate entities “learning” within IR and ask whether what is being proposed is, in fact, genuine learning by the organizations themselves. Second, I will attempt to construct a robust account of institutional learning that departs from these conceptions and acknowledges instead the self-reflection and structural transformation that I will argue learning at the corporate level requires. Third, for the purpose of illustration, I will turn briefly to the UN following the 1994 genocide in Rwanda and the slaughter of more than eight thousand men and boys outside Srebrenica in 1995 to identify examples of each stage of institutional learning. Finally, I will offer three provisional claims about my proposed conception of institutional learning that warrant attention in future work. Namely, I will suggest that institutional learning: (1) cannot be equated with moral progress; (2) is possible despite formal organizations being incapable of emotional responses such as shame or regret; and, perhaps most controversially, (3) can occur at the level of the IGO without prior or parallel learning taking place at the level of the state or individual human actor.
References to Corporate Agents “Learning” in the Discipline of IR
A range of studies in IR make explicit reference to learning in relation to corporate agents. For example, employing the conceptual categories of “learning” and “teaching,” Martha Finnemore provides a pathbreaking constructivist account of how international organizations can redefine state interests and behavior. Although her ultimate focus is on teaching, she makes repeated reference to states learning and acknowledges that this involves a process that “changes and reconstitutes” states.Footnote 4 Yet, the process of learning remains largely unexplored, aside from her observation that the targeted pedagogical pursuits in which international organizations engage result in states’ transformed preferences. International organizations, she maintains, are “active teachers with well-defined lesson plans for their pupils [states].”Footnote 5 Richard Price aptly describes Finnemore's teaching as a form of “socialization”—one that “encompass[es] the critical pedagogical ingredients of information, persuasion, shame, and discipline that are the tools available to the otherwise underpowered agents of transnational civil society.”Footnote 6 He refers to the same process as “moral proselytizing,”Footnote 7 thereby underlining the point that, while learning on the part of states is assumed, it is deemed to be a largely passive process that involves being socialized in the course of concerted attempts at norm promotion and diffusion.
In a very different type of analysis, G. John Ikenberry invokes the concept of learning in the context of “the spread of new information,” whereby governments “copy policies that are developed elsewhere.”Footnote 8 In an observation that reveals as much about her own understanding of the learning experienced by states as it does about Ikenberry's, Finnemore describes Ikenberry's conception of learning as entailing that states are “self-taught,” as his account does not rely on agents external to the state assuming the role of teachers.Footnote 9 The understanding of institutional learning that I aim to explore here may involve corporate agents being influenced by external factors (such as censure in response to perceived moral failure), but there is an important sense in which I also see their learning as “self-taught.” The conception of institutional learning that I want to propose is an internal process—although it may have external impetuses. It is not merely a response to persuasion, social influence, or lobbying. In other words, it is necessarily reflexive and not merely reactive. Yet, this conception also diverges in important ways from Ikenberry's conception of states learning. Ikenberry ties learning to emulation, whereas I see learning as necessarily involving critical reflection on past experiences. Both Ikenberry and Finnemore rely on conceptions of learning that ignore requisite internal cognitive processes. Ikenberry neglects critical self-reflection, whereas Finnemore sidesteps the question of what constitutes the internal processes of learning altogether.
In an important example of learning in foreign policy analysis, Jack Levy both defends an account of what he calls “organizational learning” and embraces a conception of learning that resonates with my own in entailing reflection on past experiences.Footnote 10 However, he sees learning as the exclusive purview of individual human actors—leaders and policymakers—rather than occurring at the level of the organization itself. For Levy, “the reification of learning to the collective level—and the assumption that organizations or governments can be treated as organisms that have goals, beliefs, and memories—is not analytically viable.”Footnote 11 What he promisingly labels organizational learning is, in fact, “the institutionalization of individually learned lessons into organizational routines and procedures.”Footnote 12 Only individual human beings learn. Learning occurs when individual human actors observe and interpret their own experiences in a way that results in their acquiring skills or in changed, confirmed, or new beliefs.Footnote 13 These individual human learners then impose changes on the organization.
Ernst Haas's seminal cognitive evolution approach, which focuses on learning to improve the performance of IGOs such as the UN and the World Bank, also understands learning as an internal process that is transformative and reflexive. Moreover, for Haas, learning is ostensibly not the exclusive purview of individual human actors. This seems to bode well for a corporate conception of learning. Yet, like Levy, he also asserts that it is not the organization as a whole that learns. According to Haas, saying that “an international organization learns” is merely
a shorthand way to say that the actors representing states and members of the secretariat, working together in the organization in the search for solutions to problems on the agenda, have agreed on a new way of conceptualizing the problem.Footnote 14
He argues that learning belongs to collective entities—“not individuals, entire governments, blocs of governments, or entire organizations,” but “clusters of bureaucratic units within governments and organizations.”Footnote 15 We might describe his conception as “shared learning” within informal associations—learning that is ultimately distributive among its individual human constituents. In other words, it cannot take us as far as what I am calling institutional learning.Footnote 16
Institutional Learning: Beyond Socialization, Emulation, and Aggregated Individual Learning
The terminology that I have chosen in labeling both “institutional moral agents” and “institutional learning” is significant—and helps to convey features that are fundamental to both. The label “institution” can, of course, mean different things.Footnote 17 Even though I am focusing explicitly on institutions in the sense of formal organizations, another connotation of the term draws attention to the norms, rules, procedures, practices, and cultures that frame and channel the intentions and actions of the purposive agents within the organization in question. The label institutional moral agent thereby also highlights why certain types of group can reach decisions and act in ways that cannot be described simply in terms of the sum of the decisions and actions of their constituents. One might argue that groups with certain types of decision-making structures can have something analogous to intentions, or wills, that are not reducible to the intentions of their constituents. In the case of IGOs, states (and sometimes other IGOs)Footnote 18 are the purposive agents whose individual decisions, actions, and intentions can be framed and channeled within a broader intergovernmental structure and decision-making procedure to contribute to a distinct corporate agent with something resembling a will of its own.
Institutional learning, by the same logic, should be understood as more than merely the sum of individuals’ learning within an organization at any one point in time. The individual learning of human agents within a group can, of course, affect the aggregate behavior of the group, and in potentially significant ways, but this does not seem enough to count as institutional learning. Such learning remains tied to a determinate membership and is lost with changes to the composition of the group over time. Levy gets around this by having his individual human learners make changes to organizational procedures; yet, this results in institutional change as a result of individual learning, not institutional learning. This conceptual distinction is crucial.
Institutional learning, to be worthy of the label, must involve two processes most accurately described as occurring at the level of the organization rather than the individual. The first is the process of reflection and deliberation on past experiences, and on the consequences of previous acts and omissions, resulting in a commitment to pursue particular actions or outcomes in the future. The second process that must also take place at the corporate level of the IGO for institutional learning to be possible is the actual implementation of change. If this is to be an enduring change (as I understand genuine learning to require) rather than a temporary alteration in policy, for example, it must somehow affect the future conduct of the organization independently of shifts in attitudes, beliefs, or commitments on the part of its individual human members, and not be subject to reversal simply when there are changes to the membership of the group. I will address each in turn.
Institutional Self-Reflection, Deliberation, and Commitment to Change
An organization can deliberate in the sense of accessing and processing information—and generally has more sophisticated capacities to do both than any individual human actor.Footnote 19 Moreover, with a formal decision-making structure that can commit the group to a policy or course of action that is different from the individual positions of some (or all) of its members, it can qualify as a purposive agent in itself.Footnote 20 IGOs can meet this criterion. Of course, different IGOs have different decision-making models. The most obvious model for meeting this criterion is decision-making by majority vote, the model found in the UN's General Assembly and in the Security Council (when the veto is not invoked).Footnote 21 By contrast, IGOs with decision-making models that require unanimity (for example, the Arab League with respect to decision-making on procedural issues) cannot arrive at positions that can convincingly be described as the will of the IGO as a whole, as distinct from the aggregate preferences of its members, and are thereby unable to meet this criterion. IGOs that reach decisions by consensus (such as NATO) provide an interesting case. At first glance, they might also seem to only represent the aggregate wishes of the constitutive members of the IGO. Yet, the institutional culture that channels and molds the intentions and preferences of the IGO's member states, through the process of consensus building, means that decisions reached via this model are not straightforwardly reducible to the positions of their members.Footnote 22 As part of the institutionalized process of arriving at a consensus, states (and sometimes other IGOs) can arrive at positions that are different from those they would defend independently of this decision-making structure. In sum, it is eminently plausible for an IGO employing either majority voting or consensus-based decision models to reflect, at the corporate level, on the consequences of previous acts and omissions and to decide, at the corporate level, to promote internal change.
Structural Reform
The conscious change requisite to institutional learning—prompted by reflection on past experiences and an accompanying decision to alter future actions and outcomes—must involve revisions to (or reinforcement of) the information-gathering, decision-making, or executive mechanisms, procedures, and structures of the formal organization, or transformation (or bolstering) of the norms, rules, and practices that contribute to the organization's functioning at the corporate level. In this way, institutional learning is the imprint on the organization's very structure and culture of its self-reflection, deliberation over past experiences, and corresponding commitment to change. Such change then persists even if none of the members of the corporate entity at a certain point in time shared in the experience that prompted the reform. This is neither a passive reaction to external pressure, nor a process of imitation, nor considered change that is the result of learning that can be reduced to individual human cognitive processes. Rather, it is a purposive, systemic, inward-looking reconfiguration of the institutional agent in order to negotiate its external landscape and achieve its goals—which may include any number of often overlapping objectives, including the accrual of power and influence, consistency between self-image and reputation, and survival.
Indications of Institutional Learning? The UN after Rwanda and Srebrenica
I have suggested that charges of moral failure directed at institutional agents often appear to be accompanied by the assumption that they can learn. The underlying intuition seems to be that if the body itself were to realize that it had failed to discharge some responsibility, for example, and reflect on the consequences of its previous acts and omissions, this could trigger some internal process of change. Do IGOs learn in the context of perceived moral failure? Although a detailed illustrative example is beyond the scope of this short essay, it might be useful to highlight a case in which some degree of institutional learning has taken place in such circumstances. Internal change within the UN following genocides in Rwanda and Srebrenica arguably provides just such a case: the preliminary self-reflective and deliberative stage of institutional learning is readily apparent, and the subsequent stage of resulting structural reconfiguration has been gestured toward and, in some respects, enacted.
Before proceeding with this case, two qualifications are in order. First, I am not suggesting that institutional learning is only likely to occur in the wake of perceived moral failure. The institutional learning that I outlined above could occur in any context in which the corporate agent seeks to better realize its goals by reflecting on the past outcomes of its acts and omissions and engaging in internal structural reform as a result. If it is possible for IGOs to learn, then learning can occur in a number of different contexts—and in response to success as well as failure. The perception of moral failure simply provides one potentially powerful prompt for institutional learning. In such cases, external condemnation and censure may be a powerful contributing factor. Yet, it is the institution's goal of achieving consistency between its espoused values and self-image, on the one hand, and its actions and external reputation, on the other, that is likely to trigger internal self-reflection and reform. Second, nothing here relies on an objective measure of what constitutes moral failure—or, relatedly, on a definitive account of the source or substance of the moral responsibilities of IGOs. The focus, instead, is on one possible response to the perception of moral failure. Yet, importantly, when it comes to what we understand to be moral responsibilities in international politics, some principles represent near universal agreement. There is widespread consensus that the UN abrogated its moral responsibilities in failing to prevent or mitigate genocides in Rwanda and Srebrenica. Indeed, this judgment was expressed by the UN itself.
“Remarkable Self-Criticism”
In 1994, almost a million Tutsis and moderate Hutus were slaughtered in Rwanda over a period of just one hundred days. The killings were meticulously planned and orchestrated by those in power. Despite having forces on the ground, the UN failed to prevent or mitigate the genocide. The following year, during the Bosnian War, more than eight thousand Muslim Bosniaks, mainly men and boys, were massacred in and around the town of Srebrenica—a town that had been declared a “safe area” under UN protection. In the wake of both mass atrocities, in addition to blame being apportioned to those who had choreographed and carried out the killings, the UN was the object of condemnation for failing to act and avert great harm when it had the capacity to do so. While blame for some acts and omissions in the context of these failures might be best apportioned to the UN's member states, discrete parts of the IGO (such as the Security Council), or its individual human constituents, significant failures were compellingly attributed to, and acknowledged by, the organization as a whole. The UN's own assessment and acceptance of its failure is particularly significant in the context of the conception of institutional learning outlined above.
For institutional learning to have taken place on the part of the UN, the organization itself must have first engaged in self-reflection and deliberation over its failures in Rwanda and Srebrenica, resulting in a commitment to change. There is clear evidence of this first requisite stage of institutional learning. In 1999, the UN commissioned respective detailed analyses of its role in each genocide.Footnote 23 These reports provide clear assessments of the consequences of the organization's acts and omissions, powerful statements of institutional failure, and bold recommendations for reform.Footnote 24 The inquiry into the genocide in Rwanda judges that “the failure by the United Nations to prevent, and subsequently, to stop the genocide in Rwanda was a failure of the United Nations system as a whole.” It asserts that “acknowledgement of responsibility must also be accompanied by a will for change: a commitment to ensure that catastrophes such as the genocide in Rwanda never occur anywhere in the future.”Footnote 25 According to the separate report on Srebrenica:
The United Nations experience in Bosnia was one of the most difficult and painful in our history. It is with the deepest regret and remorse that we have reviewed our own actions and decisions in the face of the assault on Srebrenica. Through error, misjudgement and an inability to recognise the scope of the evil confronting us, we failed to do our part to help save the people of Srebrenica from the Serb campaign of mass murder.Footnote 26
These are important statements. However, meeting the first criterion for institutional learning requires not only an assessment of UN failure and responsibility but also a demonstration that self-reflection, acknowledgment of failure, and a commitment to change genuinely occurred at the level of the IGO. Significantly, each report was not only commissioned but also discussed and endorsed by one or both of the organization's two main decision-making bodies.Footnote 27 At the formal meeting of the Security Council during which the findings of the inquiry into the UN's role in Rwanda were deliberated over and extolled, and its recommendations supported, the representative of the United States aptly described the Rwanda and Srebrenica reports, together, as representing a “remarkable self-criticism.” He observed that the very fact “that these reports were sponsored and generated”—and I would add endorsed—“by the United Nations itself is testament to our collective commitment to work with the Secretary-General to reform the United Nations, to overcome and avoid the failures of the past and do better in the future.”Footnote 28
Incremental, Imperfect Structural Reform
Conscious, internal structural transformation is also requisite to institutional learning. Specific recommendations that have accompanied and followed critical self-reflection by the UN in the wake of Rwanda and Srebrenica gesture passionately toward such reform, give us a sense of what it could look like, and, in some instances, have resulted in imperfect incremental change that satisfies the second criterion for institutional learning. Although a comprehensive study of the reform proposed and partially realized by the UN is beyond the scope of this essay, I will touch on a couple of areas of internal change prompted, in part, by the UN's acknowledgments of failure just surveyed.
Analyses of the UN's role in Srebrenica and Rwanda highlighted weaknesses in its capacity to access and process information in conflict situations and effectively disseminate this information to different parts of the organization.Footnote 29 What was described as an “institutional weakness in the analytical capacity of the United Nations” represented a significant limitation.Footnote 30 That this capacity was circumscribed meant that the UN necessarily struggled to meet responsibilities that it had explicitly assumed, including the responsibility to respond to genocide. In the ten years following the Srebrenica and Rwanda reports, the UN implemented internal reforms to bolster existing “information, assessment, and early warning” functions.Footnote 31 These changes included, in 2004, the establishment of the Office of the Special Adviser on the Prevention of Genocide as a new mechanism for early warning and information coordination.Footnote 32 Although this reform has been slow, and remains incomplete, it is one example of structural change that has arisen from critical self-reflection and constitutes institutional learning.Footnote 33
Another variation on structural reform was prompted by UN self-assessments and deliberations over the genocides in Rwanda and Srebrenica that repeatedly attributed institutional failure, at least in part, to a lack of political will to act.Footnote 34 These judgments were accompanied by calls to reinforce expectations that the UN, and, concurrently, its member states, act in accordance with their responsibilities to prevent and mitigate mass atrocity.Footnote 35 In other words, recommendations were made for structural change in terms of strengthening existing norms. Bolstering the normative imperative to intervene to prevent and mitigate genocide could counter both the likelihood of decision-making paralysis in the Security Council and the problem of insufficient material support from member states in cases where the UN does decide to act. Incremental progress in achieving this normative shift was evident when all member states endorsed the “responsibility to protect” (RtoP) at the 2005 World Summit, thereby promoting the principle that “the international community, through the United Nations . . . has the responsibility . . . to help to protect populations from genocide, war crimes, ethnic cleansing and crimes against humanity.”Footnote 36 Such a groundbreaking consensus was achieved after RtoP was placed on the agenda following its adoption by the UN's High-Level Panel in 2004 and its endorsement by Secretary-General Kofi Annan in his plan for UN renewal.Footnote 37 This principle—whose justification has been explicitly linked back to the wish to avoid futures tragedies, such as the UN's failures in Rwanda and SrebrenicaFootnote 38—was then endorsed by the Security Council and the General Assembly.Footnote 39 Admittedly, there is reason to question the enduring force of this norm, particularly when observing recent inaction by the UN in response to mass atrocity crimes in Syria. Nevertheless, the normative change that was championed, endorsed, and—albeit imperfectly—implemented by the UN in response to self-acknowledged failures, including in Rwanda and Srebrenica, constitutes another clear (and still evolving) example of institutional learning.Footnote 40
Conclusion: Three Conceptual Claims for Further Study
Recognizing the possibility of institutional learning should guide how we think about the design of IGOs, the significance of mechanisms and opportunities for self-reflection, and what we can reasonably expect of such institutional agents in terms of internal reform. It should also guide how we respond to IGOs that have failed to discharge what we understand to be their responsibilities, specifically prompting us to consider the potentially rehabilitative, and not simply deterrent, retributive, or performative, functions of censure in cases of moral failure. Institutional learning requires further attention and analysis as an important, and neglected, example of peaceful change in international politics.
To conclude, I offer a few preliminary, and perhaps provocative, conceptual claims. These will serve to further distinguish my proposed account of institutional learning from other references to learning on the part of corporate entities in international politics, highlight some of its complexities, and raise questions that require study in future work on institutional learning.
Institutional Learning as an Ethically Neutral Process
I have begun to sketch an argument for how an institution that fails to discharge what we understand to be its moral responsibilities might make changes that would reduce the likelihood of such failure in the future—by, for example, strengthening the capacities that allow it to meet specific objectives, adopting new codes of conduct, and consciously transforming the culture of the organization. Institutional learning is a unique and powerful category of peaceful change in international politics. Yet, institutional learning itself does not connote moral progress.Footnote 41 As proposed above, institutional learning occurs when the IGO draws on its past experience of the consequences of its acts and omissions to pursue internal reform in order to better achieve its own goals and interests, however these are defined and however these map onto the evaluative criteria that we invoke to make moral judgments. If, for example, an organization is motivated to avoid external censure and the internal cognitive dissonance that arises when its self-image clashes with convincing portrayals of its behavior, then it could also learn to evade responsibility for genocide rather than improve its mechanisms for responding to it. This would nevertheless constitute institutional learning. If we want to encourage the institutional learning of IGOs to proceed in a direction that makes what we understand to be morally praiseworthy behavior more likely, then somehow the internal process of learning that I have focused on here would need to be accompanied by the separate “socialization” or “moral proselytizing” of Finnemore and Price.
Shameless Agents
In his description of Finnemore's constructivist account of “teaching” and “learning” in international politics, Price refers to shame as a central “pedagogical ingredient.” Although my account of institutional learning does not rely on external triggers (or “teachers”)—and, indeed, could constitute a wholly internal process—institutional learning in response to perceived moral failure may often be prompted by external censure. Moreover, in the UN report on the fall of Srebrenica, the UN experience in Bosnia is described as one of “regret and remorse.” Yet, I hesitate to invoke notions of shame or regret in the context of institutional learning, as this problematically imbues IGOs and other corporate agents with emotions, and thereby anthropomorphizes them in a way that I argue is unnecessary for recognizing them as purposive agents and unhelpful in practice. Emotions have a collective dimension, but only individual human beings (and not corporate agents) can experience emotions.Footnote 42
“Remainders of feeling” are emotional responses to unmet obligations—such as guilt, remorse, regret, and shame—that can be contrasted with “remainders of action,” or responses such as reparation and rehabilitation.Footnote 43 While individual human beings are capable of remainders of feeling, it is unrealistic to expect that institutions can be similarly moved. How, if at all, does this affect our understanding of institutional learning as a response to charges of moral failure? This is a question for careful consideration in later work. The outline of a future answer might be that we can coherently talk about unintended harm and accompanying censure informing the character and identity of an institutional moral agent that caused the harm without venturing into the contentious territory of ascribing emotions to the agent that are not reducible to its individual human constituents.Footnote 44 There will, after all, be consequences for the corporate agent in such cases, both in terms of reactions from other agents—condemnations and demands for redress—and in terms of the dissonance between how it presents and perceives itself and the effects of its actions. In short, IGOs and other institutional agents might experience the functional equivalent of remainders of feeling.
Individual, State, IGO: Mutually Exclusive Levels of Learning
Finally, to what extent can the institutional learning of IGOs be disaggregated from the learning of their constitutive agents? Institutional learning, to qualify as such, must not be distributive among the members of the IGO. The acquisition of new knowledge arising from self-reflection and deliberation over past experiences, as well as the decision-making and structural reform that supports enduring change in behavior, must take place at the intergovernmental corporate level. One perhaps counterintuitive implication of this argument is that it is conceivable that an IGO, such as the UN, could learn—in response to perceived moral failure, for example—without its member states or individual human constituents experiencing prior or parallel learning. It is also conceivable that the member states and individual human constituents might, gradually, be socialized by the evolving norms and practices embodied by the reformed organization. This, of course, would radically reverse the common IR narrative of organizations only “metaphorically” learning when institutional change is a by-product of the learning experienced by its individual human members, or aggregates of individual human members. Exploring the theoretical potential and practical implications of this observation will be an important exercise.
