Agency : Sharon Krause

Ofer Wolberger / The Blue Dress
Ofer Wolberger / The Blue Dress

Agency / Sharon Krause

The concept of agency is a fundamental one in political theory because agency is crucial to the coordinated activity that is a constitutive component of political life. Agency is especially central in theories of democratic politics because it is a precondition for collective self-rule, political contestation, and the pursuit of justice. As a political concept, agency has both descriptive and normative functions. Descriptively, it enables us to identify and distinguish among different sources of activity within complex political dynamics, to make sense of causal relationships, and hence to understand political outcomes. For example, agency is often described as an important causal factor in social movements, institutional reform, international diplomacy, leadership, and the policy process, among other things. Normatively, the concept of agency makes it possible to distinguish sources of activity that are properly subject to ethical and political accountability from those that are impervious to accountability. Hurricane Katrina created a great deal of misery in New Orleans, for instance, but no one attributes ethical or political accountability for this impact to the hurricane. We reserve accountability for sources of activity that are more than simply the causes of outcomes, sources that are also agents. Agents are sources of activity that can be responsive to notions of what ought to be – to principles of justice and conceptions of the good, to the claims of others, to their own dreams and aspirations.

The most common way to conceive agency in both liberal and democratic theory has been to identify it with intentional choice and control over action – in short, with a kind of personal sovereignty. Yet the sovereigntist view fails to reflect the realities of human action, and it tends to undermine the quality of political life that we create. Indeed, it contributes to some of the most pressing political problems of our time, including persistent racial and gender-based oppression, injustices in the global economy, and ecological destruction. At the same time, agency as a source of accountable activity is a necessary condition of politics. It enables us to respond to collective problems such as persistent oppression and ecological destruction, and to bring more emancipatory possibilities into being. Despite not being reducible to personal sovereignty, agency is often remarkably robust. Understanding the non-sovereign but robust character of agency is an important part of making progress toward a more just and sustainable political life.

Part one of this article diagnoses the difficulties inherent in conventional conceptions of agency and identifies the problems they pose for democratic politics; part two lays out an alternative, non-sovereign account of agency, showing how agency is both socially and materially distributed and hence not reducible to intentional choice and control over action; part three briefly examines the emancipatory implications of understanding agency in non-sovereign terms.

Agency and personal sovereignty

The common identification of agency with personal sovereignty locates agency in the exercise of will and treats it as an inner faculty of the individual. This view associates agency above all with intentional choice and control over action.1 No one believes that perfect control is ever possible, of course; all sorts of external and internal conditions are acknowledged to constrain agency. Yet, many of us do believe that the closer one comes to the ideal of spontaneous, self-generated action, the more agentic one will be. Common as it is, however, this way of conceiving agency fundamentally misunderstands agency’s distributed structure, the ways it is constituted through social and material processes that exceed not only the inner faculty of will but the very boundaries of the self.2 Agency requires the exercise of will but it does not end there, for agency involves actually doing things, making things happen, with real impact and effect. To be an agent is to be a source of activity, which means that efficacy—or impact on the world—is as much a part of agency as is individual initiative.

The dual character of agency as involving both initiative and efficacy means that agency cannot be contained within the individual. Our effects depend as much on how the world receives and responds to our initiatives as they do on our intentions. This is not to say that agency exists only in social practices as distinct from the intentions and initiatives of individuals. The point is that our deeds are a function of how our intentions and initiatives interact with the responses they generate. Agency is a socially and materially distributed phenomenon in this respect; it depends in a constitutive way on other people’s uptake and the things that help shape our impact on the world. Individual agency is a dynamic, interactive phenomenon with widely distributed sources that include but are not limited to the individual subject. It involves intersubjective and intercorporeal exchanges, and so cannot be reduced either to social practices alone or to faculties such as the will that are internal to a particular person.

While many people acknowledge that social and material factors condition agency, few accept the idea that such factors actually help constitute agency. The idea that human agency is a fundamentally non-sovereign experience, as Hannah Arendt put it, disrupts convictions at the heart of democratic theory.3 Yet, coming to terms with the non-sovereignty of agency is important for democratic politics in at least two ways. First, when we equate agency with intentional choice and control over action, we make it difficult to establish accountability for social dynamics to which individuals contribute without intending to do so and without controlling the outcomes. This applies to dynamics such as systemic oppression, implicit bias, and the frequently impersonal exploitation involved in globalized labor markets. It also applies to outcomes such as climate change and other forms of environmental degradation. In all these cases, outcomes exceed the control of any particular individual, and they often result from effects that are unintentional and even unconscious at the individual level. If we did not intend to discriminate or exploit or pollute, we think, and if we did not have personal control over these outcomes, then they could not be a function of our agency, and so they could not be our fault. Yet dynamics such as these persist only because many people contribute to sustaining them, however unintentionally, and they will not change unless we hold ourselves and one another accountable for acting differently. We cannot do this so long as we believe that our agency, which is the source of our accountability, lies solely in activity that we intended and controlled. Thus our sovereigntist way of thinking about agency tends to undercut the sense of personal accountability needed to generate effective political change in the face of systemic injustice and other harms.

A second consequence of this way of thinking about agency is to undermine our sense of possibility, our confidence in being able to make things happen, our faith that we can affect the conditions of our own coexistence. Because we do not appreciate the socially and materially distributed character of agency, we do not recognize the importance of nurturing the social and material formations that help constitute it. Even as the forces governing our lives have grown larger, more complex, and more impersonal, we have neglected the infrastructure of associational life and social movements that individual agency needs to flourish. The result is a widespread sense of disempowerment. Nowhere is this sense more vivid than in the context of the global crises posed by climate change and other ecological problems. However much we may wish to live more sustainably, many of us feel impotent in the face of forces that so clearly dwarf us and exceed our control. Wherever we turn, whatever we do or eat or buy, we find ourselves simultaneously participating in the despoiling of nature and suffering its effects—in superbugs and rising seas, or tapped-out water sources and air that is too polluted to breathe. We know that our individual efforts to resist these forces—through recycling, for example, alternative energy use, or animal advocacy—have negligible effects and sometimes generate additional, unintended damage. We feel weak and ineffectual because we are weak and ineffectual in the face of forces such as global capitalism and modern states. And we are weak and ineffectual as individuals because the social and material conditions needed to sustain our agency have been eroded concurrently with the immense growth in the scale and power of impersonal political and economic forces. Because we think of agency in terms of personal sovereignty, we have failed to protect the sources that nourish it, and the result is a lost sense of possibility.

This loss goes together with our impaired sense of accountability. They are the twin casualties of our sovereigntist assumptions about agency, and they represent the political stakes of our conceptual confusion. A healthy democracy depends on citizens who have both a robust sense of personal accountability and a confident sense of political possibility, but neither one can flourish without the right conditions. Creating the right conditions requires understanding the structure of human agency better than we do. We need to rethink the meaning of agency in more non-sovereign ways, both in the interest of conceptual clarity and because cultivating the conditions of a more enlivened, emancipatory democratic politics requires it.

Non-sovereign agency

Agency as a socially distributed phenomenon.

The notion of non-sovereign agency was first introduced by Hannah Arendt.4 As she puts it in The Human Condition, there are “two parts” to any action, “the beginning made by a single person and the achievement in which many join by ‘bearing’ and ‘finishing’ the enterprise, by seeing it through.”5 Action begins with individual initiative, an effort to start something new, through which the agent reveals her distinctive identity.6 Yet personal initiative is not the whole of action. At least in the public sphere, action always “needs the presence of others” to have an impact.7 Others contribute to the action of the individual as “co-actors” who help bring “the actual achievement” of her “enterprise” to fruition.8 In the same way that a leader’s initiatives depend for their impact on the responses of his followers, so the efforts of any individual rely on a community of “bearers” to sustain their efficacy. The responses that other people have to our initiatives deeply affect the impact we have on the world; they shape the unfolding narrative of enterprises and outcomes that constitutes the story of what we have done.9

Much of the time, agency’s dependence on social uptake is invisible to us because uptake is regularly forthcoming, at least for those whose identities and initiatives are recognizable within prevailing norms of discourse. In our interactions with one another we regularly function as the bearers for one another’s agency in Arendt’s sense, meaning that we respond to one another’s initiatives in ways that help to establish their impact on the world. When I step up to the lectern on the first day of class, my students recognize this initiative as an effort to begin the course, and they respond accordingly. In the absence of their enabling response, my effort to begin the course could not come to fruition and my agency would be frustrated. Likewise, when I go to the counter to swipe my credit card at the grocery store, the clerk knows exactly what I am trying to accomplish and responds in a way that facilitates my action. Societies are set up to provide just this kind of uptake, at least for those who are privileged. In view of agency’s dependence on social uptake, Arendt described action as “non-sovereign,” meaning that it could not be reduced to intentional choice or control, or contained within the boundaries of the individual agent. Agency is socially distributed; it arises through the interaction of personal initiatives with social interpretations and responses.

Because of agency’s non-sovereignty, our deeds do not always track the intentions that motivated our initiatives. Our effects can outrun or counteract our conscious intentions, without thereby eradicating our agency.10 Consequently, the non-sovereignty of agency tends to extend the breadth of our accountability. Consider Oedipus, for example, whose tragic misery consists in the fact that he holds himself accountable for deeds that were, in an important sense, unintended. He did not mean to have sex with his mother and murder his father, and yet the reason for his misery is that he understands these actions as his deeds, however unintended they may have been. He recognizes that his agency extends beyond the activities that he intends and controls. In a similar way, we commonly acknowledge that people can powerfully perpetuate norms of racism and sexism without meaning to do so: a woman who unconsciously grasps her handbag more firmly as a young black man approaches her on the sidewalk; the male colleague who hears the good idea in a group discussion as having come from another man instead of from the woman who actually voiced it. People are regularly the agents of racism and sexism without meaning to discriminate, given the ways that their individual initiatives interact with background relations of power. We can identify with Oedipus and unwitting racists and sexists because we see, as Bernard Williams put it,  “that in the story of one’s life there is an authority exercised by what one has done, and not merely by what one has intentionally done.”11 The non-sovereignty of human agency extends the bounds of personal accountability beyond the acts that we intend and control. In this sense, agency is more potent than we often realize, or potent in ways that we do not always wish to acknowledge.12 Accepting the enlarged scope of accountability that non-sovereign agency entails is a necessary condition of emancipatory political change.13

At the same time, agency’s non-sovereignty also makes it more vulnerable than we often assume, especially in relation to social inequality. If agency’s socially distributed character is often invisible to those who are privileged, this aspect of agency is keenly felt by the marginalized. Indeed, agency’s dependence on social uptake may be clearest in cases where uptake is missing or distorted by bias, which makes these cases especially informative. Systematic inequality can disrupt the social interpretation of individual initiatives on the part of marginalized people in ways that track injustice, thereby undermining the community of bearers required to sustain effective action and generating widespread failures of agency.14 The novel Invisible Man by Ralph Ellison depicts this dynamic, showing how racial inequality makes the distinctive identities of black individuals invisible to the wider society, which causes others to misunderstand their initiatives and respond to them in ways that impede the effective exercise of their agency.15 The unnamed protagonist of the novel cannot disclose his identity in his deeds, in Arendt’s sense, because racism makes it impossible for others to really see him. As a result, the social response to his initiatives regularly fails to affirm his own understanding of them, and his impact on the world is often at odds with who he is. The novel describes a long series of failed and frustrated exertions of agency.

In a similar way, James Baldwin’s Notes of a Native Son illustrates agency’s deep vulnerability to social inequality. “To be a negro,” Baldwin says, “meant that one was never looked at but was simply at the mercy of the reflexes [that] the color of one’s skin caused in other people.”16 The result of having his initiatives interpreted through the lens of racial stigma and stereotypes, rather than in light of his actual identity, was that Baldwin consistently found himself  “doing” things that he never intended and could not identify with: “I simply did not know what was happening. I did not know what I had done.”17 Racism frustrated his agency by making it impossible for his initiatives to come to fruition in his deeds, and by attributing deeds to him that were at odds with the actions he initiated. Likewise, women whose demeanors fail to conform to conventional norms of feminine deference frequently find themselves generating effects, such as hostility and resistance, that at are odds with their initiatives and that impede their efficacy, generating similar failures of agency. Experiences like these are well known to the marginalized and the subordinate. In view of agency’s non-sovereignty, subtle dynamics of social uptake and non-uptake affect agency in ways that are extremely powerful, however invisible and difficult to articulate they may be.

It is important to see that the failures of agency depicted by Baldwin and Ellison are not simply a function of their intentions being out of step with their effects. As the examples of Oedipus and the unwitting racist demonstrate, our effects can exceed our intentions without necessarily undoing our agency. The failures of agency that we see in Baldwin and Ellison involve a background of systematic and unjust social inequality that makes it impossible for others to see them in their initiatives and to respond to the actions they begin in ways that could help bring them to fruition. The responses of others distort or impede the impact of their efforts, and consequently the deeds that they are understood to have performed are not ones that affirm their initiatives and disclose their identities. Even on reflection, they cannot find themselves in their effects because these effects are shaped by a background of racial stigma that makes their actual identities invisible and misconceives the nature of their initiatives. Consequently, against a background of social inequality, agency is often deeply troubled.18 This explains why the sovereigntist injunction to pull yourself up by your bootstraps is a losing proposition. It locates agency in the individual will, ignoring agency’s socially distributed character. It fails to acknowledge that empowering individuals means changing the background conditions that undo their agency so as to create new communities of bearers who can help bring it to fruition.

Thus even as our impact on the world can sometimes exceed our intentions, thereby extending our efficacy in ways we may not expect, so our impact also can fall short when our actions fail to manifest our initiatives and identities, thus disabling agency. These are among the dynamics that the Black Lives Matter movement is currently trying to illuminate. The backgrounds of racialized social meaning against which the actions of young black men are interpreted in police encounters have a powerful impact on their efficacy, on how they actually affect the world. These backgrounds and the forms of social uptake they foster interact with individual initiative in ways that too often turn deadly. As a college student, I did door-to-door fundraising for a Naderesque non-profit. One night when we were working in an affluent, white suburb of Boston, one of the other fundraisers (a young black man I’ll call David) was late for his pick-up. When he finally arrived, sauntering casually up to the car, the field manager said to him, “David, if you’re going to be late and keep us all waiting, you could at least hustle a little and run to the car.” David replied, “If I run in this neighborhood people will think I just stole their TV, and you’ll be picking me up at the police station.” David knew that for him to run in that context was not simply to ambulate quickly but also to arouse suspicion and invite incarceration. Arousing suspicion and inviting incarceration were not separate from the act of running, in his mind, but part of it. He understood, however inchoately, that his agency was a function of the interaction of his intentional initiatives with the social uptake they generated. In this case, that uptake was not only at odds with his intentions but blind to who he really was as an individual because it was based on racial stigma and bias. Black Lives Matter shines light on these dynamics. The resistance that the movement has generated is at least in part a function of how much we misunderstand the nature of human agency, how deep our sovereigntist assumptions run. To insist, with opponents of the movement, that “all lives matter” is to obfuscate the fact that the interaction of individual initiative and social uptake has differential effects on agency in contexts of entrenched racial stigma and bias. If agency were sovereign, we would not need to worry about what actually happens to it in the presence of entrenched inequality. Because we assume it is sovereign, we resist the transformative work that Black Lives Matter is asking us to do. A non-sovereign approach helps us see the ways that agency can be both more potent and more vulnerable than we often assume, and it presses us to think about the broader context of intersubjective dynamics that help constitute—or disable—agency.

Agency as a materially distributed phenomenon.

The non-sovereignty of agency also refers to the ways in which agency is materially or corporeally distributed, in the sense that our efficacy results in part from how our initiatives interact with the activities of other material entities.19 Most of the time, what we conventionally think of as individual agency actually reflects what Jane Bennett calls an “assemblage” of material “actants” that includes far more than the individual alone.20 If one looks closely enough, Bennett says, “the productive power behind effects is always a collectivity.”21 And the collectivities that constitute agency encompass more than just human beings. In one sense, the materially distributed character of agency is obvious. When I speak, the formation of sound that I initiate with my vocal chords, my breath, and my lips can have an impact only because of the vibration of molecules in the media (liquids, solids, gasses) through which the sound waves I generate move. Without the movement of molecules through media, my agency in this instance could not come to fruition. Likewise, when I sit down to eat breakfast, the chair at the kitchen table holds me up while gravity holds me down. The verb “holds” expresses something real here, which is that there are physical forces doing things in and around us all the time, things that interact with our initiatives in ways that can either obstruct them or support them. Matter is not inert but full of activity, and its activities interact with our initiatives to constitute our agency in ways that we do not fully understand or control.

Ecologist Aldo Leopold’s “land ethic” is informed by an understanding of this interaction. As he puts it, “many historical events, hitherto explained solely in terms of human enterprise, were actually biotic interactions between people and the land.”22 He cites as an example the settling of the American South and Midwest. Whereas historians debate about the relative influence of American pioneers, French and English traders, and Native Americans, Leopold draws our attention to the contributions of the soil. Specifically, he wants us

to ponder the fact that the cane-lands, when subjected to the particular mixture of forces represented by the cow, plow, fire, and axe of the pioneer, became bluegrass. What if the plant succession inherent in this dark and bloody ground had, under the impact of these forces, given us some worthless sedge, shrub, or weed? Would Boone and Kenton have held out? Would there have been any overflow into Ohio, Indiana, Illinois, and Missouri? Any Louisiana Purchase? Any transcontinental union of new states? Any Civil War?23

In short, he says, “the plant succession steered the course of history” as much as the initiatives of human beings.24 Or more precisely, “what the human actors in this drama tried to do” depended for its success “in large degree on the reaction of particular soils to the impact of the particular forces exerted by their occupancy.”25 What they did—and hence their agency—was a product of this interaction. Their agency exceeded their wills and the boundaries of their persons, even the boundaries of their species. When Leopold speaks of “land as a community,” he is pushing us to acknowledge that our agency is an assemblage and that the Earth provides some of its most important bearers.26

Similarly, Steven Vogel emphasizes that our ability to affect the world through action depends on material things and forces that always retain an element of “wildness” insofar as they exceed our understanding and control. One cannot “hammer without the force of gravity,” for example, or bake bread “without relying on complex biochemical processes taking place within millions of yeast cells.”27 These material forces and processes are constitutive components of our agency. While it is true that the material world also “comes to be what it is through our actions … at the same time, our actions are absolutely of the world,” meaning that our agency is dependent on “the operation of forces that [we] are … unable to master, to predict, or even fully to understand.”28 Individual agency is not located exclusively in the individual, and not reducible to intentional choice or control; it is a materially distributed phenomenon.

We tend to be oblivious to matter’s role in constituting agency in part because we are so focused on the aspects of agency that do happen inside us—the willing, thinking, and desiring. In part too, as adults we have had extensive experience in learning how to act in ways that are upheld rather than undercut by the material world, and we have adapted our initiatives to the matter we regularly encounter. Moreover, the material fields through which agency emerges tend to be relatively stable and predictable. Once adapted to them, we interact with them in ways that support our agency with a fair amount of consistency. Likewise, we construct our built environment in ways that are specifically tailored to sustain our agency. This adaptation and tailoring make the role of material forces in constituting agency recede from view and allow us to feel far more sovereign than we actually are.

To be sure, the habitual feeling of sovereignty is limited to those whose bodies conform to conventional standards of “normal.” As disability theorists point out, for those whose bodies depart from these standards, agency’s constitutive dependence on the vitalities of the material world never recedes from view. They regularly experience the differential effects that the material environment has on the agency of differently abled individuals. The stairs that invisibly (to me) support my agency as I effortlessly make my way to the front door of the campus library actually impede the agency of those who need wheelchairs to get around. The disability literature thus shines light on the often dark recesses of agency’s materially distributed character.29 This literature also brings out the ways in which failures of social uptake and failures of material uptake can interact in thwarting the agency of those who are marginalized. Our built environment is constructed in ways that regularly undermine the agency of disabled people because historically there has been little or no social uptake for their agency. Others have seen them through a lens of stigma and stereotype that denies their individual identities and shuts down their initiatives. The fact that the built environment makes it difficult or impossible for people with certain disabilities to get around in public spaces then reinforces their invisibility and deepens the absence of social uptake. This vicious circle itself recedes from view when we think about agency as an exclusively inner faculty and fail to acknowledge its distributed character.

So agency is an intercorporeal phenomenon characterized by a constitutive dependence on the material world. It has what David Abram calls a “porosity” that makes each agent like “an open circuit that completes itself only in things, in others, in the surrounding earth.”30 The open circuitry of agency not only undermines the ideal of agency as personal sovereignty but also disrupts the strict divide between human beings and the rest of nature.31 If agency arises through circuits of vitality that include and depend on non-human aspects of the material environment, then human agents can hardly claim to stand apart from and above this environment. Human beings do have some distinctive capacities not shared by most other material entities, and this distinctiveness is significant insofar as it makes us subject to ethical and political accountability in special ways, as we are about to see. But to acknowledge this distinctiveness is a far cry from insisting on a binary and hierarchical divide between human beings and everything else on earth. This imagined divide is another consequence of our sovereigntist thinking about agency. The truth is that we depend on the earth as we depend on one another, and we depend on it to provide more than just a field of opportunity for the exercise of our agency. We depend on it as a vital, contributing source in the constitution of agency, the very thing that supposedly sets us apart.

Agency, accountability, and the sense of possibility

The open circuitry of agency is sometimes thought to imply that everything in the world is agentic.32 Although Bennett acknowledges agency-relevant differences between various “ontological forms,” she does attribute agency widely. As she sees it, “there are various sources or sites of agency, including the intentionality of a human animal, the temperament of a brain’s chemistry, the momentum of a social movement, the mood of an architectural form, the propensity of a family, the style of a corporation, the drive of a sound-field, and the decisions of molecules at far-from-equilibrium states.”33Attributing agency exclusively to human beings, as sovereigntist views typically do, understates “the ontological diversity of actants.”34 In using the language of “actants,” Bennett means to convey the sense of “agentic capacities” as “the ability to make a difference, to produce effects, or even to initiate action,”35 but to detach these capacities from “figurations of agency centered around the rational, intentional human subject.”36 Actants may therefore include “electrons, trees, wind, electromagnetic fields,” and other putatively inanimate objects.37

Dismantling the hierarchical divide between human beings and non-human nature is one of the most important ethical and political challenges of our time, and a non-sovereign view of agency undoubtedly points in this direction. Still, there are good reasons to distinguish agency from material vitality or activity per se. The vitality of matter frequently manifests in certain kinds of responsiveness, as when Leopold’s soil responds to the plow, or your eardrum responds to the sound waves generated by my voice. Earth’s ecosystems are intricate webs of responsiveness in this sense. Yet, if the responsiveness of the material world helps constitute agency, responsiveness alone is not sufficient for something to count as agentic. Agents are able to be responsive in a special way, namely they can be responsive to norms—to notions of what ought to be, to standards of right and conceptions of the good, to hopes and aspirations. As we saw at the outset, the concept of agency has a normative function as well as a descriptive one. Its normative function is to distinguish sources of activity that are properly subject to ethical and political accountability from those that cannot be held accountable. We hold agents accountable for what they do, but not hurricanes or soil or sound waves, because agents have the ability to be norm-responsive. The normative dimension of political agency is crucial to politics because it makes possible a collective life that answers to standards of justice, and it allows for principled contestation, civil disobedience, and common decency. A source of activity that can respond to normative injunctions is different from material vitalities that have a causal impact without being responsive in this way. Norm-responsiveness is a threshold that makes the difference between an agent and a mere cause, or between agents and other kinds of actants. In acknowledging the importance to agency of norm-responsiveness, however, we need not limit agency to human beings. Research on non-human animals is increasingly finding this faculty in varying degrees in a range of beings beyond the human. Agency as a source of accountable activity exists on a continuum, and many beings exercise agency in some measure. Yet not all material entities that manifest vitality are also capable of agency.

Once we understand the ways that agency is non-sovereign, acknowledging its distinctiveness no longer carries the baggage it once did in terms of attributing mastery and moral superiority to human beings. A non-sovereign account of agency does not set the human apart from and above the rest of nature. It will not sustain the illusion of human control over nature, and it does not invite human beings to instrumentalize nature for their own purposes. On the contrary, a non-sovereign approach gives us reason for humility, gratitude, and respect toward nature. Indeed, it achieves precisely the reorientation to nature that Bennett seeks, but without ascribing agency to everything under the sun. Ironically, the effort to elevate the moral status of nature relative to human beings by ascribing agency to it recapitulates the old, false assumption that agency is the sole basis of moral standing. We should abandon that assumption and acknowledge instead that there are many ways to count, both morally and politically.

To insist that agency entails norm-responsiveness may seem to reproduce core elements of the sovereigntist approach. Yet the norm-responsive self of non-sovereign agency is not equivalent to rational autonomy traditionally conceived. Norm-responsiveness may be unavailable to beings and objects that lack complex cognitive functions, but it would be wrong to equate it with a form of cognition that transcends affective, bodily experience. As much recent work has shown, the sense of right and wrong involves a mix of cognitive and affective modes of consciousness.38 We feel as much as reason our way to what we ought to do. Thus the notion of a norm-responsive self need not imply an ideal of personal sovereignty grounded in transcendent reason. Still, the capacity to be responsive to normative claims and aspirations is at the heart of the important distinction between an agent and a mere cause. This capacity underlies both the sense of accountability and the sense of future possibility that agency uniquely entails.

The accountability associated with non-sovereign agency admittedly differs from conventional conceptions of responsibility, which typically insist on intentionality and control. If you did not mean to do it, or did not have control over what happened, we do not normally hold you responsible—or we hold you responsible only in a mitigated, secondary way.39 I do not mean to suggest that intentionality and control are irrelevant to accountability but only to insist that they do not exhaust its conditions. Intentional harms may be especially noxious, and we are right to hold people accountable for them in special ways. But we must also acknowledge our accountability for harms we participate in even without controlling or intending them. Although it would be a mistake to blame people for their contributions to dynamics such as implicit bias in the same way that we blame them for intentionally discriminating, we can and should hold them accountable for acquiescing to such dynamics in the sense that we ask them to acknowledge the wrong and to do things differently in the future. Accountability in this form is a necessary condition of emancipatory political change, and it is one of the most valuable things that agency offers to democratic politics.

In thinking about accountability in non-sovereign terms, it is important to approach it in a holistic way. We should look broadly for networks of agentic contributors to outcomes rather than focusing narrowly on specific individuals, who often represent just one link in a much longer causal chain. We must also be mindful of the different subject positions we inhabit and the attenuations of agency they may imply. To ask the poor and the disenfranchised to be accountable for outcomes in the same ways as the rich and powerful is not only unfair but blind to the realities of their often compromised agency. It is important for individuals, however they are positioned, to be accountable for their contributions to harm, but we must recognize that accountability, like agency, is a distributed phenomenon.

When conceived in a suitably non-sovereign way, accountability can be a source of empowerment for agents themselves. It is the flipside of the sense of possibility that opens the door to political transformation, agency’s other gift to democracy. Acknowledging our own participation in dynamics that often feel too large and impersonal for us to affect helps us see them for what they are: the products of all-too-human practices to which we contribute, practices that depend on us and that together we can help change. The sense of accountability and the sense of possibility are reciprocally reinforcing. They combine to enliven our confidence in ourselves and our ability to change the world. At the same time, however, a non-sovereign understanding of agency should make us realistic about the necessary conditions for effecting change. The impersonal structures of political economy and governance that regulate so many aspects of our lives cannot be contested effectively without rejuvenating the social and material bearers that help bring our individual agency to fruition.

Rejuvenating the conditions of individual agency is work that will require collective effort. This is not to say that the non-sovereignty of agency renders all agency a collective enterprise. We can distinguish collective agency, as action in concert where individuals come together with a shared purpose and coordinated initiatives, from the basic character of agency as a distributed phenomenon. Insofar as individual agency is always an intersubjective and intercorporeal assemblage, it depends for its effects on social and material uptake, on the responses of other people and things. Yet this dependence does not imply that every exercise of agency, to be effective, must have a shared purpose or that it must involve consciously coordinated collective efforts. Shoring up the social and material conditions of non-sovereign agency will certainly require explicitly collective action, but as these conditions are rejuvenated, they will open the door to all kinds of agentic possibilities for individuals as well.

The catch is that to achieve our agency and the political possibilities it opens up, we must begin to think of agency in non-sovereign terms. So long as we continue to equate agency narrowly with intentional choice and control over action, we are likely to neglect the actual sources of our agency. This neglect can only perpetuate the lack of accountability and the sense of entrapment that plague so many of us today in confronting challenges such as climate change, or the distant injustices that often attend globalization, or the subtle and frequently invisible dynamics that sustain racial and gender-based oppression.


Agency is not what we thought it was—or what many of us thought it was. It is not a strictly internal property of the person, and not reducible to intentional choice or control over action. In the exercise of agency we are dependent on how our initiatives interact with dynamics in the material and social worlds that we can never fully master. In addition to eluding our control, agency also regularly outruns our intentions because it is a distributed phenomenon that arises through intersubjective, intercorporeal exchanges. Agency is more than mere causality because it involves activity that is responsive to norms. It does place us in special relations of accountability that are not open to non-agentic materialities, and it gives rise to a sense of possibility that has uniquely liberatory potential for politics. Although agents are distinctive in these ways, however, this distinctiveness is not the same thing as superiority. Agents are not the measure of all things; we are particular kinds of things with our own kind of value and dignity, which is not the only kind of value and dignity. Moreover, even as agency is non-sovereign, it is nevertheless robust and full of life. Its sources are found in the initiatives of individuals, the affirming communities that sustain social uptake, and the vitalities of the material world.

Cultivating the conditions of agency is a primary function of democratic politics, society, and culture. More work is needed to develop the contours of non-sovereign agency on all these dimensions, and to guide us to the kinds of assemblages that support a just and sustainable political life. Yet one of the remarkable things about agency is that it gives us tools for creating the conditions of its own flourishing. In this respect, agency is a source of emancipatory possibility, but it is also a perennial project that demands the continuing initiative and responsiveness of us all. Our old ideas about agency are deeply implicated in the enduring injustice and ecological destruction that everywhere haunt us today, often with deadly results. We can do better. The dream of mastery is an illusion, but the special responsibility we bear for emancipation is real. Emancipation—of human and other creatures, of the world itself—is possible but it requires new ways of conceiving agency and new ways of being human. The emancipation that non-sovereign agency promises is a radical project because of all that it asks us to change. It is a political project because the only way to proceed effectively is together.


Sharon R. Krause is Royce Family Professor and Professor of Political Science at Brown University. She is the author of Freedom Beyond Sovereignty: Reconstructing Liberal Individualism (Chicago, 2015); Civil Passions: Moral Sentiment and Democratic Deliberation (Princeton, 2008); and Liberalism with Honor (Harvard, 2002). She has written widely on dilemmas of democratic politics. She is currently writing a book about ecology and emancipation.

1. Agency so conceived is the foundation of the two moral powers in Rawls. See John Rawls, A Theory of Justice (Cambridge: Harvard University Press, 1971). Philip Pettit likewise understands agency as the capacity for control over action (whether rational, volitional, or discursive). See Pettit, A Theory of Freedom: From the Psychology to the Politics of Agency (Oxford: Oxford University Press, 2001). Nancy Hirschmann defines agency in terms of intentional choice, although she emphasizes the importance for freedom of ensuring that informal social conditions protect women’s choices against domination. See Hirschmann, The Subject of Liberty: Toward a Feminist Theory of Freedom (Princeton: Princeton University Press, 2003). Sovereigntist assumptions about agency can also be found in the communitarianism of Charles Taylor and feminist work on relational autonomy. See Taylor, “What is Human Agency?” in Taylor, Philosophical Papers: Volume 1, Human Agency and Language (Cambridge: Cambridge University Press, 1985); Catriona MacKenzie and Natalie Stoljar, Relational Autonomy: Feminist Perspectives on Autonomy, Agency, and the Social Self (New York: Oxford University Press, 2000); and Marilyn Friedman, Autonomy, Gender, Politics (New York: Oxford University Press, 2003). In addition, poststructuralist views that associate agency with transgression and resistance sometimes recapitulate the sovereigntist identification of agency with intentional choice. See Mahmood’s critique of Judith Butler on this point in Saba Mahmood, Politics of Piety: The Islamic Revival and the Feminist Subject (Princeton: Princeton University Press, 2004).

2. I focus here on individual agency, not collective agency, although individual agency is conceived as an assemblage, which is a kind of collectivity. The difference between individual and collective agency, and the ways that individual agency is both distinct from and dependent on collective agency, are discussed below.

3. Hannah Arendt, The Human Condition (Chicago: University of Chicago Press, 1999), 190-2.

4. For discussion of Arendt and further development of non-sovereign approaches to agency, see Linda Zerilli, Feminism and the Abyss of Freedom (Chicago: University of Chicago Press, 2005); Patchen Markell, Bound by Recognition (Princeton: Princeton University Press, 2003), esp Chapter Three and “The Insufficiency of Non-Domination,” Political Theory 36, no. 1 (February 2008): 9-36; and Sharon R. Krause, Freedom Beyond Sovereignty: Reconstructing Liberal Individualism (Chicago: University of Chicago Press, 2015). This paragraph and the five that follow draw from Krause, Freedom Beyond Sovereignty, which contains an account of agency’s intersubjective character that is more fully elaborated than the relatively brief discussion included here. My own account of agency draws inspiration from Arendt but also departs from her view in key respects. Among other things, it insists on the importance of an enduring (as opposed to episodic) identity within agency, and it takes seriously the disabling effects that social inequality can have on agency, something Arendt seriously understated. An enduring identity is not the same thing as a fixed or essentialized identity. For discussion of the differences, see Krause, Freedom Beyond Sovereignty, 71-73. For analysis and critique of Arendt’s view, see Freedom Beyond Sovereignty, 29-42.

5. Arendt, The Human Condition, 189.

6. Arendt, The Human Condition, 177.

7. Arendt, The Human Condition, 188.

8. Arendt, The Human Condition, 190.

9. Action is not always subject to social uptake, of course. I can flip on a light switch without needing social uptake to bring the action to fruition. Yet a great deal of action in the social and political domains does depend on uptake.

10. Arendt, The Human Condition, 173, 190-91, 197, 234; and see Markell, Bound by Recognition, esp. Chapter Three; and Bernard Williams, Moral Luck: Philosophical Papers: 1973-1980 (Cambridge: Cambridge University Press, 1981), esp. Chapter Two, and Shame and Necessity (Berkeley: University of California Press, 1993), esp. Chapter Three.

11. Williams, Shame and Necessity, 69.

12. Markell uses the language of “acknowledgement” to express recognition of the ways that our non-sovereign agency makes us complicit in injustice. Markell, Bound by Recognition, 177-89. It is worth emphasizing, however, that more than mere acknowledgement is required to undo the relations of injustice that Markell identifies. We need to actively take responsibility for the generation of a more just future through concrete action in the world. Doing so requires understanding and nurturing the accountability dimensions of agency.

13. The notion that we can be held accountable for things that we did not control or intend is admittedly unnerving. The ideal of agency as sovereignty holds that no one should be responsible for what is outside their control, and there are real advantages in this way of thinking. It is an achievement of the sovereign view, for example, that in liberal democratic societies we no longer hold people responsible for deeds committed by their family members, or punish women for being raped. Intentionality and control are not irrelevant to responsibility but neither are they the whole of it. Particularly in the context of complex social dynamics such as implicit bias or climate change, when we limit responsibility to the things we intended and controlled, we make it too easy to ignore our complicity in harms that depend on our participation even as they elude our grasp and sometimes our awareness. We need to be able to distinguish between different degrees and types of responsibility, some involving intentionality and control, and some exceeding these conditions. A fully elaborated theory of non-sovereign responsibility is beyond the scope of this article, but I touch on that theme briefly below and I have pursued it elsewhere (see the discussion of “non-sovereign responsibilities” in Krause, Freedom Beyond Sovereignty, 82-97).

14. Political and economic inequalities also pose problems for agency in related ways. I focus here on social inequality, meaning inequalities of power that attach to social identities (such as race, gender, or sexual orientation). These inequalities interact with political and economic inequalities, of course, so the contexts of social inequality are often also contexts of economic and political inequality.

15. Ralph Ellison, Invisible Man (New York: Random House, [1952] 1995).

16. James Baldwin, “Notes of a Native Son,” in Collected Essays, ed. Toni Morrison (New York: Literary Classics of the United States, Inc. 1998), 68.

17. Baldwin, “Notes of a Native Son,” 69.

18. The disabling effects of inequality can be countered by relations of solidarity and mutual recognition that provide alternative communities of bearers within otherwise oppressive social contexts. This is one function of the subaltern counterpublics documented by Nancy Fraser, which are important sources of agency for those who are marginalized. See Fraser, “Rethinking the Public Sphere: A Contribution to the Critique of Actually Existing Democracy,” in Habermas and the Public Sphere, ed. Craig Calhoun (Cambridge: M.I.T. Press, 1993), 109-42; Michael Warner, Publics and Counterpublics (New York: Zone, 2002); Jeffrey Stout, Blessed Are the Organized: Grassroots Democracy in America (Princeton: Princeton University Press, 2010); and Krause, Freedom Beyond Sovereignty, Chapter Three.

19. This is another place where my view departs from that of Arendt. She associates corporeality with necessity, the antithesis of action’s spontaneity and the enemy of freedom on her account.

20. Jane Bennett, “The Agency of Assemblages and the North American Blackout.” Public Culture 17, no. 3 (Fall 2005): 446-7.

21. Bennett, “The Agency of Assemblages,” 463. See also Connolly, who insists that there are many types of agency in “forcefields.” William E. Connolly, A World of Becoming (Durham: Duke University Press, 2011), 7, 21.

22. Aldo Leopold, A Sand County Almanac (New York: Ballantine, 1966), 241.

23. Leopold, Sand County, 241.

24. Leopold, Sand County, 243.

25. Leopold, Sand County, 242.

26.Leopold, Sand County, 243.

27. Steven Vogel, Thinking Like a Mall: Environmental Philosophy After the End of Nature (Cambridge: MIT Press, 2015), 113.

28. Vogel, Thinking Like a Mall, 120.

29. See, for example, Nancy J. Hirschmann, “Feminist Thoughts on Freedom and Disability,” Politics & Gender 8(2) (2012): 216-22; “Disability as a New Frontier for Feminist Intersectionality Research,” Politics & Gender 8(3) (2012): 396-405; and “Disability, Feminism, and Intersectionality: A Critical Approach,” Journal of Radical Philosophy, 16 (1) (2013); Nancy J. Hirschmann and Beth Linker, eds., Civil Disabilities: Citizenship, Membership, and Belonging (Philadelphia, University of Pennsylvania Press, 2015); Barbara Arneil, “Disability, Self Image, and Modern Political Theory,” Political Theory 37(2) (April 2009): 218-42; Tobin Siebers, Disability Theory (Ann Arbor: University of Michigan Press, 2008); and Tom Shakespeare, ed., The Disability Reader: Social Science Perspectives (London: Cassell, 1998).

30. David Abram, Becoming Animal: An Earthly Cosmology (New York: Vintage, 2011), 254. Abram is referring to the self in this passage, or personal identity, not to agency per se. I mean to extend his concept of the distributed circuits of the self to agency.

31. In this respect, the non-sovereign approach to agency resonates with recent work in post-humanist studies, which also contests this divide. See Rosi Braidotti, The Posthuman (Cambridge: Polity, 2013), 49, 60, 67; Abram, Becoming Animal, 47; Giorgio Agamben, The Open: Man and Animal (Stanford: Stanford University Press, 2004), 12, 16, 26; Jacques Derrida, The Animal that Therefore I Am (New York: Fordham University Press, 2008), 23, 32, 126; and Brian Massumi, What Animals Teach Us About Politics (Durham: Duke University Press, 2014), 3, 52, 92-3. Jedidiah Purdy also insists on the “permeability” of the boundary between human and non-human beings. Purdy, After Nature: A Politics for the Anthropocene (Cambridge: Harvard University Press, 2015), 282.

32. Bruno Latour’s Actor Network Theory takes this view, for example. See Latour, Reassembling the Social: An Introduction to Actor-Network-Theory (New York: Oxford University Press, 2007). Bennett’s work points in this direction as well. See Bennett, “The Agency of Assemblages” and Vibrant Matter: Toward a Political Ecology of Things (Durham: Duke University Press, 2010). See also John Law and Annemarie Mol, “The Actor-Enacted: Cumbrian Sheep in 2001,” in Carl Knappett and Lambros Malafouris, eds. Material Agency: Towards a Non-Anthropocentric Approach (New York: Springer, 2010), 58; Owain Jones and Paul Cloke, “Non-Human Agencies: Trees in Place and Time,” in Material Agency, 80-82; and Alfred Gell, Art and Agency: An Anthropological Theory (Oxford: Oxford University Press, 1998), 6-7.

33. Bennett, “The Agency of Assemblages,” 447.

34. Bennett, “The Agency of Assemblages,” 453. Connolly likewise insists that we should “appreciate multiple degrees and sites of agency, flowing from simple natural processes, through higher processes, to human beings and collective social assemblages.” Connolly, A World of Becoming, 22.

35. Bennett, “The Agency of Assemblages,” 446.

36. Bennett, “The Agency of Assemblages,” 453.

37. Bennett, “The Agency of Assemblages,” 446.

38. See, for example, Antonio Damasio, Descartes’ Error: Emotion, Reason, and the Human Brain (New York: Harper Collins, 1994); George E. Marcus, The Sentimental Citizen: Emotion in Democratic Politics (University Park: Pennsylvania State University Press, 2002); William E. Connolly, Neuropolitics: Thinking, Culture, Speed (Minneapolis: University of Minnesota Press, 2002); Michael Morrell, Empathy and Democracy: Feeling, Thinking, and Deliberation (University Park: Pennsylvania State University Press, 2010); Martha Nussbaum, Upheavals of Thought: The Intelligence of Emotions (Cambridge: Cambridge University Press, 2001); Cheryl Hall, The Trouble With Passion: Political Theory Beyond the Reign of Reason (New York: Routledge, 2005); and Sharon R. Krause, Civil Passions: Moral Sentiment and Democratic Deliberation (Princeton: Princeton University Press, 2008).

39. The “control condition,” as the second criterion is sometimes called, is widely thought to be “the very essence of moral responsibility” in liberal societies. Tamler Sommers, “The Two Faces of Revenge: Moral Responsibility and the Culture of Honor,” Biology and Philosophy 24 (2009), 48. See also Joel Feinberg, Doing and Deserving: Essays in the Theory of Responsibility (Princeton: Princeton University Press, 1970); Brian Barry, “Sustainability and Intergenerational Justice,” in Fairness and Futurity, ed. Andrew Dobson (New York: Oxford University Press, 1999), 97; and Iris Marion Young Responsibility for Justice (Oxford: Oxford University Press, 2011), 96.