Who should fight domination? Individual responsibility and structural injustice

Abstract

Who is responsible for fighting domination? Answering this question, I argue, requires taking the structural dimension of domination seriously to avoid unwillingly reproducing domination in the name of justice. Having cast domination as a structural injustice that refers to structurally constituted positions of power and disempowerment, I show that the outcome-based, the capacity-based and the social connection model suggested in literature on responsibility, fail to fully meet this challenge. Drawing on insights from all of them, I propose an account that proves more sensitive towards the power dynamics at play in fighting domination. It is based on a fundamental duty of justice, which gives rise to two kinds of responsibility. Dominators, dominated and peripheral agents share political responsibility for domination in virtue of reproducing domination by occupying a position within structures of dominating power; they are required to acknowledge and undermine their position of power or disempowerment rather than simply using and thus tacitly reaffirming it. Political responsibility for domination is distinct from moral responsibility for acting within contexts of domination; in fact, ignoring this difference risks reproducing rather than transforming relations of domination. Bystanders who are not implicated in reproducing domination bear limited remedial responsibility to support this struggle.

Introduction

Who is responsible for fighting domination? Republicans hold that domination constitutes one, if not the, paramount evil with which political theory should be concerned. Conceiving of domination as the lack of a status, their focus lies on spelling out institutional requirements of realizing non-domination. However, the crucial issue of who is to establish these institutions and thus to fight domination has hardly been addressed. ¹ One reason might be that the answer seems obvious: Why not simply hold dominators responsible for putting an end to dominating others?

My aim in this article is twofold. First, I seek to show that this view is based on an oversimplified conception of domination that fails to address its structural dimension – and thus overlooks that the dominators’ unilateral attempt to abolish domination reverts into domination. In fact, I argue that any account of responsibility for domination that does not take this structural dimension into account, risks unwillingly reproducing domination in the name of justice. Second, I propose an account of responsibility for domination which proves more sensitive towards the power dynamics at play in fighting domination. Note that my focus is on filling a gap in the republican body of literature. I will neither situate my account in the broader debate on structural injustice. Nor will I engage with recent debates on structural domination. My aim is limited. I will cast domination as a structural injustice and develop an account of how to attribute responsibility for it that avoids reproducing domination and that even those who reject any stronger notion of structural domination can accept.

I proceed in three steps. First, I lay out my conception of domination which casts it as a form of structural injustice. Domination, I propose, is best conceived as a structurally constituted form of power. While I believe that this conception better captures the core idea that domination is a denial of status, I focus here on providing reason for why it should be the starting point for thinking about responsibility for domination even when my modified conception is rejected. However, it requires a model of responsibility that accounts for its structural dimension. I secondly show that the outcome-based, the capacity-based and the social connection model suggested by existing literature on responsibility, fail to fully meet this challenge and thus risk reproducing rather than transforming domination.

Drawing on insights from all three of them, I propose a structurally sensitive account of responsibility for domination. It is based on a fundamental duty of justice, which gives rise to two kinds of responsibility. While dominators, dominated and peripheral agents share relational, political responsibility for domination in virtue of their position within dominating power structures, bystanders bear secondary, non-relational responsibility to support their struggle. Responsibility for domination in this sense needs to be kept distinct from moral responsibility for acting within contexts of domination. And it calls for recognizing and undermining one’s position of power or disempowerment rather than for simply using – and thus reaffirming – it.

Domination as a structural injustice

The concept of domination has been put to the fore of contemporary political philosophy by the neo-republican theory of freedom as non-domination (Pettit, 1997a, 2012, 2014; see also Bohman, 2007 among others). It articulates a distinctive account of the freedom of the person (Pettit, 2007) which is decidedly status-oriented: A free person is defined as a person who enjoys a certain social standing vis-à-vis other persons, i.e. non-domination. This status-based, interpersonal notion of political freedom goes against the grain of liberal conceptions of political freedom as non-interference. They articulate an interactional theory of free choice that focuses on the number and relevance of accessible options. Freedom as non-domination, in contrast, is not primarily interested in the mere availability of options. It draws attention to the structure of social relations between persons. A person is deemed free so long as she retains control over her choices and actions within social relations. Domination, in turn, refers to a relation of power, which deprives the dominated of this kind of control. Hence, the normative relevance of domination does not lie merely in protecting a realm of free choice, as some of Pettit’s critics have it (Carter, 1999: Ch. 8; Kramer, 2008: 44–50). Rather, as Pettit emphasizes: ‘The terrible evil brought about by domination, over and beyond the evil of restricting choice, and inducing a distinctive uncertainty, is that it deprives a person of the ability to command attention and respect and so of his or her standing among persons’ (Pettit, 2002: 351).

As this brief introduction shows, domination is of a systematic quality. Following Pettit, the power asymmetry between dominator and dominated is characterized by the capacity of the dominator to interfere arbitrarily with choices of the dominated (Pettit, 1997a: 52). Everyone who shares the characteristics that mark them as vulnerable to interference of the powerful, for instance being a woman in a sexist labour market, is prone to interference in the same way – and thus equally subjected to their dominating power (Pettit, 1997a: 122). Similarly, everyone who bears the social marker of power, for instance as a male employer, enjoys a capacity to interfere arbitrarily and thus finds him- or herself in a dominating position. Hence, domination is not reducible to discrete interactions between specific individuals. Rather, these individuals occupy positions of power and disempowerment which structure any interaction between them. This is what it means that domination is based on a status asymmetry.

Before moving on to discuss how to assign responsibility for domination, let me highlight two ways in which I modify Pettit’s seminal conception of domination. First, I conceive of domination as the antonym of justice rather than of freedom. This might not seem a significant modification. After all, Pettit has recently cast the ideal of freedom as non-domination as defining justice (Pettit, 2014: xxii; see also Lovett, 2010). There is more to be said on this matter. Given that the free person is defined by an equal status, which in turn delimits the legitimate range of free action, I conceive of the relation the other way around: The ideal of the free person expresses what kinds of freedom of action we may justifiably ask from one another as equals. From this perspective, the requirements of justice are already built into the idea of a free person and hence, justice defines freedom, not the other way around. ² However, for this paper nothing hinges on this difference. What matters is that domination describes the denial of equal status and this constitutes an injustice.

Second, I maintain that given its status-based nature, domination is best conceived of as a structural kind of injustice. I will not be able to fully argue this point in this paper; however, it requires some elaboration as it bears, or so I will argue, on assigning responsibility for domination, even if my account is not fully endorsed. ³ On my view, domination refers to structurally constituted forms of empowerment and disempowerment, not to opportunistic forms of power. Think of the often-cited paradigmatic case of slavery. The slave master enjoys a capacity to interfere with the enslaved, by extracting labour or simply by abusing them, physically or mentally. This capacity, however, is not due to superior physical or mental skills. Even if the enslaved outclassed him in physical strength or intellectual wit, they would remain dependent on the will of the slave master. This is because the latter’s power is based on the legal institution of slavery which posits the enslaved as a personal property of the slave master. In other words: his capacity to interfere arbitrarily with the enslaved is structurally constituted; it is the legal institution of slavery which establishes the positions of power and disempowerment that we refer to as slavery.

The case of slavery is paradigmatic in that it shows how domination is based on social structures ⁴ which posit some as powerful and mark others as dependent on their will – and thus denies them standing. Hence, domination is not a dyadic relation between two (or more) identifiable persons (individuals or groups ⁵ ) who occupy asymmetric positions of power and disempowerment. These positions are constituted and reproduced by countless peripheral agents (Wartenberg, 1990: 144f) who, in their everyday interactions, act on or implicitly accept relevant norms and practices. Peripheral agents do not necessarily stand in any direct relation to the power dyad in question. They neither enjoy the actual capacity to interfere with the dominated person nor are they themselves exposed to the capacity to interfere of that very dominator. In that sense, they are neither dominators nor dominated, at least not with respect to this particular power dyad. ⁶ Nevertheless, by sustaining aligned social practices which empower some while disempowering others, they form part of that dominating relationship and co-constitute it as one of dominating power. ⁷

Taking this triadic nature of domination into account highlights how domination pervades society rather than being an isolated phenomenon between two agents – and thus helps unravel the status-based nature of domination. This systematic quality sets domination apart from mere opportunistic forms of power. Think of a mugger under otherwise just, i.e. non-dominating social circumstances, who happens to encounter me alone at night and seizes this rare occasion to rob me. The mugger may well pose a physical threat to me and certainly limits my choices. Yet, if his power is not structurally constituted, this will not amount to domination because it is not based on a systematic denial of status. Quite to the contrary: if the mugger is caught, he will be sanctioned, and my status will be vindicated. ⁸ In other words, domination is not an interactional but a structural kind of injustice, which is rooted in an asymmetry in standing between persons.

One might agree that paradigmatic forms of domination such as slavery are indeed structurally constituted; but why reduce domination to such kinds of power? Doesn’t being held up by a gunman count as domination just as the case of slavery does? In fact, Pettit’s seminal conception of domination is ecumenical in this way; it includes both structurally constituted and opportunistic capacities to interfere arbitrarily (Lovett and Pettit, 2018: 13; Pettit, 1997b: 68f). This is where my account differs. While I believe that my conception better captures the core idea that domination is a denial of status, ⁹ in this paper I focus on providing reasons for why it should be the starting point for thinking about responsibility for domination even when my conception is rejected.

First, for the issue of responsibility it is crucial to distinguish opportunistic crime or other forms of interactional wrongs from the systematic denial of status. Analysing ordinary crime in parallel terms to the social denial of status ignores the crucial difference between a socially produced injustice and the wrongdoing of an individual who deviates from otherwise just background conditions. While in the former case, the victim has effective institutionalized counter-power to at least vindicate her status after having suffered an interactional wrong, this does not hold under conditions of structurally constituted domination. And while the wrong in the former case is a direct consequence of how the offender acts, in the latter case, the issue is one of power, even if this power is not exercised. Failing to keep these cases apart, conceptually, not only risks trivializing, disguising and stabilizing a dominating status quo; it also ignores that these different kinds of wrongs give rise to different kinds of responsibility.

Second, even supporters of an ecumenical account of domination which includes both structurally constituted and opportunistic forms of power, have good reason to start from the former when developing an account of responsibility for domination. After all, even in their eyes, structurally entrenched forms of power constitute paradigmatic cases of domination. In fact, as I will argue in the following section, focusing on the opportunistic case and treating structurally constituted ones as an extension of it wrongly suggests that domination is based on individual wrongdoing – and that realizing non-domination is exhausted by sanctioning wrongdoers. This view fails to capture the republican emphasis on an institutional response to domination which restructures social relations as relations between equals. Hence, even if my proposal to conceive of domination as a structurally constituted form of power, only, is rejected in favour of an ecumenical account, cashing out how to attribute responsibility for domination will need to start from the structurally constituted form.

One important reason why a structurally informed perspective on domination is often rejected is precisely the problem of attributing responsibility. Lukes for instance holds that accumulated effects of individual actions only matter if there are ‘individuals or groups, who by acting otherwise could make a difference’ since ‘then it makes sense to see the latter as powerful because responsible’ (Lukes, 2005: 67 [original emphasis]). On this view, assumptions about responsibility constrain the conceptualization of domination; it only makes sense to speak of power in the relevant sense when there are agents who can be held responsible. This approach, however, puts the cart before the horse. Instead of approaching the analysis of domination from the perspective of cases that seem in line with standard models of responsibility, the structural dimension of domination calls for developing an appropriate model of responsibility.

Such a model will have to address three important implications of the structural quality of domination. First, domination is not itself an action but a relational position of power and disempowerment. It does not describe what someone else does to me; it refers to a position of power, irrespective of whether this power is exercised. Whether or not a police officer in a racist context ¹⁰ chooses to shoot a Black person or to pull out her gun at all does not bear on whether this is a case of domination. Hence, or so I will argue, there is an important difference between assigning responsibility for domination and assigning responsibility for one’s actions (and especially for how one chooses to use one’s power) within a structure of dominating power.

Second, given that domination refers to a position of power, not an action, it is not necessarily intended. A police officer in a racist context will find herself in a dominating position, irrespective of whether she wishes to dominate. Her actions, of course, may well be intentional, especially when she exercises power; yet, the social reality of domination itself is not necessarily intended (see Pettit, 2012: 62). An account of responsibility for domination needs to be able to capture such unwilled forms of domination. This requires rethinking the link between responsibility and guilt.

Third, domination is a power relation that is not confined to a bilateral setting. The power and disempowerment of a domination dyad like the police officer and a Black citizen in a racist context are rooted in social structures which are reproduced by countless peripheral agents through their everyday interactions. Whether consciously or not, they co-constitute the power dyad as a relation of domination. Hence, an account of responsibility for domination needs to shift attention away from the focus on the power of the dominator towards an analysis of how domination is socially produced and reproduced – and how individuals can be held responsible for such social dynamics.

Three models of responsibility

Assigning responsibility for domination requires a model of responsibility which can account for its structural dimension. I argue that although each of them provides valuable insights, none of the three models found in existing literature fully lives up to this challenge. The model of outcome responsibility focuses too much on isolating unjust actions. The capacity-based model, in turn, does not pay attention to the context of power in which fighting injustices is embedded. Thus, both run risk of perpetuating relations of domination in the name of justice. While Young’s social connection model provides a more promising starting point, it risks collapsing into one of the other two models.

Outcome-based responsibility

An obvious answer to the question of responsibility for domination is to hold dominators responsible because they produced the injustice that is to be rectified. Such a view is based on outcome responsibility (Miller, 2007: 84). Outcome responsibility is a classic model of relational responsibility in that it assigns responsibility based on a certain causal relation ¹¹ between the bearers of duties and the injustices in question. According to this view the slave-master is obliged to fight domination because and insofar the injustice of enslavement has been brought about by him. The crucial part of the argument therefore is to establish the individual contribution of a particular agent to a relation of domination. And this is why it proves unsuited for thinking about responsibility for domination.

First, outcome responsibility relies on identifying those actions that were unjust or lead to an injustice. However, domination is not itself an action; it is a position of power. In fact, there is not even a straightforward relation between actions of the dominator and the fact of domination. Whether or not a slave-master uses his power, say, to whip his slave, might be considered normatively relevant for other reasons; however, it does not change the fact that he dominates the enslaved.

Second, a dominator may not be able to give up her position of power. After all, it does not depend solely on her actions; it is co-constituted through aligned actions of numerous peripheral agents which, taken together, have the accumulated effect of reproducing the underlying power structure. As republicans have long emphasized, a genuinely benevolent slave-master who loathes the power over her slave remains a dominator. Even if a white man in a racist slaveholder society decided not to have slaves, the racist norms that underly slavery will still structure his relationship to any Black person – and bring about similar, even if merely informal forms of interpersonal domination. Fighting domination requires more than that individual wrongdoers stop committing wrongs. It calls for undermining asymmetric power positions and restructuring social relations as relations between people of equal status.

Third, this structural change needs to involve peripheral agents. If they don’t help change the social norms and practices which give rise to domination, there is no prospect of effectively fighting domination. Yet, peripheral agents stand in no direct relation to the dominated and their individual contribution to the injustice of slavery may be equally faint. Hence, the outcome-based model is not able to assign any substantial responsibility to them. And thus, it fails to bridge the gap between assigning responsibility to individuals and the required collective action towards structural change.

One might object that the outcome model of responsibility is at least suitable for opportunistic forms of power – if this form of power is indeed considered domination at all. This may be right. However, treating the outcome-based rationale as the paradigm for how to think about responsibility for domination jeopardizes the distinctiveness of the ideal of non-domination. The outcome-based rationale focuses on isolating moral wrongdoing. This focus wrongly suggests that realizing non-domination merely means stopping people from abusing their power. Domination, however, does not presuppose actual interference. In turn, non-dominating institutions do not merely render interference unlikely by sanctioning wrongdoers; they restructure social relations as relations between equals. Focusing on bringing an end to individual wrongdoing cannot account for this crucial institutional role – and thus misses the entire point of the republican notion of domination. In other words: the outcome model may well be suitable to assign responsibility for how we use our power in contexts of domination. However, this interactional responsibility is distinct from responsibility for domination.

Capacity-based responsibility

An alternative approach to attributing responsibility for injustices leaves the outcome model behind and focuses on identifying agents who can bring about a better state of affairs. Such capacity-based responsibility is non-relational; it grounds responsibility in the mere capacity to remedy injustices rather than on any particular relation to them. It is a remedial, not a corrective kind of responsibility as it does not presuppose a prior wrongdoing. Lovett bases his theory of justice as minimizing domination on such a capacity-based rationale by appealing to a prima facia duty ‘to ensure that people everywhere live under just institutions and practices’ (Lovett, 2010: 226). He explicitly casts this duty as a non-relational form of responsibility, which applies ‘regardless of the particular relationship between the relevant individuals’ (Lovett, 2010: 225), for the simple reason that justice as non-domination ‘cannot tolerate practices directly or indirectly involving domination’, no matter where they occur (Lovett, 2010: 231).

At first sight, the capacity-based model seems better suited for attributing responsibility for domination. It does not isolate actions, which produced injustice. In fact, it completely uncouples attributing responsibility from prior actions by focusing on capacities for remedying injustices. Moreover, it broadens the spectrum of agents who can be held responsible. It obliges everyone who can contribute to realizing justice to do so, irrespective of her role within a context of domination. However, the capacity-based rationale also fails to address the structural dimension of domination, albeit for different reasons.

Think of a man in a patriarchal society who decides to donate to a severely under-resourced women’s shelter to prevent it from shutting down. If the focus is only on what he does, he may appear as a generous benefactor who helps victims of domination – thus ignoring that his act reproduces the pattern of women’s dependency on men to secure their living. He may genuinely intend to help the women and he may indeed contribute to easing their plight. However, domination is a relation of power, not a mere lack of resources. Fighting domination requires addressing these relations of power rather than reaffirming them.

As the example illustrates, attributing responsibility for domination based on the capacity model proves misleading in two ways. Focusing on the capacity to help rather than on an analysis of power relations, it risks misrepresenting dominators as generous benefactors while casting the dominated as destitute persons rather than as victims of injustices. This is a normative mistake as it fails to account for what is normatively significant about this situation: that it is itself a relation of dominating power. But it also implies a practical mistake: it is precisely by shielding the normatively significant relations of power from scrutiny that these power relations are reaffirmed rather than challenged. Failing to take the structural context of power into account, in which any effort to fight domination is embedded, the capacity-based model fails to see how it may reproduce the very injustice it sets out to remedy.

One might object that the capacity model simply needs to assess the capacity of an agent to help remedy injustice in a power sensitive way. However, once power relations are considered, it becomes questionable whether dominators even bear remedial responsibility for domination. Capacity-based responsibility is premised on the assumption that the responsible agent is indeed able to remedy the injustice. Yet, dominators cannot simply dismantle domination without reaffirming their superior position of power. Does that mean that on the capacity-based rationale, dominators do not bear any responsibility to fight domination? To avoid misrepresenting what is at stake, normatively, it seems essential to reconnect the non-relational, capacity-based rationale with some version of the relational rationale and its emphasis on the genesis of injustices.

Social connection-based responsibility

Young’s social connection model of responsibility provides a promising starting point (Young, 2006, 2011). In fact, she combines an (indirect) outcome-based rationale for grounding responsibility with a capacity-based rationale for assigning specific duties. Young starts from the conviction that how injustices come about matters normatively. However, her model is based on the insight that structural injustices cannot directly be traced back to individual actions precisely because they arise as the cumulated outcome of countless actions which, taken separately, might be morally innocuous (Young, 2006: 114). Hence, what matters is not in how far particular agents deviated from normalized background conditions but rather challenging these very background conditions (Young, 2006: 120).

Instead of specifying individual degrees of guilt, Young turns her attention to identifying agents who can be held responsible, together, for transforming those social structures that produce injustice. They are not considered culpable; Young emphasizes that ‘the point is not to blame, punish, or seek redress from those who did it, but rather to enjoin those who participate by their actions in the process of collective action to change it’ (Young, 2006: 122). Their responsibility is a decidedly forward-looking, political kind of responsibility: It holds all participants in ongoing processes, which lead to structural injustices, responsible for together transforming the relevant structures. Hence, it is a shared responsibility that can only be discharged collectively.

Despite its emphasis on forward-looking political action rather than on individual guilt, the social connection model is not a non-relational one which deems the genesis of injustices irrelevant. Young emphasizes that it shares with the classic outcome-based model ‘a reference to causes of wrongs in the form of structural processes that produce injustice’ (Young, 2011: 105). It is our role in how injustices come about, which grounds our responsibility. Yet, these injustices are not direct consequences of our actions; they are mediated through the complex interplay of countless individual actions. Hence, responsibility for structural injustice is an indirect form of outcome responsibility, grounded in mere social connection to injustice (Young, 2006).

Moreover, even though it is the fact of social connection to injustices which generates responsibility, who is to do what is not premised on specifying how exactly an individual contributed to injustices. The content of our duties depends on our position within the structures in question, as this position, in turn, accounts for our power to effect change, for our capacity to adapt to new circumstances, for our interest in transforming certain structures and our capacity to organize collective action (Young, 2006: 127–130; 2011: 142–151).

Young’s social connection model provides a promising basis for attributing individual responsibility for domination. While the relational rationale for grounding responsibility redirects attention to the genesis of injustice, the indirect outcome-based rationale, coupled with capacity-based considerations for assigning specific duties, help overcome the interactional focus of the outcome model. Yet, Young’s model fails to fully avoid the problems of the other two models. First, it remains unclear what exactly a ‘social connection’ to injustice amounts to – and thus, ultimately, who is responsible for an injustice. Does connection mean an attributive cause or contribution to the unjust outcome? In fact, this is what Young suggests, emphasizing that agents bear responsibility ‘because they contribute by their action to the processes that produce unjust outcomes’ (Young, 2006: 119 [emphasis added]). ¹² But what exactly counts as such a contribution to an unjust outcome – and how does distinguishing those who contribute from those who don’t avoid falling back into the interactional picture of outcome responsibility?

Second, Young’s emphasis on not blaming individual agents seems to be self-defeating for the simple reason that time moves on: If an agent has a prospective duty to do x but fails to do so, it seems odd to deny guilt for this failure since that would amount to giving her a moral free pass (see Nussbaum, 2009: 141f). Why then should we not be blamed for failing to discharge our political responsibility? In fact, it seems with the emphasis on not blaming responsible agents, the social connection model runs risk of collapsing into a capacity-based one that refrains from analysing the genesis of injustices altogether and thus risks reproducing it.

Responsibility for domination

Building on aspects of all three different models, I propose a structurally sensitive account of responsibility for domination which seeks to avoid tacitly reproducing domination. I argue first that the fundamental duty of justice provides the normative grounds for two kinds of responsibility for domination, political and remedial; both, however, need to be kept distinct from responsibility for acting within contexts of domination. Second, while dominators, peripheral agents and dominated share political responsibility for domination, bystanders bear remedial responsibility to support them in their struggle – though under conditions of complex, multiple forms of domination, responsibility for domination is likely of a political kind. Finally, the content of political responsibility can be specified based on one’s position of power or disempowerment; however, the focus should not be on simply using these capacities, as Young suggests, but rather on acknowledging and undermining one’s own position.

How to attribute responsibility for domination?

In our capacity as moral agents, we all bear a fundamental duty of social justice. ¹³ It commits us to fight domination and realize social justice (i.e. a practice of non-domination) wherever injustice (i.e. domination) prevails. In contrast to Lovett, who focuses only on its non-relational dimension, I propose to think of the duty of justice as grounding both relational political responsibility for domination in a corrective sense, based on one’s implication in structures of dominating power as well as non-relational responsibility for fighting domination in a remedial sense, based on mere capacity. Moral (or legal) responsibility for how one acts within contexts of domination, however, is distinct and requires different grounds.

In relational terms, the duty of justice applies to one’s own power relations. It calls for fighting domination wherever one is implicated in relations of domination oneself. Relational responsibility for domination is a political responsibility in Young’s sense. It is grounded in the recognition of the fact that, whatever position of power we occupy, when acting within a system of domination we cannot help but reproduce those very positions of power. In that sense, it is responsibility for domination proper, that is, a corrective kind of responsibility. Yet, it accounts for the fact that domination is not an action but a position of power or disempowerment which we may neither wish to occupy nor be able to renounce. We are not blameworthy for occupying this position, but we are responsible for transforming it. And we share this responsibility with everyone who is also implicated in reproducing dominating positions of power and disempowerment, even if only through their habitual interactions.

This account of political responsibility for domination overcomes both problems that plague Young’s social connection model. First, it specifies the kind of relation which gives rise to political responsibility. It is not mere social connection to an injustice which generates the responsibility to fight it, as Young claims, but rather the joint reproduction of domination. ¹⁴ Second, grounding political responsibility in the fundamental duty of justice renders plausible why responsible agents cannot be blamed for failing to discharge their political responsibility. The duty of justice cannot be discharged once and for all since accumulated outcomes of individual actions cannot always be foreseen let alone be controlled. Hence, it requires engaging in an ongoing institutionalized practice of non-domination. This explains why, from the point of view of attributing political responsibility, it does not matter whether an agent tries to fight domination or fails to do so. Her political responsibility persists precisely because of her enduring implication in its reproduction as the dominating power structure simply remains intact so long as adequate institutions and practices are not in place.

Note that this account of responsibility for domination does not dismiss direct outcome-based responsibility for how we act (cf. Young, 2006: 118). It merely emphasizes that political responsibility for domination is distinct from moral responsibility for acting in contexts of domination. Both are relational, corrective kinds of responsibility. Moral responsibility is interactional in that it arises from wrongdoing and obliges the perpetrator to put an end to it. Political responsibility, in contrast, does not refer to actions but to structures, both regarding the basis of responsibility and regarding its content. Whether someone takes advantage of a powerful position to exploit others or uses it to help does not bear on her political responsibility for domination.

Hence, while in the interactional realm, grading responsibility in virtue of degrees of blame or task-responsibility (see Goodin, 1987), benefitting from injustices (Butt, 2007) or complicity (Goodin and Lepora, 2013) might be essential, political responsibility for structural injustices does not come in degrees. Merely occupying a position within a dominating power structure is sufficient to be politically responsible for transforming it and to share this responsibility with others who occupy positions within this power structure – irrespective of the particular position and of how this position is enacted. As I show in the fourth section, the specific duties to which it gives rise may vary in scope and content, but these variations are not variations in degrees.

Of course, dominating power structures do not arise without anyone willing them to be hierarchal in the way they are. Dominators may well – and often do – intentionally use their power, also collectively, to further entrench domination and their own privileges. This likely has strong implications for their moral (and possibly their legal) responsibility for acting in contexts of domination, a responsibility that does come in degrees. However, it does not affect their shared political responsibility for domination. How to evaluate what someone does within contexts of domination is a separate matter. In other words, interactional morality is not the same as social justice. The outcome-based model is complementary to the model of political responsibility; but it applies in a different domain of moral assessment. ¹⁵ Overlooking this difference means missing the structural dimension of domination and thus reproducing rather than transforming relations of domination.

The fundamental duty of justice does, however, comprise a second kind of responsibility: in non-relational terms, it applies to relations of domination in which the responsible agent is not implicated herself. It calls on all third parties who are capable to do so to assist others in fighting domination and setting up non-dominating institutions and practices. This is not a corrective responsibility for domination but a remedial responsibility for supporting the fight against domination based on the capacity rationale.

Non-relational remedial responsibility is a secondary and limited form of responsibility. It is secondary in the sense that it only kicks in if those who bear primary, relational responsibility need support in their struggle to overcome domination. And it is limited in content. The duty of justice requires establishing non-dominating institutions and practices. This involves two steps (see Bohman, 2005; Forst, 2001: 175; Laborde, 2010: 51): it calls for establishing an institutional structure that realizes non-domination in basic respects; once established, it will allow those subject to it to engage in addressing further claims against domination. Insofar as we are part of a context of domination and thus subject to the institutions of basic non-domination, our political responsibility extends to supporting, maintaining and improving these institutions and thus to help realize justice beyond basic non-domination. Non-relational, remedial responsibility by contrast is limited to support establishing institutions of basic non-domination. Once these are in place, realizing further claims of justice is the task of an autonomous practice of non-domination.

Who should fight domination?

Having clarified how to attribute responsibility for domination, I can now specify who should fight domination. Following the structural analysis of domination, I distinguish four groups of agents: dominators, dominated, peripheral agents and bystanders. In their capacity as moral agents, all of them bear the fundamental duty of justice. However, depending on their position regarding relations of domination, it gives rise to different kinds of responsibility.

All agents who are involved in reproducing relations of dominating power share political responsibility for domination. These are first, dominators. Whether they welcome it or not, they find themselves in a position that allows them to interfere unilaterally with the dominated. What matters is not whether they make use of their power; being in the position of a dominator is sufficient to constitute domination – and to generate political responsibility for fighting it.

This is not to say that dominators who abuse their positions of power are not responsible for this abuse. But holding dominators responsible, morally or legally, for how they act is distinct from attributing political responsibility for fighting the very structures that facilitate such abuses. How dominators make use of their power does not change the fact of domination. Hence, the moral responsibility for how they treat those subjected to their power does not bear on their political responsibility for domination. Even if a dominator lives up to his moral responsibility (whatever it requires exactly), this does not diminish his political responsibility to help change the very power structures which place him in a powerful position.

In contrast to any moral responsibility dominators might bear for their actions, their political responsibility for domination is shared with other agents who are involved in reproducing dominating power structures. These are not only other dominators but also peripheral agents. Even though peripheral agents do not have dominating power over a particular other, they play a crucial role in constituting and reproducing domination. By pursuing their legitimate interests within the framework of established norms and practices, they reaffirm the power of the dominators and the disempowerment of the dominated, produce normalizing effects which make these positions of power seem a given rather than a socially produced form of subjection – and share the political responsibility to transform them.

A third group of agents who bear political responsibility for domination are the dominated. It might seem utterly misplaced to hold them responsible. After all, they are victims of injustice. This gives them an interest in fighting domination, but why should they be obliged to do so? Doesn’t this place too high a burden on them and, ultimately, come close to victim-blaming? (see Gould, 2009: 203). Besides, one might think they cannot do much anyway; after all, they depend on the goodwill or indifference of the dominators. Since dominators have a strong interest in maintaining the status quo, challenging dominating structures will certainly not be among those actions that the dominated may perform unconstrained.

Nevertheless, it would be false to conceive of the dominated as passive recipients of injustice. In fact, the reproduction of dominating power also depends on aligned actions of the dominated. This point that has been forcefully carved out by Biko, the founder of the South African Black Consciousness movement. He emphasizes that ‘the most potent weapon in the hands of the oppressor is the mind of the oppressed’ (Biko, 2004: 101f). ¹⁶ The process of dehumanization under Apartheid in South Africa, Biko argues, has turned the Black man into ‘a shadow of man […], a slave, an ox bearing the yoke of oppression with sheepish timidity’; hence, the first step towards emancipation is ‘to make the black man come to himself […] to infuse him with pride and dignity, to remind him of his complicity in the crime of allowing himself to be misused’ (Biko, 2004: 31). Biko’s call for reinfusing a sense of pride among the dominated expresses the conviction that the ‘limits of tyrants are prescribed by the endurance of those whom they oppress’ (Biko, 2004: 100).

Interestingly, republicans have long emphasized a related point. They have highlighted how domination is linked to servile attitudes among the dominated. These mentalities have a major impact not just on the dominated themselves, but also on stabilizing and normalizing dominating structures (Pettit, 2002: 349; Skinner, 1998: 84–96; Viroli, 2012: Ch. 3). The emphasis of this republican theorizing lies not on emancipation but rather on a concern about the corruption of the free polity. Yet, just as Biko, this republican motif illustrates how dominated agents play a role in reproducing their own domination.

Note that to attribute political responsibility for domination, there is no need to track in how far a dominated agent exhibits a servile or rather a resistive mentality. This would mean lapsing into the rationale of the outcome model. The point is precisely not to isolate those agents who are to blame for their actions but rather to identify those who form part of a power structure which gives rise to domination. Even resistant dominated agents will, acting within unjust structures, unwillingly reproduce them to some extent. According to the model of political responsibility, this is sufficient. To what extent the dominated bear interactional moral responsibility for how they act within a context of domination, and especially for whether they resist domination or not, is a separate issue. ¹⁷

A system of domination can only be dismantled if all agents who bear political responsibility discharge it. But what if they do not or cannot effect the required social change? A fourth group of agents who bear responsibility are bystanders. Bystanders are third parties who are not themselves implicated in the structures of dominating power in question but have the capacity to help transform dominating power. Hence, their responsibility is a non-relational, remedial one relying on the capacity-based rationale. It is secondary to political responsibility insofar as it kicks in only if those who bear relational responsibility do not effectively discharge it. And it is strictly limited in content to help establishing institutions of basic non-domination.

However, in real life, the analytical distinction between bystanders and peripheral agents is not as straightforward as it may seem. Peripheral agents are part of the power structures in question while bystanders are entirely unrelated to that context of power. Think of the generous benefactor to the women’s shelter again. At first sight, he may appear as a bystander. Yet, identifying whether that is indeed the case requires a thorough analysis of his positioning vis-à-vis the relations of domination that are to be tackled. Merely assuming the position of a bystander may veil the actual normative relations at stake: especially regarding pervasive structures of dominating power, such as capitalism, sexism and racism, which may all play a role in the struggle of women living in the shelter, putative bystanders may indeed turn out to be peripheral agents or even dominators. A similar point holds when thinking about responsibility for structural injustice in a global perspective. Donors who seek to support the struggle for women’s rights in another country might appear to be bystanders. Yet, under conditions of complex modern societies and transnational dynamics of domination, responsibility for domination is likely to be based on a relational rather than a purely remedial rationale. Simply claiming the role of a bystander without further scrutiny of the relevant power structures and one’s own position within them may indeed serve to mask these complex forms of multiple domination and one’s own involvement in it – and thus perpetuate domination.

How to fight domination?

Dominators, peripheral agents and dominated agents share political responsibility for domination. But who is to do what? ¹⁸ As Young suggests, political responsibility specifies duties based on the position one occupies within a structure of dominating power. This is because one’s position of power or disempowerment largely determines one’s power to effect social change, one’s privileges and capacities to adapt to new circumstance, one’s interest in fighting domination and one’s capacity to organize collective action (Young, 2011: 144–147). However, as I will argue, the focus should be on acknowledging and undermining these very positions rather than on simply using and thus, ultimately, reaffirming them.

Let me start with the dominated. As mentioned above, one concern about holding them responsible for domination is that they usually do not enjoy a particularly wide space of action, especially when it comes to acting upon structures of domination themselves. Yet, they do have something crucial to bring into the fight against domination: Their interest and sense of urgency in fighting domination (cf. Biko, 2004: 24; Young, 2011: 145f). Of course, the dominated might not always be fully aware of the structures of domination they are subjected to. In fact, they may share a normalized view, taking domination and corresponding denigrating views about themselves for granted. However, by exchanging experiences with others, they can change their perception of themselves and thus challenge these normalized narratives. ¹⁹

As Biko highlights, this inward-looking process is an essential first step of emancipation from racial domination precisely because domination relies on the dominated enduring and accepting racialized and dehumanizing narratives and practices. If racial domination is to be transformed, and not merely reformed in a way that ultimately reaffirms its major features, the struggle needs to start from this ‘interrelationship between the consciousness of the self and the emancipatory programme’ (Biko, 2004: 53); it needs to build on a new consciousness which expresses a sense of group pride that defies the dehumanizing racist conception of being Black. ²⁰

Note however, that even on the personal level this first step of emancipation is not limited to mere reflection. Indeed, Douglass recounts his transformative experience of resisting his then-slaveholder, an act which made him recognize his own dignity, moving from ‘a servile coward, trembling under the frown of a brother worm of the dust’ to ‘an attitude of manly independence’ (Douglass, 2003 [1855]: 181) – a crucial step, in fact, as he himself acknowledges, the ‘turning point’ (Douglass, 2003 [1855]: 180f) for him and the struggle against slavery in which he subsequently engaged. While struggling against domination starts from becoming aware of the structural nature of domination and regaining a practical sense of agency, both individually and collectively, it also needs to bring the mechanisms of domination to the attention of a larger public. After all, even though the dominated have the greatest interest in fighting domination, they are not able to dismantle a dominating power structure all by themselves.

Peripheral agents do not necessarily hold a particularly powerful position regarding the power structures in question. They merely reproduce domination through playing by established rules. Yet, they are potential allies in this struggle, precisely because they have no strong stakes in maintaining the status quo. And their support is not only relevant in terms of numbers: Relations of domination crucially depend on aligned social practices that constitute them as relations of power. Once these normalized practices are challenged, disrupted or undermined, the power of dominators will diminish. Hence, peripheral agents are called upon to scrutinize and undermine the normalizing practices in which they are situated.

Dominators might not be particularly interested in changing precisely those structures that put them in privileged positions. Yet, they enjoy a comparatively wide space of action that suggests that they will be able to effect significant changes. They usually enjoy more credibility in speaking out and will find it comparatively easy to organize collective action as they are part of networks with other powerful agents and have access to institutions. This seems to suggest that the focus of well-intentioned dominators should lie on changing given institutional frameworks. However, such a conclusion would be misguided. In fact, specifying duties based on possibilities for action that come with one’s position of power, bears the risk of overemphasizing and thus, ultimately, reinforcing the power that comes with a dominating position.

To illustrate this point, imagine a slaveholder who decides to set her slave free. She may certainly be praised in terms of interactional morality; yet, her action does not fully discharge her political responsibility for two reasons. First, she may no longer be the master of the particular enslaved person she set free and she may not, in fact, enjoy a similar power to interfere at will vis-à-vis any other particular disempowered person. However, she will still, whether willingly or not, be implicated in reaffirming the institution of slavery and underlying racist norms through her everyday actions. Hence, she will retain political responsibility, even if only as a peripheral agent. ²¹ Second, her act is itself an expression of the underlying power structure. By setting her slaves free, she inevitably assumes and reaffirms the power to unilaterally define the terms of her interaction with the disempowered person. Fighting domination requires a transformation of the underlying power structures, not merely a redistribution of power on terms defined by the powerful.

This point, again, has been forcefully put by Biko. Analysing the system of Apartheid in South Africa, he stresses that, ‘no group, however, benevolent, can ever hand power to the vanquished on a plate’ (Biko, 2004: 100). Instead of trying to prove their good intentions by helping Blacks, so called ‘liberal whites’ who claim to oppose Apartheid, should realize ‘that the place to fight for justice is within their white society’ (Biko, 2004: 27); they should concern themselves with fighting white racism and white ‘superiority complexes’ (Biko, 2004: 70). Biko does not deny that dominators bear responsibility for domination in a corrective sense. Nor does he think that dominators may or even should be indifferent to the injustices their power and privilege create. Quite to the contrary. He insists that ‘white society collectively owes the blacks so huge a debt that no one member could automatically expect to escape from the blanket condemnation that needs must [sic!] come from the black world’ (Biko, 2004: 71). But he is adamant in emphasizing that without seriously addressing the underlying structural dynamics including corresponding identities, their attempts at fighting Apartheid is doomed to fail as they will simply reproduce the assumption of a ‘monopoly on intelligence and moral judgment’ which lies at the basis of the White superiority complex (Biko, 2004: 71). ²²

Biko cautions against simply mistaking the dominators’ comparatively wide space of action vis-à-vis the dominated for a call to unilaterally effect change on them. This would reproduce rather than transform extant patterns of domination so long as the underlying mentalities are not transformed. In that sense, even the dominators are constrained (and not just empowered) by the social structures they are supposed to fight. ²³ As Biko maintains, they must start from fighting ‘for their own freedom and not that of the nebulous “they” with whom they can hardly claim identification’ (Biko, 2004: 27). That means: instead of merely focusing on helping Blacks, whites need to genuinely interrogate their own sense of entitlement and White superiority, both individually as well as among them. This also holds for whites who deem themselves non-racist and seek to help Blacks. Without questioning one’s own positionality and the assumptions, interests and privileges that come with it, the underlying racist dynamics of Apartheid will not be transformed but rather reaffirmed.

Coming back to the example of slavery: This does not entail that setting enslaved people free is wrong. It might be a way of discharging one’s moral responsibility. And it might also play a role in addressing one’s political responsibility: After all, it may form part of a subversive strategy to publicly undermine normalized expectations about slaveholders and enslaved, for instance, when formerly enslaved are then employed with a fair salary and this becomes public knowledge, especially among other slaveholders. ²⁴ It is crucial however, that this does go along with the genuine interrogation of the position of power from which such actions are undertaken. In other words: Just as for the dominated and peripheral agents, discharging political responsibility as a dominator entails an important self-reflective dimension. Becoming aware of and undermining one’s own position within structures of domination, including its identity building aspects, mentalities and implicit privileges is essential to avoiding unwillingly reproducing domination in the name of justice.

Conclusion

Thinking about responsibility for domination requires, or so I have argued, an account of responsibility that captures its structural dimension. Domination, at its core, refers not to an action but to structurally constituted positions of power and disempowerment and thus to a structural form of injustice. Even those who dispute that domination should be reduced to structurally constituted forms of power, need to keep such paradigmatic forms of domination distinct from mere opportunistic forms of power in order to be able to provide a nuanced account of responsibility that captures the status-based core of domination.

Classic models of outcome- and capacity-based responsibility, however, fail to address the structural dimension and thus risk reproducing domination in the name of justice. Young’s social connection model fares better, though she does not fully avoid collapsing her model in one of the other two. The account I proposed integrates aspects from all three models. It is based on the fundamental duty of justice, which gives rise to political responsibility for fighting domination within one’s own power relations as well as to remedial responsibility for helping third parties to do so among themselves. Under conditions of complex, multiple forms of domination, however, responsibility for domination is likely of a political kind.

Such political responsibility is shared by everyone who occupies a position within structures of domination, whether as dominating, dominated or peripheral agent. Even though it relies on an indirect outcome-based rationale, it does not track individual contributions to injustice but rather holds them responsible based on the recognition of the fact that, whatever position of power we occupy, when acting within a system of domination we cannot help but reproduce those very positions of power. How to evaluate individual actions is a separate matter. Thus, political responsibility for domination is distinct from moral responsibility for acting within contexts of domination. While both are relational, corrective kinds of responsibility, they refer to different realms of moral assessment. After all, domination is a structural injustice, not an individual wrongdoing. Ignoring this difference results in failing to address the structural dimension of domination and thus risks reproducing rather than transforming relations of domination.

On a republican view, fighting domination cannot be reduced to blaming the dominated and making them change their behaviour. It is a complex social task that obliges everyone who is involved in reproducing domination, dominators, dominated as well as peripheral agents. They are not only required to use their respective spaces of action to push for political change; a power sensitive inquiry into responsibility for domination requires interrogating, acknowledging and undermining one’s own position of power or disempowerment. Without an awareness of one’s own position within structures of domination, its historical genesis and its sociological and psychological implications – and practical attempts at undermining it – any attempt to fight domination is likely to revert into domination.