Why Is Oppression Wrong?

Abstract

It is often argued that oppression reduces freedom. I argue against the view that oppression is wrong because it reduces freedom. Conceiving oppression as wrong because it reduces freedom is at odds with recognizing structural cases of oppression, because (a) many cases of oppression, including many structural ones, do not reduce agents’ freedom, and (b) the type of freedom reduction involved in many structural instances of oppression is not morally objectionable. If the mechanisms of oppression are sometimes indistinguishable from benign, or even ineluctable, processes of social shaping, arguments that oppression is objectionably freedom-reducing risk implausibly suggesting that socialization itself is objectionably freedom-reducing. I show how three strategies for explaining how oppression reduces freedom either fall into the trap of overgenerality, or end up appealing to values other than freedom to avoid it. I conclude by suggesting that oppression might be better thought of as an affront to equality.

In her canonical essay on oppression, Frye (1983) compares the oppressed individual to a caged bird. Her suggestion that oppression reduces freedom is highly intuitive. It is echoed in most major twentieth century feminist philosophical discussions of oppression¹ and in the language of contemporary activists who describe oppression as a force that blocks and thwarts.² The recent philosophical literature, including both work by mainstream political philosophers who wish their theories to address racial and gender injustice³ and work by feminists, socialists, and critical philosophers of race⁴, also includes many versions of the claim that oppression reduces freedom.

The ubiquity of such claims might lead us to believe that oppression is bad for persons because it reduces their freedom. But it is not obvious that this is so. Oppression might reduce agents’ freedom in some cases and not others. Or it might reduce their freedom in ways that are not wrong. Though the idea that oppression wrongs persons by reducing their freedom may seem like a natural way of generalizing from existing discussions, most such discussions have occurred in contexts with different theoretical aims, such as to describe the phenomenology of oppression⁵, to show how liberal theories can be amended to offer resources for responding to race and gender injustice⁶, or to tip the scales in favor of some concept of freedom over another.⁷ Even discussions that have attempted to define oppression have been ambiguous about why it is bad for individuals.⁸

There is also an important insight of oppression theory that weighs against the view that oppression wrongs agents by reducing their freedom. This is the insight that oppression can operate by shaping people’s wants rather than thwarting them, that it can make what are arguably exercises of free choice turn out to be self-undermining. Oppression is able to work in these ways partly because of how it is embedded into our everyday practices, but once we recognize this, it becomes difficult to give a freedom-based account of the wrong of oppression that does not generalize across oppressed and non-oppressed participants in these practices. In what follows, I argue that freedom dimunition is not what makes oppression wrong across cases, and that conceiving the wrong of oppression in this way undercuts another important desideratum for oppression theory: the need to acknowledge the wrong of structural oppression. Structures can oppress agents without diminishing their freedom, or by diminishing their freedom in ways whose harms can only be spelled out with reference to values besides freedom. We need to characterize the wrong of oppression as something besides freedom dimunition, because (a) many cases of structural oppression do not involve diminished freedom and (b) even when they do, the relevant form of freedom reduction is not itself objectionable.

The paper unfolds as follows: I begin by motivating the idea that we should want to characterize the wrong of oppression in a way that includes the effects of structural oppression. I then describe what I call the “structure dilemma” facing attempts to characterize oppression as wrong because it reduces freedom: we must choose between (a) denying that social structures can wrong victims of oppression or (b) adopting an implausibly broad view of what counts as objectionable freedom reduction. In the third section, I describe three distinct strategies for claiming that freedom dimunition is the characteristic wrong of oppression and show that each falls on one prong of the structure dilemma. Finally, I sketch a version of the view that oppression affronts equality that promises to explain what is wrong with instances of oppression that involve reduced freedom and instances that do not.

1 The concept of oppression

The term “oppression” has circulated in Western political thought at least since the modern period. Hobbes uses it to describe the situation of constant threat from other individuals in the state of nature. Rousseau describes the condition of having traded away freedom in pursuit of status and positional goods as oppressive. In her genealogy of oppression, Ann Cudd argues that the now-familiar use of oppression, as subjection to arbitrary or unjust rule, gained currency in the nineteenth century (Cudd, 2006, 6–7).

This paper focuses on a related but particular conception of oppression, the one that predominates in contemporary social and political philosophy, especially feminist philosophy and critical philosophy of race. This conception became differentiated from earlier conceptions through the work of twentieth century social movements, such as feminism, Black power, and the new Left. Their choice of the term “oppression” to illuminate the wrongs to which their constituencies were subject was not accidental. Activists wanted to illuminate just how much arbitrary rule remained in an era where freedom and civilization were touted as having won the day.⁹ Celebratory narratives about economic progress and the victory of liberalism in the twentieth century notwithstanding, imperialist war and economic colonialism continued, African-Americans faced mass voter suppression, and women remained the sexual property of their husbands. To call these harms “oppression” would also counteract Cold War ideology that took relations of control and repression to be the sole province of the USSR (Young, 1988, 271).

The injustices these activists sought to name were not widely thought of as injustices before. Power relations like patriarchy, institutional racism, and the persistence of imperialism after the decline of metropole colonialism, did not always take the form of overt policies and acts, and their dynamics were often difficult to cast as relations between rulers and ruled. The difficulty isolating perpetrators of wrongs with systemic characters is apparent in the emergence of the proverbial “man” in the 1960s who supposedly kept people down.¹⁰ But, as the theory and activism of these movements evolved, the need to bring wrongs that were not caused by individual agents under the scope of oppression became apparent.

A key innovation of these social movements was the view that oppression could be caused by agents because of how they were placed within groups. This was a central motivation for Kwame Ture’s and Charles Hamilton’s coinage of the term “institutional racism” in 1967 and their contrast between it and discrimination. Institutional racism constituted acts by the “total white community against the total Black community” (Hamilton and Ture 1967, 5). However, it also became clear that oppression could continue in the absence of one group intentionally acting against another. The Guinean politician and decolonial theorist Amilcar Cabral argued that colonization continued through the actions of the colonized, in spite of the fact that the colonizer had officially left, because of the mark colonialism had left on the local culture (Cabral, 1974). More recently, bell hooks argued for shifts from the term patriarchy to the term institutionalized sexism (Jhally et al., 1997, 7) and racism to white supremacy (Jhally et al. 1997), partly on the grounds that victims of oppression could perpetuate it.

This turn was part of a second, distinct, theoretical innovation, namely, the idea that oppression could be structural. The term “structural oppression” is sometimes used loosely just to indicate that intragroup harms need not involve discrimination by individuals. However, I have in mind a narrower use of the term “structural”, one that is more consistent with uses in social science and Marxism. Structures are background rules and institutions that determine how agents interact and that contribute to the production of certain social outcomes. They allow us to “explain the behavior of the [individual] by its being part of something larger whose behavior we explain” (Haslanger, 2016, 114). That larger whole is often a social hierarchy wherein structures sort individuals into their relative positions (see Marx, 1977; Reiman, 1987; Haslanger, 2016; Blau, 1977).¹¹ Structures condition the behavior of agents, but also emerge from it such that their operations are not easily reducible to individual behavior or external control. To be clear, thinking that oppression can be structural in this sense does not mean denying that oppression can be perpetrated through intentional agency; it just opens up the possibility that individual or group agency is not a necessary feature of oppression.

The upshot of recognizing structure was the need for a conception of oppression that could accommodate cases of what we might call the “diffuse causation” of oppression. The best known philosophical account of such a conception is Frye’s (1983), on which oppression requires the presence of (a) social structures that (b) select members of social groups for (c) treatment that disadvantages them relative to members of other groups. Note that this account makes dynamics of loss and benefit, rather than the dynamic of the dominant agentically harming the oppressed, definitional for the concept.¹²

Understanding instances of structurally caused oppression as genuine cases of oppression¹³ was critical for the concept of oppression to serve its desired diagnostic—or, in critical theoretical terms—emancipatory role. Anti-oppressive philosophers have long sought to generate normative concepts that reveal injustices that are occluded by our habitual, and ideological, ways of viewing the world. Critical theorists call this unmasking role an emancipatory one (Horkheimer, 1933; Geuss, 1981; see also Haslanger, 2020), but the basic idea is also familiar from the feminist process of consciousness-raising (MacKinnon, 1989, 83–84). Prevailing ideologies long told us, for example, that sexual advances at work were “compliments.” Feminists created the concept of sexual harassment to illuminate the actual moral valence and political function of such advances (MacKinnon, 1989).

The emancipatory potential of many concepts, including oppression, depends on their ability to group together seemingly unrelated phenomena. The concept of sexual harassment illuminated the way “compliments” worked together with things like meetings in strip clubs and requests to make the coffee to enforce occupational segregation (see Superson, 1993). Similarly, the concept of oppression unmasks ideology partly through its unifying role. A central purpose of the concept is to reveal the pervasiveness of the forces that push members of oppressed groups into their situation. Frye analogized the condition of the oppressed individual to that of a caged bird to illustrate the shared function of an array of oppressive forces (Frye, 1983, 4–8). On the analogy, a single oppressive practice or oppressive action is one bar, whose role as part of a cage we will miss if we look at it in isolation. Similarly, the practice of some people making the food and some people watching the game on holidays seems innocuous on its own; taken together with the panoply of practices that push women into housework and care work, not so much.

Put differently, need for a concept of oppression to include structural cases emerges from the need to recognize that practices perpetrated by individual agents and those caused by diffuse social processes often have a shared function. Policies like racial residential segregation and practices of calling looks with proximity to whiteness “beautiful” were, according to Black activists of the 1960 and 1970 s, of a piece. It is not a coincidence, for example, that the Black is beautiful and natural hair movements emerged with civil rights, or that wearing an Afro became a symbol of Black power (Taylor, 2022). To publicly celebrate visual markers associated with Blackness was to make visible just how wide-ranging anti-Blackness was and how transgressive it was for a Black person to have pride. Just being as a Black person who refused to emulate a white look (in the case of hair, a look that, among other things bespoke the hierarchy of house and field slave (see Greensword 2022)), was not acceptable. The idea that practices with diffuse social causes still constituted oppression helped make salient this type of fact.

2 The structure dilemma

The idea that oppression can be diffusely caused puts pressure on the idea that the characteristic wrong of oppression is reduced freedom. To see why, beyond the fact that we often think of limitations on freedom as caused by (individual or collective) agents, it will be helpful to spell out some potential implications of the view that oppression limits freedom. Simply put, the pressure comes from the fact that social structures can negatively impact persons without reducing their freedom, and the fact that, even when structures do reduce freedom, they often do so in ways that are not objectionable.

Here are two examples of structural oppression that we have already begun to discuss:

The Gender Division of Labor

norms and institutions induce men and women to specialize in different forms of labor, where women do more housework, care, and emotional labor, and where the latter are unpaid, or more likely to be poorly paid.

Normalization of White Aesthetic

norms and institutions attribute value to facial features, skin color, hair, forms of dress and speech, and communicative styles associated with whiteness such that whites, and those who emulate or pass as them, are perceived as more credible, beautiful, entitled to basic rights, hireable in high-status positions, etc.

Philosophers who wish to treat freedom reduction as the characteristic harm of oppression face what I call the “structure dilemma.” They end up having to choose between (a) omitting the wrong of structural oppression and (b) claiming that an implausibly broad swath of socialization is objectionable. The dilemma arises, because, as poststructuralist and communitarian philosophers have often noted (see Magnus 2006; Althusser 1970; Taylor 1989) the processes that cause structural oppression are difficult to distinguish from general processes of socialization. By “socialization”, I mean not just gender or racial norm internalization, but also interpellation into practices that frame and scaffold our interaction with others. Such processes range from laws, to the acquisition of linguistic forms and conventions of dress and time use, to the internalization of ethical norms. Many of the processes that normalize the white aesthetic, for example, are identical with those that are typical parts of socialization; being exposed toys that look this way and not that way, consuming cultural objects that one did not construct, encountering some people and not others doing certain types of labor.

In fact, many of the processes that participate in constituting structural oppression seem ineluctable. Being with others usually requires heuristics and shortcuts that enable coordination, ranging from shared language to shared assumptions to behavioral rituals. Learning to speak a language (or two or three) before I have a choice in the matter enables me to connect with others, as does the expectation that I come to the classroom at the appointed time, or the ritual of exchanging money for goods (or bartering for them or giving them as gifts as the case may be). There are, of course, some important senses in which processes like these are restrictive. My enculturation into many of them happens without my understanding their ramifications as practices, often before I am too old to consent to them. Penalties, sometimes serious ones, are attached to my behaving in ways besides the prescribed ones. But these are present in many, many forms of socialization. For example, I adopted the norm of not cutting in line before I could consent, and I will surely face penalties if I start refusing now.

Avoiding absurd consequences is a desideratum for any theory, but there are also some specifically anti-oppressive reasons for avoiding the conclusion that most socialization reduces freedom in an objectionable way. One is that such a view would suggest that oppression harms the dominant and the oppressed in the same way. At the core of the injustices that constitute oppression is a dynamic of victimization, one where there are beneficiaries and losers. But if the wrong of oppression is freedom reduction, and if freedom reduction is understood broadly enough to include general processes of socialization, the dominant and the oppressed are both victims. As Frye (1983) argues, oppression cannot be identical with lack of freedom because all of us have our freedom restricted; some restrictions, like traffic regulations, benefit all of us. Applied to our examples, the forces that normalize the white aesthetic work on the preferences and activities of whites, perhaps even more so than they do on those of people of color. It is not as though the gender division of labor assigns roles to women but not men.

A second anti-oppressive reason for avoiding the conclusion that socialization is generally wrongly freedom-reducing is more contingently related to the concept of oppression. Anti-oppressive theorists have often criticized dominant theories and ideologies for their negative portrayals of relationships and groups, many of which are entered into and maintained by processes of unchosen socialization. Such criticisms take a variety of forms, ranging from the feminist claim that the ideal of a self unencumbered by relationships is androcentric (see Code, 2000; Kittay, 1999; Walker, 2007) to the decolonial claim that the ideal of freedom from culture makes cultural destruction appear necessary for anti-oppressive change (Narayan, 1997; Khader, 2019; Mahmood, 2005), to the Black feminist claim that it is a desideratum of anti-oppressive theory that it explain the wrong involved in being unable to love one’s children (Collins, 2000; Threadcraft, 1990). Theorists who assign a positive valence to some relationships that arise out of unchosen socialization have a special reason to avoid the conclusion that a large swath of socialization is objectionably freedom-diminishing.

3 How oppression might diminish freedom: three strategies

But perhaps a plausible account of the wrong of the oppression can be found by spelling out the details of the relationship between oppression and reduced freedom. In the interests of finding such an account, I turn now to developing the details of three different strategies for arguing that the characteristic harm of oppression is freedom reduction and show how each falls on one prong of the structure dilemma. Since my aim in this section is to explore the theoretical resources one might marshall in favor of the freedom-focused view, I reconstruct three lines of reasoning by which one might characterize the effects of oppression as instances of freedom dimunition.¹⁴

3.1 The option reduction strategy

A first strategy might be to emphasize the role option restriction plays in preventing persons from living the kinds of lives they want to live. Despite her reluctance to define oppression (1991), Young claimed in “Five Faces of Oppression” that all forms of oppression share the feature of “inhibit[ing] the exercise of capacities and express[ing] thoughts, feelings, and needs” (1991, 38). One straightforward way to capture this insight might be to claim that oppression inhibits its victims by making some options unavailable to them.¹⁵

If oppression reduces options, it might interfere with people’s abilities to achieve their aims by directly taking courses of action they desire off the table, or in the long-term, by impacting their agency. The former, direct, link between option reduction and freedom dimunition is easily arrived at by positive liberty and capability views.¹⁶ G.A. Cohen’s argument that capitalism reduces freedom by removing, from each proletarian, the option to leave the proletariat might be thought of as one version of such a view (Cohen, 1983), as might Amartya Sen’s view that gender injustice reduces women’s freedom by making it difficult for them to earn incomes and participate in social decisionmaking (Sen, 1999, 190–95). Though she does not refer to a specific conception of freedom, Frye’s claim that oppressed people face double binds might also be thought of as a claim like this with a slightly different twist: the only options on the table for oppressed individuals are ones that expose them to “penalty, censure, or deprivation,” and it is difficult to do what one wants if one is constantly punished for it.

These arguments are intuitively appealing, and do seem to capture some of the wrongs of some forms of oppression, such as employment discrimination or restrictions on sexual behavior. But there are four problems with using such lines of reasoning to understand the characteristic wrong of oppression. First, in many cases of oppression, there is no prior state from which the oppressed person’s freedom has been reduced. Many oppressed people have always lived under conditions of oppression, and many have internalized oppression or otherwise adapted their desires to it. Second, and relatedly, the fact of adaptive preference means that there will be many cases where the options oppression makes unavailable are options the oppressed individual does not want. This in turn, poses problems for claims that option restriction prevents oppressed persons from living the lives they want. The woman who specializes in caregiving and accepts its attendant disadvantages may never have had the opportunity to be something else. She may also want to specialize in caregiving, such that the gender division of labor does not frustrate her desires.

Third, attempting to apply direct option restriction arguments to cases of structural oppression forces us to appeal to views of freedom that are implausible or that cannot supply a reason to see the relevant forms of freedom reduction as harmful. For example, the option of becoming a breadwinner is not in many women’s sets partly because of the toys they played with, and their experiences of having enjoyed hanging out in the kitchen at holiday gatherings. To take these as diminishments of freedom just because they close off options, would require us to say the same of spending one’s time playing piano instead of flute, or of losing the option to specialize in housework because one has spent all of the holidays watching sports instead of hanging out in the kitchen. Even if we are ready to admit that there is freedom dimunition in all of these (and to bite the bullet that much socialization is freedom diminishing since it always reduces options), it is difficult to see how the freedom theorist can supply the additional claim they need: that such freedom dimunition is a wrong. Such a claim would fall on the second prong of the structure dilemma and thus elide the effects of oppression on the dominant and the oppressed. After all, you can be tracked into hanging out on the couch or in the kitchen during holiday meals, but either way you are tracked.

The natural response for the opportunity reduction theorist to all three worries is to claim that the oppressed have less freedom than they would have in some counterfactual situation. This would solve the first two problems by not needing to rely on claims about the agent’s presocial states or desires, and the third by assigning some special moral weight to the lost opportunities in the oppression cases. However, this response would require characterizing the wrong to the oppressed person as more than a loss of freedom.¹⁷ To determine that someone has less freedom than they should inevitably involves moralizing the counterfactual condition, whether that means making some claims about flourishing and weighing the relative contributions of various options to it, saying that the oppressed ought to have the same options as the dominant, or something else.¹⁸

A fourth issue for the idea that the characteristic harm of oppression is freedom reduction through option reduction is that some cases of structural oppression, especially but not exclusively¹⁹ structural ones, do not involve freedom reduction at all. Reduced options do not always translate into an inability to achieve one’s aims. I have already discussed adaptive preferences, but they are not the only reason an oppressed person might not experience frustration of her aims in spite of oppression. The content of oppressive norms may be such that they are oppressive irrespective of their effects on individual options. This point has often been made about the expressive harm of segregation (see Hellman 2001, Anderson 2005), but is also translatable to the instances of structural oppression we are discussing. A Black woman may fail to notice that that all the advertisements of Black women are hypersexualizing, or she may embrace the idea that Black is beautiful so successfully that she is unaffected, while also happening to be insulated from the effects of employment discrimination by contingencies about her life and community. In these cases, her ability to achieve her aims is not reduced, but she is oppressed. Additionally, oppression can work through the introduction of options, not just their removal. This might happen when one gets a better job conditional on the expectation that one straighten one’s hair, or—as actually happened to me—one’s employer starts to offer parental leave, but contingent on one’s declaring oneself “a primary parent.”

Similar issues arise for the argument that option reduction diminishes freedom indirectly, by impacting oppressed persons’ capacities for self-directed agency. For a number of feminist theorists of agency, especially relational autonomy theorists, the absence of the right opportunities causes what Lugones would call a “blocking between intention and action” (1990, 503). Sukaina Hirji interprets Frye’s idea of the double-bind along these lines, arguing that the harm oppressed individuals suffer is a sort of damaged agency caused by a life of having to consistently act against their interests (664). A host of relational autonomy theorists and critical philosophers of race argue that believing one’s projects are not worth pursuing affects one’s ability to do what one wants (see Moody-Adams 1994; Mackenzie 2008; Superson 2005; Bartky 1990).

To avoid denying the effects of social structures on our wants and desires, or to avoid suggesting that social shaping of our desires is itself a wrong, these indirect arguments must refer to values besides freedom. Being unable to get what one wants because of constraints in the world is a routine part of social life (see Khader, 2011; Hijri 2021), and may even be sometimes required for justice (Khader, 2011, 88), so the frustration of aspirations cannot on its own be the wrong of the double-bind. (And if it is, the dominant, whose aspirations are higher, may be harmed worse by this wrong than the oppressed.) If the mechanism of indirect freedom reduction is instead socially induced low self-esteem, two issues arise. First, many cases of oppression do not cause reduced self-esteem. Second, complying with oppressive norms can increase self-esteem and social recognition. Think, for example, of praise for the traditional homemaker. (see Khader 2014, 2021; Lepold 2018). To say that one is wronged by having to find self-respect in such norms is clearly to refer to values besides freedom.

3.2 The social force strategy

But perhaps the need to appeal to values other than freedom to characterize the wrong of oppression might have been avoided by closer attention to the processes by which oppression impacts its victims. Young suggested on occasions in her work that the harms of oppression are wrong because they are symptoms of pernicious social processes that need to be diagnosed in their own right (Young, 2001, 16). Young’s own remarks about this are limited²⁰, but we can draw on other accounts of the mechanisms of oppression to develop a second strategy for demonstrating that the characteristic wrong of oppression is freedom dimunition. We might claim that oppressive social forces coerce. Where most work on interpersonal coercion picks out evident forms of force (the gun to the head, the threat to your family), philosophers who wish to characterize oppression as coercion face the task of showing how unnoticed, social-level factors can compel.

Analytic Marxist arguments for the claim that laborers are forced to sell their labor often include the claim that oppressive social forces coerce (Ehring, 1989; Reiman, 1987; Cohen, 1983). Since exploitation, as understood by Marxists, is a form of oppression as we understand it here²¹, perhaps their accounts might be extensible to oppressions like sexism and white supremacy. Jeffrey Reiman argues that the requirement that one must sell the surplus value of one’s labor is generated by human institutions, and since the content of this arrangement requires that some portion of one’s work be uncompensated, capitalism is tantamount to slavery (Reiman, 1987, 34–35).

Two different thoughts about coercion seem to be at work in this claim. The first is that theft is a form of coercion, and capitalism requires theft. The second is that subjection to institutional structures to which there are less harmful alternatives is coercion. I will focus on the second, since the analogy to slavery in the first seems more about the injustice of uncompensated work than freedom dimunition. Reiman develops the first argument in a way that seems extensible to other forms of oppression, and that seems helpful for tackling the structure dilemma, by developing an idea he calls “structural force.” According to Reiman, rules and patterns of action can shape individual behavior such that there is no possible outcome besides the division of society into advantaged and disadvantaged groups, even if a few individuals are capable of moving from class to class. He offers the metaphor of a “human bottleneck” where the exits to a stadium are shaped like a bottleneck, and individuals from the stadium must fight to displace individuals in the bottleneck in order to get out. A few may move, but others will fight to take their places, and given the shape and location of the stadium exits, only a few will ever get out (Reiman, 1987, 13).

Reiman designs his view to be structural in the sense of allowing that the dominant group need not be understood as the source of structural force (Reiman, 1987, 12). But the sense in which Reiman thinks structures coerce is a bit puzzling, since the argument is made largely through the bottleneck metaphor. For Reiman, the difference between the force exerted on the worker and the force in interpersonal “your money or your life” cases is largely a matter of degree. To make the case that the worker is forced in spite of having more “play” than the person who is physically coerced (Reiman, 1987, 17), Reiman suggests two different things: (a) that the group is forced even if the individuals within it have play and (b) that the individuals within it are restricted to an “array of choices” that make it most likely that they will end up trapped within their social group (Reiman, 1987, 39).

Both a and b risk defining coercion in a way that makes too many forms of socialization count as wrongly freedom-reducing. Though they may not go so far as to characterize all socialization in this way (not all socialization sorts people into classes), the claims still stigmatize too broad a swath of socialization. Some forces, like the force that causes people in different countries to speak different primary languages, shape the fates of groups, and trap them in certain relations with each other, but do so in a way that is relatively innocuous. Moreover, both dominant and oppressed group members are trapped in their social roles.

Still, it might be argued, though Reiman does not, that the worker is more trapped in their role than the bourgeois, because all the bourgeois has to do to stop being bourgeois is stop profiting from their capital (see Cohen, 1983, 22–23). But this attempt to prevent the social force account from falling on the second prong of the structure dilemma is implausible about forms of oppression besides exploitation. The white person can try to renounce their privileges but they usually cannot do so in any thoroughgoing way. They certainly cannot just become Black. Another strategy for the social force theorist to avoid overgenerality might be to focus, not narrowly on the force that divides society into classes, but more broadly on the force that is entailed in being part of a social order one did not shape. This is the view advocated by Nancy Hirschmann (Hirschmann, 1999), who claims that women lack freedom because others shape the representational contexts that surround them. This seems plausible about corporate media and government in contexts where these are controlled by elites, but these do not exhaust social and cultural processes (especially not in the era of social media), and many oppressive norms predate the laws and media that institutionalize them– and more generally seem to have an existence that is partly independent of them. Once we allow informal cultural processes into the picture, it is far less clear that the dominant are shaping a process others are subject to. For example, most of us, irrespective of our genders or gender identities, find ourselves subject to gender expectations over which we have little control.

One way to refine the social force strategy might be to say, not that the oppressed group faces conditions they did not shape, but that social conditions reduce their options to a degree that counts as force. Perhaps lacking options that do not force one to contribute to one’s own oppression, or lacking options that one is owed as a matter of justice, constitutes coercion. Ann Cudd, who offers the most extended discussion of why oppression is wrong to date, claims that lacking certain options can constitute coercion. Just as the options of the person in a “your money or your life” case have been engineered to coopt their will into producing the outcome the coercer desires, so, too have the alternatives of an oppressed person reduced so as to ensure the continuation of oppression. Cudd offers the example of the woman who becomes a stay-at-home parent because the gender wage gap makes it more economically rational for her to forego her income than for her husband to.

However, Cudd defines coercion in a way that builds in values besides freedom. Cudd defines an institution as coercive if “it unfairly limits the choices of some group relative to others” (Cudd, 2006, 131). Cudd justifies this definition of coercion, with its appeal to values such as fairness and equality, by drawing on a well-known claim by Nozick that the presence or absence of coercion always depends on a background theory of moral entitlements. In order to have been coerced, one has, not just to be deprived, but to be deprived of something to which one has a moral claim. This may or may not be true, but surely having been deprived of something is not the same thing as having been coerced into lacking it. Nozick’s (1969) definition of coercion includes a number of other features beyond the mere presence of deprivation of an entitlement, the presence of a coercer who makes a threat, the worsening of a person’s previous state of affairs, evidence that the threat has affected the will of the victim. None of these is central to Cudd’s picture– and for good reason for a theorist interested in oppression who wishes to take the presence of structural oppression seriously. Oppression is usually a longstanding state of affairs, the oppressed may know no “before” state and may not will that things be different. And, of course, oppression need not be caused by intentional action.

But because the will of the coerced and the intention of the coercer disappear from Cudd’s account, it is unclear to what extent the type of lacking opportunities she has in mind is well-characterized as force. We might try to supplement Cudd’s view with the claim that the oppressed are coerced in the sense that their will is coopted. In other words, we might say that they are coerced in the sense that their options are limited so as to produce the same type of subordinating outcome no matter what they do; being a stay-at-home mom will contribute to the wage gap, but working outside the home will contribute to the devaluation of women’s work. A key problem with this revision is that it misses the fact that it can simultaneously be true that all of the possible outcomes will contribute to subordination of the group and that some contribute more than others to the oppressed person’s well-being (see Khader 2014, 2021). For example, working outside the home is more beneficial for women in mixed-gender households that cannot survive without two incomes (and where the woman’s income is not identical with, or less than, the cost of childcare) even if it does also participate in the devaluation of women’s labor. The oppressed individual’s will is not inert; they are just forced to decide among bad alternatives. We might try to solve this problem, using a strategy Marxists have often used (see Reiman 1987, Cohen, 1983), by claiming that women as a group and not individual women, are coerced. But if we pursue this line of thought, Cudd’s view either dissolves into Reiman’s and takes on board the implication that all social processes that assign fates to groups are problematically freedom-reducing, or into the option reduction view described in the previous section.

If Cudd’s view, or a view that oppression coerces, is simply the view that coercion is a special case of lacking opportunities, it becomes unclear what normative work the terms “force” and “coercion” are doing, and so we no longer have a freedom-based account. We instead have an account that takes on the downsides of the option reduction account—especially those of allowing the evaluative work to be done partly by concepts besides freedom and excluding forms of oppression that do not involve freedom reduction—that is instead characterized as an account of how the mechanisms of oppression undermine freedom. If being coerced is the same thing as lacking opportunities others have, coercion seems to be a denial of equality or fairness (the opportunity lacks are relative to someone), in addition perhaps to a denial of flourishing or well-being. Cudd does not deny this; she just says that this is the background theory that supports her theory of freedom. My point here is not that Cudd’s understanding of coercion is incoherent—just that, it redefines coercion such that coercion is better understood as an affront to equality or fairness than freedom.²²

3.3 The nondomination strategy

Perhaps a better route to establishing that oppression characteristically diminishes freedom is to choose a different theory of freedom, one that does not take coercion to be its paradigmatic infringement. This is what the nondomination strategy attempts to do. Theories of freedom as nondomination hold that the antithesis of freedom is not interference, but rather the absence of self-rule. Alison Jaggar, who explicitly argues that oppression is a species of freedom reduction(Jaggar, 1983, 6), likely has this type of view in mind. Jaggar elaborates that the “constraint” involved in oppression consists in “one group actively subordinating another group to its own interest” (Jaggar, 1983, 6).²³

These nondomination views are plainly inapplicable to cases of structural oppression.²⁴ The gender division of labor is not imposed by men, even if they benefit from it. But there is an understanding of freedom as nondomination that seems more amenable to characterizing the harm of structural oppression, one that draws on Philip Pettit’s view that nondomination is subjection to arbitrary interference, rather than actual rule by another. For Pettit, a person is free to the extent that they cannot be interfered with by those who do not track their interests. If making different choices would make a person become vulnerable to arbitrary interference, that person is not free—irrespective of whether coercion, or interference, actually occurs (Pettit, 1997, 25).

Conceiving oppression as a form of domination seems particularly promising for navigating the structure dilemma, because domination is straightforwardly not the same thing as being interfered with, or having fewer options than one otherwise might have. Because being dominated is not the same thing as being influenced, the nondomination strategy does not need to supply an argument that a broad swath of socialization is harmful; because it is not the same thing as coercion by the dominator, it may seem to leave open the possibility of explaining the wrong of diffusely caused oppression. Pettit takes it to be a strike in favor of his theory of freedom that it captures standard oppression cases (he frequently uses his theory’s unique ability to handle the example of the dominated wife as a reason to adopt it (Pettit, 1997, 57–58).²⁵ A number of feminists and critical philosophers of race have argued that nondomination captures the central insight of anti-oppressive movements, (see Roberts, 2015; Costa, 2013; Einspahr, 2010; Getachew, 2020; Coffee, 2018). The basic insight of those who seek to apply Pettit’s notion of nondomination to oppression is that oppression involves relations where some groups, or members of some groups, are in a position to arbitrarily interfere with others; the slave and the dominated housewife are unfree even if their masters are kind, because the masters are in a position to require them to bend to their wills.

Unfortunately, for our aim of capturing the wrong of structural oppression at least, Pettit’s view retains the feature of the older domination view that made it incompatible with characterizing the wrong of structural oppression: the idea that domination is a relation between two agents. The dominator is a dominator because of how they stand relative to the dominated—the slave master has a legal right to treat the slave as property, the dominating husband controls the purse strings, the world’s rich countries can plunder the world with impunity because there is no power to counterbalance them. But structural cases do not reduce to one agent having the capacity to control another. The gender division of labor is not reducible to the husband’s ability to control the comings and goings of the wife. The normalization of the white aesthetic is poorly understood as a set of acts that involve anyone—whether white people as a group or individual white persons– trying to secure specific outcomes from anyone else.

But perhaps the nondomination strategy can be refined to say something like this: oppressed agents are not wronged by the gender division of labor or the normalization of the white aesthetic; they are wronged by agents who take advantage of the enabling conditions these create. In this vein, Frederick Douglass argues that the end of slavery does not produce freedom because poverty and voter disenfranchisement still allow whites to control Blacks (see Coffee, 2018). Similarly, Jennifer Einspahr argues that structures like rape culture enable men to disrupt women’s lives by sexually assaulting or harassing them with relative impunity (Einspahr, 2010).

It would be a mistake to conclude from this enabling effect of structures that domination is the characteristic wrong of oppression. Many of the results of structural oppression are poorly understood as dimunitions of freedom. The gender division of labor may enable husbands to arbitrarily interfere with wives, and perhaps even legislators (who are overwhelmingly men) to make legislation that is unfriendly to women and families, but this is far from all it does. It denies women equal status and denigrates the work they do; it causes aggregate wage inequality, and results in health disparities.

Moreover, structures can oppress with the intermediary of an agent, but they can also oppress without one. The problem here is not just that the idea of structural conditions as domination-enabling prevents us from explaining how victims of a variety of instances of structural oppression are wronged. It is also that it renders impossible the idea that social structures can themselves be oppressive. The gender division of labor and the normalization of the white aesthetic do not just enable oppressive acts; they are themselves oppressive. As a result, the nondomination strategy falls on the first prong of the structure dilemma; it is insufficiently explanatory of the wrongs of structurally caused oppression.

4 Why is Oppression Wrong?

So it seems that freedom dimunition cannot be the characteristic harm of oppression for two reasons: (a) a number of instances of oppression, often structural ones, do not involve diminished freedom and (b) attempting to characterize these as instances of diminished freedom involves reference to understandings of freedom that cannot supply the objectionable quality of oppression, since they entail that many everyday forms of socialization are objectionably freedom-reducing.

4.1 An objection: nonliberal conceptions of freedom

Before I sketch an alternative account of the characteristic wrong of oppression, I pause to acknowledge what some will see as a limitation of my negative argument: that it criticizes only liberal notions of freedom. Since many anti-oppressive theorists are critics of liberalism, and draw on nonliberal theoretical lineages with distinctive concepts of freedom, it may seem that I have left open the possibility that oppression is wrong because it infringes some nonliberal conception of freedom. Exploring and precissifying these other conceptions of freedom is a topic well worth pursuing, though it is beyond the scope of this paper.

However, my analysis in this paper does give some initial reasons to doubt that the characteristic wrong of oppression is the infringement of even some nonliberal conception of freedom. I have argued that some instances of oppression, especially those that involve only disrespect or derogation (such as the case of the Black woman who must live with hypersexualizing ads) are poorly understood as involving freedom dimunition of any sort. Moreover, many of the positive conceptions of freedom advocated by feminists under care ethical (Kittay, 1999; Weir, 2013) or Arendtian auspices (Zerilli, 2005), are likely vulnerable to the criticisms of the option reduction strategy I made above, because they build substantive values into freedom. Similarly, conceptions of freedom that take groups rather than individuals to be the subject of freedom²⁶ are likely prey to the criticisms of the idea that oppression restricts classes that I I discussed in the section on structural force, or my criticisms of the nondomination view. A particularly common nonliberal view of freedom among radical theorists and activists emphasizes collective self-determination, or what the Black Panther party called “freedom to determine our destiny” (Black Panther Party 1966). But if what I have said about nondomination is correct and there are cases of oppression that do not involve some usurping the agency of others (individual or collective), many cases of oppression do not involve dimunition in this form of self-determination. ²⁷

4.2 An alternative account of the wrong of oppression

If freedom dimunition is not what makes oppression wrong across cases, what does? Our discussion up to this point suggests we need an account that (a) identifies a normatively significant feature of all or most instances of oppression that is (b) not a feature of socialization in general (or a very broad swath of socialization). I want to suggest that the characteristic wrong of oppression can be framed in terms of its inegalitarian effects. The wrong of oppression is being subordinated because one is a member of a group. Much socialization reduces freedom, but much less of it subordinates.

Equality is also well-suited to capture the wrong of oppression, because of similarities between the type of concept it is, and the type of concept oppression is. Equality, like oppression, is a relational concept. Though they also describe oppression as freedom-reducing, Frye and Cudd both note that oppression has to involve comparing the statuses of groups to those of other groups. Frye (1982) understands oppression to consist in a relation between groups, where some are advantaged and others are disadvantaged. Cudd (2006) indexes the presence of oppression to the options available to dominant groups; whether an option deficit is oppressive is determined by what and how much the dominant group has relative to other groups.

The insight that oppression is a form of inequality has of course not been absent from oppression theory. Instead, the role of equality in such accounts has been ambiguous. In some cases, freedom has been something of a red herring; Frye, Young, Cudd, and Jaggar all claim that oppression coerces or reduces freedom while also referring to ideas about inequality, like subordination to the interests of others or lacking what the dominant group has. Freedom language has also sometimes obscured what is morally at stake, since oppression theorists sometimes (re)define freedom so that it requires equality—as in Cudd’s argument that one is coerced if one lacks equal opportunities, nondomination theorists’ arguments that power over abridges freedom, and relational autonomy theorists’ claims that one cannot do what one wants if one thinks one does not have equal worth. These theorists likely introduce freedom to try to make sense of the fact that oppression is a humanly-caused wrong²⁸, but as we know from cases like disrespect and wrongful beneficence; it is possible to talk about wrongs without talking about freedom dimunition. Even to say that a person is wronged by having worse institutions in place than it might otherwise be feasible for them to have is to make a claim about a wrong without a claim about dimunition of freedom by any actual person.

The tendency to discuss freedom and some other value at once has led to a state of affairs where it is simply unclear whether freedom or equality (or something else) is the normative difference-maker. I am suggesting that focusing on inequality allows us to avoid important theoretical downsides of seeing the characteristic wrong of oppression as unfreedom. One of these is the inability to capture the wrong of structural oppression. But this paper has also revealed other important theoretical downsides, such as the inability to make sense of cases where oppression does not involve reduced freedom, including cases where oppression just degrades or denies individuals what they are due, and cases where the oppressive order works through individuals’ choices.

My claim that equality has always been the normative difference-maker may seem surprising, since many oppression theorists have avoided, or even criticized the langauge equality. This, I think, has been a response to the limited ways U.S. political discourse, and liberal political philosophy, have conceived equality. Historian Richard King argues that the Civil Rights movement preferred freedom to equality discourse because they did not want to capitulate to a discourse that treated their status as humans as something that had to be proved (King, 1996, 14). Some feminists, such as Young and hooks, have criticized equality talk for suggesting that justice comes from trying to achieve the same status as the dominant within existing structures, rather than challenging the system itself (see Young 1991; hooks 1984). Much existing equality theory, even that written by feminists, also accepts the public/private distinction (Anderson, 1999).

These critiques of equality are compatible with the idea that oppression is fundamentally a situation of getting less, counting less, mattering less, being treated as less. And feminists, antiracists, and socialists have offered strategies for upending oppressive hierarchies that push toward alternative theorizations of equality. Young argues that we need to move away from distributive equality (1991), but her work is today widely understood as laying the groundwork for today’s relational egalitarianism. Okin argued that egalitarian distribution needed to obtain in both the private and public spheres. It is true that concern with structural oppression may justify more interventionist public policy than many liberals hope for, but, as I mentioned above, it is up for debate the extent to which satisfying liberals is a desideratum for an anti-oppressive theory. If the cost of restricting the demand for equality to relations the liberal recognizes as public is acceptance of background relations of oppression, it is open to the anti-oppressive theorist to conclude that this is so much worse for liberalism—or at least some strains of it.

Conceiving of inequality as the characteristic wrong of oppression is also fully compatible with the view that oppression sometimes reduces freedom. The inequalities oppression causes are often inequalities in freedom. The civil rights movement emphasized freedom partly because they were trying to point out that Blacks had not been fully liberated from slavery (King, 1996). Abortion restrictions offer differential degrees of negative freedom to men who cannot get pregnant than they do to women. The advantage of the equality-focused view, however, is that it does not restrict the purview of oppression-theorizing to these types of cases. It is also oppressive to live in a world where one is tracked into doing the less valuable labor, where others cannot see one (or one cannot see oneself) as beautiful or professional as members of the dominant race, or where one is dehumanized or regarded with disgust. We should not be satisfied with oppression theories that cannot make sense of these facts.

These brief remarks of course do not constitute a complete defense of the view that inequality is the characteristic wrong of oppression. But I hope to have at least planted the seed of an alternative view, and along the way, to have made the case for closer attention to the normative claims in anti-oppressive theory. Such attention is important for clarifying the types of wrongs such movements identify, and to which they attempt to respond, and ultimately for diagnosing and responding to real-world injustices.