Global Gender Justice and Epistemic Oppression: A Response to an Epistemic Dilemma

Critiques of Western feminists’ attempts to extend claims about gender injustice to the global context highlighted a dilemma facing Western feminists, what I call the global gender justice dilemma. In response to this dilemma, Alison M. Jaggar argues that Western feminists should turn our attention away from trying to resolve it and, instead, toward examination of our own complicity in the processes that produce injustice. I suggest that this kind of approach is helpful in responding to an additional dilemma that confronts the Western feminist, namely the epistemic dilemma . Western feminists can speak for women of the global South and run the risk of distorting those women’s experience and further silencing their voices , or we can refuse to speak and abdicate our responsibilities to address injustice. I argue that we should address this dilemma not by trying to resolve it but by examining our role in the reproduction of epistemically unjust practices. To explain this response, I offer a preliminary account of epistemic injustice as epistemic oppression. I conclude by claiming that our own epistemic complicity in epistemically oppressive social practices is a weighty reason for us to work to transform those practices.


Introduction
Crisis and dilemma have confronted feminist philosophy at a number of stages. The rapid and revolutionary development of feminist philosophy in the 1970s and 1980s shaped philosophical projects that would face serious challenges to their central categories of analysis. Postcolonial and postmodern feminist criticisms, in particular, questioned feminist philosophy's ability to deliver on its promising radical aim: to utilize philosophical tools to empower women throughout the world to fight for gender justice. A series of seemingly intractable theoretical debates appeared to leave feminist philosophers little hope to develop responsible theory for understanding and working toward global gender justice.
But in two insightful pieces published in 2005, Alison M. Jaggar, in addition to mapping the dilemma confronting Western 2 feminists, argued for an alternative approach to developing philosophical work on global gender justice. Jaggar argued that Western feminist work on global gender justice often omitted the contributory role of Western social practice in shaping processes that cause the gender injustices to which Western feminists were apt to pay special attention. Jaggar urged Western feminists to begin our theorizing by examining our own power and complicity in global processes that cause gender injustice.
Building globalized feminist philosophy from this methodological starting point, Jaggar, Iris Marion Young, Linda Martín Alcoff, and a number of other feminist philosophers developed philosophical frameworks for understanding global gender injustices and Western feminists' responsibilities to work to remedy those injustices. For example, Jaggar and I have argued for a conception of structural complicity as a basis of individual responsibility to work to remedy injustices involved in migrant domestic labor (Aragon and Jaggar 2018). But this work has paid less attention to 2 I use the term "Western" here to mark out the privileged members of the affluent nations of Western Europe and North America (specifically, the United States of America and Canada). In contrast, I use the term "global South" to mark out the subjugated members of nations primarily in sub-Saharan Africa, Southeast Asia, Eastern Europe, and Latin America. I recognize the seeming arbitrariness to this terminology, and I also recognize the potential pitfalls in employing this terminology. However, there is no widely agreed upon set of terms to describe the various places in the world I wish to pick out, and I find my own terminology less problematic than many widely used dichotomies, such as West/East, first world/third world, and rich/poor. I also utilize the terminology of "Western" and "global South" to remain consistent with the philosophical work I reference. I include myself among the group of "Western feminists" to whom this paper is directed. Consequently, the "we" and "us" of this paper are fellow Western feminists and myself. the ways in which global epistemic processes shape relationships that bridge national borders in ways that harm women of the global South as knowers. 3 In other words, recent work on responsibility for global gender justice has yet to sufficiently theorize responsibility for epistemic injustice.
This paper offers a preliminary contribution to this project. To historically situate the project, I first provide a brief history of the development of a global gender justice dilemma that confronts Western feminists, and, drawing on Jaggar's work, I explain this dilemma. Following Jaggar, I suggest that this dilemma can be avoided, to some degree, by focusing on the complicity of Western feminists in processes that cause gender injustice. However, I also suggest that there is a parallel epistemic dilemma, highlighted most prominently in Alcoff's "The Problem of Speaking for Others" (1991), that creates barriers to conceptualizing and working toward global gender justice. To address this epistemic dilemma, I argue that Western feminists need to examine our epistemic relationships to the women for whom we speak, utilizing the concept of epistemic injustice to locate our own complicity in global epistemic injustices. Finally, I sketch a conception of epistemic oppression as a theoretically fruitful conceptualization of epistemic injustice when working to understand what is required of Western feminists to remedy epistemic injustice.

A Global Gender Justice Dilemma
North American feminist philosophy emerged in the early 1970s out of leftist activism and liberal consciousness-raising activities with a radical aim: "to use theory to critique male dominant power relationships in order to empower all women" (Ferguson 1994, 199). "First-generation feminist philosophers in the academy such as Kathy Addelson, Alison Jaggar, Iris Young, Linda Nicholson, Nancy Fraser, and [Ann Ferguson]" drew insights from Marxism and radical feminism to begin to develop theories that would help in combating gender injustice (Ferguson 1994, 201). Jaggar's Feminist Politics and Human Nature (1983) is one of the most important early contributions to the development of feminist philosophy. In this book, Jaggar mapped the philosophical claims made by four distinct camps of feminist theories, formulated these four major (clusters of) theories as fully formed political theories, and critically analyzed each of these views. This extremely influential book exemplified the ambitious aim of early feminist philosophy by providing detailed and nuanced political worldviews for understanding women's oppression and fighting for women's liberation.
However, this early feminist philosophy came under considerable scrutiny in the 1980s. Though feminist philosophy has, from the outset, been characterized by disputes about the meaning of its central concepts (such as gender, equality, liberation, and justice) and the appropriate political means for seeking liberation, the central aim of empowering women in their fight for liberation still served to unite these divergent feminisms. But in the 1980s, the central category of analysis, "woman," was called into question, making the uniting aim seem, at best, untenable and, at worst, oppressive. In particular, emerging anti-essentialist and intersectional critiques demonstrated that the conceptualizations of the category "woman" typically universalized the experiences and politics of a small, relatively privileged (by class, race, nationality, and so on) subset of women, further marginalizing the voices and experiences of women of color and women of the global South. These criticisms presented feminist philosophy with a seeming internal contradiction: the liberatory aim of feminist philosophy appears to require the articulation of a category whose definition must be continually contested (Alcoff 1988). In the late 1980s and early 1990s, a number of prominent feminist philosophers took note of and responded to what they describe as an "identity crisis" (Alcoff 1988), an "impasse" (Ferguson 1994;Fraser 1997), and a "crossroads" (Ferguson 1994).
Postcolonial and postmodern feminist criticism specifically challenged Western feminist philosophers working to theorize the character and moral demands of global gender justice by questioning their authority to make universal claims about justice for women. Western feminists, in extending universalist theories to concrete issues of global gender justice, often focused on scrutinizing non-Western cultural practices that were seen as exemplary cases of injustice, such as female genital cutting, dowry murder, or veiling. Critics charged these attempts to globalize Western feminist theory with mischaracterizing cultural practice, distorting affected women's experience, exoticizing non-Western cultures, and, ultimately, reinforcing some of the oppressive structures Western feminists wished to combat (Mohanty 1984(Mohanty , 2003Narayan 1997). Many of the Western feminist responses to these criticisms invoked the philosophical specters of moral relativism and false consciousness to answer charges of Western feminist colonialism, imperialism, and orientalism (Jaggar 2005a). Jaggar (2005a) explains that a series of debates in feminist philosophy emerged among Western liberal feminists, advocating universal conceptions of wellbeing that transcend culture, 4 and postcolonial 5 and postmodern 6 feminists, who deem the impulse to transcend misguided, distortive, and possibly even oppressive.
Jaggar helpfully parses these debates as rooted in disputes over four distinct philosophical issues: "essentialism, relativism, intervention, and false consciousness" (2005a, 195). These disputes leave Western feminists in a tenuous position. In a candid moment, Jaggar describes this position: These philosophical battles, fueled by political passions, have left some Western feminists, including me, wondering which side we are on. Morally and politically, I have strong sympathies with the anti-imperialist feminists, since it is clear to me that many Western feminist criticisms of third world practices are objectifying, patronizing, and self-congratulatory. At the same time, I do feel responsibility to help women (indeed, all citizens) in the poorer countries of the world. Is it possible to help while avoiding a colonialist stance? (2005a, 188) The disputes within feminist philosophy "often [seem] to imply that [Western feminists] who are concerned about the well-being of women across the world confront a stark and unwelcome choice between colonial interference and callous indifference" (Jaggar 2005a, 186). Let's call this the global gender justice dilemma.
One possible response to this dilemma is to continue to develop and defend new positions within the four kinds of disputes Jaggar highlights; this response tries to escape the dilemma by treading the same lines of argumentation that gave rise to the dilemma in the first place, leading us back into a seeming impasse. However, Jaggar recommends for Western feminists an alternative approach to theorizing global gender justice: "we need to enlarge our philosophical and political frameworks to scrutinize our own complicity and power" (2005a, 195). Jaggar argues that much of the discourse around Western feminist responsibility for global gender justice has "been preoccupied with the question of how Westerners should respond to the perceived injustices of non-Western cultural traditions" (2005a, 189). By focusing in on this question, Western feminists obfuscate the many ways in which we participate in the imposition of these unjust social conditions.
There are a number of ways in which this obfuscation occurs. First, Western feminists often ignore or minimize the impact of Western colonialism on the creation or ascension of illiberal cultural traditions and practices in the global South (Jaggar 2005a, 190). Second, "Western philosophers who complacently contrast our 'liberalism' with their 'illiberalism'" give insufficient attention to the fact that Western nations have often "supported or even imposed" some of the most genderoppressive state regimes (190). Third, Western criticism of non-Western cultural practices "often assumes a sharp opposition between tradition and modernization," an opposition that is strongly challenged by empowerment theorists (191). Fourth, the Western imposition of neoliberal economic policies and practices on poor nations throughout the world has increased inequality both within as well as among nations, and this "has had a devastating effect on the livelihoods of women in the global South by destroying traditional industries" (191). Fifth, women have had a disproportionately difficult time gaining access to the gains in wealth afforded by globalization (192). And finally, women suffer from the "interlocking problems of sexualization, militarism, and environmental destruction" (192).
Jaggar makes a compelling case that: (a) we need to expand our conceptualization of the causes of global gender injustice, and (b) "expanding our understanding of the causes of women's poverty in poor countries requires that we also expand our conception of our responsibility toward such women" (2005b, 70). Jaggar claims, "Once we acknowledge that we share past, present, and future connections with poor women in poor countries, we see that we inhabit with them a shared context of justice" (2005b, 70). By attending to issues of how to structure our shared context of justice through intercultural dialogue, we can shift away from narratives about Western feminists "'saving' poor women in poor countries by proselytizing supposedly Western values or raising consciousness about the injustice of non-Western practices" (70-71). Instead, we can address the impact of Western neoliberal global economic policy, militarism, and environmental practices and begin to discuss Western responsibility for past and ongoing injustice suffered by women of the global South because of those practices (70-72).
In this theoretical vein, Iris Marion Young (2003Young ( , 2004Young ( , 2006aYoung ( , 2006bYoung ( , 2007Young ( , 2011) offers a promising account of moral responsibility for injustice: the social connection model. Young explains: The social connection model of responsibility says that individuals bear responsibility for structural injustice because they contribute by their actions to the processes that produce unjust outcomes. . . . All who dwell in structures must take responsibility for remedying injustices they cause, though none is specifically liable for the harm in a legal sense. Responsibility in relation to injustice thus derives . . . from participating in the diverse institutional processes that produce structural injustice (2011, 105). For Young, the social connection model avoids the theoretical pitfalls of the more prominent model, what she calls the liability model of moral responsibility. The liability model encourages reasoning about responsibility to seek an isolated agent or set of agents who have caused the injustice at hand. The social connection model, in contrast, does not aim to isolate individual perpetrators but, rather, morally assesses the background social conditions that influence individual action. It also conceptualizes responsibility for structural injustice as remedial and forwardlooking, as opposed to retributive and backward-looking, and this remedial responsibility is shared by all who are socially connected to the injustice. Consequently, the social connection model avoids humanitarian impulses to "save" or "rescue" those who suffer injustice by refocusing moral reasoning on our own participation in the social processes that cause injustice. 7 Following Young, I have developed my own account of individual moral responsibility for structural injustice, or structural responsibility. I argue that the only theoretically adequate moral basis of one's structural responsibilities is her structurally mediated social connection to specific injustices. When an individual is socially connected to specific injustices, she is structurally complicit in those injustices and bears weighty moral responsibilities to work toward their remedy. In our coauthored work, Jaggar and I explain this idea: "people are structurally complicit when they exercise their agency in ways that reinforce the unjust structures in which they participate, regardless of their conscious intentions" (Aragon and Jaggar 2018, 449). We explain that structural complicity is a salient type of social connection for grounding our responsibility to remedy injustice and offer this idea as a supplement to Young's social connection model (Aragon and Jaggar, 2018).
However, in Jaggar's call to examine our complicity in injustice, in Young's theorizing about social connection, in my own work on structural responsibility, and in my and Jaggar's coauthored work on structural complicity, the theoretical frameworks that we utilize do not adequately address the ways in which social epistemic practices systematically harm members of oppressed groups. Even after we move forward the discourse on global gender justice to take account of Western complicity and power, we still need to extend this discourse to examine our own complicity in social processes that systematically harm women, particularly in the global South, as knowers.

An Epistemic Dilemma
To direct our focus toward epistemic injustice, I want to first take note of a parallel epistemic dilemma that accompanies the global gender justice dilemma. Alcoff best captures this epistemic dilemma in her article "The Problem of Speaking for Others." Alcoff explains that "the problem of speaking for others" comes from the growing recognition of two premises: Premise 1: The "ritual of speaking" . . . in which an utterance is located, always bears on meaning and truth such that there is no possibility of rendering positionality, location, or context irrelevant to content. (1991,14) And, Premise 2: Certain contexts and locations are allied with structures of oppression, and certain others are allied with resistance to oppression. Therefore, all are not politically equal, and, given that politics is connected to truth, all are not epistemically equal. (1991,15) Because one's social location within a discursive context affects the meaning of what one says, there is no way to speak for others without one's positionality having an epistemic impact. Moreover, since discursive contexts are embedded in systems of social power, one's position within these contexts might lead their acts of speaking for others to further entrench oppressive social systems.
Alcoff's "problem of speaking for others" identifies both epistemic and political dangers of Western feminists speaking for all women. Given the relative privilege afforded many Western feminists, we run the risk of both falsely representing the experience of women subordinated along other axes of social difference as well as furthering some of the very same oppressive practices we are aiming to challenge with our speech. If the two premises above are true, it appears that the problem of speaking for others is inescapable. Thus, it also appears to present an irresolvable dilemma for Western feminists: refuse to speak for others and abdicate one's responsibilities to address gender injustice; or speak, potentially silencing those one aims to empower and potentially doing violence to their experience. Let's call this the epistemic dilemma.
There are two primary ways to respond to the epistemic dilemma. The first is what Alcoff calls the "retreat" response: "to retreat from all practices of speaking for and assert that one can only know one's own narrow individual experience and one's 'own truth' and one can never make claims beyond this" (1991, 17). Alcoff convincingly argues that the retreat response is both ontologically and politically mistaken. Ontologically, the retreat response presumes "an ontological configuration of the discursive context that simply does not obtain," as there is no way to retreat into one's own experience and "truth" external to the discursive context from which she attempts to retreat (1991,20). Politically, retreat both undercuts political effectivity and allows the speaker to abdicate her political responsibilities. As Alcoff puts it, avoiding speaking for others "may result in a retreat into a narcissistic yuppie lifestyle in which a privileged person takes no responsibility for her society whatsoever" (1991,17).
The second way to respond is to argue that the problem is not actually a problem. For example, one might argue that the problem only results from an essentialist or reductionist epistemology-a theory of justification that reduces meaning entirely to an evaluation of one's social location-that ought to be rejected in the first place. This response, what Alcoff calls the "charge of reductionism" (1991,16), claims that the problem only arises if we adopt the epistemological view that one's social location determines the meaning and truth of their speech, which is an essentialist or reductionist view of individual epistemic agency. Though Alcoff quickly dismisses the charge of reductionism by asserting that one's social location can influence without determining the meaning of her speech (1991,(16)(17), this second kind of response appears to venture back into the same debates that arose in response to the global gender justice dilemma. The charge of reductionism exemplifies an approach to resolving the epistemic dilemma that pushes us back into debates around essentialism, relativism, Western intervention, and false consciousness, though the debates will now be more explicitly epistemological.
Reprising these debates leads us only further into the epistemic dilemma and provides Western feminists with little recourse in responding to the problem of speaking for others. Alcoff argues that we cannot escape the problem of speaking for others but are still required to speak. Rather than trying to avoid the problem, Alcoff provides four "interrogatory practices" to help ameliorate the negative effects of speaking for others: 1. We should carefully analyze "the impetus to speak," even fighting against it in many cases (1991,24). Specifically, we should examine whether our impetus to speak stems from a "desire for mastery and domination" or an "impulse to teach rather than listen to a less-privileged speaker" (24). 2. We should "interrogate the bearing of our location and context on what it is we are saying" (1991,25). This interrogation works best in dialogue with others, "by which aspects of our location less highlighted in our own minds might be revealed to us" (25). 3. We should be accountable and responsible for acts of speaking for others, which requires "a serious and sincere commitment to remain open to criticism and to attempt actively, attentively, and sensitively to 'hear' (understand) the criticism" (1991, 26). 4. We should "analyze the probable or actual effects of the words on the discursive and material context" in our acts of speaking for others (26). In other words, we "must look at where the speech goes and what it does there" (26).
These practices are helpful reflexive reasoning exercises that aim to ameliorate some of the potentially oppressive effects of speaking for others. While Alcoff's interrogatory practices do not give us a clear approach to resolving the epistemic dilemma-actually, they concede that the dilemma cannot be resolved-they focus our attention on the fact that we regularly distort, marginalize, or even silence the voices of the oppressed in our attempts to call attention to and combat oppression. Consequently, Alcoff's analysis of the problem of speaking for others illuminates that our common responses to the epistemic dilemma are likely to further entrench us within the confines of the dilemma. And while her interrogatory practices can help us to avoid some of this entrenchment, we might also try to find a way to address the conditions that give rise to the dilemma in the first place.
I suggest that we should address the epistemic dilemma by shifting the discussion away from the consideration of responses (retreat, denial, interrogatory practices) and toward examination of the ways that we actively participate in unjust epistemic practices. This shift will allow us to look at (a) the ways we act out epistemic practices that distort, marginalize, and silence some voices while amplifying others, and (b) our complicity in the injustice of those practices. In other words, an alternative way of addressing the epistemic dilemma is to examine our complicity in global epistemic injustice.

"Border-Crossings" and "Death by Culture"
At this point, it would be helpful to offer an example of the kind of unjust epistemic practice that I have in mind. Consider Uma Narayan's careful and nuanced discussion of gender violence in different cultural contexts in "Cross-Cultural Connections, Border-Crossings, and 'Death by Culture'" (1997). Specifically, Narayan analyzes the way in which the practice of dowry murder garnered the attention of Western feminists, was distorted in the process of crossing over to the Western feminist context, and ultimately reinforced exoticized images of the cultural Other.
Narayan identifies "two sorts of problems that often beset the general project of 'learning about Other cultures'" within the context of agenda-setting in Western feminism (1997,84). Narayan explains, "the first set of problems has to do with features of context that 'bring' particular issues onto feminist agendas, mold the information that is available on the issue, and shape as well as distort the ways in which they are understood when the issue 'crosses borders '" (1997, 84). Narayan argues that Western feminist attention tends to focus on "the problems of women in Other cultures" (1997, 100) when those problems "seem 'Different,' 'Alien, ' and 'Other'" (100). For this reason, Narayan asserts that problems that will cross borders are most likely to be those issues of gender injustice that are marked by their "Otherness" (101). With respect to dowry murder, Narayan highlights two features of the practice that mark it out as "Other," namely dowry and burning by fire (101). When the issue of dowry murder crosses borders, the issue becomes distorted: Western feminists come to understand the issue independent of the contextual information that is needed to properly understand it, recontextualize it by drawing on background assumptions about cultural "Otherness," and ultimately, transform the issue (Narayan 1997, 104). These kinds of border-crossings distort the issues toward which Western feminist attention is directed and undermine Western feminists' ability to actually understand the issue they are trying to address (104).
The second set of problems "has to do with the ways in which 'culture' is invoked in explanations of forms of violence against Third World women, while it is not similarly invoked in explanations of forms of violence that affect mainstream Western women" (Narayan 1997, 84). According to Narayan, an invocation of culture to explain fatal violence against women in the global South implies that these women suffer "death by culture" (84). Moreover, Narayan claims that Western explanations of fatal forms of gender violence often invoke a largely imagined and exoticized culture, religious views, or traditional values (1997,112). Dowry murder, for example, often gets attributed to Hindu mythology, religion, and traditions, even though they provide very poor explanations of the practice (1997,107). And Narayan contends that, in the parallel case of fatal gender violence in the United States, we would not offer up a parallel "death by culture" explanation for the violence.
While I cannot do full justice to Narayan's insightful analysis in the space here, I want to highlight two features for the purposes of my argument. First, the problems of border-crossing and death by culture found in the case of Western feminist attention to dowry murder serve as a paradigmatic example of the problem of speaking for others. When Western feminists turn our attention to issues of gender injustice that affect women in the global South, we will often misrepresent those women's experiences by offering our own explanations and solutions to their problems. We will also often provide the kinds of simplified and exoticized cultural explanations for these issues that we would not find convincing for parallel issues in our own cultural context.
Second, Narayan's discussion provides a clear case of epistemic injustice and, moreover, a case of epistemic injustice that is best understood as a form of epistemic oppression. I will reference this example below to argue that, in responding to the epistemic dilemma, we should turn our attention to our own epistemic complicity in oppressive epistemic practices.

Epistemic Injustice
To develop this idea, I will first introduce the concept of epistemic injustice. In Epistemic Injustice: Power & The Ethics of Knowing, Miranda Fricker (2007) names the concept of epistemic injustice. 8 Epistemic injustice is, according to Fricker, "a kind of injustice in which someone is wronged specifically in her capacity as a knower" (2007, 20; emphasis in original). Fricker argues that epistemic injustices are based in the exercise of identity power: "an operation of power that depends in some significant degree upon . . . shared imaginative conceptions of social identity" (2007,14). Identity power can be "agential," exercised by some particular agent(s), or operate "purely structurally," wholly a feature of social systems (2007, 10). Fricker articulates two distinct forms of epistemic injustice: Testimonial injustice occurs when prejudice causes a hearer to give a deflated level of credibility to a speaker's word; hermeneutical injustice occurs at a prior stage, when a gap in collective interpretive resources puts someone at an unfair disadvantage when it comes to making sense of their social experience. (2007, 1) Let's take a closer look at each of these types of epistemic injustice.
Testimonial injustice involves a "prejudicial credibility deficit" on the part of the hearer, and the "central case" of testimonial injustice with which Fricker is concerned is found when this credibility deficit is based in "identity prejudice" (2007,27). The relevant identity prejudices, such as gender or racial prejudices, are both systematic and persistent. They are systematic insofar as they "track the subject through different dimensions of social activity-economic, educational, professional, sexual, legal, political, and so on" (27). They are persistent insofar as they are "repeated frequently" (2007, 29). As Fricker explains, "'Persistent' labels the diachronic dimension of testimonial injustice's severity and significance, whereas 'systematic' labels the synchronic dimension" (29). The central cases of testimonial injustice for Fricker involve credibility deficits on the part of the hearer because of systematic and persistent identity prejudice.
Hermeneutical injustice, in contrast to testimonial injustice, "is a purely structural notion" (2007,159). According to Fricker, hermeneutical injustice is "the injustice of having some significant area of one's social experience obscured from 8 Jaggar (2015) has noted that feminist philosophers have long discussed the ways in which women's knowledge, from a wide variety of feminized social standpoints, has been dismissed, discounted, and marginalized, and, indeed, much of the work in feminist epistemology begins from or is centered around this fact. In other words, the concept of epistemic injustice has been discussed by feminist philosophers for most of the existence of academic feminist philosophy. However, Fricker's groundbreaking work gave a new name to this concept and provided a robust articulation of the epistemic character of this injustice. social understanding owing to a structural identity prejudice in the collective hermeneutical resource" (2007,155). The collective hermeneutical resource provides our means for interpreting and sharing our experiences to both make sense of our own experience and to share the meaning of that experience with others. Identity prejudice in the collective hermeneutical resource causes "unequal hermeneutical participation with respect to some significant area(s) of social experience," leaving those targeted by this prejudice "hermeneutically marginalized" (2007,153). Hermeneutical marginalization disallows those who are marginalized from developing the resources to meaningfully understand and share their own experience. This is what Fricker calls "a situated hermeneutical inequality," and it is the primary wrong of hermeneutical injustice: "the concrete situation is such that the subject is rendered unable to make communicatively intelligible something which it is particularly in his or her interest to be able to render intelligible" (2007,162). Fricker continues to explain, "the primary harm of (the central case of) hermeneutical injustice concerns exclusion from the pooling of knowledge owing to structural identity prejudice in the collective hermeneutical resource" (162).
Fricker's conception of epistemic injustice provides a different approach to addressing the epistemic dilemma. Rather than focusing in on the parameters of the familiar debates, we should rethink our own role in perpetuating epistemic processes that systematically harm women of the global South as knowers. We should examine the ways in which Western epistemic practices structure epistemically unjust interactions among citizens of affluent Western nations and members of oppressed groups in the global South. Focusing in on epistemic injustice avoids the problematic assumptions about the internal epistemic failures of non-Western cultures, the autonomy of different cultural groups' epistemic practices, and the privileging of Western epistemic practices.
Fricker's conceptualization of epistemic injustice, though philosophically nuanced and an important introductory step, suffers theoretical limitations for accomplishing this task of reframing. 9 First, Fricker's account of epistemic injustice at times conflates injustice, which is a moral property of systems of social interaction, and interpersonal wrongdoing, which is a moral property of the interpersonal conduct of individual agents. 10 Second, Fricker's account, though 9 The theoretical limitations noted here are informed by recent critiques of Fricker's account by Anderson (2012), Dotson (2012), and Medina (2013). 10 This distinction between social justice, on one hand, and interpersonal ethics, on the other, follows from John Rawls's (1999) view that social justice is about the character of the "basic structure" of social organization and not about the character explicitly claiming that epistemic injustice is not of a distributive character, often utilizes distributive language that obscures the relational character of injustice. 11 Third, Fricker's account gives insufficient attention to nondominant testimonial and hermeneutical practices in a manner that potentially furthers the very injustice it aims to address. 12 Finally, Fricker's account, in focusing on the epistemic character and behavior of the hearer, undertheorizes the exercise of epistemic agency of those that suffer epistemic injustice.
In mentioning each of these theoretical limitations, I do not mean to suggest that any of these are debilitating criticisms of Fricker's view, nor do I take it that I have established any kind of theoretical inadequacy or fatal flaw in her conception of epistemic injustice. Instead, I offer the above limitations as methodological reasons for developing an alternative conception of epistemic injustice. The account I offer is merely a preliminary attempt to offer such an alternative.

Epistemic Oppression
Given these four theoretical limitations, I, following recent work by Kristie Dotson (2012Dotson ( , 2014, conceptualize epistemic injustice as epistemic oppression. 13 Dotson defines epistemic oppression as "epistemic exclusions afforded positions and communities that produce deficiencies in social knowledge," where "an epistemic exclusion . . . is an infringement on the epistemic agency of knowers that reduces her or his ability to participate an epistemic community" (2012,24). While this definition of epistemic oppression provides a strong starting point for analysis, I wish to articulate an account of epistemic oppression in a manner that comes apart from the approach to this issue found in Dotson's project. Taking up the foundational ideas of epistemic oppression, I supplement them by articulating four conditions of epistemic oppression.
Oppression is the defining feature of structural injustice. Social structures are enduring yet dynamic systems of social relationships that significantly shape the or conduct of individuals. For more on my understanding of this distinction, see Aragon and Jaggar (2018, 442). 11 A distributive understanding of social justice will focus on whether or not the distribution of social opportunities and resources to participants in some system of social cooperation satisfies the moral principles that should govern that distribution. In contrast, a relational account of social justice will focus on the moral character of the relationships among social groups. 12 See Dotson (2012). 13 In an early article, Fricker (1999) conceptualized epistemic injustice in terms of epistemic oppression, but she largely moved away from this conceptualization in her 2007 book. options for building a life for all who participate in them. Social structures position individual agents in relation to one another in a manner that influences the options for action open to them, their share of social resources for acting on those options, and the sets of social incentives and disincentives for acting on some options rather than others. In other words, social structures are the background social architecture within which individual agents shape their lives.
Even if all social structures are coercive, not all structures are unjust. Rather, structural injustices occur when social structures systematically, but avoidably, harm members of some social group by positioning them within oppressive social relationships. Oppressive social relationships make some people vulnerable to domination or deprivation, while at the same time enabling members of related social groups to prosper. This vulnerability is the result of one's agency being placed under severe social constraints by social processes that are largely outside of one's control. And these severe constraints wrongfully harm the agents who suffer them. 14 Young famously claimed that there are no unifying features shared by all cases of oppression, even if minimally all members of oppressed groups "suffer some inhibition of their ability to develop and exercise their capacities and express their needs, thoughts, and feelings" (1990,40). But I disagree with Young. Following Ann Cudd (2006), 15 I identify four necessary and jointly sufficient conditions that, when met, classify a structure as oppressive and thus, unjust: 1. The avoidable social harm condition: A system of social relationships avoidably harms some of the individual agents who are situated within it by placing upon them severe social constraints for the exercise of their agency.
14 My own accounts of social structure and structural injustice are heavily shaped by Young's conception of structural injustice. (See, for example, Young, 2000Young, , 2003Young, , 2004Young, , 2006aYoung, , 2006bYoung, , 2007Young, , 2011. Also, Jaggar (2009) has led me to include social vulnerability in my understanding of oppression. 15 These conditions parallel those developed by Cudd, who offers the harm condition, the social groups condition, the privilege condition, and the coercion condition as her four necessary and jointly sufficient conditions (2006,25). My conditions are distinguished from hers more in the articulation of the specific character of the condition than in the general feature of oppression each condition is meant to pick out. My view of oppression, thus, is deeply informed by, and indebted to, Cudd's work but departs from her conception in the articulation of the conditions of oppression.

The social group condition:
The avoidable social harm is also systematic insofar as the severe social constraints are placed upon individuals because of their social group membership. 3. The social privileges condition: There is some other social group in relationship to the harmed group that enjoys a wide range of opportunities and resources to shape their own lives and even flourish. 4. The violation of equal moral respect condition: There is a violation of a basic principle of equal moral respect.
On my account, oppressive social structures systematically harm some while privileging others, based on their social group memberships, in a manner that violates some general principle of equal moral respect. I offer this account to begin to map out a conception of epistemic injustice based in epistemic oppression. I contend that epistemic injustices occur when social epistemic processes systematically, yet avoidably, harm members of some social group as knowers, by positioning them in epistemically oppressive social relationships. Epistemically oppressive social relationships position members of some social group to suffer vulnerability to epistemic domination or deprivation, while at the same time enabling members of related social groups through a wide range of opportunities and resources to exercise their epistemic agency. This vulnerability to epistemic domination or deprivation is the result of having one's epistemic agency placed under severe constraints by epistemic processes that are largely out of one's control. Epistemic agency is an individual agent's capacity to come to know things about herself and her social world and to contribute to processes of social knowledge production. 16 Social processes shape one's epistemic agency by (a) influencing epistemic opportunities (personal experiences, interpersonal dialogue, public forum), resources (conceptual schemas, written and oral language, grammars, contextual clues, institutional access), and sets of social incentives and disincentives (social interest, social policing, social costs and rewards) to (b) understand one's experience, meaningfully articulate that experience to self and others, and contribute to the social accumulation of knowledge. When an agent suffers severe constraints on the exercise of her epistemic agency, these severe constraints wrongfully harm her as a knower.
From this conception of epistemic oppression, we can articulate four necessary and jointly sufficient conditions for a social epistemic process to be oppressive: 1. The avoidable epistemic harm condition: The epistemic process avoidably harms some of the individual agents who rely upon it to produce knowledge of self and world by placing upon them severe constraints for the exercise of their epistemic agency. 2. The social group condition: The epistemic harm is also systematic insofar as the severe constraints on epistemic agency are placed upon the individuals because of their social group membership. 3. The epistemic privileges condition: There is some other social group in relationship to the harmed group that are provided greater epistemic opportunities, resources, or sets of social incentives and disincentives to more fully exercise their epistemic agency. 4. The violation of equal epistemic respect condition: There is a violation of a basic principle of equal epistemic respect.
These four conditions offer a preliminary sketch, rather than a fully articulated account, of epistemic oppression. Let's return to Narayan's discussion of border-crossings and death by culture to help illustrate these four conditions. First, the distortions and misrepresentations of dowry murder provided by Western, decontextualized interpretations of that practice avoidably harm Indian women by placing constraints on their ability to produce knowledge of their own experience. Specifically, Indian women must negotiate the distorted meanings mapped onto their own experience to attempt to render their experiences intelligible to the many others who have internalized those distorted meanings. This will make it even more difficult for Indian women to convey to Western feminists their own understandings of the role that gender violence plays in the Indian national context. Second, Indian women suffer these constraints on their epistemic agency because of their social group membership. More concretely, distorted Western explanations of dowry murder that explain this particular form of gender violence by appeal to culture must posit that there is a group that is targeted by these cultural forms of violence. In doing so, these explanations write cultural interpretations of gender violence onto the experience of the members of the group thought to be targeted by that violence. Third, there is a co-relative group that is enabled in the exercise of their epistemic agencynamely, the group of Western feminists that are provided with the epistemic resources to shape which "issues" will garner Western feminist attention, whose interpretations of those issues are shared and celebrated, and who set the conditions for critical exchange on those issues. Finally, the relationship between the Western feminists that directed their attention to issues like dowry murder in a manner that distorted the issue and the Indian women who had to suffer the resulting constraints on their epistemic agency is one of unequal epistemic respect. At a general level, this just means that this relationship has the character of treating one group as though they are not equally capable as knowers and thus are not worthy of the equal epistemic respect.
Again, this is meant to provide only a preliminary sketch of an alternative approach to epistemic injustice. But, even if only preliminary, we can draw on this account to reinterpret the two paradigmatic forms of epistemic injustice. On my account, a hearer's giving a credibility deficit to her interlocutor because of her own identity prejudice does not, on its own, constitute an epistemic injustice, even if it is wrong. Rather, the injustice of testimonial injustice is located in the epistemic processes that systematically discount the testimony of members of some social groups while privileging the testimony of members of other, related social groups. Hermeneutical injustice, on my account, is not located in identity prejudice being built into the collective hermeneutical resource; rather, it is located in epistemic processes that systematically constrain members of some social group in the development and exercise of their hermeneutical capacities while at the same time privileging the hermeneutical practices of members of some other, related social group. What makes these paradigmatic cases of epistemic injustice unjust, on my account, is that they harm the individual agent as a member of a social group by positioning her in epistemic relation to others in a manner that treats her as a lessthan-equal epistemic agent. Consequently, the wrongfulness of epistemic injustice, on my view, is found in the character of the epistemic relations among social groups and not merely in the harms suffered by individuals. For this reason, my conception of epistemic injustice offers an alternative, relational account of epistemic injustice.

Conclusion: Responding to the Epistemic Dilemma
With this sketch of epistemic oppression in hand, we can now return to the epistemic dilemma I discussed above. Recall, the epistemic dilemma confronting Western feminists is found in the conflict between the imperative to speak out against gender injustice suffered by women throughout the world, and the inherent political and epistemic dangers involved in the act of speaking for others. I offer as a suggested alternative approach to this dilemma that we focus on our own participation in epistemic oppression. This approach begins by asking us, Western feminists, to examine our epistemic relationships to the women who suffer the injustices to which we direct our attention. We should first ask ourselves if the way we currently exercise our epistemic agency-the way that we participate in social processes of knowledge production-restricts the agency of others. Additionally, we should ask if we derive and enjoy epistemic privileges as a consequence of the restriction of someone else's epistemic agency. Do we silence, marginalize, exploit, or render powerless the voices of women in the global South through the exercise of our epistemic agency in discussing their situations? Do we enjoy greater epistemic opportunities and resources to discuss their situations by restricting their agency? Or do our actions of speaking for relieve, to some extent, the constraints on their agency? The concept of epistemic oppression can help us to examine, descriptively and normatively, our epistemic relationship to differently situated women across the world.
Moreover, the concept of epistemic oppression helps to identify an alternative basis of Western feminist moral and epistemic responsibilities to women of the global South. Much of the debate in Western feminism has assumed that the basis of our responsibilities to women of the global South is either (a) an abstract humanitarian or liberal cosmopolitan duty to combat gender injustice wherever we find it and have the capacity to prevent it, or (b) a kind of culpability for continuing colonial and imperial processes that cause greater injustice. These two bases of responsibility-capacity and culpability-frame, respectively, Western feminists as entirely disconnected from the gender injustices they "discover" in the world or as the primary cause of those injustices. If one takes capacity as the appropriate basis of responsibility, then Western feminists ought to do whatever they can to aid or assist women suffering injustice. If one takes culpability as the appropriate basis of responsibility, then Western feminists ought to discontinue their actions that are thought to cause the injustice.
For reasons I cannot go into here, neither of these bases of responsibility are adequate for grounding Western feminist responsibility for global gender justice. But by beginning analysis of this responsibility from the perspective of epistemic oppression, we locate an alternative basis: epistemic complicity. Epistemic complicity is a state of being in which our agency is bound up in unjust epistemic processes. When our agency is bound up in unjust epistemic processes, the exercise of our agency becomes implicated in the harms that those processes cause. In other words, we become accomplices to the wrongs committed by our social epistemic processes. When complicit, we are neither completely disconnected nor fully culpable for the unjust processes in which we participate; rather, we share in the wrongfulness with all those who participate.
Epistemic complicity generates weighty responsibilities to change-or, more precisely, transform-the unjust set of epistemic relationships in which we participate. Alcoff's sets of interrogatory practices help us to think about our motives of speaking for others and the effects of our acts of speaking for. But, while these reflexive practices are important reasoning tools for deciding when and how to speak for others, they do not identify how we can work to transform unjust systems of epistemic relations. For this, we need to also examine the features of our own social location within epistemic processes to locate transformative resourcespower, privilege, collective ability, interest (Young 2011)-afforded us by those same processes. We need to examine our epistemic habits to change epistemic vices built into the character of our own individual epistemic processes (Fricker 2007;Medina 2013). We need to develop networks of epistemic solidarity to address injustices at the structural, rather than purely individual or interpersonal, level. And we need to focus less on "washing our hands" of the epistemic injustices that confront us and more on the arduous task of transforming our social world.