Feminist philosophers have had a complicated relationship with autonomy, as the ideal of personal autonomyFootnote ¹ has been consistently associated with qualities of nonrelational self‐sufficiency, individualism, and “hierarchical control” (Reference BensonBenson 1990, 51; see also Reference Friedman and MeyersFriedman 1997; Reference Mackenzie and StoljarMackenzie and Stoljar 2000, 5). This led some feminist philosophers (Reference JaggarJaggar 1983; Reference CodeCode 1991) to criticize autonomy for being inveterately “masculinist” (Reference Stoljar, Mackenzie and StoljarStoljar 2000, 94). In the last couple of decades, however, feminist philosophers have attempted to overcome the limits of the mainstream conceptions of autonomy and reclaimed autonomy as a feminist value while retaining its core idea as “self‐government” (Reference Mackenzie and StoljarMackenzie and Stoljar 2000, 5).Footnote ² The reconceptualized autonomy has been dubbed “relational autonomy,” an “umbrella” term to encompass approaches to autonomy predicated on a “shared conviction” that “persons are socially embedded and that agents’ identities are formed within the context of social relationships and shaped by a complex of intersecting social determinants, such as race, class, gender, and ethnicity” (4). Feminist philosophers of relational autonomy argue that the concept of relational autonomy can overcome the masculinist biases of mainstream conceptions of autonomy and serve as a plausible universal normative conception of agencyFootnote ³ that would be conducive to feminist goals (Reference Friedman and MeyersFriedman 1997; Reference Mackenzie and StoljarMackenzie and Stoljar 2000; Reference MeyersMeyers 2000; Reference Stoljar, Mackenzie and StoljarStoljar 2000; Meyers 2002; Reference Oshana, Paul, Miller and PaulOshana 2003; Reference WestlundWestlund 2003; Reference Benson, Christman and AndersonBenson 2005; Reference Benson and TaylorBenson 2005; Reference OshanaOshana 2006; Reference WestlundWestlund 2009; Reference NedelskyNedelsky 2011; Reference Mackenzie, Veltman and PiperMackenzie 2014; Reference Meyers, Veltman and PiperMeyers 2014). At least in the Western context, theories of relational autonomy seem to have succeeded in bridging autonomy and feminism.Footnote ⁴

A thorny question remains, however, regarding women in traditionally and currently nonliberal third‐worldFootnote ⁵ nation‐states. Certainly, women in nonliberal third‐world nation‐states are not monolithic. They are often divided along the lines of not only ethnicity, race, and class, but also ideology; some third‐world women subscribe to liberal values and Western feminism, whereas others defend their traditional nonliberal cultural/religious values and practices. In this article, I reserve the term third‐world women to refer to the second group of third‐world women who embrace their nonliberal cultural/religious values and practices.Footnote ⁶ Third‐world women in this sense have posed a conundrum to Westerners, including feminist theorists, as these women's choices, motivations, and actions cannot be explained in terms of liberal individualism. Consequently, some Western feminists have claimed that third‐world women's agency is compromised due to “false consciousness” (Reference OkinOkin 1994) or “adaptive” preferences (Reference NussbaumNussbaum 2000).Footnote ⁷ Against such a backdrop, feminist debates have intensified in the last decade regarding the agency of a subgroup of third‐world women—“religious women”Footnote ⁸ (Reference SinghSingh 2015; see also Reference AvishaiAvishai 2008; Reference BrackeBracke 2008; Reference BilgeBilge 2010; Reference BucarBucar 2010; Reference BurkeBurke 2012; Reference WeirWeir 2013; Reference KhaderKhader 2016)—often alleged to be the most subjugated even among third‐world women.

If relational autonomy is to serve as an inclusive universal ideal of human agency, then it should also be applicable to women and men of all societies, including nonliberal third‐world nation‐states, and provide a normative standard by which to assess their agency. As feminist conceptions of relational autonomy have upgraded mainstream conceptions of autonomy, the claim that relational autonomy is at the core of normative agency with universal applicability seems to have prima facie validity. But is this conception of normative agency universally applicable? This article considers this question in relation to IslamistFootnote ⁹ women whose “oppressed” status has been wildly exaggerated and sensationalized in the West (see Reference Abu‐LughodAbu-Lughod 2002). To date, two feminist philosophers of relational autonomy have applied their theories to a particular instance of Islamist women—“the Taliban woman”Footnote ¹⁰—and came up with conflicting conclusions (Reference Oshana, Paul, Miller and PaulOshana 2003; Reference OshanaOshana 2006; Reference WestlundWestlund 2009). This article examines this debate closely in order to determine whether feminist theories of relational autonomy succeed in illuminating Islamist women's agency.Footnote ¹¹

In the process of scrutinizing this debate, I hope to bring two subfields of feminism—anthropological cum social‐scientific feminism on the one hand and feminist philosophy on the other—into conversation by assessing Islamist women's agency, intensely debated by feminist anthropologists and social scientists since 9/11, in terms of relational autonomy, a unique contribution by feminist philosophers. My hope is that the interdisciplinarity of this article will shed light on Islamist women's agency by combining the best of both worlds, the sound empirical research of the former and the conceptual clarity and logicality of the latter. To that end, the article proceeds as follows: The first section revisits a debate about a hypothetical instance of Islamist women—the Taliban woman—by two feminist philosophers of relational autonomy. Since I suspect that philosophers’ tendency to focus on hypothetical examples is likely to distort rather than clarify the issue under discussion, I introduce in the second section the case of female participants of the Women's Mosque Movement (WMM) in Egypt (Reference MahmoodMahmood 2005) as a pertinent, real‐life instance of Islamist women. In the third section, I elaborate on why the two feminist conceptions of relational autonomy fail to illuminate WMM participants’ agency. Although this may seem to entail a pessimistic diagnosis of WMM participants’ agency, I conclude by proposing and briefly examining an alternative normative conception of agency that may elucidate WMM participants’ agency.

I. Relational Autonomy and the “Taliban Woman”

Feminist philosophers of relational autonomy have debated whether Islamist women who embrace their nonliberal religious values and practices can be considered autonomous by focusing on the hypothetical Taliban woman. Marina Oshana, who first introduced the example, characterizes the Taliban woman as a woman living under the Taliban regime in pre‐2001 Afghanistan who has “embraced the role of subservience and the abdication of independence” that she thinks is mandated by the Quran (Reference Oshana, Paul, Miller and PaulOshana 2003, 60). Perhaps she believes that a life of abject subservience conforms to her spiritual values or endows her with a sense of worth or satisfies her idea of well‐being (104). According to Oshana's “social‐relational” approach (Reference OshanaOshana 2006), the Taliban woman is unambiguously nonautonomous. This is so, according to Oshana, because autonomy is “having authority over one's choices and actions whenever these are significant to the direction of one's life” (Reference Oshana, Paul, Miller and PaulOshana 2003, 100).

An autonomous person is able to set goals for her life among a range of feasible options, which conform to her desires and values that have been formed in an “uncoerced” manner and which she would affirm as important upon critical reflection. Whether a person is autonomous is a “global” phenomenon pertaining to her entire life rather than a “local” phenomenon relating to individual actions (Reference Oshana, Paul, Miller and PaulOshana 2003, 100). An autonomous person is someone who is “in control” of her choices, actions, and will throughout her life and has “the power to determine how she will live” (101). Autonomy understood in this sense can be impaired not only by “inner,” psychological obstacles but also by “external” or social impediments, such as “[m]anipulation and intimidation” (102). An autonomous individual must not only be able to withstand external impediments but also not “be disposed to impose impediments upon herself” (103).

Oshana is willing to allow that the Taliban woman may not be suffering from an internal, psychological handicap, such as weakness of the will or low self‐esteem (102). The Taliban woman may be psychologically competent, capable of critical reflection, and fully voluntary in her endorsement of women's subservient status in her society. The Taliban woman is devoid of autonomy, however, in that she willingly accepts the notion that women's subservient status “espoused in certain passages of the Quran” is sacred and embraces the sense of purpose in acquiescing to a life of utter dependence on others, which she believes is implied by this notion. As a result, the Taliban woman cannot support herself financially, nor have legal custody of her children. She must remain costumed in “cumbersome garb” that covers her entire face and body—a burqa—when in public and is not allowed to travel by herself. She is also aware that “any transgression, any show of independence counts as heretical defiance and invites punishment both swift and harsh” (Reference OshanaOshana 2006, 60). The Taliban woman who willingly accepts a life in which she cannot exercise “self‐directed agency” (Reference Oshana, Paul, Miller and PaulOshana 2003, 102) or “practical authority” is not autonomous in the global sense, although her initial decision to embrace women's subjugated status might be autonomous in the local sense (104).

In contrast, we find in Andrea Westlund's theory of “dialogical answerability” an attempt to recognize the Taliban woman's relational autonomy (Reference WestlundWestlund 2009). According to Westlund, autonomy consists in “the disposition to hold oneself answerable to external critical perspectives on one's action guiding commitments” and requires “an irreducibly dialogical form of reflectiveness and responsiveness to others.” In other words, the key to autonomy is the dialogical disposition to hold oneself answerable to others—a disposition for “dialogical answerability” (Reference WestlundWestlund 2009, 35). Although this disposition is internal to the agent, it is “engaged by what is external to the agent” (33; original emphasis), and to that extent this conception of autonomy is “constitutively relational” (27).

Westlund rightly recognizes that not all Taliban women voluntarily accept their subordinate status. Yet she intends to show that even those who do accept such a status are not necessarily devoid of autonomy. If a Taliban woman is “prepared to take up and respond to the critical perspectives of others,” even if she does not find their arguments convincing (29)—Westlund calls her the “responsive” Taliban woman (33)—then she is autonomous. In other words, if the responsive Taliban woman is disposed to answer for her unwavering commitment to her subservient role in response to critical challenges posed by her critics, she cannot be considered as lacking in autonomy (34). At first glance, dialogical answerability expected of the responsive Taliban woman may seem to require not only critical self‐reflection but also actual discursive engagement with critics. Indeed, in her previous article, Westlund emphasizes the importance of “critical reflection” (Reference WestlundWestlund 2003, 485, 492–93; original emphasis). In her 2009 work that discusses the Taliban woman, however, Westlund argues that dialogical answerability does not require the agent to have gone through “a suitably rigorous process of critical scrutiny” (Reference WestlundWestlund 2009, 34). In order to be autonomous, all that is required is that the agent have a certain kind of “self‐relation” to hold herself answerable to external critical perspectives regarding her commitments (35); it may not involve a face‐to‐face conversation and may simply take place internally. Even then, the requisite dialogical answerability does not have to involve “a willingness or ability to cite reasons on demand” (39).

Furthermore, the disposition for dialogical answerability is applicable only against “legitimate challenges” that meet two necessary conditions: “relational situatedness” and “context‐sensitivity.” In the first, the challenge must be “situated in a way that makes relational sense of the intervention” (39); and in the second, the challenge must be “context‐sensitive with respect to the kind of response it invites and tolerates” (40). Here I focus on the first condition, which is more relevant to our discussion. The condition of relational situatedness requires that the critic be in a relationship to the challenged in a way that the concern raised by the critic makes sense in relational terms. For example, it must be clear “why it matters to the critic why I think and act the way I do, and it must matter to her in a way that she can reasonably expect to matter to me.” Such “sense‐giving” relationships can be broad, such as being members of the same moral community, citizens of the same nation, or it can be narrow, such as being family members or neighbors. If such a relationship does not exist, Westlund argues, challenges posed by critics may be rejected or ignored by the challenged as “inappropriate” or even “outrageous” (39; original emphasis).

Westlund argues that the responsive Taliban woman, characterized in this way, is different from “deeply deferential agents,” whom she considers compromised in their autonomy. Deeply deferential agents “endorse their deference” without providing any basis that is “not itself deferential” (32). They exhibit “self‐abnegating” deference, which involves “the systematic subordination of oneself to another whose interests, needs, and preferences are treated as pre‐emptively decisive in one's own practical reasoning” (Reference WestlundWestlund 2003, 485). The “Deferential Wife (DW),” who is a paradigm example of the deeply deferential agent, subordinates her interests and desires to those of her husband and “organizes her will around them.” DW's deliberation itself takes a “distinctly deferential form” (487). The attitude expressed in self‐abnegating deference is “self‐undermining” as it involves “denying” or “effacing” oneself (486). When pressed to answer why they always defer, DWs will “simply persist in referring their interlocutors to the perspectives of those to whom they defer.” Westlund argues that such agents are “merely in the grip of the concerns that motivate that reasoning” (Reference WestlundWestlund 2009, 32) and thereby exemplify “passivity.” The responsive Taliban woman willing to engage in dialogue with her critics, although firmly committed to her perhaps mistaken values, is, by contrast, “not just passively in [the] sway” of her values (34).

II. Women's Mosque Movement Participants in Egypt

It is undeniable that Oshana and Westlund take pains in their discussion to present the Taliban woman in the best possible light. Yet it is worth reminding ourselves that the Taliban woman analyzed by Oshana and Westlund is at best a sympathetic and charitable figment of Western feminists’ imagination. She is intended to capture a paradigmatic third‐world religious woman who wholeheartedly subscribes to and lives by her nonliberal religious values and practices. Yet, given the extremely coercive conditions for women under Taliban rule, it is difficult to imagine that many women would have willingly submitted themselves to an ideology of thorough and excessive gender subordination. Indeed, Elaheh Rostami‐Povey argues that an overwhelming majority of Afghani women exerted their agency by resisting, rather than conforming to, the Taliban impositions in both subtle and explicit ways. A minority of women who genuinely identified with Taliban ideology were not helpless victims, but rather enjoyed the elevated status as enforcers of Taliban ideology and “ruthlessly suppress[ed]” other women (Reference Rostami‐PoveyRostami-Povey 2007, 35). Although it is possible that some instances of “the Taliban woman” may have existed, the number would be too few, if there were any at all, to represent real religious women under the Taliban regime. Hence, taking this largely imaginative reconstruction as representative of third‐world religious women risks flattening the lives of such women into a monolithic stereotype that seems to confirm the Western “imperialism of the imagination” that views third‐world religious women as either “dupes” or “prisoners” of Islamic patriarchy (Reference Narayan, Antony and WittNarayan 2002, 419). It is thereby more likely to distort rather than shed light on such women's agency.

I believe that the only way to confer basic respect on those who are our cognitive and moral equals and do justice to third‐world religious women's agency is to be attentive to the ways that they themselves describe their motivations and actions and to be as charitable as possible to their perspectives. In order to do so, I focus on the case of Women's Mosque Movement (WMM) participants in Egypt meticulously portrayed in Saba Mahmood's Politics of Piety (Reference MahmoodMahmood 2005). The reason for my choice is twofold: First, the perspectives of WMM participants come alive in a richly detailed ethnography presented in a balanced and even‐handed way by a first‐rate feminist anthropologist with a keen eye for subtlety and nuance. Second, and perhaps more important, WMM participants superbly exemplify third‐world religious women who wholeheartedly embrace and live by their nonliberal religious values and practices.

Mahmood's subjects are female participants in the “da‘wa”Footnote ¹² or Women's Mosque Movement (WMM) in Egypt. The da‘wa movement is associated with the Egyptian Islamist group Muslim Brotherhood's political goal of establishing an Islamic state regulated by sharia laws. Islamist movements in North Africa and the Middle East have gained legitimacy in the eyes of ordinary Muslims, as they advocate “social justice, equity, and solidarity” (Reference SalimeSalime 2011, 10) not only in opposition to policies and actions of the government but also “the power of the religious establishment or mainstream ulama” (9). When the Muslim Brotherhood's leadership explicitly repudiated violence as a means of achieving the Islamist goal in the 1970s and adopted “a gradualist approach,” they took it upon themselves to transform society from the ground up, by engaging in collective efforts of outreach, charity work, and, most important, education. The idea is that once Egyptians became Islamically educated, they would willingly accept the Islamic government and sharia “out of their own convictions” (Reference AhmedAhmed 2011, 73). Therefore, da‘wa is understood as “a religious duty” of all Muslims to urge fellow Muslims to “greater piety” by teaching one another “correct Islamic conduct” (Reference MahmoodMahmood 2005, 57).

The WMM is an offshoot of this larger da‘wa movement (4). As the da‘wa movement became popular in Egypt, neighborhood mosques had “dramatic” increases in attendance by both women and men. To meet the demands for religious instruction, women became actively involved in these movements, as they began to organize weekly religious lessons to read the Quran and other religious literature, first at their homes and then within mosques (3). In the process, many women took on the role of “da‘iya”—literally meaning “one who practices da‘wa” (57)—who provides religious teachings in mosques and helps others cultivate “those bodily aptitudes, virtues, habits, and desires that serve to ground Islamic principles within the practices of everyday living” (45). The WMM is very “broad‐based” (43) and involves women from a variety of socioeconomic backgrounds (2). The reason is that qualifications for becoming da‘iya exclude doctrinal expertise on religious texts, but include “one's moral uprightness and practical knowledge of the tradition.” This has enabled women to participate in the movement, although they mostly lack formal training in religious doctrine (65).

Mahmood recognizes that Egyptian women's participation in the WMM presents a “dilemma” for feminists (5). On the one hand, their activism seems to exemplify progress toward gender equality, as the WMM has enabled Egyptian women to hold public meetings in mosques for the first time in Egyptian history. Consequently, this movement has changed the “historically male‐centered” character of mosques and Islamic pedagogy (2). On the other hand, WMM participants occupy an “uncomfortable” place in feminist scholarship, because they are firmly embedded in a patriarchal tradition of Islam and seem to perpetuate its patriarchal practices and ideals (4). The idioms that Islamist women adopt within this tradition are grounded in discourses that “secured their subordination to male authority” not only historically, but also currently (6). Consequently, women's participation in the movement has certain limits. For instance, WMM participants have not challenged gender hierarchy within Islamist organizations; women are encouraged to carry out da‘wa only among other women and are not allowed to do so among men. Hence the roles of “khatib (one who delivers a sermon)” and “imam (one who leads the prayers)” are reserved for men (65).Footnote ¹³

Furthermore, women's improved public role in religion and politics seems to be predicated on women's acceptance of “feminine” virtues, such as “al‐haya,” “sabr,” and fear. Al‐haya means shyness, modesty, and humility (6), and is one of the most “feminine” of Islamic virtues (155), highly encouraged for all pious Muslims. For women in particular, it is considered “necessary” for piety (156). Sabr means to “persevere” without complaint in the face of difficulty (171), which WMM participants consider “an essential attribute of a pious character” (172). Fear is “the dread one feels from the possibility of God's retribution (such as, fires of hell)” (140), and Muslims consider it one of “the critical registers by which one monitors and assesses the progress of the moral self toward virtuosity.” Its absence, on the other hand, is taken to signal “an inadequately formed self” (141). Muslims view fear as a necessary condition for the “felicitous” performance of a pious act (145), as it deters the believer from engaging in actions and thoughts that may “earn His wrath and displeasure” (140).Footnote ¹⁴

These virtues advocated by WMM participants may seem profoundly problematic to Western feminists. Both al‐haya and fear are antithetical to autonomous agency as self‐government, as the cultivation of such virtues will deter women from actively pursuing their “own interests and agendas” (2) and render them submissive to externally imposed goals and values. Cultivating sabr seems to complete the circle of female submission, as it engenders a “defeatist and fatalist” attitude in women (173). The defense of such patriarchal virtues by WMM participants, then, may seem to exemplify “deplorable passivity and docility” from a feminist point of view (15).Footnote ¹⁵

III. Relational Autonomy and WMM participants

How might the two feminist philosophers of relational autonomy assess WMM participants’ agency? Recall that Oshana's approach determines that the hypothetical Taliban woman lacks autonomy, because she chooses to accept “the role of subservience,” which she believes is entailed by the Quran. WMM participants’ voluntary commitment to a life of da‘wa may seem reminiscent of the Taliban woman's acceptance of women's subservience “anchored in religious piety” (Reference OshanaOshana 2006, 60).

Yet the similarities end here. WMM participants do not subject themselves to a life of utter subservience that will entail the severely restricted life of the Taliban woman described by Oshana. At least two reasons, among others, can be adduced in support of my assessment. First, the da‘wa movement to which WMM participants are committed does not promote women's abject subjugation. As mentioned earlier, the Muslim Brotherhood in Egypt has gained legitimacy in the eyes of ordinary Egyptian Muslims, women included, as it advocates “social justice, equity, and solidarity” (Reference SalimeSalime 2011, 10). A major part of the appeal of Islamism to WMM participants, therefore, has been its social ideals and programs that promote social justice. Second, WMM participants have in fact benefited from their participation in the WMM, as it has actually opened up opportunities for women and expanded their role within their religious institutions. Not only are women participating in previously “male‐defined spheres” on a relatively equal footing (Reference MahmoodMahmood 2005, 5), but, more important, their presence is viewed by many religious scholars as consistent with verses of the Quran that enjoin women and men to undertake the duty of da‘wa “equally” (65).

Oshana, however, may still reach a pessimistic conclusion about the agency of WMM participants. An autonomous individual must not only be able to withstand inner or psychological obstacles but also external or social impediments, which include the imposition of “unreasonable conformist attitudes and role expectations” (Reference Oshana, Paul, Miller and PaulOshana 2003, 102). Moreover, she should not be disposed to impose impediments on herself (103). Despite their expanded role in the da‘wa movement, WMM participants are still steeped in a tradition that promotes “subordination to a transcendent will (and thus, in many instances, to male authority) as its coveted goal” (Reference MahmoodMahmood 2005, 2–3). Indeed, they regard “submission to certain forms of (external) authority” as a condition for realizing the subject's potentiality (31). WMM participants thereby seem to exemplify “an obsequious deference to social norms that both reflects and reproduces women's subordination” (157). Since Oshana believes that someone who imposes on oneself unreasonable conformist attitudes and role expectations is not autonomous, then WMM participants are not autonomous.

How would Westlund's account of relational autonomy deal with WMM participants? Westlund argues that as long as the Taliban woman is disposed to answer for her unfaltering commitment to her subservient role in Muslim society in response to critical challenges, she is “responsive” and ought to be considered autonomous (Reference WestlundWestlund 2009, 34). Would WMM participants be considered similarly autonomous? This conclusion may seem logical from the fact that WMM participants often engage in critical exchanges with their da‘iyat regarding lessons received from the latter. For example, a group of young female students between the ages of fifteen and twenty‐two participating in Islamic lessons organized by a conservative mosque in Cairo partook in critical and even “contentious” dialogue with the da‘iyat about the practical implications of ikhtilat (“mixing and blending” of unrelated men and women) (Reference MahmoodMahmood 2005, 100). The da‘iyat did not attempt to reject or restrict the challenges coming from the students, and the discourse was quite “equitable” despite the age differences among the interlocutors. The “contentious character” of pedagogical conversations in mosque lessons, according to Mahmood, is quite common (105) and is consistent with the protocols of religious debate established by Muslim jurists (106).

It must be borne in mind, however, that the critical exchanges between the da‘iyat and young female students regarding ikhtilat have been predicated on “a substrate of assumptions and presuppositions” that these interlocutors share (109). For instance, all parties to the debate, including young female students and the female da‘iyat, agree that women's physical appearance is a “threat to the integrity of the Muslim community” and that men are “more libidinal and sexually charged” than women (111). Further, the values of “feminine chastity and modesty” are accepted by all parties as “divinely ordained” (113). Most Western feminists would find these assumptions problematic and might want to critically interrogate WMM participants’ acceptance of them as “divinely ordained.” Yet WMM participants may refuse to respond to the challenges from outsiders to what they consider to be core presuppositions of their religious/cultural belief system.

It seems that Westlund would support the WMM participants’ stance, as one of her two necessary conditions of “legitimate” challenges to which agents need feel obligated to answer is the condition of “relational situatedness.” Recall that this condition requires a legitimate critic to be in an appropriate relationship—a family member or a neighbor or at least a member of the same moral community—to the challenged in a way that compels the challenged to answer. In the absence of such a relationship, Westlund's position implies, WMM participants may rightly feel that the challenges posed by Western feminists, who are neither immersed in their religion nor sympathetic to their religious values, are “inappropriate” or “outrageous” (Reference WestlundWestlund 2009, 39). In other words, Westerners, including Western feminists, may be viewed as disqualified from being legitimate critics of WMM participants, as they lack the requisite insider knowledge or emotional attachment to the latter's culture and religion. If Western feminists are willing to accept this seemingly counterintuitive conclusion, it may seem that Westlund succeeds in providing a theory of relational autonomy that successfully accommodates WMM participants’ agency.

This judgment, however, would be hasty: Westlund's account of autonomy predicated on one's willingness to engage with others’ critical perspectives (35) is ultimately at odds with WMM participants’ wholehearted commitment to obey God and “secure God's pleasure” (Reference MahmoodMahmood 2005, 123). As Mahmood explains, these women's agency does not belong to themselves, but rather to “authoritative discursive traditions,” the logic and power of which “far exceeds the consciousness of the subjects they enable.” The subject‐centered conception of agency that presupposes “self‐reflexivity” is inapplicable here, as the subject is only “contingently” produced in its conformity to the discursive logic of the authoritative moral tradition (32). Accordingly, WMM participants fully embrace the goal of transforming themselves into pious subjects of God. They strive to embody Islamic virtues, such as fear, al‐haya, and sabr—all of which are “socially prescribed” (148)—by assiduously practicing “disciplinary acts” (126). The aim is to incorporate these seemingly “defeatist and fatalist” virtues (173) into their “natural” disposition (130). Social conventions are eagerly accepted as prerequisites for the realization of the pious self, rather than deemed as external impositions calling for critical scrutiny (149). WMM participants, then, seem to exhibit “self‐abnegating deference,” which involves the “systematic subordination” of oneself to God whose interests, needs, and preferences are regarded as “pre‐emptively decisive” in one's own practical reasoning (Reference WestlundWestlund 2003, 485). Therefore, WMM participants approximate “deeply deferential” persons, such as DW, whom Westlund considers paradigmatically nonautonomous (Reference WestlundWestlund 2009, 32).

One may question the analogy I make between WMM participants and DW, as dedicating oneself to what one considers an omniscient God seems qualitatively different from dedicating oneself to another fallible human. This skepticism may seem to be supported by instances of WMM participants who, contrary to DW, complain about their “disobedient” husbands (Reference MahmoodMahmood 2005, 184). Further, Westlund herself differentiates DW from the “Anti‐Feminist (AF)” who is DW with some religious or secular “creed” to justify her deference; AF, according to Westlund, is relationally autonomous, because AF, unlike DW, is “disposed to enter into justificatory dialogue about her deference” with her critics (Reference WestlundWestlund 2003, 512). This may seem to imply that women who attempt to justify their actions by invoking religion, such as WMM participants, are relationally autonomous. Are WMM participants AFs who are relationally autonomous after all?

It all depends on how extensive is WMM participants’ willingness to subject their religious creed to critical scrutiny. In order to qualify as AF, according to Westlund, DW ought not to “begin and end with the assertion that the Bible instructs her to defer to her husband” and must remain “disposed to take up further challenges.” If she “simply repeats pat responses” (513), then she is AFʹ who “does not really hold herself answerable” and “really does not differ significantly from DW.” Still, AFʹ may seem distinct from DW in that DW “abdicates responsibility for self… to a specific other person,” whereas AFʹ does so in deference to “an ideology or dogma.” Yet Westlund herself acknowledges that the two cases are merely two instantiations of “a more general pathology” that undermines “one's functioning as one's own representative” by thwarting “engagement in justificatory dialogue” (514; original emphasis). Further, Westlund continues, AFʹ is not that different from DW in that in deferring to scripture she may be thought of as “deferring to the will of a personal entity (God) and not just to a dogma impersonally conceived” (523, note 35; original emphasis). Then, similarities between DW and AFʹ are endorsed by Westlund herself.

The closest among Westlund's characters to WMM participants is AFʹ, not AF. By Westlund's own admission, however, AFʹ is on a par with DW, who is nonautonomous. We may therefore conclude that WMM participants are nonautonomous. Although this is a valid argument, I need to show that it is indeed the character of AFʹ that is equivalent to WMM participants in order to make the argument sound. I argue that WMM participants are AFʹs because WMM participants are “tightly gripped” by their devotion to God (Reference WestlundWestlund 2003, 502) and unwilling to engage in “potentially open‐ended self‐evaluative dialogue” (514). WMM participants may be open to engaging in dialogue regarding matters of interpretation of their religious texts with those standing in “sense‐giving” relationships and sharing their religious commitment (Reference WestlundWestlund 2009, 39). Yet if WMM participants were pressed by critics to give reasons for why they always defer to God, they, as pious subjects of God who are constantly “desirous of God” (Reference MahmoodMahmood 2005, 125), may simply persist in referring their interlocutors to the perspective of the one to whom they defer, God (Reference WestlundWestlund 2009, 32). In fact, their unwavering dedication to God seems to imply a commitment not to critically scrutinize their deference to God, whether in response to external critics or in internal self‐reflection (Reference WestlundWestlund 2003, 494).

Take, for instance, WMM participants’ pursuit of automaticityFootnote ¹⁶ in their daily performance of worship. The purpose of striving to embody Islamic virtues—such as fear, al‐haya, and sabr—through diligent and scrupulous self‐cultivation, discipline, and practice, and to transform these virtues into their natural disposition (Reference MahmoodMahmood 2005, 130) is to be able to orient all acts toward securing God's pleasure (126). In other words, they aim to obey God's will by spontaneously expressing the right attitudes on appropriate occasions without any conscious effort or thinking (129–30). This exemplifies WMM participants’ resolute commitment not to engage in critical reflection regarding their obedience to God. Further, the deference to God that WMM participants exhibit is “self‐undermining” in that it requires “deny[ing]” or “effac[ing]” oneself (Reference WestlundWestlund 2003, 486; original emphases). Attaining automaticity by embodying Islamic virtues is not natural and must be created through disciplinary acts. WMM participants’ repeated practice of orienting all acts to secure God's pleasure is intended to create and strengthen a second‐order desire—which does not exist naturally—to enact obedience to God's will, which will contribute to the realization of a pious self (Reference MahmoodMahmood 2005, 126). This disciplinary process requires overcoming and subduing one's first‐order desires and tendencies for pleasure, comfort, or disobedience, and constantly “guarding against disobedience and sins.” The objective is to tenaciously strive only for God “against” oneself and one's desires (125), which requires denying and effacing oneself. WMM participants, then, are Westlund's deeply deferential persons. The assessment that Islamist women can be relationally autonomous, therefore, contradicts Westlund's position that deeply deferential persons are nonautonomous.

IV. Illuminating WMM Participants’ Agency

If my argument above is plausible, then neither Oshana's nor Westlund's theory of relational autonomy succeeds in illuminating the agency of Islamist women. Oshana's and Westlund's theories are not the only relational autonomy theories that were applied to third‐world religious women.Footnote ¹⁷ Their theories, however, showcase two contrasting applications of relational autonomy to the same instance of third‐world religious women, and Westlund's theory in particular is perhaps the most charitable to third‐world religious women's agency as far as autonomy theories go. Therefore, their failure to explicate Islamist women's agency may be taken to indicate the inapplicability of relational autonomy in many third‐world contexts.Footnote ¹⁸ Does their failure, then, imply that the agency of Islamist women, and third‐world religious women more generally, is hopelessly compromised? I reject this suggestion, as it presupposes the notion that the concept of autonomy captures the essence of normative human agency cross‐culturally and the concomitant assessment that those who do not meet the conditions of relational autonomy fail to realize their full potential as human agents. In this concluding section, I propose and provide a brief overview of an alternative conception of normative agency compatible with Mahmood's own analysis of WMM participants. This account may do justice to third‐world religious women's agency, as it is predicated on what I consider to be the cross‐cultural core of normative human agency—the moral capacity—distinct from self‐government presupposed by accounts of personal autonomy.

Mahmood's analysis has been accused of implying “cultural relativism,”Footnote ¹⁹ and Mahmood herself objects to the construal of her 2005 project as offering an “alternative conception of agency to the liberal one” (personal email correspondence, 2014; see also Reference MahmoodMahmood 2005, 188). Yet I believe that it is possible to identify a cross‐cultural core of normative agency operative in WMM participants’ motives and actions that may be crucial for elucidating Islamist women's agency. To do this, a good place to start is with Mahmood's own statement that the agency of WMM participants comprises first and foremost “the capacities and skills required to undertake particular kinds of moral actions” (29; added emphasis). I take this to imply that WMM participants are first and foremost moral agents who have the capacity to make distinctions between right and wrong, good and bad, higher and lower, and noble and base: the moral capacity. The contrastive evaluative language presupposes fundamental moral values and ideals—what Charles Taylor calls “hypergoods”—that are “incomparably more important than others but provide the standpoint from which these must be weighed, judged, decided about” (Reference TaylorTaylor 1989, 63). Subscribing to and living in accordance with certain hypergoods is not only important for our sense of self‐worth, but is essential to our identity (27–28, 44). From this perspective, the core of normative human agency lies in our moral capacity; we are essentially moral beings who make moral evaluations based on deep and powerful moral intuitions concerning the treatment of others, which may be “rooted in instinct” (5).Footnote ²⁰ For WMM participants who wholeheartedly embrace Islamic values as hypergoods, these are not social impositions on the subject but are constitutive of the very substance of their “intimate, valorized interiority” (Reference MahmoodMahmood 2005, 23). Therefore, WMM participants believe that they can become their authentic moral selves only by fully habituating Islamic virtues and practices in their daily lives.

Yet, as Mahmood astutely points out, normative agency that comprises the capacities and skills required to undertake particular kinds of moral actions is inevitably intertwined with “the historically and culturally specific disciplines” that form the subject (29). This conception of agency can be explicated further through Taylor's “dialogical” self (Reference Taylor, Hiley, Bohman and ShustermanTaylor 1991)Footnote ²¹ whose very selfhood is constituted by a shared understanding among those who co‐inhabit a culture (Reference Taylor, Hiley, Bohman and ShustermanTaylor 1991, 311).Footnote ²² This understanding is often “inarticulate” (308) and embodied to the extent that certain pervasive features of my attitude toward the world and others is “encoded in the way I carry myself and project in public space” (309). The body then is not a “medium of signification” but rather “the substance and the necessary tool” for forming the embodied subject. Given the constitutive role that the body plays in the construction of the self, training the body to embody one's norms and standards is not only possible but crucial for human agents who exist “inescapably in a space of ethical questions” (305). Take, for instance, the virtuoso pianist, who is central to Mahmood's analysis. The virtuoso pianist committed to the goal of excellence in piano performance thereby submits herself to the “often painful regime of disciplinary practice as well as to the hierarchical structures of apprenticeship, in order to acquire the ability—the requisite normative agency—to play the instrument with mastery” (Reference MahmoodMahmood 2005, 29). WMM participants are similarly unfaltering in their commitment to becoming God's pious subjects. Consequently, they willingly take on the “often painful regime of disciplinary practice” to embody pious virtues in their daily lives. It is in this context that WMM participants embrace the very processes and conditions that secure their subordination to God as the means by which they become “self‐conscious” moral agents (17).

Theories of autonomy, including relational autonomy, are inadequate in making sense of the conception of normative agency embraced by WMM participants, because they presuppose a conception of the selfFootnote ²³ that governs itself from an “inner” space separate from those external to the self (Reference Taylor, Hiley, Bohman and ShustermanTaylor 1991, 307). As stated at the outset, feminist philosophers of relational autonomy acknowledge the social embeddedness of the self (Reference Mackenzie and StoljarMackenzie and Stoljar 2000, 4) and emphasize “relationality” to recognize the importance of human relations in both the development and the exercise of the capacity for autonomy. Yet their acceptance of self‐government as the core of relational autonomy—whether predicated on individual freedom of choice (Reference Oshana, Paul, Miller and PaulOshana 2003; Reference Oshana2006) or on the individual ability or willingness to defend one's choices against actual or potential critics (Reference WestlundWestlund 2009)—still presupposes a self capable of making choices against a backdrop of culture and social conditions external to the self (see Reference Narayan, Antony and WittNarayan 2002, 424). The concept of relationality embraced by feminist theorists of relational autonomy, then, does not go deep enough to account for the thoroughgoing nature of the self's social embeddedness.

My argument that relational autonomy cannot fully elucidate the agency of Islamist women, however, does not imply that feminist theories of relational autonomy, including those discussed here, are without merit. Autonomy is an enduringly powerful ideal, especially in the Western context since the Enlightenment, and I stand by my earlier statement that feminist theories of relational autonomy have made significant progress in the evolution of autonomy in Western philosophy. What is problematic, however, is the claim that autonomy, whether relational or not, is a universal ideal of human agency that provides the standard by which to judge a person's status as a full agent and a fortiori his or her deservingness of our respect as an equal. The presumption that autonomy as self‐government constitutes the core of normative human agency is of relatively recent provenance and is culturally specific to the modern liberal West. Autonomy therefore may be an ideal, but only for those enculturated in liberal societies who embrace autonomy as their hypergood. Those socialized in other cultures would not necessarily find autonomy an attractive or even familiar ideal. As such, this normative conception of agency cannot illuminate the ideal of human agency across cultures (Reference Taylor, Hiley, Bohman and ShustermanTaylor 1991, 311). By failing to recognize the cultural specificity of autonomy and insisting on its status as a universal ideal, even well‐meaning feminist philosophers of relational autonomy risk alienating those who are our equals in their moral capacity, with whom we must build transnational feminist solidarity necessary for creating a world in which women are empowered according to their own moral frameworks in their particular cultural communities.