Toward an Expressivist View of Women's Autonomy

1. Introduction

To feminist philosophers, the ideal of autonomy presents both a promise and a challenge. On standard views, autonomy consists not in the absence of constraint, but in the ability to reflectively endorse the desires, ends, and values that underlie one’s action as ‘one’s own.’³ Gendered socialization, feminists have long argued, shapes women’s desires to align with patriarchal demands. Not only does this naturalize women’s oppression, but women’s willing participation in sexist practices may be thought to immunize such practices from critique. As the concept of autonomy shines a critical light on an agent’s relationship to her own desires as the site of self-determination or lack thereof, it can articulate how oppression undermines women’s autonomy, even absent external constraints. To reap this concept’s critical promise, however, it is necessary to transform a traditional, individualistic conception of autonomy into a robustly social one.⁴ Feminist philosophers have taken up this philosophical task: reconceptualizing autonomy in a way that captures (1) the socially constituted nature of the self – the way agents are embedded in social contexts and relationships – and (2) how an oppressive social context, in particular, can undermine women’s autonomy.

It is widely agreed that autonomy must be understood in a more social register. But the question of what it means to do so is controversial. The debate often begins with disagreement about so-called ‘hard cases’ of women who embrace gendered subordination. Thomas Hill’s ‘Deferential Wife,’ for instance, happily plays a subservient role in relation to her husband, and “tends not to form her own interests, values, and ideals; and, when she does, counts them as less important than her husband’s” (Hill 1973: 89). Other oft-discussed cases include Marina Oshana’s ‘Taliban Woman’ who chooses a life of extreme dependence out of religious piety (Oshana 2006; Westlund 2009), women who get plastic surgery (Chambers 2008), or who accept gender-based injustices like lower wages without complaint (Nussbaum 2001).

Some interpret these as cases in which oppression has insidiously compromised women’s autonomy, and seek to capture this intuition with substantivist views, on which autonomy is not content-neutral. Strong-substantivists place direct constraints on the content of women’s preferences or relationships (Stoljar 2000: 94–111; Oshana 2006; Hill 1973), while weak-substantivists build ‘content’ into their accounts indirectly: autonomy may require certain normative competencies, or traits like self-respect and imagination (Benson 1990; 1991; 1994; Wolf 1987; MacKenzie 2008).

Proceduralists, however, worry that content-laden views of autonomy raise difficulties: they risk imposing a conception of the good on agents, rule out women who embrace traditional roles as non-autonomous, and can justify coercion or disrespect. They argue that autonomy must be understood as content-neutral (Christman 1990; 2001; 2004; Friedman 1997; 2003; Meyers 1987; 1989), such that autonomy does not depend on the content of a woman’s desires, but on the way in which she comes to identify with them.⁵ Women can, on this approach, autonomously choose gendered subordination.⁶

Questions of social ontology about the nature of the social self underlie this debate. Yet these questions have remained implicit.⁷ This article brings this philosophical terrain to light in order to motivate an expressivist view of the social self as a promising foundation for an account of women’s autonomy. To shift the debate onto new conceptual ground, I focus on a pervasive but less discussed set of cases I term ‘thwarted autonomy,’ in which women value and pursue autonomy, but in ways that reinforce their own gendered subordination. By revealing how an oppressive social context can channel women’s attempts to actualize their autonomy in ways that come to undermine it, these cases highlight the need for a novel conception of the social self to underpin an account of autonomy. Drawing on Hegelian ideas of agency, I argue that the process of self-understanding and self-discovery that often constitutes autonomy’s core requires agents to first express their initially inchoate desires and values in a social context. Social meanings and norms, and the practices in which they inhere, mediate this expressive activity, and the way an agent interprets herself and her desires. This framework allows cases of thwarted autonomy to appear as a result of the limited and problematic vehicles available to women, through which they can express a desire for autonomy. This article motivates and explicates this view of the social self, and show how it illuminates the cases. While a full account of an expressivist view of women’s autonomy must await future elaboration, I describe some of its features in the final section.

The structure of the argument, in brief, is that if our concept of autonomy is to capture (1) the socially constituted nature of the self and (2) the way oppression can undermine women’s autonomy, an expressivist account furnishes an understanding of (1) in order to capture an under-theorized phenomenon that falls into the category of (2).

The article is organized into three sections. In Section 1, I motivate the expressivist view by describing and drawing on two cases of thwarted autonomy to reveal the limitations of proceduralist and substantivist views. As proceduralists rely on a causal model of the social self, they fail to bring into view the way an oppressive culture can undermine autonomy. Substantivist views are equipped to capture this, but their emphasis on internalized norms renders them unable to account for the ambiguity of thwarted autonomy: the fact that women pursue the right value.

Section 2 sketches the expressivist view’s key insight: that agents gain self-understanding through the activity of expressing their initially unformed desires and values in a social context. The expressivist view borrows the proceduralist insight that autonomy is an activity that yields self-understanding, but recasts it as the practical, reflective articulation of desires and values in the world. It also borrows the substantivist insight that ‘content’ matters, but emphasizes the way a social context partially constructs this expressive activity.

In Section 3 I show how this framework accounts for the dynamics of the cases from §1. I also consider what an expressivist view of autonomy entails by engaging with one class of relational views proposed by Westlund (2009) and MacKenzie (2008).⁸ One crucial implication is that a social conception of autonomy must expand beyond the interpersonal realm to include collective hermeneutic resources and the normative structure of social practices.

I. Thwarted Autonomy

To motivate the expressivist view, I begin by considering the phenomenon of what I shall call thwarted autonomy and the challenge it poses to proceduralist and substantivist approaches. Cases of thwarted autonomy are ones in which women value and pursue autonomy, but in ways that reinforce their own subordination. A case in point is that of Miriam Weeks who, as a student at Duke University in 2014, was revealed by a male classmate to have participated in physically abusive, misogynistic pornography. The media narrative seized not only on Weeks’s status as a student at an elite university, but also on the purportedly feminist motivations she elaborated in interviews. As she told one interviewer, participating in the porn industry was not only her choice, but one that made her feel like a “strong, independent woman” (Van 2014). As she explained, “everything is on my terms. I can say no whenever I want to…I am in control.” Even more directly, she stated that “[w]hat porn has done for me is it has given me back my agency” (Van 2014). Yet these claims were hard to square with the misogynistic content of the pornography, in which she was filmed being sexually degraded, verbally abused, and crying while being choked. Feminists were split between those who saw Weeks as a victim of a predatory industry, and those who saw her as a woman shamed for publicly expressing her sexuality.

This case shares some features with the more familiar ‘hard cases.’ Weeks is not obviously coerced into participating in pornography, but does so of her own volition. As she can articulate her reasons for doing so, her decision does not seem entirely unreflective. If autonomy consists in an agent’s reflective identification with her desires and values, Weeks can be autonomous: indeed, she may exemplify the possibility that women can autonomously choose gendered subordination. Yet, as Stoljar has argued, the ease with which standard views can reach this conclusion raises the ‘feminist intuition’ that preferences that have been shaped by oppression cannot be autonomous (Stoljar 2000: 95). This case both raises and complicates this intuition. As Weeks’s choice mirrors oppressive norms that encourage sexual submissiveness in women, the case does, indeed, raise the feminist intuition that her autonomy has been undermined by her social context. But while women in ‘hard cases’ are often described as having a preference for subordination, this description is not quite apt. For in describing her motivations, Weeks cites a cluster of values diametrically opposed to subordination, such as strength, agency, control, and independence – values that, taken together, look a lot like an ideal of autonomy.⁹ While this does not mean there is no sense in which Weeks may be said to prefer subordination, it does indicate the need for a view that can capture the way oppression can compromise women’s autonomy, even as they pursue it as a central value.

A second case will, shortly, furnish a more detailed example of this puzzling dynamic. But first, let us consider how a proceduralist approach fares in capturing the phenomenon. Proceduralist views are content-neutral: they do not specify what an agent must desire to be autonomous, but attend instead to the way in which an agent comes to identify with her desires. There is a diversity of proceduralist views but, generally, they take autonomy to have two conditions. First, agents must engage in a process of critical self-reflection, in a way that is conducive to self-awareness and self-appraisal: an agent may, for instance, have to display minimal rationality, and be free from delusion, paranoia, and self-deception. Second, this process must yield a form of identification with desires and values, whether understood in a minimal sense as a failure to be alienated from them (Christman 2004: 153), or in a stronger sense, as a capacity to affirm them as deeply significant (Friedman 2003: 14). By virtue of the proceduralist commitment to content-neutrality, women who meet these two conditions can, in principle, choose gendered subordination autonomously.

The socio-ontological foundations of the proceduralist approach, I argue, make it ill-equipped to identify the full range of ways oppression undermines autonomy.¹⁰ While traditional proceduralist views can be criticized for relying on a notion of a ‘true self,’ purged of social influence, the charge does not stick against feminist proceduralists. Rather than deny that the self is social, they argue that a proceduralist approach is perfectly compatible with this claim and, by extension, the claim that autonomy has social conditions. Still, this approach’s commitment to content-neutrality places limits on the way it can render this idea intelligible.

It is useful to distinguish here between the claim that social conditions have a causal effect on autonomy, and the claim that they are constitutive of autonomy. On the former, egalitarian relationships – to take just one example – impact the extent to which agents can develop the skills they need for autonomy. Social conditions are “background requirements for the development of autonomy” (Christman 2004: 158), or “causal conditions” (Friedman 2003: 14) for the development of autonomy-enhancing skills like understanding one’s desires, acting on them appropriately, and reflecting on alternatives. On the latter, constitutive version of the claim, egalitarian relationships can define autonomy, such that one could not be autonomous in their absence. Proceduralists largely conceptualize the relation between autonomy and social conditions as causal because a constitutive model risks violating their core commitment to content-neutrality.¹¹ If one makes egalitarian relationships definitive of autonomy, for instance, this automatically renders a life lived in deference to others as non-autonomous, a move that Christman, characteristically, rejects as “dangerous” and “ultimately problematic” (Christman 2004: 158).

Few deny the causal version of the claim that proceduralists grant. Nevertheless, if one is primarily concerned with the causal impact of social conditions on an agent’s abilities, certain threats to autonomy appear more salient than others. The model privileges, in particular, social forces that are external to agents and have a coercive impact on their ability to engage in the kind of critical self-reflective process proceduralists take as central to autonomy. This tacit focus can be seen in the kinds of examples to which proceduralists turn when reflecting on how a social context can undermine autonomy. Imagining a woman whose autonomy has been diminished by oppression, for instance, Christman pictures one who has been “denied all education…systematically punished when expressing curiosity about alternative conditions” and whose “skills have been narrowly fashioned to accept only one role” (Christman 2001: 206). Friedman’s discussion similarly reveals an overriding concern with coercive, external threats: an agent cannot be “overly subjected to coercive pressures or the controlling power of another person” while engaged in self-reflection and, once agents have reflectively affirmed their values, the social world cannot severely constrain their ability to act according to those values (Friedman 2003: 18). While Christman and Friedman consider different scenarios, the most relevant social forces for both— whether punishment, another’s controlling interference, or severe constraints on one’s ability to act—are largely external to agents and possess a coercive dimension.

While the causal model the proceduralist favors can show how a certain class of oppressive social forces diminishes autonomy, it cannot easily articulate how cultural meanings and norms do so. For these furnish agents with a symbolic, normative framework within which they make sense of who they are and what they value, and so appear as more ‘internal’ and less coercive than the kinds of threats on which proceduralists concentrate. It is also hard to account for how they compromise autonomy without referring to their content.¹² Consider Weeks again. Alert only to threats a causal model makes visible, one is hard-pressed to say how her autonomy is compromised: she is not denied an education, subject to coercive control, or facing severely restricted opportunities. And while it is hard to draw firm conclusions about her critical, reflective abilities, it is no glaring deficiency in this sphere that raises the ‘feminist intuition,’ but the way her choice reflects misogynistic ideas that pervade her hyper-sexualized culture. Joel Anderson and Axel Honneth highlight the particular relationship between culture and autonomy when they observe that the “self-interpretive activity central to autonomous reflection” is “fundamentally framed by the semantic and symbolic field in which that reflection occurs” (Anderson & Honneth 2005: 136). They point out, for instance, that it will be hard for a man to value being a stay-at-home-dad if this is, socially, a euphemism for unemployment. Similarly, the case of Weeks primarily raises the question of how her ‘semantic and symbolic field’ has framed what it means to be a woman, powerful, or sexually liberated, and the effect this has had on her self-interpretive activity. Yet this will remain invisible if one is attuned only to causal threats, which take an external, coercive form.

Before turning in §2 to an alternative conception of the social self, I wish to consider the substantivist approach. Many feminist versions of this approach are motivated by a similar concern that attending solely to a contentless reflective process conceals the way oppressive socialization compromises autonomy. Hence, substantivists deny that autonomy is content-neutral, while invoking ‘content’ in different ways: strong-substantivists directly constrain the content of desires or relationships, while weak-substantivists incorporate content indirectly. Most germane to our purposes are views, such as those Stoljar and Benson develop, that seek to remedy proceduralist shortcomings in capturing the threat of an oppressive culture to women’s autonomy. While Stoljar and Benson put forth strong-substantivist and weak-substantivist views, respectively, both focus on cases in which women meet all proceduralist requirements for autonomy, but have internalized false, oppressive norms, whether those that surround sex and pregnancy (Stoljar 2000: 109) or beauty (Benson 1991). As they have internalized these norms, they cannot perceive them as false, and so develop preferences that reflect their oppressive socialization. Stoljar argues that in such cases, a woman’s autonomy is diminished by virtue of the content of the preferences she develops, while Benson argues that it is the way false, internalized norms undermine her ability “to be aware of applicable normative standards, to appreciate those standards, and to bring them competently to bear in one’s evaluations of open courses of action” (Benson 1990: 54), as well as her ability to appreciate the genuine source of her “real strength and value” (Benson 1991: 396). Despite these differences, both agree that the root of women’s diminished autonomy in such cases is that they have internalized oppressive norms they cannot identify as false.¹³

In order to bring out the strengths and limitations of the substantivist approach, I turn to a second case. This is drawn from Susan Bordo’s analysis of anorexia nervosa, an epidemic of self-starvation that emerged among young women in the mid-20th century (Bordo 1993).¹⁴ Bordo situates this phenomenon in a wider cultural context to complicate the assumption that the act of self-starvation is solely a symptom of objective pathology or a desire to emulate thin models. As these interpretations cast women as passive and their activity as lacking deeper significance, they tacitly obscure what Bordo wants to bring to the fore: the sense in which anorexia represents an “unconscious, inchoate” (Bordo 1993: 175) protest against gendered oppression and a striving for autonomy. Bordo observes that in first-personal narratives, anorexic women betray a visceral fear of inhabiting a mature female body, imagined as an “archetypal image of the female: as hungering, voracious, all-needing, and all-wanting” (Bordo 1993: 160). Their narratives also mirror hegemonic ideas about gender, and the nature of the body and mind. One theme that emerges, for instance, is an association between, on the one hand, values of purity, control, transcendence and freedom, and, on the other hand, the mind—both coded as masculine. Against this backdrop, inhabiting a mature female body appears to necessarily entail dispossession of valuable, ‘masculine’ traits, such as dignity, self-control, and freedom. And the achievement of these traits appears to depend on the mind’s ability to triumph over the body – coded feminine.

Self-starvation can appear to be irrational when analyzed at an individual level. But Bordo reveals its intelligibility as a way in which young women try to “find honor on the ruinous terms of…[their]…culture” (Bordo 1993: 65). By denying bodily desires, and eradicating signs of female embodiment such as breasts, menstruation and fat, young women aim to embody “an ethics and aesthetic of self-mastery and self-transcendence, expertise, and power over others through the examples of superior will and control” (Bordo 1993: 178). By crafting their bodies, women seek to publicly demonstrate their possession of esteemed ‘masculine’ traits, and to dissociate from the negative traits that attach to female embodiment. In this case, too, women value and pursue autonomy: they want to define themselves in opposition to oppressive, alienating ideas about who they are. Yet the way in which they do so is self-undermining and, in certain respects, reinforces patriarchal ideas about women. In “disturbingly concrete, hyperbolic terms” (Bordo 1993: 169), anorexia colludes with traditional gender norms: women deny their own desires, remain physically fragile, and take up little room in the public sphere. And although young women may feel “deeply attracted to the aura of freedom and independence suggested by the boyish body ideal” (Bordo 1993: 160), their pursuit of this ideal leads to serious physiological problems, and the expenditure of time, energy and attention on a narrow aim that impedes the development of their individual talents. Paradoxically, the act of self-starvation can be seen to re-inscribe the very normative ideal of femininity that, on another level, it aims to counter.

A substantivist approach is well-equipped to articulate what goes wrong in this case: women perceive and appraise themselves in light of oppressive gender norms such that, even with reflective abilities intact, the ends they formulate bear the imprint of such norms. Yet the substantivist approach also risks obscuring the ambiguity that cases of thwarted autonomy exhibit—and so what ‘goes right.’ For inquiry into the content of the women’s desires and values yields an ambiguous result: not only do they latch onto and pursue the right value of autonomy, but they also display a deep ambivalence about the oppressive meanings and norms that imbue their culture. It seems, then, one-sided to say that they have internalized them and so cannot perceive them as false. If the women were entirely uncritical of the idea that freedom and independence are male traits, for instance, it would be hard to make sense of how their activity could also be a nascent rebellion against norms that cast women as unable to achieve such traits. Second, if one imagines that the women have internalized oppressive norms, one risks inadvertently reinforcing the very image Bordo wants to avoid, of women as merely passive repositories of the oppressive meanings and norms that circulate in their culture. Indeed, on the basis of such an image, it is hard to see how women could gain critical purchase on oppressive norms or how, in attempting to fashion an ‘honorable’ self out of a ‘ruinous’ culture, they could recognize and pursue the value of autonomy itself.

Capturing the case’s ambiguity and critical edge requires one to reject the assumption that women conform to patriarchal practices because they have internalized, and so uncritically accept, sexist norms. To bring this into view, one must attend to women’s practical, embodied expressions of opposition to oppressive meanings and norms that, in fact, characterize their social world. For while it is tempting to see the women Bordo discusses as unable to appreciate normative standards in the right way, the culture in which they live really does connect masculinity to the mind and vaunted traits, and femininity, to the body and degraded traits. Rather than see the women as normatively incompetent, one can see their activity as reflective of an acute awareness of prevailing normative standards in their social environment and, on that basis, reasons they have for acting as they do.¹⁵ Uma Narayan argues, along similar lines, that women’s conformity to sexist practices need not reflect their acceptance of those practices, but their attempts to pursue what they want in light of realistic judgments they make about their options.¹⁶ This activity of ‘bargaining with patriarchy’ lies in between a “dupes of patriarchy” model (Narayan 2002: 425), on which women uncritically accept oppressive norms, and a “prisoners of patriarchy” model (Narayan 2002: 419), on which women are coerced into compliance. Similarly, women’s outward conformity to patriarchal norms in cases of thwarted autonomy need not indicate a deficiency in their normative competency. Rather, one can interpret these cases as ones in which women try to get what they want—autonomy—in light of accurate judgments they make about the nature of their social world, and the normative standards at play within it.¹⁷

This section’s aim has been twofold. First, I introduced a novel set of cases, in which women pursue autonomy, but in ways that reinforce patriarchal norms. Second, I drew out underlying socio-ontological assumptions of two dominant approaches and considered the extent to which they capture the cases. The upshot is that a positive view must accommodate both (1) the way an oppressive culture furnishes agents with with a symbolic, normative frame through which they engage in self-interpretive activity, as well as (2) the sense in which agents can engage in meaningful, practical opposition to oppressive norms, even if not evident from outward behavior. An expressivist conception of the social self, I shall now show, meets both desiderata.

II. Expressive Activity

Autonomy involves self-understanding: the ability to identify with or feel alienated from desires, ends, and values requires some sense, even if not entirely transparent, of who one is. An expressivist view, which intervenes at the level of social ontology to inquire into the necessary conditions for self-understanding, argues that it is only by expressing desires and values in a social medium that we come to discover who we are. This view enables one to bring the wider sphere of ‘expressive activity’ into view when assessing an agent’s autonomy. In this section, I briefly sketch this idea before applying the framework to the cases of thwarted autonomy in §3.

The expressivist view begins from a broadly Hegelian insight into the nature of self-understanding: that agents are not immediately transparent to themselves, but must engage in a certain kind of social activity to achieve this state.¹⁸ As Charles Taylor glosses this idea, self-understanding is not immediately given but an achievement: it is, he writes, the “fruit of an activity of formulating how things are with us, what we desire, think, feel, and so on” (Taylor 1985: 85). Our ideas, emotions, and desires are initially inchoate and elude our full grasp. Reflection alone, however, does not yield clarity. To be able to identify with or feel alienated from desires and values we must outwardly articulate our previously inchoate sense of what these desires and values are within the social world itself. A paradigmatic example of this expressive activity is that of writing an essay or a journal entry. One often begins with a hazy sense of what one thinks or feels. As the idea of ‘writing to find out what you think’ captures, however, it is by engaging in the activity itself – with its demand that one find the right words – that one develops a more lucid and concrete sense of one’s thoughts and emotions. It is through this activity of articulation that one arrives at self-understanding.

Expressive activity can aid self-understanding in a passive sense, as when one’s action contradicts one’s self-image. I may fervently avow commitment to a political cause but, if I never attend a meeting and spend all my time playing music instead, I may conclude that, contrary to what I wish to believe about myself, I in fact value music more than politics. I can also, however, actively strive to articulate my desires and values in the world, and gradually refine this articulation as my sense of myself emerges over time. As Taylor notes, at first expressive activity is “relatively unreflecting” (Taylor 1985: 91). But through the process of reflecting on and realizing deficiencies in the form my activity takes, I may seek out more refined modes of expression. I might initially express the value I place on justice by posting on social media, for instance, but come to feel that this mode of expression is too superficial, detached from the world, or egocentric. Realizing that its form is not adequate to what I want to express, I may instead join an in-person activist group, as a better way in which to embody this value. This activity, of expressing, reflecting on, and revising the public articulation of our desires and values, yields self-understanding.

What matters for our purposes is the image of the social self that corresponds to this picture. First, as expressive activity requires a social medium, the social world will partially shape the form one’s outward expression takes. The way one expresses anger in a private journal, for instance, will differ from how one does so on an online platform set up to incentivize attention-grabbing and polemical forms of expression. More broadly, the intelligibility of my activity, and the form it takes, will depend on social meanings, norms, and others’ interpretations. In order for an act of rebellion to be intelligible as such, for instance, it must be visibly at odds with status quo values, behaviors, and modes of appearance that are themselves culturally and historically specific. Silicon Valley’s casual attire can only be read as reflective of an ethos that valorizes ‘disruption’ and rejects ‘business-as-usual,’ not just an arbitrary fashion choice, against the backdrop of a status quo characterized by staid suits; it could not have this same meaning in 15th-century France. The way others interpret my expressive activity also matters to what it can signify. If what I take to be an act of rebellion is widely interpreted as a shallow attempt to stand out, for instance, there is a sense in which I have failed to express what I wanted to; this in turn may lead me to reexamine and refine my activity. If this points to the way in which the social world constructs expressive activity, highlighting the sense in which individuals are not “self-sufficient bearers” (Anderson 2010: 18) of social meaning, it does so only partially. For agents also play an active role in expressing their desires and values in a way that is not merely determined by the social world. The particular desires and values I express, and how I choose to do so will reflect my individuality; as Anderson observes, an agent’s “character, history, mood, energy, actions, and reading of her predicament play a profound part in influencing what she values, especially in influencing which practices, roles, and relationships she will make her own” (Anderson 2010: 25).

The outward form of one’s expressive activity is partially constituted by one’s social context. But insofar as this activity furnishes us with self-understanding, it makes little sense to imagine that one can achieve such self-understanding by standing wholly apart from one’s social context, or that this context has no effect on who one becomes. In particular, the practices in which one engages to express some desire or value provide a ‘semantic and symbolic’ frame with which to interpret one’s desires, emotions, relationships, and so forth. Through this activity, then, what are initially inchoate desires and values come to be imbued with socially mediated content. To express my rejection of mainstream values in a way intelligible in terms of my social context, for instance, I may gravitate towards a subculture whose practices represent the most visible rejection of such values. Engaging in the subculture’s practices, however, will leave neither my initial desire nor my conception of myself unchanged. Perhaps I come to strongly identify as a member of the subculture, and cannot imagine who I would be apart from it, or view outsiders to the group as shallow or morally corrupt. And my initially inchoate desire to reject mainstream values may acquire a specific content: to wear all-black, for instance, follow a vegan diet, or become an adherent of an extreme political ideology.

With this brief sketch of the expressivist view of the social self, we can see one way of conceptualizing the relation between agents and the social world. Although agents are not determined by their social context, the expressive activity through which they arrive at self-understanding is socially mediated. Socially available meanings and norms therefore shape not only the outward form their expressive activity takes, but also their conception of themselves and what they desire. I turn now to show how this expressivist view of the social self can illuminate the cases of thwarted autonomy from §1, and what it reveals about a social conception of autonomy.

III. Toward an Expressivist View of Women’s Autonomy

This view invites inquiry into the wider sphere of expressive activity. What is an agent trying to express? In what ways do cultural meanings and norms shape the form her expression takes? How does her expressive activity structure the relationship she has to herself? In this section, I consider what asking these questions of the cases from §1 can reveal about what it means to conceptualize autonomy in a social register. In these cases, women do express a desire for autonomy: their activity reflects a desire for self-definition and direction over their own lives.¹⁹ But initially this desire is inchoate. It may manifest only as a pressing discomfort with how others perceive them, or frustration at the way this restricts their capacity for self-definition. Reaching, at first relatively unreflectively, for a way to express this as-yet unformed desire in the world, the women encounter a set of meanings and norms that structure its intelligible expression. Following Marilyn Frye’s (1983) definition of oppression, these meanings and norms are oppressive: they set up double-binds that penalize women for any choice they make. The result is that women face particular challenges in trying to express a desire for autonomy.

Artistic, cultural, and everyday practices often portray women as sexual objects who exist for the pleasure of others, and dispossessed of traits like independence, powerfulness, and self-assertion.²⁰ When women do exhibit such traits, they often face explicit or implicit forms of punishment, or these traits are conflated with sexual objectification as when female celebrities claim to be empowered by virtue of their conformity to highly sexualized roles. This puts women in a double-bind. If they accept the status of sexual object, they can avoid social censure, but they become complicit in their own self-effacement. But if they reject such complicity in favor of a more robust autonomy, they expose themselves to stigma, critical scrutiny or, at best, a tenuous recognition. Part of what perpetuates this double-bind is the way that female embodiment is freighted with social meanings—of being passive, object-like, or ungovernable—that put it in tension with autonomy. The prevalence of such meanings does not determine the relation a woman will have to her body, or block her expression of autonomy entirely. It does, however, make it challenging for women to escape or re-signify such meanings, or to express a desire for autonomy in a straightforward way.

Cases of thwarted autonomy can be understood as attempts to express a desire for autonomy in a social context that problematizes this act of expression for women. As this context casts the female body as at odds with autonomy, it is unsurprising that the women cultivate a relationship to their own bodies that can, in a way that is intelligible within their social context, signify their opposition to oppressive meanings that would otherwise be simply imposed on them. Instead of mounting a collective challenge to these meanings by rejecting broader assumptions about what it means to exist in a female body, women try to express their desire for autonomy as individuals within this symbolic and normative framework. By drastically shrinking her body, the anorexic woman defines herself in opposition to negative meanings that would otherwise attach to it. If the female body is cast as an obstacle to autonomy, she rejects and reshapes her body to publicly signal her achievement of this value. Weeks does not reject her body so much as try to actively appropriate the meanings that attach to female embodiment. By actively positioning herself in relation to sexually objectifying meanings, she can feel she has chosen them for herself – bearing out Simone de Beauvoir’s perceptive analysis of the ‘woman in love’ who tries to “overcome her situation as inessential object by radically assuming it” (De Beauvoir 2011: 684).

On the one hand, seeing these cases through an expressivist lens allows us to heed Khader’s recommendation to look for “signs of agency in oppressed people – even when that agency is not readily visible” (Khader 2012: 313). By foregrounding the question of what women are trying to express, and how cultural meanings and norms shape the outward form this activity takes, this framework enables a reinterpretation of women’s apparent acquiescence to gendered subordination. Women need not be mere ‘dupes of patriarchy,’ in thrall to images of beauty or blindly acquiescent to their own sexual objectification, but can appear instead as seeking a way to express their desire for autonomy in a world that gives them few good options for doing so. This can be understood as a version of ‘bargaining with patriarchy,’ albeit transposed onto the hermeneutic sphere. Moreover, as the view focuses on the whole sphere of expressive activity, it also permits consideration of how the women are engaging in such activity. It allows us to understand autonomy as graded: not as something one can possess more or less of, but as an activity one can do more or less well. Generally, for instance, shying away from expressive activity altogether can lead one to develop only an attenuated sense of self. But doing so in an overly confident, unreflective manner may, equally, mean that one fails to reap the rewards of self-understanding such activity can provide. Because the women in cases of thwarted autonomy are making an active attempt to express a desire for autonomy in the social world, we can recognize them as engaged in the kind of activity that could, in principle, form the basis of an autonomous life.

On the other hand, the expressivist view can also articulate the way in which oppression thwarts women’s autonomy. It allows us, that is, to recognize positive ‘signs of agency,’ without abandoning the task that feminist social critique presents: that of pointing out the ways an oppressive social world undermines women’s autonomy. The view takes the expressive activity and the self to be socially mediated, and so can reveal how oppressive social forces will distort the attempt to express autonomy. While my discussion of how it does so is necessarily truncated, I wish to proceed by considering how an expressivist view will both borrow and depart from a class of relational views developed by, among others, Andrea Westlund (2003; 2009) and Catriona MacKenzie (2008). The distinctive claim of these views is that one’s relationship to oneself, and one’s ability to be autonomous, is constituted in and through intersubjective relationships to others—an intersubjective realm that, as Westlund notes, does not “map onto any straightforward distinction between what is internal and what is external to the agent” (Westlund 2009: 33). The expressivist view makes a similar claim: that insofar as social meanings and norms structure the outward form of expressive activity, and the self-understanding one develops through such activity, the social world mediates and constructs the relationship one has to oneself – the site of autonomy. Westlund and MacKenzie’s insights into the way oppression can undermine autonomy leads us to see how the scope of the social world must be expanded.

Westlund argues that autonomy is a critical, self-reflexive stance towards one’s own commitments, constituted through dialogue with others. It depends, in particular, on the extent to which one is ‘answerable’ for one’s commitments in the face of others’ critical challenges.²¹ Westlund’s proposal sheds light on what expressive activity must entail if it is to lead to autonomy. In trying to make sense of my own, initially incoherent or self-defeating activity, for instance, I look to the interpretations and challenges of others (Taylor 1985: 89). How I engage with the reactions of others will matter. If others tell me, for instance, that the way I express the value of justice is alienating or merely reflects my narcissism, I may ultimately reject these interpretations. But being receptive to, and willing to engage with them, indicates that I do have the right kind of self-reflexive relationship to my own commitments. It also affects the extent to which I can revise my activity to better reflect what I am trying to express. There may be aspects of my own activity to which I am blind, but that others recognize, and so being open to, and willing to engage with the interpretations of others will help me amend my own activity.

Focusing solely on the interpersonal realm, however, can obscure systemic forms of oppression that do not take a dyadic, relational form. Hence, the scope of the social world must be expanded beyond the dialogical realm to include dimensions not easily characterized in relational terms. The social world may, for instance, lack hermeneutical resources such that a willingness to engage in justificatory dialogue is insufficient for an increase in autonomy. The dominant interpretation of homosexuality used to be that it was a psychological disorder in need of treatment, for instance, which would have appeared in the form of a critical challenge, from analyst to patient, along with a narrative that the latter’s activity should be revised—namely, towards heterosexuality. The interaction can be described as having an ideological dimension: the analyst’s interpretation serves to naturalize unjust social arrangements. In this instance, a receptivity to the critical challenges of others risks mystifying the nature of an agent’s activity to herself, leading her to attempt to revise it in ways which are not genuinely emancipatory.²² Similarly, women’s reactions to misogyny have often been construed as individual, psychological defects, disconnected from socio-political conditions. If a woman gets the message that her activity is rooted in an individual flaw, she may ‘work on herself.’ But if it is rooted in oppressive social conditions, she will have difficulty fully understanding her own activity or revising in liberating ways. The point is not that by engaging in ideologically laden forms of justificatory dialogue, an agent’s autonomy is necessarily diminished. Rather, it is that wider social forces structure the dialogical realm, which must also be considered in reflecting on how oppression undermines autonomy.

MacKenzie, relatedly, argues that autonomy consists in a normative authority over one’s life, grounded not just in answerability to others, but in attitudes of self-esteem, self-respect, and self-trust, as well as in processes of self-interpretation. It is within intersubjective relationships of recognition that we figure out who we are, what matters to us, and come to see ourselves “as beings whose needs, beliefs, and abilities are worth being realized” (Honneth 2012: 41). Hence, autonomy is sustained through relationships of recognition, and may be undermined by the disrespect of others or violations of one’s physical integrity.²³ Again, this captures one sense in which autonomy is thwarted in the cases from §1. For what raises the ‘feminist intuition’ is not just the way misogynistic pornography and anorexia mirror patriarchal ideas about women, but the fact that both involve forms of self-abnegation, physical harm, and, in Weeks’s case, degradation. One may think that expressing a desire for autonomy through such practices will, over time, erode the affective attitudes women need in order to see themselves as “beings whose needs, beliefs, and abilities are worth being realized” (Honneth 2012: 41). However, while MacKenzie connects recognitive relationships to social practices and institutions, the concept of recognition can still incline our thought towards the interpersonal. I wish to conclude, therefore, by pointing to two reasons the normative structure of practices themselves must be equally at the fore.²⁴

A brief, stipulative definition of a practice: an activity governed by constitutive rules that must be followed if the activity is to be performed correctly, out of which emerge social meanings, norms, roles, and relationships. By emphasizing the structure of practices, one can, first, capture cases in which failures of recognition in the interpersonal realm are not immediately apparent. While anorexia does involve forms of self-abnegation and physical harm, for instance, these do not inhere in intersubjective relationships; this is, rather, an activity individuals carry out alone. These features show up more clearly when we see this phenomenon as a practice. For instance, Bordo’s observation that young women strive for ethical and spiritual qualities, not just aesthetic ones, by crafting their bodies in a certain way is borne out by online forums dedicated to ‘pro-anorexia’ discourse. A central theme is the pursuit of ‘perfection,’ a state that is at once physical and imbued with quasi-spiritual properties, and the worthlessness of those who fail to achieve it. This is a pursuit whose goal is never reached, such that participating in it entails viewing oneself in a lowly light, forever falling short of an ideal. In conjunction with the denial of basic bodily needs the practice demands, participating in it may erode the affective attitudes that are the necessary ingredients of autonomy. This erosion may not appear, however, if one looks only for failures of recognition in the interpersonal sphere; hence, one must look also at the structure of the activity itself.

The second reason to foreground the structure of practices is that it helps orient critique in the right way. MacKenzie argues that when social institutions entail relationships of mutual recognition that sustain autonomy, they meet the demands of justice. When the rules of practices license failures of recognition, however, these must also be brought into view as a root cause of diminished autonomy. In the case of Weeks, for instance, there is a clear failure of recognition in the interpersonal sphere. But if one just targets these relationships of disrespect without also looking at their enabling conditions in the structure of practices, one’s social critique will be insufficient. Feminists have argued, for example, that the constitutive rules of pornography as a practice eroticize a submissive feminine role and dominant masculine role, and prevent women’s refusal of sex from having the right force (MacKinnon 1987; Langton 1993). Hence, it is the normative structure of the practice that establishes the specific dynamic in which participating as the female partner involves adopting a sexualized, submissive role and being dominated by others. To do otherwise would be to perform the practice ‘wrong,’ or be engaged in a different practice altogether. In this case, then, it is the structure of a practice that licenses failures of recognition in the interpersonal sphere, and the transgression of physical boundaries. Hence, this structure, which shapes what happens at the interpersonal level, must be an object of social critique.

Conclusion

Autonomy must be understood in a social register if the concept is to be useful in diagnosing and redressing patriarchal oppression. Little consensus exists about the way the social world must figure in autonomy, however. In this article, I introduced the under-theorized phenomenon of thwarted autonomy to clarify the socio-ontological issues at stake in this debate and to motivate an expressivist view of the social self as a foundation for an account of autonomy. On this view, we develop self-understanding by practically embodying our inchoate desires and values in the social world. Hence, the social world shapes the form one’s expression takes and the way one understands oneself. This framework allows us to reinterpret outward conformity to gendered subordination as a way in which women try to actualize autonomy, in a social context that gives them deficient vehicles through which to do so. It can also articulate how an oppressive social world undermines women’s autonomy. Without the right hermeneutical resources, women may struggle to make sense of and revise their activity in genuinely liberating ways. Or expressing their desire for autonomy through practices that encourage a submissive or abject self-conception may erode the affective attitudes they need to exercise autonomy. While a full account of an expressivist view of autonomy must await future elaboration, I have highlighted the way it accommodates the agency women do exercise under oppression, without losing sight of the way oppression diminishes women’s autonomy – or the wider sphere of what matters in making sense of oneself.