Book Search:  


Google full text of our books:


The Subject of Liberty:
Toward a Feminist Theory of Freedom
Nancy J. Hirschmann

Book Description | Reviews | Table of Contents

COPYRIGHT NOTICE: Published by Princeton University Press and copyrighted, © 2002, by Princeton University Press. All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher, except for reading and browsing via the World Wide Web. Users are not permitted to mount this file on any network servers. Follow links for Class Use and other Permissions. For more information, send e-mail to

This file is also available in Adobe Acrobat PDF format

Chapter 1


A TWENTY-THREE-YEAR OLD, unemployed single mother in West Virginia became pregnant as a result of date rape. Due to recent cuts in federal funding, "Meg" had trouble locating an abortion clinic but finally found one four hours away in Charleston. She was told that she was 17 weeks pregnant, and that the clinic performed abortions only up to 16 weeks, so she was referred to a clinic in Cincinnati which would perform an abortion up to 19 1/2 weeks for a cost of $850. When she went there 1 1/2 weeks later, however, she was told that she was actually 21 weeks pregnant and referred to a clinic in Dayton that would perform the abortion for $1,675. She refinanced her car, sold her VCR, borrowed money, and went to Dayton. There she learned that she was a high-risk patient because of an earlier cesarean delivery and would therefore have to go to Wichita, where the procedure would cost $2,500. "I just didn't think I could manage that," she says. "Now that I know I have to have it, I'm trying to get used to the idea . . . I'm not thinking about adoption, because I've never understood people having a baby just to give it away. So I've been thinking a lot about trying to love this baby the way I love my daughter."1 Can we say that this woman has freely chosen her role as mother?

Susan is beaten by her husband and is admitted to a hospital. This is the second time she has been severely beaten this year. An advocate from the local battered women's shelter visits her and gives her information about the resources they have, as well as information about pressing charges and prosecuting her husband for assault. Susan is angry at her husband, and very frightened of him, but she is reluctant to press charges with the police because she doesn't want to put her children through such an ordeal. Her husband, meanwhile, sends her a dozen bouquets of lovely flowers and comes to the hospital with lavish gifts and profuse apologies, declaring his love and promising never to hit her again. The battered-women's advocate tells her that she should not believe him, that he will do it again, but Susan decides to return to him. She says that he has apologized and that she forgives him; that he is basically a good person and has promised to change; that he loves her and is a good father; that she loves him; that it was partly her fault anyway. Is she free if she returns to her husband?

In the novel Mrs. Bridge, the title character is a woman who seems to subordinate herself completely to her husband. She defers to him, is extremely self-deprecating, rarely ventures a political opinion, and has so effectively effaced herself that at the end of the novel, she risks freezing to death while trapped in her car because she won't yell for help; the implication is that she doesn't want to disturb anyone and simply waits passively, until her husband gets home to rescue her. She doesn't appear to be holding these views out of fear or coercion. Her husband is somewhat overbearing but not violent; he doesn't overtly seek to control her, he clearly loves her, and demonstrates his consideration and respect for her in various ways. Is Mrs. Bridge free or not?2

When Greta got married, her husband and she agreed that she would quit her job as a secretary so that they could raise a family. Greta's mother always worked when she was a child, and Greta has always believed it is better for children to have a full-time mother. She had three kids over the course of the next seven years. Shortly after the birth of the third child, her husband left her. It turns out he was having an affair, but he told Greta that he felt smothered by the routine of their domestic life and all the kids. Although she eventually was able to file for divorce, she was never able to collect child support or win any other financial settlement from her husband, and so she was forced to find a job. But her former secretarial skills were sorely out of date, so the only jobs she could obtain were barely above minimum wage. After paying for child care and transportation, she has less money than she would receive on welfare. Greta grew up in a family that always scorned those on public assistance, however; and her mother, who is horrified by the prospect of her daughter's becoming "one of those welfare mothers," argues that Greta brought this on herself by quitting her job when she got married. She will, after all, receive a raise every six months if she continues to work. Greta knows that economic security is years away at this rate, and she feels trapped by her situation. Is she trapped, or is her mother right that freedom requires taking responsibility for one's choices?

Charlene, a lesbian, is an attorney with an extremely conservative Wall Street firm that has never had a woman partner. Charlene wants very badly to become a partner. Accordingly, she is not open about her sexuality. Her lover, Sally, believes this is a mistake, not only tactically, but from the perspective of personal cost as well. Though Charlene declares that their relationship is more important to her than anything, she has become so fearful about colleagues finding out about it that she and Sally have virtually stopped going out of the house together; and the stress is affecting not only Charlene's health but the relationship as well. Sally is beginning to contemplate "outing" Charlene. She believes that this would liberate Charlene from her fears, her anxiety, her extra stress, and save the relationship. Is she right?3

These scenarios are not particularly special or unusual; they are common examples of the everyday dilemmas many women (and some men) face. And precisely because of their familiarity, many of us probably have immediate, perhaps even gut-level, reactions to them. For instance, most people would probably say--at least initially--that the pregnant woman is unfree4 and that Sally is wrong. The other stories might give us more pause. It is difficult, for instance, to imagine that a victim of domestic violence really knows what she is doing when she "chooses" to return to her abuser; Greta could not know that her husband would abandon her; and Mrs. Bridge might seem rather hopelessly repressed to many but quite normal to others. I want to suggest, however, that there is really no simple answer to the question of freedom in any of these cases.

This complexity is partly due to the amazing ambiguity of the term "freedom" in its popular usage, not to mention the vast disagreements among political philosophers over the meaning of the term, as well as over individual instances of "freedom" and "unfreedom." But it is also partly due to the fact that the dominant discourse of freedom in philosophy and political theory--which founds as well as reflects popular, everyday conceptions--is inadequate to fully encompass this complexity. Moreover, it is a central contention of this book that feminism--which many might assume would maintain that the women in all four stories are unfree--highlights both this complexity and this inadequacy.


Implicit in these introductory questions is the more fundamental issue of what the term freedom means. This is a central bone of contention among liberty theorists;5 but most, if not all, conceptions of liberty have at their heart the ability of the self to make choices and act on them. The contested terrain, therefore, generally covers differences about what constitutes the process and activity of choosing and what constitutes the product, or an "actual" choice. Theorists disagree also on what constitutes a "restraint on" or "barrier to" choice, what prevents certain options from being made available, or what prevents me from taking a particular choice that is normally available. Many of these questions--and indeed, much freedom theory--are arguably "semantic" rather than normative; that is, concerned with distinguishing freedom from other terms, such as equality, justice, and obligation, or with the features that constitute a "restraint" (as opposed to an "inability"). Subsidiary to and implicit in these debates, however, though often not addressed, is the more normative and political question of what or who the "self" is that makes these choices; in other words, what constitutes the "choosing subject" of liberty.

While definitions and conceptions of freedom can be quite varied, ranging from neo-Hobbesian descriptivist accounts of behavior to the most value-laden prescriptive accounts of actions,6 most formulations still divide along the lines offered by Isaiah Berlin in his famous 1958 essay "Two Concepts of Liberty," of "negative" and "positive" liberty.7 According to Berlin, negative liberty consists in an absence of external constraints. The individual is free to the extent that she is not restrained by external forces, primarily viewed as law, physical force, and other overt coercion. So, for instance, if I wanted to leave the house but my husband broke my leg to prevent me, he would be restricting my freedom. "By being free in this sense I mean not being interfered with by others. The wider the area of non-interference, the wider my freedom."8 Berlin's general notion that restraints come from outside the self, they are alien to the self, or "other," is an important basic feature of negative liberty; specifically, other humans' direct (or, in some cases, indirect) participation "in frustrating my wishes" is the relevant criterion in determining restraint.9 pAnd these "wishes," or desires, preferences, interests, and needs which I must be able to pursue unimpeded if I am to be free, are seen as coming from me and from me alone. Desires in negative liberty do not necessarily need to be "brute," that is, immediate, physical, and compelling, as some theorists have maintained.10 Although brute desire is an important concept in negative liberty, desire can just as easily be seen as long term, well thought out, and rational. The point for negative liberty, however, is that whether the desires are long term or immediate, "brute" or "rational," what matters is that they are desires that the agent has formulated by herself. That is, a desire may be formed in reaction to external stimuli--I may want to leave the house because I can no longer stand listening to the televised football game my husband is watching in the next room--but this desire is mine, say the negative libertarians, and I am responsible for acting on or resisting it.

Similarly, the desire in question must be conscious: I must know that I have it. Certainly, desires may be responses to unconscious feelings--perhaps my aversion to televised football stems from repressed childhood memories of fear of my father, who would yell angrily at the television as his team lost yet again--but the relevant point for negative freedom is that I want it, and that I know I want it, not why I want it. Thus, negative liberty draws clear-cut lines between inner and outer, subject and object, self and other. This kind of freedom, as Taylor puts it, is "toughminded," because of the strict notions of individual responsibility and accountability that it finds conceptually necessary to "choice."11 It is also "tough-minded," however, in the way it starkly differentiates between freedom and various other political concepts, such as equality and justice. As Berlin says, "Everything is what it is: liberty is liberty, not equality or fairness or justice or culture, or human happiness or a quiet conscience."12 Thus, negative liberty defines itself in opposition to concepts such as obligation and authority; these things, while perhaps necessary to human society, or even to individuals' pursuit of their desires and possibly even to greater freedom in the future, are nonetheless limitations on freedom. As John Rawls argued, while equality, or wealth, or other factors may affect the "worth" of an individual's liberty by enhancing or inhibiting her ability to pursue opportunities, these factors are distinct and separate from the liberty itself, which is measured by the absence of external restraints, such as laws.13 The central question for negative liberty, according to Berlin, is "What is the area within which the subject--a person or group of persons--is or should be left to do or be what he is able to do or be, without interference by other persons?"14 For this reason, Taylor calls negative liberty an "opportunity concept": the significant factor in determining whether I am free is that no other person or thing is actually preventing me from doing what I want, nothing or no one is barring me from taking advantage of opportunities that I could otherwise pursue but for this restraint.15 Berlin similarly says that "Freedom is the opportunity to act, not action itself."16 What is important is that I be allowed to make choices, rather than that I make a particular choice. Freedom is "constituted by the absence of obstacles to the exercise of choice."17

Thus, Berlin holds that freedom is determined by "the number of doors open to me"; the more doors that are open, regardless of whether I go through any of them, or even want to go through any of them, the freer I am. Berlin does concede that "the extent to which [doors] are open," as well as the relative importance of these various doors and paths, are relevant to freedom; but given the necessarily subjective dimensions of such evaluations, "the number of doors" is ultimately determinative. Freedom is thus in an important sense quantitative on the negative-liberty model, even quantifiably measurable.18 More importantly, freedom operates from an "objective" rather than subjective notion of choice. Recognizing the "adaptive preference" phenomenon--that when faced with a limited range of options, I can increase my freedom by simply accommodating my desires to availability--Berlin insists that freedom requires "a range of objectively open possibilities, whether these are desired or not."19 Freedom "consists in the absence of obstacles not merely to my actual, but to my potential choices . . . it is the actual doors that are open that determine the extent of someone's freedom, and not his own preferences."20 The presence of options themselves, objectively defined, is key to freedom; I may not want many, or even any, of the alternatives available, but I am nevertheless freer than if there was only one option. If choice is paramount in the definition of freedom, then the more choices I have, the freer I am.

Defining freedom in the objective terms of available options rather than the subjective expression of desire is supported by many contemporary theorists, such as Stanley Benn, W. L. Weinstein, Joel Feinberg and Christine Swanton,21 because it allows us to circumvent the fact that subjective desire is almost always contingent on social circumstances. For instance, one could not claim that an African slave was free simply because she said she did not want to leave the plantation, because this desire could be seen as the final effects of colonization. But at the same time, this "available options" conception is somewhat counterintuitive: if there are only two options, one of which is the one I want, I would seem to be less free than if what I want is not available at all amidst dozens of other options. Accordingly, other negative-liberty theorists, such as Richard Flathman, Robert Nozick, and William Parent, reject this account and insist on the relevance of the more traditional, Hobbesian notion of being able to make the choice I want.22 But even for the latter theorists, such ability presupposes coherence between subjective desire and objective circumstance--I cannot have what I want if it is unavailable--so they favor the availability of more rather than fewer options.

Such arguments may appear to be little more than semantic hairsplitting (as indeed, they often are), but it is this need for preserving opportunities that delineates negative liberty in a political context. Berlin and almost all negative libertarians recognize the need for laws if liberty is to be preserved; absolute negative liberty would result in chaos, thus jeopardizing not only other things of importance to humans--security, justice, fairness--but freedom as well, in that I must spend so much time defending what I have that I am prevented from seeking other things that I want. Laws can restore a balance between my liberty and others' equal liberty. But even given the need for law and political society, negative liberty requires "a certain minimum area of personal freedom which must on no account be violated . . . A frontier must be drawn between the area of private life and that of public authority."23 Thus, regardless of the technical or semantic differences among negative-liberty theorists over the definitional requirements of "restraint" and the role and meaning of "choice," the strong individualism of negative-liberty theory ensures that restraints on liberty are seen as external to the self and, moreover, that they contain an inherently conflictual character, because individuals' desires and interests inevitably collide with them.24

Positive-liberty theory challenges, or at least "expands," the negative conception of liberty in at least three ways. Most obviously, it concerns itself with the "positive" provision of the conditions necessary to take advantage of negative liberties, such as providing wheelchair access to buildings or scholarships for education. The definition of barriers as external impediments is too narrow; for instance, freedom of education is rather hollow if you cannot afford tuition or get into the building where instruction is offered. Adopting a more contextual and communal notion of the self, positive liberty is able to view individual conditions such as disability, as well as social conditions such as poverty, as barriers to freedom that can be overcome by positive action, that is, the provision of conditions the individual cannot create on her own. This might not seem like a noteworthy enhancement of negative liberty. After all, since Berlin wrote his original article, various negative-liberty theorists have endorsed different versions of this idea and incorporated it into their own conceptions of negative liberty. Particularly popular is the effort to expand the notion of "external barrier" to include generalized social conditions such as poverty. I will discuss some of these arguments in the following section; but even if this first challenge of positive liberty has been taken up successfully or even appropriated by negative liberty, it sets the stage for the remaining two ways in which positive liberty differs from its counterpoint, and these are far more significant.

One of these is positive liberty's focus on "internal barriers": fears, addictions, compulsions that are at odds with my "true" self. According to positive liberty, we can have "second order desires," or "desires about desires. We experience our desires and purposes as qualitatively discriminated, as higher or lower, noble or base, integrated or fragmented, significant or trivial, good and bad."25 Because of these conflicting capacities, it is not enough to experience an absence of external restraints, because the immediate desires I have may frustrate my true will. For instance, while trying to quit smoking or to combat an eating disorder, an argument with my department chair triggers a craving for a cigarette or an entire bag of Oreos. Positive liberty says that if I were to give in to these cravings, I would be not just weak-willed but unfree, because I am violating my true desire, on which I have reflected at some length (and after all, smoking or binging will not resolve the dispute with my colleague, so we cannot even argue for a trade-off between competing goals).

It follows from the notion of internal barriers, however, that others may know my true will better than I, particularly when I am in the grip of these self-destructive desires: as you snatch the cigarette or Oreo from my lips, you are preserving my true self from false desires, and enhancing my liberty. This is often called the "second-guessing" problem, because others claim to know what you want better than you do yourself. It is the most troubling aspect of positive liberty: determination of the will by others, and specifically by the state. The classic instance is Rousseau's general will; since the laws embody the true will, he says, then by forcing me to obey the law the state is only "forcing me to be free," that is, to follow my true will, whether I know it or not. This is the nightmare that Berlin particularly argued against in "Two Concepts of Liberty," with good reason. But what Berlin and other negative-liberty advocates seriously underplay is the idea of an individual having conflicting desires and a divided will. That is clearly the sentiment behind the claim that the compulsive binger or addicted smoker is unfree: I really want to quit smoking, but the stress is pushing me to cheat. Similarly, perhaps my husband broke my leg because he was trying to keep me from meeting my cocaine supplier; I've been struggling for months to kick my addiction, but she called me at a particularly vulnerable point. In both of these cases, I want two mutually exclusive things--to smoke or take drugs and to quit--and most people would probably have to agree that quitting would be a better choice, a choice that would be more consistent with my continued freedom; indeed, one that would liberate me, and hence the choice I really prefer to make. So it is at least an open question whether the person who prevents me from pursuing my addictions impedes my liberty or enhances it. For this reason, Taylor calls positive liberty an "exercise concept" of freedom: people must exercise their full capacities if they are to be considered free.26

The notion of internal barriers to freedom generally raises the greatest hostility to the idea of positive liberty, precisely because it seems to indicate that others can prevent me from doing certain things which they think will violate my deepest desires and highest purposes, or force me to do other things which they think will help me realize these purposes. Berlin thus argues that the two key notions in positive liberty are self-mastery and rationality: in order to be my own master, I (or my intellect) must control, or at least not be under the influence of, anything that is "not me." But what is "not me" may often include my own emotions, feelings, desires, physical sensations, and so forth: freedom is defined as what I would want if I were perfectly rational. Since people frequently betray rationality and desire the wrong things, however, an independent external force, usually the state, must ensure that I follow my true will. The idea, subscribed to by Hegel and (many believe) Rousseau, of an objectively true, higher-order desire stemming from a true self which others can know better than I entails that I do not need to be able to identify the desire as mine for it to be genuinely mine; because it is the state's will, the common good, the Geist or general will, it is my higher will by definition, whether I realize it or not. Berlin thus defines positive liberty as a paradox: to coerce X to do A because it is for her own good is a logical impossibility, because if it is for her own good, then she really wants it and therefore is not being coerced.27 He therefore maintains that internal factors should not be considered relevant to the concept of freedom.

However, most positive-liberty theorists do not share Berlin's interpretation. Charles Taylor, for instance, argues that a focus on internal barriers to realizing my higher desire does not entail external mechanisms to direct me to it. His account of the "divided self" is indeed quite individualist: in his examples, the subject always seems to know that he has a higher and lower desire, and is struggling to achieve the former. He offers this example: "spiteful feelings and reactions which I almost can't inhibit are undermining a relationship which is terribly important to me . . . I long to be able not to feel this spite. As long as I feel it, even control is not an option, because it just builds up inside until it . . . bursts out."28 In this example, he is aware that he is destroying himself, and the higher-order desire is one he can identify and identify with: "I long to be free of this feeling . . . as I lash out at her with my unbridled tongue." In what Taylor calls the "self-realisation" view, direction toward the higher desire must be done individually, because individuals have different higher purposes and goals.

Taylor's account has the merit of avoiding the "totalitarian menace" of second guessing. But just like Berlin's argument that positive liberty's emphasis on higher desires inevitably leads to state control, Taylor's argument that I can and must be able to identify higher desires as mine over-simplifies positive liberty. Its individualism brings us full circle to the very atomism that Taylor critiques in negative-liberty theory. For instance, even though Taylor's spite is expressed in the context of a personal relationship, and particularly when he is interacting with a specific other person, rather than alone in his study, he describes the feeling as emerging spontaneously from within himself, the highly individuated subject, without any attention to the context within which it has developed or exists: Where does the spite come from? Why does he feel it? And why spite rather than, say, withdrawal or selfpity or sadness? What are the criteria for determining the "importance" of the relationship? What does the "almost" mean when he says he "almost can't inhibit" his spiteful feelings? Is there a specific reason, hidden in that space between "almost" and "can't," that prevents him from controlling his spite? For instance, perhaps the relationship isn't really "good" for him, not in accord with his "true" interests, and spite is his subconscious's way of telling him this.

On this view, the question of which desire is "higher" and which "lower" is somewhat up for grabs. Just as Taylor sees that the relationship is more important to him and wants to stop his spiteful behavior, so could he then identify the spite as in fact protecting him from the relationship; but then what's to stop him from saying that such justification of spite is simply a more sophisticated way to rationalize his fear of commitment? And so on, in an endless stream of second-guessing.29 Such a possibility certainly allows for, and even requires, articulation of his preferences and desires by the individual himself, as Taylor argues; but it also demands a working-through of history, relationship, and context, all of which are subjectto and even necessitate the assessment and input of others. Understanding the self presupposes the existence of language, of a conceptual vocabulary, a system of signs with which to formulate and represent my own experience to myself; and it requires others with whom I can be in conversation, to analyze and determine what desires are really mine, and really better for me.30 This raises the question of where to draw the line between the internal self and the external world, because our self-understandings, our desires and choices, as well as the barriers we experience, always need to be understood in context. If individuals exist in contexts, then they--their feelings, desires, thoughts, wills, preferences--cannot be understood outside of those contexts, as abstract and self-contained units. Without such specificity of context, the individual too is unspecified, an abstraction. In this regard, Taylor's argument displays, even as he himself ignores, the inherently social dimension of internal barriers and the relationship between internality and externality.

Acknowledging this relationship leads to the third way that positive liberty challenges negative liberty, though it is an aspect that most theorists of freedom do not recognize: namely, the "social construction" of the choosing subject of liberty, of the person who is supposedly having desires and making choices. How is it that I have the desires I have? Why do I make the choices I do? If it is possible to say that we can have conflicting desires, and if it is possible to rank these desires as better or worse, more and less valuable, then the issue of who I am comes inevitably into play: What is my true will? What do I really want? Such questions invite us to consider the social construction of the choosing subject, of the individual agent who has desires and makes choices within specific social, historical, and institutional contexts. The idea of social construction is that human beings and their world are in no sense given or natural but the product of historical configurations of relationships. Our desires, preferences, beliefs, values--indeed, the way in which we see the world and define reality--are all shaped by the particular constellation of personal and institutional social relationships that constitute our individual and collective identities. Understanding them requires us to place them in their historical, social, and political contexts. Such contexts are what makes meaning possible; and meaning makes "reality."31

But of course the context that women and men live in is one of patriarchy, sexism, and male privilege--as well as racism and white privilege, capitalism and class privilege, and so forth. If we are socially constructed, feminists have argued, male domination has played an important part in that construction; its laws, customs, rules, and norms have been imposed by men on women to restrict their opportunities, choices, actions, and behaviors. Hence, to those who say that women are more "caring" than men, we could reply that that is because they have always been required to take care of children and men; to those who maintain that women are naturally irrational, we could respond that that is because they are denied access to education; to those who assert that women and girls are not naturally assertive in the classroom, we could also say that that is because they are often not called on or are ignored when they speak.

This construction of social behaviors and rules comes to constitute not only what women are allowed to do, however, but also what they are allowed to be: how women are able to think and conceive of themselves, what they can and should desire, what their preferences are, their epistemology and language. If a man beats his wife repeatedly and tells her that she's stupid and ugly and that she deserves to be beaten; if she calls the police but they just tell the man to cool off and walk around the block; or if she has left him, but he has hunted her down, or threatened to hurt her family, or come to her office and behaved violently there, resulting in her dismissal, she may eventually start to act in ways that facilitate the batterer. For instance, she may begin to believe that there is no escape, so why bother trying? She may find that going limp during a beating reduces the injuries she receives; but at the same time it makes it easier for him to hit her. Or when she perceives that he's getting angry, she may trigger an outburst by picking a fight, so that the attack does not catch her by surprise. It is difficult for most nonbattered people to make sense of such choices--it looks like she is not really doing anything to help herself--but the choices she has made have been constructed for her through a system of male privilege that tacitly condones domestic violence even though we claim that it is unacceptable. In this sense, these rules and norms of patriarchy are not simply external restrictions on women's otherwise natural desires; rather, they create an entire cultural context that makes women seem to choose what they are in fact restricted to.32

Social construction thus shares some features with "adaptive preferences": the notion, as Elster in particular develops it, that my preferences become shaped and formed by the options that are available.33 Certainly that could explain the woman with an extremely controlling husband watching her every move who has convinced herself that she does not really want to leave the house. Similarly, "oppressive socialization"34 is an apt description of women who have come to associate violence with love because they have never experienced any other kind of relationship. Like theories of socialization and adaptive-preference formation, the notion of social construction allows us to engage questions that most contemporary theorists avoid, for example, the possibility of "internal" barriers to liberty. It simultaneously opens up the possibility that the inner self--our preferences, desires, self-conceptions--is constructed by and through outer forces and social structures, such as patriarchy, colonialism, capitalism, and so forth.

However, social construction is more complicated and deeply layered than either socialization or adaptive preferences. Socialization, for example, is seen as something that is done to people by particular other people; it is seen as at least quasi-conscious and intentional, and often damaging. Socialization is thus conceived as specific psychological and behavioral responses to conditions that could be changed or avoided. It thereby assumes an essential, natural self underlying these oppressive conditions which would emerge if they were removed. The concept of adaptive preferences similarly assumes this natural self by positing that the preferences that have been adapted to circumstances are not what the individual truly wants. Hence, every one of Amartya Sen's examples of adaptive preference involves someone who is clearly oppressed--the "hopeless destitute," the "subjugated housewife"--and who would have other desires but for the conditions oppressing them.35 Granted, Sen most frequently writes about freedom in the context of dire poverty and oppression in Asia and North Africa, the millions of "missing" women resulting from a disproportionately large female mortality rate, which he attributes to sexist neglect of female children "in terms of health care, hospitalization, and even feeding."36 I by no means wish to deny the force of such arguments, nor the possibility that people can adapt their preferences to extremely oppressive conditions. But the idea of social construction is aimed at understanding much less overt forms of social production; it is something that happens to everyone, men as well as women, rich as well as poor, at all times and in multiple ways. By suggesting that people are produced through social formations, and not simply limited by them, the idea of social construction thereby calls into question the assumption of what is genuine or true to the self and what is false. As Kathy Ferguson puts it, "It is not simply that [we are] being socialized; rather, a subject on whom socialization can do its work is being produced."37 The social constructivist argument that the context in which women's desires and preferences are produced, expressed, evaluated, and either granted, denied, or ignored altogether, is a patriarchal one does not mean that women are simply "unfree." In developing a critique of naturalism and a theory of how women are who they are because they are socially constructed by patriarchy, feminists have been able to note the ways in which desires, preferences, agency, and choice are as socially constructed as are the external conditions that enable or restrain them. But at the same time, if social construction characterizes our entire social identity and being, if everyone is always and unavoidably socially constructed, then not only our restrictions but our powers as well must have been produced by this very same process. Who we are--the "choosing subject"--exists within and is formed by particular contexts; the ideal of the naturalized and unified subject utilized by most freedom theory is thus deeply problematic and simplistically overdrawn. The contexts in which we live, patriarchal, sexist, racist, and classist though they may be, have produced women's, indeed everyone's, agency. Yet, while they "empower" women and make their agency possible, they often simultaneously put restraints on women's freedom not suffered by men. This duality of social construction permits, even requires, a more complicated engagement of the question of freedom.

This might suggest something of a paradox, for it seems to imply that nothing need--or can--be done about social construction. If women or the poor happen to be constructed in ways that make them different from wealthy white men, so what? How can feminism claim that this construction is worse than any other construction of identities? Do we not end up with the helpless relativism of a deconstructive poststructuralism? I believe not. Social construction suggests that the dichotomy between negative and positive liberty, and between internal and external restraint, is itself artificially constructed; and moreover that this construction can be seen as motivated by particular power structures that favor men over women. The fact that these power structures themselves were socially constructed, in a seemingly endless devolution, does not prevent feminists from acknowledging the ways in which power operates within any given social context and to make political evaluations of those operations. For instance, in a society where resources for battered women are a low funding priority, where courts and police openly disbelieve women who report abuse, where relatives and friends fail to help and protect them, shame, guilt, feelings of helplessness and/or desert are predictable responses. But they are not just "internal" feelings that come from the self as an isolated entity. Rather, they are social expressions and manifestations of public policy and attitudes, political structures and decisions that reflect particular, even if frequently unrecognized, values and relationships of power that privilege some and disadvantage others. Similarly, in a society that pays poor women a bare subsistence subsidy for raising children, that requires them to reveal the most intimate aspects of their personal and sexual lives in order to gain such help, that demands that they enter the labor force even at the cost of foregoing education, that considers them as "cheats" if they try to rise above subsistence poverty by working on the side, "public assistance" is likely to have an impact on the self-conceptions of recipients, in the form of shame, anger, feelings of powerlessness, or victimization, or some combination of these. But to "blame" welfare mothers for these self-conceptions ignores the generative source of such feelings, which are, arguably, reasonable responses to extreme situations produced by a set of power relations and values that are constructed to privilege some over others.

Social constructivism thus reveals that a focus on external barriers will be weakened without attention to the internal ones, as well as to the larger social, institutional, and cultural context in which such barriers are created and operate. We must acknowledge the interaction of "inner" and "outer" and see them as interdependent in meaning and in practice, in order to interrogate the social construction of the choosing subject, the subject of liberty. This interaction does not result in determinism--the view that since there is no way not to be socially constructed, there is no way to change ourselves, because humans cannot control social formations like "patriarchy" or "racism"--but rather provides the means for identifying not only the ways in which power relations are structured but why it is so difficult to see those relations and that structure.


The issue of social construction will be developed more fully in chapter 3; the perhaps more controversial question of whether social constructivism actually informs positive-liberty theory, and the relationship of social constructivism to the modern history of liberty theory, will be considered in chapter 2. But for now, I merely want to note that the possibilities of social construction and the focus on the internal aspects of liberty--desire, will, and identity--and how they both facilitate and block freedom is, in my opinion, the most important contribution of Berlin's typology: it offers two different conceptions of the free self and a conceptual language of the "external" and "internal" factors of freedom. Although Berlin rejected the latter, and although many freedom theorists place internal and external in a dichotomous relationship and deny or ignore the more complicated potential that positive-liberty theory offers us concerning social construction and the inner self, Berlin's conceptualization makes it possible to see this potential and offers a vocabulary for a more complex understanding of freedom that is more conducive to feminist concerns.

Certainly, Berlin's typology has been challenged by many, and some would suggest that couching a discussion of liberty in its terms is misdirected. For instance, Gerald MacCallum argues that there are not "two concepts of liberty" at all but rather only one, consisting of a triadic relationship between agents, objects of desire (including desired actions and conditions), and preventing conditions. "[F]reedom is thus always of something (an agent or agents), from something, to do, not do, become, or not become something."38 Theorists of negative and positive liberty are each "attending to, or emphasizing the importance of only one part of what is always present in any case of freedom,"39 and what they are really arguing about are various values which they believe are important to political society and social relations,40 not about freedom per se. Freedom as a political practice can be structured with a particular content according to particular values, but that can never negate the basic philosophical truth that its conceptual meaning is constituted in a triadic relation.

MacCallum's argument is important for the ways in which it identifies, with brutal honesty, the politicalness (or normativity) of the supposedly philosophical (often semantic) debate about freedom. Neither positive nor negative liberty is usually embraced simply on the grounds of its philosophical consistency or compelling logic. Rather, politics is always at issue. Indeed, as Gray implicitly suggests, MacCallum himself obscures his own political leanings: the triadic formula that he sets up is already biased in favor of negative liberty, because his idea of "reasonableness" and the role he ascribes to "rationality" all presuppose a negative-liberty framework. His negative-liberty bias allows him to dismiss variants of positive liberty as confusions of the question of liberty rather than as genuine dimensions of it.41 But even if we grant MacCallum's point, this does not counter the notion that positive and negative liberty are two significantly different visions of freedom. It is precisely because of the political nature of liberty that the positive/negative typology provides an important theoretical division, one that defines general parameters for two orientations toward ontology and epistemology. To borrow from Rawlsian terminology, MacCallum may be correct that we have one concept of liberty, but we do have two conceptions, negative and positive, and they differ at fundamental levels.

This suggests three reasons for retaining the typology, not to perpetuate the often semantic arguments about how to define freedom, but as an important normative tool for framing its central issues. The first reason is that, debates and challenges notwithstanding, the typology of positive and negative liberty has in fact dominated theoretical discussions of freedom. Prominent theorists as diverse as Stanley Benn, Ian Carter, Diana Coole, Richard Flathman, John Gray, Quenten Skinner, Hillel Stiener, and Robin West all acknowledge the centrality of the typology to liberty theory.42 Though individual theorists such as MacCallum may disagree with Berlin's typology, they seem unable to escape it, and it has retained a powerful grip on philosophical thinking about liberty. Second, this grip is due, to a significant degree, to the decidedly--if often overlooked--political character of the typology. I agree with some critics that Berlin's formulation is overdrawn and simplistic. He clearly had Cold War political motivations for his categories: he wanted to ally positive liberty with bad-guy communist dictatorships and negative liberty with good-guy Western democracies. But that Berlin manipulates philosophy to the ends of politics should not lead us into the trap of separating the two, and of missing the impact of the concepts as political and not "just" philosophical. The two concepts of liberty, as he articulates them, reflect two different--though perhaps equally problematic--conceptions of a person: one as innately separate, individualistic, unconnected, rights oriented, even antagonistic; the other innately connected, communitarian, even selfless, concerned with responsibility. Depending on which view of the subject one takes--or more precisely, where on the continuum between these two extremes one's conception of the subject lies--a variety of conclusions follow about the relation between state and society, society and individual; in short, political values. In this, the typology suggests that freedom is not just about "who we are" but also about "what kind of world we want to live in." Many critics such as MacCallum who want to reject the typology altogether fail to recognize these issues as central to freedom, and this failure may explain why political philosophers keep getting drawn back into Berlin's framework.

Finally, and perhaps most importantly, both positive and negative conceptions inform popular understandings of liberty: after all, most of us can understand, in an everyday sense, how the cheating smoker and cocaine addict are both free and not free. This is because the typology does in fact say something very important about freedom. Both variants of freedom are centrally about making choices. Choice is a complex process of negotiation and relationship between what we commonly call "internal" and "external" factors; between will, desire, and preferences, on the one hand, and forces that not only inhibit or enable the realization of such desires but also contribute to or influence the formation of these desires, on the other. It is precisely in this notion of internal and external barriers to liberty that I think the positive-negative typology is the most powerful, and at the same time most problematic. Negative liberty emphasizes the role of external barriers, while positive liberty highlights the internal; and indeed, I believe that this is the key dividing line between the two models. Those, like MacCallum and Orlando Patterson, who focus on Berlin's differentiation between "freedom from" and "freedom to" to illustrate its incoherence (because every "freedom from" is automatically a "freedom to") miss the more important point of this external/ internal divide.43

Indeed, a great deal of contemporary freedom theory fails to acknowledge the politics involved in arguments over freedom and instead tends to focus on somewhat semantic debates over the meaning of freedom, and particularly what constitutes a "barrier" or "restraint." The distinction between being "unfree" and "unable" sits at the heart of contemporary theorists' consideration of what counts as a barrier. A "restraint" by definition is something that makes me unfree, but "inability" is seen as something intrinsic to myself or the world, part of nature or naturally caused. As Phillip Pettit argues, freedom involves more than "nonlimitation," because intentionally placed limits on freedom are of greater moral and political concern than are those that occur naturally or through no one's fault.44 Accordingly, negative-liberty advocate Richard Flathman requires that barriers must be "obstacles deliberately placed or left by other intentional or purposive agents."45 The requirements of intentionality and purposiveness, Flathman suggests, eliminate from consideration impediments that are not deliberately placed but are nevertheless socially generated and constructed, such as poverty and lack of employment opportunity or inhibitive social norms for women. Rather, these are part of our "form of life," basic background conditions that must be taken as given. In a footnote, Flathman does potentially allow "social arrangements as impediments to liberty if they are alterable or remediable by human agency," but only if they are conducive to a "charge of responsibility for deliberate interference, not merely . . . causal influence."46 There must be an identifiable moral agent who can be held accountable, and that agent must have used her agency--that is, acted deliberately, intentionally, purposively--in erecting the barrier to others' freedom.

In Berlin's initial formulation, of course, restraints had to be seen as deliberate in order to count as such; the mere incapacity to do something did not constitute unfreedom to do it unless someone else caused the incapacity. Though Berlin later expanded this notion to include "alterable human practices" which bar possible choices,47 thus suggesting that obstacles need not be deliberate, they still had to be attributable to specific humans who could be held accountable for them. Parent, however, argues that Berlin may go too far in this modification. For Parent, physical force or other overt coercion is required to constitute a restraint.48 Though Parent rejects Flathman's requirement of purposive action, and thereby allows for natural events to serve as barriers, this apparent expansion is counteracted by his strict requirement that unless something is impossible, I am free to do it. Barriers to freedom must be objectively assessable, not subjectively contingent; a battered woman, for instance, may be afraid to leave her abuser, but this fear does not in itself make her unfree to leave. After all, "consider the many individuals who have performed actions, started movements, and initiated reforms knowing full well that their doing so would cause others to inflict severe, sometimes unjustified punishment on them."49

Ian Carter, by contrast, suggests that intentionality is not relevant to the existence of a barrier, though identifiable agency is.50 This might allow us to identify a rapist who claims he "didn't mean to hurt" his victim, or a sexual harasser who thought he was simply paying his secretary a compliment, as interfering with women's freedom even if such men did not intend to do so. Another way of expanding the notion of restraint, "the responsibility view," involves focusing less on how barriers are set in place and more on their possible removal: I may not have closed the door to the room in which you are now locked, but the fact that I can let you out without serious risk to myself means that I am morally responsible for doing so, even though I am not causally responsible for your predicament. By not letting you out, I interfere with your freedom.51 This view rejects the notion that obstacles need to be deliberately placed in order to count as restraints on freedom, for the notion of moral responsibility transcends causal responsibility: just because I did not create an obstacle does not mean I may not be morally responsible for removing it. This view of barriers to liberty can be seen as extremely expansive as well as highly political, for what constitutes a barrier will turn on our views of moral responsibility. So as David Miller suggests, a socialist will view a capitalist as responsible for the unfreedom of workers.52 One could, therefore, see poverty as a restriction on liberty if one developed an economic theory that identified a definite class of people who were able to alleviate the poverty.

Indeed, the question of poverty and whether "economic inequalities constitute constraints on freedom"53 is one that a number of freedom theorists take up. It holds potential for feminist inquiries into the possibility that sexism or patriarchy presents a barrier to women's freedom, but this potential is largely unrealized, because of the individualist framework freedom theories deploy. For instance, because Kristjansson separates moral from causal responsibility in the responsibility view, social conditions such as poverty can be viewed as a constraint on freedom, but only "to the extent that identifiable people can be held accountable for it through their negligence, violation of positive duties, etc."54 Such precision cannot be achieved, however, even within a socialist framework: at best, a generalized class of people can be shown to benefit disproportionately from the existing structure, and even they are only responsible for participating in it, not for changing it, which as individuals they may not be able to do. But this undercuts the socialist's claim at the root.

Similarly, Miller maintains that the standard of moral responsibility should be construed fairly narrowly, and he retains the individualist idea of assignable agency in two senses. First, he rejects the idea of collective responsibility except insofar as we can isolate and identify individual agents within a collective "who contribute to the outcome intentionally or in dereliction of duty."55 Noting that the person who adds the final straw that breaks the proverbial camel's back is responsible insofar as he knows others have added their straws, he concedes that "I bear my share of responsibility of the resulting injury to the camel." However, "the difficulty in such cases is likely to be finding out who could reasonably be expected to have foreseen which outcomes," a dim prospect at best, according to Miller. Hence, one cannot hold a collective responsible to a greater degree than the individual members. At the same time, however, the assignable-agency criterion can also restrict the scope of a barrier; for instance, a capitalist will not see his own activities as restricting his workers' freedom but rather as augmenting it, by offering them various choices.56 Thus, Miller limits responsibility to a relatively direct relationship; otherwise, he says, the notion of restraint grows far too large.57

Benn and Weinstein, whom Kristjansson credits as originators of the responsibility view, similarly consider poverty as a potential barrier to freedom,58 only to take away with one hand what they give with the other. Their well-known example of a shopkeeper who sells eggs rather than giving them away is used to counteract notions of social responsibility that extend beyond purposive individual agency. Because "the general framework of property relations is taken to define the normal conditions of action," and the shopkeeper is an individual agent who has not caused the background conditions that lead to the eggs' price--as would be the case if he owned the only shop in the neighborhood and raised his price with the express purpose of exploiting his customers--then the shopkeeper cannot be seen as interfering with liberty.59 Because of this individualism, the shopkeeper is the focus of inquiry and therefore let off the hook. It may be true that the egg merchant in this example is not himself a "restraint" on my desire to eat eggs, but the lack of money is, and the possibility that poverty is a barrier to freedom requires inquiry into the social conditions that create that lack. By directing our focus to the price of eggs and rejecting the claim that it is a barrier, Benn and Weinstein can claim that there is no barrier at all, when it may instead be the case that the blockage exists elsewhere, and would be revealed by broader notions of economic and social responsibility.

In Benn's subsequent book on freedom, he seeks to pursue the issue of economic and social structures as barriers to liberty by considering economic exploitation. Exploitation requires both "that there is no reasonably eligible alternative and that the consideration or advantage received is incommensurate with the price paid. One is not exploited if one is offered what one desperately needs at a fair and reasonable price." This definition is certainly compatible with a broad understanding of barriers and social responsibility. But Benn goes on to argue that restraints must be the product of "some rational being or beings" who "can be held responsible for them, for bringing them about or permitting them to continue."60 For his example, however, he uses Esau and Jacob, not worker and capitalist, that is, a relationship of two specific individuals rather than of two social groups and categories. This presents an almost willful denial of the broader potential of his argument. Indeed, in my view, Benn misconstrues the point of the parable altogether, which is not to illustrate exploitation; insofar as he limits Esau's freedom by stealing his birthright, Jacob betrays the love and trust that a brother is supposed to feel and display and violates his responsibility to his brother.

So even the responsibility view of restraints persists in its dependence on the overly individualized self and insists on an assignable agency of particular and identifiable individuals who, if they do not purposely place barriers in the agent's way, are at least morally responsible for failing to remove them. By contrast, Gray argues that restrictive conditions such as poverty do not have to be caused intentionally, nor by identifiable agents, as long as they are avoidable (e.g., if "mass unemployment . . . resulted from misguided monetary policies whose application was in no way inevitable"), or at least remediable by human action.61 This allows for collective, rather than merely individual, responsibility. But at the same time, Gray also maintains that determining whether social conditions are avoidable or remediable is extremely difficult, if not impossible, and gives up on the idea fairly quickly. For him, as for the other theorists discussed so far, even when negative-liberty theory seeks to accommodate positive-liberty theory's expansion of the concept of barrier to include generalized social conditions, the strong individualism that persists in these efforts dooms them to failure. For when considering large social forces such as poverty in a capitalist economy, intentionality and assignable agency are generally impossible to locate; indeed, they are in some ways definitionally excluded, because a much broader set of theoretical assumptions is required, such as those found in socialism. Indeed, the focus on defining terms such as "barrier" rather than on the political and normative implications that follow from particular social relations is what allows this methodological individualism to persist. Poverty is a political and social condition, not a semantic one, and it affects classes of people, not isolated individuals.

Gerald Cohen seems to adopt this more political understanding when he argues that poverty restricts individuals' freedom. Wealth is not simply a resource that some people happen to have, and which happens to make those people freer, according to Cohen. Rather, he seeks a more systematic understanding of poverty and wealth. "My argument overturns the claim that a liberal capitalist society is, by its very nature, a free society," he says, for all societies and economic systems "have structurally different ways of inducing distributions of freedom." In capitalist societies, that distribution is determined by "the distribution of money."62 Indeed, the notion that wealth does not determine freedom is an "illusion" forged by the very ubiquity of capital's determination of freedom, much as the supposed "naturalness" of property that the social-contract theorists posit (as will be shown in the following chapter) creates the illusion that inequalities of wealth are the result of individual choices and abilities rather than of power. But according to Cohen, "a property distribution just is . . . a distribution of rights of interference." Hence, poverty is deeply political; indeed, it constitutes "social relations of constraint" whereby some are free to act and others are not.63

Such a broad understanding of the constraints that poverty produces as a social force, and particularly as a social relation, has the potential to broaden the concept of a "barrier" to liberty to include general social forces that lack identifiable agency and intentionality. But like the other theorists discussed here, Cohen is drawn back into individualist ideals and qualifies his arguments. He maintains that capitalism per se does not interfere with poor people's freedom; rather, "to lack money is . . . to be prey to interference" by particular individuals who have more power (relative to the people whose liberty they interfere with) within a capitalist system.64 For instance, if I cannot afford to buy a sweater, he suggests, the department store clerk (or, more accurately, the store security guard) will interfere with me if I try to leave the store with the sweater without paying for it. The "interference" with freedom is located in, or exercised by, particular individuals; once again, it is identifiable agents, not economic systems, that interfere with my action.65 What Cohen's account leaves out, however, is the notion that the clerk's or security guard's actions are themselves constructed and determined by the contextual framework of capital in which the possibilities for action are set. Just as the shopper is prevented by the security guard from taking the sweater without paying, so the security guard is prevented by his supervisor from letting the shopper take the sweater (he would be fired, at the least). And his supervisor is prevented from overlooking the guard's infraction by the store owner, who would likely dismiss the agency if he discovered its laxity. And the store owner could even be seen as restrained by a moral imperative to keep the store going so that people like the sales clerk can continue to have jobs and feed their families. All of the actions and choices made by these individuals may be self-consciously "theirs," in the sense that the guard does not want to let the shopper take the sweater, the supervisor wants a vigilant employee, the store owner wants a well-run security force guarding his store. It is not necessarily the case that capitalism is overtly forcing these individuals into certain choices. However, these choices and actions are, nevertheless, conditioned and constructed by and within the context of capital; and insofar as capitalism is not a force of nature but an economic and political system that distributes freedom along with economic resources, as Cohen maintains, then that context itself may be conceptualized as at least a significant, if partial, source of the shopper's unfreedom.

Certainly, as Cohen himself argues, the conclusion that poverty restricts freedom does not in and of itself mean that capitalism is "bad," or that it must be rejected in favor of socialism; socialism can produce its own set of constraints on freedom. But even if we decide that capitalism is desirable, Cohen would say, we still need to acknowledge the ways in which freedom for the poor is limited. As he admits, however, this final assertion is a conceptual claim, not a normative one--"no normative claim has been defended, or even asserted" in his paper--much less a political one. It is a claim about the meaning of freedom, and whether poverty fulfills the criteria of "interference." By contrast, I suggest that the politics of unfreedom--whose freedom is limited and whose is not, whether certain groups of people (for instance women or racial minorities) are systematically poorer than other groups (such as white men)--are central to conclusions about the concept of freedom. Like the other theorists considered here, Cohen's individualist focus undermines the ability to accomplish what he--and they--originally set out to do: namely, consider poverty as a barrier to freedom.

Amartya Sen is a notable exception to this individualism, for he pays a great deal of attention to the ways in which poverty as a social force inhibits people's liberty. Poverty involves "deprivation of basic capabilities, rather than merely . . . low income." Wealth and income are not goods in themselves but rather means to other things, such as food (freedom from starvation) and health care (freedom from morbidity). Sen argues that by negatively affecting the capabilities of individuals, poverty can deny substantive freedoms to do specific things and thus undermine any formal or process-oriented freedom they may in theory possess. Thus "economic unfreedom," though not the only freedom, is nevertheless key, for "extreme poverty can make a person a helpless prey in the violation of other kinds of freedom," such as political freedoms.66 Other kinds of freedoms he identifies, such as health care, are of course directly "economic" only in some contexts; in the United States, for instance, health care is privately administered and therefore requires health insurance, usually through employment benefits or out-of-pocket expenditures (or sometimes both). But even socialized medicine, which guarantees rich and poor alike access to care, requires a certain level of GNP for a nation to afford the technology and medicines necessary for adequate health care; and as Great Britain has demonstrated, those rich enough to go outside the public system generally receive better and prompter care. Sen argues for the importance of evaluating substantive freedoms on a "factual base" that considers the "real opportunities" individuals have: not just the doors that are open, as Berlin puts it, but whether one can actually go through them. Thus famine should be considered "in terms of the loss of entitlement--a sharp decline in the substantive freedom to buy food"; illness and mortality resulting from inadequate (access to) health care should be seen in terms of a denial of entitlements to life.67 I will return in chapter 7 to Sen's arguments about the role of equality in freedom, but here it is enough to note that Sen's consideration of poverty as a barrier to freedom, in contrast to the other theories I have discussed, raises the question as to whether the negative-liberty concept of external barrier can be expanded to include broad social conditions such as poverty. He clearly demonstrates a strong alliance with liberal principles, and the conception of freedom to which poverty poses a barrier is expressed in the negative-liberty terms of making choices and expressing preferences that are not the product of or adapted to oppression. And of course he also gives a great deal of attention to political freedoms. Yet Sen also asserts that substantive freedom is as important as procedural freedom, and an important concept he develops along with "agency freedom" is "well-being freedom," which shares positive-liberty values.68 But Sen is much less interested in whether he should be labeled a positive or negative libertarian than in the political and economic conditions that set the terms for people's lives, their capabilities and freedom to achieve their own welfare. He is concerned with understanding the political conditions in which particular situations of freedom evaluation are made; and poverty, he maintains, is a decidedly political condition.


Sen's argument is significant for feminism, because if generalized conditions like poverty can be established as a barrier to freedom, perhaps patriarchy can be similarly established. And indeed, it is not coincidental that Sen is one of the very few mainstream theorists to develop a theory of freedom that gives a central place to the specific forms of oppression and unfreedom that women experience. He identifies these forms of oppression as the political products of cultural, historical, and economic forces rather than as simply part of the cultural landscape, a "normal background condition."69 Feminists frequently point out ways that customs, practices, and beliefs that men and women have accepted as "normal" in fact encode deeply sexist attitudes that restrict women--and often men as well--in illegitimate, unjustified, and unnecessary ways. If context sets the terms for understanding claims of freedom, and if women's choices, opportunities, desires, and options exist within a context of patriarchy or sexism, there is good reason to believe that sexism itself can be a barrier to freedom. That is, not just individual sexist acts perpetrated by particular individuals, but the entire cultural construct that assigns greater value to men than to women, that provides more options to men and supports men's pursuit of choice more than women's--in short, what many feminists alternately call male privilege or patriarchy--can restrict women's freedom. But this is a restriction which the mainstream conception of restraint and freedom, focused as it is on intentionality, identifiable agency, and purposiveness, can never see. For instance, the theorists discussed here would most likely agree that an individual sexual harasser or abuser or rapist interferes with individual women's freedom but would not regard as barriers to freedom the social conditions that allow such things as rape or abuse to happen in the first place, and make the arrest and conviction of the perpetrators of such actions difficult. As Benn maintains, "the mere possession of . . . power [what he calls "power against"] without its being exercised [what he calls "power over"] would not be enough to make the other unfree";70 the fact that men can rape women and beat up their wives, frequently with impunity, does not matter, it is only when a woman is raped or beaten up that her freedom can be said to be limited.

But consider a woman's decision to go out at night. Most men have the power to rape at least some women, a "power against" women which women do not reciprocally have against men. Most men do not exercise this power, and many would be horrified at the suggestion that they do so, but it is a power that pervades the culture nonetheless, making many women afraid to go out at night alone. Even if they have never themselves been raped or attacked, the fear still inhibits them. Such women are made unfree by such power and the fear it generates; the existence of power against, though not actually exercised at a given moment, has translated into a generalized and constant power over. Could we say that while women may "feel" unfree, they are wrong, that they are "in fact" free, that their fear is irrational and misplaced? Some, like Katie Roiphe and Christina Hoff Sommers, suggest that statistics on rape are inflated, and that women's perception of unsafety is not justified.71 But even granting their claims--which are at best controversial, given that the rape figures used in many studies are not made up by feminists but are provided by the FBI and other state-run, and supposedly patriarchal, institutions--the fact remains that rape is an act perpetrated against individual women, and there is no way to predict accurately whether a particular woman will in fact be raped in a given situation.72 The notion that women simply need to avoid overtly unsafe situations is further belied by the experience of acquaintance rape; and even stranger rape often occurs in women's homes and during the day, rather than in dark alleys at night. And even if it did only occur in such stereotypical contexts, doesn't telling women to simply avoid those situations--say, jogging in Central Park at night--restrict women's liberty, and do so more than men's?

We could say, following William Parent, that women are "in fact" free because no one is physically restraining women, no laws bar them from going out, no one literally waits outside the door with a gun. After all, the famous Central Park jogger did go jogging in Central Park, so it is contradictory to say that she was unfree to do so.73 Women go out all the time, afraid or not, so it is absurd to say that the fear of rape inhibits them. A related position is taken by Camille Paglia; she says that in order to be free, women have to face up to the risks of adult life. Just as a man risks being mugged in New York City at night, a woman risks rape. Indeed, she suggests, risking rape is part of the deal of liberation: you cannot have one without the other.74 As Benn asserts, even if "power against" is actually used, it is not determinative of unfreedom:

If Alan has the power adversely to affect Betty's interests and successfully threatens her, he has exercised his power over her. But if the power he has against her is insufficient to support her claim that she acted under duress, her excuse that she was in Alan's power would fail. For when assessing excuses, one looks not only at Betty's preferences but at standards governing the course that she might reasonably be required and expected to choose, whatever her actual preferences.75

On a feminist reading, however, this sounds like the "reasonable man" view often drawn on in rape cases. For instance, if Alan said he had a knife but in fact did not have one, should Betty have then fought back to prevent her rape? Does the fact that she did not fight mean that, in fact, she wanted the sex and hence was not raped? Or let's say Alan and Betty were on a first date and drinking heavily; and Alan forced Betty to have sex, or she was too inebriated to resist effectvely. Do we say that Betty should not have been drinking and therefore is responsible for what happened to her? Did she "ask for it" by getting drunk in the first place, and compromising her ability to resist? If so, then is Alan not guilty of raping her, by definition? Or do we say that rape is excusable because Betty was in fact free to avoid it?76

As many theorists note, the simple existence of cost does not make one unfree: if Betty has a gun and could use it to prevent her assault, she must choose between two terrible options, shooting Alan or being assaulted, but the point is that she has a choice, can exercise a counterpower, and is thus free. And "power against" can be countered by other "powers against": self-defense courses and Take Back the Night marches can empower women; advocacy for tougher sentencing and more woman-friendly evidentiary procedures in rape trials have made prosecution more possible for women; sexual harassment policies can give women rights of redress against harassers. But the difficulty of building such counterpowers in the first place, not to mention using them effectively, suggests something more systemic about barriers to women's freedom than is captured by the requirement of assignable individual agency. Indeed, such difficulty stems precisely from the normalization of masculine power, which ensures that the dominant culture will resist women's counterpowers, and indeed depends on such resistance. The semantic focus of Benn's argument prevents us from seeing the politics of the situation: supposedly neutral standards of reasonableness are stacked against women in a systematic way because of the background condition of patriarchy that provides men with greater power.

Philip Pettit potentially offers a more feminist-friendly approach to liberty with his argument that freedom should be defined in terms of "non-domination." In contrast to Benn, Pettit holds that a person can be dominated by others without their actual interference. "Even if the others don't interfere in his or her life," Pettit maintains, "they have an arbitrary power of doing so: there are few restraints or costs to inhibit them." Such domination is a function not merely of individual behavior but of background social conditions as well; "an employee may be dominated by an employer in a tough labor market, a wife by a husband in a sexist culture, or an illegal immigrant by the citizen who gives them a job and a living."77 In fact, Pettit says, in such examples, the employer, husband, or citizen does not actually need to interfere, because those background conditions pressure the employee, wife, or immigrant to limit their own behavior. Freedom, therefore, cannot be defined as merely "noninterference," because people can dominate others without interfering; freedom also requires "nondomination."

Despite the apparent social understanding of power that his recognition of domination offers, however, and despite his acknowledgment of the restrictions such power may impose on liberty, Pettit, like the other theorists discussed here, undermines the potential of his argument by falling back on an individualistic framework. There are two particular ways in which this individualism stands out. In the first instance, Pettit posits a conscious awareness of unfreedom; in his examples, the dominated person always seems to be aware of her or his domination, and such awareness affects that person's choices and actions. Accordingly, a dominated person "will not be able to speak out in a forthright and free way--or act on a basis that such speech might justify--but must always have an eye out for what will please the powerful and keep them sweet." Hence, "The price of liberty is not eternal vigilance but eternal discretion"--a term that particularly connotes conscious effort and control.78 For instance, battered women often seek to placate the batterer by doing what he demands and saying things to soothe his anger. But of course, nonbattered women may similarly seek to please their spouses and in doing so may frequently subjugate their own desires to their husbands'. This leads to the second instance of Pettit's individualism, though it is even subtler than the first. As I mentioned previously, Pettit claims that domination can occur without any interference at all; for instance, Mrs. Bridge is not beaten, but because of social norms of masculinity and femininity, she restrains her own behavior in ways that her husband expects. But wait: how does my example of Mrs. Bridge attest to Pettit's individualism? Indeed, does it not attest to the power of patriarchy, which does its work regardless of what Mr. Bridge does?

Not in Pettit's concept of domination, which ironically pays little attention to the social forces that enable such domination. It is as individualistic as his claim that discretion is the price of liberty. For Pettit assumes that since the individual does not notice the interference, and since no particular individual is engaging in identifiably interfering behavior, then interference simply does not exist. Such an assumption is unwarranted, however. Considering domination as a social force suggests that domination is often effective precisely because the interference that power makes possible is not noticed. Indeed, the more effective social ideologies are in creating subjects who conform to restrictive norms--for instance, women who stay with their abusive partners because they believe that women are responsible for making relationships work--the less aware of interference individuals, both the dominated and the dominator, will likely be. But such lack of awareness does not mean that interference is not exercised. Even within the terms Pettit offers us, there is a certain illogic and counterintuitiveness to his assertion that domination can exist without interference. For unless there is some actual interference, some material instances of the power the dominant figure (what he calls the "dominus") has to interfere, the power of domination will weaken. If the dominus never interferes with the actions of the dominated, the vigilance and self-restriction of the dominated will wane, boundaries will be tested and stretched. If such tests are not met with resistance by the dominus, then domination can no longer be said to exist. Hence (to return to the example I used in my discussion of Benn), if all men refrained completely from assaulting women, men's "power against" women would diminish, if not vanish entirely.

The fact that some men do assault some women, however, ensures that interference occurs all the time in patriarchal society; just because an individual woman has never been raped does not mean that reading about other women's assaults in the newspaper will not "interfere" with her freedom to go out at night. Like Benn, Pettit seems committed to saying that women's fear of going out at night is a function of domination ("power against"), but not of interference ("power over"). But this misses the feminist point that the "power against" women that patriarchy gives to men constitutes a kind of interference. This interference is not individually located, nor does it even have to be expressed by any given individual man, and yet it works to restrict women both as a group and individually. Consider, for instance, that women in Afghanistan continue to wear the burqa even after the Taliban have been deposed, because they fear that violence against them may continue. Though the religious police have been removed, the possibility that others may seek to interfere with women's freedom by criticizing or assaulting them for uncovering their faces causes them to continue to wear it. The effectiveness of their domination is a result of the past interference with women's movement, and an expectation of possible future interference as well, even when the interference is not actually occurring. (As of January 2002, for instance, there have been no reports of violence against women who have removed the burqa, though many women are still wearing it because they fear the assaults will be renewed).79 The trauma of such past violence creates a framework for perception that continues to interfere in women's lives and choices even though such interference is not exercised by particular individuals (indeed, even though such individuals are presumably gone). But at the same time, such fears are not unchangeable; if the women who have removed the burqa continue to be unmolested, if women gain more social status and economic power, presumably such fear will diminish, and more women will abandon the burqa for less restrictive (and uncomfortable) forms of veiling.

Thus, we need a more complicated understanding of the relationship between domination and interference, and between social forces and individual action, than Pettit gives us. I am arguing, against Pettit, that domination always requires interference; there has to be a reason for the fear that motivates the self-vigilance of the dominated, even if the dominated is not fully aware of that fear, or of being dominated. Social norms do not come into being by themselves, and they cannot persist without people's actively calling on and deploying them. So the domination that might result from such norms cannot persist without individuals' acting within larger frameworks of cultural meaning to interfere with other individuals' self-conceptions, desires, and choices. But at the same time, the power of norms, practices, and meaning far exceeds the grasp and control of any individual, so that this power can be used and called into play without the explicit awareness of anyone, either dominus or dominated. I believe that Pettit's individualism keeps him from recognizing this feature of domination. That is, he wishes to hang onto the individualist conception of freedom, wherein interference always requires an agent who acts intentionally and purposefully. Because of that, rather than seeing that interference is often systematic and socially produced, and that individual actions take place within larger social structures that make those actions possible and give them meaning, Pettit maintains that we can have domination without interference. That is, the only way for him to claim that freedom can be restricted without intention and purpose being exercised--a claim which I obviously support--and yet retain individualism is to reject "noninterference" as the defining feature of freedom. While such a solution makes a welcome advance over some of the other theories I have critiqued here, in the end it really begs, rather than resolves, the question of freedom from a feminist perspective. It fails to provide a framework for understanding how it is that social forces like patriarchy are able to restrict women's freedom. Such a framework requires the rejection of individualism as a guiding principle.80

In a negative-liberty model, the systematic nature of power, domination, and interference could be revealed by expanding the concept of a barrier, as various theorists discussed here suggest; but this would then require a reformulation of "agency" to acknowledge the ways in which individuals operate within systems and structures and are responsible for contributions to power imbalances without necessarily intending or realizing it. Sexual violence is a good example of this problem, for it is on the one hand pervasive and systemic (it is overwhelmingly women who are targeted for rape) and on the other hand arbitrary and random (all kinds of women are raped, young and old, regardless of race or class, in a broad variety of circumstances, thus making formulaic solutions like "avoid X to avoid rape" impossible). Although rape is a specific and individual act--even when a woman is "gang raped," we tend to consider her raped repeatedly by a number of individuals, she is not seen to be raped by a collectivity, much less by patriarchy--individual rapists live and operate within a cultural context that makes rape a conceivable and even acceptable act, that in effect "enables" the action. The failure of police to arrest and courts to convict is compounded by the stigma of shame that plagues rape victims; this shame stems in part from the intimate violation of the assault, and particularly from the fact that the assaulter's pleasure comes directly from the victim's pain, disempowerment, and humiliation.81 But the acceptability of rape also stems in part from patriarchal ideas of sexuality, both men's (rape is an inevitable outcome of men's sex drive) and women's (the twin images of angel and whore, which are invoked to justify consideration of women's moral character and sexual history in judging whether to blame the victim).82 Indeed, the vanishing point of rape comes at the pinnacle of patriarchy, in that the rapist's belief that what he is doing is simply "having sex," that "normal" sexuality for him involves women's subordination, humiliation, and injury, and that women "really" want such treatment depends on and operates from the obliteration of women's human subjectivity. If women are not people, if they are only objectivized projections of male desire, if masculine perspectives dominate the account of "what actually happened," then "rape" by definition does not occur. Such views of women fail to challenge--indeed, are unable to challenge--the very existence of rape as a practice; they do not question why men rape or make that question the focus of women's unfreedom. Feminists thus seek to challenge patriarchy by claiming that rape is unacceptable, to make the streets safe for all. Without such wide-sweeping challenges, the very real barriers to women's freedom are invisible.

Thus, many "expanded" negative-liberty theories have serious flaws that may stem in part from a failure to concretize and specify the experiences of people who are inhibited by political and social conditions such as sexism, or racism, or poverty. Yet it is for this very reason that an expanded notion of external barriers is necessary to a feminist understanding of freedom. In considering gender, once we expand the notion of an external restraint we open the door to asking how external factors shape the internal self, how restraint and opportunity form and influence desire, preference, and choice. Accordingly, I will be operating from an understanding of freedom that advocates the need to see the relation between inner and outer factors of freedom. Like classic negative-liberty theorists, I maintain that the ability to make choices and act on them is the basic condition for freedom. However, like positive-liberty theorists, I maintain that choice needs to be understood in terms of the desiring subject, of her preferences, her will, and identity. For subjectivity exists in social contexts of relations, practices, policies, and institutions that affect and shape desires, will, and identity. Recognizing the inadequacies of the negative/positive typology, I maintain that it can nevertheless yield useful insights into freedom from a feminist perspective.


That, however, may well depend on what one means by "feminism." If "freedom" is a concept that is difficult to define, because of the unrecognized but implicit political battles underlying it, "feminism" is even more difficult to define, because it is the locus of explicit political and theoretical debate. I follow bell hooks and define feminism very basically as a political and philosophical devotion to ending the oppression of people on the basis of gender and sex. Hooks argues that the more common definition of feminism, as women's equality with men, traps feminism within the terms of not only masculinist discourse--because men become the neutral standard used to evaluate women's experience, thus erasing gender difference--but racist and classist discourse as well. For which "men" do such feminists want to be equal to? Poor African-American men? Or economically privileged white men? She maintains that a genuinely inclusive feminism should be defined as a struggle dedicated to "ending sexist oppression."83 Of course, what constitutes "sexist oppression" is itself in need of clarification, but people who are assigned to the social category "women" are the primary, although not the only, victims of such oppression. Hence, feminism is a political value system that has at its heart the empowerment of women to direct their lives. Though this may seem individualistic--and perhaps terminally Western--the notion of "self-direction" here depends on and presupposes the context of community: feminism as I use it here does not entail an abstract notion of choice, but rather acknowledges that all choices occur in contexts. The issue, then, turns on how power operates in various contexts. The focus on context also requires attention to other aspects of identity, such as ethnicity, culture, race, and class; since "women" belong to multiple identity categories, their "gender" and "sex" are defined in terms of these other categories as well. But what makes something "feminist" is its primary devotion to women, regardless of how that category is defined by and within different contexts.

In focusing on the relationship between internality and externality, and in considering positive and negative liberty as political and normative categories, not just philosophical and semantic ones, a feminist approach to liberty can suggest a way to hold onto the political usefulness of the differing models of self, subject, and politics, and yet in the process develop a somewhat different conceptualization of freedom that transcends the duality even as it borrows from it. Why should feminism be particularly useful to this undertaking? To begin with, despite the fact that feminists would not agree on whether the women in any of the scenarios with which I opened my discussion are free, or in what ways they are unfree, most feminists would agree that the question of freedom is extremely important, because women have been denied freedom in most societies throughout the world and throughout history. In the control of their bodies and reproduction, property rights, participation in politics, law- and policymaking, definition of moral values, family and childrearing decisions, and even the construction of language and epistemology, women have historically been restricted--not totally or uniformly by any means, but far more than males of the same race and class. This history of women's oppression thus gives feminism a particularly appropriate edge on the concept of freedom.

Furthermore, feminists should support my desire to hang onto certain aspects of the positive/negative typology, because both positive and negative liberty models inform feminist concerns. For instance, negative liberty's emphasis on individual choice is vital to issues such as reproductive freedom, sexual harassment, and employment discrimination. Feminists who have struggled against centuries of "second-guessing" by men about what it is that "women really want" are not eager to give up on the ideal of negative liberty. Similarly, positive liberty's emphasis on context and community, as well as the internal restrictions on liberty, are also important to feminist issues such as affirmative action and other means to positively provide for the conditions that enable women to exercise choices, as well as efforts to identify the effects of patriarchy on women's identity, psyche, and self-conceptions that may interfere with their ability to formulate choices in the first place.

While the dualistic typology of positive and negative liberty is useful to understanding freedom, however, it is also theoretically inadequate to deal with many questions raised by women's historical and material experience. Throughout history, theorists of both persuasions generally denied women both the "opportunities" and the "exercise" of freedom. Theorists such as Hobbes and Locke, commonly seen as negative libertarians, barred women from public life on the basis of their "natural" inferiority, which meant that they were incapable of rational choice. Women's diminished humanity disqualified them from taking advantage of the "opportunities" of liberal society and required that they be ruled by men. Indeed, women's restraint in the private sphere was one of the things that made negative liberty in the public sphere possible for men. That is, freedom could be defined as abstract choice for men only because women were bound to the aspects of life that are not necessarily chosen.84 Even Mill, in spite of his remarks against women's subjection, limited women's freedom more than men's by failing to challenge the structural barriers to women's choice. He sometimes extended the notion of what counted as a barrier to include social customs and norms but at other times treated such customs and norms as natural and self-evident. Like Wollstonecraft, Mill suggested that women learn to be irrational by being denied education and trained to engage in trivial pursuits; yet he also seemed to believe that the social role of motherhood and wifehood was naturally ordained.85 Positive libertarians such as Rousseau, Kant, and Hegel denied women's rationality, requiring them to adopt very particularized and structured roles within the family as a means of guaranteeing the stability of the state. All people could be free only by following their true will, but women were too emotional to know what their true will was. Furthermore, their irrationality confused men and impeded their ability to know their true will. So women could "exercise" their greatest freedom only by being restrained in the private realm and allowing men to act for them in the public realm.86 Yet in both positive- and negative-liberty theories, despite these exclusions, freedom is seen as natural and universal, thus suggesting the virtual erasure of women from the concept of freedom.

In this light, women's experiences provide a powerful basis for highlighting the sexism frequently found in liberty theory, precisely because these experiences often lie at the crossroads of the Enlightenment ideology of agency and choice and modern practices that systematically restrict women. That is why I conduct my analysis of freedom not simply through standard works in political philosophy (which I discuss in chapter 2), but through specific material experiences of women. Such an approach offers a concrete understanding of patriarchal contexts and the paradoxical relation between empowerment and restraint. Accordingly, in chapters 4, 5, and 6 I examine the experiences of domestic violence, welfare, and Islamic veiling to apply the theory of social construction to the question of freedom. These three experiences are certainly not the only ones which can be seen to restrict women's freedom, limit their opportunities and choices, and compromise their self-definitions and autonomy. But by virtue of their familiarity in everyday discourse, I hope to raise questions not normally considered in discussions of the freedom or unfreedom of women. For instance, while a narrow reading of negative liberty such as Parent or Flathman offers would say that Susan, the battered woman discussed at the beginning of this chapter, is free to leave if she is not actually being restrained at the time her decision is made, positive libertarians would say she is unfree, because she does not understand what is in her best interest--and both of these responses dominate the common public view of domestic violence, which alternates between scorn and pity. This merely perpetuates the internal/external divide, however; the possibility that battered women actually express agency and choice every day, but do so within severely restrained contexts that in turn affect their understanding of their options as well as their preferences, is not generally recognized. By expanding the notion of a barrier, we may be prompted to look more carefully at the social context in which battering occurs: the (un)responsiveness of state agencies such as police and courts, resources (un)available to women who wish to leave abusive partners, the impact of psychological abuse on desire, and beliefs about the naturalness or normality of male violence.

Similarly, it is commonly asserted that the external conditions of welfare have produced a mind-set of dependency such that women who could otherwise be productive members of society are instead drains on taxpaying citizens. This situation, it is argued, is bad not just for taxpayers but for welfare recipients as well, for in order to be autonomous, one must be economically self-sufficient. What these critics fail to acknowledge is that welfare makes it possible for women to bring pregnancies to term and to raise children without dependency on individual men. At the same time, liberal democrats and many feminists have sought to defend the welfare state from Republican attacks by attempting to present poor women as downtrodden victims deserving of middle-class sympathy. Not only does this perpetuate the disempowerment of women on welfare and the view of welfare as charitable dispensation rather than a right but it also fails to recognize the fact that welfare forces women to depend on a paternalistic state. Both sides of the argument get a piece of the puzzle right but miss the larger picture, which turns on the ways in which the state both forms a backdrop for individual choice and action and yet constructs individuals to make particular choices through an elaborate system of punishment and reward. Such ambiguity is also seen in the practice of Islamic veiling. Although many Westerners view veiling as positive proof of women's oppression under Islam, in fact many women choose to wear the veil as an expression of cultural identity and resistance to Western imperialism. Moreover, women use the veil to negotiate and resist patriarchal customs and norms, such as bars to women's paid employment. Yet veiling also exists within the parameters of patriarchy, such that women's choosing it simultaneously expresses their free agency and reinscribes the terms of their oppression.

Considering these problems in light of the social construction argument articulated in chapter 3 makes a more complex reading of freedom possible. The different configurations of choice found in these various scenarios derive from, and in turn suggest, differing levels of power, control, and coercion stemming from a variety of contextual factors. If, returning to my original definition, freedom consists in the power of the self to make choices and act on them, but the self that makes those choices, including her desires and self-understanding, is socially constructed, then to analyze freedom theorists must examine specific concrete situations in which that construction takes place. Viewing freedom as a political question requires applying general theoretical conceptions not just to specific events--a sexual assault, losing a job to a less qualified male, being forbidden by religious edict to leave the house, blocked entry to an abortion clinic, coerced sex with your boss--but to the broader contexts that construct both the events and the individuals who participate in them. Women are obviously a diverse group, though perhaps also one group among many. Despite their differences, however, they are nevertheless oppressed throughout the world to varying degrees, ranging from the minimal to the horrific, but always systematically. By locating the problem of freedom in the lived experience of women, and by seeing these experiences as variably situated in contexts that may share certain elements of male power but also differ significantly from one another, the complexity of freedom becomes particularly apparent. It is true that when we pit women's real-life situations against abstract and often abstruse theoretical formulations, existing conceptions of freedom are revealed to be gender biased and inadequate for understanding women's lesser freedom, or even unfreedom. But I do not explore domestic violence, welfare, veiling, and other material experiences simply to establish that women are less free than men. Rather, the goal is to explore how the social and personal construction of gender through concrete experiences influences theoretical perceptions and representations of the world, and how these representations in turn create experience. The specific political issues and concrete conditions of women's lives, such as battering, welfare, veiling, heterosexism, sexual harassment, rape, or abortion, are less important in defining "the problem of freedom" than they are in providing touchstones for understanding what is problematic about freedom.

What this suggests is that a feminist conception of freedom requires a detailed political analysis of patriarchal power in the particular contexts in which "freedom" is in question.87 Such an analysis would include the conditions in which women live, the way power is utilized and executed by individual men, social institutions, and state agencies, and the effects of these conditions and this power on both the options that are materially available to women and their subjective understandings of those options. Individual women's desires, the choices they make, and the reasons for those choices--both the reasons they understand themselves to have and the reasons they may not be able to see, precisely because desire has been constrained and produced in particular ways--are all important to a feminist understanding of freedom. Recognizing that all choices, and our very subjectivity as choice makers, exist in contexts requires that we be self-conscious and critically aware of the contexts we live in, draw on, and utilize in making our choices, but such awareness is in circular fashion conditioned by the social formations that have constructed us. Hence, in chapter 7 I attempt to explore how to escape, or at least subdue, this apparent circularity so that awareness may expand and positive reconstruction proceed. In the process, I show how the feminist analysis of patriarchal power and contexts I have provided in the foregoing chapters is not only deconstructive but productive as well, not just critique but a positive vision of alternative possibility.


Before turning to that analysis, one question remains, and ironically (given my earlier criticisms of the semantic orientation of many freedom theorists) it brings us back to questions of definition. My criticisms of contemporary freedom theory, combined with my concern with the internal aspects of freedom, suggest the importance of autonomy to my conception of freedom. Cass Sunstein suggests that freedom and autonomy are somewhat interchangeable.88 In particular, the traditional conception of autonomy as self-rule overlaps a great deal with both negative and positive conceptions of liberty. It shares the extreme individualism of negative liberty, defining itself in terms of absolute freedom from any influence of others,89 and at the same time the extreme sense of self-control that often characterizes positive liberty, the notion of a rational, reflective, self-ruling agent who is concerned not just to choose but to make the best choice, the right choice. As Emily Gill puts it, autonomy involves "the capacity to govern oneself . . . the freedom to pursue what one judges to be good but also the ability to define this good in one's own manner."90 At the same time, several autonomy theorists explicitly identify positive liberty with autonomy; John Christman calls autonomy positive liberty's "identical twin," and Joseph Raz similarly allies autonomy with positive liberty.91 Many other autonomy theorists, though not explicitly asserting such alliances, nevertheless develop similar accounts of autonomy that focus on "the inner self," and control over the self, which echo central positive-liberty themes. To the extent that I maintain that social constructivism is suggested by positive liberty and urge the importance of the internal to a conception of freedom, this might prompt the question of why this book is devoted to freedom rather than autonomy.

It is certainly true that there are many shared concerns and features of autonomy and the conception of freedom that I develop here. Autonomy is fundamentally about capabilities, specifically about the ability to assess one's options, reflect critically about them, and make choices that allow one to exert some control over one's life. Thus, both freedom as I conceptualize it and autonomy are concerned with the inner workings of desire and will; both allow for the possibility of differentially evaluating desires and declaring some desires to be more consistent with autonomy or freedom than others; and both, hence, are forced to confront the paradox of "second-guessing." Additionally, because of the focus on such internal elements, both autonomy and my approach to freedom turn on the conception of self: a theorist's conception of "who we are" will be central to her conception of what selves need for freedom or autonomy, what conditions define the parameters of possibility for its achievement. And some recent work on autonomy, particularly in the feminist vein, has argued for a "social self," whose capacity for self-understanding must be understood in the context of community, family, and other particular and personal relationships.92

But there are also important differences between the two concepts, and these lead me to retain freedom as my subject. Some theorists argue that one cannot be free unless one is autonomous; unless one knows what one's true desire is, one cannot pursue it. For instance, Wright Neely maintains that "to be free . . . means there is an absence of restraints (positive or negative, internal or external) standing between a person and the carrying out of that person's autonomous desires."93 Others maintain the reverse, that freedom to do what one wants is a precondition of autonomy, of figuring out what it is that one wants to do. "Freedom is necessary for autonomy," says Robert Young; "that is what freedom is for," argues Jennifer Nedelsky, ''the exercise of" autonomy.94 Freedom is thus both broader and narrower than autonomy. According to some theorists, such as Flathman, freedom is narrower, because it refers only to actions and is therefore event-specific; claims of freedom cannot be made about generalized conditions, only about specific actions.95 An "unfree person" is unfree because she is prevented from doing many things, but it is those specific blockages that define unfreedom. As MacCallum argued, "[F]reedom is thus always of something (an agent or agents), from something, to do, not do, become, or not become something."96 By contrast, autonomy is a characteristic of persons and thus must be considered historically over the life span of an individual.97 As Dworkin puts it, "freedom is a local concept, autonomy a global one."98 At the same time, other theorists maintain that freedom is broader than autonomy, because freedom is seen to cover a wider array of possibilities, ranging from the absurd (freedom to eat chocolate versus vanilla ice cream) to the most serious (freedom from incarceration for criticizing the government). By contrast, autonomy is reserved for consideration of choices that really matter to the individual and her ability to exercise control over her life.99 Additionally, because freedom is concerned with external blockages to action, it enters at many more access points and involves a much wider range of social relations and institutions than does autonomy, which, being concerned with the inner workings of the will, focuses primarily on the individual in whatever context she happens to find herself.

Indeed, the issue of internality demarcates the most important distinction between freedom and autonomy. Although the absence of coercion and severe restraint is necessary to acting autonomously, autonomy itself is generally conceived in terms that refer to the inner self, the psychological, rational, moral, and emotional capabilities of the individual. Friedman defines it as "acting and living according to one's own choices, values, and identity within the constraints of what one regards as morally permissible." Nedelsky says that autonomy means "we feel that we are following an inner direction rather than merely responding to the pushes and pulls of our environment." Dworkin defines autonomy as a "capacity . . . to reflect critically upon . . . preferences, desires, wishes," an idea Raz echoes when he calls autonomy "a kind of achievement . . . a capacity." Autonomy particularly involves "reflective judgment," "the critical evaluation of desires," "self-discovery, self-definition, and self-direction," "living life from the inside."100

Because of this focus on internal aspects, autonomy arguments frequently encode value judgments about desire into their arguments. Distinctions between true and false, higher and lower, and first- and second-order, authentic and inauthentic desires, as well as frequent references to the "true self," the "authentic self," or the "unified self," pepper autonomy writings, even feminist ones.101 In this, they share a great deal with classic positive-liberty theory, the variants of it that Berlin most feared and which I agree are troubling, the strong "second-guessing" for which theorists like Rousseau are excoriated. So-called procedural accounts of autonomy such as Dworkin and Meyers put forth claim to obviate this dilemma. They maintain that autonomy says nothing about what choices we make but only requires that we are able to, and actually do, reflect critically on our choices. Hence, there can be no value judgments about the specific things I choose, only observational judgments about whether I have reached those choices after adequate reflection. But even these procedural criteria are extremely value laden, echoing the essence of Enlightenment rationality. Meyers, for instance, says that one must not merely be able to offer reasons; one must be guided in formulating those reasons by "firm goals or moral views" rather than "feelings, intuitions, and arguments of the moment." For autonomy "expresses the true self."102 Dworkin's criteria are somewhat weaker, requiring only that we be able to have reasons for our actions with which we identify; but at the same time "the conditions of procedural independence involve distinguishing those ways of influencing people's reflective and critical faculties which subvert them from those which promote and improve them." The question this begs, of course, is who decides that; Dworkin implies that this is somewhat self-evident, but only because he identifies "hypnotic suggestions, manipulation, coercive persuasion, subliminal influence" and the like as the obvious candidates for "subversion."103 Indeed, to the extent that most autonomy theorists acknowledge the impact of external influences at all, they are generally described in the same terms of purposeful and intentional agency that freedom theory relies on: manipulation, coercion, oppression, brainwashing. The more subtle influences of social construction, and the ways in which restraint is intrinsically intertwined with production, are not generally acknowledged.

These conceptions of the self, though they may adhere to a feminist conception of self-in-relationship, nevertheless pose problems for feminist theory. For instance, consider, in terms of a "true self," the woman discussed at the start of this chapter, who was trying to get an abortion. External barriers prevented Meg from acting on her preference; as a result, she decided to adjust her life plan to adapt to the reality of these constraints. Here there is certainly an effort to be autonomous, to exert her own presence and individuality on circumstances created for her; Meg has critically evaluated her remaining options and has committed herself to loving her child. Yet her self, her life, follows a different path than the one she intended; she wanted to abort the fetus and was prevented from doing so. But if she achieves success in loving her baby, then which "self" is the true one: the one who was thwarted in obtaining the abortion, or the one who now loves and delights in her baby and is glad that in the end she had it? It is difficult not to say that she is autonomous; yet we can readily make the claim that she was unfree, if we redefine subjectivity and barrier beyond the confines of existing freedom theory. Or consider Young's claim that deep conflict over which projects to pursue means that one is not autonomous, that autonomy requires a "unified self," in the context of sexual assault. Yet, as Susan Brison has argued, in the face of traumatic memory spurred by sexual assault, autonomy may in fact require us to fragment the self, to compartmentalize and avoid aspects of ourselves that would otherwise immobilize and incapacitate us from acting in the world, forging life projects, and moving ahead with life.104

As will be argued more fully in chapter 3, the idea of social construction challenges the possibility of an essential "inside," which seems so vital to autonomy theory, and demands that determinations of freedom must consider internality and externality together. Meyers notes that "Since one must exercise control over one's life to be autonomous, autonomy is something that a person accomplishes, not something that happens to persons."105 But freedom is precisely a combination of self-creation and what happens to you, the internal as well as the external, the combination of and dynamic between the two. If freedom is concerned with the capacity to choose, then social construction requires us to think about the broader conditions in which choices are made. Even if we can say that someone is "manipulated" into making a particular choice, we need to consider the social conditions that made such manipulation possible and that made this choice appear to make sense.

I will return to some of these points in chapter 3, but it is sufficient for now to note that there are significant differences between my approach to freedom and most theories of autonomy, even feminist ones. This is not to deny that feminist arguments about autonomy make valuable contributions to a feminist approach to freedom and that I draw on them here. Most importantly, they have brought attention to the need to recon-figure autonomy rather than reject it. As with freedom, the critique of mainstream accounts of autonomy as masculinist could yield the conclusion--one which some feminists have drawn--that feminists should reject the concept of autonomy altogether. By recognizing that the problem with autonomy is not the concept itself, but rather the conception of it that follows from an individualist and naturalist construction of the self, feminists have been able to begin the kind of rethinking that I here suggest is also necessary to the concept of freedom. But while the areas of concern overlap, freedom prompts somewhat different questions than autonomy. Indeed, freedom is, rather appropriately, less conceptually constrained. I agree with those who argue that freedom is a precondition for autonomy, not the other way around: if there is such a thing as a "true" self, or "authentic" desire, then in order to determine what that might be, women must be freed from the multiple, intersecting, and overarching barriers that pervade patriarchal society. In this, I seek to push freedom further than autonomy seems able to go.

Return to Book Description

File created: 8/7/2007

Questions and comments to:
Princeton University Press

New Book E-mails
New In Print
PUP Blog
Princeton APPS
Sample Chapters
Princeton Legacy Library
Exam/Desk Copy
Recent Awards
Princeton Shorts
Freshman Reading
PUP Europe
About Us
Contact Us
PUP Home

Bookmark and Share