VELTMAN, Andrea. Autonomy, Oppression, and Gender

Autonomy, Oppression, andGender
Studies in Feminist Philosophy is designed to showcase cutting-edge monographs and collections that
display the full range of feminist approaches to philosophy, that push feminist thought in important new
directions, and that display the outstanding quality of feminist philosophical thought.
STUDIES IN FEMINIST PHILOSOPHY

Cheshire Calhoun, SeriesEditor
AdvisoryBoard
Harry Brod, University of NorthernIowa
Claudia Card, University of Wisconsin
Lorraine Code, York University, Toronto
Kimberle Crenshaw, Columbia Law School/UCLA SchoolofLaw
Jane Flax, Howard University
Ann Garry, California State University, Los Angeles
Sally Haslanger, Massachusetts Institute of Technology
Alison Jaggar, University of Colorado, Boulder
Helen Longino, Stanford University
Maria Lugones, SUNY Binghamton
Uma Narayan, Vassar College
James Sterba, University of NotreDame
Rosemarie Tong, University of North Carolina, Charlotte
Nancy Tuana, Penn State University
Karen Warren, Macalester College
Recently Published in the Series:

Visible Identities:Race, Gender, and theSelf The Metaphysics ofGender
Linda MartnAlcoff CharlotteWitt
Women and Citizenship Unpopular Privacy:What Must WeHide?
Edited by Marilyn Friedman Anita L.Allen
Womens Liberation and the Sublime:Feminism, Adaptive Preferences and Womens Empowerment
Postmodernism, Environment SereneKhader
BonnieMann Minimizing Marriage:Marriage, Morality, and
Analyzing Oppression theLaw
Ann E.Cudd ElizabethBrake
Ecological Thinking:The Politics of Epistemic Out from the Shadows:Analytic Feminist
Location Contributions to Traditional Philosophy
LorraineCode Edited by Sharon L. Crasnow and Anita M.
Self Transformations:Foucault, Ethics, and Superson
NormalizedBodies The Epistemology of Resistance:Gender and
Cressida J.Heyes Racial Oppression, Epistemic Injustice, and
Family Bonds:Genealogies of Race andGender Resistant Imaginations
Ellen K.Feder JosMedina
Moral Understandings:AFeminist Study in Simone de Beauvoir and the Politics of Ambiguity
Ethics, Second Edition SoniaKruks
Margaret UrbanWalker Identities and Freedom:Feminist Theory Between
The Moral Skeptic Power and Connection
Anita M. Superson AllisonWeir
Youve Changed:Sex Reassignment and Vulnerability:New Essays in Ethics and Feminist
Personal Identity Philosophy
Edited by Laurie J.Shrage Edited by Catriona Mackenzie, Wendy Rogers,
and SusanDodds
Dancing with Iris:The Philosophy of Iris
MarionYoung Sovereign Masculinity:Gender Lessons from the
Edited by Ann Ferguson and MechthildNagel War onTerror
BonnieMann
Philosophy of Science after Feminism
Janet A. Kourany Autonomy, Oppression, andGender
Edited by Andrea Veltman and MarkPiper
Shifting Ground:Knowledge and Reality,
Transgression and Trustworthiness
Naomi Scheman
Autonomy, Oppression, andGender
Edited by Andrea Veltman
and
MarkPiper
1
3
Oxford University Press is a department of the Universityof Oxford.
It furthers the Universitys objective of excellence in research, scholarship,
and education by publishing worldwide.
OxfordNewYork
Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala LumpurMadridMelbourneMexico CityNairobi
New DelhiShanghaiTaipeiToronto
With officesin
ArgentinaAustriaBrazilChileCzech RepublicFranceGreece
GuatemalaHungaryItalyJapanPolandPortugalSingapore
South KoreaSwitzerlandThailandTurkeyUkraineVietnam
Oxford is a registered trademark of Oxford UniversityPress

in the UK and certain other countries.
Published in the United States of Americaby

Oxford UniversityPress
198 Madison Avenue, NewYork, NY10016
Oxford University Press2014
All rights reserved. No part of this publication may be reproduced, storedin a

retrieval system, or transmitted, in any form or by any means, without theprior
permission in writing of Oxford University Press, or as expressly permitted bylaw,
by license, or under terms agreed with the appropriate reproduction rights organization.
Inquiries concerning reproduction outside the scope of the above should be senttothe
Rights Department, Oxford University Press, at the addressabove.
You must not circulate this work in any otherform

and you must impose this same condition on any acquirer.
Library of Congress Cataloging-in-PublicationData

Autonomy, oppression, and gender / edited by Andrea Veltman and Mark Piper.
pages cm.(Studies in feminist philosophy)
Includes bibliographical references.
ISBN 9780199969104 (hardcover : alk. paper)
ISBN 9780199969111 (pbk. : alk. paper)
1. Autonomy (Philosophy) 2. Feminist theory. I. Veltman, Andrea.
B808.67.A985 2014
126dc23
2013044775
135798642
Printed in the United States of America
on acid-freepaper
{ Contents}
Notes on Contributors vii

1. Introduction 1
Andrea Veltman and MarkPiper
Part I Autonomy and Independence

2. Three Dimensions of Autonomy:ARelational Analysis 15
Catriona Mackenzie
3. Relational Autonomy and Independence 42
Marilyn Friedman
4. Autonomy? Or Freedom? AReturn to Psychoanalytic Theory 61
Nancy J. Hirschmann
Part II Autonomy and Normative Commitments

5. Feminist Commitments and Relational Autonomy 87
PaulBenson
6. The Feminist Debate over Values in Autonomy Theory 114
Diana TietjensMeyers
7. A Commitment to Autonomy Is a Commitment to Feminism 141
MarinaOshana
Part III Autonomy, Reasons, and Care

8. Emotions, Reasons, and Autonomy 163
Christine Tappolet
9. Autonomy and Self-Care 181
Andrea C. Westlund
Part IV Autonomy, Oppression, and Adaptive Preferences

10. Coping or Oppression:Autonomy and Adaptation to
Circumstance 201
John Christman
11. Autonomy and Adaptive Preference Formation 227
Natalie Stoljar
viContents
Part V Autonomy in Social Contexts

12. Raising Daughters:Autonomy, Feminism, and Gender
Socialization 255
MarkPiper
13. Autonomy and Oppression at Work 280
Andrea Veltman
14. The Right to Bodily Autonomy and the Abortion Controversy 301
Anita M. Superson
15. Choosing Death:Autonomy and Ableism 326
AnitaHo
Index 351
{ Notes on Contributors}
PaulBenson is professor of philosophy and dean of the College of Arts and

Sciences at the University of Dayton. He works in the areas of moral psy-
chology, action theory, and social philosophy. He has published on autonomy,
free agency, oppressive socialization, and moral responsibility in journals
such as Journal of Philosophy, Hypatia, Canadian Journal of Philosophy,
Philosophical Studies, and Journal of Social Philosophy.He has also published
in edited volumes on autonomy including Autonomy and the Challenges
to Liberalism (Cambridge, 2005, ed. Christman and Anderson), Personal
Autonomy (Cambridge, 2005, ed. Taylor), Moral Psychology: Feminist Ethics
and Social Theory (Rowman & Littlefield, 2004, ed.DesAutels and Walker),
and Relational Autonomy (Oxford, 2000, ed. Mackenzie and Stoljar).
John Christman is professor of philosophy, political science, and women's

studies at Pennsylvania State University. His is the author of various essays
and books, including The Myth of Property:Toward an Egalitarian Theory of
Ownership (Oxford, 1994), Social and Political Philosophy: A Contemporary
Introduction (Routledge, 2002), and most recently The Politics of
Persons:Individual Autonomy and Socio-historical Selves (Cambridge, 2009).
He is editor of The Inner Citadel: Essays on Individual Autonomy (Oxford,
1989)and Autonomy and the Challenges to Liberalism:New Essays (Cambridge,
2005, with Joel Anderson), and Contemporary Debates in Political Philosophy
(Basil Blackwell, 2009, with Thomas Christiano).
Marilyn Friedman is W.Alton Jones Professor of Philosophy and professor

of political science at Vanderbilt University. She works in the areas of political
philosophy, feminist theory, and ethics and has published extensively in those
areas. Her authored books include Autonomy, Gender, Politics (Oxford, 2003),
and her edited collections include Women and Citizenship (Oxford, 2005). She
has also published articles in journals such as Ethics, Journal of Philosophy, and
Hypatia, as well as numerous book chapters in edited collections.
Nancy J. Hirschmann is professor of political science at the University of

Pennsylvania. Her books include The Subject of Liberty:Toward a Feminist Theory
of Freedom (Princeton University Press, 2003), which won the 2004 Victoria
Schuck Award for the best book on women and politics from the American
viii Notes on Contributors
Political Science Association; Gender, Class and Freedom in Modern Political

Theory (Princeton University Press, 2008); andRethinking Obligation:AFeminist
Method for Political Theory (Cornell University Press, 1992). She is also co-editor
of several volumes including Revisioning the Political:Feminist Reconstructions
of Traditional Concepts in Western Political Theory (Westview Press, 1996)and
Feminist Interpretations of Thomas Hobbes (University of Pennsylvania Press,
2013). She is also the author of many articles and has held fellowships from
the National Endowment for the Humanities, Institute for Advanced Study,
American Council of Learned Societies, and Princeton University Center for
Human Values.
AnitaHo is associate professor at the W.Maurice Young Centre for Applied

Ethics at the University of British Columbia and director of ethics services
at Providence Health Care. She specializes in bioethics, social/political phi-
losophy, and disability studies she has published on themes such as autonomy,
informed consent, disability, and trust in journals including American Journal
of Bioethics, Journal of Medical Ethics, and Journal of Bioethical Inquiry. She
has also co-edited an anthology titled The Reflective Woman (Coply, 2002)and
an issue of Teaching Philosophy (Special Issue on Disability in the Classroom,
2007, with Anita Silvers). She is currently working on a project on trust and au-
tonomy in clinical and research medicine, supported by a research grant from
the Social Sciences and Humanities Research Council of Canada, and another
project on supportive decision making in health care, funded by the Canadian
Institutes of Health Research.
Catriona Mackenzie is associate dean (research), professor of philosophy, and

director of the Research Centre for Agency, Values and Ethics at Macquarie
University, Sydney. She is co-editor of several volumes, including Relational
Autonomy: Feminist Perspectives on Autonomy, Agency, and the Social Self
(Oxford, 2000, with Natalie Stoljar), Practical Identity and Narrative Agency
(Routledge, 2008, with Kim Atkins), and Vulnerability:New Essays in Ethics
and Feminist Philosophy (Oxford, 2013, with Wendy Rogers and Susan Dodds).
Mackenzie has published widely in moral psychology, ethics, applied ethics,
and feminist philosophy in a variety of edited collections and in journals
including Australasian Journal of Philosophy, Hypatia, Journal of Applied
Philosophy, Journal of Social Philosophy, and Philosophical Explorations.
Diana Tietjens Meyers is emerita professor of philosophy at the University

of Connecticut. She has held the Ellacura Chair of Social Ethics at Loyola
University, Chicago, and the Laurie Chair in Womens and Gender Studies
at Rutgers University. She works in four main areas of philosophyphiloso-
phy of action, feminist ethics and aesthetics, and human rights theory. Her
Notes on Contributors ix
monographs are Inalienable Rights: A Defense (Columbia University Press,

1985), Self, Society, and Personal Choice (Columbia University Press, 1989),
Subjection and Subjectivity: Psychoanalytic Feminism and Moral Philosophy
(Routledge, 1994), and Gender in the Mirror:Cultural Imagery and Womens
Agency (Oxford University Press, 2002). Being Yourself: Essays on Identity,
Action, and Social Life (Rowman & Littlefield, 2004)collects some of her pre-
viously published essays as well as presenting one new one. She is at work on a
new monograph, Victims Stories and the Advancement of Human Rights, and a
new edited collection, Poverty, Agency, and Human Rights (Oxford University
Press,2014).
Marina Oshana is professor of philosophy at the University of California,

Davis. Her research focuses on issues in personal autonomy, responsible
agency, and self-identity. She teaches classes in normative ethics, moral psy-
chology, philosophy of law, political philosophy, and feminism. Her publica-
tions include Personal Autonomy in Society (Ashgate, 2006)and The Importance
of How We See Ourselves: Self-Identity and Responsible Agency (Rowman &
Littlefield,2010).
MarkPiper is assistant professor of philosophy at James Madison University.

His principal research interests are in normative ethical theory, with a special
concentration on the topics of autonomy and well-being. He has published
articles and book chapters on autonomy, virtue ethics, well-being, justice, and
the value of education in venues such as Journal of Value Inquiry, International
Journal of Philosophical Studies, and Southwest Philosophical Review. He is
also the author of Autonomy: Normative in the Internet Encyclopedia of
Philosophy. He is currently working on a monograph on the principle of re-
spect for autonomy.
Natalie Stoljar is associate professor in the Department of Philosophy and the

Institute for Health and Social Policy at McGill University. She is co-editor of
the collection Relational Autonomy:Feminist Perspectives on Autonomy, Agency
and the Social Self (Oxford University Press, 2000, with Catriona Mackenzie).
She is the author of many articles on the philosophy of law, feminist philoso-
phy, and autonomy, in journals and edited collections including Philosophical
Topics, Legal Theory, and Journal of Political Philosophy. She is working on a
book on self-government under conditions of oppression.
Anita M. Superson is professor of philosophy at the University of Kentucky.

She specializes in ethics, including metaethics, moral psychology, norma-
tive and applied ethics, and feminism. Her recent books include The Moral
Skeptic (Oxford University Press, 2009)and Out from the Shadows:Analytical
x Notes on Contributors
Feminist Contributions to Traditional Philosophy (Oxford University Press,

2012, co-edited with Sharon Crasnow). She is working on a book on bodily
autonomy.
Christine Tappolet is full professor in the Department of Philosophy at the

Universit de Montral. Her research interests lie mainly in metaethics, nor-
mative ethics, moral psychology, and emotion theory. She has edited a number
of volumes, including Weakness of Will and Practical Irrationality (Oxford
University Press, 2003, with Sarah Stroud), The Modularity of Emotions
(Canadian Journal of Philosophy, 2008, with Luc Faucher). She is the author of
numerous articles on themes such as values, normativity, weakness of will, pro-
crastination, autonomy, and emotions and of two books, motions et valeurs
(Presses Universitaires de France, 2000)and Les concepts de lthique. Faut-il
tre consquentialiste? (Hermann diteurs, 2008, with Ruwen Ogien). She is
currently working on a book manuscript titled Emotions, Values, and Agency.
Andrea Veltman is associate professor of philosophy at James Madison

University. She has edited Social and Political Philosophy (Oxford University
Press, 2008), Oppression and Moral Agency (Special Issue of Hypatia, 2009,
with Kathryn Norlock), and Evil, Political Violence and Forgiveness (Rowman
& Littlefield, 2009, with Kathryn Norlock). She is also author of articles on
feminist ethics and on Simone de Beauvoir, Hannah Arendt, Jean-Paul Sartre,
Kant, Plato, and Aristotle in journals such as Hypatia and Journal of Value
Inquiry. She is currently working on a monograph on meaningful work and
human flourishing.
Andrea C. Westlund is associate professor of philosophy and womens

studies at the University of Wisconsin, Milwaukee. She works in ethics,
moral psychology, and feminist philosophy and has published in journals
including Hypatia, Monist, Philosophers Imprint, Philosophical Studies, and
Philosophical Review. Her research focuses on relational autonomy, authority,
and self-regarding attitudes, and she is currently working on a series of papers
on narrative, meaning, and theself.
{1}
Introduction
Andrea Veltman and MarkPiper
If feminism is a response to the oppression of women, and if resistance

and emancipation include living according to ones own lights, then au-
tonomy is central to issues in feminist philosophy. Classically defined as
self-determination, autonomy includes the ability to shape our own lives and
to live authentically rather than being directed by external forces that manip-
ulate or distort us.1 Some influential accounts define autonomy as requiring a
process of critical self-reflection, whereas others emphasize several agential
competencies, values, or self-regarding attitudes. Still others argue that au-
tonomy requires control over ones circumstances, a range of options that one
can hope to achieve in the development of her life, and a lack of severe con-
straint, coercion, or subordination in which one would be subject to the dic-
tates of others.2 Each of these kinds of accounts of autonomy can recognize the
social and relational character of human agency, and each can acknowledge
that autonomous abilities can be undermined by severely oppressive social
forces, for instance by stifling the development of critical intellectual faculties
or by blocking life options among the oppressed.
Autonomy provides not only an emancipatory ideal for those who cope
with systemic abuse, degradation, domination, or other forms of oppression
but also a lens for illuminating philosophical issues surrounding womens
desires, choices, and identities. Feminist philosophers working in this area
ponder, for instance, whether women can freely or authentically accept condi-
tions that support their own oppression. Should we give credence to reflec-
tively endorsed desires and choices that are the result of socially subordinate
positions? Is the pursuit of desires that issue from patriarchal norms consistent
with autonomous agency? What do we say about women who are willingly
self-abnegating or wholly deferential to the interests of others? An analysis
1
As characterized by John Christman, Autonomy in Moral and Political Philosophy, in The
Stanford Encyclopedia of Philosophy, August2009.
2
Marina Oshana, Personal Autonomy in Society, Journal of Social Philosophy 29:1 (Spring
1998):81-102; Oshana, Personal Autonomy in Society (Aldershot, UK:Ashgate,2006).
2 Autonomy, Oppression, and Gender
of autonomy serves crucially in illuminating these and related questions,

informing evaluations of women who adopt symbols of gender oppression,
who define themselves through unequal personal relationships, or who harm
themselves or others in conforming to culturalnorms.
Autonomy plays an important role not only in feminism but also, more
broadly, in ethical theory, applied ethics, political philosophy, and the phi-
losophy of education. In the area of ethical theory, it has been argued that au-
tonomy is integral in living wellthat is, that autonomy is one primary good
among others that a person needs to lead a good life or to achieve human
flourishing. Autonomy also supports such basic human values as dignity, re-
spect, truthfulness, and moral responsibility:in the vein of Kant, mature and
rational human beings are seen as free and responsible moral agents in virtue
of our capacities to control ourselves through the exercise of our autonomous
wills. In applied ethics, autonomy informs ever-bourgeoning debates on is-
sues surrounding, for example, abortion, birth, physician-assisted suicide,
and same-sex marriage. Aprinciple of respect for autonomy also lies at the
core of liberal democracies, and political philosophers often invoke autonomy
in evaluating social and political principles and political power as well as in
grounding individual rights or in criticizing paternalistic policies or practices.
Since enhancing autonomy ranks among the most important goals of a free
society, some also argue that promoting autonomy is among the most impor-
tant goals of a liberal education.3 These branches of philosophical interest in
autonomy intertwine with feminist work on autonomy, as issues involving
gender and oppression deeply permeate ethical and political philosophy.
Insofar as liberal democracies value individual autonomy, ideals of au-
tonomy provide norms for critiquing oppressive practices that stifle agency
and limit opportunities.If living autonomously requires an agent to have a
significant array of opportunities to act in ways that reflect what deeply matters
to her, as Marilyn Friedman writes, then social conditions should not so limit
her options that she cannot choose or act for the sake of any of her deep values
and commitments.4 Oppression not only limits opportunities and life options,
thus preventing an oppressed person from acting autonomously in ways that
reflect her values and commitments, but also deforms desires and infects the
conditions under which growing persons are socialized.5 Oppressive sociali-
zation can damage a persons concern for herself and stifle the development of
cognitive capacities, such as those employed in self-reflection or the critical
appraisal of socialnorms.
3
For more on the importance of autonomy in normative philosophy, see Mark Piper,
Autonomy:Normative, in the Internet Encyclopedia of Philosophy, December2010.
4
Marilyn Friedman, Autonomy, Gender, Politics (NewYork:Oxford University Press, 2003),18.
5
Friedman, Autonomy, Gender, Politics,19.
Introduction 3
Classically defined as a weighing down or as harsh dominion, oppression is

characterized in contemporary feminist philosophy as structural or systemic
in nature.In her landmark feminist analysis of oppression, Marilyn Frye writes
that it encompasses a system of interrelated barriers and forces which reduce,
immobilize and mold people who belong to a certain group, and effect their
subordination to another group.6 Others add that oppression presents mul-
tiple faces, including marginalization, exploitation, and powerlessness, and
extends beyond economic and political forces to include psychological bar-
riers that reduce, limit, or mold people as members of certain groups.7 Ann
Cudd also clarifies that, by means of physical violence, economic domination,
and psychologically coercive forces, oppression is essentially an institution-
ally structured harm perpetrated on groups by other groups, in which a privi-
leged social group benefits from the harm endured by the oppressed.8
Oppression can distort or damage the self-conception of an oppressed
person, alienating her from her authentic self and further molding her into
subordinate positions. As Sandra Bartky highlights in her work on the psy-
chological dimensions of oppression, an oppressed person can come not only
to adopt desires and values that are not her own but also to hold beliefs about
herself that reflect social positions of inferiority: to be psychologically op-
pressed is to be weighed down in your own mind; it is to have a harsh domin-
ion exercised over your self-esteem.9 The oppressed internalize a message of
inferiority, as when, for instance, women are regarded by others and come
to regard themselves as childlike, as cheap labor, or as objects for the gaze or
sexual pleasure of others.10 As Michael Walzer writes in conveying another ex-
ample from working life, When a garbage-man feels stigmatized by the work
he does...the stigma shows in his eyes. He enters into collusion with us to
avoid contaminating us with his lowly self. He looks away; and we do too. Our
eyes do not meet. He becomes a non-person.11 To feel oneself inferior or to
feel oneself worthless as a person poses a threat to autonomy by undermining
self-respect, which is necessary for the realization of autonomous agency on
some accounts.12
6
Marilyn Frye, The Politics of Reality:Essays in Feminist Theory (Freedom, CA:Crossing Press,
193),33.
7
Sandra Bartky, On Psychological Oppression and Iris Marion Young, Five Faces of Oppression
reprinted in Feminist Theory:APhilosophical Anthology, edited by Ann Cudd and Robin Andreasen
(Malden, MA:Blackwell,2005).
8
Ann Cudd, Analyzing Oppression (NewYork:Oxford University Press, 2006), 26, cf.2327.
9
Bartky, On Psychological Oppression,105.
10
Bartky, On Psychological Oppression, 106,112.
11
Michael Walzer, Spheres of Justice:ADefense of Pluralism and Equality (NewYork:Basil Blackwell,
1983), 176. Walzer here cites Stewart E. Perrys San Francisco Scavengers:Dirty Work and the Pride of
Ownership (Berkeley:University of California Press, 1978),7.
12
As Joel Anderson and Axel Honneth write, practices that confer denigration or humiliation
threaten self-esteem by making it much harder (and, in limit cases, even impossible) to think of one-
self as worthwhile. The resulting feelings of shame and worthlessness threaten ones sense that there is
In thinking about autonomy and gender oppression, it is important to

recognize at the start that autonomy has long been coded masculine, as
Jane Dryden writes.13 Given historical and ideological exclusions of women
from ideals of autonomy, some feminist philosophers have looked askance at
conceptions of autonomy, at times rejecting the value altogether. One clas-
sic criticism, in circulation since the 1980s, is that autonomy is drawn from
male biographies and bound up with socially atomistic and individualistic
conceptions of human beings, such that autonomy is antithetical to the per-
sonal connections and social bonds around which many women reflectively
form self-identities. In this earlier wave of feminist scholarship on autonomy,
basic questions asked by philosophers concern whether autonomy requires
self-sufficiency at the expense of human connections, whether women find the
ideal of autonomy alienating, and whether feminine or feminist moral con-
cerns require different conceptions of autonomy, relative to those that have
been dominant in the history of philosophy or in contemporary moral and
political philosophy.
This skeptical stance toward classic ideals of autonomy forms part of the
starting point for work on relational conceptions of autonomy, in which fem-
inist philosophers rehabilitate autonomy to accommodate the social char-
acter of human agency. Accounts of relational autonomy draw attention to
the social, cultural, and historical contexts in which agents are embedded and
to the fact that autonomy is a capability of human agents who are not only ra-
tional but also emotional, embodied, desiring, creative and feeling.14 Those
who offer accounts of relational autonomy also analyze the effect of oppressive
socialization upon human agency and underscore that autonomy should not
be cast as antithetical to human connections, including those manifested in
love, friendship, appropriate care, and even loyalty and devotion.
Feminist accounts of relational autonomy have now changed the land-
scape of autonomy studies, shifting philosophical thinking about autonomy
toward the social and interpersonal dynamics that shape agency, desires, and
choices.15 Feminist scholarship has focused attention on the need for a finer
and richer account of agency, and there is now a fair amount of agreement that
autonomous agency is saturated with selfother relations. As Friedman notes,
point to ones undertakings. And without that sense of ones aspirations being worth pursuing, ones
agency is hampered. Anderson and Honneth, Autonomy, Vulnerability, Recognition and Justice, in
Autonomy and the Challenges to Liberalism:New Essays, edited by John Christman and Joel Anderson
(Cambridge:Cambridge University Press, 2005),131.
13
Jane Dryden, Autonomy:Overview, in The Internet Encyclopedia of Philosophy, November2010.
14
Catriona Mackenzie and Natalie Stoljar, Introduction: Autonomy Refigured, in Relational
Autonomy:Feminist Perspectives on Autonomy, Agency and the Social Self, edited by Mackenzie and
Stoljar (New York: Oxford University Press, 2000), 21; Natalie Stoljar, Feminist Perspectives on
Autonomy, in Stanford Encyclopedia of Philosophy, Summer 2013.
15
See, e.g., John Christman, Autonomy in Moral and Political Philosophy.
Introduction 5
philosophical conceptions of autonomyas opposed to conceptions of au-

tonomy that may be culturally dominantnow seldom suggest that autonomy
requires a self-sufficient or self-made person.16 Adaptive preferences formed in
the context of oppressive circumstances, such as preferences for subservience
or for iconic symbols of gender oppression, also now serve as formidable po-
tential counterexamples to purely proceduralist accounts of autonomy, which
typically hold that an agents autonomy in relation to a commitment is secured
merely by the agents endorsement of it, assuming the agents reflection is suit-
ably independent.
Since the publication of the landmark collection Relational Autonomy in
2000, feminist philosophers and autonomy scholars have continued debate
over the conditions necessary for autonomous choice, the satisfactoriness of
value-neutral accounts of autonomy, and the respect-worthiness of preferences
formed in adaptive contexts, among other issues. For instance, in developing
feminist accounts of autonomy, some theorists maintain that women who act
subserviently or upon preferences formed in oppressive circumstances are not
autonomous.17 Others, however, argue that respecting the agency and delib-
erative capacities of oppressed women requires that we not characterize such
women as compliant dupes of patriarchy18 and that women living in severely
oppressive conditions find outlets for the exercise of autonomy.19 Both lines of
argument initially appear plausible:as Diana Meyers observes, value-neutral
accounts of autonomy, in which autonomy does not require choosing particu-
lar values, such as equality or independence, appear attractive partly on ac-
count of showing respect for women who choose subservience or deference.
On the other hand, value-saturated accounts appear attractive on account
of highlighting the autonomy-subverting costs of living under oppressive
systems.20
In Gender in the Mirror, Meyers notes that both value-neutral and
value-saturated accounts of autonomy are troubling:value-saturated accounts
appear to stigmatize some women as victims, to homogenize autonomous and
16
Marilyn Friedman, Autonomy, Social Disruption and Women, in The Feminist Philosophy
Reader, edited by Alison Bailey and Chris Cuomo (NewYork:McGraw-Hill, 2008). See also Friedmans
chapter Relational Autonomy and Independence in this volume.
17
Natalie Stoljar, Autonomy and the Feminist Intuition, in Mackenzie and Stoljar, Relational
Autonomy.
18
Uma Narayan, Minds of Their Own: Choices, Autonomy, Cultural Practices and Other
Women, in A Mind of Ones Own:Feminist Essays on Reason and Objectivity (Boulder, CO:Westview
Press,2002).
19
John Christman, Relational Autonomy, Liberal Individualism, and the Social Construction of
Selves, Philosophical Studies 117 (2004):143164; Andrea Westlund, Rethinking Relational Autonomy,
Hypatia 24 (2009): 2649; Serene J. Khader, Adaptive Preferences and Womens Empowerment
(NewYork:Oxford University Press,2011).
20
Diana Tietjens Meyers, Gender in the Mirror: Cultural Imagery and Womens Agency
(NewYork:Oxford University Press 2002),11.
authentic lives, and to overlook the agency women exercise even in contexts
of oppressive circumstances. Yet value-neutral accounts serve poorly as tools
of social critique and suffer on account of packing autonomy into purely pro-
cedural processes of reflection that effectively neglect the possibility that a
well-integrated, smoothly-functioning self could be in need of rigorous scru-
tiny and drastic overhaul.21 In light of potential pitfalls of both approaches,
Meyers argues that a compelling feminist account of autonomy must ac-
knowledge that oppression impedes autonomy without stripping women of
that autonomy which they have managed to wrest from a patriarchal, racist,
heterosexist, ageist, class-stratified world.22 In her contribution to this volume,
she further distinguishes ways values enter autonomy theories, demarcating
new conceptual axes along which to position accounts of autonomy.
In this collection of new papers, leading scholars carry forward exami-
nations of central theoretical and practical issues at the intersection of au-
tonomy studies and feminist philosophy. Contributors examine fundamental
components and commitments of autonomy, examining for instance the
role of reflective deliberation, reasons, values, cares, emotions, self-worth,
self-care, adaptive preferences, social and political commitments, and norms
of independence in accounts of autonomy. Some papers pursue the question
of whether autonomy is compatible with subordination, including forms of
gender subordination and class-based subordination. Others examine how
ideals of autonomy are affected by capitalism, political commitments to inclu-
sivity, and feminist emphases on the relationality of human agency. In looking
at autonomy amid oppression, the volume represents a plurality of perspec-
tives about autonomy. Some contributors examine the agency of women and
oppressed persons through the lens of value-neutral accounts of autonomy,
whereas others utilize dialogical accounts, capabilities accounts, or thicker
value-saturated accounts. Still others make meta-arguments about the merits
of different kinds of approaches relative to feminist ambitions. Anumber of
papers focus on assessing autonomy in social contexts in which agents form
adaptive preferences or internalize gendered norms, and some focus on how
autonomy bears in social and personal contexts of raising girls, working, preg-
nancy and abortion, and end-of-life decisions.
We have organized the papers in the volume into five sections, beginning
with an initial cluster that explores key dimensions of the concept of autonomy,
especially in regards to its relational character and associated notions of inde-
pendence and freedom. In Chapter 2, Catriona Mackenzie focuses on the con-
cept of autonomy itself. According to Mackenzie, one of the key reasons that
autonomy remains a contested value is because philosophers have tended to
Meyers, Gender in the Mirror,16.

21
Meyers, Gender in the Mirror,16.

22
Introduction 7
view autonomy as a unitary concept. She argues that autonomy ought instead
to be understood as a multidimensional concept consisting of three logically
distinct but causally related dimensions:self-determination, self-governance,
and self-authorization. In addition, Mackenzie provides a relational analysis
of each of these dimensions of autonomy and argues that what is required to
satisfy the conditions of autonomy in particular contexts will often fluctuate.
The result of her work is to provide philosophers with a more nuanced under-
standing of autonomy, one that will allow debate on autonomy to proceed with
greater clarity, precision, and sensitivity to context.
Chapter 3, by Marilyn Friedman, explores ways the concept of autonomy can
combine relational and individualistic elements. Focusing on the discussion of
liberal individualism in Jennifer Nedelskys book Laws Relations:ARelational
Theory of Self, Autonomy, and Law, Friedman contends that relational and in-
dividualistic aspects of autonomy need not be irreconcilable:acknowledging
the constitutive relationality of human selves is consistent with maintaining
boundaries among individuals, such as occurs when the state serves to pro-
tect individuals from threats posed by one another. She argues further that
the popular notion of the self-made manvalorized by some but criticized by
othersis not relevant to discussions of autonomy so much as to debates about
capitalism. Freidman concludes her paper with a defense of independence as
an ideal for subordinate persons:not only does an ideal of independence pro-
vide a useful goal for liberatory movements, partly because it protects against
some forms of vulnerability, but it also serves a classic feminist goal of freeing
women to shape their own lives rather than accepting confining definitions
imposed by others.
Chapter 4, by Nancy Hirschmann, further explores the theme of indepen-
dence. Hirschmann registers skepticism about relational autonomy, argu-
ing provocatively that the concept originates from a pathology in feminine
psychoanalytical development: in sexist and heteronormative practices of
childrearing, girls and women emerge with relational self-identities oriented
toward care and empathy that undermine the need to cultivate an indepen-
dent self. She suggests that if we retain the concept of relational autonomy,
then we also need a feminist concept of freedom in which a person remains
an individual responsible for her own choices. Whatever desires and abilities
we cultivate through relationships with others, she argues, we need to act by
and for ourselves.
The volume continues with a second cluster of papers focusing on the nor-
mative and social commitments of relational approaches to autonomy. In
Chapter 5, Paul Benson contends that many of the ongoing disagreements
about the normative commitments of relational autonomy can be overcome
by considering the practical question of how a conception of autonomy can
best advance the ethical, social, and political aims of feminism. He argues that
a conception of autonomy that focuses on autonomous agents authority to
speak or answer to others for their choices and actions affords a preferred way
to understand autonomys relational character. Such a conception is especially
congenial to advancing the practical goals of feminism, according to Benson,
because it captures the importance of womens attitudes toward their own ex-
perience and because it accurately takes account of the social conditions that
inform that experience. Chapter 6, by Diana Meyers, also discusses the issue of
the normative commitments of autonomy. Taking as her starting point the on-
going disputes between those who defend substantive accounts of autonomy
and those who defend content-neutral accounts, Meyers argues that values
may be implicated in autonomy theories in two distinct ways. On one hand, an
autonomy theory may prescribe or proscribe certain types of behavior or allow
that any sort of behavior might be autonomously chosen; she calls this the
Directivity Axis. On the other hand, an autonomy theory may utilize or invoke
background values to elucidate the process of autonomous choice; she terms
this the Constitutivity Axis. According to Meyers, this Double Axis Thesis has
the benefit of making room for autonomy theories that are both value neu-
tral and value utilizing. Echoing the practical concerns addressed in Bensons
paper, Meyers ends her paper by arguing that value-neutral positions on the
Directivity Axis serve feminist purposes well.
In Chapter 7, by contrast, Marina Oshana sees feminist purposes served well
with a thicker conception of autonomy in which autonomy requires author-
ity over certain choices, a lack of domination and exploitation in social rela-
tionships, and enough economic security to maintain control over important
aspects of our lives. Without meaningful economic security, she points out,
a person cannot maintain control over fundamental choices, such as choices
concerning family or life partners. Broadly, Oshana also argues biconditionally
that a commitment to autonomy entails a commitment to feminism and that
a commitment to feminism entails a commitment to autonomy. For Oshana,
respecting autonomy entails opposing forms of social domination and thus
respecting autonomy entails the core demands of feminism, which mutually
entails respecting the abilities of persons to make their own decisions and en-
gage in action by means of their own authority.
A third cluster of papers in the volume attends particularly to care, emotion,
and reason in accounts of autonomy and challenges certain influential notions
about autonomy. In Chapter 8, Christine Tappolet rebuts a notion found in
both historical and contemporary philosophical sources that emotional agents
(and by extension women) cannot be autonomous. She draws on an account
of emotions she develops elsewhere to show not only that emotions are inte-
gral in autonomous agency but also that emotions and reason-responsiveness
are not at odds. In addition to dismantling a ludicrous argument that women
cannot be autonomous, Tappolets paper thus constitutes a powerful challenge
to autonomy accounts that are predominantly rationalistic in character. This
paper dovetails with our ninth chapter by Andrea Westlund, who argues that
Introduction 9
autonomy requires an attitude of self-care, which involves taking care with

ones reasons and decisions as well as taking responsibility for ones judgments.
Clarifying distinct conceptions of care at play in the philosophical literature,
she utilizes a finessed conception of care to augment answerability accounts
of autonomy, which, for Westlund, involve maintaining an open-minded
disposition to explain ones reasons and cares and to engage in potentially
self-transformative deliberation with others. Westlund sees the relational
character of autonomy as being at least partially grounded in the autonomous
agents attitude of self-care:it is precisely the autonomous agents attitude of
care for her self that leads her to take seriously reflective deliberation with
qualified others about reasons for action.
The fourth cluster of papers in the volume engages with issues surrounding
the relation among autonomy, oppression, and adaptive preferences. In
Chapter 10, John Christman addresses the question of how to distinguish be-
tween (1)cases in which agents respond to changing life circumstances while
retaining their autonomy and (2)cases in which agents adapt to new circum-
stances in an autonomy-undermining way. He examines this distinction by
considering not only different sources of serious life constraints, including
human trafficking and paralysis through disability, but also different judg-
ments about these constraints that can be supported with various accounts of
autonomy. He argues ultimately that a proceduralist account of autonomy is
best able to account for the difference between self-governing adaptation and
loss of agency in response to oppressive constraints. In so doing he refines the
procedural account of autonomy developed in his earlier work, incorporating
a new condition he calls reflexive self-affirmation.
Chapter 11, by Natalie Stoljar, continues a focus on autonomy and adaptation
to oppressive social circumstances. Tackling the claim that some philosophers
have made that desires formed under oppressive circumstances are deformed
and hence autonomy undermining, Stoljar defends a two-sided thesis. On one
hand, she argues that adaptive preferences per se are not incompatible with
autonomy. Yet on the other hand, she contends that many of the cases of adap-
tive preference formation of concern to feminists are autonomy undermin-
ing insofar as they fail to satisfy criteria contained in both proceduralist and
substantive theories of autonomy. Like that of Christman, Stoljars discussion
offers both an illuminating analysis of preferences formed in oppressive con-
texts and a substantive engagement with the important recent work of feminist
philosopher Serene Khader.
A final fifth cluster of papers deals with autonomy as it relates to particular
social and personal contexts. In Chapter 12, Mark Piper raises the question of
how parents ought to gender socialize their daughters, given a commitment
both to feminism and to enhancement of the development of daughters future
autonomy. He joins other feminist theorists in observing that traditional fe-
male gender socialization tends to subvert the autonomy of women, such as by
teaching girls to favor subordinate roles or by undermining girls confidence

in themselves. In light of a need for alternative forms of gender socialization,
Piper examines a variety of forms of gender socialization in light of commit-
ments to autonomy and feminism, ultimately favoring an inclusive form of
feminist gender socialization that retains a positive valuation of the category
of womanhood.
Whereas Piper addresses the impact that parents and others can have on the
development of autonomy in girls and women, Andrea Veltman, in Chapter
13, turns attention to the impact that work can have on the development and
exercise of autonomous capabilities. Drawing on empirical and philosoph-
ical literatures on work and well-being, she argues that working extensively
at eudemonistically meaningless work undermines autonomy and self-respect
and that promoting autonomous agency entails respecting the agency and skills
people exercise at work. In part, Veltman also examines autonomy in relation
to economic independence, writing in agreement with Friedman, Oshana, and
others that economic independence is a condition of personal autonomy that
women have good reason to seek. Although some feminist theorists highlight
morally problematic implications of a social ethos of economic independence
(particularly for women whose need to raise young children renders ideals
of independence and self-sufficiency unattainable), Veltman says that an ap-
preciable dimension of the oppression of many women workers is that they
labor extensively at jobs that do not enable a meaningful measure of financial
independence.
A final pairing of papers in our fifth cluster examines autonomy and embod-
iment with attention to pregnancy, abortion, disability, and physician-assisted
suicide. Chapter 14, by Anita Superson, concerns the place of a right to bodily
autonomy in relation to the abortion controversy. Drawing centrally on the
work of Judith Jarvis Thomson,23 Superson argues that the right to bodily au-
tonomy is even stronger than Thomson supposed, and she underwrites pow-
erful arguments in favor of womens right to abortion. Although focusing
the majority of her essay on the abortion debate, Superson concludes by not-
ing that the right to bodily autonomy has crucial normative weight in many
other debates that are of importance to women, including rape, female gen-
ital manipulation, and woman battering. Finally, in Chapter 15, Anita Ho dis-
cusses the influence of ableist social ideology on the autonomy of those facing
end-of-life decisions. According to Ho, even if the influence of ableist social
ideology is not directly coercive, it can inform peoples deliberations about
their alternatives in an autonomy-undermining way. The upshot of her work is
a call for greater attention to the broader societal assumptions about the value
23
Judith Jarvis Thomson, A Defense of Abortion, Philosophy and Public Affairs 1:1 (Autumn
1971):4766.
Introduction 11
of life with impairments and an appreciation of how these assumptions may

negatively affect peoples quality of life, the autonomy of their decision-making
processes, and their end-of-life decisions.
This collection is a collaborative endeavor, and we would like to express
our sincere thanks to all our contributing authors, who have been a pleasure
to work with and whose papers we are proud to bring together in this volume.
We especially thank Natalie Stoljar and Catronia Mackenzie for supporting
this collection by organizing a workshop called Relational Autonomy: Ten
Years On, at which our contributors had an invaluable opportunity to present
and discuss drafts of papers for the volume. We would also like to thank the
Department of Philosophy and Religion at James Madison University for sup-
porting our work on the project and our research assistant, Sara Scherer, for
assistance in helping us prepare and streamline the collection. Thanks also to
our acquisitions editor at Oxford University Press, Lucy Randall, for her sup-
port and assistance with the project from its inception and to the anonymous
reviewers of our book project for valuable recommendations for the project.
We hope that readers of this volume will benefit from the efforts of all who
have contributed and will find the papers as important and as richly stimu-
lating aswedo.
{PartI }
Autonomy and Independence

{2}
Three Dimensions of Autonomy:ARelational

Analysis
Catriona Mackenzie
It is no secret that although respect for individual autonomy is a foundational

principle of liberal and democratic theory, just what this principle means and
how the concept of autonomy should be interpreted and applied in specific
contexts are highly contested. For example, in many debates in bioethics, it is
not uncommon for disputants on both sides of a debate to invoke the concept
of autonomy in support of their position. One way to interpret what is at stake
in these disputes is to appeal to Rawlss concept/conception distinction. While
the disputants both invoke the concept of autonomy, they are in fact appealing
to different conceptions of individual autonomy, and these different concep-
tions arise from different normative theoretical frameworks and from different
value orientations and political commitments.
I think this suggestion goes part of the way toward explaining why au-
tonomy is such a contested value. However, Idont think that the concept/con-
ception distinction provides a full diagnosis of the problem. Another aspect
of the problem is that in both bioethics and debates about autonomy in the
philosophical literature there is a lot of talking at cross-purposes because there
is a lack of clarity about what kind of concept the concept of autonomy is. Is it
a unitary concept or a multidimensional concept? And is it a context-invariant
concept or a context-sensitive concept that is employed for different purposes
in different contexts? Related to this second question, there is a lack of clarity
about what is at issue in debates about autonomy. Is the aim of such debates to
analyze the concept, or is it to clarify the role the concept plays in our social
and normative practices?
My aims in this chapter are twofold. First, Iseek to address the question of
what kind of concept the concept of autonomy is. My response to this question
is that autonomy is a multidimensional, rather than a unitary, concept. In devel-
oping this response, Ipropose a taxonomy that distinguishes three distinct, but
causally interdependent, dimensions or axes of autonomy:self-determination,
self-governance, and self-authorization. Each dimension involves distinct
conditions, which may be satisfied to varying degrees. Isuggest further that

autonomy is not a context-invariant concept. Rather, it is employed for dif-
ferent purposes in different social and normative contexts, and in these dif-
ferent contexts it may be appropriate to set different thresholds for satisfying
the conditions for autonomy.
The second aim of the chapter is to develop a relational interpretation
or conception of the three axes of self-determination, self-governance, and
self-authorization. Although relational conceptions of autonomy have be-
come increasingly influential in feminist theory, relational autonomy is often
invoked in support of feminist commitments rather than analyzed in detail,
particularly in feminist bioethics. By developing a relational interpretation of
the three dimensions or axes of autonomy, Iseek to provide a more detailed
analysis of the commitments and aims of relational autonomy. In doing so,
I point to interesting conceptual connections between relational autonomy
theory and capabilities approaches to justice, which have not hitherto been
identified in the literature.1
The chapter is in five sections. In the first section, Iprovide a brief overview
sketch of my proposed multidimensional analysis of the concept of autonomy.
In the second section, Iprovide an overview of the main motivations and com-
mitments of relational autonomy. In sections 3 to 5, Ipropose a specific rela-
tional interpretation of each autonomy axis and its relevant conditions. It is
important to clarify at the outset that this is a big picture project, an exercise
in conceptual mapping. Each of the three axes of autonomy, with their specific
conditions, refers to a large and complex literature, which Icannot discuss in
any great detail. So my conceptual map is an outline of the terrain. Filling in
its topographical details is a much bigger project than could be undertaken in
a single chapter.
1. Three Dimensions of Autonomy
A unitary concept is a concept for which there is a single set of necessary and
sufficient conditions for the correct application of the concept. Natural kind
terms such as water, gold, and elephant are unitary concepts. Ethical concepts,
such as goodness, trust, and autonomy, are also often treated as unitary con-
cepts. In the case of individual autonomy, the kernel of what is taken to be the
unitary concept of autonomy is the notion of self-governance, the idea that to
be autonomous is to be capable of making decisions and acting on the basis of
1
I also explore these connections in Mackenzie, The Importance of Relational Autonomy
and Capabilities for an Ethics of Vulnerability, in Vulnerability: New Essays in Ethics and Feminist
Philosophy, edited by Catriona Mackenzie, Wendy A. Rogers and Susan Dodds (New York: Oxford
University Press, 2014). For a related discussion, see Marina Oshanas contribution to this volume.
Three Dimensions of Autonomy 17
motives, values, or reasons that are ones own in some relevant sense. Different
conceptions of autonomy aim, in part, to analyze the necessary and sufficient
conditions for a motive, value, reason, or action to be onesown.
In the literature two broad kinds of condition for self-governance have been
identified: competency conditions, which spell out the cognitive, volitional,
normative, or other competences deemed necessary to act effectively on ones
motives, values, or reasons; and authenticity conditions, which spell out what it
means for these elements of ones motivational structure to be genuinely ones
own. Much debate about autonomy in philosophical moral psychology, in-
cluding debate among theorists of relational autonomy, has focused on advanc-
ing differing interpretations of these conditions, with debate proceeding by way
of examples and counterexamples designed to test the necessity or sufficiency
claims of rival interpretations. Not surprisingly, no agreement on a single, defin-
itive interpretation of the competence or authenticity conditions for autonomy
has emerged from the debate. Ithink there are two connected reasons for this.
First, the concept of autonomy is employed for different purposes in different
social and normative contexts. This is one reason why we need a multidimen-
sional analysis of the concept. Second, the unitary notion of self-governance,
which identifies conditions for autonomy that relate to agents practical identities
and the skills and competences required for autonomy, is inadequate to capture
the multidimensional nature of the concept of autonomy.
My suggestion is that the concept of autonomy involves three dis-
tinct but causally interdependent dimensions or axes: self-determination,
self-governance, and self-authorization. Self-determination involves having
the freedom and opportunities to make and enact choices of practical im-
port to ones life, that is, choices about what to value, who to be, and what to
do. The self-determination axis identifies external, structural conditions for
individual autonomy, specifically freedom conditions and opportunity condi-
tions. Freedom conditions identify the kind of social and political constraints
that interfere with the exercise of self-determination and the kind of political
and personal liberties that enable it. Opportunity conditions specify the kinds
of opportunities that need to be available to agents in their social environ-
ments for them to have choices about what to value, who to be, and what to
do. Gerald Dworkin, Joseph Raz, and Marina Oshana have all emphasized the
importance of freedom and opportunity conditions for autonomy.2
Self-governance involves having the skills and capacities necessary to make
choices and enact decisions that express, or cohere with, ones reflectively
constituted diachronic practical identity. Following Korsgaard,3 Iunderstand
2
Gerald Dworkin, The Theory and Practice of Autonomy (NewYork:Cambridge University Press,
1988); Joseph Raz, The Morality of Freedom (Oxford:Clarendon Press, 1986); Marina Oshana, Personal
Autonomy in Society (Aldershot, UK:Ashgate,2006).
3
Christine Korsgaard, The Sources of Normativity (Cambridge:Cambridge University Press,1996).
practical identity as a normative self-conception, which embodies a per-

sons sense of self-identity and her commitments, values, and beliefs. The
self-governance axis identifies internal conditions for autonomy, specifically
competence and authenticity conditions. Competence conditions are central
to many accounts of autonomy in bioethics. Beauchamp and Childress, for
example, understand competence in medical decision making in terms of
capacities to understand relevant information, make a judgment about that
information in light of ones values, and communicate the decision to caregiv-
ers.4 Within philosophical moral psychology, there is quite a range of different
accounts of competence, including reason-responsiveness views (e.g., Wolf),
normative competence views (e.g., Benson and Stoljar), and skills-based views
(e.g., Meyers).5 Predominantly, however, debates about autonomy within phil-
osophical moral psychology have focused on authenticity conditions. For
example, the early work of Harry Frankfurt on identification with ones will,
care-based analyses of autonomy that develop themes in Frankfurts later work
on caring6, and John Christmans nonalienation analysis of critical reflection
all provide different articulations of authenticity.7
The third axis, self-authorization, involves regarding oneself as having the
normative authority to be self-determining and self-governing. In other words,
it involves regarding oneself as authorized to exercise practical control over
ones life, to determine ones own reasons for action, and to define ones values
and identity-shaping practical commitments. The self-authorization axis has
emerged as a central concern of recent social and relational theories of au-
tonomy. This axis identifies several conditions for autonomy, relating to ac-
countability, self-evaluative attitudes, and social recognition. These conditions
are discussed, with different inflections and emphases, in recent work by a
range of theorists. Paul Benson and Andrea Westlund, for example, ground
4
Tom Beauchamp and James Childress, Principles of Biomedical Ethics, 7th ed. (NewYork:Oxford
University Press,2012).
5
For reasons-responsiveness views, see Susan Wolf, Freedom within Reason (NewYork:Oxford
University Press, 1990); for normative competence views, see Paul Benson, Autonomy and Oppressive
Socialization, Social Theory and Practice 17 (1991): 385408; Natalie Stoljar, Autonomy and the
Feminist Intuition, in Relational Autonomy:Feminist Perspectives on Autonomy, Agency and the Social
Self, edited by C. Mackenzie and N. Stoljar (New York: Oxford University Press, 2000), 94111; for
skills-based views, see Diana Meyers, Self, Society and Personal Choice (NewYork:Columbia University
Press,1989).
6
For the notion of identification, see the essays in Harry Frankfurt, The importance of what we care
about (Cambridge:Cambridge University Press, 1988); for care-based analyses, see, e.g., Frankfurt, On
Caring, in Frankfurt, Necessity, Volition and Love (Cambridge: Cambridge University Press, 1999),
155180; Agnieszka Jaworksa Caring and Internality, Philosophy and Phenomenological Research 74:3
(2007):529568; Jaworksa, Caring, Minimal Autonomy, and the Limits of Liberalism, in Naturalized
Bioethics:Towards Responsible Knowing and Practice, edited by H. Lindemann, M. Verkerk, and M.
Walker (Cambridge:Cambridge University Press, 2009), 80105.
7
John Christman, The Politics of Persons: Individual Autonomy and Socio-historical Selves
(Cambridge:Cambridge University Press,2009).
autonomy in an agents attitudes or dispositions to hold herself accountable

and answerable to others for her reasons.8 Carolyn McLeod emphasizes the
importance of agents self-evaluative attitudes, in particular self-trust, for
autonomous decision making in health-care contexts.9 And Joel Anderson
argues that autonomy is a constitutively social capacity, in the sense that it is
constituted within normative social practices of mutual recognition.10
Two clarifications are in order at this point. First, the aim of this taxonomy
is to map out the different dimensions or axes of the concept of autonomy
in a way that makes sense of our complex, and sometimes conflicting, intu-
itions about autonomy and our diverse autonomy-related social practices. Of
course, the map by itself doesnt resolve debates about how best to understand
the different conditions I have identified as relevant to each axis. Different
theories of autonomy will continue to conceptualize these conditions in dif-
ferent ways, and in later sections Ipropose a relational conception of each of
the conditions. But Ihope the map will help to resolve some of the talking at
cross-purposes in the literature that arises from understanding autonomy as
the unitary concept of self-governance.
Second, it is important to recognize that the exercise of autonomy along
these different dimensions can be assessed at different levels: locally, with
respect to specific decisions or choices; programmatically, with respect to
domains of a persons life, such as with respect to her finances, at work, or in
her family or intimate relationships; or globally, with respect to a persons life
overall.11 So it is possible that a person might exercise a high level of autonomy
in one domain, such as at work, while being subject to domination in another
domain, such as in her family life. It is also possible that a person might be
able to exercise local autonomy with respect to a particular decision, such as
the choice to accept or refuse a particular medical treatment, even though
his global autonomy may be seriously impaired, for example because he is an
addict. Within bioethics, autonomy is usually understood as local, that is, as
8
Paul Benson, Feminist Intuitions and the Normative Substance of Autonomy, in Personal
Autonomy:New Essays on Personal Autonomy and Its Role in Contemporary Moral Philosophy, edited
by James Stacey Taylor (Cambridge: Cambridge University Press, 2005), 124142; Paul Benson,
Taking Ownership:Authority and Voice in Autonomous Agency, in Autonomy and the Challenges to
Liberalism, edited by J. Christman and J. Anderson (Cambridge:Cambridge University Press, 2005),
101126; Andrea Westlund, Rethinking Relational Autonomy, Hypatia 24:4 (2009):2649.
9
Carolyn McLeod, Self-Trust and Reproductive Autonomy (Cambridge, MA:MIT Press,2002).
10
Joel Anderson and Axel Honneth, Autonomy, Vulnerability, Recognition and Justice, in
Christman and Anderson, Autonomy and the Challenges to Liberalism, 127149; Joel Anderson,
Vulnerability and Autonomy Intertwined, in Vulnerability: New Essays in Ethics and Feminist
Philosophy, edited by Catriona Mackenzie, Wendy A. Rogers, and Susan Dodds (NewYork:Oxford
11
For discussions of the distinctions between local, programmatic, and global autonomy, see espe-
cially Meyers, Self, Society and Personal Choice; Oshana, Personal Autonomy in Society.
applying to a specific decision at a specific time.12 Feminist bioethicists, how-

ever, who are influenced by relational theories of autonomy, tend to hold the
view that the exercise of local autonomy needs to be understood in the context
of a persons capacity to exercise programmatic or global autonomy.13
2. Why Relational Autonomy?
Relational theories of autonomy were originally motivated by three central con-

victions. First, in spite of feminist critiques of autonomy as a masculinist and
excessively individualist ideal, it is a mistake for feminists to jettison the value
of autonomy because the exercise of some threshold level of self-determination,
self-governance, and self-authorization is crucial for womens emancipation.
Second, gender oppression, in different ways in different contexts and to dif-
ferent degrees, constrains womens abilities to lead self-determining lives and
thus can impair their capacities for self-governance and can undermine their
sense of being self-authorizing agents. Gender oppression is here understood
as a system or pattern of hierarchically structured social relations, institutions,
and practices of gender-based domination and subordination. Third, we need
a refigured conception of autonomy that is responsive to feminist critiques
of overly individualistic conceptions of autonomy. Relational theories aim to
provide this refigured conception.14
As a conception of individual autonomy, relational autonomy is committed
to a form of normative individualismthat is, to the view that the rights, wel-
fare, dignity, freedom, and autonomy of individuals matter and impose nor-
mative constraints on the claims of social groups or collectives.15 Capabilities
theory is also committed to normative individualism in this sense.16 In what
12
See the discussion of autonomy in Beauchamp and Childress, Principles of Biomedical Ethics.
See also UK Department for Constitutional Affairs, Mental Capacity Act 2005 Code of Practice
(Norwich:Stationery Office,2007).
13
For further discussions of this claim and of the significance of relational autonomy for bio-
ethics, see Susan Sherwin, A Relational Approach to Autonomy in Health Care, in The Politics of
Womens Health:Exploring Agency and Autonomy, edited by S.Sherwin and the Feminist Health Care
Ethics Research Network (Philadelphia:Temple University Press, 1998), 1947; Catriona Mackenzie,
Autonomy, in Routledge Companion to Bioethics, edited by John Arras, Elizabeth Fenton, and Rebecca
Kukla (NewYork:Routledge, forthcoming).
14
For more extended discussion of these claims, see Mackenzie and Stoljar, Introduction:Autonomy
Refigured, in Mackenzie and Stoljar, Relational Autonomy.
15
Astronger version of normative individualism is proposed by Elizabeth Anderson, Towards
a Non-Ideal, Relational Methodology for Political Philosophy, Hypatia 24:4 (2009): 132. She holds
that the claims of individuals have normative priority over the claims of social groups or collectives.
Relational autonomy need not be committed to this stronger form of normative individualism, al-
though some relational theorists might endorse the strongerview.
16
See. e.g., Martha Nussbaum, Creating Capabilities (Cambridge, MA: Harvard University
Press,2011).
sense, then, is the theory relational? Isuggest it is relational in several ways

that can be articulated in the form of three premises or starting assumptions to
which relational theories are committed. It should be noted that, with respect
to these assumptions, there is much common ground between relational au-
tonomy theory and capabilities theory.
The first premise is that an adequate conception of autonomy must be
responsive to the facts of human vulnerability and dependency rather than
assuming a conception of persons as self-sufficient, independent, rational con-
tractors. It follows that the conditions for self-determination, self-governance,
and self-authorization cannot be specified in such a way that these dimensions
of autonomy are inconsistent with human vulnerability, dependency, or social
relations of care.17 Because relational autonomy theory starts from a conception
of persons as vulnerable and dependent to varying degrees, Iwant to suggest
that it is highly relevant to health-care contexts, where obligations to respond
to vulnerability must be balanced with respect for patient autonomy. It should
also be noted that Nussbaums capabilities theory, in particular, emphasizes
the importance for a theory of justice to be premised on the facts of human
vulnerability.18
The second premise is that persons are embodied and socially, historically,
and culturally embedded and that their identities are constituted in relation
to these factors in complex ways. Thus, although relational autonomy theory
is committed to normative individualism, it is also committed to a social on-
tology of personsthat is, a conception of persons that emphasizes the role of
embodied social practices (including linguistic and cultural practices), social
group identities, and historical contingencies in the formation of our indi-
vidual practical identities. This premise is particularly relevant to the way the
self-governance dimension of autonomy should be understood. It suggests,
first, that an adequate analysis of authenticity conditions must account for the
complex social constitution of individual identity. Second, this premise sug-
gests that an adequate analysis of competence conditions must acknowledge
17
This is not to deny that some sources of vulnerability (e.g., abusive interpersonal relationships,
or political repression) may undermine autonomy. For more detailed analysis of vulnerability, see
Catriona Mackenzie, Wendy A. Rogers, and Susan Dodds, eds., Vulnerability:New Essays in Ethics and
Feminist Philosophy (NewYork:Oxford University Press, 2014). For reflections on the complex relation-
ship between autonomy and vulnerability, see especially the essays in that volume by Joel Anderson,
Catriona Mackenzie, and Jackie Leach Scully. For extended discussion of the way that some social
relationships, including relations of care, can undermine autonomy, see especially Marilyn Friedman,
Autonomy and Social Relationships:Rethinking the Feminist Critique, in Feminists Rethink the Self,
edited by Diana Tietjens Meyers (Boulder, CO:Westview Press, 1997), 4061; Friedman, Autonomy,
Social Disruption, and Women, in Mackenzie and Stoljar, Relational Autonomy, 3551; Friedman,
Autonomy, Gender, Politics (NewYork:Oxford University Press,2003).
18
See especially Martha Nussbaum, Frontiers of Justice:Disability, Nationality, Species Membership
(Cambridge, MA:Harvard University Press,2006).
the extensive interpersonal, social, and institutional scaffolding necessary for

the development and ongoing exercise of the complex cognitive, volitional,
imaginative, and emotional skills involved in self-governance. As Iargue in
the following sections, the extent of this scaffolding also points to the causal
interdependence between self-governance and the self-determination and
self-authorization dimensions of autonomy.
The third premise is that social conditions restricting the exercise of
self-determination are unjust. These include social conditions that deny mem-
bers of social groups fundamental political or personal liberties or restrict the
opportunities available in their social environments, that thwart the develop-
ment and exercise of autonomy competences, or that sanction social relations
of misrecognition, thereby excluding some individuals or social groups from
the status markers of autonomy and compromising their sense of themselves
as self-authorizing agents.19 Asocially just society therefore has an obligation
to develop social, legal, and political institutions that foster the autonomy of all
citizens, particularly those from historically oppressed or marginalized social
groups.
It is worth clarifying two points about the understanding of justice that
underlies this third premise. First, this account of justice assumes that injustice
is not just a function of inequality in the distribution of resources; rather, it is
also a function of inequalities in opportunities, and in social relations, institu-
tions, norms, and practices. These inequalities constitute forms of domination
and subordination that track social group identities and that operate through
hierarchies of social standing, power, authority, and esteem.20 These hierar-
chies operate at multiple levels within a society, for example within the family
and in the context of intimate relationships; in the institutions of civil society,
for example in workplaces, educational institutions, hospitals and clinics, reli-
gious institutions, clubs, sporting teams, the media (including the Internet);
and through the institutions of the state, such as the police force, the army,
This third premise is implicit rather than explicit in many accounts of relational autonomy.
19
Theorists who have explicitly discussed the social justice implications of relational autonomy are
Anderson and Honneth, Autonomy, Vulnerability, Recognition and Justice; Catriona Mackenzie,
Autonomy: Individualistic or Social and Relational? in Risk, Welfare and Work, edited by G.
Marston, J. Moss, and J. Quiggin (Melbourne:Melbourne University Press, 2010), 107127; Mackenzie,
Conceptions of Autonomy and Conceptions of the Body in Bioethics, in Feminist Bioethics:At the
Center, on the Margins, edited by Jackie Leach Scully, Laurel Baldwin-Ragaven, and Petya Fitzpatrick
(Baltimore:Johns Hopkins University Press, 2010); Marina Oshana, Personal Autonomy and Society,
Journal of Social Philosophy 29:1 (1998):81102; Oshana, Personal Autonomy in Society.
20
The way I have formulated this point is influenced by Elizabeth Andersons analysis of rela-
tional inequality in Anderson, What Is the Point of Equality? Ethics 109:2 (1999):287337; Anderson,
Towards a Non-Ideal, Relational Methodology for Political Philosophy, Hypatia 24:4 (2009):130145;
Anderson, Justifying the Capabilities Approach to Justice, in Measuring Justice:Primary Goods and
Capabilities, edited by Harry Brighouse and Ingrid Robeyns (Cambridge:Cambridge University Press,
2010), 81100.
the judiciary, and the legislature. Relational theories aim to develop analyses
of autonomy that are responsive to the effects of such inequalities on agents
abilities to lead self-determining, self-governing, and self-authorizinglives.
Second, Isee relational autonomy as a form of non-ideal theorizing about
autonomy, which is aligned with non-ideal approaches to justice.21 It does not
assume that persons are ideally, or even hypothetically, fully rational agents or
that their motivational structures are transparent to themselves. Nor does it
take as its starting point a model of an ideally just society and assess the extent
to which current social relations fall short of this model. Rather, its starting
point is the individual as situated in, shaped, and constrained by her sociorela-
tional context in all its complexity; that is, its starting point is non-ideal agents
in a non-ideal world, characterized by social oppression, injustice, and ine-
quality. Given this starting point, the aim of relational autonomy theory is to
theorize the kind of autonomy that is possible for non-ideal human agents; to
diagnose how social domination, oppression, stigmatization, and injustice can
thwart individual autonomy; and to hypothesize possible solutions, in the form
of proposing how specific social relations, practices, and institutions might be
reformed in such a way as to protect and foster individuals autonomy.
In the following three sections of the paper Ielaborate the three axes of au-
tonomy in more detail and develop a specifically relational conception of each
axis and its distinct conditions.
3. Self-determination:ARelational Conception
The notion of self-governance, as Ihave already suggested, is the core or kernel

of what is taken to be a unitary concept of individual autonomy and has been
the central focus of debates in bioethics and philosophical moral psychology.
One problem with this focus, and with analyzing autonomy in terms of nec-
essary and sufficient conditions for self-governance, is that it makes it difficult
to explain how structural (social and political) constraints can undermine or
impair individual autonomy and to identify the enabling structural conditions
for autonomy.
In the recent literature, theorists of autonomy have attempted to try to ex-
plain the importance of these constraints and conditions in different ways. One
approach is that of Oshana, who develops a global analysis of self-governance
as having de facto power and authority to exercise effective practical control
21
In his critique of ideal theory, Charles Mills describes relational autonomy as a form of non-ideal
theorizing about autonomy. Mills, Ideal Theory as Ideology, Hypatia 20:3 (2005):165184. For a re-
lated discussion of non-ideal theory, see Ingrid Robeyns, Ideal Theory in Theory and Practice, Social
Theory and Practice 34:3 (2008):341361. For a different kind of critique of ideal theory, see Amartya
Sen, The Idea of Justice (Cambridge:Cambridge University Press,2009).
over ones life.22 Oshana argues that certain structural sociorelational condi-
tions must be in place for an agent to be genuinely self-governing in this sense.
Agents who stand in relations of subordination, subservience, deference, or
economic or psychological dependence, for example, cannot be self-governing
because they do not enjoy practical control over their lives. This is the case
even if the agents in question endorse (or are not alienated from) their subor-
dinate, subservient, or dependent position and even if they seem to satisfy the
competence and authenticity requirements for self-governance. Oshana uses
an array of examplesvoluntary slaves, prisoners, women subject to extreme
forms of gender oppression, members of restrictive religious ordersto sup-
port the guiding intuition behind her accountnamely, that a person cannot
lead a self-governing life if her options are severely restricted and she is effec-
tively under the control of others.
I am very sympathetic to Oshanas guiding intuition, but Ialso think that she
is forced into an overly strong position because she attempts to articulate this
intuition in terms of necessary and sufficient conditions for self-governance.
Her view is overly strong because it is easy to devise counterexamples or to find
real-life examples of persons who are subject to severe constraint but who are
nevertheless clearly self-governing. Moreover, as John Christman and others
have argued, the view seems to impugn the autonomy of persons who, despite
being subject to crushing forms of oppression, nevertheless struggle to make
plans, set goals, and have clear identity-defining commitments that are gen-
uinely their own.23 For this reason, Ithink a better way to articulate Oshanas
guiding intuition is by distinguishing self-governance and self-determination
as two distinct, but causally interdependent, axes or dimensions of autonomy.
I develop this argument in what follows. Before doing so, however, Iwant
briefly to discuss Christmans way of accounting for the structural constraints
on and enabling conditions for autonomy. Irefer here to his autonomy-based
conception of justice as democratic legitimacy, as developed in The Politics of
Persons.24 Christmans account of the competence and authenticity conditions
for self-governance is premised on a thick, sociohistorical conception of the
person, and he is highly sensitive to the way that social group memberships and
historic and ongoing injustice shape and constrain individual practical identi-
ties. But, contra Oshana, he does not think that structural conditions should
be included in the conditions for self-governance.25 Instead, he includes these
Oshana, Personal Autonomy and Society, Oshana, Personal Autonomy in Society.

22
See, e.g., his critique of Oshanas position in Christman, Politics of Persons, 167173.
23
24
Christman, Politics of Persons.
25
Christmans reason for excluding structural conditions from the conditions for self-governance
is that he thinks including them involves unacceptable perfectionism. Christman criticizes Oshanas
view on these grounds, claiming that her account requires that autonomous agents must have cer-
tain value commitments and/or must be treated in certain normatively acceptable ways, even if
the agents in question do not endorse those value commitments. Christman, Politics of Persons, 171.
For a detailed response to Christmans perfectionism argument, see Oshana, Personal Autonomy in
conditions as part of the requirements of democratic justice. Thus, Christman

claims that an adequate description of the background circumstances of jus-
tice must include reference to the fact that specific social groups have been
subject to historical and ongoing forms of social domination, inequalities of
power, and systematic exposure to unique forms of violence.26 Christman
also argues that his account of democratic legitimacy would support, as a basic
claim of justice, constitutionally guaranteed access to the basic capabilities and
to the basic rights and liberties because of their role in enabling autonomy.
Iagree with Christman that claims of justice are distinct from the conditions
for self-governance. However, autonomy as self-determination is typically a
practical precondition for autonomy as self-governance, and securing the con-
ditions for self-determination is a matter of basic justice.
I have defined self-determination as having the freedom and opportunity
to make and enact choices of practical import to ones life, that is, choices
about what to value, who to be, and what to do. I have also suggested that
the self-determination axis identifies external, structural (social and political)
conditions for individual autonomy, specifically freedom conditions and op-
portunity conditions.27 Let me now say more about how Ithink these condi-
tions should be understood.
Freedom conditions specify the political and personal liberties that are nec-
essary for leading a self-determining life. Political liberties include the kinds
of liberties that all liberal, democratic, and feminist theorists think should be
protected, such as freedom of thought and expression, freedom of associa-
tion, freedom of conscience and religious exercise, freedom to engage in po-
litical participation, and freedom from arbitrary arrest. Nevertheless, among
theorists of autonomy and political philosophers, there will inevitably be
substantive disputes about the extent of the liberties that are necessary for
self-determination. Asalient example in the context of international debates
about immigration is whether freedom of international movement should be
included among the basic liberties.28 Iwont take a stand on this issuehere.
Society; Mackenzie, Relational Autonomy, Normative Authority and Perfectionism, Journal of Social
Philosophy 39 (2008):51233.
26
Christman, Politics of Persons,224.
27
The distinction between freedom conditions and opportunity conditions overlaps to some ex-
tent with Isaiah Berlins distinction between negative and positive liberty in Berlin, Two Concepts of
Liberty, in Four Essays on Liberty (Oxford:Oxford University Press, 1969). As Iexplain in the following
text, theorists who understand the freedom conditions for autonomy in terms of negative liberty con-
strue the freedom required for autonomy primarily as freedom from the undue interference of other
persons or the state. Theorists who understand the freedom conditions for autonomy in terms of sub-
stantive freedom or opportunity think that although some degree of negative liberty is a necessary con-
dition for autonomy it is insufficient. What matters for autonomy is the extent of a persons substantive
opportunities to be and to do. Autonomy-enabling opportunities require a lot more than freedom from
interference; they require substantial support by other persons and by state agencies.
28
Joseph Carens and Phillip Cole are examples of theorists who argue that liberal political prin-
ciples entail that freedom of international movement should be included among the basic liberties.
With respect to the personal liberties, again Ithink there is likely to be sub-
stantial agreement among liberal, democratic, and feminist theorists about the
basic personal liberties that are preconditions for self-determination. These lib-
erties include freedom of movement (if not international movement), freedom
of sexual expression, and freedom from all forms of coercion, manipulation,
exploitation, and violence, including sexual exploitation and assault. However,
again, there will be substantive disputes about what counts as coercion or ex-
ploitation. In bioethics, the question of how exploitation should be understood
is at the core of debates about the commodification of reproductive tissue, body
parts, and blood products. Bioethicists of a more libertarian persuasion, for
example, insist that commodificationthat is, markets in tissue, body parts,
and blood productsneed not be exploitative and that if properly regulated
such markets enable individual self-determination. Their critics argue that in
the constrained and often oppressive social contexts in which such markets
operate, commodification is inherently exploitative.29 Similarly, while liberal,
democratic, and feminist theorists support the principle of reproductive free-
dom, there are substantive disputes about how far this liberty should extend.
Bioethicists of a more libertarian persuasion think it extends to a right to deter-
mine the sex or to enhance the characteristics of ones offspring. Their critics re-
spond that the libertarian position equates autonomy with negative liberty and
mere preference satisfaction, fails to account for the social influences on and
social impacts of individual choice, and trivializes the responsibilities and obli-
gations attendant upon the exercise of reproductive autonomy, in particular
parental responsibilities for the well-being and future autonomy of children.30
Carens, Aliens and Citizens:The Case for Open Borders, Review of Politics 49:2 (1987):251273; Cole,
Open Borders:An Ethical Defense, in Debating the Ethics of Immigration:Is There a Right to Exclude?,
edited by Christopher Heath Wellman and Phillip Cole (New York: Oxford University Press, 2011).
Liberal theorists who reject this claim and who seek to defend the rights of nation states to exclude
immigrants include Michael Blake, Immigration, in Companion to Applied Ethics, edited by R. Frey
and Christopher Heath Wellman (Malden, MA:Blackwell, 2005), 224237; David Miller, Immigrants,
Nations and Citizenship, Journal of Political Philosophy 16:4 (2008):371390. This is of course a com-
plex debate that Icannot enter into here, and Irefer to it purely for illustrative purposes.
29
Theorists who appeal to the right to self-determination to defend markets in body parts, tis-
sues, and eggs, include, e.g., J. Radcliffe-Richards, Selling Organs, Gametes and Surrogacy Services,
in The Blackwell Guide to Medical Ethics, edited by R. Rhodes, L. Francis, and A. Silvers (Malden,
MA: Wiley-Blackwell, 2007), 254268; David Resnik, Regulating the Market for Human Eggs,
Bioethics 15:1 (2001):125; James Stacey Taylor, Autonomy, Constraining Options, and Organ Sales,
Journal of Applied Philosophy 19:3 (2002):273285. For critiques, see, e.g., Margaret Jane Radin, Contested
Commodities:The Trouble with the Trade in Sex, Children, Body Parts and Other Things (Cambridge,
MA:Harvard University Press, 1996); N. Scheper-Hughes, Illegal Organ Trade:Global Justice and the
Traffic in Human Organs, in Living Donor Organ Transplants, edited by Rainer Gruessner and Enrico
Benedetti (NewYork:McGraw-Hill, 2007), 106121; Tamara Zutlevics, Markets and the Needy:Organ
Sales or Aid? Journal of Applied Philosophy 18:3 (2001):297302.
30
The most well-known advocates of the libertarian position on genetic enhancement are Nicholas
Agar, Designing Babies: Morally Permissible Ways to Modify the Human Genome, Bioethics 9:1
(1995): 115; John Harris, Clones, Genes and Immortality (Oxford: Oxford University Press, 1998);
With respect to both these issues, I disagree with the libertarian position.
Although I cannot engage in a detailed discussion of these specific normative
issues here, hopefully the basis for my stance will become clearer in what fol-
lows. What these debates point to is that although the freedom and opportunity
conditions for self-determination overlap and often converge they can also be in
tension. Theorists who understand freedom in terms of negative liberty are likely
to give more weight to freedom conditions than to opportunity conditions and
hence to interpret the freedom conditions quite expansively. Theorists who em-
phasize the importance of substantive equality of opportunity think that what is
important is not the promotion of freedom per se, since not all kinds of freedom
are either important or desirable, but those freedoms that enable equality of op-
portunity.31 According to this view, which freedom conditions count as important
will be determined by the opportunity conditions for self-determination. On my
interpretation, theories of relational autonomy prioritize opportunity conditions
and hence fall into this latter category of theory.
Opportunity conditions specify the personal, social, and political oppor-
tunities that are preconditions for individual self-determination. Among
theorists of autonomy, Raz is the most prominent exponent of the view that
self-determination, as Ihave defined it, requires access to an adequate array
of significant opportunities, or what he refers to as options.32 Although access
to options requires certain freedoms, in the form of basic rights and liberties,
Raz rejects the libertarian conception of freedom as minimally constrained
freedom of choice, arguing that this conception conflates self-determination
with license.33 Raz develops a nuanced analysis of options as those complex
and multidimensional activities, practices, and relationships that make our
lives meaningful, such as pursuing a career in a particular profession, being
a parent, being a member of a religious community, or participating in sport,
cultural activities, or politics. Options, in other words, are ways of life, dense
webs of complex actions and interactions, which depend on social prac-
tices, with their implicit rules, conventions, and shared meanings. Specific
options are available only to those who have developed the relevant skills, as
embodied in social practices and transmitted by habituation.34 Aculture,
Julian Savulescu, Procreative Beneficence:Why We Should Select the Best Children, Bioethics 15:56
(2001):413426.
I develop critiques of the libertarian position with respect to debates in bioethics in Mackenzie,
Autonomy:Individualistic or Social and Relational?; Conceptions of Autonomy and Conceptions of
the Body in Bioethics; Autonomy, in Arris etal., Routledge Companion to Bioethics.
31
This is clearly Nussbaums position as well, and it is on these grounds that she criticizes Sens in-
terpretation of capabilities theory as concerned with promoting freedom. See, e.g., Nussbaum, Creating
Capabilities,6976.
32
Raz, Morality of Freedom.
33
Raz, Multiculturalism:ALiberal Perspective in Raz, Ethics in the Public Domain:Essays in the
Morality of Law and Politics (Oxford:Clarendon Press, 1994) 17191.
34
Raz, Multiculturalism,177.
in Razs view, is a complex network of interlocking social practices, and it

is only by being socialized in a culture that the range of options that con-
stitute that culture are made available to a person. Thus, by and large ones
cultural membership determines the horizon of ones opportunities.35 Razs
analysis of options highlights the importance of the social environment for
self-determination. As he puts it, autonomy depends not on the availability
of one option of freedom of choice. It depends on the general character of
ones environment and culture.36 An autonomy-supporting culture, in his
view, is one that makes available to its citizens an adequate range of signif-
icant options, by providing the institutional infrastructurelegal, educa-
tional, economic, politicalthat supports a wide array of significant options
and by developing social policy measures aimed at making these options
genuinely accessible to members of marginalized, disadvantaged, or histori-
cally oppressed social groups.
In developing her sociorelational account of autonomy, Oshana endorses
Razs view of the importance of significant options for autonomy, and this
is why Ithink her concern with the structural conditions for autonomy is
best characterized as a concern with the self-determination dimension of
autonomy.37 While Iam broadly sympathetic to Razs view, Ihave come to
think that capabilities theory provides a better vocabulary for articulating
the opportunity conditions for self-determination, both because it helps to
make explicit the link between basic justice and self-determination and be-
cause I think there are fruitful connections, still to be explored, between
relational autonomy theory and feminist interpretations of capabilities
theory, such as those of Elizabeth Anderson, Martha Nussbaum, and Ingrid
Robeyns.38 Icannot explore these connections in any detail here, but Ido
want to provide a few reasons as to why I think capabilities theory pro-
vides a useful vocabulary for articulating the opportunity conditions for
self-determination.
Capabilities theory provides a fine-grained comparative metric of equal-
ity, which disaggregates distinct dimensions of well-being and human
development. The metric aims to focus attention on what a person is ac-
tually able to be and do in her life (her achieved functionings) and on the
Raz, Multiculturalism,177.
35
Raz, Morality of Freedom,391.

36
37
Oshana, Personal Autonomy in Society,8486.
38
Elizabeth Anderson, What Is the Point of Equality?; Anderson, Justifying the Capabilities
Approach to Justice; Martha Nussbaum, Women and Human Development:The Capabilities Approach
(Cambridge:Cambridge University Press, 2000); Nussbaum, Frontiers of Justice; Nussbaum, Creating
Capabilities; Ingrid Robeyns, Sens Capability Approach and Gender Inequality: Selecting Relevant
Capabilities, Feminist Economics 9:23 (2003):7184; Robeyns, Gender and the Metric of Justice, in
Brighouse and Robeyns, Measuring Justice, 215235.
extent of her substantive opportunities (capabilities) to achieve valuable

functionings. It also aims to focus attention on the factors that affect an
individuals ability to convert the resources available to her into valuable
functionings.39 Such factors include both individual differences in natural
endowments, personal characteristics, and values and the external circum-
stances of peoples lives, such as their natural, social, cultural, and political
environments. Because it attends to the particularities of individual cir-
cumstance and context, capabilities theory also enables a fine-grained anal-
ysis of the different sources of social injustice, inequality, and disadvantage
within a society and their impact on peoples opportunities and capacities
to lead self-determininglives.
There is some disagreement among capability theorists as to how valuable
capabilities should be identified. Sen has resisted the development of a list of
valuable capabilities, arguing that this should be settled via mechanisms of
social choice, although it is evident from his writings that he thinks educa-
tion, political participation, health, and gender equality are central capabili-
ties.40 Nussbaum, as is well known, has developed a universal list of ten central
human capabilities.41 Other feminist capability theorists argue that such lists
should be context specific and should be developed and refined through pro-
cesses of democratic deliberation. For example, Elizabeth Anderson proposes
a set of functionings necessary for equality of democratic citizenship, which
includes adequate safety, health and nutrition, education, mobility and com-
munication, the ability to interact with others without stigma, and to partici-
pate in the system of cooperation.42 Ingrid Robeyns suggests a more extensive
list of fourteen capabilities necessary for gender equality in Western societies,
which includes life and physical health, mental well-being, bodily integrity
and safety, social relations, political empowerment, education and knowledge,
mobility, domestic and paid work, time-autonomy, respect, and religion.43 Ido
not propose to discuss the details of these lists here or to weigh into the debate
about the appropriate method for arriving at any such list, although Ido favor
a more context-specific, deliberative approach. However, Iwould argue that
lists of capabilities such as those of Anderson and Robeyns help to identify the
kinds of opportunity conditions required for self-determination. Furthermore,
the idea that a just society is one that ensures access, above a certain min-
imum threshold, to the valuable capabilities, however they are defined, shows
39
See Amartya Sen, Inequality Reexamined (Cambridge:Cambridge University Press, 1992) for a
detailed analysis of the notions of capability and functioning.
40
See the chapter on social choice in Sen, Idea of Justice.
41
For a recent version of this list, see Martha Nussbaum, Creating Capabilities.
42
Anderson, Justifying the Capabilities Approach to Justice,83.
43
Robeyns, Sens Capability Approach.
why ensuring the opportunity conditions for self-determination is a matter of

social justice.
A further reason for thinking that capabilities theory provides a useful vo-
cabulary for articulating the opportunity conditions for self-determination is
that it marries a focus on the importance of individual self-determination with
attention to the social and political constraining and enabling conditions of
self-determination. Thus, on the one hand, the theory recognizes that different
people will value different capabilities, and hence argues that a just society
ought to ensure equality of access to a wide range of opportunities but leave
it to individuals to choose which particular capabilities to realize beyond the
threshold. On the other hand, capability theorists, like theorists of relational
autonomy, are sensitive to the role of the social environment and social, po-
litical, legal and economic institutions in enabling or constraining individual
self-determination. Of particular concern to both theories is the phenomenon
of adaptive preference formationthe phenomenon whereby persons who are
subject to social domination, oppression, or deprivation adapt their prefer-
ences (or goals) to their circumstances, eliminating or failing to form pref-
erences (or goals) that cannot be satisfied, and even failing to conceive how
their preferences might differ in different circumstances.44 In social situations
characterized by oppression or deprivation, the problem then is not just that
restricted opportunities constrain self-determination but also that the inter-
nalization of these constraints can shape individuals sense of who they are
and what they can be and do. Here, the causal interdependence between the
opportunity conditions for self-determination, the competence and authen-
ticity conditions for self-governance, and the conditions for self-authorization
become evident.
My aim in this section has been to provide a brief characterization of au-
tonomy as self-determination and its two conditions:freedom and opportu-
nity. Ihave argued that relational autonomy theorists prioritize opportunity
conditions, in the sense that the freedoms that count as important are those
that promote substantive opportunity. Ihave also suggested that capabilities
theory provides a vocabulary for articulating the kinds of opportunity condi-
tions for self-determination that is particularly amenable to the concerns of
relational autonomy theorists.
44
For the original formulation of the problem of adaptive preferences, see Jon Elster, Sour
Grapes:Studies in the Subversion of Rationality (Cambridge:Cambridge University Press, 1983). For
its relevance in motivating capabilities theory, see Sen, Inequality Reexamined; Nussbaum, Women
and Human Development; Nussbaum, Adaptive Preferences and Womens Options, Economics and
Philosophy 17 (2001):6788. For a recent discussion of its relevance for gender oppression, see Serene
Khader, Adaptive Preferences and Womens Empowerment (NewYork:Oxford University Press, 2011).
For discussion of its relevance to relational autonomy theory, see Natalie Stoljars contribution to this
volume.
4. Self-governance:ARelational Conception
Self-governance, as Ihave defined it, involves having the skills and capacities
necessary to make choices and enact decisions that express or cohere with
ones reflectively constituted diachronic practical identity. The self-governance
dimension of autonomy picks out autonomy conditions (competence and
authenticity) that are in some sense internal to the person, whereas the
self-determination axis identifies external, structural conditions. However,
from a relational perspective, the distinction between internal and external
conditions is complicated. If persons are socially constituted, then external
conditions, including our social relations with others, shape the process of
practical identity formationthe self of self-governanceand the develop-
ment of the skills and competences required for governing the self. On the
one hand, this is not necessarily problematic from a relational perspective
it points to the facts of developmental and ongoing dependency and to the
extensive interpersonal, social, and institutional scaffolding necessary for
self-governance. On the other hand, in contexts of social oppression, as is
shown by the phenomenon of adaptive preference formation, severely limited
opportunity conditions can deform the process of practical identity formation
and impair the development and exercise of autonomy competence.
Authenticity conditions specify what it means to be self-governing with re-
spect to ones motivational structurethat is, what it means for a choice, value,
commitment, or reason to be ones own. Relational autonomy theorists are
critical of conceptions of self-governance, such as hierarchical or endorsement
conceptions, which analyze authenticity in terms of structural features of an
agents will at the time of choice or action. The argument, in brief, is that such
synchronic accounts and criteria for authenticity that appeal to relations of
internal coherence within the agents will (e.g., identification, endorsement)
fail to account for the historical processes of practical identity formation.45
In particular, they fail to account for the internalized effects of psychological
oppression, that is, the way oppression shapes agents practical identities and
motivational structures, for example their preferences, values, and cares.46
Furthermore, criteria for coherence, such as identification, wholehearted-
ness, or endorsement, seem to rule out any kind of ambivalence or internal
45
Examples of synchronic hierarchical theories include Dworkin, Theory and Practice of Autonomy;
Frankfurt, Importance of What We Care About. Endorsement theories include Gary Watson, Free
Agency, Journal of Philosophy 72 (1975):205220; Korsgaard, Sources of Normativity.
46
For the first critique along these lines, see Marilyn Friedman, Autonomy and the Split-Level
Self, Southern Journal of Philosophy 24:1 (1986):1935. For a more detailed overview, see Mackenzie
and Stoljar, Introduction:Autonomy Refigured. Christman is the best-known proponent of the his-
torical approach, first articulated in Christman, Autonomy and Personal History, Canadian Journal
of Philosophy 21 (1991): 124 and developed in its most sophisticated form in Christman, Politics of
Persons.
psychic conflict or fragmentation as inconsistent with self-governance.47 But

this is both normatively problematic and unrealistic, since ambivalence and
some degree of inner psychic conflict or fragmentation seem to be inescapable
aspects of practical identity formation.
On the basis of these criticisms, some relational theorists have rejected any
kind of authenticity condition for autonomy, arguing that the concept of au-
tonomy should be uncoupled from notions of practical or self-identity and
understood in terms of other conditions, such as sociorelational structural
conditions, normative (or some other) competence conditions, or accounta-
bility and answerability.48 As Isaid at the outset, if we think of autonomy as the
unitary concept of self-governance, requiring a single set of necessary and suf-
ficient conditions, then these other conditions all need to be understood as ad-
ditional or competitor conditions for self-governance. My proposal is to think
of these conditions as identifying different dimensions or axes of the multidi-
mensional concept of autonomy. In my view, the concept of self-governance
cannot be uncoupled from the self who is supposed to govern itself, and Ithink
the notion of practical identity provides a plausible conception of the self. For
this reason, Idont think the authenticity condition should be rejected. Rather,
Ifavor adopting a more nuanced analysis of authenticity, such as that proposed
by John Christman, which is responsive to the criticisms that relational theo-
rists have raised against standard accounts of authenticity. Christman under-
stands authenticity as non-alienation upon (historically sensitive, adequate)
self-reflection, given ones diachronic practical identity and ones position in
the world or as reflective self-acceptance.49
Competence conditions specify the range of competences or skills a person
must possess, to some degree at least, to be self-governing. These include cog-
nitive skills, ranging from minimally specified capacities to understand and
process information, to more complex capacities for critical reflection and
reasons responsiveness as well as volitional skills, such as self-control, and
For critiques along these lines, see Benson, Taking Ownership; Oshana, Autonomy and
47
Self-Identity, in Christman and Anderson, Autonomy and the Challenges to Liberalism,7797.

48
For sociorelational structural conditions see Oshana, Personal Autonomy in Society. For norma-
tive (or some other) competence conditions see Benson, Autonomy and Oppressive Socialization.
For accountability and answerability see Benson, Feminist Intuitions and the Normative Substance
of Autonomy; Benson, Taking Ownership:Authority and Voice in Autonomous Agency; Westlund,
Rethinking Relational Autonomy.
49
Christman, Politics of Persons, 155. Much more needs to be said of course to flesh out Christmans
analysis of authenticity. For detailed discussion see Christman, Politics of Persons, 149163, and his con-
tribution to this volume. It is important to stress that Christmans reflection condition is hypothetical;
an agent is not required to reflect on every aspect of her practical identity, that is, every value, commit-
ment, preference, trait, and characteristic to be authentic with respect to that aspect of her motivational
structure. Rather, the requirement is just that were she to so reflect she would not feel alienated from it.
Note that in his contribution to this volume Christman weakens the requirement of reflection further,
now preferring to talk of reflexive rather than reflective self-acceptance.
decisiveness. These skills have been extensively discussed in the philosoph-

ical literature, along with the ways self-governance can be compromised by a
range of mundane volitional or cognitive failingssuch as weakness of will,
self-deception, failures of self-control, and prudential irrationalityas well
as by less common but more severe, volitional, and cognitive threats, such
as compulsions, phobias and neuroses, addictions, and delusions and other
forms of psychosis.
Relational autonomy theorists concur that such skills are certainly impor-
tant for self-governance. However, relational approaches extend the philosoph-
ical understanding of autonomy competence by developing a more complex
and capacious analysis of the skills involved in self-governance. Relational
theorists argue that philosophers have overly rationalized the competences re-
quired for critical reflection while neglecting an array of other skills, including
emotional skills, such as emotional responsiveness, and being able to inter-
pret ones own and others emotions; imaginative skills, which are necessary
for envisaging alternative possible courses of action, or imagining oneself
otherwise and engaging in self-transformative activities; and social or dia-
logical skills required for self-understanding or self-knowledge.50 On a rela-
tional view, although self-governance does require a degree of self-knowledge,
self-knowledge is not necessarily developed primarily through introspection.
This is not only because we are not motivationally self-transparent but also be-
cause self-knowledge, indeed the self itself on this view, is constituted through
ongoing social or dialogical interaction with others. This focus on the dialog-
ical dimensions of self-governance echoes Charles Taylors dialogical concep-
tion of the self. 51 This kind of dialogical approach does recognize, of course,
that some kinds of social interaction can thwart self-understanding and de-
form the formation of a self-governing practical identity. But the corrective is
not to think that we can be self-governing only if we are freed from social in-
fluence. It is rather to promote social relationships that foster self-governance.
It is worth noting here that the autonomy skills identified by relational the-
orists and the relational emphasis on the constitutive role of social interaction
overlap with four of the items on Nussbaums list of capabilities:52
1. Senses, imagination and thought, which she characterizes as being
able to use the senses, to think, imagine and reasonand to do
these things in a truly human way, a way cultivated by an adequate
education.
50
For discussion of the importance of these skills for autonomy, see especially Meyers, Self, Society
and Personal Choice; Mackenzie, Imagining Oneself Otherwise, in Mackenzie and Stoljar, Relational
Autonomy, 124150; Mackenzie, Critical Reflection, Self-Knowledge, and the Emotions, Philosophical
Explorations 5:2 (2002):186206.
51
Charles Taylor, Sources of the Self (Cambridge:Cambridge University Press,1989).
52
Nussbaum, Creating Capabilities,3334.
2. Emotions, which she characterizes as being able to have

attachments to people and things outside ourselves.... Not having
ones emotional development blighted by fear and anxiety.
3. Practical reason, which she characterizes as being able to form a
conception of the good and to engage in critical reflection about the
planning of oneslife.
4. Affiliation, which involves two components:(a)Being able to
live with and towards others, to show concern for other human
beings, to engage in various forms of social interaction; to be able to
imagine the situation of another; and (b)Having the social bases
of self-respect and non-humiliation; being able to be treated as a
dignified being whose worth is equal to that of others.
Although Iwould include affiliation (b)under the axis of self-authorization,
the overlap between these four items on Nussbaums list and the autonomy
skills identified by relational theorists suggests that these items are centrally
focused on the capabilities required for self-governance. Nussbaum also
claims that practical reason and affiliation play an architectonic role in orga-
nizing and pervading the other capabilities.
This overlap raises a puzzle about why Nussbaum insists that her capabili-
ties approach does not assume the value of autonomy and why she is at pains
to distinguish the capability of practical reason from autonomy. She gives two
reasons for this position.53 First, although she regards the capability of practical
reason as central to a dignified human life and thinks that this capability raises
claims of justice against the state to provide the kind of educational opportuni-
ties that enable the development of this capability, she argues that individuals
who have acquired the capability may choose not to exercise it. Autonomy,
on the other hand, is an exercise concept.54 Nussbaum does concede that it
is difficult to develop the capability for practical reason without exercising it,
but she seems to think that this is possible. I think this is implausible and
that it makes more sense to say that individuals may exercise the capacity for
practical reason, or self-governance, to greater or lesser degrees, sometimes
by choice, but often because their social environment has not supported the
development of the relevant competences.
Nussbaum provides the most extensive discussion of these reasons in Nussbaum, Political
53
Liberalism and Respect: A Response to Linda Barclay, SATS: Nordic Journal of Philosophy 4:2
(2003):2544.
54
Nussbaum also seems to think that those who support the value of autonomy, such as Raz, think
that a life in which a person does not exercise this capacity is less valuable and less worthy of respect.
Idont think that Raz holds a derogatory view of persons whose lives are not autonomous. His view is
rather that individual autonomy is crucial for living a decent human life in contemporary democratic
societies and that therefore the state has an obligation to support its development and exercise by
ensuring that individuals have an adequate range of significant opportunities.
Nussbaums second reason, which Ithink underpins the first, concerns the
relationship between autonomy and epistemic and normative authority. Itake
up this issue in the following section.
5. Self-authorization:ARelational Conception
Self-authorization, as I have defined it, involves regarding oneself as having

the normative authority to be self-determining and self-governing. In other
words, it involves regarding oneself as authorized to exercise practical control
over ones life and to determine ones own values and identity-defining com-
mitments. In the recent literature on social and relational autonomy, a number
of theorists (myself included) have characterized what I am here calling
self-authorization as a necessary condition for self-governance.55 Inow think
this is a mistake and that self-authorization is a separate axis or dimension of
autonomy, which involves three conditions:that the person regards herself as
the kind of agent who can be held accountable and answerable to others for
her reasonscall this the accountability condition; that the person stands in
certain self-affective relations to herself, in particular relations of self-respect,
self-trust, and self-esteemcall this the self-evaluative attitudes condition; and
that such attitudes in turn presuppose that the person is regarded by others as
having the social standing of an autonomous agentcall this the social recog-
nition condition.
In explaining the accountability condition it is helpful to contrast it with
one way in which the notion of self-authorization might be interpreted.
Self-authorization is sometimes understood as requiring substantive epistemic
and normative independence from authority. That is, the self-authorizing
person does not look to any external authoritywhether another person or
Godin determining what to believe, value, or commit herself to. I think
Nussbaum equates autonomy with such a conception of self-authorization, as
does Oshana. This is the second reason why Nussbaum insists that the capa-
bility of practical reason is different from autonomy:because she thinks a com-
mitment to the value of autonomy is inconsistent with religious commitment.
So she says:Autonomy thus means something positive, that one gives one-
self laws and engages in critical reflection; but it also means something nega-
tive, that one denies that God is a necessary part of the justification of moral
claims.56 In my view, this is a mistaken interpretation of self-authorization,
55
See, e.g., Benson, Feminist Intuitions and the Normative Substance of Autonomy; Benson,
Taking Ownership; Mackenzie, Relational Autonomy, Normative Authority and Perfectionism;
Westlund, Rethinking Relational Autonomy.
56
Nussbaum, Political Liberalism and Respect,41.
and in fact Nussbaums capability of affiliation is centrally concerned with

self-authorization, as Icharacterizeit.
Self-authorization involves, first, regarding oneself as responsible to oneself
and as answerable and accountable to others, for ones beliefs, values, commit-
ments, and reasons. Taylors notion of autonomy as involving responsibility for
self articulates this kind of accountability condition, as does Stephen Darwalls
second-person standpoint.57 In the recent literature on autonomy, Benson and
Westlund are its most eloquent exponents.58 The basic idea is that as social
agents we are subject to others demands that we account for our decisions,
explain the reasons for them, and take responsibility for those decisions. To
be self-authorizing, a person must not only be capable of understanding and
responding to this social demand but must also regard herself as a valid source
of self-authorizing claims.
It is important to clarify that being accountable does not mean that we
must be accountable to certain specific others or that others must agree with
our beliefs, values, commitments, reasons, or decisions. Nor does it mean
that we must actually be held to account for each and every belief, value, and
so on. What it requires is what Westlund refers to as a disposition to be an-
swerable:that we are prepared to take responsibility as agents, for our beliefs,
values, and commitments. This means being responsive to others requests for
explanation, being prepared to provide reasons for our beliefs and values, and
being willing to defend or revise them in light of others critical questioning.
It also requires, in turn, that we regard ourselves as authorized to hold others
accountable in the same way. Iwant to stress that the accountability condition
is fundamentally social and dialogical, not only because its structure is dyadic
or second-personal but also because our reasons, values and commitments,
indeed our sense of our selves, emerge only through this kind of dialogical
interaction. Iwould also suggest, contra Nussbaum, that accountability, thus
understood, is not inconsistent per se with religious commitment. But it is in-
consistent with dogmatic forms of religious commitment, or any other form
of dogmatism, for that matter, which involve appeals to authority that bypass
a persons reflective agency.
Regarding oneself as accountable involves having a sense of ones epistemic
and normative authority with regard to ones life and ones practical commit-
ments. This in turn seems to require the agent to have certain self-evaluative
attitudes, in particular attitudes of appropriate self-respect, self-trust, and
57
See Taylors distinction between strong and weak evaluation in What Is Human Agency?
in Taylor, Human Agency and Language: Philosophical Papers 1 (Cambridge: Cambridge University
Press, 1985), 1544; Taylor, Sources of the Self. For Darwalls account of the second-person stand-
point, see Darwall, The Second-Person Standpoint:Morality, Respect and Accountability (Cambridge,
MA:Harvard University Press,2006).
58
Benson, Taking Ownership; Westlund, Rethinking Relational Autonomy.
self-esteem. This second condition for self-authorization has been articu-

lated in different ways in the literature on social and relational autonomy.59
Imention here Joel Anderson and Axel Honneths version of the claim, which
brings out the connections between self-authorization and self-governance.60
Anderson and Honneth characterize self-respect as an affectively laden
self-conception that underwrites a view of oneself as the legitimate source of
reasons for acting.61 To regard oneself with appropriate self-respect is to re-
gard oneself as the moral equal of others, as having equal standing to have
ones views and claims taken seriously. Self-trust or basic self-confidence is
the capacity to trust ones own convictions, emotional responses, and judg-
ments. Anderson and Honneth argue that this attitude underpins the kind
of reflexive self-interpretation that is central to self-governance. Because our
practical identities are complex and dynamic, deliberating about what to do
involves self-interpretationfor example, working out which desires should
constitute reasons for us, which commitments are most important, which
emotional responses we should attend to, and how to reconcile conflicts aris-
ing from different social roles. The attitude of self-esteem or self-worth is a
fundamentally evaluative stance toward oneself; it involves thinking of ones
life, ones commitments, and ones attitudes as meaningful, worthwhile, and
valuable.
It is always possible to find or think up examples of heroic persons who hold
appropriate self-evaluating attitudes even in situations where they are despised
and humiliated by others. However, psychologically these self-evaluative atti-
tudes are typically dependent on intersubjective social relations, that is, on
whether a person is regarded as a respect-worthy, autonomous agent by others.
These attitudes are also constituted within normative structures and practices
of social recognition.62 Hence, they are vulnerable to others failures, or refusals,
to grant the person appropriate recognition in a range of different spheres:for
example, in her interpersonal relations; at work and in the institutions of civil
society; and as a citizen. Such failures of recognition are quite typical in social
relations involving domination or inequalities of power, authority, or social
and economic status, especially when these are inflected by gender, race, eth-
nicity, or disability. Relational autonomy theorists claim that the internaliza-
tion of non- or misrecognition can corrode an agents sense of self-regard,
59
For different variants, see, e.g., Trudy Govier, Self-Trust, Autonomy, and Self-Esteem, Hypatia
8 (2003):99120; Paul Benson, Free Agency and Self-Worth, Journal of Philosophy 91 (1994):650
668; Carolyn McLeod, Self-Trust and Reproductive Autonomy; Mackenzie, Relational Autonomy,
Normative Authority and Perfectionism.
60
Anderson and Honneth, Autonomy, Vulnerability, Recognition and Justice.
61
Anderson and Honneth, Autonomy, Vulnerability, Recognition and Justice,132.
62
See Anderson, Vulnerability and Autonomy Intertwined for a detailed analysis of the inter-
relation between the psychological and normative dynamics of social recognition with respect to the
self-evaluative attitudes.
thereby undermining her sense of herself as self-authorizing. For example, unless

she is incredibly resilient, an employee who is subject to constant sexist and racist
abuse by her boss and co-workers may find it very difficult to maintain an appro-
priate sense of self-regard.63
It might be thought that the self-authorization axis of autonomy is unduly de-
manding and likely to be met by relatively few people. For example, the account-
ability condition for self-authorization may seem to assume an unrealistically
high level of articulacy and open-mindedness, while the self-evaluative attitudes
condition may seem to assume an unrealistically high level of confidence in one-
self. Who among us, for example, does not doubt the worth or meaningfulness
of our commitments? And if we did not would that make us more autonomous
or just arrogant and insufferable? In response, it is important to emphasize that
this condition requires that a person have appropriate self-evaluative attitudes.64
Iwould also reiterate that each of these conditions should be understood as a
matter of degree rather than an all-or-nothing affair. Different persons are likely
to meet these different conditions to different degrees. Indeed any one individual
is likely to meet them to different degrees in different contexts and domains.
It is also important to counter here a common, but I think misguided,
charge against supposedly overdemanding conceptions of autonomy, namely,
that they can be used to sanction objectionable forms of paternalistic inter-
ference with the choices of persons who may already be socially marginal-
ized:those with limited capacities for self-governance or who do not regard
themselves and are not recognized by others as self-authorizing agents.65 In
response I would argue that if the social relations, institutions, norms, and
practices to which a person is subject fail to recognize her social standing as
a self-authorizing agent and that if this kind of misrecognition impairs her
capacities for self-governance it does not follow at all that paternalistic inter-
ference is warranted. Rather, first, it provides grounds for criticizing those
social structures as unjust and reforming them, and second, it provides rea-
sons for providing targeted social support and resources to scaffold the per-
sons capacities for self-governance.66
The impact of work on personal autonomy has been curiously neglected in the literature. This
63
neglect is a serious omission, given the importance of work for our lives. For valuable analyses of work
and autonomy, see Beate Roessler, Meaningful Work:Arguments from Autonomy, Journal of Political
Philosophy 20:1 (2012):7193; Andrea Veltmans contribution to this volume.
64
The question of how to distinguish appropriate from misplaced self-evaluative attitudes raises
complex psychological and normative issues that do need to be addressed by theorists who emphasize
the importance of these attitudes for autonomy. However, it is beyond the scope of my discussion in
this chapter to address this question.
65
For criticisms along these lines, see, e.g., John Christman, Relational Autonomy, Liberal
Individualism, and the Social Constitution of Selves, Philosophical Studies 117 (2004):143164; Jules
Holroyd, Relational Autonomy and Paternalistic Interventions, Res Publica 15:4 (2009):321336.
66
For a more detailed argument for this claim, see Mackenzie, Importance of Relational
Autonomy and Capabilities for an Ethics of Vulnerability.
The conditions for self-authorization, then, although they may be met to

varying degrees, are underpinned by a fundamentally normative conception
of persons as moral equals who ought to be treated as respect-worthy, or dig-
nified beings, and as self-authorizing sources of valid claims. This notion of
persons as respect-worthy or dignified beings is categorical; it is neither con-
text specific nor a matter of degree.67
Conclusion
I have argued that autonomy should be understood as a multidimensional con-

cept, with three distinct but causally interdependent axesself-determination,
self-governance, and self-authorizationeach with its distinct conditions,
which may be realized to different degrees. There are two obvious objec-
tions to this multidimensional analysis. One objection might be to insist that
there must be a unitary idea underlying these different dimensions of au-
tonomy. The other objection might be to suggest that if my analysis is correct,
then it is not clear why we should hold onto the concept of autonomy at all.
Why not just stop talking about autonomy and start talking instead about
self-determination, self-governance, and self-authorization?
In response to the first objection, my reply is that if our goal is to under-
stand the role played by the concept of autonomy in our social and normative
practices, then insisting on the necessity of an underlying unitary idea is mis-
taken. Although philosophers are attracted to tidy conceptual boundaries, and
for that reason are also attracted to the project of determining the necessary
and sufficient conditions for the correct application of our concepts, many
social concepts, such as the concept of autonomy, are not readily amenable to
such analyses.
In response to the second objection, my reply is that although Ihave argued
that the concept of autonomy is multidimensional rather than unitary, it
would be a mistake to stop talking about autonomy and just talk instead about
self-determination, self-governance, and self-authorization. This is because
the concept of autonomy plays an important role in diverse social and norma-
tive practices. The advantage of the multidimensional analysis is that it enables
us to tease apart the different dimensions of the concept while still holding
onto the term autonomy.
67
This categorical notion of respect is what Stephen Darwall refers to as recognition respect, as dis-
tinct from appraisal respect, which involves positive evaluation of a persons character or accomplish-
ments and attributes and which is context dependent, earned, and a matter of degree. Darwall, Two
Kinds of Respect, Ethics 88 (1977):3649. It should be noted that having appropriate self-evaluative
attitudes seems to be dependent on both recognition and appraisal respect.
There are several further advantages of the multidimensional analysis. First,

distinguishing self-determination from self-governance and self-authorization
enables us to give due weight to the social and political preconditions for au-
tonomy and to underscore that securing these conditions is a matter of social
justice. But it also accounts for our important intuitions about autonomy
namely, that a person may be self-governing and self-authorizing to some de-
gree even in situations of crushing oppression where these preconditions are
severely curtailed, and conversely that a person may enjoy all the freedoms
and opportunities necessary for self-determination but nevertheless fail to be
self-governing or have a sense of herself as a self-authorizing agent. I think
many of the examples and counterexamples in the literature on relational au-
tonomy appeal to these conflicting intuitions and that the unitary concept
of self-governance is inadequate to account for them. Rather than devoting
our energies to thinking up yet more examples and counterexamples, theo-
rists of autonomy should instead focus our attention on analyzing in more
detail and depth the different dimensions of autonomy and the interrelations
betweenthem.
Second, in doing so, we may find that in different contexts certain dimen-
sions of autonomy, and the conditions that fall under them, may be more sa-
lient than others. Further, since the conditions for autonomy may be satisfied
to differing degrees, it may be important to set different thresholds for meeting
these conditions in different contexts. For example, with respect to democratic
citizenship, it might be important to set high thresholds for self-determination,
in the form of requiring extensive opportunities and the liberties that enable
them, but quite low thresholds for an agent to count as self-governing. It might
also be important to place a great deal of emphasis, in interpersonal relations,
in civil society, and at the level of the state, on ameliorating the social condi-
tions that undermine self-authorization. In the context of medical decision
making it may be the case that the most salient dimensions of autonomy are
self-governance and self-authorization. Thus, health professionals ought to
focus on whether patients meet the competence and authenticity conditions
for self-governance to a sufficiently high degree and whether they seem ca-
pable of meeting the accountability condition. In cases where individuals do
not meet the required threshold for one or other of these conditions, then
extensive social scaffolding in the form of support and resources provided by
carers may be required to enable autonomous decision making, except in cases
where it is more appropriate for a legally appointed guardian or family mem-
bers to make the decision on the patients behalf.
Third, even if it is the case that the salience of the different dimensions
and conditions varies with context, the importance of the multidimensional
analysis is that it also draws attention to the interrelations among these dis-
tinct dimensions and conditions. For example, even if the self-governance
and self-authorization dimensions of autonomy are most relevant to making
specific decisions in medical contexts, it is nevertheless crucial, both in medical

ethics and in bioethics more generally, to be attentive to the self-determination
dimension of autonomy. In particular, we need to be attentive to the ways that
constrained opportunities and restricted liberties may increase peoples vul-
nerability, making them more open to exploitation, coercion, and manipula-
tion and less able to regard themselves as self-authorizing agents.
Acknowledgments
I would like to thank the editors, Mark Piper and Andrea Veltman, for ex-
tremely helpful comments on an earlier version of this chapter. Thanks also to
Natalie Stoljar for extensive and ongoing discussions about autonomy. Earlier
versions of the chapter were presented at the conference on Thick (Concepts of)
Autonomy, Institute for Advanced Study in Bioethics, University of Muenster,
and to a seminar at the University of Amsterdam. Thanks to audiences on both
occasions for helpful comments.
{3}
Relational Autonomy and Independence

Marilyn Friedman
As Catriona Mackenzie and Natalie Stoljar write in their introduction to

Relational Autonomy, a relational approach to autonomy involves emphasiz-
ing the social nature of the self and the social relations and conditions that
are necessary for the realization of autonomy.1 Defenders of a relational ac-
count might think their approach rules out any individualistic claims about
autonomy. That would be a mistake. In this chapter, Iexplore how conceptions
of autonomy can combine relational and individualistic aspects.
These possibilities are worth exploring because individualistic aspects of
autonomy, such as independence, can form worthwhile goals and should
therefore not be disregarded. Iframe my discussion with special reference to
the discussion of liberal individualism by Jennifer Nedelsky in her recent book
Laws Relations: A Relational Theory of Self, Autonomy, and Law.2 Nedelsky
champions a relational account of autonomy and strongly disavows liberal in-
dividualism. Yet her own account of autonomy includes numerous ideas that
seem individualistic. Section 1 summarizes key ideas in Nedelskys approach.
In the second section, I discuss what Nedelsky regards as a popular, lib-
eral individualist model of autonomy, namely, self-made men and women.
According to relational autonomy theorists such as Nedelsky, self-made
persons are a myth and an illusion. Iargue by contrast that the debate over
self-made persons is not relevant to the question of how to conceptualize
autonomy. Instead, it is better understood as part of the debate over capi-
talism. Furthermore, in the debate over capitalism, self-made persons are
not necessarily mythic or illusory. They can be interpreted as genuine empir-
ical possibilities so long as the term is taken nonliterally. This is not to deny
that self-made persons are controversial. However, as Ishall suggest, even in
1
Catriona MacKenzie and Natalie Stoljar, eds., Relational Autonomy: Feminist Perspectives on
Autonomy, Agency, and the Social Self (NewYork:Oxford University Press, 2000),4.
2
Jennifer Nedelsky, Laws Relations: A Relational Theory of Self, Autonomy, and Law
(NewYork:Oxford University Press, 2011). Some parts of this essay draw on my review of Nedelskys
book in the University of Toronto Law Journal, Volume 63, Number 2 (April2013).
Relational Autonomy and Independence 43
popular discourse, the controversy has to do with economic possibilities in

capitalism, not with personal possibilities of autonomy.
In the third section, Idelve more deeply into the concept of independence.
Idiscuss ways in which the concept of independence can cohere with an oth-
erwise relational account of autonomy. Ialso explain reasons why members of
subordinated or dominated groups might want to retain independence as an
aim of action.
1. Liberal Individualism
Some of the defenders of a relational conception of autonomy are avid critics

of individualistic interpretations of autonomy, which they regard as antithet-
ical to relational interpretations. Jennifer Nedelsky, for example, particularly
targets liberal individualist conceptions of autonomy in her recent book Laws
Relations.3 This prevalent Anglo-American mode of thought, according to
Nedelsky, conceptualizes human beings as essentially disconnected from one
another and fails to treat relationships as constitutive of persons. Nedelsky
maintains that, for liberalism, human beings are fundamentally independent
rational agents.4
Nedelsky does not completely reject liberalism. She accepts some liberal
goals, such as equality. However, she thinks that egalitarian projects will be
better advanced in relational terms. What she challenges is the dominance of
liberal individualism. She insists that her relational approach offers a deeper
and more insightful way of considering what liberals may value, such as equal-
ity. Nedelsky emphasizes that relationships enable individuals to develop the
capacities valued by liberalism, such as reason, autonomy, and liberty.5
Nedelsky acknowledges that egalitarian liberals such as Joseph Raz, Jeremy
Waldron, and Will Kymlicka have already paid attention to the social nature
and conditions of human beings.6 Yet Nedelsky maintains that, in the work of
academic political theory and philosophy, attention to the social co-exists
with a still powerful individualism. She claims that the liberal attention to
relationships remains peripheral, and relationships are not treated as consti-
tutive of things like cognitive capacities. Nedelsky sometimes treats liberal
3
Nedelsky, Laws Relations, e.g.,7.
4
Nedelsky, Laws Relations, 3, 86. Ihave argued by contrast that mainstream theorists of autonomy,
whether specifically liberal or not, do not generally reject, and often mention, relational or social
connections in their accounts of personal autonomy. See Marilyn Friedman, Autonomy and Social
Relationships:Rethinking the Feminist Critique, in my Autonomy, Gender, Politics (NewYork:Oxford
University Press, 2003),8197.
5
Nedelsky, Laws Relations, 7,9,86.
6
Nedelsky, Laws Relations, 382, note 14. Nedelsky also lists Ronald Dworkin, Owen Fiss, Frank
Michelman, and Cass Sunstein.
individualism and relationality as if they were mutually exclusive. Yet at other

times she treats them as mutually compatible, the theoretical difference being
a matter of which one is more heavily emphasized.7
Nedelsky writes that liberal individualism is exemplified by the liberal con-
ception of autonomy as independence. Nedelsky regards independence as
mythic and an illusion. Models of autonomy as independence are not really
possible for human beings. In her view, the possibility of autonomy always
requires dependence on others. Dependence and interdependence are perva-
sive throughout all human lives. Arelational approach denies most conven-
tional claims to independence and denies that autonomy is fundamentally
about independence.8 Nedelsky thus writes:
Relations (at all levels) are then not simply conditions under which a free-
standing self emerges. It is the very nature of human selves to be in inter-
action with others. In important ways, they do not exist apart from these
relations.... Autonomy is a mode of interacting with others. Relations, in-
cluding those with collectives of all sorts, become not just potential threats
to autonomy but its source.9
Nedelsky claims, in addition, that when people strive to be independent, this
can happen only...at the cost of subordinating the others who do the (unac-
knowledged) work made necessary by dependence.10
It is not always clear what Nedelsky is rejecting under the rubric of inde-
pendence although, at one point she gives what sounds like a definition of
independence:the capacity to make decisions without being subject to an-
yone elses preferences, judgments, or choices....11 Yet Nedelsky often con-
strues autonomy in ways that sound very similar to this idea. For example, she
writes that autonomy is about finding ones own law. The laws in question are
to include commitments to oneself that will foster ones capacity for creative
interaction with others, a capacity that Nedelsky regards as encompassing au-
tonomy. For Nedelsky, autonomy captures that aspect of creative interaction
that involves acting on new patterns of engagement that alter received pat-
terns. Laws are ones own when they involve values and demands that arise
from within each person and are not imposed from outside. Nedelsky writes
that an autonomous person shapes her own life and does not accept a defini-
tion of herself that is conferred by other persons. Individual values should not
be superseded by collective powers. The essence of autonomy, according to
Nedelsky, Laws Relations, 9, 36,8687.

7
Nedelsky, Laws Relations, 5, 118, 43, 45, 46,134.

8
9
Nedelsky, Laws Relations,55.
10
11
Nedelsky, Laws Relations, 142. Nedelsky claims that in a 1964 essay Charles Reich associates this
sort of autonomy with property. Reich, The New Property, Yale Law Journal 73(5):733787.
Nedelsky, may well be norms that a person recognizes as stemming from the
significance or aims of the life she is living.12
What is noteworthy is that Nedelsky gives all of these individualistic
sounding claims a relational interpretation. As quoted already, she asserts that
autonomy is a mode of interacting with others. She argues that, for a rela-
tional self, nothing is ever simply ones own. She holds that the capacity to find
ones own law must be nurtured by relationships and that the content of ones
own law is meaningful with reference to shared norms, values, and concepts.
Indeed, her whole project includes giving a relational interpretation to key con-
cepts that liberalism supposedly interprets individualistically, concepts such as
autonomy, equality, dignity, individuality, and individuation. For Nedelsky, all
of these values arise out of community and constitutive relationships.13
Also significant is Nedelskys insistence that her relational interpretation
does not deny the significance of the experience of making or claiming some-
thing as ones own.... As well, recognizing the centrality of relationships does
not entail denying the value of privacy or solitude. Rather, it reveals how these
values are sustained by practices of allowing people to step back from each
other in certain ways.14 In general, Nedelsky insists that her approachis
very attentive to and concerned with the particular individual. Indeed,
Iwould say that Ihave insisted on such attention. Much of the relational
approach is a call to attend to the individual in her particularity, which must
include her particular context.15
Nedelsky writes that her approach
participates in the vision of self-defining interiority as the core of what is
human, the ground and purpose of the liberal commitment to freedom. At
the same time, Iurge a return to relationality as constitutive of self.... One
might say that Iam trying to claim the best of both worlds, to propose a kind
of synthesis. Iembrace the notion of the unique, infinite value of each indi-
vidual, and the value of interiority, and the value of the ability of individuals
to shape their own lives. But Ireject the liberal variants of these values that
fail to see the central role relation plays in each of them.16
Thus, Nedelskys relational interpretation, by her own account, does not
undermine the apparent individualism of many of her own claims about au-
tonomy. Instead she accommodates their apparent individualism. She either
reconceptualizes those claims in relational terms or simply combines them
as they stand with relational ideas. This prompts a question: Why not treat
12
Nedelsky, Laws Relations, 46, 166167, 123124, 121,33.
13
Nedelsky, Laws Relations, 55, 49, 124,36.
14
Nedelsky, Laws Relations, 49,33.
15
16
independence in the same inclusive manner? If autonomy can be thought of as

a mode of interacting with others,17 why not think of independence as a mode
of interacting with others? In that way, individualistic claims about autonomy
would take their place alongside relation claims as two aspects of the same
underlying reality.
However, we might wonder whether a simple juxtaposition of individual-
istic and relational claims wouldnt be self-contradictory. How could both be
true together? Nedelsky recognizes this puzzle and regards it as one that she
must solve or dissolve. The puzzle, as she puts it, is that autonomy seems to be
both grounded in and opposed to relationality.18 Let us consider how Nedelsky
resolves this apparent inconsistency.
Nedelsky notes that there are indeed cultural narratives in which autonomy
seems to involve opposing ones community and changing or rising above
its traditional roles and practices. Her examples include stories that honor
women for exerting nontraditional, heroic efforts. Nedelsky thinks that such
oppositional accounts would make it self-contradictory to regard autonomy
also as being inseparable from relationships and requiring relationships over
the course of ones life. Nedelsky argues that if the prevalent conception of
autonomy is as opposition to the collective, a relational approach will remain
puzzling and seem exceptional. Nedelsky claims that her approach involves
changing the framework that treats autonomy in this oppositional manner.
She asserts that her changes make the self-contradiction disappear.19
Sometimes, Nedelskys solution seems to be simply to deny certain indi-
vidualistic sounding claims. For example, she frequently simply rejects inde-
pendence as a conception of autonomy. Yet such claims are in tension with
Nedelskys synthesis between individualistic and relational ideas. In her
synthesizing mode, Nedelsky herself expresses individualistic ideas such as
the value of the ability of individuals to shape their own lives.20 Granted,
she interprets these ideas relationally. Nedelsky thus often retains both indi-
vidualistic and relational ideas in her account, although the deepest level of
understanding is relational. Individual selves are individuals, but they are con-
stitutively relational in the last analysis. Autonomy can be described individu-
alistically, say, as resisting collective norms, but those feats can be realized only
in virtue of relational aspects of selves.
Does this synthesized approach succeed as an account of autonomy? Ibe-
lieve it does and, in fact, will rely in Section 3 on the method of synthesis to
argue against Nedelsky that independence, too, can be retained in accounts of
autonomy. However, some reservations about Nedelskys exact position seem
17
18
19
20
called for at the outset. First, Nedelsky is mistaken to suggest that the dom-
inant picture of autonomy is opposition to the collective. That is not the
dominant picture among philosophers, at any rate.21 No major philosophical
account of autonomy holds that autonomy requires opposition to any col-
lective norms, roles, or practices. Philosophical accounts generally include
procedural requirements that a person reflect on her choices, values, or com-
mitments. When philosophical accounts specify the sorts of commitments a
person must hold to be autonomous, these commitments usually have to do
with valuing autonomy or regarding oneself in a certain way, such as trusting
oneself.22 Opposition to collective norms or traditions is not typically a philos-
ophers defining feature of autonomy. Also, although philosophical accounts
do not always emphasize the collective origin of an individuals perspective,
this is usually not precluded either.23
It is possible that Nedelsky is aiming her critique at popular conceptions
of autonomy rather than those in academic philosophy. However, philosoph-
ical works are still legitimate sources of counterexamples since there is a sub-
stantial amount of work in philosophy on autonomy and, especially, on liberal
theories in which autonomy figures as a fundamental value. Also, Nedelsky
makes clear that academic philosophy is among her targets; for example, she
points to a still powerful individualism within academic political theory and
philosophy.24
A second reservation about Nedelskys combined individualized and rela-
tional approach is this. There is need for clarification of how the individu-
alism that remains in Nedelskys account coheres with the relationality that she
claims as fundamental. It seems that, on Nedelskys account, individualistic
ideas about autonomy arise when one first thinks about autonomy, ideas such
as finding ones own law. Relationality seems to enter Nedelskys account
when those starting points are explained. One interpretation of Nedelskys
approach is that the relational explanations of autonomy replace and thereby
eliminate the individualism of the initial understandings of autonomy. On this
interpretation of Nedelsky, the deeper explanatory understanding gets at what
is really true about autonomous selves. Individualistic appearances of au-
tonomy would turn out to have been mere appearances without any deeper
reality.
21
This is important because it is not clear that many people outside philosophy actually deal in any
depth with the concept of autonomy.
22
On self-trust, see, e.g., Joel Anderson and Axel Honneth, Autonomy, Vulnerability, Recognition,
and Justice, in Autonomy and the Challenges to Liberalism:New Essays, edited by John Christman and
Joel Anderson (Cambridge:Cambridge University Press, 2005), 127149.
23
Nedelsky seems to direct her critique toward at least three different social sectors: academic
theorists (including philosophers), law, and government. My comments deal with academic theorists
unless otherwise indicated.
24
However, another interpretation is possible according to which the deeper

relational explanations of autonomy do not necessarily undermine individ-
uality but rather explain how it emerges. On this interpretation, the deeper
relational understanding of autonomy need not undermine the apparent in-
dividualism of notions such as finding ones own law.25 Both relational and
individualistic claims would be simultaneously true of autonomy on this in-
terpretation because they would be regarded as applying to different modes or
levels of understanding. Consider, for example, Nedelskys claim that the ca-
pacity to find ones own law (an individualistic attribution) must be nurtured
by relationships26 (a relational, causal explanation). Both relationality and in-
dividuality can thus be (different) aspects of autonomy, expressing different
modes of understandingit.
Nedelsky does not clarify whether her relational account of autonomy is
meant to eliminate its initially apparent individuality or to explain and pre-
serve it. This is an issue that merits further elaboration in regard to what we
might call the metaphysics of her account.
A third reservation regarding Nedelskys combined account is that she
exaggerates the problem that arises when autonomy is construed as oppo-
sition to ones collective. Opposition to a collective to which one belongs is
not necessarily individualistic. Opposition to a collective of ones own does
not entail that one must be free of affiliation with any and all collectives. For
one thing, there are usually multiple collectives in anyones life. Someone may
gain the perspective to defy the conventions of her own limited community
by associating with another collective that is committed to differing values
or norms. The relationships that enable a particular persons autonomy need
not be the very same ones she defies. Even if they were the same relationships,
there is nothing contradictory in a persons criticizing or repudiating, say, the
very people or doctrines that endowed her with the capacity for critical re-
flection that she now exercises against them. Why think that a masters tools
are necessarily so effectively master oriented that they cannot be used by their
own community members to destroy the masters house?27 Both opposing
and reaffirming the norms of a collective to which one belongs each constitute
a relational engagement with ones collective.
In the next section, Itake a temporary detour around what Nedelsky con-
siders to be a particularly troubling exemplar of liberal individualist accounts
of autonomy, namely, the idea of self-made men and women. Iargue that
this idea does not in fact challenge the conception of relational autonomy, as
Nedelsky thinks, but instead plays a different role in popular debate. Because

25
Nedelsky, Laws Relations.

26
27
Cf. Audre Lorde, The Masters Tools Will Never Dismantle the Masters House, in Audre Lorde,
Sister Outsider:Essays and Speeches (Berkeley, CA:Crossing Press, 1984), 110113.
of that different role, challenges to the idea of self-made persons do not un-
dermine any particular conceptions of autonomy. My conclusion to the next
section will be that the idea of self-made persons should be dropped from
debates over autonomy.
2. Self-made Men andWomen
As part of her critique of liberal individualism, Nedelsky directs attention to

the liberal vision of human beings as self-made and self-making men. She
rejects the idea of the self-made man as an exemplar of autonomy. This image
of autonomy provides an ugly caricature of autonomy. People are not simply
self-made, writes Nedelsky; they are constituted socially.28
One drawback of Nedelskys rejection of self-made men is that she does not
take account of what is meant by those who invoke this image. In ignoring its
defenders, Nedelsky risks attacking a straw person. In this section, Iattend to
what some ordinary people mean by self-made person and to what some social
thinkers have written about this image. Some of its defenders in particular
use the image in an empirically plausible manner. However, this usage does
not vindicate any particular conception of autonomy, for, as Iwill suggest, the
debate over self-made men or women is not really a debate about autonomy
in any of the usual philosophical senses of autonomy. Instead, it is part of a
debate over capitalism. The self-made person, Iargue, is supposed to exem-
plify an important kind of success in a capitalist economy that, if it can really
be achieved by participants in the economy, would undermine certain criti-
cisms of capitalism. Most importantly, even a successful critique (or defense)
of the idea of the self-made person would contribute nothing to an account of
autonomy.
Are self-made persons empirically possible? One of my maternal uncles
was regarded as a self-made man by my mothers family. The uncle Ihave in
mind, like the rest of my mothers generation in her family, was a Lithuanian
Jewish immigrant to the United States. My uncle had had little formal edu-
cation in Lithuania and had arrived in the United States with few resources.
Also, he was partially disabled by the loss of sight in one eye due to a child-
hood illness. Once he arrived in the United States, he got a job as a cutter in a
clothing factory, a ground-floor-level job that involved some physical danger.
Cutters in the garment industry sometimes lost fingers using power tools to
cut long, stacked rows of fabric. My uncle worked hard, went to night school,
and attracted the attention of the factory owner, who eventually promoted
my uncle up the ranks. After several decades, he reached the position of vice
28
president of the company. From this position, he eventually bought out the
original owner. At the time of his death in 1967, my uncle was a millionaire
(when that was still a lot of money!) who owned the company in which he had
started out as a low-ranking laborer. He reached his successful position from
humble origins that involved no inherited wealth, immigrant status, English as
a second language, and few if any special privileges.
Nedelsky claims that terms such as independent and self-made man refer to
myths and illusions. In her view, no one is really independent or a self-made
person.29 Nedelsky would be right about self-made man if this expression were
taken literally. Aliteral interpretation would treat a self-made man as one who
created himself out of nothing. This idea is, of course, metaphysically ridicu-
lous. (I can assure you that my family did not think my uncle created him-
self out of nothing. My grandmother would have had a thing or two to say
aboutthat.)
There is another interpretation of self-made person that may not be meta-
physically impossible but would come very close to it. According to this sec-
ond interpretation, a self-made man is one who depended on no other persons
in any way on his path toward workplace success. This notion would also be
wrong because any successful career person had to have depended on other
persons and on social institutions for the care, support, technology, means of
communication, and so on that allowed the successful person to grow from
childhood and do what he or she did to succeed.
To be sure, several contemporary social commentators agree with Nedelsky
in rejecting the image of the self-made person. Brian Miller and Mike Laphams
book The Self-Made Myth:The Truth About How Government Helps Individuals
and Businesses Succeed30 shows the central role played by government in creat-
ing the conditions that enable people to gain economic prosperity. The gov-
ernment provides public education, a regulatory environment that protects
businesses, copyright and intellectual property laws, an infrastructure in-
cluding roads and airports, and support for the Internet. Miller and Lapham
also observe that success is partly due to luck, fortunate timing, and sources of
social capital such as race, gender, appearance, and upbringing.31 Social capital
is obviously relational.
29
Nedelsky, Laws Relations, 43, 118, 120. Asimilar position is defended by Iris Young, Autonomy,
Welfare Reform, and Meaningful Work, in The Subject of Care:Feminist Perspectives on Dependency,
edited by Eva Feder Kittay and Ellen K. Feder (Lanham, MD:Rowman & Littlefield, 2002),4060.
30
Brian Miller and Mike Lapham, The Self-Made Myth: The Truth About How Government
Helps Individuals and Businesses Succeed (San Francisco: Berrett-Koehler Publishers, 2012); cited
favorably by Sara Robinson, Self-made men debunked, http://www.salon.com/2012/04/30/
self_made_men-debunked_salpart/print/.
31
Miller and Lapham, Self-Made Myth,23.
Also Mike Myatt, contributing to Forbes Magazine online, disparages the

concept of the self-made man or woman.32 He regards as obviously false and
ridiculous the self-made man propaganda floating around business circles.
He agrees that there are inspiring stories such as that of Bill Gates but believes
that every success involves the contributions of many people, such as family,
friends, advisors, and teachers. Myatt writes that mature leaders tend to rec-
ognize and seek out the contributions of others.33
Yes, a self-made person would be a ridiculous idea if interpreted literally or
close to it. It is possible that some defenders of self-made persons do under-
stand this idea in the impossible, literal sense. However, to criticize only this
weakest version of the idea is to attack a straw person.
It is more worthwhile to search for a defensible, nonliteral version of the
idea of self-made person and see where that idea leads. This figurative version
of the idea turns up in some defenses of it. Aself-made man or woman figur-
atively is one who entered the workforce with minimal resources, few special
privileges, and few or no influential contacts but who nevertheless worked his
or her way up to a position of economic or career success. Aself-made person
is the main author of her successful labors; her success is not thought to be due
merely to luck. My uncles narrative fits this picture. He entered the workforce
at a working-class level with few resources and the difficulties of an immigrant
trying to survive in a foreign language and culture, and he ended his work life
as a wealthy businessowner.
On the interpretation Iidentify as plausible, to call someone self-made is to
assert not a near impossibility but rather an empirically possible narrative of a
distinctive sort. It is a narrative that credits someone with a life of hard work
at an income-producing job, starting with few if any special privileges and ul-
timately achieving economic success. To be sure, this narrative is a debatable
one (see note 46); however, the debate is not about autonomy.
A noteworthy feature of self-made persons in the empirically plausible
sense is that, even though this idea is not about autonomy, it is still compatible
with a relational analysis. The person without special economic advantages
doubtless acquired habits of hard work from family, teachers, and friends. She
entered an economy created by others using means of technology and com-
munication also created by others. And so on. Defenders of the plausible con-
ception of self-made persons rely on the fact that, while everyones life depends
on relationships, not everyone has a life story that conforms to the self-made
narrative in its empirically possiblesense.
The empirically possible conception of self-made persons that Ihave iden-
tified appears in a nineteenth-century essay called Self-Made Men by the
32
MikeMyatt,Self-MadeManNoSuchThing,http://www.forbes.com/sites/mikemyatt/2011/11/15/self-
made-man-no-such-thing/print/.
33
Myatt, Self-Made Man,1,2.
ex-slave Frederick Douglass.34 Douglass interpreted self-made man in a rela-

tional manner yet still believed the concept had a legitimate use. Relationally,
he wrotethat
Properly speaking, there are in the world no such men as self-made men.
That term implies an individual independence of the past and present which
can never exist.... [N]o possible native force of character, and no depth
of wealth and originality, can lift a man into absolute independence of his
fellow men, and no generation of men can be independent of the preceding
generation.35
Thus, Douglass rules out the existence of literally self-made men. However, he
still insists that the expression self-made men is eminently descriptive of a
class.36 He explains:
Self-made men are the men who, under peculiar difficulties and without the
ordinary helps of favoring circumstances, have attained knowledge, useful-
ness, power and position and have learned from themselves the best uses to
which life can be put in this world, and in the exercises of these uses to build
up worthy character. They are the men who owe little or nothing to birth,
relationship, friendly surroundings, to wealth inherited or to early approved
means of education, who are what they are, without the aid of any favoring
conditions by which other men usually rise in the world and achieve great
results.37
Douglass surely has ex-slaves in mind when he adds that self-made men some-
times rise in the world in defiance of all the efforts of society and the tendency
of circumstances to repress, retard and keep them down.38
Douglasss initial explanation of why some men succeed in unfavorable cir-
cumstances while other men do not is this: allowing only ordinary ability
and opportunity, we may explain success mainly by one word and that word
is WORK! WORK!! WORK!!! WORK!!!! Douglass later qualifies this point
to take account of the changed nature of work under slavery:Men cannot be
depended upon to work when they are asked to work for nothing.39 With a few
qualifications such as these, however, Douglass states his final messagethus:
Frederick Douglass, Self-Made Men, in The Frederick Douglass Papers, Series One, vol. 4, edited
34
by John Blassinghame and John McKivigan (New Haven, CT:Yale University Press, 1992), 545575.
See also http://www.monadnock.net/douglass/self-made-men.html. References here are to the online
version.
35
Douglass, Self-Made Men,4.
36
37
38
39
Douglass, Self-Made Men, 12,14.
My theory of self-made men is, then, simply that they are men of work....
[H]onest labor faithfully, steadily and persistently pursued, is the best, if not
the only, explanation of their success.... Other agencies co-operate, but this
is the principal one and the one without which all others would fail.40
Douglass is clear that self-made men need relationships with others. However,
this need for relationships does not prevent self-made man from having a
qualified, nonliteral meaning, namely, as that of a man (today:person) who
comes from humble origins and achieves great success due principally to her
or his own hardwork.
Another example of my interpretation of self-made persons can be found
in an article on an otherwise objectionable website on The Art of Manliness.
The article is entitled 25 of the Greatest Self-Made Men in American History
and is coauthored by Brett McKay and Kate McKay.41 The authors explain what
they mean by self-mademan:
anyone who attains far greater success than his original circumstances would
have indicated was possible. The self-made man often has to overcome great
obstacles to achieve his goals. Self-made men attain their success through
education, hard work, and sheer willpower. While no man is an island, its
not external help or special relationships that make the crucial difference in
the self-made mans rise.... Nor is luck the deciding factor.... While there
are always many factors to success, all are subordinate to work, which is the
great key to success.42
McKay and McKay do not deny that education, relationships (no man is an
island), and luck play a role; what they claim is that the deciding factor, the
great key to success is how hard the self-made man works. The concept of a
self-made person is, in their view, the concept of someone who achieves ca-
reer success that is unusual in light of his humble origins and is based, among
other things, most importantly on his hard work. Their qualified terminology
again allows that relationships could be included among the conditions that
together explain success in these cases. However, the authors single out hard
work as the most important factor. Understood in this way as a comparative
concept, the idea of a self-made person becomes harder to dismiss as mythic
or illusory.43
40
41
Brett McKay and Kate McKay, 25 of the Greatest Self-Made Men in American History, http://
artofmanliness.com/2008/12/28/self-made-men/print.
42
McKay and McKay, 25 of the Greatest Self-Made Men,1.
43
Readers might disagree with the twenty-five examples of self-made men that McKay and McKay,
25 of the Greatest Self-Made Men, list. However, the list does span something of a political and eco-
nomic range. It includes Clarence Thomas, Barack Obama, Ronald Reagan, Harry Truman, Harry
Reid, Frederick Douglass, Booker T.Washington, Henry Ford, Ralph Lauren, Sean Combs, and Ben
& Jerry (326).
Recall that Miller and Lapham challenge the concept of the self-made
person on a very specific point. They are not making a general claim that eve-
ryone is dependent on other persons in many ways. They are instead making
the more specific claim that every successful businessperson depends on gov-
ernment.44 Their argument is with those who reject government programs
and regulations in particular. At least two sorts of government programs and
regulations are relevant:those that intervene in business activities and those
that provide welfare services for the poor and needy.45 Let us consider welfare
programs more specifically.
Critics of welfare programs might agree that everyone is dependent on some
others in various ways but argue that we are not all dependent on government
handouts; they might cite self-made persons as examples of those who do not
depend on government programs. Miller and Lapham, as noted already, argue
specifically that successful business people are dependent on government sup-
port in one form or another. This argument makes it even clearer that debates
over self-made persons are irrelevant to the debates over autonomy because
relational theorists of autonomy are not trying to make the case that autonomy
requires dependence on government in particular.
However, Miller and Laphams argument is very relevant to what I be-
lieve the concept of self-made person is really about. As I argued earlier, a
self-made person, as understood by those who defend the concept in a plau-
sible way, is a person who achieves great career success despite humble origins
and principally by hard work. If people can be self-made in this sense, then
being self-made is not humanly impossible, and human impossibility would
be the wrong reason for excluding the idea of self-made persons from accounts
of autonomy.
However, there is a right reason for excluding the image of self-made person
from debates over autonomy. The right reason is that the image of self-made
persons is not an image of autonomyat least not an image that necessarily
meets philosophical definitions of autonomy. For example, it is not about
self-reflection or acting authentically according to ones own values. Defenders
of the concept of self-made person are not assuming that a self-made person
reflects on her own values or commitments in any way, let alone endorses
them. Aself-made person may or may not exemplify autonomy in any philo-
sophical definition of theterm.
A self-made person is one who is (presumed to be) a success in con-
ventional capitalist terms. The debate over self-made persons, I suggest, is
Miller and Lapham, Self-Made Myth,78.

44
For a discussion of the concept of dependency in relation to welfare programs, cf. Nancy Fraser
45
and Linda Gordon, A Genealogy of Dependency: Tracing a Keyword of the U.S. Welfare State, in
Kittay and Feder, Subject of Care,1439.
implicitly part of debates over the nature and value of capitalism and the re-
lated debate over the justification of welfare programs. Praise for a self-made
person tends to reflect a capitalist, anti-welfare-program mentality. A suc-
cessful businessperson who started her work life in the paid labor force with
few resources (and no friends in high places) and worked her way up the
business ladder is an epitome of what capitalist defenders say is possible in
capitalism.46
Perhaps there could be a sense of self-made person that is relevant to au-
tonomy. If so, we should not simply speculate on what that sense would be
like. We should first find expressions of such an account by those who would
actually defend it. Otherwise, we would risk conjuring up, and then attacking,
a straw person. Thus far, my admittedly nonsystematic researches into public
debates over the idea of self-made persons do not reveal any important mean-
ings that are relevant to the concept of autonomy.
It thus appears that the insistence that self-made persons are a mythic illu-
sion fails to discredit either capitalism or autonomy. It fails to discredit capi-
talism because there are examples of self-made persons, in the nonliteral sense
its defenders have in mind. If a self-made person is someone who starts her
income-producing work life from humble origins, works hard in that capacity,
and becomes successful in conventional capitalist terms, then there do seem to
be some of those persons around. Ibelieve my uncle was an example. McKay
and McKay, as noted earlier, give famous contemporary examples. However,
and more importantly for this essay, even if self-made persons are genuine pos-
sibilities, this tells us nothing about how to understand autonomy. As Ihave
been arguing, the idea of self-made persons, even in the plausible, nonliteral
sense, is not about autonomy.
46
The economically successful person who rose up from humble origins is supposed to be evi-
dence that the inequalities of capitalism are not absolute barriers to success in the workplace. If some
people can be self-made, that is, can achieve economic success principally by hard work even from
humble origins, then, so the story goes, anyone who works hard can do so as well; poverty is no neces-
sary barrier to workplace success. From this perspective, welfare programs serve only those who refuse
to work hard or to work at all at an income-producing job. Refuting this claim is part of the significance
of the argument in Miller and Lapham, Self-Made Myth, that all businesspeople depend on government
in particular.
Thus, the existence of self-made persons does not bring the critique of capitalism to an end. Far
from it. Those who oppose capitalism have various options. They can argue that, even if a few people
can achieve career success from humble origins by working hard, such successes are so few in number
that their existence does not show that the system is inherently fair. They can debate the extent to which
such lives are still possible today. They can ask what costs and trade-offs are imposed on those who
dont succeed, and, especially, on their young children and other dependents? Critics of capitalism do
not yield any ground by conceding the possibility of self-made persons, in the sense that Douglass and
others definethem.
3.Independence
As noted in Section 1, Nedelsky criticizes the notion of independence as a mis-

guided model of autonomy. However, as in the case of the idea of self-made
person, the idea of independence also can be used in a nonliteral sense that
is genuinely possible for people to achieve. To understand this plausible sense
of independence, let us begin by noting that dependence is a scalar notion, a
matter of degree in some specific mode of acting or interrelating with others.
One can be more or less dependent on others in some particular way. Isuggest
that independence be understood as a relatively low degree of dependency, but
one that is still humanly possible.47 Also, my proposed notion of independence
allows for varied forms, just as there are varied ways in which people can be
dependent. Accordingly, independence should be understood as referring to
relatively low levels of dependency in some particular form of dependence.
Isuggest that this notion is reasonable and empirically plausible. Furthermore,
it is consistent with relational approaches such as that of Nedelsky.
Nedelsky generally rejects theories and concepts that deny or disregard
relationships that are constitutive of human beings and lives. While Nedelsky
does not define constitutive relationship, she does make clear at least two things
about what it means. First constitutive relationships shape people, both in
childhood and throughout life; they are formative relationships. Nedelskys
examples include relationships with parents, teachers, friends, and neigh-
bors but also relationship patterns such as gender norms and employment
practices. Second, constitutive relationships are not determinative. Nedelsky
appears to mean that constitutive relationships do not exhaustively determine
what people are or how they behave. Constitutive relationships, she insists,
leave room for freedom and autonomy.48
The meaning of independence that Ipropose, namely, that of being a rel-
atively low but still humanly possible form of dependence is consistent with
Nedelskys idea of selves as the product of constitutive relationships. Consider
someone who is said to be financially independent in virtue of having wealth
of her own and not needing to depend on financial support from someone
else. This person exhibits a low degree of financial dependence on others, but
her financial dependence is not nonexistent. She may rely on a bank or finan-
cial institution to protect her wealth, and she probably depends on a function-
ing economy to sustain her wealth. However, she legally owns resources that
are uncontested under her name, and these resources are more than sufficient
47
The idea that independence is a form of relationship involving dependency appears in
Nancy Chodorow, The Reproduction of Mothering: Psychoanalysis and the Sociology of Gender
(Berkeley:University of California Press, 1978), e.g., 187188.
48
to cover her financial needs. Why be reluctant to say that she is financially
independent?
This notion of financial independence leaves open the possibility that the
person in question is relatively dependent in nonfinancial ways. The finan-
cially independent woman may rely on other persons to grow her food, build
her dwelling, and provide her health care. Also the notion of financial inde-
pendence is quite consistent with the idea that the person in question was
constituted by early relationships with parents and teachers and continues
in adulthood to be constituted by relationships with friends and by gender
norms and employment practices. Thus, my proposed use of independent is
easily consistent with a relational approach to selves and to autonomy. On my
proposal, independence, interpreted as a relatively low form of some type of
dependency, need not be a myth or an illusion.
Why does the idea of independence matter to an account of autonomy? Are
there any special reasons for wanting to reclaim the idea of independence?
Ibelieve there are such reasons. Independence can identify a useful ideal for
disadvantaged or subordinated persons to aim at. It can constitute one legiti-
mate goal among many for emancipatory movements. Onora ONeill, for one,
cites financial independence as a valuable attribute for women worldwide.49
Poor women worldwide are often dependent on the economic terms and con-
ditions set by others. As ONeill writes:
They are vulnerable not only to low wages, low standards of industrial safety,
endemic debt and disadvantageous dependence on those who provide
credit, but also to disadvantageous patterns of entitlement within the family.
Debtors who need further loans for survival cannot make much fuss about
the terms creditors offer for purchasing their crops; the most dependent
womendaughters-in-law and younger daughters in some societiesare
acutely vulnerable both to market forces and to more powerfulkin.50
The term independence can serve to stand for a capacity that Nedelsky her-
self lauds, namely, the human capacity for creation in the shaping of ones life
and self. It can serve to guide a feminist concern that Nedelsky cites, namely,
freeing women to shape their own lives and to define themselves rather than
accepting the definition given by others (men and male-dominated society, in
particular).51
49
See Onora ONeill, Justice, Gender, and International Boundaries, in Onora ONeill, The
Bounds of Justice (Cambridge:Cambridge University Press, 2000), 162167. See also my discussion of
ONeill on these issues in Marilyn Friedman, Independence, Dependence, and the Liberal Subject,
in Reading Onora ONeill, edited by David Archard, Monique Deveaux, Neil Manson, and Daniel
Weinstock (Abingdon:Routledge, forthcoming), 111129.
50
ONeill, Justice, Gender, and International Boundaries,164.
51
To be sure, Nedelsky worries about independence for several reasons. She

worries that the use of the idea of independence is not only illusory but also
positively harmful. She writes that when people strive to be independent, this
can happen only...at the cost of subordinating the others who do the (un-
acknowledged) work made necessary by dependence.52 We are familiar with
the idea that, for example, successful careerpeople, usually men, are highly
dependent on family partners, usually women, whose domestic work is cru-
cial for allowing the career people to concentrate on building their careers.
However, if, say, a womans job outside the home requires her husband to
do more housework than he did before, this might not cause his subordina-
tion at all. And if the woman must hire someone to do child careoften a
working-class womanshe (the employer) can pay a decent wage, social se-
curity, and workers compensation as well as providing fair conditions of em-
ployment, thus minimizing the subordinating effects of paid child-carelabor.
If subordination were made necessary by dependence, as Nedelsky states,
then all dependency relations would subordinate someone. Independent striv-
ings would not be any different from dependent strivings in that respect. The ar-
gument has yet to be made, however, that either peoples dependencies or their
strivings to be independent necessarily result in subordinating others. Arela-
tionship to someone who is striving to be independent might make someone
more vulnerable to subordination. Or it might not; it is a contingent matter
whether subordination actually occurs in any given case. When it does, this
probably means that the person striving to be independent has actually become
more exploitative and thus more dependent in some ways on the now more
subordinate other person, so the problem is not independence per se. Thus,
these complex outcomes should not discourage oppressed persons from seek-
ing relatively greater independence in those areas in which their own indepen-
dence would reduce their vulnerabilities to subordination or abuse by others.
Nedelsky claims a second way the cultural esteem for independence can be
harmful. She suggests that when independence is culturally celebrated, people
who do not fit the image are devalued.53 Because different groups have une-
qual abilities to maintain the illusion of autonomy conceptualized as inde-
pendence,54 some groups will probably be stigmatized whenever autonomy
is revered as independence. Nedelsky is right to worry about scorn directed
against persons regarded as dependent. Those who fail to become independent,
that is, to reduce dependency to a relatively low level, may have lacked the ap-
propriate, constitutive relationships or material circumstances that could have
promoted their independence. They should therefore not be stigmatizedjust

52
Nedelsky, Laws Relations, 4243. On this point, Nedelsky cites Margaret Walker, Moral Contexts
53
(Lanham, MD:Rowman & Littlefield, 2003),192.

54
as people should not be stigmatized for failing to develop any capability for
which they lacked appropriate relationships or material circumstances.
However, the faulty application of an ideal of independence does not entail
that independence is a faulty ideal.55 The key point for discussions of autonomy
is that if independence is defined as the lesser form of some specific type of
dependence, then independence is humanly possible. In addition, persons can
be independent in some ways while being (more) dependent in otherways.
We can also ask whether any sort of independence is especially impor-
tant for autonomy. The notion of independent-mindedness comes to mind.
Saying that someone is independent-minded seems highly relevant to saying
she is an autonomous person, even if it is not the only relevant considera-
tion.56 Independent-mindedness need not be interpreted as a narrow kind of
rationality or rational agency. Mindedness can encompass reason, emotion,
desire, will, and so on.57 Nedelskys own conception of autonomy fits well with
a notion of independent-mindedness. As discussed earlier, Nedelsky writes
that autonomy involves finding ones own law, which consists of values and
commitments that arise from within the self and are not imposed by others.58
Nedelsky does not use the term independent-mindedness for her account, but
there is no need for her to reject it. As Ihave been suggesting, independence
can be understood consistently with a relational account of autonomy, and it
offers a valuable aspirational ideal for women and other subordinated persons.
One final point about liberalism, independence, and autonomy: Nedelsky
criticizes liberalism for focusing on the way people are threats to each other and
ignoring the positive or constructive ways they affect each other relationally.
Nedelsky suggests that regarding individuals as independent and as threats to
each other is at odds with regarding them as constitutive of each other.59
However, the two perspectives are not incompatible at all. First, an emphasis
on the threats posed by (some) others does not entail a belief that all others are
55
One widely scorned sort of dependency is dependency on government programs by those who
are poor and need those programs to meet their basic needs. Nedelsky is concerned that the esteem for
independence becomes part of a discrediting of the poor for their (welfare) dependency. Nedelskys
strategy for contesting this approach is to discredit the standard of independence so that it is not avail-
able for condemning people who depend on welfare services.
However, criticizing the idea of independence does not succeed in defending welfare programs.
One key question is whether welfare programs are required or justifiable as a matter of justice.
Unfortunately, discrediting independence does not make a positive case for the justice of social wel-
fare programs. Arguing that no one is literally independent and everyone is dependent does not by
itself justify government programs that redistribute wealth to those dependents who arepoor.
56
For a strong defense of the importance of favorable external social conditions, see Marina
Oshana, Personal Autonomy in Society (Aldershot, UK:Ashgate,2006).
57
The notion of independent mindedness may seem to exclude embodiment. This could be rem-
edied, however, by building embodiment into the accounts of reason, emotion, desire, and will. Most
scientific accounts of mindedness today are accounts of brain function.
58
59
nothing more than threats to the self. An emphasis on persons as (sometimes)

threats is consistent with regarding persons also as (sometimes) constitutive of
each other.60 Second, the notion of protecting individuals from the threats of
others does not presuppose that individuals are completely independent of all
other individuals. Indeed, if individuals were completely independent of each
other, they would not need any protections against each other. Only someone
in some sort of relation to another person can be a threat to that other person.
Regarding individuals as (sometimes) threats to each other thus exemplifies a
relational approach. Nothing about a relational approach requires us to deny
that some relations are harmful.
Third, it is obvious that people sometimes harm each other, and stopping
these harms is a legitimate purpose of the state. Alarge part of feminist theory
and practice consists of opposition to harms inflicted on women. It is therefore
odd for feminists such as Nedelsky to dismiss the liberal concern with prevent-
ing interpersonal harm. Indeed, most of Nedelskys examples throughout her
book, as she herself states, focus on harm suffered by women.61 Chapters5
and 8 in particular focus on violence against women and against children.
Nedelsky writes that women are always at risk of violence and...the violence
is tolerated, condoned, and not taken seriously.62 Perhaps Nedelsky should
have said that liberalism has not paid enough attention to threats to persons
especially to women.63 At any rate, the focus on persons as threats to each other
and as independent does not undermine liberalisms account of autonomy and
is consistent with a relational approach to autonomy.
I have argued that a relational approach does not preclude the ideas of
self-made person and independence. The idea of independence can easily
be incorporated into a relational account of autonomy. However, the idea of
self-made person is not really relevant to debates about autonomy. Instead, it
picks out a humanly possible sort of career narrative that should be assessed in
debates about the injustice of capitalist systems.
Acknowledgments
I am grateful to the editors, Andrea Veltman and Mark Piper, for helpful sug-
gestions on an earlier draft of this chapter.
60
People are no more always constitutive of each other than they are always threats to each other.
Also, lives can vary in the proportions of each sort of relationships.
61
62
63
Nedelsky, Laws Relations, 22, does say, at one point, that liberalism has not paid the right kind of
attention to threats to persons.
{4}
Autonomy? Or Freedom? AReturn to

PsychoanalyticTheory
Nancy J. Hirschmann
The idea of autonomy as developed by Kant, pertaining to the lone indepen-

dent self who is able to deduce and follow universal moral principles through
the use of a priori reasoning, has been the subject of much feminist critique,
even leading some feminists to argue that we should reject the idea of au-
tonomy altogether as a relic of a sexist, racist, and classist liberal individu-
alism.1 But relational autonomy feminists have brought attention to the need
to reconfigure autonomy rather than reject it. By recognizing that the problem
with autonomy is not the concept itself but rather the conception of it that fol-
lows from the individualist and naturally free and equal self that dominates
Enlightenment liberal thought, feminists have contributed to an important
rethinking of the concept.
The basic idea of relational autonomy may appear to be commonsensical to
most feminists:the powers, capabilities, talents, life plans, and visions of self
that are essential to our abilities to guide and shape our own lives come from
relationships. We are not isolated individuals but acquire our very sense of
self, our self-understandings, through living in relationship with others. Ihave
been an advocate of relational autonomy ever since my first book, in which
I traced the idea of relational autonomy to object relations theory, a school
of psychoanalysis made popular by Nancy Chodorows The Reproduction of
Mothering and embraced by Carol Gilligan as a foundation for the gendered
dimensions of her care model in In a Different Voice.2 Chodorows theory
was similarly deployed by feminist political theorists Nancy Hartsock, Jane
Flax, and Christine DiStefano as well as feminist philosophers like Sandra
1
See particularly Martha Fineman, The Autonomy Myth:ATheory of Dependency (NewYork:New
Press, 2004). But also see Sarah Lucia Hoagland, Lesbian Ethics: Toward New Values (Palo Alto,
CA:Institute of Lesbian Studies, 1989); Catharine Keller, From a Broken Web:Separation, Sexism, and
Self (Boston:Beacon,1986).
2
Nancy J. Hirschmann, Rethinking Obligation: A Feminist Method for Political Theory (Ithaca,
NY:Cornell University Press,1992).
Harding and Evelyn Fox Keller, not to mention social theorists like Dorothy
Dinnerstein and Jessica Benjamin.3 This feminist deployment of object rela-
tions theory arguably ushered in a revolution within the ongoing revolution of
feminist theory by kick-starting the argument that, rather than rejecting the
traditional work of care and relationship that women have historically been
assigned to, we should value its contributions and importance to the human
condition.
Yet I have become more ambivalent about the concept of relational au-
tonomy in recent years. Part of this ambivalence stems from the work on the
concept of freedom that I have done, in which I have increasingly come to
believe that whatever desires, preferences, and abilities our relationships may
have led us to have, we need to act by and for ourselves. Even when we act
wrongly, perhaps even against our own interests, we must be free to so act.
Ibelieve that this position diverges from the views of many (though not all)
relational autonomy theorists because freedom is still relevant even if an agent
fails a variety of autonomy tests and measures. Consider, for instance, a driver
who causes an accident while drunk:even if we excuse her decision to drive
(on the argument that her judgment was impaired), the choice to drink in the
first place when she knew that she had to drive home afterward clearly fails var-
ious autonomy tests of self-reflective judgment and rational self-interest. But
we consider her action free, which is why we hold her responsible for the harm
she causes. I may want to act recklessly, irrationally, even self-destructively,
but the freedom to choose to act in such ways is nonetheless important to our
conception of a person:we want to be able to make our own choices.
Part of my ambivalence, however, stems from going back to where rela-
tional autonomy started:object relations theory. The psychoanalytic origin of
relational autonomy is something that we tend to forget about, and this forget-
ting suppresses the dark side of relational autonomy, a suppressing that Ihave
been as guilty of as anyone.
That worry is the starting point for this paper. Ibegin by describing object
relations theorys account of relational autonomy and why Ithink it is prob-
lematic for feminism. Ithen turn to the ways feminist accounts of freedom and
agency can help us get beyond this problematic point of origin. My goal is not
to argue against autonomy per se, much less against relational autonomy. Indeed,
3
Nancy Hartsock, Money, Sex and Power: Toward a Feminist Historical Materialism (Evanston,
IL: Northeastern University Press, 1983); Jane Flax, Political Philosophy and the Patriarchal
Unconscious: A Psychoanalytic Perspective on Epistemology and Metaphysics, in Discovering
Reality: Feminist Perspectives on Epistemology, Metaphysics, Methodology, and Philosophy of Science,
edited by Sandra Harding and Merrill B. Hintikka (Dordrecht:D. Reidel, 1983); Christine DiStefano,
Configurations of Masculinity:AFeminist Perspective on Modern Political Theory (Ithaca, NY:Cornell
University Press, 1991); Sandra Harding, The Science Question in Feminism (Ithaca, NY: Cornell
University Press, 1986); Evelyn Fox Keller, Reflections on Gender and Science (New Haven, CT:Yale
University Press, 1985); Dorothy Dinnerstein, The Mermaid and the Minotaur:Sexual Arrangements and
Human Malaise (NewYork:Harper and Row, 1976); Jessica Benjamin, The Bonds of Love:Psychoanalysis,
Feminism, and the Problem of Domination (NewYork:Pantheon,1988).
Autonomy? Or Freedom? 63
feminists such as Drucilla Cornell, Wendy Brown, Linda Zerilla, and Ihave de-
veloped many shared concerns and features between autonomy and conceptions
of freedom, even as different as those theories are from one another.4 Rather,
Iwant to suggest that recognizing the problems of relationality may lead us to
the view that it is appropriate to conceive of autonomy only in tandem with a
theory of freedom. Indeed, perhaps relational autonomy should be better thought
of as relational freedom. As a more open-ended concept, Isuggest that freedom
may allow for more flexibility in how we theorize the choice-making subject, and
Iclose the paper with some speculative notes on that suggestion.
1. The Origins of Relational Autonomy
Object relations theory is a school of post-Freudian psychoanalytic theory

holding that boys experience different psychosexual development from girls
because women, rather than men, are the primary caretakers of children.
This psychosexual development entails different gender identification and
self-identification, which results in different conceptualizations of the self
Other relationship. The central tenet of object relations theory is one that
seems self-evident to feminist theorists and philosophers today, namely, that
human beings are created in and through relations with other human beings.5
The object in object relations is the caretaker, usually the mother; the mother is
called the object both in the sense that the mother is objectified by the infant
and in the sense that the self is formed by the otherthe subject by the object.
In contrast to Freuds focus on the oedipal period, object relations theory
focuses on the first three years of life, the pre-oedipal period. When it begins
life outside of the womb, an infant is not conscious of its existence as an entity
distinct from those who care for it. This realization of distinctionand along
with that, the realization of its complete helplessnesscomes at approximately
three months, as the infant starts to recognize its caretaker in recognizing its
own dependence. This means that the infant forms an intense attachment
a symbiotic one, from its perspectiveto the caretaker. Margaret Mahler, in
fact, names this first stage of life, from birth to six months, symbiosis.6 It is
common in this work to use it to refer to infants of both genders, and I follow
that custom here. Formation of this attachment is the most important task of
infants first three years.7
4
Drucilla Cornell, At the Heart of Freedom:Feminism, Sex, and Equality (Princeton, NJ:Princeton
University Press, 1999); Wendy Brown, States of Injury: Power and Freedom in Late Modernity
(Princeton, NJ:Princeton University Press, 1995); Linda Zerilli, Feminism and the Abyss of Freedom
(Chicago:University of Chicago Press, 2005); Nancy J. Hirschman, The Subject of Liberty:Toward a
Feminist Theory of Freedom (Princeton, NJ:Princeton University Press,2003).
5
Flax, Political Philosophy and the Patriarchal Unconscious,250.
6
Margaret Mahler, On Human Symbiosis and the Vicissitudes of Individuation (New York:
International University Press,1968).
7
Flax, Political Philosophy and the Patriarchal Unconscious,251.
In this period, infants form their primary identification, which involves

their growing perception of a demarcation between the self and the other. An
essential task of infantile development, it [primary identification] involves the
development of ego boundaries (a sense of personal psychological division
from the rest of the world) and of a bounded body ego (a sense of the perma-
nence of ones physical separateness and the predictable boundedness of ones
own body, of distinction between inside and outside).8 Since in most societies
caretakers of infants tend to be women, this primary attachment is almost al-
ways to a woman, and the infant develops its sense of self mainly in relation
toher.9
As infants develop motor skills, begin to interact with the world around
them, and most importantly learn language, they start to develop a sec-
ondary identity, which is their adult identity as an individual person. This
is also when the acquisition of gender and sexual identity takes place, which
are with rare exception firmly and irreversibly established for both sexes by
age three and often as early as eighteen months.10 The acquisition of gender
identity overlaps and coincides with the acquisition of received language, as
Merleau-Ponty calls ita language that already embodies sexist understand-
ings of gender difference.11 As children learn how to assign the abstract con-
cept of boy and girl to different persons, such as the self and the mother, they
learn about what gender is, as a concept and category in the structuring of
family relations.12
The reason that this is important is that when children learn which
gender they are, they also learn which gender their mothers are. In the pro-
cess, their primary identity (again, usually with their mother) is retroactively
assigned a gender (at the subconscious level, of course). This creates a different
set of problems for girls and boys as they emerge from the pre-oedipal to the
oedipal period. The bottom lineskipping over a great deal of the theoryis
that when children learn gender identity, girls learn that they are the same
as mother and boys that they are different from mother. Experienced up to
this point in an intense and exclusive relationship as an ungendered entity who
supplies all of the infants needs and wants, the mother represents all things
outside the self to the infant. But once children learn that their mother is a
8
Nancy Chodorow, Feminism and Difference:Gender, Relation, and Difference in Psychoanalytic
Perspective, Socialist Review 46 (JulyAugust 1979):54.
9
Nancy Chodorow, The Reproduction of Mothering: Psychoanalysis and the Sociology of Gender
(Berkeley:University of California Press, 1978),78.
10
Chodorow, Reproduction of Mothering, 150; Mahler, Human Symbiosis; see also Flax, Political
Philosophy, 251; Jane Flax, The Conflict between Nurturance and Autonomy in MotherDaughter
Relationships and within Feminism, Feminist Studies 4:2 (1978):171189.
11
Maurice Merleau-Ponty, The Prose of the World, edited by Claude Lefort, translated by John
ONeill (Evanston, IL:Northwestern University Press,1968).
12
Again, operating within a heteronormative framework in which there are only two genders.
woman, girl infants see themselves as the same as, or connected to, the entire
outside world, whereas boy infants sees themselves as separate from, or even
opposed to, thatworld.
That is, as a girl identifies with her mother and perceives a basic definitional
connection and continuity between herself and her mother, she also sees the
world itself, as represented by her mother, as continuous and connected with
her. Her otherthe not-me, the boundary of the selfis very much con-
nected to and part of the self. Self and other constitute a continuum, a move-
ment from the core self to external objects that she perceives as connected to
her core self. The self, to a significant degree, is the other.13 In contrast, for the
boy, self and other constitute a dichotomy; since gender is viewed as a mu-
tually exclusive category and it is culturally forbidden for him to identify as
female, he is culturally and psychically forced to expel that part of himself that
is embodied by his mother. But that is an impossible task. Under conditions
of patriarchy, the boy copes with this task by denial (of having been related),
by projection (women are bad; they cause these problems), and by domination
(mastering fears and wishes for regression by controlling, depowering, and/
or devaluing the object [viz. women]).14 The other is radically separated and
different from the self:other is female; self is male. In doing this, masculinity
is defined by the boy in negative terms, that is, as that which is not feminine
and/or connected to women.15
This gender difference leads to two different models of autonomy. For the
boy, autonomy is conceptualized reactively, as a reaction against the mother.
According to Evelyn Fox Keller, reactive autonomy confuses autonomy with
separation and independence from others.16 Reactive autonomy is static, ac-
cording to Keller, because it is locked into a reductive and negative conception
of the self as non-mother. If autonomy is defined as, in her terms, the psy-
chological sense of being able to act under ones own volition instead of under
external control and turns on individuality and the integrity of the self, then
reactive autonomy is self-defeating. In fact, it robs the individual of self-creative
agency, for such autonomy is premised on an artificial separateness that cannot
be sustained without repression and on abstract roles and rules, which are
13
Chodorow, Reproduction of Mothering,93.
14
Flax, Political Philosophy and the Patriarchal Unconscious, 253. Other feminists, like Hartsock
in Money, Sex and Power, further maintain that violence is another tool men use to establish their dis-
tinction from the mother. Indeed, erotic violence and the linkage of sex with domination is, Hartsock
argues, the logical outcome of this situation, a conclusion with which other object relations theorists
at least implicitly agree, such as Dinnerstein, Mermaid and the Minotaur; Benjamin, Bonds of Love;
Harding, Science Question in Feminism.
15
Chodorow, Reproduction of Mothering, 174, 175. By negative Ido not imply a normative judg-
ment; Imean only that it is defined as not female.
16
Keller, Reflections on Gender and Science,97.
external to the self.17 In David Bakans words, reactive autonomy manifests

itself in self-protection, self-assertion and self-expansion...in the formation
of separateness...in isolation, alienation and aloneness...in the repression
of thought, feeling, and impulse.18 This is the conception of autonomy that
coheres with the dominant notion that we find in the philosophical literature
beginning with Kant.19
By contrast, the girl cannot engage in the sort of repression and denial
that the boy engages ineven if she wants tobecause it is precisely at this
age that recognition of her female gender identity becomes important. She
thus depends on a close psychic tie with her mother even as the boys identity
depends on severing that tie. Moreover, because her identity with the mother
is ongoing, concrete, and continuous, the girl retainsand is encouraged to
retainher close psychic connection with the mother. The girls gender iden-
tity grows out of a personal, concrete relationship involving a diffuse iden-
tification with someone elses general personality; primary and secondary
identity become merged.
It is out of this deep psychic connection that relational autonomy devel-
ops: the self is conceived in terms of, and draws strength from, relation-
ships with others. Bakan says this notion of the self gives rise to a communal
personality; in his words, girls exhibit the sense of being at one with other
organisms...the lack of separateness...contact, openness, and union...non-
contractual cooperation.20 This is seen by feminist object relations theorists as
a preferred model, a source of strength, what Keller even calls the true self 21
and clearly superior to masculine experience. Boys and men are seen as inca-
pable of genuine love, empathy, and relationship:because boys are pushed out
of the preodedipal relationship and forced to curtail the empathic tie with the
mother, their understanding of love between two people becomes confused
with the fear of loss of self. Indeed, Chodorow goes so far as to claim that men
look to relationships with women for narcissistic-phallic reassurance rather
than for mutual affirmation and love.22 Women, by contrast, can maintain
genuine, deep, empathic relations. Indeed, according to the theory, that is why
they become mothers.
Ibid.
17
David Bakan, The Duality of Human Existence: Isolation and Communion in Western Man
18
(Boston:Beacon Press, 1966),55.

19
Though not all readers of Kant interpret his conception of autonomy in this way; see Nancy
J. Hirschmann, Gender, Class and Freedom in Modern Political Theory (Princeton, NJ: Princeton
University Press, 2008), chap.4, for my views of Kant on freedom and autonomy.
20
Bakan, Duality of Human Existence,56.
21
Keller, Reflections on Gender and Science,99.
22
2. Whats Wrong with Relational Autonomy?
This valorization of the feminine model of development has particularly

characterized work that brought this psychoanalytic framework into polit-
ical and moral philosophy. Political theorists and philosophers such as Jane
Flax, Nancy Hartsock, Christine DiStefano, and Iall deploy object relations
theory to critique central figures in canonical political philosophy as well as
the conceptual frameworks that emerged out of Enlightenment thought in the
structure of concepts such as obligation, freedom, and power.23 Other feminist
philosophers such as Sandra Harding and Evelyn Fox Keller used object rela-
tions theory to develop feminist philosophies of science and epistemology.24
All of these figures, including myself, made value judgments about the supe-
riority of the feminine model of relational autonomy. It is arguably the case
that Keller was the first in this group to focus on the notion of autonomy; she
offered an extremely positive, perhaps even romantic, vision of girls relational
autonomy as far superior to boys reactive autonomy, which cohered more to
the rule-governed conception that we have learned from Kant. But Kellers
ideas came from Chodorow, who in turn got them from David Bakans The
Duality of Human Existence as well as other object relations psychoanalysts
such as D.W. Winnicott and Melanie Klein. In these sources, the further back
one goes, the less feminist friendly the ideas underlying relational autonomy
appear.
One might well ask: so what? Why cannot philosophers today adopt the
idea and reject its baggage? As a political theorist, Ibelieve that history mat-
ters, and particularly the history of a concept like relational autonomy, because
Iam worried that this uncritical favoring of the feminine model unconsciously
persists. Feminists have still failed to acknowledge that relational autonomy
stems from a pathology exhibited by feminine development operating within
a sexist model of childrearing. If the boys problem is fear of connection and
the need to differentiate the self radically from the other, the girls difficulty
comes from an inability to separate adequately, which she needs to do to form
an adult identity. If the mother represents her infantile, primary self, the devel-
opment of a secondary, adult self is compromised by the need to shift her
identity back to the very same person it is from. Just as the theory posits that
the boys reaction against the mother and identification with an abstraction
makes it difficult for him to maintain connection and trust in later life, so does
the girls difficulty in breaking away from her primary identification inhibit
her sense of separateness. In Bakans notion of the feminine communal self, the
individual as such becomes merged with the other and never realizes a sense
23
Flax, Political Philosophy and the Patriarchal Unconscious; Hartsock, Money, Sex and Power;
DiStefano, Configurations of Masculinity; Hirschmann, Rethinking Obligation.
24
Harding, Science Question; Keller, Reflections on Gender and Science.
of individual identity. Mahler and Winnicott similarly show how the girls in-
ability to separate adequately from the mother creates serious problems of de-
veloping self-confidence, self-awareness, and even an understanding of the self
as an individual.25 And Flax notes that girls have the most serious difficulty in
psychically separating themselves from their mothers, a difficulty that carries
through into adulthood and into their own roles as mothers. She cites clin-
ical evidence that female therapy patients commonly express a lack of sense
where they end and their mothers begin, even in a literal, physical way.26 This
finding is supported by Macoby and Jacklins work on the gender-specific ways
mothers discipline boys and girls.27
As Chodorow put it in probably the most famous passage from The
Reproduction of Mothering: Girls emerge from this period with a basis for
empathy built into their primary definition of self in a way that boys do not.
Girls emerge with a stronger basis for experiencing anothers needs or feelings
as ones own (or of thinking that one is so experiencing anothers needs and feel-
ings).28 Ihave added emphasis to that conditional phrase because it is signifi-
cant that Chodorow puts it in parentheses, as if to obscure it for the reader. But
what it describes is the pathology of the lack of separation, the notion that we
think we experience others feelings as our own. We know that Icannot liter-
ally feel what you feel; even Hume acknowledged that sympathetic vibrations
were less pronounced in the string next to the one plucked.29 Parents who have
had a seriously ill or injured child and who have desperately wished or prayed
that they could take on their childs suffering know that we cannot feel the
others feelings; we can only imagine and mourn. We suffer at the suffering of
others, but it is of a qualitatively different nature. Believing that you literally
feel anothers feelings as your own, as Flaxs clinical patients reported, is a sign
of pathology called borderline personality syndrome.30
But many relational autonomy theorists tend to forgetor indeed per-
haps were never awarethat this is where the idea of relational autonomy
25
Bakan, Duality of Human Existence; Mahler, Human Symbiosis; D. W. Winnicott, The Family and
Individual Development (NewYork:Tavistock,1965).
26
Flax, Conflict between Nurturance and Autonomy,174.
27
Eleanor Macoby and Carol Jacklin, The Psychology of Sex Difference (Stanford: Stanford
28
29
David Hume, An Enquiry Concerning Human Understanding, in Enquiries Concerning
Human Understanding and Concerning the Principles of Morals, edited by L. A. Selby-Bigg, 3rd ed.,
revised by P. H. Nidditch (Oxford:Clarendon Press, 1975); Nancy J. Hirschmann, Sympathy, Empathy,
and Obligation: A Feminist Rereading, in Feminist Interpretations of David Hume, edited by Anne
Jacobsen (University Park:Pennsylvania State University Press,1999).
30
Jane Flax, Thinking Fragments: Psychoanalysis, Feminism, and Postmodernism in the
Contemporary West (Berkeley: University of California Press, 1990). Dinnerstein, Mermaid and the
Minotaur, which Ihave not brought into the discussion because even Ifind it too gloomy in its por-
trayal of the disastrous gender inequality that womens mothering produces, provides an even stronger
account of this pathology.
originates. Even those who have never read this literature, however, should be
aware that the idea of relational autonomy emerged out of this trajectory, from
nonfeminist analysts like Mahler and Winnicott to feminist nonphilosophers
such as Chodorow, Dinnerstein, Gilligan, and Benjamin, to feminist philos-
ophers and political theorists such as Keller, Harding, and Hartsock. From
there, the concept of relational autonomy was taken up by feminist theorists
and philosophers who may not have even read this literature, despite the effec-
tive critiques that leading feminist philosophers, such as Diana Meyers, have
made ofit.31
None of this is to deny that there are valid objections to object relations
theory. It is often difficult to recognize the family forms it depends on, namely,
the heterosexual two-parent family where the woman has primary responsi-
bility for child care. It may at times seem more appropriate to an episode of
Mad Men than to the complicated lives we live today. After all, feminists have
struggled for the past three decades to change this family structure, and many
of us are very proud of the fact that, in our own families, male and female
partners share equally (more or less) in childrearing. We are also rightly proud
of the fact that different family forms, such as gay and lesbian families, have
succeeded in part because of our political advocacy.
But let me offer a reminder: we are not the majority or the mainstream.
Studies show that although the sexual division of labor has shifted somewhat,
women still perform the bulk of labor associated with childrearing, and partic-
ularly early childrearing (the first six months of life).32 Although some women
have male partners who also took parental leave, it is still rare, even among
feminists, for men to be as actively engaged in early infant care as women.33
31
Diana T. Meyers, The Subversion of Womens Agency in Psychoanalytic Feminism:Chodorow,
Flax, Kristeva, in Revaluing French Feminism:Critical Essays on Difference, Agency, and Culture, edited
by Nancy Fraser and Sandra Lee Bartky (Bloomington:Indiana University Press, 1992). Idisagree with
her that Chodorow deploys a traditional feminine tendency to reduce care and nurturance to conflict
minimization and uncritical support (140), because Ithink Meyers downplays the role of capitalism
in Chodorows argument. However, Ithink she is correct that Chodorow ends up valorizing feminine
pathology in spite of herself. See also Meyers, Gender in the Mirror: Cultural Imagery and Womens
Agency (NewYork:Oxford University Press, 2002), esp. chap.1.
32
Rhona Mahoney, Kidding Ourselves:Breadwinning, Babies, and Bargaining Power (Boston:Basic
Books, 1996). According to research funded by the National Science Foundation, women are still doing
two to three times more household work:seventeen to twenty-eight hours per week for women versus
seven to ten hours per week for men. Bobbie Mixon, Chore Wars:Men, Women and Housework,
http://www.nsf.gov/discoveries/disc_summ.jsp?cntn_id=111458.
33
See, e.g., the symposium on Nancy J. Hirschmann, Mothers Who Care Too Much, Boston
Review, JulyAugust 2012, particularly the response of Lane Kenworthy, Time for Public Childcare,
and my response to him, The Sexual Division of Labor is the Problem. Particularly relevant to
readers of this volume is an American Association of University Women study showing that among
tenure-track faculty, parental leave is associated with positive outcomes at tenure for men but negative
outcomes for women, the implication being that many men use such leaves to advance their scholar-
ship whereas women use them to care for their infants exclusively. Mary Ann Mason, Title IX and
Babies:The New Frontier? Chronicle of Higher Education, November 29,2012.
But even so, object relations theory posits that the solution to patriarchy is
precisely what feminists strive for todayshared parenting and redistributing
the sexual division of labor. That is, although object relations theory may de-
scribe a natural process of human developmentin the sense that the devel-
oping infantile brain responds to the external stimuli of its relationships with
other people who shape its encounter with the physical worldthe particular
form those reactions take will differ by the specific cultural practices of child-
rearing. Within the profoundly sexist late-capitalist Western sexual division of
labor, we end up with gender dichotomy. Object relations theory shows that
the sexual division of labor is not just unfair to women:it is damaging to boys
and mens psychic and emotional well-being as well. Accordingly, feminists
who mother in heterosexual relationships where men share equally, as well as
gay and lesbian shared parenting families that radically break the heteronor-
mative pattern, offer significant hope for the feminist future we all seek. This
does not disprove the theory but rather supportsit.
It is also important to remember that psychoanalytic theory is best under-
stood not as a series of reductive empirical claims about infantile psychology
but is rather an interpretive framework. Certainly, part of the logic of psy-
chology and psychoanalysis entails the attempt to understand the meaning
that adheres to real social relations, hence its reliance on case studies and
interviews. But gender psychology also provides a symbolic language that can
help us understand these meanings in a new way by offering a larger theoret-
ical framework:larger because it allows questions that a liberal individualist
framework forbids; symbolic because the unconscious gains its most impor-
tant expression at the cultural level, in the symbols and structures that a cul-
ture adopts. As Chodorow says, psychoanalysis is a deeply embedded cultural
discourse, not simply a scientific one.34 Indeed, given that this literature was a
response to the decidedly sexist tendencies of Freudian psychoanalysis, its pri-
mary contribution from a feminist theoretical perspective is precisely to point
out that what has been taken to be truesuch as the classic Freudian concept
of penis envyin fact represents a particular masculinist reading of masculine
experience. Yet it is out of the sexist, heteronormative model of childrearing
that relational autonomy arises.
The grounding of the idea of relational autonomy in this pathology of fem-
inine development is what worries me. In our arguments critiquing Kantian
individualist morality we are at least tacitly endorsing the feminist object re-
lations view that masculine individuation and separation is less valuable than
womens connectedness. Ibelieve that this is problematic. For all our attention
to the need to stress relationship and connection, we have failed to stress the
34
Nancy Chodorow, Feminism and Psychoanalytic Theory (New Haven, CT:Yale University Press,
1989),179.
need for individuation and separateness, at which girls and women too often
fail. It is because of my unease about these aspects of relational autonomy, par-
ticularly the critique of individualism, that Ithink that freedom needs to be an
important part of the discussion. Its greater stress on individuality, combined
with its more flexible and perhaps less well-defined conception of the subject,
can usefully complement theories of relational autonomy.
3. The Role of Freedom
The meaning of freedom might seem to be fairly straightforward:doing what

Iwant without interference from anybody else, or as John Stuart Mill put it,
pursuing our own good, in our own way.35 This is clearly the dominant con-
ception of freedom we find in both law and political philosophy since at least
the seventeenth century. Thomas Hobbes famously declared, By liberty, is un-
derstood, according to the proper signification of the word, the absence of
external impediments:which impediments, may oft take away part of mans
power to do what he would.36 John Locke similarly noted that we must con-
sider what State all Men are naturally in, and that is a State of perfect Freedom
to order their Actions, and dispose of their Possessions, and Persons as they
think fit, within the bounds of the Law of Nature, without asking leave, or
depending upon the Will of any other Man.37
Early feminists similarly followed this line of thought:as Mary Astell acer-
bically asked in reference to Locke, if all Men are born free, how is it that all
Women are born slaves?38 Mary Wollstonecraft argued for the abolishment
of laws restricting womens control of property as well as for increased access
to education, urging Talleyrand to expand his new proposal for national edu-
cation to girls, whom he excluded. And Harriet Taylor argued not merely for
womens enfranchisement and for their control of property but also for the ex-
treme liberalization of divorce laws and even the abolishment of marriage, all
in the interest of enhancing womens freedom from control bymen.39
35
John Stuart Mill, On Liberty, in On Liberty and Other Essays, edited by John Gray
36
Thomas Hobbes, Leviathan, edited by C. B. Macpherson (NewYork:Penguin, 1985),189.
37
John Locke, Two Treatises of Civil Government, edited by Peter Laslett (NewYork:New American
Library, 1963), sec.2.4.
38
Mary Astell, Reflections on Marriage, in Astell:Political Writings, edited by Patricia Springborg
(NewYork:Cambridge University Press, 1996), 18. See also Astell, A Serious Proposal to the Ladies,
for the Advancement of Their True and Greatest Interest, edited by Patricia Springborg (Brookfield,
VT:Pickering and Chatto,1997).
39
Mary Wollstonecraft, A Vindication of the Rights of Women, edited by Miriam Brody
(NewYork:Penguin Books, 1992); Harriet Taylor, The Enfranchisement of Women and Marriage
and Divorce, in Essays on Sex Equality, John Stuart Mill and Harriet Taylor, edited by Alice Rossi
(Chicago:University of Chicago Press,1970).
In the mid-twentieth century, Isaiah Berlin labeled this freedom negative

liberty to communicate that freedom entails an absence of interference or con-
straints.40 But he noted that another conception of liberty also pervaded the
history of philosophy, namely, positive liberty. Positive liberty entails the pos-
itive provision of resources to help the individual take advantage of negative
libertiessuch as education for girls, which would enable them to support
themselves economically. More significantly, however, positive liberty also
involves the assessment of desire. Is my desire the mere impulse of appetite,
responding to which is slavery as Rousseau maintained? True freedom, ac-
cording to Rousseau, consists in obedience to a self-prescribed law.41 Berlin,
following this line of argument, maintained that positive liberty entails the
foundational premise that pursuing rational desires is freedom but pursuing
irrational ones is not. According to this view, we can and should make qualita-
tive assessments among various desires.
This might sound to some like autonomy, for as Kant noted autonomy
involves the following of imperatives that Ilay down for myself but that are
universal in applicationthe categorical imperative.42 But Berlin maintains
that positive liberty is something less than autonomy. If Iam in the grip of irra-
tional desires and others prevent me from acting on those desiresthereby in-
terfering with my negative freedomthose others are actually preserving my
positive freedom by preventing me from acting on a desire that is irrational,
false and therefore not what Ireally want. As innocent as this may sound
after all, when I stop you from engaging in harmful addictive behavior, am
Inot helping to liberate you?Berlin argued that it led to something more,
even something ominous:because who decides what desires are rational or
irrational? What Charles Taylor called second guessing means that others
claim to know what my true interests and preferences are better than Iknow
myself.43 Rather than autonomy, however, this coheres with the classic defi-
nition of heteronomy:doing what another wills, because they will iteven if
they will it because they want you to willit.
Shaded by the Cold War, Berlin sought to align negative liberty with Western
democracies and positive liberty with communist totalitarianism. And femi-
nists, aware of the long history of mens deciding for women what is in their true
interests, have reason to be wary of this form of freedom as well. But Berlins ac-
count of positive liberty was not accurate. What he misses is that positive libertys
40
Isaiah Berlin, Two Concepts of Liberty, in Four Essays on Liberty, Isaiah Berlin
41
Jean-Jacques Rousseau, The Social Contract and Discourses, translated by G. D. H. Cole
(London:J.M. Dent and Sons, 1973), bk. 1, chap.8.
42
Immanuel Kant, Groundwork of the Metaphysics of Morals, translated by H. J. Patton (NewYork:
Harper and Row,1964).
43
Charles Taylor, Whats Wrong with Negative Liberty? in The Idea of Freedom:Essays in Honor
of Isaiah Berlin, edited by Alan Ryan (NewYork:Oxford University Press,1979).
attention to internal barriers to liberty, the possibility of conflicting desires within

an individual self, and even the notion of second guessing that Berlin despises
all tie to an idea that most feminists today take for granted:namely, the idea that
many of our desires and preferences are socially constructed and that we often
need the help of others to see how that has happened. The desires and preferences
we have as well as the choices we make must be understood in and seen as com-
ing out of the particular society, culture, and time in which we live as well as out
of the specific locations of gender, sexuality, race, and class that we occupy within
those contexts. Womens choicesas well as their ability to make the choices and
pursue their preferencesall need to be understood within these contexts that
shape options, and even shape desires themselves. These social contexts determine
which barriers or obstacles we will face, and which enabling structures and sup-
ports we may have at our disposal. In other words, these social contexts provide
the conditions for both negative and positive freedom.
But because those contexts are importantly shaped by patriarchy, sexism,
racism, classism, and heterosexism, social construction is not simply innocent
or neutral:as feminist theorists have shown, cultural norms and social practices
can produce desires within women that arbitrarily limit their abilities to engage
in the world. Women may appear to be free in negative liberty terms, but our
own desires may restrict us. Social constructivism thus presents a paradox, for the
very social structures that limit women are the same structures that produce us
and our abilities to choose. Feminists want to be able to claim that some contexts
produce greater freedom for women than others; but social construction turns
that desire back on itself, because the framework for evaluating contexts, and the
values that inform our judgments, are themselves constructed by and through
a patriarchal context. So how can feminists judge which choices are better or
worse forwomen?
A feminist theory of freedom needs aspects of both positive and negative lib-
erty to analyze this paradox of the choosing subject. Indeed, from a feminist per-
spective, the opposition between negative and positive liberty is mistaken. For
key to social constructivism is the idea that what is inside the self (i.e., desire,
preference, will) is to a significant degree produced by what is outside the self
(i.e., social relations, including institutions, customs, practices, and laws). Indeed,
language, the essential tool for thinking about who Iam and what Iwant, oper-
ates at this very juncture of inner and outer, for it comes to me from others. It is
received:Ido not invent it but learn it as a way to communicate with others who
already have it. Yet Iprocess it through and incorporate it into my conscious and
unconscious mind to form and express my desires, will, and identity. This inter-
action of inner and outer realms, and of positive and negative liberty, extends
through the history of modern political thought.44
44
Idocument this at considerable length in the theories of Hobbes, Locke, Rousseau, Kant, and
Mill in Hirschmann, Gender, Class and Freedom.
But it is my view that, just because the inner self, the individual, is socially
constructed does not mean that it is not real.45 And fundamental to the very
concept of the individual is that she is able to exercise choice:she is able to
make choices that affect, shape, and direct her own life. This is the essential
value that the Enlightenment bequeathed to us, a value that has been vitally
important to feminists throughout history.
Rather than proving the superiority of negative liberty over positive, how-
ever, recognizing the value of individual choice leads us to the need to con-
sider the two conceptions together. Both positive and negative liberty share
the notion that choice is central to freedom:both conceptions of liberty have
at their heart the ability of the self to make choices and act on them. The con-
tested terrain between them generally covers differences about what consti-
tutes the process and activity of choosing and what constitutes the product
or an actual choice. For both models, choice is constituted by a complex
relationship between so-called internal factors of will and desirethe prefer-
ences one has, which of the available options one prefers, and how one chooses
among themand factors outside the self that may inhibit, enable, or en-
hance ones ability to pursue ones preferences, including the kind and number
of choices available, the obstacles to making the choice one prefers, and the
variable power that different people have to make choices.
The terms internal and outside, however, point out that implicit in these
debates is the more fundamental question of what or who the self is that makes
these choicesa question particularly relevant to relational autonomy. The
social constructivist thesis posits that what is internal to the self, such as de-
sire, is socially produced through culture, practice, law, custom, history, and
languageall ostensibly external to the individual. Yet those very cultures,
practices, laws, customs, history, and language are themselves produced by
individuals who participate in them every day. Social constructivism chal-
lenges the duality of inner and outer, and Ibelieve that this challenge is illus-
trated in the concept of freedom. Social construction shows that a feminist
conception of freedom is as relational as autonomyis.
Yet this understanding of freedom that feminist social constructivism
yields us allows more space for the individual and for the indefinable, unde-
tectable (to others), and perhaps even incommunicable aspects of human ex-
istence than do many versions of relational autonomy. Feminist freedom does
not require an essential core self, but it does require an identifiable self who is
Nancy J. Hirschmann, Feminist Standpoint as Postmodern Strategy, in The Feminist

45
Standpoint Theory Reader, edited by Sandra Harding (New York: Routledge, 2003); Hirschmann,
Subject of Liberty, chap.3. See also Judith Butler, Bodies that Matter:On the Discursive Limits of Sex
(NewYork:Routledge, 1994) for a critique of linguistic monism and the possibilities of the real
coming out of social constructivism.
the final arbiter of her own choices.46 That is, even if my desires are ultimately
opaque to myself, in that Icannot truly know where they came from or why
Ihave them, it is important for me to be able to identify them as mine and
there is value in my being able to assert and pursue them. In keeping with
negative liberty, a feminist theory of freedom demands that the individual self
must make her own choices. Yet that self, its status and content, is continually
in doubt and requires context and relationship with others to provide and sus-
tain its meaning. The relational aspects of freedom mean that others may see
me, or at least think they do, better than Isee myself, much like positive liberty.
And this in turn means that my freedom may in part depend on their engage-
ment with my understanding of my own desires.
Understanding the relationality of freedom requires a subtle distinction.
Specifically, granting that positive libertys second guessing must be rejected
by a feminist conception of freedom as antithetical to womens agency and
self-determination does not mean that the interrogation of desirewhich is
what often leads to second guessingshould itself be avoided. On the con-
trary, it is vitally important to freedom that critical questioning about desire,
about who we are and what we want, be engaged. The issue is where we draw
the line between asking questions and providing answers. Autonomy theorists,
like positive liberty theorists, may often provide external standards for what
answers are legitimately autonomous.47 Feminist freedom holds that others
can, and indeed should, ask me questions. However, only Ican come up with
the answers; nobody else can answer those questions for me. Indeed, freedom
does not even require that Icome up with any reasoned answers at all; Ican
say I dont really know, Ijust want to do it and still be free, even if doing
so fails the autonomy test. If you dont like or disagree with my answers
as a shelter worker might when a battered woman decides to return to her
abuseryou can only persist in asking more questions, which Imay refuse to
answer.48 Feminist freedom requires relationship, much as autonomy does, but
46
Admittedly, my assertion may point to a disciplinary fracture in the concept of autonomy, per-
taining to the methodologies and modes of inquiry that we engage in as philosophers; as an interna-
tionally known colleague in an English department said in response to a draft of this paper, in my
field, the subject who is opaque to herself is simply a given, we dont even argue about that anymore.
By contrast, philosophersas well as many political theorists, including myselfdo not wish to give
over entirely to the claim of opacity. Many of us still believe that the point of philosophy is to clarify
our understanding of ourselves. But the social construction thesis requires us to acknowledge that the
project of knowing ourselves, not to mention knowing others, is ongoing, in process, always partial,
and never complete.
47
Procedural autonomy theorists suggest that autonomy is measured not by the outcome of a de-
cision but rather by the procedures that the decision maker follows. However, these theorists, too,
must indicate the standards that the decision maker must meet. See Diana T. Meyers, Self, Society,
and Personal Choice (NewYork:Columbia University Press, 1989); Gerald Dworkin, The Theory and
Practice of Autonomy (NewYork:Cambridge University Press,1988).
48
Idiscuss this theory in relation to domestic violence at greater length in The Subject of Liberty,
chap.4.
the obligations of critical engagement lie on others, not on the self, for the self
must be free to reject such engagement at alltimes.
Such rejection does not mean that Ihave somehow placed myself outside
of the processes of social construction, of course, like the abstract individual
often associated with negative liberty. As Foucault suggests in one of his last
interviews, freedom is not simply resistance or liberation. Nor is freedom a
quality one possesses or a condition that one could attain by meeting preset
criteria. Rather, freedom is a mode of activity in which people participate by
partaking in practices that create the self. That self is both passive and active;
we are produced through social formations over which we have little control,
yet we act upon these formations in daily and ongoing processes through spe-
cific actions. Such actions involve individualsin some ways intentionally, di-
rectly, consciously, in some ways notin the creation of who they are. Central
to the practices of freedom is the exercise of self upon self by which one
tries...to transform ones self and to attain a certain mode of being.49 But
this care for the self...implies also a relationship to the other...in order to
really care for self, one must listen to the teachings of a master. One needs a
guide, a counselor, a friendsomeone who will tell you the truth.50 Given
that Foucault is most well known for his rejection of the idea of truth,51 not to
mention his attention to the ways relations of power can produce oppression,52
this claim may sound odd, but Ibelieve he means only critical observation.
The counselor is to ask hard questions that challenge the answers we have
come up with for ourselves in defining the self. The point of this questioning is
to engage the process of social construction, to entangle oneself in the messy
paradox of desires that are constituted and produced through social forma-
tions, out of which one often needs someone to help pull one out in order to
understand and express theself.
4. What Does FreedomAdd?
I suggested earlier in this paper that relational autonomy should be considered

in tandem with freedom. If feminist freedom is relational, however, requiring
others to help produce and articulate the self and its desires, how does this
differ from relational autonomy and what does itadd?
49
Ral Fornet-Betancourt, Helmut Becker, and Alfredo Gomez-Mller, The Ethic of Care for the
Self as a Practice of Freedom:An Interview with Michel Foucault on January 20, 1984, translated by
J.D. Gauthier, Philosophy and Social Criticism 12:23 (1987):124, 123,114.
50
Fornet-Betancourt etal., Ethic of Care,118.
51
See, e.g., Michel Foucault, Two Lectures, in Power/Knowledge:Selected Interviews and Other
Writings, 19721977, edited by Colin Gordon (NewYork:Pantheon Books,1980).
52
Michel Foucault, Discipline and Punish: The Birth of the Prison, translated by Alan Sheridan
(NewYork:Vintage,1979).
The first thing my account of freedom adds is the relationship of the passive
subject to the active subject, both of which Foucault insists on. Autonomy pri-
marily involves an active subject; as Raz says, autonomy is a kind of achieve-
ment.53 Relational autonomy theorists suggest that autonomy requires some
kind of conscious work, as Nedelsky argues. Friedman defines autonomy as
acting and living according to ones own choices, values, and identity within
the constraints of what one regards as morally permissible.54 Autonomy par-
ticularly involves reflective judgment, the critical evaluation of desires,
self-discovery, self-definition, and self-direction, living life from the inside.55
Meyers puts it best:Since one must exercise control over ones life to be auton-
omous, autonomy is something that a person accomplishes, not something
that happens to persons.56 All of these ideas cohere with the activeself.
But a key feature of freedom is that it involves a combination of what indi-
viduals do and what happens to them: both the active and the passive. As
Foucault notes, even if the subject constitutes himself in an active fashion,
by the practices of self, these practices are nevertheless not something that the
individual invents by himself. They are patterns that he finds in his culture and
which are proposed, suggested and imposed on him by his culture, his society
and his social group.57 As I read Foucault, because we are situated in such
patterns that preexist us, what is relevant to understanding freedom is not
simply the act of making choices within these patterns but rather the degree to
which a choosing subject has the capacity to participate in the creation of the
options from which she must choose. This is essential for feminism, because
the choices that women make every day are constitutedto varying but al-
ways important degreesby and through patriarchal social relations that de-
fine even the terms of relationships between women. The paradox of relational
autonomy is that autonomy is defined by relationships that are created and
shaped by patriarchywhich has womens nonautonomous status as one of its
foundations. According to social constructivism, women cannot just say no
to patriarchy and form new relationships because the new is constituted by
and through the old.58 Instead, women need to be able to participate in the
53
Joseph Raz, The Morality of Freedom (Oxford:Clarendon Press, 1986),204.
54
Jennifer Nedelsky, Laws Relations: A Relational Theory of Self, Autonomy, and Law
(New York: Oxford University Press, 62); Marilyn Friedman, Autonomy, Social Disruption, and
Women, in Relational Autonomy:Feminist Perspectives on Autonomy, Agency, and the Social Self, edited
by Catriona Mackenzie and Natalie Stoljar (NewYork:Oxford University Press, 2000),37.
55
Robert Young, Personal Autonomy:Beyond Positive and Negative Liberty (NewYork:St. Martins
Press, 1986), 42; John Christman, Introduction, in The Inner Citadel:Essays on Individual Autonomy,
edited by John Christman (New York: Oxford University Press, 1989), 10; Emily R. Gill, Becoming
Free:Autonomy and Diversity in the Liberal Polity (Lawrence:University Press of Kansas, 2001),29.
56
Meyers, Self, Society, and Personal Choice,96.
57
Foucault, Ethic of Care,122.
58
See, e.g., The Milan Womens Bookstore Collective, Sexual Difference:ATheory of Social-Symbolic
Practice, translated by Patricia Cicogna and Teresa de Lauretis (Bloomington:Indiana University Press,
1990). Idiscuss this work at greater length in The Subject of Liberty, chap.7.
social formations that structure, both actively and passively, the framework
within which relationships are formed and exist, options are made available,
and desire is formed. They need to participate in the processes of social con-
struction that produce their freedom and restraint.
A second contribution that freedom can make to relational autonomy
involves a shift in the dynamics of relationship. That is, in my previous ac-
count Isaid that freedom requires engagement with others and critical ques-
tioning, which might seem to be identical to the requirements of autonomy.
Autonomy theorists, including many relational autonomy theorists, stress the
importance of the subjects ability to articulate reasons for choice: Dworkin
defines autonomy as a capacity...to reflect critically upon...preferences,
desires, wishes, an idea Raz echoes.59 And many of the passages I cited to
demonstrate the role of the active subject in autonomy theory also involved
critical thinking and reflective judgment. Moreover, relational autonomy theo-
rists maintain that such judgment cannot be an isolated activity. Nedelsky, for
instance, maintains that although one can exercise freedom simply by choos-
ing, choosing autonomously requires the exercise of judgment. But judg-
ment emerges out of relations of dependency:We are dependent on others for
the social world that enables us to develop all of our core capacities, including
reason and judgment.60 We must critically engage and interrogate our own
desires within a context of relationship because relationship does not merely
make it possible to exercise those capacities:it is also part of those capacities.
By contrast, note that my formulation requires others to ask questions, not
the self. Questioning by others operates from a position outside my particular
context and therefore potentially challenges the assumptions and givenness
that experience has produced in my self. It is such challenges that open up
possibilities for my thinking and choosing differently. At the same time, how-
ever, because Iam the one who has the experience, Imust be the filter and
arbiter of such questions, challenges, and information, even to the point of
deciding to reject them. Also, Imust be the one to come up with the answers,
no matter how imperfect or imperfectly formulated they might be. Autonomy
theorists might agree with that. However, where a theory of freedom diverges
is that, as Isaid earlier, Ido not have to answer others questions at all:Ican
simply say this is what Iwant, Ido not need to engage in the kind of critical
self-reflection that autonomy usually requires. Nedelsky suggests that what
this means is that people can choose to use their freedom in a weakly auton-
omous manner,61 but Ithink that even that is too strong. The value judgment
Dworkin, Theory and Practice of Autonomy, 20; Raz, Morality of Freedom,204.

59
Jennifer Nedelsky, Reconceiving Autonomy:Sources, Thoughts and Possibilities, Yale Journal

60
of Law and Feminism 1:1 (1989):24; Nedelsky, Laws Relations, 62,28.

61
Nedelsky indeed says this explicitly in response to a similar point Imade in Subject of Liberty;
see Laws Relations,61.
that that imposes on my freedom, it seems to me, echoes the paradox of pos-
itive libertys forcing you to be free. The choices that we make are not nec-
essarily self-conscious, objectively rational, or even well thought out, but to
deny or devalue the role and power that individuals have to make choices and
that help constitute the self is equally mistaken. Freedom must lie in the self s
ability to make choices and act on them, regardless of what othersincluding
philosophersthink of those choices.
This leads to the third, and perhaps most important, contribution that
thinking of freedom in tandem with relational autonomy can make:namely,
helping us rethink the role of the self. Because a feminist theory of freedom
requires us to hang on to negative liberty and not just positive liberty, it
reminds us of the value of the liberal individual. Relational autonomy theory
is built on a rejection of the classical model of the isolated self thought to sit at
the heart of traditional autonomy theory from Kant onward. The rejection of
the idea of a true or strong core self may have been gradual among some rela-
tional autonomy theorists (who nevertheless argued for a different, relational
way of understanding that core),62 but the relational self that others promote
seeks to reject altogether the liberal, atomistic self. Nedelsky, for instance,
spends much of her excellent book Laws Relations struggling with the idea
of the self at the heart of relational autonomy. She explicitly targets liberalism
as the model that she wishes to reject.63 Although she claims to acknowledge
human difference, self-creation, and individual choice, she spends much of
the book arguing for a conception of relationality that is so all-encompassing,
and so deep, as to allow very little room for individuality.64 Indeed, she says
that well-known communitarian Alistair MacIntyre understates the nature of
our dependence and interdependence claiming that relationships are part of
autonomy, rather than simply providing the condition for autonomy.65 Indeed,
she draws on object relations theory explicitly, particularly Jessica Benjamin
and to a lesser degree Chodorow and Evelyn Fox Keller, to shore up her cri-
tique of the idea of autonomy as control that is associated with the male
model of reactive autonomy.66 Moreover, in a section titled Understanding
and Overcoming Pathology, she says the most promising model, symbol, or
62
In Self, Society, and Personal Choice, e.g., Diana Meyers said not only that one must be able to
offer reasons but also that what guides those reasons must be firm goals or moral views rather than
feelings, intuitions, and arguments of the moment, for autonomy expresses the true self a view that
she no longer shares according to her essay in the present volume. See also Diana T. Meyers, Personal
Autonomy and Feminine Socialization, Journal of Philosophy 84:11 (1987):619.
63
Nedelsky, Laws Relations, esp. chap.3.
64
In this, although Iagree with much of Marilyn Friedmans critique of Nedelsky in the present
volume, Idisagree that Nedelsky is sympathetic to some aspects of liberalism. Rather, Nedelsky seeks
to claim some concepts that liberalism first gave us, such as equality, to transform them for her own
project.
65
Nedelsky, Laws Relations, 28 (emphasis added),41.
66
metaphor for autonomy is...child rearing and identifies liberal individualism

as the pathology in question.67
Yet in a curious moment, she approvingly quotes Richard Flathman, one
of the strongest contemporary advocates of liberal individualism this side of
Robert Nozick, regarding individual creativity as coming from a space beyond
relationship.68 Isay this not to disparage Nedelskys book as incoherent but as a
testimony to the books complexity and the intractability of the subject matter.
However, this last invocation of one of my own mentors crystalized some of
my dissatisfaction with her argument, and my anxiety about relational au-
tonomy more generally, which attention to freedom can help address, namely,
that the liberal self needs to be recuperated. Decades of scholarship refining
our readings of canonical Enlightenment thinking have shown that abstract
individualism is not an accurate representation of the canonical texts.69 The
liberal self may be exaggerated, an exaggeration made possible in part by the
oppression of women in the private sphere but it is nonetheless vital to the
feminist project.70
What a feminist perspective on freedom suggests is that the liberal indi-
vidual could be treated as a metaphor for the dimensions of human life that
make social relations less important. That is, rather than taking the liberal in-
dividual as a literal ideal for us to emulate, we should take it as a theoret-
ical signpost and reminder that autonomy needs to be formulated not just in
relationship with others but also against others. Taking it as a metaphor is
made possible by the fact that freedom tolerates a less well-defined self than

67
Nedelsky, Laws Relations, 62. Since the days when Flathman advised my doctoral disserta-
68
tion, he has been in strong disagreement with feminist arguments for relationality. See particularly
Richard Flathman, The Philosophy and Politics of Freedom (Chicago:University of Chicago Press, 1984),
where he endorses a strongly individualist conception of negative liberty as the only philosophically
and politically coherent conception; Reflections of a Would-Be Anarchist (Minneapolis:University of
Minnesota Press, 1998), which, to quote the publisher, warns of the individualism-limiting potential of
even liberal efforts to promote social justice; Thomas Hobbes:Skepticism, Individuality, and Chastened
Politics (Lanham, MD:Rowman and Littlefield, 2002), where he holds up Hobbesian individualism as
a model for contemporary political theorists to follow. He thus opposes almost everything Nedelskys
book arguesfor.
69
See, e.g., Quentin Skinner, Hobbes and Republican Liberty (New York: Cambridge University
Press, 2008); John Dunn, The Political Thought of John Locke:An Historical Account of the Argument of
the Two Treatises of Government (NewYork:Cambridge University Press, 1983); Gordon Schochet,
The Authoritarian Family and Political Attitudes in Seventeenth Century England:Patriarchicalism in
Political Thought (New Brunswick, NJ:Transaction, 1988); Joanne H. Wright and Nancy J. Hirschmann,
Introduction:The Many Faces of Mr. Hobs, in Feminist Interpretations of Thomas Hobbes, edited by
Nancy J. Hirschmann and Joanne Wright (University Park:Pennsylvania State University Press,2013).
70
In Rethinking Obligation, Ilinked the liberal subject of social contract theory to the male model
of object relations theory. Istill believe my analysis of the masculinism of the natural self in obliga-
tion theory is correct, because contemporary theories deploy an abstract model that the historical theo-
ries did not. However, Ialso believe that Ishould have attended to the greater subtleties of relationship
in those canonical theories, a failure Ibelieve Iaddressed in Gender, Class and Freedom.
autonomy requires. In keeping with the idea that it is important to view the
passive self in conjunction with the active self and that freedom is a function
not just of what the subject does but what happens to the subject, the self nec-
essary to freedom theory is more fluid and porous. Yet it is decidedly present.
Granting that nothing escapes social relations, a feminist theory of freedom
reminds us that some things at the least make relations beside the point. If
everything and everyone is always already socially constructed, if there is no
way not to be socially constructed, there is no way to exist outside social re-
lations, outside social contexts, outside language. We can grant that. But then
relationality becomes so expanded as to make it almost meaningless:even re-
active autonomy is relational because it entails a reaction against the mother.
In other words, if we take the concept of relationality and social construc-
tion to its logical end, we are led to an impasse, and autonomy itself starts to
look incoherent. This is not a problem Iattribute to any of the arguments in the
present volume as much as it is a conceptual problem that exceeds the grasp
of all of us. Autonomy is fundamentally about capabilitiesas several papers
in the present volume suggestspecifically the ability to assess ones choices,
to reflect critically about them, and to make choices that allow one to exert
some control over ones life. Thus, there must be a self. But if we are always
and already socially constructed, then how can autonomy actually exist? What
purpose does the notion of self rule really serve if the self is always already
constructed and produced through complex webs of relationships, which were
in turn constructed by other social relations, all the way down? Indeed, how
can there be a self necessary to the idea of self rule? We always insist that
the self exists, yet that acknowledgment is a challenge for the logic of relation-
ality. Our alternative, to reject social construction, can return us only to the
naturalism of social contract theory that gave birth to the liberal individual.
Treating that liberal individual as a metaphor, however, can help us recog-
nize that this deep embeddedness in relationship that social construction gives
us does not forestall self-definition and acting for the self by acting against
context. Returning to where Ibegan this paper, my anxiety about relational au-
tonomy is that relationship is not only complex, as many relational autonomy
theorists note, but also is as problematic for autonomy as it is helpful. Surely,
can some relationships be harmful, as many relational autonomy feminists ac-
knowledge. However, even the good relationshipslike that between mother
and daughter in object relations theorycan harm us, by hiding the ways in
which oppressive power relations of patriarchy are replicated and reinforced
in our daily interactions.
By calling for a recuperation of the liberal self as a metaphor, Iam in a sense
suggesting that feminist autonomy theorists should reassess the importance
and value of reactive autonomy. What reactive autonomy can contribute is
the recognition that autonomy importantly must develop against others and
against prevailing norms and customs, such as Hirsi Ali discussed in Marina
Oshanas essay for this volume. Iagree with Oshana and Friedman that it may
be time for the liberal self to be recuperated, though, Iwould argue, recuper-
ated within a context of freedom, specifically relational freedom.71
5. Conclusion:Recuperating the LiberalSelf
Perhaps ironically, Ifind a model for what Iam suggesting in John Stuart Mill,
with whom Iopened my discussion of freedom. Often seen as the icon of clas-
sical liberalism and of negative libertynot to mention a problematic femi-
nist72Mill actually deployed both negative and positive liberty elements as
well as the idea of social construction in his theory. In the chapter Of Liberty
and Necessity in his System of Logic Mill agrees with determinists that a per-
sons character determines his desires and causes him to act on his prefer-
ences: given...the character and disposition of the individual, the manner
in which he will act might be unerringly inferred.73 In turn, an individuals
character is formed by his circumstances: by the experiences he has, how he
was brought up, the kind of influences and education he was exposed to, and
so forth. In true social constructivist fashion, Mill says that we cannot stand
outside of ourselves to create ourselves ab initio, because we are who we are
through the social conditions, institutions, practices, relationships, language,
and frameworks in which we cometobe.
However, this does not foreclose the power of self-creation or free action.
That a persons character is formed for him, is not inconsistent with its being,
in part, formed by him as one of the intermediate agents. That is, he has, to
a certain extent, a power to alter his character. He does this by changing his
circumstances:We, when our habits are not too inveterate, can, by similarly
willing the requisite means, make ourselves different. If [others] could place
us under the influence of certain circumstances, we, in like manner, can place
ourselves under other circumstances. We are exactly as capable of making
our own character, if we will, as others are of making it for us.74 When the
Ithink it is significant that Oshana draws on Nussbaums capability approach, which is a theory
71
of freedom, to develop her own theory of autonomy. This is a point that would require another paper,
but Ioffer the provocative suggestion that a number of the essays in this volume are writing about
freedom rather than autonomy or at least are developing a theory of autonomy that tacitly depends on
the understanding of freedom that Iam articulatinghere.
72
See Nancy J. Hirschmann, Mill, Political Economy, and Womens Work, American Political
Science Review 102:2 (May 2008):199213, on Mills ambiguity over married womens participation in
paid labor that, for all but women of property, was essential to their independence and equality.
73
John Stuart Mill, A System of Logic Ratiocinative and Inductive, Being a Connected View of
the Principles of Evidence and the Methods of Scientific Investigation (Books IVVI), in The Collected
Works of John Stuart Mill, edited by J. M. Robson, vol. 8 (Toronto:University of Toronto Press, 1974),
836837.
74
Mill, System of Logic,840.
determinist asks where the desire to change our character comes from, Mill
responds that the desire to alter our character, and thereby our circumstances,
comes from the experience of the painful consequences of the character we
previously had; or by some strong feeling of admiration or aspiration, acciden-
tally aroused.75 In other words, experience is not only backward-looking, to
consider how we got to be the way we are, but also forward-looking, to pos-
sibilities of what we couldbe.
Despite his strong social constructivism, then, his argument also invokes
a strong sense of self that produces its own construction:the self-made man
that Friedman discusses in this volume becomes literally self-made, in that he
(to continue the metaphor) engages in the construction of his self. What Mill
gives us, Ithink, is a sense of how freedom depends on a notion of the self that,
although deeply constructed and socially related, is nevertheless independent,
capable of rejecting specific relationships, and capable of forging the self in
and through new contexts. The passive subject, who is formed by his context,
can act to change that context.
When one is considering the kind of wide-ranging circumstances that
feminists confront, howeverpatriarchy, including the heterosexism, classism,
and racism that it is in part built onthis task becomes proportionately more
difficult. But that is why feminism is so vital to freedom and autonomy:if we
are socially constructed by our contexts, and if those contexts are oppressive
to women, then the only way for women to achieve freedom is by partici-
pating in the processes of social construction to change that context to pro-
duce new meanings and possibilities. Greater participation in the processes
of social construction allow greater freedom not for self-imagining per se but
for group imagining within which individuals can define and construct them-
selves. Without the discursive categories defining the larger context, the indi-
vidual has no vocabulary with which to imagine the self. But changing those
categories, and the struggle to change the context of patriarchy, is beyond the
grasp of any given individual; it is multigenerational and ongoing.
In a sense, of course, this notion of relational freedom can be taken to sug-
gest that autonomy is a fiction after all, as some postmodern theorists assert.
But Ibelieve that the picture of relational freedom Ihave put forth supports
relational autonomy by recognizing that, insofar as we are produced by rela-
tionships, we can participate in those relationships in ways that are shaped by
our feminist beliefsbeliefs that have come from relationships both positively
and supportively and negatively and reactively. That is, relational freedom can
help us see that we need not just relational autonomy; we may also need a
little reactive autonomy as well to achieve the sense of individual selfhood that
many autonomy theorists desire.
Mill, System of Logic,841.

75
Acknowledgments
Thanks to the editors of this volume, the other contributors who participated
in the Montreal conference related to this volume, and participants in the
Gender, Sexuality, and Womens Studies Faculty Works-in-Progress workshop
at the University of Pennsylvania for their comments and suggestions, partic-
ularly Sally Scholz and HeatherLove.
{ PartII }
Autonomy and Normative

Commitments
{5}
Feminist Commitments and Relational Autonomy

PaulBenson
One of the notable features of Catriona Mackenzie and Natalie Stoljars in-
fluential volume Relational Autonomy1 was the contributors clear commit-
ment to investigate personal autonomy from explicitly feminist perspectives.
Mackenzie and Stoljar announced this commitment on the opening pages of
the volume, asserting that the collections aim was to challenge feminist theo-
rists suspicion of the ideal of personal autonomy and to affirm that the notion
of autonomy is vital to feminist attempts to understand oppression, subjec-
tion, and agency.2 Mackenzie and Stoljar employed the term relational au-
tonomy as a refiguring of the concept of personal autonomy in the wake of
feminist concerns about the concepts allegedly masculinist, individualist, and
rationalist baggage.3 In particular, they saw feminists attention to the inter-
subjective and social dimensions of selfhood and identity as offering fruitful
standpoints from which refiguration of individual autonomy could proceed.4
Thus, the dual aims of the broad set of projects to construct relational con-
ceptions of autonomy were, at the start, (1)to craft theories of personal au-
tonomy that could do the work that feminist ethical and social theories need
from accounts of autonomy without succumbing to the pitfalls that seem to
lurk in many contemporary accounts of autonomy and, thereby, (2)to make
important, new contributions to the philosophical literature on autonomy,
a literature that had undergone significant resurgence in the wake of Harry
Frankfurts and Gerald Dworkins groundbreaking scholarship in the 1970s
and 1980s. This latter aim of bringing feminist investigation of autonomy into
active dialogue with so-called mainstream theorizing about autonomy was es-
pecially noteworthy and has been particularly fruitful.
1
Catriona Mackenzie and Natalie Stoljar, eds., Relational Autonomy: Feminist Perspectives on
Autonomy, Agency, and the Social Self (NewYork:Oxford University Press,2000).
2
Mackenzie and Stoljar, Relational Autonomy,3.
3
Mackenzie and Stoljar, Relational Autonomy, 34. Mackenzie and Stoljar attribute the first such
feminist articulation of relational autonomy to Jennifer Nedelsky, Reconceiving Autonomy:Sources,
Thoughts and Possibilities, Yale Journal of Law and Feminism 1 (1989):736.
4
Mackenzie and Stoljar, Relational Autonomy,4.
My purpose in this chapter is to examine more carefully the relationship

between some of the accounts of autonomy that have been proposed since
Mackenzie and Stoljars volume and the feminist normative commitments that
have shaped those accounts. One of the interesting developments in the liter-
ature on relational autonomy over the past decade has been the emergence of
what appear to be increasingly sharp and unbridgeable boundaries among dif-
ferent conceptions of autonomy and their respective evaluative commitments.
As conceptions of autonomy influenced by feminist commitment have flour-
ished, we have seen at the same time that more disputes have opened up about
the relational and normative dimensions of autonomy and that the grounds
for disagreement among competing accounts seem to be hardening.
This is hardly a surprising development in philosophy, of course. Yet many
of us would have supposed at the time of Relational Autonomys publication
that the sharing of some fundamental feminist convictions about the nature
of womens subordination and the constriction of womens agency in patriar-
chal modern societies, as well as about the character of the harms sustained
through such domination, would go a long way toward overcoming philo-
sophical disagreement about the constituents of personal autonomy. The bear-
ings afforded by feminist moral and political solidarity, it was hoped, would
make for deeper and more ready agreement on the ways previous accounts of
autonomy should be refigured. One sign of such hope was Natalie Stoljars ap-
peal to the feminist intuition regarding the poisonous relationship between
the influence of oppressive social norms of femininity and the autonomy of
women so influenced.5 Stoljars important paper implied that widespread fem-
inist critiques of norms of femininity could afford a philosophically powerful
platform on the basis of which prevalent procedural, or content-neutral, ac-
counts of autonomy could be found wanting. In other words, it appeared that
familiar and widely shared feminist convictions could, in a straightforward
manner, be put in the service of establishing a more adequate understanding
of personal autonomyand that the emerging understanding would involve
intrinsically relational constituents.
Yet progress among feminist philosophers working to develop and refine
relational accounts of autonomy has not been so straightforward. Numerous
distinct camps have emerged, and the lines of argument that separate them
appear, if anything, to be solidifying, making it increasingly less likely that
strong substantive accounts of relational autonomy, such as Stoljars or Marina
Oshanas, or largely procedural, value-neutral accounts, such as Marilyn
Friedmans or John Christmans (constitutively) nonrelational theories or
Andrea Westlunds relational (and value-neutral) conception, will eventually
5
Natalie Stoljar, Autonomy and the Feminist Intuition, in Mackenzie and Stoljar, Relational
Autonomy, 94111.
Feminist Commitments and Relational Autonomy 89
be able to present arguments that hold sway decisively among most feminist
theorists of autonomy. The ensuing discussion seeks to diagnose this polem-
ical situation by drawing critical attention to some of the appeals to feminist
normative commitments that various influential theories of autonomy have
made recently. The diagnoses proposed here will suggest that some of the pri-
mary disagreements concerning relational accounts of personal autonomy
can be understood and addressed effectively by considering how a conception
of personal autonomy can best advance feminist ethical, social, and political
commitments. Future progress in work on relational autonomy, Ishall suggest,
will depend in large measure on greater attention to what feminists really need
a conception of personal autonomy to do in relation to the normative social
purposes of confronting and overcoming womens subordination.
Our discussion will begin by analyzing the ways Marina Oshana and Natalie
Stoljar, respectively, invoke feminist normative commitments to construct their
particular accounts of autonomys constitutively relational character. Ithen ex-
amine Marilyn Friedmans and John Christmans respective appeals to certain
feminist principles of social and political inclusion to support their nonrela-
tional theories of autonomy. While Friedman and Christman both appreciate
ways in which social relations influence persons capabilities for autonomous
agency, neither accepts, as do Oshana and Stoljar, that certain kinds of inter-
personal or social relation are conceptually requisite components of personal
autonomy. Consideration of the difficulties that arise with the ways these four
theorists envisage personal autonomy to function within feminist social theory
leads me to contend that a conception of personal autonomy may best serve
feminist commitments by integrating considerations of autonomous agents
voices, which nonrelational theories characteristically highlight, with consid-
erations of the sort of agential authority that distinguishes autonomy, which
relational theories characteristically highlight. Iuse this contention to propose
that conceptions of autonomy that focus on autonomous agents authority to
speak or answer to others for their choices and actions afford a preferred way
to understand autonomys relational character. Such conceptions may take a
value-laden form (as Ihave argued elsewhere), or they may offer purely formal
conditions of autonomy (as Andrea Westlund has argued).6
I conclude by surveying some of the primary lessons that this approach
affords about the role that personal autonomy should serve within accounts
of social and political life informed by feminist commitments. As will become
clear, this paper does not aim to present a detailed account of the necessary
and sufficient conditions of autonomous action.
6
As explained in Section 5, this chapter does not seek to resolve the debate between substantive
and formal, or content-neutral, theories of autonomy.
1. Autonomy and Social-Relational Power and Authority
I begin diagnosing the relationship between relational theories of autonomy

and feminist commitments by considering Marina Oshanas important efforts
to develop a substantive, social-relational theory of personal autonomy.7
Oshanas theory offers a useful starting point, as she is more explicit than
are many autonomy theorists about the relationship of her understanding of
personal autonomy to wider normative commitments in social and political
theory. Oshanas work regularly appeals to core intuitions she has about the
kinds of power and authority that personal autonomy demands. First, auton-
omous persons are beings in actual control of their own choices, actions, and
goals....The person is in possession of the de facto power to govern herself.
Here we find the familiar idea that a self-determining person faces minimal
interference in her actions and choices....The autonomous individual is not
forced to do the will of another.8 Second, Oshana continues, autonomous
persons are in a kind of authoritative control of their own choices, actions,
and goals. To have authoritative control is to own the management of ones
choices, actions, and goals....Having the relevant kind of authority guarantees
that a persons life is free of the domination of others.9 Elsewhere she states,
autonomy guarantees that will and action exist under the de facto authority
and de jure entitlement of the agent.10 Oshana believes that the power and
authority that constitute autonomy require both procedural independence and
negative liberty. But the sort of self-governance in which autonomy consists
entails much more, she argues. Among other things, autonomy requires that
the person is substantively independent so that she enjoys minimum social
and psychological security as well as a certain level of material self-sufficiency
and is empowered to pursue whatever life plan she authentically judges to be
valuable.11 According to Oshana, an autonomous persons choices must not
merely be unobstructed...but, where realistic, these choices must be socially,
politically, and economically within his or her reach.12
See Marina A. L. Oshana, Personal Autonomy in Society (Hampshire, UK: Ashgate, 2006).
7
Oshanas important essays on autonomy include Oshana, Personal Autonomy and Society, Journal of
Social Philosophy 29 (1998):81102; How Much Should We Value Autonomy? Social Philosophy and
Policy 20 (2003):99126; Moral Accountability, Philosophical Topics 32 (2004):255274; Autonomy
and Self-Identity, in Autonomy and the Challenges to Liberalism:New Essays, edited by John Christman
and Joel Anderson (Cambridge: Cambridge University Press, 2005), 7797; Autonomy and Free
Agency, in Personal Autonomy:New Essays on Personal Autonomy and Its Role in Contemporary Moral
Philosophy, edited by James Stacey Taylor (Cambridge:Cambridge University Press, 2005), 183204.
8
9
10
Oshana, Autonomy and Free Agency,196.
11
12
Oshana, How Much Should We Value Autonomy?104.
These broad claims about the social and psychological conditions required
for personal autonomy are reflected in Oshanas judgments about various case
studies in which autonomy may appear to be under threat. Oshana surveys a
range of cases in which agents are subject to the domination, rule, or authority
of others and, in each case, contends that the agent in such circumstances has
been stripped of autonomy, regardless of how the person came to be in that
state and independently of the persons reflective appraisals of her situation
and its history.13 For example, by Oshanas lights, the willing, contented slave
and the reflectively subservient woman both lack autonomy to a substantial
extent. Notwithstanding the alignment between the agents life and her values
or her reflective endorsement of the way she arrived in her circumstances,
lives of slavery or subservience compromise self-governance by obstructing
the persons powers of self-determination and rendering the person substan-
tively dependent on others who rule major portions of their lives.14
In a similar vein, Oshana elsewhere presents the case of courageous Iranian
womens advocate and Nobel laureate Shirin Ebadi as an illustration of why,
given the subjugation of women in the postrevolution Iranian legal system,
even an accomplished jurist, influential writer, and defiant womens advocate
like Ebadi fundamentally lacks autonomy.15 Although Ebadi is a vocal and
highly visible proponent for the rights of women in Iranian society, Oshana
regards her as being deprived of personal autonomy because the Iranian state
stripped her of many of her basic political rights and unfairly deprived her of
her economic and professional livelihood as an attorney and judge. Persons
in Ebadis circumstances, on Oshanas analysis, lack the social security that
is forthcoming only when others with whom a person dwells and works and
interacts are dispossessed of arbitrary power over the person.16
Oshanas appeal to intuitions about the restriction of autonomy in cases
such as these clarifies why Oshana regards autonomy as a global feature of
persons, as opposed to an episodic, or local, characteristic of an agents atti-
tudes, values, decisions, or actions.17 Because she believes that autonomy must
guarantee persons a reasonable measure of the social, political, and economic
power and authority they need to determine the course of their lives, apprais-
als of autonomy cannot, in her judgment, be confined narrowly to the local
13
Oshana, Personal Autonomy in Society, 5268. See also her discussion of earlier versions of some
of these cases in Oshana, Personal Autonomy and Society,8693.
14
15
Oshana discussed the situation of Ebadi at a workshop on autonomy and social transformation
held on March 12, 2012, at the University of California, Davis. In a discussion elsewhere of Rosa Parkss
protest, Oshana writes, Any person who lives in an environment she cannot challenge, change, or
contribute to without the say-so of others lives in a place where impediments to self-government are
present. This person lives in the shadow of others, insecure of what her steps toward self-government
might incur from others. See Oshana, Personal Autonomy in Society,175.
16
17
For instance, see Oshana, Personal Autonomy in Society,23.
circumstances of a persons actions. This feature of Oshanas theory can make

it seem as if its departure from other, less demanding relational conceptions
of autonomy is purely a terminological matter. If, as is the case, many other
relational theories attend to the local characteristics of agency while Oshanas
conception concerns only certain global features of agency, and if those global
features are not reducible solely to local characteristics, then it may appear as
if Oshana simply is mapping out a different concept of autonomy than what
other relational theorists investigate.
This is not, however, an illuminating explanation of the disagreement about
how properly to understand the autonomy of the persons sketched in Oshanas
case studies. It is not enough just to say that, on some accounts of local au-
tonomy, the reflectively subservient woman or the committed Iranian womens
advocate both count as acting with a significant measure of autonomy in a
number of specific circumstances, while, globally, these women suffer radically
diminished autonomy because they live under the domination of other per-
sons and practices that deprive them of far-reaching self-determination. For
there seems to be a more substantive disagreement at issue, and it is not merely
a disagreement about whether a local or a global concept of autonomy is pri-
marily at stake in Oshanas case studies.18 Rather, the disagreement concerns
the role that a conception of personal autonomy should play within a norma-
tive social theory that is sufficiently robust to support resistance to womens
subordination, along with other liberatory ethical and political stances.
For Oshana, it is the role of a conception of personal autonomy (among
other things) to underwrite and explain what is wrong when women are sys-
tematically deprived of power and authority in patriarchal societies.19 Oshana
appears to be correct that purely procedural, psychologically constituted con-
ditions of authenticity or reflective endorsement are not entirely sufficient to
explain what is wrong when womens social, political, and economic freedoms
are obstructed by systems of patriarchal domination. Yet she takes a far more
controversial and, in my judgment, unwarranted step in maintaining that it
falls among the proper functions of a theory of personal autonomy to afford
a comprehensive explanation of such wrongs. This is a matter about which
theories of relational autonomy genuinely can disagree, above and beyond the
In her review of Oshanas Personal Autonomy in Society, Diana Tietjens Meyers, Review of
18
Personal Autonomy in Society, Hypatia 23 (2008): 202206, makes a similar point, contending that
such disagreement over Oshanas case studies isnt merely semantic. What is at stake is the point and
purview of autonomy theory....The underlying issue here is the relation between political desiderata
and personal autonomy(205).
19
While Oshanas theory of autonomy certainly is not motivated solely by feminist commitments,
Oshanas repeated and sustained attention to the psychological and social domination of women makes
it clear that explaining the wrongs involved in such domination is one central purpose of her concep-
tion of personal autonomy.
question of whether our primary notion of personal autonomy carries global

or episodicscope.
I have three basic concerns about Oshanas understanding of the place that
personal autonomy should take within broader social theory. First, there are
other equallyif not moreattractive and justifiable ways to analyze the
wrongs of the social and political domination of women than to say that those
wrongs must rest on assaults on womens personal autonomy. Especially in
light of the complexity and multidimensional nature of agency under circum-
stances of oppression, it seems more fitting theoretically not to place the pri-
mary burden of explaining what is wrong with such oppression on the concept
of autonomy alone. If a conception of autonomy does not need to bear the
entire weight of analyzing systems of oppression, then we would not be com-
pelled, as Oshana appears to be, to deny that persons such as Shirin Ebadi ex-
ercise a significant degree of autonomy. In any case, Oshana has not presented
arguments to establish that this must be the explanatory burden of a concep-
tion of autonomy. Her appeals to contestable intuitions about the various case
studies she presents do not, by themselves, constitute such arguments.
Moreover, it strikes me as problematic to assume straightaway that ap-
pealing to a conception of autonomy is the preferred way to ground an evalua-
tion of womens subordination when feminists themselves have done so much
to uncover the suspect masculinist and excessively individualist associations
of the notion of autonomy. It is true that Oshana, along with other relational
theorists, is proposing that those suspect associations can be supplanted by
refiguring autonomy in more socially and relationally constituted terms. But
the very fact that feminists have led us to see that healthy suspicion about the
place of autonomy within social theory is warranted indicates that Oshana
has more work to do if she is to establish that the weight of a liberatory theory
of social-relational agency must rest squarely on a conception of personal
autonomy.
Third, it is problematic to link autonomy as closely as Oshana does to more
just and ideal social circumstances of self-determination, for this has the con-
sequence of making autonomy a far rarer possession or achievement than it
actually is. The nature of systemic, institutionalized injustice and unfreedom
in human societies is such that a large proportion of humanity will lack a sig-
nificant measure of autonomy, from Oshanas perspective. Iagree with Oshana
that autonomy is terribly precious and more fragile than we may like to think.
Yet the possibility of significantly autonomous agency is not nearly as rare as
Oshana would have it. Even within the restrictions imposed upon persons
lives by unjust systems of domination, most people who lack the full social,
political, and economic powers to which Oshana points still possess the re-
sources to take ownership of many of their values, goals, choices, and actions
and so exercise authority over their agency. Once more, there is a tendency
to regard this type of disagreement about autonomys normal prevalence and
availability as merely terminological. I reject that assessment. It is a matter

of much importance to our commitment to the value of autonomy and our
prospects for achieving and sustaining autonomous agency whether we regard
it as a generally accessible, adult human attainment or whether it is a rarified
achievement rendered possible only within relatively idealized social and po-
litical circumstances.
2. Autonomy and Internalized Oppression
Like Oshana, Natalie Stoljar also appeals to an intuition about personal au-
tonomy that is rooted in specific feminist social and political commitments
to develop her conception of autonomy. As noted earlier, Stoljar explicitly
advances the idea that an adequate conception of autonomy must serve sub-
stantive feminist convictions about the bearing that gender socialization has
upon womens agency. She asserts that an acceptable theory of autonomy must
conform to and be able to explain the belief she terms the feminist intuition,
according to which women whose preferences and decisions are motivated
by oppressive and misguided norms that are internalized as a result of femi-
nine socialization cannot be autonomous.20 While most of Stoljars influential
paper, Autonomy and the Feminist Intuition, is devoted to arguing that only
a strong substantive account of autonomy can properly respect the feminist
intuition, examining her arguments against procedural, content-neutral con-
ceptions and so-called weak substantive conceptions is not my main interest
here.21 Instead Iwant to consider whether Stoljar has given us compelling rea-
sons to hold that any satisfactory conception of personal autonomy must sus-
tain the social and political commitments that the feminist intuition reflects.
Icertainly feel the force of Stoljars position, as Ionce held a view much like
hers.22 Yet Ido not think that Stoljar succeeds in showing that a suitable con-
ception of autonomy must advance certain feminist convictions in the manner
she proposes.
I have two main concerns about Stoljars understanding of the role of an
account of personal autonomy within feminist social theory. First, Stoljars
position fails to afford sufficient room for the legitimate diversity of feminist
convictions about how best to understand internalized feminine socializa-
tion across a wide range of social, political, economic, historical, and cultural
Stoljar, Autonomy and the Feminist Intuition, 95,98.

20
For extended discussion of those arguments, see Paul Benson, Feminist Intuitions and the
21
Normative Substance of Autonomy, in James Stacey Taylor, Personal Autonomy: New Essays on
Personal Autonomy and Its Role in Contemporary Moral Philosophy (Cambridge:Cambridge University
Press, 2005), 124142.
22
See especially Paul Benson, Freedom and Value, Journal of Philosophy 84 (1987): 465487;
Benson, Autonomy and Oppressive Socialization, Social Theory and Practice 17 (1991):385408.
circumstances. There can be strong reasons for some women to accept the in-
fluence of stereotypical and incorrect norms of femininity23 even as they take
genuine ownership of their decisions and actions. Such womens rejection of
feminist social ideals need not be taken by those of us with feminist sensibilities
as showing that their actions are not meaningfully their own. As Diana Meyers
observes, conceptions of autonomy, such as Stoljars, that embrace strong sub-
stantive conditions on what may count as autonomous agency homogenize
authentic selves and autonomous lives. The paradoxical effect of ahistorically,
acontextually foreordaining what individuals can and cannot autonomously
choose is to deindividualize autonomy.24 Similarly, Marilyn Friedman points
out the need for feminist accounts of personal autonomy to accommodate
another feminist intuition. Friedman observes that, even within oppressive
social conditions that ought to be resisted and overcome, traditionally subor-
dinate feminine lives nevertheless can and do often nonslavishly embody and
express values worth caring about.25
Admitting the appropriate diversity of autonomous responses to the in-
fluence of internalized lessons of subordination is especially important for
feminist conceptions of autonomy because this diversity opens up conceptual
space for autonomous criticism of and resistance to oppressive social practices
and institutions from within. In other words, such space permits feminists
to acknowledge the autonomous agency of those women who come to criti-
cize traditional feminine norms even as they initially accepted the wrongs and
injustices to which those norms subjectedthem.
A second concern about Stoljars understanding of the role that a concep-
tion of autonomy should serve within feminist social theory emerges from
the manner in which Stoljars account seeks to do justice to the lived experi-
ence and perspectives of women who have internalized oppressive norms of
femininity. On Stoljars account, the familiar feminist dictum that we should
take womens experiences seriously cannot mean that womens sense of their
own agential authority must be respected. That is, for Stoljar, womens sense of
their own authority should not necessarily be respected on the basis of their
autonomy. For the feminist intuition entails that women who regard them-
selves as competent and worthy to answer for their actions can nevertheless
suffer diminished autonomy if their attitudes and decisions are the products
of internalized oppressive socialization that they do not see to be misguided.
Hence, the social commitments that shape Stoljars conception of autonomy
would lead her to regard the significance of womens experiences as being re-
flected in the degree to which those experiences evidence their freedom from
23
Stoljar, Autonomy and the Feminist Intuition,98.
24
Diana T. Meyers, Feminism and Womens Autonomy:The Challenge of Female Genital Cutting,
Metaphilosophy 31 (2000):469491,480.
25
compulsory femininitys harmful influences. But there are more natural and
straightforward ways to take seriously womens experiences of their agency.
We can respect their attitudes toward their status as potential answerers for
their actions, even when they have been influenced by their social training to
adopt unjustified or harmful attitudes about what they should do or who they
shouldbe.26
Not only does this way of understanding the value of womens experiences
of their agency expand the sphere within which women in patriarchal socie
ties may achieve some autonomy, but it also permits us to distinguish in an
appealing way between the self-governance in which autonomous agency con-
sists and the ability to discern the true and the good, which may be termed
orthonomy.27 By contrast, Stoljars normative commitments require her to
align autonomy much more closely with the ideal of right rule.28 Autonomous
agency, for Stoljar, is a concept that aims, in part, at capturing the conditions
under which persons practically can escape the pervasive influence of false or
unjustified social norms. As we have seen, however, this view leads to unpal-
atable implications, irrespective of how valuable it would be to elude the reach
of oppressive socialization.
3. Autonomy and Social Commitments to Inclusion
One may be tempted to think that only proponents of explicitly substantive

and constitutively relational conceptions of autonomy, such as Oshana and
Stoljar, guide their theories by feminist social or political commitments in
ways that give rise to difficulties for their accounts of autonomy. While it tends
to be the case that substantive, relational conceptions of autonomy issue more
transparently from particular sorts of feminist commitment, it is by no means
the case that procedural, or content-neutral, nonrelational conceptions of au-
tonomy emerge independently of efforts to sustain specific kinds of normative
commitment. And as we have found to be true of some value-laden theories
of personal autonomy, the ways content-neutral theories are informed by fem-
inist social commitments also can suggest grounds for interrogating those
theories.
While the theories Idiscuss in this and the following sections, those con-
structed by Marilyn Friedman and John Christman, respectively, reject the view
that personal autonomy is relational by conceptual necessity, both Friedman
and Christman recognize the profound ways social circumstances, practices,
roles, and relationships shape persons autonomous agency. Friedman reminds
26
Ireturn to and expand on this point in Section5.
27
Cf. Benson, Feminist Intuitions and the Normative Substance of Autonomy,132.
28
See Stoljar, Autonomy and the Feminist Intuition, 111n51.
us, for example, that autonomy cannot emerge except out of social relation-
ships. Although autonomy is individuating in its effects on persons, it never
loses its social rootedness.29 Friedman goes on to say, however, that what
distinguishes an autonomous self from those who are not autonomous but
who are equally the products of social context is the degree of individuated
distinctness and coherence that an autonomous self achieves by acting in ways
that accord with her reaffirmation of certain deeper wants and commitments
that characterize her in particular.30 Christman notes more explicitly the dis-
tinction between theories of autonomy according to which social conditions
are contributory factors and theories in which social relations are concep-
tually necessary requirements of autonomy.31 With Friedman, Christman
accepts that social conditions that enable us to develop and maintain the
powers of authentic choice and which protect the ongoing interpersonal and
social relationships that define ourselves are all part of the background require-
ments for the development of autonomy.32 Yet he rejects views that see social
conditions as not only supportive of autonomy but [also] definitive of it....33
It is Friedmans and Christmans rejection of the latter type of position that
marks their theories as nonrelational for the sake of the present discussion.
Friedman, like Oshana, initiates her examination of personal autonomy by
invoking the notion of being self-determining in ones choices and actions. For
Friedman, to choose or act autonomously the self as a whole, as the particular
self she is, must somehow (partly) determine what she chooses and does.34
Friedman understands the particularity of an agents selfhoodthe agents
differentiated distinctness35to reside in her reflective endorsement of cer-
tain of her deep and pervasive wants and values. Determination of the persons
choices and actions in part by those self-reflectively affirmed attitudes then ren-
ders those choices and actions at least minimally autonomous. Autonomous
choices and actions are those that mirror wants or values that an acting person
has reflectively reaffirmed and that are important to her.36 Autonomy is a
matter of degree, on Friedmans account, because (among other things) reflec-
tive endorsement of ones motives may be more or less probing and extensive.
Practically any self-reflective reaffirmation will do for minimally autono-
mous action, Friedman maintains.37 She also contends that minimal autonomy
29
30
Friedman, Autonomy, Gender, Politics.
31
(Cambridge:Cambridge University Press, 2009),166.
32
33
Christma, Politics of Persons,185.
34
35
36
37
requires that one have significant alternative options38 from among which to
choose and that one have the competency necessary to act on ones reflectively
endorsed wants and values in the face of some obstacles.39
Friedman develops two especially important claims about substantive
conceptions of autonomy, which she takes to be conceptions that require
that autonomous agents choose and act to promote the value of personal au-
tonomy itself in their own lives or, at least, not to undermine or conflict with
the value of autonomy.40 First, Friedman contends that, from the perspective
of the content-neutral conception she develops, this sort of substantive au-
tonomy comprises more autonomy. That is, even if we understand autonomy
to be constituted by conditions of reflective endorsement of ones deep motives
and values, independently of the content of those attitudes, a person counts
as being more autonomous if she chooses and acts in accordance with the
value of autonomy in her life. Hence, Friedman contends that some substan-
tive conceptions of autonomy should be regarded as being continuous with
content-neutral conceptions. She writes, Substantive autonomy involves
more autonomy because with it, autonomy seeking becomes a stable and en-
during concern of the agent....She tends to choose intentionally according
to that ideal and so helps to secure its ongoing importance as a feature of her
character.41
While this is a provocative claim, Ido not find Friedmans reasoning for it
to be sound. It is trivially true that, if one cares about affirming the place of au-
tonomy among ones values and realizing its value through ones conduct, then
autonomy has a larger role within ones motives, values, and actions than it
would have otherwise. But that does not mean that acting on a central, reflec-
tively affirmed concern for autonomy is itself necessarily more autonomous
than acting on a similarly affirmed concern for some other values, including
values that may conflict directly with that of personal autonomy. The degree
of an agents autonomy, according to Friedmans own content-neutral crite-
ria, should be independent of the extent of autonomys importance within the
agents motivational and valuational systems.
More germane to this discussion, however, is a second claim that Friedman
makes about substantive conceptions of autonomy in relation to the inclu-
siveness of the moral and political commitments that substantive, value-laden
accounts of autonomy can sustain. Because such accounts advance more re-
strictive requirements for autonomy, Friedman worries that they could limit

38

39
40
Friedman, Autonomy, Gender, Politics, 1920. In my judgment, this is an unorthodox way of
defining what makes a theory of autonomy substantive. By comparison, see the account in Mackenzie
and Stoljar, Relational Autonomy,1921.
41
the practical attainability of autonomy to an advantaged minority and thus be

politically untenable in this respect.42 This concern appropriately reflects femi-
nists initial skepticism about the value of personal autonomy as a concept that
can serve feminist normative commitments. Friedmans concern also mirrors
the worries Ihave raised about the overly restrictive character of both Oshanas
and Stoljars accounts.
But Friedmans concern about the potentially exclusionary implications of
substantive accounts of autonomy runs deeper. She maintains that autono-
mous choices and actions warrant a special form of respect, beyond the fun-
damental moral respect owed to agents as persons (whether or not they act
autonomously). Because substantive conceptions of autonomy are more re-
strictive, they will entail narrower restrictions on the scope of the special
respect due to those who act autonomously. By contrast, content-neutral
conceptions will support more permissive grounds for appropriately treat-
ing others with such respect since those conceptions set a lower threshold for
autonomous action. Thus, if a conception of personal autonomy is to serve a
normative commitment to expand the respect owed specially to autonomous
agency, then a content-neutral conception, according to Friedman, should be
preferred to a substantive theory of autonomy, other things being equal. She
maintains that ...an account of autonomy with fewer requirements has, inde-
pendently of other considerations, the advantage of promoting a more inclu-
sive sense of equal worth.43
This argument is noteworthy in part because it demonstrates how a nonrela-
tional, content-neutral account of autonomy can be shaped by practical social
commitments to inclusion and to the defeat of unjust advantage, commitments
that naturally comport with feminist politics. While Friedman grants that this
argument is not grounded in a conceptual analysis of autonomy considered in
itself, she indicates that the likely social implications for the ways we identify
respect-worthy choices and actions are relevant to the evaluation of a theory
of autonomy. Reducing the requirements for autonomy may minimize the
number of persons whose autonomy goes publicly unrecognized because of
their social marginalization.44 Friedman builds on this argument later in her
book, contending that personal autonomy is a necessary condition for being
entitled to live free of domination by others, as the disenfranchisement or
subordination of those not readily seen to act autonomously is a predictable
concomitant of their failure to manifest readily in their agency a claim to the
special respect owed to autonomous agents.45
42
43
44
45
I have three kinds of concern about Friedmans claim that the promotion of
a more inclusive sense of respect-worthiness for autonomous choices is a war-
ranted social criterion for the acceptability of a conception of autonomous agency.
To begin with, the fact that there may well be potential obstacles to the social
recognition or appreciation of persons autonomy, such that the autonomy of so-
cially marginalized persons may regularly be rendered nearly invisible to others,
is not necessarily a ground for rejecting a particular conception of what consti-
tutes personal autonomy. In discussing Stoljars account, Ialready have noted that
conceptions of autonomy that would fuel widespread skepticism about when, if
ever, anyone acts autonomously are objectionable for that reason. But Friedman
is making a much bolder claim. She is proposing that conceptions of autonomy
that have the likely practical consequence that some persons autonomy may be
difficult (though not impossible) for others to recognize can be criticized on that
basis. This seems mistaken. For the difficulty of persons actual autonomy gaining
broad social visibility may reflect the regrettable social fact that, owing to differ-
ences in social power and privilege, some persons capabilities for autonomous
agency may be less transparent than others powers of autonomy. Iconcur with
Friedman that this is a very serious social barrier to be identified, resisted, and
surmounted. However, it is not, in itself, a reason to reject a particular account of
what autonomy consists in. The very fact that one conception of autonomy would
make persons autonomy more readily visible to others than would be the case
on another account does not automatically lend weight to the case for the former
conception, even if the expansion of the social visibility of and appreciation for
autonomy would be a good thing in general.
Second, if this argument of Friedmans against strong substantive theories of
autonomy were sound and if such theories could be criticized on the grounds
that they set the bar too high for the special respect-worthiness that properly
accompanies autonomous action, then it also should be possible that, on social
or political grounds, a theory of autonomy could be criticized for setting the
bar for respect-worthiness too low. That is, such a theory could be faulted for
being unreasonably or excessively inclusive. Friedmans account runs precisely
this risk, I believe. Agents who reflectively endorse the deep and pervasive
wants and values that motivate their actions and thus who satisfy Friedmans
minimal conditions for autonomy nevertheless may lack autonomy if, at the
same time, they are afflicted by psychosocial forms of manipulation that assail
their fundamental sense of their competence and worthiness to speak for their
decisions and actions. Ihave argued at length for this claim elsewhere and re-
turn to it in Section5.46
46
See, e.g., Paul Benson, Free Agency and Self-Worth, Journal of Philosophy 91 (1994): 650
668; Benson, Taking Ownership: Authority and Voice in Autonomous Agency, in Christman and
Anderson, Autonomy and the Challenges to Liberalism, 101126; Benson, Feminist Intuitions and the
Normative Substance of Autonomy.
The political liability for Friedmans position here is that, in virtue of embrac-
ing an overly inclusive understanding of autonomy, her account becomes less
able to detect politically significant ways social practices and institutions ac-
tually do undermine autonomy without disrupting persons ability to act on
their reflectively affirmed or endorsable desires and values. As a consequence,
Friedmans view risks making invisible the wrongs that such practices or insti-
tutions inflict. If, as Friedman contends, it is politically objectionable for a
theory of autonomy to foster the invisibility of certain persons autonomy, then
it should also be politically objectionable for a theory to foster the invisibility
of certain kinds of assault on personal autonomy. Friedmans account seems to
have just such a consequence.
Finally, Iam concerned that Friedmans inclusiveness argument obscures
some critically important distinctions. It is important to be able to distinguish,
first, between the conditions under which a person actually acts autonomously
and the conditions under which a person possesses the capacity to act auton-
omously. The former conditions incorporate the latter, but the contrary is not
the case. Someone can fail to act autonomously without losing the capacity to
have acted autonomously. Typically, the special kinds of respect-worthiness
that attach to autonomy do so in relation to the persons capacity for autono-
mous agency47 so that failing to act autonomously on a particular occasion
does not undermine the special respect or consideration the person deserves
as an autonomous agent. For example, a person may fail to act on her deep-
est, reflectively endorsed wants or values for reasons of temporary akrasia,
inattention, or fleeting ambivalence without thereby losing the entitlement
to live free of domination or without losing claim to others respect as an
autonomousagent.
Similarly, it is important to distinguish between having the capacity to act
autonomously and simply possessing the potential to acquire, strengthen, or
regain that capacity. As in the former distinction, a transitory undermining
of someones capacity for autonomy does not entirely strip away the grounds
to respect the persons autonomous agency if she has the potential to regain
that capacity, perhaps as readily as she lost it. In many cases, persons who
have suffered diminished capacity for autonomous decision making through
illness or disability retain the potential for autonomy. Through modest sup-
port from others, they could regain or strengthen their capacity for autonomy
sufficiently well to render their decisions worthy of respect. The impairment
of the capacity to carry out certain functions required for autonomous action
need not entail the destruction of the potential to act autonomously and the
respect-worthy status that accompaniesit.
47
Friedman seems to acknowledge this, in fact. Cf. Friedman, Autonomy, Gender, Politics,23.
These distinctions, along with others related to themincluding the dis-

tinction between having the potential for autonomy and simply being a
personillustrate the fact that the bases for various kinds of special respect,
consideration, or social recognition are many. In addition, these grounds for
respect are realized by degrees; they are hardly all-or-nothing matters. Yet
Friedmans account suggests that the conditions for actually acting autono-
mously are bound in a relatively straightforward manner to the conditions
under which agents deserve valuable forms of respect so that accounts that
narrow the range of autonomous action necessarily restrict the bases for such
respect. This implication need not follow, however, if we bear in mind the dis-
tinctions sketched here and others related to them. The bases for valuable forms
of respect-worthiness can be realized in many ways even when persons do
not act autonomously or suffer limitations in their capacities to do so. Hence,
the respect-worthiness of persons autonomy as agents can be preserved even
within conceptions that circumscribe more narrowly than Friedmans account
does the conditions under which persons actually act autonomously.
4. Autonomy and Commitments to Democratic Participation
Like Friedman, John Christman also advances a content-neutral, or proce-

dural, conception of personal autonomy that does not hold that any particular
sorts of personal or social relation belong constitutively to persons autonomy
as agents. Also like Friedman, Christman defends the content neutrality of his
conception and his rejection of constitutively relational elements of autonomy
by appealing explicitly to certain social and political values, especially to the
foundation of egalitarian democratic politics,48 as he understands it. The par-
ticular components of Christmans conception of personal autonomy are gen-
erally well known and have been discussed at length elsewhere.49 These details
of his account are not my interest here. Instead, Iwant to highlight the way he
invokes a commitment to social and political inclusiveness in order to defend
his understanding of autonomy. Ithen will note that his position succumbs to
one of the serious weaknesses that we have identified in Friedmans account.
Christmans invocation of social and political inclusion as the ground for
embracing a purely procedural and nonrelational account of autonomy is not
48
Christman, Politics of Persons, 175. Christman does not follow Friedman, however, in thinking
that a substantive conception of autonomy can be accepted at the same time, as a matter of being
autonomous to a greater degree than the minimal, procedural conditions of autonomy would entail.
49
For the purposes of the present discussion of relational autonomy, see especially Christman,
Politics of Persons; John Christman, Relational Autonomy, Liberal Individualism, and the Social
Constitution of Selves, Philosophical Studies 117 (2004):143164; Christman, Procedural Autonomy
and Liberal Legitimacy, in Taylor, Personal Autonomy, 277298.
equivalent to Friedmans appeal to inclusion, although they are related. As we

have seen, Friedman is especially concerned that we construct a conception of
the minimal conditions of autonomy so that persons autonomy has the best
chance to be socially visible, irrespective of their various social positions. She
wants the autonomous actions of the socially marginalized to have the greatest
opportunity, within reason, to be recognized socially so that the special moral
and political status that attaches to autonomous agency will not be withheld
systematically from socially disadvantaged persons. Christman would not at
all be unconcerned with the conditions under which personal autonomy is
likely to be recognized and duly esteemed by others. However, his argument
for a content-neutral, nonrelational conception of autonomy does not turn on
what those conditions might be. Rather, he appeals directly to the conditions
for persons full and equal entitlement to participate in democratic practices
and institutions as a primary reason to eschew substantive or relational ac-
counts of autonomy.
According to Christman, persons may have the status of genuine par-
ticipants in democratic social practices only if they are autonomous agents.
Substantive or constitutively relational conceptions of autonomy are unaccept-
able, he maintains, because they limit the possibility of autonomous agency in
ways that conflict directly with the value pluralism and fundamental neutrality
among competing, comprehensive conceptions of the good that liberal de-
mocracy must sustain. Christman writes, ...An overly restrictive conception
of autonomy will fail to recognize the moral status of reflective, self-defining
persons whose participation in hierarchical or other kinds of social relations
that appear to us (perhaps rightly) as oppressive should not deprive them of
the social and political status that autonomy underwrites in democratic prac-
tices and institutions.50
Christman sees both normatively substantive accounts of autonomy and
relational theories of autonomy as failing by this standard of democratic inclu-
siveness. Both sorts of accounts hold open the possibility that an agent could
fail to be autonomous for reasons having to do with the specific values she
holds or the particular kinds of relationship she sustains, even when the agents
values, choices, and actions satisfy suitable conditions of authentic and com-
petent reflective acceptance.51 And that would mean, according to Christman,
that the agent should be denied, on the grounds of her lack of autonomy, full
participation in democratic practices. Yet a democratic social and political
order, properly understood, cannot deny full participant status to persons
simply on the basis of the specific values they hold or the particular kinds of
50
Christman, Politics of Persons, 173174.
For Christmans specification of the pertinent authenticity and competence conditions, see
51
Christman, Politics of Persons, 142156, esp.155156.

relationship in which they participate. Disagreements about the justifiability

of such values or the legitimacy of such kinds of relationship are themselves
precisely matters for democratic deliberation, not disagreements to be settled
preemptively by way of autonomys constitution.52 This is why Christman is
especially intent on criticizing conceptions of autonomy that, like those devel-
oped by Oshana or Stoljar, are at once value laden and constitutively relational.
Such accounts prescriptively infuse certain substantive social or political com-
mitments into the kinds of relationship that they believe are conceptually nec-
essary for autonomous agency. The ways Oshanas and Stoljars conceptions
would narrow the scope of autonomous agency are, for Christman, tanta-
mount to silencing a voice in the moral landscape that a fully inclusive democ-
racy...should include.53
As we found with Friedmans position, however, Christmans argument
risks conflating the actual autonomy of an agents action with the agents ca-
pacity for autonomous agency and further conflates the capacity for autonomy
with the conditions requisite for a person to have the status of a full participant
in democratic society. Christman couples too tightly the conditions of actually
autonomous agency with the grounds of entitlement to democratic participa-
tion. As we noted earlier, there are many bases for respect-worthy social and
political status; these bases hardly are all-or-nothing matters that align per-
fectly with the degrees of autonomy realized in our decisions and actions or
with the degrees to which we possess the capacity for autonomous choice and
action at any particular time. Thus, even if some robust, substantive account
of autonomy, such as that reflected in Stoljars feminist intuition, were to be
warranted, it would not follow that persons influenced strongly by false and
misguided gender norms would thereby be precluded from legitimate partic-
ipation in democratic social practices. If actual, episodic autonomy need not
be bound tightly to the conditions required for democratic participation, then,
pace Christman, it is not the case that the degree of inclusiveness in a con-
ception of autonomy translates directly into the inclusiveness of a genuinely
democratic society.54
52
Christman recognizes, of course, that certain values and forms of relationship will be necessary
for the very realization of democratic procedures. He is concerned about substantive or relational ac-
counts of autonomy that would impose normative constraints on autonomy that are far more restrictive
than these basic prerequisites for democratic politics. See, e.g., Christman, Politics of Persons, 175n16.
53
54
Another worry about Christmans argument from egalitarian democratic inclusion is that his
account appears implicitly to adopt substantive and constitutively relational elements of autonomy by
way of the broad and incompletely specified competency conditions that he presumes to be necessary
for autonomy. See the discussion in Christman, Politics of Persons, of the place that interpersonal rela-
tions may have within autonomys competency conditions, 177, 182183, 184185. If this is so, then his
own account may fail to preserve the value-neutral, nonrelational character that heavows.
5. Agential Voice and Authority in Conceptions of Autonomy
Before proceeding to sketch a more constructive way to conceive the rela-

tionship between feminist normative commitments and autonomy, it would
be useful to take stock of our examination of Oshanas and Stoljars relational
theories and of Friedmans and Christmans nonrelational theories. We have
found that Oshanas and Stoljars respective ways of situating personal au-
tonomy in relation to feminist social and political ideals are unduly restrictive.
Their accounts require that the conditions of autonomous agency be tied di-
rectly to certain political idealsor set directly in opposition to certain forms
of systemic social injustice. For Oshana, personal autonomy (characterized
globally) must secure agents power, opportunity, and authority to pursue
lives free of unjust domination. Oshana can grant that, within limits, autono-
mous agents may not choose to act in accordance with fully liberatory and
well-justified conceptions of their ends. But if they fail to act in such ways,
this will be by their own choice, not due to externally imposed constraints on
their power, opportunity, or authority. Stoljar goes still further in the direc-
tion of binding autonomy to feminist ethical and political ideals, prohibiting
autonomous agents (characterized locally) from adopting values or making
choices that depend upon endorsing false, misguided, or unduly oppressive
social norms, at least not without possessing the capability to revise and cor-
rect their values and judgments in this regard.
Given the pervasively non-ideal social conditions in which we actually live,
requirements such as these, according to which personal autonomy necessarily
secures fundamental conditions of social justice in the agents circumstances
or precludes certain sorts of ill-justified or misguided normative judgment,
carry the implication that personal autonomy is far rarer than we ordinarily
would believe.55 Oshanas and Stoljars accounts also imply that some illiberal,
antifeminist convictions may well defeat a persons autonomy simply in virtue
of obstructing the allegedly requisite relationship between the persons values,
choices, and actions and the relevant social ideals. Such severe constriction of
the domain of personal autonomy stands in need of far more justification than
either Oshana or Stoljar provide. Their appeals to worthy social and political
ideals for curtailing gender-based subordination cannot, by themselves, afford
such justification.
55
Mark Piper suggests that Oshana and Stoljar might respond to this criticism by invoking the
scalar character of autonomy. This would allow that many persons could exhibit some degree of au-
tonomy in their actions even if few persons act with full autonomy by Oshanas or Stoljars lights. To
my knowledge, neither Oshana nor Stoljar gives any indication of taking such a stance. Nor do their
theories indicate how the minimal conditions for acting with any degree of autonomy at all should be
determined. Neither theory adopts Friedmans strategy in this regard.
We have found, as well, that both Friedman and Christman explicitly appeal
to ideals of social and political inclusion in ways that extend the scope of personal
autonomy too broadly. One reason for the overly permissive character of their
theories is that they presume that locally autonomous agency must be coupled
tightly to the conditions for basic entitlement to equal respect and democratic
participation; autonomy is a condition of a certain respect-worthy social or polit-
ical status, they believe. Ihave argued that this approach is mistaken because the
conditions for such basic social recognition and political participation are insuffi-
cient to ensure that persons act autonomously, or have the capacity to act autono-
mously, in whatever they choose to do. The conditions for entry into the arena of
democratic deliberation are weaker than those demanded by personal autonomy.
This diagnosis of some of the most influential lines of recent argument about
whether or not feminist normative commitments favor relational conceptions
of autonomy over nonrelational conceptions can be deepened by considering
the differences in importance these competing approaches give to the role of
autonomous agents voices in their choice making and conduct compared with
the importance they assign to persons agential authority over their values,
choices, and actions. The more inclusive of these theoretical approaches to au-
tonomy typically assign greater emphasis to the respects in which autonomous
agents can give expression, or voice, to what they care aboutor at least to
what they would find acceptable, or nonalienating, upon reflection. The more
restrictive of these approaches typically assign greater emphasis to the ways
persons are authorized to embrace values, make choices, and perform actions
autonomously, normally with an eye toward identifying some objective con-
straints on such authority. It is important for present purposes to recognize
that this distinction is a matter of emphasis and degree. All reasonable accounts
strive, at least implicitly, to acknowledge and understand the roles of both
voice and authority in autonomous agency. Iwill propose that consideration
of the ways competing theories of autonomy emphasize these different, though
closely intertwined, dimensions of agency reveals how we might constructively
improve our understanding of the role that a conception of personal autonomy
should serve in relation to feminist social and political commitments.
Notice, to begin with, that nonrelational theories such as Friedmans and
Christmans that invoke political arguments as grounds for expanding their
inclusiveness do so, in large measure, to protect the voices of individual auton-
omous agents, or the expressive capacities of their motivation and conduct.
For such theories of autonomy, persons who act on authentically formed and
sustainedor hypothetically acceptable (or nonalienating) and sustainable
values, wants, and choices are in an appropriate position to give expression, or
voice, to what they care about.56 And what such agents care about may clash
56
For stylistic simplicity I will ignore, for the most part, the particular differences between
Friedmans and Christmans respective conditions for personal autonomy. While those differences are
deeply with prevailing social norms or with the substantive constraints impli-
cated in certain interpersonal relationships or social practices in which they
participate, however liberatory or oppressive those norms, relationships, or
practices may be. For such nonrelational conceptions of autonomy, the im-
perative of protecting agents capabilities for voice means being willing to em-
brace the widest possible range of legitimate social or political disagreement
and difference. Any substantive restrictions on the evaluative commitments,
relationships, or decisions of autonomous agents could, from this perspective,
threaten persons capability to give voice through their actions to their most
important wants or deepest valuesor, at least, to what they would accept
upon reflection. Theories of autonomy should be as inclusive as they reason-
ably can be, according to this approach, because this is the best way to ensure
that they respect appropriately persons authentic voices as realized in their
decisions and conduct. That is, what is at risk in overly restrictive, insufficiently
inclusive conceptions of autonomy is the silencing, or exclusion, of authentic
voices, whether this occurs in the context of interpersonal relationships, social
practices, or the arena of politicallife.
By contrast, relational conceptions of autonomy of the sort surveyed thus
far typically embrace more restrictive conditions on autonomy as a way of
ensuring that autonomous agents are appropriately authorized to govern their
effective motives, values, choices, and actions (and the processes by which they
arose). On the approaches taken by theorists such as Oshana and Stoljar, au-
thentically formed and sustained motives, values, and actions may arise within
social conditions that diminish, circumvent, or undermine agential authority
and, in so doing, obstruct the persons autonomy. External social relations and
normative practices can be relevant to determination of a persons autonomy,
these theorists maintain, because such factors can interfere directly with the
persons ability to exercise authority over her choices and actions, however
authentically those choices and actions may express what she cares about or
wants or would reflectively accept. Thus, whereas nonrelational conceptions of
autonomy tend to highlight the dimension of self in agential self-governance
and therefore emphasize the agents capacities for voice, relational concep-
tions tend to highlight the dimension of governance in autonomous agents
self-governance and therefore emphasize the conditions of the agents author-
ity. Most importantly for the present discussion, these divergent emphases are
reflected through the ways theorists of autonomy appeal to social and political
commitments in the course of defending their preferred theories.
considerable, they do not affect the fundamental proposal Iam advancing, even though Friedmans
account lends itself somewhat more readily to the interpretation offered here than does Christmans
account.
The arguments we have presented in the foregoing sections of this paper

offer no reason to hold that either of these emphases is uniformly and without
exception to be preferred over the other. Both voice and authority are essential
facets of autonomous agency.57 We have encountered good reasons to protect
the scope of agents voices in a conception of autonomy. We also have encoun-
tered good reasons for thinking that agential authority requires conditions on
autonomy that extend beyond forms of procedural authenticity or reflective
self-acceptance alone. Both dimensions of autonomy are vulnerable and sub-
ject to threat in social and political circumstances that elicit feminist attention,
critique, and advocacy.
What we have found is that inclusive conceptions of autonomy designed
to protect voice tend to be excessively inclusive. We have seen, as well, that
restrictive conceptions of autonomy designed to secure more robust agen-
tial authority also tend to risk excess. Our task, then, is to identify a role for
autonomous agency within broadly feminist normative commitments that can
incorporate and balance properly both voice and authority. While Icannot de-
velop a full account of autonomy in the space remaining, Iwill present reasons
for maintaining, first, that a certain type of constitutively relational conception
of autonomy is favored by the desideratum to balance voice and authority and,
second, that this sort of relational conception resonates with feminist social
and political commitments in meaningful and promisingways.
It may be misleading to think of balancing the dimensions of voice and
authority in a conception of autonomy as if these were inherently independent,
competing considerations between which we should settle for some trade-off
or strike a compromise. In fact, there is a natural way to conjoin these dimen-
sionseven to understand them as being internally related to one another. If
the authority that is important for personal autonomy necessarily includes the
agents authority with respect to her voice, then an account of autonomy may be
able to protect and promote both agential voice and authority in a manner that
meshes with our considered convictions about what it is to act autonomously.
More specifically, we can think about the autonomous agents authority over
her choices and actions as consisting in her authority to speak or answer to
others for her choices and actions. As Ihave maintained elsewhere, autono-
mous agents specially own what they do in that they are properly positioned
to give voice to their reasons for actingto speak or answer for their acts, or
to give account of themshould others call for their reasons....Autonomous
agents are authorized to stand at the nodal point defined by the targeting of
57
Cf. Benson, Taking Ownership, esp. section II, 106110, for more extended discussion of the
normative, social, and discursive aspects of autonomy. That paper approaches the relationship be-
tween authority and voice in theories of autonomy rather differently, however, interpreting these con-
cepts against the background of the idea that autonomous agents take ownership of their actions in a
distinctiveway.
potential criticisms and the voicing of reasons in response.58 Or, to use Andrea
Westlunds related terminology, this authority may turn on the disposition to
hold oneself answerable to external critical perspectives59 or ones disposi-
tions for dialogical answerability.60
The autonomous agents voice, on this approach, does not consist simply
in her capability to express through her actions what she most cares about or
values or would reflectively accept. Instead, the agents voice is composed of
the persons regarding herself as being in an appropriate position to speak for
her decisions and actions in response to potential criticisms, irrespective of
whether her actions actually issue from her deepest or reflectively acceptable
motives or values. And the agents authority is not a matter of her deciding
and acting in social circumstances that are largely free of unjust subordination
or of acting on the basis of judgments that have not been influenced inappro-
priately by misguided or harmful social norms. The agents authority arises,
rather, from her taking up the position of a potential answerer for her actions,
from holding herself potentially answerable to others for what she does. Such
authority need not depend on the agents being able to see through misguided
gender norms or having the unimpeded social opportunity and power to ad-
vance her fundamentalaims.
Forging internal connections between agential voice and agential
authority in this manner intuitively may seem too weak, as it appeals to
self-regarding attitudes that could be wholly unfounded or may result from
autonomy-undermining processes. Agents regard for their own competence
and worthiness to answer for their acts can arise under conditions that plainly
disrupt their autonomy. Thus, some objective constraints on these attitudes
are necessary, including, first, autonomous agents attitudes toward their own
capabilities and worthiness to function as answerers must be formed in a suit-
ably rational way, on the basis of their evidence for such attitudes.61 Second,
autonomous agents must be capable of gaining otherwise socially available in-
formation that would be practically germane to the formation of these attitudes
and to the making of their decisions. Such capability is impaired, for instance,
in societies dominated by Orwellian propaganda. Third, autonomous agents
attitudes must not be modified by processes that circumvent their capacities
for rational consideration, as in cases of forcible mind control. Finally, the
norms in relation to which agents regard themselves as capable of presenting
their reasons must be publicly shareable, in principle.62
58
Benson, Taking Ownership, 108109. Also see Benson, Free Agency and Self-Worth.
59
Andrea C. Westlund, Rethinking Relational Autonomy, Hypatia 24 (2009):2649,28.
60
Westlund, Rethinking Relational Autonomy,35.
61
John Martin Fischer and Mark Ravizza, Moral Responsibility and Control (Cambridge:Cambridge
University Press, 1998), 213214, 235236, propose a similar condition in their account of taking respon-
sibility for acting from a particular kind of action mechanism.
62
These four conditions also are presented in Benson, Taking Ownership, 117118.
Conceiving in this manner the distinctive forms of voice and authority that
characterize personal autonomy should enable the conception of autonomy
that results to avoid the respective excesses in inclusiveness and restrictive-
ness that we identified in the theories surveyed in this chapter. Understanding
voice in this way does not restrict substantively the particular values, motives,
choices, or actions that autonomous agents can adopt, possess, make, or per-
form. Agents who take up the position of potential answerers for their actions
are not thereby constrained in what they care about, hold to be worthwhile,
or aim to do. Simply in virtue of being in a position to speak for their actions,
they are not subject in any evident way to direct, substantive constraints on
their attitudes or values. Hence, this approach should be capable of realizing to
a significant (though not unlimited) extent the inclusiveness that procedural,
nonrelational theories typicallyseek.
At the same time, this approach to conceiving of autonomous agents
authority does not carry unduly restrictive or skeptical implications for the
scope of autonomy. While agents will not attain the authority to speak or
answer for their actions simply because they meet procedural conditions of
authenticity and competence, or the like, the requirements for having such
authority do not incorporate the notably idealized social or evaluative condi-
tions that relational theorists like Oshana or Stoljar envisage. Most healthy,
socially functional adults are able to sustain the requisite dispositions for di-
alogical answerability notwithstanding routine social injustices to which they
may be subject or the limitations in their apprehension of how best to live
their lives or structure human social relations. Hence, this way of relating the
dimensions of voice and authority in our understanding of personal autonomy
seems neither to be overly restrictive nor excessively inclusive.
Thinking of the distinctive dimensions of authority and voice that comprise
autonomy as being internally related in this way, and recognizing that this ap-
proach avoids some of the most prominent deficiencies that have surfaced in
other influential conceptions of relational and nonrelational autonomy, pro-
vides good reasons for holding that personal autonomy is constitutively re-
lational. We cannot make sense of agents holding themselves to be potential
answerers for their choices and actions unless we understand persons to be po-
tential participants in relationships and social interactions in which others can
criticize their conduct. We cannot conceive of agents who have the requisite
authority and voice to act autonomously apart from a web of relationships and
normative domains within which persons are permitted to hold others to ex-
pectations and are disposed to present reasons, where appropriate, in response
to others expectations. As we have seen, however, this form of constitutive
relationality does not mean that autonomous agents are subject, inherently
and directly, to specific, substantive normative constraints on what they want,
value, or aim to do. Ishall not explore further here how best to understand
the normative dimensions of this sort of relational autonomy in the context of

distinctions between substantive and content-neutral theories of autonomy.63
What does this approach to understanding autonomys relationality tell
us about the role that autonomy should serve in relation to feminist social
and political commitments? First, this approach takes seriously womens atti-
tudes toward their own relational position as agents in a natural way and is
well suited to detect assaults on womens autonomy that consist in subverting
womens sense that they are competent and worthy of speaking for, or giving
account of, their own agency. It has long been a defining priority in feminist
ethics and politics to take seriously womens experiences, beginning with the
ways women interpret and articulate what they experience. When women in-
ternalize oppressive social norms and practices, this feminist priority can have
the unintended consequence of covering up womens psychological subordina-
tion, for the internalization of oppression usually means that women will not
interpret their experiences of hurt, limitation, or diminished efficacy as expe-
riences of undue subordination. However, if one key to autonomous agents
requisite self-regard is that they treat themselves as having the authority to
speak or answer for their reasons and actions, and if internalized oppression
commonly targets womens sense of their own competence and worthiness to
answer others potential criticisms, then this approach is well designed to iden-
tify particularly sinister forms of harm characteristically inflicted through the
processes by which oppression is internalized.
Indeed, it has been one of the strengths of other relational theories of au-
tonomy that they recognize the threats that internalized oppression poses
to autonomy. Because the approach I propose concentrates on autonomous
agents actual, subjective regard for their own answerability, however, it can
respond more directly, on a psychological level, to womens own experience of
underdeveloped or diminished authority than theories, such as Oshanas and
Stoljars, that appeal entirely to external, more idealized political conditions.
Second, the approach to relational autonomy that has emerged in this dis-
cussion characterizes in a more plausible way than other theories the social
situation in which autonomous agents frequently find themselves, located in
between stark, pervasive injustices that threaten to destroy autonomy, on one
63
For further discussion of whether or not to interpret this approach as weakly, or indirectly,
substantive, as I have argued previously, or whether to interpret the approach as yielding only
formal, content-neutral conditions on autonomy, see Paul Benson, Narrative Self-Understanding
and Relational Autonomy, Symposia on Gender, Race, and Philosophy 7 (2011), http://web.mit.edu/
sgrp/2011/no1/Benson0511.pdf; Andrea C. Westlund, Reply to Benson, Christman, Rocha, and Stoljar,
Symposia on Gender, Race, and Philosophy, http://web.mit.edu/sgrp/2011/no1/Westlund0511.pdf. Also
relevant to this issue is Diana Tietjens Meyers important contribution to this volume, The Feminist
Debate over Values in Autonomy Theory, in which Meyers argues that the usual ways in which debates
between value-laden and value-neutral conceptions of autonomy have been framed conflate two inde-
pendent axes along which theories of autonomy may incorporate normative substance.
hand, and ideal social and political practices and institutions, on the other, that
are only distantly attainable. In contrast with Oshanas theory, my approach
makes room for the autonomy of courageous crusaders, like Shirin Ebadi,
who plainly assert their authority to speak and answer for their own reasoning
and conduct, notwithstanding the grave injustices inflicted upon them. At the
same time, my approach allows that, for many other women in postrevolu-
tion Iran, crushing restrictions on womens political, professional, and eco-
nomic rights may have the additional effects of undermining their disposition
to hold themselves answerable for what they do, of engendering passivity, and
of diminishing their sense of authority to take ownership for their choices.
This contrasts with procedural theories that would ask only that autonomous
agents act in ways they would reflectively accept, even when their own sense
of agential authority has been retarded or destroyed. Any politically useful
approach to autonomy must be able to distinguish between the autonomous
agency of the defiant resister of oppression (even when her resistance fails to
achieve many of its goals) and the damage done to other womens autonomy by
that very oppression. It is an appropriate social and political priority that our
understanding of womens agency carefully track these differences among in-
dividual womens social and psychological locations within the same systems
of domination.
Third, the approach proposed here is capable of distinguishing clearly be-
tween persons actual autonomy and their capabilities for autonomous ac-
tion, on one hand, and their status as full moral and political agents, on the
other. For example, women whose autonomy is compromised by profound
self-doubt or the pervasive sense that they are socially invisible, and hence
are not in a position to speak for their own decisions and actions, are not
thereby to be exempted from moral accountability or barred from political
participation. It is one thing to be capable of claiming the authority and voice
that distinguish autonomous agents; it is quite another to possess the stand-
ing of a person with equal moral worth and full rights to political participa-
tion, whether or not that standing actually is honored. The latter conditions
are necessary for the former, but the converse is not the case. Thus, the view of
relational autonomy that emerges in this paper avoids one of the primary dif-
ficulties with Friedmans and Christmans nonrelational theories.
6.Conclusion
This chapter opened with the observation that the increasing influence that
feminist commitments have had in constructing new theories of personal au-
tonomy has been accompanied by the emergence of increasingly rigid bound-
aries among feminist accounts of autonomy. Our exploration of the role that
a conception of autonomy should have in relation to feminist normative
commitments has clarified significant strengths and weaknesses in competing

accounts of autonomy that appeal explicitly to feminist perspectives on social
and political life. Moreover, our discussion has yielded an approach toward
conceiving personal autonomy that points the way toward overcoming some
of the fundamental divisions among feminist theorists of autonomy. The fa-
vored approach, Ihave argued, incorporates relational elements as part of the
very constitution of personal autonomy. This paper validates, in this respect,
the guiding impulse of Mackenzie and Stoljars volume.
At the same time, the approach favored here conceives of autonomys
relationality in a manner that respects the fundamental notion that autono-
mous action consists in agential self-governance and thus must be accessible
to agents whose wants, values, decisions, and actions vary widely from one
person to another. The variability among manners of living that autonomous
agents may accept and pursue is sufficiently broad that autonomous agents
may reject feminist commitments to resist and overcome womens subordi-
nation even as our guiding conception of autonomy is designed to respond to
feminist ends. In this regard, feminist refiguration of personal autonomy as
relational at once yields valuable theoretical advances in our understanding of
self-governing, self-determining action and also confirms some of the limits
of autonomous agency in helping us to develop values, decide on courses of
action, construct social practices, and achieve lives that are free of hurtful and
unjust domination.
Acknowledgments
I am grateful to Mark Piper and Andrea Veltman for their many insightful
comments on an earlier version of this chapter.
{6}
The Feminist Debate over Values in AutonomyTheory

Diana TietjensMeyers
The starting point of this paper is a pair of clashing feminist intuitions about
the relationship between values and autonomy. For Natalie Stoljar, desires
influenced by oppressive norms of femininity cannot be autonomous.1 For
Marilyn Friedman, however oppressive their conditions might be and how-
ever much change is morally required, traditionally subordinate feminine
lives nevertheless can and do often nonslavishly embody and express values
worth caring about.2 Prioritizing resistance to oppression as an objective for
an adequate feminist theory of autonomy, Stoljar calls for a strongly substan-
tivethat is, value-saturatedaccount. Such an account must deem actions
conforming to false conceptions of womanhood nonautonomous regardless of
whether the agent has reflected on and endorsed them. In contrast, Friedman
prioritizes recognizing and respecting the values and agency of women in
patriarchal societies, and her value-neutral view of autonomy allows for the
possibility that womens compliance with oppressive feminine norms may be
autonomous. Like Stoljar, Ithink that a feminist account of autonomy must
render social critique and oppositional agency intelligible and achievable.
However, like Friedman, Idoubt that a womans nonautonomy can be inferred
from her compliance with subordinating femininenorms.
Paul Benson and Andrea Westlund have recently taken on the challenge of
reconciling the competing intuitions Stoljar and Friedman defend. Eschewing
his earlier value-saturated theory of autonomy, Bensons bid to end the Stoljar
Friedman stalemate proposes what he calls a weak substantive account of
autonomy. Iargue, however, that Bensons proposal does not offer a compro-
mise between value-saturated and value-neutral accounts.3 Instead it adum-
brates a previously unnoticed way theories of autonomy incorporate values.
1
Natalie Stoljar, Autonomy and the Feminist Intuition, in Relational Autonomy, edited by
Catriona Mackenzie and Natalie Stoljar (NewYork:Oxford University Press, 2000),95.
2
3
In email correspondence, Benson tells me that he didnt intend to identify a midpoint between
value-neutral and value-saturated theories of autonomy. However, I believe that use of the term
substantive to cover all autonomy theories that incorporate values regardless of how the values are
Feminist Debate over Values in AutonomyTheory 115
Westlund advocates an alternative resolution of the two feminist intuitions.

Staunchly defending a value-neutral view of autonomy, Westlund argues that
her self-answerability theory shows that some forms of subordination are in-
compatible with autonomy but also that its a mistake to assume that no tra-
ditional feminine life can be autonomous. Although I agree that Westlund
avoids the trap of value saturation, Iargue that there is a sense in which her
position is not valuefree.
To develop my discussion of Benson and Westlund, Idistinguish two ways
values may be implicated in autonomy theories:(1)An autonomy theory may
prescribe or proscribe certain types of behavior, or (2)an autonomy theorys
account of the process of autonomous choice may invoke a set of background
values. Iuse this distinction to demarcate two conceptual axes. The Directivity
Axis reveals whether an autonomy theory preempts or honors the judgment
of individual agents, whereas the Constitutivity Axis lays bare the normative
gears that drive competing accounts of autonomous choice and action. Icall
the claim that these axes represent philosophically distinct and significant
ways autonomy theories deploy values the Double Axis Thesis.
Theories in the value-laden and value-saturated sectors of the Directivity
Axis, such as Natalie Stoljars and Susan Babbitts, require that autonomous
individuals act on emancipatory values or refrain from acting on oppres-
sive disvalues. In contrast, theories on the Constitutivity Axis are value uti-
lizing, for they invoke values other than autonomy to explicate autonomous
choice and action. Whereas value-prescriptive theories enunciate a set of
action-guiding directives, value-utilizing theories describe modi vivendi for
making choices. My own view of autonomy is an example of a value-utilizing
account, as is Westlunds, for both adduce subsidiary values to characterize
autonomy. Nevertheless, our positions stand at the value-neutral pole of the
Directivity Axis, for they do not mandate that autonomous individuals engage
in any particular type of behavior other than an autonomy-constituting choice
process.
Having established that it is possible for a theory of autonomy to be val-
ue-neutral and value-utilizing, Iargue that value-neutral positions better serve
feminist purposes. As well, Iillustrate the philosophical fruitfulness of expos-
ing the values that a theory of autonomy utilizes. This interpretative exercise
not only provides a new angle from which to assess the merits of theories, it
also helps to clarify how particular autonomy theories connect with different
philosophical questions about autonomy.
incorporated leads readers to assume that Bensons weakly substantive account is on a continuum be-
tween value-neutral (nonsubstantive) theories and strongly substantive theories. As well, Ibelieve that
this use of the term substantive masks the distinction between the two axes that Ipresenthere.
1. Two ValueAxes
In this section, Idistinguish two axes that represent different ways autonomy
theories incorporate and exclude values. Along one axis, autonomy theories
vary with respect to their degrees of value prescriptivity. Along the other axis,
autonomy theories vary with respect to what Ill call their constitutive value
sets. In my view, Bensons conception of a weakly substantive autonomy theory
appears to collapse the Constitutivity Axis into the Directivity Axis with con-
fusing results.
The Directivity Axis is scalar, and like a scale for measuring weights its
values start at zero and range upward. On the Directivity Axis, theories of
autonomy range from value-neutral theories to value-saturated theories with
more or less value-laden theories in between. Value-neutral theories place no
limits on the values or disvalues that people may autonomously elect to live
by. Instead, they assess autonomy on the basis of the motivational structure
of the agentic subject and/or on the basis of the procedure behind an individ-
uals choices and actions.4 Value-saturated theories demand that autonomous
individuals repudiate particular disvalues or fulfill particular values. Susan
Babbitts value-saturated position maintains that autonomous individuals
must conduct themselves in a manner that satisfies their objective interest in
fully flourishing.5 Falling short of standards of self-respect, dignity, and libera-
tion defeats autonomy, and satisfying these standards forbids compliance with
oppressive norms and capitulation to oppressive circumstances.6 Thus, women
cannot autonomously defer to their husbands, and members of marginalized
social groups who have the talent to be medical researchers cannot autono-
mously opt to become pharmacists instead.7
It is clear that Natalie Stoljar rejects value neutrality because it fails to sup-
port her feminist intuition. However, it is less clear just how prescriptive she
thinks an adequate feminist theory of autonomy needs to be, for she offers
two versionsone considerably less potent than the otherof her feminist
4
Harry Frankfurt, The Importance of What We Care About (Princeton, NJ:Princeton University
Press, 1988); Harry Frankfurt, Reply to Lear, in The Contours of Agency, edited by Sarah Buss and
Lee Overton (Cambridge, MA: MIT Press, 2002); Friedman, Autonomy, Gender, Politics; John
Christman, The Politics of Persons (Cambridge, UK: Cambridge University Press, 2009); Andrea
Westlund, Selflessness and Responsibility for Self, Philosophical Review 112:4 (2003): 483523;
Westlund, Rethinking Relational Autonomy, Hypatia 24:4 (2009): 2649; Diana Tietjens Meyers,
Self, Society, and Personal Choice (NewYork:Columbia University Press, 1989); Meyers, Gender in the
Mirror: Cultural Imagery and Womens Agency (New York: Oxford University Press, 2002); Meyers,
Being Yourself:Essays on Identity, Action, and Social Life (Lanham, MD:Rowman and Littlefield,2004).
5
Susan Babbitt, Feminism and Objective Interests: The Role of Transformation Experiences
in Rational Deliberation, in Feminist Epistemologies, edited by Linda Alcoff and Elizabeth Potter
(NewYork:Routledge, 1993), 246248.
6
Babbitt, Feminism and Objective Interests,261.
7
Babbitt, Feminism and Objective Interests, 249, 251252.
intuition. Again, Stoljars feminist intuition is that preferences influenced by

oppressive norms of femininity cannot be autonomous.8 As Imentioned ear-
lier, this formulation affirms a value-saturated position that rules out autono-
mous conformity to oppressive feminine norms. Yet Stoljar seems to dilute
her intuition a bit when she claims that conduct overly influenced [by] stere-
otypical and incorrect norms of femininity cannot be autonomous.9 It is in-
structive to explore what she might mean by this qualified claim, for it seems
to place her position somewhere in the value-laden sector of the Directivity
Axis. Occupying the zone between value saturation and value neutrality, val-
ue-laden accounts of autonomy must hold (1)that a person cannot autono-
mously comply with oppressive norms too much or too often, (2)that a person
cannot autonomously comply with the most egregiously oppressive norms but
can autonomously comply with norms that are oppressive in relatively minor
ways, or (3)both of the preceding.10
It is possible to interpret Stoljars claim that a woman loses autonomy only
when oppressive feminine norms carry excessive weight in her life from an
objective or a subjective standpoint. An objective reading might hold that ex-
tensive gender conformity or compliance with the most damaging feminine
norms would entail nonautonomy, but spotty gender conformity or trivial
gender conformity would not compromise autonomy. Awoman resembling
the Stepford fembots might be deemed overly influenced by oppressive fem-
inine norms in virtue of her complete and incessant compliance with those
norms.11 Women who consent to be reinfibulated after giving birth might
be deemed overly influenced by oppressive feminine norms in virtue of the
grievous harms to their sexuality and the high risk of harm to their health that
this procedure involves. Whereas the fembot cousin lacks autonomy because
she conforms to gender norms too much, the woman who goes along with
female genital mutilation lacks autonomy because of the egregiousness of the
feminine norm she adheresto.
A subjective reading of Stoljars claim that only conduct overly influenced
by oppressive feminine norms is nonautonomous would put the issue of
motivation in play. Conduct wholly or largely motivated by acquiescence to
8
9
Stoljar, Autonomy and the Feminist Intuition, 98. Emphasisadded.
10
My thinking about value-laden theories has been influenced by Jason Stiglianos work in my
Spring 2012 Philosophy of Action seminar. It is important to note that in feminist philosophy of ac-
tion substantive theories of autonomy, which Iparse as value-laden and value-saturated theory, were
introduced specifically to address the problem of internalized oppression. In this paper, Iconfine my
discussion to feminist autonomy theories. Although Stoljar centers her thinking on the way oppressive
gender norms undermine womens autonomy, Iformulate my account of value-laden autonomy theo-
ries to include compliance with the norms imposed by any type of oppression, for feminist philosophy
is centrally concerned with intersections of multiple systems of oppression and privilege.
11
I take this analogy from the 1975 body-snatcher film The Stepford Wives, directed by Bryan
Forbes.
oppressive feminine norms cannot be autonomous, whereas conduct moti-

vated in small part and tangentially shaped by oppressive feminine norms
may be autonomous. Awoman who complies with oppressive feminine norms
solely because she desires to fit into a feminine mold cannot be autonomous.
However, a woman whose conduct is congruent with oppressive feminine
norms but who chooses to act this way mainly for distinctive personal rea-
sons may be autonomous. In other words, a preponderance of individualized
motivations in a womans volitional amalgam would count as insufficiently
influenced by oppressive feminine norms to run afoul of Stoljars feminist
intuition. Bearing in mind, though, that the motivational springs of human
action are typically multiplex and polyvalent and that distinctive individual
concerns normally blend with nods to social convention in human motiva-
tional systems, taking motivation into account implies that acting in a way
that coincides with oppressive gender norms would raise questions about the
agents autonomy but would not settle them. To settle the question, the fealty
to oppressive gender norms built into the agents motivational system would
need to be assessed.
I have generated these interpretations of Stoljars qualified feminist intuition
in order to indicate what sorts of autonomy theories belong in the stretch of the
Directivity Axis between value neutrality and value saturation.12 Value-laden
theories, as Icall them, do not categorically proscribe all conduct that accords
with oppressive norms, nor do they categorically prescribe types of action that
betoken repudiation of oppressive norms. Rather, they may curb the extent to
which a person can autonomously go along with oppressive norms or the se-
verity of the harm a person can autonomously incur as a result of going along
with an oppressive norm, or they may regulate how motivated a person can be
to submit to oppressive norms without sacrificing autonomy. Still, they grant
autonomy to individuals who occasionally comply with oppressive norms,
who dont comply with the most egregiously oppressive norms, or whose con-
formity to oppressive norms is motivated by considerations other than a desire
to model themselves on an oppressive stereotype. In short, value-laden theo-
ries are less prescriptive than value-saturated theories and more prescriptive
than value-neutral theories.
If this account of the Directivity Axis is right, it calls into question the pos-
sibility that Paul Bensons agential ownership account of autonomy lies some-
where between the value-saturated pole and the value-neutral pole of this axis.
On the contrary, his agential ownership view of autonomy tentatively sides
12
I wont attempt to determine whether the standards I have sketched are faithful to Stoljars
overall text or whether these standards are more philosophically satisfying than the strong normative
competence theory of autonomy that she ultimately endorses at Stoljar, Autonomy and the Feminist
Intuition, 107108.
with proponents of value neutrality, for it does not affirm that the extent or
the gravity of womens compliance with oppressive gender norms is decisive
with respect to their autonomy, and it asserts that women can autonomously
comply with such norms.13 If so, Bensons view is not a value-laden position of
the sort Ihave sketched. Nevertheless, he labels his position a weak substan-
tive one, which gives the impression that his view is located on the continuum
between value-saturated positions that directly impose normative constraints
on autonomous choice and latitudinarian value-neutral positions.14 However,
I believe that his explanation of how it is possible for a woman to autono-
mously comply with an oppressive social norm alerts us to a second axis of
analysis for autonomy theories.
At stake in the debate between proponents of value-saturated accounts and
proponents of value-neutral accounts is whether it is possible to autonomously
conform to oppressive norms. I have suggested that the middle area of the
Directivity Axis is occupied by positions that place limits on an autonomous
agents absorption into an oppressive regime but that allow that oppressive
norms may have some influence in an autonomous agents life. According to
value-laden theories, autonomy is inversely related to the sway oppressive
norms exert over a persons conduct.
In contrast, Benson holds that the autonomy of gender conformists derives
from their favorable attitudes toward their own competence and worth to-
gether with their retention of a sense of their own authority as reasoning, po-
tentially answerable agents.15 Those who have these positive reflexive attitudes
and take agential ownership are autonomous; those who lack these attitudes
and deflect agential ownership are not.16 If so, autonomy is not devoid of nor-
mative content, in Bensons view, yet normative considerations do not enter
into his theory in the same way that they enter into value-saturated and val-
ue-laden theories. The latter restrict the types of action that an autonomous
agent can choose and/or the constellations of motivations that can give rise to
autonomous actions. Benson introduces values by another route. For Benson,
the psychological values of self-confidence and self-worth are constitutive of
autonomous choice and action. Self-confidence is needed to assure you of
your ability to explain your actions by citing your reasons, and self-worth is
needed to assure you of your entitlement to answer for yourself.17 Together,
13
Paul Benson, Feminist Intuitions and the Normative Substance of Autonomy, in Personal
Autonomy, edited by James Stacey Taylor (Cambridge, UK: Cambridge University Press, 2005),
128130,136.
14
Benson, Feminist Intuitions, 136,125.
15
Benson, Feminist Intuitions, 136137.
16
Benson, Feminist Intuitions,128.
17
Benson, Feminist Intuitions, 128. Paul Benson, Taking Ownership: Authority and Voice in
Autonomous Agency, in Autonomy and the Challenges of Liberalism, edited by John Christman and
Joel Anderson (NewYork:Oxford University Press, 2005), 109,116.
these capacities endow individuals with the authority to take ownership of

their actions, which Benson identifies with autonomy.
In light of the difference between the role that values play in Bensons theory
and the role they play in value-prescriptive theories, Ido not regard his view as
a value-laden theory a midpoint between value-saturated and value-neutral
theories. Instead, I propose classifying his view as a value-utilizing theory.
Whereas value-laden theories are sited on the Directivity Axis, value-utilizing
theories are sited on a separate axis. On the Constitutivity Axis, autonomy
theories do not vary by degrees as gauged by a single measure, such as how
many ostensibly oppressive options theories legislate against. Rather, the
Constitutivity Axis is composed of an array of qualitatively distinct parcels
of values, each of which is implicated in a conception of how to choose and
act autonomously. In other words, constitutive value sets are collocations of
values that different philosophical accounts of autonomy explicitly or implic-
itly invoke to explicate the reflective procedure or motivational structure that
renders choices and actions autonomous. It is difficult, perhaps impossible,
to conceive an account of autonomy that is not value utilizing. Even rational
choice theorya paradigm of value neutralityrests on a constitutive set of
epistemic values including consistency, transitivity, and knowledge of perti-
nent facts. If so, Iwould venture that any theory of autonomy can be analyzed
both in terms of its place on the Directivity Axis and in terms of the constitu-
tive value set that places it on the ConstitutivityAxis.
I have already explored the question of where Stoljars view of autonomy
belongs on the Directivity Axis. Ican now add that her view can also be posi-
tioned on the Constitutivity Axis, for it invokes a constitutive value set. To
accommodate her feminist intuition, Stoljar backs a strong normative compe-
tence theory of autonomy. On this view, autonomous individuals must be able
to criticize courses of action competently by relevant normative standards.18
Two epistemic valuestrue beliefs about correct norms and reflexive critical
acumencomprise this autonomy theorys constitutive value set. Women who
have internalized oppressive feminine norms so thoroughly that their reason
detection abilities are riddled with perniciously patterned blind spots are not
autonomous because their cognitive capacities lack these epistemic goods.
Correlatively, it is arguable that the autonomy of women who have not inte-
grated the most harmful feminine norms into their cognitive capacities or who
have also internalized countervailing norms that provide cognitive ballast may
achieve some degree of autonomy although they sometimes act in line with
oppressivenorms.
18
Natalie Stoljar, Autonomy and the Feminist Intuition, 107, quoting Benson.
Although Benson classifies my view of autonomy as a weakly substantive

view, I would resist his suggestion to reposition my view on the Directivity
Axis.19 According to my agentic competency account, autonomous people ex-
ercise a repertoire of agentic skills in choosing and acting:
1. Introspection skills that sensitize individuals to their own feelings
and desires, that enable them to interpret their subjective experience,
and that help them judge how accurate their self-understandingis
2. Communication skills that enable individuals to get the benefit of
others perceptions, background knowledge, insights, advice, and
support
3. Memory skills that enable individuals to recall relevant
experiencesfrom their own lives and also those that acquaintances
have recounted or that they have encountered in literature or other
artforms
4. Imagination skills that enable individuals to envisage feasible
optionsto audition a range of self-conceptions they might aspire to
and to preview a variety of courses of action they mightfollow
5. Analytical skills and reasoning skills that enable individuals to assess
the relative merits of different conceptions of what they could be like
and directions they couldpursue
6. Self-nurturing skills that enable individuals to secure their physical
and psychological equilibrium despite missteps and setbacks
that enable them to appreciate the overall worthiness of their
self-understandings and pursuits and to assure themselves of their
capacity to carry on when they find themselves wanting or their life
directions misguided
7. Volitional skills that enable individuals to resist pressure to capitulate
to convention and enable them to maintain their commitment to
their values andgoals
Embedded in these skills, Idiscern a panoply of epistemic and psychological
values: perspicacity, resourcefulness, creativity, rationality, self-esteem, sta-
bility, resilience, tenacity, and corrigibility, to name a few. Igrant, then, that
my view is value-utilizing, but I do not agree that it is value-laden. Indeed,
Icontend that the two value axes Ihave laid out enable me to reaffirm the value
neutrality of my view as well as my conviction that value-neutral accounts of
autonomy are preferable from a feminist standpoint.
19
Benson, Feminist Intuitions, 135. Here Irepudiate a concession to Bensons argument Imade
in Diana Tietjens Meyers, Review of Personal Autonomy in Society by Marina Oshana, Hypatia 23:2
(2008):206.
2. Are the Axes Really Distinct and Philosophically Significant?
Judging by Bensons published work, its unclear whether he would contest

my claim that a theory of autonomy that dictates or forbids certain types of
behavior as mandatory outcomes of autonomous choice differs in kind from
one that requires that autonomous agents exhibit certain values in virtue of
the process that is constitutive of autonomous choosing. On the one hand,
he might stand by his claim that both his and my views are weakly substan-
tive theories that take a moderate stance compared with value-saturated theo-
ries and value-neutral theories. Since value-saturated, value-laden, and what
Ihave termed value-utilizing theories all incorporate norms, he might con-
tend that my distinction between the Directivity Axis and the Constitutivity
Axis is nugatory. On the other hand, his defense of his own agential own-
ership theory seems to belie this criticism of my distinction. To answer the
charge that his theory is too voluntaristic, Benson rejects the idea that you
must self-consciously adopt the reflexive attitudes he considers constitutive
of autonomy confidence in your ability to answer for yourself and a sense
of your worthiness as an agent. In his view, self-authorization can be en-
tirely attitudinal, implicit, and un-self-conscious in most contexts.20 If so, the
Constitutivity Axis diverges from the Directivity Axis in not requiring that
agents embrace justifiable norms or shun unjustifiable norms. Ill argue that
the axes are separate and that distinguishing them makes conceptual room
for value-neutral autonomy theories without ruling out value-laden and
value-saturated theories.
The pictures of developing and practicing autonomy that Benson and Ipre-
sent help to explain the difference between the prescriptions and proscrip-
tions of value-saturated and value-laden autonomy theories, on the one hand,
and the constitutive value sets that undergird autonomy theories, on the other.
Benson maintains that children consolidate the appropriate reflexive attitudes
and learn to self-authorize their conduct as their parents guide them through
childhood.21 Likewise, I maintain that good childrearing includes modeling
autonomy competency, offering children opportunities to exercise agen-
tic skills, and prompting them to use these skills frequently.22 Provided that
theyve had an autonomy-friendly upbringing, children osmotically acquire
the capacities necessary for autonomy and become disposed to avail them-
selves of those capacities. Having picked up an autonomy-sustaining modus
vivendi that cannot be theorized without invoking a set of values, their pro-
cesses of choosing and acting manifest those values whether or not they ex-
plicitly subscribe to them. Just how they choose and act autonomously may be
20
Benson, Taking Ownership, 114115.
21
Benson, Taking Ownership,115.
22
Meyers, Self, Society, and Personal Choice, 195197.
a mystery to them, and the constitutive value set presupposed by their autono-
mous choices may remain in the background without deleterious effects on
their autonomy.23 They have no more need to theorize their autonomy and
commit to exercising the skills that sustain it than they have to comprehend
and commit to the physiological processes that enable them to walk around
without bumping into things or fallingdown.
Cheshire Calhouns distinction between commitments and deep psy-
chological attractions is helpful here.24 Although some philosophers re-
gard motivations that are deep-seated and impervious to change as passive
commitments, Calhoun maintains that all genuine commitments are active
in the sense that they are made, not merely discovered as facts about ones
psychology.25 Calhoun allows that decisions can be made with more or less
awareness, but she insists that commitments are authored both because they
arise from decisions for which the agent has reasons and because they require
us to make an effort to revive them should they flag and to resist temptations
to abrogate them.26 My claim is that an autonomy theory with a constitutive
value set can ascribe autonomy to people who have unwittingly absorbed the
relevant cognitive, affective, and volitional dispositions regardless of whether
those individuals have decided to deploy those dispositions when they make
choices and regardless of whether they have decided to uphold the theorys
constitutive value set. Because value-utilizing theories do not require indi-
viduals to incorporate the desiderata included in their respective constitutive
value sets into their conceptions of the good life, letalone pursue them as in-
strumental or intrinsic ends, these theories do not require that autonomous
individuals be committed to these values.
Now, it goes without saying that people who arent acquainted with the vo-
cabulary of philosophy of action dont value autonomy under that description.
Nevertheless, people commonly do value a solid sense of who they are and
what really matters to them, and they value leading a life that is truly their
own. Thus, it seems plausible to suppose that many people who have gradually
assimilated autonomy capabilities during childhood and who now regularly
avail themselves of those capabilities do come to value what philosophers call
autonomous choice and action. Moreover, some might form commitments to
23
Westlund echoes this point:To be self-governing, we dont need to have a clear understanding of
what makes us self-governing....We just need to function that way in fact. See Westlund, Selflessness
and Responsibility for Self,3.
24
Cheshire Calhoun, What Good Is Commitment? Ethics 119 (2009):616617.
25
Calhoun, What Good Is Commitment?617.
26
Calhoun, What Good Is Commitment? 617618. There is of course a familiar logical thesis
that you are committed to the presuppositions of your beliefs and to the implications of your belief set.
However, transposing this conception of commitment into the realm of practical philosophy would do
violence to the subtlety of the distinctions we use to make sense of individuals normative identities
and their conduct. Acommitment is not the same as a tentative plan; caring about X is not the same as
being committed to X, and soforth.
exercise autonomy capabilities as well as commitments to live by the values

that underwrite those capabilities. However, they need not do so to be autono-
mous. In Calhouns parlance, their dispositions to exercise autonomy capabili-
ties and thus to go about making choices in ways that coincide with the values
implicit in those capabilities can remain deep psychologicalfacts.
Discrete sets of constitutive values underwrite most, probably all, cred-
ible theories of autonomy. Consequently, individuals whom a value-utilizing
theory judges to be autonomous must manifest the theorys constitutive value
set in the way they conduct their livestheir choice processes as well as their
ability to carry out their decisions and express their values. According to my
view, for example, a person who exercises well-developed, well-coordinated
agentic skills will come across as perspicacious, innovative, rational, self-aware,
confident of her worth, stable, tenacious, and resilient. Yet nothing prevents a
value-utilizing theory from judging people to be autonomous whod be sur-
prised if someone pointed out to them that they express its constitutive value
set in the way they make choices and act on them. This point brings to mind
one of Harry Frankfurts astute observations about autonomy:
The enjoyment of freedom comes easily to some. Others must struggle to
achieveit.27
For those who achieve autonomy easily, relying on autonomy skills and
expressing an autonomy theorys constitutive values have become second
nature. For those who must struggle for autonomy, methodically cultivating
autonomous decision-making procedures and resolutely upholding the values
that subtend them may sometimes be necessary if they are to sort out their
values and stick to theirgoals.
Taken together, the claims that (1)an autonomous choice process requires
exercising a set of capacities, (2)people may regularly exercise these capaci-
ties merely as a result of a fortunate upbringing, and (3)these capacities are
underwritten by constitutive value sets suggest the need for two axes to char-
acterize autonomy theories. Whereas the Directivity Axis indicates the scope
of discretionary choice that an autonomy theory condones, the Constitutivity
Axis indicates the values an autonomy theory invokes in specifying the process
of autonomous choice.
A strand of Andrea Westlunds thought sheds some light on the idea Im
trying to convey. For Westlund, to be autonomous you must have internalized
a disposition to answer for yourself.28 She emphasizes that self-answerability
27
Harry Frankfurt, Importance of What We Care About,22.
28
Westlund, Selflessness and Responsibility, 496; Westlund, Rethinking Relational Autonomy,
3637. Please note that Westlunds view is evolving. Although she still holds that her view of autonomy is
formal, she now holds that a disposition to answer for yourself is partly constitutive of the answerabil-
ity self-relation but not the whole of it. See Andrea Westlund, Reply to Benson, Christman, Rocha, and
Stoljar, Symposia on Gender, Race, and Philosophy 7:1 (2011):23, http://web.mit.edu/sgrp; Westlund,
does not depend on favorable beliefs about or attitudes toward yourself, nor
does it presuppose that you are symmetrically situated in relation to those who
may pose challenges to your commitments or in relation to those with whom
you interact most frequently or intimately.29 It is a formal relation, constituted
by a disposition to respond to normative pressures on ones commitments.30
Having assimilated this disposition, you take queries about your value com-
mitments from other points of view seriously, and you are prepared to examine
your reasons for your commitments.31 However, satisfying this criterion does
not oblige you to engage in hard-knuckle debate with every and any opponent
of your beliefs and practices. Challenges sometimes come from left field, and
challengers are sometimes out of line.32 Moreover, telling stories or describing
exemplars are acceptable ways to justify your beliefs and actions, and thinking
through legitimate challenges privately may suffice.33 Westlunds answerabil-
ity criterion cognizes a variety of discursive styles, and it grants individuals a
measure of discretion in their encounters with would-be challengers.
Westlund concludes that properly understood, autonomy is relational and
value- neutral.34 In characterizing self-answerability as a formal relationship to
potential interlocutors, Westlund sets aside the claim that autonomy requires
egalitarian relationships with others or a self-respecting relationship to your-
self. Specifically, she is rejecting not only Stoljars and Babbitts contention that
deference to a subordinate social status is incompatible with autonomy but
also Bensons contention that the self-doubt and self-effacement promoted
by subordination are incompatible with autonomy.35 Iagree that her theory is
value-neutral. Still, Iwould argue that her account is value-utilizing, for she
grounds her conception of answerability in the epistemic value of rational
Autonomy, Authority, and Answerability, Jurisprudence 2:1 (2011):171n25. Here, Iquote from her ear-
lier work because she articulates the formality of her position in an especially clear and emphatic way
in these papers. In my discussion in Section 4 Ill shift to her modified account of the self-answerability
self-relation. Inote, though, that it is possible that her contribution to this volume may further amend
or jettison parts of the view she has previously published. Iwont attempt to take her chapter into ac-
count because Ihavent had access to it in time to do so. Since my purposes in considering her work are
(1)to help me clarify the distinction between the Directivity Axis and the Constitutivity Axis (Section
2)and (2)to illustrate the usefulness of the Constitutivity Axis (Section 4), Idont think my discussion
will do any injustice to herwork.
29
Westlund, Rethinking Relational Autonomy, 27,39.
30
Westlund, Rethinking Relational Autonomy, 37. Also see Westlund, Selflessness and
Responsibility, 520n16.
31
Westlund, Selflessness and Responsibility,497.
32
Westlund, Autonomy, Authority, and Answerability,171.
33
34
35
For related discussion, see the valuable distinction in Westlund, Rethinking Relational
Autonomy, 32, between deference and deep deference. As my purpose here is not to assess the
strengths and weaknesses of Westlunds account but rather to use her account to illustrate my distinc-
tion between the directivity axis and the constitutivity axis, Iwont take up the question of whether
Westlunds account successfully fends off substantive theories of autonomy.
dialogue and the moral value of interpersonal accountability.36 If so, her claim
that answerability is a formal relationship must do double duty. In addition
to addressing Stoljars, Babbitts, and Bensons reservations about the tena-
bility of value-neutral theories, it must provide a convincing response to the
claim that value-utilizing theories dont differ in kind from value-laden or
value-saturated theories.
I believe that Westlund underscores the formality of the answerability
relationship for the same reason that I refer to autonomy competency as a
modus vivendi. Both of us aim to preserve the value neutrality of our views.
Westlunds conception of self-answerability as a formal relation and my con-
ception of autonomy competency as a modus vivendi suggest an analogy to the
experimental method in science. The experimental method rests on a bundle
of epistemic valuesfor example, accurate data and replicatable resultsyet
it does not predict what questions scientists will pose or what results their
experiments will yield. Likewise, our value-utilizing autonomy theories and
the constitutive value sets implicit in them do not predict what autonomous
people will choose to do or become other than being autonomous.37 Just as the
answerability relationship does not preordain what propositional content your
answers will contain, so too autonomy competency does not predetermine the
selection of projects and aspirations that youll decide to organize your life
around. Unlike personal life plans or individualized value systems, answer-
ability and autonomy competency are frameworks suitable for anyone to use
in conducting alife.
That some sets of constitutive values are amenable to being diverted into
the service of a nonautonomous way of life further supports the feasibility of
conceiving value-utilizing autonomy theories that are value-neutral. Consider
some of the values implicated in the agentic skills Ispecify. Such attributes as
perspicacity, resourcefulness, creativity, rationality, self-esteem, stability, resil-
ience, tenacity, and corrigibility might be directed away from leading a life of
your own and dedicated exclusively to raising perfect children or writing great
novels. An individual might take up the project of parenting or fiction non-
autonomously, and she might assiduously devote her talent and energy to ex-
cellent parenting or writing as judged by standards set by others. Living up to
36
Westlund, Selflessness and Responsibility for Self, 494495; Westlund, Rethinking Relational
Autonomy, 3435. Paul Benson, Commentary on Mackenzie and Poltera (2010) and Westlund (2009),
Symposia on Gender, Race, and Philosophy 7:1 (2011):3 makes a similar point about Westlunds position.
And in her response to Bensons commentary, Westlund, Reply to Benson, Christman, Rocha, and
Stoljar, 3, concedes that her view invokes what Iam calling a constitutive valueset.
37
This point is on a par with David Vellemans claim about the role of your understanding in
autonomous agency. According to Velleman, your understanding is your inalienable point of view in
the following sense:You cannot suppress or abandon this capacity without sacrificing your status as
an autonomous agent. See J. David Velleman, The Possibility of Practical Reason (NewYork:Oxford
University Press, 2000), 137138.
these external standards might enlist perspicacity, resourcefulness, creativity,

rationality, self-esteem, stability, resilience, tenacity, and corrigibility. Yet if she
never considers whether her life expresses her distinctive desires, needs, and
valuesher practical identity as a unique individualshe is not acting au-
tonomously although she may be acting admirably. The same is true of the
constitutive value set of Westlunds theory. Aphysicist could scrupulously ad-
here to the value of rational dialogue in scientific exchanges and the value of
interpersonal accountability in reporting her data, but she might not be in the
least disposed to answer for herself in other respects. In other words, the sets
of constitutive values that autonomy theories invoke are not equivalent to the
practices that the theories specify to explicate autonomy. Only if you make use
of the capacities that an autonomy theory sets forth to answer questions about
who you are and what matters to you will it judge you to be autonomous. It is
not enough to display the values that comprise the autonomy theorys consti-
tutive value set in the pursuit of some project orother.
The sets of constitutive values implicit in Westlunds and my theories do
not impeach their value-neutral credentials. However, not all value-utilizing
theories are value-neutral. For example, the strong normative competence
theory that Stoljar appropriates from Bensons early work is value-utilizing in
a way that has prescriptive implications for the choices autonomous agents can
make. The epistemic values of true beliefs about correct norms and reflexive
critical acumen that are included in the constitutive value set of strong nor-
mative competence entail that an autonomous agent cannot forsake genuine
norms. Not only must autonomous individuals be able to judge which norms
are correct, but also they must be able to discern how those norms apply to
their own situations. It seems, then, that Benson is right to charge Stoljar with
conflating autonomy with orthonomy.38
Value-utilizing theories regard autonomy as a valuable way of life and in-
voke values to explicate what the process of autonomous choice and action
consists in. But whereas value-laden and value-saturated theories proscribe or
prescribe the nature of your chosen doings, some value-utilizing theories leave
the nature of your chosen doings up to you. Instead of focusing on acceptable
outcomes of autonomous choice, value-neutral autonomy theories that are
value-utilizing focus on the process of autonomous decision making and de-
cision enactment. Since this is so, pointing out that a theory is value-utilizing
really does differ from pointing out that it is value-laden or value-saturated,
for value-utilizing theories can secure value neutrality. It is clear, then, that the
Constitutivity Axis highlights a relation between autonomy and values that
doesnt map onto the Directivity Axis. To advocate distinguishing between
these two axes is to advocate what Ill call the Double Axis Thesis.
38
Benson, Feminist Intuitions,132.
I close this section as Ibegan it with a reflection on Paul Bensons work on sub-
stantive autonomy theory. Near the end of the essay in which he introduces his
conception of weakly substantive theories, Benson incisively remarks:
More interesting is our finding that the distinction between substantive and
content-neutral theories, which seemed so sharp and straightforward in ear-
lier phases of contemporary writing on autonomy, should be reconceived as
a range of theories that impute varying kinds of normative substance, through
disparate pathways, to autonomous agency.39
Although this passage might be read as affirming a single spectrum of substan-
tive theories, Ithink that Benson might be receptive to the Double Axis Thesis.40
After all, he clearly recognizes that different theories incorporate different kinds
of norms in different ways, and the purpose of the Double Axis Thesis is to pro-
vide a framework for articulating and analyzing this heterogeneity. In the next
section, Ill say why as a feminist Iadvocate the Double Axis Thesis.
3. The Double Axis Thesis and Feminist AutonomyTheory
In Section 1, I probed some well-known feminist autonomy theories to ex-

pose the various ways values figure in them. This inquiry led me to distinguish
the Directivity Axis from the Constitutivity Axis. Section 2 clarified how the
two axes differ and defended the distinction between them. Inow urge that
the Double Axis Thesis secures two major benefits for feminist philosophy of
action. First, by separating the Directivity Axis from the Constitutivity Axis,
the Double Axis Thesis avoids foreclosing the category of value-neutral au-
tonomy theories. In my judgment, feminists should prefer a conceptual ap-
paratus that permits assessment of theories that purport to be value-neutral
on their merits rather than denying their cogency by fiat. Second, in virtue of
recognizing the category of value-neutral autonomy theories, the Double Axis
Thesis avoids prejudging how feminist philosophers should theorize the vexed
topic of adaptive preferences. Ill offer what Ibelieve to be compelling practical
considerations that militate in favor of detaching feminist analyses of adaptive
preferences from feminist autonomy theory. To make my case for these claims,
Ill rely on the contrast between the Double Axis Thesis, which is conjoined to
the conception of the Directivity Axis Ilaid out in Section 1, and what Ill call
the Single Axis Thesis, which holds that all autonomy theories are somehow
prescriptive and thus to some degree substantive.41
39
Benson, Feminist Intuitions, 137. Emphasisadded.
40
In email correspondence, Benson has told me that he endorses the Double Axis Thesis as Ive
laid itout.
41
Mark Piper explicitly argues for the Single Axis Thesis, though not under that rubric. See Mark
Piper, Autonomous Agency and Normative Implication, Journal of Value Inquiry 46:3 (2012):317330.
Like most philosophers of action, Itake the idea of self-governance through

the exercise of practical intelligence to be the core of what we mean by auton-
omous choice and action. While the theme of self-reflection is pervasive in
philosophical accounts of autonomy, different theories accentuate different
dimensions of self-governance and subscribe to different accounts of prac-
tical intelligence. Some theories prioritize self-expression or self-disclosure.
Some stress self-control or self-direction. Some foreground self-ownership
or self-responsibility. By no means mutually exclusive, these themes point to
nuances that differentiate among approaches to theorizing autonomy. Still,
theories of autonomy, including feminist ones, do seem deeply divided be-
tween those that see individuality and freedom as the hallmarks of autonomy
and those that see responsiveness to reasons as the hallmark of autonomy. The
Constitutivity Axis carves constitutive value setsthe values that are implicit
in autonomous choice proceduresaway from the question of what reasons an
agent has to act in the situation she is facing.42 The Directivity Axis furnishes
a tool to conceptualize an important way in which theories that anchor au-
tonomy in responsiveness to reasons differ from those that anchor autonomy
in individuality and freedom.
No doubt autonomy theories centered on responsiveness to reasons are
heirs to Kants legacy. The idea that an autonomous person acts on objective
reasons discovered through the exercise of reason represents one version of
the value-saturated pole of the prescriptivity spectrum. Still, value-saturated
theories of autonomy need not valorize reason and marginalize other epi-
stemic faculties. For example, Susan Babbitts feminist value-saturated theory
relies on nonpropositional knowledge and experience of community as re-
sources for detecting genuine reasons.43 Value-laden theories are less strict
about compliance with correct norms. However, they do put a premium on
grasping and acting on genuine norms, for you gain autonomy as the influence
of oppressive norms wanes in your conduct. In suggesting that people who
resist oppressive norms are more autonomous than people who follow them,
Marilyn Friedman seems to take up a position in the value-laden sector of the
Directivity Axis, her defense of value neutrality notwithstanding.44 For femi-
nist proponents of value-neutral theory, any other position on the Directivity
Axis substitutes moral rectituderule by authoritative norms or guarantees of
emancipationfor autonomy. Value-neutral theories prize living by your own
lights, such as they are. Thus, they carry the banner for the view that leading an
autonomous life is principally an exercise in individuality and freedom.
42
However, I note that a constitutive value set that entails a value-saturated position on the
Directivity Axis shrinks the gap between the two axes. See my comments on strong normative compe-
tence theory in Section2.
43
Babbit, Feminism and Objective Interests, 253256.
44
Friedman, Autonomy, Gender, Politics, 2024,67.
I cannot address the question of the authority for autonomous choice and
action in all its complexity here. However, Ido want to say a little about why
it would be inadvisable for feminists to embrace the Single Axis Thesis. This
framework exacts the following trade-off:It downplays the possibility that an
individuals distinctive attributes provide the authority for autonomy in order
to highlight the fact that all credible autonomy theories incorporate values
in some way. Of course, the Single Axis Thesis does not deny that there are
stronger and weaker versions of substantive autonomy theories. Some sub-
stantive theories issue restrictive commandsfor example, an autonomous
woman must not risk pregnancy to prove that shes fertile and marriageable.45
Such strongly substantive theories are far more overbearing than weaker ones
that allow women to defer to mens sexual demands and engage in risky re-
productive behavior provided that they have distinctive personal motives for
doing so.46 Weakly substantive theories are quite permissive, and they secure
greater latitude for agents to express their particular needs, desires, and values.
This being so, it might seem that the Double Axis Thesis sacrifices theo-
retical parsimony for the sake of a mere label and a misleading label at that.
Calling an autonomy theory value-neutral, it might be objected, perpetuates
the misconception that it is possible to theorize autonomy without invoking
any values apart from autonomy itself. This objection would be sound if Iwere
maintaining an alternative Single Axis Thesisone that acknowledged only the
spectrum ranging from value-saturated theories to value-neutral theories with
value-laden theories in between. However, the Double Axis Thesis is designed
to perspicuously represent both the possibility that self-expression takes prec-
edence over reason-responsiveness in autonomous action and the insight that
all accounts of the process of autonomous choice implicitly or explicitly invoke
values. Whereas the array of constitutive value sets on the Constitutivity Axis
encodes the latter insight, the value-neutral pole of the Directivity Axis sanc-
tions autonomous expression of individuality.
For feminists, the Double Axis Thesis and its correlative conception of the
Directivity Axis are attractive on several counts. One reason to foreground in-
dividuality and freedom through value-neutral autonomy theory is to preserve
the dynamism of the feminist liberatory agenda. Feminists have repeatedly
underscored the personal and societal damage caused by silencing womens
voices.47 Value-neutral autonomy theory guards against suppressing the diver-
sity of womens perspectives and concerns because it does not preemptively
45
46
For related discussion, see Benson, Feminist Intuitions, 128130.
47
Maria Lugones and Elizabeth V. Spelman, Have We Got a Theory for You! Feminist Theory,
Cultural Imperialism, and the Demand for the Womans Voice, Womens Studies International Forum
6 (1983):573581; Maria Lugones, Playfulness, World-Traveling, and Loving Perception, Hypatia 2:2
(1987):319. Meyers, Gender in the Mirror,1722.
deny the autonomy of any womans beliefs about how she should live. Because
its zero point is value neutrality, the Directivity Axis allows for an open-ended
view of the scope of autonomous choice. Another advantage of conceiving the
Directivity Axis as Ipropose is that it creates theoretical space for an innova-
tive conception of value-laden theoriestheories that take into account the
potency or harmfulness of the influence of oppressive norms in individuals
lives. Feminist exploration of this category of theory may prove fruitful, for the
value-laden sector of the Directivity Axis invites theorists to develop more re-
fined analyses of how oppression enters into diverse individuals deliberations
and conduct. Finally, the Double Axis Thesisand specifically the Directivity
Axis it positsclarifies what is at stake in feminist debates regarding adaptive
preferences. Iturn to that issuenow.
In my view, the main appeal of value-laden and value-saturated theories of
autonomy stems from feminist concerns about adaptive preferencesprefer-
ences that are harmful to the agent, that are formed in response to an oppres-
sive context, and that help to perpetuate oppression. Feminists often affirm
that an account of autonomy should have something useful to say about adap-
tive preferences. But just what an account of autonomy should contribute to
discussions of adaptive preferences is controversial.
Value-neutral accounts address the problem of adaptive preferences
obliquely. For example, Ibelieve that by and large and in the long run women
who have full repertoires of well-developed, well-coordinated agentic skills,
who exercise these skills frequently enough to keep them from atrophying,
and who mobilize them when they make significant decisions or when they
sense trouble in their lives are less likely to submit to oppressive practices
and more likely to join with others to overcome oppression.48 Along similar
lines, Westlund adduces relational capacities for self-correction to explain how
adaptive preferences can be exposed and shunned. In her view, the process
of grappling with external critical perspectivesthat is, challenges to your
beliefs and values that are posed by someone who doesnt share your world-
viewcan prompt you to reconsider misguided self-subordinating prefer-
ences and to modify resulting patterns of behavior.49 Neither of us believes that
any capacity for autonomy provides failsafe protection against the persistence
of adaptive preferences, but both of us recognize the potential of exercising
agentic capacities for eroding adaptive preferences.
In contrast, value-saturated and value-laden accounts confront the problem
of adaptive preferences head-on. To the extent that women enact adaptive
48
Diana Tietjens Meyers, Feminism and Womens Autonomy:The Challenge of Female Genital
Cutting, Metaphilosophy 31 (2000):469491.
49
Andrea Westlund, Autonomy in Relation, in Out from the Shadows: Analytical Feminist
Contributions to Traditional Philosophy, edited by Sharon Crasnow and Anita Superson (NewYork;
Oxford University Press, 2012), 61,7576.
preferences they lack autonomy, and when they shed adaptive preferences they
gain autonomy. It is surely tempting to kill two birds with one stone in this
way. Yet there is reason to hesitate before adopting this solution.
Babbitts value-saturated theory espouses conversion experiences that
rid you of adaptive preferences as a step toward autonomous agency.50
Paradoxically, she holds that you can gain autonomy by nonautonomously
relinquishing adaptive preferences.51 Now, few would deny that you can au-
tonomously decide to temporarily suspend autonomous control in order to
achieve an autonomously adopted goalsay, undergoing hypnosis to quit
smoking. It is clear, though, and Babbitt agrees that involuntarily undergoing
a nonautonomous transformation can destroy autonomy.52 You might end up a
mindless acolyte of a crazed cult leader. But because Babbitts position presup-
poses that those who induce you to undergo a conversion experience are right
about its capacity to augment your autonomy, her position is either incomplete
or faulty. Incomplete if it does not furnish criteria for entrusting yourself to
convertors or criteria that pick out autonomy-enhancing conversion schemes.
Faulty if it leaves autonomy enhancement toluck.
Serene Khaders arguments against analyzing adaptive preferences as au-
tonomy deficits spark related concerns about value-saturated and value-laden
theories of autonomy.53 As Khader observes, building adaptive preferences into
a theory of autonomy unwisely presumes that you already know which types
of behavior are expressions of adaptive preferences.54 Khaders more circum-
spect approach posits a minimal perfectionist account of flourishingthat
is, a conception of flourishing that is limited to ensuring basic human needs
and that musters broad assent from members of diverse cultural groups.55
Where women fall short of enjoying this minimal level of flourishing, there
are grounds for suspecting that adaptive preferences have taken root, but this
suspicion can be vindicated (or quashed) only through a process of delibera-
tive inquiry with the women who lead this form of life.56 Because peoples rea-
sons for acting as they do are not transparent, two individuals may be acting
in a seemingly identical way, yet one may be acting on an adaptive preference
while the other isnt. If so, adaptive preferences cannot be singled out on the
basis of the gendered social meanings feminist theorists ascribe to different
types of behavior. It seems, then, that Khaders deliberative approach to identi-
fying adaptive preferences would counsel locating feminist autonomy theories
50
Babbit, Feminism and Objective Interests,251.
51
Babbit, Feminism and Objective Interests, 254256.
52
Babbit, Feminism and Objective Interests,251.
53
Serene J. Khader, Adaptive Preferences and Womens Empowerment (Oxford:Oxford University
Press, 2011), chap.2.
54
Khader, Adaptive Preferences, 75, 9697,103.
55
Khader, Adaptive Preferences, 21,138.
56
Khader, Adaptive Preferences, 3536, 42, 6073, 138140.
at the value-neutral end of the Directivity Axis where autonomy depends on

the process of choosing rather than on whether the actions that are chosen re-
inforce oppression or lead to emancipation.
A final worry about identifying more emancipated behavior with more
autonomous behavior is that it assigns disproportionate responsibility for
social change to individual women. It seems to me that the onus of confronting
and overcoming structural injustice should be borne by social movements and
persons in politically or economically powerful positions. Yet to avoid having
their autonomy downgraded by a value-laden or value-saturated account of
autonomy, women facing entrenched patriarchal institutions and repressive
practices would be obliged to devise individual workarounds consonant with
progressive values. In my estimation, then, such theories load too much of the
work of fighting injustice onto vulnerable individuals. Of course nothing is
likely to change for the better unless the victims of oppression band together
and take collective action. Still, theories of autonomy that make assessments
of autonomy contingent on social activism or on individual defiance of unjust
norms are in a kind of double jeopardy. Such theories risk condemning auton-
omous commitments to practices that feminists currently abhor, and they also
risk celebrating the autonomy of individuals who are nonautonomously enact-
ing values that feminists currently approve. Both types of jeopardy bespeak
a failure to keep in mind that feminist values and goals are always subject to
contestation and have undergone striking changes overtime.
For all of the preceding reasons, I believe that value-neutral autonomy
theory best serves feminist aims. Consequently, Iurge feminists to embrace
the Double Axis Thesis, which fields a conception of the Directivity Axis that
renders value-neutral theory intelligible. Still, this conclusion hardly settles
the question of which value-neutral theorythat is, which account of the pro-
cess of autonomous choice and actionis most in tune with feminist aims.
The Constitutivity Axis, Ill argue, provides a novel diagnostic tool for assess-
ing the merits of competing value-neutral theories of autonomy.
4. The Constitutivity Axis as PhilosophicalTool
It is a truism that autonomy is a cluster concept, for autonomy is a term of art,

and different theories of autonomy are designed for different philosophical
purposes. Examples of philosophical problems that theories of autonomy may
seek to solve or contribute to include:
1. When are individuals responsible for what theydo?
2. How can action be guided by reasons?
3. Who is entitled to give or withhold consent to government?
4. When is paternalistic intervention justified or unjustified?
5. What conditions secure social freedom or emancipation?

6. How is it possible to lead a life of yourown?
7. What sorts of choices and actions promote personal fulfillment or
self-realization?
Because these problems are interrelated, philosophers often undertake to solve
more than one of them with a single theory of autonomy. Some examples:David
Velleman aims to provide an account of acting for reasons that accords with
our conventions of assigning responsibility for action; John Christman aims
to provide an account of leading a life of your own that ensures that the vast
majority of adult humans qualify as citizens whose consent confers legiti-
macy on governments; Marina Oshana aims to provide an account that binds
leading a life of your own to social freedom and opportunity.57 Icannot explore
the full range of autonomy theories here. However, it is incumbent on me to
indicate how philosophers can profitably use the leverage of the Constitutivity
Axis. In my view, positioning autonomy theories on the Constitutivity Axis by
teasing out their constitutive value sets exposes the inner workings of diver-
gent autonomy theories and thereby connects each theory to a subset of phil-
osophical problems regarding autonomy.
Since Ihave argued that feminists are well-advised to adopt a value-neutral
position on the Directivity Axis, I will examine Westlunds value-neutral,
value-utilizing theory of autonomy to clarify the philosophical promise of the
Constitutivity Axis. Because Westlunds self-answerability theory aims to ac-
count both for leading a life of your own and for placing moral limits on pa-
ternalism, it needs to supply an account of autonomy as well as an account of
nonautonomy. However, Ill urge that her account of autonomy presupposes
one constitutive value set, whereas her account of nonautonomy presupposes
another. Thus, focusing on the Constitutivity Axis suggests that Westlund
overreaches in attempting to address both the problem of self-direction and
the problem of justified paternalism in one fell swoop. Nevertheless, Ill argue
that her view is particularly helpful for judging when paternalistic interven-
tions are appropriate in real-world medical and social service settings.
Recall that for Westlund self-answerabilitythat is, the self-relation in
which the agent holds herself to the expectation or demand that she respond
appropriately to legitimate criticism, experiencing herself as owing a suitable
response, at least under certain conditionsis a necessary condition for au-
tonomy.58 Westlund is surely correct that we are likely to regard people who
address earnest queries about their practical commitments in a forthright
57
Velleman, Possibility of Practical Reason; Christman, Politics of Persons; Marina Oshana, Personal
Autonomy in Society (Aldershot, UK:Ashgate,2006).
58
Westlund, Autonomy, Authority, and Answerability, 171; Westlund, Selflessness and
Responsibility for Self, 500, 503; Westlund, Rethinking Relational Autonomy,28.
fashion as autonomous and therefore immune to paternalistic interference.59

Conversely, in some circumstances we become skeptical about someones au-
tonomy when she mouths someone elses words or reasons in a tight little circle,
always sidestepping what an interlocutor who is not like-minded is pressing
her to consider.60 When the stakes are high and time is short, such resistance
to rational exchange lowers the barriers against paternalistic intervention. In
these respects, Westlunds view is clearly in sync with our interpersonal prac-
tices of autonomy attribution and our closely related practices with respect to
treating others paternalistically.
Still, Westlunds ambition is not confined to depicting how we judge
whether someone is autonomous and whether we should defer to her deci-
sions about her life. In her view, the psychological state of readiness to re-
spond to external challenges to your action-guiding commitments together
with your sense that it is imperative to do so is a crucial part of what con-
stitutes [you] as self-governing.61 How can the Constitutivity Axis help us to
think through thisclaim?
As Ipointed out earlier, Westlunds avowed constitutive value set is com-
posed of the epistemic value of rational dialogue and the moral value of inter-
personal accountability. This pair of values brings the question of the relation
between the stance of self-answerability and the practice of answering for
yourself to the fore. The appeal of the epistemic value of rational dialogue is
that it introduces an independent check on your thought processes, and the
appeal of the moral value of interpersonal accountability is that it rules out
sinking into solitary complacency. As a participant in ongoing rational dia-
logue with different-minded conversation partners who hold you accountable
for the cogency of your answers, you can be confident that you really do value
what you profess to value and that you really do want to lead the life youre
living. Yet Westlunds comments about the practice of answering for yourself
do not comport with this robustly relational interpretation of her position.
According to Westlund, autonomous agents must maintain attitudes of
humility and open-mindedness toward their values and goalsthat is, they
must acknowledge their fallibility and be prepared to recognize the possi-
bility that theyve been mistaken about whats truly worthwhile.62 However,
she regards comparatively attenuated forms of dialogical engagement and ac-
countability as compatible with humility and open-mindedness. Ihave already
mentioned that she countenances less formal discursive styles than rational
debate, such as telling pertinent stories or describing admirable persons and
their conduct. As well, she leaves it to individuals whose values and desires
59
60
61
62
Westlund, Selflessness and Responsibility, 508; Westlund, Reply to Benson,4.
are being challenged to judge whether an interlocutors objections to her

values and desires deserve a response. Here Istress that satisfying Westlunds
self-answerability criterion does not always require individuals to address crit-
ical challenges posed by an actual interlocutor, nor does it always require them
to answer such challenges in an actual conversation. Instead, you may think
up arguments against your values and desires from a point of view opposed to
your own and formulate responses in your minds eye.63 Indeed, Westlund has
recently severed self-answerability from actually answering for yourself:You
can be self-answerable if you would respond to self- or other-generated chal-
lenges should they ever arise.64
This steadily thinning view of what counts as answering for yourself raises
questions about the tie between the self-answerability criterion Westlund
defends and the constitutive value set she articulates. Since she holds that ra-
tional dialogue can take place internally, she cannot also hold that there must
be an independent check on the pointedness of the objections you raise and
the adequacy of your replies. Since she holds that interpersonal accountability
need not be instantiated in conversations with actual interlocutors, she cannot
also hold that discursive isolation is necessarily inimical to autonomy. Insofar
as your desires and values have never been called into question, you need not
answer for them at all. Neither dialogue with others nor internal dialogue is
needed to certify that they are your own. On all of these counts, the constitu-
tive relationality of Westlunds theory is compromised.65
A more fitting constitutive value set for Westlunds fully developed view
would conjoin reflexive insightfulness with accountability to yourself.66
Conceptualizing answerability according to this constitutive value set would
reflect the allowances Westlund rightly makes for self-questioning as a form
63
Westlund, Selflessness and Responsibility, 506, 507, 510; Westlund, Rethinking Relational
Autonomy, 3741; Westlund, Reply to Benson,4.
64
Westlund, Reply to Benson,5.
65
Westlund cant resolve this problem by affirming that generating internal dialogues and being
accountable to yourself for your actions presuppose that youve internalized the answerability mode
of relating to others, for this move would reduce answerability to a causally relational, as opposed to a
constitutively relational, theory of autonomy. Nor does her claim in Westlund, Relational Autonomy,
35, that the internal psychological condition of the autonomous agent...point[s]beyond itself, to the
position the agent occupies as one reflective, responsible self among many resolve this problem. If this
internal psychological condition is never activated by external challenges to your important values and
desires, an autonomous person will be indistinguishable from a nonautonomous person. For related
discussion, see Westlund, Autonomy, Authority, and Answerability, 175; also see Westlund, Reply to
Benson,5.
66
This pair of values brings out the kinship between Westlunds thought and the suggestion in
Thomas Nagel, The View from Nowhere (NewYork:Oxford University Press, 1986), 127, that the best
we can do is to try to live in a way that wouldnt have to be revised in light of anything more that could
be known about us. In email correspondence, Westlund speculates that it may be possible to theorize
reflexive insightfulness and accountability to yourself relationally. So she may seek to salvage the con-
stitutive relationality of her view by offering such an account in futurework.
of answerability. Moreover, because authorizing people to meet the answer-

ability condition intrapersonally doesnt rule out making use of interpersonal
strategies to deepen your self-knowledge and to strengthen your resolve, it
preserves a benign form of relational autonomy. Not only do the quality of
your internal dialogue and your determination to uphold certain standards
seem more germane to Westlunds avowed projectnamely, theorizing what it
takes for choices and actions to be genuinely your ownbut also these values
tally better with the linkage Westlund affirms between self-responsibility, pos-
sessing a self, and answerability.67 Unfortunately, if Westlund were to adopt the
constitutive value set Ive proposed, she would be committed to a constitutive
value set that is out of alignment with a major argument that she offers in sup-
port of herview.
Westlund takes up feminist concerns about the deleterious impact that
internalized oppression and adaptive preferences have on autonomy when she
urges that a virtue of her answerability account is that it clarifies what it is to
lack reasons of your own for avowing the values you avow or for wanting what
you want.68 Her characterizations of nonautonomy as being tightly gripped
by a value system or by a practical reasoning procedure and being utterly
impervious to objections suggest that nonautonomous agents suffer from
an intellectual form of locked-in syndrome. If this metaphor is apt, staging
internal dialogues and accounting for yourself to yourself are bound to per-
petuate your agentic malady rather than release you from it. It seems, then,
that either Westlund must embrace a requirement for autonomy that she her-
self regards as implausibly strongcommunicating your responses to critical
challenges lodged by other people in face-to-face encountersor she must
jettison her claim that her view makes sense of lacking autonomy because you
are in the grip of a dubious value system or a defective reasoning policy.
Its hardly surprising that the constitutive value set that shores up Westlunds
full account of self-answerability is at odds with the constitutive value set that
shores up her account of being tightly gripped. The elastic constitutive value
set Ive associated with her account of autonomyreflexive insightfulness to-
gether with accountability to yourselfmakes needed conceptual room for
degrees of autonomy and pockets of autonomy. Someone might be fully an-
swerableintrospectively attuned to her needs and lucid about her reasons
with respect to making medical decisions about her health but marginally
67
Westlund, Selflessness and Responsibility, 495500; Westlund, Rethinking Relational
Autonomy, 2830, 3738; Westlund, Autonomy in Relation, 7076. Although Westlund asserts that
we have good reason to treat pressures toward self-answerability...as the very form that the pres-
sures toward the self-governance of practical reasoning take, Inote that urging someone to think for
herself is no less salient a form that pressures toward self-governance take, and this advice comports
with the modified constitutive value set Ive proposed. Westlund, Selflessness and Responsibility,501.
68
Westlund, Selflessness and Responsibility, 500505; Westlund, Rethinking Relational
Autonomy, 2830, 3233; Westlund, Autonomy in Relation,7476.
answerablea bit muddled about her needs and hesitant about her reasons
with respect to her dealings with a domineering husband. If Westlunds theory
is to capture the complexity of womens autonomous agency, it must be able to
discriminate fine-grained gradations of degree as well as the qualities of rela-
tionships that are more and less conducive to answering for yourself. Perhaps
the woman who finds herself tongue-tied when answering to her domineering
husband is able to be sharply self-critical and articulate in answering criticisms
in a quiet moment alone. The constitutive value set Ive ascribed to Westlunds
nuanced answerability theory accommodates this possibility.
However, because being tightly gripped is an absolute state of nonautonomy,
a commensurately exacting constitutive value set is needed to underwrite it.
Whereas a modulated theory that accommodates internal reflection as a basis
for autonomy would confer autonomy on deeply deferential and blindly fanat-
ical individuals, a theory of autonomy premised on the constitutive value set
of rational dialogue and interpersonal accountability would not. If Westlunds
(and my) intuition that these individuals arent autonomous is correct, the less
stringent constitutive value set Ive attributed to Westlunds fully developed
position effectively abandons them to their nonautonomy.
This split between the version of Westlunds view that is needed to account
for autonomy and the version that is needed to account for nonautonomy
deprives her of a key argument for her answerability requirement. Once the
two constitutive value sets are in view, it is evident that her fully developed
theory of autonomy is too permissive to explain nonautonomy. Although it
is an open question whether a single theory might be able to explicate auton-
omous choice and action as well as nonautonomy resulting from adaptive
preferences, so far as Ican see Westlund does not provide a theory that accom-
plishes both aims.69
At this point, however, Iwould like to set aside these difficulties and con-
sider a notable strength of Westlunds position. Here, too, it is helpful to ex-
amine its underlying constitutive value set. In the course of presenting and
defending her view, Westlund describes a number of plausible scenarios in
which someones autonomy comes under scrutiny, and these discussions dem-
onstrate the tremendous practical appeal of her position. Social workers and
medical practitioners are often called on to judge whether a client or patient
has the cognitive wherewithal to competently direct her own affairs. Atheory
committed to the constitutive value set of dialogical rationality and interper-
sonal accountability suggests a sensible strategy for arriving at assessments of
autonomy in these social service and medical settings.
69
As Imentioned in Section 3, Serene Khader warns against theorizing adaptive preferences as
autonomy deficits.
Theres no simple protocol for interviewing a person in a way that elicits

reliable evidence of her self-answerability. Drawing out a stranger, attentively
tracking her idiosyncratic thought processes, and probing them gently and
sympathetically is no easy matter.70 Moreover, conclusions about a persons
autonomy based on her testimony must be regarded as controvertible judg-
ments. But detecting self-answerability is decidedly less abstruse than de-
ciding whether someone is identified with the desire she intends to act on
or whether she would be alienated from the desire if she were to reflect on it
in light of the history of its development.71 Similarly, it demands a decidedly
less tricky evaluation than trying to gauge the extent, depth, and frequency
of someones self-reflection, the intensity and certainty of her affirmation of
a desire, the depth or pervasiveness of her affirmed desire, the variety and
efficacy of her agentic capacities, and so forth or trying to gauge someones
overall proficiency with respect to autonomy competency and whether she is
successfully exercising these skills.72 In practical settings where momentous
decisions must sometimes be made quickly, these four accounts seem fanciful
if not downright comical.
Self-answerability anchored in the constitutive value set of rational dia-
logue and interpersonal accountability may be the best available criterion of
agentic competency in time-sensitive, high-stakes decision contexts. Given the
gravity of the affront to personal dignity that unwarranted paternalism inflicts
together with the superordinate importance of protecting nonautonomous
individuals from irremediable harm, self-answerability arguably strikes a fea-
sible and principled balance for assessments of agentic competence. Indeed,
my own multidimensional theory of autonomy complements Westlunds by
supplying a rationale for regarding self-answerability as a sufficient condition
for autonomy in social service and clinical settings. Surely it is reasonable to
suppose that people who command the agentic skills that make up autonomy
competency and who are able to deploy those skills in coping with their cur-
rent situations will be able to produce sensible answers to questions about their
medical treatment or public assistance options thereby satisfying professionals
charged with both respecting their autonomy and promoting their interests
that they meet Westlunds answerability requirement. Although Ihave reser-
vations about whether a conception of answerability premised on the consti-
tutive value set of dialogical rationality and interpersonal accountability is a
necessary condition for autonomy, Iam inclined to accept this conception as
70
As Westlund, Rethinking Relational Autonomy, 40, rightly observes, Part of the burden for
facilitating justificatory dialogue thus falls on the shoulders of the would-be critic, who must position
herself appropriately with respect to the agent in question.
71
Frankfurt, Importance of What We Care About, 5868; Christman, Politics of Persons,145.
72
Friedman, Autonomy, Gender, Politics, 720; Meyers, Self, Society, and Personal Choice; Meyers,
Gender in the Mirror; Meyers, Being Yourself.
a sufficient condition for acknowledging the autonomy of a person who is a

candidate for paternalistic intervention.
5. Summary of the Benefits of the Double AxisThesis
My aim in the forgoing discussions of Khaders theory of adaptive prefer-

ences and Westlunds self-answerability theory was not to reach a definitive
judgment as to the merits of their positions but rather to illustrate the philo-
sophical payoffs of the Double Axis Thesis. Here Ireview the benefits of dis-
tinguishing the Constitutivity Axis from the Directivity Axis. First, a theorys
location on the Directivity Axis signals its understanding of the authority for
autonomous agencyobjective reasons or your distinctive self. Second, the
Double Axis Thesis shows that coherent theories of autonomy can be both val-
ue-neutral and value-utilizing. Third, putting theories under the microscope
of the Constitutivity Axis can reveal sources of strain and strength in their
normative underpinnings. In other words, by exhibiting the implicit value
investments of theories of autonomy, the Constitutivity Axis highlights ways
in which autonomy theories tally (or fail to tally) with paradigmatic cases of
autonomy. Finally, because analyzing a theorys constitutive value set discloses
value commitments that might otherwise go unnoticed, the Constitutivity
Axis clarifies how and why theories of autonomy converge and diverge, and it
facilitates connecting theories to different philosophical problems concerning
autonomy. As a result, the Constitutivity Axis helps philosophers of action
avoid talking at cross-purposes.
I recognize that these are modest claims. Nevertheless, if Ihave succeeded
in providing a better map of the terrain of feminist work on autonomy and in
ending some faux controversies among feminist philosophers, Ill be satisfied
with those achievements.
Acknowledgments
Thanks to Andrea Veltman, Mark Piper, Andrea Westlund, Paul Benson, and
the participants in the Relational Autonomy: Ten Years On conference for
their invaluable comments on an earlier version of thispaper.
{7}
ACommitment to Autonomy Is a Commitment

to Feminism
MarinaOshana
1. The Problem
It is generally believed that autonomy consists, at least in part, of the ability to

choose and to act according to the direction of ones will, values, and prefer-
ences. Most of us in Western, liberal societies highly prize autonomy, to such
an extent that a legal right to autonomy is ensconced in constitutional prin-
ciples as well as in the tradition of common law. In what follows, Iargue that
if a moral and legal right to autonomy is to be respected then we must be
committed to feminism as a moral and a political ideal. In addition, if we are
committed to feminism as a moral and a political ideal then we must also be
committed to autonomy as a condition to which people are morally and legally
entitled.
To do this, Iwill use one type of life as a case in point as well as two per-
sonal histories. The type of life Iwill discuss is that of a low-wage laborer in the
United States. One of the two personal histories is the self-reported life story
of political activist Ayaan Hirsi Ali.1 The second example was recently reported
in the NewYork Times and recounts the ordeal of a young Afghani girl.2 There
are, of course, limitations to the conclusions one can draw about feminism
and autonomy from either story or from my remarks about low-wage labor.
Ido not wish to generalize the cases, but Ibelieve they are instructive. They
all lead me to wonder just how feminists construe autonomy and just how far
defenders of autonomy want to carry the banner of feminist causes. And they
make me wonder why the term feminist irks people or strikes the ear as an
outmoded moniker.
1
Ayaan Hirsi Ali, Infidel (NewYork:Free Press,2007).
2
Alissa J. Rubin, For Punishment of Elders Misdeeds, Afghan Girl Pays the Price, New York
Times, February 16, 2012, http://www.nytimes.com/2012/02/17/world/asia/in-baad-afghan-girls-
are-penalized-for-elders-crimes.html?_r=1&ref=todayspaper.
Let me begin by explaining what Imean by autonomy and by explaining how

Iunderstand the concept of feminism. Autonomy literally means self-rule or
self-governance. A person is autonomous when she is self-governing. She
is self-governing when she has a kind of authority over oneself as well as the
power to act on that authority. As Ihave argued elsewhere, the kind of authority
involved must be of a form and extent sufficient for a person to influence and
to oversee those of her choices, actions, and relationships that are definitive of
successful practical agency of a fairly comprehensive variety.3 This means that
the presence of social roles and relations that afford a person this influence and
authority are decisive if the person is to count as genuinely self-determining,
whatever her choices are for and however laudable they appear to be. Ichar-
acterize autonomy in this way because Itake seriously the idea that people are
complex beings and because I am concerned with issues having to do with
self-governance under a broad spectrum of social conditions. Ialso take se-
riously the idea that autonomy offers one dimension for evaluating human
development.4 Being autonomous implies that certain capabilities are available
to the agent, some of which are social and political in nature.
I will define feminism, as others have, as an intellectual commitment and
a political movement that seeks justice for women and the end of sexism in
all forms.... Feminism is a belief in and advocacy of equal rights for women
based on the idea of the equality of the sexes.5 Feminist ethics and politics
are concerned with issues pertinent to womens autonomy, including ending
the oppression, subordination, abuse, and exploitation of women and girls,
wherever these may arise.6 To some, the label feminist has an antiquated ring.
It conjures up images of bra burners, Gloria Steinem aviator frames, and an
eraallegedly pastwhen womens rights were in need of political support.
Others find the label distasteful, seeing it as evoking opposition to men (or
to what men allegedly represent) or to traditional values and social practices
such as marriage and a desire to raise a family. These are crude caricatures, of
3
See Marina Oshana, Personal Autonomy in Society (Aldershot, UK: Ashgate, 2006), 6467.
Psychological autonomy (in the realm of the inner citadel) is incapable of empowering a person
in many areas of life that are crucial to successful agency where conditions of this sort are present or
probable. Moreover, neither a legal right of authority over ones affairs nor a moral right of authority
shall suffice for autonomy.
4
What I will say on this point echoes some remarks made by Martha C. Nussbaum, Women
and Human Development: The Capabilities Approach (The Seeley Lectures), (New York: Cambridge
University Press, 2000); Nussbaum, The Future of Feminist Liberalism, reprinted in Cheshire
Calhoun, ed., Setting the Moral Compass:Essays by Women Philosophers (NewYork:Oxford University
Press, 2004), pp. 7288; and Nussbaum, Creating Capabilities: The Human Development Approach
(Cambridge:Belknap Press,2011).
5
In formulating this definition, Irelied on Sally Haslanger and Nancy Tuana, Topics in Feminism,
2004, http://plato.stanford.edu/entries/feminism-topics/ (original publication February 7,2003).
6
Marilyn Friedman, Feminism in Ethics:Conceptions of Autonomy, in The Cambridge Companion
to Feminism in Philosophy, edited by Miranda Fricker and Jennifer Hornsby (Cambridge:Cambridge
Commitment to Autonomy Is a Commitment to Feminism 143
course, but in the past twenty years they have been affirmed to a surprising de-
gree, especially among younger adults. Keeping these definitions of autonomy
and of feminism in mind, let us turn to the cases Imentioned:the low-wage
laborer; Ayaan Hirsi Ali; and a young Afghanigirl.
2. Autonomy and WageLabor
Autonomy-undermining social conditions that fall heavily on women are

to be found in the secular world of Western free enterprise. Certain areas of
wage-labor are illustrative of this phenomenon. Consider the circumstances of
the hotel maid, the housecleaner, the waitress, the nursing home aide, and the
Wal-Mart salesperson depicted in Barbara Ehrenreichs excellent book Nickel
and Dimed.7 These workers are disproportionately female.8 Their situations are
not uncommon. They perform work that is back-breaking, poorly remuner-
ated, lacking in benefits, and without security of employment. Each is em-
blematic of labor that is consistently exploitative and autonomy depriving.
I say this because it is a well-supported fact that persons require a minimum
level of economic self-sufficiency to minimize the likelihood that others
7
Barbara Ehrenreich, Nickel and Dimed:On (Not) Getting By in America (NewYork:Metropolitan
Books,2001).
8
Sarah Jane Glynn and Audrey Powers, See The Top 10 Facts About the Wage Gap:Women Are
Still Earning Less than Men across the Board, Center for American Progress, April 16, 2012, http://www.
americanprogress.org/issues/labor/news/2012/04/16/11391/the-top-10-facts-about-the-wage-gap/, re-
port that more than one-quarter of the wage gap is due to the different jobs that men and women
hold. Notably, women are more likely to work in low-wage, pink-collar jobs such as teaching,
child care, nursing, cleaning, and waitressing. The top ten jobs held by women include secretaries
and administrative assistants (number one); elementary and middle school teachers (number four);
cashiers and retail salespeople (number six); maids and housekeepers (number ten), and waitresses.
These jobs typically pay less than male-dominated jobs and are fueling the gender wage gap. In addi-
tion, Glynn, The Wage Gap for Women:The Consequences of Workplace Pay Inequity for Women
in America, August 16, 2012, http://www.americanprogress.org/issues/labor/news/2012/08/16/12029/
fact-sheet-the-wage-gap-for-women/, cites the familiar statistic that women who work full time year
round continue to earn only about 77percent of what men earn. The gap between the median wage
for a man and that of a woman in 2010 was a whopping $10,784 per year. That is enough money for
the average woman to fund a year of higher education and a full year of health care costs, while still
having more than $2,400 to contribute toward her retirement. The U.S. Department of Labor, Bureau
of Labor Statistics, Characteristics of Minimum Wage Workers:2011, March 2, 2012, http://www.bls.
gov/cps/minwage2011.htm, reports that in 2011, 73.9million American workers age 16 and over were
paid at hourly rates, representing 59.1percent of all wage and salary workers. Among those paid by the
hour, 1.7million earned exactly the prevailing Federal minimum wage of $7.25 per hour. About 2.2mil-
lion had wages below the minimum....About 6percent of women paid hourly rates had wages at or
below the prevailing Federal minimum, compared with about 4percent of men. The Department of
Labor, Labor Force Statistics from the Current Population Survey, February 16, 2011, notes, Women
who worked full time in wage and salary jobs had median weekly earnings of $657 in 2009. This rep-
resented 80percent of mens median weekly earnings ($819). To learn more, see Women in the Labor
Force:ADatabook (2010 Edition), BLS Report 1026, December 2010, http://www.bls.gov/cps/demo-
graphics.htm#women.
might gain control over them through their needs.9 Persons who rely on day
labor or low-income unskilled labor for their livelihood generally live in eco-
nomically precarious conditions. Often their physical and psychological states
are fragile as well. The same can be said of someone who is one paycheck away
from homelessness. Most occupy a state of vulnerability that, fortunately, is
not a feature of the actual situation of persons such as myself. Most persons
in my positionwell-educated, successful academics who enjoy the security
of a tenured job and its benefitsare autonomous. We who enjoy this kind of
security have genuine control over affairs of importance to us in large part be-
cause we are not living hand tomouth.
Of course, we must accept that some degree of insecurity and dependency is
present even in the lives of well-situated adults. Obviously, not all of instances
of insecurity and dependency compromise self-government in a serious way.
How are we to distinguish genuine threats to autonomy from more benign
and commonplace lapses of practical control and authority? Isuggest we em-
ploy a three-step assessment.10 To determine whether a persons situation in
life undermines the practical control demanded of autonomy, we must ask,
first, what is the insecurity a person encounters? Are significant areas of a per-
sons life at risk? Significant areas will encompass those in which a person has
a fundamental interest. For instance, such areas include the interest one has
in choosing ones partner, in making choices about whether to start a family,
and in ones pursuit of a career. In addition, most humans have a deep-seated
investment in the quality of their future lives, and what they can plan, hope
for, and expect depends in no small part on their current state of autonomy.
Notably, too, self-governance in these areas signals ones success as a human
agent. Being a successful agent means being capable of more than simple in-
tentional action. It means being able to own ones action. Next, we must ask
what is required to challenge the insecurity. How demanding is the type and
level of exertion a person must expend to confront the conditions of insecu-
rity and dependence that may undermine her autonomy? Finally, what effect
does this have on the person? What does it cost a person when struggle and
indefatigability are the price one must pay to have ownership over significant
domains of oneslife?
If an individual is financially insecure and this insecurity means she cannot
challenge potential encroachments on the part of those who have power
over her, then her autonomy will be diminished and of less practical value.
9
Diana Meyers, Self, Society, and Personal Choice (NewYork:Columbia University Press, 1989) 12.
As Meyers observes, If one can take care of oneself, one is beholden to no oneneither to the state
nor to any other individual(12).
10
The three-step assessment is developed in greater detail in my paper, Marina Oshana,
Is Social-Relational Autonomy a Plausible Ideal? in Personal Autonomy and Social
Oppression:Philosophical Perspectives, edited by Oshana (London:Routledge, forthcoming).
Successful agencythe cornerstone of autonomycalls for an absence of ex-

ploitation, of which financial exploitation is one variety. If those who are more
powerful are in a position to exploit the less powerful, to forcibly extract labor
from those who are economically or otherwise disadvantaged, the result is
an assault on autonomy.11 Extraction of labor is, plausibly, forcible if ones
circumstances leave one with little alternative but to acquiesce. As Marilyn
Friedman notes, Having to let the needs and desires of another person de-
termine the course of ones behavior in order to survive is a tragically heter-
onomous mode of existence.12 In contrast, the person who has the benefit of
financial security is less susceptible to these encroachments.
Control and authority can be undermined in different ways, one of which
is by the prevalence of psychological, material, and social impediments.
Archetypal examples are the conditions that black and mixed-raced persons
confronted in South Africa under the official governmental policies of apart-
heid and life for persons of color in the era of Jim Crow segregation laws in
the United States. But an impediment need not rise to the level of the systemic
oppression of apartheid or Jim Crow:dependency on some forms of low-wage
labor can be fairly characterized as an impediment of thissort.
Practical control and authority can also be undermined when they serve
a person only in activities that are of little import to human life. Again, none
of us has control and authority over every component our lives, at all times.
But we do not require muscle of such an all-encompassing variety to be
self-governing. In general, while the range of control a person requires will re-
flect the unique interests and ends of the person, the fact of control will be in-
variant. For example, while an autonomous academic will require the presence
of a unique set of practical conditions that support the variety of tasks essential
to her profession and an autonomous day-wage laborer will require that sim-
ilarly distinctive conditions be in place, neither will count as autonomous un-
less both can determine the trajectory of their lives without unduecost.
The aim of this section has been to offer a general conception of autonomy
and to shed light in a broad way on the variety of autonomy-constraining cir-
cumstances persons confront. One variety of life that illustrates the threat to
autonomy distinctively borne by women is that of low-wage and pink-collar
labor. Ihave suggested criteria for delineating circumstances that jeopardize
autonomy from more mundane conditions that are less problematic in this re-
gard. In Section 3, we will consider the cases of Ayaan Hirsi Ali and of a young
Afghani girl. Both highlight the threat to autonomy faced by females from
certain aspects of conservative patriarchy of a sort found in predominantly
11
Tommie Shelby, We Who Are Dark:The Philosophical Foundations of Black Solidarity (Cambridge,
MA: Belknap Press, 2005), 193. For a development of this claim, see Oshana, Is Social-Relational
Autonomy a Plausible Ideal?
12
Friedman, Feminism in Ethics:Conceptions of Autonomy,219.
Islamic cultures (as well as in very conservative strains of Christianity and some
ultra-orthodox sects of Judaism).
3. Cultural and Religious Patriarchy
Ayaan HirsiAli
Even while the details of her autobiographical narrative have been challenged,
facts about Ayaan Hirsi Alis life attest to a remarkable set of experiences. Hirsi
Ali spent her youth in a seminomadic Somali Muslim family, mainly in Kenya.
Her mother was a devout adherent of Wahhabi Islam, a particularly conservative,
politicized, and patriarchal variety practiced largely in Saudi Arabia. In her own
words, Hirsi Ali lived by the Book, for the Book.13 But after witnessing the frus-
tration and experiencing firsthand the misplaced anger of her mother, Hirsi Ali
came to believe that she had to escape Somali culture. Failing to do so, she feared
she would suffer her mothers fate. In Hirsi Alis estimation, her mother had sub-
limated her independent spirit to prevailing patriarchal and religious expecta-
tions, with the result that she became a devoted, well-trained work-animal.14
Had she remained, Ali remarks, I could never become an adult. Iwould always
be a minor, my decisions made for me. But Iwanted to become an individual,
with a life of my own.15
Thus, it came about that in 1992 at the age of twenty-two Hirsi Ali fled to the
Netherlands, claiming political asylum under the pretense that she was escaping
a forced marriage. (This claim was later admitted to be false.) In the aftermath of
September 11, 2001, she denounced Islam and converted to atheism. By 2002, she
was elected to the Dutch parliament on a wave of anti-immigrant sentiment.
Hirsi Ali has made it her mission to criticize Islamic customs. In particular
she has taken to task emigrant Muslim communities in Europe and the United
States for what she regards as their failure to embrace the secular political cus-
toms of their adopted nations. In 2004, Hirsi Ali wrote the screenplay and pro-
vided the voice-over for filmmaker Theo van Goghs made-for-television piece
Submission. The subject of the film was the treatment of women in Islamic cul-
ture. This was recounted by way of the tale of a Muslim woman coerced into a
violent marriage, raped by an uncle, and then viciously punished for adultery.16
13
Hirsi Ali, To Submit to the Book Is to Exist in Their Hell, Sydney Morning Herald, June 4, 2007,
http://www.smh.com.au/news/opinion/to-submit-to-the-book-is-to-exist-in-their-hell/2007/06/03/
1180809336515.html.
14
Johann Hari, My Life under a Fatwa, Independent, November 27, 2007, http://www.indepen-
dent.co.uk/news/people/profiles/ayaan-hirsi-ali-my-life-under-a-fatwa-760666.html.
15
Hari, My Life under a Fatwa.
16
Juxtaposed with passages from the Quran were scenes of actresses portraying Muslim women
suffering abuse. The film also features an actress dressed in a semi-transparent burqa who has texts
The film, which aired on Dutch television in August 2004, infuriated members
of the DutchMuslim community who viewed it as insensitive, profoundly
sacrilegious, and untruthful. On November 2, 2004, van Gogh was assassi-
nated by Mohammed Bouyeri, an Islamic radical and member of the Hofstad
Group.17 Impaled upon the knife stabbed into van Goghs body was a five-page
letter condemning Hirsi Ali to death as an apostate. After a fatwaa decree of
deathwas delivered against her in 2004, Hirsi Ali lived under armed guard.
In 2006, she moved to the United States, following uproar over the lies she had
told to obtain asylum, which grew to include a false name and birthdate. She
now holds a fellowship at the conservative American Enterprise Institute in
Washington,D.C.
Depending on your point of view, Hirsi Ali is a radical individual freedom
fighter, someone who has put her life on the line to defend women against
radical Islam,18Islam of a sort that presses for the institution of Islamic
law in government and civic lifeor she is a chameleon of a woman with a
talent for reinvention19 whose motives are to be regarded with some skepti-
cism. Notably, her refusal to distinguish mainstream Islam of a sort practiced
by roughly three million people in the United States from its politicized and
radicalized variants is empirically misguided and deeply troubling, not least
of all to feminist Muslim women.20 For instance, she charges that all of Islam
is full of misogyny.21 But whatever you think about Hirsi Aliand let me be
upfront about the fact that Iam not a fan of hersshe presents an absorbing
from the Quran written on her skin. The texts are among those often interpreted as justifying the sub-
jugation of women. Toby Sterling, Dutch Filmmaker Theo Van Gogh Murdered, Chicago Tribune,
November 2, 2004, http://metromix.chicagotribune.com; Gunman Kills Dutch Film Director, BBC
News, November 2, 2004, http://news.bbc.co.uk/2/hi/europe/3974179.stm.
17
This was an Islamic militant group of men of Dutch and Moroccan descent who met in The Hague
and who had been tried in Dutch appellate court on charges of fomenting terrorism. BBC News, Dutch
Jail Terror Group Muslims, March 10, 2006, http://news.bbc.co.uk/2/hi/europe/4793546.stm. For an
update in the case, see Court Orders Retrial of Dutch Terror Group, Radio Netherlands Worldwide,
February 2, 2010, http://www.rnw.nl/english/article/court-orders-retrial-dutch-terror-group; Wendy
Zeldin, Netherlands:Retrial of Hofstad Group, Library of Congress, February 22, 2010, http://www.
loc.gov/lawweb/servlet/lloc_news?disp3_l205401830_text.
18
Patt Morrison, Feminisms Freedom Fighter, Los Angeles Times, October 17, 2009, http://arti-
cles.latimes.com/2009/oct/17/opinion/oe-morrison17.
19
Economist, A Critic of Islam, February 8, 2007, http://www.economist.com/node/8663231.
20
Iranian American author Firoozeh Dumas is one such critic. http://firoozehdumas.com/ The
estimated number of Muslims really is rough. Statistical citations range from two to five billion. See
http://www.religioustolerance.org/isl_numb.htm, http://www.adherents.com/largecom/com_islam_
usa.html, and http://www.islam101.com/history/population2_usa.html.
21
In Bob Drogin, Book Review: Nomad by Ayaan Hirsi Ali, Los Angeles Times, June 7, 2010,
http://articles.latimes.com/2010/jun/07/entertainment/la-et-book-20100607, writes: Fired with the
zeal of a convert, Hirsi Ali insists Islam and the West are locked in a clash of civilizations...The dys-
functional Muslim family constitutes a real threat to the very fabric of Western life, and the Muslim
mind is in the grip of jihad.
picture of one womans determination to assert her autonomy against the most
resistant of forces.
Moreover, Hirsi Alis call for heightened scrutiny of political Islam does not
reflect simple paranoia or alarmism. In 1991, Sharia law, which is based on the
Quran, became authorized for use in arbitration over civil matters such as
property rights, divorce, and child custody in Ontario, Canada, ostensibly as a
way to lessen the load borne by the civil courts. Margaret Wente of the Globe
and Mail wrote of the decision:
The arbitrators can be imams, Muslim elders or lawyers. In theory, their
decisions arent supposed to conflict with Canadian civil law. But because
there is no third-party oversight, and no duty to report decisions, no out-
sider will ever know if they do. These decisions can be appealed to the reg-
ular courts. But for Muslim women, the pressures to abide by the precepts
of sharia are overwhelming. To reject sharia is, quite simply, to be a bad
Muslim.... Opponents of the new tribunals argue that the governments
imprimatur will give sharia law even greater legitimacy....What is called
sharia varies widely (in Nigeria, for example, it has been invoked to jus-
tify death by stoning). The one common denominator is that it is strongly
patriarchal.22
The decision to authorize sharia law has been hotly contested. Some of
the most vocal opponents are Muslim women who emigrated to Canada
and whose children are Canadian by birth. Two of these women are Alia
Hogben, who headed the pro-faith Canadian Council of Muslim Women,
and Homa Arjomand.23 Arjomand recounts the following story in support of
her concern:
I have a client from Pakistan who works for a bank, Ms. Arjomand tells
[Wendt]. Shes educated. She used to give all her money to her husband.
She had to beg him for money to buy a cup of coffee. Then she decided
to keep $50 a month for herself, but he said no. They took the matter to
an uncle, who decreed that because the wife had not been obedient, her
husband could stop sleeping with her. (This is a traditional penalty for dis-
obedient wives.) He could also acquire a temporary wife to take care of his
sexual needs, which he proceeded to do. Now the woman wants a separa-
tion. Shes fighting for custody of the children, which, according to sharia,
belong to the father.24
22
Margaret Wente, Life under Sharia, in Canada? Globe and Mail, May 29, 2004, http://www.
theglobeandmail.com/news/national/life-under-sharia-in-canada/article743980/.
23
http://www.nosharia.com/.
24
Wendt, Life under Sharia?
ShakilasStory
We move now to the third case, which was reported in February 2012 in the
NewYork Times. In 2010, Shakila, an Afghani Pashtun girl from rural Kunar
Province was abducted and enslaved for a year under a traditional Afghan
method of justice known as baad.25 Shakila was eight years old at the time.
Baad uses girls as compensation for moral and cultural transgressions per-
petrated by their relatives and tribal elders. In Shakilas case, the violation
occurred when one of her uncles ran off with the wife of a regional leader.
During the course of the year, Shakila was beaten, deprived of food, and con-
fined for most of the day. She managed, miraculously, to escape, but her ordeal
is hardly over. Fearing retribution, her family has fled their home, and they
now eke out a precarious living some distance from their village.
The objective of baad is to appease the anger of an injured family and
to partially reimburse them for whatever loss they suffered. In addition, by
forcing a girl into slavery, feuding families are integrated, thus minimizing the
prospect of sustained hostility. While baad is technically illegal, it is widely
accepted in rural Pashtun regions as a form of justice emanating from a Jirga,
the judicial court of tribal elders. Shakilas fathers response to her abduction
makes plain that he had no quarrel with the practice of baad. Rather, his ob-
jection was that Shakila had been promised in infancy in marriage to someone
else; she was thus someone elses property, not his to dispose of. His view is
commonplace even among members of the legislature. Fraidoon Mohmand, a
member of Parliament from Nangarhar Province, assesses the practice from a
pragmatic, utilitarian standpoint. He states that giving baad has good and bad
aspects. The bad aspect is that you punish an innocent human for someone
elses wrongdoings, and the good aspect is that you rescue two families, two
clans, from more bloodshed, death and misery. Like most men in this cul-
ture, Mohmand believes the good outweighs the bad; the girl given away may
suffer for a year or two, but when she brings one or two babies into the world,
everything will be forgotten and she will live as a normal member of the fam-
ily. His rosy assessment was challenged by the Afghan women interviewed
for the piece. One woman, Nasima Shafiqzada, who oversees womens matters
for Kunar Province states that the woman [or girl] given to a family in baad
will always be the miserable one. The girl not only represents the enemy but
also at a tender age is completely unprepared both for the brutality she will
encounter because of it and for the sexual relations often demanded ofher.
25
Alissa J. Rubin, For Punishment of Elders Misdeeds, Afghan Girl Pays the Price, New York
Times, February 16, 2012, http://www.nytimes.com/2012/02/17/world/asia/in-baad-afghan-girls-
are-penalized-for-elders-crimes.html?_r=1&ref=todayspaper. Baad is not an Islamic practice. The
Times reports that it predates Islam and can be traced to an era when nomadic tribes traveled
Afghanistans mountains and deserts.
The cases of Hirsi Ali and of Shakila discuss examples from Muslim culture.
The reader might well criticize them as being cheaply purchased by exploiting
Western liberal bias and of being bluntly inattentive to widely (and rightly)
discredited cultural imperialism. And indeed, one need exert only mild effort
to find cognate examples within certain cultural enclaves in Western societies
where non-Muslim culture does not predominate. Although low-wage labor
of the sort discussed in Section 2 is not a social practice rooted in patriarchy
or religion, it is widespread in the West and destabilizes autonomy dispropor-
tionately for women. It would seem obvious that feminists would take up the
campaign of low-wage laborers on the grounds that their economic vulner-
ability makes equal opportunities for self-governance (for women and men)
a struggle. It would seem obvious that feminists would take up the banner
of females in patriarchal societies on the grounds that they endure customs
that violate autonomy. Just as Hirsi Alis life in Kenya meant subjugation and
a loss of autonomy, Shakilas tale illustrates a socially entrenched practice that
unquestionably fails to respect the dignity and the autonomy of females. It
seems equally apparent, to my mind at any rate, that champions of autonomy
would decry the blatantly sexist (and thus arbitrary) basis on which female
low-wage workers are financially exploited. Similarly, Iwould expect propo-
nents of autonomy to denounce the manner in which girls in Shakilas situa-
tion are deprived of their rights of self-governance. Neither, however, is true.
Not all feminists have embraced the value of autonomy as readily as one might
suppose, nor have some proponents of autonomy unequivocally taken up the
banner of feminism. Let us explore why this is the case. Having done so, we
will be in a better position to see why both would be better off as allies.
4.Autonomy
Those of us in Western, liberal societies tend to construe autonomy, at a

minimum, as the desirable condition of being free to choose and to act ac-
cording to the direction of values and preferences that are ones own. At the
same time, most of us are drawn to the idea that autonomy can, in theory at
least, be realized through myriad ways of life and are wary of the idea that
only certain social arrangements befit self-governance. Some even go so far as
to claim (as does Andrea Westlund) that there is no contradiction between
self-responsibility and self-subordination.26 It is enough, Westlund argues,
that a person has freely accepted her situation and can answer for her commit-
ments in the face of critical challenges to these. The person who can do this will
count as self-responsible and so autonomous, even if these commitments are
26
Andrea Westlund, Rethinking Relational Autonomy, Hypatia 24:4 (2009):32.
self-subordinating and even if they manifest a lack of self-respect, assuming

the person has not internalized whatever social expectations might force this
commitment. Westlunds position is developed in an effort to accommodate
the belief that to respect autonomy is to permit persons to fashion and pursue
their own conceptions of thegood.
Though critical of Western feminism, the feminist philosopher Uma
Narayan argues for autonomy in a similar vein.27 Narayan contends that
Western feminists stereotype the lives of many Third World women, regarding
them either as prisoners of patriarchy, forced to accept oppressive lives, or as
compliant dupes of patriarchy, so dominated by the prevailing culture that
they have lost any capacity for critical, reflective, independent judgment. As
a result, Narayan advocates a fairly thin construction of autonomy along the
lines offered by Westlund.
Now, it would be easy to argue that the feminist commitments to womens
empowerment and to analyzing and resisting the circumstances of womens
subordination draw on some conception of autonomy as part of what such
commitments entail. I confront a more difficult taskor, rather, two tasks.
One task is that of establishing the claim that a feminist society must respect
autonomy as Ihave characterized it. That is, Ineed to establish that a commit-
ment to thick autonomy is necessary to achieve feminist goals. My view is
that such a commitment is in fact embraced far more readily than many femi-
nists and theorists of autonomy have allowed. Ihope to persuade the reader
that even those scholars who see themselves as advocating thin conceptions of
autonomy in fact advocate conceptions of autonomy that are thick, even while
they may not realize the thickness of their conceptions. In addition, Ineed to
show is that respecting autonomy means supporting feminist concerns. One
cannot claim to be a champion of autonomy at the same time one resists in a
general way feminist projects and concerns. The relation between autonomy
and feminism is one of mutual entailment.
The truth of neither assertionthat a commitment to feminism entails
a commitment to (thick) autonomy and that a commitment to autonomy
entails a commitment to feminismis obvious. Indeed, some have questioned
whether there is in fact a genuine disagreement between myself and feminists
who do not think autonomy entails feminism, and who do not even invoke
autonomy in articulating feminist concerns.28 One explanation for the putative
lack of disagreement is that the account of autonomy Iinvoke and the concep-
tion as it is deployed by feminist scholars are different conceptions. Perhaps
feminists work with so narrow an interpretation of autonomy that the thesis
27
Uma Narayan, Minds of Their Own: Choices, Autonomy, Cultural Practices, and Other
Women, in A Mind of Ones Own:Feminist Essays on Reason and Objectivity, 2nd ed., edited by Louise
M. Antony and Charlotte E. Witt (Boulder, CO:Westview Press, 2002), 418432.
28
Ithank Wilson Mendona for pressing thispoint.
that autonomy is related (in some way) to feminism will be utterly without
controversy. By contrast, the account of autonomy Iadvance is quite broad.
My position is that a person is autonomous when she has (1)the ability to su-
perintend those of her decisions, activities, and personal associations that are
central to human agency, (2)the warrant to do so, and (3)the power to act on
that ability. Together, these imply the presence of social roles and relations of a
particularly autonomy-friendly variety. Many might not interpret autonomy as
Ihave done; they may accuse my conception of being an interpretation of au-
tonomy that includes ideas of agency, self-sufficiency, individuality, empower-
ment, and human dignity. They might protest that autonomy as Ihave defined
it imports narrowly Western values or values that peremptorily designate the
good for any person, regardless of what the person professes her conception
of the good to be. In a nutshell, some would argue that we must attend to the
distinction between thin autonomy of a sort that may well be comfortable with
feminism and a thick or substantive species of autonomy of a sort that some
feminists dispute. Even if Iam correct that feminists must value the broader
conception of autonomy, feminists need not agree that this is what autonomy
is. It is certainly true, for example, that Westlund supports feminism, even if
she does not do so on the basis of a thin construction of autonomyalone.
But Ithink there is more agreement than there is disagreement regarding
the proper interpretation of autonomy for feminist concerns. Most feminists
do not advocate just a commitment to a legal or moral right to self-governance
but also a commitment to self-governance as an actual state of life. The ap-
pearance of disagreement might be traced to the belief that the conception
of autonomy I advance is normatively substantive in the sense proposed by
Natalie Stoljar and by Paul Benson. But in fact Iam not claiming, as Stoljar and
Benson have claimed, that the autonomous agent cannot internalize oppres-
sive norms or that she cannot value subordinate roles, although as a matter
of empirical fact doing so may be correlated with or give rise to diminished
autonomy.29 What I am claimingand here I suspect most feminists at any
rate agreeis that the practical control autonomy demands draws both from
sources internal to the agent and from external authority of a variety that man-
dates the absence of domination. And Iam claiming, further, that autonomy is
a matter of living in this way regardless of a persons preferences and subjective
interests. If there is convergence of opinion on this point, then there is less
daylight between the thicker, more socially demanding notion of autonomy
Iadvance and whatever account of autonomy is congenial to feminists.
For this reason, Ithink there is something perplexing about the view of au-
tonomy taken by Westlund and the view of feminism taken by Narayan. It is
29
My account is minimally substantive in that the autonomous agent cannot internalize oppressive
norms to such an extent that doing so critically undermines the possibility of effective practical control.
true that not everyone will include an autonomous life among the objectives
he or she esteems. It is also obvious (and Iwould say regrettable) that not eve-
ryone will embrace the concerns of feminism. But although neither Westlunds
construal of autonomy nor Narayans views about western feminism call for
legal rights to nondomination, and while both would grant that someone
could freely choose to be dominated, Isuspect neither Westlund nor Narayan
would claim autonomy was a feature of life for persons such as Shakila (or for
Hirsi Ali before she fled to the Netherlands). And Ithink they could deny that
Shakila and Hirsi Ali lacked autonomy because respect for feminism, along
with the practical resources feminism requires of society, was missing from
their lives.30 If we take seriously the goals of feminismnamely, advocating
for the equal rights of women and ending the oppression and exploitation of
women and girlsthen we cannot rest content with a thin variety of autonomy
premised on a persons free acceptance of and willingness to defend her cir-
cumstances. It really would not make any practical difference to Hirsi Ali if she
had accepted her life in Kenya or if Shakila had been a bit older and had agreed
to be used in baad. Baad violates autonomy owing to the patently chauvinist
foundation on which girls are divested of the power of self-governance, creat-
ing a situation in which some persons are in no way capable of freely choosing
their way of life. What this signals to my mind is that scholars such as Narayan
and Westlund may in fact construe autonomy rather thickly as a de facto or
actual state of life. Any situation where people live in common and where
some have power over others will only be congruent with feminist concerns if
thick autonomy is respected. It is one thing to acknowledge a variety of social
arrangements that comport with individual ideals and quite another to assert
that each of these arrangements will provide a life of autonomy. The three-step
assessment for autonomy serves as a filter for the satisfaction of feminist cri-
teria as gauged by the details provided by the assessment:(1)what areas of life
are insecure, (2)what is called for to diffuse the insecurity, and (3)what effect
does this have on the lives of the parties?
5. Autonomy and Human Development
The previous discussion suggests that a commitment to feminism involves a

commitment to thick autonomy. In this section, Iwould like to explore the
other side of the equation, namely, the idea that one who is committed to au-
tonomy must also be committed to feminist goals such as the empowerment
of women and girls. Let us begin by noting that respecting the autonomy of
30
The same could be said of some low-wage pink-collar workers whose employment conditions
are precarious.
women requires that we take seriously the centrality they accord interpersonal
relations in their lives. Aclassic feminist insight into the nature of womens
moral experience is that women are more likely than are men to value relational
interdependency; the moral perspective of women reflects this. Women (it is
argued) have a relational conception of moral ideals and appear less inclined
than do men to dismiss the emotions as a legitimate and essential component
of moral judgment. This does not mean that women care less about their own
powers of self-governance than do men or that the concept of autonomy must
demand less of the social and relational conditions women encounter for their
autonomy to be valued. What it means is that taking seriously any womans
autonomy involves honoring her interdependent, social personality.
What does this require? The capabilities approach to human and moral
development articulated by Martha Nussbaum and Amartya Sen offers an an-
swer to this question. The approach is congenial to the account of autonomy
Iadvance and offers a reason to treat thick autonomy as an essential element
of womens development.31 Let me say a bit more about why Ithink this is so.
The cornerstones of classical liberalism and, more specifically, of the idea of
liberal justice are the concepts of individual choice and liberty. Any liberal
ideology will begin with these two cornerstones and from these describe the
public and institutional arrangements best tailored to guarantee their pres-
ence in human life. Nussbaum argues, however, that we need to recast lib-
eral theory in a way that preserves these cornerstones while also capturing
what it means to be a flesh and blood human functioning in the real world.
To this end, liberalism would do well to begin with a conception of persons
as immersed in relationships of caring and dependency. These relationships
reflect the diverse intellectual, emotional, and bodily abilities persons possess.
What we must do, counsels Nussbaum, is start with a conception of the person
that corresponds to this human experiencethe person as both capable and
needyas opposed to a conception of persons as the impartial moral delib-
erators or self-interested contractors that populate the philosophical imagina-
tions of Kant, or Rawls, or contractarian theory.32 Having done so, we will see
that assessing who qualifies as well-off according to liberal justice calls for a
measurement that includes, among primary goods, the ability of persons to
take part in an extensive range of human endeavors. With a suitable list of the
central capabilities as primary goods, we can begin designing institutions by
asking what it would take to get citizens up to an acceptable level on all these
capabilities.33
31
Ithank Maria Silvia Possas for suggesting the connection between my conception of autonomy
and the liberal capabilities view of Nussbaum andSen.
32
Nussbaum, Future of Feminist Liberalism,79.
33
Nussbaum, Future of Feminist Liberalism,79.
Nussbaum is not prepared to say just what that level would involve, but she
does offer hints. With Sen, she defines capabilities as substantial freedoms
of the sort that include combinations of opportunities to function in various
ways. Importantly, capabilities are not just abilities residing inside a person
but also the [totality of] freedoms or opportunities created by a combination
of personal abilities and the political, social, and economic environment.34 For
instance, she would include among the basic goods the need for carewhich
is something that all of us, given our animality, will require at somepoint.
The NussbaumSen approach suggests that autonomy as Ihave character-
ized it is a capability, located among the list of primary goods founded in the
dignity of human need itself,35 such that any measurement of liberal justice
calls for a measurement that includes autonomy. The principal questions posed
by Nussbaum and Sen for the more comprehensive assessment of justice they
advocate are, What are people capable of doing? and How are people able
to live? Both Nussbaum and Sen note that what people seek is meaningful
lives for themselves; what Iwould add is that being self-governing is an ele-
ment of a meaningful life and, more generally, that people are most capable of
accomplishing this if they are self-governing in a robust way.36 To put the point
somewhat differently, autonomy is, borrowing from terminology employed by
Jonathan Wolff and Avner de-Shalit, a particularly fertile capability, that is,
a capability that is essential to removing corrosive disadvantages of the sort
that undermine other dimensions of a persons experience that spell out how
the person is able to livedimensions such as health, well-being, social re-
spect.37 Autonomy lends security to these other capabilities, promising that
one can bank on their presence in the future. And this type of security is an
objective matter...for each capability we must ask how far it has been pro-
tected from the whims of the market or from power politics.38
If my appropriation of elements of the capabilities approach in defense of
robust autonomy survives inspection, then the strain of liberalism behind
which some feminists and autonomy theorists seek the protective shelter of
neutrality is not nearly as neutral as might be hoped with respect to the social
34
35
Paraphrased from Nussbaum, Future of Feminist Liberalism,79.
36
Some people might quarrel with the idea that a life bereft of autonomy is a meaningless life.
Let me be clear that Iam not claiming that meaning can only be found in lives that are rich with au-
tonomy. In other work, Ihave taken pains to show that autonomy is not the sole good and that there
are other measures of a meaningful life, such as flourishing or satisfaction. Oshana, Personal Autonomy
in Society. Nonetheless, Ithink ample evidence can be marshaled in support of the claim that where
people do lead meaningless lives, the reason for the meaninglessness can be a lack of opportunities
for growth, development and relationship, which can be due to oppressive social relations (Andrea
Veltman in editorial commentary). Arobustly autonomous person avoids these circumstances.
37
Jonathan Wolff and Avner de-Shalit, Disadvantage (New York: Oxford University Press, June
2007). Their work is discussed in Nussbaum, Creating Capabilities,44.
38
goods and capabilities all persons are assumed to need. Autonomy is in some
measure constitutive of a meaningful life for persons. And, in truth, other phi-
losophers whose scholarship often addresses explicitly feminist concerns but
who publically have challenged (indeed, disavowed) portions of the account
Idefend nonetheless speak of the conditions needed to realize a feminist soci-
ety in terms that echo the capabilities approach. Marilyn Friedman, for in-
stance, argues that in order to challenge the formidable structures of male
domination that remain in this world, it is crucial for feminism to insist that
mature women are as capable as men of being full moral agents in their own
rights and should no more be dominated or controlled over the course of their
lives than men are.39 Being as capable as men entails having a commensurate
array of opportunities and a network of support at the ready, as championed
by the capabilities approach.
We might press this point from a different direction. Aright of autonomy
is found in the moral idea that persons have an inherent dignity in virtue
of which they are entitled to a certain standing that constrains both how
persons are to be treated and what they are permitted to do and expected
to do vis vis one another. Ithink it is clear that if many of the rights that
we indisputably have are to be respected, then autonomy must be respected
as well. And for autonomy to be respected, the social institutions that per-
meate our affairsand this includes institutions of law, whether they are
found in the United States or in Afghanistanmust afford us de facto and
de jure power to counter attempts to intrude capriciously in our lives. This
requirement is especially pressing in cases such as Shakilas, where attempts
to deprive a person of her rights are rampant. No person who is in a society,
and particularly those who are most vulnerable, can determine how she
shall live unless she is part of enduring social relations and shared traditions
that afford her authority over the most fundamental areas of her life, in law
and in practice, and without undue cost. If we are committed to something
more than the legal promise of autonomy and if we are committed to more
than a moral right of self-governanceif, that is, we are committed to an
idea of autonomy as a practical condition in which a person is empow-
ered to choose what to do by means of her own authoritythen it seems
straightforward that a commitment to autonomy entails a commitment to
feminism as a moral and a political ideal. This is certainly a conclusion to
which we are led if we follow Nussbaum and Sen in drawing on a richer
variety of social goods deemed essential to the promises of liberal justice.
39
Friedman, Feminism in Ethics, 216. Emphasisadded.
6. Some Objections
Let us turn to some objections that my account might face. To begin, what
would be the consequences for my account if a commitment to autonomy
entailed a commitment to a social structure that is de facto supportive of
womens equality and equal rights while lagging in a moral commitment to
the equality of women? A number of modern affluent societies might meet
this description, including, perhaps, the United States. Indeed, this possibility
is reminiscent of the criticism some feminists have registered against classical
liberalism as an adequate political philosophy, one attentive to the status of
women.40 Mark Piper worries that this would be enough to undermine the
notion that a commitment to thick autonomy entails a commitment to the
moral aspects of feminism dealing with a commitment to the equal value of
women.41 Iam not sure Ishare Pipers worry. It may be that such societies are
only committed to autonomy of a thinner variety. In the United States there is
certainly ample lip service to the moral dimensions of feminism; it is in the lack
of a tangible commitment to the institutional and practical dimensions of fem-
inismwhere this would call for a commitment to thick autonomythat the
true nature of this commitment is revealed. Such societies do not really sup-
port feminist goals in a functional wayif they did, the United States would
have passed an equal pay for equal work amendment long ago. The failure to
do so belies not just a commitment to the moral claim that men and women
have equal value but also a commitment to thick autonomy. It may also be the
case that a commitment to thick autonomy is incumbent on a commitment
to de facto support of equality but less so to the moral dimensions of equal-
ity. Perhaps the moral commitment is less urgent than is the commitment to
practical equality. Imyself think that de facto support is unlikely absent an at-
tendant moral commitment and that a commitment to moral equality in itself
is cold comfort for one whose autonomy is elusive or denied altogether infact.
A second worry for the account Idefend was raised by John Christman in
conversation. Christman suggested that it is an implication of my position
that only feminists can be autonomous. It would certainly seem a stretch to
deny that nonfeminists can be autonomous, and if the view Iwas defending
yielded this result the account would be less plausible. Fortunately, this is
not an implication of my account. That a person manages to live according
to ideals and allegiances that are recognizable as her own, without trouble-
some interference, is a real possibility even for people who repudiate any
40
See, e.g., Susan Moller Okin, Justice, Gender, and the Family (New York: Basic Books, 1989);
Catherine Mackinnon, Reflections on Sex Equality under the Law, Yale Law Journal 100:5 (March
1991): 12811328; Mackinnon, Feminism Unmodified: Discourses on Life and Law (Cambridge,
MA:Harvard University Press,1987).
41
In private commentary.
allegiance to feminist ideology. The would-be surrendered woman whose

desire to give herself over to the direction of her husband is frustrated given
the plainly self-governed life she leads is a case in point. In such cases, the re-
pudiation of feminism is accompanied by a repudiation of robust autonomy
as a desirable condition of life. But rejecting an idea will not amount to
rejecting the state that the idea captures or to divesting oneself of that state.
Ican reject the ideals of feminism even while (and perhaps because) Iam
autonomous.
A third complaint may be pressed by feminists who urge a more commu-
nitarian than classical liberal political ideology and who allege that personal
autonomy is less desirable than Ihave portrayed it, certainly less desirable
than Western liberalism would have us believe. They argue that autonomy is
a value opposed to social virtues such as caring, empathy, loyalty, regard
for and commitment to others and as antithetical to cooperative projects.42
These social virtues characterize the lives of women. By contrast, autono-
mous agents are said to be excessively self-absorbed, self-created individu-
als, dependent on none but themselves, bound by no limitations other than
those they levy on themselves. But this style of life is drawn from the expe-
rience of men and neglects what is of importance to women from a womans
perspective.
This is not, however, a sensible depiction of autonomy. The problem with
this picture does not reside in the idea that an autonomous individual is,
loosely speaking, sovereign over her decisions and actions. The worrisome as-
pect of this story is that it caricatures autonomy, presenting it as a value that
commends a form of individualism that is hostile to social goods and rela-
tionships. To portray persons as radically independent individuals ignores the
social dimension of our lives and discounts the importance of interpersonal
relationships. But we all know that a theory of autonomy demanding radical
independence of this variety would be implausible and undesirable. None of
us is self-created; each of us is shaped by his environment, just as each of us
if we are lucky, at any rateis supported by interpersonal relations. (Oddly,
Isuspect it is this confused caricature of autonomy that springs to the minds
of some persons when they recoil from the label feminist. They equate au-
tonomy independence with denigration of traditional ties. The irony is that the
caricature is one that has gained some traction among some feminist critics
of autonomy.) In any case, feminists who interpret autonomy in this fashion
commit a mistake, as do antifeminists whose rejection of feminism arises from
this misleading interpretation of autonomy.
42
The discussion that follows is from Marina Oshana, The Autonomy Bogeyman, Journal of
Value Inquiry 35:2 (June 2001):209226.
7.Conclusion
If we accept what I have said about autonomy, then I think we can expect
feminism of at least a modest variety. The point is not that in some cases the
concerns of feminism and of autonomy go hand-in-hand. The point is that
respecting autonomy means respecting feminist concerns and vice versa.
Proponents of autonomy must unequivocally uphold feminism. And being
committed to feminist concerns leads one more closely to a robust relational
account of autonomy than might initially seem the case. Recall my initial def-
inition of feminism as a belief in and advocacy of equal rights for women
based on the idea of the equality of the sexes. People who are autonomous
have a characteristic type of social standing. This is the social standing people
have when they live in the presence of other people and when, by virtue of
social design, none of those other dominates them.43 This characterization is
consistent with the initial definition Ioffered, namely, that a person is autono-
mous when she makes decisions and engages in actions by means of her own
authority. Iwould also suggest that it is a definition of autonomy that ought to
be accepted by feminists, even those who worry about an excessively individ-
ualistic conception of thatstate.
To avoid misunderstanding, let me state clearly that the point I wish to
make is that one cannot be committed to autonomy unless one is also com-
mitted to feminism and, conversely, that one cannot be committed to femi-
nism unless one is also committed to autonomy. This does not mean that any
society in which autonomy happens to be realized by its members will be a
society committed to feminism. Asociety may be inconsistent in upholding
those goals to which it professes commitment or may fail to give sufficient
weight to these goals. What Ihave attempted to do in this paper is show why,
to the extent that autonomy is prized, feminism ought to be prized as well
and vice versa. Thus, despite criticisms of certain interpretations of autonomy
from feminists and despite the idea that under certain conditions autonomy
and oppression may be compatible, I have uncovered no discussion in the
philosophical literature or in feminist commentary that would cast serious
doubt on the idea that a commitment to autonomy is a commitment to fem-
inism and vice versa. Despite the interminable debate over thick and thin
conceptions of autonomy, my position is that most theorists of autonomy are
not as thin as theythink.
43
Philip Pettit, Republicanism:ATheory of Freedom and Government (NewYork:Oxford University
Press, 1997),122.
Acknowledgments
Earlier versions of this paper were presented at a workshop on Relational

Autonomy organized by the Department of Philosophy at McGill University
in Montral, September 2012, at the Federal University of Rio de Janeiro, June
2012, and at the Center for Practical and Professional Ethics, California State
University, Sacramento, February 2012. Ithank the members of the workshop
and the audience on these occasions for their helpful remarks.
{ PartIII }
Autonomy, Reasons, andCare

{8}
Emotions, Reasons, and Autonomy

Christine Tappolet
As Alison Jaggar notes in her paper Love and Knowledge:Emotion in Feminist

Epistemology, not only has reason been contrasted with emotion, but it has
also been associated with the mental, the cultural, the universal, the public
and the male, whereas emotions have been associated with the irrational, the
physical, the natural, the particular, and of course, the female.1 If we add the
quite common thought that autonomy and reason are closely associated, the
question can arise whether women are even capable of autonomy.
This train of thought can be expressed in the form of what is likely to seem
a rather ludicrous line of thought.
The Ludicrous Argument:
1. Women are emotional.
2. Emotional agents cannot be autonomous.
Therefore, women cannot be autonomous.
Even though there is surely no lack of proponents of this view in the history of
ideas, it is clear that few contemporary thinkers would be inclined to embrace
such an argument, which is likely to strike most of us as both repugnant and
factually unsubstantiated.2
One notable complication comes from feminist criticisms, ac-
cording to which autonomy is a dubious masculinist ideal. Lorraine Code
writes: Autonomous man isand should beself-sufficient, independent,
and self-reliant, a self-realizing individual who directs his effort towards maxi-
mizing his personal gains....In short, there has been a gradual alignment of
autonomy with individualism.3 On such a view, which equates autonomy with
1
Alison Jaggar, Love and Knowledge:Emotion in Feminist Epistemology, Inquiry 32 (1989):166.
2
Diana Tietjens Meyers, Personal Autonomy and the Paradox of Feminine Socialization, Journal
of Philosophy 84 (1987):621.
3
Lorraine Code, Second Persons, in What Can She Know? Feminist Theory and the Construction
of Knowledge, edited by Lorraine Code (Ithaca, NY:Cornell University Press, 1991),78.
self-centered individualism, that autonomy might not be an option for women

and other beings who undergo emotions should appear welcome.
However, as many feminists have convincingly argued, autonomy need not
be conceived in terms of such a dubious ideal. Autonomy can, and indeed has
been, reconceptualised from a feminist perspective.4 What lies at the heart of
the feminist reconfiguration of the concept of autonomy is the undeniable fact
that agents are socially embedded. This is the thought that the label relational
autonomy aims at capturing. Whatever the details this new conception of au-
tonomy is taken to have, it holds the promise of an attractive ideal, which few
would want to forsake. This reconfigured conception of autonomy thus gives
good reasons to resist the Ludicrous Argument.
The question that arises is how exactly the Ludicrous Argument should be
resisted. One route is to argue that even though there is a grain of truth in the
hackneyed association of emotions with women, it would be a mistake to infer
that women cannot be autonomous. Another route is to claim that there are no
important gender differences at the affective level in the sense that all normal
human beings undergo emotions, while denying that there is any incompat-
ibility between being emotional and being autonomous. Whichever route is
preferred, the crucial question is that of the relation between emotions and au-
tonomy.5 This is the question Iwant to focus on in this paper, thereby leaving
aside the complex empirical issue of gender differences regarding emotions.6
It will be useful to begin by saying a few words about the concept of au-
tonomy. In the sense of this term at stake here, to be autonomous is for an
agent to steer her own course through life rather than being the mere play-
thing of winds and currents. The thought is that an agent is autonomous when
she governs herself. Thus, what can be considered her authentic self and not
some foreign force, be it external (such as an abusive husband or degrading
socialization) or internal (such as a compulsive desire), directs her thoughts
and actions. One prominent idea is that personal autonomy requires critical
4
See Jennifer Nedelsky, Reconceiving Autonomy: Sources, Thoughts and Possibilities,
Yale Journal of Law and Feminism 1 (1989): 736; Marilyn Friedman, Autonomy and the Social
Relationships:Rethinking the Feminist Critique, in Feminists Rethink the Self, edited by Diana Tietjens
Meyers (Boulder, CO: Westview Press, 1997); Catriona Mackenzie and Natalie Stoljar, Relational
Autonomy:Feminist Perspectives on Autonomy, Agency, and the Social Self (NewYork:Oxford University
Press, 2000); Marina Oshana, Personal Autonomy in Society (Aldershot, UK:Ashgate,2006).
5
Interestingly, this is a question that has been on the agenda of relational autonomy theorists.
See Mackenzie and Stoljar, Introduction, Autonomy Refigured, in Mackenzie and Stoljar, Relational
Autonomy,2122.
6
One difficulty comes from the fact that there could be gender differences concerning (1)the kinds
of emotions that tend to be felt, (2)what tends to trigger emotions of the same kind, (3)the different
aspects of emotions, such as expression and motivation, and (4)the regulation of emotions. Another
difficulty is that generalisation of this kind are in danger of downplaying individual as well as socio-
cultural differences.
Emotions, Reasons, and Autonomy 165
reflection. It is insofar as the agent subjects her motivations and action-guiding

principles to critical reflection that she can be taken to be autonomous.7
This notion of autonomy is thought by many to be closely related to that
of agency, or at least to the kind of agency, sometimes called full-fledged or
full-blooded, which is often considered to be characteristic of human beings.8
In the introduction to his collected papers, Gary Watson asks:What makes
us agentsthat is, individuals whose lives are attributable to them as some-
thing they (in part) conduct, not just something that occurs?9 To answer this
question, Watson appeals to the notion of a capacity for critical evaluation.
As he puts it, we are agents because (and insofar as) we shape our lives by the
exercise of normative intelligence, which according to him involves a capacity
to respond relevantly to reasons.10 On the assumption that autonomy can be
spelled out in terms of a capacity for critical evaluation that involves reason-
responsiveness, the question of the relation between emotions and autonomy
becomes, at least partly, that of the relation between emotions and the capacity
to respond to reasons.
Reason-responsiveness accounts of autonomy, which can be considered
broadly Platonic, are not the only accounts of autonomy on the market.
Drawing on the Humean tradition, some accounts emphasize the importance
of the motivational cum affective in autonomy. The foremost example of this
kind of accounts is Harry Frankfurts hierarchical model of free will and au-
tonomy, which is exclusively framed in terms of conative states of different or-
ders.11 Building on Frankfurts model, David Shoemaker proposes an account
that takes emotions to be essential to autonomy. According to Shoemaker, au-
tonomy involves acting on ones cares, understood not as motivational states
but as emotional dispositions. What he claims, thus, is that the emotions we
have make us the agents we are.12 The idea is that the emotions we feel are inti-
mately related to who we are, for they reveal what we deeply care for. Emotions,
rather than potentially distorting reflection, would be essential to authenticity
and thus also to autonomy. By contrast with reason-responsiveness accounts,
7
See, e.g., Marilyn Friedman, Autonomy and the Split-Level Self, Southern Journal of Philosophy
24 (1986):1935.
8
For the term full-blooded, see J.David Velleman, What Happens When Someone Acts? Mind
101 (1992):462.
9
Gary Watson, Introduction, in Agency and Answerability:Selected Essays (NewYork:Oxford
10
Watson, Introduction,2.
11
See, e.g., Harry Frankfurt, Freedom of the Will and the Concept of a Person, Journal of
Philosophy 68 (1971): 520. Note that insofar as Frankfurt suggests that reflection is the source of
higher-order volitions, the contrast is less stark. See, e.g., Frankfurt, Freedom of the Will and the
Concept of a Person,7.
12
David Shoemaker, Caring, Identification, and Agency, Ethics 114 (2003): 94; see Tappolet,
Autonomy and the Emotions, European Journal of Analytic Philosophy:Special Issue on Emotions and
Rationality in Moral Philosophy 2 (2006):4559.
an account in terms of care might appear to be more congenial to a feminist

rejection of the Ludicrous Argument. However, Ishall argue that it would be
wrong to believe that the two accounts are fundamentally opposed.
My plan is the following. In the first section, Ipresent an account of emo-
tions according to which emotions are a kind of perceptual experience and
suggest that emotions can grant us a privileged access to values and reasons.
In the second section, I turn to the relation between emotions and reason-
responsiveness. In particular, Ifocus on cases of akrasia in which emotions
conflict with the agents better judgment and argue that even in such cases we
are able not only to track reasons but also to be reason-responsive when we
act on our emotions. Put simply, what Ipropose is that reason-responsiveness
requires the exercise of agential virtues. In the last section, I examine the
implications of this account of emotions for both reason-responsiveness and
care-based accounts of autonomy. The first point Ispell out is that emotions
and autonomy need not be at odds. The second point is that a better under-
standing of emotions reveals that the two approaches are not as opposed as
they might at first appear.
1. Emotions and the Perception ofValues
Emotions make for a complicated field of inquiry. A first source of compli-

cation is that emotional, and more generally affective, phenomena are highly
variegated. It is thus common to distinguish between actual episodes of emo-
tions, such as the fear you feel when you realize that someone is following
you on your way home, and the disposition to undergo a range of emotional
experiences, such as arachnophobia. Asecond source of complication is that
emotions are naturally taken to be divided into different kinds. We distinguish
not only among fear, disgust, and anger but also among pride, joy and hope,
to name only but a few. Finally, a third source of complication is that emo-
tions are complex phenomena. Atypical episode of fear at someone who is
following you involves a number of different elements. You hear someone fol-
lowing you, you appraise the situation as a threat, you undergo physiological
changes such as an increased heart rate, your face will adopt a typical expres-
sion; in addition, you will feel a specific pang. Your attention will focus on the
supposed threat, your mind will be filled with thoughts related to the situation
and how to escape it, and you will be motivated to take up a specific course
of action, if only to walk as fast as you can. Such complexity makes it difficult
to develop a satisfactory theory of emotions. Acentral question any theory of
emotion has to address is which elements, if any, are essential to emotions.13
13
The best introduction to emotions theory Iam aware of is Julien Deonna and Fabrice Teroni, The
Emotions:APhilosophical Introduction (NewYork:Routledge,2012).
To introduce the account Ifind most plausible, let me consider an example

offered by Marilyn Friedman of a woman who has been socialized to believe
that a womans place is in the home and who has thus embraced principles that
prevent her from critically questioning her way of life but who experiences
emotions that conflict with her overall assessment of her life. As Friedman
notes, her frustration, grief, and depression...may be her only reliable
guides to a more autonomous path.14 More generally, Friedman suggests that
emotions or more generally affective states constitute touchstones of a sort for
the assessment of the adequacy of ones principles.15 The claim that affective
states in general and emotions in particular constitute what can be considered
to be epistemic touchstones in the assessment of our normative judgments is
grounded in a conception of emotions, according to which emotions have two
key characteristics:(1)emotions have cognitive content; and (2)emotions can
conflict with conceptually articulated states, such as beliefs and judgments.
As recent debates in emotion theory make clear, these two features are cen-
tral to an account of emotions according to which emotions are, in essence, a
kind of perceptual experience.16 What is specific to these perceptual experi-
ences is that they represent their object in evaluative terms.17 When appro-
priate, fear would thus consist in the perception of something as fearsome,
and the same would hold for disgust and the disgusting, admiration and the
admirable, and so forth. On what Itake to be the most plausible version of
such an account, emotions have representational, albeit not conceptually artic-
ulated, content. Emotions represent their object as having specific evaluative
14
Friedman, Autonomy and the Split-Level Self, 3031. Cf. Alison Jaggar, Love and
Knowledge:Emotion in Feminist Epistemology; Karen Jones, Emotion, Weakness of Will, and the
Normative Conception of Agency, in Philosophy and the Emotions, edited by Anthony Hatzimoyis
15
Friedman, Autonomy and the Split-Level Self,3031.
16
See Alexius Meinong, Ueber Emotionale Prsentation, Kaiserliche Akademie der Wissenschaft
in Wien 183, Part 2, On Emotional Presentation, translated with an introduction by Marie-Luise
Schubert Kalsi (Evanston, IL: Northwestern University Press, 1917), 1181; Ronald de Sousa, The
Rationality of Emotion (Cambridge, MA:MIT Press, 1987); de Sousa, Emotional Truth, Proceedings
of the Aristotelian Society 76 (2002):247263; Christine Tappolet, Les motions et les concepts axi-
ologiques, in La Couleur des penses, Raisons Pratiques, edited by Patricia Paperman and Ruwen Ogien,
(Ville:dition, 1995), 237257; Tappolet, motions et Valeurs (Paris:Presses Universitaires de France,
2000); Tappolet, Emotions, Perceptions, and Emotional Illusions, in Perceptual Illusions:Philosophical
and Psychological Essays (Palgrave-Macmillan, 2012); Sabine Dring, Explaining Action by Emotion,
Philosophical Quarterly 53 (2003): 214230; Dring, Seeing What to Do: Affective Perception and
Rational Motivation, Dialectica 61 (2007):361394; Jesse Prinz, Gut Reactions:APerceptual Theory of
Emotion (NewYork:Oxford University Press, 2004); Prinz, Is Emotion a Form of Perception? The
Modularity of Emotions, Canadian Journal of Philosophy 32 (2008):137160; Julien Deonna, Emotion,
Perception and Perspective, Dialectica 60 (2006):2946.
17
The account is often presented in terms of perceptions instead of perceptual experiences, but
because emotions can misfire, the perceptual account claims only that emotions are the perception
of evaluative properties unless defeated. By contrast with perceptions, perceptual experiences are not
factive. If you perceive that the cat is black, then it is black, but you can of course have the perceptual
experience of a gray cat asblack.
properties, as fearsome or disgusting, even though the agent who undergoes

the emotion need not possess the relevant evaluative concepts (e.g., fearsome,
disgusting).18 Thus, an emotion of fear with respect to a dog will be appropriate
just in case the dog is really fearsome.
What makes what can be called the Perceptual Theory of Emotions attrac-
tive is that it steers a middle course between two opposed accounts of emo-
tions, each of which has some plausibility but both of which are ultimately
unsatisfactory. At one end of the spectrum, there is the so-called Feeling
Theory, according to which emotions consist in states, such as bodily sensa-
tions, that are characterized by the way they feel but that have no representa-
tional content.19 At the other end of the spectrum lies the Judgmental Theory,
according to which emotions are or necessarily involve conceptually articu-
lated judgments so that to fear something would amount to judge that the
thing in question is fearsome.20 The main objections to the Feeling Theory are
that it cannot take into account that emotions have intentional objects (e.g.,
we are afraid of a dog, angry at someone) and that it fails to make room for the
fact that we assess emotions in terms of how they fit their object, such as when
we say that it is inappropriate to feel fear at an innocuous spider. Apart from
the fact that it does a poor job at accounting for the fact that emotions are felt
states, what plagues the Judgmental Theory is that it is incompatible with the
observation that one can undergo an emotion without possessing the relevant
conceptsone can be afraid of something without possessing the concept of
fearsomeness, for instance.21
By contrast, the Perceptual Theory can readily acknowledge that emotions
are felt states, in the sense that there is something it is like to undergo an emo-
tion of fear or of admiration. But the Perceptual Theory is in a better position
than the Feeling Theory in that it can make room for the fact that emotions
have intentional objects and allows for cognitive assessment in terms of how
well emotions fit their object. At the same time, since the content of emotions
is not taken to be conceptually articulated, the Perceptual Theory is not com-
mitted to the claim that emotions require concept possession.
18
For concepts and nonconceptual contents, see, inter alia, Gareth Evans, Varieties of References
(Oxford: Clarendon, 1982); Christopher Peacocke, A Study of Concept (Cambridge, MA: MIT
Press,1992).
19
See William James, What Is an Emotion? Mind 9 (1884): 188204; Carl G. Lange, Om
Sindsbevaegelser:Et Psyki-fysiologisk Studie (Kjbenhavn:Jacob Lunds, 1885), translated in The Emotions,
edited by C. G. Lange and William James (Baltimore:Williams and Wilkins,1922).
20
See Robert C. Solomon, The Passions (Indianapolis:Hackett Publishing Company, 1976); Martha
Nussbaum, Upheavals of Thought: The Intelligence of Emotions (Cambridge: Cambridge University
Press,2001).
21
The same problem arises for so-called Judgmental Theories, according to which the cognitive
states are taken to be thoughts (Patricia S. Greenspan, Emotions and Reasons (NewYork:Routledge
and Kegan Paul, 1988)) or construals (Robert C. Roberts, Emotions:An Essay in Aid of Moral Psychology
(Cambridge:Cambridge University Press, 2003)).
The main argument for the perceptual account is an argument by analogy.

It is based on the observation that emotions and sensory perceptions, such as
visual or tactile experiences, share a number of important features.22 Like sen-
sory perceptions, emotions appear to have a characteristic phenomenology;
both emotions and sensory perceptions are in general caused by things in our
environment; both fail to be directly subject to the will; both appear to have
correctness conditions, in the sense that they can be assessed in terms of how
they fit the world; and finally both can conflict with judgment. As is made clear
in the Mller-Lyer illusion, we can see two lines as being of different lengths
while judging they are of the same length, and we can feel fear at something
that we judge to be harmless.23 As will be clear, these last two features are the
ones that are central in an explanation of the role of emotions as epistemic
touchstones spelled out by Friedman.
Of course, there are also a number of differences between emotions and
sensory perceptual experiences.24 For instance, there are no sensory organs
underlying emotions, and unlike sensory perceptions emotions themselves
rely on what are often called cognitive basesyou need to see or hear, or else
remember something, to be afraid of it. Even though there is no space to argue
for this claim here, Ibelieve that taking these points into account makes for a
more nuanced picture of emotions than the analogies with sensory perceptions
initially suggest, but they do not impugn the core of the perceptual account.25
To better to understand what the Perceptual Theory entails with respect
to the epistemic role of emotions, consider again an episode of fear. If it is
true that fear is, for example, the perception of a dog as fearsome, it will
also be true that fear will inform an agent about what she has normative
reasons to do.26 For hardly anyone would deny that somethings being fear-
some is a pro tanto reason to do or refrain from doing certain things. This
22
See Meinong, On Emotional Presentation; Tappolet, motions et Valeurs, chap. 6; Prinz, Gut
Reactions. A Perceptual Theory of Emotion, chap. 10; Prinz, Is Emotion a Form of Perception?;
Deonna and Teroni, Emotions, chap.6; Michael S. Brady, Emotions, Perception and Understanding,
in Morality and the Emotions, edited by Carla Bagnoli ( New York: Oxford University Press, 2001),
chap.2.
23
See, inter alia, Tappolet, Emotions, Perceptions, and Emotional Illusions, in Perceptual
Illusions: Philosophical and Psychological Essays, edited by Clotilde Calabi (Houndsmill,
UK:Palgrave-Macmillan,2012).
24
See de Sousa, Rationality of Emotion; Bennett Helm, Emotional Reason:Deliberation, Motivation,
and the Nature of Value (Cambridge:Cambridge University Press, 2001); Mikko Salmela Salmela, Can
Emotions e be Modelled on Perceptions? Dialectica 65 (2011):129; Deonna and Teroni, Emotions;
Jrme Dokic and Stphane Lemaire, Are Emotions Perceptions of Values, Canadian Journal of
Philosophy 43 (2013):227247; Brady, Emotions, Perception and Understanding.
25
See Tappolet, Emotions, Perceptions, and Emotional Illusions; Tappolet, Emotions, Values, and
Agency (Oxford:Oxford University Press, forthcoming).
26
Iunderstand normative reasons to be considerations that speak in favor of beliefs or actions.
See Thomas Scanlon, What We Owe to Each Other (Cambridge, MA:Harvard University Press,1998).
is particularly difficult to deny if one assumes, as appears plausible, that

fearsomeness supervenes on dangerousness.27 That the dog is fearsome and
thus dangerous is plausibly seen to be a consideration that speaks in favor
of doing something to avoid the threat, at least under the assumption that
the agent desires to avoid physical harm. What is important to emphasize
is that the emotion informs the agent of the fearsomeness and danger that
threatens her independently of what she judges. Thus, she might mistakenly
judge that the dog is a harmless puppy. There would thus be a conflict be-
tween what she feels and what she judges. What is noteworthy is that in such
a case, by contrast to the cases that first come to mind, it is the judgment
and not the fear that gets things wrong. As a result, acting on the emotion
would be more advisable than acting on the judgment. It would amount to
acting on correct information about the reason the agent has for acting in a
certainway.
As this example makes clear, what follows from the Perceptual Theory is
that insofar as emotions are perceptions of evaluative properties, and thus
of the practical reasons that depend on evaluative properties, emotions pro-
vide privileged epistemic access to practical reasons. To use Karen Joness apt
phrase, emotions can key us to real and important considerations that speak
in favor of acting in certain ways, without always presenting that information
to us in a way susceptible of conscious articulation.28 That emotions have this
role explains why they can constitute touchstones against which to test our
action-guiding principles. There might be a debate about how to understand
the nature of deliberation, but whatever the details it follows from the account
of emotions proposed here that they may have a crucial role to play in the as-
sessment of our normative judgments.29
Let us accept, then, that emotions can afford us a privileged access to eval-
uative properties and practical reasons. As such, this makes emotions indis-
pensable in an account of agency. This is not a small feat, but at the same time
one might wonder whether this claim does full justice to emotions and their
role in agency. An important question is whether being moved to action by an
emotion, be it fear or compassion, for example, amounts to acting for a reason,
or in the light of a reason. In other words, the issue is how emotions relate to
reason-responsiveness.
27
For the claim that the fearsome supervenes on the dangerous, see de Sousa, Rationality of
Emotion.
28
See Jones, Emotion, Weakness of Will,181.
29
The main options are a coherentist model according to which emotions and normative prin-
ciples have to be brought into reflective equilibrium (see Friedman, Autonomy and the Split-Level
Self ) and a foundationalist conception, which grant emotions a defeasible but foundational justifica-
tory power (see Tappolet, motions et Valeurs).
2. Emotions and Reason-Responsiveness
In terms of Joness useful distinction between reason-tracking and reason-

responding, the question is whether emotions merely allow us to track reasons
to register reasons so that the agent can behave in accordance with themor
whether acting on the basis of an emotion may involve responding to a reason.
Jones summarizes what is required to be a reason-responder, which she char-
acterizes as someone who responds to reasons as reasons. Jones explains, To
be able to respond to reasons as reasons, an agent requires critical reflective
ability, dispositions to bring that ability to bear when needed, and dispositions
to have the results of such reflection control [her] behaviour.30 As we saw in
the introduction, the ability to respond to reasons is often considered to be
essential to full-fledged agency and hence of autonomy. Accordingly, the im-
portant question is whether emotions merely allow for the tracking of reasons
or whether we can be genuine agents who are guided by reasons when we are
moved by our emotions.
The majority view is that our emotions do not, as such, allow us to respond
to reasons.31 When we undergo an emotion, we, or maybe more accurately
some mechanism in us, might at best be tracking reasons, but we do not re-
spond to reasons in virtue of our emotions. What is taken to be necessary for
responding to reasons is deliberation (i.e., reflection about practical reasons),
and this, it is thought, requires judgments about the agents practical reasons.
Such reflective responsiveness is explicitly related to full-fledged agency. Thus,
Jay Wallace writes:
The motivation and actions of rational agents are guided by and responsive
to their deliberative reflection about what they have reason to do. Unless this
guidance condition (as we may call it) can be satisfied, we will not be able
to make sense of the idea that persons are genuine agents, capable of deter-
mining what they shall do through the process of deliberation.32
On a natural understanding of what rational guidance involves, it is only in-
sofar as an agent acts in accordance with her judgment as to what she has most
reason to do that she can be said to respond to reasons.
However, as Jones argues, there is an alternative account of what is required
for reason-responsiveness that is more favorable to emotions. According to
Jones, acting on the basis of emotions can involve genuine agency, provided
30
Jones, Emotion, Weakness of Will,190.
31
See Velleman, What Happens When Someone Acts?; Christine M. Korsgaard, The
Normativity of Instrumental Reason, in Ethics and Practical Reason, edited by Garrett Cullity and
Berys Gaut (Oxford:Clarendon Press, 1997); R. Jay Wallace, Three Conceptions of Rational Agency,
Ethical Theory and Moral Practice 2 (1999):219; Kieran Setiya, Reasons without Rationalism (Princeton,
NJ:Princeton University Press,2007).
32
Wallace, Three Conceptions of Rational Agency,219.
that the action expresses the agents commitment to what she calls regulative
guidance. Regulative guidance involves the on-going cultivation and exercise
of habits of reflective self-monitoring of our practical and epistemic agency.33
An action caused by emotions, or for that matter by any other reason-tracking
subsystem, thus manifests genuine agency just in case the agents dispositions
to reflective self-monitoring are such that she would not rely on that first order
sub-system were it reasonable for her to believe that it failed to reason-track.34
Thus, an agent who acts on, say, her anger might be responding to reasons
even if she acts against her judgment concerning what to do. The angry agent
is responding to reason if she has well-tuned self-monitoring habits, such that
she would not have relied on her anger if there had been reason for her to be-
lieve that her anger was misleading her. In other words, what is required for
reason-responsiveness is nothing but the exercise of well-tuned epistemic and
practical habits, that is, of what can be described as agential virtues.
This Agential Virtue Account, as I shall call it, is highly attractive. In the
remainder of this section Iwill provide what can be considered as a partial
defense of the account. More precisely, Iwill argue that it is not threatened
by a number of worries that have been or might be raised. The first volley of
worries is due to Sabine Dring. Dring explicitly criticizes Joness account,
and it is instructive to start with her arguments. According to Dring, the
main problem of Joness analysis is that it remains obscure how, in the case
of conflict between two reason-tracking sub-systems, the agent may decide
which sub-system is to be given preference.35 What Dring has in mind, just
as Jones, are cases of akrasia, that is, cases in which there is a conflict between
the conclusion of the practical reasoning of an agenther better judgment
on one hand and her motivation and intentional action on the other hand.
More specifically, she is thinking of cases of akrasia in which the practical con-
clusion of an agent and what she feels conflict, and it is the feeling that appears
to better fit the agents reasons.36 The classic example is that of Huckleberry
Finn, who is torn between his judgment that he ought to denounce Jim to the
slave hunters and his feelings for Jim. Dring asks:If Huck is mistaken in
the example, how could he have avoided the mistake?...What Huck needs to
know is whether he should here and now follow his sympathy or had better
stick to his judgment.37 Now, it clearly would have been useful for Huck to
know whether to follow his judgment or his feelings. The question, however,
33
Jones, Emotion, Weakness of Will,194.
34
Jones, Emotions, Weakness of Will,195.
35
Sabine Dring, Why be Emotional? in the Oxford Handbook of Philosophy of Emotion, edited
by Peter Goldie (Oxford:Oxford University Press, 2010),290.
36
See Alison McIntyre, Is Akratic Action Always Irrational? in Identity, Character and Morality,
edited by Owen Flanagan and Amlie Rorty (Cambridge, MA:MIT Press, 1993); Nomi Arpaly and Tim
Schroeder, Praise, Blame and the Whole Self, Philosophical Studies 93 (1999):161188.
37
Dring, Why be Emotional?290.
is whether a general account of agency should be expected to help Huck in

that matter. What the Agential Virtue Account says, simply, is that insofar
as Hucks self-monitoring habits are well tuned, he can be considered to be
reason-responsive. This might appear disappointing as advice to Huck, but
what has to be underlined is that no general account of agency can ensure that
people in Hucks position arrive at the correct answer. It might be useful make
a comparison to the moral case:what has to be distinguished in the moral case
is the advice to an agent who has to take moral decision, on one hand, and a
general account of right action, on the other hand. The former pertains to the
theory of deliberation and decision making, whereas the latter requires a sub-
stantive moral theory.
According to Dring, the problem that she sees in Joness account has a
deeper source. It emerges from the fact that Jones fails properly to distinguish
between objective and subjective reasons, a distinction she takes from Niko
Kolodny.38 Objective reasons are defined as reasons that depend on objective
features of the agents situation, while subjective reasons are claimed to be what
the agent justifiably sees as reasons, given the content of her mental states. In
essence, subjective reasons are the agents take on her objective reasons, some-
what idealized by the justification condition. According to Dring, the guid-
ance condition implies that an account of practical rationality...must be given
in terms of subjective reasons:an agent is rational to the extent to which he is
guided by his subjective reasons.39 Given this claim, it does not come as a sur-
prise that she disagrees with the Agential Virtue Account. Rational guidance
requires that the agent be guided by what she explicitly takes to be her reasons,
that is, by her practical judgments.
Let us look at the specific form the objection against Joness account takes.
Spelling out Joness proposal in terms of objective and subjective reasons,
Dring claims that it is objective reasons that we track while it is subjective
reasons to which we respond. She spells out the problem as a conscious and
explicit conflict between two objective reason-tracking systems and claims
that in such a case of conflict it is it unclear what subjective reasons the agent
has.40 Given this way of construing the conflict, Dring claims that there is
only one option, namely, appealing to the agents higher-order reflective and
self-monitoring capacities:In the end, it will come down to a judgment about
what one ought to do in the given situation.41 It is on these grounds that
Dring infers that acting on ones feelings instead of following ones judgment
is irrational.
38
See Niko Kolodny, Why Be Rational? Mind 114:181200.
39
40
41
A first, merely terminological, difficulty is that, as far as Ican see, talk of

objective versus subjective reasons is misleading in that it gives the false im-
pression that there are two kinds of reasons.42 Introducing subjective reasons is
like introducing subjective cats and poppies next to cats and poppies. It seems
far simpler to replace talk of objective versus subjective reasons with the dis-
tinction between reasons on one hand and the agents (idealized) take on rea-
sons on the other hand. On this distinction, both tracking and responding
concern the same thing, namely, reasons. But while tracking need not involve
judgments, responding typically doestypically, but not necessarily, because
responding can involve emotions and other subsystems, provided that they
are welltuned.
Yet the real problem with Drings suggestion is that we need to follow our
practical judgment to be reason-responsive. It might be correct that from the
point of view of the agent the only thing that she can envisage in the event of
a conflict between the verdicts of two subsystems is to appeal to a judgment
about what do to. But this is a far cry from showing that acting on that judg-
ment is what an agent ought to do. This is so because as is made clear in the
case of Huck there might be a gap between the agents take on her reasons and
her actual practical reasons.43 Given this, it is only prima facie that an agent
ought to follow her better judgment.
This is good news for the Agential Virtues Account. However, there are
other worries that one might have with respect to the suggestion that acting ac-
cording to an emotion in the face of ones judgment might consist in respond-
ing to a reason. One important question is whether or not the akratic action
can really be attributed to the agent. One might think that it is not Huck who
is in charge when he fails to denounce Jim against his better judgment. This
is exactly what David Velleman suggests regarding the case of someone who
loses his temper when meeting with an old friend for the purpose of resolving
a minor conflict.44 What Velleman suggests is that in such a case it is not the
agent who is in charge. He writes, It was my resentment speaking, not I.45
This is clearly something we might be tempted to say, would it be only to avoid
blame. However, we may also doubt whether a person who argues on these
lines is necessarily right to say that she was not in charge. According to the
Agential Virtue Account, whether or not it is the agent who is in charge does
not depend on the agents take but on whether the dispositions to reflective
42
To be fair, it has to be noted that Dring explicitely states that subjective reasons are derivative
of objective reasons, and that they do not constitute a second class of normative reasons beyond objec-
tive reasons. Dring, Why be Emotional?287.
43
As McIntyre, Is Akratic Action Always Irrational? argues that this point can be made without
postulating external reasons, for an agent can be wrong about his own motivationalset.
44
See Velleman, What Happens When Someone Acts, 464465.
45
Velleman, What Happens When Someone Acts, 465. Cf. Wallace, Three Conceptions of
Rational Agency, 222, for a similarpoint.
self-monitoring that constitute agential virtues are in place. What the account
tells is that if the conditions that it spells out are satisfied, then the agent is in
charge, despite his being reluctant to admit it. This is, Ibelieve, an attractive
feature of the Agential Virtue Account.
Finally, let me consider a last worry. On the story proposed here, emo-
tions can inform us about our practical reasons, depending on how reliable
these emotions are. Moreover, undergoing an emotion and acting on its basis
can constitute responding to a reason, and it does constitute responding to
a reason when the relevant emotional disposition or dispositionsthe sub-
systemare well tuned. Accordingly, acting on the basis of an emotion can
amount to acting in light of ones reasons. But how could this be the case, one
might ask, if the agent is not aware of the reasons she has? And how could the
agent be aware of her reasons otherwise than by having beliefs or judgments
regarding those reasons? Another way to express the same worry is to claim
that reasons need to be transparent to the agent:to be motivated by a reason,
that is, to act for a reason, an agent needs to be aware of that reason, where
awareness is understood to require belief or judgment.46
In reply, one may be tempted to deny that awareness requires judgment
and appeal to the fact that emotions are conscious experiences. One could
thus claim that emotions allow us to be aware of values and hence of the
corresponding reasons. After all, nobody would deny that there are both
judgmental and perceptual ways to be aware of poppies and their shapes
and colors. So why not say that when fearing something we are aware of that
things fearsomeness? One worry here is that emotional experiences appear
to lack the phenomenal transparency that characterizes sensory perceptions.
The best way to describe my experience of red poppies is to talk about red
poppies. But it seems that this is not so with emotions. Arguably, the best way
to describe what it is like to undergo fear or anger or regret is not to refer to
the objects of these emotions and their properties. This suggests that when
undergoing an emotion we are not stricto sensu aware of the relevant evalu-
ative properties.
There is some leeway here, for in fact we quite often find ourselves describ-
ing what our emotions are about to describe what it felt to experience them.
Thus, we describe what were afraid ofthe huge hairy spider that suddenly
fell from the tree, for instanceto explain what we felt.47 Moreover, it should
be noted that the worry in question is mitigated by the fact that we commonly
46
See Melissa Barry, Realism, Rational Action, and the Humean Theory of Motivation, Ethical
Theory and Moral Practice 10 (2007):231242, 323; Maria Alvarez, Kinds of Reasons:An Essay in the
Philosophy of Action (Oxford:Oxford University Press, 2010), 25; Setiya, Reasons without Rationalism,
40. Setiya takes it as obvious that whenever we act for a reason, we not only believe but we also know
that we are acting for that reason:The second insight is that we know without observation not only
what we are doing, but why(40).
47
Thanks to Mark Nelson for suggesting thispoint.
form evaluative judgments on the basis of the emotions we feel. For instance,
when angry at someone because of some remark, we tend to judge that this
persons remark was offensive. Thus, emotions typically come with a dispo-
sition to make corresponding evaluative judgments. Finally, the claim that
the putative phenomenal transparency of emotions indicates that there is no
awareness of evaluative properties appears committed to the assumption that
the awareness of something needs to be explicitly articulated in terms of con-
cepts; it would only be when we form judgments about things that we can be
said to be aware ofthem.
Be that as it may, the very claim that an agent needs to be aware of a reason
whether that entails judgment or another kind of awarenessto be motivated
by that reason already appears too strong to be plausible. Switching to a case
that does not involve emotions, habitual actions can certainly be motivated
by reasons even though we are not aware of the reasons for which we perform
these actions when we perform them. Indeed, the whole point of the Agential
Virtue Account is to make room for actions that are not explicitly guided by
articulated reasons but nonetheless appear to exhibit full-fledged agency.
It might be replied that cases of habitual actions are quite different from
emotional actions.48 In the case of habitual actions, agents can readily recon-
struct the reasons behind what they are doing. Consider my habit of having
an espresso at ten oclock. It is easy for me to see that having an espresso at
ten oclock is not only pleasant but also useful in terms of productivity. By
contrast, we often are at loss when we try to reconstruct the reasons behind
actions motivated by emotions. This, it can be argued, is particularly striking
in Hucks case. From his point of view, it is bound to appear utterly unjustified
not to have denouncedJim.
It is important to recognise the difference between habitual and emotional
action. However, the contrast should not be exaggerated. Afirst point is that
there are cases of habitual actions in which it is less easy to reconstruct the
reasons behind what we are doing. We are all aware of some strange habits
we have, in favor of which we find nothing to say even on reflection. Indeed,
in some cases, acting out of habit can be in conflict with our better judgment.
This is a natural way to understand Davidsons famous teeth-brushing case.49
You are lying in your bed when you realize that you have forgotten to brush
your teeth. On reflection you conclude that you should just stay in bed and do
nothing about itthere is no harm in not brushing your teeth just this one
time, and getting up would spoil your calm. But off you go and brush your
48
Thanks to Karen Jones for raising this question.
49
See Donald Davidson, How is Weakness of the Will Possible? in Davidson, Essays on Actions
and Events (Oxford:Oxford University Press, 1980; original publication1969).
teeth. It is to be expected that in such a case you will find it hard to reconstruct
the reasons behind your action.
The second point to make is that in fact we are often in a position to recon-
struct the reasons behind emotional action. Consider again Vellemans case
of the person who loses his temper when meeting with his friend.50 In the
case Velleman imagines, later reflection leads the protagonist to realize that
accumulated grievances had crystalized in his mind into a resolution to end
the friendship. Thus, the reasons for the burst of anger are in fact accessible
to the agent. According to Velleman, the fact that the reasons are accessible
to the agent is not sufficient to make it the case that the action is done for a
reason. In fact, the anger case is his main example of an action that is less than
full-blooded.
The Agential Virtue Account has a very different, and Ibelieve more plau-
sible, take on cases like this. The third and most important point, thus, is that
according to the Agential Virtue Account what counts is not so much the pos-
sibility of reason reconstruction as the exercise of agential virtues. The point
to emphasize is that the description of the case leaves open whether or not the
dispositions of reflective self-monitoring were in place. It is true that in the
absence of any reason to believe that the agent in question has such disposi-
tions, Velleman appears right when claiming that the action failed to involve
reason-responsiveness. However, we might easily imagine a case in which
being moved by anger expresses well-tuned dispositions, such that the agent
would not have relied on her anger if there would have been reason for her to
believe that her anger was inappropriate.
More would have to be to be said to spell out and fully defend the Agential
Virtue Account.51 Questions arise regarding the nature and functioning of the
dispositions of reflective self-monitoring. What exactly are these epistemic
and practical habitswhat Ihave called agential virtuesand how do they re-
late to epistemic and practical norms and goals? Also, one might wonder how
much room there is for an agent to cultivate and exercise such dispositions,
assuming that basic aptitudes and socialization are bound to be decisive fac-
tors. In particular, one might wonder whether the counterfactual description
in terms of what would the agent would have done had there been reasons
for her to believe something makes for a problematically circular account of
reason-responsiveness in that it explicitly refers to reasons. There are impor-
tant and difficult questions, but for now Iwill leave them aside and turn to
what the proposed account entails for autonomy theories.
50
See Velleman, What Happens When Someone Acts, 464465.
In this appeal to virtues, the account proposed is similar to Setiyas suggestion that good practical
51
reasoning can be specified only in terms of ethical virtues. See Setiya, Reasons without Rationalism.
3. Emotions and Autonomy Theories
As Inoted in the introduction, it is often taken for granted that emotions and
autonomy are at odds. Giving voice to this intuition, Laura Ekstrom writes:
We do not act autonomously in acting on passions, whims, and impulses be-
cause these overtake us; we are generally passive with respect to them, they
do not engage our understanding or capacity for reflective evaluation. We
make our lives more our own by examining such impulses and by acting in
accordance with our evaluations.52
If the picture Ihave sketched is on the right lines this negative conception
of emotions is not warranted. It can surely be acknowledged that in some cases
the undergoing of an emotion is a barrier to autonomy. Our emotions some-
times get things wrong and thus fail to track reasons. And acting on an emo-
tion might fail to consist in responding to reasons because agential virtues are
not in place. Quite generally, given their influence on action and thought, the
effect of emotions can be particularly pernicious. However, the wholesale re-
jection of emotions is not warranted. Insofar as emotions are perceptual expe-
riences of evaluative properties, they can, and sometimes do, inform us about
our practical reasons and hence play a crucial role in the assessment of our
normative principles. So it is deeply mistaken to suggest that critical reflection
is something that is independent of the emotions we feel and incompatible
with the fact that an emotion isfelt.
Furthermore, the wholesale rejection of emotions is misguided because
acting on an emotion can be acting in light of reasons we do have. Thus, to the
extent that reason-responsiveness is a central aspect of autonomy, emotions are
far from being at odds with autonomy. Autonomy accounts that place reason-
responsiveness at the core of autonomous agency have to accept emotions as
a potential source of autonomous actions. There might be some disagreement
as to what other ingredients autonomy requiressuch as psychological inte-
gration or a sense of our status as agents. But whatever the details of ones full
account of autonomous action, it has to be conceded that insofar as emotions
and reason-responsiveness are not at odds emotions and autonomy need not
be either.
Another implication of the account of emotions proposed here is that care
accounts of autonomy, or at least the version of such accounts according to
which emotions are essential to autonomy, should not be taken to be antithet-
ical to reason-responsiveness accounts.53 This is because on this version cares
52
Laura W. Ekstrom, Autonomy and Personal Integration, in Personal Autonomy, edited by J. S.
Taylor (Cambridge:Cambridge University Press, 2005),160.
53
In his later work, Frankfurt suggests that love is an essential feature of the self, but it is far from
clear that what he calls love involves emotions. See Harry Frankfurt, Necessity, Volition and Love
are complex emotional dispositions. As Shoemaker puts it, Talk of caring is

simply a way of referring to the range of emotional reactions one is expected to
have with respect to the fortunes of the cared-for object.54 To care for someone
or for something consists in being disposed to undergo a number of emotions,
depending on the good or bad fortune of whom or what (e.g., persons, ideals)
one cares for. We feel sadness when things go bad for the object of our care,
joy when things go well, hope that things will go well, fear that things will
go badly, and so forth. Now, according to Shoemaker, freedom and thus au-
tonomy depend on acting on our strongest cares.55 Shoemaker expresses this
thought in the following manner:
If we attach free agency to willing action, then, and willing action consists in
the action Igenuinely want to do, and what Igenuinely want to do depends
on what Icare most about in any particular situation, then free agency is
grounded in care. To the extent that what Ido does not ultimately depend
on my strongest care(s) at the time of action, Iam unfree.56
The question is what kind of states the emotions that constitute our cares
are. On the Perceptual Theory Ihave sketched, cares are dispositions to un-
dergo states that are perceptual experiences of evaluative properties. Thus,
they are dispositions to undergo states that can, when things go well, inform
us about the practical reasons we have. To the extent that our emotions fit
their objects, they happen to inform us correctly about the reasons we have.
Thus, in such favorable cases, our dispositions to feel these emotions are dis-
positions to be keyed to practical reasons. In such cases, acting on our cares
consists in acting in ways that reflect the reasons we have. Given this, it has
to be acknowledged that our cares can be essential to reason-responsiveness.
According to the Agential Virtue Model, this will be the case when the agent
displays well-tuned epistemic and practical dispositions.
Consider what happens when you care for a friend. To care for a friend is to
be disposed to be happy if she thrives, unhappy if she is not well, and fearful
when something threatens her. Thus, on the account of emotions proposed
here, caring involves being disposed to perceive the friends happiness as good
and her unhappiness as bad. And a friends happiness being good surely is a
reason you have to do a number of things, such as helping her when she needs
and asks for it. Thus, if you help your friend, what you do corresponds to the
reasons you have that are tracked by the emotions that constitute your caring
for your friend. Moreover, in the light of the Agential Virtue Account, it must
be allowed that acting on the basis of your cares can well consist in responding
54
David Shoemaker, Caring, Identification, and Agency, Ethics 114 (2003):94.
55
Shoemaker, Caring, Identification, and Agency, 103104.
56
Shoemaker, Caring, Identification, and Agency.
to reasons. Thus, if your dispositions are well tuned, your acting on the basis
of your cares consists in responding to reasons.
The lesson, then, is that insofar as emotions are central to care accounts of au-
tonomy based on care can make room for the intuition that reason-responsive-
ness is central to autonomy. Even though there is likely to be disagreement about
other putative ingredients of autonomy, the two kinds of accounts need not be
in disagreement as to the importance of reason-responsiveness in autonomy.
4.Conclusion
To conclude, let me return to what Ireferred to as the Ludicrous Argument.

In a nutshell, what is wrong with the argument is the idea that being liable to
undergo emotions and, in particular, that experiencing an emotion, is incom-
patible with autonomous agency. Emotions have traditionally been taken to be
dubious and unreliable elements in our psychology, which have to be submit-
ted to some taming process, such as critical scrutiny and reflective endorse-
ment. There has recently been a general tendency to see emotions in a more
positive light, something that echoes feminist concerns. Following that ten-
dency, what Ihave tried to show in this paper is that the negative conception
of emotions as they relate to autonomous agency is severely misguided. Even
though emotions do not involve conceptually articulated contents, there are
reasons to consider them to be perceptual experiences of a kind, in that they
have intentional objects and representational content. This is why to be prop-
erly reason-responsive and autonomous we should listen to our emotions.
Acknowledgments
Earlier drafts of this chapter have been presented at the round-table on the ra-
tionality of emotions, Pacific APA, Seattle, 2012, organized by Marina Oshana;
at the conference on relational autonomy, organized by Catriona Mackenzie
and Natalie Stoljar, McGill, September 2012; at the Philosophy Institute,
Stockholm University; at the Department of Philosophy, York University; at
the Centre Interfacultaire en Sciences Affectives, University of Geneva; and
the Congress of the CPA, Victoria, 2013. For discussions and comments, Iam
grateful to Rachel Barney, sa Carlsson, John Christman, Julien Deonna,
Vincent Duhamel, Marilyn Friedman, Bruno Guindon, Nancy Hirshmann,
Louis-Philippe Hodgson, Karen Jones, Catriona Mackenzie, Colin MacLeod,
Diana Meyers, Mark Nelson, Graham Oddie, Jonas Olson, Marina Oshana,
Antoine Panaioti, Gopal Sreenivasan, Natalie Stoljar, Sarah Stroud, Jennifer
Szende, Fabrice Teroni, Joey van Weelden, and especially to the editors of the
volume, Andrea Veltman and MarkPiper.
{9}
Autonomy and Self-Care

Andrea C. Westlund
For more than two decades, feminist philosophers have been developing im-
portant new conceptions of autonomy that take the sociality of human agency
appropriately seriously. Interestingly, this relational turn has at the same time
ushered in a new debate about the nature of autonomous agents attitudes to-
ward themselves as individual agents. Some relational theorists have argued,
for example, that a sense of individual self-worth1 or self-trust2 may be neces-
sary for autonomy. One important question this debate has raised is whether
or not the resulting conception of autonomy will be normatively neutral.3 For
purposes of this paper Iwill bracket that question and focus instead on the
relationship between autonomy and self-regarding attitudes themselves. What
has begun to emerge in the feminist literature is a conception of autonomy
in which the autonomous agent must stand in a special kind of relation to
herself.4 Ithink there is something deeply right about this idea. In this paper
Itry to flesh out one kind of self-relation that is required for autonomy and to
explain what it has to do with relationality of the more familiar, interpersonal
variety.
1
See Paul Benson, Feeling Crazy: Self-Worth and the Social Character of Responsibility, in
Relational Autonomy:Feminist Perspectives on Autonomy, Agency, and the Social Self, edited by Catriona
Mackenzie and Natalie Stoljar (NewYork:Oxford University Press, 2000),7293.
2
See Carolyn McLeod and Susan Sherwin, Relational Autonomy, Self-Trust, and Health Care for
Patients Who Are Oppressed, in Mackenzie and Stoljar, Relational Autonomy, 259279.
3
See Paul Benson, Taking Ownership: Authority and Voice in Autonomous Agency, in
Autonomy and the Challenges to Liberalism: New Essays, edited by J. Christman and J. Anderson
(Cambridge: Cambridge University Press, 2005), 101126; Benson, Feminist Intuitions and
the Normative Substance of Autonomy, in Personal Autonomy, edited by James Stacy Taylor
(Cambridge: Cambridge University Press, 2005), 12442; John Christman, Relational Autonomy,
Liberal Individualism, and the Social Constitution of Selves, Philosophical Studies 117:12 (2004):143
164; Christman, Procedural Autonomy and Liberal Legitimacy, in Taylor, Personal Autonomy, 277
298; Andrea C. Westlund, Rethinking Relational Autonomy, Hypatia 24:4 (2009):2649.
4
Of course, self-governance is itself a kind of self-relation. The idea Ipursue in this paper is that
the autonomous agent is self-governing at least in part in virtue of standing in another, conceptually
distinct sort of relationship to herself, namely, a relation of self-care.
The self-relation Ihave in mind is not exactly one of self-trust or self-worth

(at least as these are usually understood), since it does not involve any kind of
positive self-assessment. But Iargue in this paper that we have reason to con-
sider it a form of self-care. In the first section of the paper, Iset out three re-
lated conceptions of care, developed by Harry Frankfurt, Agnieszka Jaworska,
and Jeffrey Seidman. In the second section, I raise some worries about the
way these conceptions have figured in an authenticity-based approach to au-
tonomy and begin to sketch out an alternative, answerability-based approach.
In the third section, Iput Jaworskas and Seidmans conceptions of care to new
use and develop an account of self-care that emphasizes care for ones reasons.
Finally, in the fourth section Ielucidate the two-sided relationality of this kind
of self-care. Iargue that caring about ones reasons implies caring about in-
tersubjective assessments thereof and that the relation of self-care therefore
implies openness to the reasons of others.
1.Care
The recent literature on care has grown along two almost entirely separate
lines. One line focuses primarily on the place of care in morality and tends to
concern itself almost exclusively with care as an attitude toward other persons.
This strand has developed out of pioneering work in feminist ethics by Carol
Gilligan, Nel Noddings, Sarah Ruddick, and others and by now includes much
interesting work on dependency and disability as well.5 For the most part, this
body of work does not bear directly on issues of autonomy, though the strand
of Gilligans argument Idevelop later in this paper is an important (and gener-
ally overlooked) exception.
The other line of work is more immediately relevant to my project, focusing
as it does on the role of care in human agency or autonomy. This literature
tends to range more widely over care as an attitude toward objects, places, ide-
als, and so forth in addition to persons. It has been inspired at least in part by
Harry Frankfurts groundbreaking work on personhood and freedom of the
will, particularly in its later manifestations.6 In addition to Frankfurt himself,
5
See Carol Gilligan, In a Different Voice (Cambridge, MA:Harvard University Press, 1982); Nel
Noddings, Caring: A Feminine Approach to Ethics and Moral Education (Berkeley: University of
California Press, 2003); Sara Ruddick, Maternal Thinking:Toward a Politics of Peace (Boston:Beacon
Press, 1995). On dependency and disability, see Eva Feder Kittay, Loves Labor (NewYork:Routledge,
1999). I would place Stephen Darwalls care-based analysis of welfare in this category as well. See
Darwall, Welfare and Rational Care (Princeton, NJ:Princeton University Press, 2002). Though it does
not directly address feminist issues, Darwall himself describes his work on welfare and care as inspired
partly by Noddings and other feminist theorists.
6
Harry Frankfurt, On Caring, in Necessity, Volition, and Love (Cambridge:Cambridge University
Press, 1999), 155180.
Autonomy and Self-Care 183
Agnieszka Jaworska and Jeffrey Seidman have made important contributions

on this front.7 In this section Itrace out the account of care that emerges from
their work. It is a compelling account, and I will ultimately argue that care
(thus understood) does play an important role in self-governmentjust not
exactly the same role that Frankfurt, Jaworska, and Seidman have assignedit.
Frankfurt, Jaworska, and Seidman all frame questions about autonomy in
terms of the distinction between what is internal and what is external to the
agents will. In doing so, all three focus on what many philosophers now refer
to as authenticity conditions of autonomy. These conditions specify what it
is that makes an attitude, motive, or desire truly ones own or, as Frankfurt
and Jaworska put it, internal to the self. On authenticity-based views,
self-governance is quite literally a matter of governance by the self. We need
to be able to distinguish between what belongs to the self and what does not
to know if an agent is self-governing with respect to any particular choice or
action. We must, as it is often put, be able to identify the agents standpoint
or those attitudes that have the authority to speak for theagent.
In its earlier versions, Frankfurts hierarchical account of internality focused
on a thin notion of reflective endorsement, involving higher-order desires to
be moved by particular first-order desires. Over time, however, a species of
care has come to play an increased role in his account. To care about some-
thing, for Frankfurt, is to have and identify with a higher-order desire in favor
of the persistence of a lower-order desire for that thing. The persistence of the
desire in question cannot be a matter of mere volitional inertia; to count as
an instance of caring, it must be actively willed by the agent. This means that
the agent must not merely endorse but actually be disposed to support and
sustain his desire [to ] even after he has decided that he prefers to satisfy
another desire instead.8 Caring, on this view, is not primarily (or merely) a
matter of having certain feelings or beliefs but rather of having a certain con-
figuration of the will. Caring about something is a matter of being committed
to ones first-order desire for that thing, where being committed entails having
the structure of aforementioned higher-order desires and dispositions.9
So what role does caring, thus understood, play in autonomy? For Frankfurt,
individuals govern themselves to the extent that the commands that they
obey, whether based upon rules or not, are their own commands.10 Frankfurts
7
See Agnieszka Jaworska, Caring and Internality, Philosophy and Phenomenological Research 74:3
(2007):529568; Jaworska, Caring, Minimal Autonomy, and the Limits of Liberalism, in Naturalized
Bioethics: Toward Responsible Knowing and Practice, edited by H. Lindemann, M. Verkerk, and M.
U. Walker (Cambridge: Cambridge University Press, 2009), 80105; Jeffrey Seidman, Valuing and
Caring, Theoria 75 (2009): 272303. Though I do not examine it in this paper, David Shoemaker
develops a similar, care-based account in Shoemaker, Caring, Identification, and Agency, Ethics 114:1
(2003):88118.
8
Frankfurt, On Caring,160.
9
10
most distinctive move, in his later work, has been to argue that the commands
of love are at least as authoritative for us as the commands of duty or reason.
Indeed, since Frankfurt takes love to define our essential nature as persons
and to settle our final ends, it is reasonable to read his account as one on which
autonomy is ultimately grounded in love rather than in reason. On his view,
what we love has a distinctive claim to mark out our standpoint as persons or
at least what is most central to that standpoint. Since love is a species of caring,
the result is a view that links autonomy very clearly to care, via the internality
of cares to the self.11
In Caring and Internality, Jaworska accepts that an attitude is truly ones
own if it is internal to the will and also that our cares are invariably internal.
But she departs from Frankfurt by insisting that care is first and foremost an
emotional attitude and not just a commitment of the will. Jaworska argues
that Frankfurts volitional treatment of care and internality is too reflective:in
requiring agents to have a structure of attitudes about attitudes, Frankfurts
view (and other similarly hierarchical views) puts care beyond the reach of
agents who are not yet, or are no longer, cognitively sophisticated enough to
have such higher-order attitudes but who seem obviously to care about things
and to have, in that sense, a standpoint of their own. Young children and other
marginal agents (e.g., those in early stages of dementia) seem to be capable of
care, and their cares (like anyone elses) seem invariably to be internal tothem.
Jaworska draws on Michael Bratman for a revised Frankfurtian conception
of internality and on Bennett Helm for a more congenial account of care. From
Bratman she adopts the idea that an attitude is internal if it has as part of its
function to support the psychological continuities and connections that con-
stitute the agents identity and cohesion over time.12 Where she differs from
Bratman is in the range of attitudes that she thinks may perform this func-
tion. While Bratman focuses on plans, policies, and other distinctively reflec-
tive attitudes, Jaworska argues that any attitude, reflective or not, that plays the
relevant role should count as internal. At least some emotions, she argues, fit
thebill.
To understand her point, we must distinguish between primary and sec-
ondary emotions. Primary emotions, Jaworska explains, involve more or less
fixed patterns of emotional responses to specific, immediately sensed features
of ones environment, such as stereotypical reactions of fear, disgust, rage,
or surprise.13 These do not, in her view, play any special role in unifying our
11
What we love is a subset of what we cannot help caring about, in the aforementioned sense of
caring. When we love something, Frankfurt claims, we have a disinterested desire for its flourishing.
This desire, since we cannot help but have it, and cannot help but want to have it, sets boundaries on
our will that we willingly embrace. It determines what we can and cannot bring ourselves to choose and
do and in this way defines the contours of our will and our essential natures as persons.
12
Jaworska, Caring and Internality,552.
13
agency over time. A secondary emotion, by contrast, is a response to ones

conscious, deliberate understanding of a situation. She cites gratitude, envy,
jealousy, guilt, hope, and grief (among others) as examples of emotions of this
sort, which presuppose...high-level processing of the situation.14 There is
good evidence, elaborated by Antonio Damasio, that brain injuries that com-
promise secondary emotionality also cause severe deficits in agents abilities
to plan and coordinate their activities over time. As Jaworska puts it, sec-
ondary emotions support entire networks of psychological continuities and
connections and appear to be a necessary substratum of the planning-
and intention-based continuities and connections on which Bratman has
focused.15 While it is easy to imagine feeling a primary emotion as an alien or
overwhelming force, Jaworska claims, such secondary emotions as joy, grief,
or gratitude strike us as invariably internal.
Jaworska argues that care, too, is properly understood as a secondary emo-
tion that supports a network of psychological continuities and connections.
She defines care as a structured compound of emotions, emotional predis-
positions, and desires that all construe the same object as a source of impor-
tance commanding emotional vulnerability.16 The object of care is, in Bennett
Helms terms, the unifying focus of this compound of attitudes; it is what the
complex emotion is about. In addition to their joint connection to a single ob-
ject, Jaworska notes, Helm has shown that the attitudes involved in caring are
subject to two distinct patterns of rational requirement. On one hand, tonal
commitments require agents to respond with positive or negative emotions
(joy or satisfaction versus sadness or grief) as the focus of their care fares
well or badly; on the other, transitional commitments require agents to feel
forward- or backward-looking emotions (fear or hope versus relief or frustra-
tion) depending on the agents temporal perspective on the fate of the object.
Jaworska points out that these networks of continuities and connections not
only directly contribute to the support of an agents identity over time but also
imbue objects with importance in a way that can and typically does support
the development of other, higher-order unifying attitudes, such as the inten-
tions, plans, and policies on which Bratman focuses.
In Valuing and Caring, Seidman largely accepts the JaworskaHelm con-
ception of care but argues that it does not succeed in explaining what ties
together the various elements of caring (emotions, emotional dispositions,
and desires) such that they jointly constitute a single attitude. To fill this
14
15
16
Jaworska, Caring and Internality, 560. Object is to be understood broadly, as including not only
material objects and particular individuals but also states of affairs, sequences of events, places, ideals,
relationships, andmore.
explanatory gap, he argues, we must recognize that caring includes not just
emotional dispositions but also cognitiveones.
The various components of caring are, of course, linked in that they are
focused on a single object. But a caring agent might have a range of other
attitudes toward that same object without those attitudes being part of his or
her caring. Nor, Seidman argues, do the tonal and transitional commitments
identified by Helm suffice to pick out all and only the right attitudes, either.
Seidman points out that a sons irritation at his fathers humming may be ro-
bustly interconnected, both tonally and transitionally, with other attitudes to-
ward the father yet have nothing to do with the fact that he cares forhim.
According to Seidman, what does help unify the attitudinal patterns that
constitute care is the agents perception of the focal object as important. More
precisely, the fact that an agent perceives an object as important stands in an
explanatory relation to the attitudes that are constituents of her concern for
that object. To see an object as important (to oneself), Seidman argues, is to
see it as a source of practical reasons (for oneself).17 Seeing some object X as a
source of practical reasons does not entail actually believing it to be a source
of practical reasons. Sometimes our perceptions and judgments come apart.
But Seidman argues that seeing something in a certain way does, in normal
circumstances, where no countervailing beliefs defeat this disposition dispose
one to believe it is that way.18 Seeing some object X as importantseeing X as
a source of practical reasonsdisposes the agent to believe that X is a source
of practical reasons.
The fact that an agent sees X as a source of practical reasons, and is thus
disposed to believe that she has the reasons in question, explains certain of her
emotions, emotional dispositions, and desires as well as the rational patterns
among those emotions, and those emotions (the ones linked together by their
common explanation in perceptions of importance) are the ones that consti-
tute concern for the object in question. Other emotions (such as irritation at
ones fathers humming) are not explained by ones perception of the father as
important and thus are not part of caring for him. Seidmans point, in short, is
that we cannot identify the emotions, emotional dispositions, and desires that
constitute caring without appeal to perceptions of importance and the cog-
nitive disposition (to believe one has reasons) that these perceptions involve.
Thus amended, Seidman agrees with Jaworska that our cares are invariably
internal to us and have a legitimate claim to constitute the agents standpoint.
17
Jaworska seems to agree with at least this part of Seidmans claim, since she argues in Caring,
Minimal Autonomy, and the Limits of Liberalism that in acting on ones cares one acts on the percep-
tion of a reason. But as far as Iunderstand her position, she would resist the further idea that perceiving
something as a reason also disposes one to believe it is a reason. This disposition would be beyond the
cognitive capacities of some of the marginal but still (on her view) caring agents that she considers
minimally autonomous.
18
Seidman, Valuing and Caring,286.
Indeed, Seidman argues that the emotional and cognitive dispositions in-
volved in caring establish a more robust form of diachronic unity than the
policies and plans on which Bratman focuses. In addition to unifying agency
across time, Seidman argues that our concerns generate a diachronically uni-
fied subjectivity, or, as he also puts it, a temporally extended subject...with
a cognitive and emotional take on the world.19 This is important, Seidman
thinks, because when we ask what attitudes speak for the agent we really have
more in mind than just a locus of deliberation and action.20 Our concerns
may conflict with one another and with our self-governing policiesbut ac-
tion motivated by our concerns will nonetheless always be expressive of who
we are as subjects.21
2. Mental Freedom
There is much that seems promising about this view, as a view of what Ive
been calling authenticity. It gives a neat explanation of what unifies raw psy-
chic materials into a well-defined (even if complex or conflicted) subject,
with a take of its own, that endures across time. If autonomous choice or
action is fundamentally choice or action that expresses who the agent is, this
care-based account seems like a strong contender for telling us how to identify
a who in the first place. But there is a significant worry about this whole ap-
proach to the concept of self-governance:such views do not explain how being
just exactly who we are can sometimes constitute a failure of autonomy.
Within the framework of a purely authenticity-based account of autonomy,
this objection will not sound like a sensible one, for there is no conceptual
space between authenticity and autonomy on such views. But Ithink we must
make space if we are to do justice to pretheoretical intuitions about an im-
portant set of cases.22 The cases Ihave in mind are ones in which agents iden-
tify so thoroughly with their concerns that they are unable to entertain the
possibility that they might be mistaken or that their concerns might require
defense. Highly dogmatic agents are sometimes like this, and Ive argued that
deeply deferential agents share a version of the same agential pathology.23 Such
agents are impervious to ordinary forms of critical dialogue that most of us
19
Seidman, Valuing and Caring, 296. Concern is Seidmans term for the mental state we ascribe
to an agent when we say that she cares about something. See Seidman, Valuing and Caring,282.
20
21
22
Andrea C. Westlund, Reply to Benson, Christman, Rocha, and Stoljar, Symposia on Race,
Gender, and Philosophy 7:1 (2011):16, http://web.mit.edu/sgrp.
23
Andrea C. Westlund, Selflessness and Responsibility for Self: Is Deference Compatible
with Autonomy? Philosophical Review 112:4 (2003): 483523; Westlund, Rethinking Relational
Autonomy,2649.
take for granted. We might aptly describe them as being in the grip of their
concerns, or perhaps as under the influence of their concerns. Acting under
the influence of a concern for which one does not hold oneself answerable is
intuitively incompatible with self-governance, even if it is consistent with the
demands of authenticity. In other words, autonomy and authenticity (at least
as the latter is usually conceived) may sometimes come apart.24
Jaworska seems to share this intuition, since she begins, in subsequent
work, to move away from a purely authenticity-based view.25 The autonomous
agent, she argues, guides her action in light of the reasons she sees (in virtue
of caring about certain things) to pursue or promote relevant ends. But some-
thing more is required:Our theory, she writes, must stave off the possibility
that seeing a reason would simply amount to being in the grip of the reason.26
Jaworska argues that autonomy requires what she calls mental freedom, and
that mental freedom is incompatible with being rigidly stuck in some in-
flexible emotional view or assessment of facts or reasons.27
I would agree with Jaworska that being mentally unfree is intui-
tively incompatible with autonomy and that being rigidly stuck in ones
identity-constituting concerns is incompatible with mental freedom. To be
stuck in a certain view of ones reasons (either for action or for emotion)
implies not just a benign stability or absence of change but also a problematic
inability to change. An agent unable effectively to revise or change her con-
cerns under any conditions seems intuitively unable to govern herself, even if
her choices and actions are determined by those very concerns. Perhaps she
has a self, but she does not seem to be doing much in the way of governing.
Of course, an inability to change ones view of ones reasons could be
explained in a variety of ways. One might lack certain critical reasoning or
other straightforwardly cognitive skills, and this would, plausibly, undermine
autonomy. Such skills are surely amongst what many philosophers now refer to
as competence conditions for autonomy. But straightforwardly cognitive skills
do not seem to be all that Jaworska has in mind. She claims, for example, that
the agent must be capable of imaginatively entertaining alternatives28 and
that she must be open to the possibility that further reflection would change
her view of what reasons she has. By reflection, Jaworska clarifies, she does not
24
Westlund, Reply to Benson, Christman, Rocha, and Stoljar. Later in the paper Iconsider a dif-
ferent conception of authenticity, proposed by Marina Oshana, which would be more closely aligned
with autonomy as Iunderstandit.
25
Jaworska, Caring, Minimal Autonomy, and the Limits of Liberalism.
26
Jaworska, Caring, Minimal Autonomy, and the Limits of Liberalism,95.
27
Jaworska, Caring, Minimal Autonomy, and the Limits of Liberalism,97.
28
Jaworska, Caring, Minimal Autonomy, and the Limits of Liberalism. The ability to imagine
oneself or ones views as other than they are is a competence condition on autonomy that has been con-
vincingly defended by Catriona Mackenzie, Imagining Oneself Otherwise, in Mackenzie and Stoljar,
Relational Autonomy; Diana Meyers, Self, Society, and Personal Choice (NewYork:Columbia University
Press, 1989), 259279.
mean Frankfurtian, higher-order reflection on lower-order motivational states

but direct reflection on considerations relevant to the choice of a particular
course of action. To have mental freedom, the agent must not only be capable
of such reflection but must also be open to the possibility that engaging (or
re-engaging) in it would change what she sees as a reason.
What does it mean to be open to this possibility? Merely acknowledging
that reflection could, in principle, lead one to embrace an alternative perspec-
tive is not enough. For one might be unshakably indifferent to that possibility
or even adamantly (perhaps unconsciously) determined not to engage in re-
flection, precisely because it could lead to some uncomfortable change. True
openness of the sort Jaworska has in mind seems to include some readiness
to engage (under at least some conditions) in reflection that is understood, by
the agent, to be potentially transformative. One intuitively relevantand com-
moncondition calling for engagement is that in which one finds ones con-
cerns, or the actions they motivate, challenged (either directly or indirectly)
by the points of view of others. One is frequently called upon to answer for
oneself, or to account for ones motivating concerns, in response to the con-
cerns of others.
Thus understood, the attitude that Jaworska takes to be required for mental
freedom is very close to what Irefer to as an attitude of responsibility for self
or self-answerability, an attitude that Itake to be required for autonomy.29 But
Ithink the inclusion of this condition represents a significant departure from
the authenticity-based model that Jaworska otherwise seems to embrace and
that it points to a further dispositionbeyond the emotional and cognitive
dispositions that she and Seidman identifythat is characteristic of auton-
omous agency. To be appropriately open-minded, it seems to me, one must
have a disposition to answer for ones concerns when called upon to do so,
unless there are defeating conditions that render the call itself, or an effort to
answer it, inappropriate in the circumstances.30 Caring is partly constituted
by a disposition to believe one has reasons to feel and act in certain ways,
but autonomy (insofar as it requires mental freedom) requires that we also
be disposed to suspend such beliefs, or at least hold them at arms lengthin
dialogically appropriate contexts.
Of course, a too easy surrender of the beliefs one is disposed to hold is
not intuitively compatible with autonomy, either. A social chameleon who
abandons and replaces her identity-constituting concerns anytime she comes
under pressure from without hardly seems to exercise enough control over
her concerns to count as self-governing. Mental freedom may be incompatible
29
Westlund, Selflessness and Responsibility for Self ; Westlund, Rethinking Relational
Autonomy.
30
Idiscuss some defeating conditions in Westlund, Rethinking Relational Autonomy.
with being rigidly stuck, but it seems equally incompatible with a complete
lack of stability in ones concerns.
Fortunately, readiness to engage in potentially transformative reflection
does not imply readiness to capitulate at the least provocation. Our attempts
to answer for our concerns often, at least at first, take the form of self-defense
or self-advocacy. When faced with a challenge, we tend to want to commu-
nicate the value of what we take to be important, and the appropriateness of
giving it the place we do in our deliberative life. Though it is relatively easy to
veer into excessive defensiveness, there is a principled reason for self-advocacy
as a default strategy:we cannot be willing to give up our cares at the drop of
a hat, for this is incompatible with truly caring about the things we think we
care about. This imperative reflects an important kernel of truth in Frankfurts
thesis that care is a commitment of the will: in its desiderative dimension,
care requires more than simply having a desire (even a very strong one) and
more also than wholeheartedly endorsing that desire. It also means wanting to
continue to have the desire, even when one decides that it would be better, all
things considered, not to act on it. The impulse to self-advocacy reflects this
stabilizing dimension ofcare.
At the same time, though, for the agent genuinely to be mentally free, the
impulse to advocate for oneself must be moderated by an appropriate atti-
tude of humility. One must recognize that it is sometimes appropriate for
self-advocacy to give way to self-scrutiny and be prepared to let self-scrutiny
take its potentially transformative course when warranted. If one is not pre-
pared to do this, one is (once again) rigidly stuck in a way that cannot amount
to self-governance. It might help to think of the autonomous agent as possessed
of certain characteristic virtues, in a broadly Aristotelian sense:she must avoid
both an excess and a defect of answerability and openness and must know how
to be open and answerable to the right degree and in the right circumstances.
3.Self-care
In this section, Idevelop a conception of mental freedom as entailing an at-

titude of self-care. The conception of self-care Idefend is inspired in part by
Carol Gilligans vision of moral maturity, as laid out in her classic work In a
Different Voice, and in part by Michel Foucaults notion of care of the self,
as developed in some of his later writings. This might seem like an unlikely
pair of influences for a view of autonomy. It is not clear that either Gilligan or
Foucault even accepts autonomy as a concept in good standing, nor do they
have much theoretically in common with one another. But the view Idefend
is only loosely derived from certain of their ideas, which are (admittedly)
taken out of the context of their wider bodies of work. Ultimately, Iargue that
self-care is a higher-order attitude of care (a compound of emotions, desires,
and dispositions) focused on the first-order cares that Jaworska and Seidman
take to make up the agents standpoint orself.
Gilligan is best known for her path-breaking and controversial views about
gender and moral development. But the aspect of her work that is most rel-
evant to my argument has been much less widely discussed:it is the picture
she paints of moral maturity and, in particular, of the morally mature agents
relationship to her own choices. Many of the troubled girls and women she
interviews, particularly in her chapter on abortion decisions, seem at first to
speak from a position of compromised autonomy. They feel helplessly con-
strained by their understanding of what others want from them and often de-
scribe themselves as having no choice about what to do when faced with a
moral dilemma.31 Gilligan points out that, while the feeling of powerlessness
is very real for these women, it stems in large part from a confused abdication
of responsibility for their own responses to their situations. Instead of seeing
themselves as agents with a decision to make, they feel pushed and pulled by
the demands of others and can assert themselves in only relatively inchoate
ways. They do not see themselves as choosing whattodo.
But a number of Gilligans interviewees have transitioned out of this stage
of moral development, and what they say about the experience of coming to
terms with their agency is very interesting. Several of her subjects, for example,
dwell on the importance of taking care (or, as we sometimes say, taking pains)
with their decisions. One subject, for example, claims that when deciding what
to do one must take a critical view and be as conscious or awake as possible,
to consider all thats involved.32 Another stresses the importance of being
conscious of ones power or influence as a chooser and taking responsibility
for the consequences of ones choices, both for oneself and for others.33 Athird
emphasizes coming to terms with the fact that not all conflicts have a tidy or
obviously right resolution and that one must exercise judgment to the best of
ones ability and accept responsibility for the consequences of doing so.34 In
these and other similar passages, a common theme begins to emerge:moral
maturity requires recognition of ones power to influence outcomes by exercis-
ing choice, acceptance of what we might (in Rawlsian terms) call the burdens
of judgment, and willingness to take responsibility for exercising judgment in
31
For these women, any course of action other than the self-sacrificing one feels unthinkable and
the chosen course of action feels not so much chosen as inevitable. Notice, however, that the passivity
they experience does not seem to be an instance of Frankfurtian alienation from their motives. Though
Gilligans subjects are distressed by their situations, many of them do seem to identify with their self-
lessness, understood in terms of an overriding imperative to maintain relationships and serve others
needs. The problem, as Igo on to emphasize, is that they do not see themselves as choosing what to do,
even when they act in accordance with this imperative.
32
Gilligan, In a Different Voice,99.
33
34
such conditions. The essence of moral decision, Gilligan concludes, is the

exercise of choice and the willingness to accept responsibility for that choice.35
I want to argue that taking care with ones decisions, and accepting respon-
sibility for ones judgments, is in fact a form of self-care. This might sound
like an odd claim. But think of the way Foucault uses the phrase care of the
self in his later writings about ancient Greco-Roman culture.36 He does not,
by this phrase, mean concern for ones material or bodily well-being or ones
own interests or desires, at least not in any straightforward sense. The care
he describes is targeted instead at the soulmeaning, as Foucault empha-
sizes, the activity of the soul, and not the soul-as-substance.37 Foucault shows
how the ancients exhibited care of the self-as-soul through various activities
of self-examination, self-articulation, and self-management. Of particular
interest to Foucault is their engagement in various forms of self-writing, in
which they both recorded their own thoughts and actions and drew on pub-
licly accepted discourses to monitor and shape their principles of choice and
action. Greco-Roman self-writers took notes on themselves, compiled frag-
ments of advice and wisdom from public sources (sometimes to be shared),
narrated their activities in correspondences with friends, and, in some cases,
produced quasi-confessional journals that were private but explicitly intended
to encourage the kind of self-examination and self-correction that is otherwise
prompted by the critical gaze of others.
Foucaults discussion of these practices brings into focus a notion of
self-care that bears a clear connection to answerability and responsibility for
self. Foucault, tellingly, uses not only the phrase care of the self but also
government of oneself to describe the practices to which these self-writing
activities contributed.38 Like some of Gilligans subjects, Foucaults self-writers
manifest strong concern for their own agency. In reflecting on and monitoring
their principles of choice and action, they take pains over the practical per-
spectives or standpoints that define them as subjects. They care, in a practical
or volitional sense, about who they are, and this care clearly involves a sense of
responsibility, or answerability, for their deliberative activities.
Care of the self in the practical or volitional sense is clearly just one thing
that we might have in mind when we speak of self-care, and Ido not mean it
to subsume or replace all other uses of the phrase. Concern for ones material
or bodily well-being, or for ones own interests or welfare, are also forms of
self-care (arguably more familiar forms), and there is nothing inherently sus-
pect or problematic about these attitudes. Gilligan, in fact, argues that the mor-
ally mature agent must treat herself as a proper object of care, in what Itake to
35
36
Michel Foucault, Ethics:Subjectivity and Truth (NewYork:New Press,1994).
37
Foucault, Ethics,231.
38
be one of these alternative senses. She argues that to achieve moral adulthood
a woman must shift from a perspective of selfless goodness to one that empha-
sizes truthby which she seems to mean honesty about ones own needs and
desires, along with recognition of oneself as a proper object of caring attitudes.
The morally mature agent cares for herself (among other things) in the sense
that she treats her own interests as worthy of consideration, and she will seek
to minimize hurt to all involved, not just to others at the expense of herself.
But notice that the agent who cares for herself in this sensetreating her
own interests as worthy of considerationwill still face the challenge of inte-
grating her various concerns with one another and exercising judgment with
respect to their apparent claims. Even if one thinks ones own interests and
needs matter and must be taken into account, one can easily imagine being
paralyzed in cases of conflict or being rigidly stuck in some particular con-
ception of how competing claims should be handled. (Consider a well-known
example from Gilligans own research. When asked how to resolve a conflict
between self and other, eleven-year-old Jake doesnt miss a beat:You go about
one-fourth to the others and three-fourths to yourself.)39 Gilligans most
thoughtful interviewees, and Foucaults self-writers, recognize and thematize
the need to exercise judgment and to take responsibility for doing so. Their
attitudes, Isuggest, point to a distinctive kind of self-care that is focused on the
self in its deliberative or practicalmode.
What is it, then, to care for oneself in the practical sense? If we accept some-
thing like the JaworskaSeidman conception of care, then care of the self, like
care of anything else, will be constituted by a complex of emotions, desires,
and cognitive and emotional dispositions that are bound together by a percep-
tion of their focal object as a source of practical reasons. In this case, instead
of being focused on other persons, objects, or ideals, the attitudes involved
in self-care will be focused on elements of the caring agents own delibera-
tive perspective or standpoint and will be bound together by a perception of
that standpoint itself as a source of reasons (for the agent). Suppose we also
accept that ones standpoint is constituted by ones first-order cares (though
not, to be clear, the further idea that having ones choices determined by such
a standpoint suffices for autonomy). Self-care, then, would turn out to be a
higher-order attitude of care focused on the first-order attitudes of care that
make up ones subjective practical perspective. Caring about oneself would
mean, in a manner of speaking, caring about onescares.
One might object that while everyone cares about what she cares about (this
is a mere tautology), not everyone cares about herself. So the definition of
self-care Ive just offered must be wrong. But my claim is not that the self-caring
person, call her S, cares about herself simply in virtue of caring about X, Y, and
39
Z, things that (by hypothesis) she happens to care about anyhow. It is instead
that S cares about herself (in the sense at issue) if and only if she cares about
her caring attitudes: she has certain higher-order attitudes toward, and dis-
positions concerning, the clusters of emotions, desires, and dispositions that
constitute her first-order concerns for X, Y,andZ.
Even this claim could be interpreted in more than one way. The relevant
question might, for example, just seem to be whether ones emotions are apt
and well-proportioned, given that one sees X as important, or whether one
actually has the dispositions and desires that one who cares about X should
have. These questions take for granted the focus of the care and address the
quality of ones attitudes toward that object, targeting the possible discrepancy
between what one thinks one cares about and what one really does care about.
It opens up a space for assessing someones care as defective on its own terms,
and it makes possible the judgment that, if someones attitudes are defective
enough, they might not qualify as care atall.
This might be part of whats involved in care of the self, in the loosely
Foucaultian sense Ive been putting forth. But this interpretation of what it
is to care about caring doesnt yet address Jaworskas concern about mental
freedom. For even if someone cares about X badly rather than welleven if
she cares so badly that we are tempted to conclude that she doesnt truly care
about X at allit is possible for that person to be rigidly stuck in her view of
X as a (purported) source of reasons. But one cannot be rigidly stuck in such a
view of X if one is also concerned about that very way of viewing it, and this is
surely part of what it is to care about ones cares. Caring about ones first-order
concerns includes caring not just about the fit between ones emotions and
dispositions and what one sees as important but also about the question of
what one sees as important itself. More simply put, the agent who cares about
herself, in the sense Iwant to isolate, cares about her reasons.
So what is it to care about ones reasons? Itake it that a practical reason is
a consideration that favors some particular course of action or emotional re-
sponse on the part of the agent to whom it applies and that to see something
as a reason is likewise to see it as favoring some action or emotion. Favoring,
of course, is not an ideally perspicuous notion, but Ithink that a reason can be
(slightly) more precisely characterized as a consideration that may enter into a
viable defense or justification of a course of action or response. One who cares
about her reasons, then, is subject to a constellation of emotions, desires, and
dispositions focused on considerations she sees in this light, attitudes that are
jointly explained (following Seidman) by her perception of those consider-
ations as important. One who cares about her care for X, and thus about her
X-based reasons, sees her care for X as important and as the source of a dis-
tinctive set of reasons.
When an agent cares about her care for X, she will want to protect that
care from being too easily eroded. She will not, however, want to maintain
her perception of Xs importance (and her related sense of reasons) at any cost.
As Seidman points out, the kind of reasons one takes oneself to have, in virtue of
an objects perceived importance, will depend on the kind of thing that object is.
Apractical reason, Ive claimed, is a consideration that can enter into a viable de-
fense or justification of a particular action or emotional response. Ipropose that
one who cares about her reasons must be attuned to the suitability of her apparent
reasons for playing this justificatory role. More intuitively speaking, part of what
it is to care about our concerns is to be concerned for their aptness. Think of it this
way:only someone who doesnt care what she cares about could be untroubled
by a charge that a given care has been misplaced. Caring about ones concerns
counterbalances, to some degree, the cognitive disposition to take appearances of
reasons at face value. Perhaps more accurately, it defines some of the conditions
under which that disposition may be defeated.
One who does care what she cares about will be heartened by evidence that
supports her perception of X as important and troubled by signs that this per-
ception may instead be misleading. In the face of challenges to her X-based rea-
sons, she will be disposed either to advocate on their behalf or to revise or reject
them, depending on the perceived merit of the challenge. One who cares about
her reasons will desire that she see as reasons only things that are well suited to
playing their intended justificatory role, will have positive and negative emotions
in response to evidence that this is or is not the case, andperhaps most impor-
tantly, for our purposeswill be disposed to be rationally responsive to critical
perspectives on her (purported) reasons. In this way, she takes responsibility for
her reasons. Let us call this self-relation one of practical self-care.
4. Self-care and Relational Autonomy
Thus far Ihave followed Jaworska in treating mental freedom as a condition

of autonomy and have argued, further, that mental freedom requires practical
self-care. Does this picture have any bearing on the feminist point that au-
tonomy must be understood as relational? Self-care is clearly a self-relation.
But standing in some sort of relation to oneself is not, at least at first glance,
the kind of relationality that relational theorists of autonomy have in mind.
Relational conceptions of autonomy are meant, first and foremost, to establish
the role played by social relations in autonomous agency.
Some assign to social relations a purely causal role in the development of au-
tonomy competencies. Others regard social relations as, in one way or another,
partly constitutive of autonomy. Ifall into the latter camp, having argued else-
where that responsibility for self is a dialogical disposition that requires seeing
oneself always as potentially part of a deliberative dyad.40 The self-responsible
40
Westlund, Rethinking Relational Autonomy.
agent holds herself to an expectation or demand that she respond appropriately

to legitimate requests for justification, experiencing herself as owing a suitable
response, at least under certain conditions. Answerability for oneself is, on
my view, a special kind of (normative) competence condition on autonomy,
resembling what Stephen Darwall calls second-personal competence.41
Here Ive been arguing that the autonomous agent must care about her rea-
sons and that caring about ones reasons opens one up to potentially trans-
formative deliberative engagement with others. That autonomy should be so
closely associated with openness to changeand particularly with change that
is instigated by othersmay sound counterintuitive, since we have become so
accustomed to associating autonomy with authenticity, and authenticity with
the idea of being true to oneself. But Marina Oshana argues convincingly that
the form of authenticity most relevant to autonomy might involve something
more like honesty with oneself.42 (Gilligan might be pursuing a similar intui-
tion when she describes the transition to moral maturity in terms of a commit-
ment to truth.) Being honest with oneself about ones reasons is a way of being
honest about who one is, but not one that will always leave the self unchanged.
Whereas changing who one is in response to brute physical or psychological
force, browbeating, or emotional manipulation would obviously be incompat-
ible with autonomy, adjusting ones practical perspective or agential standpoint
in response to reasons given by others is another matter. Self-government does
not intuitively require insulation from this kind of external influence; even
pretheoretically, self-government would seem likely to require honesty about
ones reasons, and honesty about ones reasons would in turn seem to require
sensitivity and appropriate responsiveness to critical perspectives.
One might object that it is not always incumbent upon us to care what oth-
ers think of our commitments and, moreover, that sometimes we positively
should not care what others think. Adevoted parent, for example, need not be
open to abandoning or substantially modifying his commitment to his child,
and he appears lacking neither in self-care nor in autonomy when he refuses
to entertain doubts about the aptness of his care. Ido not dispute this claim,
at least at the level of abstraction at which it is here posed. Most of us would,
if pressed, have at least something to say in defense of caring for our children,
but a challenger to whom we actually had to articulate such a defense would
strike us as obtuse to the point of irrelevance. It is hard to imagine a chal-
lenge to parental care (as opposed, say, to a criticism of a particular way of
41
See Darwall, Second Person Standpoint. I have more to say about the relationship between
self-answerability and second-personal competence in Andrea C. Westlund, Autonomy, Authority,
and Answerability, Jurisprudence 2:1 (2011):161179.
42
Marina Oshana, Autonomy and the Question of Authenticity, Social Theory and Practice 33:3
(2007):411429. Oshana calls this an epistemic conception of authenticity. See Oshana, Autonomy
and the Question of Authenticity,412.
interpreting or enacting parental care) that would, in ordinary circumstances,

warrant sustained attention and scrutiny. Not every challenge is a serious chal-
lenge, and the agent who is appropriately responsive to critical perspectives
will recognize this fact. Of course, our assessments of seriousness are them-
selves fallible, and one can no more avoid the burdens of judgment in this
domain than in anyother.
It is worth emphasizing that, even where one takes a challenge to be serious,
self-care does not always lead to self-transformation. Imaintained earlier that
answering for oneself does not always mean acceding to the claims of others
and that the autonomous agent knows when to advocate for as well as when
to scrutinize and revise her commitments (or at least she exercises good judg-
ment with respect to such matters and takes responsibility for doing so). The
relation of self-care, as Ive described it, embodies this complexity. There is a
kind of built-in stability in the attitude of caring for ones reasons, since caring
about ones reasons means responding to them as important and not simply to
be trifled with. But stability is not rigidity, and what saves it from rigidity is a
degree of flexibility.
If Iam right, this flexibility is a product of our (appropriately moderated)
openness to the perspectives of others. The disposition to answer for oneself
is perhaps best characterized as a disposition to be engaged by other practical
reasoners, with their own take on what matters, in a kind of shared delibera-
tion regarding what one has reason to do or feel. Though it would take further
argument to make this case, Isuspect that the relational core of Gilligans view
might usefully be cashed out in terms of an orientation toward such shared de-
liberation. The idea is meant to resonate with the social dimension of Foucaults
notion of care of the self as well. For Foucault, care of the self...implies a re-
lationship with the other insofar as proper care of the self requires listening to
the lessons of a master. One needs a guide, a counselor, a friend, someone who
will be truthful with you.43 Even the more private forms of self-writing seem,
in Foucaults rendering, to depend on a model of interpersonal engagement
and responsiveness.44
So, Icontend, there is at least one important respect in which autonomy is
relational in an interpersonal (and not purely intrapersonal) sense. Indeed, on
the view Ive been defending in this paper, the relationality of autonomy and
the self-regarding attitudes on which it depends are nonaccidentally bound up
with one another. Self-care and self-responsibility are two facets of the same
self-relation, involving the same normative and dialogical competence to an-
swer for oneself.
43
Foucault, Ethics, 219221.

44
Acknowledgments
I would like to thank Andrea Veltman and Mark Piper for their very helpful
comments on earlier drafts of this paper. Iam also grateful for valuable feed-
back from participants in the Relational Autonomy Workshop at McGill
University in the fall of 2012, as well as participants in the 2013 Summer Retreat
on Love and Human Agency at Lake Tahoe. Ialso benefitted a great deal from
discussion with audiences at my Brady Lectures at Northwestern University in
the spring of 2013, and Ithank the Brady Program in Ethics and Civic Life at
Northwestern University for its support during the time in which Idid much
of my work on thispaper.
{ PartIV }
Autonomy, Oppression, and Adaptive

Preferences
{10}
Coping or Oppression:Autonomy and Adaptation

to Circumstance
John Christman
Feminists have been rightly concerned with questions surrounding the agency
status of people who appear to adapt to oppressive circumstances by alter-
ing their character in ways that seem to internalize that oppression. Much
fruitful discussion has taken place about those people, women in particular,
who seem to accept (or even claim to value) life situations that appear to the
critical observer to be stultifying and dominating and in that way incompat-
ible with autonomous agency.1 This vexing problem focuses our attention on
the seeming conflict in feminist sensibilities between, on the one hand, re-
specting the choices of actual women as well as being open to difference, and
on the other hand, decrying the oppressive social conditions that bear down
on women in thisway.
Interestingly, however, far less attention has been given to the ways that
people routinely reshape themselves in response to unforeseen and uncontrol-
lable circumstance in ways that sometimes undercut autonomy but at other
times do not. We have accidents, take risks that turn out badly in ways we are
unprepared for, undertake radically indeterminate projects such as having a
child or entering a relationship, and often fall victim to the decisions of others,
the operation of impersonal institutions, and simple natural events that deter-
mine the available course of our lives. Changes in life options can give rise to
changes in character, for example, learning to be accommodating to a partners
1
See, e.g., Paul Benson, Autonomy and Oppressive Socialization Social Theory and Practice 17
(1991):38408; Diana T. Meyers, Feminism and Womens Autonomy, in Gender in the Mirror:Cultural
Integrity and Womens Agency (New York: Oxford University Press, 2002), chap. 1; Marina Oshana,
Personal Autonomy in Society (Aldershot, UK:Ashgate, 2006); Marilyn Friedman, Autonomy, Gender,
Politics (NewYork:Oxford University Press, 2003); Anita Superson, Deformed Desires and Informed
Desire Tests, Hypatia 20:4 (2005):109126; Uma Narayama, Minds of Their Own:Choices, Autonomy,
Cultural Practices, and Other Women, in A Mind of Ones Own: Feminist Essays on Reason and
Objectivity, edited by Louise M. Antony and Charlotte E. Witt (Boulder, CO:Westview Press, 2001);
Andrea Westlund, Selflessness and Responsibility for Self:Is Deference Compatible with Autonomy?
Philosophical Review 112:4 (October 2003):483523.
unexpected habits or training ourselves to like childrens games to play with

our children.
Among these benign cases of adjustment to lifes challenges are instances where
the process of adaptation is more extreme:for example, a person has a severe ac-
cident or has a special needs child or an unplanned pregnancy, which completely
alters her life trajectory. These are cases where relatively dramatic adjustments to
a persons life ambitions and even sense of herself must take place, yet in at least
some such cases the person retains or regains her autonomy. The interesting
question, then, is what exactly marks off the cases of agency-undermining adap-
tation from those where self-government is maintained.2
Consider, for example, two cases. The first involvesAbby.
Abby was always an athletically active girl and young woman. She was a runner and
even trained for marathons on more than one occasion. One of her dreams was to
run in the Boston marathon because she grew up near there. She apparently dis-
played the characteristics typical of a self-governed life. At the age of twenty-five,
however, she was in a very serious car accident that left her permanently paralyzed
from the waist down. As a result she was to be wheelchair-bound for the rest of her
life. While completely despondent for a time, eventually Abby came to accept her
condition. She continued to exercise, though she became less interested in outdoor
activity. Instead, she picked a long-abandoned interest in writing and came to de-
vote herself to fiction. After a time she became an accomplished short story writer
andpoet.
The second case is that of Kaew, a modest and demure Thai woman who at the
age of twenty-two realized that to be a good provider for her family she had to look
for employment where she could find it. As a result, she agreed to go to Japan to
work at what she thought was a factory. When she arrived she was told she had to
work at a snack bar (a bar and brothel) and that she had to pay off an exorbitant
debt for the transport from Thailand. She worked at the bar and repaid the debt
after four years, serving clients daily without a break and under constant threat
of violence if she ever thought of attempting to leave. However, when her debt was
paid off, she stayed on at the bar and continued to work as a prostitute, sending
money back to her family. Eventually, she became one of the managers of the bar
and facilitated the trafficking and sex work of other young women.3
In both cases the person changes radically over time in response to severely
altered (and constrained) circumstance. Both individuals abandon earlier
2
Aclassic discussion of this issue is in Jon Elster, Sour Grapes (Cambridge:Cambridge University
Press,1983).
3
In both these cases Idiscuss only the subject herself, but of course the broader social and inter-
personal network within which these events and processes take place is crucial to our understanding
of the case. Imention such factors in the upcoming discussion. Also, all of the cases described in this
paper are distillations of actual stories.
Coping or Oppression 203
ideals about how their lives would go; in both cases, they rearranged their (pre-
sumably self-directed) projects and plans, indeed their senses of themselves,
in response to changes in life options that they did not choose and would not
have chosen. While of course there are numerous important differences be-
tween the cases, it is striking that we may have no hesitation in responding to
the first with admiration and the second with pity and perhaps disdain. More
to the point, we would not think that anything about the story of Abby (as told
here) indicates a lack of autonomy after her adjustment, but we may well see
Kaew as a straightforward case of the loss of autonomy or at least as problem-
atic. What account of autonomy can be given that picks out the key differences
between thesecases?
I will say more about these and related cases as we proceed. My plan here is
to examine a variety of approaches to diagnosing when alterations in character
and values mark a loss of autonomy and when such alterations count merely
as a healthy adaptation to constrained circumstances. After critically consid-
ering these approaches Iwill try to defend a broadly procedural approach to
self-government that, Iwill suggest, can account for processes of adaptation
of the sort being considered. However, my point is not to fully defend such an
account of autonomy but merely to refine this approach to include a condi-
tion that Iwill call reflexive self-affirmation, a notion akin to what some other
theorists have insisted on under the guise of self-worth and self-trust, though
not exactly in the form they have doneso.
One more important preliminary:the cases Idescribe are roughly drawn
to be sure, but more importantly judgments about these cases will perforce
be made from the equally underdescribed vantage point of detached philo-
sophical analysis. This way of proceeding, it must be emphasized, leaves out
two important factors in actual determinations of peoples self-government (or
lack thereof):one is their own voices in reporting their condition (and their
judgments about it); and another is a specification of the contexts in which
attributions of autonomy are in fact made, including the effect that those attri-
butions might have on the autonomy of the agent in question. Indeed, in this
latter category we should include reference to the dynamic interaction be-
tween those ascribing autonomy (as a classification) and the persons whose
autonomy is at issue. The contexts where such judgments are made, expressed,
and perhaps filed in official documents may have a direct effect on the state
of the person herself relative to her self-government. In my conclusion Iwill
return to these neglected aspects of such cases and discuss how attributions of
autonomy involve dynamic exchanges between observers and observed (or,
e.g., aid workers and clients, officials, and victim/survivors).4
4
It is also manifestly true that the picture of a person like Abby (and otherssee following) woe-
fully underdescribes the conditions of disability. See, e.g., Carolyn Ells, Lessons about Autonomy from
the Experience of Disability, Social Theory and Practice 27:4 (October 2001):599615.
1. The Range of Cases and Asymmetries of Judgment
The contrasts between Abby and Kaew are numerous, but my point was not
to draw any quick conclusions from our reactions to them. However, if we lay
out a certain array of types of cases of this sort, Ithink we will see a certain
asymmetry in our reaction that will prove instructive in refining our theoret-
ical accounts of self-government.
Consider two variations on the previouscases:
Bernice was also an active, athletic person who relished physical activity and had
aspirations similar to Abbys. She also had an accident that left her permanently
disabled. But unlike Abby, she never adjusted to her life in a wheelchair; she turned
to alcohol and never gave up the intense pain medication she was prescribed. She
was also bitter and resentful at losing out on the life she had always imagined.
At age forty she has now isolated herself from friends and support networks that
would have attempted to steer her into a lifestyle commensurate with her capaci-
ties. She is, in a word, consumed by resentment.
Parallel to this is the case of Irina, who like Kaew was forced into prostitution
and debt bondage. Like Kaew she was forced to remain in this life for an extended
period. Irina, however, never stopped resisting her condition. She always held a
fierce, though necessarily concealed, hatred of her captives and her life. She looked
for any means of escape, making complex and elaborate plans to hide money away,
venture out from the bar, and hopefully befriend a sympathetic and resourceful
friend who could help her escape. Alas, however, during the time we are consid-
ering here, she never foundone.
All of these cases are, again, terribly underdescribed.5 And because of their
brevity these descriptions cannot support very confident intuitive judgments
about how to classify these individuals in terms of their status as self-governing
agents. But my aim is to point out a general asymmetry that can be noticed,
I think, in our reactions to them that will motivate our theoretical discus-
sion:in the case of the two accident victims, it appears that acceptance, adjust-
ment, and adaptation in the case of Abby points toward seeing her as a case of
recovery of self-government, while Bernices resistance and refusal to adjust
points away from that judgment.6 But in the cases of Irina and Kaew the di-
rection of classification seems to be reversed. At first glance, at least, it seems
that that Irina retains a modicum of self-government just because she resists
5
One assumption about all four cases that should be mentioned is that all these individuals are
relatively autonomous when the events in the stories begin to unfold.
6
Again, Iam not saying with any confidence that we would agree that Bernice lacks autonomy
much more needs to be known about her to saybut clearly her resentment and resistance to her con-
dition make her less self-governing thanAbby.
and resents her constraining condition, while Kaew might be seen as losing
her autonomy just because she has adapted to her condition and altered her
identity in responsetoit.
The array of cases of this sort is variegated in innumerable ways. However,
Iwant to point to two dimensions of contrast that suggest dividing them up
into four categories. First, there is adaptation or not. That is, Abby and Kaew
adapted to their circumstance in the sense that at least some aspects of their
identity and value scheme changed as a result of their constraining circum-
stances.7 On the other hand, Bernice and Irina did not adapt to their condi-
tion; they remained resentful and resistant, or so Iam describingthem.
Second, the sources of constraint are radically different in the two sets of
cases. In one, Iam stipulating that the events that gave rise to the constraint are
normatively neutral:the accidents that caused Abby and Bernices disabilities
were not anyones fault, nor can they be described as unjust in any straight-
forward sense. In contrast, the conditions of Kaew and Irina are manifestly
unjust, and oppressive as Iuse theterm.
What is interesting, and what Itake now as the pivot point for the remainder
of this discussion, is that our reaction to whether the persons self-government
is retained vis--vis her adaptation (or not) shifts depending on whether the
source of the constraining condition is unjust (or not). Or to put things the
other way around, our judgment about whether she is self-governing vis--
vis oppressive circumstances shifts depending on whether she adapts or not.
What Iwill argue, however, is that it is not the oppressive nature of the cir-
cumstances itself that determines this asymmetrical reaction; rather, it is the
particular ways that radically constrained life options operate on the capacity
for self-government and self-affirming practical identity of the agent herself.
Before discussing various approaches to these issues, it might be helpful
to set the stage with a brief discussion of the model of agency and practical
reason Iam relying on here and, by extension, what we mean by adaptation.
Autonomy involves basic skills and competences related to choice, delibera-
tion, action, and interaction. Several accounts of such skills have been given,
and clearly a broad range of competences is necessary for deliberate choice and
action to take place.8 Indeed, some theorists claim that autonomy requires only
competency skills, ones involving introspection, communication, memory,
imagination, analytical reasoning, self-nurturing, and resistance to pressures
7
Ilabel Abbys circumstance constraining for illustrative purposes. When adaptation is complete,
and especially when the physical and social environment is sufficiently accommodating, the word con-
straining may not be an apt descriptor for her life situation.
8
See, e.g., Diana Meyers, Self, Society and Personal Choice (NewYork:Columbia University Press,
1989); Paul Benson, Feminist Intuitions and the Normative Substance of Autonomy, in Personal
Autonomy:New Essays on Personal Autonomy and Its Role in Contemporary Moral Philosophy, edited
by James Stacey Taylor (Cambridge:Cambridge University Press, 2005), 124142.
to conform.9 Others, however, have insisted that more is needed, in particular

to distinguish autonomous agency from mere practical rationality. Specifically,
it is claimed that autonomy requires the authenticity of the values and motives
involved in action.10
Moreover, it can be claimed that self-governing agents choose on the basis
of a value orientation that both justifies those choices and orients their un-
derstanding of the options they face and the relative importance of the con-
siderations bearing on action. The most basic set of commitments and values
that structure deliberation and choice in this way can be called our practical
identity, following Christine Korsgaard.11 Such identities guide choice but, as
Isaid, also orient our perception of the options before us. Practical identities
function over time to organize both an ongoing life narrative and memories of
past decisions and experiences. We always think as a certain kind of person,
even if that identity is necessarily inchoate, subject to alteration and renegotia-
tion, and perhaps never fully transparent to our introspective reflection. Even
for trivial and quotidian choices, the way we perceive and evaluate the options
open to us reflects our practical orientation to the world, and this perspective,
in turn, reflects our social identities to some degree (though of course the ex-
tent of the salience and functional force of such identities varies significantly).
This way of understanding practical reason in general, and autonomy in
particular, is certainly not uncontroversial. But here Imerely want to identify
components of self-governing choice to fix ideas of interest. For in the cases
we are imagining here, what we are calling adaptation involves fundamental
shifts in key aspects of persons practical identities, where they are forced by
circumstance to renegotiate their sense of themselves, their value priorities,
and their plans and projects. Over sufficient time, some of the most basic (even
self-defining) values of the persons we are considering have faded into the
background, hardly functioning to motivate thought or orient reflection. Abby
no longer much thinks about marathons, daily morning jogs, or exercises in
aerobic endurance of the sort that so preoccupied her in the past. Of course
in moments of nostalgic musing she thinks of the life she lost and perhaps
wistfully recalls wanting to train for her beloved Boston marathon. But this
is similar to the way we all reflect on lives we once wished for but that no
longer guide our choices. After a time, Abbeys pangs of longing have faded
and such imaginings dont figure at all in her motivational complex; she is
9
See, e.g., Meyers, Gender in the Mirror (NewYork:Oxford University Press, 2002),2021.
10
For an argument to this effect, see John Christman, The Politics of Persons:Individual Autonomy
and Socio-historical Selves (Cambridge:Cambridge University Press, 2009), chap.7.
11
Korsgaard, The Sources of Normativity (Cambridge: Cambridge University Press, 1996). Marilyn
Friedman also sees autonomy as relative to our basic values. See Friedman, Autonomy, Gender, Politics,
chap.1.
now completely overtaken by her writing routine, her public readings, literary
discussions, and various other of her new lifes passions.12
But Ifocus here, somewhat myopically perhaps, on the changes in these per-
sons practical identity as a mode of mental functioning, orienting perception
and deliberation and grounding evaluative reflection, in the hopes that a fuller
picture that includes bodily components would be consistent with this level of
description. When the circumstances radically changed for people like Abby
and Kaew, a process began that resulted in a reordering of their values and
priorities, culminating in a new way of looking at and experiencing theworld.
2. Adaptation and Autonomy
What we have before us is an initial puzzle, namely, that adaptation of ones

practical identity to new constraints suggests, in one case, that autonomy is
maintained (Abby) but that the very same adaptation in another case marks a
move toward heteronomy (Kaew). Moreover, resistance to adaptation suggests
the dilution of self-government in one case (Bernice) but not in another (Irina).
What approach to autonomy best helps us account for these asymmetries?
I divide responses to such cases into two categories: structural and pro-
cedural.13 Structural accounts make central reference to aspects of the social
circumstance of the agent, independent of her perspectival take on those ele-
ments. Procedural accounts view autonomy with reference only to the skills
and reflective dispositions of the agent herself from her perspective (perhaps
making reference to the processes by which such skills and reflections came
to be exercised). These are rough categories, but Iwill give particular exam-
ples presently. The key idea is that procedural accounts refer exclusively to the
capacities and perspective of the agent and mention structural aspects of their
condition or the content of their values only derivatively, as a means of filling
out the procedural conditions themselves.
12
It must be noted here, as many other theorists have emphasized, that the scenario Iam sketch-
ing also involves radical changes in forms of embodiment and that, insofar as intelligence, memory,
and identity are in large part constituted by somatic elements, Abbys bodily change has changed her-
self in quite literal ways (though Iam not claiming in metaphysical ways necessarilyshe is still the
same person in that sense). As many feminists have claimed, our bodies and their capabilities and
habits form who we are and help form our values and modes of reflection and action. See, e.g., Diana
Meyers, Decentralizing Autonomy: Five Faces of Selfhood, in Meyers, Being Yourself: Essays on
Identity, Action, and Social Life (Lanham, MD:Rowman and Littlefield, 2004); Catriona Mackenzie,
On Bodily Autonomy, in Handbook of Phenomenology and Medicine, edited by S. K. Tombs (The
Netherlands:Kluwer, 2001), 417439.
13
These classifications echo, though do not precisely mirror, distinctions in the literature between
externalist and internalist accounts. For discussion, see, e.g., Marina Oshana, Personal Autonomy and
Society, Canadian Journal of Philosophy 29:1 (Spring 1998):81102.
Some relational accounts of autonomy count as structural in my sense.

Marina Oshana, for example, requires not only procedural requirements for
autonomy but also external, sociorelational conditions. On her account, the
conditions of autonomy include various internalist requirements of the sort
procedural accounts feature, such as epistemic competence, rationality, and
procedural independence as well as certain normatively substantive condi-
tions such as self-respect. The specifically sociorelational conditions Oshana
lists include the following:the person enjoys social and psychological security
(others cannot deprive the person of de facto or de jure power and author-
ity characteristic of global autonomy); the person can pursue goals dif-
ferent from the goals of those who have influence and authority over her; the
person is not required to take responsibility for anothers needs unless rea-
sonably expected in light of her particular function; the person enjoys finan-
cial self-sufficiency adequate to maintain independence from others; and the
person is not deceived.14
I have discussed this view elsewhere,15 but Iam not concerned to be crit-
ical of it here as a plausible account of a kind of autonomy. It is an ideal of an
independent life that, in some contexts, could well serve as a component of the
requirements of social justice. Imention it, however, because the purely struc-
tural (sociorelational) elements Oshana mentions will not suffice to make the
distinctions in judgment we began with, in particular the contrasting reactions
to Abby and Bernice, on one hand, and to Kaew and Irina on the other. For
both pairs of individuals exist in similar sociorelational settings. Yet we view
their mode of adaptation to those settings as different in crucial ways. This
suggests that insofar as the structural, external conditions of the cases dont
explain the contrasts correctly, then these contrasts will best be drawn by the
procedural components of the account, a point to which Iwill return.
Other theorists insist on structural (nonprocedural) elements for autonomy
that refer to the interpersonal and social dynamics of agency but focus the
effect of such relations on the agency of the person, especially in regard to
vulnerabilities she may experience in exercising that agency. Consider, for ex-
ample, several views that all require that interpersonal recognition or acknowl-
edgment of the persons normative authority (or self-worth, or answerability)
is necessary for autonomy to obtain.16 Catriona Mackenzie is typical of these
14
Oshana, Personal Autonomy and Society,86.
15
Christman, Politics of Persons, chap.8.
16
These different accounts Ihave in mind are from Catriona Mackenzie, Relational Autonomy,
Normative Authority and Perfectionism, Journal of Social Philosophy 39:4 (Winter): 512533; Paul
Benson, Taking Ownership: Authority and Voice in Autonomous Agency, in Autonomy and the
Challenges to Liberalism:New Essays, edited by J. Christman and J. Anderson (Cambridge:Cambridge
University Press, 2005), 101126; Andrea Westlund, Selflessness and Responsibility for Self:Is Deference
Compatible with Autonomy? Philosophical Review 112:4 (October 2003):483523; Joel Anderson and
Axel Honneth, Autonomy, Vulnerability, Recognition, and Justice, in Autonomy and the Challenges
views in insisting that autonomy requires more than merely the individual en-
joyment of a sense of normative authority, more also than granting oneself that
authority; it also requires external recognition by others of that authority, in
addition to being competent and reflective. Mackenzie claims that an agents
sense of herself as having a rightful claim to normative authority...[is] based
on intersubjective recognition by others. In this way, an agents status as hav-
ing normative authority over her values and decisions (what others have called
self-trust) has both first-personal and relational aspects.17 For views like
this one, the question of whether adaptation involves a failure of autonomy in
a structural sense is more nebulous.
Let us consider, then, how adding such an interpersonal recognition require-
ment may help us in our differential responses to the cases we have been consid-
ering. Specifically, let us focus on cases of what we can call resistant slaves. What
Imean are people like Irina who never cave in to the oppressiveness of their
conditions either by losing their oppression-independent self-understanding
or by losing their will to resist, even though no (nonsuicidal) opportunities to
express that resistance afford themselves. There are any number of variations on
such cases and variations along several dimensions. For instance, slave narra-
tives from the American South describe people who never lose their desire to be
free, but within the seemingly permanent circumstances of servitude they make
elaborate plans and develop projects they can truly call their own, for example,
their efforts to avoid being sold away from their children and spouses.18 As just
one example, consider a description by one Charity Bowery, a slave in North
Carolina, of her plans to save money to buy her children from herowner:
From the time my first baby was born, Ialways set my heart upon buying
freedom for some of my children. Ithought it was more consequence to them
than to me; for Iwas old and used to being a slave. But mistress McKinley
wouldnt let me have my children. One after anotherone after anothershe
sold em away from me. Oh, how many times that woman broke my heart!19
Such stories include descriptions of elaborate and complex struggles to
achieve such important goals, and they abound in accounts of slave life, a life
that surely fails to meet the structural desiderata of views like Oshanas as well
as the interpersonal requirements of some relational accounts.
to Liberalism: New Essays, edited by John Christman and Joel Anderson (Cambridge: Cambridge
University Press), 12749.
17
Mackenzie, Relational Autonomy,514.
18
Most slaves had no ultimate power over these processes. But many had some, and even if they
had no ultimate sway over whether their wishes were carried out the fact that they maintained them
throughout their captivity is what is relevant here. See, e.g., John Blassingame, ed., Slave Testimony:Two
Centuries of Letters, Speeches, Interviews, and Autobiographies (Baton Rouge:Louisiana State University
Press,1977).
19
Blassingame, Slave Testimony,262.
In all such cases, two important factors remain in play. First, these agents
maintain a practical identity that is continuous with their pre- or nonoppres-
sion selves. Their value priorities and senses of themselves are not crushed by
their captivity, even if they are prevented from acting on them. Second, such
persons are without most if not all of the social conditions that many rela-
tional theorists require for autonomy. Iam not claiming that such people are in
any way fully autonomous, and externalist relational theories, such as Oshanas
and relational views like Mackenzies, may explain why not. But I do think
there is another sense of minimal autonomy that they do maintain, one that
importantly distinguishes them from those who are crushed by circumstance
into either internalizing the oppressive values structuring their domination or
losing completely any effective power to act on their own at all. My point so
far is merely that models of autonomy that stress structural injustice and those
that feature interpersonal recognition have not quite located that difference.
This last point may be too hasty, however, for we must ask further whether
and in what sense resistant slaves might experience the recognitional acknowl-
edgment that theorists such as Mackenzie and others say is necessary for au-
tonomy. This is a difficult question for many reasons. One reason is that it
may be difficult to discern what modes of recognition and respect are being
shown such people or what level of such respect they are recognizing, even in
cases where we may know a great deal about their situation and psychology.
Moreover, their situation picks up on an ambiguity in some recognitional ac-
counts, namely, whether acknowledgment of the normative authority of these
agents (for example) must be public or if it can be simply a kind of internal voice
that the persons in question can listen to and gain strength from. And if it must
be outwardly displayed, if signs and behaviors must show such recognition, who
must show it? The powerful people that have the most influence over the person
and her social environment? Or can it be similarly oppressed compatriots? If
the latter, or if (as some writers claim) the recognition can be a purely imagina-
tive one involving internalized voices by the agent herself, one wonders if such
recognition is too weak, in that even young children can experience a recogni-
tional acknowledgment from their dolls and invisible playmates in developing
a sense of themselves. And in other cases, recognition of deliberative capacities
is not yet deserved, no matter how much people claim it for themselves. Apre-
cocious eleven-year-old, for example, may claim full powers of independent
judgment for herself and even insist on being recognized as such, but others
may nevertheless recognize that such confidence is not quite merited asyet.20
Now many structural theorists have made it clear that internalized recogni-
tion, where a postulated community of compatriots that one engages with in
20
Paul Benson mentions this last point as a possible objection. See Benson, Taking Ownership,
n.63. Also, both Benson and other theorists in this literature have discussed the ambiguities Imention
imaginative dialogue can provide the kind of recognition of normative author-

ity being described.21 Charles Taylor states forthrightly that it can, in saying
that even the isolated artist who speaks a language that is not understood or
acknowledged by her age still engages in dialogic interaction with the voices
of other internalized listeners that gives her a sense of agency and effectance.22
But notice that the requirement that such internalized recognition is neces-
sary for self-government is that it is effective: it gives the agent the sense of
identity and self-worth needed to give force to her plans and endeavors. This
implies that what matters, in the end, is the individuals power of agency and
that whatever form or mode of social interaction that is necessary to maintain
such a power is required as a causal condition forit.
The idea of the resistant victim of oppression raises issues for other ac-
counts as well. For example, in the context of a discussion of informed desire
accounts of the good, Anita Superson analyzes the phenomenon of deformed
desires.23 On her view, having a sense of equal self-worth is required for us
to be able to say of a person that her desires are not deformeddesires, say,
to conform to oppressively patriarchal norms. In this discussion, she consid-
ers the account offered by Uma Narayan concerning the way that women (for
example) reflect a kind of authentic agency in negotiating their place in a (to
an observer) stultifying and oppressive environment. Narayan believes that
bargaining with patriarchy makes a womans deformed desires her own. She
understands Muslim womens choice to wear a veil as a bundle of elements,
some of which they want [for example, ones reflecting their commitments to
various aspects of their own religious, social, and communal identities] and
some of which they do not want, and where they lack the power to undo the
bundle so as to choose only those elements they want.24
This scenario parallels the picture of the strategizing and resistant slave:one
who lives within crushing and unjust constraints yet shows genuine signs of
agency and self-directedness. What Superson says about Narayan, however,
indicates that she would resist such a classification. In considering the case of
veiling by Muslim women, she writes:
It seems to me that the woman who chooses veiling after bargaining over
her conflicting desires, in part because she believes that she needs to send
here. See, e.g., Benson, Taking Ownership; Andrea Westlund, Replies, Symposia on Race, Gender,
Race, and Philosophy 7:1 (Winter 2011), http://sgrp.typepad.com/sgrp/winter-2011-symposium-macken-
zie-poltera-and-westlund-on-autonomy.html.
21
See, e.g., Mackenzie, Relational Autonomy; Andrea Westlund, Selflessness and Responsibility
for Self. See also Andrea Westlund, Rethinking Relational Autonomy, Hypatia 24:4:2649.
22
Charles Taylor, The Ethics of Authenticity (Cambridge, MA: Harvard University Press,
1991),3233.
23
Superson, Deformed Desires and Informed Desire Tests.
24
Narayan, A Mind of OnesOwn.
a message to men about modesty and propriety, lacks sufficient belief in

her self-worth to insist that men be the ones to change their attitudes to-
ward women who do not veil.... A person who exhibits full agency and
self-direction cannot choose to have desires that are coercive, ones making
her act in ways she otherwise would not. She cannot freely make these
desires her own, and if she has them, she cannot be fully self-directing.25
The word coercive in the second to last sentence is ambiguous, since calling
desires coercive may already assume a categorization that we are trying to dis-
cern. More to the point, however, Iconclude from Supersons observation here
that she simply doesnt believe that a woman can maintain a sense of self-worth
in negotiating conditions of this sort, which she (Superson) takes to be so
dehumanizing. She may well be right. But this shows that her criticism turns
on skepticism that the women actually have a sense of self-worth, not that
they cannot be called self-governing agents in such conditions even if they
did so. Iwont comment on women who wear traditional dress, but my read-
ing of many first-person reports of survivors of slavery and oppression leave
me little doubt that the women and men in many cases maintained a strong
sense of self-worth while struggling against such horrific constraints. Similar
judgments are made by other observers and participants concerning decisions
to adapt to traditional and arguably oppressive conditions, by making them
in some way part of ones identity.26 This is not to say that these observations
are obviously correct in the end, only that the possibility that they are correct
shows that neither relations of equal status nor overt recognition of ones status
(worth) by surrounding others is always required for authentic agency, which
is the crucial conceptualpoint.
One of the most detailed recent accounts of the phenomenon of adaptive
preferences has been undertaken by Serene Khader who is also highly crit-
ical of procedural (and in general, autonomy-based) attempts to define what
she calls inappropriately adaptive preferences (IAPs).27 Khader considers sev-
eral versions of procedural autonomy and argues in each case that they fail to
adequately identify IAPs. The version that comes closest to what Iwill defend
here is what she calls the capacity to live according to a life plan.28 She claims,
however, that those with IAPs do not uniquely fail to live according to an iden-
tifiable life plan, if indeed one could be articulated (which is doubtful). At least
25
Superson, Deformed Desires,423.
26
See, e.g., Saba Mahmoud, The Politics of Piety (Princeton, NJ:Princeton University Press,2005).
27
Serene Khader, Adaptive Preferences and Womens Empowerment (NewYork:Oxford University
Press, 2011). For her criticism of proceduralist views, see Khader, Adaptive Preferences and Procedural
Autonomy, Journal of Human Development and Capabilities 10:2 (July 2009):169187.
28
Khader, Adaptive Preferences and Procedural Autonomy, 177ff. See also Khader, Adaptive
Preferences and Womens Empowerment, chap.2. Khader cites Diana Meyers, Personal Autonomy and
the Paradox of Feminine Socialization, Journal of Philosophy 84:11 (1987):619628, as the source of the
view she discusses.
they do not fail to do this in any way different from how those with nonadap-
tive preferences might. Moreover, there is no single life story that uniquely
picks out the personal history against which reflective judgments about cur-
rent desires are to be made, both in the normal case and for those with IAPs.29
First, however, it is not clear that life stories framed by practical identities
are as inaccessible as Khader suggests, as most people can readily express the
kind of person they are, their basic values, and the way theyve developed in
light of these values. Giving reasons for our actions demands as much. More
importantly, however, Khader argues against using a notion of autonomy that
relies on maintaining a sense of self-esteem. She argues that many who adapt
their preferences to conform to oppressive circumstances do not lack global
self-esteem in any sense. It is implausible, she writes, that most persons
with APs think they are unworthy human beings who cannot make claims on
others.30
However, in the view developed herein, where I argue that being unable
to be guided by a practical identity that one can affirm the value of through
ones actions and reflections indicates IAPs, does not rest on the assumption
that global self-esteem is the marker of nonadaptation. Iaccept that a persons
identity is segmented into overlapping practical self-conceptions (e.g., being a
father, a survivor, a caring person, a professional). What is required, however,
is that one must be guided by a self-understanding that reflectively expresses a
way of being one can accept without alienation. Khaders criticisms of proce-
dural views do not undercut that approach.31
What Ihave tried to show here, then, is that structural accounts that re-
quire just social relations or expressions of recognition do not adequately cap-
ture the difference between autonomy and adaptation we are seeking, however
powerful they may be for other purposes. Before moving on to a positive view,
however, let us consider one final discussion of autonomy that may shed light
on such judgments. Sarah Buss has recently argued that accounts of autonomy
that require self-reflection and internal control over first-order desires all fail
to fully account for what makes such desires ones own.32 She takes to task ac-
counts of autonomy (which can serve to confer accountability of the right sort)
that see it as an idealized form of agency control. Such views all fail to cap-
ture what is distinctive about autonomous agency because we can be authentic
29
Khader, Adaptive Preferences and Procedural Autonomy,178.
30
Khader, Adaptive Preferences and Procedural Autonomy,181.
31
Khaders arguments are quite thorough, and my discussion may well not adequately reflect their
nuance so the comments in the text must be taken as a summary judgment rather than a painstaking
argument. Also, she goes on to develop a view of IAPs based on the idea of flourishing which deserves
more discussion. Khader, Adaptive Preferences and Womens Empowerment. For reasons laid on the rest
of this paper, however, it will be clear why Iwould reject such a perfectionist approach.
32
Sarah Buss, Autonomous Action: Self-Determination in the Passive Mode, Ethics 122:4 (July
2012):647691.
relative to habitual, unreflective, automatic, and other actions over which we

exhibit no real agentic control, she argues.33 Moreover, those views that require
higher-order reflective endorsement all fall prey to what has been labeled else-
where as the regress problem: that is, no noncircular account of what gives
those higher-order reflections special status can be given without postulating
ever higher levels of reflection. Referring to procedural independence simply
gives a name to what needs to be explained and pointing to examples such as
brainwashing and manipulation do not explain exactly why and how these
things undermine agency.34
These criticisms of procedural accounts are well taken, and in the final sec-
tion Iwill try to address them in part. However, Buss also urges us to take a dif-
ferent tack in thinking about autonomy, to forget about seeing self-government
as a kind of supped-up agency, where we have the greatest reflective control
over ourselves. Rather, she thinks that crucial aspects of agencymental,
emotional, and other statescausally produce our actions so we are essen-
tially passive in regards to them. As she says, Nothing active can sustain its
[the agents] activity without passively relying on everything that makes this
activity possible.35 Apersons character causes her action, and this is why we
can attribute those actions to her; however, she is essentially passive regarding
her character in important respects.
In explicating this view, Buss also looks at a certain asymmetry of judgment
concerning actions and their causes, noting that factors we regard as positive
that drive a person to act are thought of as commensurate with agency while
negative factors are thought to undermine it. To explain this asymmetry, she
posits the view that factors incommensurate with human flourishing under-
mine agency but that those supportive of such flourishing do not. She says
little to illuminate what she means by flourishing and how we are to delineate
the factors she has in mind, but hers is generally a structural account in the
nomenclature Iused earlier. An external, objective appraisal of the agents en-
vironment determines the quality of her agency and not an account of the
internal workings of her reflection and judgment.
I will comment on the usefulness of this account to explain the different
asymmetry that we are considering in a moment, but Iwant to note first that
an alternative explanation for the different reactions we have toward positive
and negative factors she describes is available to us. Namely, we can easily say
that positive compulsions are not out of synch with the normative frame or
schema that structures ones identityunless of course they are:imagine the
33
Buss, Autonomous Action.
34
Buss, Autonomous Action. The views she is referring to most directly are those of Harry
Frankfurt and Gerald Dworkin. For Dworkins views on procedural independence, see Dworkin, The
Theory and Practice of Autonomy (Cambridge:Cambridge University Press, 1990), chap.1.
35
Buss, Autonomous Action,658.
person in the manic phase of what used to be called manic depression (bi-
polar disorder). When a person is so overtaken with an overflow of powerful
positive emotion that she appears not to be herself it is because that level of
positive intensity is seen to be (and lets imagine in the case it is) a force out of
synch with the normative frame definitive of the persons identity. When your
manic sister wants to hug you too much, too often, going on about how won-
derful you are, it can (sadly) become clear that she is in the grips of an episode.
Consider also a person on the drug ecstasy:he is intensely explaining to you
how awesome you are. None of these are counterflourishing actions or emo-
tions per se. What makes them suspicious is that they run counter to what the
person would normally do. They reveal the workings of psychic forces that are
out of synch with the persons identity.
More pointedly, however, the distinction between agency sustaining and
agency undermining causal factors, for Bussa distinction she describes as
turning on whether such factors contribute to basic human flourishingwont
help to explain our different reactions to people like Abby and the others. For
clearly the factors that caused both Abby and Bernice to readjust their aspira-
tions were harmfula terrible accidentbut the difference in their reactions
was, Iwould insist, a result of the methods they used to come to understand
and process those factors. Similarly, both Kaew and Irina are reacting to neg-
ative (counterflourishing) environmental factors, but they differ in their re-
action to them, a difference that marks their different status as autonomous
agents, Iwould suggest.
So although Buss may be right in what she says about a certain notion of
agencythere may well be multiple takes on agency available to us depending
on the purpose of assigning that label to peopleit wont help in our present
task. However, in the final section, where I make a general suggestion that
distinguishes the cases in question along the lines of a procedural account of
autonomy, it will be necessary to again take into consideration the challenges
Buss raises for similar such accounts in the literature.
3. Procedural Autonomy and Self-affirmation
Throughout this discussion I have been trying to shed some light on the
daunting problem of internalized oppression by focusing on more common-
place cases of adaptation and adjustment, even when that involves significant
shifts in ones value commitments and practical identity. In what follows Iwant
to discuss a procedural approach to autonomy that, with some amendments,
helps us focus on the aspect of those adjustments that allow self-government
to maintain or reestablish itself, in contrast with cases where it is undermined.
What Ipropose is to take onboard the claim many have made in similar theo-
retical contextsnamely, that autonomy must involve the (weakly substantive)
requirement of a sense of self-worthbut motivate that requirement in a

slightly newway.
The point is to defend an approach to autonomy that helps distinguish these
different modes of adaptation and to explain why some of them allow agents
to retain self-government while others do not. Ihave briefly argued in the pre-
vious section that structural accounts of this difference are not likely to be suc-
cessful, though of course my discussion was far from exhaustive. But my rough
conclusion is that it is not merely the injustice of the circumstances under
which adaptation takes place that classifies that process as autonomy under-
mining. Something more must be said. To say more, let me develop an argu-
ment for including self-worthor what Iwill call reflexive self-affirmationas
a condition of autonomy.
Proceduralist accounts of autonomy are those that attempt to explicate
the notion of self-government without essential reference to the evalua-
tive content of the desires, values, and character traits of the autonomous
person but rather to specify the competences and processes of the formation
of those desires and values that mark them out as autonomous. Now it has
been claimed that proceduralist accounts that claim value neutrality are ac-
tually driven to what are called weakly substantive views. This can be said to
maintain the spirit of neutrality and sensitivity to difference that motivates
proceduralism in this context but more adequately capture the way some
value commitments will be, by their psychological function, required for
self-government. The commitment most often mentioned is some variation
on self-trust.36
I am sympathetic with this move, as long as it is described in the proper way,
namely, that certain value commitments are required for autonomy because
the core conditions of autonomy (e.g., reflective, competent, self-acceptance)
cannot be achieved without those commitments but not because such com-
mitments are inherently valid or justified.37 I will return to this point in a
moment.
36
The argument concerning weakly substantive views of autonomy can be found in Sigurour
Kistinsson, The Limits of Neutrality:Toward AWeakly Substantive Account of Autonomy, Canadian
Journal of Philosophy 30:2 (2000):257286. For discussion of self-trust as a requirement of autonomy,
see, e.g., Benson, Autonomy and Self-Worth, Journal of Philosophy 91:12 (1994): 650668; Carolyn
McLeod, Self-Trust and Reproductive Autonomy (Cambridge, MA: MIT Press, 2002); Trudy Govier,
Self-Trust, Autonomy and Self-Esteem Hypatia 8:1 (1993):99119.
37
That is, we do not go from the assumption that certain values are objectively valid to the con-
clusion that failure to grasp these values is a mark of a lack of competence, as some theorists do. See,
e.g., Susan Babbitt, Feminism and Objective Interests: The Role of Transformation Experiences
in Rational Deliberation, in Feminist Epistemologies, edited by Linda Alcoff and Elizabeth Potter
(NewYork:Routledge, 1993), 2464. See also Natalie Stoljar, Autonomy and the Feminist Intuition, in
Relational Autonomy, edited by Catriona Mackenzie and Natalie Stoljar (NewYork:Oxford University
Press, 2000), 94111.
The historical approach to autonomy that Iand others utilize emphasizes the
diachronic nature of the agent and the procedures that mark self-government.38 My
attempt to defend an historical account utilized the condition of reflective accept-
ance of ones motives (without alienation) in light of ones history and in relation
to ones diachronically structured practical identity.39 The problem Buss and others
raise with this account is that acts of reflection, even the hypothetical reflection
Irequire, can themselves be manipulated and no account of adequate reflection
can be given that doesnt assume that we already know what self-governmentis.
I now see that the account of self-acceptance upon reflection is in cru-
cial ways undertheorized. On one hand it, from the critical self-reflection
views from which it arises, it may inherit the challenge of a regress and the
specter of voluntarism (despite attempts to avoid it). That is, either the acts of
self-reflection are themselves in need of self-ratification, or we must assume a
self-evaluation of its own attitudes without a basis. In addition to weakening
the attitude that confers agential status on lower-order motivations to aliena-
tion (rather than endorsement), Ialso insisted that the reflection need only be
hypothetical. Moreover, Ialso claimed that the reflection need not be deep in
the sense of considering a world where one did not have the attitude (since that
may be psychologically impossible for many autonomous agents) but only gen-
erally evaluative, where one considers the importance of the attitude, its impli-
cations for oneself and others, and its connections to other aspects of theself.
These moves add up to a picture that is really not a hierarchical view at all,
Inow see, at least not in the sense of postulating levels of mental activity that are
ordered in terms of their centrality to the self. Rather, saying that the autono-
mous person acts from motives that she would reflectively accept as part of her
practical identity is not to imply that the reflective self is more central somehow
to our agency but rather that acting from motives that can be self-reflectively
endorsed in light of ones practical identity themselves have a special status
(vis--vis authenticity). That is, in acting deliberately, one reflexively (not nec-
essarily reflectively) affirms the value of the practical identity that motivates and
justifies that act. In a manner akin to a recursive function, when one acts inten-
tionally one engages a practical identity that affords value to that choice and in
so doing reflexively reinforces the value of having that practical identity.40 Iact
as a certain type of person when Iact, and in so doing Irepeatedly reinforce the
value of being such a personat least when things are goingwell.
Reflection does play a role in this process because we are often called on to
reconsider our actions in light of our basic values. In particular, because of the
38
Other historical accounts include Al Mele, Autonomous Agents (NewYork:Oxford University
Press, 1991); John Martin Fischer and Mark Ravizza, Responsibility and Control: A Theory of Moral
Responsibility (NewYork:Cambridge University Press,1998).
39
This is a highly truncated summary of the position. Christman, Politics of Persons, chap.7.
40
Cf. Keith Lehrer, Self-Trust:AStudy of Reason, Knowledge and Autonomy (NewYork:Oxford,1997).
socially embedded nature of actions, we must coordinate our acts and reasons
with others, either by being called on to account for our actions and values in
ways that force us to reflect on them or by listening to the reflections of others
in acting with them or in response to them.41
But deliberate action for the self-governing agent may well be passionate,
committed, unflinching, and hence nonreflective in any given instance. This
does not mean, however, that such action is unthinking or not underwritten
by ones basic values (often quite the contrary). But it does mean that in acting
one engages the value orienting function of ones practical identity and in so
doing reaffirms that identity as worth having and acting upon. In this way
Iagree with Buss when she underscores the ways we are often passive in rela-
tion to the elements of our character that motivate and justify our autonomous
actions.
The model of intentional action Iutilize here relies a great deal on Korsgaards
self-constitution account of agency.42 On her view, when one acts intention-
ally one is guided by a standard of success or failure, which is given both by
the teleological structure of action (one chooses an action with a purpose in
mind) and by the evaluative organization of the practical identity that grounds
the choice. In this picture, acting is always value conferring in two directions.
When our actions are guided by our practical identities, we both project value
onto those actions and, reflexively, inherit their value in being their source.
Now Korsgaards picture of action and identity is much more detailed than
this and is embedded in a Kantian framework that Idont want to commit to
here (necessarily). Iwant merely to extract the central element of her view of
action, namely, that we constitute ourselves in action and reflexively affirm the
value of the identities that underwrite those actions when we choosethem.
There are, as has been suggested, important interpersonal and social ele-
ments to this process. I agree, but in a particular way that should be clari-
fied to retain the procedural status of the view. The nomenclature of practical
identities, the terms in which value orientation expresses itself, involves a rich
language of roles, social categories, and standards of behavior. We see our-
selves and act as a certain kind of person, such as being a mother or a teacher
or a Muslim or a woman, and those self-descriptors carry with them socially
structured expectations and motivations. These terms do not merely refer to a
set of values or propositionally structured commitments that one can say one
believes in (though those surely figure in our identities). Rather, these catego-
ries orient our way of seeing the world and prioritizing values and options to
41
These observations can be read as a partial response to some critics of my view. See, e.g. Diana T.
Meyers, Review of The Politics of Persons, Hypatia 27:1 (March 2011):227230.
42
Christine Korsgaard, Self-Constitution:Agency, Identity, and Integrity (Oxford:Oxford University
Press,2009).
act.43 The core of the postulate of self-worth, Isuggest, is that those descriptors
can be understood as naming a life worthy of pursuit from ones own perspec-
tive. They are often self-validating in practical reflection because they stand in
for a host of interconnected values that, in turn, justify action. One need only
think, As a mother, Iput my childrens needs above my own and not feel the
need to further reflect (in the normal case) on whether it is a good thing to be
a mother of this sort.44
I say in the normal case because Iam describing here cases of nonalienated
self-affirmation. Alienation occurs when these feedback mechanisms break
down or cease to function, when affective concomitants to acting from ones
practical identity are in conflict or misaligned, when one feels shame or em-
barrassment, or when feelings of inner conflict undercut the usual motiva-
tional function of acting on your values. When these motivational feedback
effects do not line up this is a signal that the values are not yours in the proper
sense, that you are hiding something from yourself or feeling resistance to
an imposed or artificial (to you) set of values. But when alienation is not an
issue, acting from ones settled identity requires no overt reflective endorse-
ment from a detached viewpoint.
This is important because it encapsulates the way that social recognition
functions in the reflexive operation of self-affirmation. Intentional action is
motivated in part by a self-understanding under an evaluative identity cat-
egory; such a category is recognized in some publicly manifested way, as
self-validating, as worthy of pursuit. There need be no actual spokesperson,
as it were, for that value, but it must be publicly available in some way. This is
because a purely private category reduces to simply a set of desires, a nexus of
motives that are merely how one is but are not reflexively self-justifying in this
way. Of course these publicly available categories of affirmation need not be
instantiated in actual relations among peoplethis is the aspect of the struc-
tural accounts from which we are departing. Rather they must be available to
the agent to use to understand herself and to feel motivated to continue acting
in the way she understands herself to be. No one needed to actually tell Emily
Dickenson that she was a wonderful poet, but the idea of successful poet was
available to her to evaluate her commitment to her way of life and to help her
affirmit.
In the present context, it is important to note also some of those identity
categories might include things like rebel, outcast, prisoner, pariah, and slave.
43
Cf. David Velleman, Identification and Identity and The Centered Self in Self to Self
(Cambridge:Cambridge University Press, 2007), 253283.
44
This is akin to what Stephen Darwall, Two Kinds of Respect, Ethics 88:1 (October 1977):246
261, labels recognition self-respect. Also, I do not take up the details of these similarities or the
complexities of self-respect itself. For discussion see Robin Dillon, Self-Respect:Moral, Emotional,
Political, Ethics 107:2 (January 1997):226249.
Under these descriptors, the person is guided by an evaluative identity that

precisely rejects or is rejected by her social context, yet by the very nature of
such a social location this fact fails to undercut the motivational power of that
identity. Testimonials from slaves and imprisoned persons suggest how such
persons seem to be able to continue their struggle for meaning despite a com-
plete lack of affirming feedback from surrounding others, in particular those
who directly exercised power or whose perspective so influenced the domi-
nant symbolic structure of their dominated lives.45
The actual expression of social recognition of an agent by others often uti-
lizes a social category that captures the object of that recognition, the type of
person one is, but it is often quite contentious both what category of identifi-
cation is appropriate in a given instance and what such categorizations mean.
In cases of adaptation to change, one could well abandon a social identity that
one had been recognized under in the pastAbby was known as an outdoors
person, a runner, an athlete but now would not use those terms to describe
herself. But one may use terms that make sense of that very change:Iam a sur-
vivor, a teacher with disabilities, a resilient person able to change in response
to challenges. These descriptors carry with them a sense of self-affirmation that
other terms might not allow. Bernice, for example, may see herself as a (crip-
pled) runner and hence unable to pursue a self-affirming life. The dynamics of
recognition, if they are manifested in actual interpersonal exchanges, may dis-
tort the value frames that allow, in some cases, the capacity for self-governing
agency.46
Now it might be thought that Ihave taken a needlessly serpentine route to
a very familiar destination, namely, that attitudes of self-regard are required
for autonomy. Iaccept that the destination is familiar, and Iam happy to join
company with many insightful theorists of autonomy and agency here; how-
ever, Ialso want to draw some minor contrasts.
As an example we can look at Paul Bensons subtle account of the require-
ment of self-authorization for autonomy, authorization that requires that
autonomous agents must be accountable to others, to be able to respond to
potential challenges which, from [the persons] own point of view, others
might appropriately bring to his view.47 For Benson, this need not be a re-
flective act:we may exhibit the attitudes in question by way of our activities
45
Again, cf. Velleman, Centered Self, when he says, one must be able to intend to act in way one is
understood by others. The view Itake here resembles the approach in Benson, Taking Ownership, in
claiming that autonomy involves being in a position to answer for oneself see Taking Ownership,
101126. Aparallel similarity can be seen with Andrea Westlunds account.
46
Relevant here is the discussion in Susan Brison, Aftermath (Princeton, NJ:Princeton University
Press, 2002), chap.3, of how survivors of trauma must recreate self-narratives that include the trauma
itself.
47
Benson, Taking Ownership. For a similar account, see Westlund, Rethinking Relational
Autonomy.
that effectively place ourselves in a position to answer for our actions. Benson
contrasts this with identity-based accounts of autonomy claiming that authen-
ticity of actions is secured when they relate to my identity as a caring, reflec-
tively willing creature. Benson argues that such theories are either too weak
(when reflective self-endorsement is merely apathetic or undeveloped) or too
strong (since many trivial or habitual actions would not pass the test but they
are clearly autonomous).
I will not reply to the criticisms Benson wields against my own view in this
context,48 but Iwant to point out how the procedural account Idefend here can
take onboard the requirement of reflexive self-acceptance that Benson insists
upon. Recall that the question we are trying to ask, in effect, is when a persons
responses to the constraints of her circumstances, whether caused by oppres-
sion or accident, are such as to undercut the self-affirmation (e.g., answerabil-
ity, self-ascribed normative authority) that is inherent in autonomous action.
Iam not sure how Bensons condition does that since it will depend greatly on
what we take to be the appropriate modes of answerability. If Bernices atti-
tude toward herself can be expressed as, Sure Idesire to get into this sodding
wheelchair each day, but that is only because of that damned accident; its a
situation Ideeply hate, then it seems clear that she fails to value her normative
position adequately, and she does so because of the origins of that position.
But we do not want to demand complete acceptance of the constraints that
circumscribe ones motivational structure, for then none of us would pass the
test for autonomy:we all can say that we would aspire to different things if
there were no (or fewer) barriers in front of us. The question is:when does the
weight of the constraints on our existence effectively undercut the authenticity
of our self-endorsements; when are we simply slaves of our condition?
This question points to the mysterious core of many accounts of
self-governing action:the phenomenon of owning the motivational and eval-
uative structure that issues in action, of seeing it as ones own. One element of
this phenomenon is clear enough, namely, that one sees ones motives as worth
having, but another is left obscure, namely, seeing them as mine. But what is
this my having that is operatinghere?
Bensons claim is that being answerable is an active process. However, de-
spite his guidance on what should be meant by this, it merely may end up
restating this mystery, Ithink, for he admits that the activity in question need
not involve reflective choice (though it must be purposeful).49 Reflecting on
the condition of Bernice forces us to ask where the point is when the fact that
48
See also Paul Benson, Autonomy and Oppressive Socialization, Social Theory and Practice 17
(1991):385408. For a version of my view that attempts to avoid these criticisms, see Christman, Politics
of Persons, chap.7.
49
Benson, Taking Ownership,113.
ones motivational set is so structured by the contingencies of ones condition

(as all our motives are to some degree) that one can no longer see it as ones
own but simply as the result of (in this case) an accident that happened tome.
What Ihave to offer here is the following:an agent lacks this kind of own-
ershipis alienated in my sensewhen the feedback mechanism Ihave been
describing between deliberating from within a practical identity and con-
ferring (recursively) value back to that identity does not function. Bernice
acts out of an identity of a disabled person but cannot accept the reflexive
self-affirmation that might come from so doing. She lacks a key component of
self-respect, namely, a self-understanding as a being with dignity whose delib-
erative perspective is worth having as onesown.
Irina maintains this loop but in a different way, for she understands herself
under the label of survivor, one who does what is required of her circumstance
even when she begrudgingly acts from within an externally imposed identity
of prostitute, which she rejects even when acting as one. The feedback mech-
anism of reflexive self-affirmation operates for her under the description of
survivor so her effective self-worth is intact. This explains our willingness to
see her as at least minimally autonomous in her struggles.
So the condition of alienation can be stated this way:one is alienated when
ones practical identity fails to provide motivations that carry with them re-
flexive self-affirmation of a sort that issue in further solidification of that iden-
tity and subsequent decisions of the same sort. I tried to get at this idea in
earlier work but expressed it in terms of reflection:adequate reflection is what
would issue in similar self-acceptance over a variety of conditions. Here Iam
using the language of reflexive self-affirmation:the nonalienated agent acts in
a way that sheds positive light on her identity as a person acting that way, giv-
ing her further motivation to continue todoso.
Now despite these observations Ido not want to exaggerate my differences
from theorists who insist on social recognition in their relational accounts of
autonomy, and Ireadily accept that the points Iam making here are highly
resonant with points made by others in the literature. But even if this is what
was meant all along by these thinkers, Ihope to have provided even further
argument for its necessity, at least as described in the narrow and qualified way
Idohere.
Moreover, Ihope to have shown that such a postulate aids in distinguishing
modes of adaptation that maintain autonomy from those that do not. To see
this, let us now return to the cases of Abby etal. Our intuitive judgment is
that Abby, who alters her values in response to her accident, and Irina, who
refuses to adapt to her circumstance, retain a kind of autonomy that the other
people Idescribed seemed not to. The reason Iam suggesting here for why
this obtains is that in these cases the person was able to maintain a diachronic
practical identity under which her life was one she could regard as worthy of
pursuit. They were able to retain a sense of reflexive self-affirmation despite
needing to negotiate through stultifying constraints. Notice that the language

of self-affirmation is not one of celebration, and often thinkers who tread sim-
ilar paths exaggerate the degree to which we must value the lives we lead to
be self-governing in leading them.50 But lives so often involve deep ambiva-
lence and profound compromise in merely negotiating through the perilous
challenges of a modern existence. What must be the case is that one affirms
ones identity to a degree that allows the motives stemming from that identity
to be reason giving. But in so doing valuing oneself must merely be seen as
self-acceptance but not valuing ones life all things considered.
The threshold of what I elsewhere called nonalienation (from oneself) is
being able to reflexively confer value on actions guided by ones identity (under
the relevant category term), actions which by that token reflect value back on
that very identity. Bernice, the person who resists change after her accident
and leads a life of bitterness and regret, has lost her autonomy (to some signif-
icant degree) insofar as she can no longer see her decisions as arising from a
self-affirming understanding of herself. She decides to use her wheelchair be-
grudgingly, and she fights through the psychological pain of her existence with
resignation and bitterness. Since she cannot reflexively affirm her outlook on
her options, the choices she makes do not reflect a self-governing mode of ex-
istence. At least, insofar as we think self-government is lost or severely diluted
here, it is this inability to reflexively affirm her identity that explains it. It is
not, Iargue, the fact that external circumstances have altered her option set,
for that is true of others who did not lose their autonomy.
Of course, much could be said to explain differences in these cases along
different lines. For example, living through a severe accident and being forced
into sex work are deeply traumatic, and surviving trauma requires its own
process of reestablishing a functional self-narrative.51 Moreover, there is every
reason to think that those who internalize oppressive value schemes such as
Kaew also lose many basic competences that are central to autonomy:they may
suffer from drug addiction or lose emotional and psychological capacities re-
quired for self-directed choice, empathy, and imagination. Surely in most such
cases this is the way to understand the loss of autonomy. But Imerely want to
add that in those perhaps rare cases where at least basic competences are es-
tablished or restored but where persons adapt their values to the requirements
of their constrained circumstances, a necessary condition of their retaining
autonomy is that they are able to reflexively affirm the practical identity (under
50
Korsgaard, for example, says that a practical identity is a normative conception of the self
containing ones guiding principles: it is a description under which you value yourself. Christine
Korsgaard, Sources of Normativity (Cambridge:Cambridge University Press, 1996), 101. This is ambig-
uous between self-acceptance and self-celebration Ithink.
51
For a rich account of such a process, see Brison, Aftermath.
a relevant description) that provides reasons for actions, plans, and ongoing
projects.
In cases of internalized oppression, the actual diagnosis of the retention of
autonomy may be difficult or impossible to make, but that may be due to the
indeterminacy of what practical identity category is operative for the person.
Is a person like Kaew, for example, best seen as one who embraces the identity
of a resilient survivor and who engages in prostitution because her life options
limit her to such choices as a method of, say, providing for her family? If so,
then if other conditions of autonomy are met we should not withhold that
label to her. However, if we are suspicious, as Superson is in the case of veiling,
that acceptance of the identity category that requires such behavior could not
but be the result of a self-abnegating identity, then we can label the person
heteronomous. But this is not due to anything inherently degrading or undig-
nified about that behavior itself, independent of its effect on the persons sense
of self-worth.
4. Conclusion:The Dynamics of Autonomy Attribution
To sum up, to help sort out whether those who adapt to their circumstances by
seemingly internalizing oppressive conditions in which they live are autono-
mous or not, Iasked us to consider contrasting cases among those who have
and have not made such self-alterations. Those contrasts, Iclaimed, suggested
an asymmetry in our judgments about whether adapting to constraining cir-
cumstances indicates a loss of autonomy or its reestablishment. Iargued that
the structural character of those conditions alonewhether the source of the
constraint was unjust actions or conditions or merely an accidentwas not
the crucial determinant of whether autonomy is preserved. Though Idid ac-
knowledge that relational and interpersonal dynamics of the situation are rel-
evant to that determination; I merely insist that this necessity is contingent
upon the contribution to the full functioning of the procedural requirements
of autonomy.
I claimed, however, that a procedural account of autonomy could help ex-
plain these asymmetries by including a requirement of reflexive self-affirmation
along the lines that other relational theorists in the literature have suggested,
but Iclaimed that such a view is not constitutively relational except in the in-
direct manner in which value categories (in terms of which a person develops
a self-affirming practical identity) are acknowledged somehow in the social
landscape as marking a life worthy of pursuit. I argued that this social ele-
ment is required not because of the intrinsic value of those relations or those
pursuits but because of the social-psychological effect that such public accept-
ance has on the ability of a person to be reflexively self-affirming. But it is
this latter condition that marks autonomy, and it admits of great variability in
the dynamic of public affirmation, as the cases of the rebel and resistant slave
indicate.52
I want to close here by emphasizing the way accounts of autonomy should
include reference to the need for persons themselves to participate in the very
determination of their autonomy status. That is, the sites where the attribu-
tion of autonomy operates often exhibit power dynamics that make not only
the expression of a persons practical identity (for example) but also its very
determination difficult. A person may be unclear what value category best
describes the struggles she is engaged in, and whether or not she is autono-
mous may turn on whether she can make that determination and secure a
sense of self-worth that comes withit.53
This is more than an epistemological issue. It is not merely that our theo-
retical account of autonomy is complete, and we merely must be careful when
attempting to discover if its conditions are met in a given case. Rather, the
dynamics of interaction at the site where autonomy determinations are most
crucial are noninstrumentally relevant to whether or not the person is in fact
autonomous. This is because people often need helpful interlocutors (e.g.,
carers, therapists, aid workers, friends) to help establish the practical orien-
tation that organizes their reflective practices. In this way, autonomy is also
relational.54
How theoretically described conditions must be established by actual
social practice is a complex topic that I want merely to gesture at here.55
However, I want to close this discussion with an insistence that models of
self-government include desiderata that rely on persons actual participation
and self-expression to determine the presence of those desiderata. In this way,
52
This last point echoes views developed by Diana Meyers in a different context, when she argues
that what she calls value-saturated accounts of autonomy fail to accommodate cases where people
in oppressive circumstances can sometimes exercise autonomy in strategically negotiating their lives
within those circumstances. Moreover, she argues that such views must be sensitive to the ways that
the voice with which a person may try to express her reaction to her condition is not heard or under-
stood by the listening public or surrounding others, making a condition of external recognition of the
persons value orientation a problematic requirement. See Diana T. Meyers, Feminism and Womens
Autonomy:The Challenge of Female Genital Cutting, Metaphilosophy 31:5 (2000):469491.
53
For a sensitive account of this dynamic, see Catriona Mackenzie, Christopher McDowell, and
Eileen Pittaway, Beyond Do No Harm: The Challenge of Constructing Ethical Relationships in
Refugee Research, Journal of Refugee Studies 20:2 (May 2007):299319.
54
In this way Iagree with Elizabeth Ben-Ishai when she argues that ascribing autonomy to per-
sons affords them a status that is required for self-government, even in cases where their capacity for
self-government is apparently lacking or marginal. Idisagree, however, that such ascription is constitu-
tive of autonomy (rather than causally contributory to it) for reasons similar to my departure from the
aforementioned relational accounts. See Ben-Ishai, Sexual Politics and Ascriptive Autonomy, Politics
and Gender 6:4 (2010):573600.
55
For one attempt to elaborate on this point, see John Christman, Relational Autonomy and the
Social Dynamics of Paternalism, Ethical Theory and Moral Practice (Special Issue, forthcoming).
the persons voice, aided by a cooperative social context that elicits that voice
in effective ways, is central to the determination of her autonomy.
Acknowledgments
The author is grateful to the participants and organizers of the workshop

Relational Autonomy 10 Years On, where an earlier version of this chapter
was read. Special thanks is also due to Mark Piper and Andrea Veltman for
comments on an earlier draft of this chapter.
{11}
Autonomy and Adaptive Preference Formation

Natalie Stoljar
At a Sunday Bible-study group Iattended for teenage girls, the mother

who was teaching had the girls hold hands, march in a circle and
say:My husband will treat me like the princess that Iam. He will be
the head of my household. But the girls own ambitions seemed at
odds with that vision. One girl earlier confessed that her biggest earthly
temptation during her college years was likely to be pursuing too many
higher degrees. Another was known to her friends in the group as the
future president. Igot the sense that relying on a man was not what
they considered their best option.1
Many philosophers think that the inculcation of gendered norms is harmful to
autonomy:for instance, gendered socialization may damage the skills and com-
petencies required for autonomy; oppressive stereotypes may block the ability
to imagine alternatives thereby limiting and constraining life plans; hostile
and sexist responses to womens aspirations may impair ambition and damage
self-esteem so that even if nonsexist options are institutionally available girls
do not pursue them.2 It is encouraging that some of the young girls described
at the outset of this paper articulate ambitions for themselves that do not re-
quire deference to men, but it remains to be seen whether any of them find the
resources to resist their socialization and put these ambitions into effect.
In this chapter, Ifocus on an argument claiming that the inculcation of op-
pressive norms damages autonomy in a particularly insidious way. Agents who
are oppressed come to internalize their oppression:they come to believe in the
ideology of oppression and to make choices, and form preferences and desires,
1
Hanna Rosen, Who Wears the Pants in This Economy? New York Times Magazine, August
30,2012.
2
See, e.g., Diana T. Meyers, Self, Society and Personal Choice (New York: Columbia University
Press, 1989); Paul Benson, Free Agency and Self-Worth, Journal of Philosophy 91 (1994): 650668;
Catriona Mackenzie, Imagining Oneself Otherwise, in Relational Autonomy: Feminist Perspectives
on Autonomy, Agency and the Social Self, edited by C. Mackenzie and N. Stoljar (NewYork:Oxford
University Press, 2000), 124150 (and in general all articles therein).
in the light of that ideology. Ann Cudd says that oppression creates deformed
desires, in which the oppressed come to desire that which is oppressive to
them...[and] ones desires turn away from goods and even needs that, absent
those conditions, they would want.3
Sandra Lee Bartky describes an example of this process in her analysis of
the phenomenology of oppression. Through what she calls the interiorization
of the fashionbeauty complex, agents come to believe oppressive norms that
tie appearance to self-worth.4 Bartky says that repressive satisfactionsthat
is, deformed desiresfasten us to the established order of domination, [to]
the same system...produces false needs and that false needs are produced by
the denial of autonomy.5 These remarks suggest that desires that reinforce
ones own oppression are morally problematic because they are formed by
agents with impoverished autonomy. The oppressive conditions are respon-
sible for the desires, not the agent herself.6
Theorists of oppression propose that oppression is distinctive because it a
group harm. Oppression occurs when a group suffers systematic injustice due
to institutional structures or background social practices.7 Paradigm examples
of oppression are systematic injustices suffered by groups whose members
share social identities, for instance those of class, gender, race, sexuality, or dis-
ability. Although oppression employs the notion of group disadvantage, agents
who are members of oppressed groups are also harmed as individuals. Cudd
identifies both material and psychological harms of oppression.8 For instance,
racial segregation in the United States is a material and economic injustice that
3
Ann E. Cudd, Analyzing Oppression (NewYork:Oxford University Press, 2006),181.
4
Sandra L. Bartky, Femininity and Domination: Studies in the Phenomenology of Oppression
(NewYork:Routledge, 1990),39.
5
Bartky, Femininity and Domination,42.
6
Anita Superson, Deformed Desires and Informed Desire Tests, Hypatia 20 (2005): 109126;
Superson, Feminist Moral Psychology, The Stanford Encyclopedia of Philosophy, (Fall 2012 Edition);
<http://plato.stanford.edu/archives/fall2012/entries/feminism-moralpsych/>.
7
Ifollow authors such as Iris M. Young, Five Faces of Oppression, reprinted in Rethinking Power,
edited by T. Wartenberg (NewYork:State University NewYork Press, 1992), 17495; Sally Haslanger,
Oppressions: Racial and Other, in Racism in Mind, edited by M. P. Levine and T. Pataki (Ithaca,
NY:Cornell University Press, 2004), 97123; Cudd, Analyzing Oppression. These authors give accounts
of oppression that focus on systematic harm to groups; on these accounts, individuals who suffer harms
of oppression are harmed in virtue of their group membership. Haslanger points out that oppression
can be perpetrated by agents, or it can be structural; that is, systematic disadvantage to a group can
occur as a result of social practices even in the absence of agents or legal institutions that intentionally
perpetrate the oppression. Although here Itake systematic harm to a group and concomitant harm to
individuals in virtue of their group membership as sufficient for oppression, Iwish to leave open the
possibility that single agents or nongroups could suffer oppression. Hence, Ileave open the question
of whether systematic harm to a group is necessary for oppression. Ialso do not employ the notion
of autonomy in the definition of oppression. Cf. Daniel Silvermints position on which he claims that
oppression obtains when an individuals autonomy or overall life prospects are systematically and
wrongfully burdened. Silvermint, Oppression without Group Relations, unpublished manuscript,
Montral, Canada,5.
8
Cudd, Analyzing Oppression.
Autonomy and Adaptive Preference Formation 229
victimizes a group. Black Americans suffer systematic disadvantage through

residential, job, and school segregation. Individual agents who are black are
harmed in virtue of their group membership; they are prevented from attend-
ing better schools, doing better jobs, and living in the more affluent areas.
Similarly, the unjust stereotyping of a social group can cause members of the
group to suffer psychological harms such as humiliation andshame.
In addition to direct material and psychological harm, Cudd describes
indirect psychological harms that originate within the psychology of the op-
pressed themselves and hence are often responsible for the self-perpetuating
and entrenched nature of oppression. On her account, deformed desires are
indirect harms that can be mistaken for legitimate expressions of individual
differences in taste when in reality they are formed by processes that are
coercive: indoctrination, manipulation and adaptation to unfair social cir-
cumstances.9 If Cudd is right, we have a neat argument for the conclusion
that oppression undermines autonomy:gender oppression leads to deformed
desires and deformed desires constitute autonomy impairments.
One version of the argument is that desires for ones own oppression are
ipso facto deformed due to their contents. For example, Anita Superson claims
that deformed desires fail a properly fleshed-out informed desire test. They
are irrational and impair autonomy because desiring ones own oppression
is inconsistent with the agent having an appropriate sense of her own moral
worth.10 For the purposes of this paper, I set aside Supersons position and
focus on a second possible argument that deformed desires are autonomy
impairments, namely, deformed desires are adaptive preferences and for that
reason paradigmatically nonautonomous.11 In Jon Elsters canonical example
of sour grapes, a fox, after finding that he cant reach some grapes, decides
9
Cudd, Analyzing Oppression,183.
10
Superson, Deformed Desires and Informed Desire Tests. Superson claims that choosing
or preferring ones own oppression is analogous to choosing slavery; it is making a special kind of
moral mistake. Although Superson characterizes her position using a Kantian interpretation of an
informed desire test, her argument suggests that she is committed to a strong substantive account
of autonomy in which the contents of the preferences or values that agents can form or act on au-
tonomously are subject to direct normative constraints. Paul Benson, Feminist Intuitions and the
Normative Substance of Autonomy, in Personal Autonomy:New Essays on Personal Autonomy and Its
Role in Contemporary Moral Philosophy, edited by James Taylor (Cambridge:Cambridge University
Press, 2005), 133. As Benson points out, in Natalie Stoljar, Autonomy and the Feminist Intuition,
in Mackenzie and Stoljar, Relational Autonomy, I conflated this strong substantive account with a
strong normative competence condition of autonomy. Although deformed desires would be counted
as nonautonomous due to their contents on strong substantive accounts, it is less clear whether agents
who have deformed desires would fail a normative competence condition. For a fuller explanation of
different versions of substantive account, see Natalie Stoljar, Feminist Perspectives on Autonomy,
The Stanford Encyclopedia of Philosophy, (Summer 2013 Edition); http://plato.stanford.edu/archives/
sum2013/entries/feminism-autonomy/.
11
James S. Taylor, Practical Autonomy and Bioethics (NewYork:Routledge, 2009), 71. Indeed, Cudd
says that adaptive preference is another term for deformed desire. Cudd, Analyzing Oppression, 180181.
that he doesnt want the grapes after all. Elster takes this to be an unconscious
and nonautonomous process that he calls adaptive preference formation.12
The unconscious accommodation of desires to feasible options often occurs in
conditions of oppression. For example, a girl raised in a patriarchal household
may come to prefer domestic chores because other nontraditional options are
not feasible forher.13
Scholars of development ethics also employ the notion of adaptive prefer-
ence formation.14 Martha Nussbaum and Amartya Sen focus on agents habit-
uation to conditions of severe deprivation.15 On the face of it, deprivation and
oppression are not the same injustice. Oppression occurs in affluent societies
in which people are not typically victims of severe economic deprivation. The
young girls described in the opening quote are members of a group that is
subject the patriarchal oppression but they live in conditions of relative afflu-
ence.16 It is plausible, however, that agents in conditions of severe deprivation
in the developing world are members of a group that is subject to economic
oppression. In their own cultural context, they may also suffer oppression as
women or as members of particular ethnic minorities. As we will see, there are
sufficient similarities among the two sets of examples to treat them together
for purposes of an analysis of autonomy.
There are two main challenges to the position that preferences that are
adapted to the circumstances of oppression are deformed and constitute au-
tonomy impairments. The first, articulated recently by Serene Khader, claims
that even if desires for oppressive conditions are morally problematic in some
sense, this is not because they are autonomy deficits.17 Khader identifies a
category of inappropriately adaptive preferences that are morally problem-
atic mainly because they are inconsistent with the flourishing of the agent
who forms the preference.18 However she argues that adaptive preferences
are not autonomy deficits on either procedural or substantive accounts of
12
Jon Elster, Sour GrapesUtilitarianism and the Genesis of Wants, in Utilitarianism and Beyond,
edited by A. Sen and B. Williams (Cambridge:Cambridge University Press, 1992), 219238; Elster, Sour
Grapes:Studies in the Subversion of Rationality (Cambridge:Cambridge University Press,1983).
13
Donald Bruckner, In Defence of Adaptive Preferences, Philosophical Studies 142
(2009):307324,309.
14
Serene J. Khader, Adaptive Preferences and Womens Empowerment (NewYork:Oxford University
Press, 2011),10.
15
See, e.g., Amartya Sen, Gender Inequality and Theories of Justice, in Women, Culture, and
Development:AStudy of Human Capabilities, edited by M. Nussbaum and J. Glover (Oxford:Clarendon
Press, 1995), 259273; Martha Nussbaum, Women and Human Development:The Capabilities Approach
(Cambridge: Cambridge University Press, 2000); Nussbaum, Adaptive Preferences and Womens
Options, Economics and Philosophy 17 (2001):6788. Note that the phenomenon of adaptive prefer-
ence formation was invoked by Elster, Sen, and Nussbaum to critique preference utilitarianism.
16
I am grateful to Daniel Silvermint for discussion of this point and to Daniel Silvermint and
Daniel Weinstock for helpful conversations about Khadersbook.
17
Khader, Adaptive Preferences and Womens Empowerment.
18
Khader, Adaptive Preferences and Womens Empowerment,51.
autonomy. Procedural theories claim that critical reflection is sufficient for

autonomy:agents are autonomous when they critically reflect on their motiva-
tions, beliefs, and values in the right way. Procedural approaches seek to define
autonomy using morally neutral conditions such as reflective endorsement or
nonalienation.19 Substantive theories build in normative conditions, for in-
stance moral attitudes to oneself or background moral conditions.20 Khader
considers examples of adaptive preferences including womens preferences to
malnourish themselves to feed their male relatives or their endorsement of
harmful practices like genital cutting. She argues that neither the tests employ-
ing critical reflection required by procedural accounts nor the tests employing
moral conditions required by substantive accounts show these adaptive prefer-
ences to be autonomy impairments.
A second challenge claims that preferences that are adapted to the circum-
stances of oppression are not deformed; rather, they are a rational response to
difficult and distressing circumstances in which options are curtailed. These
preferences have been described as the products of a rational process of bar-
gaining with patriarchy21 or one in which they reflect a rational compromise
because they correspond to the second best alternative that is available to the
agent under the circumstances.22 The challenges are united by various com-
mon underlying concerns, for instance that treating oppressed and deprived
people as lacking autonomy in effect extinguishes or disrespects their agency,
their choices of values and conceptions of the good; that it amounts to an em-
pirically inaccurate generalization about their psychologies; or that it poten-
tially licenses coercive government intervention in theirlives.
My goal in this paper is to argue against both these challenges and to dispel
some of the worries underlying them. I agree with Khader that adaptive
19
For reflective endorsement see Marilyn Friedman, Autonomy, Gender, Politics (NewYork:Oxford
University Press, 2003). For nonalienation see John Christman, The Politics of Persons: Individual
Autonomy and Socio-historical Selves (Cambridge:Cambridge University Press,2009).
20
For a detailed explanation of this distinction, see Catriona Mackenzie and Natalie Stoljar,
Introduction: Refiguring Autonomy, in Mackenzie and Stoljar, Relational Autonomy, 334; Stoljar,
Feminist Perspectives on Autonomy. There are additional ways of dividing up theories of autonomy.
For example, procedural theories may be either internalist, requiring only internal psychological con-
ditions to spell out autonomy, or externalist, requiring additional historical conditions that are external
to agents present internal psychological states. Alfred Mele, Autonomous Agents: From Self-Control
to Autonomy (New York: Oxford University Press, 1995). Other externalist theories are not proce-
dural and employ, for instance, the notion of adequate options (Joseph Raz, The Morality of Freedom
(Oxford: Oxford University Press, 1988)) or socio-relational conditions (Marina Oshana, Personal
Autonomy in Society (Aldershot, UK:Ashgate, 2006)). The latter externalist positions may also require
a background moral theory to spell out precisely the external conditions that are incompatible with
autonomy. Thus, externalist theories are often substantive as well. Ileave these complexities asidehere.
21
Uma Narayan, Minds of Their Own:Choices, Autonomy, Cultural Practices and Other Women,
in A Mind of Ones Own. Feminist Essays on Reason and Objectivity, second edition, edited by L. Antony
and C. Witt (Boulder, CO:Westview, 2002), 418432.
22
H. E. Baber, Adaptive Preference, Social Theory and Practice 33 (2007):105126.
preferences should not be defined using the notion of autonomy. They are
formed through psychological processes and should be defined using psy-
chological criteria. Nevertheless, I argue that on several prominent theories
of autonomy, both procedural and substantive, the adaptive preferences of
concern to feminist theorists (deformed desires) will count as impairments of
autonomy.23 Idistinguish two models of adaptive preferences, both of which
originate in Elsters work. The first, the psychological processes model, proposes
that adaptive preferences are produced by a distinctive causal and psycholog-
ical process. On this model, it is the flawed process that is responsible for the
autonomy impairment. The second, the freedom to do otherwise model, char-
acterizes adaptive preferences as adjustments that occur when options that
agents would choose under other or better conditions are excluded from their
feasible set. On this model, it is the limitation of free agency or freedom to do
otherwise that is responsible for the autonomy impairment.
Sections 1 and 2 address the claim that adaptive preferences are not au-
tonomy impairments. Ifirst consider adaptive preferences construed on the
psychological processes model and argue that historical, procedural theo-
ries of autonomy have the prima facie resources to count deformed desires
as autonomy impairments through the device of procedural independence.24
However, I claim that procedural independence cannot fully explicate why
deformed desires are autonomy impairments without importing a back-
ground moral theory into the procedural account. Thus, deformed desires are
counterexamples to procedural accounts of autonomy that employ a morally
neutral test. Section 2 addresses the second model that claims that deformed
desires are adaptive in a problematic sense because they fail a freedom to do
otherwise test. Iargue that this test is a moral one and moreover that agents
whose preferences fail this test are precisely those agents who count as hav-
ing impaired autonomy on well-known substantive accounts of autonomy.
23
Khader seems to have recently modified the objection saying that adaptive preferences sometimes
revealthough need not revealcompromised autonomy. Serene Khader, Must Theorising about
Adaptive Preferences Deny Womens Agency? Journal of Applied Philosophy 29 (2012):302317,312.
24
Ileave aside structural and internalist forms of procedural theory, such as that offered by Harry
Frankfurt, The Importance of What We Care About (New York: Cambridge University Press, 1988).
For Frankfurt, a sufficient condition of local autonomy is (appropriately understood) endorsement
or wholehearted identification at a time with a preference or desire. There are objections to this kind
of theory. See, e.g., Stoljar, Feminist Perspectives on Autonomy. Moreover, as David Zimmerman,
Making Do:Troubling Stoic Tendencies in an Otherwise Compelling Theory of Autonomy, Canadian
Journal of Philosophy 30 (2000):2553, 25, points out, Frankfurts position risks being committed to an
unpalatable consequence, namely, that acting freely is a matter of making do, that is of bringing one-
self to be motivated to act in accordance with the feasible, so that personal liberation can be achieved
by resigning and adapting oneself to necessity. Zimmermans interesting point needs further exami-
nation, but it does suggest that Frankfurts theory is the wrong place to look for an explanation of the
nonautonomy of adaptive preferences.
Iconclude therefore that Khader is wrong to claim that the adaptive prefer-
ences of concern to feminists are not autonomy deficits.
Sections 3 and 4 turn to the second challenge claiming that what are called
deformed desires are not deformed but instead are rational accommodations to
bad options. Iargue in Section 3 that, even if many adaptive preferences are the
products of a rational costbenefit analysis that is endorsed by the agent, this
does not settle the question of whether the decision is autonomous. Autonomy
theorists often distinguish between the competency and the authenticity
dimensions of autonomy25 or the mental capacities required for autonomy and
additional conditions such as adequate significant options.26 The ability to en-
gage in a costbenefit analysis may be sufficient to establish some version of
rational competency, but it does not follow that the product of this process sat-
isfies the further criteria necessary for autonomy.27 Section 4 attempts to dispel
the worries underlying arguments against classifying adaptive preferences as
autonomy impairments.
1. The Psychological ProcessesModel
As Elster points out, sour grapes is a common psychological phenom-

enon:people tend to adjust their aspirations to their possibilities.28 For ex-
ample, a wage laborer may initially resist a move from agricultural work in
the country to factory work in the city but may come to prefer city life as a
result of the experience of factory work. Although the outcome of this process
is an adjustment of the laborers preferences to his current feasible options
(those offered by city life), Elster claims that if the preference for city life is
a stable and irreversible preference due to a process of learning and experi-
ence it should be distinguished from the problematic phenomenon of adaptive
preference formation that is reversible and probably involves habituation and
resignation.29 Similarly, adaptive preference formation is to be distinguished
from the deliberate adjustment of desires to possibilities in character planning:
Adaptive preference formation differs from deliberate character planning.
It is a causal process that takes place behind my back not the deliberate
shaping of desires...The psychological state of wanting to do a great many
things that you cannot possibly achieve is very hard to live with. If the escape
25
See, e.g., Christman, Politics of Persons.
26
See, e.g., Raz, Morality of Freedom.
27
Iargued for this position in Stoljar, Autonomy and the Feminist Intuition.
28
Elster, Sour Grapes, 219. See also David Zimmerman, Sour Grapes, Self-Abnegation and
Character Building, Monist 86 (2003):220241; Bruckner, In Defence of Adaptive Preferences; Ben
Colburn, Autonomy and Adaptive Preferences, Utilitas 23 (2011):5271, for explications of thisidea.
29
Elster, Sour Grapes,221.
from this tension takes places through some causal mechanism, such as
Festingers reduction of cognitive dissonance, we may speak of adaptive
preference change. The process then is regulated by something like a drive,
not by a conscious want or desire.30
On this model, adaptive preferences are formed by a distinctive psycho-
logical process. The agent unconsciously turns away from a preference that
they would otherwise have had to avoid the unpleasant sense of frustration
or cognitive dissonance that accompanies having preferences for inaccessible
options.31 The process can occur both in cases of preference change, such as that
of the fox, and preference formation, such as girls in patriarchal households
who unconsciously turn away from inaccessible nontraditional options. The
model does not employ the concept of autonomy in the definition of adaptive
preference formation. Rather, adaptive preferences are thought to be nonau-
tonomous because a blind and unconscious causal mechanism appears to
be incompatible with autonomy; it is not intentional or under the agents con-
trol.32 The psychological processes model also does not justify the claim that
feminists often seem to endorse, namely that desires for oppressive conditions
automatically count as deformed and nonautonomous. If a girl who is raised
in a patriarchal household comes to prefer domestic roles over other options,
her preference could be the result of learning and experience (autonomous) or
the result of adaptive preference formation (nonautonomous):one cannot tell
from the preferences alone whether they have been shaped by adaptation.33
The question therefore is whether the process of formation of the preferences
corresponds to an autonomy-undermining process.
Critics of Elster point out that it is implausible that blind and unconscious
processes of preference formation are always incompatible with autonomy.
Donald Bruckner considers an agent whose spouse dies. After a period of
mourning, she gradually relinquishes the preference to have significant expe-
riences with the spouse and acquires a preference to have these experiences
with a new spouse.34 It seems that, preferences that are the products of un-
conscious causal mechanisms are not necessarily nonautonomous despite the
fact that they are formed behind the agents back. Further, Bruckner employs
empirical evidence to argue that adaptive preferences can help to promote
a valuable life because adapting to ones circumstances can be conducive to
30
31
Zimmerman, Sour Grapes, 221222, notes that the process could be sub-personal rather than
unconscious. Inote this possibility but do not consider ithere.
32
If it is not intentional action, it seems that it cannot be autonomous because even if intentional
action is not sufficient for autonomous action, it is necessary.
33
Elster, Sour GrapesUtilitarianism and the Genesis of Wants,225.
34
Bruckner, In Defence of Adaptive Preferences; see also Colburn, Autonomy and Adaptive
Preferences,57.
subjective well-being, which itself is partially constitutive of a valuable life.35

He concludes that preferences that are formed through a process of uncon-
scious accommodation to feasible options should be considered presump-
tively rational and autonomous. The presumption will be defeated if the agent
on examination fails to endorse (repudiates) the preference or, if she did not
examine it, would fail to endorse it had she examined it.36 Bruckner thinks
that the causal mechanism of formation of a preference is irrelevant to its au-
tonomy (or nonautonomy); rather, the key is the way a preference is retained.37
John Christmans notion of the socio-historical self also suggests that un-
conscious processes of preference-formation are compatible with autonomy.38
There are many aspects of the self that are not the product of voluntary choice
or deliberation. Consider a passage from Shane Phelan: I was a lesbian.
Iexperienced that moment partially as discovery:so this was the difference
Ihad always felt but never had a name for.39 For Christman, preferences or
desires will be nonautonomous only if they fail either a competency condi-
tion or a hypothetical reflection condition. Competency corresponds to the
capacity of the agent to form effective intentions relative to a desire as well
as to reflect critically about the desire. The hypothetical reflection condition
employs the notion of nonalienation to characterize authenticity and hence
autonomy.40 Alienation is a combination of judgment and affective reaction.
To be alienated is to feel negative affect, to feel repudiation and resistance...41
On Christmans account, if one is (or would be) alienated from a desire but
does not succeed in repudiating it, one is inauthentic and heteronomous with
respect to the desire. In Phelans case, her unconsciously formed preferences
about her sexual identity are autonomous because she is not (or would not be)
alienated fromthem.
Neither the test suggested by Bruckner nor Christmans hypothetical reflec-
tion condition would classify deformed desires as nonautonomous. Consider
agents whose preferences for traditional feminine roles are deeply ingrained
through oppressive socialization. Due to the effects of the oppressive ideology,
the agent treats false stereotypes as natural and formulates desires and plans
based on the stereotype. Such agents are unlikely to experience alienation
from either the norms that they have internalized or the preferences formed
on the basis of those norms. In the absence of actual alienation, is it plausible
to conclude that alienation would obtain were the agent to reflect on how her
35
Bruckner, In Defence of Adaptive Preferences, 314315.
36
Bruckner, In Defence of Adaptive Preferences, 318319.
37
Bruckner, In Defence of Adaptive Preferences,319.
38
Christman, Politics of Persons.
39
Christman, Politics of Persons, 124125. Christman quotes from Shane Phelan, Getting
Specific:Postmodern Lesbian Politics (Minneapolis:University of Minneapolis Press, 1996),5253.
40
Christman, Politics of Persons, 155156.
41
preferences were formed? No doubt young women who adopt traditional fe-
male roles would think it entirely appropriate that they were taught to do so by
their mothers. As Paul Benson argues, in cases in which oppressive norms are
very deeply ingrained it is not plausible to think that the agent would repudiate
or feel alienated from desires based on the norms even after reflecting on their
process of formation.42
Moreover, on Christmans account, forming a different preference in a
situation of reduced options to avoid or eradicate actual (or counterfactual)
cognitive dissonance may well be a paradigm case of autonomous preference
formation. By hypothesis, the foxs desire for the grapes, combined with their
unavailability, led to alienation. Hence, in resolving the cognitive dissonance
by repudiating the desire for the grapes, the fox eradicates his alienation and
seems to achieve autonomy on Christmans account. Similarly, the adaptive
preference of girls for traditional female roles (by hypothesis) is the result of
the inaccessibility of nontraditional options in the patriarchal context. Girls
turn away from inaccessible nontraditional options in part to avoid the frus-
tration of desiring the inaccessible. Hence, alienation would accompany the
inaccessible, rejected preference, not the one the agent actually adopts. On
Christmans view, the preference for the traditional, feasible, option is the
autonomousone.
One option for Christman would be to bite the bullet:since alienation is not
or would not be experienced by agents whose preferences are unconsciously
adapted to oppressive conditions, these preferences are not autonomy impair-
ments. However, there may be further resources available on his historical and
procedural account that will help to demarcate autonomous adaptive prefer-
ences from nonautonomous ones.43 The hypothetical reflection condition, al-
though it purports to offer a historical condition of autonomy, actually requires
only that the agent reflect in the present about the historical formation of her
desires. Procedural theorists have noticed that critical reflection in the present
is not sufficient for autonomy and claimed that present reflection must not
be the product of a distorted causal mechanism leading to its formation.44 As
Khader herself points out, an independence of mind or procedural inde-
pendence condition has been considered a necessary condition of autonomy
on standard procedural accounts.45 For instance, Gerald Dworkin argues that
42
Paul Benson, Autonomy and Oppressive Socialization, Social Theory and Practice 17
(1991):385408.
43
Iam grateful to Mark Piper for suggesting this possibility.
44
Gerald Dworkin, The Theory and Practice of Autonomy (Cambridge: Cambridge University
Press, 1988); Mele, Autonomous Agents.
45
Khader, Adaptive Preferences and Womens Empowerment, 74. Khader goes on to ask whether
various conditions mentioned in procedural theories of autonomy would count adaptive preferences as
nonautonomous, but she does not explore procedural independence. She does discuss Elsters idea that
adaptive preference formation is an unconscious process that operates behind the agents back under
the heading life-planning as personal history (85). However, her objection to this possibility seems
certain influencessuch as hypnotic suggestion, manipulation, and coercive per-

suasioncan subvert agents critical faculties and undermine their procedural
independence.46 Christman himself mentions that hypothetical critical reflection
must not be constrained by reflection-distorting factors such as constriction,
pathology, or manipulation or being denied minimal education and exposure
to alternatives.47 Indeed, in early work, Christman compares happy slave cases
in which the slave has expunged her desires for freedom only as a result of the
oppressive presence of the restraints she faces with slaves who continue to resist
their circumstances and desire their freedom.48 He claims that if preference for-
mation resulted from the very presence of a...restraint, bearing down on the
agent and forcefully causing a change in the desire, it is implausible that the de-
sire change is an increase in freedom.49 In other words, the accommodation of
a happy slave to her circumstances does not make her autonomous; rather, the
presence of coercive, external restraints that inhibit the free formation of her pref-
erences constitutes an autonomy impairment.50
Thus, if we understand adaptive preferences using the psychological pro-
cesses model, the test of procedural independence could be used to iden-
tify adaptive preferences that are also nonautonomous. The question then
becomes whether the causal and psychological mechanisms that produce de-
formed desires render them autonomy impairments. Khader considers several
examples of the adjustment of options to possibilities that have been labeled
deformed in development ethics. One is that of intrahousehold food distri-
bution.51 Some women in South Asia starve or malnourish themselves to feed
their husband and male children. Khader points out that such preferences are
related in complex ways to religious and gender norms that link self-deprivation
and self-discipline. She describes the case of an elderly Javanese woman [who]
recalled being told as a child that women needed to discipline themselves, be-
cause they were superior to men who could not control themselves.52 Khader
off the mark. She writes, Thinking of [adaptive preferences] this way commits us to a dubious meta-
physical position...that there is one authoritative narrative about why a person forms a preference at
the moment that it happens (85). Ido not believe that Elster is presupposing this metaphysical com-
mitment. He is claiming that there is a psychological and theoretical difference between unconscious
(adaptive) causal mechanisms and deliberate, conscious and plannedones.
46
Dworkin, Theory and Practice of Autonomy,18.
47
Christman, Politics of Persons, 147,155.
48
John Christman, Liberalism and Individual Positive Freedom, Ethics 101 (1991):343359,354.
49
Christman, Liberalism and Individual Positive Freedom,353.
50
Arecent discussion of adaptive preference formation explicitly adopts the position that adaptive
preference formation is an unconscious process that violates procedural independence. Ben Colburn,
Autonomy and Adaptive Preferences, argues that an independence test fails when the formation of a
preference is subjected to covert influence. Like subliminal influence, covert influence is a mechanism
that produces desires in agents through a process in which the agent is not aware of the causal expla-
nation of her preferences.
51
52
notes that this woman seems committed to her values and reflective about the
ways her values inform her decision to malnourish herself. Hence, Khader
thinks there is no autonomy deficit on procedural accounts that employ a
critical reflection condition. However, on the account Iam now considering,
being reflective in the present is not sufficient for autonomy. The question is
whether the preference to deprive oneself of food due to the internalization of
misogynistic norms is procedurally independentthat is, whether the influ-
ence on the formation of the preference constituted coercive persuasion or was
reflection-distorting due to insufficient exposure to alternative possibilities or
some other factor.
Whether or not a preference fails a test of procedural independence is
in principle a case-by-case question. However, if theorists of oppression are
correct, we can make some general claims about the desires of members of
oppressed groups:they are often formed by processes that are coercive:in-
doctrination, manipulation and adaptation to unfair social circumstances.53
Thus, prima facie, the desires of members of oppressed groups for their own
oppression fail a test of procedural independence such as that of coercive
persuasion.
Procedural accounts of autonomy therefore potentially employ condi-
tions that would classify adaptive preferences as autonomy impairments. The
problem for such accounts is whether the processes that violate procedural
independence, like coercive persuasion, can be spelled out in a satisfactory
way without importing background moral conditions. Cudd argues that a
purely empirical account of coercion is unsatisfactory because it cannot dis-
tinguish between a hard choice and a forced choice.54 Similarly, to flesh out
the reflection-distorting influences that correspond to lack of minimal educa-
tion or inadequate exposure to alternatives, we need some normative account
of what counts as minimally satisfactory education or adequate alternatives.
Therefore, although procedural theorists can explain how desires for oppres-
sive conditions are nonautonomous, the explanation comes at the expense of
maintaining the moral neutrality of their own theory.
2. The Freedom to do OtherwiseModel
In the last section we considered a psychological processes model of adaptive

preference formation that was derived from Elsters distinction between adap-
tive preference formation and the similar processes of learning and deliberate
53
54
Cudd, Analyzing Oppression, 126129. Iam therefore in broad agreement with Khader when she
says that procedural theories need to be [supplemented] with a theory of the good. Khader, Adaptive
Preferences and Womens Empowerment,95.
character planning. On that model, adaptive preferences are accommodations

to feasible options that lack autonomy due to a distorted underlying causal
mechanism. Elster also offers an analysis of adaptive preferences that employs
the notion of freedom to do otherwise:
We can exclude operationally one kind of non-autonomous wants, viz adap-
tive preferences, by requiring freedom to do otherwise. If I want to do x,
and am free to do x, and free not to do x, then my want cannot be shaped
by necessity.55
On this second model, preferences are adaptive and nonautonomous because
they fail the freedom to do otherwise condition and are shaped by necessity.
The Javanese woman forms a desire to malnourish herself only because she is
not free to flout the cultural and religious norms that lead to the desire. The
girl raised in a household in which she is always expected to help her mother
with the traditional chores comes to prefer these chores only because other
options are not available toher.
Nussbaum applies the freedom to do otherwise model in her analysis of
the preferences of poor working women in India. One of her examples is that
of Vasanti who chooses to remain in an abusive marriage because she thinks
that although the abuse was painful and bad, but, still, a part of womens lot
in life, just something women have to put up with as part of being a woman
dependent on men, and entailed by having left her own family to move into
a husbands home.56 Nussbaum thinks that Vasantis preference is adaptive;
it is accommodated to her feasible options and lacks the condition of free-
dom to do otherwise.57 Nussbaum points out, however, that there will be many
examples of the accommodation of preferences to feasible options that fail
the test of freedom to do otherwise yet are reasonable and conducive to the
agents well-being. Being a basketball player is not a feasible option for a short
person, and being an opera singer is not a feasible option for someone with a
weak singing voice. Adjusting ones expectations to accommodate the feasible
options in these cases seems perfectly reasonable and a good thingtodo.
In Vasantis case, the desire to remain in an abusive relationship is by hypo-
thesis not a good thing for her. Her choice is the result of her freedom being
constrained in a morally problematic way. Thus, Nussbaum claims that Elsters
morally neutral account of adaptive preferences is inadequate. To distinguish
the basketball player and the opera singer from Vasanti and to explain why
agents like Vasanti have adaptive preferences that lack autonomy, we need a
substantive theory of justice and central goods:Peoples liberty can indeed
be measured, not by the sheer number of unrealizable wants they have, but by
55
56
Nussbaum, Adaptive Preferences and Womens Options,6869.
57
the extent to which they want what human beings have a right to have.58 The
freedom to do otherwise condition must therefore be understood as a moral
condition. If agents preferences are accommodated to external options only
because the options are limited by unjust social circumstances, their prefer-
ences are shaped by necessity in a way that restricts their freedom.
For Nussbaum, adaptive preferences are morally problematic in situations
in which agents face a morally inadequate set of options. In such circum-
stances, when preferences are accommodated to the options, and agents do
not desire what they have a right to have, their autonomy is impaired. Thus,
for Nussbaum, adaptive preferences are morally problematic precisely because
they are (substantive) autonomy deficits. In response to this kind of position,
Khader proposes a perfectionist definition of adaptive preferences. Aprefer-
ence is inappropriately adaptive when it is inconsistent with a persons basic
flourishing, formed under conditions non-conducive to basic flourishing,
and would not have been formed under conditions conducive to basic flour-
ishing.59 Agents like Vasanti satisfy the definition and have inappropriately
adaptive preferences. But Khader claims that such preferences do not corre-
spond to autonomy impairments even on substantive accounts of autonomy.60
Let me address Khaders argument by focusing on her discussion of Joseph
Razs substantive theory.61 Raz proposes that autonomy requires adequate ex-
ternal options in addition to mental ability and independence from coercion
and manipulation by other agents.62 For example, he describes a hounded
womana woman on a desert island who is hounded by a wild animal and
has to spend all her time and energy planning for survival. The woman has a
variety of options in addition to the mental abilities necessary for planning,
but the options she has are inadequate. For one thing, they are dominated
by her one overpowering need and desire to escape being devoured by the
beast.63 For another, having adequate options is understood by Raz as a moral
58
59
60
Khader, Adaptive Preferences and Womens Empowerment, 95106.
61
Khader, Adaptive Preferences and Womens Empowerment, 99102, also discusses other possible
substantive approaches to spelling out the notion of autonomy such as substantive autonomy as being
motivated by good norms. This label is misleading because being motivated by good norms does not
name a conception or theory of autonomy but rather a piece of evidence that a theory of autonomy
will have to explain. For example, Stoljar, Autonomy and the Feminist Intuition, claims that many
feminists think that, when agents are motivated by false and oppressive norms, the agents autonomy is
called into question. Igo on to make a conditional claim:if we accept this intuition, then a strong sub-
stantive theory of autonomy will need to be invoked to explain it. As mentioned already, my discussion
at the time did not properly distinguish between two possible substantive theories:a strong substantive
account that is content based; and a (strong) normative competence account. As Iargue here, Istill
think that a substantive account of autonomy is needed to explain why desires for ones own oppression
seem to be autonomy impairments. But Ido not now think this has to be a strong substantive account
that is contentbased.
62
63
Raz, Morality of Freedom, 374376.
requirement: autonomy requires that many morally acceptable options be

available to a person.64
Khader interprets Razs test as follows:autonomous preferences are prefer-
ences that reflect an agents own preferences and are consistent with her flour-
ishing.65 She considers the case of a person who chooses to be a bullfighter
and argues that, since this a dangerous sport, it is not compatible with the
agents flourishing. Khader suggests that choosing bullfighting would count
as an autonomy impairment on Razs account and further that, since choos-
ing bullfighting is not intuitively an adaptive preference, the category of adap-
tive preferences cannot coincide with the category of autonomy impairments
delivered by Razs account.
This argument is too quick. Choosing bullfighting may be dangerous and
incompatible with flourishing (in some sense), but if the agent has a variety
of morally adequate options in addition to bullfighting the agent and her
choice to pursue bullfighting are nevertheless autonomous on Razs account.
The agent must have the capacity to choose the good, that is, a choice among
morally adequate options, but there is no requirement that particular choices
are consistent with flourishing. Suppose we assume that bullfighting consti-
tutes a morally bad option because it is dangerous and undermines the agents
well-being. (Note that this is not a very plausible assumption because not all
dangerous life choices are morally impermissible.) Even this assumption does
not negate the agents autonomy but only the value of the exercise of autonomy
in the particular case. Raz argues that autonomy is valuable only if it directed
at the good....Autonomy is consistent with the presence of bad options, [but]
they contribute nothing to its value.66 Choosing a bad option in a situation
in which the agent has a variety of morally acceptable ones is an autonomous
choice, though not a valuable one.67 Conversely, agents like Vasanti who are
living in conditions of severe deprivation lack morally acceptable options and
hence are obvious cases of nonautonomous agents on Razs theory. Coming to
64
65
Khader, Adaptive Preferences and Womens Empowerment, 102. Emphasisadded.
66
67
Ihave based my discussion on one possible reconstruction of Khaders argument. Her remarks
are very brief, so Imay have misunderstood what she has in mind. Here is another possible reconstruc-
tion:she seems to take Razs position as equivalent to a strong substantive view that employs normative
constraints on the content of preferences. She claims that adaptive preferences cannot be defined using
contents alone. (This is correct; a preference for bullfighting may be or may not be adaptive and non-
autonomous.) Hence she may be saying that adaptive preferences do not correspond to the category
of nonautonomous preferences on Razs account because the latter is a content-based account. Ithink
however that the analysis of Raz as offering a content-based account is mistaken. As we saw, he thinks
that autonomous agents can have preferences with bad or immoral content. Hence his account is sub-
stantive in a different sense from strong substantive accounts. The moral constraints are derived from
the notion of adequate options, not from the immorality of the content of agents preferences.
prefer ones severely limited options (i.e., adapting ones preferences to ones
feasible option set) corresponds to exercising impaired autonomy.
In general, adaptive preferences as Khader defines them would count as
autonomy deficits on Razs adequate options test. On her account, basic flour-
ishing or well-being is defined using a minimal, vague and cross-culturally ac-
ceptable conception of the good.68 It is clear that, if an agent lacks the options
required for minimal flourishing, she would also be deprived of morally ad-
equate options. Thus, a preference that is formed only because better options
are inaccessible to the agentthat is, a preference that would be classified as
adaptive on Khaders definitionalso would be classified as nonautonomous
on Razs approach to autonomy.
Khader acknowledges that adaptive preferences on her model are likely to
be ruled nonautonomous on Razs account because they are shaped by inad-
equate options. But she claims that calling [adaptive preferences] autonomy
deficits and incorporating a conception of the good into autonomy leads us to
policies that are decidedly illiberal.69 She worries that if we characterize adap-
tive preferences in response to oppression, poverty, and deprivation as lacking
autonomy, this will license coercive governmental policies that will override
the voluntary choices of such agents, thereby compounding their oppression
and deprivation. Thus, Khaders main objection to characterizing adaptive
preferences as autonomy deficits is a moral one.70 I postpone discussion of
moral objections to Section4.
Up to now, the discussion has focused on the intersection of Razs account
of autonomy and the adaptive preferences of agents in conditions of severe
deprivation. It may be more difficult to invoke Razs theory to account for the
deformed desires of agents in affluent societies. Notice first, however, that
while oppressed people in Western countries often live in conditions in which
minimal flourishing is possible, it does not follow that the conditions required
for minimal well-being rise to the level of adequate options. Second, as Raz
and others point out, autonomy comes in degrees. The options of members of
oppressed groupsfor education, housing, social services, medical care, em-
ploymentare often compromised by comparison with the options of mem-
bers of groups that are not oppressed.
Consider a phenomenon that Cudd describes as oppression by choice.71
Cudd observes that in a labor market in which the average wage for women is
significantly less than it is for men, it might be rational for a mother rather than
68
69
70
Cf. Ann Cudd, Review of Khaders Adaptive Preferences and Womens Empowerment,
Notre Dame Philosophical Reviews, 2012, http://ndpr.nd.edu/news/27280-adaptive-prefe
rences-and-women-s-empowerment/.
71
Cudd, Analyzing Oppression, 146153.
a father to choose to be the primary caregiver of children, that is, to choose

either to work part-time work or not to work at all. At the individual level, it
is a rational choice because overall the family will be financially better off. But
Cudd points out that when the choice is made by many individuals it leads
to a vicious cycle in which structural inequalities are reinforced. By choosing
to stay home and become the primary caregiver of children, the mother will
become a domestic specialista specialist in work that is unpaid and under-
valued as a contribution to societywhereas the father will acquire experience
and seniority in whatever area of paid employment he takes on. When many
women make the same individual rational choice, this in turn will reinforce
social perceptions that women are primarily unpaid domestic workers rather
than potential wage earners. The cycle is oppressive because as a result of the
individual womens choices, employment opportunities are continually de-
graded both for the individuals and for women as a group.72
Cudd argues that although women in affluent Western societies are thought
to have occupational free choice, in fact their options are morally constrained
(relative to those of men); moreover, their options become progressively more
constrained as a result of the choices that are initially made rationally in re-
sponse to the structural inequalities. The argument suggests that women
and members of oppressed groups in generalhave fewer morally adequate
options and hence less autonomy than comparable members of nonoppressed
groups. The phenomenon of oppression by choice is not equivalent to adaptive
preference formation because the former involves deliberate planning whereas
the latter does not. Nevertheless, in both cases agents freedom to do otherwise
is compromised relative to that of members of groups who are not oppressed.
3. Choosing among Bad Options
I argued in sections 1 and 2 that prominent theories of autonomy, both pro-

cedural and substantive, have the resources to classify many adaptive prefer-
ences as autonomy impairments. In this section Iturn to a differently framed
challenge to deformed desires, namely, the position that preferences adapted
to the circumstances of oppression are not deformed. It is claimed that, rather,
they are a rational (and therefore autonomous) response to circumstances in
which agents have limited options. For example, Uma Narayan argues that
they are the outcomes of a rational process of bargaining with patriarchy,73
and Harriet Baber proposes that deformed desires correspond to the second
72
73
Narayan, Minds of TheirOwn.
best alternative that is available to the agent under the circumstances. When
the best alternative is inaccessible, it is rational to choose second best.74
The different versions of the challenge diverge in an important respect.
Baber, whose focus is to provide a defense of preference utilitarianism against
Nussbaums critique, argues that many preferences that are called adaptive
are not in fact genuine preferences at all. Choices in response to what she calls
a raw deal may not be what agents really prefer. On the other hand, feminist
critics like Narayan (and Khader), whose focus is to defend womens agency,
claim that preferences for sexist norms or bad circumstances are genuine pref-
erences. In many cases these preferences are reinforced by other goals that the
agent takes to be valuable. Khader discusses women who opt for genital cut-
ting or clitoridectomy. She points out that often women are motivated by the
wish for community belonging and social recognition, in particular by the goal
to promote the marriageability of their daughters.75 Similarly, the aforemen-
tioned preference for malnourishment of the Javanese woman is reinforced by
religious conviction. Womens preferences seem genuine when considered in
conjunction with these other factors.
Narayan and Baber do, however, offer parallel arguments for the ration-
ality of putative deformed desires. Agents living under oppression or deprived
conditions are capable of utility maximization; they are capable of engaging
in costbenefit analyses that weigh up the options available and of making
a choice among options on the basis of this analysis. In particular, they note
that most agents rank bundles of options, not options in isolation. Consider
Narayans example of a community of women in India, the Sufi Pirzadi, who
live in relative purdah (seclusion) within the home and are expected to veil
when they are in public.76 These women acknowledge that purdah severely
limits their education and mobility and has the effect of making them de-
pendent on male members of the community. But they also explicitly recognize
benefits, for instance, that veiling signifies womanly modesty and propriety
and their superior standing vis--vis other Muslim women.77 She argues that
due to their cultural and religious context these women cannot separate out
preferences that limit their movement and promote their dependence from
those that promote the piety and modesty that they value:the options come
as a bundle.
Baber makes the same point in her analysis of Vasanti. She points out that
Vasanti may prefer a bundle of options including abuse and having a roof over
her head over a bundle that contains no abuse and being on the street. It is
not irrational to choose the former bundle but rather an example of rational
74
Baber, Adaptive Preference.
75
Khader, Adaptive Preferences and Womens Empowerment, 100101.
76
Narayan, Minds of Their Own,420.
77
Narayan, Minds of Their Own, 420421.
calculation under oppressive conditions in which the agent gets some advan-
tages and forgoes others:
We might with equal justification understand Vasantis decision as the result
of a utility calculation given a reasonable assessment of her options and the
probabilities of various outcomes. Vasanti recognizes that given her circum-
stances, staying in an abusive marriage is her best bet if she wants to have
a home and basic necessities:even if she would rather avoid getting beaten,
she is prepared to take on that cost in order to avoid her least preferred out-
come:homelessness and destitution.78
Thus, for both Baber and Narayan, putative deformed preferences are not de-
formed but rather instances of the exercise of rational agency in which agents
rank bundles of options under difficult circumstances.
Baber offers a second argument as well:although Vasanti and other agents
may appear to have preferences for oppressive conditions, in fact we cannot
infer that they prefer what they choose:making the best of a raw deal when
no other alternatives are available is not the same as preferring it.79 Consider
the fox. According to Baber, the foxs preferences have not changed; he is only
pretending to himself that he does not value or want the grapes. If a bunch
of grapes suddenly became accessible to him, he would jump at them, and
hence, his preference for the grapes persists.80 The fox engages in a rational
process of settling for second best. The foxs possible options are (1)grapes and
no felt frustration, (2)no grapes and no felt frustration, and (3)no grapes and
felt frustration.81 Although the fox cannot have the best option (1), he pre-
fers serenity over felt frustration and thus eradicates the felt frustration that
comes with continuing to want the grapes. Similarly, Vasanti does not really
prefer domestic abuse. She would jump at a better situation were it accessible
to her. But staying under her husbands roof and putting up with the abuse is a
rational choice because it is better than the alternative, namely, no abuse and
homelessness.82
These arguments contain two basic ideas. The first denies that the desires
of agents who endorse apparently harmful practices such as clitoridectomy or
practices in which their rights are curtailed such as purdah are deformed. On
the contrary, these agents are rational choosers with complex motivations who
are making the best of the circumstances in which they find themselves. The
second idea is that there is a difference between what the agent chooses and
78
Baber, Adaptive Preference, 113114.
79
Baber, Adaptive Preference,114.
80
Preferences are understood as behavioral dispositions, not occurrent feelings. Baber, Adaptive
Preference,312.
81
82
seems to prefer, and what they really prefer. Since many agents like Vasanti, if
offered better options, would jump at them, what they choose is not what they
really prefer. It is a compromise in the face of a terrible situation in which what
they really prefer is inaccessible tothem.
One can agree with both of these ideas. Yet neither establishes that agents
like the Sufi Pirzadi, Vasanti, or for that matter women in Western contexts
who appear to adapt to oppressive conditions are autonomous. Engaging in
a costbenefit analysis may be sufficient for formal rationality yet insufficient
for autonomy.83 As we have seen, theories of autonomy typically distinguish
between competency conditions and additional conditions such as authen-
ticity or availability of adequate options. Thus, it is possible that the agents
under consideration satisfy competency and formal rationality conditions yet
nevertheless fail to be autonomous. For the same reasons, Babers argument
that agents like Vasanti make rational choices that might enhance their sub-
jective well-being given their options is also not incompatible with attributing
lack of autonomy to Vasanti and agents like her. The ability to improve ones
well-being under conditions in which ones options are curtailed is not equiv-
alent to exercising autonomy.
Indeed, not only is Babers analysis not incompatible with attributing an
autonomy impairment, it also provides an implicit argument that the choices
of agents like Vasanti are not autonomous. In effect, Baber is distinguishing
between apparent and true preferences.84 She thinks that reversible adaptive
preferences are not true preferences and that the true preferences are re-
vealed by the behavioral dispositions to jump at alternative better options.
Hence, the fox and Vasanti are deceiving themselves about what they truly
prefer. Many theories of autonomy, however, would consider self-deception
and autonomy incompatible. For instance, Diana Meyers argues that one of
the skills required for agents to be autonomous is that of self-discovery; agents
who are blind to their own true preferences are not exercising this skill and
are not autonomous.85 The reversibility of a preference therefore would be evi-
dence that it is not autonomous because reversibility suggests that it is merely
apparent not real. (Indeed, this may be the intuition behind Elsters original
classification of adaptive preferences as reversible and nonautonomous and
hence unlike the irreversibleand autonomouspreferences that come about
through learning.)
Other theories suggest that apparent self-deception is compatible with
autonomy, but only if self-deception occurs in the service of right reasons.
For example, Henry Richardson argues that an agent is autonomous if she
83
Stoljar, Autonomy and the Feminist Intuition.
84
Nussbaum, Adaptive Preferences and Womens Options, 73, discusses this line of thought while
referencing John Harsanyis distinction between manifest and true preferences.
85
See, e.g., Meyers, Self, Society and Personal Choice,47.
acts on a conception of objective moral reasons.86 He compares the foxs rea-

soning with that of the character Bully Stryver in Charles Dickenss novel A Tale
of Two Cities. Mr. Stryver selects the beautiful and kind Lucie Manette as his
future wife. When she refuses, he persuades himself and others that he never
loved her or wished to marry her. In other words, he adopts the foxs strategy of
self-deception. He denies that he ever wanted to marry Lucie; he shields himself
from significant loss and eliminates the frustration and embarrassment at being
refused. Richardson claims that nevertheless Mr. Stryvers adaptive preference
formation in this case is autonomy preserving because it promotes self-respect
and thereby corresponds to acting on a conception of (actual) moral reasons.87
Richardsons analysis could be applied to Khaders examples of adaptive prefer-
ence formation, such as that of the Javanese woman who undernourishes herself
in part because she thinks it promotes spiritual enlightenment. Given the cultural
context, the preference may be an instance of acting on actual moral reasons such
as self-respect. But in other cases, including perhaps that of Vasanti, if agents
preferences are not consistent with acting out of self-respect they would count as
autonomy impairments.
Babers discussion raises a further issue. It is not typical that preferences for
oppressive conditions can be shown to be unstable and hence for that reason
to be merely apparent. It is not the case that women, if given the opportunity,
would always jump at the chance of alternative nonsexist roles. For example,
there is evidence that the norm that only men are appropriate heads of house-
holds is remarkably resistant to shifts of context. Arecent article in the NewYork
Times Magazine discusses the situation of families in a small town in the United
States in which due to a factory closure many men who had been employed
and relatively affluent lost their jobs.88 In several of these families, the wives,
although they had not previously worked, had become successful primary
breadwinners for the family. At the same time, the men struggled with adjust-
ment to being financial dependent on their wives, and both husbands and wives
clung to traditional roles including the designation of the husband as head of
household. Hence there is evidence that cases in which agents adopt oppressive
norms (or social roles that are formed by oppressive conditions)89 are instances
86
Henry Richardson, Autonomys Many Normative Presuppositions, American Philosophical
Quarterly 38 (2001):287303.
87
Richardson, Autonomys Many Normative Presuppositions, 292. One could interpret
Richardson as claiming that agents who act in the service of right reasons are not really self-deceived be-
cause acting on right reasons is doing what rational agents really want to do. Sarah Buss, Autonomous
Action:Self-Determination in the Passive Mode, Ethics 122 (2012):647691, 666, captures the gist of
this kind of analysis of autonomy when she writes:Every rational agent wants to do what she (really)
has reason to do; so, insofar as her action is not (adequately) responsive to (the real) reasons for and
against acting this way, it does not (adequately) express her defining desire; so there is an important
respect in which this action is not attributable to her [i.e., not autonomous].
88
Rosen, Who Wears the Pants in This Economy?
89
Cudd, Analyzing Oppression, 181, points out that there is a difference between desires to subject
oneself to oppression and desires for social roles that are the effects of oppression:It is not that [the
of stable preferences in which agents have adapted to patriarchal social struc-

tures. Baber considers an analogous case of a Cambodian prostitute, Srey Mom,
whose freedom was bought from the brothel owner by a NewYork Times jour-
nalist.90 Although subject to appalling conditions in the brothel, Srey Mom had
come to prefer some if its material advantages such as jewelry and a cell phone.
When given the opportunity to return to her village, she does so, but after a
while she returns to prostitution. Baber proposes that the preference for a life
of prostitution is a stable and genuine adaptive preference; moreover, given the
other options available to Srey Mom, it is too quick to conclude that returning
to prostitution is not in her interests or that it fails to promote her well-being.91
These examples reinforce the point that although the reversibility of a
preference may be evidence that it is merely apparent, it does not follow that
stable and irreversible preferences are always autonomous.92 As we saw sec-
tions 1 and 2, even stable and irreversible preferences might fail the tests of
procedural independence or adequate options. Similarly, even if a choice pro-
motes an agents subjective well-being because it is the best of the bad options
available (as in the case of Srey Mom), this does not imply that the choice is
autonomous.
4. The Possibility of Coercion and Other Worries
Many of the objections to treating deformed desires as autonomy impairments

are not conceptual but moral, pragmatic, and empirical. Khader often claims
that characterizing agents with adaptive preferences as nonautonomous would
license coercive policies. For example, she writes that if people whose pref-
erences do not manifest a value for their own independence are not autono-
mous, public institutions may reasonably coerce those people into changing
their preferences.93 She also thinks that the characterizations of agents desires
as deformed and nonautonomous are problematic empirically because it tends
to overlook the complexity of their motivations and that this may result in
policies with an ineffective focus.94 Baber emphasizes the point that over-
simplified accounts of the motivation of agents in deprived circumstances can
oppressed] will prefer oppression to justice or subordination to equality, rather they will prefer the
kinds of social roles that tend to subordinatethem.
90
The Srey Mom example is taken from Nicholas D. Kristof, Bargaining for Freedom, NewYork
Times, January 21,2004.
91
Khader, Adaptive Preferences and Womens Empowerment, 130, argues that preferences such
as that for malnourishment could enhance the interests or well-being of the agent, for example,
self-depriving behavior often elicits actual rewards.
92
Cf. Elster, Sour Grapes,221.
93
94
be empirically inaccurate and morally problematic. For example, she criticizes

Nussbaums reference to Vasanti being in a slumberous state before being
exposed to programs in which she came to be aware of her rights.95 Indeed,
many theorists are concerned that denying that womens adaptive preferences
are autonomous is equivalent to denying their agency. In this section, I at-
tempt briefly to dispel these worries.
I do not believe that denying autonomy is tantamount to denying agency.
Autonomy and agency are not equivalent concepts. As we have seen, agents
that are judged to have autonomy impairments often retain the full gamut
of complex mental capacities. They retain the mental abilities required for
planning and weighing up competing options. They retain abilities that are
closely related to autonomy such as forms of rationality and the capacity for
self-control.96 Khader articulates a notion of adaptive preference that does not
rely on the notion of autonomy. Yet she acknowledges that various aspects of
agency may be damaged in agents who have adaptive preferences and that
noncoercive interventions may be justifiable to attempt to improve agents de-
liberative capacities.97 Her argument itself suggests that agency and autonomy
are distinctideas.
The wish to defend a robust conception of womens agency under condi-
tions of oppression is underwritten by a very legitimate concern. Focusing
on deficiencies in agents psychologies may end up psychologizing the struc-
turalthat is, putting causal responsibility for oppression inappropriately on
the agent rather than appropriately on the unjust social conditions.98 Remarks
such as that womens agency is pulverized by patriarchy or that Vasanti was
in slumberous state or even that desires for ones own oppression are de-
formed are rhetorical forms of words that potentially reinforce the problem.99
Iagree that these labels are misleading and better avoided.
The negative labels also obscure the subtleties of the autonomy analysis of
adaptive preferences. For instance, theories of autonomy distinguish between
local and global autonomy. When theorists of autonomy claim that a pref-
erence or desire is nonautonomous, this usually means that the conditions
for local autonomynamely, what is required for particular choices, prefer-
ences, or desires at particular times to count as autonomousdo not obtain.
Vasantis decision to stay in an abusive marriage may well be locally nonauto-
nomous. If it is, her degree of overall autonomy is reduced, but her agency is
relatively unaffected. Global autonomy corresponds to an agents ability to lead
95
Nussbaum, Adaptive Preferences and Womens Options, 74; Baber, Adaptive Preference,
114,126.
96
Mele, Autonomous Agents.
97
See, e.g., Khader, Adaptive Preferences and Womens Empowerment,33.
98
99
Narayan, Minds of Their Own, 422, refers to a remark of Catherine MacKinnon that womens
agency is pulverized.
an autonomous life. Global autonomy theorists typically introduce external

conditions such as inadequate moral options or socio-relational conditions.100
For instance, Marina Oshana argues that autonomy is a temporally extended,
global condition of agents in which they have de facto power and authority
over choices and actions significant to the direction of [their lives].101 Severely
constraining external conditions remove the de facto power required for au-
tonomy. Hence, global theories do not usually focus on agents psychological
capacities at all; they do not impugn agency or psychologize the structural.
Indeed, on global theories, it is precisely the structural conditions that impair
autonomy.
Let us now turn to the question of coercion. Khader notes that for liberal
political theory, autonomy [is] the capacity that exempts people from being
subject to coercion.102 She claims that, once agents are judged to be lacking
autonomy, liberals do not have the resources in their theory to block policies
that override agents voluntary choices. Let us suppose it is true that for liber-
als respect for agent autonomy is the most important moral reason for treat-
ing coercive interference with agents as illegitimate. This does not entail that
coercive policies are always morally permissible in cases of nonautonomy or
lack of autonomy:it does not entail that public institutions may reasonably
coerce...people into changing their preferences. Indeed, coercive policies
aimed at changing peoples preferences may well be self-defeating if they harm
or damage agents autonomy further under the guise of promotingit.
Khader imagines a policy that denies poor women health care unless they
stop the practice of malnourishing themselves to feed their male relatives.103
Putting aside the fact that denying malnourished people health care would
be criticizable on the grounds that it would exacerbate the physical harm they
are already suffering, such a policy would be self-defeating. Autonomy is a
matter of degree and there are different dimensions of autonomy. It would be
self-defeating for a policy whose aim is to augment autonomy to further un-
dermine what little autonomy agents have or to undermine one dimension of
autonomy while promoting another. On Razs account, for instance, a policy
that subjected people to direct coercion would inhibit an important dimension
of autonomy.
The default position for liberals, in part due to respect for the value of au-
tonomy itself, is that coercive policies are not justified unless the goal is to
prevent harm.104 In the West, this issue can arise for example in cases of mental
100
For inadequate moral options, see Raz, Morality of Freedom. For socio-relational conditions, see
Marina Oshana, Personal Autonomy in Society.
101
102
103
104
See, e.g., Raz, Morality of Freedom.
illness or religious values. Suppose a person with a mental illness lacks the
competencies required for full autonomy. From a liberal perspective, there is
no justification for subjecting this person to any sort of coercive policy unless
she is likely to harm herself or others. Similarly, preferences against certain
forms of life-saving medical intervention can be based on religious beliefs.
Jehovahs Witnesses think that blood transfusions are contrary to biblical
injunctions. Even if a theory of autonomy judges this religious conviction to
be the result of indoctrination and hence (for instance) to be incompatible
with the requirement of procedural independence, there is no justification for
subjecting Jehovahs Witnesses to coercive policies to change their convictions.
Coercive intervention such as enforced blood transfusions may be justifiable,
but only as a last resort to preventharm.
Khader seems to accept that the harm principle potentially could provide a
justification of a coercive policy. She writes that we should focus on changing
cultural practices only when they cause some sort of [serious] harm or wrong
to individuals.105 In the case of genital cutting, for example, there is wide var-
iation among physical practices as well as among social benefits that ensue.
The practice does not always constitute serious harm.106 Khader is right to say
that the mere fact that an agents endorsement of the practice of genital cutting
is driven by a cultural belief that is false and misogynistic does not provide a
justification for coercive intervention. But there is stronger position running
through Khaders critique: namely, that noncoercive policies, if their aim is
to change peoples cultural beliefs or conceptions of the good, are never per-
missible because cultural beliefs and conceptions of the good are delivered by
autonomous agency.107
This position is implausible. Although liberals (especially proponents of
procedural theories of autonomy) put a lot of weight on respecting individual
agents conceptions of the good, nevertheless they endorse the value of au-
tonomy and some version of the harm principle. Since some conceptions of the
good are harmful or inconsistent with agents own autonomy, not all concep-
tions of the good are equally valuable. As Catriona Mackenzie argues, ruling
out coercive political means for promoting autonomy...does not entail ruling
out other political means for encouraging citizens to pursue valuable goals
for example, incentive and reward schemes...; health promotion campaigns;
funding subsidies for the arts, and so on.108 Indeed, since autonomy is a value
105
106
107
Khader seems to thinks that noncoercive interventions are permissible to promote, for example,
deliberative aspects of agency, but these would fall short of policies whose aim is to change conceptions
of thegood.
108
Catriona Mackenize, Relational Autonomy, Normative Authority and Perfectionism, Journal
of Social Philosophy 39 (2008):529.
liberals endorse, noncoercive policies that attempt to promote the value of au-
tonomy may be a requirement of justice on liberal political theories.109
Conclusion
In this chapter, I set out to examine the consequences for autonomy of the
phenomenon of adaptive preference formation in which the oppressed come
to desire that which is oppressive to them.110 Feminists often seem to sug-
gest that internalized oppression ipso facto impairs autonomy and that the
notion of adaptive preference offers support for that conclusion. The preced-
ing discussion has shown that the situation is more complicated. Adaptive
preferences are not autonomy deficits by definition, and neither is it the case
that all unconscious accommodations to feasible options count as autonomy
impairments. On the other hand, many of the adaptive preferences of concern
to feminists are autonomy impairments. Iargued that even if adaptive prefer-
ences are not deformed but rather rational and reasonable decisions in the
face of oppression, this does not show that they are autonomous. Preferences
adapted to oppressive conditions fail the tests introduced by both procedural
and substantive theories of autonomy. The key reason that they fail these
testswhether it is a test of procedural independence or an adequate options
testis that the moral constraints faced by members of oppressed groups due
to their oppression reduce their psychological freedom.
Acknowledgments
For helpful discussion and comments, Iam indebted to Mark Piper, Andrea
Veltman, my fellow contributors to this volume, and the other participants in
the workshop Relational Autonomy:10 Years On, held at McGill University
in September 2012. Iam also grateful to the Social Sciences and Humanities
Research Council of Canada for a grant that supports my research on autonomy
and oppression, and for a second grant that made the workshop possible.
109
Iam grateful to Catriona Mackenzie for many discussions that have helped me think through
theseideas.
110
{PartV }
Autonomy in Social Contexts

{12}
Raising Daughters:Autonomy, Feminism, and

Gender Socialization
MarkPiper
Hence it is no small matter whether one habit or another is inculcated

in us from early childhood; on the contrary, it makes a considerable
difference, or, rather, all the difference.
ARISTOTLE
In August 2011, my wife gave birth to our first child, our daughter, Helena. We
soon found that we had little time or energy to reflect on much beyond the
question of how best to manage the learning curveoften joyful, just as often
overwhelmingforced upon new parents. But when we could reflect, our
thoughts revolved frequently around the central question of how best to meet
the immense responsibility, not just of raising a child, but also specifically of
raising a daughter in a society still saturated with gender-based inequalities of
many powerful kinds. Although not under any illusions that parental instruc-
tion alone, however enlightened, is sufficient to counteract the widespread
and entrenched institutional obstacles to gender equality, we were (and re-
main) convinced that proper parental instruction can go a long way toward
putting girls on the best footing for their future struggle for personal develop-
ment and social equality. But how best to go about this? How best to raise our
daughter in a manner that encourages her future autonomy but at the same
time informs her understanding of her gender in a way that is supportive of
the wider aims of feminism itself? As preparations for this volume proceeded,
these concerns gradually came to the fore of my reflections and prompted me
to write on precisely these topics. My goal in this paper is to argue for a form
of gender socialization for daughters, mediated by parental instruction, that
is best suited to achieving the twin goals of enhancing daughters future au-
tonomy and instilling in them a view of what it means to be a woman in such
a way that they will grow to be strong advocates of the social cause of gender
equality. In the remainder of these introductory comments, Iwish to set out
the problems discussed in this paper more carefully. I will then proceed to
canvass and consider three strategies for a feminist theory of female gender
socialization and argue for the form of female gender socialization that Ibe-
lieve holds out the best promise of meeting the vital goals mentionedabove.
Traditional female socialization processes tend to be autonomy subverting
for women.1 Women are often taught to be subordinate to men, to focus on
feeling rather than thinking, and to confine their attentions to the home.
Girls are presented with stories and images in entertainment and the media
that support the notion that women should be largely focused on gaining ac-
knowledgment from men and should identify their true selves with their
outward appearance. These factors, and many more like them, combine to
tend to make women more submissive, less confident in themselves, and faced
with diminished opportunities for self-development and self-expression. The
result of these factors is the creation of an inequality between the autonomy
prospects for men and women that is disadvantageous for women. Let us
call this the autonomy gender inequality problem. Overcoming this problem
requires initiatives at several levels.2 Among these initiatives, one central chal-
lenge for feminists is to outline a strategy for parentally guided gender so-
cialization that will attenuate the autonomy gender inequality problem. Girls
should be gender socialized in a way that supports and enhances their future
autonomy rather than hindersit.3
It is clear that a solution to the autonomy gender inequality problem should
be part of a feminist theory of gender socialization (henceforth GS). But it is
also the case that a theory of GS that is autonomy supporting does not consti-
tute a complete feminist theory of GS. Afeminist theory of GS also requires
an account of how girls should be raised to view being female in a manner
that is supportive of feminism. In short, it is not evident that raising a girl in
an autonomy-enhancing manner will, by itself, lead her to view her femaleness
in a feminism-supporting way. Hence, in addition to containing a solution to
the autonomy gender inequality problem, a feminist theory of GS must also
include resources for teaching girls to be supportive of the values contained
in feminism in a manner that advances feminist aims in society. Assuming
that both of these aspects are necessary for a feminist theory of GS, however,
1
Simone de Beauvoir, The Second Sex, translated and edited H. M. Parshley (London: Lowe
and Bridone, 1953); Janet Shibley Hyde and Nicole Else-Quest, Half the Human Experience: The
Psychology of Women, eighth edition (Belmont, CA:Wadsworth, 2013); Diana Meyers, Self, Society, and
Personal Choice (NewYork:Columbia University Press, 1989); C. Renzetti and D. Curran, Sex-Role
Socialization, in Feminist Philosophies, edited by J. Kourany, J. Sterba, and R. Tong (Upper Saddle
River, NJ:Prentice Hall,1992).
2
Beauvoir, Second Sex, 680682.
3
It should be noted that many theorists who have dealt with what I am calling the autonomy
gender inequality problem have not described it as a problem of autonomy per se. Often the pres-
ence or absence of other qualities will be pointed out: qualities such as subordination, oppression,
and reduced opportunities. But all of these negatively affect the future autonomy of women and are
included as part of the autonomy gender inequality problem in the presentwork.
Raising Daughters 257
it remains unclear (1) whether a feminist theory of GS can combine these

two aspects at all, and, if they can be combined, (2)what the resulting theory
would looklike.
In this paper Ifirst argue that there is no inherent contradiction in a fem-
inist theory of GS that is both autonomy enhancing and supportive of femi-
nism. Ithen argue for a particular feminist theory of GS that best satisfies the
demands of promoting autonomy and the normative demands of feminism
itself.
1. Definitions and Assumptions
By autonomy Iunderstand personal autonomy, or self-determination:a com-

plex psychological property composed of a constellation of enabling prop-
erties whose possession enables agents to reflect critically on their natures,
preferences, and values; to locate and endorse their most authentic commit-
ments; and to live consistently in accordance with these.4 As many theorists
have noted, the internal capacities associated with autonomy possession can
be organized into two categories: authenticity conditions and competency
conditions. Authenticity conditions refer to the cluster of agential capacities
that allow for accurate introspection of ones authentic self. Competency con-
ditions refer to the cluster of capacities that make it possible to live out ones
authentic self-conception consistently in the face of potential interference by
recalcitrant inner urges and external resistance of various kinds. Autonomy is
thus self-determination in the sense that an autonomous agent determines her
life in accordance with her (authentic) self.5 Authenticity conditions provide
access to the authentic self, and competency conditions allow for the mainte-
nance and promotion of ones authentic self-conception as one goes through
life. Agents have the capacity for autonomy to the degree that they possess
these kinds of supporting capacities and are actually autonomous to the degree
that these capacities are employed to bring about autonomous living within
external circumstances (relations to family and friends, economic conditions,
or sociopolitical arrangements) that are autonomy supporting.6 Those who do
4
Gerald Dworkin, The Theory and Practice of Autonomy (Cambridge: Cambridge University
Press, 1988); Meyers, Self, Society, and Personal Choice; Marilyn Friedman, Autonomy, Gender, Politics
5
For discussion of the question of whether autonomy possession requires normative commit-
ments, see, e.g., Friedman, Autonomy, Gender, Politics; Marina Oshana, Personal Autonomy in Society
(Aldershot, UK: Ashgate Publishing, 2006); John Christman, The Politics of Persons: Individual
Autonomy and Socio-historical Selves (Cambridge:Cambridge University Press,2009).
6
See Joel Anderson and Axel Honneth, Autonomy, Vulnerability, Recognition, and Justice, in
Autonomy and the Challenges to Liberalism:New Essays, edited by John Christman and Joel Anderson
(Cambridge:Cambridge University Press, 2005), 127149; Oshana, Personal Autonomy in Society.
not possess autonomy are heteronomous; these persons lack the aforemen-
tioned capacities and as a result generally live according to determinations that
do not stem from their authentic selves. Such persons include those who are
blindly conforming, brainwashed, or easily manipulated. While not all persons
are actually autonomous, virtually all personswith the exception of the irre-
deemably pathological or handicappedpossess the potential for a minimum
level of autonomy.7
I understand feminism to be an intellectual and social movement with
the goal of achieving justice for women in relation to all social institutions.8
Feminism fundamentally seeks equality for women. It is premised on the view
that women possess status and importance at least equal to men and hence
deserve equal rights and opportunities. Given that there are many forms of
feminism, however, and that some of these forms are at odds with one another,
it is necessary to further specify that Iam predominantly conceiving feminism
at it has been developed in the liberal tradition, and specifically in the tradition
of egalitarian liberal feminism.9 The upshot of this conception is that women
are entitled to the conditions that allow for full social equality as men, and
that achieving this goal requires both that women are enabled to develop per-
sonal autonomy and that women are empowered to be coauthors of the social
conditions under which they develop and live.10 Anatural concomitant of a
commitment to feminism of this kind is the view that girls should be raised in
such a way both as to secure the best chances for their personal autonomy en-
hancement and to be prepared to be vigorous social advocates for the equality
and freedom of all women.11
7
I assume that autonomy of this kind is metaphysically possible. See Catriona Mackenzie and
Natalie Stoljar, Introduction:Autonomy Refigured, in Relational Autonomy:Feminist Perspectives on
Autonomy, Agency, and the Social Self, edited by Mackenzie and Stoljar (NewYork:Oxford University
Press, 2000), 331; Friedman, Autonomy, Gender, Politics. This possibility involves the assumption that
it is meaningful to speak of an authentic self and that it is possible to know ones authentic self (although
it is not necessary to assume that it is possible to know it perfectly). Ialso assume that autonomy of this
kind is not especially uncommonthat most people experience at least marginal increases in this kind
of autonomy as they mature from childhood to adulthood.
8
Sally Haslanger, Nancy Tuana, and Peg OConnor, Topics in Feminism, in Stanford Encyclopedia
of Philosophy (Summer 2013), edited by Edward N.Zalta, http://plato.stanford.edu/archives/sum2013/
entries/feminism-topics/.
9
For an overview of the many forms of feminist theory, see Josephine Donovan, Feminist
Theory: The Intellectual Traditions (New York: Continuum Publishing, 2012). For different feminist
approaches to political theory specifically, see Noelle McAfee, Feminist Political Philosophy, in The
Stanford Encyclopedia of Philosophy (Winter 2011), edited by Edward N. Zalta, http://plato.stanford.
edu/archives/win2011/entries/feminism-political/.
10
For more details, see Amy R. Baehr, Liberal Feminism, in The Stanford Encyclopedia of
Philosophy (Fall 2012), edited by Edward N. Zalta, http://plato.stanford.edu/archives/fall2012/entries/
feminism-liberal/.
11
For the sake of economy, in what follows Iwill use the term feminism to refer specifically to egal-
itarian liberal feminism.
I understand gender to refer to the social roles, positions, behaviors, kinds

of identity, interests, habits, and statuses that are considered as being associated
with a person possessing a particular biological sex.12 This characterization is
intentionally ambiguous between accounts of gender that focus on character-
istics that are typically associated with a biological sex and accounts that focus
on characteristics that ought to be associated with the same. Importantly, this
characterization of gender does not contain the idea that gender is principally
a matter of occupying a subordinate or privileged social position.13 Of course,
gender relations exhibit a tremendous amount of inequality, but Itake this to
be a contingent fact about gender.14
I understand gender socialization as the process whereby children are
instructed on how to understand the nature of, value of, and practices associ-
ated with gender as an institution in society and as a form of life.15 The gender
socialization process usually takes years, and its vehicles of instruction are
manifold. In the present paper, as already emphasized, parental instruction
will be the principal form of gender socialization under consideration.
2. The Coherence of Autonomy-Enhancing GS Geared toward

Inculcating FeministIdeals
Is autonomy-supportive gender socialization compatible with raising girls to

view and value being female in a way that is supportive of egalitarian liberal
feminism and its ideals? On the surface, it might seem as though it is not:for
how could girls be raised for robust autonomy and at the same time be raised
to view and value female gender in a particular, predetermined way? It might
12
Cf. Kate Millet, Sexual Politics (London:Granada Publishing,1971).
13
Contra Catherine MacKinnon, Toward a Feminist Theory of the State (Cambridge, MA:Harvard
University Press, 1989). It might be said that I am also arguing against Sally Haslanger, Gender
and Race:(What) Are They? (What) Do We Want Them to Be? Nous 34:1 (2000):3155, insofar as
she has defined gender with basic reference to facts of inequality. But this critique would be incor-
rect: Haslangers work contains the caveat that she is defining gender to aid in the achievement of
a particular practical purpose. Her conception of gender is pragmatic and ameliorative; she is not
attempting to provide a real definition of gender.
14
Although I deny strict gender realism, I assume that the category of gender is relatively uni-
fied insofar as it satisfies the demands of resemblance nominalism. See Natalie Stoljar, Essence,
Identity and the Concept of Woman, Philosophical Topics 23 (1995):261293; Stoljar, The Politics of
Identity and the Metaphysics of Diversity, in Proceedings of the 20th World Congress of Philosophy,
Vol. VIII, edited by Daniel Dahlstrom (Bowling Green, OH:Bowling Green State University, 2000).
I assume that sex and gender are distinct, contra Judith Butler, Gender Trouble: Feminism and the
Subversion of Identity (New York: Routledge, 1999), 43. I assume that the sexgender distinction
is useful, contra Genevieve Lloyd, The Man of Reason: Male and Female in Western Philosophy
(Minneapolis: University of Minnesota Press, 1984); Elizabeth Grosz, Volatile Bodies: Toward a
Corporeal Feminism (Bloomington:Indiana University Press,1994).
15
Beauvoir, Second Sex; Millet, Sexual Politics; Sally Haslanger, Ontology and Social Construction,
Philosophical Topics 23 (1995):95125.
seem that we can have one ideal or the other but not both. We can raise girls
to be autonomous (but then we cant predetermine key aspects of their value
systems), or we can raise girls to view gender in a way that is supportive of
feminism (but such inculcation seems to rob them of autonomy, at least on
some accounts).
I believe, though, that this is a pseudo-problem. First, the objection may rest
on a misunderstanding. It is not being suggested that girls ought to be raised
in a manner that is autonomy supportive and will result in the inculcation of
a particular value system no matter what. This would constitute a straw man
of the theory being sought. The desire is rather to see if it is possible to have
autonomy-supportive gender socialization coupled with instruction in the
nature and value of gender that tends to be supportive of feminism or presents
feminism in a reasonable, plausible light. Second, the incoherence charge is
vulnerable to reductio problems. If the development of autonomy were incom-
patible with the teaching of any particular values, then no one would be auton-
omous, since everyone is raised (more or less by social necessity) to believe
that some things are to be valued and that some things are not. But that cant be
right:autonomy might not be automatically achieved by all, but it is achieved
by many, to some degree at least. So it must be possible to raise someone in
an autonomy-supportive way while teaching them that particular values are
plausible or reasonable. What is required to have GS that is supportive of both
autonomy development and feminism is a commitment to teaching the com-
petencies that comprise autonomy and to presenting a feminist conception of
gender as a plausible view that has a number of strong reasons in its support.
Although young girls do not possess the autonomy capacities to reflect criti-
cally on the conception of gender they are being taught, these abilities grow
with their autonomy competency, until they are able to decide for themselves
whether to endorse feminist-supportive teachings. To deny that this is possible
is, in effect, to deny the possibility of autonomy. Again, everyone begins life
as entirely heteronomous and gradually obtains greater autonomy potential.
People go on to actualize this potential to varying extents. The move toward
greater self-government and independence of thought is difficult, but it does
happen, even against the background of forms of upbringing that are stringent
or stifling.
3. The Options
Even if we assume that autonomy-supportive gender socialization is com-

patible with teaching girls about the nature and value of gender in a
feminism-supportive way, there are likely different ways to satisfy these two
normative demands in a theory of GS. The challenge is to determine which GS
strategy best satisfies both of these demands.
In what follows Iexamine three different forms of GSby no means an ex-

haustive listthat can be differentiated according to the associated underlying
views of the nature and normative importance of female gender:
1. Non-gendered socialization(NGS)
2. Traditional masculine gender socialization(TMGS)
3. Revisionary feminine gender socialization(RFGS)
The rest of this section is devoted to clarifying these positions. In the fol-
lowing section Ianalyze the positions to determine which one best satisfies
the demands of autonomy enhancement and the promotion of feminism itself.
Non-gendered socialization
NGS has a classic standing in feminist studies. Gayle Rubin argues that femi-
nists should seek to create a genderless (though not sexless) society, in which
ones sexual anatomy is irrelevant to who one is, what one does, and with
whom one makes love.16 According to Richard Wasserstrom, a properly sex-
ually assimilated society would be one in which the gender system is repudi-
ated. In such an ideal society, persons would not be socialized so as to see
or understand themselves or others as essentially or significantly who they
were...because they were either male or female.17 And Susan Moller Okin
argues in Justice, Gender and the Family that a feminist reading of Rawlss
Theory of Justice suggests that if those in the Original Position didnt know
their sex they would probably opt for a genderless society: A just society
would be one without gender.18
Although various forms of NGS are possible, at present I wish to focus
on a particularly robust form of NGS. According to this conception, NGS
holds that gender has no normative importance of any kind. On this view, the
psychological-social aspects of gender are nothing more than arbitrary social
constructs, yet they have become infused with normative importance and
have been used to promote unjust endsmost tellingly, the subordination of
women. Recognizing this arbitrariness, this view holds that we should do away
with the notion that the norms associated with gender generate any demands
for maintenance, loyalty, or uptake of any kind. The very notion of gender dif-
ference is to be abolished. Some humans have different organs than others, but
that is as far as the matter oughttogo.
16
Gayle Rubin, The Traffic in Women:Notes on the Political Economy of Sex, in Toward an
Anthropology of Women, edited by Rayna R. Reiter (NewYork:Monthly Review Press, 1975),204.
17
Richard Wasserstrom, On Racism and Sexism, in Todays Moral Problems, edited by Richard
A.Wasserstrom (NewYork:Macmillan Press, 1979),26.
18
Susan Moller Okin, Justice, Gender, and the Family (NewYork:Basic Books, 1991),171.
NGS holds that childrens socialization should, as far as possible, be denuded

of the notion that boys are (or ought to be) one way, and girls are (or ought to
be) another way, in relation to the social-psychological roles associated with
having a particular sex. It focuses on the idea that all children should be so-
cialized in an extremely open, tolerant, broad way, with no reference to par-
ticular roles, duties, norms, or practices that attend membership in a gender
group. Women who have been raised in accordance with NGS or who have
subsequently embraced its values will have little patience for the notion that
gender underwrites normative demands of any kind or has any intrinsic value.
They will avoid attempts by others to normatively typecast them according to
gender, and their self-conceptions will mirror this:Forget about the category
girl. Iam not to be identified with this arbitrary social construct that has been
projected on to me. Ihave female organsthats as far as the matter goes, or
ought togo.19
Traditional Masculine Gender Socialization

TMGS holds that being a female is, as such, no less valuable than being a male
but that the traditional forms that being female have taken are weak and invite
just the sorts of oppressive treatment that women have historically received. On
this view, women have been kept in a subordinate position throughout human
history by two factors:male oppression and unwitting female complicity. Yes,
women have been oppressed by men, but women have also wrongly accepted
the view that their place in society is secondary, merely supportive (rather
than dynamic and active), pliant, domestic, and weak. Yet being a woman need
not come with such entailments. In all areas except the morally insignificant
domain of physical strength, women can be the equals of men and ought to be.
But actualizing this potential requires repudiating the notion that one should
act like a womanindeed, accepting that dictum is precisely one of the chief
causes of womens oppression. On this view, the very notion of acting like a
woman is tarnished:it is indelibly associated with the idea of self-constraint
and subordination. Being female is just as valuable as being male, and hence
females deserve equality with males. However, to realize this equality, it is nec-
essary to deny the idea that female gender has a positive normative role to
19
An example of NGS in action can be found at Egalia, a preschool in Stockholm, Sweden. Cordelia
Hebblewaite, Swedens Gender-Neutral Pre-school, BBC News Europe, http://www.bbc.co.uk/news/
world-europe-14038419. The teachers there actively work against instilling in children expectations
based on gender in several ways. They avoid using the pronouns him or her when speaking to chil-
dren, instead referring to them as friends, by their first names, or as hen (a genderless pronoun bor-
rowed from Finnish). The schools books are devoid of traditional presentations of gender roles. All the
children are encouraged to choose from toys of all kinds, including those not traditionally associated
with their gender. Boys are free to dress up and play with dolls; girls are free to play with toy tractors.
Egalia is state funded and is proving popular in a country known for its passion for social equality.
play on womens behalf. Rather, women should seek as far as possible to live,
think, and behave like men, for this is the path to empowerment, equality,
self-expression, dynamic creativity, robust self-government, and flourishing.
In TMGS, girls are taught to be direct and willful, to play aggressively, to
seek out and be comfortable in positions of social leadership, to be creative,
to have social ambitions, and to expect to be part of the machinery of social
power. In short, according to this approach all children should receive the
same GS, both in terms of methods and content. This is the only way to pro-
vide all children, girls and boys, with the greatest chance of living the best life
possible.
Revisionary Feminine Gender Socialization

RFGS holds that being a woman is something that is distinctive, special, and
worthy of being honored as such:female gender is a way of being in the world
that is valuable and ought to be not only sustained but also celebrated and, in
ways appropriate to womens life paths, fought for. The adjective revisionary is
applied to this version of GS to indicate a split from traditional ways of con-
ceiving gender roles, in at least two senses:(1)revisionary approaches do not
rest on a sexist axiology that involves notions of an inherent value hierarchy
between males and females; and (2)revisionary approaches are open to the
possibility that being a woman is not necessarily associated with the traditional
way of conceiving itas a matter of being care oriented and the like. Rather,
being a woman can take many different forms, some of which are considerable
departures from traditional notions of femininity.
According to RFGS, almost all of the many different forms that being a
woman can take are legitimate ways of being a woman, as long as they do not
essentially involve accepting a sexist axiology. According to this view, certain
general facts have characterizedand continue to characterizethe category
woman. However, these facts do not constrain in any substantive way the pos-
sible legitimate ways of living that women can choose. On the basis of her
writings, it is reasonable to hold that Judith Butler, for example, would endorse
this form of GS. According to Butler, we should understand woman as a term
in process, a becoming, a constructing that cannot rightfully said to begin or
end....It is open to intervention and resignification.20 It might also be sur-
mised that those who espouse resemblance nominalism in relation to gender
would espouse RFGS.21
On this view, being a woman certainly possesses value of its own sort:it
is important for women to identify themselves as women and to engage in
Butler, Gender Trouble,43.

20
Stoljar, Essence, Identity and the Concept of Woman; Stoljar, Politics of Identity.
21
projects that speak to their gender. Living as a woman, it could be said, has
definite prudential value for the women who do so. Yet on this view woman-
hood is not the most significant fact about a womans normative status. More
important still is the basic condition of being human, or of being a person, or
of being a rational being. The normative grounding for womens basic rights
to equality in RFGS is the stance that one would expect to be associated with
feminists who espouse a sameness approach to feminism.22 On this approach,
women may be distinct in various ways, and embracing this distinctiveness
can be important for the quality of womens lives. However, the ground of
womens rights to equal opportunities and treatment is not to be found in the
value of being a woman as such; it is to be found in a quality that women share
with men alike. Girls raised in accordance with RFGS will acknowledge their
gender but place deeper ethical import elsewhere:Yes, Im a woman, and this
has a value of its own, but more importantly, Ideserve equal treatment because
Iam a person [or, e.g., rational individual or autonomous agent].
As we have seen, this approach allows for significant diversity amongst
the life paths that particular women may take. As such, general commonal-
ties among women are not conceived as binding on what a woman must be.
In RFGS, a woman may choose to become just about whatever she wants.
Following from this inclusivity, parents employing RFGS will raise their chil-
dren to have skills, habits, and resources that have wide formal value, in facili-
tating the successful pursuit of as many different forms of life as possible.
Last, RFGS holds that girls ought to be taughtproudlyto be girls and
ought to be trained to be effective advocates for women. This does not imply
that all girls raised in RFGS should be taught that they must become public
social figures leading the fight for equality, though, for this would run counter
to the very inclusiveness of this approach to GS. Part of the inclusiveness of
this approach is to teach girls that there are many ways to advocate for women.
Certainly leadership in feminist social activism is one path, but there are oth-
ers as well. The teacher or professor who is not intimidated by aggressive male
students or a largely male administrative body, the businesswoman who does
not laugh at misogynistic jokes from her coworkers, the aspiring carpenter
who is not deterred by the absence of women in her chosen trade, or the
homemaker who firmly insists that her unpaid work is at least as important as
paid forms of laborall of these are forms of advocacy for women. If there are
general constraints associated with RFGS, they are only these:that girls will be
raised to be widely competent and prepared for many different possible forms
of life, that they will be taught that there are many valuable ways of being a
woman and that they will be instructed in such a way as to foster the strength
Cf. Martha Nussbaum, Human Capabilities, Female Human Beings, in Women, Culture and
22
Development: A Study of Human Capabilities, edited by Martha Nussbaum and Jonathan Glover
(NewYork:Oxford University Press, 1995), 61104.
and self-esteem to advocate for women in whatever fashion is appropriate to

their natures and vocations.
This brings us back to the principal line of inquiry. We have seen that
autonomy-supportive gender socialization is consistent with raising girls in a
manner that is supportive of feminist values, but which approach to GS best
satisfies the demands of autonomy and feminism?
4. Evaluating the Options

Non-gendered Socialization
Is NGS supportive of girls future autonomy? The answer depends on what
autonomy development requires. As we have already seen, autonomy con-
sists in the possession of internal authenticity and competency conditions.
Authenticity conditions are those capacities that allow an agent to introspect,
to accurately perceive her authentic self, and to form a coherent authentic
self-conception (the self in self-determination). Competency conditions are
those capacities that allow an agent to maintain and live according to this au-
thentic self-conception in the face of various kinds of potential internal and
external interference. Such conditions include, for example, self-control, log-
ical aptitude, proactivity, and reasonable skepticism (to forestall easy accept-
ance of potentially undermining forms of external influence). Such conditions
also arguably include self-reflexive attitudes such as a sense of self-worth or
self-esteem, without which one would not feel oneself worthy of living ac-
cording to ones lights.23 Finally, autonomy requires the presence of external
enabling conditions of various kinds. These include access to a range of options
from which to choose, a minimum amount of security with which to prose-
cute ones life plan24 as well as a number of close relationships through which
one can better come to know oneself.25
One of the greatest strengths of NGS is in relation to the presence of a
wide range of options from which to choose. Without predetermined gender
roles, psychological profiles, or social associations, girls raised according to
NGS would likely enjoy highly increased latitude in relation to the options
open to them. Of course, at present many social institutions are influenced by
established notions about what is proper for a woman to choose and would
23
Paul Benson, Free Agency and Self-Worth, Journal of Philosophy 91:12 (1994):650658.
24
Marina Oshana, Personal Autonomy and Society, Journal of Social Philosophy 29 (1998):81102.
25
Charles Taylor, Philosophy and the Human Sciences:Philosophical Papers 2 (Cambridge:Cambridge
University Press, 1985), 187210; Jurgen Habermas, Individuation through Socialization:On George
Herbert Meads Theory of Subjectivity, in Habermas, Postmetaphysical Thinking:Philosophical Essays,
translated by William Mark Hohengarten (Cambridge, MA:MIT Press, 1992); Anderson and Honneth,
Autonomy, Vulnerability.
bring their pressure to bear on women, but this could be mitigated by a con-
sistent and systematic parental policy in support of diversity. Moreover, there
is nothing in NGS that necessarily works against parental inculcation of com-
petency conditions such as personal discipline, logical aptitude, moderate
skepticism of outside influences, or proactivity in the prosecution of ones life
plan. NGS is hostile to the system of gender roles, associations, and statuses
that characterize traditional Western society, but this hostility does not essen-
tially involve skepticism about the capacities mentioned. Indeed, it is likely
that NGS would promote a more robust inculcation of the same, since girls
raised in such a mannerone that radically opposes an entrenched aspect of
social lifewould require training in just these sorts of abilities to hold fast
to their convictions against the potent general currents of gender conformity.
There are strong objections to NGS, however. For one thing, there is reason
to believe that girls raised in NGS face dangers in relation to developing a
strong sense of self-worth. Such girls run sharply against the conventional
grain and would likely be belittled and combated on several fronts for their
repudiation of the significance of gender roles and norms. There is empirical
evidence to suggest that the distress caused by the experience of gender non-
conformity can lead to self-doubt, which, if internalized, can prompt uncer-
tainty about ones right to flaunt convention and can eventually lead in some
children to a reduced sense of self-esteem.26 And there is also evidence to sug-
gest that girls whose behavior is not geared toward social approval will face
reduced popularity.27 If a sense of self-esteem or self-worth is important for
autonomy possession, as many authors have argued, then such a result would
likely undermine autonomy.28
However, the force of this objection should not be overestimated. First, on
the plausible view that traditional forms of GS reliably lead to women having
reduced autonomy capacity, forms of GS that may hold dangers for the devel-
opment of self-worth but that also have resources for addressing the autonomy
gender inequality problem might still be overall preferable, as containing
better chances for girls to develop autonomy. Second, the influence of parental
26
J. Block and R. W. Robins, A Longitudinal Study of Consistency and Change in Self-Esteem
from Early Adolescence to Early Adulthood, Child Development 64 (1993): 909923; S. K. Egan
and D. G. Perry, Gender Identity:AMultidimensional Analysis with Implications for Psychosocial
Adjustment, Developmental Psychology 37 (2001): 451463; R. A. Josephs, H. R. Markus, and R. W.
Tafarodi, Gender and Self-Esteem, Journal of Personality and Social Psychology 63 (1992):391402; A.
Thorne and Q. Michaelieu, Situating Adolescent Gender and Self-Esteem with Personal Memories,
Child Development 67 (1996):13741390; Meenakshi Menon, Does Felt Gender Compatibility Mediate
Influences of Self-Perceived Gender Nonconformity on Early Adolescents Psychosocial Adjustment?
Child Development 82:4 (2011):11521162.
27
Hunter College Womens Studies Collective, Womens Realities, Womens Choices
(NewYork:Oxford University Press, 1983),153.
28
See, e.g. Benson, Free Agency and Self-Worth; Catriona Mackenzie, Imagining Oneself
Otherwise, in Mackenzie and Stoljar, Relational Autonomy, 124150.
instruction should not be underestimated. Girls raised according to NGS by

committed parents can be trained to be prepared for precisely such assaults on
their sense of self-esteem. In full awareness of the likelihood of pushback from
society, parents committed to NGS would be highly motivated to take special
care to instill in their girls the inner fortitude required to maintain their sense
of entitlement to cast off the traditional bonds of gender association. Indeed,
there is empirical evidence to support the claim that strongly supportive pa-
rental involvement with children has a direct and positive bearing on chil-
drens self-esteem.29
The primary worry concerning the autonomy potential of NGS, however,
is likely to come from those who hold that biology is destiny. Even if we as-
sume, on the strength of research that supports social constructivism about
gender, that this claim is false, a challenge might be raised by those who hold
that biology has some influence on the shape of a womans life. Empirical evi-
dence for this thesis can be found. The sociologist Laura Lindsey, for example,
has argued that hormones predispose the sexes to different behavior while
still claiming that societal factors will ultimately activate this behavior.30
Sociologist Richard Udry has supported this claim by arguing that empirical
research suggests that humans form their social structures around gender
because males and females have different and biologically influenced behav-
ioral predispositions.31 This and other research in the social and hard sciences
does not debunk the social constructivist view of gender; indeed, both Lindsey
and Udry call for an integrated theory of gender that combines biological de-
terminism with social construction. The present concern is whether biology
plays any role at all in gender development. If it does, the worry arises that at
least some girls raised in accordance with NGS may feel alienated from them-
selves to some extent. Given their commitment to the idea that traditional
gender associations are nothing more than arbitrary social constructs, some
girls raised in accordance with NGS may come to feel conflicted and uncer-
tain about womanhood or even about who they really are. And this could lead
to a weakened ability to satisfy the authenticity conditions of autonomy, ac-
cording to which autonomy requires that one have a firm, coherent, and com-
mitted sense of ones authentic self. The presence of empirical support for the
influence of biology should be sufficient to undermine the worry that this is
nothing more than a slippery slope fallacy masquerading as a real problem.
If biology has any influence on gender dispositions, at least in some persons,
29
M. H. Richards, I. B. Gitelson, A. C. Petersen, and A. L. Hurtig, Adolescent Personality in Girls
and Boys:The Role of Mothers and Fathers, Psychology of Women Quarterly 15 (1991):6581.
30
Linda L. Lindsey, Gender Roles:ASociological Perspective (Upper Saddle River, NJ:Prentice-Hall,
1997),27.
31
Richard Udry, Biological Limits of Gender Construction, American Sociological Review 65:3
(2000):454.
there is a real concern that raising children in a robust NGS mannersuch as

that practiced at Egalia preschool in Stockholmmay have the effect of sun-
dering them from core, biologically grounded aspects of their identities, with
concomitantly adverse effects on their potential for identifying with at least
some aspects of their authentic selves.
Let us suppose that this challenge needs to be met. More specifically, let us
assume, for the moment, that biology, while not absolute in its influence, does
exert a pull in women toward certain traditional gender associations, espe-
cially those surrounding reproduction, mothering, and caregiving. Girls who
are taught by NGS that gender associations are ultimately arbitrary may de-
velop a damaging split in their self-conceptions that would undermine their
ability to form a consistent self-identity, and this could lead to a weakening
of their ability to be self-governing. If one isnt sure who one is, one cannot
govern oneself effectively. This worry would not preclude the development of
a consistent authentic self-conception in all women, of course, but it would
likely undermine the formation of such conceptions in many women. And if
biology in fact is destiny, then this concern is conclusive; in this event, NGS
would fail the test of autonomy support, by severing women from a core com-
ponent of their authentic selves. At this point, judging the strength of this chal-
lenge becomes an empirical matter on which I will not pass judgment. The
empirical evidence seems to be conflicting, and I do not have the expertise
to adjudicate the matter. If it is true that biology exerts even a reliableif not
absoluteinfluence on women, then many women will find it difficult to form
a firm and consistent self-conception, thereby threatening their possibility of
autonomy. If biology exerts no such influence, however, then this challenge
disappears.
We come, then, to the question of the feminist credentials of NGS. It seems
clear that at least one core aspect of NGS is favorable in this respect. NGS
undermines the traditional system of gender associations that have been such
a reliable vehicle of womens oppression. The traditional gender system views
women as being of less worth than men in many important respects. NGS, by
repudiating the gender system entirely, and by espousing the formation of a
social order based on the equal worth of all persons regardless of sex, well sat-
isfies the feminist project of overcoming womens subordination and creating a
just and equal society. Iwould suggest that these considerations constitute the
strongest specifically feminist argument that can be made in favor of NGS. It is
precisely this virtue, it seems, that has led many feminists to support the erad-
ication of the gender systemand, by implication, to support NGS for girls.
There is, however, at least one objection that deserves mention.
It could be argued that there is a kind of practical value in maintaining the
idea of a real difference between the genders. The practical value that Ihave
in mind relates to the work on gender done by feminists like Sally Haslanger.
According to Haslanger, one way to approach the normative problem of what
ought to be done in relation to the gender system is pragmatic in character.

The basic idea is that conceiving gender in particular ways might best support
the feminist project. Haslanger holds that gender ought to be conceived pri-
marily in terms of subordination. On this viewroughlywhat it means to be
a woman is to be in a subordinate social position in relation to men. Haslanger
does not claim that this subordination constitutes a real definition of gender,
however. She holds that conceiving gender in this way holds the best chance
of furthering the practical aims of feminism itself, by motivating social action
on behalf of justice for women.32 This practical or ameliorative approach to
conceiving gender can be taken in other directions, however, and that is what
Ihave in mind here. The basic idea that Iwould like to tender is that the goals
of feminism might best be supported by continuing to conceive the gender cat-
egory woman as denoting a real difference from the gender category man. The
justification for this claim is twofold. First, maintaining the category woman
would have the beneficial effect of not alienating the many women who con-
tinue to accept the idea that this category is real and distinct. The majority of
the worlds women would likely consider the eradication of the gender system
to be radically alien to their way of conceiving the world. If feminism begins
with this very eradication, however, it risks the danger of alienating the great
majority of the very constituency whose improvement constitutes the basic
goal of feminism itself. In consequence, scads of possible allies to the feminist
cause would likely find themselves at variance with the feminist movement
itself. Since it is entirely possible to maintain the category woman without suf-
fering this effect, it seems reasonable to do so and hence to avoid pinning ones
hopes on such an alienating policy.
Second, maintaining the category woman could have the greatly beneficial
effect of establishing a rallying point for the feminist cause. By maintaining
the claim that the category woman is real and different from the category man,
while continuing the focus attention on the problem of womens inequality
and, of course, while suitably modifying the conception of women in such
a way as to fight the traditional (if often implicit) notion that women are in
some normative sense inferior to menthe feminist movement could give
women a home worth fighting for. Rather than being asked to fight to erad-
icate the very category of which most women consider themselves a member,
women could be invited to fight for the rights of that very category. The emo-
tional and social connotations of femininity are already exceedingly strong;
with the proper modifications, the category woman could serve as an effec-
tive and motivating rallying call for women:fight for your rights as a woman!
In accordance with the aforementioned considerations, Iwould submit that a
32
Haslanger, Gender and Race, 33,36.
policy of this kind would have the best practical chance of securing wide con-
sent and motivating womens action on a widefront.
It may perhaps be objected that enacting this suggestion might hurt the
cause more than it helps. After all, in the face of widespread social injustice,
sometimes it is best to buck even the deepest points of convention. Wouldnt
the kind of reasoning tendered previously, applied, say, to slavery, suggest that
we ought to maintain the category slave for similar reasons? Yet while Iagree
that a radical repudiation of conventional wisdom might be necessary in some
circumstances, Iam not convinced that this is appropriate here. For one thing,
the analogy with slavery is strained at best. The category slave is inherently
subordinate, while the category woman is not. Hence, maintaining the cate-
gory woman would not entail the continuation of an inherently unjust social
category. More broadly, decisions regarding the solution of social ills ought to
be determined by the particulars of the case. In the case of social inequality
toward women, Iwould argue, in accordance with the considerations given
already, that the goals of the feminist movement are best satisfied by main-
taining the category woman as real and distinct. Doing so would by no means
require maintaining the traditional role associations, psychological profiles, or
differential claims regarding status that have undermined the feminist cause,
but it would increase the likelihood that fewer women would feel alienated
from the movement devoted to their very liberation and that women could
productively rally around their very gender identity in furtherance of their
just aims. For these reasons, Iconclude that NGS, while theoretically poised to
support the feminist cause, holds considerable practical dangers for feminism
and therefore fares poorly in relation to practical support for the feminist ide-
als that it was designed to facilitate.
Although these considerations are institutional in scoperelating as they
do to the establishment of a policy for broad and effective feminist social ac-
tivismit should not be thought for that reason that they are unsuited to the
far more local decision faced by parents regarding how to raise their daugh-
ters. If the practical arguments given already have bite, there is no reason to
think that they would not inform parents judgments about the best form of
GS to apply when raising their daughters or the particular manner in which
daughters are raised. Parents committed to the goals of autonomy and fem-
inism wish for their daughters to flourish in the world as it exists, and this
requires providing sound practical advice for understanding and interacting
with the broad currents of social activism in which their daughters will live. It
makes no small difference to the nature of a daughters upbringing to be told
that gender categories have no significance whatsoeveror, for that matter,
that there is real practical value in maintaining that they do. Such instruction
will certainly play a powerful role in how a girl understands and interacts with
her peers and with gendered social conventions and norms and will influence
a girls judgments about what a commitment to feminist ideals demands of her
and other women. Parents who wish for their daughters potentially to become
strong advocates for feminism cannot afford to overlook broader consider-
ations such as those given before when they will have a direct bearing, not
only on how they raise their daughters but more importantly on what their
daughters cometobe.
In sum, then, Ihold that NGS faces strong challenges as a feminist theory
of GS, both because of concerns about its ability to be supportive of womens
autonomy and because of concerns relating to its practical effectiveness in
underwriting the desirable feminist goal of securing widespread and effective
practical action in support of womens equality.
Traditional Male Gender Socialization

Perhaps the most prominent advocate of TMGS as a feminist theory of GS is
Simone de Beauvoir. Given the advantages associated with male socialization,
de Beauvoir concludes that girls ought to be raised with the same demands
and rewards, the same severity and the same freedom, as her brothers, taking
part in the same studies, the same games, [and] promised the same future.33
De Beauvoir does not, it is true, think that this alone will solve the autonomy
gender inequality problem. She is well aware that institutional changes must
also be made in relation to education, employment, and financial indepen-
dence for women. And she also implies that boys should be raised to see girls
as equals:To gain the supreme victory, it is necessary...that by and through
their natural differentiation men and women unequivocally affirm their
brotherhood.34 But it is also clear that de Beauvoir viewed raising girls in a
like manner as boys to be an essential instrument in the achievement of the
goals of feminism. Diana Meyers, in support of this contention, argues that
de Beauvoir held that universal masculine socialization coupled with social
and economic reform holds out the only hope of freeing women from imma-
nence.35 Following de Beauvoir, other feminists have supported the same con-
clusion. And one can well understand why this policy was formed. If womens
subordination is best explained by the fact that men have been raised to be
superior and women have been raised to be inferior, it would seem clear that
the solution is to raise both sexes to be superiorthat is, to raise girls in the
same manner asboys.
How does TMGS fare in relation to the development of autonomy? As a
start, it looks quite promising. Girls raised in accordance with TMGS will
enjoy all of its benefits:they will be taught to be more assertive and proactive;
they will be taught that they are fully entitled to direct their own life paths,
33
Beauvoir, Second Sex, 681682.
34
Beauvoir, Second Sex,687.
35
Meyers, Self, Society, and Personal Choice,190.
giving them a greater range of meaningful alternatives from which to choose;

they will be raised to develop a stronger sense of entitlement to direct their
lives, which should accordingly lead to a greater sense of self-worth; they will
be taught to be more discriminating in their judgments, which may lead to
an increase in logical aptitude; they will see themselves more as the rightful
arbiters of their futures, which may lead to an increase in moderate skepti-
cism in relation to others plans for them; and they will be less concerned with
the importance of their outward appearance, placing importance instead on
the trajectory of their life plans, as determined by their own judgments. All
of these features of TMGS will likely lead to an increased possibility for the
development of autonomy.
Yet the autonomy-supporting benefits of TMGS ought not to be exagger-
ated. For one thing, TMGS places a premium on self-assertiveness and com-
petitive success. While in moderation such a focus does not undermine future
autonomy, excessive self-assertiveness can lead persons raised in accord-
ance with TMGS to become too concerned with pushing through their own
agendasdespite the presence of external obstacles. As Diana Meyers argues,
The emotional isolation central to masculine socialization obliges boys to be-
come independent, but it also heightens their aggressive tendencies because
it makes them feel insecure.36 From this point of view, TMGS could easily
translate into closed-mindedness and obstinacy. Being autonomous requires a
sensitivity to ones authentic self, yet this sensitivity could be lost if one trades
careful reflection upon possible ways of living for a stubborn adherence to
whatever one happens to will. Excessive self-assertiveness can close the doors
to insightful introspection.
By contrast, traditionally feminine traits such as sensitivity and empathy
can lead persons to be more introspective, more responsive to their feelings,
and more adept at understanding themselves.37 Such traits are also important
for the ability to foster and flourish in interpersonal relationships, which, as
autonomy theorists increasingly agree, is crucial for autonomy development.38
According to Iris Marion Young, the trajectory of feminism since de Beauvoir
has been excessively dismissive of the notion that traditional femininity con-
tains anything of worth, andwhat basically comes to the same thinghas
been excessively supportive of the idea that valuable ways of living are to be
found exclusively in the traditional male ideal.39 Young and other gynocentric
36
Meyers, Self, Society, and Personal Choice, 191192.
37
Beauvoir, Second Sex,338.
38
Taylor, Philosophy and the Human Sciences, 187210; Evelyn Fox Keller, Reflections on Gender
and Science (New Haven, CT:Yale University Press, 1985); Patricia Huntington, Toward a Dialectical
Concept of Autonomy, Philosophy and Social Criticism 21 (1995):3755.
39
Iris Marion Young, Humanism, Gynocentrism, and Feminist Politics, in Theorizing
Feminisms:AReader, edited by Elizabeth Hackett and Sally Haslanger (NewYork:Oxford University
Press, 2005), 174187.
feminists argue that traditionally female forms of lifecentered on relation-

ships of loving care, mothering, nurturing, and the likeought in fact to be
celebrated and should form one of the key bases for the feminist movement.
Traditional female virtues form a highly desirable counterpoint to the overly
aggressive, overly competitive, and often callous ways of interacting with oth-
ers (and the world) associated with traditional masculinity. Instead of dispar-
aging what is traditionally female, feminists ought to hold feminist virtues
up as ideals of social interaction and emotional development. From this per-
spective, TMGS, by lionizing traditional masculine forms of life, both unjustly
slights what is valuable in traditional femininity and does a disservice to femi-
nism by undercutting the legitimate values that inhere withinit.
Moreover, TMGS can reduce ones perceived life options as well. Raised to
view oneself as a significant mover in social matters, one may come to dis-
parage ways of life that are more focused on service or caregiving. Such ways of
living may be lost on those raised in TMGS, even if those forms of life are ac-
tually more in accord with their authentic selves. Many males are raised to be-
lieve that it is fitting for them to be fighters, for instance, and this has doubtless
led many to engage in acts of violence that are at odds with their real natures.
And there is a final potential problem, one that we have seen before. If
biology exerts an influenceeven if not absoluteon womens lives, then
women raised in accordance with TMGS may find themselves alienated from
aspects of womanhood, with the result that their ability to form clear and uni-
fied authentic self-conceptions may be undermined. As mentioned already, if
it can be shown that biology has no such pull, then these issues dissolve, but
if not they are considerable. In sum, then, TMGS presents mixed results in
relation to autonomy support. It fares best in relation to the inculcation of the
competency conditions of autonomy but presents problems in relation to the
development of authenticity conditions, through its valorization of masculine
traits that may actually undermine authenticity competency and its dismissal
of valuable traditionally feminine traits that support those competencies. It
also presents difficulties in relation to the presence of external alternatives
considered as live options.
How well does TMGS stand in relation to support for feminism? In at least
one key respect, it seems that it fares quite well. Girls raised in accordance with
TMGS stand a greater chance of becoming women who are independent and
powerful, and this is a notable attraction for parents raising girls in a sexist soci-
ety. Secure in their conviction that women are equal to men in all key norma-
tive respects and raised to be assertive, forthright, aggressive in the prosecution
of their life plans, and with a firm sense of self-worth, women who have been
raised in accordance with TMGS will stand as strong advocates for the rights of
women. They will take positions of social leadership and oppose social injus-
tices on a broad front; they will not be cowed by spurious notions of rightful
male authority or the overwhelming history of womens subordination. In
short, TMGS will likely result in the development of strong leaders of social,
political, and economic institutions and of feminism generally.
Yet there are some respects in which one might be reasonably critical of
the extent to which TMGS supports feminism. Some of these issues have been
explored already in relation to NGS. First, TMGS is premised on the idea that
traditional notions of femininity are to be avoided. Lives devoted exclusively to
mothering, or more generally to the loving care of others, or to homemaking,
or indeed to any kinds of living that are characterized by what might be called
quiet compassion are judged by the background normative commitments of
TMGS as worthy of disparagement and even hostility, as precisely the ways
of living that have historically worked to facilitate womens subordination to
men. TMGS repudiates these ways of living as forms of collaboration in wom-
ens oppression. Yet while it cannot be doubted that these forms of life have
contributed to womens oppression, values and practices based on the repudi-
ation of traditional forms of femininity might work to undermine feminism by
alienating broad swaths of women who associate with these ways of living. It is
important to avoid understating the potential effects of propagating the view
that the feminist movement is hostile to traditional forms of femininity. Told
that their commitment to these forms of life constitutes an insidious form of
unwitting collaboration with the forces of social injustice, many women may
elect to keep their distance fromor even fightthe feminist movement out
of a desire for ideological self-preservation.
Moreover, TMGS implies that women ought to assimilate their value sys-
tems, social expectations, and behaviors to traditional forms of masculinity.
The implicit upshot of this recommendation is the idea that the category
woman, while not empty of distinctive content, is empty of any desirable or
commendable content. Woman becomes a symbol for an unworthy form of
weakness that is to be avoided. By contrast, man becomes a symbol for forms
of life and valuation that are truly valuable. Such a consequence contains the
practical problem that women are deprived of basing possible activism on be-
half of women in anything that is valuable about being a woman. One may still
fight on behalf of the abstract claims of justice or perhaps on behalf of women
qua individuals or persons; but the very category woman whose weal is sought
through social action threatens to become little more than a symbol of an ob-
stacle that needs to be overcome. In this way, a powerful potential resource for
securing communal accession and social action is lost:the category woman is
not a rallying point for women; rather it becomes a threat.
The upshot of these concerns is practical in character. A more effective con-
ceptual practice in garnering wide and effective support for the feminist cause
might be instead to maintain the notion that the category woman is distinctive
and valuable in its own right. One is far more likely to fight effectively when
one is fighting for ones kin and community, conceived as worthy of fighting
for on its own merits.
In addition to practical worries, there is a second and related concern that

TMGS would undermine feminist ideals by propagating a form of normative
exclusion. Women who endorse traditional forms of femininity are judged by
the value system within TMGS as not living lives that are valuable for women
(properly conceived). Rather, their life choices and life experiences are
viewed as being unworthy. This issue is a variant of the normativity argument
made by Judith Butler.40 The difference is that whereas Butlers argument is
directed against feminist attempts to defend some form of gender realism,
the current argument is directed against attempts to defend what might be
called gender idealism:a view of what all women ought to be. The normative
problem in the present case is that such a theory would delegitimize ways of
being a woman that are not necessarily worthy of such condemnation. It is
at least arguable that it is possible for a woman to be a committed feminist
with all the attendant normative commitments to equal status as men and a
repudiation of gender injusticeyet autonomously and knowingly to choose
a life that, in its outward character at least, adheres to more traditional femi-
nine social roles. From this point of view, TMGS seems to propagate its own
form of exclusion, and thereby to commit an injustice against some women.
In this respect as well, then, TMGS seems to be potentially at odds with fem-
inist ideals.
In sum, TMGS faces many strong challenges as a contender for a feminist
theory of GS. While it doubtless satisfies the demands of autonomy support to
a greater degree than traditional forms of female GS, aspects of TMGS under-
mine certain enabling conditions for robust autonomy, especially in relation
to authenticity competency. Moreover, TMGS, like NGS, seems to hold prac-
tical problems for the possibility of establishing feminism as an effective and
normatively inclusive social movement, and its disparagement of traditional
femininity can be criticized as failing to do justice to the virtues of feminine
ways ofbeing.
Revisionary Feminine Gender Socialization

RFGS is similar to NGS and TMGS in raising girls to be strong advocates for
equality but differs from both in maintaining a clear commitment to the im-
portance and positive value of the category woman. Living in accordance with
a conception of womanhood can have great prudential value for women. Being
a mother, for example, can be viewed as a significant source of felt satisfaction
in a womans life. In giving birth, or in breastfeeding a newborn baby, a woman
takes part in an activity that binds women of all cultures and ages together; as
such, it is an activity that enlarges and enriches a womans experience of part
40
Butler, Gender Trouble.
of what it can mean to be a woman. Moreover, as we have seen, RFGS holds

that what it means to be a woman can differ depending on the social, familial,
psychological, and political circumstances in which women find themselves
worldwide. Although these different forms of womanhood will often share
common featuressuch as widely shared histories, biological features, and
challengesthe variability of womens experience and circumstances speaks
against defending privileged conceptions of womanhood. Girls raised ac-
cording to the tenets of RFGS will be taught that they are entitled to choose
almost any life path they wish, as long as it does not essentially involve accept-
ance of a sexist axiology. And in accordance with this, they will be raised to
be well equipped with skills and resources that will be of use to them in many
different forms of life. In this way, RFGS is exceedingly flexible, especially in
contrast toTMGS.
How does RFGS fare in relation to support for autonomy? Iwould argue
that general indications are very positive. As we have seen, RFGS is premised
on the idea that girls ought to be raised to be effective advocates for wom-
ens rights. A key part of this education is to raise girls to possess forms of
self-control, logical aptitude, and reasonable skepticism toward attemptses-
pecially by mento impose their conceptions of what is proper. This suggests
that girls raised in accordance with RFGS, like girls raised in accordance with
NGS, will tend to receive an upbringing that well satisfies autonomy compe-
tency conditions. Moreover, girls raised in RFGS will be educated to take great
pride in their status as women, which suggests that the condition of self-worth
will be quite well satisfied.
Furthermore, RFGS has the virtue of not predetermining acceptable forms
of life for women. The result of this inclusiveness is that girls raised in RFGS
will experience a much wider array of live options in life from which to choose
(assuming the proper supporting institutional conditions are in place, of
course). As we have already seen, this is an important external enabling con-
dition of autonomy. In accordance with this freedom, girls raised in RFGS will
be open to explore different forms of authentic self-expression. In this way, a
wider allowance for external options is matched by a wider allowance for in-
ternal expression of the kind that is important for authentic flourishing. This is
a considerable advantage for RFGS as against TMGS especially. This flexibility
also insulates RFGS from concerns stemming from the claim that biology has
some influence on ways of being a woman. By remaining open to almost all
possibilities, RFGS makes theoretical room for the possibility of choosing
and celebratingways of living that are in accordance with the influence of
biology, if such influences are inplay.
In general, then, RFGS seems to fare extremely well in relation to autonomy
support. Girls raised in RFGS will tend to receive strong training in autonomy
competency conditions, will tend to find parental support for largely unre-
stricted authentic introspection and the possibility of choosing from a wide
variety of available life options, and will tend to develop a sense of self-worth
grounded in an appreciation both of the value of their gender and their basic
moral standing as persons.
RFGS has many advantages in relation to support for feminism as well. First,
as mentioned already, one of the goals of RFGS is to produce committed, com-
petent advocates for feminism (with the understanding, noted already, that
this advocacy can take different forms, depending on a womans nature and
life choices). Second, RFGS maintains and celebrates the category woman
considerably broadened in its sense, of coursewhich serves the practical
purpose of rallying support for feminism. RFGS favors as inclusive a home
for women as possible, provided it is grounded in gender equality. As argued
earlier, allowing for this has the considerable practical virtue of giving women
a sense of kin and community that is seen as positive and worth fightingfor.
However, unlike TMGS, RFGS will not introduce practical or normative
worries associated with exclusion of conceptions of womanhood (beyond the
exclusion of normative conceptions of womanhood that do not underwrite the
basic commitment to womens equal worth and entitlements, of course). In its
broad inclusive purview, women of many kinds and natures will be welcome.
Of all of the forms of GS thus far considered, in fact, RFGS will have the broad-
est appeal and hence will have the potential for organizing and maintaining
the widest base of support.
All of these advantages speak in favor of RFGS as a promising strategy in
relation to engendering positive, broad, and sustained support for feminism.
However, the primary objection to RFGS relates precisely to its inclusive-
ness. It might be argued that this form of GS, in its broad acceptance of a wide
variety of different valid conceptions of what it means to be a woman, lacks a
unified conception of womanhood that delimits the parameters of woman-
hood itself. Feminism, it could be argued, has practical and theoretical need
of a conception of womanhood that has a relatively clear and definable exten-
sion. Without this, the feminist cause loses its distinctiveness as a unified cause
on behalf of women. From this point of view, RFGS engenders a degenerate
form of feminism, one that collapses into a mealy, confused, and potentially
contradictory congeries of all of the above that precludes coherent female
self-understanding and diminishes the possibility of unified action on behalf
ofwomen.
This objection would be problematic, Ibelieve, if RFGS had this effect. But
it can be plausibly argued that it does not. RFGS has resources to provide the
kind of unity to the category of womanhood that is necessary to ground a
broadly encompassing understanding of what it means to be a woman and to
motivate unified social action. The key here is to remember that RFGS contains
and propagates the idea that women are to be understood as grouped together
by general shared features, including biological similarities, a shared history of
similar ways of living, and a shared experience of social oppression, largely at
the hands of men. These identity conditions are specific enough to delimit the
vast majorityif not allof those who consider themselves women but, im-
portantly, are general enough to allow for a wide variety of more specific ways
one can conceive what women are or ought to be. In addition, the normative
basis of RFGS serves as a unifying ground for social action:however various
are the ways women live and experience the world, all women are clearly called
to unite on behalf of resisting, in various ways, the social institutions that con-
tribute to womens oppression. On the basis of these features of RFGS, then,
Iwould suggest that the worry that RFGS is too inclusive can bemet.
5. The Final Verdict
On the basis of these considerations, Iconclude that RFGS stands as the most
attractive option for a feminist theory of GS. RFGS fares exceedingly well both
in terms of its potential for raising girls with the best chance for autonomy
development and in terms of its worth as underwriting a broadly inclusive,
practically effective, widely unified feminist movement. It supports a concep-
tion of what it means to be a woman that engenders solidarity while recogniz-
ing and celebrating difference and remains strong and unapologetic in its call
for reform of the social institutions informed by gender conceptions.
6. Concluding Remarks
One of the most important tasks for any parent who espouses feminist ideals
is the task of raising daughters. Although a proper upbringing is not perfect
insurance against problems stemming from institutionalized gender inequali-
ties, parental instruction is the beginning of a girls experience of the world
and has pride of place as the oldest root of a girls theoretical and normative
commitments. One of the most important parts of this upbringing is gender
socialization itself. Improper forms of GS hold the risk, not only of damaging
a girls prospects for future autonomy but also of undermining, through the
personal and social implications of its teachings about what it means to be a
woman, feminist causes themselves. If we are to discharge our commitments
to raising our daughters in a manner supportive both of their future autonomy
and of feminism, we need a feminist theory of GS that pays due heed to both
of these commitments and justifies its prescriptions in a manner that is both
rationally compelling and realistic, given the state of the society as we findit.
In this chapter Ihave argued that both the goal of womens autonomy en-
hancement and the normative commitments and social goals of egalitarian
liberal feminism are best satisfied by a form of GS that involves raising girls
firmly in the commitment that women deserve equal treatment as men, that
contains the notion that the category woman is to be maintained in its own
right as identifying and celebrating a distinctive form of life and yet remains
widely inclusive in terms of what it means to be a woman, and that locates the
fundamental ground for gender equality in womens membership in a nor-
matively significant category that is shared with men:the category of rational
agents, or autonomous individuals, or simply persons. If this conclusion has
merit, the remaining task assumes all the more importance:to apply theory to
life, and raise our daughters conscientiously according to the theorys prescrip-
tions, secure in the conviction that they deserve our best efforts to prepare
them to advocate for a cause whose nobility and urgency cannot be reasonably
doubted.
Acknowledgments
I would like to thank Andrea Veltman and all of the participants in the McGill
University workshop Relational Autonomy:Ten Years On for their extremely
helpful comments and feedback on this chapter. Iwould also like to thank my
wife, Pia Antolic-Piper, for discussing these ideas with me, for giving birth to
our daughter, and for being an amazingly devoted and inspiring mother.
{13}
Autonomy and Oppression atWork

Andrea Veltman
A central aspect of autonomy, highlighted not only by feminist accounts of

relational autonomy but also by other accounts of personal autonomy, is the
exercise of reflective and agential capacities that develop in social and inter-
personal contexts. Whereas theorists of relational autonomy often emphasize
that capacities necessary for autonomy develop on account of our relation-
ality, including our initial dependency on parents or other caregivers, theo-
rists of autonomy rarely explore work as a social context that can support or
stifle the development and exercise of autonomous capacities. In this paper
Iturn attention to the impact of work on autonomy and on related goods of
self-respect and self-worth, which are required for full personal autonomy on
some accounts. Drawing on empirical and philosophical literatures on work
and well-being, Ifirst draw together a case that eudemonistically meaningless
work undermines autonomy and self-respect. By eudemonistically meaning-
less work, Imean work that does not develop or exercise human capabilities,
permit independent judgment, integrate conception and execution, or oth-
erwise facilitate expressions of agency.1 These forms of work are not neces-
sarily meaningless altogether, but working extensively at eudemonistically
1
Some of the literature on meaningful work focuses on eudemonistic dimensions of meaningful
work, that is, on the potential of work to contribute to human flourishing by developing or exercising
capabilities or skills, by fostering independent judgment in performance of tasks, or by integrating con-
ception and execution for a feeling of personal satisfaction at work. See, e.g., James Bernard Murphy,
The Moral Economy of Labor:Aristotelian Themes in Economic Theory (New Haven, CT:Yale University
Press, 1993); Adrian Walsh, Meaningful Work as a Distributive Good, Southern Journal of Philosophy
32 (1994): 233250. In a monograph I am now working on, provisionally titled Meaningful Work,
Iargue that meaningful work has several dimensions, and eudemonistic dimensions of meaningful
work are integral but not exhaustive in an account of what makes work meaningful. In addition to
being eudemonistically meaningful, work can be meaningful in serving a purpose, creating something
of enduring value, reflecting personal life goals or values, or integrating otherwise disparate elements
of a workers life. Developing or exercising human capabilities in eudemonistically meaningful work
exhibits agency, but as Iunderstand it agency extends beyond developing or exercising human capabili-
ties to encompass, for instance, expressions of values, principled commitments, character, personality,
creativity, or individuality; agency at work stands in contrast with observation, passivity, merely fol-
lowing orders, or feeling likeacog.
Autonomy and Oppression atWork 281
meaningless work stifles the flourishing of a worker and, in particular, dimin-

ishes her drive toward self-determination and her sense of self-worth.2
I submit that eudemonistically meaningless work should be counted among
other autonomy subverting social influences, which on feminist accounts
range broadly from internalized oppressive norms to entanglements in abusive
personal relationships to practices of childrearing that thwart development of
autonomous skills. Although eudemonistically meaningless work is not alone
in subverting personal autonomy, it merits particular attention in light of the
impact that work has on a person, even outside the workplace. Work molds a
person and stands to impart a cluster of moral and personal goods and virtues
that are integral in a good life, including self-respect, honor, dignity, pride and
intellectual development; work is also a primary avenue through which people
achieve recognition for the exercise of skills, make contributions in communi-
ties, and achieve a sense of purpose and personal identity, among other goods.3
With respect to autonomy and freedom, not only does an erosion of autono-
mous agency in work stand to damage skills and proclivities needed for full
personal autonomy, as I focus on here, but escaping meaningless work also
bears an intuitive association with achieving freedom. As E. F. Schumacher
observes in Good Work of wanting to avoid the rat race, to not be enslaved by
machines and bureaucracies, to avoid becoming a moron, a robot, a commuter
or a fragment of a person, to do ones own thing, to deal with people, to re-
spect that people, nature and beauty and wholeness matter and to care about
what mattersthese strivings are simultaneously a longing for freedom and a
longing for an escape from work without meaning.4
Work also merits attention from feminist philosophers for whom oppres-
sion is a quintessential topic of interest and for whom an ideal of autonomy can
serve as a normative standpoint for critically assessing oppressive social con-
ditions that suppress or prevent the emergence of autonomy.5 Appreciating the
ways women often lack autonomy in relation to work illuminates key elements
2
In Meaningful Work, Iam broadly interested in work in its relation to human flourishing, which
requires the realization of human capabilities and the possession of a plurality of goods. My focus
here is work in relation to autonomous agency, which I understand to be a component of human
flourishing. For a good discussion of the components of human flourishing, see Douglas Rasmussen,
Human Flourishing and Human Nature, in Human Flourishing, edited by Ellen Frankel Paul, Fred D.
Miller, Jr., and Jeffrey Paul (Cambridge:Cambridge University Press, 1999), especially121.
3
For a fuller list of goods that attach to work, see Andrea Veltman, Is Meaningful Work Available
to All People? in Philosophy and Social Criticism, forthcoming. Consider also what one occupational
psychiatrist writes:It is possible that no single activity defines adulthood more specifically than work.
To a large extent work influences how and where an individual lives, it affects social contacts and
family activities, and it provides a title, role, and environment that shape and reinforce an individuals
identity. Nick Kates, Barrie Greiff, M.D., and Duane Hagen, M.D., The Psychosocial Impact of Job Loss
(Washington, DC:American Psychiatric Press, 1990), 185. See also the work of Al Gini, such as A. Gini
and T. Sullivan, Work:The Process and the Person, Journal of Business Ethics 6 (1987):649655.
4
E. F. Schumacher, Good Work (NewYork:Harper and Row, 1979),50.
5
of the oppression of women, particularly given the preponderance of time that

oppressed women spend at work. Awoman whose days (and nightsoften
enough) are spent slaving in a factory assembly line, a sweatshop, or a Foxconn
plant may suffer several faces of oppression, but her oppression as a worker
is paramount in her life and transcends mere exploitation as an underpaid
employee. Her work may be oppressive first in respects of being heterono-
mous:she may enter work under conditions of constraint; her work may bear
no part of reflectively held life goals; and she may not even have the freedom
of bodily movement at work.6 Her work may also fail to permit a meaningful
measure of economic independence or to help her support herself or her fam-
ily, which she identifies as the very purpose of her working. And her work
may undermine her autonomy furthermore in the respect that her employer
requires that she only mechanically follow goals set by others, in the precise
manner in which others determinea lack of autonomy in work that yet fur-
ther erodes her well-being.
A central issue for those who value autonomy is that aspects of autono-
mous agency remain a privilege for those fortunate to work in enlightened
workplaces that value employee decision making and promote the develop-
ment of employee skills. Many current workplace management structures
treat adult workers as though they lack competency to exercise intelligence,
skill, and autonomous capacities, thus relieving employees of the need to make
decisions, design goals, or determine methods by which to accomplish goals
at work. Such ugly skepticism about the abilities of working people is a hall-
mark of the fairly influential principles of scientific management forwarded
by Fredrick Taylor,7 which stand in a basic tension with modes of working life
that support autonomous development and self-realization. The Tayloristic as-
sumption that employees enter workplaces with settled levels of intelligence
and ability also merits rejection in light of more recent empirical literatures on
work and well-being, which Idiscuss further herein and which indicate that
work itself affects the development of a range of capabilities, including capabil-
ities for self-direction, which can be nurtured or stifled by working conditions.
6
Consider, e.g., what Ruth Cavendish, Women on the Line (London:Routledge & Keegan Paul,
1982), writes about her experience working in a car parts factory in England: on the assembly line
we couldnt do the things you would normally not think twice about, like blowing your nose or flick-
ing hair out of your eyes; that cost valuable secondsit wasnt included in the layout so no time was
allowed for it. In any case, your hands were usually full (41). The women ran the line, but we were
also just appendages to it. The discipline was imposed automatically....We just slotted in, like cogs in
a wheel. Every movement we made and every second of our time was controlled by the line...(107).
7
Taylorism is the idea that workplace managers increase efficiency, productivity, predictability,
worker accountability, and control over working processes by extracting knowledge and skills from
workers, subsequently reducing worker skill and knowledge to simple and discrete formulate so that
production can be performed by men who are of smaller caliber and attainments and who are therefore
cheaper than those required under the old system. F. W. Taylor, Shop Management (NewYork:Harper
& Brothers Publishers, 1912),105.
In essence, a commitment on the part of a community to promoting human

flourishing and autonomous agency entails a commitment to respecting the
agency and dignity of people at work and a commitment to organizing work
so that people can exercise agency and skills in occupational life. This goal
does not necessarily entail Marxism, but it does entail opposing Taylorism
and oppressive forms of work in which employers take dim views of the
decision-making abilities of workers, in which workers consequently have
scant opportunities to exercise thought, skill or judgment on the job, in which
workers become effectively reduced from human agents into tools or append-
ages of machines, or in which workers are treated as cheap, expendable, inter-
changeable resources. Structuring working life around a goal of developing
and exercising human capabilities harmonizes with a number of basic ethical
principles and traditions, including eudemonistic ethical traditions originat-
ing with Aristotle and continuing with John Stuart Mill, which emphasize
developing human capabilities as part of achieving happiness or flourishing.
Outside of eudemonistic traditions, the Dalai Lama, for instance, expresses a
principle of prioritizing people over profit with his typical simplistic elegance,
writing that in modern life human beings act like machines whose function is
to make money. This is absolutely wrong. The purpose of making money is the
happiness of humankind, not the other way round. Humans are not for money,
money is for humans.8
In looking here at autonomy and work, Iconnect the development and ex-
ercise of human capabilities in work not only with eudemonistic ethics but
also with the principle of humanity formulation of Kants categorical imper-
ative, which, as some scholars emphasize, requires treating the humanity in a
person as an end and never as a mere means.9 The imperative of never using
humanity as a mere means is typically taken to mean that we should never vi-
olate the autonomy of rational human beings. But, as Thomas Hill observes,
the imperative to treat humanity in a person as an end is more than a quaint
way of saying that we must respect people or respect the ability of people to
make their own choices concerning their lives.10 Rather, in referencing the
humanity in a person, the imperative means that we must treat the rational
8
His Holiness the Dalai Lama, How to Practice the Way to a Meaningful Life, translated and edited
by Jeffrey Hopkins (NewYork:Atria Books, 2002), 35. From the Catholic tradition, Pope Leo XIII gives
a similar thought in writing that it is shameful and inhuman, however, to use men as things for gain
and to put no more value on them than what they are worth in muscle and energy. John Budd, The
Thought of Work (Ithaca, NY:Cornell University Press, 2011),59.
9
Thomas Hill, Dignity and Practical Reason in Kants Moral Theory (Ithaca, NY:Cornell University
Press, 1992),3846.
10
As Hill, Dignity and Practical Reason, 39, writes, A review of Kants repeated use of humanity
in a person in The Metaphysics of Morals and elsewhere strongly suggests that, contrary to the usual
reading, Kant thought of humanity as a characteristic, or set of characteristics, of persons.... Humanity
is contrasted with our animality; and it is said to be something entrusted to us for preservation.... Its
distinguishing feature is said to be the power to set ends.
and autonomous capabilities of persons as ends. As Kant himself writes in

Groundwork for a Metaphysics of Morals, There are in humanity [Menschheit]
capacities for greater perfection which form part of natures purpose for hu-
manity in our person. Failing to develop these capacities is not compatible
with treating humanity as an end in itself; thus, there is a duty to develop ones
talents.11
In the context of working life, this interpretation of the categorical imper-
ative is nothing short of revolutionary, for it entails not merely that we have a
duty to respect an individuals choice of employment but also that the develop-
ment of human capabilities should be a goal of the provision of work. That is,
if work were structured to treat the humanity in a person as an end and never
as a mere means, then it would not be morally permissible to treat people as
objects, machines, or appendages of machines, and diminishing the rational,
autonomous and agential capabilities of a person for the sake of profit, pro-
ductivity or efficiency would also run outside the bounds of the ethical. On
the account Iforward here, it is fundamentally a matter of ethics, and in par-
ticular a matter of respecting people as autonomous beings, to treat workers
with dignity and to promote modes of working life that provide opportunities
for people to flourish, develop agential capacities, and reinforce self-respect.
Ibegin with some key distinctions concerning autonomy and work, pulling
together a case that laboring extensively at eudemonistically meaningless work
undermines workers autonomous abilities and self-respect.
1. Work, Autonomy, and Self-respect
In considering elements of autonomy in relation to work, let us make a basic

distinction at the start among autonomously chosen work, achieving autonomy
as economic independence through work, and exercising autonomous agency
in work. In the literature on work, some thinkers focus on the first concept,
characterizing autonomous work as work that is freely entered into, or as work
that is chosen because its end product constitutes a major life goal.12 On this
view, if someone reflectively determines that her personal well-being includes
educating the young, creating art, or promoting health in her community, then
choosing to teach, paint, or provide health care constitutes autonomous work.
By contrast, heteronomous work has been characterized as work whose end
11
Kant, Groundwork of the Metaphysics of Morals, translated by H. J. Paton (NewYork:Harper &
Row, 1964), 9798; 97 Hill, Dignity and Practical Reason,40.
12
John White, Education and the End of Work: A New Philosophy of Work and Learning
(London: Cassell 1997), 48, defines autonomous work as a form of activity whose end-product is
chosen as a major goal of an autonomous agent; cf. 510. Acondensed version of the argument of
this book appears in White, Education, Work and Well-being, Journal of Philosophy of Education 31:2
(1997):233247.
product has not been chosen as a major goal; heteronomous work is work
that one is constrained to perform for any number of reasons that do not stem
from ones own personal goals, such as needing money for other life goals,
satisfying others expectations, or even following Gods directive.13 As John
White observes, most work done in the world is heteronomous. People work
because they must, and what most people do at work does not meet reflectively
held personal life goals. In Section 2, Ireturn to the concept of autonomously
chosen work in addressing the notionwhich Ifind dubiousthat an autono-
mous choice to enter an agency-depriving occupation lends moral credence to
the working arrangement.
Particularly for those whose work lacks internal rewards or whose work
in itself is not a personal life goal, the hope of earning a livelihood or pro-
viding for a family provides purpose to work and a point to what one endures
on the job. Although earning an income does not itself entail full economic
independence, an income and associated benefits can provide a measure of
independence and a source of pride, self-respect, and dignity. These virtues
issue from work both in the respect that work represents a social contribution
and in the respect that work enables a person to avoid relying on others, which
places her under the will of others whose goodwill could potentially cease at
any time. Among others, Paul Gomberg notes that lacking recognition as an
economic contributor undermines personal dignity and self-esteem and that
in the United States this moral and psychological pain has not been distributed
equally across races:In the United States for the past 50years, black people
have suffered twice the rate of unemployment as whites. The scars of this as-
sault on peoples dignity are deep in many neighborhoods.14
The concept of autonomy as economic independence is entangled with
notions of autonomy no longer fashionable, as feminist philosophers have
duly critiqued conceptions of autonomy as independence as drawn from
male biographies and bound up with socially atomistic conceptions of human
beings. Some also observe that an ideal of autonomy as economic self-reliance
is manipulated in political rhetoric and used to justify denying welfare assis-
tance to poor women, whose need to raise young children renders ideals of
independence and self-sufficiency unattainable. Lorraine Code writes that in
the politics and rhetorics of social welfare...an assumed equality of access to
social goods, that requires no advocacy, underwrites the belief that failure to
achieve autonomy is a social sin.... Reliance on social services slides rhetor-
ically into a weakness, a dependence on social advocacy that, paradoxically,
invitesand receivesjudgments of moral turpitude.15 Insofar as a regulative
13
White, Education, Work and Well-being,234.
14
Paul Gomberg, How to Make Opportunity Equal: Race and Contributive Justice (Malden,
MA:Blackwell, 2007),70.
15
Lorraine Code, The Perversion of Autonomy and the Subjection of Women: Discourses of
Social Advocacy at Centurys End, in Relational Autonomy:Feminist Perspectives on Autonomy, Agency
ideal of autonomy as economic independence or self-sufficiency expresses a

political ideology that is simultaneously oppressive and unattainable, the fem-
inist thinker may be inclined to simply shelve the notion in a dustbin of dated
ideas or to work to supplant the concept of self-sufficiency with one of sup-
portive interdependence.
I will not spend much time here on the concept of autonomy as economic
independence, but Iwould pause to note, first, that we should observe a dis-
tinction between personal autonomy and economic autonomy:16 some feminist
philosophers see economic autonomy as a dimension of personal autonomy
that women have good reason to seek,17 although others lay emphasis on mor-
ally problematic implications of a social ethos of financial independence.18
Iwould also emphasize that in the present time, in which Wal-Mart, the most
powerful corporation on the planet has been likened to a profiteering monster,
it is not irrelevant to the oppression of workers that they toil at jobs that do
not enable financial independence. At the same time that a purpose of work
deeply felt as such among many everyday peopleis to achieve a measure of
financial independence to support oneself and ones family, one of the major
scandals of our time is that many employees of profitable corporations cannot
manage a living above poverty lines. Some Wal-Mart employees require pub-
lic assistance to survive,19 and women work full-time in garment factories,
assembly lines, call centers, and fast food operations and other restaurants,
among other places, yet still cannot adequately feed their children because of
low wages and diminishing benefits that exploit workers as human resources
and the Social Self, edited by Catriona Mackenzie and Natalie Stoljar (NewYork:Oxford University
Press, 2000),194.
16
Diana Meyers, Self, Society and Personal Choice (NewYork:Columbia University Press, 1989),
12, distinguishes personal autonomy from economic autonomy, where the latter represents an ideal of
financial self-sufficiency that people seek to prevent the possibility that others might gain power over
them through their needs. If one can take care of oneself, one is beholden to no oneneither to the
state nor to any other individual. Thus, one is at liberty to live as one chooses. Friedman, Autonomy,
Gender, Politics, 4749, also writes that, although there is a superficial resemblance between philo-
sophical conceptions of personal autonomy and conceptions of independence and self-sufficiency in
popular understanding, personal autonomy and financial independence are distinct notions. On her
account, financial independence is related to personal autonomy as a condition that can promote the
realization of autonomy, but financial independence is no constitutive part of autonomy, nor is it
causally sufficient forit.
17
Marina Oshana, Personal Autonomy in Society (Aldershot, UK: Ashgate, 2006), 87, identifies
financial self-sufficiency as one of several social-relational properties of autonomy, arguing that a level
of economic autonomy that enables a person to be independent of others is a requirement of personal
autonomy. Simone de Beauvoir similarly acknowledges economic independence as a component of
womens liberation.
18
See Eva Feder Kittay and Ellen K. Feder (eds.), The Subject of Care: Feminist Perspectives on
Dependency (Lanham, MD:Rowman and Littlefield, 2002), particularly Iris Marion Young, Autonomy,
Welfare Reform and Meaningful Work,4060.
19
As given in social commentary such as Robert Greenwald (dir. and prod.), Wal-Mart:The High
Cost of Low Price (Brave New Films,2005).
in the most degrading sense of the term; that is, workers become tools of profit
for others.20 So it is not the case that a social sin of failing to achieve self-reliant
independence falls on women or men who fail to work enough but rather that
a social sin falls on corporations that net enormous profits and that can afford
to pay adequate wages and benefits but instead place profit ahead of people.
Finally, if the notion of autonomy as economic independence remains dated
and problematic, it also remains perversely relevant in illuminating an appre-
ciable dimension of oppression in relation to workone distinct from autono-
mous work as freely chosen and from autonomous agency in work itself.
For a picture of work that permits neither economic autonomy nor autono-
mous agency in work, the reader may consider the work portrayed in Fast
Food Women, in which filmmaker Anne Lewis records women employees
whose nearly every move behind the counter and in the kitchen is predeter-
mined by management, including the number of times a skillet is shook and
the number of times pieces of chicken are rolled in batter. At KFC, the func-
tion of the female employee is to count to seven as she shakes a skillet and to
count to ten as she rolls chicken in batter, on the assumption that it is best for
quality control, and ultimately for company profit, that fast food women be
relieved of the need to think or make judgments about cooking.21 Managers of
a Druthers restaurant in Whitesburg, Kentucky, comment in the film that the
work is not the sort that a married man would seek out, in part because it does
not pay a living wage and carries no benefits. Rather, the work is suitable for
a woman who will derive a sense of accomplishment from completing a job
while following orders and whose father or husband perhaps has good pay and
benefits through his coal-mining occupation or other work.22
20
Consider here the work of journalists such as Barbara Ehrenreich, Nickel and Dimed:On (Not)
Getting By in America (NewYork:Henry Holt and Company,2001).
21
When interviewed in Anne Lewis, Fast Food Women (Appalshop Film, 1991), Barbara Garson
contrasts the skill and flair exhibited in the work of a short order cook in a local diner of the 1950s with
the Tayloristic principles of the late twentieth-century fast food industrywhich essentially extract
skills and knowledge from an original cohort of workers and transfers them into machines, systems,
programs, and sets of rules for new employees, so that workers who are cheaper and more expendable
need only follow beeps and buzzers, pull knobs, and turn cranks or perform other insignificant and
mechanical movements of limbs. The short-order cook might whistle at work or swear at work but in
any case exhibited a personality while at work; even if his or her work were largely routine, its details
and execution were not predetermined and regimented by management. Similarly, in giving a portrait
of his mother as a waitress, Mike Rose, The Mind at Work: Valuing the Intelligence of the American
Worker (NewYork:Penguin Books, 2004), chap.1, suggests that waitressing permits skill, flair, and
judgment, as waitresses develop and rely on complex memory and navigation skills and create order
amid lunchtime chaos. But in the chain restaurant of the twenty-first century, even flare itself can be-
come regimented when the dialogue and attitudes used for taking customer orders becomes scripted
and when waiters and waitresses are asked by management to select a dozen pins and buttons to wear
at work to exhibit flair.
22
Lewis, Fast FoodWomen.
Autonomous Agency inWork

Work can also support or stifle autonomous agency in the respect that work
itself permits opportunities for conceiving and carrying out projects, making
decisions, exercising judgment, taking responsibility for decisions, forming
goals, planning methods by which to accomplish goals, adjusting goals and
methods in light of experience, and other aspects of autonomous agency. This
aspect of autonomous agency in work is crucial to consider for a few reasons.
First, empirical literatures on work and well-being indicate that the intrinsic
features of work (i.e., what happens in work itself) have a more profound effect
on worker well-being than the extrinsic features of work, such as wages, job se-
curity, and equality of opportunity for positions.23 Opportunities within work
to exercise abilities, learn new skills, take initiative, and make independent
judgments have more influence on mental health and happiness than extrinsic
aspects of work for the reason that, as James Bernard Murphy summarizes,
personal happiness and well-being are produced more by what people do
than by what they possess:above a certain minimum income, differences in
the enjoyment of work are more important than differences in income for
overall happiness.24 In examining the impact of full-time unskilled, routine
occupations on mental health, Arthur Kornhauser writes in a classic study that
conditions of work and accompanying modes of life at lower skill levels do,
in fact, impose more severe deprivations, frustrations and feelings of hopeless-
ness....Workers in better positions experience a greater degree of fulfillment
of their wants and enjoy correspondingly greater feelings of satisfaction, ade-
quacy, and self-regard.25
Second, as indicated in empirical and philosophical literatures on the im-
pact of work on the person, a lack of opportunities for autonomous agency
within work undermines the personal autonomy of workers even off the job.
In a study identified as marking the beginning of the contemporary study
of industrial-organizational psychology,26 Kornhauser examines mental
health among workers in the Detroit automobile industry, finding that
23
Murphy, Moral Economy of Labor,2.
24
Murphy, Moral Economy of Labor, 4.According to recent empirical studies on income and hap-
piness, in the contemporary United States $75,000 is the level of income above which increases in
income cease to correlate with increased experiences of happiness. Below $75,000 (which is still quite
high relative to what many workers earn), a lack of money brings both emotional misery and low life
evaluation, according to Princeton University professors Angus Deaton and Daniel Kahneman. See D.
Kahneman and A. Deaton, High Income Improves Evaluation of Life but not Well-Being, Proceedings
of the National Academy of Sciences of the United States of America 107:39:1648916493.
25
Arthur Kornhauser, Mental Health of the Industrial Worker:ADetroit Study (NewYork:John
Wiley & Sons, 1965)269.
26
As characterized by M. J. Zickar, Remembering Arthur Kornhauser: Industrial Psychologys
Advocate for Worker Well-Being, Journal of Applied Psychology 88:2: 363369; M. Tausig and R.
Fenwick, Mental Health in Social Context (NewYork:Springer, 2011),3.
individuals working extensively at routine production jobs tend to have poor

mental health and, in particular, diminished drives toward accomplishing
self-determined life goals. Kornhauser finds that this diminishing of personal
autonomy occurs with respect not only to personal work or career goals but
also to nonworkinglife:
The general level of purposive striving and active orientation to life is rela-
tively low among factory men and lower in the routine production groups
than among those doing more skilled and varied work.... Factory employ-
ment, especially in routine production tasks, does give evidence of extin-
guishing workers ambition, initiative, and purposeful direction toward
lifegoals.
Our own results present repeated illustrations; the unsatisfactory mental
health of working people consists in no small measure of their dwarfed
desires and deadened initiative, reduction of their goals and restric-
tion of their efforts to a point where life is relatively empty and only half
meaningful.27
In contrast with those whose jobs require skill and involve a variety of tasks
and responsibilitieswho have the highest mental health scoresfactory
workers whose jobs are automated, regimented, and repetitive tend to expe-
rience a debilitating grind, lower self-esteem and weakened personal ambi-
tion and initiative, among other mental health problems.28 Insofar as extensive
employment in routine unskilled work undermines purposeful direction
toward life goals, not to mention deadening initiative and dwarfing desires,
such work undermines an important aspect of autonomous agency. As Marina
Oshana characterizes a basic intuition about autonomous personhood, the
autonomous person formulates certain goals as relevant to the direction of
her life, and is able to pursue these goals and make them effective in action.29
Kornhauser notes that it is not only factory workers who suffer from dimin-
ished drives to accomplish personal goals, as many types of jobs can damage
workers psychologically and as mental health has many roots, ranging from
personal background and upbringing to present life circumstances. But the
evidence of his study shows genuine effects of job conditions on mental health,
particularly with respect to opportunities to for workers to exercise skill and
abilities on thejob.30
27
Kornhauser, Mental Health of the Industrial Worker, 252, 269270.
28
Kornhauser, Mental Health of the Industrial Worker, 266268.
29
Marina Oshana, Personal Autonomy in Society, Journal of Social Philosophy 29:1 (Spring
1998):82.
30
As Kornhauser, Mental Health of the Industrial Worker, writes, Many interrelated characteris-
tics of jobs contribute jointly to the comparatively high or low average mental health of occupational
groups.... By far the most influential attribute is the opportunity the work offersor fails to offerfor
use of the workers abilities and for associated feelings of interest, sense of accomplishment, personal
growth and self-respect(263).
Additionally, in longitudinal studies of workers in a variety of occupations

conducted over ten years, Melvin Kohn and Carmi Schooler show that dimen-
sions of work not only reflect but also affect the personalities and intellectual
abilities of workers, effectively overturning assumptions that personalities and
abilities of workers emerge in final form before careers begin or that workers
remain psychologically unaffected by what happens at work.31 According to
Kohn and Schooler, a reciprocal relationship between work and personality
pertains both to levels of intelligence and to aspects of autonomous agency,
including initiative and self-direction:
In industrial society, where occupation is central to mens lives, occupa-
tional experiences that facilitate or deter the exercise of self-direction come
to permeate mens views, not only of work and of their role in work, but
also of the world and of self.... The conditions of occupational life at lower
social-stratification levels...foster a narrowly circumscribed conception
of self and society, and promote the valuation of conformity to authority.
Conditions of work that foster thought and initiative tend to enlarge mens
conceptions of reality; conditions of constraint tend to narrow them.32
In brief, complex work that permits self-direction fosters intellectual flexibility
and autonomous self-conceptions among workers, whereas work that is rou-
tine, that is closely supervised, and that does not permit exercise of skill, intel-
ligence, or self-direction encourages both conformity to authority and narrow
conceptions of self among workers. Forms of work at lower social stratification
levels often preclude autonomy in work in the respect that workers pursue
goals determined by others, in ways that others deem appropriate, where
often, as Adina Schwartz adds, even the order in which [workers] perform
operations, the pace at which they work, and the particular bodily movements
they employ are largely determined by others decisions.33
31
Melvin Kohn and Carmi Schooler, Work and Personality (Norwood, NJ:Ablex Publishing, 1983),
esp.103.
32
Kohn and Schooler, Work and Personality, 33. Rather than using the term autonomous agency,
Kohn and Schooler use the concept of self-direction, looking at occupational self-direction in relation
to several facets of workers personalities and social positions. Although the central concepts and topics
of Kohn and Schoolers work do not fit squarely into the philosophical literature on autonomy, Ibelieve
there is enough conceptual overlap that the work of Kohn and Schooler bears relevance to certain ques-
tions that concern philosophers of autonomy.
33
Adina Schwartz, Meaningful Work, Ethics 92 (1982):634. Iam indebted to Schwartz partic-
ularly for her work in drawing together a case that work bears a formative influence on the worker.
As I discuss at length in my book on meaningful work, however, I disagree with Schwartz on the
role that the state should play in creating opportunities for meaningful work. Schwartz argues that we
should ask for government measures to effect rearrangements in industrial employment and to enforce
a moral imperative that no one should be employed in purely routine occupations that stunt autono-
mous development. See, e.g., Schwartz, Meaningful Work, 645. In contrast, Iargue that the formative
thesis need not entail that we call upon the state to minimize meaningless work or promote meaningful
work, and Iexamine other ways social institutions can promote meaningfulwork.
Work and Self-respect

But why might a lack of autonomy in work foster a lack of autonomy in the
worker herself? Some philosophers writing on work appeal here to connec-
tions among work, identity, and self-respect. For instance, in arguing that lib-
eral political philosophers and recognition theorists owe greater attention to
work, Beate Roessler highlights that work bears upon practical identity:work
has an influence on how we live, on who we are, and how we see ourselves.34
Work affects our relations with others and with ourselves, and in light of the
impact of work on identity it can hardly be pretended that we can switch
easily between different roles and forget...the sort of work we have been doing
for hours. In a social context in which work dominates peoples lives and plays
a central role in identity formation, it is implausible to think that alienated
work would have no impact on a persons self-conception and her relations
with others in other spheres of life, and that she could simply choose, without
cost, to undertake alienated work in order to autonomously pursue other
interests.35
Some also argue that a lack of autonomy in work can undermine personal
autonomy by undermining the self-respect or self-worth of employees. When
workplaces preempt employee decision making, script employee attitudes, de-
termine the precise ordering and pace of workplace operations, monitor em-
ployee activities, and subject employees to close, intrusive supervision and
constant correction (or the threat of it), Richard Lippke writes, workplaces
become inimical to individuals developing and maintaining a sense of them-
selves as worthy of autonomy.36 By contrast, workplace practices that convey
trust for employees and that give employees latitude to make decisionswhich
entails giving employees a chance to make mistakes or do wrongconvey that
employees are worthy of autonomy. And encouraging an individual to believe
he is worthy of autonomy may be, in the end, the most effective way of protect-
ing his autonomy.37
Along similar lines, John Rawls notes that meaningful work provides a key
source of self-respect, writing more than once in his later works that the lack
of...the opportunity for meaningful work and occupation is destructive...of
citizens self-respect and suggesting for this reason that society serve as an
employer of last resort.38 Self-respect derives partly from the esteem of others
34
Beate Roessler, Meaningful Work:Arguments from Autonomy, Journal of Political Philosophy
20:1 (2012):82.
35
Roessler, Meaningful Work,83.
36
Richard Lippke, Work, Privacy and Autonomy, Public Affairs Quarterly 3:2 (April 1989):44.
37
Lippke, Work, Privacy and Autonomy,43.
38
John Rawls, Political Liberalism, paperback edition (NewYork:Columbia University Press, 1996),
lix. Rawls repeats the idea in Law of the Peoples (Cambridge, MA:Harvard University Press, 1999),
50. See also Rawls, A Theory of Justice (Cambridge, MA:Harvard University Press, 1971), 440ff. On
the importance of self-respect and self-worth for autonomy, see, e.g., Paul Benson, Free Agency and
and, in particular, from the judgments of our associates concerning the worth
of our activities: unless our endeavors are appreciated by our associates it
is impossible for us to maintain the conviction that they are worthwhile.39
Although in A Theory of Justice Rawls suggests that nonworking social life can
provide social esteem for worthy endeavors, in turn providing a social basis
of self-respect,40 he shifts in The Law of the Peoples and in the introduction
to the paperback edition of Political Liberalism to suggest that, in particular,
opportunities for meaningful work are needed to provide a social basis for
citizens self-respect. As Jeffrey Moriarty interprets this shift, Rawls comes to
reject an assumption at play in A Theory of Justice that meaningful work pro-
vides but one avenue for a social basis of self-respect, which presumably could
be achieved in leisure activities for those whose work fails to provide a sense
of self-worth; at the time of his later works Rawls instead believes that we
cannot merely hope that if people cannot find meaningful work, they can get
self-respect from other activities, such as chess or softball.41
To be sure, work is not the only avenue by which a person can achieve
self-respect, enjoy the exercise of realized capacities, or experience autono-
mous self-expression. It is possible that some people may acquire these basic
goods in leisure activities, although the empirical literature indicates that such
a possibility is slim:people tend to apply the habit developed at work to their
leisure: mindless work leads to mindless leisure whereas challenging work
leads to challenging leisure.42 But in any case a mere possibility that a person
can live a richly autonomous life without eudemonistically meaningful work
is hardly sufficient ground for an argument concerning questions of work and
social justice, which beckon us to consider what social structures are likely
to produce or encourage in human persons. The possibilityadvanced by
Whitethat rich self-centered fainants can live autonomously and find per-
sonal fulfillment in a round of leisure activities avails very little in my mind.43
Self-Worth, Journal of Philosophy 91 (1994):650668; Joel Anderson and Axel Honneth, Autonomy,
Vulnerability, Recognition and Justice, in Autonomy and the Challenges to Liberalism: New Essays,
edited by John Christman and Joel Anderson (NewYork:Cambridge University Press,2005).
39
Rawls, Theory of Justice, 440, cf.544.
40
Rawls, Theory of Justice,442.
41
Moriarty, Rawls, Self-Respect, and the Opportunity for Meaningful Work, Social Theory and
Practice 35:3 (July 2009):450.
42
Murphy, Moral Economy of Labor, 4. Likewise, Kornhauser, Mental Health of the Industrial
Autoworker, 267, notes that the leisure activities of factory workers in routine jobs tend to be narrow
and routine, with little indication of self-development and self-expression or devotion to larger social
purposes:many appear to be groping for meaningful ways to fill their spare time but with little con-
ception of the possibilities and with inadequate preparation or stimulation.
43
White, Education, Work and Well-being, 241. On the whole, White advances an argument
against writers like myself that autonomously chosen work is not a central element of the good life;
he believes the possibility of living well without autonomous work should be reflected in social policy,
which should encourage a wide variety of ways of life in which autonomous work mightor might
notfind a place(241).
Alife lived meaningfully in relation to others involves work, whether within or

outside the home, and for most of us it is work that absorbs a predominance of
productive energies and permits key opportunities for others recognition and
esteem, without which self-respect is liable to collapse.44 For the reason that, as
Rawls suggests, work provides an important source of social esteem that serves as a
social basis of self-respect, those who perform work that does not merit esteem
such as work that is viewed as dirty, disgusting, or degradedcan in turn develop
a sense of themselves as degraded and lowly people. Michael Walzer writes in
conveying the work of sociologist Stewart E.Perry, When a garbage-man feels
stigmatized by the work he does...the stigma shows in his eyes. He enters into
collusion with us to avoid contaminating us with his lowly self. He looks away;
and we do too. Our eyes do not meet. He becomes a non-person.45
2. The Dehumanization of Eudemonistically MeaninglessWork
A lack of autonomous agency in work undermines the flourishing of a worker

in yet another crucial respect, namely, that work is often experienced as dehu-
manizing when it fails to permit development or exercise of autonomous capa-
bilities or skills, or expression of individual agency or identities. This experience
occurs particularlybut not exclusivelyin forms of automated, mechanized,
or regimented work, which depletes the humanity of a person by making the
person feel like an automated thing. At issue here is the Tayloristic outsourcing
of thinking and skill that a person would otherwise put into working, thereby
determining that workers will not act like thinking, skill-exercising human
beings while on the job and thereby that workers are not valued as special and
irreplaceable people but rather that people can be treated as interchangeable
cheap human resources.46
44
An integral connection between work and self-respect is argued for in a number of empirical
and philosophical literatures. Moriarty, Rawls, Self-Respect, and the Opportunity for Meaningful
Work, 457n30, compiles a helpful list of literatures.
45
Michael Walzer, Spheres of Justice:ADefense of Pluralism and Equality (NewYork:Basil Blackwell,
1983), 165. Walzer is here giving a point made by Stewart E. Perry, San Francisco Scavengers:Dirty Work
and the Pride of Ownership (Berkeley:University of California Press, 1978),7.
46
For but one example, consider the words of a former McDonalds griddle man interviewed in
Barbara Garson, The Electronic Sweatshop (NewYork:Simon and Schuster, 1988), 17, 20:They called
us the Green Machine, says Jason Pratt, recently retired McDonalds griddle man, cause the crew
had green uniforms then. And thats what it is, a machine. You dont have to know how to cook, you
dont have to know how to think. Theres a procedure for everything and you just follow the proce-
dures.... You follow the beepers, you follow the buzzers and you turn your meat as fast as you can. Its
like Itold you, to work at McDonalds you dont need a face, you dont need a brain. You need to have
two hands and two legs and move em as fast as you can. Thats the whole system. Iwouldnt go back
there again for anything. For a classic critique of dehumanizing aspects of automated and unskilled
work in the twentieth century, see Harry Braverman, Labor and Monopoly Capital, 25th anniv. ed.
(NewYork:Monthly Review Press,1998).
Some readers will see this point as Marxist since Marx critiques industrial
forms of work as mutilating human persons into fragments and calls on the po-
tential of work to offer a person an objectification of his individual humanness
in the world. But the previous point would be better catalogued as a Kantian
critique of Taylorism, and it is important to observe that the basic idea that work
bereft of individual agency dehumanizes the person derives from a variety of
intellectual traditions,47 and one does not need to be schooled in Marxism to
appreciate it. When Studs Terkel interviewed workers for his tome Working, he
compiled a collection of experiences of dehumanization in jobs that diminish or
restrict the potential of workers to conceive, plan, imagine, solve problems, or
otherwise think creatively or constructively or move about freely atwork:
Im a machine, says the spot-welder. Im caged, says the bank teller, and
echoes the hotel clerk. Im a mule, says the steelworker. A monkey can do
what Ido, says the receptionist. Im less than a farm implement, says the
migrant worker. Im an object, says the high fashion model. Blue collar and
white call upon the identical phrase:Im a robot.48
I would argue that what is fundamentally at play here is not merely that work-
ers are discontented, as Terkel himself suggests, but rather also that as a human
activity working has a basic ambiguity of agency and utility, insofar as work
simultaneously allows one to conceive and carry out projects, thus exercis-
ing agency, as well as to feel useful in serving needs and desires, thus exer-
cising utility. When elements of human agency and freedom are taken from
work, work degenerates from a meaningful experience of feeling oneself useful
through an engagement of ones mind or body in the world, into an experience
of feeling like a cog in a machine, or like a robot, an animal, or an implement.
Work can feel meaningless for several reasonsincluding a futile outcome,
an apparent lack of purpose, or a failure to engage an individuals talents, intel-
lectual capabilities, or artisan skillsbut among the several facets of mean-
ingless work, dehumanization and degradation stand out in undermining a
meaningful experience of work. This dehumanization issues partly from a pro-
liferation of automation in working life, which requires not that an employee
act as a person exercising human capabilities but only that she use her voice,
her hands, or her legs, as determined by a system.49 But dehumanization and
47
Consider the arguments of Adriano Tilgher, Homo Faber:Work through the Ages (Chicago:Henry
Regnery Company, 1930); Schumacher, Good Work; or Pope John Paul II, Laboren exercens, Encyclical
Letter, 1981.09.14.
48
Studs Terkel, Working:People Talk about What They Do All Day and How They Feel about What
They Do (NewYork:W.W. Norton & Company, 2004 [1972]), xixii.
49
Even as automation creates jobs that are hardly set up for the exercise of individual worker in-
genuity, pockets of ingenuity, creativity, and accomplishment can nevertheless emerge even in the con-
text of performing automated work. Factory workers interviewed in Barbara Garson, All the Livelong
Day:The Meaning and Demeaning of Routine Work (Garden City, NY:Doubleday, 1975), ixxvi, report
degradation also occur outside of repetitive automated work in work in which

people act or serve as implements, parts or objects, as when women are used
in prostitution or pornographic films and other media, and when women or
men serve as towel holders, toiletry dispensers,50 or signposts. The employ-
ment of people to stand on street corners with signs reading hot and ready to
attract consumers into pizza parlors (which is a standard marketing strategy of
Little Caesars in the area in which Ilive) does not cohere with human dignity.
This kind of employment reduces the worker to functioning as a thing, makes
a mockery of the human capacities of a person, and fails to permit virtues
associated with work, such as honor or pride, on account of a lack of agency
required for thework.
Some may argue that characterizing oppressed workers as tools is mis-
taken and overlooks an indelible element of autonomous agency exercised in
an individuals choice to enter a particular employment. To be sure, choice
of employment does represent a basic dimension of autonomy in relation to
work and, more broadly, reflectively made choices figure prominently in many
accounts of autonomy. But even if we often have reason to respect individual
choices, we also have reason to be skeptical when appeals to the value of indi-
vidual choice serve to justify the distribution of limited goods in competitive
social and economic environments. In particular, we have reason to be skep-
tical of attempts to justify someones working at an oppressive job by appeal
to the fact of her having chosen to work at such a position. As a philosophical
or rhetorical maneuver intended to sanctify an employment relationship, an
appeal to individual choice or consent appears dubious if systemic inequalities
limit available opportunities in the first place, and this maneuver flounders
when the forms of work in question are inherently unchoiceworthy, such that
a certain amount of coercion must be present to force workers with few or no
alternatives into such occupations.51
on varied creative maneuvers they intersperse throughout the workday to achieve moments of inge-
nuity and feelings of purposefulness and fulfillment, such as allowing work to pile up to experience a
few minutes of purposeful exertion in catching up, which creates opportunities for minor goals and
fulfillments. But upsurges of worker creativity and purposeful exertion amid conditions of speed,
heat, humiliation, [and] monotony likely demonstrates not that working on an assembly line pro-
vides meaning or fulfillment but, rather, that the human need for exercising agency, for reaching goals,
for displaying some measure of individuality, and for feeling that one accomplishes a task creatively
are basic enough in human well-being that workers find opportunities for these needs even on an
assemblyline.
50
As in the job of a washroom attendant, whose function is to wait on people in restrooms and
to dispense towels and toiletries. One washroom attendant employed for fifteen years at the Chicago
Palmer House, Louis Hayward, describes the physical work of waiting on men in restrooms as an
automatic thing....It doesnt require any thought. Its almost a reflex action. Iset my toilet articles up,
towelsand Im ready. Terkel, Working, 106. In its social function, he believes his work serves to bol-
ster the egos of bathroom visitors:when a man visits the restroom and receives an acknowledgment
from the attendant, it builds his ego up a little bit.... Im building him up(107).
51
The point is suggested by Michael Walzer, Spheres of Justice, 165, who writes of jobs that are hard
in the sense of being harsh, unpleasant, cruel, difficult to endure that they are like prison sentences
As Gomberg notes in criticizing Ronald Dworkin, asserting that a person

occupies a social position as a result of past choices primarily serves an ide-
ological purpose as a moral sanctification of a social order, but there is a
basic error in transferring moral concepts of praiseworthiness or blamewor-
thiness into social contexts in which basic goods exist in limited supply.52 As
Gomberg has it, in a competitive economic system in which employment is a
limited good, the functioning of state bureaucracies that handle unemploy-
ment depends on a large percentage of the unemployed living in a state of
discouragement. In such a context, rhetoric to the effect that those who are un-
employed choose to avoid the unemployment lineand thus have only them-
selves to blame for their conditionprovides an appearance of justification
of the state of things but obscures social forces, institutions and policies that
cause a certain percentage of unemployment in the first place.53
Some readers may reach here for a stock distinction made in the context
of Kantian ethics, wherein it is commonly noted that it is not wrong to treat a
person as a means (as happens in work of all sorts) but wrong to treat him as
a mere means, which disrespects his capabilities as an autonomous, rational
agent, who as such must make his own decisions free from manipulation and
coercion. In illuminating this notion, some ethicists quickly clarify that it is
not morally objectionable, for example, to use a plumber to fix a broken pipe
drain, as long as the plumber understands the situation and chooses, from
his own will, to participate in the employment.54 This clarification and illus-
tration is, in fact, crafty in avoiding pervasive moral ambiguity, insofar as the
worker in this example is one who uses intelligence and agency while on the
job, making decisions and judgments in the context of exercising compe-
tency as a craftsman. The plumber who is employed freely and fairly and who
exercises developed skills to solve problems, to install materials, or to make
repairs is not exploited as a tool in the manner in which a factory hand or
a sweatshop worker are oppressed as relatively expendable, interchangeable
tools of production.55 Even if all were to exercise agency in an initial choice of
in that people do not look for them and would not choose them if they face minimally attractive
alternatives:This kind of work is a negative good, and it commonly carries other negative goods in its
train:poverty, insecurity, ill health, physical danger, dishonor and degradation. And yet it is socially
necessary work; it needs to be done, and that means someone must be found to doit.
52
Gomberg, How to Make Opportunity Equal,23.
53
Gomberg, How to Make Opportunity Equal,2324.
54
As in the widely used ethics textbook James Rachels and Stuart Rachels, The Elements of Moral
Philosophy, 7th ed. (NewYork:McGraw-Hill, 2012), 138139.
55
For a discourse on the intelligence and agency required for nonroutinized manual work, see
Matthew Crawford, Shop Class as Soulcraft: An Inquiry into the Value of Work (New York: Penguin
Press, 2009); Mike Rose, The Mind at Work: Valuing the Intelligence of the American Worker
(NewYork:Penguin Books, 2004). For Crawford, the satisfactions of useful work accrue particularly
to skilled practitioners of manual arts, like carpentry, plumbing, and motorcycle and car repair. Since
these forms of work are necessarily situated in a particular context, they are inherently resistant to
forms of external managerial or corporate control that undermine human agency and make work
employment, not all exercise agency in employment, and not all are equally
exploited. And focusing on an exercise of autonomous agency in an initial
choice of employment obscures ways work can itself undermine or enhance
autonomous agency.
To be clear then, it is not the fact of being utilized as an instrument of pro-
duction that is itself at issue. Being instrumental in producing, serving, creat-
ing, and fabricating lies in the nature of workinga fact also revealed through
meditation on the meaning of employmentwherein one often finds satisfac-
tion in making oneself useful, being purposeful, or being a means of achieving
something of value. Rather, it is being used extensively as a tool, an implement,
an object or a body, wherein a person is not also simultaneously expressing
agency, which is dehumanizing. It is also dehumanizing to be treated as an ex-
pendable resource or to toil in a system as a nonthinking tool of production for
several hours a day, which after time damages the body and mind and depletes
intelligence and other human capabilities. When the working day is limited
and one has opportunities to flourish outside of eudemonistically meaningless
work, perhaps being a tool of production at work need not be felt as altogether
dehumanizing. However, there are some for whom oppressive work effectively
predominates life, as when one labors at eudemonistically meaningless work
for sixteen hours a day, whether at one job or two, and here a person is likely to
feel depleted of humanity or human vitality, and effectively transformed into a
means for others. Oppression at work clearly comes in degrees.
It is also interesting to consider, as Imentioned above, that some interpret-
ers of Kant argue that in giving an imperative to treat humanity as an end
an never as a mere means, Kant himself lays emphasis on respecting the hu-
manity in persons, that is, on respecting the rational and autonomous abilities
within persons. Thomas Hill observes Kant repeating that respecting people
as ends requires treating the humanity in a person as an end and never as
a mere means.56 The familiar dictum of treating persons as ends represents
an abbreviation of treating humanity in persons as ends, Hill argues, and for
Kant humanity represents a characteristic of persons, whose distinguishing
features include the rational capabilities of setting ends and forming goals. In
interpreting the principle of humanity formulation of the categorical impera-
tive, Allen Wood also writes that the basic issue for Kant is that we disrespect
vulnerable to dehumanization and degradation. Building and fixing are embedded in a community
in which the individual worker remains responsible for his or her own work and in which excellence
at work comes with the exercise of judgment, the making of a social contribution, the feeling of pride
in ones work, and the transformation of objective reality by ones own hands. Nonroutinized manual
work demands intelligence; the physical circumstances of the jobs performed by carpenters, plumbers
and auto mechanics vary too much for them to be executed by idiots. One feels like a [person], not like
a cog in a machine (5253).
56
Hill, Dignity and Practical Reason, 3846,10.
humanity in placing things of lesser value ahead of our rational nature, which
constitutes our humanity and which is an end in itself.57
What would it require of a human community to treat the humanity in a
person as an end? In the context of work, it would entail that a goal of work be
the development or exercise of the rational, autonomous, and agential capa-
bilities of a person and that diminishing these gifts as a means for achieving
economic values of profit or productivity or increasing efficiency is outside
moral bounds. That is, the full realization of a moral imperative to treat the
humanity in a person as a end would entail a revolution in modern economic
life, which as presently structured often demands not that work serve the end
of developing and exercising human capabilities but that the worker adapt
himself to the needs of the workwhich means, of course, primarily to the
needs of the machine.58 It may be hard to fathom such a moral departure from
present economic realities. But, as we know, implementing the categorical im-
perative can entail such radical transformations in human practices that, in a
specific historical moment in which people are enveloped and enculturated
in immoralities, meeting moral demands can appear bewilderingly near im-
possible. As Wood writes, the formula of using humanity as an end and never
only as a means is rather like the Sermon on the Mount...whose demands
require such a radical departure from our customary practices and accepted
attitudes toward ourselves and others that we are at first perplexed when we
try to apply them.59
Some may see this call for ethical transformation in economic life as simply
tantamount to rejecting capitalism and calling for communism or Marxism
in its stead. In response, Inote that Marx provides one rich source of criti-
cism of meaningless work in capitalist economies, but pigeonholing criticism
of meaningless work as Marxist overlooks a few facts. First, it overlooks the
convergence of a plurality of philosophical and religious traditions around the
importance of work in realizing part of the human good. Second, this pigeon-
holing overlooks the possibility for meaningful work in the context of capi-
talist employment relationships. The question of whether capitalism contains
an inherent proclivity toward depriving people of autonomous agency depends
partly on the ethical values paired with it and partly on whether the pursuit
of profitability, productivity and efficiency is pure and unbridled, or limited
and tempered by a respect for autonomy and other human values. Third, iden-
tifying meaningful work with communism overlooks the variegated realities
of alienated labor under communist conditions: as James Bernard Murphy
observes, in both capitalism and in communism the worker is often treated
as a mere instrument, a factor of production, rather than as the subject of his
57
Allen Wood, Kants Ethical Thought (NewYork:Cambridge University Press, 1999),143.
58
Schumacher, Good Work,3.
59
Wood, Kants Ethical Thought,139.
or her work.... When Charlie Chaplin satirized the mindless monotony of the
assembly line in Modern Times (1936), he was denounced in the United States
as an enemy of capitalism and in Russia as an enemy of socialismand in a
sense he was guilty on both counts.60 In essence, promoting modes of working
life that provide opportunities for people to flourish reaches beyond tradi-
tional bifurcations between capitalism and communism, implanting ethics at
the seat of economic life and requiring respect for the exercise of agency in
workinglife.
If it is bewildering to imagine economics structured around a goal of
human development, I think it is also edifying to consider the bounds of
moral progress that workplaces have already achieved in some quarters of
the world in the twenty-first century. Consider, for instance, that it is now
commonplace to maintain as workplace idealsand to instantiate in practice
in varying measuresrational and fair hiring processes, nondiscriminatory
and harassment-free workplace environments, equitable wages and freedom
from threats, and abuse and profanity while on the job. According to histo-
rian Stanford Jacoby, not one of these ideals was in place in the United States
over a century ago, when the dominant mode of the production of commodi-
ties was the factory system, in which foremen used close supervision, abuse,
profanity, and threats to motivate faster and harder work and in which work
was highly insecure, very poorly paid, fraught with pay inequities and ethnic
discrimination, and not uncommonly secured through nepotism, favoritism,
and bribery.61 Some of us live in a workplace utopia in comparison with the
factories of the late 1800s, at which time it would have been difficult to see
possibilities for the sort of change that is now a becoming reality, and it should
not therefore be said in thinking about working life as we know it that work
just is what it is. Appreciating moral progress in working life highlights abili-
ties of human communities to transcend and reinvent workplace structures
and leads us toward a position of open-mindedness in entertaining long-range
60
Murphy, Moral Economy of Labor, 34. The prologue to Murphy, Moral Economy of Labor pro-
vides important clarification on the relationship between Taylorism and communism. Murphy notes,
for instance, that the detailed fragmentation of skilled labor into monotonous routine that once sym-
bolized the horrors of capitalism became the basis of Soviet industry from Vladimir Lenin through
Leonid Brezhnev. Indeed, Taylorism was more pervasive in Soviet Russia than it ever was in the United
States (3). Further, Murphy observes that the similar quality of the experience of work for the worker
in both capitalist and socialist systems leads apologists for both systems to emphasize distribution
and exchange rather than the dignity of work. For example, one leading Marxist theoretician, John
Roemer, says that if we were to focus on the labor process we would be forced to the bizarre conclusion
that socialist countries exploit workers just as much as do capitalist countries (3). See also Braverman,
Labor and Monopoly Capital.
61
Jacoby The Way It Was: Factory Labor before 1915, in Employing Bureaucracy: Managers,
Unions, and the Transformation of Work in the 20th Century, rev. ed. (NewYork:Columbia University
Press,2004).
possibilities for transforming elements of working life that stifle autonomous

development or undermine human dignity.
Elsewhere, Ifurther explore work and human flourishing, addressing ad-
ditional ethical, social, and political implications of the formative influence
of work on people.62 In bringing this chapter to a close, Ihighlight that work-
place structures and practices are not invariable elements of a natural order
but rather change over time, and an understanding of the impact of work
on autonomous development and self-realization can guide transformations
of workplace goals and structures so that working can become more mean-
ingful, or at least less oppressive, for more people. Ialso underscore in closing
that an initial autonomous choice of occupation by no means exhausts the
intersections between autonomy and work, and focusing only on autonomous
choice of occupation constrains perceptions of ways Tayloristic forms of work
undermine the development and exercise of autonomous capacities. Given
that autonomous capacities are developed, or stifled, within social contexts
including work, a democratic commitment to supporting the development
and exercise of peoples autonomous capacities entails an opposition to eude-
monistically meaningless forms of work, which stifle worker autonomy and
well-being.
Acknowledgments
I would especially like to thank Mark Piper and Erin Tarver for helpful com-
ments on an earlier draft of this paper. This paper has also been enriched
by discussions at the Relational Autonomy Workshop organized by Natalie
Stoljar and Catriona Mackenzie, held in Montreal in September 2012, and at
the conference on Work and Human Development organized by Nick Smith
and Jean-Philippe Deranty, held in Sydney in September2011.
62
Veltman, Meaningful Work, manuscript.
{14}
The Right to Bodily Autonomy and the Abortion

Controversy
Anita M. Superson
1.Introduction
Laurence Tribe persuasively argues in his outstanding book, Abortion: The

Clash of Absolutes, that abortion is not just about having control over ones
body but also about having the choice to be a mother:Pregnancy does not
merely inconvenience a woman for a time; it gradually turns her into a mother
and makes her one for all time.1 Being a mother goes well beyond gestating a
fetus for nine months but involves being prepared to make many personal and
financial sacrifices for the well-being of ones child to its development as a full
and productive member of societyin short, motherhood is life-changing.
Pregnancy for most women involves this choice, as only less than 3percent in
the United States put up their children for adoption.2
Tribes powerful argument affirms the seriousness of enforced pregnancy
brought about by strict antiabortion legislation that is trying to take hold
in the United States. For example, Republican governor Jack Dalrymple of
North Dakota recently enacted a law banning nearly all abortions once a fetal
heartbeat is detectable, which could be as early as six weeks into pregnancy.
Dalrymple also approved a statute, originating in Mississippi, requiring doc-
tors who perform abortions to get admitting privileges at a local hospital,
which could effectually shut down North Dakotas only abortion provider.3
1
Laurence H. Tribe, Abortion: The Clash of Absolutes (New York: W.W. Norton & Company,
1992),104.
2
According to the Guttmacher Institute, 3 percent of unmarried white women gave up their
babies for adoption from 1982 to 1988; fewer than 2percent of unmarried black women gave up their
babies for adoption during this time and earlier; and none of Hispanic women in the sample reported
giving up their babies for adoption. Tamar Lewin, Fewer Children Up for Adoption, Study Finds,
New York Times, February 27, 1992, http://www.nytimes.com/1992/02/27/US/fewer-children-up-
for-adoption-study-finds.html.
3
The Campaign to Outlaw Abortion, New York Times, March 29, 2013, http://www.nytimes.
com/2013/03/30/opinion/the-campaign-to-outlaw.abortion.html?hp&_r=18.
Dalrymple signed a bill that bans abortions based solely on genetic abnormalities
or on the gender of the fetus.4 Personhood bills, which would extend legal protec-
tion to zygotes, have been introduced in Oklahoma and Mississippi. In Kansas,
Republican governor Sam Brownbeck indicated that he would sign virtually any
measure declaring that life begins at fertilization, and then he signed a person-
hood bill.5 An abortion rights advocacy group, the Guttmacher Institute, which
classifies states as hostile, middle ground, or supportive of abortion rights, reports
that in the southern United States while only a handful of states were classified
in 2000 as hostile there is now a solid wall of hostility, from Virginia down to
Florida and over to Texas and Oklahoma.6
Despite the power of Tribes argument, if we argue in favor of abortion rights
on the grounds he suggests alone, then it will be difficult to address the National
Center for Men, for instance, which believes that men are entitled to some say
over decisions that profoundly affect their livesnamely, the freedom to choose
to be a father. In a 2007 court case, Dubay v.Wells, the Center backed Matthew
Dubay, who sued his girlfriend after she became pregnant when the two had sex.
Prior to the pregnancy, Dubay told her that he was not ready to have children,
and she said that was fine, since she was infertile and using birth control just in
case. When she discovered that she was pregnant, she wasnt willing to have an
abortion. After giving birth to a girl in 2005, she obtained a court order requir-
ing Dubay to pay $500 a month in child support. Dubays lawsuit contended that
men have a constitutional right to avoid procreation. Mel Feit, the Centers di-
rector, said, Theres such a spectrum of choice that women haveits her body,
her pregnancy. Im trying to find a way for a man also to have some say over
decisions that affect his life profoundly.7 This Michigan case was called the Roe
v.Wade for men. Although it was a case about whether child support laws vio-
late Equal Protection by applying only to men, conservatives hoped to seize on
it to gain control over abortion in both directions and to overturn Roe v.Wade.
As things turned out, the judge dismissed the case in 2006. The National Center
for Men appealed, but the appeals court said that the 14th Amendment does not
deny the State the power to treat different classes of persons in different ways.8
4
Dave Thompson, North Dakota Governor Signs Heartbeat Abortion Ban, Chicago Tribune, March
26, 2013, http://articles.chicagotribune.com/2013-03-26/news/sns-rt-us-usa-abortion-northdakota.
bre92pOua-20130326_1_heartbeat-bill-supporters-of-abortion-rights-restrictive-abortion-law.
5
DeAnn Smith and Stephen Mayer, Kansas Lawmakers Could Vote Friday on Tough
Anti-abortion Bill, April 5, 2013, http://www.kctv5.com/story/21891968/Kansas-lawmakers-could-v
ote-Friday-on-tough-anti-abortion-bill.html. The bill states that life begins at fertilization and bans
sex-selection abortion and was signed on April 5, 2013 (CBS EveningNews).
6
Abortion Laws in the South:Cutting away at Roe v.Wade, March 16, 2013, Economist http:www.
economist.com/news/United-States/21573594-rest-south-not-far-behind-strict-new-law-arkansas
-cutting-away-roe-v.
7
Judith Graham, Unwilling Father Tests Mens Rights, Chicago Tribune, March 10,2006.
8
Dubay v.Wells, 506 F.3d 422 (6th Cir.; 2007); Wikipedia, Dubay v.Wells, 506 F.3d 422 (6h Cir.;
2007), http://en.wikipedia.org/wiki/Dubay_v._Wells.
Right to Bodily Autonomy and the Abortion Controversy 303
Ironically, this case for so-called mens rights is not one prohibiting womens
right to abortion but enforcing abortion when the father does not want children.
Whatever the motive, and even though the lawsuit failed, the case makes clear
the feminist objection that the state, through its institutions and practices, uses
womens bodies as vessels for its own or mensends.
Significantly, this case makes it very clear, or so Ishall argue, that the right
to bodily autonomy must be central in any argument for abortion rights. Tribe
himself makes a great deal about bodily autonomy. But he seems to lean in
the direction of nonliberal, or less individualistic, feminists, who advocate the
centrality of motherhood, and away from liberal feminists, who advocate the
centrality of bodily autonomy and other individualistic values.9 One advantage
of the nonliberal feminist view is that if parenthood is the supreme value, then
a woman has a right to decide whether she is to become a parent even if her
doing so does not involve her body in any significant way. Any argument about
abortion rights would apply equally to in vitro development of a child as well
as to the mothers rights over the child once it is born. But if bodily autonomy
is the supreme value, then we will need a separate argument about a mothers
right over the fetus in such cases. The key issue in the standard case of fetal
development is the mothers rights over the fetus when the fetus is in her body.
If we leave bodily autonomy out of, or make it secondary in, our arguments
about abortion, we jeopardize its significance in arguments about rape, sexual
harassment, female genital mutilation, woman battering, and fetal harm cases.
Yet a significant part of our analysis of why these behaviors are wrong is that
they deny womens bodily autonomy. It is important, then, for feminists to
supplement the argument about the right to become or not to become a parent
with an argument about the right to bodily autonomy.
Interestingly, if supporters of the prolife10 or antichoice position really
do oppose enforced abortion, Dubay v.Wells should force them to examine
9
According to Martha Nussbaum, Feminist Critique of Liberalism, in Sex and Social Justice,
edited by Nussbaum (NewYork:Oxford University Press, 1999), personhood, autonomy, rights, dig-
nity, and self-respect are the terms of the liberal Enlightenment (56). One of the charges cited by
Nussbaum and made by feminist against the liberal tradition is that it is too individualistic in fo-
cusing too much on the dignity and worth of the individual and thereby diminishing the value of
community and social entities such as families, groups, and classes (56, 58, 59). Nussbaum defends
liberalism against this charge, arguing that the separateness of persons is a basic fact of human life
(62), and is at the root of making the individual the basic unit for political thought:each person has a
course from birth to death that is not precisely the same as that of any other person.... Each person is
one and not more than one....Each feels pain in his or her own body....(62).
10
Iput this term in quotes because most people who claim to be prolife about abortion are not
prolife about other issues. Only 11percent of Americans hold a consistent ethic of life position, op-
posing legalized abortion and capital punishment, for example. Seventy-nine percent of prolife
Republicans and 85percent of prolife Tea Party identifiers who say abortion should be illegal in all or
most cases also support the death penalty. See Robert P. Jones, Like Perry, Most Pro-Life Americans
OK with Death Penalty, Washington Post, September 15, 2011, https://www.washingtonpost.com/blogs/
figuring-faith/post/like-rich-perry-most-pro-life-americans-ok-with-death-penality/2011/09/15/
whether it is just the abortion part that they oppose (which makes this case no
different from that of a woman willingly undergoing an abortion) or whether
their opposition is also to the use of womens bodies in this way (which means
that they actually do see the value of the right to bodily autonomy and that
to be consistent they should acknowledge it in other cases as well). Tribe and
many other feminists believe that the prolife position is really about control-
ling women and about race,11 not about the rights of fetuses, since its advocates
are more stirred up when women gain advances in society, or when women
voluntarily have sex (i.e., are guilty) and contraception fails,12 and so on, or
when more white women than black women have abortions.13 This is because
the prolife position is inconsistent on its position on the value of fetal life.14
What does it mean to say that someone has a right to bodily autonomy?
What does this right consist in, and when does it trump other rights in con-
flict cases? These are difficult questions to answer, and in this paper I hope
only to tease out a bit what this right might amount to, by taking as my starting
point an account offered by Judith Jarvis Thomson in her famous article on
abortion.15
It is noteworthy that standard accounts of personal autonomy or autonomy
of the will tend to leave out discussion of the body and focus instead on condi-
tions that are necessary and/or sufficient for self-determination. Sarah Buss
gIQA06XUIX=blog.html. The data come from the Millenials, Religion, and Abortion Survey conducted
between April 22 and May 8, 2011, using a random sample of three thousand adults.
11
In Nazi Germany, the Third Reich banned the production and distribution of contraceptives in
1941. By 1943 the penalty for performing an abortion on a woman who was genetically fit was death.
But by 1938, Jews were defined as falling in the category of being genetically defective, and the state
readily granted abortions for women in this group. See Tribe, Abortion,5960.
12
Tribe, Abortion, notes that some people believe that while abortion should be allowed in cases of
rape, it should not be in cases of contraceptive failure because the woman volunteered in the sense
that she chose to have sex and run the risk of pregnancy. He finds it curious that there is not a wide-
spread sentiment among those who generally oppose abortion rights that abortion should be allowed
in cases of contraceptive failure, since taking contraception means that the woman does not want to
be pregnant. Tribe asks, Does this not suggest that such opponents of abortion come to their views
about the immorality of abortion not in response to the voluntary nature of the womans pregnancy but
in response to the voluntary nature of the sexual activity in which she has engaged? And does this not
in turn suggest that such antiabortion views are driven less by the innocence of the fetus...than by the
supposed guilt of the woman?(132).
13
Tribe, Abortion, notes that in the 1800s there was an increase in abortion rates among mar-
ried, white, middle-class Protestant women and that by 1860 the birth rate among white Americans of
British and northern European descent declined significantly in comparison to new immigrants who
were predominantly Catholic. Physicians who supported abortion restrictions warned of race suicide,
and fear of this was more widespread than religious antiabortion sentiment(32).
14
Tribe, Abortion, 231, 232, 236237, notes that virulent opponents to abortion such as Randall
Terry, the originator of Operation Rescue, explicitly connect their opposition to abortion with their
desire to put women back in traditional roles. He notes that our society is inconsistent in its thinking
on fetuses and frozen embryos and on fetal development and childabuse.
15
Judith Jarvis Thomson, A Defense of Abortion, Philosophy & Public Affairs1:1 (Autumn
1971):4766.
notes that most autonomy theories have in common the notion that autonomy
consists in the agents endorsing or identifying with her action.16 Much of the
current debate centers on the benefits and disadvantages of procedural and
substantive theories of autonomy. Yet these theoretical debates emphasize the
will, not the body. Discussions about bodily autonomy appear mostly, as one
might expect, in the medical ethics literature (e.g., on reproductive technolo-
gies, organ donation and sale, euthanasia, and abortion), but we can also find
them in feminist critiques of traditional autonomy theories (e.g., on deformed
desires or adaptive preferences) and the feminist literature on rape, woman
battering, and the like and how these acts stifle womens autonomy. In her
feminist critique of notions of bodily autonomy found in the medical ethics
literature, Catriona Mackenzie argues that this literature has overwhelmingly
supported what she calls the maximal choice view, which in essence amounts
to the view that autonomy reduces to choice, which is a matter of our sub-
jective preferences, no matter their content.17 This is an individualistic view
of autonomy, one that favors maximizing control over ones body and having
freedom to dispose of it as one chooses. This view, Iwould add, fits squarely
into traditional accounts of autonomy by incorporating the body in terms of
the agents preferences and desires. Bodily autonomy, then, can be viewed as
an extension of autonomy of the will through satisfaction of preferences and
desires relating to thebody.
But this conception of bodily autonomy might not satisfy feminists, who
have critiqued traditional accounts of autonomy, and have developed what
has been called relational autonomy. Relational autonomy theories are distin-
guished by their focus on the autonomous agent, in particular, on the rich and
complex social and historical contexts in which the agent is embedded.18 They
aim to develop an account of autonomy that is characteristic of agents who are
emotional, embodied, desiring, creative, and feeling, as well as rational crea-
tures.19 Along these lines, Mackenzie develops an account of bodily autonomy
that espouses the liberal and libertarian values found in the medical ethics lit-
erature yet grounds these values not in bodily ownership but in what she calls
ones bodily perspective. These values are (1) the right to noninterference
16
Sarah Buss, Autonomy Reconsidered, Midwest Studies in Philosophy 19 (1994): 95121; Peter
A. French, Theodore Ed. Uehling, Jr., and Howard K. Wettstein (eds.), Midwest Studies (Notre
Dame:University of Notre Dame Press, 1994),95.
17
Catriona Mackenzie, Conceptions of Autonomy and Conceptions of the Body in Bioethics, in
Feminist Bioethics:At the Center, On the Margins, Jackie Leach Scully, Laurel E. Baldwin-Ragaven, and
Petya Fitzpatrick, eds. (Baltimore, MD:Johns Hopkins University Press, 2010),7190.
18
Catriona Mackenzie and Natalie Stoljar, Introduction: Autonomy Refigured, in Relational
Autonomy: Feminist Perspectives on Autonomy, Agency, and the Social Self, edited by Catriona
Mackenzie and Natalie Stoljar (NewYork:Oxford University Press, 2000), 21. The workshop on re-
lational autonomy for which this paper was written was held to celebrate the ten-year anniversary of
this anthology.
19
Mackenzie and Stoljar, Introduction,21.
(the right to be free from intrusion on ones body without ones consent), and
(2)the right to bodily self-determination (the right to decide what happens
in and to your body, which, Mackenzie notes, cashes out as a right to deter-
mine what you can do with your bodyincluding sex selection, genetic en-
hancement, and organ salemore extensively, and she believes incorrectly, on
liberal and libertarian medical ethics accounts). Mackenzie, borrowing from
Paul Ricouer, Maurice Merleau-Ponty, and Kant, argues that it is because my
body is the medium through which Ilive my life that Imust have a right to de-
termine what happens in and to my body.20 Our consciousness is not separate
from our body; rather, we see the world from the perspective of our body.21
Iwould add two points. First, our bodies as much as our minds shape who
we areoften others react to us and treat us in certain ways because of some
feature of our bodies, including our perceived gender, ethnicity, and ableness
and whether we are obese or thin, attractive or unattractive, and so on, and we
see ourselves in terms of how others see us and react to us because we often
internalize social perceptions. Second, we come into the world and go out of it
with our body. Because your body is yours alone in the sense that only you in-
habit it,22 there is a way your life and your death are necessarily and inevitably
solitary events, no matter who surrounds you. Martha Nussbaum defends the
individualism of liberalism on the grounds of the necessary separatism of per-
sons and the fact that we each feel pain and pleasure in our own bodies only.
These preliminary points speak to the significance of the body for morality in
general and specifically for any full account of autonomy.
The account of the right to bodily autonomy that Idefend here is relational
in the sense that it is nuanced by feminist concerns. However, my starting point
in defending this right is Thomsons article, and many of us would characterize
her view of the right to bodily autonomy as falling squarely in the liberal or lib-
ertarian tradition that Mackenzie criticizes. Iactually do not think these views
are at odds with each otherindeed, liberalism/libertarianism and feminism
share a concern for the body that seems to come from the same place. When
many feminists discuss rape, woman battering, female genital mutilation, pre-
natal harm, and abortion, to name a few, they espouse a hands-off view that
might be defended with the libertarian assumption of self-ownership of the
body.23 The underlying view is that women are not fetal containers, nor do men
own womens bodies.
20
See Catriona Mackenzie, On Bodily Autonomy, in Handbook of Phenomenology and Medicine,
S.K. Toombs (ed.) (The Netherlands:Kluwer Academic Publishers, 2001), 417439.
21
Mackenzie, Conceptions of Autonomy,8081.
22
Iam going to avoid discussing controversial personal identity theorieshere.
23
For a discussion of the libertarian principle of self-ownership, see, e.g., David Sobel, Backing
away from Libertarian Self-Ownership, Ethics 123 (October 2012):3260.
Since my interest is to explain what a right to bodily autonomy might con-

sist in, Iwill take as my starting point Thomsons article, as it has insights on
an issue near and dear to feminists and medical ethicists alike. Along the way,
Iwill propose a few principles that tweak out the right to bodily autonomy
sensitive to concerns raised by relational autonomy theorists. Iintend for these
principles to hold ceteris paribus. Ido not believe that the right to bodily au-
tonomy holds absolutely (nor does the right to self-defense). Ileave for another
occasion an exploration of the conditions under which it may be overridden.
Thomson appeals to and elucidates the right to bodily autonomy in at
least four ways. First, she characterizes this right as a fundamental right, ei-
ther grounded in, similar to, or the same as the right to self-defense. Thomson
states, My own view is that if a human being has any just, prior claim to any-
thing at all, he has a just, prior claim to his own body...the woman has a right
to decide what happens in and to her body.24 Asecond and related point is
that this right, for Thomson, is grounded in the fact that the mother owns her
own body, making claims about it prior to other claims.25 Ownership of ones
own body grounds a right to defend ones body against things happening in
and to it.26 Third, the right to bodily autonomy also seems, for Thomson, to
be tied up with the very status of being a person. Regarding what the mother
may do as opposed to what a third party may do when it comes to abortion,
Thomson says, So the question asked is what a third party may do and what
the mother may do, if it is mentioned at all, is deduced, almost as an after-
thought, from what it is concluded that third parties may do. But it seems to
me that to treat the matter in this way is to refuse to grant to the mother that
very status of person which is so firmly insisted on for the fetus. For we cannot
simply read off what a person may do from what a third party may do.27 In
other words, a persons right to bodily autonomy is connected to, and perhaps
even stems from, her status as a personthat is, that someones status as a
person in the moral community forms the basis of at least some of the rights
she has. Finally, Thomson asserts that no one has a right to use of your body
unless you give him such a right,28 suggesting that a right to bodily autonomy
is so fundamental that only the bearer of such a right can legitimately waive it.
24
Thomson, Defense of Abortion,54.
25
26
Thomson, Defense of Abortion, 53. Even though ownership of the body grounds a right to
defend ones body against things happening in and to it, Thomson argues elsewhere that a right to
self-defense is not absolute. For example, you cannot deflect a trolley heading toward you onto a by-
stander, nor can you shoot a bystander who will then fall on the trolley and stop it from killing you,
nor can you do something that results in a bystanders death to spare your own life. All of these acts
display a lack of respect for persons, treating them as if they were not there. Thomson, Self-Defense,
Philosophy & Public Affairs 20:4 (Autumn 1991):289, 291. It is beyond the scope of this paper for me to
argue when the right to bodily autonomy, as part of the right to self-defense, can be overridden.
27
Thomson, Defense of Abortion, 52. Emphasisadded.
28
And strictly speaking, Thomsons view in this paper is that you do not waive
the right to determine what happens in and to your body; rather, you just
let someone use your body for atime.
The right to bodily autonomy plays a pivotal role in Thomsons three
analogies about pregnancy, and she argues that in certain cases it trumps
other rights, including even the right to life. In the violinist analogy, which
compares enforced pregnancy resulting from rape to having to stay hooked
up to a violinist who is attached to you without your consent, the mothers
right to bodily autonomy overrides the right to life of the fetus. Because
pregnancy was the result of rape, in no sense did the mother give the fetus a
right to use her body. In the house analogy, which compares enforced preg-
nancy in cases where the mothers life is at stake to being trapped in a house
with a rapidly growing child who will crush you to death, the mothers right
to life overrides the fetus right to life because the mothers prior existence
and the fact that the fetus uses her body for its existence break the tie be-
tween the two innocent lives. The mothers prior existence and prior right
to her body is decisive in just the same way that ones owning property is
decisive over someone elses wanting it:an older brother who owns a box of
chocolates that he does not share with his younger brother has a prior claim
to them. In the people-seed analogy, which compares enforced pregnancy
in cases of contraceptive failure to having to allow the people-seed that takes
root in your upholstered furniture to develop in your house, the mothers
right to bodily autonomy overrides the fetus right to life even when the sex
is voluntary but when the mother has taken precautions against pregnancy.
Since she has done so, she does not give the fetus a right to use her body.
Thomson concludes that in cases of rape, the mothers life being at stake,
and contraceptive failure, abortion is morally permissible.
I find Thomsons analogies entirely persuasive, but I think that her re-
liance on the right to bodily autonomy allows her to draw even stronger
conclusions than she does. Here Iwant to focus on the following two cases
where Ithink that Thomsons conclusion is weaker than it need be:(1)the
woman whose pregnancy lasts only one hour and is the result of rape; and
(2)the woman who in her seventh month of pregnancy wants an abortion
so she wont have to postpone a trip. These cases are Thomsons own, but
Ibelieve that her conclusions about them are weaker than the ones she can
draw by invoking the same premise about the right to bodily autonomy.
Ibelieve that this right has more trumping power than Thomson defends.
Iuse these cases as a springboard for developing further the right to bodily
autonomy, though of course Icannot defend a full-fledged theory here. My
aim is to put forward and then offer some precursory defenses of some
principles about this right. Thomson herself does not defend this right but
merely appeals to it in her paper, though elsewhere she argues that it is one
of the fundamental rights that a plausible moral theory ought to acknowledge,

since without such a right we are not even in the moral game.29 Iwill return to
this pointlater.
2. Case 1:When Pregnancy Lasts One Hour and Is the Result

ofRape
Regarding case 1, Thomson says: ...suppose pregnancy lasted only an

hour, and constituted no threat to life or health. And suppose that a woman
becomes pregnant as a result of rape. Admittedly she did not voluntarily do
anything to bring about the existence of a child. Admittedly she did nothing
at all which would give the unborn person a right to use of her body. All the
same...she ought to allow it to remain for that hourit would be indecent
in her to refuse.30 And ...we should conclude that she is self-centered, cal-
lous, indecent, but not unjust, if she refuses.31 Thomsons remarks about the
mothers merely being indecent in refusing to let the fetus use her body show
that whether the mother invited the fetus in isnt decisive in determining the
permissibility of abortion. The mothers situation is compared to that of the
older brother who owns a box of chocolates but does not share them with his
younger brother. The older brother is indecent, but not unjust, in not sharing
them, since the younger brother has no right to them. However, the older
brother should, out of a sense of decency, share his chocolates because doing
so doesnt present much of a hardship. Prior ownership makes it the case that
the older brother has a right to the chocolates and that the mother has a right
to bodily autonomy, but the degree of burdensomeness determines whether
the brother is indecent in not sharing the chocolates and whether the mother
is indecent in not letting the fetus use her body. In the case at hand, the length
of time of pregnancy and the fact that this pregnancy doesnt pose a health risk
to the mother make it indecent, though not unjust, for the mother not to carry
the pregnancy toterm.
But I think that when we examine what at least certain violations of the
right to bodily autonomy involve, we will see that appeal to this right is suffi-
cient to override even the indecency charge. Thomson believes that since the
fetus is a person and pregnancy in this case poses no health risk, the mother
should (in the sense that shed be indecent not to) carry it to term, since
one hour of her time is no big deal when another persons life is at stake. Of
course, as Thomson admits, conservatives wrongly extend this logic to normal
29
See Judith Jarvis Thomson, Trespass and First Property, in The Realm of Rights (Cambridge,
MA:Harvard University Press, 1990), 205226.
30
31
pregnanciesthey say that its someones life, its no big deal to be pregnant
for nine months, most women are fine during pregnancy, so you ought to do
it. But consider what Thomson says in her Henry Fonda example, that if all
Ineed to save my life is for Henry Fonda to touch his cool hand on my fevered
brow, Ido not have a right against him that he fly in from the West Coast to
do this for me, though it would be frightfully nice of him to do so. But if all
he has to do is walk across the room and touch my fevered brow, then it would
be indecent of him to refuse.32 Why would the mothers refusing to carry a
fetus to term, even for an hour, when it is the result of rape, be self-centered,
callous, and indecent, when Henry Fondas refusing to hop on a plane and
cross the country to touch my fevered brow not be indecentin fact, it would
be frightfully nice of him to do so? Surely carrying a fetus that is the result
of rape even for one hour is much more onerous than flying across country.
As Margaret Little perceptively argues, forced gestation is an evil of unwanted
occupation, or an invasion of the self, which is neither reducible to the evil of
medical risks of pregnancy nor merely a different way of talking about the evil
of interference with the mothers plans.33 Pregnancy, whether or not it is the
result of rape, involves a sharing of ones very body, heart, and soul, whose sac-
rifices are measured not in degrees of risk but in degrees of intertwinement.34
No such intimacy is involved in the Henry Fonda case, and forced intimacy
is obviously not something it is indecent to reject. My first principle about
bodily autonomy is this:You shouldnt be required to do anything with your
body when the prior act that put you in the position you are in was against
your will. Iwant to support this principle with a Kantian analysis of respect
and degradation.
As is familiar, Kant believed that all rational, autonomous beings possess
dignity and are deserving of respect, which for Kant means that they are to
be treated as ends in themselves and never merely as means to an end.35 Our
rationality, or better, our capacity for rationality, is what makes us persons and
distinguishes us from inanimate objects and nonhuman animals. Our capacity
for rationality is marked by the fact that we have the ability to make plans
and have goals, interests, and reflective desires. We ought to respect these in
other rational beings by, for example, not putting our own interests and re-
flective desires ahead of those of others (e.g., by not aiding the needy when
we are able to), or discounting the plans of others (e.g., by falsely promising
to repay borrowed money and not letting the lender decide whether to give
32
33
Margaret Olivia Little, Abortion, Intimacy, and the Duty to Gestate, Ethical Theory and Moral
Practice 2 (1999):304.
34
Little, Abortion, Intimacy, and the Duty to Gestate, 305. Little rightly notes that her view is
noticeably absent from the philosophical literature and urges the development of an ethic of intimacy.
35
Immanuel Kant, Grounding for the Metaphysics of Morals, translated by James W. Ellington
(Indianapolis:Hackett, 1981 [1785]), 36, AKA 428429.
it to you in the first place), and the like. In virtue of our humanity, we ought
to respect each person as a potential co-legislator of morality and engage in
conduct only on which we could expect all reasonable people to agree. To do
otherwise is to make an exception for yourself by treating another rational
being as a mere means to your ends, behavior to which another would not au-
tonomously consent.
Lets apply this analysis first to rape and then to enforced pregnancy.
Acommon feminist analysis of rape explains rape as an act in which the rapist
attempts to degrade his victim by considering her to be an inferior being,
merely an object to be used for mens sexual pleasure or control.36 According
to Jean Hampton, rape conveys the attitude that women are even lower than
chattel; they are mere objects that are there to be used whenever the male feels
the need to do so but not beings with the capacity for rationality who ought
to be respected as ends in themselves. Hampton classifies rape as a moral in-
jury, which she defines as damage to the realization of a victims value through
behavior that diminishes, or, attempts to lower, the victims value.37 She agrees
with Kant that a persons intrinsic value, the value she has as a person, cannot
actually be lowered or degraded or for that matter raisedwe are equal in
virtue of our humanity, as ends in ourselves.38 Rape is morally injurious not
just to its direct victim but to all women. This is because rape sends the mes-
sage to all women, that they are the kind of human beings who are subject to
the mastery of people of the rapists kind, namely, men.39 Rape thus diminishes
all womensvalue.
One way to parse this analysis of rape is along the lines of an individual
account of autonomy. Kant ties respect for persons to their capacity for ra-
tionality and not anything bodily, but the body might factor in in terms of a
persons having interests or reflective desires (e.g., not mere bodily appetites,
but desires about whether to eat or drink and for which reason) relating to her
body that are indicative of the capacity for rationality. Because rape is an act
against a persons will, it disrespects her desires or interests having to do with
36
See Jean Hampton, Defining Wrong and Defining Rape, in A Most Detestable Crime, edited by
Keith Burgess-Jackson (NewYork:Oxford University Press, 1999), 118156; Susan Griffin, Rape:The
All-American Crime, in Feminism and Philosophy, edited by Mary Vetterling-Braggin, Frederick A.
Elliston, and Jane English (Totowa, NJ:Littlefield, Adams, 1981), 313332. Griffin says that rape is an
act of aggression in which the victim is denied her self-determination and is a form of mass terrorism
because the victims are chosen indiscriminately. See also Catharine MacKinnon, Toward a Feminist
Theory of the State (Cambridge, MA:Harvard University Press, 1989), esp. chap.9; MacKinnon, Desire
and Power, in Feminism Unmodified: Discourses on Life and Law, edited by Catharine MacKinnon
(Cambridge, MA: Harvard University Press, 1987). MacKinnon says that men have been knowers;
mind; women have been to-be-known, matter, that which is to be controlled and subdued, the acted
upon(55).
37
Hampton, Defining Wrong and Defining Rape,132.
38
39
how her body is treated, specifically, her desires and interests about whether to
have sex, with whom, and under which conditions.
Alternatively, we might parse this account of rape along the lines of a rela-
tional autonomy and make the Kantian case stronger, for it is not just interests
you have regarding your body that are at stake, but bodily integrity. Iwant to
suggest that all of your interests are interests of you as a personyour interests
attach to youand your person is housed in your body.40 This is why you have
bodily integrity. Rape is not just an act of ignoring or discounting or thwarting
your interests about your body; rather, it is a violation of your very person,
through your body, and as such it is a violation of your bodily integrity. This
analysis explains in a better way than the earlier one why rape is an act against
your will, specifically, using your body against your will. Rape is not just an act
that goes against an interest you have regarding how your body is treated, but
it is about how your whole person is treated.41 It sends a message not just that
a certain interest you have does not count but that you are the kind of being
whose interestsany and all of themdo not count. This analysis allows for a
deeper sense of the account of rape being against yourwill.
Rape is a violation of the right to bodily autonomy in a way that attempts to
degrade its victim. It seems that if rape attempts to turn a woman into an in-
ferior beingnothing more than a body or body part or sex object to be used
for mens pleasure or controlthen a womans right to bodily autonomy, if it
means anything, allows her to affirm, through the right to abortion, that she is
not such a being. It allows her to affirm what someone has attempted to take
away from her. When someone uses your body against your will, he treats you
as not being a member of humanity. Enforced pregnancy, some feminists have
argued, treats women as if they were mere fetal containers, not full-fledged
members of humanity.42 If the woman does not want to carry to term a fetus
resulting from rape, then she should not be required or deemed indecent if
she does not subject her body to this intrusion against her will. By being free
from morally bad judgment when she exercises her right to bodily autonomy
through abortion, she is able to assert her bodily integrity, herself as a whole
person, at base, her humanity. Whats more, since rape sends a message of deg-
radation to all women, when a woman asserts her humanity in this way, she is
40
This account was inspired by Helga Vardens explanation of why, for Kant, from the point of
view of justice we have to consider the relation between my person and my body as analytic, as a nec-
essary unity. See Helga Varden, A Feminist, Kantian Conception of the Right to Bodily Integrity:The
Cases of Abortion and Homosexuality, in Out from the Shadows:Analytical Feminist Contributions
to Traditional Philosophy, edited by Sharon L. Crasnow and Anita M. Superson (NewYork:Oxford
41
This analysis might reflect Catriona Mackenzies account of relational autonomy.
42
See Laura M. Purdy, Are Pregnant Women Fetal Containers? in Reproducing Persons:Issues
in Feminist Bioethics, edited by Laura M. Purdy (Ithaca, NY:Cornell University Press, 1996), 88105.
Purdys paper is mainly about whether and the extent to which women have duties to their fetuses, but
presumably she would argue along similar lines about enforced pregnancy.
at the same time attempting to shrug off the degrading stereotypes associated
with women:that they are nothing more than sex objects or mere bodies for
mens use. The time that pregnancy lasts has no relevance to this argument:if
someone uses your body against your will in ways that attempt to reduce your
humanity, morality should allow you to reassert your humanity, to in effect
send the message that neither you nor members of your kind are the kind of
beings that can be treated this way.43 This is true whether pregnancy lasts nine
months or one hour. Thus, there is nothing indecent or callous about a woman
who seeks to abort a fetus that reminds her of the rapists attempt to degrade
her. At best, the woman would be frightfully nice to carry the fetus toterm.
3. Case 2:Late-Term Abortion for Morally Questionable Reasons
Regarding case 2 of the woman who, in her seventh month of pregnancy,

wants an abortion to avoid postponing a trip, Thomson says, ...while Ido
argue that abortion is not impermissible, Ido not argue that it is always per-
missible. There may well be cases in which carrying the child to term requires
only Minimally Decent Samaritanism of the mother, and this is a standard we
must not fall below...It would be indecent in the woman to request an abor-
tion, and indecent in a doctor to perform it, if she is in her seventh month, and
wants the abortion just to avoid the nuisance of postponing a trip abroad.44
Again, Thomson is not saying that the fetus in this case has a right to use the
mothers body and that the mother has a corresponding obligation to allow it
to do so, so in some sense Thomson still protects the right to bodily autonomy.
Yet this is another case where Thomson levels the charge of moral indecency
on the woman who has an abortion in these circumstances, which supports
her belief that abortion is not morally permissible in this case. The woman
thus ought not to abort, but the ought is grounded not in the fetus right
to use the mothers body but in the womans being indecent not to let it use
herbody.
In her paper, Thomson seems to set up a hierarchy of moral acts. At the top
are the most morally onerous acts, ones in which a great deal of sacrifice is re-
quired of the agent. Thomson mentions Splendid Samaritanism, in which the
sacrifice is onerous, and Very Good Samaritanism, in which the sacrifice is
very large, in connection with this class of acts.45 Good Samaritanism, ranking
just below Very Good Samaritanism, also involves large sacrifices, where the
43
We might even argue that you have an obligation, out of self-respect, to reassert your humanity.
For this kind of view, see Thomas E. Hill, Jr., Servility and Self-Respect, in Dignity, Character, and
Self-Respect, edited by Robin S. Dillon (NewYork:Routledge, 1995), 7692 (reprinted from Monist 57
(1973):87104).
44
45
Thomson, Defense of Abortion, 65,64.
agent goes out of his way, at some cost to himself, to help one in need of it.46 As
an example, Thomson cites the Kitty Genovese case, where Genovese was mur-
dered while twenty-eight people watched or listened without helping hera
Good Samaritan, Thomson notes. Then she immediately corrects herself and
says that a Splendid Samaritan would have rushed out and assisted Genovese
against the murderer.47 Thomsons shift to what a Splendid Samaritan would
have done is explained by the size of the sacrifice involved, which in this case
is a risk of death for yourself.48
For any act involving Good Samaritanism or better, there is no require-
ment that the agent do it. Thomson notes that Henry Fonda would be a Good
Samaritan, performing an act of kindness but not anything required that you
can claim against him as your due, were he to fly in from the West Coast to
touch your fevered brow.49 She says the same about the violinist who needs to
stay hooked up to your body to use your kidneys for his continued life:it is a
kindness on your part, and not something you owe him.50
Next on the hierarchy come acts that are indecent not to perform yet do not
correspond to any rights violation and thus are not owed by you to another
agent. These are acts of Minimally Decent Samaritanism. Iunderstand moral
requirements to rank just below acts of Minimally Decent Samaritanism in
the hierarchy. Iwill address moral requirements first, since Thomson draws
out their significance by comparing them with acts of Minimally Decent
Samaritanism. Morally required acts, unlike the others so far mentioned, cor-
respond to rights. One example of a morally required act comes in with the
case of the two brothers who are jointly given a box of chocolates, but the
older brother takes them all and refuses to give any to the younger brother. The
younger has a right to them, so the older one is unjust and violates an obliga-
tion in not giving his brother his fair share.51 Sharing the chocolates is morally
required because the younger brother had a right to the chocolates. In con-
trast, you are not unjust if you unplug yourself from the violinist because he
had no right against you to use of your body,52 even if unplugging him means
that he dies. Unless you gave him a right to use your body, you are not morally
required to let him use it. Similarly, the woman who becomes pregnant from
rape or due to contraceptive failure does not have an obligation to carry the
fetus to term because she did not invite it in or give it a right to use her body,
and it has no right against her for use of her body. Obligations corresponding
46
47
48
49
50
51
52
with rights fall below acts that are indecent in you not to performyou ought to
perform the former; morality requiresit.
Now consider Minimally Decent Samaritanism, acts that it should be morally
indecent not to perform. Carrying a one-hour pregnancy to term, sharing your
chocolates with your brother when they were given only to you and he wants
some, Henry Fondas walking across the room to touch your fevered brow, let-
ting the violinist use your kidneys for one hour, and not requesting an abortion
in the seventh month to avoid postponing a trip are cases of Minimally Decent
Samaritanism.53 Thomson says that you ought to perform such acts in the sense
that it would be indecent to refuse.54 Thomson distinguishes these cases, and this
sense of ought, from cases of moral obligations, which correspond with rights.
She argues against those who believe that the ought of minimal decency gener-
ates a right:...it seems to me to be an unfortunate loosening of what we would
do better to keep a tight rein on.55 Rights are generated differently, independent
of the oughts of minimal decency, but on the basis of someones having a clear
entitlement to something (e.g., ownership of ones body or a box of chocolates).56
The ought of obligation follows from this, whereas the ought of indecency
follows from how easy it is to provide someone with something.57 Thomson notes
that, except where there is a right to demand it, nobody is morally required to
make large sacrifices of health and other interests and concerns, even to keep an-
other person alive.58 Still, oughts of minimal decency are oughtswe cannot
claim that the fetus who comes into existence from rape has a right to use the
mothers body even for a one-hour pregnancy but only that the mother ought
to let it use her body for this short time because she would be self-centered, cal-
lous, indecent to refuse. Thomson notes that the complaints are not less grave;
they are just different.59 That is, they are both ways you ought to act, but the
ought is grounded in a requirement in one case, and in minimal decency, in
the other case. The ought of minimal decency is binding in a way that makes
the act impermissible, but its impermissibility does not derive from its being a
moral requirement or duty. Although Thomson is not explicit about this, we
can discern from the cases that Minimally Decent Samaritanism is in the class
of acts of Samaritanism whose degree of goodness depends on the amount of
sacrifice involved. Obligations corresponding with rights do not:indeed its a
rather shocking idea that anyones rights should fade away and disappear as it gets
harder and harder to accord them to him.60 But since Minimal Moral Decency is
53
Thomson, Defense of Abortion, 60, 61, 5960,6566.
54
55
56
57
58
59
60
farther down on the scale from the supererogatoriness of Splendid Samaritanism

but closer to the moral requiredness of obligations that correspond with rights,
Thomson assigns it the force of an ought.
Return to Thomsons case of the woman who wants the late-term abor-
tion to avoid postponing a trip. What drives Thomsons view that Minimally
Decent Samaritanism means that the woman ought to carry the pregnancy
to term is the amount of sacrifice involved in doing so.61 Thomson must be
weighing the sacrifice in postponing a trip against the value of the fetus life.
But Ithink the more apt comparison is the sacrifice in carrying the pregnancy
to term versus the fetus life. And here I am not convinced that Minimally
Decent Samaritanism demands that the woman carry the pregnancy to
term. Pregnancy is a risky undertaking. Tribe notes that as many as 30per-
cent of pregnant women have major medical complications and 60 percent
have some kind of medical complication.62 The Centers for Disease Control
and Prevention (CDC) lists some of the common complications of preg-
nancy as anemia, urinary tract infections, mental health conditions (e.g., sad
mood, problems concentrating, feelings of worthlessness), hypertension, pre
eclampsia (which occurs after the twentieth week of pregnancy and causes
high blood pressure, problems with the kidneys and other organs, swelling of
the hands and face, too much protein in the urine, stomach pain, blurred vi-
sion, dizziness, headaches, and seizures and whose only cure is delivery), ges-
tational diabetes mellitus (which increases the risk of preeclampsia and having
a big baby, which can complicate birth), obesity, infections, and hyperemen-
esis gravidarum (nausea and vomiting).63 Amy Mullin, in her philosophical
book on pregnancy, notes that vomiting two to three times a day in the first
four months is considered normal, as is extreme fatigue.64 It has been reported
that 70 to 80percent of pregnant women have some type of morning sickness
and that although 60,000 cases of hyperemenesis gravidarum have been re-
ported by hospitals, the numbers are much higher because many women are
treated at home.65 There is no prevention of this condition, and it causes se-
vere dehydration, inability to keep down any food, extreme fatigue, headaches,
confusion, fainting, low blood pressure, rapid heart rate, and secondary anx-
iety or depression. The CDC reports that in the United States 650 women die
each year from pregnancy or delivery complications.66
61
Ido not take issue with Thomsons criterion for determining degrees of Samaritanhood.
62
Tribe, Abortion,103.
63
Centers for Disease Control and Prevention (CDC), Pregnancy Complications, February 5,
2013, http://www.cdc.gov/reproductivehealth/MaternalInfantHealth/PregComplications.htm.
64
Amy Mullin, Reconceiving Pregnancy and Childcare:Ethics, Experience, and Reproductive Labor
(NewYork:Cambridge University Press, 2005),61.
65
American Pregnancy Association (APA), Pregnancy Complications, 2013, http://american-
pregnancy.org/pregnancycomplications.
66
CDC, Pregnancy Complications.
In addition, Mullin notes that pregnancy exacerbates ongoing illnesses,

and most health-care providers are unsure of how it affects such disabilities.67
Some of the dozen or so preexisting health conditions listed by the CDC
that can affect pregnancy include asthma (increases the risk of preeclampsia,
Caesarean birth, and other complications), high blood pressure (causes pre-
eclampsia, placental abruption, and preterm birth), and thyroid disease
(causes heart failure).68 Aside from these serious physical and psychological
complications of pregnancy, Ann Cudd notes other harms, and jointly these
dangers push pregnancy to the status of at least Very Good Samaritanism on
Thomsonsscale:
Pregnancy is a dangerous and onerous task for a person. She may feel that
she loses bodily integrity, freedom of movement, suffers physical pain and
discomfort, and risks serious illness or death. If she is relatively wealthy, she
must either submit herself to the frequent, often intrusive examinations by
physicians and their restrictive instructions, or risk an alternative, less so-
cially acceptable, kind of care. And if she is poor then she may not be able
to afford the necessary care to lessen the risk of pregnancy. In any case she
is likely to be discriminated against in employment and education, find it
more difficult to be taken seriously, and be given unsolicited advice from
all directions. All this suggests outstanding heroism but pregnant women
are not accorded heroic status because it is expected of them, and what they
really want anyway.69
Thus, factually speaking, I conjecture that Thomson underestimates the
level of Samaritanhood in pregnancy. Since the sacrifices necessary to carry
the pregnancy to term are great, the mother who would rather not do so,
to take a trip or whatever, should not be deemed indecent if she opts to abort
she is nowhere in the league of Henry Fonda who has to walk across the room
to touch your fevered brow.70 For Thomson to say otherwise goes against her
view that Minimally Decent Samaritanism does not require a lot of sacrifice.
I want to make a more general point about Minimally Decent Samaritanism.
It has been objected that the oughts of decency should be even stronger than
Thomson argues for. Peter Singer argues convincingly that we have an obliga-
tion to help the needy when we are able to do so without suffering any com-
parable harm to ourselves.71 If we agree with Singer, the demands of morality
67
Mullin, Reconceiving Pregnancy,60.
68
CDC, Pregnancy Complications.
69
Ann Cudd, Enforced Pregnancy, Rape, and the Image of Woman, Philosophical Studies 60:12
(SeptemberOctober 1990):4759,53.
70
To be clear, Iam not saying that the indecency charge is overridden by competing moral claims
but that it should not be leveled to beginwith.
71
Peter Singer, Famine, Affluence, and Morality, Philosophy & Public Affairs 1:3 (1972):229243.
Singers principle is that If it is in our power to prevent something bad from happening, without
are even more stringent than Thomson sets out. They would require both the
woman in the seventh month of pregnancy who wants to take a trip and the
rape victim whose pregnancy lasts only one hour to carry their pregnancies to
term as well as other pregnancies when the mother does not suffer harm com-
parable to the fetus death.72
Although many of us find Singers general principle plausible, Ibelieve that
when it comes to using our bodies to carry out the prevention of something
bad, the demands of morality are less stringent than this principle sets out.
And the more intimately our bodies must be used, the less stringent the moral
demand. Of course, pregnancy is exactly the kind of case where a womans
body is used most intimately in carrying out any moral demand made by the
fetus situation. As Margaret Little puts it, there are significant qualitative dif-
ferences between gestating and giving money to Oxfam: There are special
facets to a decision about charity when the beneficence is a matter of sharing
ones body, heart, and soul, not just ones pocketbook or general energies, when
the sacrifice is measured, not in degrees of risk, but in degrees of intertwine-
ment.73 For this reason, pregnancy is more sacrificial than being hooked up to
a violinist, which is more sacrificial than donating blood or getting vaccinated
against contagious diseases.74 Any of these is more sacrificial than donating
money to Oxfam, simply because of the degree of bodily involvement.
Why does use of the body matter so much to the demands of morality?
Thomson makes several remarks that speak to its significance:
No doubt the mother has a right to decide what shall happen in and to her
body; everyone would grant that.75
If anything in the world is true, it is that you do not commit murder, you
do not do what is impermissible, if you reach around to your back and un-
plug yourself from that violinist to save your life.76
...the mother owns the house.77
Women have said again and again This is my body.78
My own view is that if a human body has any just, prior claim to anything
at all, he has a just, prior claim to his own body.79
For nobody has any right to use your kidneys unless you give him such
a right...80
thereby sacrificing anything of comparable moral importance, we ought, morally, to do it (231). Iowe
this example and the objection it makes to my view to Samantha Brennan.
72
Ido not know if Singer would say this because he does not address this in hispaper.
73
Little, Abortion, Intimacy, and the Duty to Gestate,305.
74
Samantha Brennan and Heather Douglas provided the examples.
75
76
77
78
79
80
Because you own your own body, have a just prior claim to it, and are the
only one who can give some a right to use it, the right to bodily autonomy
has high trumping power. Again, Ibelieve that the significance of the body
lies with the fact that you are your body:it is the vehicle through which you
live your life.81 Susan Brison cites Jean Amry, who claims that the boundar-
ies of my body are also the boundaries of my self.82 Both Brison and Laurence
Thomas, in their work on bodily trauma experienced in rape or child abuse,
further explain this insight. According to Thomas, a rape victim suffers the
psychological scar of having a profound sense of diminished personal agency.83
The sense of personal agency, which Thomas believes is acquired under favor-
able conditions in ones adult familial environment rather than at birth, is
absolutely central to being an adult, and at least partially captures the force of
my in my body.84 It is the sense in which one has a great deal of control over
things that happen to ones body in ones social interactions:To have a sense
of personal agency is to have the conviction that, well beyond cases of sheer
bodily harm, there are things which a person ought not, and so will not, do to
or with ones body, or observe ones doing with ones body, without ones con-
sent.85 According to Thomas, rape, childhood sexual abuse, and other similar
bodily traumas shake the victims sense of personal agency at its core in such
a way that the victims beliefs about her self become skewed, causing her to
suffer a radical diminution of her self.86 Although Thomas denies that the pain
of rape is conceptually related to having suffered a bodily injury, the reality
is that the diminution of the sense of self is carried out through an attack on
onesbody.
I would add to Thomass account that having a sense of personal agency
is to see oneself as a valid member of a moral community, a being who is
owed a certain treatment including respect for her body. Along these lines,
Brison argues that recovering from bodily trauma experienced in rape is a
matter of being reconnected with humanity in ways that you value.87 Physical
trauma inflicted by another can change ones perception of ones own body, by
81
Thomson speaks of owning your own body, but I think you are your body. She might agree,
explaining that the fact that you are your body underlies your ownership of it. Ileave aside the debate
between being your own body and owning it. For a discussion of ownership, see Jennifer Church,
Ownership and the Body, in Feminists Rethink the Self, edited by Diana Tietjens Meyers (Boulder,
CO:Westview Press, 1997), 85103.
82
Susan Brison, Outliving Oneself:Trauma, Memory, and Personal Identity, in Meyers, Feminists
Rethink the Self,18.
83
Laurence Thomas, The Grip of Immorality:Child Abuse and Moral Failure, in Reason, Ethics,
and Society: Themes from Kurt Baier, edited by J. B. Schneewind (Chicago, IL: Open Court, 1996),
144167.
84
Thomas, Grip of Immorality,152.
85
86
87
Brison, Outliving Oneself,2829.
reducing the victim to flesh or a mere object,88 causing her to perceive her body as
an enemy or a site of increased vulnerability,89 to disassociate from her body or to
have an intense awareness of her embodiment,90 to experience traumatic memo-
ries through her body (e.g., racing heart, crawling skin, or being immobilized),91
and to distance ones self from ones bodily self.92 Victims of bodily trauma need
to remake their selves, to reconnect their self that existed prior to trauma with
the self they come to be through their trauma. This cannot be accomplished in
isolation from others but only in a community of trustworthy persons, particu-
larly those who are also survivors of trauma and who will listen to their story and
engage in mutual empathy with them.93 Through this process, the victim finds
some aspects of her lost self in another person and is able to reconnect with it and
remake her self as one, surviving as an autonomous self.94 This process confirms
the victims humanity.95
I conclude from this discussion that the body is the vehicle through which you
live your life, it plays a significant role in the identity of the self, and it is integrally
bound up with having a sense of personal agency or being a member of humanity
or the moral community. These factors give backbone to Thomsons belief that
the body is so important that only you can give someone else a right to use your
body and in turn to why the demands of morality are lessened accordingly with
the degree of involvement of the body required to carry them out. Ioffer the fol-
lowing principle:
The stringency of the demands of morality are dependent on the degree of the
involvement of the body in carrying them out; an act that is otherwise a duty
can be an act of at least Good Samaritanism commensurate with the degree of
bodily involvement necessary to fulfillit.
Lets return to the indecency charge mounted against the woman who wants
a late-term abortion to avoid postponing a trip, since this still needs defense.
Perhaps it is not just the amount of sacrifice but also a kind of commitment the
mother makes to the fetus that grounds Thomsons indecency charge.96
88
89
90
91
92
93
94
Brison, Outliving Oneself, 30. Since regaining autonomy involves dependence on others,
Brison calls hers a relational account of autonomy.
95
96
Whether this is Thomsons view is unclear, and she certainly does not state it as such. The un-
derlying issue is under what conditions the mother gives the fetus a right to use her body. We know
that in cases of rape and contraceptive failure the mother does not give the fetus this right. We might
infer, then, that when sex is consensual and when the woman (or couple?) does not take contraceptive
measures because she intends to get pregnant, then if she does become pregnant, she gives the fetus a
right to use her body. But the case at issue (which does not involve rape or contraceptive failure) is one
I think that it is odd to say that a person ever makes a commitmentan

unbreakable commitmentto carry through with a bodily process, especially
when it involves an intimate use of ones body. Suppose you agreed with your
dentist to have a root canal or with your doctor to undergo chemotherapy, but
during the course of the procedure you could no longer deal with the anxiety.
Surely if you have a right to bodily autonomy, this right entails that you get to
stop a medical procedure on your body at any point, even if doing so is not
medically or practically wise. The maximal choice approach to health-care de-
cision making, the predominant model in the medical ethics literature, would
insist on it, as it endorses a right to refuse treatment, to die with dignity, to
refuse experimentation, not to be sterilized, and even to commit suicide.
But maybe the maximal choice view does not apply when it comes to preg-
nancy because another life is at stake. Janet Gallagher discusses a number of
cases in which fetal rights advocates use the viability line or cutoff point em-
ployed by the US Supreme Court in Roe v.Wade as part of its argument to show
that once a woman decides to forgo abortion and the state chooses to protect
the fetus the woman loses the liberty to act in ways that would adversely affect
the fetus.97 Gallagher cites legal cases in which pregnant women were forced
to undergo Caesarean sections despite their refusal or were detained or incar-
cerated to prevent harm to the fetus.98 The assumption is that once a woman
carries the fetus to the point of viability, she is committed or has contracted to
carry it to term and not engage in behavior that might harm it, even when this
entails severe restrictions on her liberty or bodily autonomy.
One wonders whether it is the fact that another life is at stake that makes
the contract model override the maximal choice model or whether the max-
imal choice model really does not hold for womens bodies. It used to be the
case that a womans consent to marriage meant that she consented to sex with
her husband at his discretion throughout the marriage. Indeed, there was
a legal statute, which was finally repealed in all US states and Washington,
D.C., since 1993 but still infrequently prosecuted, that denied the existence
of rape of a wife by a husband on these very grounds.99 Some people believe
that women should not behave sexually unless they are prepared to carry
through to intercourse, as if the womans sexual behavior commits her to
of mere indecency, not one where the fetus can claim a right against the mother, so there is something
else besides consent and intention that gives the fetus a right to use the mothersbody.
97
Janet Gallagher, Prenatal Invasions and Interventions: Whats Wrong with Fetal Rights,
Harvard Womens Law Journal 10 (1087): 31, quoting John Robertson, Procreative Liberty and the
Control of Conception, Pregnancy, and Childbirth, Virginia Law Review 405 (1983):437.
98
Gallagher, Prenatal Invasions, esp.4648,910.
99
See Sara Ann Ketchum, Liberalism and Marriage Law, 264276. In July 1998, the National
Clearinghouse on Marital and Date Rape reported that in seventeen states there are no exemptions
from rape prosecution granted to husbands under the law. But thirty-three states still have some
exemptions from prosecuting husbands for rape usually with regard to the use of force, http://ncmdr.
org/state_law_chart.html.
carry through in a way defined by her assailant in date rape.100 The idea is
that the woman forms a contract or makes an agreement through her be-
havior and that the contractor is entitled to demand that she uphold it. Lois
Pineau argues against this view, noting that casual, nonverbal behavior, such
as eye contact, smiling, and blushing are imprecise and ambiguous and can
be misinterpreted and thus do not constitute a contract.101 Such a contract,
according to Pineau, is grounded in the myth that men cannot turn off their
sexual arousal but women can, so it is womens responsibility not to provoke
the irrational in men.102 Men can rightly expect the contract to be fulfilled
by using their natural aggression to see to it that it does. Pineau contrasts
the sexual contract with legal contracts that are normally upheld only if the
contractors are clear on what they agreed to and have enough time to think
about whether this is what they want, the terms are usually written out or the
expectations are well known due to tradition and are enforced only by the
law, not private individuals. Consider, for example, the contracts we make
when buying a car or ahouse.
Pineau is making two points here. First is an epistemological point, that
behavioral cues should not constitute the basis of any contractnodding your
head doesnt mean you contract to buy this house, kissing doesnt mean you
contract to sex, and carrying a fetus for seven months doesnt mean you con-
tract to carry it to term. Second, a moral point, is that Pineau doubts whether
sexual contractsand, Iwould add, any bodily contractscan even be made
because behavior in fact does not entail a commitment but believes that even if
they could, the terms would not be enforceable.103 The right to bodily autonomy
should protect against any such contract involving ones body. Pineau dislikes
the contract model because it requires a strong act of refusal to overcome the
presumption of consent, which essentially allows another person to decide
what happens in and to your body. Pineau favors a communicative model that
is based on the notion that there are noncommunicative sexual encounters
that women would not find reasonable to consent to. Pineaus model puts the
burden on the man to show that he got the womans consent when the sex
was of the kind that it would not be reasonable for women to consent to. It
presumes the right to bodily autonomy because the man would have to get the
womans consent to further stages of foreplay and to intercourse by checking
with her throughout rather than assuming that her consent to some sexual be-
havior entails a commitment or contract to have intercourse.
100
Lois Pineau, Date Rape:AFeminist Analysis, Law & Philosophy 8 (1989):217243.
101
Pineau, Date Rape,229.
102
103
We can apply Pineaus objection to the contract model of sexthat it

requires a strong act of refusal to overcome the presumption of consentto
the case of pregnancy. When it comes to sex, according to the contract model,
once the woman displays certain behavior like kissing or smiling, she has con-
tracted to intercourse, and the only thing that gets her out of the contract is a
strong act of refusal on her partthe presumption is that contracts should not
otherwise be broken. When it comes to pregnancy, according to the contract
model, once the woman has carried the fetus to a certain point she has com-
mitted herself to carrying it to term (or, the longer she carries it, the stronger
her commitment to carry it to term). The only thing that releases her from
her commitment is a good and strong reason for terminating the pregnancy.
The presumption is that once she becomes pregnant and does nothing to stop
the pregnancy, she has committed to carrying it to term absent a good reason
not to. Wanting to avoid postponing a trip is not a good reason for breaking
her commitment. There are other bad reasons, like negligence and malicious
deceit. George Harris discusses the fictitious case of Michelle and Steve, who
have been married five years, and he wants children but she does not and does
not tell him because confrontation is unpleasant.104 Michelle allows Steve to
believe that it is just a matter of time when they will have children but seeks
an abortion when she becomes pregnant. Another case from Harris is that of
Anne, the man hater who vents her rage about men against Mark, seduces
him, and makes him give up his business and house for a family with her. She
pretends at first to enjoy pregnancy but then has an abortion.
The first thing to be said is that these cases are probably the exception rather
than the rule. Also, these cases assume that women use sex and pregnancy as a
weapon, that they are evil, and that they have options about abortion. Even if
some women do use sex or pregnancy for these reasons, doing so is likely out
of desperation to have children to stay with a man or make a man want them,
and the like, for lack of control over their bodies and from desires deformed by
patriarchy. The reality is that women do not have the upper hand in abortion
because, though legal in the United States, it is not widely available, its legality
is constantly under attack, and it is still stigmatized.105
But more to the point about the right to bodily autonomy, the contract
model gives another person determination of what happens in and to your
own body. On the contract model of sex, the contract is in place absent the
womans strong act of refusal, which gives the man the right to continue onto
104
George W. Harris, Fathers and Fetuses, Ethics 96 (April 1986):594603.
105
It has been estimated that in 87percent of counties in the United States it is not possible to ob-
tain an abortion because it is heavily restricted or banned in publicly funded facilities or because coun-
seling for it is banned. Rachel Weiner, No Choice:87% of U.S. Counties Have No Access to Abortion
Clinic, Huffington Post, July 3, 2009, http://www.huffingtonpost.com/2009/06/02/no-choice-87
-of-us-counti_n_210194.html.
intercoursethe contract is the default position. On the contract model of

pregnancy, the contract to carry the fetus is presumed to be in place absent
the womans giving a good and strong reason for aborting the fetus. But this
gives a person other than her the right to determine what happens in and to
her body by giving that person the power to assess her reasons. Certainly the
right to bodily autonomy does not entail this. If a woman owns her body, and
her ownership of her body is prior to anyone elses use of it, she, not someone
else, should have the right to determine whether she consents to every stage
in a process involving it. The right to bodily autonomy should protect against
someone elses deciding on her behalf, which risks having her body used for
someone elsesends.
Note that all of this is consistent with admitting that the woman in the preg-
nancy case has bad reasons for wanting to terminate her pregnancy. She might
be morally indecent in virtue of her character, but Iam arguing that her act of
ending the pregnancy is not morally indecent in virtue of her violating some
kind of commitment she makes about the use of her body. The contract model
enforces pregnancy as punishment for a womans bad moral character or ir-
rationality and constitutes an illegitimate waiver of ones right to use of ones
body. The right to bodily autonomy disallows this, given its significance to
being a member of humanity.
In conclusion, Ioffer the following principle:
Consenting to early stages of a process involving (especially intimate uses
of) your body does not entail a commitment to the entire process; not con-
senting to the rest of the process, even if your reasons for not consenting are
bad, is not morally indecent.
This principle and the others offered previously are intended to support and
give substance to the right to bodily autonomy, which is a right that Ibelieve
is crucial to feminist positions on issues such as abortion, rape, obligations to
fetuses, female genital mutilation, and woman battering, to name just a few.
Much more needs to be said to articulate fully the boundaries of this right and
its trumping power over other moral considerations. Afull articulation and
defense of this right will give feminists purchase against the current backlash
against abortion rights.
Acknowledgments
For very useful comments, I thank the audiences at the following venues
where Ipresented earlier versions of this paper:Society of Value Inquiry at
the American Philosophical Association, Relational Autonomy Workshop
held at McGill University, Society for Analytical Feminism conference held
at Vanderbilt University, Southwestern Ontario Feminist Philosophers
Workshop, graduate seminar at the University of California at Davis, and phi-

losophy department at the University of Waterloo. I especially thank Mark
Piper and Andrea Veltman for their very detailed comments and Catriona
Mackenzie and Natalie Stoljar for inviting me to the Relational Autonomy
Workshop where Ibenefited from comments from top scholars working on
autonomy.
{15}
Choosing Death:Autonomy and Ableism

AnitaHo
In recent decades, the values of autonomy and well-being have been very in-
fluential in western bioethics. Discussions regarding autonomy often focus
on two moral dimensions: the internal capacity of the patient to form and
make decisions consistent with her values and goals; and the freedom from
external coercion and other undue interference in making such decisions. In
discussions of end-of-life care, many have argued that a commitment to re-
spect autonomy and to promote beneficence requires health-care professionals
(HCPs) to facilitate the peaceful or dignified death of competent patients who
find their terminal or disabling conditions intolerable. Since treatment refusal
and withdrawal of life support can often hasten a patients death but are le-
gally and ethically acceptable when requested by competent patients, the more
recent controversies focus on the question of allowing other forms of medi-
cally assisted death, such as physician-assisted suicide or active euthanasia, for
similarly competent patients. Legalizing such procedures, proponents argue,
would support or further patients rights to make decisions about their lives
and well-being in the most compassionate and benevolentways.
Informed by feminist accounts of relational autonomy, this paper will argue
that the individualist and minimalist conceptions of autonomy that are often
used to support a right to medically assisted death often neglect the broader
contextual factors contributing to peoples loss of hope and may thus be inad-
equate in ensuring that end-of-life decisions truly promote their agency and
well-being. In particular, this paper will examine how peoples sociopolitical
environment and interpersonal relationships affect or even frame the way they
experience their impairments and related sufferings. While many bioethicists
focus their concerns on third-party involvement in facilitating suicide, Iwill
argue that we need to look at the broader societal assumptions about life with
impairments and how these attitudes and our social structure may affect peo-
ples quality of life, their decision-making processes, and their desire to die.
In examining various forms of professional assistance that would likely lead
to the death of the person with impairment, this paper will explore how a
Autonomy and Ableism 327
society that has not yet overcome ableism should consider such socioenvi-
ronmental issues in determining peoples potential right to medically assisted
death. Even if lingering oppressive powers of an ableist society are not directly
coercive, they can affect peoples thoughts about their alternatives in such a
way that certain options such as living with mechanical or human assistance
are not considered as viable and other decisions about ending ones life must
be made. In adopting an individualist and minimalist approach to autonomy,
the common philosophical approach regarding medically assisted death may
be prematurely neglecting many complex factors that contribute to peoples
decision-making process. In critically evaluating how the ableist ideology may
impose various forms of oppressive influence that can restrict peoples ability
to reflect on their value system and choose their desired life-sustaining sup-
port and end-of-life treatments accordingly, this paper nonetheless cautions
against treating people with impairments as a separate class of people who are
categorically less capable to assess their situation.
1. The Centrality of Autonomy in Bioethics and End-of-Life

Decisions
In western moral and political philosophy, particularly since the Enlightenment,

there is an explicit acknowledgment of the importance of respecting the au-
tonomy of moral agents. Personal independence and control have been held
up as ideals, and conditions of dependency are often seen as misfortunes that
should best be avoided. Immanuel Kants principle of respect for autonomy
and John Stuart Mills principle of individualism, which have shaped the con-
temporary discussion of autonomy in bioethics and end-of-life care, focus on
the inherent capacity and rights of self-determining agents to make their own
decisions. For Kant, the capacity for rationally determining ones own ends
or destiny is the locus and origin of ones unique and unconditional value.1
Respect for autonomy requires that we recognize peoples right to personal
self-governance, that is, a right to be in control of themselves and to choose
their own way. Adult autonomous moral agents presumably have the epistemic
and moral privilege to determine their own good. Even if we may disagree
with their decisions or believe that they are making mistakes, whether in re-
lation to health care or other situations, we cannot coercively override their
self-regarding decisionswe can only advise or persuade them to reevaluate
their situation with more accurate information when they hold false views so
that they are better equipped to reconsider their decisions accordingly.2 To
1
Immanuel Kant, Foundations of the Metaphysics of Morals, translated by Lewis White Beck (Indi
anapolis:Bobbs-Merrill,1959).
2
John Stuart Mill, On Liberty (Peterborough:Broadview,1999).
coerce rational beings even for their own good is to paternalistically treat them
as if they lacked the capacity to shape their own livesit is to deny them their
moral status as persons.
The primacy of respect for autonomy in Western bioethics was initiated
as a response to research atrocities such as the Nuremberg experiments and
the Tuskegee syphilis study, particularly in relation to how the individuals in-
volved were subjected to harm without their knowledge, understanding, or
agreement. The ideas of self-determination and well-being have subsequently
been adopted in the literature on health-care delivery in the face of evolv-
ing individual/civil rights and rising consumer empowerment. Concerns have
been raised regarding medical paternalism, where HCPs withhold informa-
tion from or make value judgments and clinical decisions on behalf of patients
allegedly for the latters own good. The principle of respect for peoples au-
tonomy, particularly when expressed in terms of formal consent requirements
and privacy regulations, establishes parameters and expectations for HCPs in
their conduct with patients. As modern medical science and technology come
with risks, limitations, and errors, these legal mechanisms structure HCPpa-
tient relations and regulate what professionals can or cannot do to patients.
Protection of autonomy is considered particularly crucial in health-care
settings because illnesses and injuries are physically and emotionally chal-
lenging for many patients, especially when the diagnoses are unexpected or
grim or when the potential for adverse effect from medical treatment and
abuse of power is high. Such protection ostensibly helps to promote patients
agency by ensuring that they understand the risks and benefits of available
therapeutic options and have the freedom to choose between or refuse these
interventions according to their own values.3 As feminist bioethicist Susan
Sherwin points out, patients are often worried about their situation and are
ignorant of the particulars of various treatment alternatives, which gener-
ally make them dependent on the care and goodwill of others.4 HCPs, on
the other hand, are presumably more knowledgeable about their conditions,
and their professional recommendations often determine whether patients
would have (affordable) access to diagnostic and therapeutic procedures that
can provide further information, minimize pain, restore health, or improve
functioning and extend life. Given that HCPs are inadvertent gatekeepers of
information and resources, there is an inherent power hierarchy in the HCP
patient relationship, making patients particularly vulnerable to manipulation
or even coercion by their caregivers. While many conscientious HCPs would
3
Leigh Turner, Bioethics and End-of-Life Care in Multi-Ethnic Settings: Cultural Diversity in
Canada and the USA, Mortality 7 (2002):285301.
4
Susan Sherwin, A Relational Approach to Autonomy in health Care, in The Politics of Womens
Health:Exploring Agency and Autonomy, edited by Susan Sherwin (Philadelphia:Temple University
Press, 1998),1947.
traditionally treat patients according to the clinicians judgment, with the be-
nevolent assumption that patients who lack medical expertise would not know
what clinical alternative is best for them, it is now generally acknowledged in
Western bioethics that health-care decisions are not simply clinical decisions.
They have important implications on various aspects of the patients personal,
professional, social, and family life. This is especially the case for patients who
are terminally ill:that is, they are expected to die within six months from their
conditions. For some of these patients, aggressive interventions may sustain or
slightly prolong their life without restoring or maintaining their functioning
or quality of life. In some situations, the interventions may even aggravate the
patients pain and suffering. As more medical options are now available, each
with its own set of benefits and burdens for the patients and their loved ones,
it is increasingly difficult for HCPs who have limited contact with patients
under very specific circumstances to determine which available option is most
compatible with the latters value system and priorities. This is especially so in
diverse societies, where patients may have different cultural values and beliefs
regarding what causes illness, how it can be cured or treated, and who should
be involved in the process. Astrong principle of respect for patient autonomy
is thus necessary to counter medical paternalism, particularly toward those
who are most vulnerable or socially disadvantaged. Consent requirements for
treatment, advance directives, and hospital policies regarding resuscitation
and other aggressive interventions are formalized measures that allow patients
more control in health-care and end-of-life planning.
2. Minimalist and Value-Neutral Approaches to Autonomy in

Treatment DecisionMaking
One important aspect of the mainstream discussions and practices relating to

respecting patient autonomy is that they focus on individual patients making
specific decisions regarding their health care, that is, the making of autonomous
choice, or the actual governance itself.5 In contemporary health-care settings,
particularly in Western countries that take individual rights for granted, re-
spect for patient autonomy is often manifested in an individualistic manner by
obtaining informed consent, one treatment or procedure at a time. Attending
to separate and individual cases, respect for self-determination translates into
using a primarily cognitive approach to assess that particular persons im-
mediate capacity to make the relevant health-care decision,6 providing that
5
Tom Beauchamp and James Childress, Principles of Biomedical Ethics, 4th ed. (NewYork:Oxford
6
Catriona Mackenzie and Wendy Rogers, Autonomy, Vulnerability and Capacity:APhilosophical
Appraisal of the Mental Capacity Act, International Journal of Law in Context 9 (2013):3752.
patient relevant information regarding various available alternatives for his or

her particular condition and then allowing the patient (or a substitute deci-
sion maker) to make decisions among these options according to his or her
values, whatever those values may be. In fact, in most jurisdictions, a person
is presumed to have the capacity and authority to make health-care decisions
according to his or her own reasons and values, unless there are good reasons
to question and formally assess such a capacity. Neither a persons rejection
of the HCPs recommendation nor an unusual communication style proves
incapacity. Such value-neutral and procedural approaches refrain from assess-
ing the content of the motive behind the patients decisionthese approaches
consider a choice to be autonomous if a sufficiently capable patient, construed
as a normal chooser, has and understands the relevant information about
the available options and makes a reasonable and intentional choice without
coercion from others.7 The focuses here are on how HCPs may influence the
patients ability to make a particular decision and whether the patient tempo-
rarily fails to comprehend his or her situation because of illness or psycho-
logical affliction such as depression or severe anxiety in a particular moment.
According to a minimalist notion of autonomy, if the patient had no serious
cognitive or emotional impairment and was not subject to direct coercion by
others, the persons health-care decisions, including choices regarding with-
drawing life-sustaining interventions, should be considered autonomous.8
Others are obliged to respect such decisions, even if these patients may not be
exercising their autonomy wisely.9 The minimalist notion of autonomy seeks
to ensure that any potential power hierarchy that may exist between HCPs and
patients is not used unjustly to pressure or coerce the latter in their decision
making.
In the age of patient-centered care, this individualistic focus on each patients
decision-making process appears appropriate, given that patients in diverse
societies have varying values and priorities. This approach underlies many
mainstream discussions of end-of-life care, even though ones dying process
often has tremendous familial and social impact, both in emotional and eco-
nomic terms. Interestingly, such an approach also dominates discussions of
life-sustaining supports and interventions for people with impairments who
are not actively dying, particularly around continuing various interventions
that may be considered by the patient or others to be physically burdensome.
Most people are very concerned about pain and suffering that may accompany
7
Sherwin, A Relational Approach to Autonomy in Health Care, 26; Tom Beauchamp and James
Childress, Principles of Biomedical Ethics,123.
8
Uma Narayan, Minds of Their Own:Choices, Autonomy, Cultural Practices and Other Women,
in A Mind of Ones Own: Feminist Essays on Reason and Objectivity, edited by Louise Antony and
Charlotte Witt (Boulder, CO:Westview Press, 1993),429.
9
Mackenzie and Rogers, Autonomy, Vulnerability and Capacity,39.
certain impairments and degenerative conditions as well as the experience to-

ward the end of their lives. Many have a desire to retain dignity and control
during this journey, and some find the impairments and burdens in the last
stage of their lives at some point sufficiently great to make life no longer worth
living.10
Advances of modern medicinein association with its overwhelming ten-
dency to treat various conditions and symptoms aggressivelyhave engen-
dered increasing fear among many of losing control over their own dying
process. The right to forego extraordinary interventions or even requests for
medically assisted death have been seen as potential means for patients of di-
verse values, experiences, and beliefs to maintain or regain power over their
own life, their dying process, and their ultimate death.11 Some patients may
determine that their life is unduly burdensome or wrongful, that is, it is of
sufficiently poor quality that it is worse than no further life at all.12 Under the
legal and moral right to informed consent, patients are presumably the rightful
final decision makers to determine whether they want to start or continue
life-sustaining measures either through direct consent or refusal or advance
directives. These formal documents allow capable adults to give instructions
regarding their desired health care, particularly in terms of what types of inter-
ventions they would not want, in the event that they are unable to make such
decisions lateron.
In acknowledging peoples right to consider their wishes and well-being in
the health-care decision-making process, court cases and legislations in var-
ious countries in the last few decades have considered the right of people who
are terminally ill or have significant impairments not only to forego various
treatments but also to seek physician-assisted suicide or euthanasia. While
many people with impairments are not terminally ill, interestingly, several
jurisdictions around the world have considered their requests to die by var-
ious assisted methods similarly as end-of-life cases. Some philosophers argue
that the right to various forms of medically assisted death is a mere specifica-
tion of the moral right to self-determination.13 The Canadian examples of Sue
Rodriguez in the early 1990s and Gloria Taylor in 2012, the American case of
Dax Cowart in 1973, and discussions regarding Dr.Kevorkian from the late
1980s onward are often presented to demonstrate how it is sometimes mor-
ally permissible for HCPs to facilitate the death of competent patients who
10
Dan Brock, Life and Death:Philosophical Essays in Biomedical Ethics (Cambridge:Cambridge
11
Tania Salem, Physician-Assisted Suicide: Promoting AutonomyOr Medicalizing Suicide?
Hastings Center Report 29 (1999):3036.
12
Brock, Life and Death,280.
13
Ronald Dworkin, Thomas Nagel, Robert Nozick, John Rawls, Tim Scanlon, and Judith Jarvis
Thompson, The Philosophers Brief, NewYork Review of Books 27 (1997):4147.
consider their lives to be not worth living due to the suffering brought on by
their impairments or terminal illnesses.
The autonomy argument for medically assisted death generally appeals
to patients rights to refuse any procedure that counters their value sys-
tem and priorities, even if that may result in the patients death. Refusal of
life-sustaining interventions, especially in situations where a person is termi-
nally ill, is now generally considered noncontroversial and is widely accepted
as a fundamental component of patients rights to informed refusal. Taking
such a position for granted, many now extend the autonomy argument to
challenge the alleged moral distinction between actively killing a terminally
ill or disabled patient and letting such an individual forego life-sustaining
procedures, both of which can foreseeably lead to imminent death.14 While
withdrawing or refusing these interventions may facilitate a swift and pain-
less death in some cases, it may not offer comfort for others with degenera-
tive conditions who would have to endure a slowly deteriorating and dying
process. Some argue that, when death is accepted or even desired by patients
as a way to end their burdensome existence, whether due to their terminal
illness or impairments, the autonomy argument that allows him or her to
forego life-extending measures also supports giving them life-ending assis-
tance. The assumption is that quality of life determination is subjective, such
that only the patients themselves can determine if continued life in severely
compromised and debilitated states is acceptable. According to this view, if
people believe that their impairments or conditions are rendering their lives
unbearable and if medical technology cannot cure their defects or halt
their deterioration, they should be allowed to end their intolerable suffering
and have control over the timing and circumstances of death in ways they
see fit, including via assisted suicide or active euthanasia. While some have
argued that these procedures require third-party involvement and thus are
not simply matters of self-determination, they do not generally question the
capacity of the person in forming such desires and the voluntariness of the
decision itself, unless there are clear signs that the patient suffers from severe
mental illness (e.g., depression).15 They only question the claim that others
have the obligation to assist in bringing on or hastening the dying process
according to the patients wishes.16
14
Lance Stell, Physician-Assisted Suicide:To Decriminalize or to Legalize, That Is the Question,
in Physician Assisted Suicide:Expanding the Debate, edited by Margaret Battin, Rosamond Rhodes, and
Anita Silvers (NewYork:Routledge, 1998), 225251.
15
Stell, Physician-Assisted Suicide, 225251.
16
Daniel Callahan, When Self-Determination Runs Amok, Hastings Center Report 22
(1992):5255.
3. The Impoverished Notion of Individual Autonomy in

Bioethics
In the age of patient-centered care, respect for autonomy and privacy requires
that we take the wishes of patients seriously in facilitating their decisions re-
garding life-sustaining interventions. It is often assumed that state sanctions
and individual coercion are the main barriers to peoples autonomy and that
removal of such forces will allow people to freely reflect on their priorities and
values, to form their own preferences, and to realize their life plans in ways
they deem appropriate.
However, Iwish to argue that the individualist and minimalist framework
is too narrow and misses the significance of other external powers. It does not
address how many subtle and yet powerful forms of influenceparticularly
the social structure and institutional framework that promote certain ideolo-
gies and reject othersdetermine peoples available options and shape their
decision-making processes. The individualist view tends to take restriction of
autonomy as a dyadic matter between two individualsone who is dominant
(e.g., physician) and another who is subordinate (e.g., patient). It presupposes
that decisions that are not unduly restricted by the dominant agents actions
are autonomous and thus should be respected. It also focuses on individual
actions or interactions rather than social practices, unless those social prac-
tices have been formalized or codified by law, such as sanctions against eutha-
nasia. In the context of medically assisted death, the individualist-minimalist
framework focuses on whether people have all the information to make their
end-of-life decisions rather than how various forms of medically assisted
death may reflect larger structural issues such as health-care delivery, the pro-
fessionalpatient relationship, the ethos of the medical profession, and the
definition of extraordinary care. In arguing for lifting prohibitions on certain
forms of medically assisted death, many take it for granted that competent
people with all the relevant information should be allowed to make their own
decisions regarding whether or how todie.
The minimalist notion of autonomy is attractive in a liberal democratic soci-
ety because it diminishes the possibility of paternalism, especially in situations
where power hierarchy dominates the relational structure, such as health care.
In respecting peoples expressed choices as the default position, it adopts a rel-
atively value-neutral position and prevents premature intervention in peoples
lives even when we disagree with their decisions. Nonetheless, in health care
and other arenas, power and domination are not simply or always elements of
individual actions.17 Rather, they are also structural phenomena, the intended
17
Iris Marion Young, Justice and the Politics of Difference (Princeton, NJ: Princeton University
Press,1990).
or unintended product of the actions of many people that are value laden and
shape others choices. The dyadic modeling of power and autonomy misses
the impact of the larger social structure and ideology in determining patients
value frameworks and available options, including what constitutes quality of
life, what technologies are considered mainstream, and what risks are deemed
acceptable.
While the popular rhetoric regarding patient autonomy leads many to as-
sume that patients determine their desired procedures, health-care systems
are structured in such a way that patients do not have much control over their
health-care pathway. Patients diets, diagnostic or check-up schedules, access
to specialists and technologies, and discharge plans are determined not based
on patients preferences or values but primarily on lab or bed availability, cost
benefit considerations, professionals convenience, and insurance coverage.18 It
is often the accumulation of these quotidian decisions, about which permis-
sion is rarely requested, that predetermines the subsequent clinical options
and outcomes wherein patient consent is finally sought under tight time con-
straints. Patients routinely behave as they do in the health-care setting partly
because of how medicine is practiced or delivered, as determined by clinicians,
administrators, politicians, funding agencies, and various regulatory bodies.
They make decisions in various manners because of the presumed epistemic
privilege and bureaucratic power that professionals hold and according to
these clinicians expectations, even though medical staffs generally do not do
anything special to cause patients to adopt or change their actions. Patients
recognize quickly that acting against professionals recommendations even out
of careful self-reflection can get one labeled as being noncompliant or difficult
rather than being autonomous and acting in self-determiningways.
In other words, patients actions and decisions are embedded within a com-
plex set of social relations, practices, expectations, and policies that structure
their selfhood and can significantly affect their ability to exercise autonomy
with respect to their choices.19 Given that patients decision-making processes
and considerations often incorporate intrinsically relational or social content,
it is impossible to assess patient autonomy without critically evaluating how
or whether the interconnected social, political, and health-care structural
frameworks may foreclose or expand certain opportunities or predetermine
how individuals approach various health-care situations.20 Marilyn Friedman,
for example, cautions that social conditions can affect a persons ability to act
according to ones reflectively affirmed values and that the individualist view
neglects how the collective action and ideology often shape the way people
18
Richard Friedenberg, PatientDoctor Relationships, Radiology 226 (2003):306308.
19
Sherwin, Relational Approach to Autonomy in Health Care,32.
20
Catriona Mackenzie and Natalie Stoljar (eds.), Relational Autonomy: Feminist Perspectives on
Autonomy, Agency, and the Social Self (NewYork:Oxford University Press,2000).
evaluate their options by making some alternatives more costly than others.21
The individualist and minimalist approach does not ask how the social system
and our economic resources may need to be (re)organized and (re)distributed
to ensure that people have meaningful opportunities to critically reflect upon
their priorities, freely develop attitudes toward them, and make health-care
decisions that would realize their life plans accordingly.
4. Seeking Medically Assisted Death:Sociorelational Contexts
Some notable cases may help shed light on how the individualist and mini-
malist approach to autonomy fails to capture the complexities of a patients de-
cision to seek medically assisted death. Larry McAfee and Kenneth Bergstedt,
who became quadriplegic after a motorcycle accident and swimming accident
in 1985 and 1969, respectively, sought court authorization to turn off their res-
pirators years after their accidents.22 Neither McAfee nor Bergstedt was termi-
nally ill or experiencing abrupt physical decline, and they both could live for
many more years with respiratory support that had become an integral part of
their existence. Nonetheless, their respective courts presumed that the quality
of life brought on by these individuals impairments was poor and thus found
that it was reasonable for them to think of their situation as hopeless, useless,
unenjoyable, and frustrating. Since McAfee and Bergstedt were presumably
competent adults who were not directly coerced by anyone to choose death,
their respective courts appeared to have adopted the individualist-minimalist
approach of autonomy and determined that the plaintiffs had the right to
refuse artificial methods to extend their lives, which they presumed were full
of suffering.
Another example, while not battled in court, brought up similar social is-
sues around peoples alleged desire to die. Dan Crews, an Antioch man who
became quadriplegic from a car accident at age three, made national news in
2010 when hein his mid-20sasked his health-care providers to remove his
ventilator. Despite having been paralyzed from the neck down for over twenty
years, Crews reported having a happy childhood. The family won a lawsuit and
received a $4million trust to take care of Crewss medical expenses; they also
built an accessible home.23 Crews was an honors student in high school and
21
Marilyn Friedman, Autonomy, Gender, and Politics (NewYork:Oxford University Press,2003).
22
Vicki Michel, Suicide by Persons with Disabilities Disguised as the Refusal of Life-Sustaining
Treatment, HEC Forum 7 (1995): 121131; Anita Silvers, Protecting the Innocents from
Physician-Assisted Suicide:Disability Discrimination and the Duty to Protect Otherwise Vulnerable
Groups, in Battin etal., Physician Assisted Suicide, 133148.
23
Susan Donaldson James, Quadriplegic Dan Crews Swamped with Letters: Dont Die, ABC
News, December 7, 2010, http://abcnews.go.com/Health/quadriplegic-swamped-letters-begging-p
ull-ventilator-die/story?id=12324809#.UWz0dcpvAR9.
earned an associates degree. However, the trust money ran out because he out-
lived his life expectancy. When his medical bills of more than $300,000 threat-
ened foreclosure on the familys home, Crews feared he would be moved to a
nursing home. Short of having someone generous enough to help pay his medical
bills, he wanted help to die so that his family could sell the house and be settled
financially.24 Crews requested his health-care team to turn off his respirator, but
he was deemed clinically depressed and incapable of making such a decision after
he made threats toward his care team. He explained that anyone in his situation
would be depressed. (There has not been further media report of his status in the
last two years.)
A right to refuse life-prolonging interventions is important in modern health
care and ought to be respected when demanded by competent individuals who
have reflected on all their options and found such interventions too burdensome.
Such requests by terminally ill patients and their families have become a routine
part of discussions about the goals of care at the end of life. While many simply
consider these situations matters of basic civil liberty, Icontend that even these
situations are often fraught with value-laden and stereotypical assumptions about
life with impairments. Missing in these discussions are the larger social contexts
that fail people with impairments and contribute to their suffering. In acknowl-
edging peoples despair and their desire to seek death, we need to explore the var-
ious factors that contribute to peoples hopelessness. The desire to die on the part
of McAfee, Crews, and Bergstedt was embedded within a complex set of social re-
lations, policies, and circumstances that foreclosed preferred independent living
options most people without impairment take for granted. While an altered life
is presumably most difficult soon after one becomes impaired, it is important
to note that none of these men sought to die soon after becoming quadriplegic.
McAfee, Bergstedt, and Crews all lived with their impairments for years and
wanted to die only when their support resources became so severely restricted
that they had no feasible means to continue living what they considered to be a
minimally decent life, making death appear the only plausible means to end their
despair and suffering. With good nursing support covered under his insurance
plan and a van customized with a lift and locks for his wheelchair, McAfee was
able to rejoin societyhe could ride to the grocery store, the occasional movie,
or a basketball game. However, a few years later his insurance ran out and he was
put into institutional care out of state and then shuffled into a hospital because of
his restrictive Medicaid coverage. With no hope of ever living in the community
and retaining some control over his life, McAfee wanted todie.25
24
Christian Farr, Quadriplegic Prefers Death to Nursing Home, NBC Chicago, April 12, 2011,
http://www.nbcchicago.com/news/health/dan-crews-119740714.html.
25
Peter Applebome, An Angry Man Fights to Die, Then Tests Life, NewYork Times, February
7, 1990, http://www.nytimes.com/1990/02/07/us/an-angry-man-fights-to-die-then-tests-life.
html?pagewanted=all&src=pm.
The Georgia Supreme Court granted McAfees request presumably based

on a right to refuse extraordinary interventions, but the state failed to facilitate
what he really wished for and what was available to most people without im-
pairmenthis right to independent living and self-determination. What was
more noteworthy in this case was McAfees decision to keep living after the
court ruled in his favor. When the publicity of his ordeal prompted advocates
to help make it possible for McAfee to move into an independent care home
and work toward getting occupational training so that he could use his engi-
neering talents, he no longer wanted to die, even though his physical condi-
tion remained the same,26 suggesting that it was not the impairment itself that
caused his despair.
By contrast, Bergstedts case had a different ending. Becoming quadri-
plegic as the result of a swimming accident at age ten, Bergstedt lived another
twenty-one years of an apparently satisfactory life under his fathers care and
wrote poetry. However, when his father became terminally ill, he wanted to
die, worrying that the society would cast him adrift in a sea of indifference
and force him into a nursing home after his fathers passing.27 The Nevada
court recognized that Bergstedts desire to die was closely connected to his
fear of lack of social support and that if there was an appropriate substitute
caregiver he might not have wanted to die. Nonetheless, the court focused
on the right to have his respirator disconnected rather than on facilitating his
caregiving needs. They granted Bergstedts petition, and before the Nevada
Supreme Court reviewed the case his respirator was disconnected and hedied.
Were these individuals petitions for medically assisted death autonomous?
An individualist notion of autonomy would likely render these petitions auton-
omous and thus respectworthy, even though such a minimalist view neglects
how oppressive social conditions limit the availability of desired alternatives
and might contribute to despair and low quality of life in the first place. The
individualist model ignores the multiple ways ones autonomy and well-being
can be compromised by existing institutional arrangements and practices in
the health-care and social system. It mistakenly casts that respect for autonomy
is mainly or even solely about complying with patients isolated medical deci-
sions. While some may consider the courts willingness to consider withdraw-
ing life-sustaining nourishment and interventions as the promotion of civil
liberty and individual autonomy, the courts neglected to consider the possi-
bility that the plaintiffs desire to die was more of a symptom of other social
problems rather than a true desire to end their lives. McAfee and Bergstedt
went to court not because their impairments themselves imposed too much of
a burdenrather, it was the lack of social support for independent living that
26
Michel, Suicide by Persons with Disabilities,126.
27
McKay v.Bergstedt, 801 P2d 617 (Nev. 1990),628.
made their situation unbearable. There were no other desirable arrangements

available, leading them to think that, even though they were not dying, death
was the only option to escape suffering.
A relational approach to autonomy that pays attention to the impact of
oppressive social conditions on agents motivational structures and their for-
mation can help to explain how the courts decisions regarding McAfee and
Bergstedt, while ruling in favor of the plaintiffs, may have failed to truly pro-
mote their overall agency. Arelational approach to autonomy looks not only at
the agents expressed desires but also of the social and environmental contexts
in which such desires are formed and considered. It examines how the inter-
twining contexts of social relationships, sociohistorical circumstances, and the
range of options available to the agent may affect the reflective process and the
development of the agents capacity to engage in such a process. It also explores
how these desires are labeled, interpreted, and responded to by others in social
and legal contexts.
5. Ableism, Relational Autonomy, and the Alleged DesiretoDie
A look at another court case, brought forth by Elizabeth Bouvia, can shed
further light on how a relational approach to autonomy is more equipped in
addressing the complexities around the formation of a desire to seek death
in an ableist environment. In 1983, Bouvia, a twenty-six-year-old social work
graduate student with severe cerebral palsy and degenerative arthritis, admit-
ted herself to a psychiatric hospital in California as a suicidal patientshe
wanted professional help to starve to death. Bouvia had lost her motor func-
tions and was dealing with a miscarriage, financial hardship, and a failing mar-
riage. Despite her graduate training in social work, she was told that she would
never be employable.28 As people with depression have a tendency toward
global negative thinking, Bouvias psychiatrist thought that the young woman
was making a bad decision at a very bad time and that her wish to die could
diminish or change with time and treatment.29 When the hospital staff refused
to abide by her wishes and threatened to force feed her, Bouvia went to court
with the help of the American Civil Liberties Union (ACLU). The court deter-
mined that the state had viable interests in preserving life, preventing suicide,
protecting third parties, and maintaining the ethical standards of the med-
ical professions. It denied her judicial assistance to starve herself to death and
authorized feeding via a nasogastric tube.30 Bouvia checked out of the hospital
28
Michel, Suicide by Persons with Disabilities,127.
29
Mary Johnson, Right to Life, Fight toDie.
30
Bouvia v.County of Riverside, No. 159780 (Riverside Super. Ct.1983).
to starve herself, but she changed her mind later after discussions with friends
and was admitted to a county rehabilitation facility.
In 1986 Bouvia went to court again to have her feeding tube removed. The
court denied her request, but the Appellate Court overturned the decision.
While the latter court insisted that it did not assess the motive behind Bouvias
decision to refuse nourishment, it agreed that Bouvia, lying helplessly in bed,
unable to care for herself, may consider her existence meaningless.31
It is interesting to note that, while Bouvia expressed that she would rather
be dead than live like this, she also explicitly said she never wanted to die.32
She stated that physically she was feeling all right, fine, OK, or that people
could say [she is] doing OK.33 Nonetheless, the Appellate Court and main-
stream commentators continued to consider Bouvias assertions of a desire to
die, albeit inconsistent, as evidence that the woman was making a reflective
and autonomous choice. They expressed sympathy towards Bouvia, whom
they thought was in a pitiful state and living a life of helpless dependency.34
There was no discussion of the social and relational contexts within which
Bouvia formed her judgments, nor was there investigation of why or how
Bouvias wishes to die wavered at times. The Appellate Court determined that
given Bouvias low quality of life, her right to refuse life-extending interven-
tions superseded the states interests. Nonetheless, as was the situation a few
years prior, after Bouvia was granted the right to starve to death she aban-
doned her plan, this time claiming that she could not stand the pain and was
worried that her slow starvation would bring too much grief and guilt to the
staff at her facility.
Bouvias ultimate decision to continue living in her condition begs for the
question of whether the rhetoric around respect for her autonomy is an ableist
ideology in disguise. The Appellate Court and various commentators readily
accepted her inconsistent desire to die at face value, perhaps because that de-
sire harmonizes with their own perception that the primary problem for such
individuals is the unbearable experience of a permanent disability (and/or
dependence on life aids) and that death is the solution.35 If Bouvia did not
31
Bouvia v.Superior Court, 179 Cal. App.3d 1127, 1135, 225 Cal. Rptr. 297, 299300 (1986).
32
Mary Johnson, Right to Life, Fight to Die:The Elizabeth Bouvia Saga, Ragged Edge, January
February 1997, http: www.raggededgemagazine.com/archive/bouvia.htm; Beverly Beyette, The
Reluctant Survivor:9 Years After Helping Her Fight for the Right to Die, Elizabeth Bouvias Lawyer
and Confidante Killed HimselfLeaving Her Shaken and Living the Life She Dreaded, Los Angeles
Times, September 13, 1992, http://articles.latimes.com/1992-09-13/news/vw-1154_1_elizabeth-bouvia;
Elizabeth Bouvia, The Desire is Still There to Die, Lodi News-Sentinel, April 14, 1987, http://news.
google.com/newspapers?nid=2245&dat=19870414&id=7Bc0AAAAIBAJ&sjid=qDIHAAAAIBAJ
&pg=6843,5967547.
33
Beyette, Reluctant Survivor.
34
William Raspberry, Quadriplegics Life is Hers Alone to LiveAnd Also Hers to End, Spokane
Chronicle, April 7, 1986, http://news.google.com/newspapers?nid=1345&dat=19860407&id=_vpLAAA
AIBAJ&sjid=vvkDAAAAIBAJ&pg=2623,1367027.
35
Gill, Suicide Intervention for People with Disabilities.
have impairments but confirmed that her miscarriage, failing marriage, and em-
ployment and financial hardship contributed to her loss of hope in life, many
would likely have taken her previous psychiatrists testimony more seriously and
considered her intent to die as signs of depression or social problems and thus po-
tential grounds for support or intervention. The courts neglect of these consider-
ations in its decision reflects the prevailing ableist attitude that having a disability
is a sensible reason for committing suicide that requires no further investigation.
Ironically, when Bouvias own ACLU lawyer committed suicide a few years later,
she said that she was not convinced that he couldnt have gotten help. Like many
others holding an ableist ideology, Bouvia also seemed to think that impairments
were more legitimate than other reasons for one to desire death. She said she
could understand if her lawyer had tried everything and there was no way out
or if he had cancer or some physical ailment. But he had a lot to live for.36
Bouvias view of her own situation in comparison to that of her attorneys
coincides with mainstream evaluations of life with and without impairments,
respectively. Her different reactions to her and her lawyers respective death
wishes suggests that the young woman may have internalized societys rejection
and devaluation of people with impairments, such that she did not recognize
or resist the oppressive forces of ableist ideology.37 Even as dependency is a nat-
ural and unavoidable part of the human condition, dominant social structure
continues to espouse individual self-sufficiency as a norm and an ideal, such
that technological and human assistance to daily living is often interpreted by
HCPs and others to indicate an unacceptably low quality of life. McAfee and
Bergstedt relied on respiratory support as part of their daily livingsuch me-
chanical assistance had become commonplace for them. Nonetheless, as Carol
Gill points out, in a society where professionals attach words such as radical,
extraordinary, and even futile to breathing and feeding supports people use
each day, medical professionals and health-care administrators often decide
themselves what people with impairments need and define what constitutes
excessive needs.38 International epidemiological measures that quantify the
functional, financial, and other impact of various health problems, such as the
disability-adjusted life years (DALY), incorporate as purported fact the view
that disability and its consequences lower quality of life.39 Not only do these
quasi-scientific measures and HCPs opinions influence intervention recom-
mendations as well as resource and service eligibilities,40 they also impact how
36
Beyette, Reluctant Survivor.
37
Carol Gill, Suicide Intervention for People with Disabilities:ALesson in Inequality, Issues in
Law and Medicine 8 (1992):3753.
38
Carol Gill, Disability, Constructed Vulnerability, and Socially Conscious Palliative Care,
Journal of Palliative Care 22 (2006):183191.
39
Jerome Bickenbach, Disability and Life-Ending Decisions, in Battin etal., Physician Assisted
Suicide, 123132.
40
Anita Ho, Trusting Experts and Epistemic Humility in Disability, International Journal of the
Feminist Approaches to Bioethics 4 (2011):103124.
the rest of us assess and in turn affect the quality of life of people with various
kinds of impairment.
When such structural considerations frame the daily experience of people
with impairments and their interactions with the health-care system, the rhet-
oric around autonomy in medically assisted death may not capture the full
context of many peoples decision-making processes, their subsequent choices,
and social responses to such requests. The aforementioned court cases suggest
that peoples perceptions about their quality of life and their corresponding
wishes regarding their preferred forms of end-of-life care are not isolated med-
ical decisions. While many bioethicists and advocates for access to medically
assisted suicide frame the debate as matters of clinical decisions that should fall
under the rubric of informed consent and refusal, Icontend that preferences
regarding medically assisted death are not purely, or even primarily, medical
decisions. They are part of broader evaluations and decisions determined in
terms of peoples general access to various opportunities to flourish, including
their ability to maintain a multidimensional existence composed partly of sig-
nificant relationships, goals, and values. As Tom Shakespeare points out, even
individual choices are situated in social contexts, and a duty to promote au-
tonomy requires assurance that those contexts are supportive of peoples deci-
sions in the widest possible sense.41 McAfee and Bergstedt both fought to stay
out of restrictive institutional care that further limited their social relationships
and employment opportunities. Crews was worried that his family members
would lose all their possessions. Aright to discontinue life-sustaining inter-
ventions, especially when these individuals were not terminally ill, does not
address these underlying issues of restricted opportunities that are essential to
promote autonomy. Force feeding as a form of suicide prevention also did not
address the social conditions and ableist ideology that contributed to Bouvias
belief that her life was so burdensome that death was perceived to be the only
or best means to end suffering.
The intertwining social and relational factors that frame peoples desire to
die remind us that we need to clarify the cause of peoples suffering to pro-
vide appropriate relief. Studies show that hopelessness is an essential element
of unbearable suffering, and patients perceive their suffering to be unbear-
able not solely because of their medical conditions or symptoms. Nonclinical
concerns regarding the loss of social significance, communicative problems,
living arrangements, quality of care, being a burden on others, and loneliness
also contribute to peoples suffering.42 The question is partly whether the larger
41
Tom Shakespeare, The Social Context of Individual Choice, in Quality of Life and Human
Difference:Genetic Testing, Health Care, and Disability, edited by David Wasserman, Jerome Bickenbach,
and Robert Wachbroit (NewYork:Cambridge University Press, 2005), 217236.
42
Marianne K. Dees, Myrra J. Vernooij-Dassen, Wim J. Dekkers, Kris C. Vissers, and Chris van
Weel, Unbearable Suffering: A Qualitative Study on the Perspectives of Patients Who Request
Assistance in Dying, Journal of Medical Ethics 37 (2011):727734.
system sees life with impairment and dependency as equally worthy and pro-
vides appropriate opportunities for people to flourish and retain or regain hope
according to their particular contexts.43 Even when patients are afflicted with
terminal illnesses, hope and flourishing do not necessarily require reversing the
medical conditions. In the context of mortal beings, hope and flourishing are also
about being supported in various interdependent relationships and recognizing
ones full humanity regardless of ones level of functioning. As Jerome Bickenbach
points out, when an individual chooses death as the only viable way to escape
an intolerable situation partly brought on by the social environment, it seems
perverse and unfair to say that this is an expression of self-determination or
autonomy.44
6. Relational Autonomy and the Problem of Oppressive

Socialization
The social context in which people make end-of-life decisions does not simply
affect the feasible options available to them. It also affects the agents development
of her capacity to engage in a reflective process in which decisions are formed.
As many feminists have noted, social relationships and historical conditions can
either facilitate or stunt the development of autonomy.45 When there is a lack
of a supportive environment to facilitate and encourage such development, it
is difficult to ascertain whether peoples choices, including decisions to forego
life-extending interventions or to seek assistance to die, are the results of op-
pressive socialization. Like oppressive gender socialization that can curtail some
womens ability to develop the capacity for critical reflection, ableist ideology that
treats a life with impairment as categorically worse off than one without impair-
ment can impede some peoples capacity to form the self-trust and self-confidence
that are essential to possessing and exercising autonomy, especially if they do not
have other strong social or familial support networks.46 When people devalued by
the mainstream society are also deprived of the opportunities to develop the nec-
essary level of self-trust to gain and use their reflective skills effectively, they may
not be able to exercise autonomy even when they are invited to make an unco-
erced choice regarding their care goals. Such devaluation may hinder peoples
ability to critically explore their positive commitment to their particular beliefs
43
Eva Kittay, Loves Labor:Essays on Women, Equality, and Dependency (NewYork:Routledge,1999).
44
Bickenbach, Disability and Life-Ending Decisions,128.
45
Diana Tietjens Meyers, Self, Society, and Personal Choice (NewYork:Columbia University Press,
1989); Friedman, Autonomy, Gender, and Politics.
46
Friedman, Autonomy, Gender, and Politics, 97; Carolyn McLeod and Susan Sherwin, Relational
Autonomy, Self-Trust, and Health Care for Patients Who Are Oppressed, in Mackenzie and Stoljar,
Relational Autonomy, 259279.
and value systems,47 highlighting two related difficulties of determining whether

we should accept at face value peoples expressed desire to die, be it via assisted
suicide or other means.48
First, there are questions about whether ones commitment to a belief in
itself affirms the autonomy of a decision made in accordance with that be-
lief. Many proponents of legalizing various forms of assisted death, including
mainstream commentators and the Appellate Court judge in Bouvias case,
assume that patients consistent requests show that they identify with such
desires after long reflection. However, even if we assumed for the sake of ar-
gument that Bouvia and other people with impairments never wavered re-
garding their desire to die, would that imply that their processes of reaching
such a decision, and the decision itself, were autonomous? Harry Frankfurt,
for example, famously argued that when an agent reaches inner equilibrium
after reflecting on her first-order preferences and accepts them wholeheart-
edly, with no interest in making changes regarding them, the agent is choosing
these preferences autonomously.49
Many feminists have already explained how Frankfurts ahistorical view
neglects the context within which an agent considers her feasible options
and forms desires, which might skew her vision of what options are realistic
and desirable. I contend that even some historical approaches, such as that
espoused by John Christman, do not fully appreciate the extent to which an
oppressive environment can rob some people of the ability to form preferences
in an autonomous manner. In his earlier works, Christman argues for an his-
torical and counterfactual approach to examine if an agents desire is authentic
by examining whether the individual resisted the development of her prefer-
ence or would have resisted it if she had attended to its development.50 In his
more recent work, Christman clarifies that if an agent would not be alienated
from her desire if she were to critically reflect on the historical process, that is,
if she would not hold a negative judgment or emotional reaction to that pref-
erence, the preference could be considered authentic.51
On the surface, Bouvias preference to seek death could meet Christmans
authenticity requirements. Bouvia had insisted that the real issue for her was
the right to make an individual choice regarding her own future and that she
47
Andrea Westlund, Rethinking Relational Autonomy, Hypatia 20 (2009):2649; Paul Benson,
Taking Ownership:Authority and Voice in Autonomous Agency, in Autonomy and the Challenges
of Liberalism: New Essays, edited by Joel Anderson and John Christman (Cambridge: Cambridge
University Press, 2005), 101126.
48
McLeod and Sherwin, Relational Autonomy, Self-Trust,262.
49
Harry Frankfurt, The Faintest Passion, Proceedings and Addresses of the APA 66 (1991):316.
50
John Christman, Autonomy and Personal History, Canadian Journal of Philosophy 20
(1990):124; Christman, Liberalism and Individual Positive Freedom, Ethics 101 (1991):343359.
51
(Cambridge:Cambridge University Press, 2009), 155156.
reached her personal and private decision after long and careful thought.52
She did not claim to have resisted the process of forming her desire to die.
She explicitly acknowledged an awareness of services and options available but
stated that she did not want to utilize all of them. Nonetheless, even if Bouvias
choice was possibly rational given her dismal outlook or that she did not psy-
chologically resist or struggle with her decision, it does not necessarily follow
that her decision was autonomous. Her decision might have been an adaptive
one, formed as a result of internalized oppression that was so deep-seated that
it prohibited the agent to recognize its presence and force, letalone resist it. In
other words, her wish to die might have been a deformed desire.53
Echoing many feminists focus on care and interdependent relationships in
identity and character formation, many disability activists have attempted to
counter the stereotypical mainstream view that dependency and vulnerability
are abhorrent or undignified forms of living. Some of these activists believed
that it was the social contexts that made death seem desirable to Bouvia when
such an option would not have been preferable under circumstances of social
equality.54 Bouvia might have come to see the ableist environment as the limits
within which she could make her choices or become accustomed to whatever
she saw as her lot in life. The fact that her reasoning coincided with many
ableist norms when she acknowledged feeling physically fine during the
court battle demands a deeper exploration of her motivational system. Even
if Bouvia truly had adopted such values as her own, we still need to ask if
that might have been the result of her treating ableist norms as natural and
formulating desires based on such norms. As Diana Tietjens Meyers argues,
acting in a way that coincides with oppressive norms raises questions about
the agents autonomy but does not settle them.55
Carolyn McLeod and Susan Sherwin also remind us that members of op-
pressed groups are inclined to accept societys devaluing of their personal
worth on an unconscious level and to doubt their own worth.56 It is worth not-
ing that various ableist norms seemed to be built into the workings of Bouvias
motivational system and sense of self-worth. Bouvia held many beliefs that
conform to various stereotypical norms regarding life with impairments. More
importantly, she did not appear to be fully aware of or have had critically eval-
uated such external norms and their potential impact on her own assessment
of her life. She did not resist the thought of herself as being trapped in a use-
less body and found the constant use of a machine or help from another
person at times humiliating and disgusting.57
52
Johnson, Right to Life, Fight toDie.
53
Ann Cudd, Analyzing Oppression (NewYork:Oxford University Press, 2006), 180183.
54
55
Diana Tietjens Meyers, The Feminist Debate over Values in Autonomy Theory, in this volume.
56
McLeod and Sherwin, Relational Autonomy, Self-Trust,262.
57
Johnson, Right to Life, Fight toDie.
In other words, the lack of resistance in the development of Bouvias pref-

erence may not imply the autonomy of her desire to die, especially when such
preference was developed within an ableist environment, one that she did not
appear to resist. While Bouvia insisted that her decision to seek medically as-
sisted death was made in a reflective manner, the fact that the young woman
considered her life to be humiliating and disgusting suggests that she may have
suffered from the indirect effects of oppression.58 Bouvia apparently thought
of her life not only as one full of sufferingshe thought it had inferior worth
and did not want to participate in society as a result. Her lack of resistance to
the development of a desire to die could thus be the result of pervasive ableist
ideology that robbed her of the power to defy such norms. She wholeheart-
edly identified her life as one with little value, but her attachment to such a
belief makes her decision-making process even more questionable. Perhaps
more importantly, the question is whether Bouvia recognized and resisted the
process through which she arrived at her sense of estrangement and a loss of
self-worth. Declining to talk with disability activists, Bouvia did not express
an awareness of the oppressive norms around disability and how that might
or might not have affected her evaluation of her options. By all appearances,
she also did not acknowledge, assess, or develop alternative norms regarding
living with impairments and in dependent conditions. Through her legal rep-
resentative, she denied that other personal and relational factors played a role
in her desire to die, emphasizing that her quality of life was poor due to the
lack of control brought on by her impairments.
Bouvias public statement released after refusing to meet with disability
activists raises a second question about her autonomy in choosing death. What
must answerable agents do to prove that their desires formed within an ableist
environment are nonetheless autonomous, and what should the society and
professionals do when these self-regarding desires may be deformed?59 On one
hand, in providing statements to respond to disability activists who reached
out to her, Bouvia appeared to have not only internalized a disposition to an-
swer for herself but also expressed a willingness to explain her reasons for
her commitments. As Andrea Westlund argues, autonomy is tied to a com-
mitment to hold oneself answerable to external, critical perspectives.60 Such
a disposition to answer for oneself, or what Westlund calls self-representing
in justificatory dialogue, constitutes the self-responsibility that is required for
genuine self-government of choice and action.61
58
Cudd, Analyzing Oppression; Paul Benson, Free Agency and Self-Worth, Journal of Philosophy
91 (1994):650668.
59
Cudd, Analyzing Oppression, 180183.
60
Andrea Westlund, Selflessness and Responsibility for Self: Is Deference Compatible with
Autonomy? Philosophical Review 112 (2003):483523.
61
Westlund, Selflessness and Responsibility for Self,485.
On the other hand, the fact that Bouvia considered her life as one without
dignity raises questions of whether her decision-making process demonstrated
self-governance. Bouvias responses expressed an attitude that appeared to
deny or efface herself. Despite her legal quest, she did not appear to have a
strong sense of trust in her own social standing as a person who should be
treated with respect and dignity62it was unclear that she respected herself
as a moral agent. She seemed partly impervious to the disability activists line
of questioning regarding how life with impairments can still be fulfilling, only
repeating her message that her life had little to no value. Bouvia appeared to
derive no confidence or happiness from her identity. She lacked what John
Christman called self-acceptance and viewed her life with disgust and con-
tempt. She could not maintain a diachronic practical identity under which she
considered her life as worthy of pursuit while negotiating through stultifying
constraints. She fought through the psychological pain of her existence with
resignation and bitterness and could not reflexively affirm her outlook on her
options. Her responses thus seemed to undercut the self-affirmation that is in-
herent in autonomous action or a self-governing mode of existence.63
7. Choosing Death in an Oppressive Environment:Lessons for

PublicPolicy
The context within which Bouvia came to see her life as one without dig-
nity raises questions about the autonomy of her requests to seek medically
assisted death. Bergstedts decision to remove his respiratory support due to
his worry of having no means to live outside of an institutional setting may
be reasonable given his wish to remain in the community, but it did not rep-
resent his true values, goals, and desires. Their respective situations remind
us that peoples desire for medically assisted death is shaped by broader fa-
milial, social, economic, historical, and cultural contexts, some of which have
reinforced the idea that a life with impairments is burdensome or even not
worth living. An ableist social context and its associated environmental bar-
riers contribute to an inequality of autonomy among people,64 suggesting that
a minimalist approach to autonomy may neglect various relational and sys-
temic factors that can contribute to peoples alleged desire to die. There is no
sufficient statistical evidence to suggest that impairment and low quality of
life are inevitably linked or that a life with impairment is globally worse than
one without impairment. In fact, when asked about their quality of life, people
62
Mackenzie and Rogers, Autonomy, Vulnerability and Capacity,45.
63
John Christman, Coping or Oppression:Autonomy and Adaptation to Circumstance, in this
volume.
64
Bickenbach, Disability and Life-Ending Decisions,126.
with impairments often report a quality much higher than that projected by
people without impairments.65 As our examples have shown, some people
with impairments who seek assisted death would want to live if social support
and opportunity-enhancing arrangements were available, even if their physi-
ological condition were to remain unchanged, suggesting that factors beyond
ones physical conditions can have significant impact on ones quality of life.
Granting their request to die may end their suffering, but it does not neces-
sarily respect their true desire or deal with the social causes of their despair.
Despite these ableist concerns, the question remains as to whether we should
treat people with impairments differently from people without impairments
regarding their life-ending preferences. In discussions of medically assisted
death, some prominent philosophers reject the application of the autonomy
argument specifically for people with impairments, arguing that these people
constitute a vulnerable group that requires special protection.66 While people
without impairment are presumed to have had the appropriate environment to
develop the capacity to make their own reasoned choices regarding end-of-life
care, those with impairments are presumed to be vulnerable. They are believed
to be at an increased risk of harm to self and others and thus allegedly re-
quire special procedural safeguards. As a lawyer representing the Canadian
federal government warned the British Columbia Court of Appeal in March
2013 regarding physician assisted suicide, people with impairments and other
vulnerable patients may be at risk of being coerced to kill themselves. The fed-
eral lawyer appeared concerned that people with impairments are, as a class,
incompetent to assess and protect their own well-being.
Putting aside the issue of whether it is always possible clearly to distinguish
between two classes of patients, particularly in end-of-life cases, there are ques-
tions about whether correcting an ableist sociocultural framework requires
special protection for all people with impairments regarding medically assisted
death. As Ihave argued elsewhere,67 we need to acknowledge the impact of an
ableist social structure on peoples despair, and to ensure that the autonomy
language used in support for medically assisted death does not mask various
barriers of oppression. McAfee and Bergstedt did not want to diethey simply
did not want to live with severely restricted social support. The ableist socio-
cultural framework often precludes certain options from being considered and
reshapes a persons value system. Any autonomy-based argument for support-
ing medically assisted death must consider such complexities.
65
Ron Amundson, Disability, Ideology, and Quality of Life: A Bias in Biomedical Ethics, in
Quality of Life and Human Difference: Genetic Testing, Health Care, and Disability, edited by David
Wasserman, Jerome Bickenbach, and Robert Wachbroit (New York: Cambridge University Press,
2005), 101124.
66
Ronald Dworkin, Thomas Nagel, Robert Nozick, John Rawls, Thomas Scanlon, and Judith Jarvis
Thomson, The Philosophers Brief, NewYork Review of Books 27 (1997):4147.
67
Ho, Individualist Model of Autonomy, 204205.
In addressing systemic issues contributing to peoples despair and

decision-making framework, we need to lift social barriers and promote a
motivational system within which people can make informed and voluntary
treatment decisions, including those regarding life-sustaining procedures. At
the system level, increased availability of home support and personal care, as-
sistive devices (e.g., ceiling lifts, electric wheelchair), and accessible transpor-
tation services can help to maximize peoples functioning, provide respite, and
allow them to live integrated lives to the greatest extent possible. These services
generally cost less than hospital stays and can contribute to peoples psycho-
logical well-being. At the individual level, clinicians also need to go beyond a
cognitive test in determining whether a persons decision to seek medically as-
sisted death is autonomous. Given the finality of such decisions, which may be
processed in a depressive state or within an internalized oppressive state, it is
important for clinicians to fully explore the context within which patients as-
sess their clinical and overall situations and to offer any resources and support
that can foster autonomy and well-being. In addition to the aforementioned
system resources, referral to peer support, advocacy groups, or psychological
counseling may help patients get more information about living with impair-
ments and navigating the system to access various services.
My suggestions of paying attention to the impact of oppressive social struc-
ture on peoples identity and decision-making framework do not imply that
people with impairment are inherently incapable of self-government. It would
be extraordinary to contend that oppression destroys the status of all individu-
als with impairments as moral agents, rendering them categorically disquali-
fied to evaluate their overall situations and exercise judgment accordingly,
when their nonimpaired counterparts have the right to make many unwise
decisions that may also have significant and irreversible consequences. The
disability movement, the feminist movement, and other similar civil rights
movements are prime examples of how, under the right circumstances and
when given the opportunity to explore various factors of oppression, many
people in historically marginalized positions can critically reflect on their
social surrounding and adopt or even promote alternative norms. They can
also empower each other and educate those in dominant positions of the un-
just situations and motivate change. Prevention of medically assisted death
particularly for people with impairment, through denial of self-determination
based on speculation about social manipulation, treats the targets of protection
as less than persons. It violates their moral agency, isolates them, and perpetu-
ates their inferior status by allowing dominant nondisabled agents to over-
ride their expressed wishes and act paternalistically toward them.68 Behind the
faade of the care paradigm, such protective measures may ironically reinforce
68
Silvers, Protecting the Innocents, 135.
yet conceal epistemic oppression and self-serving relationships of power and

domination.
When the concern of an individuals autonomy is partly a problem of the
oppressive impacts of the social environment, respect for autonomy should
focus on removing such barriers or empowering all through social restruc-
turing rather than paternalistic protection. Instead of giving people with and
without impairment different treatment regarding medically assisted death,
which reinforces the symbolism of otherness and perception of vulnerability,
it is more important to carefully assess the cultural framework that defines and
shapes peoples life experiences. To ensure that we do not further marginalize
or patronize people with impairments, we should collaboratively examine the
context within which they assess their quality of life, attend to our hidden and
explicit assumptions about living with impairments, develop and explore al-
ternative norms, provide additional information and resources as appropriate,
and advocate for removing systemic barriers that may negatively affect peoples
life experiences. However, after such comprehensive exploration, respect for
peoples agency requires that we accept their subjective interpretation of their
own situation. It is only when a system acknowledges the input of people with
impairments and guarantees a fair process to determine adequate access to ef-
fective and affordable palliative, home-care, and hospice services that people
can be genuinely free to construct their own identity and make health-care
and other decisions according to their value system.
Acknowledgments
For critical and constructive feedback on prior versions of this paper, Ithank
participants from the Relational Autonomy: 10 Years On workshop organ-
ized by Natalie Stoljar and Catriona Mackenzie. In particular, I would like
to express my deepest gratitude toward the coeditors of this volume, Andrea
Veltman and Mark Piper, for their insightful comments and suggestions. All
remaining errors are mine and minealone.
{ Index }
ableism, 338342 142, 172; rational, 59, 245; of rape victims, 319;
abortion: bodily autonomy and, 10, 301309; reason-responsiveness and, 178; reflective,
late-term, 313324; length of pregnancy and, 36, 217; respect for, 5, 10, 99, 101, 106, 231, 349;
309313; pregnancy complications, 316317; self-constitution account of, 218; self-creative,
rape, 309313 65; self-respect and, 3; self-writing and, 192;
Abortion: The Clash of Absolutes (Tribe), 301 social and relational character of, 1, 4, 6, 75,
abstract individualism, 80 89, 93, 96, 181, 195, 208, 338; successful, 145;
accountability, 19, 3536, 38 womens, 96, 111112, 114, 138, 211, 244, 249;
ACLU (American Civil Liberties Union), 338339 work and, 280285, 287290, 293297, 299;
active self, 77, 81 unified across time, 187
adaptation: defined, 206; oppression, 207215 agential ownership, 118119
adaptive preferences: deformed desires as, Agential Virtue Account, 172177, 179
229; Double Axis Thesis and, 128, 131132; agential voice and authority, 105112
internalized oppression and, 137138; agentic skills, 121124, 126, 131, 139
Khaders theory of, 9, 132, 140; in oppressive akrasia. See weakness of will
circumstances, 5, 9, 201203 alienation, 213, 217, 219, 222, 235236
adaptive preferences formation: bad options, Ali, Hirsi, 81
choosing among, 243248; coercion, 248252; American Civil Liberties Union (ACLU), 338339
freedom to do otherwise model, 238243; American Enterprise Institute, 147
overview, 9, 227233; psychological processes Amry, Jean, 319
model, 233238 analytical skills, 121
adoption, 301 Anderson, Elizabeth, 28
affective states, 167 Anderson, Joel, 19, 37
affiliation, defined, 34 answerability, 9, 19, 32, 3536, 109112, 115, 119,
agency: adaptation and, 202, 249; authentic, 124126, 134140, 188190, 192, 196, 221, 345
211212; authority and, 93, 106, 108, 140; applied ethics, 2
autonomous, 8, 11, 89, 9396, 99101, 103106, Aristotle, 255, 283
108, 112113, 128, 132, 138, 140, 152, 178, 180, Arjomand, Homa, 148
189, 195, 205206, 213215, 217, 220, 249, 251, Astell, Mary, 71
281284, 287290, 293298; brainwashing and attitudes, 184187
manipulation and, 214; capitalism and, 298; authentic self, 3, 32, 95, 106, 164, 257258, 265,
care and, 182, 191; control over ones body and, 267268, 272273
319; dialogic interaction and, 211; distinguished authenticity: conditions of, 1718, 21, 92, 103,
from autonomy, 249; emotions and, 8, 170172, 257, 265, 343; conditions of autonomy, 183,
178, 180, 184185, 189, 214; end-of-life decisions 206, 233, 235, 246, 257, 267; conditions of
and, 326, 348; environment and, 214; epistemic, self-governance, 24, 3032, 40; honesty with
172; feminist scholarship and, 4, 62, 87, 95, oneself and, 196; mental freedom and, 187190
201; free, 179, 232; full-fledged, 165, 171, 176, authority, 8, 9094
212, 281; gender socialization and, 94; general autonomous agency in work, 288290
account of, 173; global characteristics of, 92, autonomous personhood, 44, 59, 90, 216217, 289
250; human flourishing and, 214215, 283; autonomously chosen work, 284285
local characteristics of, 92; membership in a autonomy: in bioethics, 333335; of body, 10;
moral community and, 319320; moral, 348; care-based analyses, 18; choices and, 8;
oppositional, 114; oppression and, 2, 46, 9, competency, 122, 126, 139, 260, 276; concept of,
88, 93, 103, 201, 211, 214, 231, 244, 249250, 338; 39, 61, 164165; defined, 12, 45, 78, 141, 257, 265;
passivity and, 214; of patients, 328; practical, dimensions of, 56; economic independence,
352Index
autonomy: in bioethics (Cont.) Brown, Wendy, 63

10, 285286; in end-of-life decisions, 327329; Bruckner, Donald, 234235
as finding one's own law, 4445, 4748, 59; free bundles of options, 244245
will and, 165; gender inequality problem, 256; Buss, Sarah, 213215, 304305
government's role in, 54; human development Butler, Judith, 263, 275
and, 153156; independence and, 57;
individualistic aspects of, 4243; internalized Canadian Council of Muslim Women, 148
oppression, 9496; levels of, 19; in medical capabilities theory, 2830
decision making, 329332; minimalist notion capacity and capabilities: approach to human
of, 333, 335; multidimensional, 1516, 3941; and moral development, 154156; conditional
negative liberty, 25; normative commitments, environment for, 101; independence and, 5859;
8; oppression and, 207215, 224226; relational autonomy, 2021, 3334, 81, 131, 280;
overview, 67; philosophical conceptions, work and, 280, 282284, 292295, 297298, 300
5; philosophical interest, 2; role of, 2; capitalism, 7, 4243, 49, 55, 298299
self-governance conditions, 31; social capacity, care-based analyses of autonomy, 18
19; social-relational power and authority, care of the self, 190, 192194, 197
9094; thick autonomy, 151; value-neutral, cares and caring, 178180, 182187
56; value-saturated, 56. See also feminism Caring and Internality (Jaworska), 184
commitments; relational autonomy; specific categorical imperative, 72, 283284, 298
types of autonomy Centers for Disease Control and Prevention
Autonomy and the Feminist Intuition (Stoljar), (CDC), 316317
88, 9496, 104105, 114118, 120, 125 Chaplin, Charlie, 299
autonomy-based conception of justice, 2425 character: abortion and, 324; altered in response
to oppression, 201, 203; as a component
baad (Afghan method of justice), 149, 153 of gender, 259; formation of, 8283, 344;
Babbitt, Susan, 116, 129, 132 John Stuart Mill on, 8283; of autonomous
Baber, Harriet, 243248 agents, 98, 214, 216, 218; of self-made men, 52;
bad options, choosing among, 243248 planning, 233, 239
Bakan, David, 66, 67 childrearing, 10, 6970, 80, 122, 126, 230, 234, 239,
bargaining with patriarchy, 243 255279
Bartky, Sandra Lee, 3, 228 Childress, James, 18
Beauchamp, Tom, 18 Chodorow, Nancy: men's relationship with
Benjamin, Jessica, 62, 69, 79 women, 66; object relations theory, 79; on
Benson, Paul: oppressive norms, 236; relational psychoanalysis, 70; relational autonomy, 61, 67,
autonomy, 78, 18, 87; on self-authorization, 69; women's basis for empathy, 68
220221; social agents, 36; on value axes, choices, 1, 8, 17, 25, 62, 7375, 7779, 81, 90, 97,
118120 105108, 115116, 191, 206, 241244, 295296,
Bergstedt, Kenneth, 335338 333334, 341342
Berlin, Isaiah, 7273 Christman, John: authenticity, 32; autonomous
Bickenbach, Jerome, 342 decision making, 343; changing life
bioethics: autonomy in, 333335; commodification circumstances, 9; consent confers legitimacy
of body parts, 26; end-of-life decisions and, on governments, 134; nonalienation analysis
327329; facilitating suicide, 326, 331332, of critical reflection, 18, 223; nonrelational
335338, 341; local autonomy, 20; reproductive theories, 8889; Oshana's response to, 157;
freedom, 26 self-acceptance, 346; socio-historical self,
biological influences, 267268, 273 235237
bipolar disorder, 215 Code, Lorraine, 163, 285
bodily autonomy, 10, 301309 coercion, 212, 229, 248252, 295, 326328, 330
borderline personality syndrome, 68 coercive desires, 212
Bouvia, Elizabeth, 338341, 343346 collective norms, 4648
Bouyeri, Mohammed, 147 commitments: to autonomy, 6, 35, 151153,
Bowery, Charity, 209 156159, 283, 326; to fetus, 320324;
Bratman, Michael, 184 identity-defining, 2, 18, 24, 35, 37, 59, 97, 123125,
Brison, Susan, 319 206, 218; normative, 78, 8890, 96, 99,
Brownbeck, Sam, 302 105106, 108, 215216, 278; to oneself, 4445;
Index 353
to promoting human flourishing, 283, 300; designer babies, 26

to respecting agency of workers, 283; social, desires: coercive, 212213; conflicting, 73, 211;
96102; tonal, 185, 186; transitional, 185, 186; of critical evaluation of, 7778, 136; deformed, 2,
the will, 190. See also feminism commitments, 9, 211, 228234, 237, 242245, 248249, 305, 323,
religious commitments. 345; deliberate adjustment of, 233; determined
commodification of body parts, 26 by character, 82; dwarfed, 289; first-order,
communal self, 67 183184, 213; historical formation of, 236; as
communication skills, 121 non-autonomous, 235, 249; as ones own, 75, 127,
communism, 298299 136, 211, 213; oppressive norms or conditions
compassion trait, 274 and, 114; 118, 211, 228230, 234238, 243245,
competence conditions: for autonomy, 103, 110, 249; irrational, 7275; as socially constructed,
188, 196, 205, 216, 227, 233, 235, 246, 251, 257, 73, 76; as reasons, 37; reflective, 1, 101, 213,
2656, 273, 276; of self-governance, 1718, 310311, 343
2122, 24, 3034, 40 dialogical accounts of autonomy, 6, 33, 36, 135
competency skills, 121, 205206 dialogical answerability, 109110, 195, 197
constitutive relationship, 56 dialogical rationality, 138139
Constitutivity Axis, 8, 120, 129, 133140 dialogical skills, 33
content-neutral conceptions of autonomy, 96, Dickens, Charles, 247
9899 Dinnerstein, Dorothy, 62, 69
contractarian theory, 154 Directivity Axis, 8, 116120, 129
contract model of pregnancy, 322324 disability activists, 344, 345
conversion experiences, 132 DiStefano, Christine, 61, 67
Cornell, Drucilla, 63 dogmatism, 36
corporations, , 286287 Dring, Sabine, 172174
Cowart, Dax, 331 Double Axis Thesis, 8, 127133
Crews, Dan, 335336 Douglass, Frederick, 5253
critical reflection, 1, 18, 3235, 48, 78, 81, 165, 171, Druthers Restaurant, 287
178, 180, 217, 231, 235238, 257, 260, 335, 342343, Dryden, Jane, 4
348 Duality of Human Existence, The (Bakan), 67
Cudd, Ann, 3, 228229, 238, 242243, 317 Dubay, Matthew, 302
cultural beliefs, 251 Dubay v. Wells (2007), 302304
cultural images of independence, 5859 Dworkin, Gerald, 17, 78, 87, 236237, 296
cultural opportunities, 2728, 46
cultural patriarchy, 146150 Ebadi, Shirin, 91, 93, 112
economic (or financial) independence, 10, 5657,
Dalai Lama, 283 208, 271, 282, 284287
Dalrymple, Jack, 301302 economic security, 8
Damasio, Antonio, 185 egalitarian liberals, 43
Darwall, Stephen, 36, 196 Ehrenreich, Barbara, 143
date rape, 322324 Ekstrom, Laura, 178
day-wage laborers, 145 Elster Jon, 229230, 233234
death. See end-of-life decisions emotional actions, 176177
de Beauvoir, Simone, 2712 emotional agents and autonomy: autonomy
decision making in medical treatment, 329332 theories, 178180; overview, 8, 163166; reason
deference, 5 responsiveness, 8, 165166, 171177; value
deformed desires, 9, 228232, 243245 perceptions, 166170
degradation of persons, 287, 293, 295, 310311 emotional skills, 33
dehumanization of work, 283300 emotions, defined, 34
deliberate character planning, 233234 employment. See work
democratic participation, 102104 end-of-life decisions: ableism and, 338342;
dependence, 21, 24, 31, 44, 54, 5659, 63, 7879, autonomy in, 327329; autonomy in treatment
144145, 154, 182, 244, 280, 285, 327, 339340, decision making, 329332; bioethics and, 327
342, 344. See also independence 329, 332335; medically assisted death, 335338;
deprivation, 230, 241242, 288 oppressive socialization, 342346; overview,
de-Shalit, Avner, 155 1011, 326327; public policy and, 346349
354Index
epistemic values, 120, 125126, 127, 135 freedom to do otherwise model, 232, 238243
equal pay for equal work, 157 free will and autonomy, 165, 185
equality: in capabilities theory, 29; disability Freudian psychoanalysis, 70
activism and, 344; gender, 29, 142, 157, 159, Freud, Sigmund, 63
255, 258, 262, 264, 271, 275, 279; as a goal of Friedman, Marilyn: autonomy, defined by,
liberalism, 43, 45; moral, 157; of opportunity, 27, 77; concept of autonomy, 7; on emotions,
30, 285, 388; value of, 5, 43 167; financial insecurity, 145; limitations of
ethical theory, 2 opportunities, 2; nonrelational theories, 8889;
eudemonistically meaningless work, 280281, oppressive norms, 129; personal autonomy,
293300 95; philosophical conceptions of autonomy,
evaluative identity categories, 219220, 224 45; relational autonomy and independence,
exploitation: bioethics and, 26; economic or 42; self-made men and women, 83; social
financial, 145, 282; feminism and, 153 conditions and reflective values, 334; women as
external critical perspectives, 131 capable of own rights of full moral agency, 156
Frye, Marilyn, 3
Fast Food Women (Lewis), 287 full-fledged or full-blooded agency, 165, 171
Feeling Theory, 168
Feit, Mel, 302 Gallagher, Janet, 321
feminine model of relational autonomy, 67 Gates, Bill, 51
feminine psychoanalytical development, 7 gender conformity, 117
feminism: autonomy and, 2, 710, 83, 141143, gender, defined, 259
150153, 156159, 270, 278; embodiment gendered socialization: autonomy-supportive,
and, 306; defined, 142, 258; liberal, 258, 259260; defined, 259; definitions and
278; sameness approach to, 264; relational assumptions, 257259; evaluation of options,
autonomy and, 78, 62, 77 265278; forms of, 260265; gender norms and,
feminism commitments: agential voice and 227233; non-gendered socialization, 261262,
authority, 105112; autonomy, 150153; cultural 265271; overview, 910, 255257; revisionary
and religious patriarchy, 146150; democratic feminine gender socialization, 263265, 275
participation, 102104; gender socialization, 9, 278; traditional masculine gender socialization,
255261, 265, 2704, 277; human development, 262263, 271275
153156; internalized oppression, 9496; gender identity, 6466
objections to, 157158; overview, 78, 8789, Gender in the Mirror (Meyers), 5
141143; social commitments to inclusion, gender oppression, 2, 4, 5, 20, 24, 227, 229. See
96102; social-relational power and authority, also oppression
9094; wage labor, 143146 gender psychology, 70
feminist intuition, 9495, 116118 gender-specific discipline of boys and girls by
financial insecurity, 144145 mothers, 68
first-order cares, 191, 193, 195 gender system, 268270
first-order desires, 183, 213 Genovese, Kitty, 314
Flathman, Richard, 80 Gill, Carol, 340
Flax, Jane, 61, 67, 68 Gilligan, Carol: on care, 182; moral maturity, 190
flourishing, 116, 132, 214215, 230, 240242, 263, 193, 197; object relations theory, 61; relational
276, 281, 283, 293, 300 autonomy, 69
Forbes Magazine, 51 global autonomy, 1920, 2324, 249250
Foucault, Michel: care of the self, 190, 192194, global self-esteem, 213
197; on freedom, 7677 Globe and Mail (Canadian newspaper), 148
Frankfurt, Harry: autonomous decision making, Gomberg, Paul, 285, 296
343; autonomy and, 87; commitment of the Good Work (Schumacher), 281
will, 190; hierarchical model of free will, 165; government, role of, 50, 54
identification of will, 18; personhood and Groundwork for a Metaphysics of Morals
freedom of the will, 182184 (Kant), 284
freedom: concept of, 62; conditions of
self-determination, 17, 2527; defined, 76; of habitual actions, 176
international movement, 25; negative liberty, Hampton, Jean, 311
26; relational autonomy and, 7, 7678; role of, Harding, Sandra, 6162, 67, 69
7176 harm principle, 251
Index 355
Harris, George, 323 internalized oppression, 3, 31, 9496, 111112, 120,

Hartsock, Nancy, 61, 67, 69 137138, 152, 201, 215, 223224, 227, 252, 266, 281,
health care providers (HCPs), 326, 328331, 340 340, 344, 348
health-care system structures, 334, 340 introspection skills, 121, 272
Helm, Bennett, 184, 185 intuition, 9495
heteronomous persons, 258 Islamic culture, 146149
heteronomous work, 284285
heteronomy, 72 Jacklin, Carol, 68
hierarchy of moral acts, 313316 Jacoby, Stanford, 299
higher-order desires and attitudes, 183184 Jaggar, Alison, 163
Hill, Thomas, 283, 297 Jaworska, Agnieszka: authenticity and, 188; on
Hirschmann, Nancy, 7, 61 care, 182, 183, 184; first-order cares, 191, 193;
Hirsi Ali, Ayaan, 143, 146148, 153 mental freedom, 194195
Ho, Anita, 10, 326 Jones, Karen, 170, 171172
Hobbes, Thomas, 71 judgment, 78, 191192, 204207
Hogben, Alia, 148 Judgmental Theory, 168
honesty with oneself, 196 justice: autonomy-based conception of, 2425;
Honneth, Axel, 37 and inequalities, 2223
Huckleberry Finn example, 172173, 174, 176 Justice, Gender and the Family (Okin), 261
human agency. See agency
human development, 153156 Kant, Immanuel: on autonomy, 61, 66, 72,
human flourishing. See flourishing 79; bodily autonomy and, 306; categorical
Humean tradition, 68, 165 imperative, 283284; conception of persons,
154; on degradation, 310311; on ethical theory,
identifiable self, 7475 2; individualistic morality, 70; on respect,
identification of will, 18 310311, 327; responsiveness to reason, 129;
identity formation, 291 rule-governed autonomy, 67; Taylorism and,
imagination skills, 33, 121 294; treatment of workers and, 296297
impaired persons: ableism and, 338342; Keller, Evelyn Fox: object relations theory, 62, 67,
oppressive socialization and, 342346; public 79; reactive autonomy, 65; relational autonomy,
policy and, 346349 69; true self, 66
implied contracts, 322324 Kevorkian, Jack, 331
In a Different Voice (Gilligan), 61, 190 KFC, 287
inappropriately adaptive preferences (IAPs), Khader, Serene: adaptive preferences, 9, 132,
212213 140; coercion, 248251; deformed options,
inclusion, 96102 237238; desires for oppressive conditions, 230;
independence: as an aspect of autonomy, 42, 44, inappropriately adaptive preferences, 212213
46, 5660, 65, 158, 240, 260, 285; defined, 44, Klein, Melanie, 67
56; dependence as a degree of, 567; ideal of, Kohn, Melvin, 290
7, 579, 327; as faulty ideal, 589, 158, 285287; Kolodny, Niko, 173
of mind, 59, 236; normative, 35; procedural, Kornhauser, Arthur, 288289
90, 208, 214, 232, 2368, 248, 2512; relational Korsgaard, Christine, 17, 206, 218
autonomy and, 7, 424344 5660; and Kymlicka, Will, 43
self-made persons, 56, 60. See also economic
independence labor: division of, 6970; financial independence
independent-mindedness, 59 and, 10, 5657; low-wage labor, 150; stigmatized
interdependence of human lives, 44, 79, 286 by feelings toward, 3; wage labor, 143146. See
individualism: relational aspect of, 7, 4349; social also self-made men and women; work
ontology of persons, 21 Lapham, Mike, 50, 54
inequalities, 2223, 243, 255, 259, 269270, 295, late-term abortion, 313324
346 Law of the Peoples, The (Rawls), 292
injustice: autonomy and, 111112; function of, 22; laws, one's own, 4445, 4748, 59
social, 105, 110 Laws Relations (Nedelsky), 7, 42, 43, 79
inner self, 734 legislative bills on personhood, 302
intellectual abilities, effect of work on, 290 levels of relational autonomy, 1920
intentional action model, 218 Lewis, Anne, 287
356Index
liberal democracies, 2 Narayan, Uma, 151, 152153, 211, 243244

liberal individualism, 4349 National Center for Men, 302
liberal justice, 154 Nedelsky, Jennifer: choices, 78; independence, 56
liberals and liberalism, 43, 5960 60; liberal individualism, 7, 4349; relational
liberal self, 8283 autonomy, 77; on self, 7980; self-made men
life circumstances, responding to changing in, 9 and women, 4955
Lindsey, Laura, 267 negative liberty, 25, 2627, 7276
Lippke, Richard, 291 New York Times (newspaper), 149
Little Caesar's, 295 Nickel and Dimed (Ehrenreich), 143
Little, Margaret, 310, 318 Noddings, Nel, 182
local autonomy, 1920, 249 nonalienated self-affirmation, 219
Locke, John, 71 nonalienation analysis of critical reflection, 18, 223
Love and Knowledge (Jaggar), 163 nonautonomous agents, 137
lower-order motivational states, 189, 217 non-gendered socialization (NGS), 261262,
low-wage labor, 150 265271
Ludicrous Argument, The, 163164, 180 non-ideal theory, 23
normative authority over values, 209
MacIntyre, Alistair, 79 normative commitments, 78
Mackenzie, Catriona, 6, 42, 87, 208, 251, 305306 normative individualism, 20, 21
Macoby, Eleanor, 68 Nozick, Robert, 80
Mahler, Margaret, 63, 68, 69 Nuremberg experiments, 328
manic depression, 215 Nussbaum, Martha: on abortion, 306; adaptive
Marx, Karl, 294, 298 preferences, 239240; capabilities theory, 21, 28,
McAfee, Larry, 335338 3334, 154155; human and moral development,
McKay, Brett, 53, 55 154155; self-authorization, 35; severe
McKay, Kate, 53, 55 deprivation and, 230, 249
McLeod, Carolyn, 19, 344
meaningless work, 280281, 293300 objective reasons, 173174
medical ethics, 305 object relations theory, 6171
medically assisted death and suicide, 326, 331332, obligations corresponding with rights, 314315
335338, 341 Of Liberty and Necessity (Mill), 82
memory skills, 121 Okin, Susan Moller, 261
mental freedom, 187190, 194195 ONeill, Onora, 57
mental health, effect of job conditions on, open-mindedness, 9, 38, 135, 182, 188190, 196197,
288289 262, 276, 299
Merleau-Ponty, Maurice, 64, 306 opportunity conditions of self-determination, 17,
Meyers, Diana Tietjens: autonomous decision 2730
making, 344; on autonomy, 77, 95; male oppression: adaption and autonomy, 207215;
socialization, 271272; relational autonomy, 69; by choice, 242243; circumstances of, 5, 9,
self-discovery, 246; values in autonomy theory, 201203; defined, 23; dynamics of autonomy
5, 8, 114 attribution, 224226; female complicity in, 262
Miller, Brian, 50, 54 263; feminine norms, 117118; internalization
Mill, John Stuart, 71, 8283, 283, 327 of, 9496, 111112; internalized, 9496, 111112,
minimalist notion of autonomy, 333, 335 137138; judgment and, 204207; procedural
model of intentional action, 218 autonomy and, 215224; psychological
Modern Times (film), 299 dimensions of, 3; psychological harms of,
Mohmand, Fraidoon, 149 228229; self-affirmation, 215224; social
moral acts, hierarchy of, 313316 oppression, 31
moral maturity, 190193 oppressive socialization, 342346
moral perspective of women, 154 options theory of self-determination, 2728
Moriarty, Jeffrey, 292 orthonomy, 96
Mller-Lyer illusion, 169 Oshana, Marina: authority over choice, 8;
Mullin, Amy, 316317 autonomous agency, 104; autonomous
multidimensional autonomy, 1516, 3941 personhood, 289; autonomy and feminism,
Murphy, James Bernard, 288, 298 141; coercion, 250; freedom and opportunity, 17,
Myatt, Mike, 51 134; honesty with oneself, 196; liberal self, 82;
Index 357
personal autonomy, 105; relational autonomy, rape, 309313, 322324

8889; self-authorization, 35; socio-relational rational choice theory, 120
account of autonomy, 28, 9094, 208 rational guidance, 171, 173
Rawls, John: concept/conception distinction,
parents and parenting: inculcation of competency 15; contractarian theory, 154; gendered
conditions, 266; responsibilities for autonomy socialization, 15; meaningful work, 291292
of children, 26; sexual division of labor and, Raz, Joseph: choices, 78; coercion, 250; on
6970. See also gendered socialization freedom, 77; liberal individualism, 43;
paternalism, 38, 134135, 329 opportunity conditions, 17; self-determination,
patient-centered care, 330 2728
patriarchal norms, 221 reactive autonomy, 6566, 81
patriarchal social relations, 7778 reason responsiveness, 8, 165166, 171177
patriarchy, autonomy and, 151 reason tracking, 171174
Perceptual Theory of Emotions, 168170, 179 reflective acceptance, 217
Perry, Stewart E., 293 reflective endorsement, 9192, 9798
personal autonomy: agential voice and authority, reflexive insightfulness, 136137
105112; critical reflection and, 164165; reflexive self-affirmation, 9, 203, 216218,
democratic participation, 102104; economic 222, 224
autonomy and, 286; feminism and autonomy, regulative guidance, 172
158; internalized oppression, 9496; social relational accounts of autonomy, 208
commitments, 96102; social-relational theory, relational autonomy: feminine psychoanalytical
9094 development, 7; freedom and, 7, 7678; gender
personalities, effect of work on, 290 oppression, 20; independence and, 7, 4243,
personal liberties, 26 5660; levels of, 1920; liberal individualism,
personhood: freedom of the will and, 182; 4349; normative commitments, 78; object
legislative bills, 302 relations theory and, 6171; oppressive
persons: conception of, 21, 154; threats to, 60. See socialization and, 342346; origins of, 6366;
also respect for persons overview, 45; self-authorization, 3539;
Phelan, Shane, 235 self-care and, 195197; self-determination,
philosophical moral psychology, 18 2330; self-governance, 3135; self-made men
physician-assisted suicide, 331332 and women, 7, 4243, 4955; social character
Pineau, Lois, 322 of human agency, 4; unitary conception, 1619;
Piper, Mark, 1, 9, 157, 255 vulnerability, 21. See also autonomy
Political Liberalism (Rawls), 292 Relational Autonomy (Mackenzie & Stoljar), 5,
political liberties, 25 42, 8788
Politics of Persons, The (Christman), 24 relational freedom, 63
positive liberty, 7273, 75 religious commitment, 35, 36
power, 9094 religious patriarchy, 146150
practical identity, 206207, 210 Reproduction of Mothering, The (Chodorow),
practical rationality, 173174, 206 61, 68
practical reason, 34, 35 reproductive autonomy or freedom, 26
preferences. See adaptive preferences; desires resistant slaves, 209211, 237
pregnancy, complications of, 316317. See also respect: for autonomy and autonomous choices, 2,
abortion 8, 15, 99102, 107, 153, 156, 159, 201, 231, 250251,
primary emotions, 184185 283284, 295, 297298, 310311, 326328, 333,
procedural accounts of autonomy, 56, 9, 47, 349; for the body, 319; for feminism, 153, 159;
88, 94, 96, 102, 108, 110, 112, 203, 207208, for oppressed women, 5, 95, 114, 139, 150151,
212224, 230232, 236238, 243, 248, 251252, 141; for patient autonomy, 21, 329330, 333334,
305, 330 336337; for persons, 37, 39, 99, 102, 210, 281,
procedural independence, 232, 236 2834, 297, 310311, 346; for preferences formed
programmatic level of autonomy, 19 in adaptive contexts, 5, 96; for rights, 156; for
psychoanalytic theory, 7, 6166 agency in working life, 10, 283, 299. See also
psychological processes model, 232238 self-respect
psychosexual development, 6366 responsibility: for action, 134; for aspects of
public policy and medically assisted death, oneself, 9, 36, 129, 137, 150, 189, 191193, 195, 197,
346349 345; for childcare, 69, 255; for others needs,
358Index
responsibility: for action (Cont.) options theory, 2728; as relational concept,

108; for oppression, 249; for social change, 133; 2330; structural constraints on, 24
moral, 2 self-direction, 290
reversibility of preferences, 246248 self-discovery, 246
revisionary feminine gender socialization (RFGS), self-esteem, 3, 37
263265, 275278 self-evaluative attitudes, 19, 38
Richardson, Henry, 246247 self-governance: authenticity conditions, 3132;
Ricouer, Paul, 306 autonomy conditions, 31; capabilities, 3334;
rights: to bodily autonomy, 306308; of full moral competence conditions, 17, 3233; defined, 16
agency, 156; obligations corresponding with, 17, 31; honesty and, 196; relational conception
314315 of, 3135; unitary concept, 17, 32
Robeyns, Ingrid, 28 self-identities, 7475
Rodriguez, Sue, 331 self-knowledge, 33, 137
Roessler, Beate, 291 self-made men and women, 7, 4243, 4955, 83
Roe v. Wade (1973), 302, 321 Self-Made Men (Douglass), 5153
Rousseau, Jean-Jacques, 72 Self-Made Myth, The (Miller & Lapham), 50
Rubin, Gayle, 261 self-nurturing skills, 121
Ruddick, Sarah, 182 self-reflexive attitudes, 265
self-representing, 345
Schooler, Carmi, 290 self-respect, 3, 10, 3437, 116, 125, 151, 208, 222, 247,
Schumacher, E. F., 281 280281, 284285, 291293
Schwartz, Adina, 290 self-responsibility, 150151, 195197
secondary emotions, 185 self-transformative activities, 33
second guessing, 72, 73, 75 self-trust, 37, 203, 209
second-person standpoint, 36 self-worth, 37, 119, 181182, 203, 208, 211212,
Seidman, Jeffrey: autonomous agency, 189; on 216, 219, 222, 224225, 228, 265266, 272273,
care, 182183; first-order cares, 191, 193, 195; 276277, 280281, 291292, 344345
valuing and caring, 185187 Sen, Amartya, 29, 154156, 230
self: active self, 77; dialogical conception of, 33; senses, defined, 33
inner self, 73; liberal self, 8283; outer self, 73; Sermon on the Mount, 298
responsibility for, 189; role of, 7982 severe deprivation, 230
self-acceptance, 217, 346 sexual contracts, 322
self-affirmation, 215224 Shafiqzada, Nasima, 149
self-answerability theory, 134139, 189 Shakespeare, Tom, 341
self-assertiveness, 272 Shakila story, 149, 153, 156
self-authorization: accountability, 3536; Sharia law, in Canada, 148
for autonomy, 122, 220221; defined, 18, Sherwin, Susan, 328, 344
35; self-evaluative attitudes, 3537; social Shoemaker, David, 165, 179
recognition, 35, 3738 Singer, Peter, 317
self-care: care as attitude towards others, 182187; Single Axis Thesis, 128, 130
concept of, 190195; mental freedom and, skills: for autonomy, 17, 2022, 3134, 205, 207,
187190; overview, 89, 181182; relational 227, 246, 281, 342; cognitive, 188; damaged by
autonomy and, 195197 meaningless work, 281; developed by infants
self-conception, 265 and children, 64, 264, 276; exercised at work,
self-confidence, 37, 119 10, 281283, 288, 294, 296. See also agentic skills
self-constitution account of agency, 218 slaves, 209211, 237, 270
self-deception, 246247 social capacity, 19
self-defense, 307 social commitments to inclusion, 96102
self-definition of self, 57 social conditions, 22
self-determination: autonomy as, 1; social constructivist and constructivism, 7378,
autonomy-based conception of justice, 81, 8283
2425; capabilities theory, 2830; cultural social injustice, 105, 110
opportunities, 2728; defined, 17, 25, 257; social ontology of persons, 21
freedom conditions, 17, 2527; global analysis, social oppression, 23, 31, 227
2324; opportunity conditions, 17, 2730; social recognition, 3738
Index 359
social-relational theory of personal autonomy, 25 of the Greatest Self-Made Men in American

9094 History (McKay & McKay), 53
social skills, 33 two value axes, 116121
socio-historical self, 235237
socio-relational account of autonomy, 28, 9094, Udry, Richard, 267
208, 335338 Understanding and Overcoming Pathology"
sour grapes, 229230, 233234 (Nedelsky), 7980
Srey Mom, 248 unitary concept, 1619, 39
Stoljar, Natalie: adaptive preference formation,
9, 227; autonomous agency, 104; feminist value-neutral autonomy, 56, 88, 114122, 125140,
intuition, 116118; personal autonomy, 8789, 329330
9496, 105; social relations, 42; value axes, value-saturated and value-laden autonomy, 56,
116118, 120 96, 98, 101, 104, 114122, 126133
structural accounts of autonomy, 207208 values: agential skills and, 121, 126; alienation
structural conditions of autonomy, 17, 2325, 28, and, 219; of autonomous agents, 2, 17, 54, 978,
3132, 208209, 224 103, 10610, 114122, 152; constitutive, 116, 120,
structural constraints on self-determination, 234 122124, 126127, 129130, 134140; distinctions
structural injustice and inequality, 133, 210, 243 in, 6, 8, 114115, 122128; double axis thesis, 8,
subjective reasons, 173174 127133; emotions and, 166170, 175; liberalism
Submission (television film), 146 and, 4345; overview, 8, 114115; supported by
subordinate persons, 58 autonomy, 2; two value axes, 116121
subservience, 5 Valuing and Caring Seidman, 185
substantive autonomy, 89, 88, 90, 9496, 98100, van Gogh, Theo, 146147
103104, 111, 114115, 128, 130, 152, 208, 2302, Vasanti (Indian woman), 239241, 244246, 249
240, 243, 252, 305. See also weakly substantive Velleman, David, 134, 174, 177
views Veltman, Andrea, 1, 10, 280
successful agency, 145 virtues: agential, 166, 172, 175, 177178; of
suicide, medically assisted, 326, 331332, 335338, autonomous agents, 190; female, 273; feminist,
341 273; social, 158; work and, 281, 285, 295
Superson, Anita, 10, 211, 229 voice: authentic, 107; of autonomous agents, 89,
System of Logic (Mill), 82 105112, 203, 226; internal, 210211; silencing
of, 130
Tale of Two Cities, A (Dickens), 247 volitional skills, 121
Tappolet, Christine, 8, 163 vulnerability, 7, 21, 41, 5758, 133, 144, 150, 185, 208,
Taylor, Charles: dialogical conception of self, 33; 320, 328329, 344, 347, 349
internalized recognition, 211; responsibility, 36;
second guessing, 72 wage labor, 143146
Taylor, Fredrick, 282 Waldron, Jeremy, 43
Taylor, Gloria, 331 Wallace, Jay, 171
Taylor, Harriet, 71 Wal-Mart, 286
Taylorism, 283, 293294, 300 Walzer, Michael, 3, 293
Terkel, Studs, 294 Wasserstrom, Richard, 261
The Art of Manliness (website), 53 weakly substantive views, 94, 114, 116, 119, 121122,
Theory of Justice (Rawls), 261, 292 128, 130, 215216
thick autonomy, 151, 157 weakness of will (akrasia), 33, 101, 166, 172
Thomas, Laurence, 319 welfare programs, 54, 55, 285286
Thomson, Judith Jarvis: on abortion, 10, 304; on Wente, Margaret, 148
late-term abortion, 313324; on rape, 309310; Western values, 150151
on right to bodily autonomy, 306308 Westlund, Andrea: accountability, 18; feminist
tonal commitments, 185, 186 view, 152153; Meyers' critique of, 134139;
traditional masculine gender socialization relational autonomy, 8889; relational
(TMGS), 262263, 271275 capabilities, 131; responsibility, 36; self-care,
transitional commitments, 185, 186 89, 181; self-representing, 345; self-respect,
Tribe, Laurence, 301, 303304, 316 150151; self-responsibility, 150151
Tuskegee syphilis study, 328 White, John, 285, 292
360Index
will: commitment of, 190; freedom of, 182184; work: autonomous agency in, 288290; autonomy
free will, 165, 185; identification of, 18 and, 280284; dehumanization of, 283300;
Winnicott, D. W., 67, 68, 69 equal pay for, 157; heteronomous, 284285;
Wolff, Jonathan, 155 identity and, 291; impact on autonomous
Wollstonecraft, Mary, 71 capabilities, 10; as key to success, 5253;
womanhood: NGS view of, 267; RFGS view of, meaningless, 280281, 293300; self-respect
264, 275; TMGS view of, 273, 274275; value and, 291293; social recognition, 38
system of, 114
women: harm suffered by, 60; Ludicrous Young, Iris Marion, 272
Argument and, 163164
Wood, Allen, 297, 298 Zerilla, Linda, 63

VELTMAN, Andrea. Autonomy, Oppression, and Gender

Uploaded by

Copyright:

Available Formats

VELTMAN, Andrea. Autonomy, Oppression, and Gender

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

VELTMAN, Andrea. Autonomy, Oppression, and Gender

Uploaded by

Copyright:

Available Formats

Autonomy, Oppression, andGender

STUDIES IN FEMINIST PHILOSOPHY

Recently Published in the Series:

Oxford is a registered trademark of Oxford UniversityPress

Published in the United States of Americaby

Oxford University Press2014

All rights reserved. No part of this publication may be reproduced, storedin a

You must not circulate this work in any otherform

Library of Congress Cataloging-in-PublicationData

Notes on Contributors vii

Part I Autonomy and Independence

Part II Autonomy and Normative Commitments

Part III Autonomy, Reasons, and Care

Part IV Autonomy, Oppression, and Adaptive Preferences

Part V Autonomy in Social Contexts

PaulBenson is professor of philosophy and dean of the College of Arts and

John Christman is professor of philosophy, political science, and women's

Marilyn Friedman is W.Alton Jones Professor of Philosophy and professor

Nancy J. Hirschmann is professor of political science at the University of

Political Science Association; Gender, Class and Freedom in Modern Political

AnitaHo is associate professor at the W.Maurice Young Centre for Applied

Catriona Mackenzie is associate dean (research), professor of philosophy, and

Diana Tietjens Meyers is emerita professor of philosophy at the University

monographs are Inalienable Rights: A Defense (Columbia University Press,

Marina Oshana is professor of philosophy at the University of California,

MarkPiper is assistant professor of philosophy at James Madison University.

Natalie Stoljar is associate professor in the Department of Philosophy and the

Anita M. Superson is professor of philosophy at the University of Kentucky.

Feminist Contributions to Traditional Philosophy (Oxford University Press,

Christine Tappolet is full professor in the Department of Philosophy at the

Andrea Veltman is associate professor of philosophy at James Madison

Andrea C. Westlund is associate professor of philosophy and womens

If feminism is a response to the oppression of women, and if resistance

of autonomy serves crucially in illuminating these and related questions,

Classically defined as a weighing down or as harsh dominion, oppression is

In thinking about autonomy and gender oppression, it is important to

philosophical conceptions of autonomyas opposed to conceptions of au-

Meyers, Gender in the Mirror,16.

Meyers, Gender in the Mirror,16.

autonomy requires an attitude of self-care, which involves taking care with

teaching girls to favor subordinate roles or by undermining girls confidence

of life with impairments and an appreciation of how these assumptions may

Autonomy and Independence

Three Dimensions of Autonomy:ARelational

It is no secret that although respect for individual autonomy is a foundational

conditions, which may be satisfied to varying degrees. Isuggest further that

1. Three Dimensions of Autonomy

practical identity as a normative self-conception, which embodies a per-

autonomy in an agents attitudes or dispositions to hold herself accountable

applying to a specific decision at a specific time.12 Feminist bioethicists, how-

2. Why Relational Autonomy?

Relational theories of autonomy were originally motivated by three central con-

sense, then, is the theory relational? Isuggest it is relational in several ways

the extensive interpersonal, social, and institutional scaffolding necessary for

The notion of self-governance, as Ihave already suggested, is the core or kernel

Oshana, Personal Autonomy and Society, Oshana, Personal Autonomy in Society.

conditions as part of the requirements of democratic justice. Thus, Christman

in Razs view, is a complex network of interlocking social practices, and it

Raz, Morality of Freedom,391.