Philosophical Studies Series

James F. Childress
Michael Quante Editors

Thick
(Concepts of)
Autonomy
Personal Autonomy in Ethics and
Bioethics
Philosophical Studies Series

Volume 146

Editor-in-Chief
Mariarosaria Taddeo, Oxford Internet Institute, University of Oxford, Oxford, UK

Editorial Board
Patrick Allo, Vrije Universiteit Brussel, Brussel, Belgium

Advisory Editors
Lynne Baker, Department of Philosophy, University of Massachusetts,
Amherst, USA
Stewart Cohen, Arizona State University, Tempe, AZ, USA
Radu Bogdan, Dept. Philosophy, Tulane University, New Orleans, LA, USA
Marian David, Karl-Franzens-Universität, Graz, Austria
John Fischer, University of California, Riverside, Riverside, CA, USA
Keith Lehrer, University Of Arizona, Tucson, AZ, USA
Denise Meyerson, Macquarie University, Sydney, Australia
Francois Recanati, Ecole Normale Supérieure, Institut Jean Nicod, Paris, France
Mark Sainsbury, University of Texas at Austin, Austin, TX, USA
Barry Smith, State University of New York at Buffalo, Buffalo, NY, USA
Nicholas Smith, Department of Philosophy, Lewis and Clark College
Portland, OR, USA
Linda Zagzebski, Department of Philosophy, University of Oklahoma,
Norman, OK, USA
Philosophical Studies Series aims to provide a forum for the best current research in
contemporary philosophy broadly conceived, its methodologies, and applications.
Since Wilfrid Sellars and Keith Lehrer founded the series in 1974, the book series
has welcomed a wide variety of different approaches, and every effort is made to
maintain this pluralism, not for its own sake, but in order to represent the many
fruitful and illuminating ways of addressing philosophical questions and
investigating related applications and disciplines.
The book series is interested in classical topics of all branches of philosophy
including, but not limited to:
• Ethics
• Epistemology
• Logic
• Philosophy of language
• Philosophy of logic
• Philosophy of mind
• Philosophy of religion
• Philosophy of science
Special attention is paid to studies that focus on:
• the interplay of empirical and philosophical viewpoints
• the implications and consequences of conceptual phenomena for research as well
as for society
• philosophies of specific sciences, such as philosophy of biology, philosophy of
chemistry, philosophy of computer science, philosophy of information, philoso-
phy of neuroscience, philosophy of physics, or philosophy of technology; and
• contributions to the formal (logical, set-theoretical, mathematical, information-­
theoretical, decision-theoretical, etc.) methodology of sciences.
Likewise, the applications of conceptual and methodological investigations to
applied sciences as well as social and technological phenomena are strongly encouraged.
Philosophical Studies Series welcomes historically informed research, but
privileges philosophical theories and the discussion of contemporary issues rather
than purely scholarly investigations into the history of ideas or authors. Besides
monographs, Philosophical Studies Series publishes thematically unified antho­
logies, selected papers from relevant conferences, and edited volumes with a well-
defined topical focus inside the aim and scope of the book series. The contributions
in the volumes are expected to be focused and structurally organized in accordance
with the central theme(s), and are tied together by an editorial introduction. Volumes
are completed by extensive bibliographies.
The series discourages the submission of manuscripts that contain reprints of
previous published material and/or manuscripts that are below 160 pages/88,000 words.
For inquiries and submission of proposals authors can contact the editor-in-chief
Mariarosaria Taddeo via: [email protected]

More information about this series at https://link.springer.com/bookseries/6459

James F. Childress • Michael Quante
Editors

Thick (Concepts of)

Autonomy
Personal Autonomy in Ethics and Bioethics
Editors
James F. Childress Michael Quante
University of Virginia University of Münster
Charlottesville, VA, USA Münster, Germany

ISSN 0921-8599     ISSN 2542-8349 (electronic)

Philosophical Studies Series
ISBN 978-3-030-80990-4    ISBN 978-3-030-80991-1 (eBook)
https://doi.org/10.1007/978-3-030-80991-1

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Switzerland AG 2022
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

Many ethical theories and practices, particularly in domains such as biomedical eth-
ics, incorporate minimal, formal, procedural conceptions of personal autonomy and
autonomous decisions and actions. Over the last 30 years or so, concerns about the
deficiencies and limitations of these “thin” conceptions have led to “thick” concep-
tions that highlight certain mental, corporeal, biographical, and social conditions of
what it means to be a human person and that seek to enrich concepts of autonomy,
with direct implications for ethical requirements to respect personal autonomy. The
chapters in this book offer a wide range of perspectives, largely within an analytic
philosophical approach, on both the elements of and the positive and negative rela-
tions between “thin” and “thick” concepts of autonomy as well as their relative roles
and importance in ethics and bioethics.
It may be useful for readers to understand the logic of the overall structure of this
volume even if, with that understanding, they autonomously chose to read the chap-
ters in a different order. The first three chapters attend to key contested issues in
conceptions of autonomy with particular reference to major philosophical contribu-
tors who have sought to frame and resolve these issues. Featured in all three of these
chapters is the influential work of philosopher Harry Frankfurt who taught at
Princeton University. Understanding and assessing Frankfurt’s work is crucial for
engaging in contemporary debates about conceptions of autonomy.
In the opening chapter, John Martin Fischer (University of California, Riverside)
clarifies what autonomy is by examining what moral responsibility is not. Paying
particular attention to Frankfurt’s work, along with others, Fischer argues that fail-
ure to draw the appropriate distinction between “moral responsibility” and “auton-
omy” creates conceptual and normative difficulties. According to Fischer, “moral
responsibility” refers to a minimal condition for rendering an agent an “apt target”
for reactive attitudes, including moral blame. By contrast, “autonomy” indicates “a
more robust notion of self-governance.” It presupposes moral responsibility but is
also more demanding. Although there is debate about how to specify what else is
demanded, this roughly includes governance by the “real” or “true” self (or a desig-
nated set of goals or aims) in practical reasoning. An agent’s failure to act in accor-
dance with his or her true or real or designated self, perhaps because of weakness of

v
vi Preface

will, can still be a free act in the sense required for moral responsibility, even though
it falls short of autonomy. In short, the mental elements, particularly identification,
that link choice and action are more robust for autonomy than for moral
responsibility.
Joel Anderson (Utrecht University) rejects a common picture of Frankfurt’s work
on autonomy, contending that it does not seek to defend a moral theory of the person
or of autonomy. Instead, what Frankfurt offers is a normatively thin ontology of
autonomy, not a normative hierarchy of desires or wholeheartedness as a moral
criterion. He distinguishes “persons” from “wantons,” creatures who don’t really
care what they desire. According to Anderson, Frankfurt’s central idea is that a nec-
essary feature of the ontology of the will of persons is the capacity to be troubled by
their desires. Anderson calls this the predicament of volitional reflexivity. In this
interpretation, Frankfurt’s goal is to provide an account of what would count as a
successful response to this predicament, captured in the idea of volitional coherence
and expressed in the language of “wholehearted” identification with some desires,
that is, what one cares about. Refusals to face this predicament represent the passiv-
ity of non-persons (“wantons”). Volitional incoherence may threaten autonomy
through ambivalence or disorientation. Anderson argues that Frankfurt’s language
of “wholeheartedness” and, more recently, “self-satisfaction” is too passive and
excessively thin because it does not adequately incorporate self-understanding or
making sense of the position one takes.
In Chap. 3 entitled “Determining Oneself and Determining One’s Self,” Thomas
Schramme (University of Liverpool) exploits an ambiguity in the concept of self-­
determination. On the one hand, self-determination can mean to determine oneself
in choices and actions; on the other hand, it can mean to determine one’s self. In the
second kind of self-determination, agents imagine alternative self-ideals to be actu-
alized. Such a self-determination means not only to actualize a self but also to criti-
cally examine whether this self is the one that he or she wants to be or should be,
where “should” is understood hypothetically, not categorically, because it is bound
to the perspective of the person him- or herself. This normative conception of self-­
determination thus incorporates ideas of a right or good way to determine oneself.
Developing his normative interpretation of self-determination by reference to three
theorists (Harry Frankfurt, Thomas Hill Green, and John Stuart Mill), Schramme
concludes by pulling together the “best elements” in their theories: “the proper way
of self-determination consists in developing one’s own human potentials as far as
one oneself would like, in addition to realizing these potentials in decisions and
actions.”
Some internal and external conditions are important and even necessary for
autonomy and its exercise in certain circumstances, as the next two chapters argue.
Johann Ach (University of Münster) and Arnd Pollmann (Alice Salomon Hochschule
Berlin) identify “the most fundamental personal condition for deciding or acting
autonomously.” They contend that, in contrast to elementary competence, this con-
dition is too often neglected or, when recognized, rendered too demanding. In their
formulation, the crucial self-relation in autonomous decisions and actions is the
“triple-S-condition,” consisting of “self-confidence,” “self-assertiveness,” and
Preface vii

“self-esteem.” They argue that this triple-S-condition is necessary for autonomy and
that any decision or action lacking it cannot appropriately be considered “autono-
mous.” After describing the genesis and development of the triple-S-condition
beginning in childhood and triple-S-susceptibility in later life, Ach and Pollmann
draw some normative implications for ensuring and protecting the social relation-
ships indispensable for acquiring and preserving the triple-S-condition.
In stressing that “autonomy is a cooperative enterprise in many ways,” John
Christman (Pennsylvania State University) focuses on the implications of respect
for autonomy in relations in which mutual interactions help to establish or reestab-
lish the autonomy of one or both of the agents and may constitute in part the exer-
cise of their autonomy. An example of symmetrical power relations, featuring
co-deliberation, involves romantic partners considering whether to have children,
while an example of asymmetrical power relations is an aid worker helping a trauma
victim regain autonomy. These “partnerships,” in which interpersonal dynamics
have a causal and constitutive relation to the establishment or exercise of autonomy,
pose a challenge for the liberal sensibility of respect for autonomy. In such partner-
ships, the duty to respect another’s autonomy should also take account of the ways
that the interactions are constitutive as well as expressive of autonomy. Christman
offers guidelines based on his socio-historical conception of autonomy. For instance,
the obligation to respect the autonomy of others in relations that may influence the
development of that autonomy must be diachronic in facilitating “the development
of a practical identity which can be the basis of future judgments that would not be
disturbed when considered in light of that very process of interaction and
facilitation.”
The next two chapters consider whether and to what extent mistaken beliefs or
limited understanding can prevent decisions and actions from being autonomous.
Viewing autonomy within “the family of metaphysical freedom terms” (including
free action, free will, etc.), Alfred Mele (Florida State University) examines “actions
distinctive of self-ruled or self-governed individuals” with attention to whether cer-
tain facts about beliefs might render an action non-autonomous. He uses several
examples, including his well-known case of King George raising taxes 5% on the
basis of the his advisors’ provision of false information to him, to indicate the
uncertainty about moving, along various dimensions, from apparently paradigmatic
cases of different kinds of action, such as autonomous action, to other supposed
instances of those kinds of action. Sometimes it is difficult to be confident about
whether an action is autonomous or not (or intentional, or voluntary, or free or not).
Nevertheless, Mele holds that an agent does not act autonomously if another agent
has virtually complete control over the information going to the first agent and pro-
vides false information to the first agent who then acts on it. The first agent’s action
is heteronomous.
While there is widespread agreement that persons must understand what they are
doing in order for their decisions or actions to be considered autonomous, there is
little agreement or even much discussion in ethics or bioethics or other areas about
how much understanding is required. In contrast to more robust, thicker accounts
that require substantial understanding, James Stacey Taylor (The College of New
viii Preface

Jersey) concludes that “very little” understanding is actually required for autono-
mous decisions and actions. After first arguing that the moral value of autonomy
does not undergird the ethical requirement to obtain a patient’s informed consent in
medical care, he then develops, against critics of this view, theoretical points about
the sufficiency of a thin understanding for autonomous decisions and actions. In
making his arguments, Taylor examines situations in which the acting agent’s inten-
tional description of the action is not how a neutral third party would describe it, and
the agent’s autonomous decisions and actions are of little value to him or her. He
concludes that the minimalist account of understanding in autonomy is superior to
more robust accounts.
The next two chapters consider negative and positive relations between certain
“thick” conceptions of personal autonomy and respect for personal autonomy in
decisions and actions. In Chap. 8 entitled “Is ‘autonomy talk’ misleading?” Thomas
Gutmann (University of Münster) contends that the ongoing debate about autonomy
would be greatly improved if the distinction between two concepts of autonomy
were taken more seriously: (1) personal autonomy as a graded and at least moder-
ately perfectionist ideal of the autonomous person, and (2) autonomy as right. He
argues that these two concepts are independent and have different and even conflict-
ing normative functions. Hence, it is important to treat them as separate, rather than
to use quantitative characterizations such as “thinner” and “thicker” or “minimum”
and “full” to distinguish them. To avoid confusion, Gutmann suggests that we call
“autonomy as right” the “right to self-determination” or “authority,” rather than
“autonomy.” This involves a threshold conception of competence, recognizes a
binary, ontic status of persons who can make claims as moral equals, and focuses on
the priority of right over the good (the latter including ideal concepts of autonomy).
In questions about autonomy as authority in biomedical ethics or medical law,
Gutmann insists, there is no place for “thick” concepts of autonomy, however
important they may be in other contexts. No moral or legal consequences follow
from not meeting thick concepts—a “bad” achiever of “thick” personal autonomy
still has authority. According to Gutmann, any moral conception of autonomy is
inadequate if it fails to account for the respect owed to autonomous agents whose
equal moral right to hold views, make choices, and act in accord with their own
values and rights should be acknowledged.
Focusing on respect for personal autonomy in bioethics—what Gutmann calls
“autonomy as right” or as “authority”—James Childress (University of Virginia)
explores several complexities of respecting autonomous choices by reference to the
case of a 14-year-old boy who died after being allowed to refuse a necessary blood
transfusion on religious grounds (the Jehovah’s Witness tradition). Childress argues
that some thicker concepts of autonomy, closely connected with relational auton-
omy, can direct our attention to aspects of respect for autonomy that are often
neglected or underappreciated in bioethical theory and practice. In particular, they
illuminate the interpretive complexity of respecting autonomous choices by attend-
ing to the relational context of individuals’ decisions, respect for individuals’ social-­
cultural beliefs and values, and the temporality of selves, as expressed in their past
consent or future consent as well as in—and sometimes in contradiction to—their
Preface ix

present consent or dissent. However, Childress’ conclusion overlaps with Gutmann’s:

it is crucially important not to convert the illuminating and valuable insights from
theories of relational autonomy into necessary conditions of autonomous choices
and actions and hence of respect for autonomy.
The final two chapters consider whether, in certain contexts, different vulnerabil-
ities to risk should affect judgments about whether an agent’s decisions and actions
are autonomous or non-autonomous. In autonomous authorization in bioethics, the
notion of competence serves a “gate-keeping” function; it provides “an entrance
ticket” that entitles patients to make authoritative treatment decisions in light of
their non-medical values and preferences. Bettina Schöne-Seifert (University of
Münster) considers whether the concept of mental competence, a key component in
autonomous decision making and in informed consent to and refusal of medical
procedures, varies according to the level of risk (i.e., the probability and magnitude
of net harmful consequences) involved in the decision. A risk-related conception of
competence, which assigns a “sliding scale” for competence based on the level of
risk involved in the decision, is now widely used in practice and accepted by a num-
ber of bioethicists. However, Schöne-Seifert contends that this conception has two
fatal flaws. On the one hand, paternalism can be buried in the concept of compe-
tence and can mask whether a patient’s decision is overridden because of genuine
incompetence or because of the risk of harm. On the other hand, a patient’s self-­
authorization is devalued by attributing illusory autonomy to incompetent patients
in low-risk situations. By contrast, Schöne-Seifert argues for a fixed concept of
patients’ decisional competence—the capacity to understand and appreciate—in the
context of informed consent; this identifies a sufficiency threshold, independent of
the level of risk involved in the decision.
In “Vulnerability, Exploitation, and Autonomy,” Catriona Mackenzie (Macquarie
University) challenges certain bioethical defenses of what intuitively appear to be
exploitative commercial transactions, particularly organ sales and commercial sur-
rogacy. These defenses combine a liberal analysis of exploitation with a libertarian
analysis of autonomy. Mackenzie contends that the liberal analysis of exploitation
sees these transactions as involving only two parties, thereby occluding the struc-
tural dimensions of exploitation. This occlusion then allows the transaction to be
viewed in terms of libertarian autonomy. Mackenzie’s approach instead pairs a vul-
nerability analysis of the transaction with a multidimensional, relational theory of
autonomy, which together highlight the structural dimensions of exploitation and
expose the myth-making that underlies claims that these forms of exploitation are
actually autonomy-enhancing.
The volume concludes with “Outlook,” some editorial reflections on its impor-
tant contributions in illuminating, especially from the standpoint of analytic phi-
losophy, the complexities and tensions in the distinctions and relations between
“thick” and “thin” conceptions of autonomy and respect for autonomy. These reflec-
tions also note the volume’s limits, which present challenging opportunities to
develop broader perspectives on the autonomy of persons, decisions, and actions in
at least three contexts. A first context is to broaden the range of philosophical tradi-
tions, beyond both analytic philosophy and western philosophy to encompass
x Preface

intercultural philosophy including dialogue with Asian and African traditions. A

second context is the changing environment of autonomy, which includes chal-
lenges of an aging society, care for persons with dementia, information technolo-
gies, autonomous artificial systems (such as autonomous vehicles or weapons), and
various kinds and means of personal enhancement. For many of these it is unclear
and debatable whether proposed interventions strengthen, transform, or destroy per-
sonal autonomy. A third context includes tools from within and across particular
sub-disciplines of philosophy, for example, not only ethics but also epistemology
and philosophical anthropology, to name a few. This philosophical intra-­
disciplinarity and interdisciplinarity also needs to be accompanied by rigorous
research findings related to autonomy from relevant social, biological, and other
sciences.
Most of the chapters in this volume were initially prepared for an international
conference on “Thick (Concepts of) Autonomy?” organized by the Centre for
Advanced Study in Bioethics at Westfälische Wilhelms-Universität in Münster,
Germany, in October 2012. These original papers were then thoroughly revised
based on the discussions at and after the conference, and supplemented by addi-
tional contributions. This rich examination of autonomy within the broad analytic
philosophical tradition is an international undertaking, with contributors from
Germany, the United States, the Netherlands, the United Kingdom, and Australia.
As editors, we thank the authors of these chapters for their important contribu-
tions to on-going debates about how best to characterize personal autonomy and
autonomous decisions and actions, the elements and the advantages and disadvan-
tages of “thin” and “thick” versions of these concepts, the various positive and nega-
tive relations between these versions, and their implications for respect for autonomy
in both ethics and bioethics.

Charlottesville, VA, USA  James F. Childress

 
Münster, Germany  Michael Quante
May
  2021
Contents

1 What Moral Responsibility is Not����������������������������������������������������������    1

John Martin Fischer
2 The Passivity of Self-Satisfaction:
A Critical Re-appraisal of Harry Frankfurt’s
Normatively Thin Ontology of Autonomy ��������������������������������������������   17
Joel Anderson
3 Determining Oneself and Determining One’s Self��������������������������������   33
Thomas Schramme
4 Self-Confidence, Self-Assertiveness, and Self-Esteem:
The Triple S Condition of Personal Autonomy ������������������������������������   53
Johann S. Ach and Arnd Pollmann
5 Autonomy, Respect, and Joint Deliberation������������������������������������������   67
John Christman
6 Autonomy and Beliefs������������������������������������������������������������������������������   87
Alfred R. Mele
7 How Much Understanding Is Needed for Autonomy?�������������������������� 101
James Stacey Taylor
8 Is “Autonomy Talk” Misleading? ���������������������������������������������������������� 117
Thomas Gutmann
9 Respecting Personal Autonomy in Bioethics:
Relational Autonomy as a Corrective?�������������������������������������������������� 135
James F. Childress

xi
xii Contents

10 Patients’ Decision-Making Competence:

Discontents with a Risk-Relative Conception���������������������������������������� 151
Bettina Schoene-Seifert
11 Vulnerability, Exploitation and Autonomy�������������������������������������������� 175
Catriona Mackenzie
12 Outlook����������������������������������������������������������������������������������������������������� 189
James F. Childress and Michael Quante

Index������������������������������������������������������������������������������������������������������������������ 193
Chapter 1
What Moral Responsibility is Not

John Martin Fischer

Abstract Moral responsibility and autonomy are closely related structurally and
contentwise: they are both members of the “freedom family”. Here I argue that
because of these similarities, they are often conflated or at least not carefully sepa-
rated, and that this has resulted in confusions in important contemporary debates.
Autonomy and moral responsibility involve the agent’s identification with the
sources of her actions; but autonomy-identification is more robust than
responsibility-identification.

1.1 Introduction

Over the years I have noted that there appear to be two separate–but often conflated–
notions: moral responsibility and autonomy. In previous work, I have sought to
distinguish these notions and to sketch some difficulties that can arise if they are not
distinguished appropriately.1 Here I shall further explore aspects of the distinction
between moral responsibility and autonomy. I shall argue that Harry Frankfurt’s
focus has shifted subtly from moral responsibility to autonomy over the course of
his career, and I wish to identify some additional examples–from the work of such
influential philosophers as Gary Watson, Charles Taylor, David Velleman, and
Michael Bratman in which the failure (in at least some authors) to keep these closely
related notions distinct has vitiated the analysis of the phenomena.2

1
See John Martin Fischer (2010, 2012).
2
I do not mean to claim that all of these philosophers themselves have been insufficiently attentive
to the distinction; some of them have been explicit in making the distinction. My main point is that

J. M. Fischer (*)
University of California, Riverside, CA, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 1

Switzerland AG 2022
J. F. Childress, M. Quante (eds.), Thick (Concepts of) Autonomy, Philosophical
Studies Series 146, https://doi.org/10.1007/978-3-030-80991-1_1
2 J. M. Fischer

1.2 Frankfurt

In his early articles, Harry Frankfurt’s fundamental focus was on distinguishing two
kinds of freedom—“freedom to do otherwise” and “acting freely”—and arguing
that acting freely (properly understood) is both compatible with causal determinism
and all the freedom required for moral responsibility. He thus construed his contri-
butions in these early papers as squarely within the realm of the traditional debates
in metaphysics about the relationships between causal determinism, freedom, and
moral responsibility.
In his 1969 essay, “Alternate Possibilities and Moral Responsibility,” Frankfurt
is clearly addressing those who have thought that freedom to do otherwise is
required for moral responsibility. He argues that such freedom is not required for
moral responsibility, and thus we might be able to make progress on the traditional
metaphysical debates, which tended to focus on the relationship between causal
determinism and freedom to do otherwise. If Frankfurt is correct here, we can side-
step these debates and attend to whether “acting freely” is compatible with causal
determinism.3
Now Frankfurt goes on in his 1971 paper, “Freedom of the Will and the Concept
of a Person,” to give an account of “acting freely” and to explore aspects of its rela-
tionship to causal determinism. He first argues that his theory of freedom of the will
can meet two conditions that various other accounts cannot: it can account for our
disinclination to attribute such freedom to members of other species and it can show
why freedom of the will is desirable. Frankfurt goes on to say:
It is generally supposed that, in addition to satisfying the two conditions I have mentioned,
a satisfactory theory of the freedom of the will necessarily provides an analysis of one of
the conditions of moral responsibility. The most common recent approach to the problem of
understanding the freedom of the will has been, indeed, to inquire what is entailed by the
assumption that someone is morally responsible for what he has done. In my view, however,
the relation between moral responsibility and the freedom of the will has been very widely
misunderstood. It is not true that a person is morally responsible for what he has done only
if his will was free when he did it.
… For the assumption that a person is morally responsible for what he has done does not
entail that the person was in a position to have whatever will he wanted. This assumption
does entail that the person did what he did freely… It is a mistake, however, to believe that
someone acts freely only when he is free to do whatever he wants…(1971, pp. 18–19)

Frankfurt reiterates in the concluding section of the paper the point that the sort of
freedom that involves access to alternative possibilities (freedom to will or do oth-
erwise) is not required for moral responsibility. Rather, according to Frankfurt, the
freedom-relevant requirement for moral responsibility is acting freely, and–in the
view presented in “Freedom of the Will and the Concept of a Person”–it suffices for

a proper evaluation of some of their work will require care in distinguishing moral responsibility
from autonomy.
3
For further discussion and a defense of Frankfurt’s strategy here, see: John Martin Fischer (2002),
reprinted in John Martin Fischer (2006, pp. 124–142).
1 What Moral Responsibility is Not 3

acting freely that one identifies with the particular first-order desire that moves one
to action via the formation of a second-order volition (i.e., a second-order prefer-
ence to act in accordance with a certain first-order desire). Here it is in virtue of the
second-order volition that one “identifies” with the relevant first-order desire, in the
sense of “identification“ involved in moral responsibility. We can say that Frankfurt
is here answering the question of what must be added to merely acting on desire to
get to the kind of freedom linked to (and, more specifically, entailed by) moral
responsibility; the answer is that one must in a distinctive way “identify” with the
desire, and this identification consists in forming an appropriate second-order
volition.
Frankfurt concludes the paper with ruminations on the relationship between his
account of freedom of the will and the traditional metaphysical worries about the
relationship between such freedom and causal determinism:
My conception of the freedom of the will appears to be neutral with regard to the problem
of determinism. It seems conceivable that it should be causally determined that a person is
free to want what he wants to want. If this is conceivable, then it might be causally deter-
mined that a person enjoys a free will…(1971, p. 20)

Whatever else Frankfurt had in mind in putting forward his “hierarchical” account
of acting freely, he certainly sought in his early papers to make contributions to the
traditional metaphysical debates about determinism, free will, and moral
responsibility.4
But I would contend that, beginning with Frankfurt’s “Identification and
Externality,” there is a subtle shift in Frankfurt’s aim—a shift that Frankfurt himself
does not explicitly note.5 Perhaps, although this is difficult to ascertain, Frankfurt
did not even realize that he was shifting his target of analysis from the notion of
freedom that is a condition for moral responsibility to a rather different notion of
freedom. But whether or not Frankfurt recognized this shift, I wish to argue that he
did indeed change his focus of attention. And others followed suit.
As noted above, in “Freedom of the Will and the Concept of a Person,” Frankfurt
employs the term, “identification”, to refer to a psychic element that must be added
to action on a first-order desire to get to “acting freely”, the freedom-relevant condi-
tion for moral responsibility. Further, this condition is arguably consistent with
causal determinism; at least it is not in tension with causal determinism in the way
that freedom to do otherwise is. But in “Identification and Externality,” I do not find
any uses of the terms “causal determinism” and “moral responsibility”; nor do I find
any instances of “freedom”. The focus here is not on the traditional metaphysical
problems concerning causal determinism, freedom, and moral responsibility; the

4
In his 2002 article “Reply to John Martin Fischer,” Frankfurt further elaborated and explained his
views on the traditional problem of the relationship between causal determinism and moral respon-
sibility. Here he explained that his goal has been to address certain objections to compatibilism,
although he remains officially neutral about the compatibility of causal determinism and moral
responsibility insofar as he is not sufficiently confident that causal determination is consistent with
being active (rather than passive): p. 29.
5
Frankfurt (1976); reprinted in Frankfurt (1988, pp. 58–68).
4 J. M. Fischer

focus is on something else—something perhaps closely related to the traditional

issues, for which Frankfurt also employs the term, “identification”.
Roughly speaking, Frankfurt says that corresponding to the distinction between
action and mere movement of the body, there is a distinction between being identi-
fied with a “passion” (or mental element) and being alienated from a passion (or
mental element); that is, there is a distinction between a mental feature’s being
“internal” to one’s mind or in a certain special sense “one’s own” and its being
“external” to oneself in this special sense, although also in a different sense one’s
own. As Frankfurt puts it:
The problem of attribution or of externality with regard to bodily movements is to explain
what differentiates a movement that a person makes from a mere happening in the history
of his body. The corresponding problem with regard to psychic phenomena concerns the
nature of identification: what does it mean for a person to be identified with one rather than
with another of two passions (or whatever), both of which are in the gross literal sense his?
Each of these problems is, in its own realm, both fundamental and frustratingly recalcitrant.
(1976, p. 251)

In this paper Frankfurt makes a distinction that is important to the subsequent devel-
opment of his views—a distinction between two desires that fit into the same order-
ing, as it were, and two desires that do not fit into the same ordering. When two
desires—say, a desire to see a concert and a desire to see a film—fit into the same
ordering, the agent’s challenge is simply to figure out which desire should have
priority within the single ordering. On the other hand, one might have a conscious
desire to compliment a friend on a recent accomplishment but also a (perhaps)
unconscious desire to insult him out of jealousy. As Frankfurt puts it:
When a person suffers a conflict of [this] kind, his problem is not to assign one desire that
he feels to a lower position than another on a single scale. It is to reject one of the desires
that he feels altogether. Conflicts between conscious and unconscious desires are typically
of this kind, but both of the conflicting desires may well remain conscious. … The conflict
is not one to be resolved by ordering the conflicting desires, in other words, but by rejecting
one of them. In rejecting the desire to injure his acquaintance, presuming that this is what
he does, the person withdraws himself from it. He places the rejected desire outside the
scope of his preferences, so that it is not a candidate for satisfaction at all. Although he may
continue to experience the rejected desire as occurring in his mental history, the person
brings it about in this way that its occurrence is an external one. The desire is then no longer
to be attributed strictly to him, even though it may well persist or recur as an element of his
experience. (1976, p. 250)

Here, then, Frankfurt’s target of analysis seems to be something a bit different from
the notion of freedom linked to moral responsibility. Here “identification” is sup-
posed to pick out a process whereby certain of one’s goals or aims are thought to be
more closely associated with oneself—these goals are “internal” to oneself in a
special way—a way in which other goals, although one’s own, are not. In using
“identification” to pick out the process whereby some goals or aims become more
truly one’s own, Frankfurt is subtly shifting his target of analysis.
In the early papers, “identification” was employed to denote the mental element
that is to be added to mere action on the basis of a particular first-order desire to get
to “acting freely” (the sort of freedom implicated in moral responsibility). Again, I
1 What Moral Responsibility is Not 5

do not know if Frankfurt himself was at the time of writing “Identification and
Externality” aware of the shift, and the fact that the target of analysis has subtly
changed can easily be missed by a reader, especially given that the same term,
“identification”, is employed by Frankfurt in the 1971 paper and also “Identification
and Externality”, published in 1976.
In his subsequent work, Frankfurt seems to be increasingly interested in the
notion of identification that plays a role in specifying a set of aims or goals more
closely associated with the self, rather than the more minimal notion of identifica-
tion that is (putatively) part of the freedom-relevant condition for moral responsibil-
ity. I introduce a bit of technical terminology here to simplify the exposition.6 That
is, I shall use the term, “moral responsibility”, to refer to a more minimal condition,
in which an agent is (for example) an apt target for the reactive attitudes (on the
Strawsonian way of interpreting the nature of moral responsibility). And I shall use
the term “autonomy” to refer to a more robust notion of self-governance, which
presupposes moral responsibility but requires more. What more is required is con-
tentious and difficult to specify, but roughly it involves governance by the “real” or
“true” self as regards practical reasoning, that is, governance by a special or desig-
nated “part” of the self or (slightly less metaphorically) set of aims and goals.
Although Frankfurt does not make this distinction explicitly, or employ the terms,
“moral responsibility” and “autonomy” as above, I contend that the distinction is
important. Further, I claim that in his later work Frankfurt’s main (although not
exclusive) focus is on autonomy, rather than mere moral responsibility.
In his later work, beginning perhaps with his essay, “The Importance of What We
Care About,” Frankfurt is primarily interested in giving an account of “caring” and
explicating the role of caring in understanding our agency and in explaining norma-
tivity (the way in which considerations become reasons for action for agents, on one
way of conceptualizing these matters).7 On the view of caring defended by Frankfurt,
identification—understood as the formation of a second-order preference to act in
accordance with a certain first-order desire–plays a role, but there are additional
requirements (that is, requirements in addition to those for acting freely). These
requirements involve the idea that one cares about things and persons only if one has
a relatively stable and long-term commitment to the targets of the attitude of caring.
So, for example, if one finds the first-order desire in question “flagging” or becom-
ing less strong, one is disposed to take measures to refresh it. As Frankfurt puts is:
A person who cares about something is, as it were, invested in it. He identifies himself with
what he cares about in the sense that he makes himself vulnerable to losses and susceptible
to benefits depending upon whether what he cares about is diminished or enhanced.
(1988, p. 83)

On Frankfurt’s view, when one cares about something, one has a higher-order atti-
tude that the relevant first-order desire “not be extinguished or abandoned”

6
Here I follow Fischer (2012).
7
Frankfurt (1982); reprinted in Frankfurt (1988, pp. 80–94). Subsequent references will be to the
reprinted version.
6 J. M. Fischer

(Frankfurt 1999a, p. 161). About someone who cares about attending a concert,
Frankfurt says:
… his caring about going to the concert implies that he is disposed to support and sustain
his desire to go to it even after he has decided that he prefers to satisfy another desire
instead. Forgoing the concert would frustrate his first-order desire to attend the concert. In
itself, on the other hand, it would fail to touch the higher-order desire—which he might or
might not have—that this first-order desire not be extinguished or abandoned. His caring
about the concert would essentially consist in his having and identifying with a higher-order
desire of this kind.
… Whether the person cares about the concert is not fundamentally, then, a question
either of how enthusiastic he feels about going to it or of how beneficial he believes or
expects going to it would be. No doubt beliefs, feelings, and expectations of those kinds do,
in many situations, more or less reliably indicate whether a person considers something to
be important to him. What is at the heart of the matter, however, is not a condition of feeling
or of belief or of expectation but of will. The question of whether a person cares about
something pertains essentially to whether he is committed to his desire for it in the way that
I have suggested, or whether he is willing and prepared to give the desire up and to have it
excluded from the order of preferences. (1999a, p. 161)

What, then, is the importance of what we care about? In my view, Frankfurt takes it
that what we care about specifies the “true” or “real” or “designated” self, as regards
practical reasoning: the set of aims and goals that specify what is more closely asso-
ciated with the agent and thus would help to define autonomy (self-governance). As
Frankfurt puts it, one identifies oneself with what one cares about; caring is an ele-
ment that helps to specify what one identifies oneself with in the sense of a set of
aims or goals that are more truly one’s own.
Frankfurt goes on to develop an account of love that builds on his account of car-
ing; on this view, love involves caring about someone’s interests for their own sake
and also not being able to avoid or resist so caring. Thus, for Frankfurt, love is a
species of caring that is characterized by “volitional necessity” (Frankfurt 1999b).
Another role played by caring in Frankfurt’s overall framework is that it generates
normativity; we have reasons for action—including reasons for being moral—only
to the extent that we care about the relevant action and, indeed, about morality. The
“true self” is the self that explains our reasons for action, and for Frankfurt that it is
a matter of what you care about.8
To take stock, I claim that Frankfurt was not always careful to distinguish
between moral responsibility, which was the primary focus of his early work (in
particular, the famous 1969 and 1971 papers in The Journal of Philosophy) and
autonomy, which was the primary (although not exclusive) focus of his later work.
The crucial notion of identification plays a role in Frankfurt’s accounts of both
moral responsibility and autonomy, and this common element (and term) might
have (at least in some places) made it more difficult—for Frankfurt and other theo-
rists—to separate the phenomena. But it is analytically important to distinguish the
notions of moral responsibility and autonomy.9 For example, it seems clear that one

8
See, for example, Frankfurt (2004) and Frankfurt (2006).
9
For additional argumentation for this point, see Fischer (2012).
1 What Moral Responsibility is Not 7

could act in a “weak-willed” manner, that is, against the dictates of one’s “true” or
“real” or “designated” self. In such a case, one would (or could) be morally respon-
sible and yet one would not be acting autonomously. Moral responsibility is the
“gateway” to autonomy, but it is less demanding; one can be morally responsible but
not self-governing in the way required by autonomy.
I am not sure whether Frankfurt was aware of the shift in his focus. Here it might
be helpful to consider the beginning of the entry on “Mission Creep” in Wikipedia:
Mission creep is the expansion of a project or mission beyond its original goals, often after
initial successes. The term often implies a certain disapproval of newly adopted goals by the
user of the term. Mission creep is usually considered undesirable due to the dangerous path
of each success breeding more ambitious attempts, only stopping when a final, often cata-
strophic, failure occurs. The term was originally applied exclusively to military operations,
but has recently been applied to many different fields. The phrase first appeared in articles
concerning the UN Peacekeeping mission during the Somali Civil War in the Washington
Post on April 15, 1993 and in the New York Times on October 10, 1993. (Wikipedia 2011)

A salient example of mission creep relates to the United States policy surrounding
the war in Iraq (launched by George W. Bush). Originally, the mission was to find
and destroy the “weapons of mass destruction” that were allegedly possessed by
Iraq (under the leadership of Saddam Hussein). When such weapons were not
found, the United States explicitly changed the mission to one involving the elimi-
nation of a tyrant and the building of a reasonably stable “democracy” in the region
of the Middle East.
Mission creep can then come in (at least) two varieties. In one form, the mission
is explicitly (or fairly explicitly) reevaluated and changed openly. In another form,
the mission does in fact change, but the change is not explicitly recognized and/or
acknowledged. This kind of mission creep might be called “mission drift”, and I
would suggest that mission drift arguably, at least, characterizes Frankfurt’s work
on agency (and related topics) over the years.
I must concede that there are elements even in Frankfurt’s early work on these
topics that suggest that he might have in mind as the target of his analysis something
closer to autonomy than moral responsibility. Also, even in his later work—work
that focuses primarily on developing accounts of caring, love, and normativity—
Frankfurt does indeed come back to the role of identification in the traditional meta-
physical disputes. He says:
But suppose that we are doing what we want to do, that our motivating first-order desire to
perform the action is exactly the desire by which we want our action to be motivated, and
that there is no conflict in us between this motive and any desire at any higher order. In other
words, suppose we are thoroughly wholehearted both in what we are doing and in what we
want. Then there is no respect in which we are being violated or defeated or coerced.
Neither our desires nor the conduct to which they lead are imposed upon us without our
consent or against our will. We are acting just as we want, and our motives are just what we
want them to be. Then so far as I can see, we have on that occasion all the freedom for which
finite creatures can reasonably hope. Indeed, I believe that we have as much freedom as it
is possible for us even to conceive.
Notice that this has nothing to do with whether our actions, our desires, or our choices
are causally determined. The widespread conviction among thoughtful people that there is
8 J. M. Fischer

a radical opposition between free will and determinism is, on this account, a red herring.
(2006, pp. 15–16)

Whereas I contend that the thrust of Frankfurt’s efforts in his early work seeks to
analyze moral responsibility (rather than autonomy) and that a subtle shift—a “mis-
sion drift”—occurs over the course of Frankfurt’s career, Benjamin Mitchell-Yellin
(2012) has suggested that Frankfurt’s work has always had elements that address
each of these related notions. That is, Mitchell-Yellin claims that there has not so
much been a change in focus over the course of Frankfurt’s career, but that Frankfurt
has always been concerned with both notions, although he has not always been suf-
ficiently careful about distinguishing them. For my purposes here it does not really
matter whether the Mission Drift thesis or some alternative, such as a Mish-mash
Thesis, is true; what is most important is the contention that there are two closely
related but importantly different notions—moral responsibility and autonomy—that
are not always distinguished with sufficient care in Frankfurt’s work.
I should point out that Frankfurt has indeed shown an awareness of the distinc-
tion in question (which I have marked by “moral responsibility” and “autonomy”).
In the Preface to his 1988 collection, The Importance of What We Care About,
Frankfurt says:
So far as freedom is concerned, it is of course true that freedom is commonly understood to
be a necessary condition of moral responsibility. Moreover, the nature of freedom is often
investigated, as in several efforts of my own, in the light of that particular connection. On
the other hand, I do not think that it is mainly for the sake of moral responsibility that we
care as much as we do about being free. Nor am I convinced that it is possible to illuminate
what it means for us to be free only if we begin by construing freedom specifically as a
condition that being morally responsible requires. (1988, p. viii)

I agree with Frankfurt that it is not only in light of our interest in moral responsibil-
ity that we are concerned about freedom, and it might even be correct that it is not
“mainly” for the sake of moral responsibility that we are interested in freedom. And
I certainly agree with Frankfurt here that there are various routes to illumination
about freedom; it is not true that the only such route is via reflection on the condi-
tions for moral responsibility. No worries thus far. But I would still insist that it is
analytically important to distinguish moral responsibility from autonomy; freedom
as it relates to moral responsibility need not be the same as freedom as it relates to
autonomy, and failing to keep the notions of freedom separate can lead to confu-
sions and difficulties.
As above, note that one can fail to act in accordance with one’s “true”, “real”, or
“designated” self and yet act freely in the sense relevant to moral responsibility.
That is, one can presumably freely fail to pursue any of the set of aims and goals
specified by one’s account of autonomy as more closely associated with oneself;
some such cases might well be cases in which one is morally responsible for one’s
behavior. Thus, it is important to distinguish moral responsibility from autonomy.
An account of autonomy that is problematic—that in some way fails to capture the
“true” or “real” self as regards practical reasoning—would not thereby be problem-
atic as an account of moral responsibility. This is my point, and it is completely
1 What Moral Responsibility is Not 9

compatible with the contentions that we do not care about freedom solely or even
mainly because of its relationship to moral responsibility and that it is not true that
the only route to illumination of freedom is through reflection on the conditions for
moral responsibility.

1.3 Other Theorists: Taylor, Watson, and Velleman

1.3.1 Taylor

In his famous paper, “Responsibility for Self,” it emerges that Charles Taylor inter-
prets Frankfurt as primarily seeking to give an account of what I have called,
”autonomy”, rather than moral responsibility (even though Taylor employs the term
“responsibility” for the target phenomenon). He begins the paper as follows:
What is the notion of responsibility which is bound up with our conception of a person or
self? Is there a sense in which the human agent is responsible for himself which is part of
our very conception of the self?
This is certainly a commonly held idea, among ‘ordinary men’ as well as among phi-
losophers. … Frankfurt has made the point (J. Phil., Jan. 1971) that a person is more than
just a subject of desires, of choices, even of deliberation; that we attribute to persons the
ability to form ‘second-order desires’; to want to be moved by certain desires, or ‘second-­
order volitions’: to want certain first order desires to be the ones which move them to action.
If we think of what we are as defined by our goals, by what we desire to encompass or
maintain, then a person on this view is one who can raise the question: Do I really want to
be what I now am? (i.e., have the desires and goals I now have). In other words, beyond the
de facto characterization of the subject by his goals, desires, and purposes, a person is a
subject who can pose the de jure question: is this the kind of being I ought to be, or really
want to be? There is as Frankfurt puts it a ‘capacity for reflective self-evaluation’… mani-
fest in the formation of second-order desires (J. Phil., 7). (1976, p. 281)

Taylor contends that Frankfurt’s apparatus, as it stands, is too weak to capture the
pertinent kind of reflective self-evaluation, of which he seeks to give his own
account. Taylor’s account of the putatively required reflective self-evaluation
involves what he calls “strong evaluation”, rather than “simple weighing” of alter-
natives. For Taylor, a strong evaluator can offer articulate explanations of the supe-
riority of one course of action over another, where these courses of action are
experienced as incommensurable. It is perhaps not too much of a distortion to say
that, on Taylor’s view, the required kind of reflective self-evaluation involves the
capacity to employ and articulate something like deontological, and not mere con-
sequentialist, evaluations of courses of action.
I claim that Taylor interprets Frankfurt’s 1971 account as seeking to specify what
I above called “autonomy”, rather than moral responsibility. That is, Taylor invites
us to “think of what we are as defined by our goals”, and then gives what he takes
to be a more refined and plausible account of how to pick out these goals—the goals
more closely associated with the person in question. Here Taylor is interpreting the
capacity to step back from one’s first-order desires and to identify with one of them
10 J. M. Fischer

via the formation of a second-order volition as part of a project of specifying the

“designated” set of aims or goals—the “true” or “real” self, as it plays a role in
autonomy.10
As above, it is perhaps unclear just how Frankfurt was interpreting his account,
even in his 1971 paper. But it is at least worth noting that one of Frankfurt’s avowed
aims in that paper was to address the traditional metaphysical issues pertaining to
the relationship between casual determinism, freedom, and moral responsibility. It
is then not so surprising that Frankfurt gave a rather “minimalist” account of identi-
fication in the 1971 paper, and emphasized that the relevant second-order volition
need not be based on any specific, substantive sort of reflection, including rational
reflection of any sort.
Frankfurt’s claim here would be more appropriate if one’s fundamental aim were
to give an account of moral responsibility, rather than autonomy. Further, it seems
to me that Taylor’s suggestion is more appropriate as an account of autonomy than
of moral responsibility. For certainly one could be morally responsible for a choice
that does not in any way spring from strong evaluation; here one might well be fail-
ing to act autonomously and failing to act in accordance with some norm pertaining
to our agency and even personhood, broadly conceived; but one would not thereby
fail to be morally responsible. Or so I would suggest, and it seems best to leave this
open as a theoretical possibility.
Perhaps Taylor would wish to maintain that the capacity for strong evaluation is
a requirement for moral responsibility, even when strong evaluation is not exhibited
in a particular case (or in the causal sequence leading to a particular choice and bit
of behavior). This may be true, but it is rather delicate and difficult to evaluate such
a claim. I do not here wish to dispute this sort of position. That is, it may well be that
an agent is morally responsible only if he has at least the capacity to engage in a
certain sort of rational self-reflection.11 But I would simply point out that acting
freely, in the sense relevant to being morally responsible for a particular choice and
bit of behavior, does not require that one exhibit or display such reflection either at
the time of the behavior in question or along the causal sequence to it.

10
In his work on Frankfurt, it is obvious that Jan Bransen (1996, 2000) interprets Frankfurt in a
similar fashion. Bransen emphasizes the distinction between “identifying with” and “identifying
as”, and he contends that the proper interpretation of Frankfurt employs “identifying as”. According
to this approach, when an agent identifies as such and such, he picks out such and such as an “alter-
native of oneself”. So, on this picture, identification is not about selecting a certain designated
mental element, but, rather, about defining oneself in a certain way.
11
For a very thoughtful exploration of this (and related issues), see: Andrew Eshleman (2004).
1 What Moral Responsibility is Not 11

1.3.2 Watson

Gary Watson has offered an insightful and influential critique of Frankfurt, and also
an alternative account of free action, in his paper, “Free Agency”.12 Watson criti-
cizes Frankfurt’s hierarchical model, and he endorses a non-hierarchical model in
which there are different “sources” of behavior—motives and values. Put simply, on
Watson’s approach, one acts freely to the extent that one harmonizes these different
sources; that is, an individual acts freely insofar as he is motivated by his values. On
this view, unfreedom stems from a discordance between values and motivation.
Watson says:
… It is not easy to give a nontrivial account of the sense of ‘to value’ in question. … We
might say that an agent’s values consist in those principles and ends which he—in a cool
and non-self-deceptive moment—articulates as definitive of the good, fulfilling, and defen-
sible life. That most people have articulate ‘conceptions of the good,’ coherent life-plans,
systems of ends, and so on, is of course something of a fiction. Yet we all have more or less
long-term aims and normative principles that we are willing to defend. It is such things as
these that are to be identified as our values. (p. 91)

Unfortunately, Watson does not say much more about the crucial notions, “good”
and “defensible”; for example, from what perspective and to whom is the concep-
tion of the good and fulfilling life to be defended? In any case, Watson continues:
The valuational system of an agent is that set of considerations which, when combined with
his factual beliefs (and probability estimates), yields judgments of the form: the thing for
me to do in these circumstances, all things considered, is a. To ascribe free agency to a
being presupposes it to be a being that makes judgments of this sort. To be this sort of being,
one must assign values to alternative states of affairs, that is, rank them in terms of worth.
The motivational system of an agent is that set of considerations which move him to
action. We identify his motivational system by identifying what motivates him. The possi-
bility of unfree action consists in the fact that an agent’s valuational system and motiva-
tional system may not completely coincide. Those systems harmonize to the extent that
what determines the agent’s all-things-considered judgments also determines his
actions. (p. 91)

Watson at least appears to claim here that one acts freely if and only if one’s values
and motivations are in alignment. (He does not here discuss the possibility that
one’s values do not specify a unique action or that perhaps one’s valuational system
contains elements that are incommensurable.) Perhaps a more accurate way of for-
mulating Watson’s claim is: one acts freely if and only if one is moved to act by a
desire because of the worth of its object.13
As with Taylor, Watson criticizes Frankfurt’s account of freedom as being too
“minimalist” or austere. Watson holds that merely invoking desires, even desires of
higher orders, does not give one enough resources adequately to analyze the rele-
vant kind of freedom. According to Watson, one must add values to the

12
Watson (1975); reprinted in Moral Responsibility (1986), ed. Fischer, pp. 81–96. All references
will be to the reprinted version.
13
I am indebted to Benjamin Mitchell-Yellin for this suggestion.
12 J. M. Fischer

psychological mix. But now note that Watson faces an obvious problem of accom-
modating weak-willed agency (here understood as free action against one’s values).
Insofar as Watson appears to define free action as simply a harmony between value
and motivation, and unfree action as a lack of such a harmony, he would appear to
leave no room for genuinely free but weak action. And this is a deep problem for
Watson’s view.14
The problem here stems from requiring too close a connection between value
and motivation. Of course, as in the analysis of Taylor above, we could interpret
Watson as seeking to give an account of autonomy, rather than moral responsibility.
Alternatively, again as in the discussion of Taylor, we could loosen the connection
posited by Watson to simply require, for free action, that the agent have the ability
or power to achieve a harmony between value and motive, even if he does not actu-
ally achieve this harmony. This would be parallel to the interpretation of Taylor as
requiring, for moral responsibility, merely the capacity to engage in strong evalua-
tion, rather than the actual exercise of such evaluation in the sequence leading to the
behavior under consideration.
In evaluating Watson’s approach, as in the case of Taylor’s model, it is crucial to
distinguish between moral responsibility (and its associated freedom) and auton-
omy (and its associated freedom). One can exercise a robust kind of freedom, in
virtue of which one is fully morally responsible, even when one does not act in
accordance with one’s strong evaluations or one’s values. We have to leave concep-
tual space for this possibility. If we do not distinguish moral responsibility from
autonomy, we are in jeopardy of collapsing the space of possibilities, and in so
doing, conflating issues and vitiating our analyses of related—but distinct—agency
phenomena.

1.3.3 Velleman and Bratman

In “What Happens When Someone Acts?”, David Velleman offers an interesting

critique of Frankfurt.15 He says:
Suppose that I have a long-anticipated meeting with an old friend for the purpose of resolv-
ing some minor difference; but that as we talk, his offhand comments provoke me to raise
my voice in progressively sharper replies, until we part in anger. Later reflection leads me
to realize that accumulated grievances had crystallized in my mind, during the weeks before
our meeting, into a resolution to sever our friendship over the matter at hand, and that this
resolution is what gave the hurtful edge to my remarks. In short, I may conclude that desires
of mine caused a decision, which in turn caused the corresponding behaviour; and I may
acknowledge that these mental states were thereby exerting their normal motivational force,

14
In contrast to Frankfurt, Watson has admitted that weakness of the will constitutes a challenge for
his view, and he has grappled with this problem in numerous subsequent papers, including: Watson
(1977, 1987, 1999, and 2003).
15
Velleman (1992); reprinted in Perspectives on Moral Responsibility (1993), eds. John Martin
Fischer and Mark Ravizza, pp. 188–210. Subsequent citations will be to the reprinted version.
1 What Moral Responsibility is Not 13

unabetted by any strange perturbation or compulsion. But do I necessarily think that I made
the decision or that I executed it? Surely, I can believe that the decision, though genuinely
motivated by my desires, was thereby induced in me but not formed by me; and I can
believe that it was genuinely executed in my behaviour but executed, again, without my
help. Indeed, viewing the decision as directly motivated by my desires, and my behavior as
directly governed by the decision, is precisely what leads to the thought that as my words
became more shrill, it was my resentment speaking, not I. (pp. 191–192)

Velleman is here criticizing the notion that one can give a proper analysis of action
that reflects the true self—where the agent really stands—simply by employing the
resources of desire, belief, and “standard causation”, even supplemented by
Frankfurt’s notion of “decisive commitment” (Frankfurt 1987). It should be evident
that Velleman is interested in something different from mere moral responsibility, or
so at least I would suggest.16 Indeed, it seems to me that the individual depicted in
Velleman’s story can legitimately be held morally responsible for his “decision” to
sever his relationship with his friend, just as one can be held morally responsible for
various habitual, unreflective, and or even subconsciously-motivated behavior. One
is morally responsible for severing one’s relationship with one’s friend, even if it
does not stem from one’s “true self”.
Insofar as we share Velleman’s intuition about such a case that the individual is
not fully and robustly involved, we might say that, although the individual is fully
morally responsible for severing his relationship with his friend, he does not do so
autonomously. His behavior in severing the relationship thus does not reflect a more
robust participation or involvement of the self. It is not evident to me, from the story
as told by Velleman, that this is indeed so. Perhaps Velleman’s view about the case
stems from an excessively intellectualist theory of the relevant aspects of the self.17
But we can imagine that the story is filled in so that it is more congenial to Velleman’s
conclusion. Even so, such a case would at best show that a Frankfurtian model
would not provide a fully adequate account of autonomy; it would not follow that
such a model could not provide an adequate account of the more minimal notion of
moral responsibility. Failing to distinguish these two potential targets of analysis
could then lead to analytical difficulties. If one concluded from a story such as
Velleman’s that Frankfurt has given an inadequate account of moral responsibility,
this would be a mistake.
Similarly, consider Michael Bratman’s example:
Perhaps I think it strictly better to be a person who forgives and turns the other check [sic]
but nevertheless, in a kind of self-indulgence, allow into my life a willingness to express
reactive anger. Though this role of my desire to express my anger diverges from my relevant
evaluative judgments, it is not a desire I reject or disown. … So, value judgment is one
thing, and ownership another. (2003, reprinted in 2007; quotation on p. 144 of 2007)

Bratman is here offering a critique of Watson’s “Platonic” model of the “true” self,
according to which one is identified (as this notion figures in an analysis of

16
Velleman agrees: see note 11, p. 192 of paper as reprinted in Fischer and Ravizza, eds.,
Perspectives on Moral Responsibility.
17
I am indebted to Justin Coates for this point.
14 J. M. Fischer

autonomy) with those desires endorsed by one’s evaluative commitments. My con-

tention is that, whatever is correct about the true self and autonomy, we can consider
Bratman morally responsible for acting from anger (in the case as Bratman describes
it). So if one concluded from a story such as Bratman’s that Watson has given an
inadequate account of moral responsibility, this would be a mistake.

1.4 Conclusion: What Moral Responsibility is Not

In the last 40 years or so there has been a tremendous explosion of literature in what
has come to be called “agency theory”. Some of this work has focused on free-
dom—freedom to do otherwise, free action, and acting freely. Freedom is thought
(by many, although not all) to be a condition on moral responsibility, although the
specific notion of freedom in question is a matter of controversy. But freedom—of
some kind or another—is also a key feature of something related to, but different
from, moral responsibility; I have called this more robust thing, “autonomy”,
because it involves a kind of self-governance, where the “self” is perhaps a more
robust notion than the corresponding element in moral responsibility.
Various philosophers who have written on these subjects have used different
terms for their target of analysis. Some have indeed employed “moral responsibil-
ity” to pick out the target of their inquiry, and others have used “autonomy”. Still
others have used terms such as “agency par excellence” and “agency at its best”.
Some have talked of behavior that is guided by mental elements that have “agential
authority” or that “speak for the agent”.
Whereas the targets of the analyses have been picked out by different terms,
philosophers have often—although not invariably—employed the notion of identi-
fication as an element of the analysis. I have argued here that it is important to dis-
tinguish different targets—in particular, moral responsibility and autonomy. Further,
I would point out that it is easy to get mixed up, given the similarity of the language
employed in the various analyses—the hierarchical framework shared by many, and
the element referred to by “identification”.
Despite the confusing terminology, I think we can usefully distinguish two
underlying phenomena that are closely related but importantly different. Understood
in terms of identification, the two kinds of phenomena are structurally similar in that
they essentially involve uniting different parts of the self, so to speak; that is, they
both involve bringing together different elements of the mind or perhaps connecting
one’s action to certain elements of the mind.
On this sort of picture, when one acts in such a way as to be morally responsible,
one identifies with one’s action in the sense that one chooses and acts in accordance
with a certain mental element or structure—the element or structure that specifies
moral-responsibility identification. On Frankfurt’s 1971 view, this element or struc-
ture is a second-order volition. And when one acts autonomously, one also identifies
with one’s action; but here one chooses and acts in accordance with a different
mental element or structure—the element or structure that specifies
1 What Moral Responsibility is Not 15

autonomy-identification. These items (relevant to autonomy) may be a set of aims

or goals specified by one’s theory of autonomy, or they may be set of judgments
based on value (on Watson’s approach).
So acting freely in the sense relevant to moral responsibility and acting freely in
the sense relevant to autonomy are structurally parallel; they both involve linking
specified elements with one’s choices and/or behavior. But the elements in question
may well be different. On Frankfurt’s 1971 view, the highlighted element related to
moral responsibility is a second-order volition. In his later work, the specified aims
and goals are ones the agent cares about or loves. These are interpreted in part in
terms of second-order volitions, but Frankfurt adds elements to the analysis, which
suggests autonomy-identification. Also, the aims and goals relevant to Taylor’s view
of autonomy require a certain sort of reflection involving strong evaluation, and the
relevant elements in Watson’s view somehow stem from “reason” or reflection on
value. In general, then, there is both a structural similarity between moral responsi-
bility and autonomy at an abstract level and an important difference: the mental
elements with which choice and action are to be linked are more “robust” in the case
of autonomy than moral responsibility.
Given the similarities between the two notions, and the similar analytical frame-
works and terminology employed in them, it is perhaps easy to conflate the two
separate ideas. But it is important to keep them separate. In particular, it might well
turn out that analyses that are inadequate to give an account of autonomy are quite
promising with respect to moral responsibility. In seeking to understand what moral
responsibility is, it is crucial to remember what moral responsibility is not.18

References

Bransen. 1996, January. Identification and the Idea of An Alternative of Oneself. European Journal
of Philosophy 4: 1–16.
———. 2000, January. Alternatives of Oneself: Recasting Some of Our Practical Problems.
Philosophy and Phenomenological Research 60: 381–400.
Bratman, Michael. 2003. A Desire of One’s Own. Journal of Philosophy 100: 221–242; reprinted
in Bratman 2007, 137–161.
———. 2007. Structures of Agency. New York: Oxford University Press.
Eshleman, Andrew. 2004. Responsibility for Character. Philosophical Topics 32: 65–94.
Fischer, John M. 2002. Frankfurt-Style Compatibilism. In Contours of Agency: Essays on Themes
from Harry Frankfurt, ed. S. Buss and L. Overton, 1–26. Cambridge, MA: MIT Press.
———. 2006. My Way: Essays on Moral Responsibility. New York: Oxford University Press.
———. 2010. Responsibility and Autonomy. In A Companion to the Philosophy of Action, ed.
T. O’Connor and C. Sandis, 309–316. Oxford: Wiley/Blackwell.

18
I have benefitted from very helpful comments on previous versions of this paper by Justin Coates
and Benjamin Mitchell-Yellin. Also, I delivered the paper at the conference on “Thick (Concepts
of) Autonomy” at the University of Münster in October, 2012, and I am especially grateful to help-
ful discussions on that occasion with Michael Quante, Alfred Mele, James Taylor, and John
Christman.
16 J. M. Fischer

———. 2012. Responsibility and Autonomy: The Problem of Mission Creep. Philosophical
Issues 22, Action Theory: 165–184.
Frankfurt, Harry G. 1969. Alternate Possibilities and Moral Responsibility. Journal of Philosophy
66: 829–839.
———. 1971. Freedom of the Will and the Concept of a Person. Journal of Philosophy 68: 5–20.
———. 1976. Identification and Externality. In The Identities of Person, ed. A.O. Rorty, 239–251.
Berkeley: University of California Press.
———. 1982. The Importance of What We Care About. Synthese 53: 252–272.
———. 1987. Identification and Wholeheartedness. In Responsibility, Character, and Emotions,
ed. F.D. Shoeman, 27–45. Cambridge: Cambridge University Press; reprinted in Perspectives
on Moral Responsibility, eds. J. M. Fischer and M. Ravizza, 170–187. 1993. Ithaca: Cornell
University Press.
———. 1988. The Importance of What We Care About. Cambridge: Cambridge University Press.
———. 1999a. On Caring: I. Caring and Necessity. In Necessity, Volition, and Love, deliv-
ered as part of the Kant Lectures at Stanford University, 155–166. Cambridge: Cambridge
University Press.
———. 1999b. On Caring: II. The Necessities of Love. In Necessity, Volition, and Love, deliv-
ered as part of the Kant Lectures at Stanford University, 167–180. Cambridge: Cambridge
University Press.
———. 2002. Reply to John Martin Fischer [“Frankfurt-Style Compatibilism”]. In Contours
of Agency: Essays on Themes from Harry Frankfurt, ed. S. Buss and L. Overton, 27–31.
Cambridge, MA: MIT Press.
———. 2004. The Reasons of Love. Princeton/Oxford: Princeton University Press.
———. 2006. In Taking Ourselves Seriously and Getting It Right, The Tanner Lectures in Moral
Philosophy, ed. D. Satz. Stanford: Stanford University Press.
Mitchell-Yellin, Benjamin. 2012. Self-Governance, Moral Responsibility, and Weakness of the Will:
In Defense of the Platonic Model. PhD Dissertation, Department of Philosophy, University of
California, Riverside.
Taylor, Charles. 1976. Responsibility for Self. In The Identities of Persons, ed. A.O. Rorty,
281–299. Berkeley: University of California Press.
Velleman, J. David. 1992. What Happens When Someone Acts? Mind 101: 461–481; reprinted
in Perspectives on Moral Responsibility, eds. J. M. Fischer and M. Ravizza, 188–210. 1993.
Ithaca: Cornell University Press.
Watson, Gary. 1975. Free Agency. Journal of Philosophy 72: 205–220; reprinted in Moral
Responsibility, ed. J. M. Fischer, 81–96. 1986. Ithaca: Cornell University Press.
———. 1977. Skepticism About Weakness of the Will. Philosophical Review 86: 316–339.
———. 1987. Free Action and Free Will. Mind 96: 145–172.
———. 1999. Disordered Appetites. In Addiction: Entries and Exits, ed. J. Elster, 3–28. New York:
Russell Sage Publications.
———. 2003. The Work of the Will. In Weakness of Will and Practical Irrationality, ed. S. Stroud
and C. Tappolet, 172–200. Oxford: Oxford University Press.
Wikipedia. 2011. Mission Creep. Last modified April 29, 2020. http://en.wikipedia.org/wiki/
Mission_creepm. Accessed 1 June 2011.
Chapter 2
The Passivity of Self-Satisfaction:
A Critical Re-appraisal of Harry
Frankfurt’s Normatively Thin Ontology
of Autonomy

Joel Anderson

Abstract This chapter attempts to “re-boot” the discussion of Harry Frankfurt’s

approach to autonomy, in the service of a new diagnosis of the strengths and weak-
nesses of his satisfaction-based ontology of the will. Criticisms of Frankfurt’s work
have tended to focus on a lack of normative foundations, often missing Frankfurt’s
aim of shifting discussions of autonomy towards a focus on avoiding passivity in
how one cares about what one cares about, while still acknowledging the central
role of volitional necessity and, especially, self-satisfaction in autonomous agency.
Although this approach provides Frankfurt with interesting ways of deflecting criti-
cisms that are ethical in character, the lack of attention to the hermeneutic and lin-
guistic dimensions of our inner life, leaves his approach with insufficient grounds
for identifying the passivity of those who are unable or unwilling to engage in genu-
ine self-interpretation.

2.1 Re-booting the Debate Over Second-Order Desires

There is a familiar depiction of the development of Harry Frankfurt’s approach to

autonomy that goes something like this. Frankfurt started out arguing that one is
autonomous only if one has reflected critically on one’s desires and figured out
which desires one really wants to have. But then it was pointed out that these
“second-­order desires” of self-assessment are themselves mere desires, and thus
also in need of further assessment – “third-order desires” – and a regress of justifica-
tion ensues. According to the standard story, Frankfurt struggled with this regress
problem and then ended up settling for a normatively anemic view, according to
which all that matters is the coherence and harmony within the will. This turn to

J. Anderson (*)
Utrecht University, Utrecht, The Netherlands
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 17

Switzerland AG 2022
J. F. Childress, M. Quante (eds.), Thick (Concepts of) Autonomy, Philosophical
Studies Series 146, https://doi.org/10.1007/978-3-030-80991-1_2
18 J. Anderson

wholeheartedness, it is then often argued, is often thought of as deeply problematic.

Some view too strong a requirement, in the sense that it seems to rule out a variety
of perfectly acceptable forms of ambivalence. For others, the problem is that it is not
strongly critical enough, in that it is value-neutral approach that can’t rule out the
possibility that one might be wholeheartedly evil.1
Unfortunately, this way of framing the discussion of the approach misses much
of Frankfurt’s originality, and he has spent much of the last 40 years trying to make
clear that his aim is not to defend a moral theory of the person or of autonomy.
However rich his account of the person and the will, it is intended to be normatively
thin. With regard to the “regress objection,” Frankfurt made the point himself
already in 1976. As he put it later, “The mere fact that one desire occupies a higher
level than another in the hierarchy seems plainly insufficient to endow it with
greater authority or with any constitutive legitimacy” (Frankfurt 1988d, 166–7).2
Clearly, the view he has been promoting in not simply one of a nested hierarchy of
attitudes.
Moreover, the moralistic objections to Frankfurt’s approach similarly miss their
target. I recall Frankfurt saying, in a discussion after delivering a paper (“The
Faintest Passion”) at Northwestern University in 1992, that he is not making claims
about morality but about “the ontology of the will” and that, from that perspective,
there is no reason to expect wholeheartedness to serve as a moral criterion. As he
put it then, “Stormtroopers can be wholehearted.” Or, as he puts it more recently,
“Being wholehearted is quite compatible….with being dreadfully and irredeemably
wicked” (Frankfurt 2004, 98). For many, this is rather too thin, in terms of normativ-
ity. For Frankfurt, this neutrality regarding the autonomous will is not a bug; it’s a
feature.
In light of these misunderstandings, then, it is time for what Hollywood calls
a “re-boot”.3 It’s time to go back to the core motivating problem that Frankfurt is
concerned with, to reconsider his proposed solution to the problem that he is
actually addressing, and then – but only then – to point out where it runs aground.
The point is not to search through the archives of the early work of Frankfurt, but
to take a fresh look at the core ideas, without the obsession with the regress
problem.

1
The concern with moral rectitude and the regress of justification is was central to early discus-
sions of Frankfurt’s position, including (Benson 1987), (Christman 1991), (Friedman 1986),
(Watson 1975), and the more recent (Wolf 2002).
2
This is already clear in the 1976 essay, (Frankfurt 1988c, 58–68, esp. pp.65–6).
3
The concept of a “re-boot” is used to describe a recent trend in comic books and film or TV series,
in which goes back to the beginning of a serial development and makes a fresh start. See http://
en.wikipedia.org/wiki/Reboot_(fiction)
2 The Passivity of Self-Satisfaction: A Critical Re-appraisal of Harry Frankfurt’s… 19

2.2 The Problem Our Desires Pose

One of the fascinating things about animals and very young children is that they are
not troubled by their desires. There is an appealing purity to the way in which they
pursue what they want. They may strategize about how best to realize their goals or
adjust their plans when these efforts are frustrated. But they don’t worry about their
desires or question their goals.
This is not a purity that is available to us as persons. It is a defining, constitutive
characteristic of persons that we can and do ask ourselves whether our desires,
urges, cravings, etc. deserve to be heeded. Often we don’t stop to think about our
desires, and often we are perfectly happy with the desires we have. But not only is
it the case that we can and do question our desires, it is something that we expect of
one another. It is on these grounds that Frankfurt famously distinguishes “persons”
from what he calls “wantons,” creatures who don’t really care what they desire
(Frankfurt 1988b). His central idea is that the capacity to be troubled by these con-
flicts is a necessary feature of the ontology of the will of persons.
There is admittedly something distasteful about the whole business of dividing
“genuine persons” from “mere wantons”, animals, or whatever. We can easily get
wrapped up in debates about what the best terms are here, or about what makes a
creature “really human”. But I think Frankfurt is correct to say that there is a funda-
mental watershed that exists between those creatures who ask the question of
whether their desires are undesirable and those who don’t. In what follows, I will
continue to refer to the first set of creatures by employing the technical philosophi-
cal term, “we”.
For us, then, desires (like beliefs) can and do pose questions. Frankfurt recently
put this in a particularly stark fashion:
[T]he mere fact that a person has a desire does not give him a reason. What it gives him is a
problem. He has the problem of whether to identify with the desire and thus validate it as
eligible for satisfaction, or whether to dissociate himself from it, treat it as categorically
unacceptable, and try to suppress it or rid himself of it entirely (Frankfurt 2006: 11, empha-
sis added).

Our desires raise issues that we must resolve. But the option of solving the problem
by simply treating all our desires as automatically reasons is not an option that is
available to us – not to anyone who is still one of us. There is no way back. We must
move forward. Call this “the predicament of volitional reflexivity.”
The predicament of volitional reflexivity is a bit like the Garden of Eden after the
Fall. Having eaten from the tree of critical self-knowledge, our situation is trans-
formed, and we must find a way to move forward. Seen this way, Frankfurt’s philo-
sophical work is, at a most fundamental level, about the question of what it means
to handle this predicament in a mature way. Whether one refers to it as “maturity”,
“genuine personhood,” “the highest good,” “freedom of the will,” or “autonomy,”
the primary normative goal of Frankfurt’s writings is to develop an account of what
can count as a successful response to the predicament of volitional reflexivity – a
predicament that emerges from the internal structure of the will.
20 J. Anderson

Implicit in Frankfurt’s writings is the claim that the predicament is a general

phenomenon, part of the human condition. But it presents itself in specific cases.
Suppose you find yourself increasingly drawn to play a more active role in politics.
Suppose, in particular, that you find yourself with a strong desire to run for local
political office. The predicament is that you have to figure out what to do with this
desire. Do you really stand behind this desire, or is it a distraction from what you
really care about, something that snuck up on you via a flattering remark from a
neighbor or a wave of panic about a current political crisis? Engaging with these
questions – “taking ourselves seriously” as Frankfurt puts it in the title of his Tanner
Lectures – involves taking a stand with regard to one’s will and facing the predica-
ment squarely.

2.3  assive Failures to Face the Predicament

P
of Volitional Reflexivity

Before continuing, it is important to get clear on what is so problematic about fail-

ing to face the predicament. Why should we let our desires bother us? For Frankfurt,
the answer lies in understanding how such refusals represent the passivity that char-
acterizes non-persons (or “wantons”).
The most extreme form of passivity is not to care at all what desires one has and
what one’s will is like. As we saw earlier, this degree of passivity puts one under the
threshold of full personhood, on Frankfurt’s account. There are interesting debates
to be had here, about the nature of personhood, but my concern here is with cases in
which we are passive about inner conflicts, cases in which there is a real danger of
remaining a mere spectator to the inner struggles between our desires.4 Take the
example of someone who has a strong desire to retaliate when slighted. Suppose he
receives an email circulated to his department at work, in which it is suggested that
he was responsible for a situation in which the entire department was embarrassed.
Suppose this gives rise, as it so often does, to two competing inclinations: to lash out
with counter-accusations in an email CC’ed to the entire department, or to demon-
strate to his skeptical colleagues that he is, in fact, capable of acting like a grownup.
To be passive here is not really care how this internal conflict plays out. It is, in that
sense, to be a bystander.
It is important to emphasize that overcoming passivity, in Frankfurt’s sense, does
not guarantee that the rejected desire will disappear or cease to both one (Frankfurt
1988d, 175). Rather, when we take an active stand regarding which side of the con-
flict we want to win – in Frankfurt’s language, when we “identify” with one side –
we make it the case that the failure of that desire to win the conflict would be felt as

4
It is worth noting that Frankfurt here departs from much of contemporary empirical psychology,
as well as the work of philosophers such as Daniel Dennett, builds on the idea that we are, in fact,
bystanders to the process in which action is generated as the result of a competition between dif-
ferent subpersonal agents (e.g., Dennett 2003).
2 The Passivity of Self-Satisfaction: A Critical Re-appraisal of Harry Frankfurt’s… 21

a loss. The conflict is not thereby terminated, but its character is fundamentally
altered. Frankfurt’s way of putting this is to say that the conflict ceases to be a con-
flict within oneself. For the conflict can now be described as being one between
“…what the person really wants and other desires…that are external to the voli-
tional complex with which the person identifies and by which he wants his behavior
to be determined” (Frankfurt 1988d, 165). What this means is that although one is
no longer torn or divided or confused about what one wants, one may still find one-
self continuing to have both of the conflicting desires. It is thus not Frankfurt’s
claim that the wholehearted person has no internal conflicts. The desire to send the
spiteful email may continue to pester and fester. In wholeheartedly rejecting one
desire and endorsing the other, what one resolves is not the conflict but oneself
(Frankfurt 1988d, 100; Frankfurt 1998b, 164).
Often, however, we struggle with which stand to take, and these uncertainties
raise deeper or higher-order questions with regard to which we must also take a
stand. And here, too, like bystanders who unsure of how to respond to a crime in
progress, we can end up rendered passive by our volitional incoherence.
In Frankfurt’s writings, one can distinguish between two ways in which voli-
tional incoherence can threaten the self. On the one hand, there is ambivalence, in
which one is torn between committing oneself and not committing oneself to a
given person, project, or ideal; on the other hand, there is the paralyzing disorienta-
tion of being pulled in so many different directions that one does not know what one
wants. Both forms of volitional incoherence undermine one’s autonomy, but in dif-
ferent ways.
Ambivalence is a matter of one’s being at cross-purposes at the highest level of
one’s reflective endorsements (Frankfurt 1988d; Frankfurt 1998b; Frankfurt 2004;
for a discussion, see (Swindell 2010) and (Velleman 2002)).
For example, someone considering running for political office could be deeply
and authentically committed to improving society and, equally, to sparing his family
the added strains of his taking on extra responsibilities, but circumstances make
these two commitments incompatible. In cases of ambivalence, the agent is seri-
ously divided with regard to this internal conflict. Frankfurt describes the phenom-
enon this way:
The disunity of an ambivalent person’s will prevents him from effectively pursuing and
satisfactorily attaining his goals. Like conflict within reason, volitional conflict leads to
self-betrayal and self-defeat. The trouble is in each case the same: a sort of incoherent
greed – trying to have things both ways – which naturally makes it impossible to get any-
where. The flow of volitional or of intellectual activity is interrupted and reversed; move-
ment in any direction is truncated and turned back. However a person starts out how to
decide or to think, he finds that he is getting in his own way (Frankfurt 1998b, 99).

Frankfurt not denying here, I take it, that it may be essential for psychological health
to be able to acknowledge conflicting feelings or to grieve the losses incurred even
in making the right decision.5 The point is simply that, if we are to move forward,

5
Here I am in agreement with critics such as Rössler (Rössler 2009) who argue that Frankfurt’s
concept of ‘wholeheartedness’ is incompatible with this sort of awareness of mixed feelings. In
22 J. Anderson

we must not be passive bystanders but must face up to the predicament of being
aware of the incoherencies in our will.
There is also a second form of volitional incoherence about which one might be
passive, something we could call “disorientation”. Whereas the phenomenology of
ambivalence is one of being pulled in two directions, the feeling associated with
disorientation is of being aimless. Frequently, according to Frankfurt, this is the
result of having more options and few constraints:
…[Someone with more options] may become, to a greater or lesser degree, disoriented with
respect to where his interests and preferences lie. Instead of finding that the scope and vigor
of his autonomy are augmented as the ranges of choices open to him broadens, he may
become volitionally debilitated by an increasing uncertainty both concerning how to make
decisions and concerning what to choose (Frankfurt 1998a, 109; see also Frankfurt 1998b,
101–03; Dworkin 1988).

Disorientation involves a lack of “flow” and vigor, which leaves one feeling frus-
trated or just apathetic, but also usually means that one, here again, gets in one’s
own way. These negative characteristics add to the rhetorical force of Frankfurt’s
discussion, but the core argument I believe is again one about passivity. The prob-
lem is not that one lacks some objectively good degree of integration but that one is
negligently passive about the fact that one’s volitional makeup leaves one incapable
of resolving the conflict effectively. Seen this way, I think it is a mistake to interpret
Frankfurt as obsessed with volitional coherence. His reasoning is ultimately rather
instrumental: the threat posed by disorientation (and other forms of volitional inco-
herence) lies in the way it leaves one unable to resolve conflicts. Failure to figure out
where one stands thus involves, indirectly, negligent passivity with regard to inter-
nal conflicts, since we need to know where to stand to know how to resolve them.
And it passivity about internal conflicts, I am arguing, that constitutes the basis of
Frankfurt’s implicit claims about the structure of a good will.
Ambivalence and disorientation both represent, then, forms of volitional inco-
herence that undermine a person’s active, purposeful agency. Frankfurt’s way of
making his point is typically in terms the phenomenology of its opposite: we feel
most freely self-directing, he suggests, when our minds are made up, when we are
so decisively resolved to follow a certain course of action that nothing distracts us
from the flow of our activity. In this sense, making up one’s mind has a liberating
effect (Frankfurt 1988e, 88; Frankfurt 1998a, 110). But his claim has less to do with
how one experiences oneself than with the objective character of one’s will. When
we take a stand, we may still be engaged in a struggle with desires, temptations, and
so, but we are no longer passive bystanders.

response to a similar objection from David Velleman, Frankfurt writes: “Resolving ambivalence by
taking a decisive stand against certain feelings does not require (or even permit) that a person
misrepresent those feelings or that he conceal them from himself” (Frankurt 2002c, 126).
2 The Passivity of Self-Satisfaction: A Critical Re-appraisal of Harry Frankfurt’s… 23

2.4  acing the Predicament and Resolving One’s Will:

F
Structural Hierarchicalism of the Will

In general, then, Frankfurt’s position is that one can avoid being a passive bystander
towards one’s desires and, more broadly, towards the health of one’s will by taking
a stand with regard to one’s will. By itself, this can sound rather arbitrary, and
indeed, there was a decisionistic strain in Frankfurt’s early claim that what one
needs most in identifying with or rejecting one’s desires is to do so decisively
(Frankfurt 1988b, 21). Since a position like that is unable to rule out cases of people
being decisive from moment to moment, and yet flip-flopping and vacillating con-
stantly over time, Frankfurt has moved towards a position that emphasizes the ways
in which the structure of one’s will anchors one’s actions, feelings, responses,
and so on.
As I noted at the outset, however, having the right structure, for Frankfurt, is not
a matter of tracking some independent reality of a true self or even of moral values.
His normatively thin approach centers on structural features of the will. He fre-
quently uses the term “wholeheartedness” for the requisite state of the will, in which
one responds effectively to the predicament of volitional reflexivity. But my sense is
that what has emerged most recently as his preferred language for talking about this
is the language of “satisfaction” with oneself. Since self-satisfaction is rarely treated
as the hallmark of a mature, autonomous person, this requires some explanation.
What we are looking for, according to Frankfurt, is a state of the will in which
one has actively taken a stand on the inclinations one finds oneself with, but where
this stand is not arbitrary or fickle or otherwise inherently unstable. This is the ideal
towards which persons strive when they respond to predicaments. It is, in his view,
the only way to resolve the predicament with which choice presents us.
Avoiding arbitrariness requires that one’s identification with a desire be accom-
panied by a certain degree of integration and equilibrium within the volitional
makeup of the person, and the term “satisfaction” is actually not such a strange
choice here:
Hierarchical accounts of the identity of the self do not presume, however, that a person’s
identification with some desire consists simply in the fact that he has a higher-order desire
by which the first desire is endorsed. The endorsing higher-order desire must be, in addi-
tion, a desire with which the person is satisfied (Frankfurt 1998b, 105).

The central requirement of this condition of satisfaction is that one be free from
uneasiness about this identification, that one have no particular inclination to recon-
sider the identification. Being satisfied with endorsing a desire to perform a particu-
lar action is compatible with the counterfactual claim that, had one endorsed a
desire not to perform that action, one would have been satisfied with that (Frankfurt
1998b, 103–04). Crucial to this argument is the further claim that not going to a yet-­
higher level of reflection is not a matter of refusing to do so. As a deliberate act or
decision, such a refusal would be open to the suspicion that it might be motivated by
a desire that one would not endorse, and trying to establish the authority or whole-
heartedness of this refusal itself would spawn an endless series of further
24 J. Anderson

reflections. But one might also simply be satisfied with one’s identification with a
given first-order desire. And this, Frankfurt argues, avoids the problem entirely:
Satisfaction with one’s self requires, then, no adoption of any cognitive, attitudinal, affec-
tive, or intentional stance. It does not require the performance of a particular act; and it also
does not require any deliberate abstention. Satisfaction is a state of the entire psychic sys-
tem – a state constituted just by the absence of any tendency or inclination to alter its condi-
tion” (Frankfurt 1998b, 104).

More recently, has articulated his position regarding self-satisfaction, in connection

with Spinoza’s claim that the highest good is “acquiescentia in se ipso.” (Frankfurt
2006, 17). In answer to the question, why “acquiescing to oneself” is the highest
thing one can hope for, Frankfurt says this:
Perhaps because it resolves the deepest problem. In our transition beyond naïve animality,
we separate from ourselves and disrupt our original unreflective spontaneity. This puts us at
risk to varieties of inner fragmentation, dissonance, and disorder. Accepting ourselves rees-
tablishes the wholeness that was undermined by our elementary constitutive maneuvers of
division and distancing. When we are acquiescent to ourselves, or willing freely, there is no
conflict within the structure of our motivations and desires. We have successfully negotiated
our distinctively human complexity. The unity of our self has been restored (Frankfurt
2006, 18).

This, it seems to me, is as close to a definitive statement of Frankfurt core position

as one can find in any of his writings. There is the idea of what I am calling the
“predicament,” which is clearly part of the human condition. It is resolved at a level
that is both volitional and, as he puts it in the next sentence, “purely structural”. And
there is the equation of “willing freely” with not being internally divided. Self-­
satisfaction and self-acceptance thus represent a mature way of facing the predica-
ment that our desires pose.

2.5 How Critical Must Volitional Reflexivity Be?

At this point the key question arises regarding the implications of this normatively
thin ontology of the will: is it not possible that people are too easily satisfied with
themselves? Isn’t this the ultimate form of passivity? What ever happened to the
idea that autonomous agency is a matter of reflecting critically on one’s first-order
desires?
In a way, however, this line of objection misunderstands Frankfurt’s project. As
I have been emphasizing throughout, what he is offering is an account of how we
can successfully respond to the predicament that arises from our natural inclination
to be critical. Self-critical doubt is thus the problem to be solved, and the question
for Frankfurt is how one can solve it while still being a person who actively cares
about his will.
The notion of self-satisfaction actually captures the approach quite well. As
Frankfurt argues, “Being satisfied with the way things are is unmistakably an excel-
lent reason for having no great interest in changing them” (Frankfurt 1988a, 155). It
2 The Passivity of Self-Satisfaction: A Critical Re-appraisal of Harry Frankfurt’s… 25

is, moreover, a normatively thin and value-neutral reason, based on a formal, struc-
tural ontology of the will of persons. Thus, in a situation in which one is genuinely
settled on what one cares about, desires, and loves, one has successfully faced the
questions raised by having those cares, desires, and loves. In that regard, more criti-
cal reflection contributes nothing further. Despite his emphasis on the capacity for
reflection as a condition for personhood, Frankfurt is deeply concerned with the
way in which demands for critical reflection can run amok and end up blocking
individuals’ capacity ever to resolve to do anything.
This is where it becomes clear that the ideal of mature agency that Frankfurt aims
to capture with “self-satisfaction” is closely tied to the notion of confidence. Indeed,
he concludes his Tanner Lectures with the claim that it is confidence – rather than,
one might add, intersubjective confirmation – that “provides us with the deepest and
most secure foundation for practical reason” (Frankfurt 2006, 52). But this is an old
theme for Frankfurt. In the spirit of a “re-boot,” it is worth recalling Frankfurt’s
origin story as a Descartes scholar and the interpretation of Cartesian epistemologi-
cal foundations in terms of a feeling of confidence:
Descartes cares less about the correspondence of his beliefs to “reality” than he does about
their permanence and constancy. What he wishes above all to avoid is not error, in the sense
of non-correspondence, but betrayal. What might be found out to be false is what he wishes
to guard against. If a belief can confidently be expected to remain unshaken by any further
inquiry, that is all the truth he cares to demand” (Frankfurt 1970, 180).6

At this point, a critic could challenge Frankfurt on the cogency of grounding epis-
temic warrant in a feeling of confidence. Indeed, I have done so elsewhere (Anderson
2003). Given the prevalence of over confidence, after all, felt confidence would
appear to provide appropriate grounding for judgments on how to live only when it
is warranted confidence. Ultimately, I think that following this question (as to the
justification for trusting our feelings of confidence) will reveal the limits of a purely
subjective and individualistic approach such as Frankfurt’s. Without public criteria
or procedures that transcend subjective consciousness, it is impossible to make a
principled distinction between what seems to be justified and what is justified.
But this is not the line of criticism that I wish to pursue here. Instead, I want to
identify a different sort of dissatisfaction that I have with Frankfurt’s way of fram-
ing the predicament of volitional reflexivity.

6
See also Frankfurt, (Frankfurt 1988d, 169), where Frankfurt mentions that his understanding of
unreserved commitment or identification “owes much to Descartes’s discussion of clear and dis-
tinct perception.” For a fuller discussion, see (Anderson 1996, ch. 3.2), and (Herman 2002, 259),
together with Frankfurt’s reply (Frankurt 2002a, 275–78).
26 J. Anderson

2.6  elf-Opacity and the Importance of Understanding What

S
We Care About

Suppose I can respond confidently to a predicament of volitional reflexivity. The

“problem” posed by my desires is resolved wholeheartedly by my taking a stand on
whether I identify with or reject these desires. I could, in principle, continue to
probe whether I really care about what I think I care about, but if I am satisfied with
where I stand, it’s not clear what the point of doing that would be. My life may not
work out as I hope, but at least I can get on with it.
That, then, is Frankfurt’s basic picture of a mature and healthy will, and as I have
suggested throughout, it is not easily dismissed. Nevertheless, I do believe that there
is something missing. The problem, in short, is that Frankfurt overlooks the extent
to which resolving the problem of volitional reflexivity involves making sense of the
position one takes. For someone might be wholeheartedly and confident and yet not
really understand her desires. One might be satisfied with oneself but opaque to
oneself. To illustrate the sort of wholehearted self-opacity that I have in mind, con-
sider the following case, from a National Public Radio profile of the late Rev. Robert
Shields of Dayton, Washington (Isay 1994, accessible at https://storycorps.
org/stories/worlds-longest-diary/).
From 1972 to 1997, Shields kept an extraordinarily meticulous diary, typing up
the minute-by-minute occurrences in his life. It totalled 37.5 million words, and the
detail is overwhelming. At periodic intervals during the day, Shields recorded how
many ounces of juice he drank at breakfast, what the morning newspaper weighed,
what he read, when he went to the bathroom, what the barometer reading was, and
how much time he spent speed-typing his journal. Even in retirement, his journal
kept him very busy. Indeed, the diary usually took up so much time that he had no
time to read it – and apparently little interest in doing so. In the radio interview,
there is a rushed tone to his voice, as he keeps typing rather than taking time to
answer the interviewer’s questions. At one point in that interview, the following
exchange occurs:
Isay: “Why are you doing this?…What are you trying to do?”
Shields: “I don’t know. I don’t really – I really can’t answer that. [He reads the text
aloud as he continues typing] Five-forty-five to 6:15 I read more from the Oxford Dictionary
of Quotations. I ate half a dozen large Archway sugar cookies while I drank two cups of
milk.” …..
Isay: “How does your family feel about this?”
Shields: “Never asked them.” [laughs]
Isay: “What about leaving town?”
Shields: “I don’t leave town. I haven’t left town since 1985, to visit my brother in
Tennessee. I don’t like to be away overnight because it gets me behind. If I travel to Walla
Walla to do shopping, it puts me behind in the diary. I have to take notes all the time, and…it
takes almost a day to catch up with the notes. So I avoid going out. I avoid being away. [He
begins typing again.] Three-O-five to 3:30, I read the Tri-City Herald….”

In terms of Frankfurt’s criteria, Rev. Shields is a model of wholehearted caring. He

explicitly endorsed his desire to continue with the diary. And he obviously cared
2 The Passivity of Self-Satisfaction: A Critical Re-appraisal of Harry Frankfurt’s… 27

about it. Indeed, he had a half-dozen typewriters set up to guard against the loss that
would otherwise threaten were one of the them to break down. And his aversion to
leaving town is grounded in the fact that he did not want to betray his project or
himself.
Yet Rev. Shields had very little idea why he was doing what he was doing, and
this is the point of the example for my purposes. When pressed, during the inter-
view, to explain why he keeps the diary, he draws a blank. He does not really under-
stand his own actions. He is unable to provide even a minimally plausible account
of what he is up to, incapable of rendering his way of life – and the choices that
reproduce it – intelligible. Although it is possible that he was just shy, he appears to
have been unable to make sense of his life, even to himself. And then it is hard to see
how he wasn’t somehow failing to face fully the predicament of volitional reflexiv-
ity. His failure to make sense of himself illustrates yet another form of passivity.
At this point, in the concluding two sections, I would like to take up two ways in
which Frankfurt could respond to the charge that his account overlooks the impor-
tance of understanding oneself. One response is to say that intelligibility does mat-
ter, but that those who are wholehearted have all the intelligibility that is required
for active personhood. Alternatively, one can deny that the predicament of volitional
reflexivity requires intelligibility at all. In the remainder of this essay, I will try to
say briefly where I think that both of these responses run aground.

2.7  rankfurt’s Hermeneutically Impoverished Ontology

F
of the Will

The first response would be for Frankfurt to argue that there is something like “voli-
tional intelligibility” and that this is sufficient for an agent to count as being able to
make sense of his responses. To develop this line of argument, one could start with
David Velleman’s discussion of intelligibility as a necessary condition for action.
After all, we continue acting only if we understand what we are going to do. If, to
use his example, you are walking up Fifth Avenue and you suddenly realize you
don’t know where you are going, this pulls you up short (Velleman 1989, 15–17). In
Frankfurt’s language, you don’t know where you stand with regard to where you
want to go. Building on this idea, one could defend Frankfurt’s approach by saying
that he has a volitional understanding of “intelligibility.” As Frankfurt says in
another context, “the fact that someone cares about something, or that something is
important to him, means just that the person is disposed to behave in certain ways”
(Frankurt 2002b, 161). What we get, then, is an account of what it means for our
actions to “make sense” to us that is cashed out in terms of its covariance with a
confident, wholehearted exercise of the will. In this sense, it is sufficient evidence
for Rev. Shields’s diary project being intelligible to him that he pursued it whole-
heartedly, which he most certainly did. As Frankfurt would likely say, if he was
28 J. Anderson

fully satisfied with his caring, it’s unclear what more he could have reasonably
needed in the way of evidence for thinking that it made sense for him to care.
In effect, however, this response involves answering the objection by deflating
the notion of “intelligibility” beyond recognition. After all, the starting intuition of
the objection was that someone like Rev. Shields might be perfectly wholehearted
but lack any way of fitting his commitments into a larger context in which it makes
sense. Mere behavior cannot provide sense of “fitting in”, because bodily move-
ments by themselves make no reference to a larger constellation of meaning. The
interesting question is whether this sense of fitting in can be provided by the reflex-
ivity of the will central to Frankfurt’s account, according to which caring about
something entails, for example, a commitment to ensuring that one’s will is so struc-
tured that one continues to care. But here Frankfurt seems to face a dilemma. On the
one hand, he could stick to the largely dispositional account, in which case he has to
deny any necessary role for the inferential links between possible courses of action.
Seen this way, Rev. Shields would count as “understanding” what he was up to, even
though he had nothing to say about how the various things he cared about fit together.
On the other hand, Frankfurt could allow that caring involves a sensitivity to these
inferential connections, such that Shields’s caring about updating his diary needs to
fit with his response to specific situations as well as his caring about other connec-
tions. Frankfurt could indeed argue that, as a matter of fact, we won’t be satisfied,
confident, and wholehearted if we aren’t on the attuned to the discordance between
different things we care about. Seen in this way, Frankfurt could say that his theory
does not actually force him to deny that there are problems with the Rev. Shields’s
volitional coherence, for Shields may have been rigidly cutting himself off from
dimensions of his caring and thus revealing himself to be a “wanton” with regard to
those dimensions. That said, it’s entirely unclear how this emphasis on self-­
understanding, much as I would welcome it, is something that Frankfurt could inte-
grate into this theory. For I see no way of developing such an approach without
making judgment and reasons-responsiveness a central part of a richer conception
(and thus requirement) of self-understanding. In other words, what is needed is a
thicker conception of the forms of hermeneutic agency that are necessary for being
able to make sense of even our most mundane commitments. And that is something
that Frankfurt has consistently viewed as unnecessary to a healthy, mature, confi-
dent will.

2.8 The Costs of Frankfurt’s Commitments to Thinness

Alternatively, Frankfurt could defend his normatively thin ontology of the will by
rejecting the requirement of intelligibility. However important intelligibility might
be to some people, he could argue, it is not necessary for the kind of volitional
coherence that concerns us here. It is, thus, an optional part of a mature response to
the predicament of volitional reflexivity – and perhaps a part that reflects a particu-
lar, substantive value-commitment on the part of members of cultural elites. On a
2 The Passivity of Self-Satisfaction: A Critical Re-appraisal of Harry Frankfurt’s… 29

parallel with Frankfurt’s separation of questions of ethics from questions of caring,

one could say that, if you care about having a deep and articulate understanding of
your motives, that’s fine; it may help you win friends and influence people; but it
would be a mistake to see self-understanding and wholeheartedness as linked in any
essential way.
That is, I think, the most consistent position for Frankfurt to take, since it fits
with his general suspicions regarding the normative overburdening of concepts
employed in discussions of how to lead one’s life. Frankfurt effectively and repeat-
edly puts the burden of proof on those who would say that personhood, autonomy,
genuine caring, or what have you somehow requires “more”. And, while I can
appreciate the wisdom of his caution against an endless pursuit of “the better” at the
cost of “the good,” I believe that his culture-critic’s aversion to spoiled adults with
limitless expectations (Frankfurt 1982, 319) has distracted him from the importance
of avoiding the form of passivity involved is acquiescing to self-opacity, to not
understanding oneself.
The passivity I have in mind is a form of disregard that comes into view when
one takes a broader view of the way in which our inner life engages us when we
reflect. Frankfurt operates on the assumption that the mode in which desires present
themselves as requiring a response is the mode of a problem that calls for volitional
resolution. In this sense, the predicament of reflexivity we face as persons is to be
seen in terms of problems to be solved, in the way a pain in one’s back calls for
adjusting how one sits. But there is actually a much wider range of ways in which
our inner life calls for a response. Indeed, part of what is distinctive about the pre-
dicament of volitional reflexivity that we face is this. We find our desires, emotions,
aversions, and so on not just troubling and unsettling but perplexing. It is not merely
problems that our desires pose; they also pose questions.7 And the grammar of
answering questions adequately is different from that of resolving problems ade-
quately. In particular, the success conditions are different. We know when volitional
problems are resolved; we can feel it. But questions are more complex, and, although
there are better answers and worse answers, there is in principle always more to say.
This, it seems to me, is why Frankfurt’s model of mature, confident, whole-
hearted self-satisfaction is – despite its nuance, elegance, and prudence – ultimately
rather unsatisfying. It fails to acknowledge the way in which our inner life presents
us with ongoing, open-ended possibilities for self-interpretation, and thus with the
possibility of a richer and deeper form of selfhood. We may well have reasons not
to pursue all these possibilities, but Frankfurt’s commitment to thinness forces him
to underestimate the importance of a form of openness that is crucial to avoiding
passivity. Frankfurt’s satisficing emphasis on volitional stability leaves him with
little to say about the form of active agency that is required for being alert to rigidity
and receptive to transformation. Frankfurt has a point in warning against a

7
This is a point that Charles Taylor has been particularly adept at expression: “We [as opposed to
other animals] experience our pre-articulate emotions as perplexing, as raising a question” (Taylor
1985, 74). Regarding the complexities of the ongoing task of finding an adequate answer to the
questions posed, see especially (Taylor 1989, ch. 1–4).
30 J. Anderson

compulsive, constant reexamination of all one’s feelings all the time, chasing after
the next shiny possibility. But avoiding that extreme does not require adopting the
approach that Frankfurt seems to take, in which we wait, like passive bystanders, for
internal conflicts to arise. An active engagement with one’s desires and opportuni-
ties, with the questions they pose, and with the perspectives of those with whom we
share our lives not only fits naturally with mature and robust ontology of the will,
but it may even be required for the will’s dynamic character.
My diagnosis of where Frankfurt’s approach comes up short here is that it
neglects language and meaning (as well as experimentation and exploration) in its
characterization of the predicament that arises when – often enough, because of our
engagement with the world and others in it – we turn inward. We can, of course,
decide to leave these issues of meaning and self-interpretation to one side. If we do
so, we might well be able to live less troubled and more volitionally robust lives. But
we will lack what I have elsewhere called the hermeneutic or disclosive dimension
of autonomy (Anderson 2014, 149–50). We will be bystanders to these deeper,
language-­dependent dimensions of our lives. In refusing to engage actively with this
further dimension of the predicament of volitional reflexivity, we will be, in that
sense, wantons.8

References

Anderson, Joel. 1996. A Social Conception of Personal Autonomy: Volitional Identity, Strong
Evaluation, and Intersubjective Accountability. Dissertation, Northwestern University.
———. 2003. Autonomy and the Authority of Personal Commitments: From Internal Coherence
to Social Normativity. Philosophical Explorations 6 (2): 90–108.
———. 2014. Autonomy and Vulnerability Entwined. In Vulnerability: New Essays in Ethics and
Feminist Philosophy, ed. Catriona Mackenzie, Wendy Rogers, and Susan Dodds, 134–161.
Oxford: Oxford University Press.
Benson, P. 1987. Freedom and Value. Journal of Philosophy 84: 465–486.
Christman, John. 1991. Autonomy and Personal History. Canadian Journal of Philosophy 20: 1–24.
Dennett, Daniel C. 2003. Freedom Evolves. New York: Viking.
Dworkin, Gerald. 1988. Is More Choice Better Than Less? In The Theory and Practice of
Autonomy, 62–81. New York: Cambridge University Press.
Frankfurt, Harry G. 1970. Demons, Dreamers, and Madmen: The Defense of Reason in Descartes’s
“Meditations”. Indianapolis: Bobbs-Merrill.
———. 1982. Comments on Macintyre. Synthese 53: 319–321.
———. 1988a. Equality as a Moral Ideal. In The Importance of What We Care About, 134–158.
New York: Cambridge University Press.
———. 1988b. Freedom of the Will and the Concept of a Person. In The Importance of What We
Care About, 11–25. New York: Cambridge University Press.

8
For valuable discussion and feedback, I would particularly like to thank Thomas Nys. An earlier
version of this essay appeared in Dutch as “Is tevredenheid met jezelf genoeg?” in Vrijheid,
noodzaak en liefde: Een inleiding tot de filosofie van Harry Frankfurt, ed. Katrien Schaubroeck
and Thomas Nys (Kapellen: Pelckmans, 2011), pp. 75–95.)
2 The Passivity of Self-Satisfaction: A Critical Re-appraisal of Harry Frankfurt’s… 31

———. 1988c. Identification and Externality. In The Importance of What We Care About, 58–68.
New York: Cambridge University Press.
———. 1988d. Identification and Wholeheartedness. In The Importance of What We Care About,
159–176. New York: Cambridge University Press.
———. 1988e. The Importance of What We Care About. In The Importance of What We Care
About, 80–94. New York: Cambridge University Press.
———. 1998a. On the Necessity of Ideals, 108–116. Cambridge: Cambridge University Press.
———. 1998b. The Faintest Passion, 95–107. Cambridge: Cambridge University Press.
———. 2004. The Reasons of Love. Princeton: Princeton University Press.
———. 2006. Taking Ourselves Seriously and Getting it Right. Stanford, CA: Stanford
University Press.
Frankurt, Harry. 2002a. Reply to Barbara Herman. In Contours of Agency: Essays on Themes from
Harry Frankfurt, ed. Sarah Buss and Lee Overton, 275–278. Cambridge, MA: MIT Press.
———. 2002b. Reply to Gary Watson. In Contours of Agency: Essays on Themes from Harry
Frankfurt, ed. Sarah Buss and Lee Overton, 160–164. Cambridge, MA: MIT Press.
Frankurt, Harry G. 2002c. Reply to J. David Velleman. In Contours of Agency: Essays on Themes
from Harry Frankfurt, ed. Sarah Buss and Lee Overton, 124–128. Cambridge, MA: MIT Press.
Friedman, Marilyn. 1986. Autonomy and the Split-Level Self. Southern Journal of Philosophy
24: 19–35.
Herman, Barbara. 2002. Bootstrapping. In Contours of Agency: Essays on Themes From Harry
Frankfurt, ed. Sarah Buss and Lee Overton, 253–274. Cambridge, MA: MIT Press.
Isay, David. 1994. Robert Shields, World’s Longest Diary. Morning Edition, National Public Radio
Rössler, Beate. 2009. Autonomie und Ambivalenz. In Sozialphilosophie und Kritik, ed. Rainer
Forst, Martin Hartmann, Rahel Jaeggi, and Martin Saar, 359–383. Frankfurt: Suhrkamp.
Swindell, J.S. 2010. Ambivalence. Philosophical Explorations 13: 23–34.
Taylor, Charles. 1985. Self-Interpreting Animals. In Human Agency and Language, 45–76.
Cambridge: Cambridge University Press.
———. 1989. Sources of the Self: The Making of the Modern Identity. Cambridge: Harvard
University Press.
Velleman, J. David. 1989. Practical Reflection. Princeton: Princeton University Press.
———. 2002. Identification and Identity. In Contours of Agency: Essays on Themes from Harry
Frankfurt, ed. Sarah Buss and Lee Overton, 91–123. Cambridge, MA: MIT Press.
Watson, Gary. 1975. Free Agency. The Journal of Philosophy 72 (8): 205.
Wolf, Susan. 2002. The True, the Good, and the Lovable: Frankfurt’s Avoidance of Objectivity. In
Contours of Agency: Essays on Themes from Harry Frankfurt, ed. Sarah Buss and Lee Overton,
227–244. Cambridge, MA: MIT Press.
Chapter 3
Determining Oneself and Determining
One’s Self

Thomas Schramme

Abstract In this essay, I exploit an ambiguity in the concept of self-determination.

Self-de Green termination can mean to determine oneself in choices and actions or
to determine one’s self. The second kind of self-determination leads to our capacity
to imagine alternative selves of ourselves, which are to be actualized. This creates
the basis for a normative conception of self-determination, i.e. a conception that
incorporates the aspect of a right or good way to determine oneself. I defend a nor-
mative interpretation of self-determination, which is illustrated by reference to three
theorists: Harry Frankfurt, Thomas Hill Green and John Stuart Mill.

In this essay, I exploit an ambiguity in the concept of self-determination. Self-­

determination can mean to determine oneself in choices and actions or to determine
one’s self. I will call the first interpretation self-actualisation and the second self-­
formation. The second kind of self-determination leads to our capacity to imagine
alternative selves of ourselves, which are to be actualized. This creates the basis for
a normative conception of self-determination, i.e. a conception that incorporates the
aspect of a right or good way to determine oneself. I defend a normative interpreta-
tion of self-determination, which is illustrated by reference to three theorists: Harry
Frankfurt, Thomas Hill Green and John Stuart Mill.
In the first section (3.1), I set the context by describing in general terms how I
conceive the connection between freedom and human beings’ idea of themselves.
There is a conception of self-determination that sets no limits so that any possible
uncoerced way to express oneself is supposed to be self-determined. There is a sec-
ond conception that introduces norms of proper forms of self-determination. I call

T. Schramme (*)
University of Liverpool, Liverpool, UK
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 33

Switzerland AG 2022
J. F. Childress, M. Quante (eds.), Thick (Concepts of) Autonomy, Philosophical
Studies Series 146, https://doi.org/10.1007/978-3-030-80991-1_3
34 T. Schramme

the former voluntarist and the second perfectionist accounts of self-determination.

In the second section (3.2), I introduce the general distinction between self-­
actualisation and self-formation and connect them with the two accounts from sec-
tion one, so that I end up with a classification of four ideal-types of theories. The
following three sections discuss three versions (one is disregarded for being implau-
sible) with historical illustrations. A version of voluntarist self-actualisation is rep-
resented by Harry Frankfurt in Sect. 3.3. In the fourth section (3.4), I discuss Thomas
Hill Green’s version of perfectionist self-formation, before moving on to John
Stuart Mill’s account of perfectionist self-actualisation in Sect. 3.5. Mill’s theory
seems to me the most plausible in that it allows for normative constraints of proper
self-actualisation but also insists on one’s own individual versions of self-­
determination. Section 3.6 concludes.

3.1 Introduction

The value of our freedom to act does not merely stem from the fact that it enables
us to do things that we may, at some specific point in time, strive to do. The value of
individual freedom especially derives from the desire of human beings to express
themselves in the world. Human beings do not merely want to satisfy their relevant
preferences, but also to realise their individual identity – the way they are – in their
decisions and actions.
This necessitates a capacity to see oneself as a being that possesses and com-
mands certain continuous attributes, such as evaluative beliefs, ideals, and a charac-
ter. Restrictions to our options seem to us to be problematic wherever they bring us
to do something with which we do not identify ourselves, or when we are not able
to do something for which we stand. This is, in a word, the value of
self-determination.
If we did not have the capacity to identify with our actions – or to distance our-
selves from them – self-determination would not seem to have value to us; for what-
ever we would do, it would be impossible to evaluate it from our own perspective.
We would simply actualise that which currently impels us to do so. If we were for
instance automata, then we would not need to be concerned about whether we have
actually realised ourselves or not. We would simply act, and would either be capable
or incapable of doing that which is given to us as a goal, e.g. by a computer pro-
gram. An automaton can have freedom to act or not, but it cannot be self-­determined.
Self-determination, rather, has to do with freedom of choice. Only when beings can
be this or that, when they have alternative ideas of themselves arises the idea of a
self and thereby the significance of whether one can realise oneself in the world or not.
The idea of self-determination is not commonly understood such that every way
we express ourselves through our actions actually counts as an acceptable form of
self-determination. There are indeed many examples, especially in literature and
film, in which human beings do not reach their potential or “waste their lives”. For
instance, should we consider the life of Jeff Lebowski, also known as “The Dude”,
3 Determining Oneself and Determining One’s Self 35

protagonist of the film The Big Lebowski, as a self-determined one? The Dude does
not do much else than to consume drugs, to go bowling, and to take care of his
shag rug.
Many philosophers believe that we human beings have a natural need to cultivate
our individual capabilities as much as possible. According to this understanding of
the human life the Dude would not live up to his own nature. In this manner John
Rawls for instance has made the claim: “Human beings enjoy the exercise of their
realized capacities (…), and this enjoyment increases the more the capacity is real-
ized, or the greater its complexity.” (Rawls 1999, 364). We could paraphrase this
so-called “Aristotelian principle” of Rawls’s in that human beings have a perfectible
nature, and that they are interested in improving this nature. Here the connection
between self-determination and happiness or the good life of human beings is
clearly visible. Indeed, many would claim that only a self-determined life can be a
good life. In this way it seems important how and to what end we determine
ourselves.
But not to see the Dude as at all self-determined is perhaps demanding too much
of a self-determined life. In the end, for all we know he has chosen his life without
coercion. Is it not enough for the goal of self-determination that certain formal cri-
teria for free choice are fulfilled? The Dude may fall short of real happiness or he
may simply foster some sort of unconventional interpretation of the good life, but he
lives his life in self-determined fashion all the same. From this point of view there
would be no prescribed external norms of proper self-determination whatsoever. A
person would then be self-determined where she voluntarily lives a life according to
her free choice; only where she does not choose freely, her self-determination would
be in doubt.
The discussion about the concept of self-determination accordingly seems to
oscillate between two philosophical poles: On one side, the ideal of perfectionism,
which takes the development of potentials as a criterion for self-determination, and,
on the other side, the ideal of voluntarism – as I call it – which only focuses on an
individual‘s own selection of options. As it often happens, or so it seems to me, the
most plausible approach lies somewhere between the two extremes.
In my analysis, I would hence like to illuminate the notion of self-determination
a little more, not in pursuing any sort of conceptual analysis, but rather by discuss-
ing actual philosophical interpretations of this concept and weighing their argu-
ments. First and foremost, I will defend a normative conception of self-determination.
I believe the most plausible interpretation of self-determination necessarily encom-
passes an aspect of the right or good way of self-determination. In pursuing this
argumentative goal my point of departure will be an equivocation that engenders
two central aspects of the idea of self-determination: Self-determination can either
mean to determine oneself in the world or it can mean to determine one’s self. The
possibility of the second sort of self-determination – which I will call self-­
formation – implies envisaging alternatives for one’s own self that can be realised
(cf. Bransen 2002). The mentioned alternatives of understanding self-­determination,
combined with the distinction between voluntarism and perfectionism, leads me to
a four-fold categorisation of different theories of self-determination. In the
36 T. Schramme

following I will illustrate this classification with the help of philosophical theories
devised by Harry Frankfurt, Thomas Hill Green and John Stuart Mill.
When I refer to “a self” in this paper, I merely mean to refer to the attributes,
capabilities, value beliefs and characteristics of a person. I do not want to add a
metaphysical claim about the ontology of selves (cf. Velleman 2006, 1). As far as I
am concerned, selves are assemblages of aspects of an individual being. In addition,
I want to presume that there is a self only where there is a relationship of a being to
itself; where it can understand itself this or that way (cf. Taylor 1985). To this end it
is important that such a being is conscious of itself and acquires knowledge about
itself. The peculiar form of self-relation is an old theme of philosophy (cf. Taylor
1989), of course, but for my analysis it is only relevant in terms of representing a
precondition of self-determination. My main aim is to clarify what it means to
determine oneself and one’s self.

3.2 Self-Actualisation and Self-Formation

As mentioned before, the concept of self-determination has two dimensions: The

determination of decisions and actions by oneself and the determination of one’s
self. In most contexts self-determination is understood in the first sense, for instance
when applied to the choices of patients in a medical context. They determine them-
selves in cases where they decide between therapeutic options and where this deci-
sion is actually implemented. Talk of self-determination as a determination of the
self, i.e. self-determination in the second sense, is predominantly discussed in con-
nection with the metaphysical problem of free will: In order for freedom of the will
to be possible, some might claim, it is required that aspects of the self, for instance
the character of a person, have been determined by the person herself and not arisen
through some deterministic mechanism. For the sake of simplicity, I will from now
on refer to self-determination as action under the aegis of oneself – self-­determination
in the first sense – as self-actualisation and to the determination of one’s own self –
self-determination in the second sense – as self-formation.
To illustrate the two ideas, one could claim that a student only self-determinedly
or freely chose to study philosophy if the underlying desire itself was in turn deter-
mined or generated by that person herself. The willing student of philosophy might
actualise herself by acting on her choice, but her respective will might not have been
formed by herself. I believe many quarrels about people being or not being self-­
determined are due to cross-purposes regarding the two different ideas of
self-determination.
Still, should we really accept this differentiation between two meanings of self-­
determination? Does is not lead to various philosophical dead ends, especially in
terms of the metaphysical problem of free will? It seems to me that we can indeed
make sense to the relevant distinction in a way that does not simply merge into the
problem of free will. It appears to me that a focus on the debate regarding free will
would put us on the wrong course. For it seems to pose the problem of
3 Determining Oneself and Determining One’s Self 37

self-­determination in the following way: Self-determination is the actualisation of

aspects of the self – desires, value judgments, personal characteristics, etc. – in the
world; but how does the self, which determines itself here, arise? Was it determined
by itself or “externally” determined? The problem with this point of view lies in the
fact that existence of the self is postulated prior to its own actualisation.
This is a problematic point of view: according to it, first a self is developed, and
this is then actualised through actions. The problem is that within such a framework
we can ask the question whether one’s self has been determined by oneself, which
makes it obvious that a doubling of the self has arisen. There would have to be in
turn someone, a self, who forms the self. The structure of self-determination as an
actualisation in the world is transposed onto the self within this dubious framework;
the self becomes the result of the forming actions of a – probably another – self.
This view therefore produces a kind of homunculus that forms the person.
To be sure, these problems are by no means new, and especially in relation to the
related philosophical problem of self-consciousness they have been well analysed
(Tugendhat 1989). The questionable aspect stems from the translation of a subject-­
object structure into the idea of self-formation. In order to be able to “form” one-
self – an object – there would have to be already to be a subject in place, which
precedes the self to be formed. But here subject and object are of course identical,
and herein lies the philosophical oddity.
The problem of the subject-object structure is relatively easy to solve. We should
simply say that a person forms and expresses herself through her actions (Schechtman
2004). This framework circumvents the problematic implication of a split between
a forming self, the “former”, as it were, and that which is to be formed. There is
accordingly no prior self, no homunculus. Insofar as we have different options there
are also different possible self-formations.
In addition, this way of understanding self-formation allows “quality control”; in
other words, assessments of right and wrong or good and bad self-formations. As
we shall see, this is a central element of the philosophical discussion about self-­
determination: Human beings have an “evaluative self-relation” (Quante 2007, 29).
Yet understood in this way the difference between self-actualisation and self-­
formation indeed seems to collapse, for to actualise oneself now simply seems to
mean to express one’s self. To be sure, one could form oneself in various ways so
long as one can act in varying ways, but there would seem to be no scope here for
an incongruence between self-actualisation and self-formation. One would simply
be what one expresses, or what one does.
This would be an unsatisfactory result in my opinion, since the relevant phenom-
enon of actualising oneself in a way that one is not, hence in a way that does not
correspond to one’s true self, is very real. One can be a person who does not express
herself in the actions she commits. The person’s actions in these cases are, as it is
said, alienated or inauthentic (cf. Oshana 2010). Here the person is therefore not
expressing herself qua self-determined actions; nonetheless she acts self-­
determinedly, in the sense that she actualises a self freely.
It seems implausible to me to completely abandon the concept of self-­
determination in those cases in which actions are alienated in order to maintain a
38 T. Schramme

conceptual identity between self-actualisation and self-formation. There are defi-

nitely forms of self-actualisation in which a person does not express her self and
still acts self-determinedly. Indeed, the converse also seems possible: A person
might express her own self in that she actualises herself; these are cases in which the
person identifies with her actions. Here she determines herself (really), that is, here
the congruence between self-formation and self-actualisation is present. This con-
gruence seems to be the primary goal of self-determination: To express oneself
through actions in the way that one sees oneself; which in turn means to act in har-
mony with oneself. Collapsing the idea of self-actualisation and self-formation
seems to undermine our conceptual resources to describe and value these cases.
In my opinion we should therefore not level the difference between self-­
actualisation and self-formation, for we would remain on the level of factual descrip-
tion if we see congruence in each case. We could then do no more than indicate that
a person is the one who actualises herself in free action; that is, we would have
generated a purely descriptive concept of self-determination. The only relevant cri-
terion of self-determination would be the exhibition of voluntary decisions and the
corresponding options to act. Naturally I do not want to claim that this purely
descriptive use of the expression “self-determination” would be pointless or useless.
It does for instance function well in political philosophy, where the limits of legiti-
mate influences on decisions are involved. Yet it seems to me that such a restricted
view of the concept of self-determination neither fully captures the phenomena nor
takes account of all ways of looking at related ethical problems.
The idea of self-formation implies the idea of alternative selves. We are not sim-
ply those who actualise themselves in decisions and free actions, but rather ones
who have the possibility – however limited – to be this or that way. In brief, we can
form ourselves within certain limits (cf. Glover 1988). We can also conceive of dif-
ferent alternatives of ourselves; we can for example imagine how it would be like to
be more selfish or to live the life of an artist in the South Seas instead of with a fam-
ily in the big city. We can also evaluate these alternatives. We can try to bring these
hypothetical, alternative selves – let’s call them self-ideals – into reality. In this way
a potential incongruence arises between the actual self-actualisation and possible
alternative forms of self-actualisation, that is to say, the actualisation of alternative
selves. Here it is important to note that the difference between the actual self, which
actualises itself in actions, and the self-ideal, which could be actualised in actions,
does not lead to a problematic doubling of the self – criticised above – since here no
synchronically present multiple selves are postulated.
With this differentiation we succeed in gaining a normative point of view in
application to the concept of self-determination, for now we can ask how and
towards what end we should determine ourselves. The goal in this pursuit is to reach
a congruence between self-actualisation and self-ideal, that is, to express one’s ideal
self in self-determining actions.
Now we can return to the extreme positions of voluntarism and perfectionism
that were mentioned at the beginning, so that four possibilities of conceiving self-­
determination arise:
3 Determining Oneself and Determining One’s Self 39

1. Voluntarist Self-Actualisation: Self-determination means to actualise one’s true

self. This self is not created by the person herself, but is rather revealed through
her actions. The point is to be what one is. The goal of self-determination is to be
authentic.
2. Voluntarist Self-Formation: Self-determination means to actualise one’s chosen
self. The point is to make oneself that which one has chosen oneself to be. The
goal of self-determination is to actualise one’s own self-conception.
3. Perfectionist Self-Actualisation: Self-determination means to become one’s best
possible self, to actualise one’s potentials. The point is to become the most devel-
oped variation of oneself that one can be.
4. Perfectionist Self-Formation: Self-determination means to actualise the best pos-
sible self. The point is to become the most developed variation of a self that is
possible.
It barely needs emphasising that this categorisation has its own problems. It will
nevertheless serve to bring order into the upcoming examples of philosophical theo-
ries and hopefully to help making progress in the discussion on self-determination. I
refer to the philosophical theories I discuss as theories of self-determination, even if
the concept of self-determination is not central to all the respective theories. As shall
be seen, I refer to Harry Frankfurt’s approach as a version of voluntarist self-­
actualisation. The problem of his reflections seems to me primarily to lie in that he
leaves no room for a self-ideal that departs from the mere being-identical-with-­
oneself. Thomas Hill Green, who represents the perfectionist self-formation version,
has, as it were, a little too much of this idea of the ideal self. He lays out an ideal self
that is not one’s own self, but is rather attained from the outside by reference to the
idea of reason. In John Stuart Mill we find an ideal of one’s self that is thoroughly
one’s own self; a self which does not lie outside of the self-relation of the person,
hence is not self-formed, yet can be improved. Mill therefore represents the perfec-
tionist self-actualisation version. This approach seems to me most plausible and I
will defend it up to a certain point. The voluntarist self-formation version, which can
perhaps be connected most closely to Jean-Paul Sartre, will not be discussed here. To
me this approach is a non-starter, because it either reproduces the homunculus-prob-
lem or, in any plausible variation, in which a person actualises herself as the self she
has formed, collapses into either of the two self-actualisation variants.

3.3  elf-Identification: Harry Frankfurt’s Theory

S
of Voluntarist Self-Actualisation

According to Frankfurt, a specific capacity of persons is to turn desires into objects

of our volitions (Frankfurt 1971). We can accordingly take a stance with respect to
ourselves in that we may desire to be effectively motivated through specific desires.
Desires in this model are hierarchically structured; we can evaluate desires of the
first order and thereby generate desires of a second order – desires with respect to
40 T. Schramme

desires. This is in no way a strange philosophical idea, but is rather in line with
everyday experience, for instance when we, starting from a desire to eat chocolate,
distance ourselves and desire instead that the particular action may not be initiated.
In Frankfurt’s account, a person acts from her own will when the effectively
motivating desires are the ones that she desires to be motivated by; hence if a per-
son’s so-called second-order volitions are actualized. For our purposes we might
want to say that second-order desires represent an element of the self-ideal of a
person. This self-ideal does not actualize itself in each case of action. For numerous
reasons an incongruence can arise between the way in which one actually deter-
mines oneself and the way in which one desires to determine oneself. If a person
however identifies herself with the desire that has a motivating effect on actions –
that is, the “will” in Frankfurt’s terminology – then she acts on account of her own
will. This, according to Frankfurt, means to enjoy free will (Frankfurt 1971, 17).
But on which basis does one identify with certain desires? Structurally, within
Frankfurt’s model, it is second-order volitions that serve this function of identifica-
tion, but from where do these stem? Are they self-chosen and thereby, potentially
anyway, brought under the hierarchical model? This would require a possibly infi-
nite regress in the structure of desires; that is, even higher-order volitions than the
second-order ones. Frankfurt’s answer is that the continuation of the volitional hier-
archy ends where one wholeheartedly identifies with a desire (Frankfurt 1987).
Accordingly, action-controlling desires – representing the will – are not always
themselves objects of wilful identification. In other words, in the case of decisive
identification there is no considered choice between different first-order desires,
rather, when acting wholeheartedly, the will is simply based on the personal charac-
teristics one happens to have.
For a person, it can lie at the heart to act one way and not another; and then it
appears unthinkable to her to have another will. Although this almost sounds like a
description of compulsive behaviour, according to Frankfurt a person is neverthe-
less not acting unfreely in this case. On the contrary, she then acts on her own free
will – on what she cares about – because she identifies herself wholeheartedly with
the motivating desire. Such an identification does not lead her to any further weigh-
ing or reasoning – a kind of undecidedness – but rather to a satisfaction with a given
configuration of the will (cf. Seel 2002, 291f.)
When developing his theory, Frankfurt later went even further and introduced the
notion of volitional necessities (Frankfurt 1982). In a supposedly paradoxical way,
volitional necessities become the epitome of freedom, because through them we act
wholeheartedly on the basis of our own will. The necessitating character of these
special volitions in turn inhibits the ascent into higher spheres of volition – it ends
all questioning and balancing about motivating desires’ adequacy or justification.
Since no assessment of desires is possible in these circumstances – after all a person
is then driven by a form of compulsion – there is no scope for an infinite progress of
hierarchies of volition.
Frankfurt even describes the phenomenon of love within this model: Its will-­
defining necessities signify culmination points within the structure of caring.
“Loving is a mode of caring. Among the things that we care about there are some
3 Determining Oneself and Determining One’s Self 41

that we cannot help caring about; and among the things that we cannot help caring
about are those that we love” (Frankfurt 1999, 165).
However, Frankfurt has created a new problem with his analysis, one that has
garnered much criticism: Volitional necessities cannot really be objects of a per-
son’s own evaluation, because they are not really available to the person; they have
no control over volitional necessities. Yet this means that the very possibility of
subjective critical engagement with any of one’s wholehearted desires has appar-
ently been ignored by Frankfurt. Such a model seems to leave out something rather
important: the critique of oneself as regards one’s basic commitments. Frankfurt
insists on a model of self-determination that builds on reflexivity – desires regarding
desires – without reflection – a critical stance towards oneself (cf. Christman 1991b;
Herrmann 2000, 159).
Accordingly, in Frankfurt’s model, whether a person acts in unity with her own
will is ultimately determined merely by a specific volitional relation, namely a
strong identification between volitions. Self-determination is hence independent of
what would be intelligent, rational or morally imperative – independent of any
subject-­external normative standard. Frankfurt indeed insists that it is the very
unavailable subjective will that forms reasons for actions, and not the other way
around – that is, reasons forming the will. In an alternative model, where a person
would weigh and assess reasons for or against a specific action, she would require,
according to Frankfurt, prior evaluative criteria. Yet, these criteria themselves can
only be based on the person’s own wholehearted commitments.
This means that the most basic and essential question for a person to raise concerning the
conduct of his life cannot be the normative question of how he should live. That question
can sensibly be asked only on the basis of a prior answer to the factual question of what he
actually does care about (Frankfurt 2004, 26; emphasis in original).

For Frankfurt, the concern about something is the source of reasons. We cannot
justify what we care about, only uncover it, for we are “beings, who do not create
themselves” (Frankfurt 2004, 20). To be sure, Frankfurt does in no way want to
exclude that the worthiness of an object can causally produce or influence caring or
love. But in order to really be able to love something or someone, this behaviouristic
basis is not sufficient to produce worthy volitions, since a person must individually
be able to actualise – to live, as it were – a particular object of love or concern. “It
is not necessarily as a result of recognizing their value and of being captivated by it
that we love things. Rather, what we love necessarily acquires value for us because
we love it” (Frankfurt 2004, 38f.; emphasis in original). The desirable or valuable
object has to possess an attractive power for the lover. Therefore, an unavoidably
personal element is found in love. We show ourselves in what we love.
It seems to me that there are very significant insights regarding the concept of
self-determination within Frankfurt’s model of the will structure, even in the sup-
posedly overstretched idea of volitional necessity. The capacity for self-­determination
touches upon our capacity to enter a position towards ourselves. We act self-­
determinedly if we can actualise the conception of our self in the world with which
we identify ourselves. Our freedom appears exactly at that point, when we are
42 T. Schramme

determined in a special way from the bottom of our hearts, or through aspects about
which we are most concerned. The point of self-determination is not to actualise a
supposedly free, self-produced will, but rather our own will.
Yet Frankfurt’s strict rejection of any reflection about the question as to how one
should live or rather what kind of person one should be, independently of the ques-
tion of what one is content with, indeed seems excessive. He seems to be driven by
a concern, which is internal to his theory, to avoid the potential infinite regress of his
hierarchical model of volition. Yet it is a common experience of human beings that
we are not only able to take a stance towards our motivating desires, but also in rela-
tion to deep-rooted aspects of ourselves. When we determine ourselves, then we are
always – however indirectly – concerned with the right ways, or at least with alter-
native ways, of self-determination.
Few people would generally be satisfied with a simple “I am what I am” – even
if that might be enough for their daily lives. Our self, that is, our attributes, evalua-
tive beliefs, motivations, etc., is not clearly predefined. We can always think of new
alternatives to it. Even if Frankfurt rightly emphasises that we should be concerned
with realising ourselves in the form in which we are contented with ourselves, this
does not mean that we could not very well have reasons nonetheless to imagine
ourselves differently and perhaps even to hold these images as being better for us.
We can foster desires for alternatives to what kind of person we are. Frankfurt’s
philosophical concern regarding the hierarchical structure of his model seems to be
unfounded: The decision in what way or to what end we should actualise ourselves
does not require a third order of volition, although it does require some kind of
consideration.
To be sure, Frankfurt rightly questions the idea of a wilful formation of one’s
own self. Perhaps we merely delude ourselves when we assume that we would have
the possibility to form ourselves in our deep-rooted aspects. To illuminate this prob-
lem further, I will elaborate on the idea of the formation of one’s own self in the
following section.

3.4  elf-Realisation: Thomas Hill Green’s Ideal

S
of Perfectionist Self-Formation

Before discussing Green, it helps to introduce a famous critic of his account, who
has unfortunately dented Green’s philosophical reputation for some time. Isaiah
Berlin, in his famous Oxford inaugural lecture Two Concepts of Liberty, objects
to an ideal concept of liberty, and he ascribes this idea mainly to Green. Even if in
this work Berlin does not explicitly use the concept of self-determination, it is clear
that his discussion is pertinent to the interpretation I have pursued so far (Berlin
2002, 53). For Berlin the problematic idea of positive freedom results from asking
the already mentioned question whether a person is indeed herself that person who
3 Determining Oneself and Determining One’s Self 43

forms herself. In Berlin’s own words: whether one is master over one’s own self
(Berlin 2002, 178).
In this way Berlin also seems to reject any normatively conceived conception of
the correct or best way of self-determination or of the exercise of freedom, because
such a normative account seems to come with an ideal of rational or moral self-­
determination. For instance, he endorses Bentham’s rhetorical question: “The lib-
erty of doing evil, is it not liberty?” (Berlin 2002, 194). To be sure, I do not believe
that Berlin therefore has to accept every possible form of actual self-determination
as unproblematic and as the right form for the particular person; this does not seem
for instance to be the case if a person is ignorant about what is in reality contributing
to her own good.
But Berlin does not seem to allow for there to be incongruence between the self
that is actually expressed in actions and another form of self-actualisation that could
result from an alternative account of one’s self. Berlin is quite careful not to commit
to postulate alternative selves, even where he wants to allow criticism of some exer-
cises of freedom (Berlin 2002, 41, 177). In my terminology, it might be said that
when he explains the possibility of criticising self-actualisation, he avoids any pos-
sibility of self-formation. Yet in this regard he seems to be going too far, for he
thereby dismisses the possibility of differentiating between the actual self-­
actualisation in particular actions and a possible alternative self-actualisation in line
with a self-ideal.
This is a problem for Berlin, because not every conception of ideal self-­
actualisation seems to necessarily take on the problematic form that Berlin dis-
misses, namely to be determined to a degree by a conception of the “‘real’ self, of
which the poor empirical self in space and time may know nothing or little” (Berlin
2002, 180). To this extent the differentiation between various ways of self-­
actualisation and their evaluations would not be problematic as such, even accord-
ing to Berlin, but rather the specific form of this differentiation, which is assumed in
theories that take as a basis a subject-external ideal, for instance on grounds of an
account of specific human purposes. Berlin therefore seems to accept forms of
“quality control” of self-determination, but he dismisses normative considerations
of how – or in which way – we should determine ourselves. Yet exactly here lies the
flaw of Berlin’s argument, for it means to dismiss the underlying practical question
of how one is to live. To omit this very aspect seems inadequate when it comes to a
theory of self-determination (cf. Gerhardt 1999).
British philosopher Thomas Hill Green, who was strongly influenced by Hegel,
is one of the best-known proponents of a conception of positive freedom. As men-
tioned, he is also one of the antagonists of Berlin’s defence of negative freedom,
which merely requires the absence of coercion. Green, in contrast to Berlin, is of the
belief that freedom consists in determining oneself in a particular way.
It needs to be stressed that Green does not commit the previously discussed error
of assuming a kind of homunculus in order to differentiate between actual and right
ways of determining oneself. His point is accordingly not to iteratively ask whether
the will that actualises itself in actions could itself count as free. Locke had already
brushed off this question as meaningless in his Essay Concerning Human
44 T. Schramme

Understanding, and Green shares this view. The will as such is by definition free,
because human beings express themselves in executions of their will; indeed, Green
describes the figure of speech “free will” as a pleonasm (Green 1879, 308). In this
respect Berlin’s critique of Green has to be challenged, since it depends on the
assumption that when using the concept of positive freedom one need to presume
that there is an agency – the “dominant self”– that rules over the will, or the “empiri-
cal self” (Berlin 2002, 179f.). Green however maintains: “The will is simply the
man” (Green 1883, 158). Accordingly, Green postulates no synchronic reduplica-
tion of the self, even when he talks about being “master of himself” (Green 1879,
322). To be the master of oneself does not mean to be a self that has a master, who
is identical to oneself.
There is no such thing as a will which a man is not conscious of as belonging to himself, no
such thing as an act of will which he is not conscious of as issuing from himself. To ask
whether he has power over it, or whether some other power than he determines it, is like
asking whether he is other than himself. Thus the question whether a man, having power to
act according to his will, or being free to act, has also power over his will, or is free to will,
has just the same impropriety that Locke points out in the question whether the will is free
(Green 1879, 317f.).

Green distinguishes between freedom in the “juristic” or civil sense – the negative
freedom of Berlin’s – and a freedom that represents a certain form of self-formation
(Green 1879, 315), whereby this form allows different interpretations: The forma-
tion of oneself according to nature, as the Stoics assumed; according to God’s will
like in Paul; in the sense of a universal lawgiver as in Kant; or pursuant of the inter-
ests of a well-regulated state like in Hegel. For Green himself, the best result of
self-determination consists in realising a specific version of an ideal self. To discuss
this ideal, he indeed uses the notion of positive freedom, but for him this positive
freedom is based on negative freedom. In this regard positive freedom therefore
does not stand in opposition to Berlin’s approach.
[T]o the grown man, bred to civil liberty in a society which has learnt to make nature its
instrument, there is no self-enjoyment in the mere consciousness of freedom as exemption
from external control, no sense of an object in which he can satisfy himself having been
obtained. […] ‘freedom’ is the natural term by which the man describes such an object to
himself, − describes to himself the state in which he shall have realised his ideal of himself
[…] (Green 1879, 323f.).

The will therefore has to lead to a specific goal in order for a person (in this positive
sense) to be able to be free: The improvement, and eventually the perfection, of
herself. There is nothing philosophically strange at all in this idea. Berlin was of
course concerned about the possible political consequences of such a conception,
which could come close to forcing people to achieve the relevant perfection. But
nothing in Green’s account as such seems to imply these potentially coercive mea-
sures (cf. Geuss 1995). His approach is indeed in harmony with my previous obser-
vations, for Green’s point is to hold firm that the idea of self-determination always
includes possibilities or potentials and that a driving force of human beings consists
in realising these potentials: “[Man] has the impulse to make himself what he has
the possibility of becoming but actually is not […]” (Green 1883, 182).
3 Determining Oneself and Determining One’s Self 45

The problems that Berlin pointed out begin where such an ideal is forced onto
persons and does not even represent a person’s own ideal self. To be sure, at least on
the face of it, for Green the idea consists in purely formal aspects, so does not seem
to impose a particular interpretation of the ideal self. To the extent that one can
control oneself through the operation of one’s practical reason, the perfection of
oneself consists in the best possible operation of the capabilities connected with
reason (cf. Simhony 1993).
The label Green introduces for this idea, especially in the posthumously pub-
lished Prolegomena to Ethics, is “self-realisation” (cf. Brink 2003, 41). Yet the
realisation of the self in Green’s approach does not appear to be a realisation of
one’s own self after all. It rather consists in the realisation of the best possible self
that is determined on the basis of criteria that are not our own, but stem from a
theory of the rational nature of human beings. In this respect, Berlin’s worries seem
well-founded. “Once I take this view, I am in a position to ignore the actual wishes
of men or societies, to bully, oppress, torture them in the name, and on behalf, of
their ‘real’ selves, in the secure knowledge that whatever is the true goal of man”
(Berlin 2002, 180).
In addition, the notion of self-realisation has connotations that cast doubt on the
very idea of actualising one’s own potentials. ‘Self-realisation’ seems to be refer-
ring to the realisation of that which is already present in the self. To which extent are
ideal alternatives of our selves actually available to us? To which extent is it in our
power to form ourselves? If we cannot form ourselves according to an ideal of the
self, then Green’s theory is potentially vacuous. A philosophical theory that allows
insights regarding the restricted scope for self-formation is that of John Stuart Mill.

3.5  elf-Development: John Stuart Mill’s Conception

S
of Perfectionist Self-Actualisation

It may be surprising to suggest that Mill puts in doubt the possibility of human
beings to form themselves, since he is generally supposed to be a proponent of a
fairly radical liberalism that wants to see individual freedom as broadly protected as
possible. Why protect something that cannot really be achieved? Yet liberalism
involves first and foremost a perspective on the societal and political conditions of
self-determination in the sense of self-actualisation. One can reasonably want to
protect self-determination without believing in the possibility of the independent
formation of alternative selves. Political liberalism as such is non-committal about
the feasibility of self-formation. If the idea were rejected, the point of political pro-
visions to protect self-determination would then be to model them in a way that an
already laid out or latent self can develop. Mill seems to suggest exactly this
approach when he compares self-development with the growing of a tree: “Human
nature is not a machine to be built after a model, and set to do exactly the work
prescribed for it, but a tree, which requires to grow and develop itself on all sides,
46 T. Schramme

according to the tendency of the inward forces which make it a living thing.” (Mill
1859, 263). The inward forces here might very well be interpreted as the specific
characteristics of a pre-conceived individual self.
In contrast to Green, Mill does not speak of self-realisation, but rather of self-­
development, although it – so it seems – could terminologically have come the other
way around, since Green’s account involves perfection and Mill refers to the realisa-
tion of inner forces. Obviously this choice of terms has, at least in Green’s case,
philosophical-historical reasons and goes back to his reference to Hegel. Be that as
it may, in the quoted passage Mill indeed sounds as if he allows self-formation only
little scope. A tree just cannot decide in which direction it wants to grow, although
its growth is influenced through its environment and not completely determined by
inherent forces. Self-determination, within this picture, would consequently be
merely the actualisation of an already existing, latent self. To be sure, a person could
still act inauthentically, for instance when she does something against her own
nature only because the majority of people do it. She then acted, according to Mill,
not as an individual, but rather as a member of a herd. Individuality requires in his
opinion having one’s own idea of how one should live. The point of self-­development
is therefore to actualise one’s own self. Yet this very self seems again to be some-
thing to be discovered, not created or formed.
Understood this way, Mill’s theory seems unpromising because the development
of one’s own potentials alone seems deficient as an ideal of self-determination.
Consider Sisyphus: His potential – though admittedly forced upon him by the
gods – consisted in again and again hauling a stone up a hill. Would the develop-
ment of this potential already be a worthwhile way to self-determination? Now,
Sisyphus could develop this single ability, which still remains for him to actualise.
Joel Feinberg speculates accordingly: “[T]he gods might have designed for him a
peculiar talent for rockpushing much like others’ talents for pianoplaying, tennis, or
chess. The new Sisyphus starts all over as a perpetual youth, and from the start he is
a veritable prodigy at rockpushing. He comes to enjoy exercising his skills, and
makes evernew challenges for himself. He pushes the rock righthanded, then
lefthanded, then nohanded, then blindfolded, then does two at a time, then juggles
three in the air all the way to the summit, eager to return for another so that he can
break his record” (Feinberg 1980, 320). Most people would likely regard the life of
a perfect stone-hauler as an absurd form of self-realisation, even if – contrary to the
actual example – Sisyphus had himself chosen the occupation of a stone-hauler.
This sentiment is likely based in the belief that the practice of certain capabilities
cannot represent a reasonable self-ideal, and that their actualisation therefore also
facilitates no proper form of self-determination.
So far, we have interpreted Mill’s idea of self-development as restricted to pre-
conceived individual characteristics. Yet Mill‘s conception of discovering one’s
own self in order to live self-determinedly allows two possible interpretations: On
the one hand the point of self-determination could be to discover in each case one’s
own nature, that is, the essence of each individual, and to actualise this in the world.
Here the concept of authenticity would be appropriate, since it is applied to describe
the relation of an individual to itself. On the other hand, it would be conceivable that
3 Determining Oneself and Determining One’s Self 47

Mill postulates a general nature of human beings, which in turn would be there to
be discovered in self-actualisation. Here we would not be speaking of authenticity,
but perhaps of self-emergence – since it is here that a person altogether comes to a
specifically human way of life.
Mill is renowned for his deliberations on individual diversity and originality, up
to a defence of human eccentricity. In this regard the first interpretation appears
pertinent, whereupon self-determination is the actualisation of a unique self; a
“quiddity” (Gray 1991, 193). Accordingly Mill emphasizes the value of pluralism
of “experiments in living” (Mill 1859, 281). In this interpretation there appears to be
no normative standard of a right self-determination: The point simply is to actualise
oneself as one is. The authentic Sisyphus or the authentic Dude, to use the former
example, would from this point of view be self-determined human beings.
If, however, we refer to other works, especially Mill‘s account of hedonism in
Utilitarianism, it is clear, at least to my mind, that also the second interpretation, i.e.
the assumption of specific human aspects of self-development, was not foreign to
Mill. For there he claimed that a genuinely human life would be preferable to a “pig-­
like” life (Mill 1861, 210ff.). In these passages Mill sounds Aristotelian and hence
perfectionist – there are according to him bad kinds of self-actualisation after all,
because they violate the ideal form of a human life. “Few human creatures would
consent to be changed into any of the lower animals, for a promise of the fullest
allowance of a beast’s pleasures; no intelligent human being would consent to be a
fool, no instructed person would be an ignoramus, no person of feeling and con-
science would be selfish and base, even though they should be persuaded that the
fool, the dunce, or the rascal is better satisfied with his lot than they are with theirs”
(Mill 1861, 211).
The two interpretations of Mill‘s account of self-determination do not in any way
have to contradict each other: One can indeed postulate a general human form of
self-development and simultaneously be of the opinion that every person qua self-­
determining being develops her own individual version of this general human
nature, i.e. forms herself into an individual person. The only case where these two
accounts conflicted would be where an individual way of life would to some extent
be unhuman. For Mill, this extreme case seems to imply that it does not involve
genuine self-determination. Whether this again is a philosophically dubious propo-
sition depends first and foremost on how narrow the perfectionist framework, which
Mill builds, actually is. Although I cannot elaborate on this question here, it seems
clear to me that the point for Mill is to reveal forms of merely alleged self-­
determination that deny the underlying human capabilities. His point was not to
establish specific ways of life as the only appropriate ones for humans; that would
also contradict his belief in the “unlikeness of one person to another” (Mill 1859,
273) in connection with his notion that to develop oneself is only possible where
one lives one’s own life.
Surely it is empirically possible, according to Mill, that one does not actualise
underlying human capabilities, or does not want to actualise them, and therefore
lives against human nature, somewhat like a pig. Yet it would just not be desirable –
Mill can make that claim, for he proposes a normative conception of
48 T. Schramme

self-determination. At the same time this framework does not involve more than the
claim that self-determination shall lead to a life that achieves the predicate “human”.
Within this frame there remains a variety of good ways of individual self-­
development. In summary, the scope for self-formation is limited by an account of
the proper form of human nature, yet the scope is still wide and allows for
individuality.
There is still the question whether Mill allows for self-formation. It seems that
certain aspects of human nature by and large elude human control in his account. To
be sure, he merely seems to imply that we cannot determine the contours of the
appropriate frame of self-determination, properly understood, for this is given to us
as human nature. To properly determine oneself it is necessary to stay within this
frame. At the same time, we individually actualise ourselves within these contours
because we understand it as a task or responsibility, not because we are forced, that
is, determined by the laws of nature. It is necessary to choose the right goals, namely
those that do not defraud a human life. This means to form oneself according to
one’s own ideal within self-endorsed limits.
This conception still allows the possibility to form oneself as a unique individual.
Individuality in this understanding does not solely consist in affirming through
choice exactly those characteristics which our individual nature specifies for us. In
his work A System of Logic, of which the sixth book is dedicated to the “logic of the
moral sciences”, Mill takes to task the fatalistic argument whereby the character of
a person “is formed for him, and not by him” (Mill 1843, 840, emphasis in original;
cf. Evans 2007, 61ff.) and that it would therefore be useless to desire that it was
cultivated in a different way. To this Mill replies: “But this is a grand error. He has,
to a certain extent, a power to alter his character. Its being, in the ultimate resort,
formed for him, is not inconsistent with its being, in part, formed by him as one of
the intermediate agents. His character is formed by his circumstances (including
among these his particular organization); but his own desire to mould it in a particu-
lar way, is one of those circumstances, and by no means one of the least influential”
(Mill 1843, 840, emphasis in original).
So according to Mill, the desire to cultivate and educate one’s self is essential for
self-formation. But how does this desire in turn come about? Is it accessible to us?
Mill grapples with this last fatalistic objection: “But to think that we have no power
of altering our character, and to think that we shall not use our power unless we
desire to use it, are very different things, and have a very different effect on the
mind. A person who does not wish to alter his character, cannot be the person who
is supposed to feel discouraged or paralysed by thinking himself unable to do it. The
depressing effect of the fatalist doctrine can only be felt where there is a wish to do
what that doctrine represents as impossible” (Mill 1843, 841, emphasis in original).
It therefore doesn’t matter whether the desire to form oneself was itself again
determined; it matters whether it is present. We have to conceive of ourselves as
self-forming beings; only then can we be satisfied or unsatisfied with ourselves. The
fatalist cannot at all account for the common distinction between those who live
under the aegis of necessity and those who live under the aegis of freedom. We have
already seen before that self-determination does not mean to have a self that is
3 Determining Oneself and Determining One’s Self 49

brought forward through one’s own choice. Self-determination means, also accord-
ing to Mill, to actualise oneself in the world in the way one would like to. Now to
still ask the question whether having chosen freely the way one would like to be
means asking too much.
However, an important problem indeed, which fatalists could draw attention to
and which is connected to the idea of authenticity, seems to be the following: Up
until now we have expressed the idea of self-determination such that a person forms
herself in that she expresses herself in actions, value beliefs etc. In addition, there
was the idea of a self-ideal, which can bring us to conceive of other actions, value
beliefs etc. to form ourselves in other ways. Yet, on this basis, how shall we explain
the phenomenon that it can happen that a person never, or at least extremely rarely,
expresses herself in actions in a way that corresponds to her self-ideal? In other
words, on what basis can someone claim to be rather someone else, an ideal version
of herself, if she never expresses this ideal version of herself in reality?
An example shall illuminate this consideration: Assume that a person who main-
tains herself to be orderly and meticulous carries out activities that are consistently
sloppy. Can this person reasonably claim that she is actually completely different –
namely orderly and meticulous – but for whatever reasons consistently acts inau-
thentically; that is that she is permanently “out of character”? In order to be able to
really act inauthentically, so it seems, there has to be at least evidence of another
possible self in a person. Whether this alternative self exists seems in turn not to be
entirely a matter of the person, but also to a substantial degree depend on the assess-
ment of others. If, for example, others never experience a person as polite, then she
is not polite. Surely she can strive to become more polite, but to claim that she is not
really herself when she is impolite seems in these circumstances inconceivable.
The idea of authenticity depends on the possible incongruence between the fac-
tual self-actualisation and another way to determine oneself, which supposedly bet-
ter actualises the self. A distinction between actual and possible self-actualisation
therefore makes philosophical sense. Where single actions contradict observably
consistent characteristics of a person, judgements of inauthenticity indeed seem to
be justified. Yet self-ideals cannot be determined fully detached from how a person
really actualises herself. I do not mean to imply, of course, that people can never
change themselves – only they have to change themselves first before they can rea-
sonably claim that they do not show their true selves when they trespass against
their own ideals. In brief: A Saul might change himself into a Paul. But Paul cannot
claim that he did not actualise himself as Saul.
In summary, Mill’s account helps us to see that the idea of self-formation needs
to be closely linked to the idea of self-actualisation. We can form ourselves, but we
always need to evidence it in expressions of our self. At the same time, Mill’s
account stands for the perfectionist idea of improvement of character. For him, there
is a shape of a human life-form, which we need to endorse if we want to really be
self-determined. Within these contours, we can become an individual; our very
own person.
50 T. Schramme

3.6 Conclusion

Let us summarise what has resulted in the course of this analysis. Self-determination
does not merely mean to actualise a self, but also to scrutinise whether someone
actualises oneself in the way that he or she wants to be or should be. This should
seems to me to always be understood in a hypothetical sense, not categorical, since
it is bound to the perspective of the person herself, insofar as the point is realising
an own self (cf. Christman 1991a). An ideal self that only comes from the outside,
with which one then does not identify, cannot fulfil this criterion. Hence self-­
determination can be understood as a normative concept. Human beings have aspi-
rations that allow them to describe alternatives to themselves and to strive towards
these alternatives (cf. Gewirth 1998). In this regard incongruences can appear
between the self that one expresses in decisions and actions and the self that one
wants to be.
Additionally the possibility arises – which Harry Frankfurt appears to ignore –
that we are the ones who we want to be – or to be more precise: that we are content
with ourselves – and yet at the same time foster alternative self-ideals, because we
see reasons to be different. There is no contradiction involved between acting
wholeheartedly and still striving to be different: In the same way that we can be
content with different lives, we have the possibility to foster different self-ideals in
ourselves, towards each of which we can attempt to strive. We do not need a sophis-
ticated theory for how to choose or how we should choose between these alternative
selves. It is enough to acknowledge that we can actualise ourselves in different ways
and that we can take a stance towards the specific way and the goal of our self-­
actualisation. We identify ourselves with our decisions and actions not merely on
the grounds of a preconceived idea of who we are, but rather evaluate this self on the
grounds of simultaneously available alternative ideas of who we could be. There
appears to be little that can be said regarding to which extent these alternative selves
are available to us. On the one hand we know that human beings can change drasti-
cally, and on the other hand a self-ideal to be actualised should still be one’s own self.
It also does not seem to me to be once and for all clear to which extent we should
better ourselves in order to determine ourselves in the right way, or in order to
become our best selves. In a negative fashion one could agree with Mill that at least
the lower threshold set by general norms of self-development – those of human
nature – restricts the proper scope of self-determination. But the assumption that
one must also actualise individual potentials as best as possible seems to me to go
too far.
I have called Mill’s version of self-determination perfectionist self-actualisation.
Labelled in this way, it presents an ideal-type position within a categorisation that
mainly serves to organise the field of philosophical theories. When discussing Mill’s
theory more closely, it becomes clear that it does not allow for any simple categori-
sation after all. It is perfectionist only in a limited sense, as it leaves a large degree
of leeway to subjective ideas as to how it is best to live. Even if it is first and fore-
most a theory of self-actualisation, its approach also remains obligated to the idea
3 Determining Oneself and Determining One’s Self 51

of self-formation. When we put together the best elements of the discussed theories,
it appears to me that the proper way of self-determination consists in developing
one’s own human potentials as far as one oneself would like, in addition to realising
these potentials in decisions and actions.

Acknowledgements I would like to thank Fabian Wendt, Andrea Esser and participants at the
“Thick (Concepts of) Autonomy” conference in Münster (2012) and at the University of Arizona
(2014) for helpful comments. This paper is a revised version of an article published in German in
Deutsche Zeitschrift für Philosophie in 2011.

References

Berlin, Isaiah. 2002. Two Concepts of Liberty. In Liberty: Incorporating Four Essays on Liberty,
ed. Henry Hardy, 166–217. Oxford: Oxford University Press.
Bransen. Jan. 2002. Making and Finding Ourselves. In Personal and Moral Identity, ed. Albert
W. Musschenga et al., 77–96. Dordrecht: Springer.
Brink, David O. 2003. Perfectionism and the Common Good: Themes in the Philosophy of
T. H. Green. Oxford: Oxford University Press.
Christman, John. 1991a. Liberalism and Individual Positive Freedom. Ethics 101 (2): 343–359.
———. 1991b. Autonomy and Personal History. Canadian Journal of Philosophy 21 (1): 1–24.
Evans, Mark. 2007. Self-Realization: Politics and the Good Life in Modern Times. New York:
Nova Science Publishers Inc.
Feinberg, Joel. 1980. Absurd Self-Fulfillment. In Freedom and Fulfillment: Philosophical Essays,
297–330. Princeton University Press 1992: Princeton.
Frankfurt, Harry. 1971. Freedom of the Will and the Concept of a Person. Journal of Philosophy
68 (1): 5–20.
———. 1982. The Importance of What We Care About. In The Importance of What We Care
About, 80–94. Cambridge: Cambridge University Press 1988.
———. 1987. Identification and wholeheartedness. In The importance of what we care about,
159–176. Cambridge: Cambridge University Press 1988.
———. 1999. On Caring. In Necessity, Volition, and Love, 155–180. Cambridge: Cambridge
University Press.
———. 2004. The Reasons of Love. Princeton: Princeton University Press.
Gerhardt, Volker. 1999. Selbstbestimmung: Das Prinzip der Individualität [Self-Determination:
The Principle of Individuality]. Stuttgart: Reclam.
Geuss, Raymond. 1995. Freedom as an Ideal. Proceedings of the Aristotelian Society,
Supplementary Volume 69: 87–100.
Gewirth, Alan. 1998. Self-Fulfillment. Princeton.
Glover, Jonathan. 1988. I: The Philosophy and Psychology of Personal Identity. London: Penguin.
Gray, John. 1991. Mill’s Conception of Happiness and the Theory of Individuality. In J.S. Mill: On
Liberty in Focus, ed. J. Gray and G.W. Smith, 190–211. London: Routledge.
Green, Thomas Hill. 1879. On the Different Senses of ‘Freedom’ as Applied to Will and to the
Moral Progress of Man. In The Works of Thomas Hill Green, ed. P. Nicholson, vol. II, 308–333.
London: Thoemmes Continuum 1997.
———. 1883. Prologomena to Ethics. In The Works of Thomas Hill Green, ed. P. Nicholson, vol.
IV, 1–427. London: Thoemmes Continuum 1997.
Herrmann, Martina. 2000. Freier Wille ohne Wunschkritik – Autonomie als Zustimmung zum
eigenen Wünschen [Free Will without Critiquing Desires – Autonomy as Affirmation of
52 T. Schramme

Own Desires]. In Autonomes Handeln: Beiträge zur Philosophie von Harry G. Frankfurt, ed.
M. Betzler and B. Guckes, 153–166. Berlin: Akademie.
Mill, John Stuart. 1843. A System of Logic. Ratiocinative and Inductive Part II. In Collected
Works, ed. J.M. Robson, vol. VIII. Toronto: Toronto University Press 1969.
———. 1859. On Liberty. In Collected Works, ed. J.M. Robson, vol. XVIII, 213–310. Toronto:
Toronto University Press 1977.
———. 1861. Utilitarianism. In Collected Works, ed. J.M. Robson, vol. X, 203–259. Toronto:
Toronto University Press 1969.
Oshana, Marina. 2010. The Importance of How We See Ourselves: SelfIdentity and Responsible
Agency. Lanham: Lexington Books.
Quante, Michael. 2007. Person. Berlin/New York: de Gruyter.
Rawls, John. 1999. A Theory of Justice. Revised ed. Cambridge, MA: Harvard University Press.
Schechtman, Marya. 2004. SelfExpression and SelfControl. Ratio 17 (4): 409–427.
Seel, Martin. 2002. Sich bestimmen lassen: Ein revidierter Begriff der Selbstbestimmung [To Let
Oneself be Determined: A Revised Concept of Self-Determination]. In Sich bestimmen lassen:
Studien zur theoretischen und praktischen Philosophie, 279–298. Frankfurt/M: Suhrkamp.
Simhony, Avital. 1993. Beyond Negative and Positive Freedom: T.H. Green’s View of Freedom.
Political Theory 21 (1): 28–54.
Taylor, Charles. 1985. SelfInterpreting Animals. In Human Agency and Language, Philosophical
Papers I, 45–76. Cambridge: Cambridge University Press.
———. 1989. Sources of the Self: The Making of the Modern Identity. Cambridge: Cambridge
University Press.
Tugendhat, Ernst. 1989. Self-consciousness and Self-Determination. Cambridge, MA: MIT Press.
Velleman, David. 2006. Self to Self: Selected Essays. Cambridge: Cambridge University Press.
Chapter 4
Self-Confidence, Self-Assertiveness,
and Self-Esteem: The Triple S Condition
of Personal Autonomy

Johann S. Ach and Arnd Pollmann

Abstract In this paper we seek to clarify the question of what exactly is meant by
an “autonomous” decision or act by focusing on the most fundamental personal
condition for deciding or acting autonomously. This basic personal requirement has
often been overlooked in recent debates; where it has been seen, it is characterized
in ways that are too demanding. What is meant here is an individual form of self-­
relation that seems to be constitutive for leading a life as a human and is likely to be
acquired (or not) very early in life. This significant personal self-relation has three
essential aspects: a primal sense of “self-confidence”, a basic form of “self-­
assertiveness”, and a fundamental form of “self-esteem”. We will call this the “triple
S condition” for autonomy. Our main thesis is that any decision or action which is
not based on this triple S condition cannot be called “autonomous”.
In bioethical discussions the normative principle of respect for the autonomy of
patients plays a key role. However, even if the great normative significance of auton-
omy of the patient can now be considered as widely accepted, there is still contro-
versy around a number of issues concerning the principle of autonomy. Especially
open to dispute is the question of what exactly is meant by an autonomous decision
or action, i.e. what makes an action autonomous. What are the descriptive criteria
for autonomous actions?
In the following contribution we concentrate on one especially basic condition
for autonomous decisions or actions that is often overlooked in discussion or, when
seen, often characterized in a way that is too demanding. What is meant is a basic
form of practical self-relation that has three essential aspects: a primal sense of
“self-confidence”, a basic form of active “self-assertiveness”, and a just as funda-
mental form of “self-esteem”. We will call this “the triple S condition” for

J. S. Ach (*)
Center for Bioethics, Münster University, Münster, Germany
e-mail: [email protected]
A. Pollmann
Alice Salomon Hochschule Berlin, Berlin, Germany
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 53

Switzerland AG 2022
J. F. Childress, M. Quante (eds.), Thick (Concepts of) Autonomy, Philosophical
Studies Series 146, https://doi.org/10.1007/978-3-030-80991-1_4
54 J. S. Ach and A. Pollmann

autonomy, and our main thesis will be that any decision or action not based on this
triple S condition cannot be called “autonomous”.

4.1 The Standard Model for Autonomous Action

What must be the case in order to say that any concrete action is autonomous? The
standard bioethical answer is, on the one hand, that the subject of analysis must
actually perform an “action”, i.e. demonstrate intentional behavior and, on the other
hand, that this action come about in the right way. With regard to this second condi-
tion, Tom L. Beauchamp and James F. Childress (2019), for example, demand that
the action in question, first, be performed by a competent decision-maker who, sec-
ond, understands what is to be decided upon, and who, third, makes this decision
without any controlling influence by other persons or any “inner constraints” due to
illness. The standard model for autonomous actions thus differentiates between con-
ditions of intent, competence, understanding, and voluntariness.
In the following, we will focus on the second component in the standard model,
the condition of competence, since what is to be discussed here as the “triple S con-
dition” for autonomy is commonly described in discussions of autonomy as
“competence”.
In order to be able to speak of a person as (sufficiently) competent to make a
decision or perform an action, it is usually required that this person be (sufficiently)
addressable in regard to supporting arguments and counter-arguments, with regard
to her convictions, as well as to her wishes. She thus has to be in a position to
arrange her wishes in proper order and to be able to envision approximately what
fulfillment or frustration of her wishes would mean for her. She should also be
capable of transforming her wishes into concrete “wanting” and thus to let herself
initiate action – and much more. Some interpreters of autonomy, such as John
Christman, demand in addition to this that the competence condition include the
“wide variety of capacities for care, intimacy, social interaction, and the like” or
that, at least, room be left for these (Christman 2011, 177). This conception is
intended to take into account – as currently discussed under the catchword “rela-
tional autonomy”1 – that personal autonomy is not only a matter of solitary self-­
reflection and fitness, but also the result of interpersonal relationships in areas of
intimacy, care, or even dependency.
The following treats one particular and especially fundamental aspect of this first
condition for competency that evolves in interpersonal relationships and is then
proven to be reliable through them: a complex and qualified form of self-relation we
sum up in the triad of self-confidence, self-assertiveness, and self-esteem. In the
end, however, we show that this complex self-relation, for its part, should not be

1
See, for example, the contributions in: Mackenzie and Stoljar 2000; critical: Ach and Schöne-
Seifert 2013.
4 Self-Confidence, Self-Assertiveness, and Self-Esteem: The Triple S Condition… 55

understood as a personal “competence” to be acquired and exercised. It is rather a

fundamental ethical-existential prerequisite for the competence at issue here,
namely, the capability for autonomous decisions and actions. In other words: this
triple self-relation has little to do with our specific personal capacity for transform-
ing our convictions, wishes, values, or intentions into corresponding actions as
autonomously as possible. It has just as little to do with any skills in interaction,
cooperation, or even care which are necessary in regard to interpersonal relation-
ships. It is rather explicitly a matter of a practical self-relation. Or, to repeat a
phrase by Paul Benson, it is not a question of “what we are capable of doing”, but
rather of “who we regard ourselves as being” (Benson 2000, 85).

4.2 Affirmation, Position, Evaluation

As previously mentioned, the fundamental self-relation for personal autonomy of

action we first want to examine conceptually a little more closely in this section
demonstrates three aspects or facets:
(a) The aspect of “self-confidence” includes not only a self-reflective, but also an
affirmative reference to the actor’s rational, motivational, imaginative, or emo-
tional characteristics and abilities (cf. Govier 1993). If a person is supposed to
be capable of autonomous decisions or actions, she has to possess not only a
basic feeling for the continuity of her person, but also a fundamental “positive”
sense for her own self, an elementary confidence in her most important abilities,
as well as a – at least vague – feeling of her own reliability and a certain “mutu-
ality” of her being and her actions. Anyone who, however, does not trust herself
or her senses decisively, who thus doubts whether she can rely on her own
memories, feelings, or even basic attitudes, or who faces them with distrust, is
already fundamentally incapable of autonomy (Anderson and Honneth
2005, 134).
(b) The aspect of “self-assertiveness” goes beyond basic affirmative self-­confidence
and means that the person acting has a positional reference towards her own
self. The issue here is that persons acting autonomously have to be able to be
aware of their own position and also of their boundaries, and to make these
limits clear, especially to other people: they take a position. The German word
for “self-assertiveness”, “Selbstbehauptung”, expresses this form or facet of
self-reference that makes autonomy possible very well since two things are
built into it: on the one hand, self-assertiveness means that one makes an “asser-
tion” (“Behauptung”) about oneself and, in this sense, takes a position. On the
other hand, it expresses that the self-positioning self raises its head (“Haupt”),
so to speak, and thus takes possession of itself. Accordingly, self-assertiveness
is always also a kind of self-affirmation in the form of experiencing one’s own
effectiveness in the world: I affirm myself through my action as the center of my
actions.
56 J. S. Ach and A. Pollmann

(c) The aspect of elementary “self-esteem”2 is ultimately an evaluative reference

by the person carrying out the action towards herself. If a person is to be capable
of autonomous decisions or actions, she has to consider herself “worthy” of
making a (certain) decision or performing a (certain) action. She has to be able
to understand herself as someone whose wishes and convictions are relevant to
the impending decision or action and whose decisions or actions therefore also
make a difference for her. This does not have to mean that the person should
have to earn the same respect as all the other persons involved in regard to the
impending decision or action in social or legal terms, which is a much further-­
reaching and nowadays customary interpretation. It is rather a basic feeling of
relevance in general. To put it pointedly, it is not a question of earning moral or
legal “respect”, but rather, first and foremost, of not considering oneself as
“nothing”. If a person does not have this minimal feeling of self-esteem, she
will not be in a position to relate her decisions or actions to herself in any way
that is relevant for her. It then makes no difference – for her – whether she takes
action at all or, if she does, how she decides or acts. In a certain sense, she does
not even appear in these decisions or actions.
The aspects of self-confidence, self-assertiveness, and self-esteem thus refer to
three forms or facets of an affirmative, positional, and evaluative reference to herself
by the person taking action and are, as such, constitutive for the possibility of auton-
omous decision or action. In other words, for autonomous decision-making and
action, a fundamental affirmation, position, and evaluation of one’s own self belongs
in actu. As will be shown below, “fundamental” means two things here: it is not only
a matter of the three facets of self-relation that have to be present in at least a mini-
mal manner and are especially fundamental in the sense that they have to be acquired
before any autonomy. Additionally important is the point that the person taking
action does not have to be explicitly aware of any of these forms of self-reference in
the act of decision-making or taking action. Instead, it is usually the case that the
person only becomes conscious of them when they purposefully direct their atten-
tion to them, for example because these self-relations are impaired or have been
violated. They are, however, an important part of a person’s self-perception, which
is comprised of various forms of confirmatory self-reference, in addition to her
convictions, wishes, values, and emotional inclinations.3

2
This elementary form of self-esteem should not be confused with more demanding forms of quali-
fied self-relations which are occasionally termed “self-esteem” (for example, Honneth 1996, chap-
ter 5), but which are only realized when the (adult) person knows she is recognized as a fully
adequate member of the social community.
3
This suggestion is similar to Joel Anderson’s and Axel Honneth’s “recognitional account of
autonomy” (Anderson and Honneth 2005; cf. Honneth 1996; Mackenzie 2000). This approach
goes too far in two respects, however. Firstly, the conviction of both authors may be debatable that
“full autonomy” presupposes not only fundamental self-confidence, but also moral-legal self-
respect, as well as socially conveyed self-esteem, namely with regard to an ethical autonomy-ideal
of the good life. It would, however, be asking too much to want this demanding self-relation to be
regarded as a prerequisite for autonomous action. Secondly, this suggestion seems to demand that
4 Self-Confidence, Self-Assertiveness, and Self-Esteem: The Triple S Condition… 57

4.3 The Genesis of Autonomy

Philosophy has even more difficulty in regard to genetic prerequisites for autono-
mous action and the question of their developmental origins. These involve pre-
sumptions about empirical circumstances at the beginning of human life, which are
also difficult to verify empirically. It can find resources in developmental psychol-
ogy, on the one hand, and socialization research, on the other – complicated by the
fact that both of these neighboring disciplines offer anything but a uniform theoreti-
cal framework.4 Furthermore, we must avoid interpreting these reconstructions
anthropologically from the outset, i.e. as a reconstruction of developments in life
everyone has gone through. The following considerations are valid only in social,
historical, or cultural contexts for which the value of personal autonomy is of cen-
tral importance. There may be societies in which this is not the case, even ones that
are hostile towards autonomy. In such social contexts, the psychological and social-
ization processes that are significant for adolescents have a different nature (see, for
example, Trommsdorff 1999).
With regard to contexts sympathetic to autonomy, it is debatable which “phase”
of personality development should be investigated for the genetic origins of auton-
omy that have been summarized here for the triple S condition. Looking back, step
by step, from the perspective of personal adulthood, life can be retraced – rather
simplistically – through the stages of adolescence, puberty, the so-called “latent
period”, early childhood, and on back to infancy. Viewed in light to the three auton-
omy factors of affirmative self-confidence, positional self-assertiveness, and evalu-
ative self-esteem, there may be dramatic experiences and developments in each of
these phases. Moreover, it is not predetermined from the outset that all these dra-
matic processes of development can be dated as postnatal. The possibility must be
calculated, not only in medicine but also in modern developmental psychology,
research on empirical socialization, and pedagogy, that a person can also have pre-
natal influences, and that birth itself is a dramatic psychological and emotional
experience that also stamps a person perinatally (after birth) (see Trommsdorff
1999; Janus and Haibach 1997).
Relevant literature presents three basic guiding premises concerning the genesis
of competence for autonomous decisions and actions.5 Firstly, genetic foundations
for the three autonomy factors of self-confidence, self-assertiveness, and self-esteem

every autonomous person have “undistorted” or “non-pathological” relationships to her social

environment. This would mean that persons who do not consider themselves integrated in such
acknowledged relationships are thus not able to make any autonomous decisions, which leads to
the reproach of perfectionism susceptible to paternalism.
4
In the psychological respect, we will refer essentially to the psychoanalytical “object-relations
theory”, as well as to the tradition of “psycho-historical I-psychology” according to Erik
H. Erikson. We are not able to elaborate on the tensions between these two traditions here. For an
initial overview: Mitchell and Black 1995. Concerning socialization, we base ourselves especially
on “symbolic interactionism” according to Georg H. Mead see Abels 2010.
5
We draw primarily on classic papers: Winnicott 1965; Erikson 1977, 11.
58 J. S. Ach and A. Pollmann

are laid substantially in the first three years of a child’s life. They may be developed
further and even changed during the subsequent course of a person’s life, but the
respective basic prerequisites have to arise very early in life. Secondly, the forma-
tion of the self-relation outlined by these three factors is the result of typical crises
in early childhood that cause some children to become stronger and others weaker.
In each case, this depends especially upon whether and how the first reference per-
sons for the child are successful in compensating for the ruptures experienced in the
relevant crises within the childhood world by building up a confirmatory emotional
structure of self-confidence, self-assertiveness and self-esteem. Thirdly, in later life,
the quality and stability of self-relation depend upon the success of these infant care
relationships, where, first of all, the relationship of the child’s self to its earliest
reference persons has to be taken advantage of and where the child is compensated,
sometimes more, sometimes less, for ruptures experienced in its own childhood
world of experiences that are absolutely necessary for its life-history. In each case,
special demands arise in view of the triple S factors.
(a) The first important form of emotional confirmation the child is able to experi-
ence from others is assuring care – especially from the mother, but also from the
father, siblings, or hospital staff – directly following birth (see Erikson 1977,
62ff.; cf. Leu 1999, 83f.). After a phase of complete intrauterine care, which is
often described in psychological literature as a “symbiosis” between mother
and child, the newborn experiences genuine social or inter-subjective care for
the first time; it immediately experiences whether it can rely on other people
(who, at first, are only recognizable as vague shadows).6 The infant needs
warmth, nourishment, and emotional care and attention, and getting this is a
confirmation that creates trust in its protective surroundings. If the infant does
not get this confirmation or only gets it insufficiently, for example, as the result
of temporary or even permanent absence of caring reference persons, a corre-
sponding fundamental feeling of social distrust may already prevail after a short
time.7 In both cases, however, the newborn experiences itself as physically and
emotionally needy and dependent, but, at the same time, also as partially
“autonomous” in the sense that it experiences itself as different from its sur-
roundings. This, in turn, is a part of the experience, important for maturity pro-
cesses and for later life, that temporary phases with absence of others, lack of
care, “cutting the umbilical cord”, and being alone are actually “survivable” (cf.
Mahler et al. 1975). More and more, this also turns the often crisis-prone trust
in others into a first form of affirmative self-confidence that, in the beginning,
might still be based on the literally organically grown feeling that the child can
trust not only its own reference persons, but also itself and the ability of its own
organs to cope with its inner drives (Erikson 1977, 70).

6
As an introduction to research on infants, still very worth reading: Dornes 1993.
7
Trust and mistrust do not exclude each other in the world of a child since each child has temporary
experiences of sometimes adequate and sometimes inadequate care. Psychologically decisive for
the process of maturing is whether it is fundamental social trust or fundamental mistrust that
remains.
4 Self-Confidence, Self-Assertiveness, and Self-Esteem: The Triple S Condition… 59

(b) Once immediate postnatal care is fairly well ensured, the child’s second experi-
enceable form of emotional confirmation by others has to do with the recogni-
tion it demands from its reference persons that it increasingly demonstrates its
“own will“. Even the needy newborn’s first screams can be understood as pre-
cursors of the demonstrative articulation of the child’s personal will to make
itself be recognized as its own person and thus already an “autonomous” being
in specific circumstances. To some extent, the child itself is aware of its own
will when it does not experience its surroundings as compliant, but rather as
emotionally and materially inadequate or even resistant to its demands.
Consequently, the self-reflective insight that one has an autonomous will seems
to be the developmental result of repeated frustration of this will. Moreover, the
child’s demonstrative “fight” for the gratification of its needs, for example by
screaming or throwing objects, increasingly becomes a “life and death” struggle
(Hegel), or at least a struggle for self-assertiveness. The child learns to assert its
will, even in face of opposition. To put this psychological situation anatomi-
cally, the latent double wish in the term “self-assertiveness” mentioned above –
the wish for positional self-affirmation, on the one hand, and for ensuing
self-appropriation by raising its own head, on the other – is soon expressed by
the child months later when it literally begins “to stand on its own two feet”.
And, while the child is gaining these first experiences about its own effective-
ness, the quality and stability of the self-assertiveness it has generated will, in
the end, depend on whether the child’s reference persons are willing not only to
tolerate this early form of autonomy, but also to be noticeably affirmative
towards it and recognize it.
(c) Thus, it is only when affirmative self-confidence has evolved in the first step
and a just as fundamental capability for positional self-assertiveness in the sec-
ond that the end result can be what has been reconstructed here as the basic
genetic form of evaluative self-esteem. Over time, defense of one’s own will on
the basis of elementary self-confidence becomes leading a life that is increas-
ingly diverse in content and also more demanding socially. Within this frame-
work the child is not only increasingly experienced as a passive recipient of
attentive benefits, but also experiences herself as an “active player” who seems
to be important. After a while, the child usually does not feel alone with her
need for interaction and closeness. The child’s first reference persons, upon
whom she may, at first, feel one-sidedly dependent, also show that they long for
commitment and interchange.8 And, in phases of temporary separation and
when faced with the subsequent “joy of reunion”, the child realizes not only that
other people are of great value, but also that it is very important for others, too.
It is this elementary appreciation that is internalized and transformed into feel-
ings of self-esteem and of one’s own “autonomous” relevance – a feeling based
on the certainty, which cannot be taken for granted and must sometimes be
sought anew, that one’s own life is relevant, for oneself and for others.

8
For the complexity of infant interaction and the child’s contribution to this, see the significant
work by Martin Dornes (for example, Dornes 1993, 2006).
60 J. S. Ach and A. Pollmann

Table 4.1 Triple S Developmental Stages in the Child’s First to Third Yearsa

Stage 1 SELF-CONFIDENCE earlier forms earlier forms

Stage 2 later forms SELF-ASSERTIVENESS earlier forms

Stage 3 later forms later forms SELF-ESTEEM

a
We adopt and modify a diagram from Erikson (1977, 60)

­ ccordingly, a child – and then later perhaps the adult, too – will partially lack
A
this evaluative self-relation if she does not see her own relevance reflected in
others (Table 4.1).9
Explanation of the table: Pertinent studies in developmental psychology surmise
that each of the three autonomy factors develops at a different time in an infant’s life
(each underlain in grey), essentially as a reaction to typical biographical crises:
inadequate care, obstacles, separation. In addition, it can be assumed that the three
stages are necessarily built upon one another. This means that without affirmative
self-confidence, there is no positional self-assertiveness, and, without this, no evalu-
ative self-esteem. Of course, this does not mean that the three aspects come about
abruptly, completely, and irrevocably in the child’s development. It is rather that
they are initiated (“earlier forms”), develop (underlain in grey), and then progress
further (“later forms”). This marks a significant gradualism in the model that extends
from preliminary stages of insufficiently developed self-relation to sufficiently
developed main stages. It also means that under beneficial surrounding conditions,
these self-relations “mature” (cf. the title of Winnicott 1965). Nevertheless, they
remain susceptible to injuries in later life, as will be shown below.
Furthermore, the arrows at the bottom and right edge of the table signal that the
three autonomy factors after early child development in the first 3 years may go
through further stages of development (↓) and that still more fundamental aspects
(→) that could be relevant for autonomous action and decision-making in these
early phases of development of individual self-relation may also follow. What these
might be remains unanswered here.10 We limit ourselves to the three, in our opinion,
central self-relations. Nevertheless, the basic openness of the model should be
emphasized since it allows for an important second gradualism: as mentioned at the
beginning, this analysis focuses on conditions for autonomous competence in
decision-­making and actions. However, the arrows signal that these only mark one
end of a scale for the development of autonomy whose end is the fully developed
ideal of totally autonomous life.11

9
For the biographical consequences of disturbances in early relationships, see the groundbreaking
work by John Bowlby (for example, Bowlby 1982).
10
Consideration should be given to whether moral or even motor abilities are also important for the
genesis of the capability for autonomous action, for example.
11
See also the eight-step model for “psycho-historical development”: Erikson 1977.
4 Self-Confidence, Self-Assertiveness, and Self-Esteem: The Triple S Condition… 61

4.4 The Vulnerability of Autonomy

The genetic triple S stages are accompanied by typical susceptibilities and vulner-
abilities that continue at least partially in later life and then – from a socio-­
philosophical viewpoint – will, in certain cases, have an effect as social “pathologies”
of autonomy, as they are sometimes termed in socio-philosophical recognition theo-
ry.12 When viewing the obvious parallels between the history of genetic develop-
ment, on the one hand, and the later course of life, on the other, recognition theory
admittedly tends to overlook the relative independence of the two. There may be
good reasons for the view that a person may be more immune to injuries to her
autonomy in later life the more extensively she has experienced recognition in earli-
est childhood. Exactly the opposite could also be true: the less self-confidence, self-­
assertiveness, and self-esteem developed in earliest childhood, the more susceptible
the interplay between autonomy and commitment for an adult in view of the recog-
nition denied her.13 However, a person who has grown up with very little self-­
confidence, self-assertiveness, and self-esteem is not “lost”, in the sense that these
early deficits cannot be compensated for in later life. Nor is a child with very pro-
nounced triple S factors completely immune from dramatic feelings of insecurity in
later life. However, the three basic autonomy factors prove to be a “foundation” in
the sense that the respective convulsions have to be fundamental. In the following,
we focus on convulsions that result from problematic or even pathologically dis-
torted relationships to others.14
(a) As stated above, a person’s elementary self-confidence is based on the feeling
of being able to rely especially on her own body and its most important func-
tions, as well as on her own senses. Since this self-confidence is genetically the
result of reflection on the trust a person has built up and placed in her first refer-
ence persons and in their qualities of care and protection, inter-subjective inju-
ries to self-confidence often assail this spot in later life – interpersonal
encroachments or assaults15 that cause the person, who feels confident that oth-
ers have a fundamentally peaceful, friendly, or even loving attitude, to lose fun-
damentally the feeling of being able to rely on herself, her own body, her
abilities, or even her own thinking processes. Since the deepest layer of practi-
cal self-relation is affected, the aggression or invasive harm is often

12
For clarification of the socio-philosophical term in pathology in connection with Axel Honneth,
see: Pollmann 2005, 30–37.
13
See the fascinating analysis: Benjamin 1988.
14
There are also non-social adversities, for example, the failure of great hopes or a serious illness.
However, we blank out any non-social negative experiences since we want to indicate the extent to
which these thoughts have normative or moral consequences and that moral consequences are only
the result if other people are involved in a demanding or hurtful way.
15
“Encroachments” or “assaults” on a person do not at all necessarily lead to partial or complete
loss of triple S factors. Whether this loss comes about depends greatly on how stable the self-
relation concerned is. Consequently, the more stable a self-relation is, the graver, more far-reach-
ing, and more aggressive the offence has to be if the person suffers lasting damage.
62 J. S. Ach and A. Pollmann

c­orrespondingly grave and occurs rather seldom, as in cases of physical

violence, sexual abuse, rape, or torture, and even gross negligence, isolation, or
brain-washing practices.16
(b) The aspect of fundamental self-assertiveness, on the other hand, comes from
the developmental fact that the small child who is frustrated in her needs
increasingly feels a gap between her will and the often resistant will of others.
Later life holds numerous such frustrations that will compel the person con-
cerned to resist the will of others and, if necessary, implement her will despite
opposition. This self-assertiveness may be successful or may fail at times, and
compromises will often have to be made. In order for self-assertiveness to be
fundamentally impossible, other people have to take away, in a very basic way,
any chance of one’s asserting one’s own will and seeing one’s own personal
function confirmed (or at least tolerated) in the social surroundings. This could
come from invasive encroachments or immoral behavior that leave fundamental
doubts about the existence of one’s own will (which it would be important to
assert), or at least make it difficult to continue to exist with one’s “head held
high”: compulsion and coercion, instrumentalization and exploitation, but also
acts of indignity, abasement, and humiliation.17
(c) Finally, the aspect of fundamental self-esteem has been interpreted here as the
developmental result or incorporation of intersubjectively conveyed awareness
that one is not only dependent upon other people, but also has significant worth
for at least some of these others. Consequently, the practical certainty sets in of
individual but also social relevance in life that will later support one existen-
tially, but can also be convulsed in later life if someone gives the person the
feeling of being “nothing” of any importance. This is especially the case in situ-
ations and through unsocial practices whereby someone not only contests that
the person has the same moral or legal worth as any other human being, but
denies her any value as a human (see, for example, Margalit 1996). In such a
case, the complete deprivation of evaluative relevance is accompanied by the
danger of incorporating this experience into a feeling of one’s own irrelevance
or inferiority. Some examples of forms of unsocial behavior showing this ten-
dency are demonstrative forms of ignoring or failing to notice a person, social
exclusion, the deprivation of liberty, and especially also severe forms of mob-
bing, disregard, or discrimination.18
The complex relationship between conditions for susceptibility to injury in later life
linked with the triple S factors is illustrated in the following table, which should be
interpreted as follows: once the three central self-relations of self-confidence, self-­
assertiveness, and self-esteem have been roughly established in infancy, they result
in three “deep layers” of practical self-relations, one on top of the other. This means,
on the one hand, that an injury to deeper layers can “break through” to the layers

16
See the analyses oriented on phenomena: Scarry 1987; Sofsky 1996.
17
See the relevant articles in: Pollmann 2010.
18
See also articles in: Pollmann 2010.
4 Self-Confidence, Self-Assertiveness, and Self-Esteem: The Triple S Condition… 63

Table 4.2 Triple S Susceptibility in Later Life

Dramatic losses →↓

Deep layer 1 SELF-CONFIDENCE →↓

Deep layer 2 ↑← SELF-ASSERTIVENESS →↓

Deep layer 3 ↑← SELF-ESTEEM

↑
← Increasing invasiveness

above (→↓). It also means, on the other hand, that invasive encroachments on exter-
nal self-relations with others have an even more serious effect the deeper they pen-
etrate into that external self-relation with someone else (↑←) (Table 4.2).

4.5 Conclusion

Among the elementary competences a person has to have in order to be capable of

autonomous decision-making and action, there is the triad of self-confidence, self-­
assertiveness, and self-esteem, in addition to various rational and social abilities.
These self-relations are formed and developed in interactive childhood relation-
ships, especially in intimate family relationships with material and emotional care
and attention. Relevant reconstructions make it questionable, however, whether it is
actually a matter of acquiring corresponding “competences” in regard to these three
self-relations, as bioethical discussion seems to suggest. On the one hand, the triple
S factors prove to be so fundamental in the history of a person’s life that they should
rather be understood as an existentially indispensable basic prerequisite of the abil-
ity to make decisions and perform actions autonomously (and thus not as a triple
competence). Moreover, the term competence suggests that the abilities in question
can be acquired, learned, and practiced individually. However, this typical assump-
tion seems to be misleading for the triad of self-confidence, self-assertiveness, and
self-esteem, at least from a genetic perspective, when it is a question of early child-
hood acquisition of this triple self-relation or its failure. In view of the early devel-
opmental history of a person, it must be concluded that the very first self-relation
that should be acquired in partially “unavailable” relationships supplying love and
care may be unavailable to the persons concerned and their reference persons. This
means that it is a question of biographical “fate” whether a child gets the care and
attention she needs very early in order to be able to make decisions and perform
actions autonomously in later life.
Even if we have spoken of only the descriptive criteria for autonomous actions in
this contribution, some normative conclusions can nevertheless be drawn from our
observations. Whoever, for whatever reasons and in whatever way, is interested in
64 J. S. Ach and A. Pollmann

assuring that persons are capable of autonomous decisions or actions will, firstly,
have to ensure that they are able to grow up in more or less intact social relation-
ships and experience more or less intact intimate relationships. It is evident that this
insight would have far-reaching social, political, cultural, and even pedagogical
consequences. This would, however, be the normative aspiration that follows from
our genetic analysis. Furthermore, whoever, for whatever reasons and in whatever
way, is interested in assuring that persons are capable of autonomous decisions or
actions will, secondly, have to ensure that the triad of self-confidence, self-­
assertiveness, and self-esteem is not infringed upon or even destroyed through (sys-
tematic) invasive encroachments or unsocial practices in later life. This is the
normative aspiration that follows from the socio-philosophical perspective of our
observations. The reason for this double aspiration is that more or less intact inti-
mate and social relationships are indispensable not only for acquiring a self-relation
enabling autonomy, but also for preserving it later, which means that the demand for
absence and abolition of certain forms of destructive social relationships is norma-
tively urgent in both respects.

Conflict of Interest The authors state that there are no conflicts of interest.

References

Abels, H. 2010. Interaktion, Identität, Präsentation. 5th ed. Opladen: Verlag für
Sozialwissenschaften.
Ach, J.S., and B. Schöne-Seifert. 2013. “Relationale Autonomie”– Eine kritische Analyse. In
Patientenautonomie. Theoretische Grundlagen – Praktische Anwendungen, ed. C. Wiesemann
and A. Simon, 42–60. Münster: Mentis.
Anderson, J., and A. Honneth. 2005. Autonomy, Vulnerability, Recognition, and Justice. In
Autonomy and the Challenges to Liberalism, ed. J. Christman and J. Anderson, 127–149.
New York/Cambridge: Cambridge University Press.
Beauchamp, T.L., and J.F. Childress. 2019. Principles of Biomedical Ethics. 8th ed. Oxford:
Oxford University Press.
Benjamin, J. 1988. Bonds of Love. New York: Pantheon Books.
Benson, P. 2000. Feeling Crazy: Self-Worth and the Social Character of Responsibility. In
Relational Autonomy, ed. C. Mackenzie and N. Stoljar, 72–93. Oxford/New York: Oxford
University Press.
Bowlby, J. 1982. Attachment and Loss. 2nd ed. New York: Basic Books.
Christman, J. 2011. The Politics of Persons. Cambridge: Cambridge University Press.
Dornes, M. 1993. Der kompetente Säugling. Frankfurt a.M.: Fischer.
———. 2006. Die Seele des Kindes. Frankfurt a.M.: Fischer.
Erikson, E.H. 1977. Wachstum und Krisen der gesunden Persönlichkeit. In E. H. Erikson: Identität
und Lebenszyklus, 55–122. Frankfurt a.M: Suhrkamp.
Govier, T. 1993. Self-Trust, Autonomy, and Self-Esteem. Hypatia 8 (1): 99–120.
Honneth, A. 1996. Struggle for Recognition. Cambridge: MIT Press.
Janus, L., and S. Haibach, eds. 1997. Seelisches Erleben vor und während der Geburt. Neu-­
Isenburg: LinguaMed.
4 Self-Confidence, Self-Assertiveness, and Self-Esteem: The Triple S Condition… 65

Leu, H.R. 1999. Die “biographische Situation” als Bezugspunkt eines sozialisationstheoretischen
Subjektverständnisses. In Zwischen Autonomie und Verbundenheit, ed. L. Krappmann and
H.R. Leu, 77–107. Frankfurt a.M: Suhrkamp.
Mackenzie, C. 2000. Imagining Oneself Otherwise. In Relational Autonomy, ed. C. Mackenzie and
N. Stoljar, 124–150. Oxford/New York: Oxford University Press.
Mackenzie, C., and N. Stoljar, eds. 2000. Relational Autonomy. Oxford/New York: Oxford
University Press.
Mahler M.S., Pine F., and Bergman A., eds. 1975. The Psychological Birth of the Human Infant:
Symbiosis and Individuation. New York: Basic Books.
Margalit, A. 1996. Decent Society. Cambridge: Harvard Univ. Press.
Mitchell, S.A., and M.J. Black. 1995. Freud and Beyond: A History of Modern Psychoanalytic
Thought. New York: Basic Books.
Pollmann, A. 2005. Integrität. Bielefeld: Transcript.
———. 2010. Unmoral. Ein philosophisches Handbuch. München: C.H. Beck.
Scarry, E. 1987. The Body in Pain: The Making and Unmaking of the World. Oxford: Oxford
University Press.
Sofsky, W. 1996. Traktat über die Gewalt. Frankfurt a.M.: Fischer.
Trommsdorff, G. 1999. Autonomie und Verbundenheit im kulturellen Vergleich von
Sozialisationsbedingungen. In Zwischen Autonomie und Verbundenheit, ed. L. Krappmann and
H.R. Leu, 392–419. Frankfurt a.M: Suhrkamp.
Winnicott, D.W. 1965. Maturational Processes and the Facilitating Environment. New York:
International Universities Press.
Chapter 5
Autonomy, Respect, and Joint Deliberation

John Christman

Abstract Respecting the autonomy of agents grounds various obligations to others

such as non-interference, deference to her authority over self-regarding decisions,
limitations on paternalism, and so on. According to a broadly liberal moral sensibil-
ity, respecting others in this way implies accepting (in some sense) the values they
autonomously hold even if they are judged problematic, immoral, self-destructive,
or otherwise non-ideal. In discussions of such respect, it is generally assumed that
persons expressing that respect (or not) have no direct bearing on whether the sub-
ject of that respect is herself autonomous. However, in many situations, persons
interact in a way that helps establish or re-establish the autonomy of one or both of
these agents themselves.
For example, the would-be paternalist may be a therapist whose professional
obligation is to facilitate the process of re-establishing autonomy for vulnerable
victims of trauma. In such scenarios, the usual lines between hard and soft paternal-
ism, as well as the standard liberal rejection of the former, do not apply. Still, it
would be wrong for the aid worker to simply impose her views of a decent life on
the struggling person. How, then, do we reformulate restrictions on paternalism and
other such normative strictures grounded in respect for autonomy in scenarios
where the relationship between client and aid worker is itself a crucial part of the
process that results in the self-governing agency of the client?
This paper explores these issues and argues that such (very typical) scenarios
indicate that conceptions of autonomy must view the self as diachronic as well as
socially constituted but also must be sensitive to the ways that autonomy can some-
times be (re-)established only with others’ help. In particular, the obligation to
remain relatively neutral toward the content of others’ values in showing respect for
their autonomy (the broad liberal sensibility) must be reformulated to take into
account the way interpersonal dynamics themselves help establish the autonomy
with which persons hold the values that they hold. These observations are applied to
certain cases of aid and care where such questions are central.

J. Christman (*)
Philosophy, Political Science, and Women’s Studies, Penn State University,
University Park, PA, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 67

Switzerland AG 2022
J. F. Childress, M. Quante (eds.), Thick (Concepts of) Autonomy, Philosophical
Studies Series 146, https://doi.org/10.1007/978-3-030-80991-1_5
68 J. Christman

5.1 Autonomy, Respect, and Joint Deliberation

The moral demand that we respect the autonomy of other agents grounds various
obligations to others such as those regarding non-interference, deference to their
authority over self-regarding decisions, limitations on paternalism, and so on.
According to a broadly liberal moral sensibility, respecting others in this way
implies accepting (in some sense) the values they autonomously hold even if they
are judged to be problematic, immoral, self-destructive, or otherwise non-ideal.
However, in discussions of such respect, it is generally assumed that persons
expressing that respect (or not) act and reflect independently of the object of respect.
But in many situations, persons interact in a way that helps establish or re-establish
the autonomy of one or both of these agents together or which partly constitutes the
exercise of their autonomy.1
For example, two romantic partners are considering whether to have children or
not. Such a choice has profound implications for each of their lives, but their com-
mitment to each other means that they want the decision and the reasoning going
into it to be mutual and collaborative. Both rest on their values and self-defining
commitments as separate individuals but also allow their thinking to be shaped by
their interaction and mutual commitment. Certainly this commitment and associ-
ated trust would be violated if one of the partners engaged in manipulation in order
to get the other to come to his way of thinking. The obligation to respect the wishes
and perspective of the other would be violated.
Another example might be an aid worker or therapist whose professional obliga-
tion is to facilitate the process of re-establishing autonomy for vulnerable victims of
trauma. In such scenarios, the standard liberal principle to respect the autonomy of
others in their decisions is based on the obligation to respect the person’s wishes if
she is autonomous or to treat her with care and benevolence if she is not. But in
either case it would be wrong for the aid worker to simply impose her views of a
decent life onto the struggling person. How, then, do we reformulate restrictions on
paternalism and other obligations grounded in respect for autonomy in scenarios
where the relationship between one person and another is itself a crucial part of the
dynamic by which the autonomy of one or both of them is established or exercised?
What further complicates this issue is that while autonomy has traditionally been
thought to apply paradigmatically to individuals considered in themselves, without
essential reference to relations with others, that presumption has been rejected
recently in theories of autonomy that either see that property as essentially rela-
tional or as applying to persons understood as constituted by social and interper-
sonal relations. In the situations being imagined here, however, the deliberating
partners are engaged in actions that either help establish or help the exercise of the
other partner’s autonomy by way of their specific interaction. The problem is how
each person can show respect for the autonomy of that other in such intertwined
situations.

1
I have discussed similar issues in Christman 2014a.
5 Autonomy, Respect, and Joint Deliberation 69

In what follows, I want to lay out the problem as I see it and propose (guardedly)
a solution. I will proceed by first sketching the broad outlines of a view of autonomy
that sees it as a trait applying to individuals but which nevertheless contains ele-
ments that relate to the social and relational dynamics between persons. I will also
outline what I mean by the “liberal sensibility”, meaning a set of pro tanto duties to
defer to the values of others out of respect for their autonomy.2 In Section III, I will
describe what I call “partnerships” where the interpersonal dynamics of two indi-
viduals have a causal and constitutive relation to the establishment or exercise of
autonomy of one or both of them. I will then describe how such partnerships pose a
challenge for the liberal sensibility I mentioned. In the closing section I will suggest
a set of guidelines for interpersonal respect in such situations which reflects the
requirements of autonomy (as I have described it) and remains faithful to the liberal
sensibility by which we generally express respect for one another. My overall con-
clusion is that the obligation to show respect for another’s autonomy remains in
force in such partnerships but must be reformulated to take into account the way that
interpersonal dynamics themselves help establish the autonomy with which persons
hold the values that they hold. Such a conclusion implies, in turn, that we see auton-
omy as not only socially structured but extended over time.

5.2 The General Contours of Autonomy

In order to consider the demand that we respect others’ autonomy we need to look
closely at the idea of autonomy itself. As we will see, that idea is hardly a simple
one and its particular components will operate differently in the mechanics of coop-
erative deliberation of the sort we will consider. To get clear on those operations, let
us take a brief overview of the standard contemporary understandings of the idea of
autonomy that populate the recent literature.
Autonomy can refer to particular decisions or to the status of the person gener-
ally. Autonomy can also be seen as a matter of degree or as a threshold property.
Insofar as autonomy marks the aspect of persons by which moral respect is grounded,
then it must be seen as an all or nothing affair, at least insofar as prescriptions of this
sort of respect do not vary in degree. Moreover, in cases of interest here, the respect
in question regards major decisions that rest on the person’s most basic value
scheme, in respect of which the person’s autonomy is in question.
Standard approaches to autonomy see that characteristic as the ability of the
person to guide her life from her own perspective rather than be manipulated by
others or be forced into a particular path by surreptitious forces. Conditions for
autonomy in this sense have typically included what have been called competency

2
As I will explain, the phrase “defer to” does not mean one must abide by or follow others’ values,
but the posture I refer to here includes obligations to pay special heed to others’ desires and values
in some way (which may include not interfering with their pursuit of them even when one thinks
they are wrong-headed).
70 J. Christman

conditions, on the one hand, and authenticity conditions, on the other. The former
refer to the agent’s ability to form intentions and engage in action voluntarily, and
with sufficient information. These conditions may also include requirements of
rationality, motivational competence, self-control, emotional stability, and other
conditions.
More complex are the requirements of what is here called “authenticity”.
Requirements of authenticity attempt to determine when the values, desires, charac-
ter traits, and so on that orient a person’s choices and move her to action are truly
her own in a way that coincides with the idea of self-government. Those who are
acting competently but on the basis of values that they have been forced or manipu-
lated into adopting, or which arise out of pathological, oppressive, or overly con-
straining conditions, are not acting as self-governing agents. Various accounts of
this set of conditions have been given. For example, many follow Harry Frankfurt
and Gerald Dworkin in requiring that the person be able to critically reflect on her
first level motives and accept them as her own or in keeping with her deepest cares.3
Lack of autonomy can stem from many conditions on this model then. Compulsion,
temporary or protracted mental disability, and addiction are most often mentioned
as examples, as are brainwashing and manipulation of the psychic economy of the
agent. These conditions undercut the person’s ability to reflectively accept her first-­
order motives and/or the way that she has come to develop them.
However, it has been claimed that such a view assumes an overly individualistic
understanding of agency and value. For example, those whose self-conception
makes reference to a cultural, religious, or social role but who are permanently
removed from the conditions in which such a role can be exercised, may lack the
capacity for self-governing agency, and not simply because of a lack of a capacity
as an individual. The goals in question may no longer make sense in the new condi-
tions because the social interactions that gave them meaning no longer take place.
The person’s practical identity by which he or she considers options, makes plans
and considers questions of value is completely fractured because of this social
breakdown.4
There are also many different accounts of the role of social interaction for auton-
omy. On my own view I have tried to stress the socio-historical nature of the self
who is doing the reflecting and the social nature of many of the values and commit-
ments that are the object of that reflection. Additionally, I stress how the autono-
mous person must accept her values and motives (and other life conditions) as part
of an ongoing historical self-narrative, so that autonomy must be seen as diachronic
in these ways. Hence, a person is autonomous if she can reflectively accept her con-
dition in light of its history and ongoing narrative and do so without deep

3
See Frankfurt 1987 and 1992 and Dworkin 1989.
4
Jonathan Lear examines the case of Plenty Coups, the leader of the Crow Indian nation in the
U.S. whose traditional tribal way of life had been completely destroyed. On Lear’s analysis, Plenty
Coups’s values, motives and social self-understanding was thereby fractured, making him unable
to find a way forward until he was able to reconstruct a sense of social identity in a new setting. See
Lear 2006
5 Autonomy, Respect, and Joint Deliberation 71

self-alienation, given her social situation.5 Others have insisted that others’ recogni-
tion of the person’s normative authority or ability to answer for herself be part of the
social conditions of autonomy.6
So for now let us say that the central test for autonomy that I favor (omitting
many details) is that the autonomous person is such that she is disposed to reflec-
tively accept her motives (or other features of herself relevant to action) without
alienation in light of her history and social condition as part of an ongoing personal
self-narrative. This dispositional reflective self-acceptance is meant to help avoid
the problems of regress or groundlessness raised against standard hierarchical mod-
els of autonomy where second-order reflection plays a crucial endorsing role for
first-order attitudes and motives. On my view, these problems are avoided insofar as
these reflections arise out of the agent’s diachronic practical identity, the broad and
deep set of evaluative structures that orient judgment for the person. This borrows
language from Christine Korsgaard but is also akin to what Frankfurt refers to as our
deepest cares and what Michael Bratman refers to as our self-defining projects
(Frankfurt 1999; Korsgaard 1996; Bratman 2007, ch. 2).
Also, however, my view accepts the idea that much intentional, self-governing
action is not guided by actual reflection. Many such actions are habitual, automatic,
or arising from such strongly held convictions where consideration of doing other-
wise might be “unthinkable” for the person, in Frankfurt’s phrase. The reflective test
I use here is dispositional and involves various additional requirements ensuring
that the reflection in question adequately expresses the viewpoint and evaluative
orientation of the agent.7 I will return to this point.
Indeed, intentional self-governing action on my view merely involves what we
can call reflective self-monitoring: this is the function of passive acceptance of the
motives that guide action but with a disposition to “step in” as it were if things go
awry. Like a driver whose mind wanders, one is still in control of oneself because
one is able to snap into concentration and change one’s direction or speed if the
occasion called for it. I call such self-monitoring “reflective” only in the sense that
the person is in a position to give reasons for her habitual action if asked. The reflec-
tion need not be “deep” in the sense of bringing into question the very value forma-
tion that defines her as an agent (her practical identity), but the person is able to give
an account to others if asked under appropriate conditions.8 That is to say, the per-
son must be able to give reasons that reflect her practical identity and which, in turn,
reflexively affirms her commitment to that identity. Such reflective reasoning, of
course, may come to an end in statements like “that is the way I am” but they must
also express statements of the form “and that is a worthwhile way to be”.

5
See Christman 2009.
6
See, e.g., Mackenzie 2008, Westlund 2003, Anderson and Honneth 2005, Oshana 2006, Benson
2005, and Christman and Anderson 2005.
7
See Christman 2009, chap. 7.
8
This is to borrow the requirement of “answerability” stressed by writers such as Andrea Westlund
and Paul Benson. See Benson 2005 and Westlund 2003.
72 J. Christman

That is to say, one accepts oneself without alienation in my sense when one
thinks one’s motives are worth having – that they are coherent and at least mini-
mally justified from one’s own point of view – and that they feel intelligible as one’s
own. The addict, the mentally impaired person, and the deeply conflicted agent does
not sense her values as worth having as her own in this way. The motives that spur
action for such a person are not ones that she can see as giving her reasons to con-
tinue so acting. In this way, she is alienated in my sense.
Being (self-) alienated means feeling and judging one’s motives to be in such
conflict (with each other or with other deeper self-defining values) that their self-­
justificational force is undercut (relative to an alternative condition one can imag-
ine), even if they in the end move one to act. The alienated person also will not
experience the appropriate affective and emotional concomitants to decisions, such
as pride and shame, in the right way. This sense of self-fragmentation is important
for it marks a lack of autonomy but is described in a way that does not postulate a
separated and disembodied point of view from which the person judges her identity
and condition. The assumption that such a completely detached point of view is
available and psychologically operative has been rightly criticized as untenable.9
Much more can be said about the details of this and other models of autonomy,
and we will return to some of these issues presently. For now, let us turn to the broad
requirement of interpersonal respect required of moral agents and how such respect
is complicated by acknowledging the way in which people often deliberate in close
interaction with others, making both the development and exercise of autonomy –
the object of such respect – dependent on that very interaction.

5.3 The Liberal Sensibility

It is no accident that in calling this perspective the “liberal” sensibility I utilize a

political notion, as the ethical principles I will refer to arise from the moral and
political framework of modernist liberal social philosophy. But the ideas I mention
here are neither inherently political nor reliant on those broader commitments. I
want only give a name to a cluster of what I hope are uncontroversial ethical pre-
scriptions that apply to interpersonal relations and involve acknowledgement of the
moral importance others’ autonomy.
Respect for others as autonomous beings is a fundamental part of the common
sense moral obligations structuring everyday social life. One need not be a Kantian
to accept the idea that others’ capacity to make self-governing decisions serves as
the object of value and respect in our interactions with them, even if obligations of
respect of this sort might be grounded in some broader moral framework (such as
utilitarianism or virtue).

9
See, e.g., Sandel 1982.
5 Autonomy, Respect, and Joint Deliberation 73

All I want to identify here is a set of pro tanto obligations that figure in some cen-
tral way in the ethical principles with which we deal with each other, even if such
obligations are often overridden or in competition with other quite different consider-
ations. But insofar as respect for autonomy has this central ethical role, it has a certain
structure that should be identified. Respect for autonomy grounds various obligations
of non-interference, limitations on paternalism, deference to others’ authority in mat-
ters relating to them, conferral of equal status to such persons in contexts of collective
deliberation, tolerating those we disapprove of, letting people have their say, and other
such obligations. In these ways, in interpersonal relations, one person’s desire to
advance her own interests or the interests of the persons she encounters must take
seriously into account or be strictly constrained by respect for the autonomous agency
of those others.
In close relations between co-deliberators of the sort we will discuss in a moment,
it may seem that respect for autonomy plays no role in the ethical considerations
guiding the interaction, as considerations of care, love, concern and cooperation
dominate the ethical scene. However, as I will emphasize, even in such relatively
intimate settings, it is clear how the directive to respect one’s partner as a person
who is able to make up her own mind, exercise her own unique point of view, and
so on, still operates at least as a background obligation framing the interaction.
Caring for another can go awry, morally, if one fails to treat her also as a person who
is able to competently and authentically formulate and act upon values of her own,
at least when she meets the standards of autonomy we are discussing.
I hope so much is uncontroversial. The point of articulating these obligations is
that they have as their object the actual capacity of persons to be self-governing.
That is, insofar as others are autonomous, we owe them the treatment these obliga-
tions describe, to some significant degree. If they are not, such obligations of respect
are suspended in favor of alternative moral postures, such as duties of care whose
object is advancing well-being, full stop. But those ethical considerations are sepa-
rable from the obligations to respect autonomy and the judgments a person makes
about her own well-being she makes as an autonomous being.
While commonplace, this set of ethical strictures also presupposes a certain
interpersonal dynamic that will be crucial for the problems we are discussing,
namely that the person who bears the obligation to show respect in question is not
typically in any way involved in the constitution or exercise of the autonomy of the
person so respected. This, as we will see, is the crux of the difficulty we will be
examining.

5.4 Partnerships

Recall that in the general schema for autonomous agency I rely on here, a person
acts for motives that are her own when she competently and reflectively accepts
them as part of an ongoing autobiographical narrative. Such a schema is meant to be
“content neutral” in the sense that no direct reference is made to the kinds of motives
74 J. Christman

or values that an autonomous person must have to count as such. Though it may
well be that various particular values will be incompatible with reflective self-­
acceptance because of the social-psychological effects of having those values on the
competence and reflective capacities of the agent, but not because of the status of
those values in themselves.10
Also, the schema in no way assumes that the motives and values that move an
agent were originally chosen by her or that they are under her control in a narrow
sense. So a person may be autonomous on these models even if the entirely inher-
ited her values and is not in a current position to alter them, psychologically speak-
ing (what Al Mele describes as “unsheddable” values) (Mele 1995, 153–73). As
long as she meets the conditions of reflective self-acceptance (with all other require-
ments intact), these factors will not negate her self-government.
And while the reflection in question is from the individual’s point of view, there
are many conditions where close interaction between people directly and indirectly
determines the content and strength of the motives a person finds herself with. For
example, long-term romantic partners may well see many decisions as fully collab-
orative ones, as I mentioned earlier, where each considers the other’s needs and
judgment not as a separable datum that must be weighed in balanced consideration
of her own desires but very much as part of the content of those very desires. In
many such cases, what we can call “cooperative reflection” takes place, where each
individual considers her values and options in close interaction with the other, either
in actual dialogue or in deep consideration of the other’s perspective as part of one’s
own self-appraisal. What “I” think becomes very much constituted by what
“we” think.
To give a more precise account of the interpersonal dynamics of partnerships,
and for the special challenges they raise for the principle of respect for autonomy,
let us briefly examine Michael Bratman’s account of shared agency and, in particu-
lar, shared deliberation. For Bratman, agents have shared cooperative intentions
when “each agent intends that the group perform [the] joint action in accordance
with sub-plans (of the intention to perform the joint action) that mesh…”11 By
“meshing” Bratman means that the particular elements of each person’s plan coheres
and compliments the corresponding elements in the other’s (sub-) plan. In this way,
they display interlocking and reflexive intentions. They are also mutually responsive
to the norms generated by their meshing (sub-)plans, for plans are hierarchically
organized desires that are grounded in the past and extend into the future and carry
with them directives to treat as a reason particular desires that form part of the plan.
In this way, plans are normative for the people whose plans they are. Further, these
intentional dynamics are commonly believed by the group members to be persisting
(Bratman 2014, ch. 3).

10
This is to accept, in part, what Kristinsson calls a “weakly substantive” account of autonomy. See
Kristinsson 2000.
11
Bratman 2007, 232.
5 Autonomy, Respect, and Joint Deliberation 75

In this way, what Bratman calls “modest sociality” obtains when there is mutual
responsiveness on the part of participants: their shared intentions lead to their
actions by way of commonly known mutual responsiveness (to shared norms) that
tracks the ends they each mutually intend (Bratman 2014, ch. 4). Of greater interest
to us here is how he applies this account of shared intentions to act to a view of
shared deliberation. For the kinds of partnerships we are considering involve mutu-
ally accepted plans to deliberate together and to allow such thinking together to
reflect complex interdependence. The question such phenomena gives rise to is how
the autonomy of each can be maintained with such mutual interdependence and,
more pointedly, how each can maintain an attitude of respect for the autonomy of
the other when the reflection they mutually engage in partly constitutes that
autonomy.
For Bratman, shared deliberation involves three elements: embeddedness of the
shared intentions of the group, meaning that the intentions of the two (or more)
agents are intentionally co-mingled; common ground of shared commitments,
meaning that certain considerations that matter to us matter to other group members
(our partner, say); and shared commitment to “weights”, meaning a shared commit-
ment to take certain normative considerations into account together in our delibera-
tions. This last element is perhaps the most important and we will come back to it
presently. The overall picture is that groups count as deliberating together when
group members share an intention to deliberate interactively in these ways and
mutually follow certain norms implicit in their guiding sub-plans and shared
intentions.
Mutually committing ourselves to “weights” involves allowing ourselves to be
guided by various exclusionary reasons (what counts as a default position in our
reasoning, for example, or what counts as a defeater to such a position, and so on).
However, mutually allowing ourselves to be guided together by the weight of these
norms does not involve valuing the objects of such norms, Bratman argues (2014,
136–7). This is crucial, in that it means that I can deliberate together with you and
in so doing let myself be guided by the common normative considerations that
structure our common thinking, but in doing this I don’t have to hold an independent
judgment that the things such norms point to as values are values to me outside of
our relationship.
An example might help: my partner and I deliberate together about whether to
both accept a new academic position. In doing so we implicitly allow ourselves to
be moved by certain considerations that are embedded in certain aspects of our
shared plans, such as only taking a job with a salary that matches our current one. I
am happy to use this consideration in our deliberation together, but it may well be
the case that I don’t much care about money, and I’d be happy to take a cut in salary
for the new job. But I realize that doing this unilaterally will upset a certain balance
in our mutual standing in negotiating as partners (if I took a cut, it would put pres-
sure on my partner to do so and I know this would make her uncomfortable, etc.). I
don’t value the object of our shared normative commitment but I’m happy to be
guided by it as constitutive of our agreement to think about this together.
76 J. Christman

This means that shared deliberation is compatible with interpersonal value plu-
ralism. Two people who (when thinking independently) have different values can be
guided by norms that presuppose such values as a constraint on their co-thinking.
We need this shared policy about weights to ensure the reliability, predictability, and
intelligibility of our deliberations, but we can commit to such a policy for divergent
reasons. Our shared commitment to these weights is based in part on the reasons we
have for our partnership to begin with (for Bratman our sociality). Valuing this soci-
ality may be based on its desirability, its feasibility or on some obligation we have
to do so (Bratman 2014, 144). But in any case, we value the partnership and the
shared deliberations that go with it in ways that give us reasons to shape our think-
ing in particular ways that we wouldn’t have if reasoning alone.
Bratman claims that such shared deliberation and activity can express the auton-
omy of the group so acting (2014, chap. 6). However, this view of collective agency
is something I want to put to the side for now, for my worry is not how the partner-
ship acts as a kind of autonomous agent but how the respective partners maintain
their autonomy as individuals even if engaged in such an intimate deliberative inter-
action. This is necessary so we can ask our question about respect.
But the example of long-term partners I just gave suggests relative equality of
power, at least as I am describing it (perhaps in a naively idealized tone). Such part-
nerships can be labeled SP partnerships, for “symmetrical power” partnerships.
Also of importance in this context are those partnership connections where one
person’s judgment and values has a constitutive and causal relation to a second per-
son’s deliberative perspective about an important decision but not vice versa. An aid
worker, therapist, physician, or care-giver might occupy this position relative to an
extremely vulnerable or needful client. These will be labeled asymmetrical power
(AP) partnerships.
Although the question of mutual respect for autonomy can be raised about part-
nerships of any type of power configuration, it has particular poignancy when asked
about partnerships of asymmetrical power, in particular those where the relatively
less powerful person is in a fragile and vulnerable position. The dimensions of
power operating in such situations are numerous, and may include differences in
intelligence, social status, resources, authority, and so on. All of these are relevant
to real-world cases. But for the theoretical purposes at hand, I want to focus on the
way that the judgments and actions of the one person can affect and in a way help
constitute the reflective deliberations of the subject. In such a situation, I will argue,
the “liberal sensibility” we discussed earlier will be complicated, to say the least.
Consider a vulnerable and struggling young woman, Christina, who has just been
traumatized because of being forced into sex work. Imagine she has been “rescued”
from such a condition in a foreign country.12 This coercive sex work forced Christina
to engage in activities that dramatically violated the self-defining value orientation
of her earlier life. In addition to the substandard conditions of her life as a sex
worker, this work involved trauma and violence that Christina must now work to

12
I discuss a similar example in 2014a.
5 Autonomy, Respect, and Joint Deliberation 77

recover from. A social worker is working with her to determine, for example, what
her options are regarding such things as whether to return home or remain in the
destination country (where she is highly vulnerable to either being forced or “choos-
ing” to return to prostitution). After medical, nutritional, and housing resources are
secured for Christina, a therapeutic process begins where the social worker helps
her gain a sense of confidence and psychological (and material) capacity to make
such decisions about her life.
Christina’s vulnerability here is pronounced, as is her dependence on the judgment
and actions of the social worker (call her Seba). Many people in Christina’s situation
return to the life of prostitution (and drugs) both out of fear of returning home, where
they may face shame and rejection, but also as a possible mode of survival in a con-
text of few feasible options. Such people can put themselves in a position, however,
to engage in prostitution in ways that allow them to survive and even send funds back
to their families as remittances, even if such activity continues to involve extreme risk.
Consider further that Christina was brought up in a religious setting that was
built around the virtues of sexual modesty. Her practical identity – the virtues and
obligations she regarded herself as defined by – structured the sense of her future
and memories of her past. However, after the experience of forced sex work, she can
no longer embrace that identity in its original form. The fact that her family and
home community would not accept her because of this drives home this point in
tragic detail.
In Christina’s case the aid worker must engage in the process of therapy in ways
that delicately negotiates a process of both recovering from trauma but also forming
a renegotiated practical identity that will contain value commitments relating to
self-respect, dignity, and perhaps religious obligation.13 The social worker, of
course, has tremendous influence over the course of this renegotiation of Christina’s
sense of herself. She might have to decide, for example, whether to counsel Christina
away from considering any options that might lead her back to prostitution or other
risky activities. In doing so, should she judge it best, Seba would be influencing
strongly the value orientation that Christina develops over the course of treatment.
In addition, the process of Christina’s regaining the competencies that are
required of autonomy is subtly guided by Seba’s therapeutic attentions. Christina’s
goal, we can assume, is both to settle on a life plan that is realistic and attainable for
her, given her situation, and to do so in ways that she can live with as her own, so to
regain her autonomy relative to such goals. Seba’s role, then, is to aid this process
cooperatively with Christina but in ways that make the outcome Christina’s own.
Such scenarios are sadly common. I simplify here grossly in order to focus on the
(for us) central normative issue at play, namely what strictures on paternalistic inter-
ventions are applicable here based on respect for the (perhaps nascent) autonomy of
the subject. Clearly questions surrounding Christina’s competence, rationality,
mental health, and so on are relevant here. But what is not often noted in discussions
of similar cases, is that the judgments and actions of the aid worker herself are also
highly relevant to whether the woman is autonomous.

13
For a sensitive and insightful discussion of recovery from trauma in this way, see Brison 2003.
78 J. Christman

Again, I simplify here by considering a dyadic relation between two persons,

while in fact cases of this sort will involve complex networks of agencies, social
connections, cultural ties, and so on that make the picture of an “individual” decid-
ing what to do overly simplistic. Second, I paint a picture of a victimized and vul-
nerable woman (the formerly trafficked woman) in order to raise issues of
paternalism, but in fact many such women have highly developed survival mecha-
nisms, defensive strategies, and other psychological resources at their disposal.
Finally, I focus on the paternalistic reasons the aid worker has in trying to decide a
course of action, but in most cases such a person will be guided by a complex array
of motives and reasons and will be acting within a policy matrix that has its own
guidelines and protocols, the prevalence of which may be impossible to sort out
precisely in particular cases.
But the fundamental components of the example include the fact that the pater-
nalist is both a potential agent of interference (motivated by concern for the subject)
but also is a person who expresses attitudes and provides resources (psychic and
otherwise) that are themselves crucial for the subject’s capacity to make decisions
for herself.
This example is obviously one of asymmetric power. In relations of (more or
less) symmetric power, similar dynamics can be in play. As we noted, people in
close relations engage in deliberations with a shared intention to allow mutually
responsive, norm-guided, reasoning to lead to a decision, without either partner
intending a partnership-independent outcome. Insofar as such decisions are central
to the life plans of the partners then these sorts of collaborations exercise the auton-
omy of each and so exhibit the characteristics that give rise to a challenge to the
liberal sensibility, namely how respect for autonomy operates when the interaction
itself is constitutive of the exercise of that autonomy.

5.5 The Challenge for the Liberal Sensibility

The crux of the liberal sensibility I outlined earlier is that the obligation to respect
the autonomy of others places limits on the ways that we can intervene into their
lives and character development, and that these limits include an attitude of tolera-
tion toward the values they (autonomously) embrace. The challenge that asymmetri-
cal partnerships such as that of Seba and Christina pose for this moral sensibility is
that while Seba owes Christina respect for the values she might autonomously
embrace, a respect that consequently limits the ways she can interact with her, Seba
is in the position of shaping the very values and character to which that respect
applies.
This challenge, it can be noted, parallels discussions in the realm of liberal edu-
cation, where a primary directive of facilitating the development of autonomy in
children is weighed against (and complicated by) the requirement that education be
broadly neutral relative to diverse conceptions of the good life. Harry Brighouse, for
example, has argued that education should aim to advance the capacity for
5 Autonomy, Respect, and Joint Deliberation 79

autonomy without constraining the person in the development of a wide variety of

ways of life (Brighouse 2000). But he claims education should aim to facilitate but
not promote autonomy, since autonomy involves the actual practice of critical self-­
reflection and such habits may well not be part of a particular person’s value struc-
ture or orientation. Facilitating autonomy merely helps a person develop the
capacities for critical self-reflection but does not induce the actual exercise of those
capacities. Indeed, his view has been criticized by those who claim that such capaci-
ties cannot be fostered without strongly encouraging actually using those abilities;
just as one cannot facilitate the disposition for kindness or open-mindedness with-
out inducing people to actually engage in those practices (see, e.g. Corngold 2012).
That worry aside, there is a structural similarity to the tension we are discussing,
in that the role of an educator is, arguably, to aid in the development of autonomy
related competences and self-defining values, but she or he must also abide by an
obligation of tolerance (if not neutrality) toward a wide array of value structures that
people might go on to develop by way of that very education. An important contrast,
however, is that the vulnerable subjects in partnerships we are imagining are adults
(albeit perhaps young adults) with an already established practical identity, frac-
tured though it might be by the trauma and circumstance. Brighouse’s directive to
facilitate but not promote autonomy may not be effective in cases where aid is
required for agents to reassemble a sense of themselves and to establish the capacity
to be self-governing in light of this sense.
Recall that the spirit of the liberal sensibility I describe implies that those who
are positioned to be instrumental in the development or exercise of the autonomy of
another (with whom they are in partnership in my sense) must be guided in that
enablement of the others’ capacities without expressed judgment of the particular
values that might be pursued by that subject. The process of aid whereby the subject
regains the capacity for self-government cannot be shaped by an independent
appraisal of the quality of the values such self-governing subjects go on to pursue.14
This tension arises particularly because the schema for autonomy we are utiliz-
ing here attempts to maintain “content neutrality”, where the conditions of self-­
government do not depend on the quality of the values agents are moved by, except
insofar as the holding of such values themselves serve to undercut autonomy-related
capacities. So appraisal of the inherent quality of the values of the agent themselves
is not required to determine her autonomy. But this tension also arises because of
the way autonomy has been seen as relational, where interpersonal and social
dynamics of particular sorts are claimed to be central to the individual’s status as
self-governing, as I discussed earlier. As a partner with you, I directly affect the
structure and content of the motives that you reflectively accept as your own, and

14
In other work I looked at this problem specifically with an eye toward the issue of paternalism,
asking what limits to paternalism are operative in cases of these sorts of “partnerships”. Here I am
asking a slightly more general question, namely what is the nature of the obligation to respect oth-
ers’ autonomy, encapsulated in the liberal sensibility, given the ways that one’s actions will likely
shape the very autonomy which is its object. See Christman 2014a. I also elaborate on partnerships
of this sort in Christman 2014b, 231–25.
80 J. Christman

our relationship may instantiate a dynamic upon which your ability to be autono-
mous itself depends. Recognition of your status as a self-governing agent able to
answer for yourself may in many cases be fundamental in your ability to compe-
tently (and confidently) accept her values as her own.
Returning to Bratman’s terminology, such shared activity must involve mutual
responsiveness, a commitment to the joint activity, and a commitment to mutual
support of that activity.15 The activity I have in mind here is the process of (compe-
tent and authentic) deliberation itself. Even in asymmetrical partnerships of the sort
I envision, both subjects have the shared intention of helping (at least) one of them
develop capabilities and attitudes that allow self-government to be re-established,
and hence for the values that she comes to embrace to reflect that self-government.
However, an aid worker such as Seba may well have strong views, for example,
on not only the inherent value of practices such as sex work and drug use, but she
also must make complex judgments about the relation between being open to going
back to such practices and the risk that doing so will undercut Christina’s ability to
maintain basic competencies of agency that the therapeutic process is designed to
re-establish.
Indeed, it has become somewhat of the standard orthodoxy in some therapeutic
settings to see the interaction between therapist and client as a dynamic and mutu-
ally participatory one. As Stephen Mitchell puts it, the “kind of participation
required of the analyst is a complex blend of listening; silently responding; giving
oneself over to the explicit and subtle interactional gambits offered by the patient;
observing the impact on the patient of one’s own ideas and emotional commitments;
and giving oneself over to a range of states of mind that allow a broad array of one’s
own feelings and imaginings” (Mitchell 1997, xii). In this way the therapist is
engaged in cooperative deliberation with the patient, one that involves asymmetrical
power but which is mutually responsive in Bratman’s sense.
The question we must answer, then, is what is the nature and scope of interper-
sonal respect, called for by the liberal sensibility, in cases of both symmetrical and
asymmetrical power when the interpersonal dynamics in question directly affect the
establishment or exercise of autonomy?
To answer that question, we should recall that the view of autonomy I described
earlier sees that characteristic as not only social but also diachronic, or as described
elsewhere “historical.” The diachronic view of autonomy sets the conditions of self-­
government in a way that is sensitive to the ongoing narrative structure of our practi-
cal identities. Moreover, autonomy seen in this manner attempts to capture the ways
that the authenticity of our current commitments relies importantly on the nature of
the processes of value formation that produced them. In short, autonomy amounts
to competence (including self-control), and reflectively accepting (without alien-
ation) one’s motives in light of their history as part of one’s diachronic practical
identity.

15
Other accounts that come close to capturing partnerships in my sense include Gilbert 1989 (spe-
cifically her account of “plural subjects”).
5 Autonomy, Respect, and Joint Deliberation 81

Now even given the complications introduced by the effect that interpersonal
power dynamics can have on autonomy, it is easy to see ways in which respect for
autonomy can be obviously violated. If Seba* (a parallel aid worker), for example,
is motivated by a religious conviction that sees prostitution as inherently sinful, so
that in addition to helping Christina* regain basic psychological competences and
social skills necessary for recovery, she makes it her mission to convince her of that
sinfulness, as well as bring her along to the (say) Christian world view that moti-
vates that idea. Such a strategy may well involve imposing a conception of value
and the good life that Christina* no longer would share if she were given the chance
to competently consider her life options without such pressure.
But if we judge this to be a violation of the respect owed to Christina*, this can-
not be based on duties rooted in respect for her established autonomous value
scheme, for such autonomy is not functioning and its development is dependent on
Seba*‘s actions themselves. The duty to avoid manipulating vulnerable people to
adopt your own world view, then, must be based on respect for what we could call
the “nascent autonomy” of the agent. But more to the point, the diachronic, or his-
torical, understanding of autonomy we are using here helps explain this expanded,
autonomy-based, duty of respect. The religious devotee that Christina* would
become if Seba*‘s ministrations were successful would have to accept that position
by way of a practical identity that develops over time and embrace that set of (reli-
gious) motivations in light of how they came about. So only if Christina* would be
able to (competently) say that she accepts being a religious person who decries the
sinfulness of prostitution, knowing it came about in large part because of the psy-
chological pressures of Seba*‘s therapeutic work, would she count as autonomous
in the end.
Of course she might so accept such a position: many converts express gratitude
toward the influential leaders who brought them to their faith. But only if such
reflective acceptance takes place in the full light of the awareness of its origins
(among other conditions) is it authentic and so autonomy conferring. In light of
such cases, some have criticized the view of autonomy at work here as too permis-
sive, since it seems to allow that a fully manipulated person such as Christina* will
count as autonomous as long as the manipulation extends to the acts of historical
appraisal itself: any self-respecting manipulator will make sure the person has come
to rationalize the process of manipulation itself as part of a grand plan of which she
is a part (or some such scenario) (see, e.g., Oshana 2006). But the requirement of
eventual self-acceptance in light of one’s history I defend is also linked with various
other conditions of competence (and competent self-reflection). Such conditions
include the ability to consider alternatives, entertain reasons, and avoid overt self-­
deception. It is unlikely that a person like Christina* will be able to do that without
reverting to the self-undercutting rationale of “I did it because that’s what I was told”.
To elaborate this last point, recall that the model of autonomy I am sketching
here demands that the person’s eventual reflective self-acceptance must enable her
to give reasons that emanate from her diachronic practical identity and which reflex-
ively affirms the values inherent in that identity. That reflective self-­acceptance must
be in light of the processes (in these cases recovery processes) that allowed her to
82 J. Christman

renegotiate her practical identity. But of course her reflective understanding of her-
self cannot merely reiterate the psychological results of those processes, so she must
be in a position to support her values and commitments independent of them. As I
put the point elsewhere, the reflection in question must be sustainable over a variety
of conditions and not merely operate because of local and ephemeral causal factors
(Christman 2009, 152–3).
So those who are engaged in these sorts of partnership relations are self-­governing
only if their reflective self-acceptance includes an awareness of those very relations.
This is especially true of asymmetrical partnerships. This means that a recovering
person like Christina must be able to give, in part at least, process-independent rea-
sons in support of the values she comes to hold (and be guided by). To maintain
self-government here requires being cognizant of the way that one’s development is
the result of cooperative reflection and a process of self-constitution with the help of
another. The independence from that other – making the resulting self-­understanding
one’s own – requires that one can give reasons for those values independent of the
mere fact that one emerged from those therapeutic processes themselves. Only then
will the person’s reflective self-acceptance meet the adequacy condition mentioned
just now, namely that such acceptance is sustainable over a variety of conditions.
Being unable to give reasons for one’s values independent of the process of forming
them (“independent” in the sense of not depending on that process for their justifica-
tory force) would mean that such reflective judgment could not be sustained over a
variety of conditions.
This rules out as heteronomous certain sorts of people who emerge from such
dependent relations, such as the pure conformist convert who can say nothing in
favor of the values she excepts that they accord with what her guru instructed; and
at the other end of the spectrum, the recalcitrant rebel who rejects all values as soon
as she realizes that some other person influenced her in adopting them.16 In both
cases, the person fails to meet the requirement of reflective self-acceptance here
because they cannot give process-independent reasons for holding the views
they hold.
Now of course many who go through processes of recovery from trauma may
develop emotional dependencies on care-givers and therapists that prevent them
from imagining paths other than those that they perceive those therapists would
approve of. Perhaps that is the best way to understand what is happening with
Christina*. But that, I contend, marks the lack of self-government precisely because
she is unable to reflectively answer for herself in ways that both accounts for the
process of recovery and is also in some ways independent of it.
Therefore, insofar as the aim of aid workers like Seba should be guided by the
(revised) liberal sensibility, then their aim should be that clients and patients come
to achieve autonomy in ways that do not manifest the imposition of values that they
themselves instigate. That is, they should have as their aim their clients’ eventual
reflective self-acceptance in light of the process of recovery, reflection whereby the

16
I am grateful to Ryan Cooke for helping me clarify this point.
5 Autonomy, Respect, and Joint Deliberation 83

clients can give process-independent reasons for their values. In this way, the respect
for autonomy underwritten by the liberal sensibility can operate in contexts where
interpersonal interaction of this sort has such deep and pervasive effects on the very
autonomy being respected.
This conclusion applies to asymmetrical power relationships, but what about
other partnerships involving collaborative deliberation that are more or less sym-
metrical? The claim I made above applies in general form here too, in that the obli-
gation to respect the autonomy of another (with whom one is co-deliberating)
amounts to a commitment to a process the end of which would result in a non-­
alienated narrative self-understanding without reference to the values being pursued
by that self. That is, my interaction with my partner involves respectful relations in
this sense when I am guided by the norm that says I should make sure our interac-
tion allows her to maintain a sense of herself over time that is undisturbed by an
awareness of how our interaction help constitute that sense of self. Diachronic self-­
acceptance in light of one’s history, including a history of interactions that partly
constituted the resulting value perspective, is the goal that the obligation to respect
the other underwrites.
So those asymmetrical relations where another’s autonomy is dependent on
one’s own actions and attitudes, one has a responsibility to respect the other as a
diachronically developing agent, with a past as well as a future, and whose future
judgments will be effectively the court of appeal for the respectfulness of one’s cur-
rent interactions. My hope is that such guidelines, even stated in these broad, rough-­
grained ways, will help reshape duties of respect in all-too-common cases of
dependence and vulnerability.
Now it may be thought that a person like Seba giving aid in such situations has
just one dominant obligation, namely to advance the well-being of her client, per-
haps even including helping her develop autonomy but on the basis of an obligation
to care, not out of respect for autonomy itself. But this approach fails to capture the
obligation to remain respectfully neutral concerning the imposition of particular
substantive value schemes expressed in the liberal sensibility. Viewing duties of
care-givers as merely the obligation to advance well-being (full stop) would be con-
sistent with such agents’ inculcating into their clients a value scheme entirely of
their own determination, assuming that it is one that will advance well-being.
More importantly, insofar as the duty to respect autonomy must be refashioned
in this way due to the interactive nature of the development and exercise autonomy
in such partnerships, the autonomy that is the object of such respect must be under-
stood in the diachronic manner I have laid out. Only then can we understand why
partners and care-givers have duties to take seriously the autonomous generation of
values and choices of the other even when their activity dynamically constitutes that
autonomy. If we viewed autonomy as a time-slice characteristic, or as a trait imper-
vious to the processes of its development, then the obligation to respect others as
autonomous would gain no foothold in the partnership interactions we have envi-
sioned, and Seba* would have done no wrong to Christina*. But surely she does.
What I have argued here, then, is that in partnerships of the sort I have described,
the duties associated with the liberal sensibility – to respect autonomy of
84 J. Christman

others – must be complicated by a provision taking account of the ways that inter-
personal and social interaction are themselves constituents of autonomy. Hence, our
obligations to respect the autonomy of others when we have power over their devel-
opment of that very autonomy must take on a diachronic mode, namely that we
must aim to facilitate the development of a practical identity which can be the basis
of future judgments that would not be disturbed when considered in light of that
very process of interaction and facilitation.
Autonomy is a cooperative enterprise in many ways, and only if we engage in
this enterprise in caring and respectful ways, sensitive to the complex manner in
which social interactions shape the very persons we are, will self-government
emerge as the morally noteworthy capacity it is thought to be.17

References

Anderson, Joel, and Axel Honneth. 2005. Autonomy, Vulnerability, Recognition, and Justice.
In Autonomy and the Challenges to Liberalism: New Essays, ed. John Christman and Joel
Anderson, 127–149. Cambridge: Cambridge University Press.
Benson, Paul. 2005. Taking Ownership: Authority and Voice in Autonomous Agency. In Autonomy
and the Challenges to Liberalism: New Essays, ed. John Christman and Joel Anderson,
101–126. Cambridge: Cambridge University Press.
Bratman, Michael. 2007. Structures of Agency. Oxford: Oxford University Press.
———. 2014. Shared Agency: A Planning Theory of Acting Together. Oxford: Oxford
University Press.
Brighouse, Harry. 2000. School Choice and Social Justice. Oxford: Oxford University Press.
Brison, Susan. 2003. Aftermath: Violence and the Remaking of Ourselves. Princeton, NJ: Princeton
University Press.
Christman, John. 2009. The Politics of Persons: Individual Autonomy and Socio-historical Selves.
Cambridge: Cambridge University Press.
———. 2014a. Relational Autonomy and the Social Dynamics of Paternalism. Ethical Theory and
Moral Practice 17: 3690–3682. (special issue edited by Annete Dufner and Michael Kuhler).
———. 2014b. Freedom and the Extended Self. Ethical Perspectives 21 (2): 225–254.
Christman, John, and Joel Anderson, eds. 2005. Autonomy and the Challenges to Liberalism: New
Essays. Cambridge: Cambridge University Press.
Corngold, Josh. 2012. Autonomy-facilitation or Autonomy-promotion? The case of Sex Education.
Theory and Research in Education 10 (1(March)): 157–170.
Dworkin, Gerald. 1989. The Theory and Practice of Autonomy. Cambridge: Cambridge
University Press.
Frankfurt, Harry. 1987. Freedom of the Will and the Concept of a Person. In The Importance of
What We Care About. Cambridge: Cambridge University Press.
———. 1992. The Faintest Passion. Proceedings and Addresses of the Aristotelian Society 49:
113–145.
———. 1999. Necessity, Volition, and Love. Cambridge: Cambridge University Press.
Gilbert, Margaret. 1989. On Social Facts. Princeton: Princeton University Press.

17
This paper is a revised version of a paper given at the conference: Thick Concepts of Autonomy.
Center for Advanced Study in Bioethics. University of Muenster, Germany. October 18–20, 2012.
I am grateful to Michael Quante, Thomas Gutmann, Johann Ach, and Bettina Schoene-Seifert for
organizing the conference and to all participants for very helpful discussion and commentary.
5 Autonomy, Respect, and Joint Deliberation 85

Kristinsson, S. 2000. The Limits of Neutrality: Toward a Weakly Substantive Account of Autonomy.
Canadian Journal of Philosophy 30 (2) (June): 257–286.
Korsgaard, Christine 1996. The Sources of Normativity. Cambridge: Cambridge University Press.
Lear, Jonathan. 2006. Radical Hope: Ethics in the Face of Cultural Devastation. Cambridge, MA:
Harvard University Press.
Mackenzie, Catriona. 2008. Relational Autonomy, Normative Authority and Perfectionism.
Journal of Social Philosophy 39: 512–533.
Mele, Al. 1995. Autonomous Agents: From Self-Control to Autonomy. New York: Oxford
University Press.
Mitchell, Stephen. 1997. Influence and Autonomy in Psychoanalysis. London: Routledge.
Oshana, Marina. 2006. Personal Autonomy in Society. Hampshire: Ashgate Publishing.
Sandel, Michael. 1982. Liberalism and the Limits of Justice. Cambridge: Cambridge
University Press.
Westlund, Andrea. 2003. Selflessness and Responsibility for Self: Is Deference Compatible with
Autonomy? The Philosophical Review 112 (4(October)): 483–523.
Chapter 6
Autonomy and Beliefs

Alfred R. Mele

Abstract In Autonomous Agents, I argued that among the obstacles to autonomous

action are facts of certain kinds about an agent’s beliefs. For example, someone who
is deceived into investing her savings in a way that results in her losing the entire
investment to the person who deceived her may correctly be said to make that invest-
ment nonautonomously. But not everyone has agreed. In this article, I return to doxas-
tic aspects of individual autonomy and argue more fully for the thesis that facts of a
certain kind about an agent’s beliefs can render some of an agent’s actions
nonautonomous.
King George just now issued a decree raising property taxes by five percent through-
out his kingdom. What would suffice for his doing that autonomously? And what
would suffice for his doing that nonautonomously? In Autonomous Agents (Mele
1995, pp. 180–82), I suggested that certain facts about George’s beliefs might ren-
der the specified action nonautonomous. My view about this has been challenged
(McKenna 2005, Weimer 2009), and I return to doxastic aspects of individual
autonomy in this article.
Before I get down to business, I should point out that philosophers use the term
“autonomy” in a variety of senses, even when only autonomous individuals (rather
than nations, for example) are at issue.1 The root idea of autonomy – from autos (self)
and nomos (rule or law) – is self-rule or self-government. My attempt to understand
autonomous action in Mele 1995 is primarily an attempt to understand the nature of
actions distinctive of self-ruled or self-governed individuals. “Autonomy,” as I use

1
Joel Feinberg usefully distinguishes among four “meanings” of “autonomy” as applied to indi-
viduals (1986, p. 28). Also see Arpaly 2003, pp. 118–2.

A. R. Mele (*)
Department of Philosophy, Florida State University, Tallahassee, FL, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 87

Switzerland AG 2022
J. F. Childress, M. Quante (eds.), Thick (Concepts of) Autonomy, Philosophical
Studies Series 146, https://doi.org/10.1007/978-3-030-80991-1_6
88 A. R. Mele

the term there (as applied to individuals) is in the family of metaphysical freedom
terms: “free will,” “free choice,” “free action,” and the like. I use it the same way here.

6.1  utonomous and Nonautonomous Actions: Setting

A
Things Up

Some background information about King George is in order (see Mele 1995,
p. 180). Through no fault of his own, the only access George has to his kingdom is
through his staff of advisors. His primary concern in life is to do what is best for his
kingdom as a whole. The staff provide him with information, and he takes legisla-
tive and other measures on that basis. They also provide George with feedback
about his measures, which information he takes into account in drafting further
legislation and in modifying previous measures. They give him monthly figures on
a variety of matters, including the gross national product, the distribution of wealth,
the percentage of the population living below the poverty level, the King’s popular-
ity, and the popular reaction to his legislative acts. However the staff have their own
agenda – namely, their own enrichment at the expense of the populace. Knowing the
King’s preferences, they systematically provide him with false information that will
lead him to make decisions that will further their aims. Of course, they are careful
not to be so obvious about this as to call their loyalty into question.
Today, George’s advisors presented him with information they knew would lead
him to conclude that his issuing a decree raising property taxes by five percent
would put an end to poverty in his kingdom while not greatly inconveniencing any-
one. George rationally drew that conclusion, and he issued such a decree. The advi-
sors kept the new tax revenues for themselves, as they planned all along to do; and
they knew that even if the proceeds had gone to the poor, poverty would have con-
tinued to be a serious problem in the kingdom. Most landowners are in debt to wage
earners – especially to the King’s advisors. A raise in income taxes (rather than
property taxes) would have helped significantly with poverty in the kingdom (pro-
vided the money actually went to the poor). In their subsequent monthly reports to
George, his advisors provide very encouraging news about shrinking poverty
figures.
In my view, issuing the tax decree is not something George autonomously does.
By controlling what George believes about the effect a five percent raise in property
taxes will have, his staff gain control over whether he will judge it best to issue the
decree and whether he will issue it (see Mele 1995, pp. 180–82). In light of the
control his advisors exert here, I see George as heteronomous in this connection.
Am I wrong about this? Michael McKenna spins a story that is intended to shed
light on George’s situation. In it, Tal tries to save Daphne’s life by giving her “the
prescribed dose” from “a bottle marked ‘The Good Stuff’” (2005, p. 209). Tal has
every reason to believe that the medication will help Daphne and no reason to
6 Autonomy and Beliefs 89

believe that it will harm her. Unfortunately, the bottle is accidentally mislabeled. It
contains a drug that is lethal for a person with Daphne’s illness: Tal kills Daphne by
giving her the drug. McKenna reports that this story “helps to confirm the need for
[an] epistemic condition on morally responsible agency” (p. 209). He also asks
whether “Tal acted nonautonomously in giving Daphne the drug” (p. 210). The
answer he seems to endorse is no: he acted autonomously “in giving Daphne the
drug” (pp. 210–11).
Consider the following true sentences:
Tal gave Daphne the prescribed dose from a bottle marked “The Good Stuff”
Tal gave Daphne a lethal (for her) drug
Tal killed Daphne
Insert “autonomously” immediately after “Tal” in each sentence. Which (if any) of
the resulting sentences are true? Some readers may experience some uncertainty
about how to answer. A brief review of a basic issue in the philosophy of action
may help.
Donald Davidson writes: “I flip the switch, turn on the light, and illuminate the
room. Unbeknownst to me I also alert a prowler to the fact that I am home” (1980,
p. 4). How many actions has the agent, Don, performed? Davidson’s coarse-grained
answer is one action “of which four descriptions have been given” (p. 4). A fine-­
grained alternative treats A and B as different actions if, in performing them, the
agent exemplifies different act-properties (Goldman 1970). On this view, Don per-
forms at least four actions, since the act-properties at issue are distinct. For example,
the property of flipping a switch is distinct from the property of turning on a light,
and the property of turning on a light (in a room) is distinct from the property of
illuminating a room. One may flip a switch without turning on a light and vice versa.
Similarly, one may turn on a light in a room without illuminating the room (the light
may be painted black) and illuminate a room without turning on a light (by setting
a dark room on fire). Another alternative – a componential one – represents Don’s
illuminating the room as an action having various components, including (but not
limited to) his moving his arm (an action), his flipping the switch (an action), and
the light’s going on (Ginet 1990; Thalberg 1977; Thomson 1977). Where propo-
nents of the coarse-grained and fine-grained theories find, respectively, a single
action under different descriptions and a collection of intimately related actions,
advocates of the various componential views locate a “larger” action having
“smaller” actions among its parts.
When actions are individuated in Davidson’s coarse-grained way, it is actions
under descriptions that are performed intentionally or otherwise – not actions
period. Similarly, it is actions under descriptions that are performed for a reason R
(or for a reason at all). For example, under the description “flips the switch,” what
Don does is intentional; but under the description “alerts the prowler,” what he does
is not intentional. Similarly, under the description “flips the switch,” Don might
have acted for a reason having to do with getting sufficient light to find his keys; but
under the description “alerts the prowler,” Don does not act for a reason at all
(Davidson 1980, p. 5).
90 A. R. Mele

Fine-grained and componential theorists make no special appeal to action-­

descriptions. Theorists of both kinds can straightforwardly say, for example, that
Don intentionally flipped the switch, unintentionally alerted the prowler, and flipped
the switch in order to have enough light to find his keys and that, although he alerted
the prowler, he did not do that for a reason.
Might “autonomously” work in a way that resembles the way in which “inten-
tionally” and “for a reason” work? Might it be the case, for example, that although
Don autonomously flips the switch, autonomously turns on the light, autonomously
illuminates the room, and alerts the prowler (by illuminating the room), he does not
autonomously alert the prowler? Framed in Davidsonian terms, might it be the case
that what Don does is autonomously done under the descriptions “flips the switch,”
“turns on the light,” and “illuminates the room” but not under the description “alerts
the prowler”?
An easy way to avoid saying a lot of things twice – both in terms of Davidson’s
coarse-grained theory of action individuation and in non-Davidsonian terms – is to
invite readers to understand the action variable “A” as a variable for actions them-
selves (construed componentially or otherwise) or actions under descriptions,
depending on their preferred theory of action individuation. The same goes for the
expressions that take the place of “A” in concrete examples. For example, propo-
nents of Davidson’s theory should read “Don illuminates the room” as “something
Don does under the description ‘illuminates the room’” and non-Davidsonians
should make no adjustments. I proceed on the assumption that readers accept my
invitation. Readers who have no preferred theory of action individuation should feel
free to adopt one temporarily.
McKenna asks (2005, p. 210): “Should we reconsider the case of Tal and, consis-
tent with Mele’s intuitions about … King George, conclude that Tal acted nonauto-
nomously in giving Daphne the drug?” His answer is no. But it is difficult to be sure
exactly what McKenna has in mind by the expression “in giving Daphne the drug.”
Suppose we change “in” to “when” in the clause just quoted. We might then say that
Tal did some things autonomously and some other things nonautonomously at that
time. In the former group, some might confidently place his giving Daphne the pre-
scribed dose from a bottle marked “The Good Stuff.” In the latter, I would confi-
dently place his giving Daphne a lethal (for her) drug, and his killing Daphne.
It might have occurred to some readers that if the three placements in the preced-
ing paragraph are correct, what is doing the work is the difference between inten-
tional and nonintentional action. It is true that Tal intentionally gives Daphne the
prescribed dose from a bottle marked “The Good Stuff” and that he neither inten-
tionally gives her a lethal (for her) drug nor intentionally kills her. Even so, the
claim about what is doing the work is open to criticism.
Consider the following case:
Unbeknownst to Bob, who intends to vote Democratic, the Democratic lever in his voting
booth is attached to a randomizing device. The only way for him to vote Democratic is to
pull that lever, but pulling the lever yields only a 0.001 chance of producing a Democratic
vote. Luckily, Bob succeeds in registering a vote for the Democratic candidate. (Mele 1995,
p. 14, n. 11)
6 Autonomy and Beliefs 91

As I see it, Bob’s producing that vote was too lucky to count as an intentional action,
even though that is what he intended to do (see Mele and Moser 1994). Even so, if
Bob autonomously decided to vote for the Democrat and autonomously pulled the
lever, I see no compelling reason to deny that he autonomously (or freely) produced
a Democratic vote.
That is why I would not infer merely from its being true that Tal did not inten-
tionally kill Daphne that he did not kill her autonomously. But, of course, there is an
important difference between the stories about Bob and Tal in this connection. Bob
intended – and was trying – to register a vote for the Democrat, and Tal neither
intended nor tried to kill Daphne. As far as I can see, the judgment that Tal did not
autonomously kill Daphne is extremely plausible, as is the judgment that he did not
autonomously give her a lethal (for her) drug.
Return to George. The following sentences are true of him:
George raised property taxes by five percent
George enriched his advisors
Insert “autonomously” immediately after “George” in both sentences. Which (if
any) of the resulting sentences are true?
The judgment that George did not autonomously enrich his advisors (by raising
taxes) is at least as plausible as the judgment that Tal did not autonomously kill
Daphne. George neither intended nor tried to enrich his advisors (by raising taxes).
He unintentionally enriched them. And, of course, he was tricked into enrich-
ing them.
Did George autonomously raise taxes by five percent? Earlier in this section I
denied that he did on grounds having to do with the control his advisors exerted over
him. McKenna contends that if “we are interested in action-theoretic questions
about whether certain sorts of persons have the capacity to guide their own conduct
in light of rules that they themselves endorse, it is hard to see how the sorts of epis-
temic pollutants presented to King George could foul up [his] having acted in such
a manner” (2005, p. 228, n. 13). Let us stipulate that George has the capacity
McKenna mentions and that he often guides his behavior in a variety of spheres in
light of rules he endorses. It is hard to see how this fact about George undermines
the claim that he did not autonomously raise taxes. Presumably, this fact about
George is no obstacle at all to the claim that he nonautonomously enriched his advi-
sors. This is so even if, at the pertinent time, George was exercising a “capacity to
guide [his] own conduct in light of rules” he endorsed – perhaps, in this instance, a
rule favoring trying to help the poor in his kingdom. So, why, exactly, should the
fact at issue be thought to undermine the claim that George nonautonomously raised
property taxes? Does the answer feature the fact that George intentionally raised
property taxes? The action-theoretic questions that interest me at present are about
the status of particular actions: for example, the question whether George autono-
mously raised taxes.
In a passage to which the note from which I just quoted is appended, McKenna
identifies agents’ “ruling themselves in acting as they do” with their acting “in light
of rules that they themselves accept as their own” (2005, pp. 210–11). Now, at a
92 A. R. Mele

time at which one is acting in light of a rule that one endorses (that is, accepts as
one’s own), one may be doing some things intentionally and other things uninten-
tionally. For example, Don endorses a rule in favor of always turning on a light
when he enters his kitchen. Just now, he acted in light of that rule. He flipped a
switch in his kitchen and turned on a light. Unfortunately, an enemy who knew
Don’s habits wired the switch to a bomb he placed in Don’s oven. When Don flipped
the switch, he detonated the bomb and blew up his house. Although he flipped the
switch and turned on the light intentionally, he detonated the bomb and blew up his
house unintentionally. And, again, it may be that just as one can do some things
intentionally while doing other things unintentionally, one can do some things
autonomously while doing other things nonautonomously.
McKenna supposes, to keep things simple, that when Tal gave Daphne the lethal
drug, he was “consciously acting” from the following “principle that he endorsed”:
“Always attempt to help those who are suffering innocently” (2005, p. 209). At that
time, then, in McKenna’s view, Tal was acting in light of a rule that he accepts as his
own and ruling himself in acting as he did. Even if this were to suffice for Tal’s
autonomously doing something or other at the time – perhaps, attempting to save
Daphne’s life – that would leave it open that Tal nonautonomously does various
other things at the same time. My concern, again, is with what is and is not done
autonomously at a time – not simply with whether an agent is doing something or
other autonomously at a time.
McKenna contends that “if there is an epistemic constraint on autonomous
agency, it differs from the sort of epistemic constraint pertinent to morally respon-
sible agency” (2005, p. 211). “It is reasonable to assume,” he writes, “that morally
responsible agency requires two sorts of conditions: an epistemic and a control con-
dition” (p. 207). He understands the latter condition “in terms of the freedom of the
will” (p. 208), and he tells us that “freely willed action” does not “involve a signifi-
cant epistemic component” (p. 211).2 McKenna writes: “Perhaps careful investiga-
tion will reveal that self-ruling agency just is, or is a component of, freely willing
agency. If so, it would be correct to conclude that autonomous action does not
involve a significant epistemic component because freely willed action does not”
(p. 211).
In a note to the sentence I quoted about the assumption that “morally responsible
agency requires an epistemic and a control condition,” McKenna (2005, p. 228, n.
7) cites “Aristotle’s conditions on the voluntary from Nicomachean Ethics …
1109b30–1111b5” and a related discussion in Fischer and Ravizza 1998 (pp. 12–14).
Attention to a pair of ironies about these citations will prove instructive.
Aristotle (1915) writes: “Since that which is done under compulsion or by reason
of ignorance is involuntary, the voluntary would seem to be that of which the mov-
ing principle is in the agent himself, he being aware of the particular circumstances
of the action” (Nicomachean Ethics 1111a22–24). Now, in ordinary English, what

2
McKenna backs away from this claim about free (or “freely willed”) action in his recent book
(2012, p. 13).
6 Autonomy and Beliefs 93

is said to be done voluntarily seems to be more tightly connected to what is said to

be done freely or autonomously than to what people are said to be morally respon-
sible for doing. In an ordinary drunk driver scenario in which the driver is too drunk
even to see the pedestrian he runs down, it is generally claimed that the driver is
morally responsible for running down the pedestrian even though he did not volun-
tarily run the unfortunate person down. But the suggestion that a person does some-
thing autonomously (or freely) only if he does it voluntarily has a very plausible
ring. Accordingly, Aristotle’s view about the voluntary seems congruent with the
claim that there are significant epistemic requirements for autonomous actions and
free actions.
The second irony concerns the discussion by John Fischer and Mark Ravizza that
McKenna cites and related work by Fischer (alone or with a coauthor). Fischer and
Ravizza report that “the focus of … attention” in their book is “the freedom-relevant
condition on moral responsibility, rather than the epistemic condition” (1998, p. 13;
also see Fischer 1994, p. 238, n. 4). The epistemic condition “corresponds to the
excuse of ignorance. It captures the intuition that an agent is responsible only if he
both knows the particular facts surrounding his action, and acts with the proper sorts
of beliefs and intentions” (p. 13). “The freedom-relevant condition specifies that the
agent must control his behavior in a suitable sense, in order to be morally respon-
sible for it” (p. 13).
As I have explained elsewhere (Mele 2010), Fischer and his coauthors have not
specified epistemic requirements for being morally responsible for A-ing that they
clearly regard as not also being requirements for freely A-ing. Furthermore, their
“freedom-relevant condition” – or “control condition” (Fischer and Ravizza 1998,
p. 13) – has some notable epistemic features, including the agent’s having at least
one mechanism that can recognize moral reasons and the agent’s not being substan-
tially deluded about the nature of reality (see Mele 2010). Even if, in McKenna’s
words, “morally responsible agency requires an epistemic and a control condition”
(2005, p. 207), work by Fischer and coauthors thus far seems to leave it open that
any necessary epistemic conditions for an agent’s being morally responsible for A-­
ing are also necessary conditions for the agent’s freely – or autonomously – A-ing.3
This is the second irony to which I referred.
A related observation is in order. Compare a kleptomaniac, Karen, who has,
through no fault of her own, an irresistible desire to steal a necklace with a man,
Dan, who is nonculpably ignorant of a certain important fact – that a party he has
been invited to is a surprise party (see Fischer and Tognazzini 2009, p. 531–32 and
Mele 2010, pp. 109–10). The former steals the necklace and the latter spoils the
surprise by talking about the upcoming party to the person who was meant to be
surprised. Karen is not blameworthy for stealing the necklace, and Dan is not blame-
worthy for ruining the surprise. The excuses are of two very different kinds. Karen’s
excuse is that she was forced to steal by an irresistible desire and therefore did not

3
I am not denying that there are necessary epistemic conditions for being a morally responsible
agent that are not necessary conditions for being a free agent. For discussion, see Mele 2010,
pp. 108–9.
94 A. R. Mele

freely steal; and what excuses Dan for spoiling the party was nonculpable ignorance
of a crucial fact. Dan’s excuse is epistemic and Karen’s is not. But it should not be
inferred from this that Dan freely ruined the surprise – that his ruining the surprise
was a free action.4 Just as Karen’s being compelled to steal makes it true that she did
not freely steal, Dan’s ignorance might make it true that he did not freely ruin the
surprise. The fact that there are legitimate “control”-based excuses and very differ-
ent “ignorance”-based excuses does not itself straightforwardly entail that there are
epistemic requirements for being morally responsible for A-ing that are not also
requirements for freely – or autonomously – A-ing.

6.2 Sorting Through Cases

Return to McKenna’s claim that Tal acted autonomously “in giving Daphne the
drug” (2005, pp. 210–11). Suppose the bottle that was accidentally marked “The
Good Stuff” contained labeled pills. Tal found three pills in the bottle – labeled
GS8, GS9, and GS10. He decided to give Daphne what he assumed was the next pill
in the sequence of pills she had been taking – GS8. One can ask whether Tal acted
autonomously “in giving Daphne” GS8, as McKenna might. Or one can ask simply
whether he autonomously gave her GS8. I choose the latter question. Whether a yes
answer to this question is true or false, it at least does not sound terribly implausible.
It does not sound as implausible as, for example, a yes answer to the question
whether Tal autonomously killed Daphne.
Consider a modified version of Tal’s story. Daphne’s mother, Molly, is in the
room when Daphne has a seizure. Molly is confined to a wheelchair and moves too
slowly to help Daphne. Tal asks Molly what he should do. She tells him to get pill
GS8 from the bottle on Daphne’s dresser and give it to her. Tal has every reason to
believe that Molly wants to save Daphne; but, in fact, she wants her dead. Molly
knows that GS8 is lethal for anyone in Daphne’s condition. Tal gives Daphne GS8,
thereby killing her.
Did Tal autonomously give Daphne GS8 in this modified story? Was his giving
her that pill an autonomous action? I say no. To be heteronomous regarding some-
thing is to be ruled or dominated by another agent (or other agents) regarding it.
Molly is in control of how Tal seeks to solve Daphne’s urgent problem; and I see
him, accordingly, as heteronomously giving Daphne GS8. He has become Molly’s

4
Readers will, of course, distinguish Dan’s saying what he does about the party from his ruining
the surprise. For example, they will notice that even if it is assumed that he intentionally says what
he does about the party, he does not intentionally ruin the surprise. Incidentally, in the paragraph
that precedes Aristotle’s assertion that “the voluntary would seem to be that of which the moving
principle is in the agent himself, he being aware of the particular circumstances of the action”
(Nicomachean Ethics 1111a22–24), he mentions an agent who “did not know” that what he
reported “was a secret” (1111a9–10). I take it that he would say that, owing to his ignorance of this
circumstance, this agent did not voluntarily reveal a secret. If it is true that he did not voluntarily
reveal a secret, then if we freely do only what we voluntarily do, he did not freely reveal a secret.
6 Autonomy and Beliefs 95

tool for killing Daphne. If Tal heteronomously gives Daphne GS8, he does not do it
autonomously (which is not to say, of course, that he does not do anything autono-
mously at the time).
There is no obvious contradiction in claiming that Tal autonomously gives
Daphne GS8 in McKenna’s story but not in mine. But is there a troubling tension?
In my discussion of a radical kind of manipulation in Mele 1995 – a kind in which
manipulators intentionally erase a person’s core values and replace them with radi-
cally different values (with a certain property that need not concern the reader
now) – I say that the resulting nonautonomy (regarding various actions) would also
result if the same transformation in a person’s values were produced by blind forces:
for example, a bizarre electromagnetic storm (pp. 168–69). Steven Weimer con-
tends that I should take a similar approach on the informational front (2009,
pp. 29–30). In his view, I should avoid endorsing a view according to which the
difference between whether someone did something autonomously or nonautono-
mously sometimes depends on whether important false beliefs of his are or are not
products of intentional deception.5 So, for example, if I say that in McKenna’s story
about Tal, where there is no intentional deception, Tal autonomously gives Daphne
GS8, I should say the same in my story.
Suppose I were to say that I am agnostic about whether Tal autonomously gives
Daphne GS8 in McKenna’s story. Would that put enormous pressure on me to say –
symmetrically – that I am agnostic about whether Tal autonomously gives Daphne
GS8 in my story?
An attempt to answer this question may benefit from attention to some “bad
information” scenarios that have a feature in common with the manipulation stories
of mine to which Weimer refers and with counterpart stories featuring blind forces.
Recall that in the pertinent stories, the agents’ core values are erased and replaced
by very different values. One might do well to consider stories in which crucial
beliefs are erased and replaced by very different beliefs.
Here is a third version of Tal’s story. In this version, Tal had learned that when
Daphne has a seizure, he needs to give her a pill is the S series. What he learned is
true; but, as part of her plan for bringing about Daphne’s death, Molly hired some-
one to knock out Tal’s true belief about this and replace it with the belief that the
correct pill for seizures is any pill in the GS series.6 She then induced a seizure in
Daphne. Tal gave Daphne GS8, thereby killing her. In this case, in my opinion, Tal
is heteronomous with respect to his giving Daphne GS8 and therefore gives it to her
nonautonomously. Here, as in my earlier story about him, Tal has become Molly’s
tool for killing Daphne.
In a fourth version of Tal’s story, a bizarre electromagnetic storm does the belief-­
replacing work. Molly induces a seizure in Daphne, believing that Tal will not be
there to try to save her; but Tal shows up in time to act on his new belief. Tal is not

5
Suzy Killmister takes the approach Weimer recommends on the informational front (2013,
p. 517 n. 4).
6
On some complications about belief replacement, see Mele 2006, pp. 177–78.
96 A. R. Mele

heteronomous with respect to his giving Daphne GS8: no other agent has taken
control of what he does. That leaves it open that giving Daphne GS8 is not some-
thing he autonomously does. But it also leaves it open that he autonomously gives
her GS8.
Suppose for the sake of argument that Tal autonomously gives Daphne GS8 in
the version of the story in which the bottle of pills is accidentally mislabeled and the
version featuring the strange storm. Is the truth of that supposition compatible with
it also being true that Tal nonautonomously gives her GS8 in the other two versions
of the story (the version in which Molly lies and the version in which Tal’s true
belief about the needed medication is intentionally knocked out and replaced by a
false belief)? At first sight, at least, the answer to this question about compatibility
would seem to be yes. There is a salient difference between the two pairs of stories:
in one pair Tal, has been intentionally manipulated (by means of intentional decep-
tion or intentional belief replacement); and in the other pair, this is not the case. That
difference may help to explain how Tal can autonomously give Daphne GS8 in two
of the stories but not in the other two.
This brings us back to the idea that just as I make symmetrical judgments about
certain value-manipulation stories and comparable value-change stories featuring
“blind forces,” I should make symmetrical judgments about the four stories about
Tal. Is this correct? To pursue this question, I need to say something about how I
understand values and tell a story about value manipulation.
As I understand valuing, “S at least thinly values X at a time if and only if at that
time S both has a positive motivational attitude toward X and believes X to be good”
(Mele 1995, p. 116). When values are understood as psychological states, I take
them to have both of these dimensions by definition. This account of thinly valuing
and the corresponding thin account of values are not intended as contributions to the
theories of valuing and values; their purpose is to make my meaning clear.
The following story about value manipulation will do. I call it “One Good Day”
(see Mele 2009, p. 471; for a similar story, see Mele 1995, pp. 164–65).
Chuck enjoys killing people, and he “is wholeheartedly behind” his murderous
desires, which are “well integrated into his general psychic condition” (Frankfurt
2002, p. 27). When he kills, he does so “because he wants to do it” (Frankfurt 2002,
p. 27), and “he identifies himself with the springs of his action” (Frankfurt 1988,
p. 54). When he was much younger, Chuck enjoyed torturing animals, but he was
not wholeheartedly behind this. These activities sometimes caused him to feel
guilty, he experienced bouts of squeamishness, and he occasionally considered
abandoning animal torture. However, Chuck valued being the sort of person who
does as he pleases and who unambivalently rejects conventional morality as a sys-
tem designed for and by weaklings. He freely set out to ensure that he would be
wholeheartedly behind his torturing of animals and related activities, including his
merciless bullying of vulnerable people, and he was morally responsible for so
doing. One strand of his strategy was to perform cruel actions with increased fre-
quency in order to harden himself against feelings of guilt and squeamishness and
eventually to extinguish the source of those feelings. Chuck strove to ensure that his
psyche left no room for mercy. His strategy worked. In hardening his heart as he did,
6 Autonomy and Beliefs 97

Chuck also ensured that he had no values at all that could motivate a charitable
deed. (He might buy some Girl Scout cookies to lure an innocent child away for evil
purposes; but a cookie-buying motivated in that way is not a charitable deed.)
A high-powered team of scientists captured Chuck and ran an experiment on
him. They had identified the system of values of an extremely sweet and generous
woman, Beth; and they wanted to learn whether they could erase Chuck’s nasty
values and replace them with values like Beth’s. The plan was to do this while
Chuck slept, turn him loose for a day, and then restore his old values at midnight.
Things worked perfectly. Shortly after he awoke, Chuck started working with a
local Habitat for Humanity crew in his neighborhood. He worked long and hard,
feeling very good about what he was doing. When the work day ended, he drove
around town for an hour and bought several boxes of Girl Scout cookies from every
Girl Scout he saw – about fifty boxes in all. Then he delivered the cookies to a local
homeless shelter. His motives were pure, as Beth’s are when she does her charitable
deeds. Naturally, Chuck wondered about the remarkable change in himself. His
explanation was that he had finally seen the light about how to live. Occasionally, he
reflected on his new values and wholeheartedly endorsed them. The scientists pre-
dicted the endorsement. After all, any process of critical reflection is conducted
from some perspective or other; and Chuck’s wonderful new values dominate the
perspective from which reflects on his values in such a way that continued identifi-
cation with them is ensured.
Does Chuck autonomously work with Habitat for Humanity, help Girl Scouts,
and feed the homeless on his one good day? The intuitive answer is no, or so it
seems to me. Presumably, if it is true that Chuck autonomously performs the chari-
table actions at issue, then he deserves some moral credit for those actions, as Beth
does for hers. Does he deserve credit for them? Knowing what you do about the
source of those actions, would you give him moral credit for performing them? I
doubt it – unless, perhaps, you endorse a theory that commits you to doing so. And
why would you not give him such credit? Because he does not deserve it, it seems.
Consider now a variant of One Good Day in which a strange electromagnetic
storm (or a weird cluster of brain tumors) has the effect on Chuck’s values that the
scientists achieved in the original story. Chuck does everything he does in the first
story – including reflectively endorsing his new values – and he feels the same way
about his activities. At the end of the day, all of the changes in his values are undone.
Chuck is back to normal – for him, that is.
Here too, it seems, Chuck deserves no moral credit for his good deeds. They are
the product of values that swept into him over night and that he had no role at all in
producing or shaping.7 Moreover, the values that guide his behavior on his one good
day are utterly contrary to values that he did play a part in shaping – the core values
of his that were erased by the strange storm. Now, presumably, if it were true that
Chuck performs his charitable actions autonomously, he would deserve some moral

7
I am not claiming that one has to be responsible for attitudes if one is to be responsible for behav-
ior in which the attitudes issue. On this topic, see Mele 1995, pp. 221–230.
98 A. R. Mele

credit for them (as Beth does for hers). From his deserving no such credit for them,
we can infer that he does not autonomously perform them.
In my first story about Chuck, the claim that he is heteronomous with respect to
his good behavior is plausible. There, the scientists seize control of (some of) what
he does by instilling wonderful Beth-like values in him. In my second story about
Chuck, no other agents seize control. The drastic, temporary change in his value
system is a product of blind forces. He does not heteronomously perform his chari-
table deeds; but he does not autonomously perform them either, if I am right.
If it is correct to judge that Chuck’s good deeds are nonautonomously performed
in both stories about him but incorrect to make symmetrical judgments about Tal’s
giving Daphne GS8 in the four stories about him, why might that be? In the absence
of a reasonable answer, those who have symmetrical intuitions about Chuck but
asymmetrical intuitions about Tal should worry that some of these intuitions are
misguided.
Where there are differences in responses to stories (even if the difference is
between agnostic and non-agnostic responses), we should look for differences in the
stories that might reasonably account for them. In the pair of stories about Chuck,
his entire moral personality, as it were, has been radically reversed – either inten-
tionally or accidentally. In the quartet of stories about Tal, the featured psychologi-
cal matter is very localized, and it is in a different sphere – the sphere of beliefs
about relatively simple matters of fact. These differences between the Chuck stories
and the Tal stories are obvious ones, but do these differences have a significant bear-
ing on autonomous action?
Suppose that value reversal of the kind and magnitude present in Chuck’s stories
precludes his autonomously performing his good deeds. It certainly does not follow
from the truth of this supposition that, for example, the changes in belief in my last
two stories about Tal preclude his autonomously giving Daphne GS8. The two kinds
of change – that is, radical change in moral personality and change in a belief about
pills – are so different that the inference at issue is, at best, far-fetched. Now, it is a
truism that if one is heteronomous with respect to one’s A-ing, one does not autono-
mously A. A reader who sees Tal as heteronomous with respect to his giving Daphne
GS8 in the story in which his relevant true belief is intentionally knocked out and
replaced with a belief that the correct pill for Daphne’s seizures is any pill in the GS
series and in the story in which Molly deceives him into believing that GS8 will
save her will deny that he autonomously gives Daphne GS8 in these stories. But
when that reader turns to the other two stories about Tal, in which he is not heter-
onomous, this particular route to the claim that Tal nonautonomously gives Daphne
GS8 is closed, and the reader is faced with a difficult question, a question related to
one raised by a remark by Aristotle that I quoted earlier: “the voluntary would seem
to be that of which the moving principle is in the agent himself, he being aware of
the particular circumstances of the action” (Nicomachean Ethics 1111a22–24).
Circumstances of what kind? Aristotle says a bit about this, and one of his illus-
trations is especially fitting in the present context: “one might give a man a draught
to save him, and really kill him” (1111a 13–14). But is he claiming merely that
agent does not voluntarily kill the man or also that he does not voluntarily give the
6 Autonomy and Beliefs 99

man the drink he gives him – does not voluntarily give the man D, for short? And,
whatever Aristotle might have thought about this, what should we think? Again, it is
plausible that one A-s autonomously only if one A-s voluntarily. So a reader who is
unsure whether to accept or reject the claim that this agent voluntarily gives the man
D may also be uncertain about whether he does this autonomously. And the very
same reader may be confident that in a version of Aristotle’s story in which, in order
to bring about the man’s death, someone deceived the agent into believing that the
lethal liquid would save the man, the agent does not autonomously give the man D.
I plead ignorance about whether, in the non-heteronomy stories, the agent just
discussed voluntarily gives the man D and Tal voluntarily gives Daphne GS8. But
my being ignorant about this does not entail that I am mistaken in holding that, in
the versions of these stories that feature deception or intentional belief replacement,
the agent nonautonomously gives the man D and Tal nonautonomously gives
Daphne GS8.
What brought me to this point was an investigation prompted by Weimer’s con-
tention that just as I take a symmetrical position regarding autonomous action in
value-change stories like my two stories about Chuck (one of which features inten-
tionally induced value change whereas the other features a matching change in val-
ues induced by blind forces), I should do the same in the informational sphere. I
have tried to motivate a rejection of this contention. My own grip on a notion of
autonomous action leaves me agnostic about whether Tal autonomously gives
Daphne GS8 in the non-heteronomy stories even though I claim that he nonautono-
mously does this in the other stories. This is an asymmetrical position, and I have
argued that it is compatible with my symmetrical position on the Chuck stories.
I started with King George. What does this have to do with him? In my view (see
Mele 1995, pp. 180–82), he does not autonomously issue the decree raising prop-
erty taxes by five percent. McKenna, partly by means of his discussion of his story
about Tal, argues that this view of mine is mistaken (2005). Here, I have explained
why McKenna’s argument is unpersuasive even if it is granted that Tal autono-
mously gives Daphne GS8 in a story like his (in which the bottle of pills is acciden-
tally mislabeled). I hope that I also have shed some light on the nature of autonomous
action and on the complexity of the subject matter.
In this article, I have had something to say about intentional action, voluntary
action, free action, and autonomous action. In each of these spheres, one properly
wonders how far one can move – along various dimensions – from seemingly para-
digmatic instances of the kind of action at issue and still find instances of that kind
of action. And in each sphere, one encounters difficult cases. Sometimes, it is diffi-
cult to make a confident judgment about whether a featured action is or is not inten-
tional, voluntary, free, or autonomous – even when we know all the other relevant
details about it. That is part of what keeps philosophers of action with a penchant for
conceptual analysis in business. But we should be confident that when another
agent’s influence on your conduct benefits from his having surreptitiously seized
something approaching complete control over whether you A or not – as in the case
100 A. R. Mele

of King George raising property taxes by five percent – you are heteronomous in
that connection and therefore do not autonomously A.8

References

Aristotle. 1915. Nicomachean Ethics. Vol. 9 of William Ross, ed. The Works of Aristotle. London:
Oxford University Press.
Arpaly, N. 2003. Unprincipled Virtue. New York: Oxford University Press.
Davidson, D. 1980. Essays on Actions and Events. Oxford: Clarendon Press.
Feinberg, J. 1986. Harm to Self. New York: Oxford University Press.
Fischer, J. 1994. The Metaphysics of Free Will. Cambridge, MA: Blackwell.
Fischer, J., and M. Ravizza. 1998. Responsibility and Control: A Theory of Moral Responsibility.
New York: Cambridge University Press.
Fischer, J., and N. Tognazzini. 2009. The Truth About Tracing. Noûs 43: 531–556.
Frankfurt, H. 1988. The Importance of What We Care About. Cambridge: Cambridge
University Press.
———. 2002. Reply to John Martin Fischer. In Contours of Agency, ed. S. Buss and L. Overton,
27–31. Cambridge, MA: MIT Press.
Ginet, C. 1990. On Action. Cambridge: Cambridge University Press.
Goldman, A. 1970. A Theory of Human Action. Englewood Cliffs: Prentice-Hall.
Killmister, S. 2013. Autonomy and False Beliefs. Philosophical Studies 164: 513–531.
McKenna, M. 2005. The Relationship Between Autonomous and Morally Responsible Agency. In
Personal Autonomy, ed. J. Taylor, 205–234. Cambridge: Cambridge University Press.
———. 2012. Conversation and Responsibility. Oxford: Oxford University Press.
Mele, A. 1995. Autonomous Agents: From Self-Control to Autonomy. Oxford: Oxford
University Press.
———. 2006. Free Will and Luck. New York: Oxford University Press.
———. 2009. Moral Responsibility and History Revisited. Ethical Theory and Moral Practice
12: 463–475.
———. 2010. Moral Responsibility for Actions: Epistemic and Freedom Conditions. Philosophical
Explorations 13: 101–111.
Mele, A., and P. Moser. 1994. Intentional Action. Noûs 28: 39–68.
Thalberg, I. 1977. Perception, Emotion, and Action. New Haven: Yale University Press.
Thomson, J. 1977. Acts and OtherEvents. Ithaca: Cornell University Press.
Weimer, S. 2009. Beyond History: The Ongoing Aspects of Autonomy. Journal of Ethics and
Social Philosophy 4: 1–31.

8
I wrote this paper for a 2012 conference at the Centre for Advanced Study in Bioethics, University
of Münster. I made some small changes while hiding out during the COVID-19 pandemic.
Chapter 7
How Much Understanding Is Needed
for Autonomy?

James Stacey Taylor

Abstract How much understanding should be required of a person with respect to

her actions and their implications for her to be autonomous with respect to her deci-
sions to perform them? I defend a thin approach to the question of how much under-
standing of her acts a person should possess for her possibly to be autonomous with
respect to her decisions to perform them: That a person could be autonomous with
respect to her decision to perform a certain action if she understood both the nature
of the act and its expected implications under the intensional description under
which she intentionally performed it. This paper is divided into three sections. First,
I argue that a concern for the moral value of autonomy does not undergird the ethi-
cal requirement to secure a person’s informed consent to her medical treatment.
Second, I respond to several objections that have been leveled against this argument
that focus on its implicit premise that only a thin understanding of one’s actions is
required for one to be autonomous with respect to one’s decisions to perform them.
I conclude by indicating why this minimalist account of autonomy that I defend is
preferable to its more robust competitors.

In his autobiographical work Surely You’re Joking Mr. Feynman the Nobel-prize
winning physicist Richard Feynman tells a story about an experience that he had as
a consultant to an engineering project. Faced with a set of plans, he was unwilling
to admit that he didn’t know what some of the symbols for the parts were, and so he
hit on an ingenious way to discover their meaning without admitting his ignorance.
Pointing to one of the mysterious symbols, he asked “What would happen if that
valve were left open?” expecting to be corrected as to the symbol’s meaning. The
engineers looked carefully at the plans, became visibly alarmed, and consulted
among themselves, before thanking him profusely (and with great admiration!) for
so quickly identifying a major flaw in the design of their system.

J. S. Taylor (*)
The College of New Jersey, Ewing, NJ, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 101

Switzerland AG 2022
J. F. Childress, M. Quante (eds.), Thick (Concepts of) Autonomy, Philosophical
Studies Series 146, https://doi.org/10.1007/978-3-030-80991-1_7
102 J. S. Taylor

Clearly, Feynman did not autonomously identify a flaw in the design of the valve
system, for he had no understanding of what he was doing. He did not, in any mean-
ingful sense, “direct himself” to decide to identify the flaw, as he might have done
had he both understood the plans he was examining and the engineering principles
involved. (Although he did “direct himself” to decide to point at a particular sym-
bol, not knowing he was identifying a flaw.) To be autonomous with respect to a
decision to perform a certain action, then, the person must understand what he is
doing.1 As such, it is not surprising that contemporary theories of autonomy have
built into them the requirement that a person understand what she is doing for her to
be autonomous with respect to her decisions (or her actions, or her effective first-­
order desires).2 Thomas Beauchamp and James Childress, for example, make this
requirement explicit when they hold that to be autonomous a person must act “with
understanding” (Beauchamp and Childress 1994, 123). It is also implicit in Gerald
Dworkin’s (early) requirement that for a person to be autonomous with respect to a
motivating desire she be procedurally independent with respect to it, and hence not
subject to “manipulation, deception, the withholding of relevant information, and so
on” (Dworkin 1976, 25).
Yet although there is widespread agreement that autonomy requires understand-
ing there has been little discussion of how much a person should be required to
understand her action and its implications for her to be autonomous with respect to
her decision to perform it.3 This oversight is surprising. It would seem that if an
account of autonomy requires that a person understand her actions and their impli-
cations for her to be autonomous with respect to her decision to perform them then
it should include an account of how much understanding is required for the sake of
theoretical completeness.4 Moreover, providing an account of how much a person
should understand her actions and their implications for her to be autonomous with
respect to her decisions to perform them would seem to be critical for persons who
wish to draw on discussions of the nature and value of autonomy to address issues
in applied ethics where this concept plays a central role. In bioethics, for example,
it is held that persons should give their informed consent to their treatment, where

1
This focus in this essay is on “local” models of autonomy, which provide an account of what it is
for a person to be autonomous with respect to some particular mental state or act, rather than on
“global” accounts, which focus on what it is to have a capacity for autonomy.
2
Here, the focus will be on what it is for a person to be autonomous with respect to her decisions—
although the arguments for the thin approach to the degree of understanding that is required can be
reprised with suitable modifications to establish the degree of understanding required for a person
to be autonomous with respect to her actions or her effective first-order desires.
3
In this paper I am addressing the question of how much understanding a person must have of her
actions and their implications for her to be autonomous with respect to decisions to perform them,
and so when I write of a person’s understanding of her actions this should be understood as encom-
passing their implications, also, since a person’s understanding of these is required for her to
understand the actions that she performs (or considers performing).
4
Beauchamp and Childress do go some way in this direction, noting that the degree of understand-
ing that is required will be context-dependent—although this is still merely the first step towards
providing such an account (Beauchamp and Childress 1994, 123).
7 How Much Understanding Is Needed for Autonomy? 103

this requirement is intended to protect patient autonomy by ensuring that they ade-
quately understand the options available to them—a vague requirement that imme-
diately raises the question of the standard of adequacy required. A person’s
autonomous consent is similarly held to be both necessary and authoritative in busi-
ness ethics—although there, there is little discussion of the need to secure the
informed consent of (for example) the contracting parties.
How much understanding, then, should be required of a person with respect to
her actions and their implications for her to be autonomous with respect to her deci-
sions to perform them? One who adopted a thin approach to this question would
hold that a person could be autonomous with respect to her decision to perform a
certain action if she understood both the nature of the act and its expected implica-
tions under the intensional description under which she intentionally performed it.
By contrast, one who adopted a thicker approach to this question would hold that a
person could only be autonomous with respect to her decision to perform a certain
action if she understood both the nature of the act and its expected implications
under that intensional description of the act that would be held by someone who
adequately understood these (where the adequacy conditions are to be specified).
For example, a proponent of a thicker approach to this question might hold that a
person could only be autonomous with respect to those of her actions that she per-
formed under an intensional description of them that correctly identified the likeli-
hood that they would achieve her ends in performing them.
In this paper I will defend a thin approach to the question of how much under-
standing of her acts a person should possess for her possibly to be autonomous with
respect to her decisions to perform them. To this end it will be divided into three
sections. First, I will outline an argument for the conclusion that a concern for the
moral value of autonomy does not undergird the ethical requirement to secure a
person’s informed consent to her medical treatment. Second, I will then outline and
respond to several objections that have been leveled against this argument that focus
on its implicit premise that only a thin understanding of one’s actions is required for
one to be autonomous with respect to one’s decisions to perform them. (The focus
of this second section will be on defending the theoretical point that one need only
have a thin understanding of one’s actions for one possibly to be autonomous with
respect to one’s decisions to perform them, rather than on defending the conclusion
of the argument concerning the foundations of informed consent in which this plays
a role.) Of course, showing that the view that only a thin understanding of one’s
actions is necessary for one to be autonomous with respect to one’s decisions to
perform them can be defended does not thereby show that this should also be
accepted. And, indeed, this view of the degree of understanding that is required for
autonomy could be criticized on the grounds that it fails to capture the best intuitive
understanding of what it is to be autonomous: That one is autonomous when one’s
decisions lead to one’s actions instantiating one’s will in the world. To respond to
this challenge, I will offer some indication in the concluding section of this paper as
to why the more minimalist account of autonomy that I defend is preferable to its
more robust competitors.
104 J. S. Taylor

7.1 Autonomy and Informed Consent

It is widely held that respect for the value of autonomy undergirds the doctrine of
informed consent (see Beauchamp and Childress 1994, 142–146). To be autono-
mous a person must direct her own life, make her own decisions, and perform her
own actions; to the extent that her life, decisions, and actions are under the control
of another she thereby suffers from a diminution on her autonomy with respect to
them.5 Given this, for a person to be autonomous with respect to her medical deci-
sions it should be she, and not her healthcare provider, who decides what course of
treatment she is to pursue.6 To the extent that a person’s healthcare provider attempts
to control her medical decisions through deliberately failing to disclose information
about them to her, or through misrepresenting the risks and benefits of each to her,
or through otherwise selectively presenting her with information in an attempt to
guide her decisions, then, if he is successful in such manipulations, he will compro-
mise her autonomy with respect to them. Hence, to ensure that healthcare profes-
sionals do not compromise the autonomy of their patients they should provide
unbiased information about their treatment options to them. In requiring that they
do this, then, the doctrine of informed consent appears to be grounded in the prin-
ciple of respect for autonomy.
As will be outlined below, however, it is possible for a person to fail to give her
informed consent to the course of treatment that she pursues, and yet still be fully
autonomous with respect to her decision to pursue it, and the actions that she takes
to do this. Since in many such cases it would still be appropriate to hold the health-
care professional who failed to secure his patient’s informed consent to her course
of treatment blameworthy for this failure, even though it did not adversely affect the
patient’s autonomy, it appears that the doctrine of informed consent rests on some-
thing other than a concern for the moral value of autonomy.7
How, then, could a person fail to give her informed consent to her treatment and
yet still retain full autonomy with respect to her treatment decisions and consequent
actions? The key to answering this question lies in recognizing that a person will be
autonomous with respect to her decisions if it is she, and not someone else, who
controls what it is that she decides to do. To see this, consider the case of Iago and
Othello from Shakespeare’s Othello. Through a series of manipulations and insinu-
ations Iago convinces Othello that his wife is having an affair with his, Iago’s, rival
Cassio, which leads to Othello smothering her from jealousy. Here, the font of the
decision for Othello to kill Desdemona was Iago, not Othello, and it is through

5
This characterization of what it is to be autonomous is compatible with a wide variety of views of
how much understanding of one’s actions is necessary for one to be autonomous with respect to
one’s decisions to perform them.
6
This claim is compatible with a person choosing to abdicate her decision to her healthcare pro-
vider, perhaps believing that he will make a better decision for her than she will.
7
Many, but not necessarily all, cases. It might be that the healthcare provider inadvertently failed
to secure his patients’ informed consent to her treatment for a reason other than negligence that
could be excused.
7 How Much Understanding Is Needed for Autonomy? 105

Iago’s manipulation of Othello that he killed his wife. Since Othello was manipu-
lated into making the decision to kill Desdemona by Iago, and hence it was not
Othello himself, but Iago, who directed him do this, Othello’s autonomy with
respect to this decision was compromised.
The lesson to be learned from Othello—and that which makes it relevant to both
discussions of the relationship between autonomy and informed consent, and dis-
cussions of the degree of understanding that is needed for a person to be autono-
mous with respect to her decisions—is that a person will suffer from compromised
autonomy with respect to her decisions to act if they are controlled by another.
Control is an intentionally characterized concept: A will only control the decisions
of B if she intends so to do. To see this, consider a (non-existent) variant on
Othello—PseudoOthello. In this version of the play PseudoIago’s communications
with PseudoOthello are precisely the same as they are in Othello, but with one cru-
cial difference: In PseudoOthello PseudoIago engages in no manipulation at all, but
simply descriptively provides PseudoOthello with the facts as he sees them, devoid
of guile. Since in PseudoOthello it was PseudoOthello, and not PseudoIago, who
was the font of PseudoOthello’s (identical) decisions, Othello’s autonomy with
respect to them was uncompromised. Thus, given that the only differences between
Othello and PseudoOthello are that in the former, but not the latter, Iago intended to
usurp the direction of Othello’s actions, and that in the former, but not the latter,
Othello’s autonomy was compromised as a result of his decisions being under the
control of Iago, it is clear that a necessary condition (but not sufficient) for one per-
son to control the decisions of another is that she intend so to do.
Let us now return to the question of the relationship between autonomy and
informed consent—and hence to the question of how much understanding is needed
for a person to be autonomous with respect to her decisions. Since a necessary con-
dition for one person to control the decisions made by another is that she intend to
do so a healthcare professional who deliberately withheld information from her
patients to ensure that they made the decisions she wanted them to make (and was
successful in her efforts so to do) would indeed be usurping their autonomy with
respect to their decisions. However, a healthcare professional who negligently failed
to provide her patients with sufficient information for them to make informed deci-
sions concerning their treatments would not be intending to exercise control over
the decisions that they made. As such, then, although her ill-informed patients
would not have given their informed consent to their treatment decisions they would
not have suffered from any diminution in their autonomy with respect to them.
However, despite this, we would still blame the negligent healthcare professional
for failing to secure her patients’ informed consent to their treatment. And, since
this is so, it appears that it is not concern for autonomy, but for something else, that
undergirds the doctrine of informed consent, with autonomy being valued primarily
for its use in enabling those who exercise it to secure this “something else” (see
Taylor 2004).
106 J. S. Taylor

7.2  efending the Thin View of How Much Understanding is

D
Needed for Autonomy

This account of why it is not the value of autonomy, but the value of “something
else” that a person’s exercise of autonomy might secure her, is of obvious interest to
medical ethicists concerned with the ethical foundations of informed consent. But it
is also of wider theoretical interest, for two reasons. First, this account of why it is
not the value of autonomy that provides the ethical foundations for the requirement
that a healthcare professional secure the informed consent of her patients to their
treatment is based on a thin view of how much understanding of her actions a person
must possess for her to be autonomous with respect to her decisions to perform
them. A full defense of this revisionary account of the relationship between auton-
omy and informed consent would thus require a defense of the view that a person
need only have a thin understanding of the acts that she performs for her to be fully
autonomous with respect to them. Second, since the above argument held that a
person’s autonomy with respect to her decisions will be compromised if it is not she,
but another, who controls what decisions she makes, and since it also held that con-
trol is an intentionally characterized concept, if the above argument is correct it
entails that the question of whether a person is autonomous with respect to her acts
cannot be decided only by reference to her own mental states. Instead, this question
could only be decided by reference to both her mental states, and those of others, to
ensure that her decision-making process is not under another’s control. To put this
colloquially, the above argument entails that whether or not I am autonomous with
respect to my decisions to perform certain actions will depend in part on the mental
states of others.
While the radically externalist implications of the above argument are striking, it
is the thin understanding of autonomy that this argument is based on that has drawn
the most criticism, with critics charging that this thin approach to understanding the
degree of understanding that is required of a person to be autonomous with respect
to her decisions is too impoverished. The aim of this section of the paper is thus to
defend the thin view of the degree of understanding that a person must have for her
to be autonomous with respect to her decisions, and consequent actions, that under-
girded the above analysis of the relationship between autonomy and informed con-
sent. In so doing the above view of the relationship between autonomy and informed
consent will also be defended—although defending this is not the primary focus of
this paper.

7.2.1 The Objection from the Best Intensional Description

The first objection to the thin account of how much understanding of her actions a
person needs to be autonomous with respect to them that the above argument is
based on begins with the claim that, occasionally, owing to the ignorance of the
7 How Much Understanding Is Needed for Autonomy? 107

acting agent the intensional description under which she performs an action is not
that which would best describe it from the point of view of a neutral third party. For
example, while the description of Oedipus’ act of marriage under which he per-
formed it was “marrying Jocasta,” an informed third party would hold that the most
appropriate intensional description of this act would be “marrying his mother”.
Since owing to his lack of understanding of what he was doing that arose from his
failing to perform his act under the intensional description that would be judged
most appropriate for it by an informed third party, Oedipus suffered from a diminu-
tion in his autonomy with respect to it. He did not “direct himself” to perform the
act that he actually performed, where the relevant intensional description of “the act
that he actually performed” that should be used to assess this claim is that which
would be deemed the most appropriate one by an informed third party. With this
claim in hand, the proponent of this objection then notes that both Othello and
PseudoOthello performed their deadly acts under the description “kill my unfaithful
wife”. However, since neither of their wives were actually unfaithful this would not
be the most appropriate intensional description of their acts (instead, their acts
would be more appropriately described by informed third party observers as “kill-
ing a faithful wife”), and so both of them would suffer from equally diminished
autonomy with respect to their acts of smothering their wives.8 Thus, the proponent
of this objection concludes, it is clear that a thicker approach to the type of under-
standing that is required for a person to be autonomous with respect to her actions
than that which underlies the above argument is required. And, moreover, since both
Othello and PseudoOthello suffered from diminished autonomy with respect to
their actions, and these diminutions are explicable in terms of their lack of under-
standing, rather than in terms of being subject to the control of another, the above
argument concerning the relationship between autonomy and informed consent
does not entail that one person’s autonomy will crucially depend on the mental
states of another.9
This objection is based on the view that an informed third party would judge that
the acts of both Othello and PseudoOthello would be most appropriately described
as “killing my unfaithful wife”. But this act description is overly rich, for it contains
both a description of the action and (implicitly) the reason why the act was done. In
the context of this argument, however, it is only the (descriptive) description of the
act that is being performed that is relevant, with the (normative) reason for the per-
formance of this action being separable from this. What is of interest is whether
Othello and PseudoOthello were autonomous with respect to the acts that they per-
formed, not the separate issue of the reason why they performed those acts. Hence,
only the description of the act is relevant. With this noted, the above objection can
be revisited. It is not clear that the act-description “killing my unfaithful wife”

8
Note that the claim that they suffered from diminished autonomy with respect to these acts is
compatible with both the claim that they were nonautonomous with respect to them, and the
(weaker) claim that while they were autonomous with respect to them, this was to a reduced degree.
9
A version of this objection appeared in the original version of this paper; this (more robust) ver-
sion was pressed on me by John Christman.
108 J. S. Taylor

should be understood as being elliptical for the act-description “killing my wife

because she is unfaithful”. If this is so, then the most appropriate description of the
actual act that is being performed will be “killing my wife” (with the adjective being
indicative of the reason why this as done). Both Othello and PseudoOthello are thus
performing the acts that they believe that they are performing; they are directing
themselves to perform the acts (killing their wives) that they believe that they are
performing. As such, then, when their understanding of the acts that they are per-
forming is alone considered, they are fully autonomous with respect to them. It is
only when we realize that Othello was subject to the control of Iago that we revise
our assessment of the degree to which he was autonomous with respect to the
smothering of Desdemona.
To support this response to the above objection consider a third case—
OutsourceOthello. In this example, Othello hires an assassin to smother Desdemona,
which the assassin duly does. The assassin has been hired to “kill an unfaithful
wife”, just as both Othello and PseudoOthello aimed “to kill my unfaithful wife”.
Like Othello and PseudoOthello, then, this assassin performs an action without
understanding that he was not killing an unfaithful wife. However, despite this it
would be implausible to hold that the assassin’s autonomy with respect to his killing
of Desdemona was impaired in any way because of this. And the reason for this is
because one can distinguish between “killing a wife” and “killing an unfaithful
wife”, with the latter description of the act being both a description of the act itself
(“killing a wife”) and an account of why this was done (“because she was unfaith-
ful”). Thus, the mere fact that the assassin did not succeed in killing an unfaithful
wife does not undermine the fact that (like Othello and PseudoOthello) he achieved
what he set out to do—namely, kill Desdemona. A proponent of the above objection
might respond at this point by claiming that the assassin was fully autonomous with
respect to his acts as he did not care whether Desdemona was unfaithful or not,
whereas Othello and PseudoOthello did. Thus, since they failed to achieve what
they intended to achieve by killing Desdemona whereas as the assassin achieved
what he set out to do, they suffer from diminished autonomy with respect to their
murderous acts in a way that he does not. But this response to the above defense
offered against this first objection confuses a person’s being autonomous with
respect to her action with her achieving what she set out to achieve by performing
it. A person (such as, here, Othello and PseudoOthello) could be perfectly autono-
mous with respect to her acts even if her performance of them does not achieve what
she intended, and so her exercise of her autonomy has little instrumental value for
her. But that this is so does not undermine the thin approach to the degree of under-
standing that a person needs to be autonomous with respect to her actions. Indeed,
this should be clear from the fact that the observation that a person’s act might have
little instrumental value for her is perfectly in accord with the original analysis of
the relationship between autonomy and informed consent: a person might be fully
autonomous with respect to her medical decisions, but, owing to negligence by her
healthcare provider, these decisions might fail to lead he to achieve what she
intended to achieve. They would thus have low instrumental value for here, and it is
on this basis—that is, on the basis of the failure to secure the value that she was
7 How Much Understanding Is Needed for Autonomy? 109

attempting to exercise her autonomy to secure—that we would hold her negligent

healthcare provider to be morally culpable for his failure to secure her informed
consent.

7.2.2 Agency, Autonomy, and Understanding

The objection that the thin view of the degree to which a person must understand her
actions to be autonomous with respect to her decisions to perform them is mistaken
because it fails to recognize that both Othello and PseudoOthello suffered from
diminished autonomy with respect to their acts of killing their wives can thus be
met. But this view of the degree to which a person must understand her actions to be
autonomous with respect to decisions to perform them faces a second serious objec-
tion: That it lacks the conceptual resources to distinguish between (mere) agents and
autonomous agents.
It is widely accepted that small children and cognitively impaired adults are not
autonomous agents.10 However, both children and cognitively impaired adults can
satisfy the thin view of the degree of understanding of their actions and their impli-
cations that is necessary for them to be autonomous with respect to their decisions
to perform those actions: They decide to perform acts under particular intensional
descriptions, and understand (to a degree) the implications of their actions. As such,
then, a proponent of the thin view of the degree to which a person must understand
her actions to be autonomous with respect to her decisions to perform them seems
committed to endorsing the implausible claim that small children and cognitively
impaired adults are autonomous with respect to their decisions to perform their
actions.11
But while this objection is initially plausible it is based on confusing the neces-
sary conditions required for an agent to be autonomous with respect to her decisions
and those that are sufficient for this. Clearly, it is necessary for an agent to be auton-
omous with respect to her decisions that she possesses the thin understanding of the
actions and their implications that she decides to perform. But more is also needed.
In particular, for an agent to be autonomous with respect to a decision that she
makes she must have made that decision “as a result of a decision-making procedure
that she is satisfied with as being her decision-making procedure for making the
type of decision that is in question” (Taylor 2009, 8).12 Since being satisfied with a

10
Widely, but not universally (see Mullin 2007).
11
I thank Jessica Flannigan for pressing a version of this objection.
12
This “Degree Condition” for autonomy is supplemented by both a Threshold Condition and a
Tracing Condition to provide an account of “practical autonomy” (Taylor 2009, 6–11). Note that
as it was explicated in Practical Autonomy and Bioethics this account of practical autonomy is
incomplete. This is because, in brief, it held that a person would be autonomous with respect to her
decision to perform a certain action if the information on which she based her decision had not
been affected in some way without her knowledge by another with the aim of leading her to make
110 J. S. Taylor

decision-making procedure requires that one have reflected upon it and accepted it
as one’s own, this additional requirement for a person to be autonomous with respect
to her decisions will preclude small children and cognitively impaired adults being
autonomous with respect to their decisions—even if they understood their actions
and their implications to the minimal degree required by the thin view of the degree
of such understanding that is required.

7.2.3 Maclean’s Objections

The final objection against the thin view of understanding that undergirded the
above argument that will be considered here has been developed by Alaisdair
Maclean. Writing of the example of Iago and Othello, Maclean argues
This example, while superficially attractive, fails to acknowledge a crucial issue. That is:
whose responsibility was it to ensure that Othello was sufficiently informed to make the
decision? If the duty is Othello’s then his capacity for autonomy has not been compromised
because it is his responsibility, and not Iago’s, to ensure that he is sufficiently well-informed;
Iago has done nothing that would constrain Othello’s ability to satisfy himself that the
information is sufficient and reliable. However, if the duty is Iago’s, and it is reasonable for
Othello to rely on the information, then Othello’s autonomy has been compromised because
he has been caused to believe that he is sufficiently informed (Maclean 2009, 43).

Noting that in modern society “it is impossible to live without relying on informa-
tion provided by experts,” Maclean continues,
Where the expert has a duty to provide sufficient information… then any failure to do so,
whether intended or not, will lead to a false reliance that undermines the individual’s auton-
omy. By inducing a false belief in the patient, that he or she knows enough to make a ratio-
nal decision, the patient’s motivation for seeking additional information is removed. Thus,
although the patient is free to seek additional information, he or she falsely believes there is
no reason to do so… This means that where a physician… misinforms the patient this
reduces the patient’s ability to make a rational—and hence autonomous—decision (Maclean
2009, 43 – 44).

Although it is not completely clear, Maclean’s argument seems to be this: (1) That
if B is responsible for ensuring that he was sufficiently informed then, if A does not

a particular decision, or particular type of decision, or, if it had, she was aware of how it had been
so affected (the Threshold Condition), and if she was satisfied that the decision-making procedure
that she used to make the decision in question was that which she was satisfied with in making
decisions of that type (the Degree Condition). But this account allows the possibility that a person
could be manipulated into being satisfied with a decision making procedure that she used to make
a decision D when she normally would not be satisfied with that procedure in making a decision of
type D. And if such manipulation was done with the intent of her making a particular decision D
without admitting the possibility that she would revise her decision then she would not be autono-
mous with respect to D. Given this possibility, the original Threshold Condition must be modified,
so that it applies not only to the information on which the person in question bases her decision
about whether or not to do X, but also to the information that the persons’ satisfaction with her
decision is based upon. I thank Steve Weimer for pressing me on this point.
7 How Much Understanding Is Needed for Autonomy? 111

constrain his ability to do so, B’s capacity for autonomy has not been compromised,
and so the thin view of how much understanding is required of B concerning her
acts is sufficient for her to be autonomous with respect to her decisions to perform
them; (2) That if A has a duty to ensure that B was sufficiently well-informed such
that if he fails to do so B’s autonomy will be compromised, then (if it would be
reasonable for B to rely on A) the thin view of how much understanding is required
of B concerning her acts is not sufficient for her to be autonomous with respect to
her decisions to perform them (for the thin view does not require that a person be
sufficiently informed for her to be fully autonomous with respect to her decisions);
(3) A physician has a duty to ensure that her patients are sufficiently well-informed
about their treatment options, and so (from (2) and (3)); (4) the thin view of how
much understanding is required of A concerning her acts is not sufficient for her to
be autonomous with respect to her decisions to perform them in a medical context.
While Maclean’s immediate conclusion is that “Taylor’s argument fails to sever
the connection between [informed] consent and autonomy” his argument also sup-
ports the more general claim that at least in some cases a person’s having only a thin
understanding of her actions is insufficient for her to be autonomous with respect to
her decisions to perform them. Yet there is no reason to accept either of these con-
clusions. As is clear from the above reconstruction of Maclean’s argument his con-
clusion rests on premises (2) and (3); (1) is redundant (for it allows that the thin
approach to the degree of understanding that is required for a person to be autono-
mous with respect to her acts is acceptable, at least under some circumstances).13
And the core of each of these two premises as they pertain to the question of how
much understanding a person must have concerning her actions for her to be auton-
omous with respect to her decisions to make them is the question of whether or not
she is “sufficiently informed”. But there’s the rub—for this raises the question of
what criterion should be used to determine whether or not a person is “sufficiently
informed”. The criterion clearly cannot be that the person is sufficiently informed to
be autonomous with respect to her decisions to perform the acts that she does, since
the underlying issue is precisely what degree of information (and hence understand-
ing) is required for her to possess this property with respect to them. That is, one
cannot reject the thin view of the relationship between a person’s autonomy with
respect to her decisions and her understanding of her actions on the grounds that it
would hold a person to be autonomous with respect to her decisions even if she was

13
This is a good thing for Maclean, for there are two problems with his first premise, both in itself
and when conjoined with the example of Iago and Othello. First, the question is not whether A’s
capacity for autonomy (i.e., his ability to be autonomous) would be compromised by the degree to
which he understands his actions and their implications, but whether the degree to which he under-
stands them is sufficient for him to be autonomous with respect to them. This, however, might be
merely a semantic issue, and there will be no violence done to the premise if this were changed.
However, when Maclean applies this premise to the case of Iago and Othello he claims that Iago
did not constrain Othello’s ability to ensure that he was sufficiently well-informed about his situa-
tion. This claim is either straightforwardly mistaken (for the very point of Othello is to show how
Iago manipulates Othello) or else Maclean has an implausibly restricted view of what would count
as one person’s constraining another.
112 J. S. Taylor

insufficiently informed about her actions properly to be so, since this would clearly
beg the question against the proponent of the thin view.
A more charitable interpretation of Maclean’s argument here, then, would hold
that the question of whether a person is “sufficiently informed” should be assessed
against something other than her autonomy. What, then, should this be? Recall here
that Maclean holds that “…where a physician… misinforms the patient this reduces
the patient’s ability to make a rational—and hence autonomous—decision”
(Maclean 2009, 44). As written, this is mistaken on two counts: A person can make
perfectly rational decisions on the basis of poor or false information, and there is
(outside the Kantian tradition) no necessary link between autonomy and rationality
(see Double 1992; Noggle 1997). Perhaps, though, what Maclean means here is not
that a misinformed person cannot make a rational decision, but that she cannot make
an advisable decision (see Gibbard 1998, 19). If this interpretation of what Maclean
intends here is correct then the question of whether a person is sufficiently informed
will become the question of whether she has sufficient information to decide to
perform that act which is likely to achieve the ends that she intends to achieve
through its performance. But if this is how we are to understand Maclean’s view
here it will face the response that was offered to the final defense of the Objection
from the Best Intensional Description—that we can distinguish between a person’s
being autonomous with respect to her decisions and the acts that she performs as a
result of them, and her making good instrumental use of her autonomy. Fortunately
for Maclean, it seems that this is not how we should understand his view. This is
because Maclean allows that a person who performs an action that is not advisable
for her could still be fully autonomous with respect to her decision to perform it
provided that it was her responsibility to ensure that she was sufficiently well-­
informed prior to making it. As such, then, the fact that B decided to perform an
action that was inadvisable for her does not ground Maclean’s view that she thereby
suffered from diminished autonomy. Perhaps, then, Maclean’s view is that B suffers
from a diminution in her autonomy with respect to her (inadvisable) decision as a
result of her ceding control to A over the decision as to when she would be suffi-
ciently well-informed to decide which act to perform, and his subsequent misuse of
this control. But this cannot be Maclean’s view, either. For Maclean, B does not
suffer from a diminution in autonomy with respect to her ceding this control to A for
if A were to perform his duty and provide her with adequate information—and so
Maclean’s objection is not based on B’s ceding control to A. The final possible
interpretation of Maclean’s view is that B will suffer a diminution in her autonomy
if she cedes control to A and A subsequently misinforms her, thus leading her to
make an inadvisable decision. But it would be odd if this were Maclean’s view, for
it is difficult to see how the combination of B’s ceding control and making an inad-
visable decision would compromise her autonomy when neither would do so alone.
Maclean, then, has no way of explaining why the issue of who is responsible for
informing B is so “crucial” to the question of whether or not B was autonomous
with respect to her decisions and consequent actions. There is thus no reason to
accept the argument that Maclean offers that supports the claim that at least in some
cases the thin view of the relationship between a person’s having only a thin
7 How Much Understanding Is Needed for Autonomy? 113

understanding of her actions is insufficient for her to be autonomous with respect to

her decisions to perform them.

7.3 Conclusion

In David Lodge’s academic satire Changing Places one of the characters, an

American careerist academic called Professor Zapp, was described as disliking
English academics, in part as they spent all their time complaining about the
American academy while at the same time freeloading off it as much as they could,
and in part because they seemed to consider their job complete if they raised ques-
tions about issues rather than actually answering them. In concluding this essay I
am sure that I, too, would annoy Professor Zapp, for this paper comes in the middle
of a research project, rather than at its end—and so no doubt there remain questions
to be asked to which I have not provided any answers. Before looking forward to
such annoyance, however, it is time to take stock of the positive conclusions that can
be drawn from the above discussion. First, the thin view of the degree of under-
standing of her actions and their implications that a person must possess for her to
be autonomous with respect to her decision to perform them can be defended against
the objections outlined above: the objection from the best intensional description,
the objection that it failed to distinguish between non-autonomous and autonomous
agents, and Maclean’s objections. Second, while the account of autonomy discussed
in this paper is based on a thin view of the degree of understanding that a person
must have of her actions to be autonomous with respect to her decision to perform
them it is, in another respect, a very thick view of autonomy indeed, insofar as the
answer to the question of whether a person is autonomous with respect to her deci-
sions will depend not only upon her mental states, but on those of others. But while
it might at first sight render this account of autonomy implausible (for it might seem
odd that my autonomy depends in part on the mental states of others) the fact that
this is an externalist, relational account of autonomy is a strength. The question of
when a person is autonomous with respect to her decisions is the question of when
it is she, and not someone else, who is the font of them. Unlike the property of iden-
tification (which is a metaphysical property ascribed to persons with respect to
those of their effective first-order desires that are truly her own, that can be properly
attributed to her as a person), autonomy (at least as it is understood above) is a
political property, one that persons have or lack on the basis of their relationships
with other persons (see Taylor 2009, chapter 3). Thus, since control is an intention-
ally characterized concept, and so one can only answer the question of whether
person one person controlled the decisions of another by reference to the first per-
son’s mental states, it is only to be expected that the question of whether a person is
114 J. S. Taylor

autonomous with respect to her decisions must be answered by reference to the

mental states of others.14
It is now time to annoy Professor Zapp. It was argued above that autonomy only
requires persons to have a thin understanding of their actions for them to be autono-
mous with respect to their decisions to perform them. The intuitive basis for this
argument was that a person was autonomous with respect to his decisions if was she,
and not someone else, who directed her to make them. But it must be admitted that
this is not the only way in which one might intuitively understand autonomy.15 One
might hold that a person was autonomous with respect to her decisions if they led
her to perform actions that enabled her to instantiate her will in the world. On this
understanding of autonomy having only a thin understanding of one’s actions and
their implications would be insufficient for a person to be autonomous with respect
to her decisions to perform them, for this would be unlikely to enable one to instan-
tiate her will through her actions. A person, for example, whose physician negli-
gently failed to inform her of some of the serious implications of her treatment
would not be able effectively to instantiate her will in the world, for she would lack
the requisite information to do so.16
Which of these intuitive understandings of autonomy should take precedence?
Although this issue cannot be conclusively settled here it should be noted that there
is at least prima facie reason to prefer the thinner approach to autonomy defended
above. First, the proponents of the more robust understanding of autonomy cannot
criticize its thinner rival simply by noting that this account of autonomy appears to
be of little value. One might, for example, note that the thinner account of autonomy

14
As part of his argument against the thin account of understanding that a person must have of her
actions and their implications in order for her to be autonomous with respect to her decision to
perform them, Maclean argues that control should not be understood as an intentionally character-
ized concept, and that one person could exercise control over the actions of another without thereby
intending to do so. In support of this view he offers an example in which children inadvertently
switch train tracks from one line to another, and thus “undermined the railway company’s control
of the train’s route”; an example, he believes, whereby one party’s control is undermined uninten-
tionally by another. But this example fails to show that control can be exercised unintentionally, for
the children do not control the direction that the train goes in; they merely affect it. To see the
distinction here, consider the difference between a leaf’s being blown by the wind as it falls, and
its being choreographed in its falling by someone with a radio-controlled device whose receiver
and engine she has attached to the leaf, so that it will weave the patterns that she desires as it
descends. Here, the wind would not control where or how the leaf falls, but would merely affect
this; the person with the radio-controlled device, however, would control it, and not merely affect
it, for she would intentionally direct its falling. More generally, note that Maclean’s example here
fails to touch the examples of Othello and Pseudo Othello, for the childrens’ switching of the
tracks does not affect (and so cannot control) the decisions of the railway company’s personnel;
nor does it affect (and hence does not control) their actions. Instead, they merely affect (but do not
control) the outcomes of these decisions and actions. As such, Maclean’s example is irrelevant to
the Shakespearian examples which focus on establishing that for one person to control the deci-
sions (and hence actions) of another she must intend to do so, for it does not address this point at all.
15
Although it might be that this would still be the most defensible approach, as is indicated below.
16
This appears to be the understanding of autonomy that Mele accepts at times (Mele 1995,
180, 181).
7 How Much Understanding Is Needed for Autonomy? 115

would hold that a person could be fully autonomous with respect to her decision to
perform some action that owing to her ignorance of its true nature would not only
fail to advance her interests, but would actively hinder them. With this observation
in hand, one might then note that it is not clear why one should value being “autono-
mous” in this sense. Hence, one might conclude, since the possession of autonomy
is supposed to be valuable, then insofar as its possession is not (obviously) valuable
when it is understood thinly thinner conceptions should be rejected.17 But this objec-
tion moves too quickly, for (as noted above, in response to the Objection from the
Best Intensional Description) a proponent of the thinner understanding of autonomy
defended here can readily distinguish between a persons’ being autonomous with
respect to her decisions simpliciter and the degree to which her autonomy is of
instrumental value to her (Taylor, 2009, chapter 10). And, with this distinction in
hand, she could then argue that the value of autonomy was primarily instrumental,
such that while a person could be fully autonomous with respect to those of her
decisions that were made in conditions of relative ignorance (for she need only have
a thin understanding of her actions and their implications for her to be autonomous
with respect to her decisions to make them) her autonomy would possess only low
(or no) value under such circumstances. Second, while the more minimalist account
of autonomy can respond to the charge that it cannot account for the value of auton-
omy the proponents of its more robust rival face a problem of vagueness, for it is not
clear what would count as “instantiating one’s will in the world”. Obviously, requir-
ing that a person’s actions must lead her to succeed in her goals for her to count as
being autonomous with respect to her decisions to perform them would be an
implausibly stringent standard. But some degree of success in a person’s decisions
leading to her achieving her goals must be required for her to be autonomous with
respect to them on this more robust account of autonomy, for unless this is so it is
unclear how one is to understand the claim that to be autonomous a person must
instantiate her will in the world.18 Given the vagueness of this more robust approach
to understanding autonomy, and given the fact that the thinner account defended
above can account for the view that in some cases a person’s autonomy will be of
little value, there is at least prima facie reason to accept the thinner account. As
such, then, the simple answer to the question of how much understanding is required
for autonomy is: Very little.19

17
This criticism of the thin account of how much understanding is necessary for autonomy is simi-
lar to that which Berlin (Berlin 1969, 124) recognizes could be leveled against the negative account
of freedom.
18
A weaker requirement might be that a person is autonomous with respect to her decisions only if
it is likely that they will lead to her will being instantiated in the world. But while this might be a
more plausible version of the more robust approach to autonomy, this plausibility comes at the cost
of even more vagueness is its application.
19
An earlier version of this paper was presented at the “Thick Concepts of Autonomy Conference”
at the Center for Advanced Study of Bioethics, Westfalische-Wilhelms-Universität, Münster,
Germany, October 2012. I thank all of the participants in that programme for their exceptionally
helpful comments on my paper which helped me to improve it greatly.
116 J. S. Taylor

References

Beauchamp, Tom L., and James F. Childress. 1994. Principles of Biomedical Ethics. 4th ed.
New York: Oxford University Press.
Berlin, Isaiah. 1969. Two Concepts of Liberty. In Four Essays on Liberty, ed. Isaiah Berlin,
118–172. Oxford: Oxford University Press. 124.
Double, Richard. 1992. Two Types of Autonomy Accounts. Canadian Journal of Philosophy 22
(1): 65–80.
Dworkin, Gerald. 1976. Autonomy and Behavior Control. Hastings Center Report 6 (1): 23–28.
Gibbard, Alan. 1998. Wise Choices, Apt Feelings: A theory of normative judgment. Oxford:
Clarendon Press.
Maclean, Alasdair. 2009. Autonomy, Informed Consent and Medical Law: A Relational Challenge.
Cambridge: Cambridge University Press.
Mele, Al. 1995. Autonomous Agents: From Self-Control to Autonomy. New York: Oxford
University Press.
Mullin, Amy. 2007. Children, Autonomy, and Care. Journal of Social Philosophy 38 (4): 536–553.
Noggle, Robert. 1997. The Public Conception of Autonomy and Critical Self-Reflection. Southern
Journal of Philosophy 34 (4): 495–515.
Taylor, James Stacey. 2004. Autonomy and Informed Consent: A Much Misunderstood
Relationship. Journal of Value Inquiry 38 (3): 383–391.
———. 2009. Practical Autonomy and Bioethics. New York: Routledge.
Chapter 8
Is “Autonomy Talk” Misleading?

Thomas Gutmann

Abstract Especially during the last three decades, there has been a fine-grained
debate on the notion of personal autonomy. This article will suggest taking one
important distinction within this debate more seriously: the differentiation between
(a) “personal autonomy” designating a family of ideal, gradual, and hence at least
moderately perfectionist conceptions about what an autonomous person is and (b)
what is called “autonomy as right” (Feinberg. 1986. 47). These are not different
dimensions of one concept, but independent concepts with disparate, even conflict-
ing normative functions and should be kept clearly separate from each other.
Treating these two concepts merely as different aspects of one and the same notion
of personal autonomy, as done by most authors, is severely misleading. For clarity,
we will ultimately need different words for divergent concepts. It will be suggested
that “autonomy as (moral or legal) right” should no longer be called “autonomy”,
but forthrightly the “right to self-determination” or, in accordance with Thomas
Hobbes, “authority”.

8.1 A Functional Approach

Autonomy—the question of whether a person is autonomous or acts autonomously

or has the authority to act as she likes—has different functions in particular norma-
tive settings. One of the basic functions of autonomy—not only in bioethics and
(medical) law—is related to a person’s claim (her moral or legal right) to be
respected as an agent entitled to choose her own path and follow it. “Autonomy as
right” is what is meant when the “principle of respect for autonomy” is invoked in
the way Beauchamp and Childress use the term in their Principles of Biomedical

T. Gutmann (*)
Faculty of Law, University of Muenster, Muenster, Germany
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 117

Switzerland AG 2022
J. F. Childress, M. Quante (eds.), Thick (Concepts of) Autonomy, Philosophical
Studies Series 146, https://doi.org/10.1007/978-3-030-80991-1_8
118 T. Gutmann

Ethics in regard to “individuals’ decision-making”: “To respect autonomous agents

is to acknowledge their right to hold views, to make choices, and to take actions
based on their values and beliefs” (Beauchamp and Childress 2013, 101 et seq., 106;
cf. Beauchamp 2005). The necessarily formal notion of having a set of rights expres-
sive of one’s sovereignty over oneself (i.e. of having a protected space in which one
may choose freely between options for action [Sect. 6]1) may well be called a ‘thin’
concept of autonomy, as there is no specific content required for the choices such an
autonomous (sovereign) person is allowed to make within such a right.2 So where
we have reason to conceive “autonomy as right” (i.e. as right of “self-­determination”
or as “authority”, respectively), ‘thick’ concepts of autonomy are out of place.

8.2 ‘ Thick’ Concepts of Autonomy as Concepts

of the Good Life

When I think about what kind person I want my 15-year old son to be or to become,
what comes to my mind are so-called ‘thick’ concepts of autonomy referring to a
certain personality structure.
I want him to be his own person, to live his life according to reasons, motives,
and principles that are his own and thus give meaning and coherence to his life. I
want him to have the second-order capacity to reflect critically upon his first-order
preferences, desires, wishes, and so forth, and the capacity to identify with these or
to attempt to change them in light of his higher-order preferences and values
(Dworkin 1988, 20, 108; see Quante 2011, 598; id. 2017, 119 et seq.).3 I want him
to be authentic in the sense that he is able to reflectively endorse (or at least not be
alienated from) his basic organizing values, motivations, and commitments in the

1
The metaphor of space was always crucial for the notion of autonomy rights, see e.g. Feinberg
1986, 28, speaking of the “sovereign authority to govern oneself, which is absolute within one’s
own moral boundaries (one’s ‘territory’, ‘realm’, ‘sphere’, or ‘domain’)”; Wellmann 1985, 95
(“dominion”) and Steiner 1994, 90 (“domain”). The metaphor goes back at least to von Savigny
1867, 6, who (in this regard following) Immanuel Kant’s notion of right) defines “the right in a
subjective sense” as “a territory in which his will rules and rules with our consent”. It was taken up
in consequentialist accounts of (quasi-) rights to self-determination, see Mill 1992, Ch. I, IV
(“Over himself, over his own body and mind, the individual is sovereign”).
2
Autonomy as right does not have to be exclusively a negative right. It may also generate reasons
for others to assist or to promote the autonomy competence of others as well, cf. Beauchamp,
Childress 2013, 101 et seq., and Mackenzie 2008. See for the needs of vulnerable persons who are
attempting to reconstruct a sense of practical identity required for their autonomy, Christman 2014.
3
Quante, in order to avoid a utopian ideal of autonomy that is unattainable for mortals, convinc-
ingly introduces a capacity for critical self-evaluation which contains a counterfactual: “a person
has personal autonomy with respect to a first-order-attitude if she would be able to exercise the
capacities necessary for personal autonomy if she noticed this first-order-attitude (or if someone
asked her to critically reflect her first-order attitude).” Thus, permanent critical reflection and iden-
tification is not required for personal autonomy. Cf. Quante 2017, 119, 131, and also Christman’s
(2009, 155) hypothetical reflection condition for the authenticity element of autonomy.
8 Is “Autonomy Talk” Misleading? 119

light of his diachronic practical identity, i.e. the historical processes (including my
own role in them) that gave rise to these characteristics (Christman 2009, 133
et seq., 143). I therefore hope that he will be able to evaluate the formation of his
desires (Elster 1983, 109 et seqq.) in a process of sustained critical reflection
(Christman 2009, 152), exercising the repertory of necessary skills for autonomous
self-discovery and self-definition (Meyers 1989, 49; id. 2000, 163, 166). I want him
to fully develop the affective attitudes—attitudes of self-respect, self-trust, and self-­
esteem—necessary for a view of himself as the legitimate source of reasons for
acting (Mackenzie 2008, 524–527; Honneth and Anderson 2005). I want him to
work through his unconscious connections of instinctual drives to a conscious lib-
erty with which he can have his self-powers at his disposal as unimpededly as pos-
sible, without being hindered by frustrating, infantile expectations later in his life
(Freud 1972a, 317; id. 1972b, 516; cf. Berofsky 1995). I want him to be responsive
to a sufficiently wide range of reasons concerning his choices (Berofsky 1995, 182
et seq., 199) and to subject his opinions and tastes to rational scrutiny. I want him to
be authentic to the extent that he can and does alter his convictions for these reasons
of his own and that he does this without guilt or anxiety (Feinberg 1986, 33).
I want him to be programmatically autonomous, posing and answering the ques-
tion “How do I want to live my life?” (Meyers 2004, 8; id. 1989), and, therefore, I
would like him to lead a life that forms an ongoing narrative, a life with a high
degree of biographical consistence and coherence which follows what Rawls calls a
“rational life plan” that “would be chosen … with full deliberative rationality, that
is, with full awareness of the relevant facts and after careful consideration of the
consequences” (Rawls 1971, § 63, 408). In this sense, I want him to be a self-­
governing, planning agent (Bratman 1987; id. 2007, 195 et seq.), but, on the other
hand, I do not want him to experience himself as a “tight identity knot” (Meyers
2000, 163), but to be able to seriously imagine alternative lives (Mackenzie 2000;
Meyers 2004, 10) and to realize that this dynamic process of autonomous self-­
discovery and self-definition will be a lifelong journey based on piecemeal change,
on improvisation, and on constant balancing of ambivalent and contradicting con-
stituents of his identity (cf. Meyers 2000, 173–175).
Finally, I want him, on the basis of an authentic interpretation of his “real” objec-
tives, to be able to make “strong” evaluations in regard to the success of his life and
to his identity as a whole (Taylor 1985a, b; cf. Benn 1976, 125 et seq.; id. 1988, 9,
178 et seq.). However, in contrast to Taylor’s overboarding 1980s communitarian-
ism, I also want my son to free himself from unreasonable conformist attitudes and
role expectations (Meyers 1989; id. 2000; Oshana 2006a; id. 2006b), as well from
cognitive limitations on his spontaneous experience (Berofsky 1995). In other
words, I want him to guide his life from his own perspective, rather than be manipu-
lated by others, and to be able to free himself from “heterarchy” (Benn 1988, 164
et seqq.) and from considerations, values, desires, and emotions imposed externally
upon him. I want him to have a “desire … for self-governance” (Oshana 2006a,
101), and yes, I even want him to be that kind of idiosyncratic eccentric John Stuart
Mill associated with genius, mental vigour, and moral courage (Mill 1992, Ch. III).
The latter being a little close to a specific male and atomistic fantasy, I do hope that
120 T. Gutmann

all this will not prevent him from experiencing himself as deeply socially embed-
ded, his capacity to exercise his autonomy depending (up to a certain point) on
social institutions and the fragile nexus of relationships of trust, loyalty, friendship,
love, caring, and shared values.4
I want all this because I care for what I think constitutes a good life for him. All
these ‘thick’ notions of autonomy are different conceptions of the same concept of
autonomy (and, although the variety of these conceptions demonstrates substantive
philosophical disagreements about the conditions and capacities necessary for
‘thick’ autonomy, we do not have to choose between them here5). This concept of
autonomy, however, is a concept of what it is to lead a ‘good’, valuable and flourish-
ing life (Mackenzie 2008, 529).6 Or at least, as long as the criteria for personal
autonomy remain largely formal and are therefore not sufficient for a comprehen-
sive definition of the good life (Quante 2017, 125), they still define a part of such a
concept. As an ideal, the notion of autonomy used to qualify the structure of a per-
son’s life belongs to the realm of value-centered reasons (Benn 1988, 10). ‘Thick’
concepts of autonomy describe a value, even “an irrefutable” one (Christman 2015,
Sub 1). Values can be realized more or less, which is one difference between values
and rules.

8.3 The Good and the Right

One basic function of the notion of “autonomy”—not only in (bio-) ethics and
(medical) law—, however, has to do not with the good, but with the right, and, as we
shall see, with the idea of the priority of right (Rawls 1988). The latter is related to
a person’s claim (to her moral or legal right) to be respected as an agent who is
entitled to choose her own conception of the good and follow it. To quote Rawls,
persons are “self-authenticating sources of valid claims” (Rawls 2001, 23 et seq.)
because identifying a natural person as a moral (or legal) person is to recognize her
as a bearer of rights (Benn 1988, 9).
This realm of rights (cf. Thomson 1990) to self-determination is a central feature
of our concept of normativity and deeply embedded in our normative (moral and
legal) practices. In a sense, respect for persons may simply be respect for their rights

4
Although, strictly speaking, some aspects of such a notion of “relational autonomy” (cf. Mackenzie
and Stoljar 2000a) are no longer a concept of autonomy at all, but of aspects of human flourishing
that should be clearly distinguished from personal autonomy.
5
We can also skip the question here of whether “the ideal of personal autonomy […] requires not
merely the presence of options, but of acceptable ones” (Raz 1986, 205) or “calls for positive
freedom”, “de facto power”, and the social, political, and economically means to realize one’s
choices (Oshana 2006a, 101 et seq.; id. 2006b, 2; id. 1998; Mackenzie and Stoljar 2000a;
Mackenzie 2014, 25).
6
In a wider sense, personal autonomy in this sense is also one of the elements of well-being (cf.
Mill 1992, Ch. III).
8 Is “Autonomy Talk” Misleading? 121

and their capacity to assert claims (Feinberg 1970). How central it is depends, of
course, on the normative framework we (have reasons to) choose. For what we can
call mainstream liberal Kantianism in moral and legal philosophy, respecting rights
is the core element of morality which deals with what we owe to each other. It is a
normative relation of (at least prima facie) lexical priority (Dworkin 1977). For a
consequentialist, rights may be of only instrumental value for individual and collec-
tive well-being (Raz 1986, 185 et seq., 203 et seq.), and, for a principlist approach,
the principle of respect for autonomy (i.e., for the right to self-determination) does
not necessarily always have priority over other considerations (Beauchamp and
Childress 2013, 101, 141).
Axel Honneth is right in pointing out that respecting rights is one basic form of
social recognition and that it is the only egalitarian one (Honneth 1996, 109 “legal
recognition”; cf. Honneth and Anderson 2005, 138). Being a moral and legal agent
with free will, able to freely decide which actions to take and which path to follow
in our life, applies to all of us equally. It is a conception of persons as moral equals.
There is no degree of being worthy of this kind of recognition,7 neither is there any
room for exceptions or privileges. Law, at least in the Western legal tradition, is
deeply embedded in this sort of egalitarian normative individualism, but for the
purpose of this article, it suffices to address the right to self-determination as a
moral claim.

8.4 I deal and Threshold Concepts of Autonomy: Equality

and Inclusion

In almost all the varieties of its philosophical use as described above (sub 2), a sub-
stantial concept of personal autonomy (autonomyi) designates an ideal8 which real
persons can, at best, only approach. Autonomy in this sense is always a matter of
degree—a person can be more or less autonomous (Mackenzie and Stoljar 2000b,
18). As a goal to be attained, “it may well be enjoyed by very few if any individuals”
(Christman 2015, Sub 1.1.). In this development-orientated and thus broadly perfec-
tionist sense, however, autonomy cannot be an essential condition for guaranteeing
persons areas of liberty for self-responsible decision-making.
“Autonomy as right” (or better, the right of “self-determination”) means that we
attribute to individuals the normative competence for decisions concerning their
own sphere of life (and corresponding duties for other persons at least not to inter-
fere). Autonomy as the ontic status of being a person that can claim a right to self-­
determination (autonomyr) must be independent of the question of how far she

7
Cf. Darwall’s (1977) notion of “recognition respect”, which is neither context dependent nor a
matter of degree.
8
It may be “at best only a partial ideal“(Feinberg 1986, 45), “an actually unattainable ideal”
(Quante 2017, 130) and “an ideal deeply rooted in the liberal tradition” (Benn 1988, 10), but it is
an ideal.
122 T. Gutmann

realizes the value of autonomous personhood (autonomyi). Therefore, when we

address “autonomy as right” (autonomyr) it is only a “threshold conception of natu-
ral competence—minimal relevant capacity for a task—that [must be] used in stipu-
lations of necessary and sufficient conditions for the sovereign right of
self-­government ascribed to individuals” (Feinberg 1986, 29 et seq; Wikler 1983;
VanDeVeer 1986, 351; Benn 1988, 10; Beauchamp 2005, 316). In order to be con-
sidered a valid decision-making body in one’s own affairs, it must suffice to be suf-
ficiently autonomous, i.e. to have basic cognitive and volitional competence
(regarding rationality, self-control, and the absence of certain mental pathologies).9
We have to be competent above a certain threshold, but we do not have to be more
of an autonomous person in order to be entitled to make use of our rights. So, in
relation to the right of “self-determination”, on grounds of its functional purpose,
the concept of autonomy designates a range property (Wikler 1983, 87; cf. Rawls
1971, 508) and is binary—either a person is entitled to act on her behalf, or she is
not.10 Autonomous actions in this sense (autonomyr) thus cannot “be autonomous to
a different degree” (Quante 2017, 139). The ontic status of being considered a valid
decision-making body in one’s own affairs is an egalitarian notion. Everyone who is
ascribed this status has it equally. Otherwise—if autonomyr were not an all-or-­
nothing characteristic—the commitment to moral equality would become problem-
atic, and the fundamental assumption of our legal systems (and of our conceptions
of moral rights as well) that, in general, individuals have normative competence for
their decisions would lose its foundation and would refer to paternalistic tutelage of
the average citizen by members of society endowed with exceptional autonomy.
Behind the functional reason, there is also a normative one for this. In order to
achieve normative inclusion and equality, the ontic status of autonomyr—having
rights to self-determination and the authority to use them—must include at least
most adults. Ascribing rights to persons (whether liberty rights or rights to demo-
cratic participation) has to be an extremely inclusive concept. In this perspective,
autonomyr has to define a default position. It is the very sense of ascribing rights
within a liberal society based on rights (or in a notion of morality based on equal
respect) for this to be as general a concept as possible. Otherwise, it would be
impossible to treat persons “as equals” or “with equal concern and respect”, as

9
Of course, some authors in the debate about autonomyi see this point clearly, e.g. Christman 2015,
Sub 1.1., and Beauchamp. 2005, 319 (We need “a way to ensure that ordinary choices qualify as
autonomous even when persons have not reflected on their preferences at a higher level and even
when they are hesitant to identify with one type of desire rather than another”).
10
It is only through such a low threshold concept of autonomy that the fundamentally egalitarian,
as well as the inclusive character of the concept (or regime) of autonomy can be ensured. It is not
until this is accomplished that it makes sense to think about autonomy-enabling policies in order
to address Joel Anderson’s critique that the focus on low threshold conditions fails to protect less-­
competent groups in society because “the central fact about individual autonomy is how limited the
competence of many individuals is and how vulnerable groups are significantly disadvantaged by
arrangements in which benefits are distributed in part on the basis of the autonomy-competence
that individuals develop ‘naturally’” (Anderson 2004).
8 Is “Autonomy Talk” Misleading? 123

Ronald Dworkin puts it (Dworkin 1977, 181 et seq., 273–292).11 This is the main
reason why the kind of ‘autonomy’ referred to here has to be a thin and binary
­concept. First, as basic competence is a status designator (Christman 2015, Sub
1.1.) and judgments about it have a normative gatekeeping function (Beauchamp
and Childress 2013, 114) by distinguishing persons whose decisions should be
accepted from those whose decisions should not be accepted, the threshold applied
must not be set too high. Second, the number of possible reasons for assuming that
a generally competent person is not acting autonomously in the sense of not volun-
tarily (i.e. not in a way that allows attribution of responsibility) is restricted to the
two forms of a lack of voluntariness: coercion, on the one hand, and deceit and
error, on the other, the first two being modes of external control. Voluntariness, too,
is a binary concept (Wertheimer 1987).
‘Thick’ concepts of autonomy are generally inadequate and dysfunctional in this
regard. Autonomy understood as an ideal and gradual notion and not as an all-or-­
nothing characteristic is an inherently unequal concept. So when it comes to respect-
ing claims to autonomy as self-determination, e.g. in biomedical ethics or medical
law, the question is not whether referring to ‘thinner’ conceptions of autonomy
might suffice. The question is whether any ‘thick’ concept of autonomy can do the
job at all. The answer is clearly negative.

8.5 Interrelations and Disruptions

There are positive evaluative interrelations, but also tensions between different
‘thick’ concepts of autonomy and the ‘thin’ concept of autonomy as right. (a) In
short, the ideal of being one’s own person, living one’s life according to reasons and
motives that are one’s own, is a basic moral and political value and might be the
very reason why we attribute rights to self-determination in the first place. Vice
versa, having rights and being able to claim them is necessary for self-respect and
gives a sense to the notion of personal dignity, both of which are prerequisites for
leading an autonomous life (Feinberg 1970, 252).12 As a political ideal, citizens’
autonomyi is the main source of legitimacy for just political institutions (Christman
2009, 135) and provides reasons for arguing against institutions that attempt to
impose a set of ends and values upon the citizens of a society (Dworkin 1988, 10)
instead of respecting their right to live their lives in accordance with their own con-
ception of the good. Having basic cognitive and volitional competence (regarding

11
The very idea of rights involves a commitment to equality, cf. Waldron 2007, 752.
12
“Having rights enables us to ‘stand up like men’, to look others in the eye, and to feel in some
fundamental way the equal of anyone [...] To respect a person then, or to think of him as possessed
of human dignity, simply is to think of him as a potential maker of claims”. Self-respect (together
with self-trust and self-esteem) serves as a necessary element of several thicker conceptions of
autonomyi as well, cf. Honneth and Anderson 2005, 132–138; Mackenzie 2008, 525.
124 T. Gutmann

rationality, self-control, and the absence of certain mental pathologies) is a prereq-

uisite for both autonomyi and autonomyr.
(b) However, “personal autonomy” and rights to self-determination are indepen-
dent concepts with different and conflicting normative functions. The very form of
protecting autonomy by granting rights to self-determination must produce heavy
tensions between acts which are protected by these rights and the overall goal of
realizing ‘thick’ concepts of autonomy. Rights to self-determination must include
the right to act in a way that misses the point or even endangers ‘thick’ ideals of
autonomy. A right to self-determination protects a domain of choice for a person
whose reasons for her behavior are necessarily linked in each case to her current
self-understanding (Williams 1981, 13). If the ideal of ‘thick’ autonomy were gen-
erally “one of those values for which freedom can be sacrificed” (Oshana 2006a,
126), this would reduce each person as a whole to the role of a mere trustee of her
future manifestations of autonomyi in a way that would be utterly incompatible with
the very idea of having rights to self-determination here and now (autonomyr). Such
a strategy would disengage a reified concept of “autonomyi” from its bearers and
turn it against them.
This is the reason why, in general, no moral or legal consequences follow from
the verdict that a sufficiently competent bearer of an autonomy right has not acted
“autonomously enough”, as measured by ‘thick’ standards. For this reason, we can
conclude that, as far as persons (following Rawls) are self-authenticating sources of
valid claims, it is not, as Christman claims, the autonomyi of the person “that
authenticates the validity of these claims” (Christman 2009, 135). If the presump-
tion is “that autonomy confers [...] the authority to make decisions of practical
importance for one’s life, for one’s own reasons, whatever those reasons might be”,
and “[a]utonomous persons are presumed to have the capacity, the right and the
responsibility to exercise this authority” (Mackenzie 2008, 512), then no conception
of autonomyi—regardless of whether it is a procedural (content-neutral) or a sub-
stantive one (Mackenzie and Stoljar 2000b, 13 et seq.)—qualifies as an adequate
criterion for this function.13 Autonomyi cannot be a necessary precondition for
allowing someone to make use of a right to self-determination, and the very reason
for this is that it would indeed be “a mistake to think of personal autonomy[i] as an
all or nothing affair and to draw a sharp distinction between autonomous and non-­
autonomous agents” (Mackenzie 2008, 523).

13
Mackenzie sees that herself, in part, cf. Mackenzie 2008, 523 (“In the case of citizens’ rights to
de jure political autonomy, the threshold level of competence required ought to be minimal. Agents
who meet this minimal level of competence ought to be treated as politically autonomous and
entitled to enjoy the rights and liberties that it guarantees, including freedom from the unwarranted
paternalistic intervention of the state. To characterize an agent as personally autonomous, however,
involves attributing to her capacities that go well beyond this minimal threshold”). Mackenzie’s
whole argument, however, seems to imply that in moral (as opposed to legal) theory, persons who
are not autonomous in a ‘thick’ sense should not have “the capacity, the right and the responsibility
to exercise […] the authority to make decisions of practical importance to one’s life, for one’s own
reasons, whatever those reasons might be” (512).
8 Is “Autonomy Talk” Misleading? 125

We face this problem with every ideal-type, (at least moderately) perfectionist
notion of personal autonomy, even if it follows a procedural account, i.e. is neutral
towards the content of a person’s values, beliefs, and desires as long as they are
endorsed autonomously. What if my upbringing left me with a neurosis that I, upon
historically sensitive, adequate self-reflection, am neither willing nor able to endorse
nor to identify with, but rather from that I am alienated? Am I allowed to act upon
it, nevertheless (although it may not be advisable)? In any case, it is my neurosis,
not yours. Am I allowed to choose a partner although my therapist tells me that I am
acting upon my unconscious Freudian repetition compulsion? And if I marry her or
him anyway, can this person demand me to stick to my valid commitments, i.e. to
my marriage contract? Should I be allowed to follow a career plan although I still
feel somewhat alienated from the basic values and commitments I am acting upon?
The answer, of course, is yes. In other words: interpersonal respect for the autono-
myr of a bearer of rights must not require her to positively identify with or not to be
alienated from the trait that motivates her to act. Even people feeling deeply alien-
ated from a personal characteristic have to be allowed to decide how to live their
lives, given the persons they are at the moment, even if they are acting out this very
characteristic. There may be a right to make alienated decisions as a subform of the
general “right to do wrong”, as Jeremy Waldron analyzed some 35 years ago
(Waldron 1981). In general, a feeling of alienation does not necessarily undermine
a person’s competence or her ability to act voluntarily. Even neurotic motives and
mildly compulsive behavior do not necessarily undercut these features (Feinberg
1986, 168). The same, again, is true for a lack of authenticity, as long as the term
“refers to the obligation to live up to one’s ‘true’ self, […] as a form of integrity”
(Feinberg 1986, 134; cf. Quante 2017, 129). Even if it comes to decisions of exis-
tential importance, we cannot require the bearer of a right to self-determination to
reveal some kernel of his true or authentic self. Having a right means not having to
give reasons of this kind (Habermas 1996, 119), and inauthentic behavior in general
may justify queries, but not interference with people displaying basic competence
(which again, for reasons of inclusion, has to be a rather low threshold concept).14
In the final analysis, the biographical dimension which must be so important in
models of personal autonomyi, is widely irrelevant for autonomyr. Therefore,
authenticity conditions can only play a very limited role in the concept of autonomy
as right to self-determination.
As far as I have a right to self-determination, I have the moral or legal right to
make critical life decisions although I am a conformist bore, and I am entitled to act
“out of character”, to change my life spontaneously in an existentialist manner,
discarding biographical consistency, coherence, and rational life plans. In other
words: I am allowed to be a bad (‘thick’) autonomy achiever. If, in one of my exis-
tentialist moods, I quit my job, sell my house, and leave my partner voluntarily, but
in a manner that rejects my biographical narrative so far, my rational life plan, and

14
In some contexts, the stability of a decision might function as a place-holder for its voluntari-
ness—something which is much less than authenticity.
126 T. Gutmann

the relational nexus of my autonomy, I nevertheless have to take moral (and legal)
responsibility for the moral (and legal) consequences of my decisions and acts. And
this is because ‘thick’ concepts of autonomy, on the one hand, and the right to self-­
determination, on the other are functionally independent concepts with different
normative functions and must be clearly distinguished.
It becomes even clearer that autonomyi cannot be a necessary precondition for
allowing someone to make use of a right to self-determination when one regards
substantive notions of personal autonomy that place restrictions on the contents of
agents’ value commitments, beliefs, and preferences in order to qualify as autono-
mous (cf. Mackenzie and Stoljar 2000b, 13 et seq.).
Some persons may deliberately forge lives in which autonomy is largely absent,
though they possess the capacity and the freedom to do otherwise. Some people
become monks leading a life in total obedience or join law firms. Marina Oshana
presents the example of a woman, who, having previously enjoyed a successful
career as a physician, has since voluntarily chosen to wear a burqa and to lead a life
of utter dependence, out of the earnest belief in the sanctity of this role and/or in
unquestioned adherence to a religious tradition. Oshana claims that “it is evident
that [this] woman is not autonomous” (Oshana 2006a, 104). According to Oshana’s
specific ‘thick’ concept, this may be the case.15 But what follows from this state-
ment? Are we entitled to interfere with her choice? Is this a case for incapacitation,
for legal tutelage, for coercive re-education, for taking away her children, for declar-
ing her contracts null and void? Or is it at least a reason to complain that the law will
unfortunately not allow us to act in this way, although moral reasons and an ade-
quate theory of ‘autonomy’ would be clearly in favor of doing so?
Some authors claim that we should not wrongly attribute autonomy to those
whose restricted socialization and oppressive life conditions pressure them into
internalizing oppressive values and norms (cf. Oshana 1998; id. 2006a. 104;
Mackenzie and Stoljar 2000b. See also Oshana 2014). If we agree for the sake of the
argument, the question must arise of whether such “non-autonomousi” people shall
be allowed to make their own decisions although their capacity for ideal autonomy
has been impaired by their oppressive social environments. Again, the answer, in
general, must be yes because making use of individual autonomy rights is very often
the only exit option from such environments, whereas denying such people the right
to make their own decisions would add insult, degradation, and exclusion to oppres-
sion. So if, e.g. according to Natalie Stoljar, “the feminist intuition [...] claims that
preferences influenced by oppressive norms of femininity cannot be autonomous”
(Stoljar. 2000. 95),16 taking this notion of autonomyi to withhold autonomyr, i.e.
rights of self-determination, from these women would not only amount to quite
extensive forms of unwarranted paternalistic interference (Christman. 2004. 157)
and to denying existing opportunities for choice, but would also imply that other

15
Although this is highly questionable. “It is one thing to publically criticize modes of social prac-
tice that denigrate their participants, but it is another to define autonomy in a way to claim that
those participants are not fully functioning agents at all” (Christman 2004, 158).
16
See Meyers 2004, 3–12, though.
8 Is “Autonomy Talk” Misleading? 127

persons may legitimately serve as decision-makers in these cases. It would amount

to specifying a huge (maybe even a majority) group of second-class citizens in need
of legal guardians, thus undermining the ontic status of basic equality of persons.17
If, however, we sever the justificatory ties between such conceptions of autonomyi,
on the one hand, and respect for autonomyr, on the other (which must clearly be
done), the claim that we should not wrongly attribute autonomy to those who have
internalized oppressive values and norms loses most of its relevance.
(c) The potential dysfunctionality of ‘thick’ concepts of personal autonomy in
certain—but not all—contexts of biomedical ethics can be further elaborated with a
look at Michael Quante’s text “In Defence of Personal Autonomy” (Quante 2011.
See also Quante 2017, 131–150). Against Tom L. Beauchamp’s arguments in favor
of the conclusion that, in biomedical ethics, we should not refer to personal auton-
omy (which qualifies the structure of a person’s life), but exclusively to action-­
autonomy (a conception of autonomyr) when we use the principle of respect for
autonomy in our ethical thinking, Quante argues that introducing a fine-grained
conception of the principle of autonomy referring to both action-autonomy and per-
sonal autonomy might be an improvement in principlism, the basic idea being “that
a person’s autonomy can express [it]self both in action-autonomy and in personal
autonomy” (Quante 2011, 597).
Quante tries to rebut the thesis that any attempt to draw on personal autonomy as
an extension of the principle of respect for autonomy leads to ethically unacceptable
consequences in the context of biomedical ethics. And he is certainly right, as “bio-
medical ethics” deal with a lot of things, including questions of human flourishing
and of the good life. There are problems in biomedical ethics which obviously touch
the dimension of personal autonomy used to qualify the structure of a person’s life
or her personal integrity, e.g. long-term care, psychiatry, neurocognitive enhance-
ment, or medical drugs, which have effects on the patient’s personality (cf. Quante
2007, 209). So Quante’s claim that “a principle of respect for autonomy that is of
use to biomedical ethics cannot manage without recourse to autonomy of agency,
but must also refer to personal autonomy or the integrity of the personality” (Quante
2017, 132) is certainly true. The question is how the division of labor between
autonomy of agency (autonomyr) and personal autonomy (autonomyi) is conceived.
There seem to be only two options. Concepts of personal autonomy, including
Quante’s approach, which try to play on the field of action autonomy (or of auton-
omy as right, as we preferably call it) either address questions of competence (which
is the basis for both concepts of autonomyr/i), in which case they are relatively triv-
ial, or they really conflate autonomyi as an ideal, gradual, and hence at least

17
The justified claim that respect for agents whose autonomy is impaired would, as a matter of
social justice, entail positive obligations of social institutions to promote the autonomy of these
citizens by fostering the social conditions for autonomy (Mackenzie 2008; Honneth and Anderson
2005) would not change anything regarding the two class-system of persons that is created until
full justice or recognition in this sense is achieved. Mackenzie’s approach does not overcome the
correlation of protection and incapacitation that is typical of substantive notions of personal
autonomy.
128 T. Gutmann

moderately perfectionist notion, and autonomyr. In the latter case, we should expect
their demands to prove inadequate or dysfunctional. Michael Quante’s approach
seems to oscillate between these two possibilities. At least in some refined sense, he
claims that “personal autonomy is a necessary condition of autonomy” (Quante
2011, 598) (meaning respect for autonomy, i.e. autonomyr in the latter case). His
example is that consent to become a living organ donor given by a person who is
completely incapable of critically reflecting upon her first-order attitude, which was
induced in her heteronomously through socialization without ever having been
made explicit or endorsed by herself, does not deserve respect. If we take this
description seriously, such a choice does indeed not deserve respect because, by all
meaningful definitions, this person is not competent to consent. Insofar, and only
insofar it refers to basic competence, Quante’s thesis is convincing that “in biomedi-
cal ethics the respect paid factually to autonomy not only refers to autonomy of
agency, but also and always to personal autonomy” (Quante 2017, 140). Referring
to basic cognitive and volitional competence (regarding rationality, self-control, and
the absence of certain mental pathologies), the capacities on which the concept of
autonomyr (autonomy of agency) draws are indeed not independent of the capacities
that constitute personal autonomy (autonomyi) (Quante 2017, 142). In short: the
concepts of personal autonomy (autonomyi) and autonomy of agency (autonomyr)
overlap only insofar as being a “personality with the constitutive capacity of acting
autonomously” (Quante 2017, 149) belongs to an adequate threshold definition of
basic competence. Quante, however, will not want to entertain a conception of
autonomyi that demands so little. On the other hand, if Quante were to say that the
fact in itself that a person is acting upon inexplicit, unreflected motives internalized
in her upbringing—or that, if asked, she would not be effectually able to reflect criti-
cally upon her first-order preferences, desires, and wishes (Quante 2011, 598)—is
sufficient to deny her to make use of her rights (and demand respect in this regard),
this would seem to be just another example of the liberty-limiting consequences of
the concept of personal autonomy when it is playing on the wrong field.
So there are reasons to assume that, when it comes to “respect for autonomy” or
to “respect for persons” guaranteed by individual autonomy rights expressive of
one’s sovereignty over oneself (autonomyr), as e.g. in the concept of informed con-
sent, the concept of personal autonomy remains misleading and a latent threat to the
concept of autonomyr. Quante’s remark that “action autonomy [...] and personal
autonomy can come in degrees” (Quante 2011, 599) demonstrates that he cannot, or
will not, in the final analysis, conceive autonomyr as the binary ontic status of being
a person who can (or cannot) claim a right to self-determination.
Quante concludes that action-autonomy and personal autonomy often “go hand
in hand, but in some situations and in some types of case both can diverge (this is
why these cases are so puzzling)” (Quante 2011, 599). The answer is that both only
go hand in hand as long as they refer to their common basic element, which is com-
petence. In other cases, they have to diverge because they have very different func-
tions within different normative settings, and there is nothing puzzling about that.
8 Is “Autonomy Talk” Misleading? 129

8.6 The Form of Autonomy Rights

The reason for distinguishing autonomyi from autonomyr is not primarily that
autonomy also functions as a political concept, as “the status marker of an indepen-
dent citizen whose perspective and value orientation get a hearing in the democratic
processes that constitute legitimate social policy” (Christman 2004, 156 et seq.; id.
2009, 219–244). The first reason is the inadequacy of all moral concepts of auton-
omy that cannot account for the respect claimed by autonomous agents, their claim
for acknowledgement of their equal moral right to hold views, to make choices, and
to take actions based on their values and beliefs.
Of course, there is some awareness for the difference between autonomyi and
autonomyr in the autonomy debate. Joel Feinberg has analysed four different mean-
ings of “autonomy” in moral and legal philosophy, among them a personal ideal, on
the one hand, and a set of rights expressive of one’s sovereignty over oneself, on the
other. John Christman stresses that “we must keep separate the idea of basic auton-
omy, the minimal status of being responsible, independent and able to speak for
oneself, from ideal autonomy, an achievement that serves as a goal to which we
might aspire and according to which a person is maximally authentic and free of
manipulative, self-distorting influences. Any plausible conceptualization of basic
autonomy must, among other things, imply that most adults who are not suffering
from debilitating pathologies or are under oppressive and constricting conditions
will be autonomous” (Christman 2015, Sub 1.1.). And as already discussed, Michael
Quante, Tom Beauchamp, and others divide the principle of respect for autonomy
into two sub-principles: personal autonomy and action-autonomy. The notion of
‘action autonomy’, however, is too weak and too fuzzy to explain the difference we
are looking for (especially as long as we follow the premise that both “action auton-
omy [...] and personal autonomy can come in degrees” (Quante 2011, 599).
At a closer look, only a certain form of protecting autonomy, i.e. protecting
autonomy by granting (moral or legal) rights to self-determination, can explain
some of our most important normative institutions and social practices. The reason
for this is the normative structure of rights as specialized normative devices (Sumner
2000, 288, 297), i.e. the distinctive kind of normative work that rights are best
equipped to do: guaranteeing “protected options to act” (Kamm 2004, 476) by con-
ferring claims (with corresponding duties for other persons), liberties (with corre-
sponding duties for other persons not to prevent the right-holder from acting within
these liberties), and normative powers to individuals. A right is a legally (or mor-
ally) “respected choice” (Hart 1982, 189 et seq.). For this task, there is no functional
alternative to individual rights. Rights to self-determination and personal liberties
are not only “preconditions for self-determination” (Mackenzie 2014), but its main
form of realization. People become agents of their lives by making choices that are
protected and respected.
In light of our practices, rights to self-determination seem to cover a lot of
ground, although they do not occupy the whole moral landscape and certainly do
not cover all the dimensions of autonomy and the evaluative power of the notion of
130 T. Gutmann

autonomy. Two final examples have to suffice. The requirement for informed con-
sent and the notion of consent in general can only be justified by the principle of
respect for autonomy if autonomy is conceived as a right (the bearer of the right,
due to a certain normative power conferred by his right, being able to abrogate the
corresponding duty of another person not to impair the bodily or personal integrity
of the bearer of the right). A right of this sort does not come “in degrees”. At the
same time, autonomyr is the main source of our obligations. Similar to the notion of
consent, the notions of promising and of contract—the possibility of binding one-
self and/or another person, thus changing the nexus of corresponding duties and
enhancing and limiting ‘action autonomy’ at the same time—can also only be
explained by the concept of (making use of) a right to self-determination (in the
form of a normative power). Again, a right of this sort does not come “in degrees”.
Consequently, the normative problems dealing with the fact that persons exist
through time, and that certain limits on autonomous self-binding for the sake of
future autonomy are thus required, can only be grasped and analyzed by distin-
guishing between autonomyr (including the right to bind oneself voluntarily, here
and now, for the future) and autonomyi (the value of being one’s own person and
living one’s life as a narrative unity according to reasons and motives that are one’s
own, at every point in time).
It is the form and intricate anatomy of individual liberty rights (which, in general,
do not require any specific content for the choices that the bearer of the right is
allowed to make within her right) that calls for the very structure of autonomyr, and
vice versa. So the difference between autonomyr and autonomyi is not one between
basic and ideal autonomy (Christman 2009, Sub 1.1.),18 it is not a distinction
between “metaphysically autonomous” persons and “morally autonomous persons”
and not about “relaxing” the standards of higher-order reflection (Beauchamp 2005,
320, 325) until they are low enough to allow for widespread autnomyr. Neither can
it be sufficient in cases of conflict between autonomyr and autonomyi, to assume that
the person opting for over-ruling the former in terms of the latter has the burden of
proof (Quante 2011, 599). It is rather that we must clearly differentiate between (a)
“personal autonomy” (autonomyi), designating a family of ideal, gradual, and hence
at least moderately perfectionist notions, on the one hand, and (b) the protection of
individual decisions and actions by upholding of rights to self-determination or
“authority” (autonomyr). This means that all the quantitative language about one
concept of autonomy that comes in “thinner” or “thicker” forms, in a “minimum” or
a “full” meaning, in a “basic” or in an “ideal”, in a “downgraded” or a “comprehen-
sive” version, is misleading. It seems especially misleading to suggest that, in order

18
In the final analysis the difference between the mildly perfectionist authenticity criterion e.g.
Christman (2004, 2009) suggests on the one hand, and the strong perfectionist view of human
values (i.e, the view that values and moral principles can be valid for a person independent of her
judgment of those values and principles) that is imported into the account of autonomy by some
relational concepts of autonomy on the other hand (see Christman 2004, 146) is also not crucial in
this regard. Both conceptions see autonomy as a matter of degree and do not clearly take into
account the normative structure of rights that defines the concept of autonomyr as an ontic status.
8 Is “Autonomy Talk” Misleading? 131

to reach an adequate notion of action autonomy (meaning a concept of what it takes

to make use of a sovereignty right), we just have to “weaken” the conditions of our
theory of personal autonomy that are “too demanding” (Quante 2011, 598) and
make them more “realistic” to get this job done. Instead, we are dealing with two
different normative concepts with different functions, and we cannot grasp the one
by making the other “thicker” or “thinner”, or “stronger” or “weaker”, respectively.
For clarity, we will ultimately need different words for divergent concepts.
Although the term “autonomy”, originally meaning “the having or making of one’s
own laws”, is close to the semantic field of “self-determination” and “sovereignty”
(Feinberg 1986, 27, 47 et seq., 52 et seq. “personal sovereignty”), we suggest that
“autonomy as (moral or legal) right” should no longer be called “autonomy”, but
forthrightly the “right to self-determination” or, in accordance with Thomas Hobbes,
“authority”.19

References

Anderson, Joel. 2004. Regimes of Autonomy. Ethical Theory and Moral Practice 17: 355–368.
Beauchamp, Tom L. 2005. Who Deserves Autonomy, and Whose Autonomy Deserves Respect?
In Personal Autonomy: New Essays on Personal Autonomy and Its Role in Contemporary
Moal Philosophy, ed. James Stacey Taylor, 310–329. Cambridge/New York: Cambridge
University Press.
Beauchamp, Tom L., and James F. Childress. 2013. Principles of Biomedical Ethics. 7th ed.
New York: Oxford University Press.
Benn, Stanley L. 1976. Freedom, Autonomy, and the Concept of a Person. Proceedings of the
Aristotelian Society 86: 109–130.
———. 1988. A Theory of Freedom. Cambridge: Cambridge University Press.
Berofsky, Bernard. 1995. Liberation from Self: A Theory of Personal Autonomy. New York:
Cambridge University Press.
Bratman, Michael E. 1987. Intention, Plans, and Practical Reason. Cambridge: Harvard
University Press.
———. 2007. Planning Agency, Autonomous Agency. In Structures of Agency: Essays, id,
195–221. Oxford/New York: Oxford University Press.
Christman, John. 2004. Relational Autonomy, Liberal Individualism and the Social Constitution of
Selves. Philosophical Studies 117: 143–164.
———. 2009. The Politics of Persons: Individual Autonomy and Socio-historical Selves.
Cambridge: Cambridge University Press.
———. 2014. Relational Autonomy and the Social Dynamics of Paternalism. Ethical Theory and
Moral Practice 17: 369–382.
———. 2015, Autonomy in Moral and Political Philosophy. The Stanford Encyclopedia of
Philosophy, ed. Edward N. Zalta. https://plato.stanford.edu/archives/spr2015/entries/
autonomy-­moral/. Accessed 4 Apr 2020.
Darwall, Stephen. 1977. Two Kinds of Respect. Ethics 88: 36–49.
Dworkin, Ronald. 1977. Taking Rights Seriously. Cambridge: Harvard University Press.
Dworkin, Gerald. 1988. The Theory and Practice of Autonomy. Cambridge: Cambridge
University Press.

19
“[T]he right of doing any action is called authority”, Hobbes 1651, Ch. XVI.
132 T. Gutmann

Elster, Jon. 1983. Sour Grapes: Studies in the Subversion of Rationality. Cambridge: Cambridge
University Press.
Feinberg, Joel. 1970. The Nature and Value of Rights. The Journal of Value Inquiry 4: 243–260.
———. 1986. Harm to Self The Moral Limits of The Criminal Law. New York: Oxford
University Press.
Freud, Sigmund. 1972a. Das Ich und das Es. [1923]. In Studienausgabe Band III, 273–330.
Frankfurt a.M.: Fischer.
———. 1972b. Neue Folge der Vorlesungen zur Einführung in die Psychoanalyse [1932]. In
Studienausgabe Band I, 448–608. Frankfurt a.M.: Fischer.
Habermas, Jürgen. 1996. Between Facts and Norms: Contributions to a Discourse Theory of Law
and Democracy. Cambridge, MA: The MIT Press.
Hart, Herbert L.A. 1982. Legal Rights. In Essays on Bentham: Studies in Jurisprudence and
Political Theory, 162–193. Oxford: Oxford University Press.
Hobbes, Thomas. 1651. Leviathan or The Matter, Forme and Power of a Common Wealth
Ecclesiasticall and Civil. London: Andrew Crooke.
Honneth, Axel. 1996. The Struggle for Recognition: The Moral Grammar of Social Conflict.
Cambridge: The MIT Press.
Honneth, Axel, and Joel Anderson. 2005. Autonomy, Vulnerability, Recognition and Justice.
In Autonomy and the Challenges to Liberalism: New Essays, ed. John Christman and Joel
Anderson, 127–149. New York: Cambridge University Press.
Kamm, Frances M. 2004. Rights. In The Oxford Handbook of Jurisprudence and Philosophy of
Law, ed. Jules L. Coleman and Scott J. Shapiro, 476–513. Oxford: Oxford University Press.
Mackenzie, Catriona. 2000. Imagining Oneself Otherwise. In Relational Autonomy: Feminist
Perspectives on Autonomy, Agency and the Social Self, ed. Natalie Stoljar, 124–150. New York:
Oxford University Press.
———. 2008. Relational Autonomy, Normative Authority and Perfectionism. Journal of Social
Philosophy 39: 512–533.
———. 2014. Three Dimensions of Autonomy: A Relational Analysis. In Autonomy, Oppression
and Gender, ed. Andrea Veltman and Mark Piper, 15–41. Oxford: Oxford University Press.
Mackenzie, Catriona, and Natalie Stoljar. 2000a. Relational Autonomy: Feminist Perspectives on
Autonomy, Agency and the Social Self. New York: Oxford University Press.
———. 2000b. Introduction: Autonomy Refigured. In Relational Autonomy: Feminist Perspectives
on Autonomy, Agency and the Social Self, 3–31. New York: Oxford University Press.
Meyers, Diana Tietjens. 1989. Self, Society and Personal Choice. New York: Columbia
University Press.
———. 2000. Intersectional Identity and the Authentic Self: Opposites Attract! In Relational
Autonomy: Feminist Perspectives on Autonomy, Agency and the Social Self, ed. Catriona
Mackenzie and Natalie Stoljar, 151–180. New York: Oxford University Press.
———. 2004. Personal Autonomy and the Paradox of Feminine Socialization. In Being Yourself:
Essays on Identity, Action and Social Life, ed. Diana Tietjens Meyers, 3–11. Lanham/Boulder:
Rowman and Littlefield.
Mill, John Stuart. 1992. On Liberty [1859]. In On Liberty and Utilitarianism, with an introduction
by Isaiah Berlin. New York: Alfred A. Knopf.
Oshana, Marina. 1998. Personal Autonomy and Society. Journal of Social Philosophy 29: 81–102.
———. 2006a. How Much Should We Value Autonomy. Social Philosophy and Policy 20: 99–126.
———. 2006b. Personal Autonomy in Society. Hampshire: Ashgate.
———. 2014. Personal Autonomy and Social Oppression. London: Routledge.
Quante, Michael. 2007. Autonomy for Real People. In Intentionality, Deliberation and Autonomy:
The Action-Theoretic Basis of Practical Philosophy, ed. Christoph Lumer and Sandro Nannini,
209–226. Aldershot: Asgate.
———. 2011. In Defence of Personal Autonomy. Journal of Medical Ethics 37: 597–600.
———. 2017. Personal Identity as a Principle of Biomedical Ethics. Cham: Springer.
Rawls, John. 1971. A Theory of Justice. Cambridge: The Belknap Press of Harvard University Press.
8 Is “Autonomy Talk” Misleading? 133

———. 1988. The Priority of Right and Ideas of the Good. Philosophy and Public Affairs 17:
251–276.
———. 2001. Justice as Fairness: A Restatement. Cambridge: Harvard University Press.
Raz, Joseph. 1986. The Morality of Freedom. Oxford: Oxford University Press.
Steiner, Hillel. 1994. An Essay on Rights. Oxford: Oxford University Press.
Stoljar, Natalie. 2000. Autonomy and the Feminist Intuition. In Relational Autonomy: Feminist
Perspectives on Autonomy, Agency and the Social Self, ed. Catriona Mackenzie and Natalie
Stoljar, 94–111. New York: Oxford University Press.
Sumner, Leonard W. 2000. Rights. In The Blackwell Guide to Ethical Theory, ed. Hugh LaFollette,
354–373. Oxford: Wiley-Blackwell.
Taylor, Charles. 1985a. Self-interpreting Animals. In Philosophical Papers, Vol. 1, 45–76.
Cambridge: Cambridge University Press.
———. 1985b. What’s wrong with negative liberty. In Philosophical Papers, Vol. 2, 211–229.
Cambridge: Cambridge University Press.
Thomson, Judith Jarvis. 1990. The Realm of Rights. Cambridge: Harvard University Press.
VanDeVeer, Donald. 1986. Paternalistic Interventions: The Moral Bounds on Benevolence.
Princeton: Princeton University Press.
von Savigny, Friedrich Carl. 1867. System of the Modern Roman Law, Vol 1
Waldron, Jeremy. 1981. A Right to Do Wrong. Ethics 92: 21–39.
———. 2007. Rights. In A Companion to Contemporary Political Philosophy, ed. Robert
E. Goodi, Philip Pettit, and Thomas W. Pogge, 745–754. Oxford: Wiley-Blackwell.
Wellmann, Carl. 1985. A Theory of Rights. Oxford: Oxford University Press.
Wertheimer, Alan. 1987. Coercion. Princeton: Princeton University Press.
Wikler, Daniel. 1983. Paternalism and the Mildly Retarded. In Paternalism, ed. Rolf Sartorius,
83–94. Minneapolis: University of Minnesota Press.
Williams, Bernard. 1981. Persons, Character and Morality. In Moral Luck, 1–19. Cambridge:
Cambridge University Press.
Chapter 9
Respecting Personal Autonomy
in Bioethics: Relational Autonomy
as a Corrective?

James F. Childress

Abstract Focusing mainly on respect for autonomy, particularly autonomous

choices and actions in bioethical decisions, I examine several complexities of enact-
ing this respect through the case of a fourteen-year-old boy who died after being
allowed to refuse a necessary blood transfusion on religious grounds (Jehovah’s
Witness). I argue that thicker concepts of autonomy, closely connected with rela-
tional autonomy, direct our attention to aspects of respect for autonomy that are
often neglected or underappreciated in much bioethical theory and practice. In par-
ticular, they illuminate the interpretive complexity of respecting autonomous
choices by focusing attention on the relational context of individuals’ decisions, on
respect for individuals’ social-cultural beliefs and values, and on the temporality of
selves, as expressed in their past consent or future consent as well as in—and some-
times in contradiction to—their present consent or dissent. However, it is important
not to convert the illuminating and valuable insights from theories of relational
autonomy into necessary conditions of autonomous choices and actions and hence
of respect for autonomy.

9.1  risis of Autonomy and Respect for Autonomy

C
in Bioethics

From its beginnings in the late 1960s and early 1970s, bioethics has stressed respect
for the autonomy of patients and participants in research. This is not surprising—
after all, bioethics emerged in part in reaction to centuries of medical paternalism
and exploitation in research, and it was infected with a strongly individualistic ori-
entation, particularly in the U.S., the site and context of much of the early work in
the field.

J. F. Childress (*)
University of Virginia, Charlottesville, VA, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 135

Switzerland AG 2022
J. F. Childress, M. Quante (eds.), Thick (Concepts of) Autonomy, Philosophical
Studies Series 146, https://doi.org/10.1007/978-3-030-80991-1_9
136 J. F. Childress

Now, however, autonomy and respect for autonomy are under attack—at least
certain conceptions are under siege. Critics charge that proponents of autonomy—
and many policies and practices in contemporary medicine, health care, and research
that are supposedly based on respect for autonomy—make too much of autonomy,
extend autonomy too far, and put too much weight on autonomy over against other
ethical principles, such as benefitting the patient or protecting the society.
Critics also contend that much work in bioethics does not adequately character-
ize and interpret autonomous choices and that it thus distorts the processes of con-
sent and refusal. From this standpoint, many conceptions of autonomous choice and
consent/ refusal—perhaps the dominant ones—are seen as
• too individualistic, in that they neglect the social context and relational aspects
of autonomy;
• too rationalistic, because they ignore the non-rational and emotional aspects of
autonomous choices;
• too legalistic, insofar as proponents devote too much attention to laws, regula-
tions, and institutional rules rather than to social practices;
• too formalistic, in view of the excessive attention paid to consent forms over
against the process of consent;
• too thin, a common critique of several of these putative deficiencies.
I will examine some key aspects of the standard thin conception of autonomy and
respect for autonomy that have come under attack and then explore what one impor-
tant thicker conception, that of relational autonomy, might add. I will argue that this
thicker conception helpfully identifies several aspects of the development and exer-
cise of personal autonomy that further complicate the practice of respect for auton-
omy. As illuminating and valuable as these insights are, it is important not to convert
them into necessary conditions of autonomous choices and actions and of respect
for autonomy.

9.2 The Principle of Respect for Autonomy (PRA)

What does it mean to respect persons’ autonomous choices and actions? This cum-
bersome formulation is more accurate and more adequate than the abstract formula-
tion “respect for autonomy,” and it’s what I mean in this chapter when I use the
phrase respect for autonomy.
What is involved in respecting autonomy, in the sense of autonomous choices
and actions? “Respect” is more than an attitude; it involves actions. Some respectful
actions can be stated negatively: not interfering with or attempting to interfere with
the autonomous choices and actions of others, and not subjecting or trying to sub-
ject those choices to controlling influences.
Two major types of controlling influence are manipulation of information and
coercion. (Beauchamp and Childress, 2019, Chap. 4) When persuasion fails, these
two types of intervention are sometimes used in efforts to control others’ actions.
9 Respecting Personal Autonomy in Bioethics: Relational Autonomy as a Corrective? 137

Manipulation of information may occur through non-disclosure or deception, while

coercion involves a threat of some kind. These interventions are incompatible with
respect for autonomous choices and actions. However, it is not inconsistent to say
that we respect another person’s autonomous choices and actions by non-­interference
while, at the same time, we try to persuade that person to act differently. It is not
even inconsistent to use modes of influence, such as modest incentives and defaults,
which shape action but do not control it. (Thaler and Sunstein 2008).
The first formulation of the PRA is negative. My version slightly modifies one
offered by Ruth Faden and Tom Beauchamp (1986): “it is prima facie wrong to
subject or attempt to subject the actions (including the choices) of others to control-
ling influences.” The modifications include the addition of “prima facie” and “or
attempt to subject” and “including the choices.” This negative principle is expressed
and specified in various concrete rights or duties—for example, the right to consent
to or refuse medical treatments and the right to liberty of action.
Neither the PRA nor the several specified rights and duties can properly be
viewed as absolute. They are at most prima facie binding—that is, they are binding,
obligatory, other things being equal, and are sometimes justifiably overridden by
other principles. An example is protecting public health in an epidemic of serious
communicable disease through forcible quarantine against the autonomous choices
and actions of recalcitrant citizens.
The PRA is not only negative. It also has clear positive implications in several
contexts. In research, medicine, and health care, for instance—where there are fidu-
ciary relations that involve significant differences in power and knowledge, such as
the doctor-patient or researcher-subject relationship—the PRA engenders a positive
or affirmative obligation to disclose information and to create the conditions for
autonomous choices and actions, for instance, by promoting communication and
comprehension. As a result, the PRA is reflected in rules of informed consent and
refusal but it is not exhausted by or limited to those rules.
I will explore autonomous choices largely through their expression in consents
and refusals. Autonomous choices are often made by autonomous persons, but they
may also be made in particular situations by persons who would not be described
generally or globally as autonomous persons.

9.3  omplexities in Respecting Autonomous Choices:

C
A Difficult Case

Even if we accept the PRA, it is neither easy nor simple to act on it. Far from being
a mechanical task, respecting autonomous choices and actions is a difficult interpre-
tive task, involving, for instance, judgments of competence or capacity, communica-
tion (disclosure of information, comprehension, and response), and voluntariness.
This interpretive task entails attentive listening, imagination, discernment, and the
like, and these are particularly important in determining what the patient or research
138 J. F. Childress

subject is really saying. The following case provides a useful point of reference as
we examine a few of the complexities of respecting persons’ autonomous choices:

9.3.1 Case: Death of 14-Year-Old Jehovah’s Witness

On November 28, 2007, Dennis Lindberg, age fourteen, died of the effects of acute
lymphocytic leukemia and the treatments he was receiving after he refused, on reli-
gious grounds (Jehovah’s Witness), a blood transfusion that, according to the physi-
cians in this case, could have saved his life and allowed him to live for a few years.
His refusal decision was approved by Superior Court Judge John Meyer, who denied
a motion filed by the state of Washington to force Dennis to undergo the blood
transfusion.
Lindberg was diagnosed with acute lymphocytic leukemia November 8, 2007,
and chemotherapy was started but then was stopped because his red blood cell count
dropped too low—a common side effect of chemotherapy. Hence the need for the
blood transfusion. Medical reports provided to the judge indicated that, with the
transfusion and other treatments, Dennis had a 70 percent chance of surviving the
next five years.
This case was complicated by Dennis Lindberg’s family circumstances. His bio-
logical parents wanted him to receive the transfusion, but they did not have legal
custody of their son. The boy’s legal guardian at the time of his death was his aunt,
Dianna Mincin, with whom he had lived for four years, after his father, now a recov-
ering addict, was jailed for drug possession. (According to some reports, both bio-
logical parents had been addicted to methamphetamine.) Ms. Mincin, who is a
Jehovah’s Witness, supported Dennis’s decision to refuse a blood transfusion.
Jehovah’s Witnesses do not want to die in such cases—therefore, this was not a
right-to-die case. Indeed, generally Jehovah’s Witnesses accept most medical proce-
dures and treatments, but reject blood transfusions for religious reasons.
Hospital administrators reported this case to the state, which then sought a court
order to force the transfusion. Courts in the U.S. recognize the rights of adults to
refuse potentially life-saving blood transfusions as well as other treatments on reli-
gious grounds, but they generally order blood transfusions when medically indi-
cated for children even against parental objections. In reaching his decision, Judge
Meyer heard from Dennis’s parents, his aunt, social workers, and doctors. Dennis’s
doctors reportedly supported his decision. Judge Meyer determined that the eighth-­
grader was a mature minor, who understood that “he’s basically giving himself a
death sentence.” The judge also said: “I don’t think Dennis is trying to commit sui-
cide. This isn’t something Dennis just came upon, and he believes with the transfu-
sion he would be unclean and unworthy.” Finally, according to observers of the
court proceedings, the judge indicated that he didn’t “believe Dennis’s decision is
the result of any coercion. He is mature and understands the consequences of his
decision.”
9 Respecting Personal Autonomy in Bioethics: Relational Autonomy as a Corrective? 139

Some critics of the judge’s decision insisted that a fourteen-year-old boy is not
competent to make such a life/death decision. According to one critic, adolescents
are both more likely to act in ways that seem heroic and very susceptible to influence
and pressure. Moreover, they often change their beliefs as they develop. Other critics
wondered whether the teenager was experiencing pressure from his immediate care-
giver and religious community. In the crisis, his biological parents came from Idaho,
where they lived, to indicate their desire for their son to have the transfusion, and to
raise questions about whether their son was unduly influenced by his aunt. The father
stated: “My sister has done a good job of raising him for the past four years, but her
religious beliefs shouldn’t be imposed on my son. I want my son to live.”
Dr. Benjamin Wilfond, who directs pediatric bioethics at Children’s Hospital in
Seattle but was not directly involved in this case, focused on several complexities.
Dennis would have needed continuing treatments and, with them, repeated transfu-
sions, but it would have been very difficult to force him to cooperate with the treat-
ments over time. As clinicians try to balance protection of children with respect for
their autonomous choices, Wilfond observes, “it’s not typical to physically restrain
children of that age for medical intervention.” Dr. Douglas Diekema, another pedia-
trician and ethicist at Children’s Hospital, echoed similar themes—but neither of
them could talk specifically abouzzt the details of the Lindberg case. Caregivers are
“nervous” in such cases because fourteen-year-olds can have an “adultlike” decision-­
making process, even though they often change their mind as they get older. Diekema
too noted the difficulty of “effectively treat[ing] a kid when he’s not going to
cooperate.”
In my mind, if there is a role of the court it would be to test a 14-year-old and see just how
intense he is about his decision. My approach would be to push it a little further. If he fights
you physically, then I’d respect that. But also, are you willing to tie him down every time he
needs a transfusion knowing he’ll need treatment for the next three years? You’ll have a
hard time finding a provider willing to do that.

The natural response, Wilfond noted, is to want to do what is necessary to save the
young boy: “But imagine if you heard a story about how this individual was strapped
down to be given a transfusion, or tackled as he tried to leave the hospital. At four-
teen, if you feel strongly about something, you’re going to fight back.” In contrast
to dealing with very young children, caregivers feel “profoundly conflicted” when
adolescents are at the center of such cases.
Dianna Mincin’s blog (subsequently deactivated) included this comment a few
days before her nephew’s death:
For those reading (about) this journey our family has been on that are not one of Jehovah’s
Witnesses, we compassionately understand your confusion and, perhaps, even your anger
at the decision that Dennis and his family have made….While we empathize with your
strong feelings, we ask that you attempt to respect Dennis’ fight for what he and his family
believe so strongly in.1

1
This case was constructed from the following media accounts: Michels 2007; Black 2007a, b;
Ostrom 2007.
140 J. F. Childress

9.3.2 Steps in Respecting Autonomous Choices

A first step in respecting persons’ autonomous choices is to determine whether they

have the capacity to act autonomously. This capacity is often referred to as “compe-
tence,” and I’ll use “competence” and “capacity” interchangeably here, although in
the United States “competence” is often used in legal contexts. I am viewing the
activity and process of determining capacity/competence in ethical rather than legal
terms, but sometimes it will be necessary to obtain a legal determination of
competence.
Competence or capacity is task specific. We ask whether people have the capac-
ity to do X, Y, or Z. They may lack the capacity to do X but have the capacity to do
Y—for example, they may have the capacity to drive an automobile but not to make
complex transfers of property. In the case of Dennis Lindberg, the judge determined
that, even though he was only fourteen years old, he was a “mature minor,” and that
he had the capacity to refuse treatments that could prolong his life for perhaps five
years (even though, legally, he lacked the competence to obtain a license to drive an
automobile). Commentators note that fourteen year olds can have an “adultlike”
decision-making process.
Suppose a person lacks competence to give valid consent at a particular time. It
is sometimes possible through various interventions to build up or restore a person’s
capacity for informed consent/refusal. For adults, in contrast to a fourteen-year-old
making a life/death decision, we proceed from a presumption of competence and
then determine when to disqualify him or her on grounds of limited capacity to
make such decisions. Here the fourteen-year-old was presumed to be incompetent,
and he had to be evaluated by health care professionals and then by a judge to deter-
mine whether he was sufficiently competent to refuse treatment. In practice, ques-
tions rarely arise about an adult patient’s capacity to decide unless he or she
disagrees with what the clinician proposes. Probing an individual’s capacity to
decide or his/her understanding and reasons for decisions in routine health care
would be unnecessary, time-consuming, and very costly, apart from a conviction the
decision in question is objectively unreasonable or irrational. (Maclean 2009, 248)
By contrast, closer attention to evidence about the adult’s level of capacity for
autonomous choice is warranted when the diagnostic or therapeutic procedures are
more invasive or risky.
For a decision to be autonomous the agent needs an adequate understanding of
his or her situation, the available options, the probable consequences of different
decisions, and so forth. The judge in the Jehovah’s Witness case determined that the
teenager understood the implications of his decision to refuse a blood transfusion—
that he was “basically giving himself a death sentence” through acting on his beliefs.
However, some critics observe that fourteen-year-olds often view themselves as
heroic and, furthermore, often change their minds about some of their earlier firm
convictions as they mature.
Next it is necessary to determine whether a patient’s choice is voluntary or made
under undue or controlling influence. In the case of Dennis Lindberg, questions
9 Respecting Personal Autonomy in Bioethics: Relational Autonomy as a Corrective? 141

were raised about the voluntariness of his choice because of concerns that he was
under undue influence or even coercion. His legal guardian was his aunt, a Jehovah’s
Witness herself, who in a blog asked for respect for “Dennis’s fight for what he and
his family believe so strongly in.” However, Dennis’s biological parents opposed his
aunt’s imposition of her religious beliefs on their son; the inculcation of those
beliefs was probably much more subtle and effective than “imposition” would have
been for a teenager, who probably would have resisted “imposition.” In any event,
the judge did not “believe Dennis’ decision [was] the result of any coercion.”
Finally, it can be difficult to discern what even clearly competent patients really
want. That is, what is the actual content of their choices? Even supposing the
Jehovah’s Witness teenager was competent, understood his situation and the conse-
quences of his decision, and was acting voluntarily, it still could have been difficult
to determine exactly what he wanted in the circumstances as they unfolded. It would
have been surprising if he had not been ambivalent. One pediatrician-ethicist rec-
ommended pushing the interaction further to test “just how intense he is about his
decision,” to determine whether he would “fight” it. And this pediatrician-ethicist
expressed his hope that the judge had done this sort of probing. It is also important
to note that sometimes practitioners of the Jehovah’s Witness faith appear to give
both health care professionals and judges conflicting signals about what they want.
I’ll return to this later when I discuss cultural beliefs and values.

9.4 Relational Autonomy: A Thicker, Richer Conception

9.4.1 Components of Relational Autonomy

Relational autonomy—also called “autonomy in a relational context,” “situated

autonomy,” “contextual autonomy,” and so forth—offers an important corrective to
some interpretations of the PRA. It directs attention beyond the immediate decision,
“the moment of medical decision making” (Sherwin 1998, 32), to the broader con-
text of relations that shape those decisions. Proponents of “relational autonomy,”
particularly (but not only) feminist thinkers, have helpfully criticized some wide-
spread interpretations and applications of the PRA. I’ll focus on relational auton-
omy rather than on other thick conceptions of autonomy because of limited space,
the wide appeal of relational autonomy in bioethics, and its incorporation of ele-
ments of some other thick notions. Even so, this is only a brief sketch, hardly repre-
sentative of the full range of ideas to be found in various conceptions of relational
autonomy, and it will concentrate on feminist rather than communitarian concep-
tions of relational autonomy. (See, for example, the following: Mackenzie and
Stoljar, 2000; Stoljar 2014; Friedman 1997; Donchian 2001; Meyers
1997; Sherwin 1992)
Relational autonomists or relationalists find very problematic the picture of the
self that sometimes dominates conceptions of autonomy, understood as self-rule or
142 J. F. Childress

self-governance: individual, self-made, independent, self-interested, isolated and

insulated, and so forth. Of course, not all proponents of thin conceptions of auton-
omy in bioethics or elsewhere accept all these characteristics—these are at most
only tendencies. But relational autonomists offer several criticisms of these
tendencies.
Self-rule, self-governance, or self-determination does not mean self-creation.
The independence emphasized by the PRA does not or should not deny interdepen-
dence and neglect the emergence and expression of the self in relationships. For
relationists, individuals’ autonomous choices are not given, individualistically
achieved, or fixed. Instead, we need to recognize how supportive social conditions
foster autonomous action. In the words of Susan Sherwin (1998, 39), “Relational
autonomy redefines autonomy as the social project it is, but it does not deny that
autonomy ultimately resides in individuals.”
In the first instance, we think of the impact of close, intimate, interpersonal rela-
tionships, particularly families, in shaping the self in both positive and negative
ways. This was certainly evident in the case of the Jehovah’s Witness teenager—(a)
the impact of his biological parents’ drug problems and his father’s imprisonment,
and (b) the impact of his relationship with his aunt, who had become his legal guard-
ian and provided him with a home. Broader social, religious, cultural, and political
contexts may also be important. In this case, the teenager’s affiliation with a particu-
lar religious community was very significant as was his school environment. Even
though health care professionals are part of the relational context, in this particular
case we don’t know much about those relations.
In some situations, the self in self-rule or self-determination may actually be an
oppressed self. Hence, it is important to examine the conditions of constraint and
limitation under which many exercise—or try to exercise—their autonomy.
Attention to social injustice is crucially important for relational autonomists. Even
culture may be oppressive; for example, some feminists argue that decisions for
breast augmentation or even for breast reconstruction following mastectomy for
cancer are unduly influenced and perhaps even, in a way, coerced by the dominant
culture.
Furthermore, in health care, medicine, and research, patients and participants/
subjects exercise their choices in relation to and with some dependence on clini-
cians, researchers, and other professionals. This is one factor that gives rise to the
positive obligations of respect for autonomy that I mentioned. Through some of
these relations, it may be possible to engender, maintain, or restore autonomous
choices.
Several themes featured in relational autonomy now appear in more or less stan-
dard accounts of respect for autonomy, in part as a result of the extensive and power-
ful depictions and defenses of relational autonomy in the literature. In the Jehovah’s
Witness case, a relational approach would have considered carefully the overall
relational context—though perhaps no more carefully than the physicians and judge
did. Even under a thin account of autonomy much of what relationalists emphasize
would have been incorporated because the case involved a minor, a fourteen-year-­
old. The same might not have been true for many interactions with adults.
9 Respecting Personal Autonomy in Bioethics: Relational Autonomy as a Corrective? 143

I have already noted the relational context that shaped Dennis’s decision making:
parental drug problems, father’s imprisonment, home with and legal guardianship
under his aunt, a Jehovah’s Witness, for four years; religious community and school
community; his physicians and other care givers; and his biological parents who
may have arrived on the scene too late to make a difference in view of how rapidly
the crisis developed. The relational context also included the judge who probed the
teenagers’ views. As previously noted, one pediatrician-ethicist indicated that the
role of the court should have been to determine the intensity of Dennis’s beliefs—
“to push back a little further” to determine whether he “fights you.” Since the treat-
ment over time would have required repeated transfusions, there would have been
on-going fights, and health professionals opposed using physical force in order to
provide the rejected blood transfusion.
In sum, valuable insights and helpful correctives to common understandings and
misunderstandings of the PRA have emerged from conceptions of “relational auton-
omy.” Some of these correctives were already implicit in or have since been incor-
porated into the standard approach. No doubt, virtually any thin approach would
attend to relational autonomy in the care of minors, but proponents of relational
autonomy contend that it should extend to all throughout health care. However, not
all approaches to or implications and applications of relational autonomy—and
there are many different conceptions—are defensible.

9.4.2  elationally Responsible Choices: A Necessary

R
Condition of Autonomy?

One conception of “relational autonomy” is problematic because it builds a notion

of relational, moral responsibility into the conception of autonomy itself. Some
interpreters hold that a relationally autonomous choice is one that recognizes its
societal repercussions and decides and acts accordingly.2 Similar issues are raised
by “principled autonomy,” which Onora O’Neill (2002, 2003) and others, following
Kant, use to build a notion of responsibility into autonomy (hence, “principled
autonomy.”) (See also Manson and O’Neill 2007) From this standpoint, we are not
choosing autonomously if we do not attend to the impact of our actions on others.3
My approach is quite different, as developed in Principles of Biomedical Ethics
co-authored with Tom Beauchamp: people’s autonomous choices may be unethical
in failing to attend to relational responsibilities and to negative impacts on others;
hence, these autonomous choices may sometimes be justifiably overridden. That
these choices are unethical does not disqualify them as autonomous; it only means
that they are not ethically justifiable all things considered. And since, like other

2
See the depiction of relational autonomy in McLean 2010.
3
It is unclear whether and to what extent the Jehovah’s Witness teenager considered the impact of
his refusal of the blood transfusion on others.
144 J. F. Childress

ethical principles (in the Beauchamp-Childress framework), the PRA is only prima
facie binding, it too can be justifiably overridden under some circumstances. Take,
for instance, a refusal to follow an order of quarantine in a public health crisis. That
refusal could be an autonomous choice but justifiably overridden to protect others.

9.4.3  espect for an Individual’s Social-Cultural Beliefs

R
and Values

Relationalists also stress the larger social-cultural context that shapes individuals’
choices, sometimes to the extent of practically constituting their identities, not
merely functioning as a causal background. Of course, bioethics needs to attend to
various and significant social-cultural contexts of choices. This is crucially impor-
tant in multicultural societies. How we should do this requires careful thought.
Several empirical studies challenge what they take to be the dominant Western
bioethical view of respect for patient autonomy and its implications for determining
how much patients should know and decide. All in all, these studies enrich our
understanding of the PRA and its applications, but their findings require careful
interpretation, both theoretically and practically.
In one study, researchers in California studied the differences in the attitudes of
elderly subjects (65 years or older) from four different ethnic backgrounds (Korean
American, Mexican American, European American, and African American) toward
the roles the individual patient and the family should play in the disclosure of a
diagnosis and prognosis of a terminal illness, such as cancer, and in decision mak-
ing at the end of life. The study enrolled 800 subjects, 200 from each of the four
ethnic groups. Following is the researchers’ summary of their main findings:
Korean Americans (47%) and Mexican Americans (65%) were significantly less likely than
European Americans (87%) and African Americans (88%) to believe that a patient should
be told the diagnosis of metastatic cancer. Korean Americans (35%) and Mexican Americans
(48%) were less likely than African Americans (63%) and European Americans (69%) to
believe that a patient should be told of a terminal prognosis and less likely to believe that
the patient should make decisions about the use of life-supporting technology (28% and
41% vs 60% and 65%). Instead, Korean Americans and Mexican Americans tended to
believe that the family should make decisions about the use of life support. (Blackhall et al.
1995, 820–825).

This study determined that the attitudes toward disclosure and toward decision mak-
ing primarily correlated with ethnicity and secondarily with respondents’ years of
education and personal experiences of illness and medical decision making.
The researchers indicate that their study shows that the “belief in the ideal of
patient autonomy is far from universal.” (Blackhall et al. 1995, 824) They contrast
that ideal with a “family-centered model,” which emphasizes the individual’s web
of relationships, and tends to place a higher value on “the harmonious functioning
of the family than on the autonomy of its individual members.” (Blackhall et al.
9 Respecting Personal Autonomy in Bioethics: Relational Autonomy as a Corrective? 145

1995, 825) Hence, there are two divergent models of medical decision making: a
family-centered model and a patient-centered model.4
The researchers recommend that physicians “ask their patients if they wish to
receive information and make decisions or if they prefer that their families handle
such matters.” (Blackhall et al. 1995, 820) That recommendation is a way both to
recognize and to respect patients’ autonomy. Rather than abandoning the ideal of
patient autonomy, as the researchers sometimes suppose, this permits individuals to
exercise their autonomy by choosing the model they prefer for disclosure of infor-
mation and decision-making—a patient-centered or a family-centered one. Even a
patient’s choice of a family-centered model can be an autonomous choice.
In places the authors seem to admit that we don’t have to abandon our commit-
ment to individual autonomy or to (the legal doctrine of) informed consent. Instead,
they say (Blackhall et al. 1995, 825), what is required is “broadening our view of
autonomy so that respect for persons includes respect for the cultural values they
bring with them to the decision-making process” (italics added). Generally I concur
with this conception of broadening the view of respect for autonomy, in contrast to
claims of abandonment—perhaps the researchers’ earlier statements should be read
as an abandonment of excessively individualistic notions of autonomy.5 However,
my concurrence includes a strong cautionary note. It is important to avoid reading
individuals’ values off the cultural values they generally affirm, or individuals’ nar-
ratives off their cultural narratives.
Each year I teach a broad, comparative course on Religion and Medicine for
fourth-year medical students, and I caution the participants as follows: when we
learn what a particular religious or cultural tradition affirms, we cannot conclude
from this that an individual affiliated with that tradition thinks or believes or acts in
a particular way. The individual is not a mere subset of a social-cultural tradition
even if that tradition has, in part, shaped the person’s identity and the person still
identifies with it. We cannot suppose that an observant religious person necessarily
holds or acts on a belief or value generally associated with, or even authoritatively
promulgated by, his or her particular religion.
This too is important for our central case. It is natural to assume that a self-­
identified and committed Jehovah’s Witness will consistently and adamantly refuse
a blood transfusion as contaminating and polluting, based on that tradition’s formal
statements of beliefs and values. Nevertheless, some Jehovah’s Witnesses, perhaps a
small minority, actually believe that consent to the transfusion, rather than the trans-
fusion itself, contaminates the individual—no consent, no contamination! And,
according to transcripts of a number of bedside interviews conducted by judges of
adult Jehovah’s Witnesses who refuse to consent to a blood transfusion recom-
mended by physicians, these patients may appear even to welcome or invite health
care professionals and courts to help them through this loophole in their beliefs—for

4
See also the critique of what the author sees as the individualistic version of respect for autonomy
in principlism, from the perspective of Japanese culture, in Traphagan 2013.
5
See McLean (2010) for a discussion of both individualistic and relational conceptions of auton-
omy in English law.
146 J. F. Childress

some, a court-ordered (but unconsented to) transfusion would not have negative reli-
gious repercussions. (Childress 2015) Health care professionals and judges are in a
difficult ethical and legal bind when in some circumstances the adult Jehovah’s
Witness does not want to die but cannot consent to a potentially life-­saving blood
transfusion without jeopardizing his or her religious standing. Some judges have
authorized the transfusions while others have declined to do so. (Childress 2015).
The suggestion above that the Jehovah’s Witness teenager’s resolve could be
tested by seeing whether he would put up a fight should be understood in this con-
nection. As a result of his probing, perhaps the kind required by relational auton-
omy, the judge understood the teenager to believe that “with the transfusion he
would be unclean and unworthy”—even if the court ordered it without his consent
or against his refusal. But, as also indicated earlier, we actually don’t know what
kind of probing the judge used.
In summary, understanding the patient’s social-cultural context is very important
for clinicians in both disclosing information and interpreting patients’ responses.
But individuals’ appropriations and expressions of their socio-cultural traditions
and values may be quite idiosyncratic—they may even reject those traditions and
values at some key points relevant to the decision at stake. This holds whether these
social-cultural traditions and values function causally, as background conditions, or
constitutively for those individuals.

9.4.4 Respect for Temporally Extended Selves

Another source of complexity in respecting autonomy stems from the fact that we
are “temporally extended selves.” (Baumann 2008) We exist in and through time
and our choices and actions occur over time. Among the several theorists of rela-
tional autonomy who have emphasized this, Susan Sherwin (1998, 35) observes:
“Under relational theory, selfhood is seen as an ongoing process rather than as
something static or fixed.” In discharging our obligations under the PRA, we not
only have to determine whether a person is choosing autonomously and what he or
she chooses, but we also have to put that person’s current consents and dissents in a
broad temporal context with attention to the past and the future as well as the pres-
ent. As temporal beings through and through, we frequently express different views
and preferences at different times.
Not surprisingly, respecting autonomy in bioethics often focuses on the present
moment—for example, do we have the patient’s or prospective research subject’s
informed consent or refusal at this time? However, respecting persons and their
autonomous choices becomes very complex when we consider their temporality:
Which choices and actions should we respect? In particular, is it ever justifiable to
override a patient’s present choices and actions in the light of his or her past or (antic-
ipated) future choices and actions? Most dramatically, which person do we respect?
Past or Prior Consent Past or prior consent/refusal poses no problem if the patient
cannot currently autonomously express his or her wishes, perhaps because of men-
tal incapacity. As in the case of advance directives, we respect personal autonomy
9 Respecting Personal Autonomy in Bioethics: Relational Autonomy as a Corrective? 147

by acting on that past or prior statement of preference, what is sometimes called

precedent autonomy. (Davis 2007) Matters become more problematic, however,
when a person’s present choices appear to contradict those prior choices. Sometimes
the person may have even made those prior choices to prevent future changes.
For example, in one case, a 28-year-old man decided to terminate chronic renal
dialysis because of his poor quality of life and the burdens his condition and its
treatment created for his family—his medical conditions included diabetes, legal
blindness, and progressive neuropathy that rendered him unable to walk. At his
request, his wife and physician finally agreed to this arrangement: when he stopped
dialysis, he would receive medication to relieve his pain while he died, and he would
not be put back on dialysis even if he requested it. While dying in the hospital, the
patient awoke complaining of pain and asked to be put back on dialysis. The
patient’s wife and physician decided to act on the man’s earlier request that he be
allowed to die, and he died four hours later. (Beauchamp and Childress 2019,
111–112)
In my judgment, the spouse and physician should have put the patient back on
dialysis in view of his current request and the irreversibility of acting in accord with
his earlier, now repudiated, statements. After putting him back on dialysis, they
could have determined whether he had autonomously revoked his prior choice.
Suppose he said, “Damn it, I told you not to put me back on dialysis whatever I
said….” Then they could have proceeded again with greater confidence.
As this case suggests, it is sometimes critically important to determine whether
people have autonomously revoked their previous consents/refusals. In such cases,
we need to assess a person’s level of autonomy over time in order to determine
whether the apparent revocation of previous consents or dissents is autonomous.
The PRA requires that we attend to both a person’s prior consent/refusal and his or
her present revocation when there is a conflict, but the present revocation takes pri-
ority if it is autonomous.
An important interpretive perspective in respecting autonomous choices is that
of narrative unity, integrity and authenticity. (See Miller 2014) One question that
often arises is how a patient’s current choices fit in with the choices he or she has
made over time and in accord with his or her character. Narrative unity and integrity,
for example, involve wholeness and constancy over time.
Authenticity, according to one philosopher, means that “an action is consistent
with the attitudes, values, dispositions and life plans of the person.” (Miller 2014)
Its intuitive idea is “acting in character.” Learning whether a person’s present choice
or action is consistent or inconsistent with that person’s life plan and risk budget as
embodied over time may help us determine, for example, whether a revocation is
genuine. On the one hand, we may wonder whether a person’s actions are autono-
mous if they are out of character (e.g., a person’s sudden and unexpected decision
to discontinue a needed medical treatment). On the other hand, we are less likely to
challenge actions as non-autonomous if they are in character (e.g., a committed
Jehovah’s Witness’s refusal of a blood transfusion).
148 J. F. Childress

These ideas of narrative unity, integrity, and authenticity are important, but it
would be a mistake to make them criteria for determining whether choices are
autonomous. At most, actions apparently out of character and inauthentic raise cau-
tion flags that warn others to press for explanations and justifications in order to
determine whether those actions are in fact autonomous. However, we cannot rule
out in advance the possibility of a change or even a conversion in basic values.
Future or Deferred Consent Suppose a patient refuses a potentially life-sustaining
treatment because of pain, suffering, and apparently limited life prospects. In some
such situations clinicians may have good reason to believe that if the patient were
kept alive by a particular treatment that she is now refusing, she would eventually
ratify the coercive or deceptive treatment on her behalf, perhaps even thanking the
professional. Such retrospective ratification does occur in some cases.
But the question is whether anticipated future ratification justifies—or helps to
justify—acting now against a patient’s current express choices in order to respect
who that person will become rather than who she currently is. I would argue that
predicted future ratification is neither necessary nor sufficient to justify paternalistic
interventions against current choices. At most, a patient’s probable future ratifica-
tion may provide supporting (but still inconclusive) evidence that the conditions for
justified paternalistic interventions have been met. It is also important to note that
the intervention may have changed the person in the process.

9.5 Conclusions

I have argued that thicker concepts of autonomy, closely connected with relational
autonomy, can enable us—and, indeed, have enabled us—to see more clearly
aspects of respect for autonomy that have often been neglected or downplayed in
much bioethical theory and practice. In the process, they help us understand better
the interpretive complexity of the PRA. This complexity is unavoidable because
people—the very people whose autonomy we are to respect—are complex, as indi-
cated by conceptions of relational autonomy.
A danger is that these valuable indicators for respecting autonomous choices
might be made into necessary conditions for autonomy, conditions that would then
ground respect for autonomy. This would set a standard for autonomous actions that
is too high and too demanding and would thus deprive ordinary decision-makers of
the protection and support needed for their autonomous choices and actions. It
could even create the specter of paternalistic interventions in the name of thicker
conceptions of autonomy. The worry is that we will be tempted to override a per-
son’s specific minimally but sufficiently autonomous choice because it fails to sat-
isfy the standards of thicker conceptions of autonomy.
9 Respecting Personal Autonomy in Bioethics: Relational Autonomy as a Corrective? 149

References

Baumann, Holger. 2008. Reconsidering Relational Autonomy: Personal Autonomy for Socially
Embedded and Temporally Extended Selves. Analyse & Kritik 30: 445–468.
Beauchamp, Tom L., and James F. Childress. 2019. Principles of Biomedical Ethics. 8th ed.
New York: Oxford University Press.
Black, Cherie. 2007a. Boy dies of leukemia after refusing treatment for religious reasons. Seattle
Post-Intelligencer, November 29
———. 2007b. Jehovah’s Witnesses: Boy’s death ends battle. Seattle Post-Intelligencer,
November 29.
Blackhall, Leslie J., Sheila T. Murphy, Gelya Frank, et al. 1995. Ethnicity and Attitudes toward
Patient Autonomy. JAMA: Journal of the American Medical Association 274: 820–825.
Childress, James F. 2015. Jehovah’s Witness Refusal of Blood Transfusions: Ethical and Legal
Controversies. Unpublished paper.
Davis, J.K. 2007. Precedent Autonomy, Advance Directives, and End-of-Life Care. In The Oxford
Handbook of Bioethics, ed. Bonnie Steinbock, 349–374. New York: Oxford University Press.
Donchian, Anne. 2001. Understanding Autonomy Relationally: Toward a Reconfiguration of
Bioethical Principles. Journal of Medicine and Philosophy 26 (4): 365–386.
Faden, Ruth R., and Tom L. Beauchamp. 1986. A History and Theory of Informed Consent.
New York: Oxford University Press.
Friedman, Marilyn. 1997. Autonomy and Social Relationships: Rethinking the Feminist Critique.
In Feminists Rethink the Self, ed. Diana Tietjens Meyers. Boulder: Westview Press.
Maclean, Alasdair. 2009. Autonomy, Informed Consent and Medical Law: A Relational Challenge.
Cambridge: Cambridge University Press.
Mackenzie, Catriona, and Natalie Stoljar, eds. 2000. Relational Autonomy: Feminist Perspectives
on Autonomy, Agency, and the Social Self. New York: Oxford University Press.
McLean, Sheila. 2010. Autonomy, Consent and the Law. Biomedical Law and Ethics Library.
London: Routledge-Cavendish.
Manson, Neil C., and Onora O’Neill. 2007. Rethinking Informed Consent in Bioethics. Cambridge:
Cambridge University Press.
Meyers, Diana Tietjens, ed. 1997. Feminists Rethink the Self. Boulder: Westview Press.
Michels, Scott. 2007. Judge Allowed 14-Year-Old Jehovah’s Witness to Refuse Potentially
Lifesaving Care Leukemia Patient Dies After Refusing Blood Transfusion for Religious
Reasons. ABC News Internet Ventures, with reporting from the Associated Press, November 29.
Miller, Bruce. 2014. Autonomy. In Bioethics [formerly The Encyclopedia of Bioethics], ed. Bruce
Jennings, 4th ed., 302–307. New York: Macmillan Reference.
O’Neill, Onora. 2002. Autonomy and Trust in Bioethics. Cambridge: Cambridge University Press.
———. 2003. Autonomy: The Emperor’s New Clothes. Proceedings of the Aristotelian Society
sup. 77: 1–21.
Ostrom, Carol M. 2007. Mount Vernon leukemia patient, 14, dies after rejecting transfusions, The
Seattle Times, November 29.
Sherwin, Susan. 1998. A Relational Approach to Autonomy in Health Care. In The Politics of
Women’s Health: Exploring Agency and Autonomy, The Feminist Health Care Ethics Research
Network, coordinator Susan Sherwin, 19–47. Philadelphia: Temple University Press.
———. 1992. No Longer Patient: Feminist Ethics and Health Care. Philadelphia: Temple
University Press.
Stoljar, Natalie. 2014. Feminist Perspectives on Autonomy. The Stanford Encyclopedia of
Philosophy (Spring 2014 Edition), ed. Edward N. Zalta, URL = http://plato.stanford.edu/
archives/spr2014/entries/feminism-­autonomy/
Thaler, Richard H., and Cass R. Sunstein. 2008. Nudge: Improving Decisions About Health,
Wealth, and Happiness. New Haven: Yale University Press.
Traphagan, John. 2013. Rethinking Autonomy: A Critique of Principlism in Biomedical Ethics.
Albany: State University of New York Press.
Chapter 10
Patients’ Decision-Making Competence:
Discontents with a Risk-Relative
Conception

Bettina Schoene-Seifert

Abstract Regarding patients’ valid informed consent to treatment, it is a still con-

tested issue, whether the level of required decisional competency should mirror the
riskiness of the treatment at stake. This chapter attempts to refute such a risk-­relative
conception of competence (RRCC) on the basis of a systematic analysis of relevant
arguments, both conceptual and normative ones. A misconception of autonomy’s
proper value and attribution as well as the danger of a hidden paternalism are among
the main objections against RRCC.

10.1 Introduction

Informed consent requirements are a cornerstone of modern biomedical ethics and

medical law. “Gate-keeping”1 for the whole procedure of soliciting a patient’s per-
sonal consent to medical interventions is his or her mental competence. Taking com-
petence as existing in degrees, some ethicists have proposed a “sliding-scale”
conception of those mental capacities (making up for competence) that are deemed
necessary for a patient’s valid consent. In a nutshell, the suggestion requires a higher
degree of decisional abilities, the higher the (potential) harm of the medical decision
at stake. Although ethicist James Drane (1985) is often regarded as the originator of
this idea, others made similar suggestions even before him, not the least the
legendary President’s Commission in 1982.2 Since then, this suggestion of a

1
To my best knowledge, the now common gatekeeping metaphor was coined by Faden and
Beauchamp 1986, p.287ff.
2
Roth et al. 1977; President’s Commission 1982, Vol 1, p.60ff. Similarly: Gaylin 1982. I owe the
first reference to Culver and Gert 1990, p.632 and the following ones to Beauchamp 1991, p.70.

B. Schoene-Seifert (*)
Institute for Medical Ethics, History and Philosophy of Medicine, Münster, Germany
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 151

Switzerland AG 2022
J. F. Childress, M. Quante (eds.), Thick (Concepts of) Autonomy, Philosophical
Studies Series 146, https://doi.org/10.1007/978-3-030-80991-1_10
152 B. Schoene-Seifert

sliding-scale or, as I prefer to call it in line with probably dominant terminology, a

risk-related3 conception of competence (RRCC) has been adopted by many clini-
cians, psychologists, and bioethicists. They welcome it as a reconstruction of com-
mon clinical practice and as allowing a “balance between the competing values of
self-­determination and wellbeing that are to be served by a determination of compe-
tence” (Buchanan and Brock 1989, p.64). Others, however, have criticized the risk-
related conception for various conceptual and normative reasons, thus arguing for
the alternative, i.e. a fixed concept of patients’ decision-making competence.
Notably, although a controversial and high-level exchange of arguments among
some bioethicists took place around the 1990s,4 nothing like an expert consensus or
explicitly adopted majority view on this practically important issue has yet been
achieved.
In this paper I attempt to provide a systematic analysis of relevant arguments,
critically reconsidering older ones and elaborating on what I consider some new
concerns. As a result, I will join the camp of RRCC-opponents, holding that only a
fixed concept of decisional competency fits the very reasons why we (should) value
autonomy – despite some initial plausibility to the contrary. More specifically, I will
argue that not only hidden paternalism in matters of risk-taking, but also a danger-
ously inflationary attribution of autonomy (e.g. in cases of severe mental distur-
bance) with a number of problematic implications are lingering in RRCC’s
background. In any case, I want to show that the RRCC issue still or anew deserves
bioethical attention.

10.2 Decision-Making Competence: Preliminary Aspects

10.2.1 A Precondition for Valid Authorization

In medico-legal and bioethical contexts ‘competence’ is a common shortcut for

patients’ decisional competence5 and as such I will use it, too. The main medical
area in which this concept plays a role is the context of informed consent, where

3
There are several important ambiguities in the term ‘risk’, well-known from other fields of deci-
sion-making analysis: (1) risks can be understood and quantified either as the mere likelihood that
some negative effect might occur, or else as a joint parameter (in standard decision theory: the
mathematical product) of the probability and quality of possible negative effects. Moreover, risks
can be taken (2) as gross or net, (3) subjective or objective, as well as (4) expected ex ante or real-
ized ex post. For the principled purpose of this paper it seems acceptable to understand ‘risk’ as
some non-negligible probability of some non-negligibly net-harmful consequences assessed from
an intersubjective ex ante perspective – and to leave all the detail questions open.
4
See, for example, the following: Macklin 1987; Buchanan and Brock 1989; Culver and Gert
1990; Beauchamp 1991; Wicclair 1991, 1997; Wilks 1997, 1999; Cale 1999; Checkland 2001;
DeMarco 2002.
5
Some authors insist that competence is a purely legal term: e.g. Ganzini et al. 2004, p.264, but
common usage proves to the opposite.
10 Patients’ Decision-Making Competence: Discontents with a Risk-Relative Conception 153

patients are asked to make personalized decisions for or against suggested medical
interventions in either clinical or research contexts.6 For the sake of brevity I will in
the following only speak of decision-making on medical ‘treatment’ options.
Following the dominant understanding, the main impetus of the informed con-
sent doctrine is to grant competent patients decisional authority with regard to their
medical treatment – i.e. choosing between professionally appropriate interventions
or refusing any of them.7 Ethically, this “autonomous authorization model of
informed consent”8 was pioneered and notably fleshed out more than 30 years ago
by US-bioethicists Beauchamp, Childress, and Faden (Beauchamp and Childress
1
1979; Faden and Beauchamp 1986) and has since become the governing model.
According to common legal, ethical, and practical standards,9 a patient’s consent
should be authoritative or binding, if it fulfils the requirements of originating from
a patient who (i) is competent, (ii) actually understands what he is consenting to,
and (iii) decides voluntarily. These triple core requirements mirror the leading
understanding of valid consent as autonomous authorization (AA) given that they
correspond to the three standard criteria of autonomous action in general.
Thus, attribution of competence in the context of medical decision-making func-
tions as an entrance ticket that entitles a patient to make authoritative self-regarding
treatment decisions and to assume responsibility for part of this choice. An impor-
tant other part of responsibility stays, of course, with caretakers who must properly
recommend and explain treatment options and refrain from interfering with volun-
tary decision-making. But the final authorization rests with the patient. This is not
the place for analyzing the underlying reasons various stakeholders have to wel-
come or deplore this partial shift of responsibility. One can admit of informed con-
sent policies sometimes being motivated by caretakers’ fear of legal liability or their
wish to personally reduce some of their own responsibility, to shorten patient con-
tact etc. (See O’Neill 2007) However, such secondary motivations do not rule out a
serious ethical concern with respecting patients’ personal autonomy. Despite the
enormous variety of different concepts of personal autonomy in ethics that I cannot
even start to discuss in this paper, the core idea of patient autonomy in bioethics is

6
I use the term ‘intervention’ in a loose way, covering therapeutic, diagnostic, and preventive pro-
cedures, hospitalizations etc. Other patient decisions requiring competence may regard specific,
often controversial interventions primarily asked for by patients such as elective abortions, assis-
tance in suicide, or enhancement.
7
Thus, in terms of absolute authority, patients’ decisional power is that of a veto power. Whenever
there are several reasonable options patients are, moreover, authorized to choose among them.
Granting respect for patients’ autonomy in this standard sense is clearly distinct from wish-fulfill-
ing medicine – in contrast to some common suspicions.
8
“Autonomous authorization” was identified as the primary function of informed consent by Faden
and Beauchamp 1986. It was proposed as a label by Miller and Wertheimer 2011. I cannot here
discuss their (and other authors’) recent objections to the AA-model for consent to biomedical
research.
9
In the following my focus will be on ethical considerations and literature. However, I certainly
acknowledge that relevant debates are taking place in the legal(-ethical) literature, e.g. Alexander
1996; Amelung 1992.
154 B. Schoene-Seifert

the notion of (optional) authority in personalizing one’s medical treatment with an

eye also on one’s non-medical values and preferences. It is this specific task that
must dictate patient competence. As we shall see, it thereby renders RRCC a ques-
tionable idea.
Since the early attempts to flesh out ethically sound criteria for valid informed
consent, scholars have had, however, to differentiate between ideal and non-ideal
approaches, where the latter take into consideration that ‘ordinary’ people often
make decisions under less than optimal conditions (inner and outer ones) and that
‘ordinary’ medical personnel most often have to judge their patients’ decision-­
making within limited time and with restricted expertise. Clearly, such real-life
obstacles to the ideal of fully autonomous consent (whatever that is, exactly) should
be remedied to the extent possible. But in addition, informed consent policies
should, according to the standard view, already accept substantially autonomous
decisions as valid. Moreover, policies must make a set of reasonable presumptions
and set certain standards of evidence for the validity of consent.10 One such element
is the presumption of competence in the case of most conscious adult patients. In the
words of bioethicist Govern den Hartogh’s:
We assume in the case of most patients, and of most persons generally, that they are com-
petent in all their decisions until a reason for doubt arises, we do not investigate their com-
petence in the case of each and every decision. That would be impossible in practice, but
there is more fundamental reason for the presumption as well: we can only take each other
seriously and treat each other as moral equals on this basis. (Hartogh 2016 p.72; likewise:
Buchanan and Brock 1989, p.57ff )

As Hartogh goes on to explain, the most prominent “reason for doubt” regarding
patients’ competence is that their decisions deviate from their physicians’ treatment
recommendations. Given that these recommendations are commonly based on pro-
fessional comparative risk/benefit assessments of available treatment options, a
deviation is likely to appear puzzling, unreasonable, or irrational. Thus, this might
justify and trigger a process of competence assessment because one of several pos-
sible explanations for a judgmental deviance between a physician and her patient is
the latter’s lack of competence to properly make the judgment and decision at stake.
Hence, according to mainstream assumptions, a decision’s irrationality is to be
taken as an indicator for potential incompetence, not as the latter’s proof – although
the conceptual relation between incompetence and irrationality proves somewhat
tricky and controversial (see next paragraph).

10
According to my understanding, the difference between a given consent’s real qualities and its
assessment by legal and institutional policies corresponds to the difference between what Faden
and Beauchamp (1986, p.277f.) have called informed consent (sense1) and (sense2). In any case,
the gap between hypothetical ethical requirements for any single case of decision-making and
actual policy requirements can account for some puzzling questions. They regard, e.g., issues of
self-induced non-understanding, robust false beliefs, or imprudent decision-making by competent
patients. In all such cases that I cannot here pursue any further, ethically legitimate policy might
accept resulting decisions as if sufficiently autonomous and ‘informed’ although they do not fulfil
the standard criteria.
10 Patients’ Decision-Making Competence: Discontents with a Risk-Relative Conception 155

10.2.2 A Set of Abilities

Obviously, competence does not consist in one single element but rather in a set of
mental capacities. One minimalist account asks for (i) a capacity to understand and
communicate11 and (ii) a capacity to appreciate various treatment alternatives. On
closer look, this “understand-and-appreciate account of competence” (Culver and
Gert 1990, p.639; likewise, Checkland 2001, p.36) can only serve as a rough short-
cut, which I will deliberately use in the following. However, for the purpose of
conceptual, ethical, psychological, or clinical analyses both additions and further
fine-graining would be necessary.
A patient’s appreciation, for instance, presupposes the capability to deliberate,
i.e. to process information in a basically consistent way and to be potentially respon-
sive to new aspects and arguments. But it also needs the capability to assess avail-
able treatment options in the light of one’s personal values and preferences, thus
making subjective comparative evaluations. Again, a background condition of such
evaluations must be what is usually called “the possession of a (stable) set of per-
sonal values”.12 This last item cannot properly be called a capability in the strict
sense. Rather, as common-sense psychology already suggests, such value set is the
result of a person’s biographical experiences over time, not the least formed by
personal relations, social determinants, and a more or less conscious reflection of
who one wants to be. In debates on how to conceptualize autonomy, this “evaluative
self-image” (Quante 2011, p.174ff) plays a key role in at least two aspects. One
aspect concerns the logical and substantial links between the focal autonomy of a
patient’s treatment decision and her personal autonomy on a more global scale. The
other aspect concerns the ways in which the acquisition of one’s value-set can be
disturbed in autonomy-violating ways (e.g. by repression, traumatization, or brain-­
washing). Both aspects can be disregarded in the context of the sliding scales dis-
pute that mainly focuses on patients’ cognitive sophistication or deficits.
Cognitive competence, too, is an issue of on-going interdisciplinary discussion
among legal and philosophical scholars as well as among cognitive scientists.
Debates concern its composite nature, its interdependence with emotional or voli-
tional capacities, and its practical assessment in bedside situations.13 All of this will

11
Communication of choice is sometimes listed as a necessary element of competence. Strictly
speaking, this is not required for decision-making as such, but rather for its operationalization
(think of a locked-in patient who might well come to a decision but might under unfortunate cir-
cumstances not be able to communicate it). However, in practically oriented lists communication
is a standard item. Moreover, communication usually has an important role in the process of patient
understanding.
12
Hence, various authors add “the possession of a set of values and goals” as a third minimalist
requirement. Compare Wicclair 1991, p.91; Buchanan and Brock 1989, p.23.
13
To quote Louis Charland, bioethicist and philosopher of cognitive sciences: “The combined theo-
retical and practical nature of decisional capacity in the area of consent is probably one of the
things that makes it so intellectually compelling to philosophers who write about it. But this is still
largely uncultivated philosophical territory. One reason is the highly interdisciplinary and rapidly
156 B. Schoene-Seifert

again be ignored in the following – in favour of a somewhat sketchy understand-­

and-­appreciate concept of competence. Whatever its more precise nature, its real-
ization surely is a gradual, not a binary phenomenon. People can memorize,
understand, deliberate, appraise, or make a choice with gradual goodness.14
(Charland 2015, sec. 1) Against this background it is commonly and reasonably
agreed, that an authoritative patient decision only needs a sufficient amount of each
of the capabilities that are constitutive of competence, thus resulting in a sufficient
degree of competence itself. It is this sufficiency threshold that proponents of risk-­
relativity in the determination of competence want to handle flexibly whereas their
opponents advocate a fixed, i.e. risk-independent threshold.

10.2.3 Competence and Rationality

Impaired decision-making capacities can lead to irrational choices. The latter can,
according to the AA-model of legitimate consent, possibly be overridden in order to
protect patients from harm due to incompetence. To characterize a decision as ‘irra-
tional’ can, however, be understood in three distinct ways – inviting for normative
misunderstanding and controversy that can only gain from more conceptual
differentiation.
In its most narrow understanding decisional irrationality results from a deficit in
the capacity to perform procedurally correct deliberations (drawing logically cor-
rect conclusions etc.). Here, irrationality does indeed correspond to a subset of
incompetence. Quite a number of authors take this way, e.g. lawyer-ethicist Larry
Alexander:
If one’s choice […] is to count as valid consent, one must presumably be of a certain age,
lack serious mental disease, irrationality, or intoxication, and have a certain degree of self-­
control. (Alexander 1996, p.166ff )

Irrational decisions in this dispositional-procedural sense lack a necessary condition

for authority. Thus, if they prove immune against consultation and are prone to
appreciably harm the patient they can be disregarded.
A very different concept of irrationality, however, refers only to the decision’s
content. In particular, well-minded physicians have long tended to regard as irratio-
nal those patient-decisions that seem incompatible with widely shared values, e.g.
decisions to refuse life-saving treatment under certain conditions. According to the
depicted standard AA-view, irrationality in this sense of substantive deviation from
what others consider objectively good or reasonable can never serve as a

changing nature of the field. Clinical methods and tests to assess capacity are proliferating. The law
is also increasingly being called upon to respond to these clinical developments. All of this makes
for a very eclectic and challenging field of inquiry.” (Charland 2015, sec. 1).
14
I do not here take issue with questions of the exact nature – gradual or binary, actual or disposi-
tional – of each capability at stake.
10 Patients’ Decision-Making Competence: Discontents with a Risk-Relative Conception 157

justification to override a patient’s decision. Otherwise, broadly accepted rights to

personalized value-judgements and to their authority would be violated. This dan-
ger of undermining patient autonomy by overstretching the concepts of irrationality
and competence has long been recognized. It has caused various authors to insist on
a strictly procedural – in contrast to an outcome oriented – understanding at least of
competence, if not also of rationality.15
Yet another conceptual strategy to block an outcome-based understanding of
competence is to describe competence assessment in the practice of informed con-
sent procedures as “gatekeeper”. (Beauchamp and Childress 2019, p.112). This
already introduced metaphor reflects the view, that competence is in principle inde-
pendent of particular decisions’ outcome: Whenever a patient is determined incom-
petent to make a pending decision, the informed-consent ‘game’ is over and some
surrogate’s decision must be sought. To be sure, in many cases, it will still be impor-
tant and necessary to inform the patient about her condition and treatment alterna-
tives, to assess her preferences, and to deliberate together (see below 10.5.3). But
the goal of seeking and relying on her treatment authorization must be dispensed
with, whenever the patient is ‘kept off the gate’.
Between the two depicted options of understanding competence/rationality as
dispositional-procedural or else as outcome-oriented in an objective sense, there is,
however, a third option that does not fit into the standard binary model of process/
outcome views. Probably, it touches the most contested issue in this context, namely
the authority of patients’ decisions whose content turns out subjectively unreason-
able, i.e. incompatible with the decision-maker’s own values. Whether or not you
call such decisions irrational,16 doesn’t – as Jason Hanna has recently unfolded and
argued for – their “imprudence” resulting from a somewhat “impaired” decision-­
making undermine their validity as much as is admitted in cases of decisions under
obvious pain or emotional distress?17 Answers to this problem require a deeper
analysis of the normative differences between a dispositional and an actual-use
understanding of competence/rationality – a task that I must leave to another
occasion.

15
So, for instance, Buchanan and Brock (1989, p. 50): “An adequate standard of competence will
focus primarily not on the conent of the patient’s decision but on the process of the reasoning that
leads up to that decision.” Feinberg (1989, pp.106ff) has insisted on differentiating between a deci-
sion’s “rationality” in the procedural sense and its “reasonableness”, obviously without having
been followed.
16
Thus e.g. Culver and Gert (1990) write: “If one requires a patient to make use of his competence
when making a decision, the one is no longer judging the competence of the person to decide, but
the rationality of the decision made” (p.622; emphasis added).
17
Compare for instance Hanna 2011. A plausible answer might refer to some surplus value of a
policy that insists on the normative difference between impairment and imprudence, even if it can-
not be justified in single cases.
158 B. Schoene-Seifert

10.2.4 Task-Specificity

Yet another agreed-upon assumption of the standard view on patients’ competence

is its task-specificity. In the words of Culver and Gert:
To discuss competence with precision one should focus on a person’s competence to do a
particular kind of action or make a particular kind of decision, rather than trying to charac-
terize the person as a whole. […] However, even this is too wide a classification, for a per-
son may be competent to make one kind of medical decision and not another kind. (Culver
and Gert 1990, p.620)

Assuming some task-specificity for patients’ decision-making competence is indeed

highly plausible for at least two reasons that I will soon discuss. Nevertheless, fine-­
graining the context-dependence of competence requirements should only go as far
as can reasonably be justified. In particular, the task-specificity ‘doctrine’ and its
prima facie plausibility should not be taken as carte blanche for RRCC by simply
claiming a medical decision’s degree of riskiness to be one of its relevant task-­
specific aspects.
These are the two arguments in favour of task-specificity: The first is a general
feature of the very concept of competence. Depending on the task in question, cri-
teria of competence gain strictly different content. Competences to (i) artistically
walk on a tightrope18 or (ii) to consent to one’s medical treatment exemplify huge
involved differences. Obviously, the first task requires athletic skills, a good sense
of balance, and freedom of giddiness whereas the second task requires understand-
ing and evaluating alternative medical risks and benefits for one’s own future life
within the framework of one’s value set.
A second impulse for requiring task-specificity when it comes to patients’ medi-
cal decision-making results from certain paradigm clinical experiences. Thus, as
often discussed, a patient suffering from severe psychotic delusions might prove
competent to play chess on a highly competitive level or to make sophisticated
financial transactions. But the same patient might be incompetent to evaluate anti-­
psychotic medication, just because she does not realize the pathological nature and
provenance of her delusional episodes. Experiences of this type have given rise to
medico-legal maxims and regulations not to judge a psychiatric patient per se as
lacking decisional competency, but to determine her competence with an eye on the
specific decisional task she is facing. (Ganzini et al. 2004)
In addition, caretakers are aware that patients’ competence may fluctuate accord-
ing to their emotional conditions: decreasing when they are in a state of pain, anxi-
ety, or excitement and increasing when they feel relaxed and taken care of. Therefore,
physicians are commonly advised to regard competency as a context-sensitive phe-
nomenon and to strive as much as possible for optimal conditions when seeking a
patient’s informed consent. Unquestionably, awareness regarding the context-­
dependent fragility of patient competence is important. But it goes without

18
Tightrope walking as a comparative example for understanding competence has been introduced
into the RRCC debate by Wilks (1997, p.419ff.). As others have consequently done, I will stick to it.
10 Patients’ Decision-Making Competence: Discontents with a Risk-Relative Conception 159

argument that context-regarding and task-regarding relativity can in principle be

kept and handled apart.19 Thus, context-dependent fragility of cognitive capabilities
is a different issue than their task-specificity, though they might look adjacent on
first sight. Moreover, to acknowledge that the capabilities making up for someone’s
competency are decisively task-specific (high wire competence differing from deci-
sional competence) does not solve the problem of how finely these tasks should be
individuated. This will turn out to be an important point for the risk-relativity issue
of decisional competence.

10.2.5 The Idea of a Risk-Related Concept of Competence

Take the following example from a recent clinical paper, meant to inform and edu-
cate physicians about the understanding and assessment of patients’ decision-­
making capacities:
[A] mildly demented patient might be able to decide that she wants antibiotic treatment for
a urinary tract infection because the treatment allows her to pursue important goals such as
feeling well or staying out of the hospital, and its burdens and risks are low. On the other
hand, the same patient might be unable to weigh the multiple risks and benefits of a com-
plex neurosurgical procedure with uncertain trade-offs between quality and quantity of life.
(Ganzini et al. 2004, p.264)

Obviously, the example is meant to illustrate the view that a patient’s competence
(here: her being “able”) to decide on medical treatments ought to be assessed rela-
tive to the “burdens and risks” involved in the decision at stake. The example sug-
gests that a given cognitively impaired patient might be competent to authorize a
low-risk antibiotic treatment while not being competent to authorize neurosurgery.
Note, however, that the example remains somewhat ambiguous on two points that
will later come up for discussion: First, it does not say whether the patient should
also be considered competent to refuse the antibiotic treatment (most probably not),
whereas her asserted decisional incompetence regarding neurosurgery seems to
refer to both consent and refusing. Second, the depicted scenario remains unclear as
to whether the patient is deemed incompetent to decide about neurosurgery because
of the complexity of the involved harms and risks of the procedure or rather because
of its degree of riskiness (probably the latter).
In any case, accepting a risk-relative conception of competence (RRCC) seems
to correspond well with widely shared intuitions and to reflect common clinical
practice. Moreover, RRCC has to the best of my impressions become an accepted
view in bioethics and a – if not the – standard model in clinical care.20 Before

19
Beauchamp seems to overlook this point when he equates “the specific competence to make a
certain kind of decision – e.g., whether to cease a certain drug therapy” with “the capacity to make
such choices in the circumstances” (Beauchamp 1991, p.61).
20
This view is shared e.g. by Checkland 2001, p.36. Most notably and influentially since Drane’s
pioneering article, Dan Brock and Allen Buchanan have at length defended RRCC (Buchanan and
160 B. Schoene-Seifert

starting to critically analyze various arguments for and against this view, I must pay
some attention to the general understanding and function of competence judgments
in the particular context of patient care. As it will turn out, this understanding
already is by no means uncontroversial, although crucial for the RCCT debate.
In his seminal paper, James Drane (1985) suggested three differently demanding
standards of patient competence that should be applied in accordance with the qual-
ity and probability of harmful consequences of the decisions to be authorized. Thus,
he argued for:
(a) A low standard of competence vis-à-vis decisions involving small risks and
high benefit. This low standard, he suggested, would already be met whenever
a patient is conscious and in possession of any ability to express consent, if only
implicitly.
(b) A middle standard of decisional abilities for choosing treatments with high
risks of harm if they are a patient’s only chance to get better or to survive. This
standard could be met by patients who can understand their medical situation
and the proposed treatment but who are unable to critically reflect and evaluate
options in light of their values and beliefs. Drane suggests that mildly cogni-
tively impaired or borderline patients be among the candidates here.
(c) A high standard of decisional abilities for refusing interventions of low risk and
high benefit. Here Drane (1985, p.19) thinks that candidates must be reflective
and self-critical and well as in possession of “mature coping devices”.
For matters of illustration and simplicity let us fill the categories of treatment in
Drane’s three standards with the treatment options involved in the example from
Sect. 10.1 above: antibiotic treatment for a urinary tract infection as a small risk/
high benefit option and complex neurosurgery as a high risk/unclear benefit deci-
sion. Thus Drane would have us, for instance, accept a severely demented patient
authorizing the antibiotic treatment, but not refusing it. Basically, Drane adduced
two arguments, though not in much detail: the moral duty to properly balance
respect for autonomy with patient benefit and common clinical practice. The last
point is easy to imagine but hard to evidence, and it is certainly not a convincing
argument for anybody who considers such practice faulty. The moral argument will
be discussed below.

Brock, Deciding for Others, Chap. 1). In contrast, Tom Beauchamp and Jim Childress have in their
common monograph Principles of Biomedical Ethics supported a fixed threshold from the first
edition (1979) to the latest one (2019). Remarkably though, Beauchamp has, in his singly authored
essay “Competence” (1992), defended RRCC. For a recent RRCC-approval in research consent
see Bromwich and Rid 2015. For a recent example from the German-speaking literature see
Hermann et al. 2016.
10 Patients’ Decision-Making Competence: Discontents with a Risk-Relative Conception 161

10.3 Arguments from Conceptual Grammar

Some authors, such as Tom Beauchamp (1991, pp. 50ff), have tried to draw concep-
tual conclusions regarding RRCC from what they see as the general concept of
competence. The underlying assumption is that there is a core concept with a gen-
eral structure that is prior and transferrable to its contextualized subtypes. In par-
ticular, this is playing a role in arguments on normativity, externalism, symmetry,
and risk-relativity. As I will try to argue, the idea of a deeper conceptual logic that
might answer the controversy on RRCC is implausible. Rather, questions have to be
answered within the context of decision-making competences (Likewise, Checkland
2001, p.51).

10.3.1 Externalism and Normativity

Some authors have disputed whether referring to competence’s generally externalist

rather than internalist concept can defend RRCC. In the words of Ian Wilks:
Competence is not an absolute, not a feature solely native and internal to a person; compe-
tence to render a decision likewise is determined only partly by the qualities of mind that
figure in the act. The nature of the options, with their benefits and risks, is also relevant; and
this is very much the insight now supplied by those who justify a risk-related standard.
(Wilks 1997, p.426)

Although some opponents of RRCC agree with this diagnosis (e.g. Culver and Gert
1990, p.635), the point seems to be poorly made: could anybody seriously deny that
capabilities can be purely internally realized and at the same time be individualized
by an external task? (Likewise, Wicclair 1999, p.151) Actually, this seems to be the
standard case – as can be exemplified by the capability to distinguish between vari-
ous shades of grey, to perform tightrope dance over a distance of 100 meters, or to
make a decision about medical treatment. Hence, the distinction between internal-
ism and externalism in this sense seems of little help. What is ultimately needed to
justify or reject RRCC will be a specification of the external task in question.
In the literature on decisional competence in medical contexts, it is often and
rightly emphasized that this concept – like many other competences: even the one
in high wire artistry – is normatively impregnated. This is by no means a general
conceptual feature. Rather, certain circumscribed competences, e.g. the one to cor-
rectly memorize 100 digits in a row, are purely descriptive. Normativity comes into
play whenever the task in question gets more abstract, allowing for gradual compe-
tence or for leeway in choosing and weighing relevant sub-competencies. In such
constellations, competence is used as an evaluative résumé of someone’s abilities or
skills to perform a certain task. It involves three steps, namely (1) the identification
of the relevant capabilities, (2) a determination of the degree to which these abilities
are required, and (3) a personalized assessment of someone’s possession of these
requirements. The first two steps are prone to controversies resulting from
162 B. Schoene-Seifert

dissenting views on the task in question. Thus, there seems to be some latitude, e.g.
in deciding whether a (sufficiently) competent high wire artist needs jumping com-
petence or whether the fact that he might be unable to perform for longer than
10 min is of relevance. Whereas athletic competences involve aesthetic values, deci-
sional competence in the context of informed consent involves ethical norms con-
sidered relevant for making authoritative decisions. Only the ascriptive third step,
certifying the absence or presence of what has been fixed by the first two steps, is by
itself non-normative: a given patient is described as meeting or not meeting agreed-­
upon standards. But, again, their nature and degree are set with an eye to their nor-
mative function. In Tom Beauchamp’s words:
[…] Any concept that serves this function is inherently normative in the way it is used to
establish the abilities and level of abilities – a normative choice of entry criteria – and to
certify a person who possesses such abilities. […] Thus, it is a mistake to infer that empiri-
cal judgments of psychological competence are free of prior evaluative commitments. The
reverse is true: they are inescapably value laden. (Beauchamp 1991, p.53)

Setting due thresholds for patients’ decision-making competence is ruled by two

ethical norms: preventing patients from taking decisional responsibility without
being ‘fit’ to do so on one side and preventing the infringement of the rights of ‘fit’
patients to authorize treatment as they like, on the other side.
Having said this, one crucial problem seems to evolve: Is there a clear distinction
to be drawn between competence assessment motivated from those two norms as
described by the last sentences, and competence assessment incorporating a bal-
ance of those two norms as RRCC proponents have it? I will come back to this point
in Sects. 10.4 and 10.5. But here is already a short anticipatory answer: The incor-
porating (RRCC-) approach seeks to give more weight to beneficence than liberals
want to, and it does so in a non-transparent, uninformative way. In contrast, the
motivated (fixed threshold) approach is bent to explain what grounds legitimate
claims to authority. Although a motivated approach can also be applied in a way that
violates authority claims, these can only be overridden by blatantly giving priority
to beneficence-concerns.
Understanding patient competence as a normative concept is part of the more
complex story of what it means to understand authorizing autonomy (AA), the over-
arching concept, as normative, too. If someone qualifies as an agent of sufficient
(task-specific) AA, third parties have good pro tanto reasons not to interfere with his
self-regarding decision-making. Thus, autonomy and its ‘ingredients’ are normative
in this first sense of constraining third party behaviour. At the same time, they are
conceptually and empirically constrained by why, depending on rival moral theo-
ries, AA is viewed as a claim right to be respected or as a value to be promoted in
people’s life, at least in our culture. The most straightforward and simplest answer
points to AA’s instrumental value in shaping one’s life according to personal values
and preferences as well as to a widespread preference for being in control as such.
Both aspects would preclude, e.g., respect for ‘authorizations’ resulting from drunk-
enness, error, or manipulation. In this second sense, competence and AA are
10 Patients’ Decision-Making Competence: Discontents with a Risk-Relative Conception 163

normative in that they inherit constraints by the very idea of what turns AA into an
ingredient of a good life.

10.3.2 Asymmetrical Risk-Related Competences?

Drane’s suggestion openly invites a remarkable asymmetry in the authoritative

power granted to a given patient: lacking the proposed high standard (see above),
she might be considered competent to consent to a certain intervention but incom-
petent to refuse it, or vice-versa. This asymmetrical competence has been repeatedly
rebutted as paradoxical by some (Cale 1999; Demarco 2002, p.234; Hartogh 2016;
Wicclair 1991, 1999), but also explicitly been defended by others (Buchanan and
Brock 1990; Wilks 1999). As I will try to argue in this and the following section,
only the former view is convincing.
It was ethicist Ian Wilks (1997, 1999) who introduced the example of two tight-
rope walkers into the debate on RRCC in order to defend the latter despite its impli-
cation of asymmetrical competence. He makes us imagine two persons facing the
challenges of tightrope walking. In Wilks’ scenario, one of them does a marvelous
artistic job and can thus be called competent to walk on a high line with or without
safety net. In contrast, the second candidate happens to still be a rather lousy high
wire performer, thus facing a substantial risk that he might fall off the line. According
to Wilks, one correctly uses the concept of competence if one describes this second
guy as incompetent to perform the high-wire act without safety net, but as compe-
tent to perform with the net in place. As Wilks asks and then concludes,
What is the basis for the fallible walker’s being competent to walk the wire with the net in
place, but not without? It is certainly not something intrinsic to the person, not a matter of
the person’s abilities – since by the terms of the example, there is an equal chance that the
fallible walker will complete the task successfully whether the net is in present or not […].
The difference lies entirely in something extrinsic to the person, in the level of risk the situ-
ation involves; and what emerges here is that the level of competence is affected not only
by degree of ability but also by degree of risk. (Wilks 1997, p.420)

Drawing from various authors’ critical answers to Wilks proposal21, it seems obvi-
ous that in his case four different tasks-specific competences can be helpfully distin-
guished, i.e. the competence to (i) tightrope walking in general; (ii) walk a particular
high line without a substantial risk of falling off; (iii) manage the same high line
with a safety net in place (unknown to him), possibly interrupted by a drop-off; (iv)
take responsibility for walking the high line without a safety net. Wilks’ imagined
lousy artist lacks (i) and (ii), but possesses (iii); and we are left ignorant about (iv).
What then does this case prove?
Competence (i) is a resumé evaluation depending on an artistic candidate’s cour-
age together with sophisticated acrobatic skills such as strength, body control, and

21
Thus, I draw insights from Cale 1999; DeMarco 2002; Checkland 2001, and Wicclair 1999.
164 B. Schoene-Seifert

balance management. Normally, possession of (i) is indicated by a comparatively

low risk of falling off the line and thus of risking injury when performing without
safety net, i.e. competence (ii). (Wicclair 1999) In contrast, if someone has only
mediocre acrobatic skills, he will surely lack (i) and (ii) in a package, but might
nevertheless possess competence (iii), as in Wilks’ example: he might possibly drop
off the line but still get credit for mediocre performance. Contrary to Wilk’s own
description, what makes the mediocre guy competent enough for (iii), but incompe-
tent for (ii) is his relative deficit in genuine (‘internal’) abilities when compared to
the other, more brilliant high wire candidate. Competence (ii) for the risky task
presupposes high artistic skills (i), while competence (iii) for the comparatively
low-risk task does not. Hence, the lesson seems to be that competence-ascriptions
might indeed depend on the whether the task in question involves running or avoid-
ing certain risks; in such cases differences in competence can then be reduced to
differences in relevant abilities. As with regard to other competences, their ascrip-
tion depends on the exact nature of the task in question. This might simply become
invisible if one attests the mediocre guy’s substantial tightrope competence when it
comes to performing with a net and insufficient tightrope competence when the net
is removed.
Moreover, Wilks’ analogy would, as others have already argued, be a more
appropriate test case for RRCC in medical decision-making if it turned on an acro-
batic candidate’s decision to perform without the safety in question. (Checkland
2001, p.49ff.; Wicclair 1991, p.97ff, and 1999, p.151ff.) Obviously, this decisional
competence would depend not primarily on acrobatic skills but mainly on the self-­
critical appraisal of one’s acrobatic potential, as well as on understanding and evalu-
ating the harmful consequences of falling including permanent injury or death. In
the context of decision-making the central question is whether Wilks is justified in
claiming that:
[…] cases do indeed exist where one can be competent to say yes but not no, or vice versa;
and that it is thus not an anomaly in the risk-related standard [i.e. RRCC] that it entails the
existence of such cases. (Wilks 1997, p.413)

Let us then take closer looks at this particular context.

10.3.3 Asymmetrical Decisional Competences?

On first sight, making an informed choice for or against a medical treatment T

seems to be an essentially comparative task within a set of options.22 In the simplest
case, the choice is only between T and non-T. In this scenario, understanding and
evaluation of risks and benefits must count on both sides, because chances of benefit

22
Epistemically this seems true even in light of a common asymmetry in consent versus refusal
practice, where the former is taken to require substantial understanding of the relevant facts and
prognoses whereas the latter goes without much ado.
10 Patients’ Decision-Making Competence: Discontents with a Risk-Relative Conception 165

from T correspond to risks of non-benefit/harm from non-T. But also in more com-
plex cases e.g. with options T1, T2, or no treatment, evaluation in any reasonable
sense involves comparison, i.e. appraisal of each option in the light of its available
alternatives.
However, as plausible as a symmetry requirement seems to be in the specific case
of patients’ informed decision-making for or against treatment options, this might
not be true for decision-making competences in other contexts. What can be made
of such arguments resulting from examples of the following kind? Take a European
politician who is asked to take over major responsibility in nuclear energy politics
right after the Fukushima catastrophe. Given her second-order knowledge that she
lacks the relevant knowledge and expertise (and maybe even the competence to ever
grasp the complex relevant facts) she decides against the offer.23 Uncontested, she is
competent to do so; but had she also been competent to decide the other way? Given
the politician’s lack of expertise on nuclear dangers, presupposed in the example,
one might be tempted to deny this.
On closer sight, this conclusion in favour of an asymmetrical decision compe-
tence seems flawed, nonetheless. First of all, the decision on taking over responsi-
bility for nuclear politics is clearly a different one than any decision on how to make
use of this responsibility. Second, whatever the outcome of the decision on assum-
ing responsibility, it must be based on the competence to roughly understand and
appraise its necessary preconditions and its possible consequences in both, magni-
tude and quality. She must, in other words be competent to compare an affirmative
and a negative answer and decide in light of this comparison. Therefore, someone
who is competent to answer in the affirmative must also be competent to answer
no – although her ignorance might be asymmetrical.
Following these arguments, RRCC so far turns implausible only in one of two
possible readings. Thus, it does seem implausible that a patient might – as Drane or
Brock/Buchanan see it – at the same time be competent to consent to treatment T,
but incompetent to refuse T. Here the alleged asymmetry is choice-dependent.
However, according to a second possible reading of RRCC the granted asymmetry
is choice-set-dependent.24 Under this version, a patient might be competent to
authorize or refuse treatments whenever all options involve only low risk and low
hazard, e.g. taking or not taking nose drops to alleviate symptoms of a simple cold.
At the same time she might be incompetent to decide for or against any available
treatment options for a severe neurological damage. Let us look at this choice-set-­
dependent reading of RRCC in the next section.

23
Compare Checkland 2001, p.41ff. on the complex role of second-order abilities in matters of
competence.
24
To my knowledge, in the RRCC literature this distinction has not been systematically pursued.
166 B. Schoene-Seifert

10.4 Arguments from the Task at Stake

Taking stock of the analysis so far, patients’ medical decision-making competence

is to be determined by the task in question. Thus, understanding and individuating
the latter is of decisive importance. Hence, what exactly is the leading idea of grant-
ing patients authority over their medical treatment?

10.4.1 Making Treatment Preferences Decisive

Granting authority rights to patients is a special case of granting rights to persons.

Rights in turn get justified in various ways – much dependent on various legal/ethi-
cal theories. Circumventing much of the involved controversies, there seems to be
common ground for assuming that patients’ authority-rights function (whether pri-
marily or not) as instruments that protect their interests in choosing treatment
options within the available choice set. The availability in question being deter-
mined by factual accessibility and professional standards, patients have to agree or
refuse according to their values and preferences.
The importance attributed to personal decision-making can be analysed as a
threefold one: (i) some medical aspects of an intervention (especially, but not only:
pain) can be experienced in personally different ways and degrees, therefore
demanding an individual assessment; (ii) generally wanted as well as generally
unwanted medical aspects of a given treatment (especially, but not only: end-of-life
issues) can have inter-individually different impact on patients’ general preferences
and thus need individual assessment. I will call this the life-shaping25 or holistic
aspects that some medical decisions might have; (iii) being treated as an authorizer
over one’s medical treatment has in our culture become as such an aspect of being
respectfully treated as a person.
Trivially true but easily forgotten, the tasks in question are very different from
physicians’ tasks to come up with treatment recommendations in the first place:
with regard to medical chances and risks, the patient’s part is usually and inevitably
that of a lay-user who has to rely on expert opinion, thus making various and again
often implicit judgments on the expert’s trustworthiness, on the information she
considers pertinent, and on the treatment’s potential impact on her life. Where this
might be a simple task at the dentist who recommends refilling a back tooth, it might
become complex in other contexts. Here, medical decision-making can, as already
emphasized, gain the importance of life-shaping – in light of a patient’s personal
fears, hopes, values, or preferences that might well turn her into a ‘non-standard’
patient.
Hence, decisions about medical options might gain what I have called a holistic
dimension. This is particularly likely, if such decisions might have substantial

25
This – somewhat self-imposing – expression is owed to Silver (2002, p.462ff.).
10 Patients’ Decision-Making Competence: Discontents with a Risk-Relative Conception 167

effects on one’s quality or length of life – as in matters of reproduction, body shape

or mental states, research settings for the treatment of major ailments, or end-of-life
scenarios. Even here, life shaping is, of course, partial, given the many natural and
social determinants of a human life. Moreover, life shaping neither needs to be
ambitious, nor explicit, nor even positively wanted. But it is always an option to
some extent, even if produced by negative decisions. Last not least, life shaping in
the explained sense is an option not supported by all cultures (cf. Beauchamp 1991,
p.59ff.).
Given this trivial reminder, it should become clear that self-regarding medical
decision-making might get a holistic, and value-laden dimension going far beyond
the professional expertise of physicians. To perform this task, patients need to
understand such core concepts as life, disease, death, and risk. They must have
‘ordinary’ computational skills, a decent memory, and the ability to deliberate. And
they must have some evaluative framework relative to which they can appreciate
and comparatively evaluate treatment different burdens, benefits, and life
perspectives.
Obviously, these deliberations add nothing really new to the standard view of
patient competence as a set of mental abilities (see Sect. 10.2). It might only make
clearer that patient competency has to differ from other decisional competences –
e.g. from physicians’ competence for identifying best treatment options or from
people’s competence for risk assessment in various other private or professional
realms contexts – mainly by its personal evaluative holism. This might also explain
some of the ongoing controversies in bioethics about the proper conceptual place of
personal values and personal value-conformity (‘authenticity’) in informed consent,
because this place is often not sufficiently elaborated in the standard view.26

10.4.2 The Ordinary-Person Standard

Holistic decision-makers about medical treatments, I have suggested, do need a set

of values and sufficient understanding of such key concepts as disease and death,
risk and probability, uncertainties, of expertise and credibility. They need the ability
to understand something about risk framing and to appreciate the potential occur-
rence of specific harms and benefits for their lives. But none of these requirements
exceeds the competences of an ordinary person. This thesis does not follow from
any empirical data showing that ordinary persons happen to have the competencies

26
See Sect. 10.2.2. above. Sometimes reference to values is made within the bundle of mental
capacities, e.g. requiring “the ability […] to evaluate [treatment] consequences in view of one’s
own values” (Hartogh 2016, p.71) – presupposing the possession of such values. Sometimes
authenticity is introduced as an additional requirement outside (narrowly understood) competency,
see (Quante 2011). Value conformity should not get misunderstood as nailing patients down to
their past values: As has often been emphasized, it should also entail leeway for new, may be ‘cou-
rageous’ evaluations.
168 B. Schoene-Seifert

in question. Rather, it follows from the normative premise that competence require-
ments standards should be governed by an ordinary person standard. After all, it is
normal people who are viewed as giving shape and meaning to their lives, not some
extraordinarily gifted sub-class of them.27
So far, nothing in this section has explicitly touched the main issue of RRCC,
namely whether risky choice-sets (e.g., decisions for or against neurosurgery) do
not, after all, require some deeper competence in grasping these decisions’ potential
life-shaping meaning than do low-risk choice sets (e.g., regular decisions at the
dentist). But now, this suggestion can clearly be dismissed: What is needed in either
scenario is a somewhat realistic assessment of the decisions’ degree of riskiness and
of riskiness as such: In order to meet AA requirements, a patient at the dentist must
be competent to understand the relative absence of risk (and the latter’s meaning).
Likewise, the patient to whom neurosurgery has been recommended must be com-
petent to understand the relative presence of risk and what this means for life and
health and maybe for his life-plans etc. Hence, to assess some treatments as trivial
and others as non-trivial for one’s life-shaping project presupposes the very same
basic competences.
On the other hand, there are many decision-makers who do not qualify for evalu-
ative holism and thus do not count as competent decision-makers, even if the issues
at stake are completely trivial. Of course, a severely demented patient might and
should be respected in as many of their self-regarding choices on trivial matters or
harmless issues as possible. But this should not blur important distinctions in justi-
fying why someone’s decisions should be respected (see Sect. 5.4 below).

10.4.3 Complex Versus Risky Decisions

A controversially answered question in the RRCC debate concerns the role of cer-
tain medical issues’ complexity. This complexity, it is sometimes argued, might in
certain cases require beyond-ordinary cognitive skills, but should be kept apart from
riskiness. In Beauchamp/Childress’ words:
It is correct to say that the level of a person’s capacity to decide will rise as the complexity
or difficulty of a task increases (deciding about spinal fusion, say, as contrasted with decid-
ing whether to take a minor tranquilizer), but the level of competence to decide does not rise
as the risk of an outcome increases. It is confusing to blend a decision’s complexity or dif-
ficulty with the risk at stake. (Beauchamp and Childress 2019, p.117)

Convincing as this seems on first sight and in principle, I tend to think that a
complexity-­relative concept of competence will rarely, if ever, be relevant and that
the opposite view might rest on a misunderstanding: to make holistic evaluations,
patients do not have to understand why and by which mechanisms a spinal fusion or

This view, I take it, belongs to the framing theses of the AA-model of informed consent. Compare
27

Beauchamp and Childress 2019, p.100ff.

10 Patients’ Decision-Making Competence: Discontents with a Risk-Relative Conception 169

rather a tranquilizer might or might not be effective. Rather, they have to understand
a procedure’s chances of cure, its uncertainties or possibly its being a non-stand
procedure. Ordinary persons, I take it, can grasp such aspects if they are communi-
cated understandably. Physicians’ communication about complex treatment and
about its expected consequences plays a major role in empowering patients to
become autonomous authorizers. The task of translating science talk into messages
that ordinary patients can understand remains a challenge. The same proves true for
risk communication, which has been shown to lead to innumerable false beliefs on
the part of patients. But ultimately this seems a matter of deficit understanding and
incompetent communication on the side of physicians rather than a lack of capabili-
ties on the part of patients (See Gigerenzer et al. 2007).

10.5 Arguments from Protection

10.5.1  allibilism: Raising Standards of Evidence or

F
of Substance?

Paying particular attention to a patient’s competence whenever she wants to choose

against recommended treatments has already been identified (see Sect. 10.2.1) as a
prudent strategy in the service of detecting incompetence at low moral cost. The
same is true for paying closer attention to a patient’s competence whenever the
choice-set involves considerable risks of harm in absolute terms – regardless of
physicians’ recommendations. In both cases not competence itself but evidence of
competence is subject to stricter requirements in case of risk-taking decisions. Many
have thus recommended this ‘ethically innocent’ precautionary policy (e.g.
Beauchamp and Childress 2019, p.117f.; Wicclair 1999, p.153).
Obviously, raising the standards of evidence of competence for risky choices
principally differs from raising the substantive standards of competence, i.e. in
degree or quality. However, Joel Feinberg, pioneering proponent of respecting per-
sonal autonomy in both law and ethics, has tried to link the two in defending RRCC
on what he sees as purely fallibilistic or epistemic ground:
The more risky the conduct the greater the degree of voluntariness required if the conduct
is to be permitted. […] [This] entails that the voluntariness required for permissible self-­
endangering actions should be determined by standards whose stringency varies directly
with the gravity of the risked harm and with the probability of the risked harm occurring.
(Feinberg 1989, pp.118/1190)

Bioethicists Brock and Buchanan (1989, p.45ff.) who accuse Feinberg of hidden,
though unwanted paternalism, have famously criticized his argument. In a nutshell
they argue that any RRCC-policy will have two simultaneous effects: it will (i) pro-
tect some patients from the harm of risky, though incompetent decisions (the wanted
effect), but it will also (ii) violate the autonomy of some other patients who falsely
get assessed as incompetent (the unwanted, but accepted side effect). Accepting (ii)
170 B. Schoene-Seifert

as the moral price for (i), so Brock and Buchanan’s argument, can only be justified
from a position that ultimately ranks the reduction of harm by effect (i) morally
higher than the violation of autonomy by effect (ii) – a (paternalist) position that
they themselves favor anyhow and explicitly.
On closer analysis,28 Feinberg’s suggestion cannot be as easily dismissed as
Brock/Buchanan have it, but it seems unconvincing, nevertheless. A Feinberg-style
strategy can indeed – and against Brock/Buchanan – be reconstructed as compatible
with giving first priority to the protection of autonomy over harm. Hence, with an
eye on policy-making the strategy might be justified in contexts with genuine uncer-
tainty about the nature of the required competence and with a sufficiently high prob-
ability that severe harm from incompetent decision-making can indeed be prevented
by effect (ii). From an ex ante view of uncertainty, a high likelihood of non-­
autonomously chosen harm can then outweigh a smaller likelihood of violating
autonomy although respect for autonomy gets clear lexical priority over the preven-
tion of autonomously self-afflicted harm.29 Ex post, an analogous balance would
hold true for collectives of persons so treated: many cases of prevented non-­
autonomously chosen harm might outweigh rare cases of violated autonomy. Given
that the latter cannot be identified in advance, even proponents of inviolable auton-
omy rights might swallow this.30 In such contexts, a policy of raising standards (and
tests) might be justified on clearly autonomy-oriented grounds. These grounds, that
is, do not require neglecting the prevention of harm, especially if unauthorized, in
the moral calculus, only to rank it second place.31
But again, granting this presupposes circumstances of genuine conceptual uncer-
tainty and appropriate proportionality. Both conditions seem to be missing in the
context of medical decision-making. Here we know that ordinary persons’ compe-
tence is sufficient and that requiring more than this will thus always result in viola-
tion of autonomous authorization and often in the prevention of ‘harm’ that, given
the choice-set available to them, patients themselves assess as minor evils. Therefore,
RRCC does not seem justifiable on fallibilistic grounds.

28
I find myself in some agreement with and owe much to ethicist Joseph Demarco (2002) who first
drew my attention to this particular debate.
29
Compare Demarco 2002, p.236ff.
30
Systematically, Feinberg’s position falls under what Fateh-Moghadan and Gutmann have labeled
the “endangerment”-variety of soft paternalism. They, too, rightly stress the importance of “pro-
portionality” in justifying such policies and their inherent danger of hidden hard paternalism
(Fateh-Moghadam and Gutmann 2014, p.392ff.).
31
This is what Buchanan/Brock seem, however, to require from a non-paternalist position: com-
pare Buchanan and Brock 1989, p.41ff.
10 Patients’ Decision-Making Competence: Discontents with a Risk-Relative Conception 171

10.5.2 Over-Exclusiveness: Hidden Paternalism

From a critical perspective, favoring RRCC can have both over-exclusive and over-­
inclusive motivations and consequences. Thus, it aims at (i) protecting patients,
even if they might be competent according to the ordinary-person standard, against
self-regarding decisions that from outside are deemed gravely harmful. And it is
ready to (ii) grant decision-making authority to patients far below that standard as
long as their decisions involve low-risk issues. Whether someone’s pro-RRCC intu-
itions are more in line with (i) or with (ii), the two sides of the coin might well
coincide as two mutually supportive lines of argument. I will briefly discuss the first
one in this section and the second one thereafter.
Anti-paternalism holds that a competent patient’s voluntary and informed deci-
sion should never be overridden so as to protect him from harm. In this paper there
is no room to discuss paternalism and anti-paternalism in their own right. However,
as several critics have warned, introducing RRCC turns
[…] the statement that a competent patient’s decision may not be overridden for paternalis-
tic reasons [into] a tautology rather than an expression of an important liberal-democratic
principle. (Wicclair 1999, p. 150)

This conclusion is correct because RRCC is masking whether a patient’s decision

has been overridden due to ‘real’ incompetence or rather with an eye on impeding
self-afflicted harm. This is especially true since RRCC-proponents do not (and
probably cannot) give full particulars about what exactly they require for sufficient
competence in cases of (seemingly) hazardous decisions. Rather, it seems conceiv-
able that patients are never viewed sufficiently competent to make certain decisions
(e.g. suicide).
Paternalists, of course, would not much mind and have explicitly suggested giv-
ing up the traditional distinction between the criteria of competence and harm (e.g.,
Hartogh 2016). But I can only confirm DeMarco’s (2002, p. 244) testimony:
“Autonomy is too important, and paternalism is too controversial to be buried in a
concept of competence.”

10.5.3 Over-Inclusiveness: Devaluating Authorization

Coming to the other end of the competence scale, including incompetent patients
into the sphere of personal respect seems to be a major motivation behind
RRCC. Many of the underlying concerns can be easily understood: judging a patient
incompetent can hurt or stigmatize her. As emphasized for instance by Vandeveer
(2014, p.415), it might even function as a ‘self-fulfilling prophecy’. Moreover, care-
takers might easily forget that the incompetent should be respected as choosers
whenever possible, thus enhancing their preference satisfaction and their self-­
esteem. It is a sad and completely unnecessary story that the attribution of
172 B. Schoene-Seifert

incompetence should lead caregivers to forget their duties of benefiting patients by

following their self-regarding choices whenever possible without afflicting
major harm.
This might be all the truer, as over-inclusion via RRCC would be limited to those
decisions that are deemed low-risk/high-benefit in the first place. Hence, only as-if-­
authority or illusory autonomy is granted to those patients. For any harmful deci-
sion physicians would jump in and take over.
But even then, three serious objections have to be made against the attribution of
illusory (or pseudo-) autonomy. First, it confounds reasonable concepts of compe-
tence, authority, and autonomy, possibly inviting caretakers to lose respect for ‘real’
autonomy. Secondly, by circumventing the need for transparent surrogate decision-­
making procedures, RRCC re-opens a back door to paternalism in medicine. And
thirdly, respectful over-inclusion might fuel the belief - wrong in my eyes - that
severely demented patients can validly revoke their former advance directives. This
is a contested issue that cannot be explicated here. Suffice it to hint at the logical
connection between RRCC and the ‘past-directive-versus-present-interest conflict’
in severely demented patients.32
Thus, granting pseudo-competence to an incompetent patient might look respect-
ful. Explicitly following her wishes in matters that involve only low risk and low
harm as well as following her whenever she agrees to a recommended low-risk
option surely is the right thing to do. However, doing this under the label of AA
intentionally obscures the locus of responsibility and ultimate decision-making
(likewise, Wicclair 1991, p.101).

10.6 Conclusions

In this paper, I have argued in favor of a fixed concept of decisional competence for
patients in the context of giving their informed consent to medical treatments.
Although its alternative, a risk-relative conception of competence (RRCC), might
well have become the governing conception among clinicians and clinical ethicists
it runs into fundamental problems. As ethicist DeMarco (2002, p.232) put it about
fifteen years ago: “[…] including risk in the definition of ‘competence’, or even
raising the standard of competence due to risk, is a fundamental mistake that
obscures proper decision making in bioethics.”
I have tried to reaffirm this view by reviewing well-known RRCC-objections
from symmetry problems and the dangers of hidden paternalism towards patients
with uncompromised competence. Moreover, I have argued that RRCC miscon-
ceives the specific task behind patient competence, i.e. optionally making medical
decisions with an eye on one’s shape of life within the boundaries of being ordinary
women and men. Finally, by attributing illusory autonomy to clearly incompetent

32
Compare Dworkin 1986; Dresser 1995; Jox et al. 2012; Schoene-Seifert et al. 2016.
10 Patients’ Decision-Making Competence: Discontents with a Risk-Relative Conception 173

patients, RRCC not only obscures the need for surrogate decision making, but it
also implies conceptual de-differentiations that turn out highly problematic in other
areas of bioethical dispute, e.g. about honoring advance directives against severely
demented patients’ expressed behavior.

References

Alexander, Larry. 1996. The Moral Magic of Consent (II). Legal Theory 2 (3): 165–174.
Amelung, Knut. 1992. Über die Einwilligungsfähigkeit (Teil I). Zeitschrift für die gesamte
Strafrechtswissenschaft 3: 525–558.
Beauchamp, Tom L. 1991. Competence. In Competency. A Study of Informal Competency
Determinations in Primary Care, ed. Mary Ann Gardell Cutter and Earl E. Shelp, 49–77.
Dordrecht: Springer.
Beauchamp, Tom L., and James F. Childress. 2019. Principles of Biomedical Ethics. Vol. 8.
New York/Oxford: Oxford University Press.
Bromwich, Danielle, and Annette Rid. 2015. Can Informed Consent to Research Be Adapted to
Risk? Journal of Medical Ethics 41 (7): 521–528.
Buchanan, Allen E., and Dan W. Brock. 1989. Deciding for Others: The Ethics of Surrogate
Decision Making. Cambridge: Cambridge University Press.
Cale, Gita S. 1999. Continuing the Debate over Risk-Related Standards of Competence. Bioethics
13 (2): 131–148.
Charland, Louis C. 2015. Decision-Making Capacity. In The Stanford Encyclopedia of Philosophy,
ed. Edward N. Zalta, https://plato.stanford.edu/archives/fall2015/entries/decision-­capacity/.
Checkland, David. 2001. On Risk and Decisional Capacity. Journal of Medicine and Philosophy
26 (1): 35–59.
Culver, Charles M., and Bernard Gert. 1990. The Inadequacy of Incompetence. The Milbank
Quarterly: 619–643.
DeMarco, Joseph P. 2002. Competence and Paternalism. Bioethics 16 (3): 231–245.
Drane, James F. 1985. The Many Faces of Competency. Hastings Center Report 15 (2): 17–21.
Dresser, Rebecca S. 1995. Dworkin on Dementia. Elegant Theory, Questionable Policy. Hastings
Center Report 25 (6): 32–38.
Dworkin, Ronald. 1986. Autonomy and the Demented Self. The Milbank Quarterly 64 (Suppl.
2): 4–16.
Faden, Ruth R., and Tom L. Beauchamp. 1986. A History and Theory of Informed Consent.
New York/Oxford: Oxford University Press.
Fateh-Moghadam, Bijan, and Thomas Gutmann. 2014. Governing [through] Autonomy. The Moral
and Legal Limits of ‘Soft Paternalism’. Ethical Theory and Moral Practice 17 (3): 383–397.
Feinberg, Joel. 1989. Moral Limits of the Criminal Law: Harm to Self. New York/Oxford: Oxford
University Press.
Ganzini, Linda, Ladislav Volicer, William A. Nelson, Ellen Fox, and Arthur R. Derse. 2004.
Ten Myths about Decision-Making Capacity. Journal of the American Medical Directors
Association 5 (4): 263–267.
Gaylin, Willard. 1982, April 1. The Competence of Children: No Longer All or None. Hastings
Center Report 12 (2): 33–38. https://doi.org/10.2307/3561805.
Gigerenzer, Gerd, Wolfgang Gaissmaier, Elke Kurz-Milcke, Lisa M. Schwartz, and Steven
Woloshin. 2007. Helping Doctors and Patients Make Sense of Health Statistics. Psychological
Science in the Public Interest 8 (2): 53–96.
Hanna, Jason. 2011. Paternalism and Impairment. Social Theory and Practice 37 (3): 434–460.
Hartogh, Govert den. 2016. Do We Need a Threshold Conception of Competence? Medicine,
Health Care and Philosophy 19: 71–83. https://doi.org/10.1007/s11019-015-9646-5.
174 B. Schoene-Seifert

Hermann, Helena, Manuel Trachsel, and Nikola Biller-Andorno. 2016. Decision-Making Capacity:
Inherent Ability or Ethical Judgment? Ethik in der Medizin 28 (2): 107–120
Jox, Ralf J., et al. 2012. Surrogate Decision Making for Patients with End-Stage Dementia.
International Journal of Geriatric Psychiatry 27 (10): 1045–1052.
Macklin, Ruth. 1987. Mortal Choices: Ethical Dilemmas in Modern Medicine. Boston: Houghton
Mifflin.
Manson, Neil C., and Onora O’Neill. 2007. Rethinking Informed Consent in Bioethics. Cambridge:
Cambridge University Press.
Miller, Franklin G., and Alan Wertheimer. 2011. The Fair Transaction Model of Informed Consent:
An Alternative to Autonomous Authorization. Kennedy Institute of Ethics Journal 21 (3):
201–218.
Quante, Michael. 2011. In Defence of Personal Autonomy. Journal of Medical Ethics 37 (10):
597–600.
Roth, Loren H., Alan Meisel, and Charles W. Lidz. 1977. Tests of Competency to Consent to
Treatment. The American Journal of Psychiatry 134 (3): 279–284.
Schoene-Seifert, Bettina, Anna Lena Uerpmann, J. Jochaim Gerss, and David Herr. 2016.
Advance (Meta-) Directives for Patients with Dementia who Appear Content: Learning from a
Nationwide Survey. Journal of the American Medical Directors Association 17 (4): 294–299.
Vandeveer, Donald. 2014. Paternalistic Intervention: The Moral Bounds on Benevolence.
Princeton: Princeton University Press. https://doi.org/10.1515/9781400854066.
Silver, Mitchell. 2002. Reflections On Determining Competency. Bioethics 16 (5): 455–468.
Wicclair, Mark R. 1991. Patient Decision-Making Capacity and Risk. Bioethics 5 (2): 91–104.
———. 1999. The Continuing Debate over Risk-Related Standards of Competence. Bioethics 13
(2): 149–153.
Wilks, Ian. 1997. The Debate over Risk-Related Standards of Competence. Bioethics 11 (5):
413–426.
———. 1999. Asymmetrical Competence. Bioethics 13 (2): 154–159.
Chapter 11
Vulnerability, Exploitation and Autonomy

Catriona Mackenzie

Abstract Bioethicists who seek to defend commercial transactions that intuitively

seem exploitative, such as organ sales and commercial surrogacy, typically pair a
liberal analysis of exploitation with a libertarian analysis of autonomy. In this paper,
I argue that the liberal analysis of exploitation, which focuses primarily on two
party transactions between individuals, occludes the structural dimensions of
exploitation. This occlusion then paves the way for the transaction to be understood
in terms of libertarian autonomy. I propose that a vulnerability analysis paired with
a multidimensional and relational theory of autonomy highlights the structural
dimensions of exploitation and exposes the myth-making that underlies claims that
these forms of exploitation are autonomy-enhancing.

11.1 Introduction

Over the last two decades a growing body of ethnographic research has sought to
investigate and give voice to the lived experience of people who have been recruited
into the global trade in body parts and reproductive services. Ethnographic research
with kidney sellers in Pakistan, Bangladesh, Egypt, Iran, Brazil and the Philippines,
amongst other countries, and with gestational surrogates in India, has helped to shed
light on these people’s socio-economic plight and family situations, their motiva-
tions, and the recruitment tactics used by brokers.1 This research has also painted a
disturbing picture of the circumstances in which they undergo surgery or perform
reproductive labour, and of their lives in the aftermath of their decisions.

1
For examples of the ethnographic research conducted with organ sellers in a range of different
countries, see Scheper-Hughes (2000, 2008), Moazam et al. (2009), Moniruzzaman (2012).
Ethnographic research conducted with surrogates in India is reported in Pande (2010).

C. Mackenzie (*)
Macquarie University, Sydney, Australia
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 175

Switzerland AG 2022
J. F. Childress, M. Quante (eds.), Thick (Concepts of) Autonomy, Philosophical
Studies Series 146, https://doi.org/10.1007/978-3-030-80991-1_11
176 C. Mackenzie

Invariably, people recruited as kidney sellers or gestational surrogates are desper-

ately poor, illiterate, indebted, suffer poor nutrition, live in inadequate housing, and
have extremely limited options for what they can be and do. The options of surro-
gates, in addition, are constrained by entrenched patriarchal gender norms. To use the
terminology of Jonathon Wolff and Avner de Shalit (2007), those recruited could be
characterised as experiencing extreme insecurity of functioning across a range of dif-
ferent functionings. Moreover, especially in the case of kidney sellers, the ethno-
graphic research provides evidence that the vast majority of those recruited regret
their decision and subsequently end up being worse off financially, as well as experi-
encing significant physical, psychological and social suffering.2 This research there-
fore suggests that the global trade in body parts and reproductive services is grossly
exploitative of extremely disadvantaged and vulnerable persons; it is a noxious market.
Over the same period in which the results of this ethnographic research have been
published, there has been a chorus of voices in bioethics arguing in favour of regu-
lated global markets in reproductive services and in nonessential or renewable body
parts.3 Defenders of such markets typically pair a liberal, transactional analysis of
exploitation with a libertarian analysis of autonomy. Liberal, transactional accounts,
which Jeremy Snyder (2012) refers to as micro fairness accounts, analyse exploita-
tion in terms of impaired consent and/or an unfair bargain, contract or offer.4 On these
views, what makes a specific transaction unfair, and therefore potentially exploitative,
is when one party A gains more, or the other party B gains less, than they should from
the transaction. The standard of fairness is what would be achieved under hypotheti-
cal ideal fair market conditions. To show that organ selling or commercial gestational
surrogacy is exploitative thus requires showing that the consent of the seller or sur-
rogate is impaired, or that the bargain or contract is unfair in the sense that the seller
or surrogate does not receive a fair share of the benefits of the transaction. It is impor-
tant to note that the fairness or otherwise of the background social and economic
conditions in which the transaction occurs is bracketed from consideration.
Libertarian conceptions of autonomy equate personal autonomy with negative
liberty and hold that to be autonomous is to be free from unwarranted external inter-
ference by other persons or the state with respect to important decisions about one’s
life. In bioethics, autonomy theorists of a more libertarian bent operate with a mini-
malist conception of autonomous decision-making as voluntary and informed
choice. They hold that autonomy is best promoted by maximising individual free-
dom of choice and minimising regulatory and other kinds of constraints on choice.5
Pairing the liberal, transactional analysis of exploitation with a libertarian con-
ception of autonomy, the defense of organ markets, for example, typically goes as

2
See Koplin (2014) for a summary of the ethnographic research documenting regret in relation to
the decision to sell a kidney.
3
See e.g. Fabre (2006), Radcliffe-Richard (2007), Taylor (2002), Wilkinson (2003).
4
Wertheimer (1996) is the classic source for the liberal, transactional account. See also
Wertheimer (2011).
5
In Mackenzie (2015), I provide a more detailed discussion and critique of libertarian conceptions
of autonomy in bioethics.
11 Vulnerability, Exploitation and Autonomy 177

follows. Individuals have an autonomy right to use their bodies however they
choose, including selling body parts for material gain. If a seller consents to the sale
of her body parts in full knowledge of the potential risks involved, and perceives the
financial offer as advantageous and as likely to make her better off relative to her
current situation, then the seller has a right to enter into the transaction even if it
appears exploitative. This is a case of what Wertheimer (1996, 2011) would call
‘mutually advantageous exploitation’. Commercial markets in organs therefore pro-
vide prospective sellers with autonomy-enhancing options and prohibiting such
sales or other high-risk opportunities for improving their material situation is unjus-
tifiably paternalistic.
My aim in this paper is to argue that the conceptions of autonomy and exploita-
tion assumed in this kind of argument occlude the structural inequalities and injus-
tices that must be central to any adequate assessment of whether so-called mutually
advantageous exploitation is autonomy-enhancing.6 My argument proceeds in three
steps. In the first step, I propose that the concepts of situational and pathogenic
vulnerability (Mackenzie, Rogers, Dodds 2014) and corrosive disadvantage (Wolff
and de Shalit 2007) can help to clarify the sense in which people who make desper-
ate choices to sell kidneys or become gestational surrogates are vulnerable to
exploitation. In the second step, I draw on Snyder’s (2012) taxonomy of exploita-
tion and Vida Panitch’s (2013, 2017) work on transnational commercial surrogacy
to show why the transactional or micro fairness account of exploitation is insuffi-
cient for addressing the moral question of whether markets in organs and reproduc-
tive services are exploitative. In the third step, I propose that a multidimensional and
relational analysis of autonomy (Mackenzie 2014) illuminates the different ways
that autonomy can be compromised by disadvantage, vulnerability and social
oppression. This analysis exposes what I regard as the myth-making in the claim
that mutually advantageous exploitation can be autonomy-enhancing. I also endorse
a suggestion by Malmqvist (2014), that a group soft paternalist approach to markets
in organs and reproductive services might be justified to avoid harms to individuals
whose autonomy is likely compromised by disadvantage.
A clarification may be useful at this point. My aim in the paper is to show why
the conceptions of exploitation and autonomy assumed by defenders of markets in
organs and reproductive services are inadequate. I do not offer a defense of either
prohibition or regulation with respect to such markets, although my argument might
point in the direction of prohibition as an appropriate policy response. However,
since organ selling and transnational gestational surrogacy are distinct practices, the
merits of prohibition versus regulation need to be debated separately in each case
and with reference to the specificities of the relevant jurisdictional contexts.7 In

6
The importance of attending to structural injustice for understanding exploitation in a range of
practical contexts, including gestational surrogacy and sweatshop labour, is a central theme of
many of the essays in Deveaux and Panitch (2017).
7
In bioethics, the literatures on organ selling and the transnational market in reproductive services
are quite distinct. While I acknowledge that these different kinds of market raise distinct moral and
policy issues, my aim in discussing them together in this paper is to highlight the structural com-
178 C. Mackenzie

addition, as Serene Khader (2013) argues, any policy recommendations need to be

attentive to an intersectional and context-sensitive analysis of the specific kinds of
class or gender harms that might be caused either by such markets or by different
kinds of regulatory intervention or prohibition.

11.2 Vulnerability and Corrosive Disadvantage

To exploit someone is to take advantage of her vulnerability. How then should vul-
nerability be analysed? To be vulnerable is to be at increased risk of suffering harm
or being wronged.8 This risk might be potential or occurrent. Hence efforts to miti-
gate or alleviate a specific vulnerability will need to attend to whether the risk of the
person being harmed or wronged is potential or occurrent. For example, poverty can
make a person vulnerable to homelessness. If she has become unemployed and has
no income, is behind on her rent, and is about to be evicted from her home, then she
is occurrently vulnerable to homelessness. If she has some form of precarious
employment, but her income is insecure and she struggles to pay the rent and keep
up with her bills, then while she is not occurrently vulnerable to homelessness, she
is potentially vulnerable to homelessness.
In bioethics there has been considerable debate between two approaches to vul-
nerability: ontological and relational or context-specific.9 The taxonomy of different
sources of vulnerability that I have developed in earlier work with Wendy Rogers
and Susan Dodds, aims to reconcile these two approaches. The example of home-
lessness is useful for helping to clarify this taxonomy, which distinguishes inherent,
situational and pathogenic sources of vulnerability.
The notion of inherent vulnerability acknowledges that vulnerability is an onto-
logical condition of our humanity. We are all inherently vulnerable because we are
finite, embodied and social creatures with needs. And it is by virtue of this inherent
neediness that we are vulnerable to harm (physical, psychological, emotional) and
to being wronged by others. Using the example of homelessness, because one of our
basic needs is the need for adequate housing or shelter, all human beings are inher-
ently vulnerable to homelessness.

monalities in the types of arguments used by defenders of transnational markets in body parts and
reproductive services.
8
Hurst (2008) analyses vulnerability in terms of increased risk of being subject to wrongful harms.
Her reasons for restricting the analysis of vulnerability to wrongful harms rather than harms more
broadly are twofold. First, because she is unsympathetic to ontological accounts of vulnerability;
and second, to make it easier to identify which duties are required to remediate specific vulnerabili-
ties and on whom those duties fall. However, an implication of her view is that persons cannot be
characterised as vulnerable unless their increased risk of harm arises from wrongful actions on the
part of others. Thus, persons at increased risk of suffering harm as a result of accidents, natural
catastrophes, or illness would not count as vulnerable on this view. This seems counter-intuitive to
me and inconsistent with the way the concept of vulnerability is employed in everyday usage.
9
For more detailed discussion of this debate and of the taxonomy discussed below, see Rogers,
Mackenzie, Dodds (2012) and Mackenzie, Rogers, Dodds (2014).
11 Vulnerability, Exploitation and Autonomy 179

However, clearly we are not all equally vulnerable to homelessness. Further,

what counts as adequate housing varies according to context. The notion of situa-
tional vulnerability draws attention to the context-specific factors that shape the
ways in which vulnerability might be manifested and that might render a person or
social group especially vulnerable, in this case to homelessness. Such factors are
many and varied and may include environmental conditions, social norms and prac-
tices, institutional structures (political, economic, legal), as well as the individual’s
specific situation or needs. A person who is mobility impaired, for example, and
who needs wheelchair accessible housing may be more vulnerable to homelessness
than a person who can easily climb several flights of stairs.
The notion of pathogenic vulnerability refers to a sub-set of situational vulnera-
bilities that are caused by morally dysfunctional, oppressive or dominating social
relationships, or by remediable structural inequality and injustice. We also use the
term to refer to vulnerabilities that are caused by (sometimes well-intentioned)
interventions to remediate vulnerability. A young teenager who was being sexually
abused at home by a relative and is now homeless, experiences pathogenic vulner-
ability. This vulnerability may be further compounded by an intervention designed
to remediate that vulnerability, such as housing in a shelter for young people if, for
example, she develops a drug addiction or is exposed to further abuse and violence
while living in the shelter. This example shows that the distinction between inher-
ent, situational and pathogenic vulnerability is not categorical and that these sources
of vulnerability can and often do overlap and interact in complex ways.
The work of Wolff and de Shalit (2007) on corrosive disadvantage is helpful for
clarifying the complex interactions between different sources of vulnerability. Wolff
and de Shalit analyse disadvantage in terms of insecurity of functioning across a
range of functionings.10 People who are disadvantaged experience not only insecu-
rity of functioning, but also lack genuine opportunities for secure functionings.
Moreover, this insecurity and lack of genuine opportunity typically affects not just
one functioning but a clustering of functionings, with cumulative effects over time.
An example of such clustering is the way over time unemployment can lead to
homelessness, and then loss of friendship networks, then ill health (Wolff and de
Shalit 2007, 120). The concept of “corrosive disadvantage” refers to the way in
which poor or insecure functioning in one domain can cause poor or insecure func-
tionings in other domains, as in this example. Wolff and de Shalit further argue that
disadvantage typically exposes people to risks that those who are advantaged do not
face and that taking these risks can generate corrosive disadvantage when the risks
taken in an attempt to secure one functioning then jeopardises other functionings.
They refer to this kind of risk taking as “inverse cross category risk” (70).
Earlier I defined vulnerability as being at increased risk of suffering harm or
being wronged. In contexts where vulnerability is associated with disadvantage, it
could also be understood as being at increased risk of insecurity of functionings,

10
Wolff and de Shalit distil Nussbaum’s list of ten capabilities into six central functionings: life,
bodily health, bodily integrity, affiliation, control over one’s environment, and sense, imagination
and thought (2007, 106).
180 C. Mackenzie

where the source of that insecurity is likely to be some combination of inherent,

situational and pathogenic vulnerability. In the example of the dynamic clustering
over time of unemployment, homelessness, loss of social networks and ill health,
the initial situational vulnerability of the unemployed person exposes her to patho-
genic forms of vulnerability (homelessness and social isolation), which then
increases her inherent vulnerability (illness).
What is abundantly evident from the ethnographic research is that those people
who are initially recruited to sell kidneys or to labour as gestational surrogates are
pathogenically vulnerable, experience significant insecurity of functionings and
lack genuine opportunities for functionings. To describe these people as pathogeni-
cally vulnerable is not to deny their agency. It is rather to characterise the structur-
ally unjust and institutionally dysfunctional circumstances in which they exercise
agency, circumstances which tempt them to take risks that they would not take if
they had genuine alternative opportunities for functioning. In the case of kidney
sellers, the risky choice of surgery then exposes them to significant inverse cross-­
category risk over time: ill health, loss of livelihood, social exclusion, psychological
suffering, and deepened financial insecurity.11
While defenders of markets in organs acknowledge that, even in a regulated mar-
ket, those recruited to sell their organs will continue to be the desperately poor, they
argue that a regulated market would mitigate some of these additional risks. Whether
or not this might be the case is an empirical question. The evidence from Iran (where
there has been a regulated market in organs since 1988) suggests otherwise.12
However, in light of the analysis of vulnerability and disadvantage outlined in this
section, it seems fanciful to claim that the availability of the option to sell a kidney
or labour as a gestational surrogate is autonomy-enhancing. I return to this issue in
section 4. In the next section I focus on the question of exploitation.

11.3 Exploitation and Injustice

I noted earlier that exploitation involves taking advantage of another’s vulnerability.

Vulnerability therefore makes a person potentially exploitable. How should exploi-
tation be understood in the context of global markets in organs and reproductive
services? In answering this question I draw on Snyder’s taxonomy of three forms of
exploitation (2008, 2012) and Panitch’s (2013, 2017) analysis of exploitation in
transnational surrogacy contracts.
In a discussion of exploitation in human subject trials of pharmaceuticals, Snyder
distinguishes between micro fairness, macro fairness and mere use or needs

11
For surrogates, who in the Indian context are typically housed in hostels for the duration of preg-
nancy, the risks are different: loss of affiliation with family and children, loss of control over their
lives for the duration of the pregnancy, and the requirement to emotionally distance themselves
from the child they are gestating (for more detailed discussion see Pande 2010, Khader 2013).
12
For a summary of the ethnographic research with Iranian kidney sellers, see Koplin (2014).
11 Vulnerability, Exploitation and Autonomy 181

exploitation. He argues that these different forms of exploitation can coexist, but
each is tied to different vulnerabilities and different forms of wrongdoing (2012,
252). Micro fairness exploitation occurs when asymmetries in the bargaining posi-
tion of the parties enables the more advantaged party to skew the benefits of the
transaction unfairly in their favour. As I mentioned earlier, some advocates of micro
fairness accounts do not think that all such transactional unfairness is wrongful,
however. In cases of so-called mutually advantageous exploitation, if the party with
less bargaining power nevertheless benefits relative to the position she was in before
the transaction, and is capable of consenting to it, then this kind of transaction need
not be wrongful. This is the view of exploitation assumed by defenders of markets
in organs and gestational surrogacy.
Macro fairness accounts of exploitation object that micro fairness accounts fail
to attend to long-term and systemic asymmetries in bargaining power arising from
background inequalities and injustices. However, these kinds of long-term and sys-
temic asymmetries create vulnerabilities that should be taken into account in assess-
ing the fairness or otherwise of transactions, and the consent of the disadvantaged
party. Further, they give rise to duties on the part of the more advantaged to mitigate
the effects of injustice and to reform unjust institutions (Snyder 2012, 256).
Macro fairness accounts are likely to be much more sceptical than micro fairness
accounts of the claim that mutually advantageous exploitation is permissible.
Panitch (2013, 2017), for example, argues that assessments of whether or not so-­
called mutually advantageous transnational surrogacy contracts are exploitative
must attend to background structural injustices that determine the bargaining posi-
tions of the parties. She argues that, given these injustices, it is misleading to focus
solely on the share of the benefits to the parties to the contract – which is difficult to
do in any case, given the incommensurability of the goods or benefits that accrue to
the parties in a surrogacy contract. Rather, “underpayment, or unfairness of benefit
sharing, must be assessed not just relative to what the other party to the contract
gains but relative to what other people performing the same type of work in other
parts of the world receive as compensation” (2103, 332). An inter-contractual com-
parison between an Indian surrogate and her US counterpart with respect to pay-
ment as well as other benefits makes it clear that the Indian surrogate’s vulnerability
is being exploited. Panitch lists benefits such as “freedom to pursue other interests
while under contract, health care, travel and dietary expenses, legal representation,
a post-birth opt-out clause, and the potential for a rewarding relationship with the
adoptive family” (333).
Additionally, the ethnographic research provides evidence that Indian surrogates
are not able to negotiate the terms of the contract. In fact, they are explicitly coun-
selled against attempting to negotiate better terms, and warned that they are replace-
able by another woman who would be willing to accept the terms (Pande 2010).
They are replaceable precisely because there are so many other poor and desperate
women; that is, women who are vulnerable, in both situational and pathogenic
senses and who experience insecurity across a range of functionings. Panitch argues
that this inability to negotiate the terms of the contract is morally equivalent to an
inability to refuse; hence their consent is not genuine or authoritative. On grounds
182 C. Mackenzie

of both inter-contractual unfairness of benefit and failure of genuine consent, the

claim that these contracts are not wrongful cannot be sustained.
Snyder’s third form of exploitation, “mere use exploitation”, is a failure of the
Kantian duty of beneficence. Elsewhere, in the context of a discussion of sweatshop
labour Snyder terms this form of exploitation “needs exploitation”: “Needs
Exploitation takes place when an exploiter gains advantage from an exploitee while
disregarding shortfalls in the basic needs of the exploitee which the exploiter has a
duty to meet” (2008, 390). In other words, rather than acting from the duty of benef-
icence, which requires us to reduce the other’s vulnerability, in needs exploitation
one takes advantage of the other’s vulnerability and disregards her basic needs.
Basic needs here refer to the different dimensions of well-being required to lead a
minimally decent human life (395). This notion serves as a normative baseline for
what is owed to others who stand to us in particular relations of dependence; thus
for example, what is owed by parents to their children, employers to employees,
researchers to research participants, the state to its citizens.
Like the macro fairness account, the needs exploitation account will be sceptical
of the claim that so-called mutually advantageous exploitation is morally permissi-
ble. In the case of transnational markets in organs and reproductive services, even if
we assume that a specific transaction does not leave a person worse off than they
were before and even if it may marginally improve their situation, the transaction is
nevertheless exploitative if, due to background structural inequality and injustice,
the person’s living conditions are such that their basic needs are not met and will not
be met subsequent to the transaction. The question of on whom the duty of benefi-
cence falls in such cases is complex, but the relevant parties will include the state,
the medical profession, and organ buyers or prospective parents.

11.4 Autonomy and Paternalism

The argument so far has made a case that transnational markets in organs and repro-
ductive services exploit persons who are vulnerable and disadvantaged. What then
of the argument made by their defenders that such markets increase the option sets
of the desperately poor, thereby enhancing their autonomy, and that prohibitions or
restrictions on these markets constitute unwarranted paternalism? In this section of
the paper, I draw on a relational and multidimensional analysis of autonomy to show
why this kind of argument is flawed. I also endorse a suggestion by Malmqvist
(2014), that a group soft paternalist approach to commercial markets in organs (and
reproductive services) might be justified.
Elsewhere I have argued that a multidimensional and relational analysis of indi-
vidual or personal autonomy is helpful for clarifying some of the complex questions
about autonomy that arise in non-ideal and practical contexts (see e.g. Mackenzie
2014, 2017). Autonomy is both a capacity and a status concept. While my account
acknowledges the normative importance of individual autonomy, it aims to clarify
the relational dimensions of both the capacity for autonomy and of agents’
11 Vulnerability, Exploitation and Autonomy 183

autonomy status. In doing so, it makes several claims about the requirements for an
adequate conception of autonomy: i) it must be responsive to the facts of human
vulnerability and dependency, and the need for social relations of care; ii) it must be
premised on a conception of individual identity as embodied, and shaped by inter-
secting social, cultural, political and historical determinants; iii) it must explain the
social constitution and scaffolding of autonomy capacities; iv) it must explain how
social oppression, inequality and disadvantage can impair individuals’ capacities to
lead autonomous lives – to different degrees and in different ways in different con-
texts; v) it must explain the ways that our status as autonomous agents is vulnerable
to the character of our social relationships and the broader social environment;
hence how some persons with the capacity for autonomy may fail to be recognised
and treated by others as having the status of autonomous agents. The account thus
combines both internalist and externalist conditions for autonomy.
In my view the concept of autonomy encompasses three distinct but causally
interconnected dimensions or axes: self-determination, self-governance and self-­
authorisation (Mackenzie 2014, 2017). Each dimension should be understood as a
matter of degree and domain: a person can be self-determining, self-governing and
self-authorising to differing degrees, both at a time and over the course of her life.
To be self-determining is to have the authority and power to exercise both de jure
and de facto control over important domains of one’s life (cf. Oshana 2006). This
authority and power is a function of the freedoms and opportunities available to a
person in her social environment and of her social standing or status. The self-­
determination axis thus draws attention to externalist requirements for autonomy. In
my view, it is difficult to lead a self-determining life in contexts characterised by
social relations of domination and subordination, where a person’s access to rele-
vant freedoms and opportunities is blocked or restricted, and where her status as
autonomous agent is constrained by her position within the social hierarchy. I sub-
mit it is also difficult to lead a self-determining life in contexts of extreme disadvan-
tage, characterised by insecurity of functioning and a paucity of genuine opportunities
for secure functionings.
To be self-governing is to be able to make and enact decisions that express or
cohere with one’s practical identity, or one’s deeply held values and commitments.
This involves introspective and emotional skills for self-understanding and self-­
definition, rational and imaginative skills to envisage and chart a path between a
range of alternative courses of action, and self-control to enact and follow through
on one’s decisions.13 Hence, the self-governance axis draws attention to internalist
requirements for autonomy. However, in my view self-governance is a relationally
constituted capacity. The skills required for self-governance are developed and scaf-
folded (or not) by an agent’s social relationships and broader social environment.
Self-determination and self-governance, though conceptually distinct, thus interact
causally: the authentic self of self-governance, and the development and exercise of

13
For detailed analysis of the emotional and imaginative skills involved in self-governing agency,
see Meyers (1989) and Mackenzie (2000, 2002).
184 C. Mackenzie

the competences required for governing the self, are to a significant degree enabled
and/or constrained by external social conditions. This is why restricted freedoms,
lack of genuine opportunities for functionings, and socially subordinated status can
(but need not always) be internalised and manifest in adaptive preference formation,
restricted imaginative horizons, and constricted psychological freedom.
To be self-authorising is to regard oneself as having the normative authority to
take ownership of, or responsibility for, one’s values, decisions and one’s life over-
all. It also involves regarding oneself as an equal participant in reciprocal account-
ability relations – as able to account for oneself to others and also to hold others to
account. Thinking of oneself in this way involves holding appropriate attitudes of
self-respect, self-trust and self-esteem.14 These psychological attitudes are inher-
ently social, because they are developed and sustained through intersubjective social
relations and normative structures and practices of social recognition. One of the
insidious effects of social subordination, pathogenic vulnerability or disadvantage is
that persistent messages of social inferiority and unworthiness, embedded in every-
day social interactions or conveyed through demeaning stereotypes and social
scripts, can be internalised in feelings of shame and humiliation, of diminished self-­
respect, self-trust and self-esteem, and of social invisibility.
The ethnographic research indicates that many of the people recruited as organ
sellers or gestational surrogates appear to exercise compromised autonomy across
these different dimensions. Their severe impoverishment, vulnerability, insecurity
with respect to significant functionings, and lack of genuine opportunities for func-
tionings, compromises their capacities to lead self-determining lives. Domination
within patriarchal social and familial structures is an additional constraining factor
on surrogates’ capacities to lead self-determining lives overall, while their enclo-
sure in hostels for the duration of the pregnancy is an enforced restriction on their
freedom of movement and affiliation.
These factors, combined with illiteracy, may also compromise these people’s
capacities to lead self-governing lives. This is not to deny their capacities for agency,
nor is it to suggest that they do not know what is of value to them. A central theme
to emerge from the ethnographic research is that one of the main motivations for
entering into these transactions is people’s sense of obligation to provide and care
for their families, as well as the hope of leading a more self-determining life. My
argument is rather that illiteracy and insecurity of functioning may impair the devel-
opment and exercise of some of the skills involved in self-governing agency. The
regret expressed by many kidney sellers about their decision, documented in some
of the ethnographies, seems to support this claim.
Finally with respect to self-authorisation, a theme that emerges from the ethno-
graphic research with both kidney sellers and surrogates is that social inequality and
subordination compromises their sense of themselves as self-authorising agents and
renders them vulnerable to coercion, deception and corrupt systems of patronage.

14
For detailed arguments concerning the importance of these self-affective attitudes for autonomy,
see Anderson and Honneth (2005), Benson (1994, 2000), Govier (2003), and Mackenzie (2008).
11 Vulnerability, Exploitation and Autonomy 185

Surrogates, as discussed already, do not feel authorised to negotiate the conditions

of the transaction. Kidney sellers report being duped by unscrupulous brokers, ill-
treated by medical professionals and socially shamed and excluded following sur-
gery. Some clearly hope that by selling their kidney to a wealthy person they will
acquire a patron who will feel an obligation to provide for them.
In light of these considerations, which are informed by the ethnographic research,
the claim that markets in organs and reproductive services add an autonomy-­
enhancing option to desperate people’s choices cannot be sustained.15 To be
autonomy-­enhancing, in my view, options need to generate fertile functionings
(Wolff and de Shalit 2007, 138), functionings that improve people’s capacities to
lead more autonomous lives overall. This requires concerted efforts on the part of
governments and the international community to create genuine opportunities for
functionings, thereby beginning to address the structural injustices that create the
conditions in which noxious markets can take root.
In response to the objection that this argument is unjustifiably paternalistic, I
want to endorse Malmqvist’s (2014) appeal to group soft paternalism in defense of
a prohibition on kidney sales. Malmqvist argues that a group soft paternalist
approach can be justified as autonomy-promoting in situations where: (i) the option
to sell a kidney is potentially harmful to sellers; (ii) the autonomy of a significant
number of sellers is compromised; and (iii) it is not desirable or practically feasible
to make determinations about whether individual sellers do or do not satisfy the
conditions for autonomous choice. Thus, the argument is that even if some indi-
vidual prospective sellers might satisfy the conditions for autonomy, the justifica-
tion for group soft paternalism is to protect non-autonomous prospective sellers
from harms they have not autonomously chosen (118).

11.5 Conclusion

My aim in this paper has been to respond to the claim by defenders of transnational
markets in organs and reproductive services that these kinds of market transactions
constitute autonomy-enhancing and morally permissible forms of mutual exploita-
tion. I have argued that the claim that this form of mutual exploitation is morally
permissible cannot be sustained in light of a vulnerability-based approach to exploi-
tation that is informed by an analysis of disadvantage and that is attentive to struc-
tural inequality and injustice. Nor can the claim that such markets are
autonomy-enhancing be sustained. Although I have not argued in defense of either

15
See also Gupta, who argues that the language “of choice” in relation to commercial reproductive
services in India is “misused and misplaced”; it is a “myth” because women’s agency in this con-
text “cannot be isolated from other areas of life that condition women’s agency, such as education,
own income, vulnerable position within marital household, gender discrimination, and violence”
(2012, 46).
186 C. Mackenzie

prohibition or regulation, the considerations adduced in this paper support the view
that a group soft paternalist policy approach to such markets is morally justifiable.

References

Anderson, Joel, and Axel Honneth. 2005. Autonomy, Vulnerability, Recognition and Justice. In
Autonomy and the Challenges to Liberalism, ed. John Christman and Joel Anderson, 127–149.
Cambridge: Cambridge University Press.
Benson, Paul. 1994. Free Agency and Self-Worth. Journal of Philosophy 91: 650–668.
———. 2000. Feeling Crazy: Self-Worth and the Social Character of Responsibility. In Relational
Autonomy: Feminist Perspectives on Autonomy, Agency and the Social Self, ed. Catriona
Mackenzie and Natalie Stoljar, 72–93. New York: Oxford University Press.
Deveaux, Monique, and Vida Panitch, eds. 2017. Exploitation: From Practice to Theory. London:
Rowman and Littlefield International.
Fabre, Cecile. 2006. Whose Body Is It Anyway? Justice and the Integrity of the Person. Oxford:
Oxford University Press.
Govier, Trudy. 2003. Self-Trust, Autonomy, and Self-Esteem. Hypatia 8: 99–120.
Gupta, Jyotsna Agnihotri. 2012. Reproductive Biocrossings: Indian Egg Donors and Surrogates in
the Globalized Fertility Market. IJFAB 5 (1): 25–51.
Hurst, Samia A. 2008. Vulnerability in Research and Health Care; Describing the Elephant in the
Room? Bioethics 22 (4): 191–202.
Khader, Serene. 2013. Intersectionality and the Ethics of Transnational Commercial Surrogacy.
IJFAB 6 (1): 68–90.
Koplin, Julian. 2014. Assessing the Likely Harms to Kidney Vendors in Regulated Organ Markets.
The American Journal of Bioethics 14 (10): 7–18.
Mackenzie, Catriona. 2000. Imagining Oneself Otherwise. In Relational Autonomy: Feminist
Perspectives on Autonomy, Agency and the Social Self, ed. Catriona Mackenzie and Natalie
Stoljar, 124–150. New York: Oxford University Press.
———. 2002. Critical Reflection, Self-Knowledge and the Emotions. Philosophical Explorations:
An International Journal for the Philosophy of Mind and Action 5 (3): 186–206.
———. 2008. Relational Autonomy, Normative Authority and Perfectionism. Journal of Social
Philosophy 39: 512–533.
———. 2014. Three Dimensions of Autonomy: A Relational Analysis. In Autonomy, Oppression
and Gender, ed. Andrea Veltman and James Piper, 15–41. New York: Oxford University Press.
———. 2015. Autonomy. In Routledge Companion to Bioethics, ed. John Arras, Elizabeth Felton,
and Rebecca Kukla, 277–290. New York/London: Routledge.
———. 2017. Feminist Conceptions of Autonomy. In Routledge Companion to Feminist
Philosophy, ed. Ann Garry, Serene Khader, and Alison Stone, 515–527. New York/London:
Routledge.
Mackenzie, Catriona, Wendy Rogers, and Susan Dodds. 2014. What is vulnerability and why does
it matter for moral theory? In Vulnerability: New Essays in Ethics and Feminist Philosophy, ed.
Rogers Mackenzie, 1–29. New York: Oxford University Press.
Malmqvist, Erik. 2014. Are Bans on Kidney Sales Unjustifiably Paternalistic? Bioethics 28 (3):
110–118.
Meyers, Diana. 1989. Self, Society and Personal Choice. New York: Columbia University Press.
Moazam, Farhat, Riffat Moazam Zaman, and Aamir M. Jafarey. 2009. Conversations with Kidney
Vendors in Pakistan: An Ethnographic Study. Hastings Center Report 39 (3): 29–44.
Moniruzzaman, Monir. 2012. ‘Living Cadavers’ in Bangladesh: Bioviolence in the Human Organ
Bazaar. Medical Anthropology Quarterly 26 (1): 69–91.
Oshana, Marina. 2006. Personal Autonomy in Society. Aldershot: Ashgate.
11 Vulnerability, Exploitation and Autonomy 187

Pande, Amrita. 2010. Commercial Surrogacy in India: Manufacturing a Perfect Mother-Worker.

Signs: Journal of Women in Culture and Society 35 (4): 969–991.
Panitch, Vida. 2013. Global Surrogacy: Exploitation to Empowerment. Journal of Global Ethics
9 (3): 329–343.
———. 2017. Exploitation and Intimate Labor. In Exploitation: From Practice to Theory, ed.
Monique Deveaux and Vida Panitch, 119–138. London: Rowman and Littlefield International.
Radcliffe-Richards, Janet. 2007. Selling Organs, Gametes and Surrogacy Services. In The
Blackwell Guide to Medical Ethics, ed. R. Rhodes, L. Francis, and A. Silvers, 254–268.
Malden: Wiley-Blackwell.
Rogers, Wendy, Catriona Mackenzie, and Susan Dodds. 2012. Why Bioethics Needs a Concept of
Vulnerability. IJFAB 5 (2): 11–38.
Scheper-Hughes, Nancy. 2000. The Global Traffic in Human Organs. Current Anthropology 41
(2): 191–224.
———. 2008. Illegal Organ Trade: Global Justice and the Traffic in Human Organs. In Rainer
Gruessner and Enrico Beneditti, ed. Living Donor Organ Transplants, 106–121. New York:
McGraw-Hill.
Snyder, Jeremy. 2008. Needs Exploitation. Ethical Theory and Moral Practice 11 (4): 389–405.
———. 2012. Exploitations and their Complications: The Necessity of Identifying the Multiple
Forms of Exploitation in Pharmaceutical Trials. Bioethics 26 (5): 251–258.
Taylor, James Stacey. 2002. Autonomy, constraining options, and organ sales. Journal of Applied
Philosophy 19 (3): 273–285.
Wertheimer, Alan. 1996. Exploitation. Princeton: Princeton University Press.
———. 2011. Rethinking the Ethics of Clinical Research: Widening the Lens. New York: Oxford
University Press.
Wilkinson, Stephen. 2003. The Exploitation Argument Against Commercial Surrogacy. Bioethics
17 (2): 169–187.
Wolff, Jonathon, and Avner de Shalit. 2007. Disadvantage. Oxford: Oxford University Press.
Chapter 12
Outlook

James F. Childress and Michael Quante

Our introduction announced that the chapters in this volume deliver important con-
tributions to ongoing debates about the characterization of personal autonomy and
autonomous decisions and actions. Without exaggeration, we can claim that the
wide-ranging papers in this book illuminatingly address questions that have been
dealt with in many ethical theories for a long time and that elicit concern in numer-
ous social contexts and practices too.
It is also noteworthy that these contributions stem from a specialized philosophi-
cal context, namely, analytic philosophy, and that they mostly focus on problems
and questions related to biomedical ethics or bioethics (in a fairly broad sense) and
ethics more generally. In the last few decades, in modern western societies, respect
for autonomy has become the most important, or at least one of the most dominant,
ethical principles. In addressing (and countering) paternalism in many dis-guises,
our practices and social contexts have changed in many ways. Another effect is that
many ethical theories rely on thin conceptions of personal autonomy and some even
limit their focus to decisions or actions. Many authors presuppose that only mini-
mal, formal, or procedural conceptions of autonomy can adequately protect indi-
viduals from paternalism.
In the last three decades, concerns have emerged about the limitations of such
thin conceptions. The dialectic of this debate has led to the reemergence of “thick”
conceptions of various kinds. These make explicit the anthropological, biographi-
cal, and social aspects of our being human persons. Because autonomy and living
one’s life as a person are interrelated in various ways, these have enriched our

J. F. Childress
University of Virginia, Charlottesville, VA, USA
e-mail: [email protected]
M. Quante (*)
University of Münster, Münster, Germany
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 189

Switzerland AG 2022
J. F. Childress, M. Quante (eds.), Thick (Concepts of) Autonomy, Philosophical
Studies Series 146, https://doi.org/10.1007/978-3-030-80991-1_12
190 J. F. Childress and M. Quante

conceptions of autonomy. As the chapters in our volume demonstrate, this has

implications for the ethical requirement to respect autonomy as an ethical principle
on the one hand and within many social practices on the other.
Against this background, this volume offers a wide range of perspectives. Its
chapters explicate the central elements of and the many relations between “thin”
and “thick” concepts of autonomy and identify normative and social consequences
these different conceptions might (or actually do) have in our practices. These chap-
ters demonstrate the importance of these matters for ethics and biomedical ethics.
Nevertheless, it remains true that the contributions in our volume are located
within a specialized discourse in analytic practical philosophy and biomedical eth-
ics. No single volume can address all the contexts in which autonomy is rooted and
important or analyze all dimensions of autonomy—needless to say, such a task
would also be far beyond our competences. Here, we only briefly outline three con-
texts to open wider perspectives and to identify areas for further research and schol-
arly debate.

12.1 Other Philosophical Traditions

“Autonomy” is a term of art; therefore, we cannot use ordinary language or our

daily uses of the term as a source and corrective for our analysis. Moreover, the term
is rooted in different scholarly and scientific disciplines. This means that the con-
cepts and conceptions of autonomy are differently framed in different disciplines
(sometimes even in different schools or traditions within a single discipline).
Because “autonomy” is not a natural kind we have to accept that different concepts
or conceptions are meaningful in light of different epistemic interests or premises
(e.g. concerning methods).
Furthermore, the concept of autonomy has a long and complex history sedi-
mented in etymology. In ancient Greece, it was used mainly in the context of politi-
cal philosophy, where it referred to the right of a political entity (e.g. a polis) to rule
its inner affairs without external influence. In modern times, and especially in the
philosophy of Rousseau and Kant, the concept of autonomy was transferred from
political entities to human agents. Since then, it has stood for a human’s capacity for
self-determination and for the perfectionist ideal to realize this capacity both in
one’s individual life and in one’s social context. In the romantic tradition, autonomy
also became important in philosophical aesthetics, where it was enriched by such
concepts as originality, authenticity, and expressing one’s (real) self. Altogether,
autonomy became embedded in at least three philosophical language games:
• autonomy and responsibility (two main foci on human agency in ethics)
• autonomy and freedom (a main metaphysical theme in action theory)
• autonomy and personhood (a central relation in both theoretical and practical
philosophy)
12 Outlook 191

The volume of literature in analytic philosophy and in the broader western tradi-
tion is unmanageable in each of these three contexts. In each one, autonomy is
juxtaposed to specific conceptual ‘counterparts’ (e.g. autonomy vs. authenticity in
the third or autonomy vs. autarky in the second). In addition, the concept of auton-
omy is analyzed against the background of specific types of philosophy; for exam-
ple, in ethics, the explication is different within a deontological framework than
within a utilitarian meta-ethics. The chapters in this volume provide evidence for
this multi-layered and complex structure of the concept of autonomy in the western
tradition of philosophy and within the narrower tradition of analytic philosophy.
Therefore, in a wider perspective, the concept of autonomy should be analyzed
in relation to alternative traditions in analytic philosophy (e.g. the late Wittgenstein
and his followers) or in western philosophy (e.g. German idealism or pragmatism).
Some of the themes and topics developed there have left their traces in contributions
to this volume that try to overcome weaknesses of thin conceptions of autonomy
(e.g. Christman or Mackenzie) by bringing into view the social dimension of agency
and personhood.
In an even wider perspective of intercultural philosophy, the western tradition (of
which this book only covers a small part) needs to be compared and contrasted with
other conceptions of agency and the self (e.g. in Asian or African traditions). This
would be relevant not only for a comprehensive analysis of concepts of autonomy
in all philosophical traditions. It will also be practically important in all pluralistic
social contexts in which respect for autonomy functions as a normative resource.
For example, the more biomedical ethics deals with diverse and multicultural set-
tings, the more urgently it will need to understand sociocultural differences in inter-
pretations of autonomy.

12.2 Changing Environments

Our societies are characterized by dynamic changes in technologies and social

structures. In both contexts, the concept of autonomy and the principle of respect for
autonomy need to be rethought and further developed. In an aging society, living
wills become an important tool to extend one’s autonomy along with diverse artifi-
cial assistance systems in long-term care or in the care of persons with dementia.
Information technologies, big data, and new forms of mass communication (e.g. via
twitter et al.) all raise questions about how to spell out autonomy in data ethics or
informational self-control. And these changes will have consequences for all of us
as politically responsible agents too.
Even more complicated issues lurk in the field of artificial systems that are
regarded as autonomous agents themselves. Our everyday conceptions of autonomy
and responsibility are prominently challenged by autonomous driving and, less
prominently but significantly, by smart weapons. We have to ask: Do such tools
preserve or strengthen our autonomous actions? Or do they replace our autonomy
by taking away our responsibility?
192 J. F. Childress and M. Quante

Finally, some developments seem to destroy autonomy from the outset by using
enhancement technologies as tools for self-optimizing our bodily and psychological
capacities. Perhaps moral enhancement is a good idea from the viewpoint of an
educator interested in the good. But can such a replacement of normal human auton-
omy still be regarded as a version of our autonomy? From the perspectives in this
volume, it is evident that the concept of autonomy itself has too many dimensions
and appears in too many contexts and traditions to function as a safe anchor.
Evaluating these dynamics requires a much broader set of philosophical tools as we
seek to determine whether these developments are strengthening, transforming, or
destroying our autonomy.

12.3 Intra- and Interdisciplinary Contexts

Facing this complexity is both philosophically fascinating and threatening.

Obviously it cannot be addressed adequately within one single book or by one phi-
losopher alone. Like agency, personhood, and free will, the conception of autonomy
is a topic for different sub-disciplines in philosophy (intra-disciplinarity) and needs
to be dealt with through interdisciplinary cooperation.
Within philosophy itself, with its widely accepted subdisciplines, it is evident
that philosophy of action and person, epistemology, and ethics need to be integrated
(even in thin conceptions). In any adequate analysis of thick conceptions, philo-
sophical anthropology and political and social philosophy also will become central.
Some of this intradisciplinary dialectic operates in the contributions of our volume.
Others would become visible if we were to widen the philosophical perspective
beyond the tradition of analytic philosophy featured in this book.
Finally, empirical research (e.g., in psychology or sociology) is important for
understanding and using the concept of autonomy in real life and social practices.
Our understanding and use also need to be enriched by scientific knowledge and
evidence from cultural anthropology, biology, and medicine. In some contexts (e.g.
illness, human embodiment, aging, or dying), medical knowledge is central, but
much more is needed. This is especially the case when we recognize the cultural
framing of these anthropologically basic features of our human existence.
Autonomy is a central and distinctive feature of our lebensform. Since human
persons are complex and dynamic entities, they must be considered from a variety
of perspectives — within and beyond philosophy. In this volume, we offer one way
into this complex phenomenon; we also aim at making accessible some of the com-
plexities and tensions surrounding autonomy. But we do not pretend to deliver an
exhaustive or even a comprehensive analysis of personal autonomy. This we couldn’t
have willed autonomously—it would be far beyond what we can do. The limits of
the philosophical debates represented in this volume should be taken as motivation,
as a challenging opportunity for further multidisciplinary and interdisciplinary dis-
cussion to enrich our understanding of what it means to be autonomous human
persons in the world.
Index

A Authenticity, 46, 47, 49, 70, 80, 125, 130, 147,

Abasement, 62 148, 167, 190, 191
Abels, H., 57 Authority, viii, 14, 18, 23, 68, 71, 73, 76, 117,
Ach, J., vi, 53–64, 84 118, 122, 124, 130, 131, 153, 154, 157,
Acting freely, 2–5, 10, 14, 15 162, 166, 171, 172, 183, 184
Action Authorization
autonomous, 57, 60, 87, 92, 99, 106, 107, autonomous, 170
117, 137 Autonomy
in character/out of character, 147 capacity for, 78, 79, 110, 111, 122, 126,
free, vi, vii, 12, 23, 36, 40, 88, 90, 99 182, 190
intentional, vii, 54, 71, 74, 99 constitution of, 73
intentional and nonintentional, 90, 99 degree condition, 109
voluntary, vii, 99 degree/threshold property, 69
Advance directives, revocation of, 172 different functions of, 117
Affection/affective, 24, 72, 119 disclosive dimension of, 30
Agency enabling policies, 122
hermeneutic, 28 establishing/re-establishing, 68
shared, 74, 76 exercise of, vii, 68, 69, 72, 73, 78, 79, 106,
Agency theory, 14 108, 136
Agents externalist account of, 113, 183
autonomous, 80, 183, 191 genesis of, 57
Aid workers, vii, 68, 76–78, 80–82 libertarian conception of, 176
Alexander, L., 153, 156 local versus global models of, 102
Alienation, 71, 72, 80, 125 minimalist account of, viii, 103, 115
Ambivalence, vi, 18, 21, 22 moral value of, viii, 103
Animality, 24 nascent, 77, 81
Animals, 19, 47, 96 necessary and sufficient conditions for, 122
Answerability, 71 pathologies of, 61
Anti-paternalism, 171 perfectionist, 121, 125, 128, 130, 190
Aristotelian principle, 35 personal, vi, viii, x, 54, 55, 57, 71, 120,
Aristotle, 92, 93, 98, 99 123, 124, 127–130, 135–148, 153, 155,
Autarky, 191 162, 169, 176, 182

© The Author(s), under exclusive license to Springer Nature 193

Switzerland AG 2022
J. F. Childress, M. Quante (eds.), Thick (Concepts of) Autonomy, Philosophical
Studies Series 146, https://doi.org/10.1007/978-3-030-80991-1
194 Index

Autonomy (cont.) Bransen, J., 10, 35

as a political concept, 129 Bratman, M., 1, 12–14, 71, 74–76, 80, 119
precedent, 147 Brink, D.O., 45
principled, 143 Brock, D.W., 152, 154, 155, 157, 159,
procedural account of, 125 163, 169
relational, viii, 54, 68, 69, 79, 113, 126, Buchanan, A.E., 152, 154, 155, 157, 159, 163,
135–148, 177, 182 169, 170
respect for, vii, viii, x, 53, 68, 73, 75–78, Bystanders, 20–23, 30
81, 83, 104, 117, 127–129, 135–137,
142, 144, 148, 160, 162, 170, 189, 191
robust versus minimal accounts of, 115 C
thick concepts of, v, viii, 84, 115, 118–120, Cale, G.S., 152, 163
123, 124, 126, 127, 141, 190 Capabilities, 35, 36, 45–47, 55, 59, 80, 155,
thin concepts of, 118, 123, 136, 156, 159, 161, 169
142, 189–191 Capacity/ies, 9, 19, 54, 70, 91, 110, 118, 137,
threshold and ideal concepts of, viii 151, 182, 190
threshold condition, 109, 122 mental, 128, 151
tracing condition, 109 Care/caring, 5, 19, 54, 71, 104, 120, 136, 158,
Autonomy and informed consent, relation 181, 191
between, 105, 106 Cartesian, 25
Autonomy as ideal, 56 Character, 21, 22, 30, 34, 36, 40, 48, 49, 70,
Autonomy as right, viii, 117, 118, 121–123, 78, 113, 125, 147, 148, 183
125, 127 Charland, L.C., 155, 156
Autonomy-enhancing, ix, 177, 180, 185 Checkland, D., 152, 155, 159, 161, 163–165
Autonomy-identification, 15 Childress, J.F., viii, 54, 102, 104, 118, 121,
Autonomy-promoting, 185 123, 135–148, 153, 157, 160, 168,
Autonomy rights, 124, 126, 128–131, 169, 189–192
170, 177 Choices
autonomous, viii, 83, 119, 136, 137, 140,
142, 143, 146–148, 185
B Christian world view, 81
Bargaining positions, 181 Christman, J., vii, 15, 18, 41, 50, 54, 68–84,
Beauchamp, T.L., 54, 102, 104, 118, 121–123, 107, 118–124, 126, 129, 130, 191
127, 129, 130, 136, 137, 143, 147, 151, Citizens, 122, 123, 127, 129, 137, 182
153, 154, 157, 159–162, 167–169 Coercion, 35, 43, 62, 123, 136–138, 141, 184
Belief(s), vii, viii, 6, 11, 19, 34, 36, 42, 49, Coherence, vi, 17, 22, 28, 118, 119, 125
87–100, 118, 125, 126, 129, 139–141, volitional, 28
143–145, 154, 160, 169 Commitments, 5, 13, 14, 21, 28–30, 41, 59,
false, vii, 25, 95, 169 61, 68, 70–72, 75–77, 80, 82, 83, 118,
replacement of, 96, 99 122, 125, 126, 145, 162, 183
Beneficence Communications, 105, 137, 169, 191
duty of, 182 Compatibilism, 3
Benjamin, J., 61, 139 Competence
Benn, S.L., 119–122 assessment of, 161, 168
Benson, P., 18, 55, 71, 184 asymmetrical decisional, 164
Berlin, I., 42, 44, 45, 115 attribution of as “entrance ticket”, 153
Bioethics, v, vii–x, 84, 100, 102, 115, 117, cognitive and volitional, 122, 123, 128
135–148, 153, 159, 167, 172, 176, complexity-relative concept of, 168
178, 189 decisional, ix, 158, 159, 161, 162, 164,
Biography, v, 60, 63, 119, 125, 155 167, 172
Blood transfusion evidence of, 169
refusal of, 137, 140, 143, 145, 147 fixed threshold conception, 160, 162
Bowlby, J., 60 of patients, ix, 153, 158, 160, 165,
Brainwashing, 70 167, 169
Index 195

a precondition for valid risky, 140, 168–169

authorization, 152 shared, 78, 156
presumption of (for adults), 140, 154 Deference, 68, 73
risk-related concept of decisional, ix, Deliberation, 9, 67–84, 156, 167
152, 159 joint, 67–84
set of mental capacities or abilities in, 155 Demarco, J., 152, 163, 170–172
sliding scale of, ix, 151 Dementia, x, 191
task-specificity of, 158 Democracy, 7
Competence and rationality Democratic processes, 129
process and outcome, 157 den Hartogh, G., 154
Competence conditions, 54 Dennett, D.C., 20
Compulsion, 13, 40, 62, 70, 92, 125 Dependence, 77, 83, 126, 142, 182
Concerns, v, 4, 20, 28, 41, 42, 73, 78, 88, 92, Descartes, R., 25
93, 95, 103–105, 122, 141, 152, 153, Description
155, 168, 171, 189 intensional, 103, 106–109, 112,
Confidence 113, 115
warranted, 25 De Shalit, A., 176, 177, 179
Conflict Desire(s), vi, 4, 7, 9, 11, 13, 17–21, 23–26, 29,
external, 4, 21 30, 37, 39–42, 69, 70, 74, 96, 102, 113,
internal, 20–22 114, 118, 119, 128
Consent conflicting, 4
future/deferred, 148 conscious, 4, 119
informed (in medicine), viii, ix, 103–108, higher-order, 5–7, 23, 40
111, 128, 130, 137, 140, 145, 146, 151, motivating, 7, 40, 42
153, 154, 157, 158, 167, 172 second-order, 3, 5, 9, 40
past/prior, 146 unconscious, 4, 119, 125
presents, ix, 146 Determinism
valid, 140, 151 causal, 2, 3
Consequentialism, 44, 60, 61, 64, 119, 124, Deveaux, M., 177
126–128, 138, 140, 141, 160, 164, 165, Development, vii, 4, 17, 35, 45, 46, 57, 60, 61,
169, 171, 190, 191 72, 78, 79, 81–84, 136, 156, 183,
Consequentialist, 9, 121 184, 192
Constraints, 22, 54, 76, 92, 142, 163, 176 Developmental stages, 60
Content neutrality, 79 Dialectic, 189, 192
Controlling influences, 54, 136, Dignity
137, 140 human, 123
Controls, vii, 37, 41, 43–45, 48, 71, 74, 88, Disability, 70
91–94, 96, 98, 99, 104–108, 112, 113, Disadvantage, x, 122, 176–185
123, 136, 137, 162, 163, 183 corrosive, 177, 179
Conversion, 148 Disciplines
Culver, C.M., 151, 152, 155, 157, 161 interdisciplinary, x, 155, 192
intra-disciplinary, 192
multidisciplinary, 192
D sub-disciplines, x, 192
Davidson, D., 89, 90 Disorientation, vi, 21, 22
Deception, 95, 96, 99, 102, 137, 184 Disposition, 71, 79, 147
Decisions Dodds, S., 177, 178
autonomous, v–x, 53–57, 63, 68, 72, Domination, 183, 184
102–106, 108, 109, 111–115, 122, Dornes, M., 58, 59
124, 126, 130, 140, 145, 147, 154, Drane, J.F., 151, 159, 160, 163
176, 189 Dresser, R.S., 94
complex, 54, 168–169 Dworkin, G., 22, 70, 102, 118, 123, 172
life-shaping, 166–168 Dworkin, R., 121, 123
196 Index

E Friedman, M., 18, 141

Education Functioning(s), 81, 126, 144, 145, 176,
liberals, 78 179–181, 183, 184
Egalitarian, 121, 122
Egalitarianism, 121, 122
Elster, J., 119 G
Emotion/emotional, 29, 55–59, 63, 72, Ganzini, L., 158, 159
80, 82, 119, 136, 155, 157, 158, Garden of Eden, 19
178, 183 Gatekeeping function (of competence
Emotional stability, 70 judgments), 123, 160
Enhancement, x, 127, 192 Gerhardt, V., 43
Epistemic conditions, constraints, 92 Gert, B., 151, 152, 155, 157, 161
Equality, 76, 121–123 Geuss, R., 44
Eshleman, A., 10 Global trade
Ethics in body parts, 175, 176
applied, 102 Good life
biomedical, v, viii, 117, 123, 127, 128, conceptions of, 120
143, 151, 189–191 Govier, T., 55, 184
Ethnographic research, 175, 176, 180, 181, Gray, J., 47
184, 185 Green, T.H., vi, 36, 39, 42–46
Evaluations Gutmann, T., viii, 84, 131, 170
holistic, 168
Evans, M., 48
Exploitation, ix, 62, 135, 175–185 H
liberal analysis of, ix Habermas, J., 125
mutual, 185 Hanna, J., 157
structural dimensions, ix Happiness, 35
taxonomy of, 177, 180 Harms, ix, 61, 89, 151, 156, 159, 160, 165,
transaction analysis of, ix, 176 167, 169–172, 177–179, 185
Health
psychological, 21
F Hegel, G.W.F., 43, 44, 46
Fabre, C., 176 Herman, B., 25
Faden, R.R., 137, 151, 153, 154 Hermeneutic, 28, 30
Fair market conditions, 176 Herrmann, M., 41
Fairness Heteronomous, vii, 82, 88, 94–96, 98, 100
macro, 180–182 Heteronomy, vii, 82, 88, 94–96, 98, 100
micro, 176, 177, 180, 181 Hierarchical accounts, 23
Fallibilism, 169–170 Hierarchicalism, 23–24
Fateh-Moghadam, B., 170 Hobbes, T., 131
Feinberg, J., 46, 87, 118, 119, 121–123, 125, Homelessness, 178–180
129, 131, 157, 169, 170 Homunculus, 37, 39, 43
Feminism/feminists, 126, 141, 142 Honneth, A., 55, 56, 61, 119, 121, 123, 184
Fiduciary relations, 137 Human beings, 34, 35, 37, 42, 44, 45, 47,
First-order desire, 3–7, 24, 102 50, 62, 178
Fischer, J.M., v, 1–15, 92, 93 Human nature, 45, 47, 48, 50
Frankfurt, H., 2, 17, 70, 96 Humiliation, 62, 184
Freedom, vii, 2–5, 7–12, 14, 19, 33, 34, 36, Hurst, S.A., 178
40–45, 48, 88, 92, 93, 115, 120, 124,
126, 158, 176, 181, 183, 184, 190
juristic/civil, 44 I
negative, 43, 44, 176 Iago, 104, 105, 108, 110
positive, 42–44, 120 Idealism, 191
Freud, S., 119 Ideal type, 50, 125
Index 197

Identification, vi, 3–7, 10, 14, 23–25, 40, 41, L

97, 113, 161 Law, viii, 48, 87, 117, 121, 123, 126, 131, 136,
Identity 145, 151, 156, 169
practical, vii, 70, 71, 77, 81, 84, 182 medical, viii, 117, 120, 123, 151
Inclinations, 20, 23, 24, 56 Lear, J., 70
Inclusion, 121–123, 125 Lebensform, 192
Incoherence Leukemia, 138
volitional, 21, 22 Liberal analysis, ix
Independence, 61, 82, 142 Liberal principle, 68
Indignity, 62 Liberal sensibility, vii, 69, 72–73, 76, 78–84
Individuality, 46, 48 Liberating effect, 22
Individual(s), vii, viii, 11, 13, 25, 34–36, Libertarian, ix, 176
45–50, 60, 62, 68–70, 74, 76, 78, 79, Libertarianism, ix, 176
87, 88, 110, 118, 121, 122, 126, Liberty/liberties, 42–44, 62, 119, 121, 129,
128–130, 139, 140, 142, 144–146, 166, 137, 176
176, 177, 179, 182, 183, 185, 189, 190 civil, 44
Infant development, 60 Liberty rights, 122, 130
Influences Life plans, 11, 77, 78, 119, 125, 147, 168
controlling, 54, 136, 137, 140 Lindberg, D., 138, 140
Information technologies, x, 191 Locke, J., 43, 44
Informed consent, viii, ix, 102–109, 111, 128, Love, 6, 7, 15, 25, 40, 41, 63, 73, 120
130, 137, 140, 145, 146, 151–154, 157,
158, 162, 167, 168, 172
connection with autonomy, 111 M
ideal and non-ideal, 154 Macklin, R., 152
valid, 153, 154 Maclean, A., 110–112, 114, 140
Instrumentalization, 62 Malmqvist, E., 177, 182, 185
Integration, 22, 23 Manipulation, 68, 70, 81, 95, 96, 102, 104,
Integrity, 125, 127, 130, 147, 148 105, 110, 136, 162
Intelligibility, 27, 28, 76 of information, 136
Intention(s), 55, 70, 74, 75, 78, 80, 93 Manipulator, 95
Interactions, vii, 55, 59, 68, 72–74, 76, 78, 80, Margalit, A., 62
83, 84, 141, 142, 179 Markets
Interdisciplinarity, x noxious, 185
Interdisciplinary, 155, 192 regulated, 176, 180
Interpersonal relations, 68, 72, 73 Maturity, 19, 58
Interventions, x, 77, 136, 137, 139, 140, 148, McKenna, M., 87, 88, 90–95, 99
151, 153, 160, 163, 166, 178, 179 Mead, G.H., 57
Intuition(s), 13, 28, 90, 93, 98, 126, 159, 171 Medical treatments, 103, 137, 147, 153, 154,
Isay, D., 26 158, 159, 161, 164, 166, 167, 172
Mental states, 12, 106, 107, 113, 114, 167
Meyers, D.J., 119, 126, 183
J Miller, B., 147
Jehovah’s Witness, viii, 137–139, 141–143, Miller, F.G., 153
145, 147 Mill, J.S., 36, 39, 45–49, 118–120
Jox, R.J., 172 Mission creep/drift, 7, 8
Mistrust, 58
Mitchell-Yellin, B., 8, 11, 15
K Models of decision making
Kamm, F.M., 129 family centered, 145
Kant, I., 44, 118, 143, 190 patient centered, 145
Kantian, 72, 112, 182 Motivations, 11, 12, 24, 42, 81, 110, 118, 153,
Kidney sales, markets, 185 171, 175, 184, 192
Korsgaard, C., 71 Motive(s), structure, content of, 79
198 Index

N Persons
Narrative unity, 130, 147, 148 autonomous, 59, 70, 71, 73, 74, 81, 93,
Needs 121, 129, 130, 183
basic, 62, 63, 178, 182 concept of, vi, 36–39, 42, 46, 118, 121,
Negligence, 62, 108 122, 124, 125, 128, 130, 183
Negligent, 22, 105, 109 equality of, 127
Negligently, 22, 105, 114 respect for, viii, ix, 68, 69, 73, 120, 125,
Neutral, viii, 3, 73, 78, 83, 107, 125 127, 130, 145, 191
Neutrality, 18, 79 Phenomenology, 22
Non-autonomous agents Philosophy
actions, 109 analytic, ix, 190–192
Non-interference, 68, 73, 137 intercultural, x
Normativity, 5–7, 18, 120, 161–163 Platonic model, 13
Political, 20, 21, 38, 44, 45, 64, 72, 113, 123,
142, 179, 183, 190, 192
O Politics, 20, 165
Obligation(s), vii, 68, 69, 72, 73, 76–79, 83, Pollmann, A., vi, 53–64
84, 125, 130, 137, 142, 146, 184, 185 Poverty, 88, 178
Oedipus, 107 Power, vii, 12, 41, 44, 45, 48, 76, 78, 80, 81,
O’Neill, O., 143, 153 83, 84, 120, 129, 130, 153, 181, 183
Ontic status, viii, 121, 122, 127, 128, 130 asymmetrical, vii, 80
Ontology, vi, 17–30, 36 symmetrical, vii, 76, 80, 83
Oppression, oppressive, 126 Pragmatism, 191
Option sets, 182 Predicament, vi, 19–30
Ordinary-person standard, 167–168, 171 Prima facie, 114, 115, 121, 137, 144, 158
Organ sales, markets, ix, 176 Principled autonomy, 143
Oshana, M., 37, 71, 81, 119, 120, 124, Principles, 11, 26, 29, 53, 72–74, 92, 94, 98,
126, 183 102, 104, 117, 118, 127, 136, 137, 143,
Othello, 104, 105, 107–110 144, 157, 159, 168, 171, 189, 190
Overconfidence, 25 Principlism, 127, 145
Principlist, 121
Professionals
P healthcare, 104
Panitch, V., 177, 180, 181 Prohibitions, 177, 178, 182, 185, 186
Partnerships Pro tanto, 69, 73, 162
asymmetrical, 76, 78, 80, 83 Psychology
symmetrical, 76 developmental, 60
Passions, 4, 18
Passivity, vi, 17–30
Paternalism, ix, 57, 68, 73, 78, 79, 135, Q
169–172, 182–185, 189 Quante, M., 15, 84, 118, 120, 125, 127–131,
group soft, 185 155, 167, 189–192
hard and soft, 67, 170 Quiddity, 47
hidden, 152, 171, 172
Patient autonomy, 103, 144, 145, 153, 157
Perfectionism, 35, 38, 57 R
Personal autonomy Radcliffe-Richards, J., 176
perfectionist ideal of, viii Rationality, 70, 77, 112, 119, 122, 124,
Personality 128, 156–157
moral, 98 Ravizza, M., 13, 92, 93
Personhood, 10, 19, 20, 25, 27, 29, Rawls, J., 35, 119, 120, 122, 124
122, 190–192 Raz, J., 120, 121
Index 199

Reasons, 5, 6, 15, 18, 19, 21, 24, 25, 29, Rössler, B., 21
39–42, 45, 46, 49, 50, 61, 63, 64, 71, Rousseau, J.-J., 190
72, 74–76, 78, 81–83, 88–94, 106–108,
110–112, 114, 115, 118–126, 128–130,
138, 140, 148, 152–154, 156, 158, S
162, 171 Sartre, J.-P., 39
Recognition, 28, 59, 61, 71, 80, 121, 184 Satisfaction, 4, 19, 23, 24, 110, 171
Reflection, ix, 8–10, 12, 15, 23, 25, 39, 41, 42, Scarry, E., 62
61, 70, 71, 74, 75, 82, 97, 118, 119, Scheper-Hughes, N., 175
125, 130, 155 Schoene-Seifert, B., 84, 151–173
cooperative, 74, 82 Schramme, T., vi, 34–51
critical, 25, 41, 97, 119 Second-order desires, preferences, volitions,
Refusal ix, 4, 6, 9, 15, 34, 40, 41, 88, 118, 122,
informed, ix, 137, 146 126, 128, 146, 154, 155, 157, 162, 166
Regress objection, 18 Seel, M., 40
Regress of justification, 17 Self, 5, 23, 54, 68, 87, 119, 141, 153, 183, 190
Regulations, 136, 158, 177, 186 alternatives, 38, 43, 45, 50
Relationship(s), vii, 2, 3, 9, 10, 13, 36, 54, 55, dominant, 44, 142
58, 60–64, 68, 75, 80, 83, 105–108, empirical, 43, 44
111–113, 120, 137, 142, 144, 179, multiple, 38
181, 183 oppressed, 142
Relations, symmetrical vs. asymmetrical, real/true/designated, v, 6–8, 10
vii, 76, 78, 83, 96, 163 temporally extended, 146–148
Reproductive labour, 175 Self-acceptance, 24, 71, 74, 81–83
Resolution, volitional, 29 reflective, 71, 74, 81, 82
Respect, 7, 22, 68, 95, 102, 118, 135–136, Self-actualisation, 36–39, 43, 45–50
160, 176, 189 voluntarist and perfectionist
Respect for autonomy, vii–x, 68, 73, 74, 76, conceptions of, 39
78, 83, 104, 117, 121, 127–130, Self-affirmation, 55, 59
135–137, 142, 145, 148, 160, 170, Self-alienation, 71
189, 191 Self as diachronic, 67, 83
negative and positive, v, viii, x Self-assertiveness, vi, 53–64
practice of, 136 Self as synchronic, 44
principle of, 121, 127, 129, 130, Self-authorization, ix
136–137, 191 Self-conception, 39, 70
Responsibility Self-confidence, vi, 53–64
moral, v, 1–15, 124, 125, 143 Self-control, 70, 80, 122, 124, 128, 156,
Responsiveness 183, 191
mutual, 75 Self-critical, 24, 160, 164
Rights, 21, 54, 72, 98, 117, 137, 157, Self-definition, 119, 183
177, 190 Self-determination
negative and positive, viii perfectionist account, 34
Risk, ix, 24, 77, 80, 104, 151–173, 177–180 right to, 121, 123–126, 130
Risk assessment, 167 voluntarist account, 34
Risk communication, 169 Self-development, 45–50
Risk-relative conception of competence Self-discovery, 119
(RRCC), 152, 154, 158, 159, Self-emergence, 47
161–165, 168–173 Self-esteem, vii, 53–64, 119, 171, 184
Risk taking, 152, 169, 179 Self-formation, 35–39, 42–46, 48, 49, 51
Risk, understandings of, 140, 152, 158, 159, perfectionist, 39, 42–45, 49, 50
164, 167 voluntarist, 39
Rogers, W., 177, 178 Self-fragmentation, 72
Romantic tradition, 190 Self-governance, v, 5, 6, 14, 119, 142, 183
200 Index

Self-governing, 7, 70–73, 79, 80, 82, 119, T

183, 184 Taylor, C., 1, 9–12, 15, 29, 36, 119
Self-government, 70, 74, 79, 80, 82, Taylor, J., 15, 101–115, 176
84, 87, 122 The Big Lebowski, 35
Self-ideal, vi, 38–40, 43, 46, 49, 50 The Dude, 34, 35
Self-identification, 39–42 The Fall, 19
Self-interpretation, 29, 30 Third-order desires, 17
Self-monitoring, 71 Tightrope walking, example for explicating
Self-narrative, 70, 71 competence, 158, 163
Self-opacity, 26–27, 29 Tolerance, 79
Self-perception, 56 Toleration, 78
Self-realization, 42–46 Transactions, ix, 176, 177, 181, 182, 184, 185
Self-regarding decisions/actions, 68, 153, 162, Trauma, vii, 68, 76, 77, 79, 82
167, 171, 172 Triple S condition (of personal autonomy), v,
Self-relation, vi, 36, 37, 39, 53–56, 58, 60–64 viii, 53–64, 120, 121, 124–128, 131,
evaluative, 37, 56, 59, 60 136, 153, 189, 192
practical, 53, 55, 61, 62 Trommsdorff, G., 57
Self-respect, 77, 119, 123, 184 Trust, 55, 58, 61, 68, 120
Self-rule, 87, 141, 142
Self-satisfaction, vi, 17–30
Self-trust, 119, 184 U
Self-understanding, vi, 28, 29, 82, 83, Understanding
124, 183 adequate, 140
Sherwin, S., 141, 142, 146 thick and thin views of degree required for
Shields, Rev. R., 26–28 autonomous decisions, viii
Silver, M., 166
Simhony, A., 45
Sisyphus, 46, 47 V
Snyder, J., 176, 177, 180–182 Value(s), 11, 23, 55, 95, 102, 118, 141,
Social interactions, 54, 70, 84, 184 152, 183
Sociality, 75, 76 change, 96, 99
Socialization, 57, 126, 128 judgment, 13, 37
Sovereignty manipulation, 96
over oneself, 118, 128, 129 structures, 79
Sovereignty right, 131 system, 98
Stoljar, N., 54, 120, 121, 124, 126, 141 Velleman, J.D., 1, 9, 13, 21, 22, 27, 36
Story/stories, 13, 14, 17, 25, 26, 88, 89, 91, Volitional necessities, 6, 40, 41
94–99, 101, 139, 162, 171 Volitional reflexivity, vi, 19–30
Strong evaluations, 9, 10, 12, 15 Volitions, 3, 9, 10, 14, 15, 39–42
Structural inequality(ies), injustice(s), 177, Voluntariness, 54, 123, 125, 137, 141, 169
179–182, 185 Voluntarism, 35, 38
Subjectivity, 25, 41, 50, 118, 152, 155, 157 Vulnerability, ix, 61–63, 77, 83, 175–185
Subject-object, 37 inherent, 178–180
Surrogacy pathogenic, 177–180, 184
commercial, ix, 176, 177 situational, 177–180
gestational, 176, 177
Surrogacy contracts
transnational, 181 W
Surrogate decision-making, 172, 173 Waldron, J., 123, 125
Swindell, J.S., 21 Wantons, vi, 19, 20, 28, 30
Systems Watson, G., 1, 9, 11–13, 15, 18
artificial, x, 191 Wertheimer, A., 123, 153, 176, 177
Index 201

Westlund, A., 71 freedom of, 2, 3, 36, 44, 92

Wholehearted, vi, 7, 18, 21, 26–29, 41 healthy, 28
Wholeheartedness, vi, 18, 21, 23, 29 mature, 28, 60
Wicclair, M., 161, 163, 164, 169, 171, 172 weakness of, vi, 12
Wilkinson, S., 176 Winnicott, D.W., 57, 60
Wilks, I., 158, 161, 163, 164 Wittgenstein, L., 191
Will Wolf, S., 18
free, 2, 3, 8, 36, 40, 42–44, 110, 119, 129 Wolff, J., 176, 177, 179