Normativity and Agency: Themes from the Philosophy of Christine M. Korsgaard 0198843720, 9780198843726

Citation preview

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Normativity and Agency

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Normativity and Agency Themes from the Philosophy of Christine M. Korsgaard Edited by

TA M A R S C HA P I R O, K Y L A E B E L S - ­D U G G A N , A N D SHA R O N ST R E E T

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © the several contributors 2022 The moral rights of the authors have been asserted First Edition published in 2022 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2022933696 ISBN 978–0–19–884372–6 DOI: 10.1093/oso/9780198843726.001.0001 Printed and bound in the UK by Clays Ltd, Elcograf S.p.A. Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Contents Editors’ Preface vii List of Contributorsxiii 1. The Horizons of Humanity David Sussman

1

2. Finite Valuers and the Problem of Vulnerability to Unmitigated Loss21 Sharon Street 3. A Question of One’s Own: Concepts, Conceptions, and Moral Skepticisms50 Kyla Ebels-­Duggan 4. The Two Normativities J. David Velleman

78

5. Self-­Consciousness and Self-­Division in Moral Psychology Richard Moran

95

6. What Makes Weak-­Willed Action Weak? Tamar Schapiro

126

7. Integrity, Truth, and Value Sigrún Svavarsdóttir

147

8. Shadows of the Self: Reflections on the Authority of Advance Directives175 Japa Pallikkathayil 9. Korsgaard on Responsibility T. M. Scanlon

197

10. Animal Value and Right Stephen Darwall

213

11. Juridical Personality and the Role of Juridical Obligation Barbara Herman

240

12. The Social Conditions for Autonomy: Kant on Politics and Religion264 Faviola Rivera-­Castro Index

287

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Editors’ Preface Christine Korsgaard decided not to go to college—at least at first. She had purchased a set of great books, and her plan was to educate herself while earning a living doing office work. And so it happened that she first stumbled upon Plato and Nietzsche while on the Illinois Central commuter train between her home in south suburban Chicago and the secretarial school she was attending downtown. The encounter was to change her life. Growing up, Korsgaard had often found herself preoccupied with big questions about human life, including questions about the objectivity of ethics. But, as she later recounted, “I am a first-­generation college student, so I hadn’t the vaguest idea that philosophy was an academic field.”1 It was only upon reading the philosophers of the past that she realized that this thing she had been doing—this thinking and wondering about fundamental questions—had a name. In ways she couldn’t possibly have anticipated, this discovery would also change the course of moral philosophy, for Korsgaard would go on to become one of the most important moral philosophers of her time. She would write pathbreaking books and articles, developing her own answers to the questions that had intrigued her early on, and she would teach and inspire countless students. Along the way, she would break barriers for women in the profession, serving as the first woman chair of the Harvard Philosophy Department, and the first woman to give the Locke Lectures at Oxford. We owe a debt of gratitude to those whose recognition and support helped set her on this remarkable path. At her first job, working for attorneys at the American Bar Association, she made new friends who gave her a picture of college that was more attractive than the one she had imagined. That, together with her sense that her efforts to learn philosophy on her own weren’t bearing fruit, changed her mind about college. She enrolled mid-­year at Eastern Illinois University, later transferring to the University of Illinois at Urbana-­ Champaign, where she earned her B.A. in 1974. Her professors at UIUC saw her extraordinary potential, and with their encouragement, she applied to graduate school. Soon she began to study toward her Ph.D. in philosophy at Harvard under the supervision of John Rawls.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

viii Editors ’ Preface She landed in fertile soil. Rawls was a generous and conscientious mentor, and though women were a marginalized minority within the profession at that time, they were unusually well-­represented among his students. Korsgaard found Rawls’s lectures on the history of moral philosophy exhilarating. She soon found herself writing on both Aristotle and Kant. Rawls “told me to pick one of them,” Korsgaard later remembered, “so I picked Kant and my dissertation became a search for the basis of the claim that the categorical imperative is a principle of reason.”2 Korsgaard went on to become one of most successful among Rawls’s many accomplished students. After graduating from Harvard in 1981, she held positions at Yale, UC Santa Barbara, and the University of Chicago. In 1991, Korsgaard returned to Harvard as Professor of Philosophy, becoming the Arthur Kingsley Porter Professor of Philosophy in 1999. Now, on the occasion of Korsgaard’s retirement from teaching, we are delighted to present this collection of essays to acknowledge and honor the tremendous impact that both her writing and teaching have had on the discipline. In her lectures, Korsgaard made the history of moral philosophy live and breathe. Her storied courses, Philosophy 168 (“Kant’s Ethical Theory”) and Philosophy 172 (“The History of Modern Moral Philosophy”), kindled a passion for moral philosophy and its history in countless undergraduates and graduate students. She demonstrated how to have substantive conversations with writers of the past by reading charitably, with careful attention to the text and its historical context, but also with an eye to how that philosopher’s system might be relevant to the philosophical questions that continue to engage us today. Like her mentor, Rawls, Korsgaard was a deeply giving and committed teacher. Her office hours—every Thursday at 2 p.m.—were always crowded, and legendary among graduate students for the way one could drop in for immediate, penetrating feedback on whatever one was thinking about. She devoted countless hours to helping her students find what was most valuable in their own thoughts. To one young graduate student, who was unsure whether she belonged in the profession, Korsgaard wrote, “This is well done, but next time I’d like to see more of you in there.” The fact that Christine Korsgaard was listening and wanted to know what you had to say was powerful and inspiring to many a student trying to find their way. Korsgaard’s published work is striking for two reasons—first, its ability to communicate the urgency and excitement of moral philosophy, and second, its sheer range and depth. To convey the scope of the work, it is useful to divide it into four stages.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Editors ’ Preface  ix In the first stage, Korsgaard published interpretive papers on specific topics in Kant’s ethical theory, along with comparative papers highlighting the structural features of the Kantian approach in relation to other ethical theories. These papers were collected in her volume Creating the Kingdom of Ends (Cambridge, 1996). While issues of Kant interpretation are never entirely settled, in this early period Korsgaard succeeded in mapping out positions of enduring influence on a range of topics: the motive of duty, the nature of universalizability, the concept of humanity, and the relationship between freedom and morality. Many of these have become standard readings of the texts, and these early positions continued to inform her distinctive version of Kantianism even as it evolved over subsequent decades. In her comparative essays, Korsgaard raised the level of the standard debate between consequentialists and non-­consequentialists. Her strategy was to reveal deeply rooted structural differences between utilitarian and Kantian approaches to ethical theory. In “Personal Identity and the Unity of Agency,” for example, she helps us to see why utilitarians and Kantians differ on the topic of whether to aggregate value across persons. Her diagnosis includes the observation that utilitarians and Kantians hold different conceptions of personal identity. But it goes even deeper. These philosophers hold divergent views of identity, she argues, because they disagree about why moral philosophy needs a conception of personal identity in the first place. The utilitarian looks for a concept of practical identity that can be employed in theoretical reasoning, whereas the Kantian needs that concept to play a role in practical reasoning. In “The Reasons We Can Share,” she shows that the consequentialist feature of utilitarianism is not simply based on an intuition that results matter more than motives. It is rooted in the idea that morality is about what we do to or for others, where this contrasts with the Kantian idea that morality is about what we do with one another. In these essays, Korsgaard put into practice a new way of doing moral philosophy. Instead of focusing on competing intuitions about, say, whether to kill one or let five die, she asks how and why we use moral concepts in the first place, mapping out different possible answers, along with their implications. The second stage of Korsgaard’s career is best represented by her contemporary classic, The Sources of Normativity (Cambridge, 1996). The volume consists of four Tanner Lectures, delivered in 1992 at Cambridge University, along with comments and replies. It is perhaps the most vivid example of Korsgaard’s ability to make moral philosophy gripping. The book opens by identifying and pressing what Korsgaard calls “the normative question”—why am I obligated to do anything at all? This question is at once deeply

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

x Editors ’ Preface philosophical—concerning the foundation of ethics itself—and yet also arrestingly practical. It is a question that, in an extreme moment, any one of us might ask. This framing of the question illustrates Korsgaard’s Rawlsian conviction that questions of moral philosophy proceed, ultimately, from the standpoint of the moral agent, rather than the observer of moral practice. Her thought is that it is only by answering the moral agent directly that we can even frame subsequent questions about the existence and nature of moral facts, the attainability of moral knowledge, or the meaning of moral terms. The book is also a signature example of Korsgaard’s engagement with the history of philosophy as part of an ongoing conversation. In the first two ­lectures, she interprets the voluntarists, rationalists, and sentimentalists of seventeenth- and eighteenth-­century Britain as providing, respectively, three candidate answers to the normative question. In the culminating third and fourth lectures of the book, she offers her own answer. Normativity arises out of our reflective nature, which puts us into the position of being governors of ourselves. It is a Kantian answer—that is, deeply informed and inspired by Kant—but an answer which Korsgaard began here to articulate, revise, and elaborate in ways that are her own. In The Sources of Normativity, Korsgaard emphasizes that the theory proposed there is only a sketch. We may think of the third stage of Korsgaard’s work as the period in which she filled in additional substance and detail. Her initial replies to her commentators grew into articles, and eventually into her Locke Lectures, delivered at Oxford in 2002. The articles and lectures were published as companion volumes in 2009 (The Constitution of Agency and Self-­Constitution, both Oxford). Drawing on metaphysical ideas from both Plato and Aristotle, she develops and defends her Kantian answer to the normative question. A central feature of this view is the claim that moral obligations are binding upon us because they are the internal standards of agency as such. To act well, in a moral sense, is just to be fully active in governing oneself. This aspect of her theory, which others have labeled “constitutivism,” has again changed the course of debates in normative ethics, metaethics, and the philosophy of action, and has become one of the most-­discussed positions in contemporary metaethics. In the fourth stage of her career, Korsgaard broadened her influence in an entirely different direction, engaging applied ethicists and policymakers. One of Korsgaard’s deepest lifelong commitments has been to the protection of non-­human animals. (She has been a vegetarian since 1975.) She began to lay out a defense of this commitment in a second set of Tanner Lectures,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Editors ’ Preface  xi delivered at the University of Michigan in 2004. While Kant himself had notoriously limited views about the moral standing of the other animals, Korsgaard argues that a proper development of Kantian commitments brings them within the scope of the moral community. After publishing Self-­Constitution, she returned to this topic. The result was Fellow Creatures (Oxford, 2018), in which she shows how her distinctive blend of Aristotelian and Kantian ideas can provide the basis for robust obligations to protect the interests of non-­ human animals. The present volume marks Korsgaard’s retirement from teaching. She will continue to write and to advise, and there will, no doubt, be a fifth stage in the development of her ideas. But this occasion provides us with the opportunity to express our gratitude and admiration by collecting essays written specifically in her honor. The contributors are former students (the three of us, Japa Pallikkathayil, Faviola Rivera-­ Castro, and David Sussman), Harvard colleagues (T. M. Scanlon and Richard Moran), and colleagues at other institutions (Stephen Darwall, Barbara Herman, Sigrún Svavarsdóttir, and David Velleman). This list is only representative of the many thinkers who have been influenced and inspired by Korsgaard’s teaching and writing. We could have easily filled additional volumes with papers by others who have been lucky enough to know and work with her. The essays engage with the kinds of philosophical problems that Korsgaard helped us to recognize and appreciate. Four of the papers assess her answer to the normative question and explore alternative versions of constitutivism (Sussman, Street, Ebels-­Duggan, Velleman). Another four focus on concepts central to her philosophical psychology—self-­consciousness, strength and weakness of will, integrity, and personal identity (Moran, Schapiro, Svavarsdóttir, Pallikkathayil). T.  M.  Scanlon offers an interpretation of Korsgaard’s account of holding someone morally responsible. Stephen Darwall evaluates her more recent work on the moral status of nonhuman animals. Barbara Herman and Faviola Rivera-­Castro explore the relation between Kant’s ethics and his political theory. The scope of the volume is broad, testifying to the remarkable range and fertility of Korsgaard’s work. We would like to thank our contributors, each of whom responded enthusiastically to our invitation. We are also grateful to Peter Momtchiloff and others at Oxford University Press for their guidance and patience, and to Evan Behrle and Bennett Eckert, for invaluable editorial assistance. Of course our deepest thanks are to Chris, for finding in us what we never could have found on our own.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

xii Editors ’ Preface

Notes 1. Katrien Schaubroeck, “Interview with Christine Korsgaard, holder of the Cardinal Mercier Chair 2009,” The Leuven Philosophy Newsletter, volume 17, 2008–2009/2009–2010, pp. 51–56, p. 51. Available at https://docplayer.net/89264337-­Volume-­17.html. 2. Schaubroeck, “Interview with Christine Korsgaard,” 51.

Bibliography Korsgaard, Christine  M. Creating the Kingdom of Ends. Cambridge: Cambridge University Press, 1996. Korsgaard, Christine  M. The Sources of Normativity. Cambridge: Cambridge University Press, 1996. Korsgaard, Christine  M. The Constitution of Agency: Essays on Practical Reason and Moral Psychology. Oxford: Oxford University Press, 2008. Korsgaard, Christine M. Self-­Constitution: Agency, Identity, and Integrity. Oxford: Oxford University Press, 2009. Korsgaard, Christine M. Fellow Creatures: Our Obligations to the Other Animals. Oxford: Oxford University Press, 2018. Schaubroeck, Katrien. “Interview with Christine Korsgaard, holder of the Cardinal Mercier Chair 2009.” The Leuven Philosophy Newsletter, volume 17, 2008–9/2009–10, pp. 51–56. Available at https://docplayer.net/89264337-­ Volume-­17.html.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

List of Contributors Stephen Darwall is the Andrew Downey Orrick Professor of Philosophy at Yale University and John Dewey Distinguished University Professor Emeritus at the University of Michigan. His books include Impartial Reason, The British Moralists and the Internal Ought, Welfare and Rational Care, The Second-­Person Standpoint, and two recent collections of his essays, published by Oxford University Press. He is a fellow of the American Academy of Arts and Sciences and, with David Velleman, a founding co-­editor of Philosophers’ Imprint. Kyla Ebels-­D uggan is Professor of Philosophy at Northwestern University. She specializes in moral and political philosophy working in a broadly Kantian tradition.  Her work has appeared in  Ethics, The Philosophical Quarterly,  Philosophical Studies, and Philosophers’ Imprint. Barbara Herman  is Griffin Professor of Philosophy and Professor of Law at UCLA. She is the author of The Practice of Moral Judgment (Harvard, 1993), Moral Literacy (Harvard, 2007), The Moral Habitat (Oxford, 2021), Kantian Commitments (Oxford, 2022), and numerous essays on Kant’s ethics, practical agency, and moral psychology. Richard Moran  is Brian  D.  Young Professor of Philosophy at Harvard University.  He is the author of  Authority and Estrangement: An Essay on Self-­ Knowledge  (Princeton, 2001),  The Exchange of Words: Speech, Testimony, and Intersubjectivity (Oxford, 2018), and The Philosophical Imagination (Oxford, 2017), a collection of essays. Japa Pallikkathayil  is Associate Professor of Philosophy at the University of Pittsburgh. She received her Ph.D.  from Harvard University. Her work focuses on issues at the intersection of moral and political philosophy. She has written about coercion, consent, and Kantian political philosophy. Faviola Rivera-­C astro  is Professor of Philosophy at Instituto de Investigaciones Filosóficas at National Autonomous University of Mexico. She is author of Virtud, felicidad y religión en la filosofía moral de Kant (2014) and currently works on the development of liberalism in Latin America as well as on the relation between politics and religion from a philosophical perspective. T.  M.  Scanlon  is Alford Professor of Natural Religion, Moral Philosophy, and Civil Polity, emeritus at Harvard University, where he and Christine Korsgaard were long-­term colleagues.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

xiv  List of Contributors Tamar Schapiro  is Professor of Philosophy at MIT. Previously she was a faculty member at Stanford and a Junior Fellow at the Harvard Society of Fellows. She is author of Feeling Like It: A Theory of Inclination and Will (Oxford, 2021) as well as articles on various problems in Kantian practical philosophy. Sharon Street  is Professor of Philosophy at New York University. She is the author of a series of articles on how to reconcile our understanding of normativity with a scientific conception of the world. In current work, she is exploring how insights from Eastern meditative traditions could shed light on questions in secular metaethics. David Sussman is Associate Professor of Philosophy at the University of Illinois at Urbana-­Champaign. His main research interests are ethics, Kant, moral psychology, philosophy of action, and German Idealism. His current work focuses on the expressive aspects of human action, their ethical significance, and their bearing on the distinction between practical and theoretical reasoning. Sigrún Svavarsdóttir  is an Associate Professor of Philosophy at Tufts University. Her research focuses mainly on issues within metaethics but extends also into theory of agency. She has published on the nature of moral judgments, value ascriptions, the link between evaluation and motivation, objectivity in ethics, practical rationality, personal integrity, and coherence of attitudes. J.  David Velleman  is an Emeritus Professor of Philosophy at New York University and William  H.  Miller III Research Professor of Philosophy at Johns Hopkins University. His papers are available in open-­access volumes published by Open Book Publishers and Michigan Publishing.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

1 The Horizons of Humanity David Sussman

One of the most striking features of the work of Christine Korsgaard is her aspiration to derive substantive moral norms from the idea of rational agency, simply as such. In so doing she joins Kant in his endeavor to legislate for all rational agents, regardless of the ways those agents might differ with respect to their biology, psychology, or culture. In Kant, this project goes beyond even the ambitions of the Critique of Pure Reason. Supposedly, the epistemological-­ cum-­metaphysical claims of the first Critique hold only for those rational beings who encounter their world, as humans must, through the sensible intuitions of space and time as they are brought under concepts in some way. Kant does not pretend to show that space and time are the only forms of sensible intuition possible. He allows that there might be other kinds of discursive thought in which such formal categories as substance, cause, and even reality are given radically different interpretations, thereby constituting realms of nature completely distinct from our own. Yet Kant shows no such reserve when it comes to the realm of freedom and, with it, morality. At least by the second Critique, Kant comes to recognize an important asymmetry between practical and theoretical thought. Theoretical cognition depends upon the deliverances of contingent forms of sensible intuition in order to give determinate content to its basic concepts. In doing so, these forms of intuition help establish just what is to count as an object in this context, how such objects are to be individuated, what sorts of powers and forms of interaction they can possess, and so on. In contrast, practical cognition does not depend on any such prior intuitions to provide determinate sense to its fundamental concepts (such as action, end, motive, intention, goodness, responsibility, etc.). Although Kant realizes that sensibility has an important role to play in practical life, he maintains that the character and significance of such feeling is consequent upon the Moral Law, rather than being a prior psychological given that could ground or constrain that law. This is what Kant’s detractors, more than anything else, find so incredible: his David Sussman, The Horizons of Humanity In: Normativity and Agency: Themes from the Philosophy of Christine M. Korsgaard. Edited by: Tamar Schapiro, Kyla Ebels-­Duggan, and Sharon Street, Oxford University Press. © David Sussman 2022. DOI: 10.1093/oso/9780198843726.003.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

2  David Sussman conceit of legislating not just for human beings, but for angels, robots, Martians, and all other possible finite beings, insofar as they are able to confront practical questions at all. Kant even goes so far as to claim that the Moral Law must apply to God, despite the fact that God supposedly has no body, exists outside of time, and knows the world immediately in a way that depends on neither sensations nor concepts.1 In his later works, Kant seems to abandon the goal of legislating morality for all rational agents. The second Critique presents our ethical life as springing from a “Fact of Reason,” the supposedly immediate recognition that we stand under some unconditional obligations and so are absolutely responsible for our actions. The appeal to such a fact (or “deed,” Faktum) suggests that Kant is no longer ruling out the possibility of other forms of rational agency that might be grounded in profoundly different understandings of how and to whom we are accountable. Here things get pretty obscure, even for Kant. In a puzzling footnote to Religion Within the Limits of Mere Reason, Kant distinguishes our essential predispositions in a way that intimates that there might be beings who manage to partake in “humanity” (finite rational agency) without participating in “personality” (autonomy, as defined by recognition of the Moral Law): For from the fact that a being has reason does not at all follow that, simply by virtue of representing its maxims as suited to universal legislation, this reason contains a faculty of determining the power of choice unconditionally . . . . Were this law not given to us from within, no amount of subtle reasoning on our part would produce it or win our power of choice over to it. Yet this law is the only law that makes us conscious of the independence of our power of choice from determination by all other incentives (of our freedom) and thereby also of the accountability of all our actions.  (Religion 6:26n)

Yet in the Doctrine of Virtue, Kant goes on to tell us that with the loss of “moral feeling” that comes with “moral death,” our humanity “would dissolve (by chemical laws, as it were) into mere animality and be mixed irretrievably with the mass of natural beings” (Metaphysics of Morals 6:400). Unfortunately, Kant never develops this cryptic suggestion, leaving unclear whether this moral mortality characterizes humanity as such, or merely our humanity, allowing the possibility of other forms of finite rational agency that might sustain themselves outside of ethical life as we know it. Sorting out these issues in Kant would require a deep dive into the murkiest regions of his metaphysics of nature and agency, where we would have to

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

The Horizons of Humanity  3 come to grips with transcendental freedom, timeless choices of an agent’s entire “empirical” character, and the inescapable but still culpable “radical evil” in human nature. Remarkably, Korsgaard manages to rehabilitate the Kantian project in a way that avoids any metaphysical entanglements that might offend a proper naturalistic sensibility. (I think Kant does so as well, but that’s another story.) Although her arguments shift, Korsgaard never wavers in her resolve to show how a recognizably Kantian form of morality realizes a fundamental commitment implicit in any sort of self-­conscious, rational agency: that is, in the self-­understanding of any being capable of reflecting upon and deciding what to do. In The Sources of Normativity, Korsgaard contends that any sort of practical reasoning on my part presupposes that I embrace a basic normative conception of myself as simply a human being, governed by a distinctively Kantian standard of universalizability. In Self-­Constitution, Korsgaard argues that the formal function of all action, regardless of its specific material goals, is to integrate the agent as an agent, and so make it possible for there to be any determinate facts about what he really wills in the first place. Supposedly, the unity of the self that makes action possible requires that our practical reasons be unconditionally “sharable” in a way that once again commits us to Kant’s Moral Law. In both of these central works, Korsgaard may appear to have succeeded in deriving substantive moral norms from an analysis of an unmoralized understanding of rational agency and action, simply as such. However, I believe that both accounts fall short of this audacious (if not quixotic) goal. In what follows, I contend that what Korsgaard actually establishes is something less dramatic than she herself thinks, although still quite significant. The proper moral of her story is not that Kantian norms of universalizability, publicity, and reciprocity are presupposed by rational agency per se. Instead, what Korsgaard really shows is only that these norms are implicit in the distinctively modern, liberal form of self-­consciousness that we happen to inhabit as the result of wholly contingent historical processes (contingent both metaphysically and rationally). Admittedly, this more humble conclusion might appear to relativize Kantian morality to a particular time and place in a way that would threaten the Moral Law’s supposedly unconditional normative authority. However, I argue that while a certain kind of moral relativity does emerge from my reading of Korsgaard’s conclusions, such relativity turns out to be almost completely innocuous for practical purposes, much as the ontological relativity that follows from the first Critique places no interesting limits on the achievements and ambitions of the natural sciences.2

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

4  David Sussman Ultimately, I believe we will have to give up the ambition of showing morality to be grounded in the fundamental character of rational agency simpliciter. This result does not mean that rational agency must have some different, non-­ moral nature. Rather, we need to see that “rational agency as such” isn’t something that has a real, determinate character at all, any more than do the ideas of an object or cause, simply as such.3 What I hope to show is that when we think in terms of any of these abstractions, we are not getting to the fundamental nature of anything, despite framing issues in the most general terms. Instead, such abstractions often turn out to be nothing more than logical placeholders, setting forth the form of an answer to some basic question, but not constituting any such response themselves. Instead, questions of nature and essence gain purchase only when these placeholders are further interpreted in the context of the right sort of conceptual background: in the case of rational agency, a particular kind of “ethical life.” At first, my reinterpretation of Korsgaard’s conclusions may seem to deprive the Moral Law of a certain kind of important objectivity. But once we see that this standard of objectivity isn’t the right one to bring to practical concerns in the first place, the corresponding (purely formal) sense of relativity will similarly lose its bite. Like Kant, Korsgaard’s defense of the Moral Law does not ultimately achieve the apodictic certainty for which they both strive; nevertheless, her defense of the Moral Law turns out to be good enough for self-­government work.

1 In The Sources of Normativity, Korsgaard presents Kantian morality as the solution to a problem pressed on us by our capacity to reflect upon our own mental life. Unlike other animals, human beings are able to direct their attention to the grounds of their own beliefs, desires, and feelings, and so consider whether those grounds adequately support such thoughts. Korsgaard observes that, as self-­reflective beings, we stand in need of reasons; that is, potential justifications for our continued adherence to a belief or intention. Such justifications allow us to recognize these attitudes as aspects of our own doing, rather than as just psychological states that we experience and observe within ourselves. Korsgaard contends that we can avoid such self-­estrangement only insofar as we embrace, at least implicitly, some “practical identity.” A practical identity is a kind of background normative self-­conception, “a description under which you value yourself ” (Sources 3.3.1, p. 101).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

The Horizons of Humanity  5 Of course, there are many different ways someone might value herself, with many different kinds of practical significance. A human being might value herself as a physical commodity, a volume of water and assorted hydrocarbons, or as a chunk of meat that fetches a certain price on the dark web. An individual could value himself merely as a tool or resource, as a self-­identified slave might. For Korsgaard, however, a practical identity is a way in which “you find your life to be worth living and your actions to be worth undertaking” (Sources 3.3.1, p. 101). Such a self-­conception specifies a way in which I can  value myself as a person, as something immediately important as the bearer of something like rights, duties, prerogatives, liberties, and liabilities. Fundamentally, a practical identity defines a way a person can esteem or respect herself as an agent, and so how she might experience pride, shame, or embarrassment. As Korsgaard’s discussion develops it becomes apparent that my practical identities also determine who else I could take seriously with respect to such practical self-­evaluation, picking out those agents whose responses would count when it comes to how I think and feel about myself. My practical identities define the ways in which I do or could matter to myself, and so who else does or could matter to me, and how. To recognize such an identity is to locate myself within a horizon of possible interpersonal relationships, oriented with respect to those whom I might be able to take seriously, those by whom I expect to be taken seriously, and just what such “seriousness” comes to within this particular social space. Korsgaard contends that only reasons that satisfy the norms of such a practical identity can adequately address what she calls “the normative question.” To answer this question, my reasons must do more than merely demonstrate that it is true that someone in my situation should act in some way, as a piece of propositional knowledge. An adequate response to the normative question must also immediately engage our concern in a non-­alienated way. Such an answer must speak not just to the question, but to the questioner, to me (rather than just someone in my circumstances, however precisely described). Only an appeal to such identities can answer the normative question in an irreducibly practical rather than a merely theoretical spirit. Such answers are addressed to me foremost as an agent thinking in a way that is irreducibly first-­personal and practical. Korsgaard contends that our reasons for action are properly grounded in our practical identities insofar as such identities receive our on-­ going “endorsement” or self-­identification. More specifically, my obligations are the distinctively practical necessities that stem from the principles that are constitutive of some such identity. For Korsgaard, our most fundamental moral

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

6  David Sussman obligations are grounded in an ur-identity (“humanity”) that we must all recognize in order to coherently affirm any other practical identity at all. Supposedly, the normativity of obligation is just a more richly conceptualized form of the normativity of life in general: the normativity that any animal’s needs have for it. For Korsgaard, all our duties are essentially matters of basic self-­maintenance and self-­preservation.4 In an animal, self-­preservation involves the continual reproduction of an individual organism, that is, a living body insofar as that body is engaged in just such a continuous activity of self-­reproduction. At the most basic level of consciousness, anything that threatens this self-­constituting activity is experienced as pain: as something that must be attended to at the risk of ceasing to fully be oneself. For Korsgaard, our sense of moral obligation is really just a fancy version of such pain, now developed and articulated in ways appropriate to a being whose defining form of life is not just conscious, but essentially self-­conscious. To feel that I ethically must do something is to recognize my situation as one in which my life (as a Catholic, or an aristocrat, or an American, or my children’s father, etc.) is somehow very much at stake. The demandingness of obligation, running together cognition and concern, is thus no more “queer” than the special urgency of pain; sadly, one the most familiar and unmysterious things in the world. Such “inherent prescriptivity” can indeed appear occult, but only because it is much too close to us to be seen clearly, rather than because it is infinitely distant.5 We may indeed feel at a loss as to how to explain such normativity, but this may only be because we cannot imagine anything easier to grasp than such experiences that might serve as a way of making them more intelligible.6 So far, Korsgaard has shown only that a person’s practical identity must be analogous to the implicit self-­awareness and self-­concern possessed by any animal, a position that leaves the content of such identities still pretty much up for grabs. However, Korsgaard observes that by solving the first problem caused by reflection, the endorsement of a practical identity engenders yet another difficulty of the same type at a higher level. Perhaps I am wondering why I shouldn’t plagiarize an obscure but useful passage, but then realize that doing so would conflict with my endorsement of myself as an academic and as a philosopher (if nothing else). This response is a fine reply to my original worry, but in so answering that initial question about plagiarism, I now find myself having to face my endorsement of this academic identity as a further act in which I am engaged. At this point, the possibility of alienation now attaches to this background activity that has now been brought to my attention. Now I must consider whether my embrace of an academic identity is to

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

The Horizons of Humanity  7 count as something that I am really doing, or if such endorsement will count as just a psychological hang-­up that I find myself gripped by, something that is merely happening to and through me (as with a compulsion). Unfortunately, my acts of identification are no more immune to such anxieties than are my more immediate desires. The normative question can just keep reinscribing itself: is this newly identified identification really an expression of das Ich, or merely das Es? This process of self-­estrangement and self-­reconciliation keeps spiraling outward to increasingly abstract conceptions of myself, until ultimately I wind up wondering why I should identify with any particular practical identity at all.7 For Korsgaard, this last challenge can be decisively answered only by appeal to the supposed practical identity of a rational agent, simply as such. For as we have seen, any rational agent, if able to make up their mind about what to do in the first place, must recognize some practical identity or other, and so the affirmation of this highest identity gives us at least some reason to adopt any more specific identity that is minimally up to the task. Admittedly, from this lofty perspective we will not find any further reason to adopt one particular identity rather than another: to satisfy the basic needs of rational agency, any coherent practical identity will do. However, the lack of such a more complete justification would be a real problem only if we approached the normative question completely uninvested in any such self-­ conception in the first place: and so, like Buridan’s ass, faced a range of options that were all equally desirable. Our basic self-­definition would then, like the ass’s choice of meal, depend on an initial moment of profound arbitrariness. Yet for good or ill, human beings can never encounter problems about themselves in this form. Instead, we confront such normative questions only insofar as we are already “thrown” into a particular practical identity that we find ourselves inhabiting and endorsing (however unreflectively). The only real issue I can face is whether to continue to affirm that endorsement as my own activity, or instead to disown that attitude as merely a psychological quirk that I am going to have to find a way to work around. The situation is a little like contemplating your love for someone. I would be at a loss if, without any prior attachments, I had to choose someone to love out of all the perfectly love-­worthy people of the world (assuming that a person can be worthy of love in some sense). Fortunately, I can only confront this question already deeply “biased” toward my spouse (or child, or parent, etc.). From this position, I might then wholeheartedly reaffirm my devotion, if only because it is a very good thing in general for a person to have love in their life (so long as there aren’t any strong reasons against loving this particular

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

8  David Sussman person (e.g., their being manipulative, or cruel, or a Yankees fan.)). A rational being needs reasons for what she does, and so she must endorse some practical identity or other. Yet although that need is very general, it can suffice to settle any real worries about the specific practical identities in which I already see myself (so long as no defeaters obtain). So far, self-­reflection has brought me to acknowledge my implicit endorsement of the practical identity as simply a rational agent. As a rational being, I accept that I need some reasons or other, and in so seeing this need as my need, I implicitly affirm my identity as a rational being, simply as such. I recognize the demands of rationality (whatever they are) to be demands on me, and so take myself to be accountable to other rational agents simpliciter. Yet while I have now become conscious of my ongoing embrace of this very abstract identity, further reflection need not alienate me from that self-­ conception as it had done before. In the case of other (more substantive) identities, once I realized that I was endorsing that identity, the question of whether or not that very act of endorsement itself is truly expressive of me was able to reinscribe itself. Yet that alienating question cannot similarly find purchase when I come to face myself simply as a rational agent. What is different in this case is that, unlike previously, I cannot see myself as performing some further “act” of endorsement that might be called into question, some optional affirmation from which I might in turn become estranged. The practical identity of “rational agent” is unique in that, unlike any of my other identities, I cannot make sense of what it would be not to see myself in this way. Admittedly, I might be able to form a coherent idea of a purely theoretical intellect that observes and interprets the doings of its body and psyche in a purely spectatorial way (Kant even proposes this form of subjectivity as a thought experiment in the opening chapter of the Groundwork8). Yet although we can allow for the logical possibility of such a purely theoretical form of self-­consciousness, I nevertheless cannot accept such a self-­understanding as a real possibility for me. As a being that already inhabits some practical perspective, I cannot recognize such a purely contemplative creature as a possible version of myself. Such a contemplative being could be at best just a replacement for me (even if identical to me in all physical respects). Because I could not become alienated from my identity as a rational agent, I cannot see myself having to perform any further act of endorsement with respect to that self-­ conception. I can make no sense of the suggestion that I might do or have done anything else (I might just as well “endorse” my own birth). Instead, my commitment to rationality is, from within a practical point of view, less a voluntary act of endorsement than the irresistible acknowledgment of a

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

The Horizons of Humanity  9 necessary truth, the horizon against which I can recognize any particular practical questions at all.9 Admittedly, there can be no (non-­question-­begging) reasons to abide by the norms of practical rationality itself.10 This limitation would be a problem if we approached the ultimate normative question from some completely neutral position, as if we were considering whether to understand ourselves merely as intellects or also as agents. But, as before, this situation is not one in which human beings could ever find themselves. Rather, we can come to such dilemmas only from already within the practical, from some substantive background understanding of ourselves as agents rather than as mere contemplators. Perhaps a purely theoretical form of mindedness might be possible for some kind of thinker, but this possibility is not open to anything I could recognize as myself (any more than the fact that it is possible for some human beings to be Bronze Age warriors shows that I could be such).11 This necessity is neither causal nor logical but rather phenomenological or existential. Here the token-­reflexive (“I,” “me,” “mine”) cannot be replaced with any name or other referring expression without profoundly changing what is under consideration, and how. At this point, Korsgaard may seem to have assembled all the materials needed to reconstruct Kantian morality. As a condition on the reason-­ giving force of any other considerations, I must first recognize myself as a rational agent, and so as bound by norms that hold for myself and all other rational agents, simply as such. Here it would be tempting to jump to the conclusion that rational agency must then be the “source” of all values, having authority over any other concerns, regardless of whether such agency is realized within myself or somebody else. If this were the case, it might seem that our reasons are any good only insofar as they can be sincerely addressed to any being that could coherently be the object of any such address. Korsgaard would then have succeeded in deriving robust ethical norms of publicity, universality, and reciprocity from the nature of practical reflection, simply as such. Korsgaard does sometimes suggest these inferences but apparently realizes that they would be too quick. After all, even if rational agency can be shown to be the “source” of all other value or values in some sense, this fact would not entail that rational agency is itself the supreme, authoritative value. As has often been pointed out, it is not generally the case that the source of F is itself supremely F (soil is a source of food, but nevertheless inedible; Picasso produced great beauty, but was not himself particularly good-­looking, etc.). There are many ways for one thing to be the source of the value of another,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

10  David Sussman and only some of these ways involve a transfer or delegation of some power from a greater authority to a lesser. To make sense of what the basic human identity involves, Korsgaard argues that all practical reasons must be “public” or “sharable” in a robust sense. Following Wittgenstein, Korsgaard observes that normative concepts (rules, reasons, meanings, etc.) only apply where ideas of violating such norms can find purchase. An agent cannot be following a rule merely because she is sincerely trying to do so; her perspective must, in principle, be answerable to some different point of view. This consequence rules out the radical privacy that Kant calls “logical egoism”: that some reason or principle applies to just me, only here, merely now. Admittedly, this conclusion by itself does not usher in the Kingdom of Ends. The rejection of absolute privacy need not drive us to the opposite extreme of universal publicity, so that norms must be taken to apply to everyone, everywhere, always. Obviously, there is a great deal of space between logical egoism and logical universality. To avoid an objectionable privacy, it is enough that an agent sees her reasons as applying to some other perspectives in ways that can be situated in some sort of normative practice. Although there cannot be a logically private language, it does not follow that the grammatical or semantic norms of English must be universal, or even that English be fully translatable into every other language (or potentially comprehensible by every other language-­user, simply as such). Korsgaard is certainly right that reasons must be “sharable” in some sense. The real question, though, is sharable how, and with whom? As I understand her, to share a practical reason is really to engage in a process of shared reasoning, in which some group of people make up their minds together, as part of what is essentially one complex deliberative act. Korsgaard need not deny that an individual can ever make up his mind all by himself, any more than the impossibility of a private language keeps me from coming up with a personal code that no one else ever learns. The point is only that the possibility of such individual decisions or meanings depends on a background of essentially public, shared normative practices. A person can come up with a secret code purely for her own use; but such a code could not be her first or only language. Similarly, the possibility of my having some completely idiosyncratic values depends on a broader ethical practice, because it is only against such a background that I can understand myself as any sort of agent or valuer at all. However, this requirement shows little about the specific content of any such background practice. Private property presupposes some sort of public,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

The Horizons of Humanity  11 political authority that makes legal and economic relations possible. Yet this condition does little more than rule out the more egregious metaphysical fantasies of libertarians (e.g., Lockean labor-­mixing). The dependence of private property on some kind of prior political institutions does not by itself entail that people must be able to sell their organs, or alienate their labor-­power, or that the means of production must or must not be owned in common. Although any of these positions might be true, none can be derived from the transcendental conditions of property in general. Instead, such claims stand in need of some more specific ethical/political justification. To be any kind of normative consideration at all, a reason must at least be sharable with some others in the sense that these others would count as having the standing to judge whether that reason has been applied correctly or not. Their agreement or disagreement with my judgment would have to matter to me in my judging. To make possible practical reasons, these other judges must in addition have a kind of distinctly practical (or ethical) standing. As Aristotle holds, to draw a practical conclusion is not to accept some sort of proposition about what would be good or right to do in some set of circumstances. A theoretical judgment about a practical subject is not itself a practical judgment. Instead, to draw a practical conclusion is simply to act with a specific intention (or, derivatively, to form an intention to act in the future). If so, then for other people to agree with my practical judgment is to act together as a result of making up our minds jointly. Insofar as I recognize these other agents as fellow participants in such a collective endeavor, I must take their attitudes toward me to have a special standing with regard to how I evaluate myself as a practical being (that is, in terms of my sense of desert, merit, or self-­worth, as matters of pride, shame, guilt, disappointment, regret, glory, etc.). In order to be in a truly practical kind of agreement with others, I must understand my individual agency in terms of what I do or might owe somebody as a fellow deliberator—so that my self-­worth is situated within some practice of holding one another responsible. It is these structures of mutual accountability that transform a group of individuals acting in similar ways (either cooperatively or competitively) into a truly joint enterprise, constituted by a “general will” prior to the decisions of any particular members. This conclusion, although broad, is hardly an “empty formalism.” Arguably, the dependence of practical reasons on some such background practice of shared reasoning and action rules out at least rational egoism. Admittedly, the rational egoist holds herself accountable to her future self with respect to her long-­term welfare, and so does not succumb to logical egoism. Nevertheless, such purely intra-­personal forms of accountability fall short of being a real

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

12  David Sussman practice simply because the ideas of mutual disagreement, argument, advice, and deliberating together cannot find adequate purchase within the life of just one being. I may now successfully anticipate my future views, and my future self may accurately recall and respect my present positions, but the arrow of time rules out the possibility of any real give and take between these points of view. At the very least, real practices need participants that can exist at the same time, but in different spaces. Similarly, it may also be the case that no agent could inhabit a completely freestanding “Kingdom of Two,” just because the relations between two people cannot have enough complexity or open-­ endedness to count as a real normative practice.12 Yet we have not yet reached a completely universalistic understanding of our most basic practical identity, in contrast to the more parochial forms of ethical life that have characterized the vast majority of human history. To get to recognizably Kantian conclusions, Korsgaard has to show that the most basic conception of this formal identity should be that of fellow citizen-­ legislators of a universal “Kingdom of Ends.” In such a Kingdom, I understand myself, with every other rational agent, as engaged in an essentially joint activity of determining the basic standards of our common life, so we are all equally accountable to one another, regardless of any of the ways in which we might differ as individuals. What is distinctive about this practice is that it has completely open borders; the only qualification of citizenship is the ability to assert a citizen’s rights, by holding others responsible (and taking oneself to be responsible to others). To defend this radically egalitarian and universalistic vision, Korsgaard appeals to the phenomenology of our moral experience. Supposedly, when we interact with others, even strangers, the structure of our reactive attitudes (such as resentment, remorse, and gratitude) reveals that we cannot help but recognize such others as beings to whom we owe justification in a richer moral sense, simply as persons to other persons. It may indeed be possible for some human beings to fail to recognize the objections of other agents to be directly addressed to them merely as such (rather than just as a Jew from fellow Jews, as a Greek aristocrat from other Greek aristocrats, etc.). Even so, there seems to be no way for you and me (given our place in history) to fail to experience this relation as that of just one person to another. As a good liberal modern, I take it for granted that such qualities as my gender, race, class, and religion are, metaphysically speaking, accidental. However important and unavoidable these categories are in my actual social life, I cannot but allow the possibility that I could have been female, or black, or Muslim (regardless of

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

The Horizons of Humanity  13 whether or not there is any remaining possibility that I might become such, given who and what I already am). Yet however inescapable for me, this Kantian form of self-­understanding is nevertheless a fairly recent invention. In the past, human beings have taken qualities such as caste, religion, or gender to be essential to the people they are, and so simply did not recognize any sorts of ethical claims addressed to them from others outside of such concrete identities.13 Korsgaard does not establish that these more limited conceptions of an agent’s basic practical identity must be intrinsically defective (assuming that these conceptions are not dependent on confusion, fantasy or self-­deception). At best, she shows only that such self-­ conceptions are no longer open to us, given the particular ways we have been already been initiated into rational agency. The possibility of practical reason may indeed require that we inhabit some kind of ethical life, taken broadly. However, a specifically Kantian conception of this life is not the consequence of a general transcendental argument. Instead, this morality (like every other) is ultimately the expression of a kind of groundless self-­confidence: a fact not of reason simpliciter, but of what reason can be in and for us, here and now.

2 I do not think there are any reasons to be dissatisfied with this more modest rendering of Korsgaard’s conclusions. However, in her subsequent Self-­ Constitution, Korsgaard redoubles her efforts to establish the absolute necessity of Kantian morality for all rational agents. Her arguments here appeal to a teleological metaphysics of action. Korsgaard contends that for a being to be able to truly act (and not just respond to inner or outer forces), that being must define and govern itself by a fundamental principle, just as a polity is defined and made capable of action by a legal constitution. Such a principle determines what activities count as the agent truly exercising her powers (in response to her desires and other incentives), and what counts as her being acted upon by something else. As such, these basic commitments give the agent’s soul a special kind of integrity. This unifying scheme establishes the distinction between what behaviors are to count as the doing of the agent’s whole self, and what responses are to be merely the products of some external influence or sub-­personal component of her psyche. Constitutional principles serve to integrate (differentiate and unify) the parts of the soul so as to make action possible, and this service is something

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

14  David Sussman that all agents need qua agents. From this necessity Korsgaard infers that the essential function of such basic commitments is to articulate and integrate the agent in this way. The more stable or durable the resulting unity, the better the constitution qua constitution. Supposedly, the principle that provides for the most unshakable integrity is the one that most fully does its job, and so is a practical principle in the fullest sense. For Korsgaard, Kant’s Moral Law is the only constitutive principle that an agent can address to any other rational agent (and so any possible version of herself) under all conditions. As a result, the agent who recognizes this law can thereby recognize what counts as her own proper choices in any possible situation, regardless of how drastically she or her environment may have changed. In contrast, the unifying principles of defective souls only render a clear verdict under a limited range of circumstances. The timocratic soul pursues honor above all else, not distinguishing between meriting the esteem of the right sort of person, and actually being so esteemed. Supposedly, such an agent would lose his bearings should these two considerations come apart: when doing the truly honorable thing would elicit only scorn and contempt from the people who matter, so that only a shameful action will be admired. As a result, the honor-­lover cannot see any way of remaining true to himself in such contexts; either way, he must betray the fundamental commitment that defines him. In contrast, the autonomous Kantian person can recognize and affirm what it would be to act as herself in any possible context: “not necessarily happy on the rack, but herself on the rack, herself even there” (Self-­Constitution, 180).14 Korsgaard is right that only the autonomous soul has such integrity needed to recognize itself in any possible context. However, what she fails to establish is that proper agency requires this maximal degree of durability. It does not follow from the fact that an agent’s self-­conception must have some degree of integrity in order to act that the stronger that unity, the more fully that being acts, and so is better qua agent. A car must be able to hold together under some stress, but that doesn’t mean that a Humvee is the best or truest kind of car, or that my old Honda Civic is defective because it can’t withstand a roadside bomb. With agents as with cars, once a certain minimal threshold of integrity is attained, other considerations may come to the fore in ways that can justify various trade-­offs with respect to further increases in durability (my Civic gets much better mileage and is easier to park just because it lacks much armor). As Korsgaard contends, a fundamental principle is needed for me to be able to reproduce and recognize myself in other practical contexts. Yet

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

The Horizons of Humanity  15 although a principle of self-­constitution enables me to recognize myself in a wide range of possibilities, this constitution does not have to guarantee that I will be able find myself in all conceivable situations. It may be perfectly all right for an agent to know that she must lose a grip on herself in some possible set of (exceptional) circumstances, so that she could not continue to recognize herself no matter what she might do (Sophie’s Choice comes to mind). Indeed, a vulnerability to coming undone this way may be a necessary concomitant of anything that can count as love or personal attachment: and so, of particularly rich kinds of agency and forms of interpersonal relationship.15 At bottom, action may be less a simple functional kind like “can-­opener,” and more like speech. Speech (and language-­use generally) serves many different functions simultaneously, and the nature and extent of these tasks are continually being redefined by linguistic practice itself. Some of the functions of language may indeed be more central than others (e.g., communication vs. exhortation). Yet such centrality does not entail that the language that best facilitates such sharing of belief is most fully a language, let alone the best language simpliciter. After all, an ideally communicative language might still make it very difficult to issue an order, make a joke, or give thanks. Of course, a language that makes basic communication very difficult is defective. However, this fact doesn’t show that ease of communication is the only or even dominant linguistic virtue, or that facilitating such exchanges always takes priority over everything else we might be trying to do with our words. Languages are constituted by norms and normative practices centered in certain tasks, but there need be no such thing as the better language as such. Past some threshold, there may just be different ways of being better or worse which do not allow for any determinate ranking overall (at least, outside of a specific context). Likewise, Korsgaard’s and Kant’s autonomous agent may enjoy much greater integrity than the minimum required for action, but this does not mean he has attained a fuller kind of agency in any absolute sense. Kantian agency is just better in one important respect, perhaps to the detriment of others. There is no reason to suppose that all the potential virtues of agency could be completely unified or even specified, any more than all potential excellences of speech could all be fully realized in a particular language. Perhaps there simply is no fullest or most real type of agency. Instead, there may be many different types of agency, realizing different virtues of action to varying degrees, without there being any neutral standpoint from which to rank them as forms of agency in itself. There would then be many ways of being an agent, involving different sorts of trade-­offs, and so indefinitely

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

16  David Sussman many ways to act well or poorly. Once again, the analysis of action only shows that an agent must identify herself in terms of some broader ethical practice. But that analysis does not specify any particular version of such a practice, not even a Kantian one. Such substantive conclusions depend not just on what we are as rational agents in general, but on who we can be, given the specific self-­understandings in which we can actually find ourselves.

3 Ultimately, Korsgaard manages to establish Kantian morality only relative to a certain modern form of subjectivity. Yet little of value is lost by this qualification. Such relativity does not entail that a person must be particularly enlightened or even decent for Kantian morality to apply to him. Even the bigot must, on some level, implicitly recognize the moral address of the people he despises and denigrates; otherwise, why would he feel such a need to deny or diminish their apparent moral standing? Here hatred may be the homage that self-­conceit and self-­deception pays to human dignity. Kantian morality does indeed bind all liberal moderns not just in their more virtuous (that is, coherent and articulately self-­aware) forms, but even in those degenerate cases presented by the avowedly illiberal (racists, reactionaries, Proud Boys, etc.). Even so, we must still allow that there can be (and perhaps have been) forms of practical self-­consciousness where the moral standing of the stranger is neither affirmed nor denied in any real way. In pre-­modern societies, human beings may have often seen themselves as standing in substantive ethical relations only with fellow members of their clan, class, or caste. These groups would define the people to whom the agent was answerable in a rich way, before whom they might feel proud, ashamed, or disappointed. The claims of complete strangers would not need to be either recognized or repudiated; being simply inaudible in the second-­person, no rationalization or self-­deception would be required to ignore them.16 In these cases, agents simply might not address themselves to, or take themselves to be addressed by, strangers, who might then appear as nothing more than potential resources or threats (as in relations of trade or war). Kantian morality would then be an answer, not to the normative question as such, but only to such queries as they could issue from my mouth, or the mouths of those with whom I could hope to form substantive interpersonal relationships. Other agents might well get by with different and perhaps weaker answers to the normative questions they confront, relating to strangers in a purely Hobbesian way.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

The Horizons of Humanity  17 Yet while a medieval samurai or Neolithic hunter-­gatherer need not be impressed by Korsgaard’s (or Kant’s) results, these conclusions may nevertheless be inescapable for anyone who might actually find themselves reading her work. This more reserved interpretation opens the door to moral skepticism only if it turns out that there is indeed some real chance of our abandoning a basically liberal self-­conception in favor of something more medieval or Neolithic, but such forms of ethical life are almost certainly no longer real possibilities for us. The problem is not simply that being a medieval samurai or Bronze Age warrior requires broader social practices and intuitions that no longer exist and are beyond our knowledge and power to restore. More important, these earlier forms of self-­understanding might well require a certain kind of unreflectiveness that, having been lost, cannot be regained without doing violence to mind and spirit. An essential part of these earlier self-­conceptions was the inability to entertain certain logical possibilities; the idea that one’s fundamental ethical identity might be independent of a particular class, gender, or tribe was something that just could never be taken seriously. In contrast, we are able to form such a conception of our ethical relations and recognize ourselves in its terms, however vehemently we might then try to convince ourselves that we can’t. As Bernard Williams observes, the attempt to recover such earlier self-­understandings would be an attempt to regain a kind of lost innocence, to forget or unlearn a basic possibility that has been central to our lives up to now. As he warns, such an effort would constitute only a futile if grotesque exercise in self-­delusion (e.g., the Third Reich).17 There is one way in which this humbler reading of Korsgaard’s conclusions qualifies their normative force. Although we need not worry about objections posed from pre-­modern perspectives, we cannot foreclose the possibility of new, distinctively “post-­modern” challenges. We are not able to say from the outset that there could never be some new ethical self-­conception that, while not demanding that we unlearn our basic reflectiveness, is nevertheless incompatible with the Kantian way of understanding ourselves as agents. If the Moral Law were an implicit commitment of valuing or rational agency per se, we could dismiss the possibility of such challenges out of hand. Instead, we can only note that no such alternatives are currently in sight, shifting the burden of argument to the skeptic to put some substantive proposal on the table. Should the skeptic manage to do so, we will then just have to consider that supposed alternative in detail: whether it is coherent, consistent with other things we know, and truly independent of the Kantian self-­understanding that it pretends to replace. Finally, even if such a proposal passes these tests,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

18  David Sussman we would then have to consider whether we can imagine actually inhabiting this new self-­conception, given who we are now, and, if so, what would be gained by doing so, and what lost. These are all high hurdles to clear, but perhaps a sufficiently bold visionary might do so (Zarathustra the godless?). Until such a figure arrives, we can only stand ready to meet these challenges as they come, with confidence rather than certainty.

Notes 1. Kant does allow that God’s experience of the Moral Law would be importantly different from our own. As discursive beings, we must confront the Law as an imperative, leaving open how we will actually act. For God, in contrast, the distinction between the normative and the descriptive breaks down; the Moral Law simultaneously spells out how God should act and also how he does act. 2. This analogy has its limits. If the first Critique is right, then the truths of the natural sciences hold for all discursively judging subjects who engage with the world through the intuitions of space and time. In a sense, we wouldn’t share a world with those who didn’t have those forms of intuitions; indeed, it’s not clear that we would be able to recognize such creatures as thinking beings at all. In contrast, it does seem that we can recognize beings outside of our basic form of ethical life as rational agents, even if we couldn’t share a social world with them. The difference may be that even when a being isn’t practically intelligible to us, we can still have some theoretical understanding of it as a thinking subject. In contrast, when even this theoretical understanding fails, there is nothing left to fall back on. My thanks to Kyla Ebels-­Duggan for pressing me on this point. 3. The move here recalls Kant’s resolution of the antinomies, which he shows to rest on a confused aspiration to grasp the “world as a whole” in some univocal sense. 4. It’s important to see that this claim is formal, not material. When I act from duty, sustaining my identity need not be part of my objectives, whether explicitly or implicitly. Rather, this sense of myself is found in the peculiar sort of necessity with which I adopt such ends as telling the truth, keeping a promise, or helping someone in need. My identity is found not in the content of my practical reasoning, but in the forms of inference I employ. As a result, moral action need not be egoistic in any interesting sense: I help the needy for their sake, although I do so as a kind of self-­affirmation. There is no greater paradox here than in the fact that a person can be self-­assertive without ever actually talking about herself. 5. See J. L. Mackie, Ethics: Inventing Right and Wrong (New York: Penguin Books, 1977). 6. Johnson was right at least this far; the way to refute the skeptic about “normative properties” is indeed to have them kick a stone (at least if it’s big and kicked really hard). 7. Here Korsgaard seems to conflate the possibility that the normative question will keep reattaching itself with the necessity that it does so. Yet my alienation from any particular demand might well disappear once I see how that claim is connected to some

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

The Horizons of Humanity  19 broader way that I understand and care about myself. It does not follow from my having lost confidence in some duty I have that I must similarly question the identity that grounds it once I become conscious of my endorsement of it. While any answer to the normative question opens up the possibility of such a regress, such alienation will typically come to an end somewhere short of the identity of rational agent simpliciter. We do not always feel a need to answer every question that reflection presents to us: arguably, experiencing such a need is not the essence of rationality, but rather a kind of pathology. 8. Immanuel Kant, The Groundwork of the Metaphysics of Morals. In Practical Philosophy, translated and edited by Mary  J.  Gregor (Cambridge: Cambridge University Press, 1996), 4:395. 9. In effect, I confront this identity as something like Kant’s “Fact of Reason.” Although it is not logically (let alone causally) necessary for a thinking being to see herself this way, this identity forms the horizon of my possibilities: that is, of all the potential ways of understanding oneself that I can recognize as ways of understanding myself. 10. The question “Why be rational?” cannot be answered, but only because it cannot really be pressed. No helpful answer could be given to anyone who asked this question in earnest, yet that question is not something a rational being needs to answer in order to properly remain confident in her commitment to rationality. Where reason ends, questioning and answering is over as well. 11. Here I have in mind Bernard Williams’s distinction between “real” and merely “notional” possibilities for how a person might understand herself. A real possibility, unlike the merely notional, is one that a rational agent can inhabit without depending on ignorance, fantasy, or self-­deception, and which makes her social world intelligible to her. See Ethics and the Limits of Philosophy (Cambridge, MA: Harvard University Press, 1985), p. 161. 12. My thanks to Nir Ben Moshe for pressing me on this point. 13. I don’t doubt that these pre-­modern agents might have had some understanding of the fact that outsiders were making claims on them or expressing emotions of resentment or disgust. My point is only that such understandings would be purely theoretical, not practical. A truly practical appreciation of resentment or anger would approach these attitudes not as objects of psychological explanation (or amusement), but as demands immediately calling for a reply of the same general kind, in the second rather than the third person. 14. Or more precisely, that there will be some fact of the matter of what it would be for her to succeed, or fail, at remaining herself on the rack. Whichever turns out to be the case depends not on her commitments, but on her subsequent actions. 15. See my “Morality, Self-­ Constitution, and the Limits of Integrity.” In Beatrix Himmelmann and Robert Louden, eds., Why Be Moral? (Berlin: Walter De Gruyter, 2015), pp. 123–40. 16. I offer this claim only as a piece of anthropological speculation, although one familiar enough from Hegel and Nietzsche. So long as this possibility is credible, not much here depends on the particular historical facts. 17. See Bernard Williams, Ethics and the Limits of Philosophy (Cambridge, MA: Harvard University Press, 1985), pp. 161–2.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

20  David Sussman

Bibliography Kant, Immanuel. Religion within the Boundaries of Mere Reason. In Religion and Rational Theology, translated and edited by Allen  W.  Wood and George Di Giovanni. Cambridge: Cambridge University Press, 1996. Kant, Immanuel. The Groundwork of the Metaphysics of Morals. In Practical Philosophy, translated and edited by Mary  J.  Gregor. Cambridge: Cambridge University Press, 1996. Kant, Immanuel. The Metaphysics of Morals. In Practical Philosophy, translated and edited by Mary J. Gregor. Cambridge: Cambridge University Press, 1996. Korsgaard, Christine  M. The Sources of Normativity. Cambridge: Cambridge University Press, 1996. Korsgaard, Christine M. Self-Constitution: Agency, Identity, and Integrity. Oxford: Oxford University Press, 2009. Mackie, J. L. Ethics: Inventing Right and Wrong. New York: Penguin Books, 1977. Sussman, David. “Morality, Self-Constitution, and the Limits of Integrity.” In Beatrix Himmelmann and Robert Louden, eds., Why Be Moral? Berlin: Walter De Gruyter, 2015. Williams, Bernard. Ethics and the Limits of Philosophy. Cambridge, MA: Harvard University Press, 1985.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

2 Finite Valuers and the Problem of Vulnerability to Unmitigated Loss Sharon Street

1. Introduction The Holy Grail of secular analytic metaethics—or, better, the “Non-­Holy Grail”—would be a clear, informative, intuitively satisfying account of the subject matter of ethics that both (1) vindicates a strong form of ethical objectivity, while (2) avoiding any hint of metaphysical or epistemological mystery. In the past, I’ve been pessimistic about the existence of the Non-­Holy Grail, having elsewhere defended a rather strongly subjectivist view.1 That being said, I believe that it would be premature to give up the search, in part because I’m convinced that in our quest for the Non-­Holy Grail, we (the “we” here being those of us who work in secular analytic metaethics in the broadly Western tradition) haven’t even begun to exhaust the available philosophical resources. In particular, I’m convinced that Eastern meditative traditions, including Advaita Vedanta and Buddhism, among others, offer a wealth of resources that are, so far, vastly underexplored by secular metaethicists. In this paper, I will mention in passing some ways in which I think meditative traditions are relevant, but that will not be the main point of the discussion.2 Instead, the main point is to summarize what I think is the most promising strategy in our search for the Non-­Holy Grail, and to get to work pursuing one part of that strategy. In the paper, I will be making two background assumptions, both of which I defend elsewhere. The first background assumption is that theism is false.3 There is no God, in the sense of an omniscient, omnipotent, and perfectly good being. More broadly, the universe has no concern for the living things that arise as a result of natural processes within it. This is not because the universe is malevolent, of course, but rather because it has no concerns at all. Only finite valuers—roughly, living beings with a conscious, evaluative Sharon Street, Finite Valuers and the Problem of Vulnerability to Unmitigated Loss In: Normativity and Agency: Themes from the Philosophy of Christine M. Korsgaard. Edited by: Tamar Schapiro, Kyla Ebels-­Duggan, and Sharon Street, Oxford University Press. © Sharon Street 2022. DOI: 10.1093/oso/9780198843726.003.0002

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

22  Sharon Street perspective on the world—have concerns, and we’re alone in the universe apart from each other. The second background assumption is that normative realism is false.4 There are no robustly mind-­independent normative truths, in the sense of truths that hold independently of the evaluative standpoints of finite valuers like ourselves. This is so, in my view, no matter what form of normative realism is proposed—whether naturalist, non-­ naturalist, or quasi-­ realist, for example. Instead, normativity is mind-­dependent in the following sense: If some fact or consideration X has any normative significance (e.g., if it is good or bad, or a normative reason for or against some belief, feeling, or action), then X has that normative significance only for a given valuer, and in virtue of that valuer’s evaluative perspective on the world.5 It is widely recognized, at least among philosophers, that the rejection of theism is compatible, at least in principle, with a strong form of normative objectivity. Secular forms of normative realism have gained enough traction to make this area of conceptual space vivid. It is less widely recognized that the rejection of normative realism (whether secular or theistic) is also compatible, at least in principle, with a strong form of normative objectivity. How so? Well, the rough idea is that even if normative significance is, as the antirealist claims, “conferred” by minds (and, in particular, by the minds of finite valuers) upon otherwise neutral facts or considerations, it might still be the case that every time you have a mind (and, in particular, a valuing mind), that mind’s state or activity of “valuing” somehow dictates or implies the same thing about the normative significance of a given fact or consideration, no matter who the valuer is. A view along such lines is a tantalizing metaethical possibility for many reasons, and in the history of secular analytic metaethics since Moore, no philosopher has done more to draw our attention to this possibility than Christine Korsgaard. In a body of work spanning nearly four decades, Korsgaard has developed a comprehensive metaethical and ethical theory which shows, both by example and in its explicit articulation, how one might simultaneously (1) understand normativity as a thoroughly mind-­dependent phenomenon (thereby avoiding the metaphysical and epistemological problems that plague realism—or so I would argue), while also (2) vindicating a strong form of ethical objectivity. Even if one is skeptical that the details of Korsgaard’s proposed vindication of morality go through, Korsgaard has— drawing on her reading of Kant—mapped out a general strategy for combining normative antirealism with ethical objectivity that is of extremely widespread philosophical relevance and interest. The strategy she identifies

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Finite Valuers and Vulnerability to Unmitigated Loss  23 provides a template not only for anyone looking to vindicate ethical objectivity within an antirealist framework but also for anyone looking to vindicate any kind of normative objectivity within an antirealist framework.6 In any case, as the above remarks suggest, I believe that for anyone who hasn’t given up on the search for the Non-­Holy Grail, there is no better stepping-­off point than the path that Korsgaard’s work in metaethics has laid out for us. I’ll now very briefly characterize that path, which I’ll call the generic constructivist strategy for vindicating the objectivity of ethics, and then begin sketching an alternative implementation of that strategy.

2.  The Generic Constructivist Strategy Glossing over many important details,7 and drawing heavily on Korsgaard’s way of articulating the constructivist strategy in terms of the search for a universal “problem” to which morality might be the universal “solution,”8 we may describe the generic constructivist strategy as consisting of three steps: Generic Constructivist Strategy for Vindicating the Objectivity of Ethics Step 1: Explain what is constitutively involved in being an agent, valuer, or will.9 Step 2: Identify a universal problem that is faced by every agent (valuer, will, etc.), by her own evaluative lights, simply in virtue of her being an agent (valuer, will, etc.) at all. Step 3: Argue that the best or only solution to the relevant problem involves adopting an ethical principle or standpoint. We may summarize Korsgaard’s more specific, Kantian-­inspired implementation this way: Korsgaard’s Kantian Implementation of the Generic Constructivist Strategy10 Step 1: Explain what is constitutively involved in being a will. Step 2: Argue that every will has the problem of “needing a law” that is not “given from the outside.” Step 3: Argue that the solution to that problem is the categorical imperative as represented by Kant’s Formula of Universal Law.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

24  Sharon Street For reasons I discuss elsewhere,11 I don’t think that Korsgaard’s specific implementation of the strategy succeeds in the end. But it would be an egregious case of throwing the baby out with the bathwater to give up on the generic strategy on that slender basis. Furthermore, we should not assume that all implementations of the generic constructivist strategy must be Kantian—or, for that matter, even committed to an exclusively ra­tion­al­is­tic account of the epistemology of ethics. On the contrary, there are any number of ways in which one might try to implement the generic constructivist strategy. In current work, I’m trying to develop one such alternative implementation— an implementation which, in its fully developed form, will draw in a variety of ways upon the resources of Eastern meditative traditions. In extremely cryptic outline, the implementation I have in mind may be summarized this way: Alternative Implementation of the Generic Constructivist Strategy Step 1: Explain what is constitutively involved in being a valuer. (The concept valuer is the one I find it most useful to start with.) Step 2: Argue that no matter what the substantive content of a given valuer’s evaluative standpoint on the world, if he is a finite valuer “in the business of action,”12 then he faces a certain problem: what I will call the problem of vulnerability to unmitigated loss. Step 3: Argue that the best solution to this problem is to seek out and occupy what I call the standpoint of pure awareness. First, point to the standpoint of pure awareness in a philosophically informative way by describing it as a “way of seeing” that may be achieved, among other ways, by way of meditation practice.13 Next, argue that the standpoint of pure awareness not only (i) provides a solution to the problem of vulnerability to unmitigated loss; but also (ii) constitutes an accurate seeing of reality; and (iii) entails the adoption of an ethical perspective. In this paper, I will be focusing exclusively on Step 2 of the Alternative Implementation, hoping to arrive at a clear statement of the relevant problem, and taking a stab at arguing for the maximally ambitious claim that the problem in question is universal to finite valuers (or, more precisely, finite valuers who are “in the business of action”). There won’t be time to say anything about Step 1, except in the most cursory fashion required for Step 2.14 Nor will there be time to say much of anything about Step 3. But as will become clear, Step 2 turns out to be more than enough to deal with in one paper.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Finite Valuers and Vulnerability to Unmitigated Loss  25

3.  Four Clarifications of the Task of “Step 2” As we just saw, the task of Step 2 of the generic constructivist strategy is to identify some problem P that every valuer has by his own evaluative lights, regardless of the particular substantive content of his set of values. Four clarifications are needed, however, before we begin the work of trying to identify such a problem. The first two clarifications concern the sense of “having a problem” that is relevant. The third clarification distinguishes between the task of Step 2 and a related task with which it is easily confused. The fourth clarification concerns argumentative strategy. First, it is consistent with a valuer’s “having a problem” in the relevant sense that she does not currently, and indeed may never, recognize the problem as hers. We are all familiar with the idea that you can have a problem (e.g., a drug problem or an anger problem) without realizing that you do, and the problem identified by the constructivist is no different in this respect. The constructivist’s claim will be that there is a problem that every valuer does indeed have, by her own evaluative lights, when these evaluative lights are combined with the non-­normative facts. But that does not imply that any given valuer will ever recognize or agree that she has the problem. There are any number of ways in which this could happen: It could happen if she is ignorant of, or incapable of recognizing, the relevant non-­normative facts, or if she is, for whatever reason, unable to see or draw out the logical and instrumental implications of her own deepest values. So it’s not an objection to the claim that some proposed problem P is a universal problem of valuers merely to point to actual valuers who don’t think they have the problem. It’s possible they’re in denial or simply haven’t realized it yet. Second, it is also consistent with a valuer’s “having a problem” in the relevant sense that they already have, in their possession, a solution to it—maybe a perfect and complete solution, or maybe just a makeshift and partial one. For example, there is a perfectly good sense in which anyone reading this paper has the “problem” of obtaining food, water, clothing, and shelter. These are ongoing challenges we face in virtue of deep (in this case, biological) features of ourselves. However, as the mere fact that we have time to do philosophy shows, we are among those lucky human beings who happen to be in possession of an ongoing solution to these biological problems. The universal problem located by the constructivist, while emphatically not a biological problem, is similar in the sense that it’s not an objection to the claim that some proposed problem P is a universal problem of valuers merely to point out that some valuers don’t have the problem because they’ve already solved it. After all, the constructivist’s hoped-­for ultimate conclusion is that a certain

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

26  Sharon Street standpoint on the world—and in particular, an ethical standpoint, cashed out in a certain way—provides a solution to the relevant problem. So it would make sense that plenty of people already have the problem in question pretty well under control. Third, it is important not to confuse the task of Step 2 with another task that one frequently sees undertaken in moral philosophy, namely the task of locating a self-­interested reason to adopt an ethical perspective. This is not what we’re after here. We’re instead on the hunt for a quarry that one might worry is even more elusive and improbable, namely a universal reason to adopt an ethical perspective, where by “universal” I mean a reason that applies to any and every valuer, by his own evaluative lights, no matter what the particular substantive content of his values. The upshot of this point is that when seeking to execute Step 2 of the constructivist strategy, the valuers we need to keep in mind aren’t just ones with mostly, or exclusively, self-­interested values, though of course this is one important and salient category of case. Rather, as best we can, we need to try to think about any and all valuers— including ones who are deeply other-­directed, altruistic, or indeed already in possession of recognizably ethical values. I emphasize this third point because the problem that I ultimately want to focus on—the problem of vulnerability to unmitigated loss—is one that for many of us (where by “us” I now mean human beings in general) is often especially acutely felt and faced “from the standpoint of ” our ethical concerns themselves—the ethical concerns that many or most of us in fact already possess. Roughly speaking, the losses we fear and face are not just personal losses; they are losses to others we care about, and potential losses to all of humanity and animal life, such as those associated with climate change. This observation might at first sound problematic for my implementation of the constructivist strategy. After all, if an “ethical standpoint” is ultimately supposed to be the solution to the problem in question, then isn’t it bad news if people who already have ethical values are still, in many cases, suffering from the problem? Here I can gesture only briefly at the reply. But the rough idea is that the “solution” in question is not going to be just any old version of the ethical standpoint, but rather a specific interpretation of it. In other words, the solution is going to be the ethical standpoint cashed out in a certain way (and in particular, in way that I think is especially clearly suggested in the Upanishads and developed more fully in the Advaita Vedanta tradition, though the same idea is present in many other literary, poetic, philosophical, and religious traditions).15

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Finite Valuers and Vulnerability to Unmitigated Loss  27 A fourth and final point. The sheer ambition of the metaethical constructivist strategy—that it aims to locate a problem that faces any valuer by his own lights—means that a successful argument is going to need to operate on multiple levels. Here I’ll mention three. First, in order to make it even remotely plausible that the proposed problem P is universal, one will need to try to show how certain formal features of being a valuer inevitably generate the problem. Arguments to this effect will proceed abstractly, seeking not to presuppose anything about the particular substantive content of the evaluative standpoints being talked about. I do some of that kind of work in this paper. Second, one will have to consider salient apparent counterexamples—valuers who seem not to have the problem—and other problematic cases as they arise. I try to do some of that too, though not remotely enough is done here.16 Third and finally, to make it plausible that the problem in question has the kind of teeth that it needs to provide every valuer with not only a reason, but a strong reason, to seek out the alleged solution, it is important to speak to the problem as it manifests for us human beings. That is what explains why, in some parts of the paper, I speak freely about the general kinds of things “we” want and dread and so forth, where the concerns mentioned are typical human concerns. Part of the idea here is that in order to get a proper grip on the nature of the problem and its strength, we cannot just consider it from a purely formal, “outside” point of view. We also need to look at the problem in the specific, substantive guise in which it presents to us. The hypothesis is that even valuers with very different substantive concerns face a formally identical problem, presented to them with all the force of whatever their own evaluative standpoint on the world happens to be. I’ll now start working to pinpoint the problem I have in mind. In the next section, I’ll begin with a rough, intuitive statement of the problem and then take a first stab at a philosophical account, noting a constitutive connection between valuing and vulnerability. In the remainder of the paper, I’ll work to develop the account in stages, moving through a series of five refinements (Sections 5–9) on the way to a sharper statement of the problem. I conclude in Section 10 with some speculative remarks about next steps.

4.  First Stab: Valuing and Vulnerability To start with a rough, intuitive statement of the problem: We are small, finite beings in a very big universe that does not, in itself, care a whit about how

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

28  Sharon Street things turn out for us. And yet, as valuers, we are invested in how things turn out—in what happens to ourselves and the people and things we love—and this makes us deeply vulnerable. Indeed, we are so vulnerable that a clear-­ sighted awareness of that vulnerability can have a corrosive effect on our ability to go forward, leading to crippling anxiety or depression, for example. Our hopes and loves, conjoined with our finitude, make us vulnerable to loss—indeed, in many cases guarantee it—and yet we don’t want to abandon our hopes and loves. How best to cope with this situation? Is there a way of coping that doesn’t involve some version or other of denial, distraction, hardening, giving up, or numbing out? Or just white-­knuckling it through and hoping for the best? How should we live with the ever-­present threat (and in many cases guarantee, for those of us who never want to die or grow old or be separated from the ones we love) of losses that we reasonably fear we might not be able to bear? To begin the process of developing this intuitive statement into a philosophical account, let’s start with the following observation. However exactly we characterize the attitude of “valuing,”17 it is plausible that constitutively involved in the attitude of valuing is taking there to be either a gap, or at least the potential for a gap, between the way the world is, as a matter of fact, on the one hand, and the way it would be good or desirable for it to be, on the other.18 This is not meant to be a substantive or particularly illuminating point in itself, but rather merely a conceptual observation—one that I hope strikes you as obviously correct. If the observation is correct, then—now shifting to put the same point in terms of a “problem” that presents from the standpoint of the valuer herself— one problem that is necessarily built into the standpoint of any valuer is the problem of either a gap, or at least the potential for a gap, between the way the world is, as a matter of fact, on the one hand, and the way it would be good or desirable for the world to be, on the other. In another shift that is not meant to introduce anything controversial, but rather just help us talk about the problem as it presents to the point of view of valuers themselves, I’ll now introduce a somewhat technical use of the term loss. I’m going to use the term loss in the following sense: Whenever there either already exists, or opens up, a “gap” between how things are and how, from a given valuer’s point of view, it would be good or desirable for things to be, and whenever that gap either persists, widens, or otherwise won’t go away in spite of that valuer’s hopes, wishes, or efforts, then the existence of that gap is what I’ll call a “loss” for that valuer.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Finite Valuers and Vulnerability to Unmitigated Loss  29 In many contexts, this use of the term loss is entirely natural and intuitive. If, for example, a beloved person, object, or way of life is lost or destroyed, then this constitutes an opening up of a “gap” in the relevant sense, and is also, obviously, a “loss” in an entirely standard sense of the word. Still, my usage is significantly broader than everyday usage. In particular, we may distinguish between “loss” in the narrow sense that requires having had something in order to be able to lose it, and “loss” in a broader sense that doesn’t require that one ever even possessed something in the first place in order to suffer a loss. In this second, broader sense, failure, disappointment, and heartbreak over things never achieved, possessed, or realized count as losses. I use the term loss in this second sense. Finally, in yet a further extension of the ordinary use of the term, loss in the very general sense I have in mind needn’t even involve objects (whether ever possessed or not) at all. All it need involve is some gap or other between the way things play out in the world and the way that the valuer in question thinks they should play out. Suppose, for example, that a valuer holds that it is never acceptable to deny someone a job on the basis of race. Next, suppose that somewhere in the world, someone denies someone else a job on the basis of race. (The people involved may be unknown to our valuer.) In my extended sense of “loss,” this event—in which the “is” of the world fails to match up to our valuer’s “ought”—counts as a “loss” from the evaluative standpoint of this valuer.19 Using the term loss in this broad sense, then, the rough idea is that to be a valuer necessarily involves vulnerability to loss. And that vulnerability to loss is, from the valuer’s point of view, a “problem.” We’re far from having adequately pinpointed the problem that I want to isolate, but it’s a start.

5.  First Refinement: The Possibility of Invulnerable Valuers What I’ve said so far might seem to suggest that every conceivable valuer faces the problem in question—roughly, the problem of vulnerability to loss. But this is not so. To see this, we need to consider the possibility of what I’ll call invulnerable valuers. There are at least two types of invulnerable valuers we might imagine.20 The first type consists of valuers who are invulnerable in virtue of being omnipotent. I won’t spend any significant time on this possibility, because I take it to be clear that the universe does not contain omnipotent valuers. Individual, conscious evaluative points of view on the world arise at

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

30  Sharon Street particular places in space and time, due to particular sorts of configurations of matter, and those configurations are continually changing and impermanent. Any being with a conscious evaluative standpoint on the world depends for its existence on a particular ongoing organization of matter, which will eventually move out of existence. In more everyday terms, valuers are finite; we are subject to illness, accident, aging, death, and more generally circumstances beyond our control. (The connection here with points emphasized in the Buddhist tradition should be clear.) The second type consists of valuers who are invulnerable in virtue of the nature of their ends. As a far-­fetched example, consider the case of a valuer who values the fact that 2 + 2 = 4. Or, as a more realistic example, consider the case of a valuer who values the fact that he already accomplished something— writing the Great American Novel, say—a fact that the rest of time can do nothing to change. Valuers who are invulnerable in virtue of the nature of their ends are important to consider. Here are a couple of brief observations. First off, notice that in order to be this type of invulnerable valuer, it is not enough that one values the above sorts of things; it must also be the case that this is all one values, and such cases are hard to imagine. Nevertheless, it seems to me that we can imagine such valuers. They wouldn’t live very long, since they wouldn’t be motivated to pursue the means necessary for living or doing anything, but we can still imagine them. If this is right, then it’s not true that every conceivable valuer is vulnerable to loss. The possibility of invulnerable valuers shows that we’ve failed to find a universal problem facing valuers as such, even when we set aside omnipotent valuers and focus on finite ones. The next question is whether this puts a damper on the larger philosophical aim that we are pursuing—namely, seeking to vindicate the objectivity of ethics by way of the generic constructivist strategy. My claim is that it does not. For notice that invulnerable valuers who are invulnerable in virtue of the nature of their ends have no normative reasons for action.21 They might be agents in the sense that they are capable of acting for normative reasons, but due to their happy combination of values and the state of the world, there is nothing at all such that they have normative reason to do it. For this reason, invulnerable valuers are not part of the audience that a vindication of ethics needs to address. If they take themselves to have reason to do anything—and a fortiori to do something unethical—they are making a mistake on their own terms (which, according to the constructivist metaethical view that I’m assuming, are the only terms there are). As I’ll also put it, this subset of finite valuers—those who are invulnerable in virtue of the nature of their ends—are not “in the business of action.” Or rather—and this is what’s relevant to a

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Finite Valuers and Vulnerability to Unmitigated Loss  31 vindication of the objectivity of ethics—they shouldn’t be. If they are seeing things clearly and not making any mistakes about the non-­normative facts, then they will see that there is no normative reason for them to act and they will simply be, evaluatively content with the world just as it is.22

6.  Second Refinement: The Problem Isn’t “Vulnerability to Loss” as Such What I’ve said so far might make it sound like I think the problem in question is being vulnerable to loss at all—such that the solution to the problem, if there is a solution, would involve becoming completely invulnerable to loss of any kind. But I take it that for many or most of us, it’s obvious that complete invulnerability to loss is not what we want. Here several points are relevant. First of all, clearly it would depend on how that invulnerability to loss is to be achieved. Dying is one way of becoming “invulnerable,” for example, but that’s not what we want. Another way to become invulnerable would be to rid oneself entirely of all hopes, loves, and desires while somehow still remaining alive. Set aside whether this is even possible; even supposing it were, I take it that for many of us, the option has no appeal. The widely shared intuition I have in mind is captured by the saying that it’s better to have loved and lost than never to have loved at all. Given a choice between (1) a life of love coupled with the vulnerability to loss that love entails, and (2) the invulnerability that never having loved would have given us, many of us would choose the former. The widely shared intuition I have in mind is also captured by the immediate negative reaction that many people have to Buddhism or Stoicism upon first, superficial encounter with those views. At first glance, both views can give the impression of recommending that we rid ourselves of vulnerability by ceasing to desire or care at all. While I think this is not the correct way to interpret either of these views, my point here is just that if that’s the suggestion, it’s one that many of us find wholly unappealing. Loss is no fun, but the solution isn’t not to play the game. A second point is this. Suppose one could remain emotionally vulnerable to loss, but then somehow order up, in advance, a life free of any loss whatsoever. Of course we cannot do that, but here too the point is that even supposing we could, it’s an option that many of us would also find unappealing. However much we all hate failure, setbacks, and losses, it would seem a shallow, immature perspective to want every single aspect of one’s life to go exactly as one wants. What would life be without ups and downs, successes and failures, and

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

32  Sharon Street opportunities to fall flat, make mistakes, and grow? We’re adults, one is tempted to stay; we don’t expect, or even want, a life that is utterly free of any loss at all. Being vulnerable to loss, and suffering it sometimes, is part of the package that we love when we love life, and life as an agent. So the problem isn’t “vulnerability to loss” per se. Still, one might say—and now we’re getting closer to the real problem—there are limits to the “package” one can love; there are limits to the degree of loss, and vulnerability to loss, that one can wittingly embrace and still retain any sense of joy or enthusiasm for life, and life as an agent. And when we pay honest attention to the degree of vulnerability that we finite valuers face, it is far from clear that it is compatible with a stable, clear-­eyed, sustainable embrace. How so? Many people won’t need persuasion on this point, I know, but for those who do, let me say a bit more. It is a fact, easily established by observation, that over and over again in this world, individual human beings, as well as large groups of them at once, suffer losses of grotesque and unthinkable proportions. Or rather, they suffer losses that would have been unthinkable—impossible to imagine in all their concrete detail and level of horror—but for the fact that they happened and are well-­documented. Sometimes these losses are at the hands of other human beings;23 other times they’re the result of pure accident combined with human stupidity;24 still other times, they’re the result of a mindless causal process such as a tsunami.25 In the context of debates about the existence of God, losses of the kind I have in mind are sometimes called “horrendous evils” and are said to raise the “problem of evil” for theism. Now of course if one is an atheist, then one doesn’t face the “problem of evil”—not, anyway, in the sense of a problem for one’s philosophical position. Unfortunately, there are worse problems than problems for one’s philosophical position, and even if one is an atheist, there is another, far worse “problem of evil” that one has to face. And that other “problem of evil” would be the problem of . . . um, evil. More precisely, the “problem of evil” faced by the atheist is evil itself, and the associated moral and philosophical question of how to live in full awareness of the fact that it happens, happens often, and that when it does, it is often utterly unmitigated—in the sense that there is nothing whatsoever redeeming about it, nothing at all that makes it “good” or “okay” in any sense at all. Call this problem, admittedly tendentiously against the theist, the real problem of evil. The real problem of evil, in short, is that horrific things happen to finite valuers; they are horrific in virtue of being horrific to the valuers themselves; and, since no

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Finite Valuers and Vulnerability to Unmitigated Loss  33 one else exists, there is nothing else to say. As my label suggests, I believe that this problem exists whether one thinks it does or not. In particular, it exists not only for atheists but for theists as well—on top of their other, more familiar “problem of evil,” which is merely a problem for their philosophical position. Talk of “evil” is unhelpful, though, because it has connotations that are in tension with the whole picture of the universe I am assuming in this paper. Instead, I will continue to talk about loss—but (and now here’s the point) not just any old loss, not run-­of-­the-­mill loss of the sort any grown-­up ought to be able to handle and embrace as part of life—but rather devastating, unmitigated loss. The category of loss I have in mind here consists of losses that are devastating on a valuer’s own terms, and unmitigated in the sense that there exists no further, unseen, compensating reality that somehow renders the loss “okay” for the valuer. This is the category of loss such that if one pays attention to it, fully registering its reality, it has the potential to undermine one’s ability to go forward as a finite valuer. Another way of putting the point is this. In his essay collection At the Mind’s Limits: Contemplations by a Survivor on Auschwitz and its Realities, Jean Améry describes the experience of being tortured by the SS, and what he calls the loss of “trust in the world”: I don’t know if the person who is beaten by the police loses human dignity. Yet I am certain that with the very first blow that descends on him he loses something we will perhaps temporarily call “trust in the world.” Trust in the world includes all sorts of things . . . [It includes] the certainty that by reason of written or unwritten social contracts the other person will spare me— more precisely stated, that he will respect my physical, and with it also my metaphysical being . . . At the first blow, however, this trust in the world breaks down . . . The expectation of help, the certainty of help, is indeed one of the fundamental experiences of human beings . . . The expectation of help is as much a constitutional psychic element as is the struggle for existence. Just a moment, the mother says to her child who is moaning from pain, a hot-­water bottle, a cup of tea is coming right away, we won’t let you suffer so! I’ll prescribe you a medicine, the doctor assures, it will help you . . . But with the first blow from a policeman’s fist, against which there can be no defense and which no helping hand will ward off, a part of our life ends and it can never again be revived.26

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

34  Sharon Street Améry describes the loss of “trust in the world” as something that happens when a person undergoes a certain kind of horrific experience: an experience in which one learns that the world (whether in human form or non-­human form) is not coming to one’s aid. The kind of trust in the world that Améry describes is so basic that even if it’s clearly unwarranted from an abstract, intellectual point of view, it’s much harder to lose at a deeper, visceral level unless one actually has the trust blown out of oneself by an experience of the kind Améry describes for us. In principle, though, we shouldn’t need a first-­hand experience to realize that “trust in the world” is not justified. The well-­documented experiences of others should be enough. After all, we have Améry’s account, and countless other such accounts from history and the news. We don’t doubt their veracity. Do we think we’re different from these other people—that the world somehow cares for us when it didn’t care for these others? Presumably not when we really think about it. Our own “trust in the world” usually isn’t, but should be, shattered simply by the evidence. What we see happen to other finite valuers makes it clear that while things can be sacred to us—that is, to finite valuers who love things and hope for things and are attached to things—nothing is sacred to reality. That is, no matter how sacred one might think something is—a mother’s love for her child, a person’s trust in her fellow human beings, someone’s dearest, most innocent hopes for his own life or that of others— that thing has been violated, violated completely, and indeed is fairly routinely violated, by reality. Let’s try this, then: The problem isn’t loss per se; the problem is the nature and extent of the losses that we know to be both actual and possible—the kinds of losses that make “trust in the world” unwarranted.27 To put it another way: We’re okay with vulnerability to loss, we might say; it’s vulnerability to devastating, unmitigated loss that is the problem. This is the kind of vulnerability awareness of which has the potential to eat away at our agency from the inside—crippling or sapping our ability to continue in effective or enthusiastic pursuit of our ends at all.

7.  Third Refinement: Mellow and Low-­Aiming Valuers In response to these points, one might agree that vulnerability to devastating, unmitigated loss is a problem—for some valuers, anyway. But then one might object that it’s a vast overstatement to suggest, as I just did, that this problem is a universal one, faced by any finite valuer, no matter what the particulars of

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Finite Valuers and Vulnerability to Unmitigated Loss  35 his evaluative makeup. More specifically, there seem to be at least two ways that a given valuer might be structured, such that while he is vulnerable to loss in a run-­of-­the-­mill, manageable sort of way, he is not vulnerable to devastating loss. First, there’s the possibility of extremely mellow valuers. A valuer might simply be extraordinarily easygoing—valuing plenty of things, and therefore vulnerable to loss, but with a low-­key attitude that makes him relaxed even about the possibility of losing everything. Second, there’s the possibility of extremely low-­ aiming valuers. A valuer might be vulnerable to devastating loss—roughly in the sense that he is deeply emotionally invested in certain outcomes, such that if the loss were to occur it would be devastating—but his ends might be so modest, so minimal, so exceptionally easy to achieve or see realized, that we might think he isn’t in any meaningful sense genuinely vulnerable to that loss. And of course one might be both extremely mellow and extremely low-­aiming at once. These seem like real possibilities. Indeed, they might even sound a bit like people we know. What to say about them? One point is that such valuers are not as common as one might think. For example, in the case of extremely low-­aiming valuers, it is important to remember that while a given end on a particular occasion might well be trivial to the point of being fail-­safe, what it will have taken to get to that point in the valuer’s life is almost surely not trivial at all. After all, getting to the point of being able to execute even the most modest aim requires having developed as a living organism and maintained oneself at least long enough to form the end and execute it. So if one takes this longer view, executing even an incredibly low aim is a tall order. (Moreover, it’s generally going to require not having been all that mellow about things like maintaining one’s bodily integrity, etc.) Still, I want to concede that extremely mellow and low-­aiming valuers are perfectly conceivable. With this in view, I also want to concede that the problem of vulnerability to devastating loss is not universal to all valuers “in the business of action.” But that doesn’t mean there isn’t an important constitutive connection. In my view, the right way to think about agency is as falling along a complicated, somewhat messy multi-­dimensional spectrum. In thinking about mellow and low-­aiming valuers, we’ve identified two traits (among many others) with respect to which a given agent may vary, independently of the particular substantive content of his values (Figure 2.1). extremely mellow

extremely low-aiming

Figure 2.1  Two dimensions of variation in valuers

extremely invested

extremely high-aiming

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

36  Sharon Street As I’ve just conceded, if a given agent falls toward the far left end of the spectrum for either trait, then she is not meaningfully subject to the problem of devastating loss. That being said, notice what happens as we approach the limit of the left end of the spectrum with regard to either trait. At the limit, an extremely mellow valuer approaches a state of not valuing anything at all. Thus she shades into the territory of not being a creature who has normative reasons to act at all.28 She therefore also shades into the territory of a form of life with respect to which a vindication of the objectivity of ethics need not be concerned. Meanwhile, at the limit, an extremely low-­ aiming valuer approaches a state of being an invulnerable valuer in the sense we looked at earlier—a state in which her values are such that she essentially cannot fail. So this agent too starts shading into the territory of not being a creature who has reasons to act at all, and therefore also not part of the audience that a vindication of ethics needs to address. Thus, although the tie between valuing and vulnerability to devastating loss is not airtight, there is a constitutive connection once we move away from the left end of these two spectra. Lastly, even a very mellow or low-­aiming valuer is subject to total loss, in the sense of total failure on his own terms, even if those terms are very modest and relaxed and unlikely. Insofar as that is true, he is also subject to whatever anxiety his nature can muster on behalf of any improbable, ­low-­stakes outcome that he cares about. Henceforth, I will use the term unmitigated loss as an umbrella term meant to encompass both devastating loss and total loss (with the idea being that not all devastating loss is total, and not all total loss is devastating). The term unmitigated is also meant to draw attention to the feature that I emphasized in the previous section, which is the atheistic point that the losses in question, whether devastating or total, are not compensated by some unseen reality that somehow makes them “okay” for the valuer. They are not okay, for that valuer, and that’s the end of it.

8.  Fourth Refinement: The Problem Isn’t Suffering As I noted earlier in passing, there are obvious similarities between some of the points about loss that I’ve been emphasizing and points about the impermanence of life that are a hallmark of the Buddhist tradition. In this section, I want to briefly highlight one important difference between the position I’m

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Finite Valuers and Vulnerability to Unmitigated Loss  37 trying to develop and “the” Buddhist position, at least as the latter is most commonly understood.29 As is often pointed out, Buddhism is a deeply practically oriented view— seeking to identify, diagnose, and then solve a very deep problem that all human beings face. This “problem and solution” structure to the view makes it potentially well suited for marriage with the generic constructivist strategy. The lead seems especially promising given that there is, according to the Buddhist tradition, a deep tie between the proposed “solution” and an attitude of compassion for all sentient beings (an ethical standpoint). I think the lead is indeed very promising, and there is a great deal from the Buddhist tradition that I ultimately want to take up in one form or another. But I also think certain adjustments to the Buddhist view (at least as most commonly understood and presented) are called for. And one such adjustment is in our characterization of the problem. According to the Buddhist position (again, at least as it is most often expressed), the most fundamental problem faced by human beings is suffering (a somewhat misleading translation of the Pali term dukkhā30). The First Noble Truth states that suffering exists; the next three seek to diagnose the source of suffering and to suggest a path for eliminating it. In my view, however, to frame the problem this way is to get things just a little bit off from the beginning. To cast suffering as the main problem is to misdescribe what the problem really is from the relevant standpoint, which is our standpoint—the standpoint of living (suffering!) beings. The problem, as it presents to us, is first and foremost the gap between what is and what would be good. Suffering is what happens when that gap gets to be too large, too hopeless, too seemingly beyond repair; but the suffering is normally the manifestation of our taking something else—not the suffering itself—to be the problem. The problem, the mother who has lost her child wants to say, is not that I’m suffering: the problem is that my child is dead and never coming back. I’d gladly suffer for a thousand years if you could just bring him back. He is what matters, she’ll say, not my state of mind.31 There is a sense in which this adjustment is cosmetic, but another sense in which it is absolutely not. To put the focus on what is most naturally understood as an experiential state is to build in from the beginning a bias or presupposition in the direction of a certain theory of value—one that I think many of us rightly regard as off-­base and ultimately shallow.32 It is to distort the normative phenomena—to misdescribe the way the world presents to a valuer. The world and other people are generally what present to us as mattering, not our own states of mind.33

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

38  Sharon Street None of this is to say that suffering itself isn’t a problem. It most certainly is, and a big one. This brings us to the fifth and final refinement.

9.  Fifth Refinement: Secondary Aspects of the Problem According to the proposal I’ve been developing, finite valuers who are “in the business of action” are subject to what I’ve called the problem of vulnerability to unmitigated loss, where “unmitigated loss” is loss that’s either devastating or total, or both. As we’ve seen, the more “invested” and “high-­aiming” the valuer, the more subject to the problem the valuer will be. Furthermore, as emphasized in the previous section, I want to insist that the primary problem, from the standpoint of the valuer, is the losses themselves, whether actual or merely potential. Before closing, however, I want to complicate the picture by pointing out several secondary aspects to the problem. In calling these aspects “secondary,” I don’t mean to imply that they are unimportant or any less serious than the primary problem. I just mean to bring out the logical structure of what we might think of as the totality of the problem (with both primary and secondary aspects combined), which we may see as starting with our vulnerability to unmitigated loss, but which can then ramify and spiral into a range of secondary problems that are related to, but analytically distinct from, the original one. To bring out this structure, we need to look again at the attitude of “valuing.” On the account I favor, the attitude of valuing (in its most sophisticated forms) involves certain loose constitutive connections with emotions such as anticipation, hope, joy, dread, relief, anxiety, fear, and sadness, to name a few.34 For example, part of what it is to value X is to feel, other things being equal, comfort and ease when X is safe; joy and relief when X is saved or preserved; anxiety when X is threatened; fear or terror when the threat to X is grave and imminent; sadness and distress at thoughts of the loss of X; grief and heartbreak in the event X is one day lost; panic, sickness, and horror if X is lost suddenly, without warning; and so on. The more “invested” a valuer is in X, the stronger the relevant emotions will be; and the more “high-­aiming” the valuer is (the more X is intricate, fragile, and therefore the harder X is to preserve and protect) the more vulnerable the valuer will be to the suite of negative emotions associated with the actual or potential loss of X. Noting this constitutive connection between valuing and the emotions, we can see how being vulnerable to unmitigated loss constitutively involves vulnerability to a suite of negative emotions that are, in a sense, a form of mental

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Finite Valuers and Vulnerability to Unmitigated Loss  39 pain—emotions that include anxiety, fear, sadness, grief, and so forth. But these negative emotions, in turn, have the potential to give rise to at least four additional problems—problems that are all too familiar to anyone who has ever done significant battle with anxiety or depression.35 First, as is well known, the negative emotions in question, when sufficiently chronic or intense, may interfere with a valuer’s general level of functioning— sapping her energy, interfering with her concentration, taking a toll on her physical health, and eroding her morale, for example. Thus, insofar as the valuer values, as an end in itself, her own general level of functioning (as presumably many, though not all, valuers do), this decline in her general level of functioning is a new, additional problem or loss, and the first of the four secondary aspects of the problem that I wish to point out. Second, if the above-­mentioned decline in functioning occurs, then this may, in turn, bring about (or render more likely) still other losses that are, by the valuer’s own lights, significant. Suppose, for example, a valuer experiences a devastating loss, such as the sudden, unexpected death of a loved one. The valuer is overcome with grief and anxiety. On the account I favor, the grief is a (partially constitutive) manifestation of the value she placed on the loved one who died, and the anxiety is a (partially constitutive) manifestation of the value she places on other loved ones, say—coupled with her new, heightened awareness of their fragility. Next suppose that the grief and anxiety are so intense that they begin to interfere with her ability to function at school, or work, and in her close relationships. This, in turn, puts other cherished ends at risk. The career she dreamed of is now threatened by her decline in functioning at work and school; the relationships she treasures are fraying and starting to fall apart; and so forth. These additional losses are not the same as the original loss, nor are they the same as the loss of her general level of functioning. They are, rather, further (and possibly devastating) actual or potential losses brought about, or made more likely, by the others. These additional losses constitute the second of the four secondary aspects of the problem that I want to mention. (And of course these additional losses may set in motion the same train of events all over again, leading to yet further undermining of functioning and yet further losses.) Third, the negative emotions in question—the ones prompted by the actuality or the contemplated possibility of devastating loss—are intrinsically unpleasant states of mind to be in. The experiences of fear, grief, and anxiety, for example, are intrinsically unpleasant. Thus, insofar as valuers tend to value not being in intrinsically unpleasant states, then simply being in the state of mind of experiencing the emotion will itself be bad by their own lights. The

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

40  Sharon Street valuer may nevertheless view his emotional state as appropriate or called for by the circumstances, and for that very reason may not wish, all things considered, to get rid of it.36 But this is consistent with the point I’m making, which is that simply being in the state, while it may well be appropriate in one sense, is nevertheless bad, by the valuer’s lights, in another sense—namely in virtue of being unpleasant to be in. Thus, our vulnerability to unmitigated loss gives rise to states of mind that are themselves intrinsically unpleasant, which is itself another loss. This is the third of the four secondary aspects of the problem. Fourth, and now arriving at a point that is one of the deepest psychological insights of the historical Buddha: Merely being in the above-­mentioned intrinsically unpleasant states very often gives rise to yet further negative emotions. Grief, fear, anxiety, and the like are a form of emotional pain that valuers naturally tend to greet as we do physical pain, which is to say, with resistance. Our immediate instinct upon encountering these emotions is, oftentimes, to fight them or push them away. When a feeling of grief threatens, for example, we may react with fear and unease. When a feeling of fear threatens, we may react with additional fear, or with anger at the thing or person that triggered the fear. This tendency—the tendency to meet intrinsically unpleasant sensations or mental states with aversion and resistance—is the problem famously described by the Buddha as the problem of “two darts,” where, roughly, the first dart is the arising of an intrinsically unpleasant state, and the second dart is the rejection, the resistance, the fighting of the first.37 As the Buddha saw and argued with great acuity, it’s the occurrence of the second dart that turns a state of pain into a state of suffering. And a state of suffering, of course, is itself yet another, even more intrinsically unpleasant state to be in. So suffering itself is the fourth of the four secondary aspects of the problem. The picture we’ve arrived at is structurally complex. To be a finite valuer in the business of action is to be vulnerable to loss. Vulnerability to loss, as such, is not a problem; but vulnerability to devastating loss is a problem, and the more a valuer tends in the direction of being invested and high-­aiming (as most conscious life forms happen to be), the more vulnerable she is to this problem. But the problem does not end there. As we’ve just seen, there are potentially vicious, spiraling aspects to it. Confrontation with one loss—whether actual or potential—can ramify in the ways we’ve just looked at, creating new (actual or threatened) losses, which can ramify in their own turn. If these ramifications aren’t somehow blocked, or short-­circuited, the end result may ultimately be the valuer’s collapse. Of course we do have ways of blocking

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Finite Valuers and Vulnerability to Unmitigated Loss  41 these ramifications; we wouldn’t be here if we didn’t. Still, to the extent we see the possibility of these systems failing, the possibility of our own collapse— due to an inability to handle the mismatch between what is, and what is good, by our lights—hovers on the horizon as a possibility, however remote. In further illustration of the potentially insidious nature of the problem, this hovering possibility can give rise to an unease that might itself, if not staunched, further precipitate the feared collapse. All of this, as I understand it, is brought about by a sort of generalized version of the problem that led Kant to think we needed God as a practical postulate.38 We have a vision of the good; we take a hard, honest look at reality, and we see that the causal structure of nature is not particularly conducive to that good;39 and this raises questions about how, as a finite valuer, we’re going to get along in this world.

10. Conclusion Suppose you grant, for the sake of argument, that I’ve succeeded in identifying a universal, or at least very deep and widespread, problem associated with being a finite valuer in the business of action. I’d like to close by addressing, very briefly, a worry that one might have in looking ahead to “Step 3” of the generic constructivist strategy. First off, it’s obviously going to be a tall order to show that adopting an ethical standpoint (cashed out in a certain way) offers a solution to the problem of vulnerability to unmitigated loss. But even supposing for the sake of argument that that point could be established, one might wonder: Aren’t there, quite obviously, all kinds of other solutions to the problem of vulnerability to unmitigated loss, such that one will never successfully argue that an ethical standpoint is the best, much less the only, solution? Examples of such alternative strategies—other ways of coping with the gap between is and ought and blocking the kinds of spirals I just described— include the following: (1) Denying the facts (imagining oneself invulnerable; believing one will be reunited with loved ones after death; telling oneself that everything, including the worst atrocities, happens for a good reason). (2) Ignoring the facts (looking away as much as possible; “focusing on the present”; distracting oneself from disturbing possibilities and eventualities with work, pleasure, entertainment).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

42  Sharon Street (3) Hardening oneself (developing a shell; becoming cynical or detached; cultivating invulnerability by shutting down emotionally). (4) Altering one’s values (changing what one cares about, or reducing how much one cares about it, so as to become less vulnerable—in other words, working to transform oneself into a “mellower” or more “low-­aiming” valuer). (5) Numbing out (treating the pain rather than the underlying problem—a familiar solution when the underlying problem is insoluble). (6) Cultivating resilience using causal methods (shoring up one’s ability to hang in there and not fall apart in the face of actual, potential, or guaranteed loss; reminding oneself that one is probably already more resilient than one thinks40). My reply, in brief, is this. Yes, all those strategies are solutions of a kind. Moreover, I think each has its place on particular occasions. However, as general policies—as one’s default, go-­to strategy for coping when there’s an unbridgeable gap between “is” and “ought” from one’s own point of view—each is sorely lacking on terms that are arguably involved in any agential perspective. Such terms include points about what it takes to continue as the agent one already is; points about an agent’s generic need to maintain a reasonably accurate picture of reality;41 and so on. The claim, in other words, is that as general policies for coping with the problem of vulnerability to unmitigated loss, each of the above strategies is in tension with needs that one has from one’s own point of view as a creature who is finite, but who cares about how reality will unfold—and who doesn’t want to give up caring and trying to  shape reality, even while knowing full well that the odds against ever satisfactorily closing the gap between “is” and “ought” (according to us) are very high indeed.42

Notes 1. I describe and defend the subjectivist implications of my view, which I call “Humean constructivism,” in “Coming to Terms with Contingency: Humean Constructivism about Practical Reason,” in James Lenman and Yonatan Shemmer, eds., Constructivism in Practical Philosophy (Oxford: Oxford University Press, 2012), 40–59; “In Defense of Future Tuesday Indifference: Ideally Coherent Eccentrics and the Contingency of What Matters,” Philosophical Issues, vol. 19 on “Metaethics,” ed. Ernest Sosa (2009): 273–98; and “Nothing ‘Really’ Matters, But That’s Not What Matters,” in Peter Singer, ed., Does Anything Really Matter? Parfit on Objectivity (Oxford: Oxford University Press, 2017), 121–48.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Finite Valuers and Vulnerability to Unmitigated Loss  43 2. I say a little bit more about the relevance of meditative traditions to metaethics in “Constructivism in Ethics and the Problem of Attachment and Loss,” Proceedings of the Aristotelian Society, Supplementary Volume (2016): 161–89. The present paper explores many of the same themes as that article, but I believe that this paper moves closer to nailing down the “problem” that I want to argue is universal, or at the very least extremely widespread, among finite valuers. 3. I defend this assumption in “If ‘Everything Happens for a Reason,’ then We Don’t Know What Reasons Are: Why the Price of Theism is Normative Skepticism,” in Michael Bergmann and Patrick Kain, eds., Challenges to Religious and Moral Belief: Disagreement and Evolution (Oxford: Oxford University Press, 2014), 172–92. 4. I defend this assumption in “A Darwinian Dilemma for Realist Theories of Value,” Philosophical Studies 127 (1) (January 2006): 109–66; “Reply to Copp: Naturalism, Normativity, and the Varieties of Realism Worth Worrying About,” Philosophical Issues (a supplement to Noûs), vol. 18 on “Interdisciplinary Core Philosophy,” ed. Walter Sinnott-­Armstrong (2008): 207–28; “Evolution and the Normativity of Epistemic Reasons,” Canadian Journal of Philosophy (35), supplemental volume on Belief and Agency, ed. David Hunter (2011): 213–48; “Mind-­Independence Without the Mystery: Why Quasi-­Realists Can’t Have It Both Ways,” in Russ Shafer-­Landau, ed., Oxford Studies in Metaethics, vol. 6 (Oxford: Clarendon Press, 2011), 1–32; and “Objectivity and Truth: You’d Better Rethink It,” in Russ Shafer-­Landau, ed., Oxford Studies in Metaethics, vol. 11 (Oxford: Clarendon Press, 2016), 293–333. 5. As this paragraph indicates, I am understanding the “realism/antirealism” debate about normativity in the traditional, “Euthyphronic” sense of a debate about the mind-­dependence of normativity. While the rough gloss in the main text is enough for my purposes in this paper, the terminological and philosophical issues are extremely complex, and readers seeking a detailed explanation of how I’m using the terms and understanding the debate should see “Nothing ‘Really’ Matters, But That’s Not What Matters,” especially section 5. The articles cited in the previous note are also relevant. 6. For example, while Korsgaard’s work has revolved around securing the objectivity of moral reasons, her general strategy for vindicating normative objectivity within an antirealist framework is transferrable to other types of reasons, including epistemic reasons. I briefly suggest what a constructivist view of epistemic reasons might look like in “Evolution and the Normativity of Epistemic Reasons.” Kenny Walden has recently proposed a broadly constructivist view of epistemic normativity in his “Foundations for Epistemic Contractualism” (unpublished manuscript). Although Walden does not explicitly endorse a “constructivist” view of normativity in that paper, his account is deeply in keeping with such a view, and elsewhere Walden defends a metaethical view that I would categorize as constructivist. See especially his “Morality, Agency, and Other People,” Ergo 5 (2018): 69–101. 7. For some of these details, see “Constructivism in Ethics and the Problem of Attachment and Loss.” 8. For this way of formulating the strategy, see especially Korsgaard’s “Realism and Constructivism in Twentieth-­Century Moral Philosophy,” Journal of Philosophical Research, APA Centennial Supplement 28 (2003): 99–122. 9. Different implementations of the generic strategy will make different choices about what concept it’s most useful to start with—agent, valuer, will, or something else.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

44  Sharon Street Regardless of the concept chosen, Step 1, properly conducted, involves purely “conceptual” work in the sense that it just involves sketching the most minimal kinds of things that a being needs to be doing in order to recognizably qualify, according to the intuitions of a competent speaker, as an agent (valuer, will, etc.) at all. In undertaking this conceptual work, a crucial point that must be respected is the extraordinarily wide range of substantive “values” that are compatible with a being’s recognizably qualifying as an agent (valuer, will, etc.). Most relevantly when it comes to the search for the Non-­Holy Grail, grossly immoral or amoral values are perfectly compatible, as a conceptual matter, with recognizably qualifying as an agent (valuer, will, etc.). To build moral values into one’s account of what is conceptually involved in being an agent (valuer, will, etc.), as some theories pretty obviously do, not only is implausible as a conceptual matter but also makes the resultant account useless for the purpose of vindicating the objectivity of ethics. Michael Smith’s recently defended version of constitutivism is guilty of this, in my opinion. See Smith’s “A Constitutivist Theory of Reasons: Its Promise and Parts,” Law, Ethics and Philosophy 1 (2013): 9–30. For a useful critique, see Michael Bukoski’s “A Critique of Smith’s Constitutivism,” Ethics 127 (2016): 116–46. 10. This is a summary of the implementation that Korsgaard develops in The Sources of Normativity (Cambridge: Cambridge University Press, 1996), and which she helpfully summarizes in “Realism and Constructivism in Twentieth-­Century Moral Philosophy.” Korsgaard develops her theory further in Self-­Constitution: Agency, Identity, and Integrity (Oxford: Oxford University Press, 2009). 11. In “Coming to Terms with Contingency.” 12. I explain what I mean by the phrase “in the business of action” in Section 5. 13. Crucially, on the view I have in mind, meditation practice is neither necessary nor sufficient for reaching the relevant standpoint. Appeal to the practice is a heuristic, a means of informative pointing. Also importantly, the standpoint in question is not subject to direct characterization in words. Nevertheless, a good philosophical explanation will be able to point helpfully to the standpoint in question, and will also be able to explain clearly (in words) why the standpoint is not subject to direct characterization in words. 14. I say more about the topic of Step 1—i.e., what’s constitutively involved in being a “valuer”—in “Constructivism in Ethics and the Problem of Attachment and Loss”; “Nothing ‘Really’ Matters, But That’s Not What Matters”; and “Coming to Terms with Contingency.” A very brief summary of some relevant points: First, I use the term valuing as a term of art meant to encompass a wide range of attitudes and experiential states. Second, in seeking to give a philosophical account of the relevant family of attitudes, what will not work, in my view, is to try to give a reductive characterization. On the contrary, there is no way accurately to characterize the relevant set of attitudes without invoking normative concepts (e.g., saying things like they’re the attitudes of taking things to be good or required and so forth). Third, the fact that we cannot accurately characterize the attitudes in question without invoking normative concepts doesn’t mean we can’t say anything informative about them and locate them perfectly well for ourselves. For one thing, we can get a fix on the relevant concepts by pointing to circumstances in which we tend to experience things as good, valuable, worthwhile,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Finite Valuers and Vulnerability to Unmitigated Loss  45 called for, etc. (Cf. color concepts.) We may furthermore characterize the relevant set of attitudes by contrasting them with the attitude of mere desire. (In “Coming to Terms with Contingency,” I discuss three points of contrast.) Lastly, we may point to loose constitutive connections with emotions such as anticipation, hope, joy, dread, relief, outrage, anxiety, fear, sadness, and regret. The resulting account of the attitude of valuing will be loose and imprecise, but this is not a problem, for two reasons. First, it simply gets the phenomenon right: for the phenomenon in question isn’t a simple and perfectly unified phenomenon; intuitively, we see a wide range of forms of “valuing” in the animal world, and the account captures this. Second, a loose and imprecise characterization of the attitude is all that’s necessary for the metaethical project we are pursuing. 15. Once it’s noted that the problem I have in mind is often felt strongly, and indeed sometimes most strongly, from the standpoint of values that are already ethical in nature, it becomes clear that the problem I have in mind is one that (at least when manifested in valuers like us, who already bring something recognizable as ethical concerns to the world) bears at least some similarity to the problem that led Kant to think we that we must, as a practical matter, postulate the existence of God. Very roughly: Kant was right, I think, to note that ethical concerns make us vulnerable to an excruciating and potentially crippling form of discouragement if the empirical facts of the world make the gap between “is” and “ought” seem too large and unlikely ever to close. But Kant was wrong to think that positing the existence of God is the solution; that is just kidding oneself—a resort to the strategy of denying the facts. The solution, rather (or so I ultimately want to argue) lies in the ethical perspective itself— again, cashed out in a certain way. 16. Most glaringly, I give no specific attention in this paper to egoists or amoralists, who are obviously going to present some of the most challenging sorts of cases for the view. This is an issue I plan to address elsewhere. 17. Again, I say more about how I understand “valuing” elsewhere; see note 14 for references. 18. While the point requires more argument than is possible here, nothing about the argument ultimately hinges on my putting things in terms of the normative concepts good and desirable. They’re just intuitive ones to talk in terms of, but ultimately any “thin” normative concept will do. For example, one might also talk about the gap between how so-­and-­so acted, on the one hand, and how so-­and-­so should or ought to have acted, or how there was most reason for so-­and-­so to act, etc., on the other. 19. In my extended sense of “loss,” even the distant past may contain losses in the relevant sense. If one cares about other human beings, for example, then one looks around not only in space, and forward in time, but also backward in space and time, and what one sees there is, in so many locations, horrific. World War II alone, a brief six years in history, faces any sensitive human being with the question of how to be, going forward, in a world where such things are possible. How to feel about such a world, where such things have happened and may well happen again? This is another example (I discuss other examples in Section  5) of how valuing and agency come apart. Our evaluative gaze stretches much farther, in space and in time, than our ability to act.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

46  Sharon Street 20. This section quickly rehashes, with a few terminological changes, some points I make in “Constructivism in Ethics and the Problem of Attachment and Loss,” 181–3. For further discussion, see those pages. 21. This is a moment when it’s very important that I’m simply assuming, in this paper, that normative realism is false. (A realist could think that an invulnerable valuer had normative reasons that he wasn’t recognizing.) 22. Here I’ve focused on finite valuers who are invulnerable in having the world perfectly match their values. However, as Andrea Westlund pointed out to me in helpful comments on this paper, there is another, very different kind of invulnerability that we can imagine befalling a finite valuer: the kind that comes when the worst has already happened, such that the valuer has nothing left to lose. In cases in which it was literally true that a valuer had nothing left to lose (or gain), she would have joined the ranks of those who have no reason to do anything. This most tragic kind of invulnerability would have been achieved due to a perfectly hopeless mismatch between the world and her values. 23. See John Hersey, Hiroshima (New York: Alfred A. Knopf, 1946). 24. See Andrea Elliott and Janon Fisher, “Two Days After  L.I.  Crash, Mother Tells of ‘Brutal’ Loss,” New York Times, July 5, 2005. I discuss this event in “Why the Price of Theism is Normative Skepticism.” 25. See Sonali Deraniyagala, Wave (New York: Alfred A. Knopf, 2013). 26. Jean Améry, At the Mind’s Limits: Contemplations by a Survivor on Auschwitz and Its Realities (Bloomington: Indiana University Press, 1980), 28–9. I am indebted to Susan Neiman for pointing me to this book. 27. To be clear: I think that trust in others can be warranted. Indeed, other finite valuers are the only things we can trust (though not all of them, unfortunately). The point is that trust in the unfolding of causal reality as a whole is not warranted. 28. This is another moment when the falsity of realism is being assumed. 29. The Buddhist tradition is far too vast, rich, and multifarious for there to be a single, uncontested position on any major question—an important point to keep in mind and the reason for the scare quotes around “the.” 30. The word dukkhā encompasses not only the obvious cases of distress and affliction that the English word suffering calls to mind but also subtle experiences of dissatisfaction, lack of ease, stress, and the like. These subtleties of translation are crucial for an adequate understanding of Buddhism, but they don’t affect the basic point of contrast I am drawing in this section. 31. In “Constructivism in Ethics and the Problem of Attachment and Loss,” I discuss Kisa Gotami, a bereaved mother whose story is famous in the Buddhist tradition. She loses her child, goes nearly mad with grief, and receives help from the Buddha. 32. To be clear: I don’t mean to be arguing for this conclusion here. I’m just saying what I think. My goal in this section is merely to point out one distinction between the view I’m proposing and the Buddhist position as it’s often understood. But I hope that many readers will immediately see the appeal of going the way I prefer. 33. Again, the Buddhist tradition is not a monolith, and there are many moments in which Buddhist descriptions of the problems we face do focus on the losses in question, as opposed to the suffering that we experience as a result of those losses. One

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Finite Valuers and Vulnerability to Unmitigated Loss  47 example: “[T]here are four kinds of loss. Because they have undergone these four kinds of loss, some people here shave off their hair and beard, put on the ochre robe, and go forth from the household life into homelessness. What are the four? They are loss through aging, loss through sickness, loss of wealth, and loss of relatives.” See Bhikkhu Bodhi, trans. and ed., In the Buddha’s Words: An Anthology of Discourses from the Pāli Canon (Boston, MA: Wisdom Publications, 2005), 208. Still, there’s an unmistakable tendency in the Buddhist tradition to suggest that the fundamental problem is the experiential state itself, as opposed to the losses, or prospect of losses, that give rise to it. 34. The connection between valuing and certain patterns of “emotional vulnerability” has been noted and explored by others. See, for example, Elizabeth Anderson’s Value in Ethics and Economics (Cambridge, MA: Harvard University Press, 1993), chapter  1; Niko Kolodny’s “Love as Valuing a Relationship,” Philosophical Review 112 (2003): 135–89; and Samuel Scheffler’s “Valuing,” in his Equality & Tradition: Questions of Value in Moral and Political Theory (Oxford: Oxford University Press, 2010), 15–40. 35. The claim is not that any of these secondary aspects will necessarily accompany the primary problem; the claim is rather that there are certain structural features of valuing that put valuers at risk for these secondary aspects, and that, in a sense, that is part of the whole problem. 36. As Korsgaard writes in The Sources of Normativity, 154: “No one, I suppose, would choose not to experience grief at the death of a loved one at all, although we are rightly afraid of finding it unbearable. We may object to a world in which our loved ones are taken away, but if they are taken away, we do not want to fail to experience that fact, to register it as an evil.” 37. On the “two darts,” see In the Buddha’s Words: An Anthology of Discourses from the Pāli Canon, 31–2. 38. See note 15. 39. For this way of putting it, I’m indebted to C. Stephen Evans’s “Moral Arguments for the Existence of God,” in The Stanford Encyclopedia of Philosophy (Fall 2018), ed.  Edward N. Zalta, https://plato.stanford.edu/archives/fall2018/entries/moral-­ arguments-­god/. 40. On this point, see the empirical findings that Dan Moller describes and discusses in “Love and Death,” The Journal of Philosophy 104 (2007): 301–16. 41. Here I have in mind the kind of point that Peter Railton highlights in his “On the Hypothetical and Non-­Hypothetical in Reasoning about Belief and Action,” in Garrett Cullity and Berys Gaut, eds., Ethics and Practical Reason (Oxford: Oxford University Press, 1997). 42. For helpful comments on earlier versions of this paper, I am indebted to Jason D’Cruz, Kyla Ebels-­ Duggan, Christopher Peacocke, Andrea Westlund, and audiences at Oberlin, SUNY Albany, Syracuse, UC Irvine, UNC Chapel Hill, the University of Oslo, and Union College. My greatest debt is to Christine Korsgaard, whose encouragement at critical moments in graduate school helped me learn to trust my own voice, and whose philosophical writings have been a source of inspiration and insight from the day I first encountered them.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

48  Sharon Street

Bibliography Améry, Jean. At the Mind’s Limits: Contemplations by a Survivor on Auschwitz and Its Realities. Bloomington: Indiana University Press, 1980. Bodhi, Bhikkhu, trans. and ed. In the Buddha’s Words: An Anthology of Discourses from the Pāli Canon. Boston, MA: Wisdom Publications, 2005. Bukoski, Michael. “A Critique of Smith’s Constitutivism.” Ethics 127 (2016): 116–46. Deraniyagala, Sonali. Wave. New York: Alfred A. Knopf, 2013. Elliott, Andrea and Fisher, Janon. “Two Days After  L.I.  Crash, Mother Tells of ‘Brutal’ Loss.” New York Times, July 5, 2005. Evans, C.  Stephen. “Moral Arguments for the Existence of God.” The Stanford Encyclopedia of Philosophy (Fall 2018), ed. Edward N. Zalta, https://plato.stanford.edu/archives/fall2018/entries/moral-arguments-god/. Hersey, John. Hiroshima. New York: Alfred A. Knopf, 1946. Korsgaard, Christine  M. “Realism and Constructivism in Twentieth-Century Moral Philosophy.” Journal of Philosophical Research, APA Centennial Supplement 28 (2003): 99–122. Korsgaard, Christine  M. The Sources of Normativity. Cambridge: Cambridge University Press, 1996. Korsgaard, Christine M. Self-Constitution: Agency, Identity, and Integrity. Oxford: Oxford University Press, 2009. Moller, Dan. “Love and Death.” The Journal of Philosophy 104 (2007): 301–16. Railton, Peter. “On the Hypothetical and Non-Hypothetical in Reasoning about Belief and Action.” In Garrett Cullity and Berys Gaut, eds., Ethics and Practical Reason, 53–79. Oxford: Oxford University Press, 1997. Smith, Michael. “A Constitutivist Theory of Reasons: Its Promise and Parts.” Law, Ethics and Philosophy 1 (2013): 9–30. Street, Sharon. “A Darwinian Dilemma for Realist Theories of Value.” Philosophical Studies 127, no. 1 (January 2006): 109–66. Street, Sharon. “Reply to Copp: Naturalism, Normativity, and the Varieties of Realism Worth Worrying About.” Philosophical Issues (a supplement to Noûs), vol. 18 on “Interdisciplinary Core Philosophy,” ed. Walter Sinnott-Armstrong (2008): 207–28. Street, Sharon. “In Defense of Future Tuesday Indifference: Ideally Coherent Eccentrics and the Contingency of What Matters.” Philosophical Issues, vol. 19 on “Metaethics,” ed. Ernest Sosa (2009): 273–98. Street, Sharon. “Evolution and the Normativity of Epistemic Reasons.” Canadian Journal of Philosophy 35, supplemental volume on Belief and Agency, ed. David Hunter (2011): 213–48.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Finite Valuers and Vulnerability to Unmitigated Loss  49 Street, Sharon. “Mind-Independence Without the Mystery: Why Quasi-Realists Can’t Have It Both Ways.” In Russ Shafer-Landau, ed., Oxford Studies in Metaethics, vol. 6, 1–32. Oxford: Clarendon Press, 2011. Street, Sharon. “Coming to Terms with Contingency: Humean Constructivism about Practical Reason.” In James Lenman and Yonatan Shemmer, eds., Constructivism in Practical Philosophy, 40–59. Oxford: Oxford University Press, 2012. Street, Sharon. “If ‘Everything Happens for a Reason,’ then We Don’t Know What Reasons Are: Why the Price of Theism is Normative Skepticism.” In Michael Bergmann and Patrick Kain, eds., Challenges to Religious and Moral Belief: Disagreement and Evolution, 172–92. Oxford: Oxford University Press, 2014. Street, Sharon. “Constructivism in Ethics and the Problem of Attachment and Loss.” Proceedings of the Aristotelian Society, Supplementary Volume 90 (2016): 161–89. Street, Sharon. “Objectivity and Truth: You’d Better Rethink It.” In Russ ShaferLandau, ed., Oxford Studies in Metaethics, vol. 11, 293–333. Oxford: Clarendon Press, 2016. Street, Sharon. “Nothing ‘Really’ Matters, But That’s Not What Matters.” In Peter Singer, ed., Does Anything Really Matter? Parfit on Objectivity, 121–48. Oxford: Oxford University Press, 2017. Walden, Kenneth. “Morality, Agency, and Other People.” Ergo 5 (2018): 69–101. Walden, Kenneth. “Foundations for Epistemic Contractualism.” Unpublished manuscript.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

3 A Question of One’s Own Concepts, Conceptions, and Moral Skepticisms Kyla Ebels-­Duggan

That venerable villain of moral philosophy, the moral skeptic, asks Why should I be moral? Why should I acknowledge the authority of morality’s purported demands? Christine M. Korsgaard uses this skeptical question to frame a conflict between constructivism and realism in moral theory and to argue for the superiority of the former. In The Sources of Normativity, she accuses moral realists of a kind of dogmatism, a refusal to answer—or even to acknowledge—an important skeptical challenge to morality, which she calls “the normative question.”1 Some wonder what, exactly, the normative question is, and whether it’s a real question or merely a pseudo-­question.2 Even supposing that the normative question is well-­formed, meaningful, and intelligible, we might still ask whether it needs an answer or whether a moral theory that does not address it falls short of a standard such a theory ought to meet. Korsgaard contends that falling into the doubt the normative question expresses is a genuine possibility.3 Thus, she presses, a successful moral theory must answer the normative question because it is, or could become, our own question. She further argues that a constructivist account like her own is the only approach capable of fully acknowledging and addressing the question. Korsgaard has characterized her constructivist view as comprising two theses. The first advances an account of what moral concepts, and normative concepts more generally, are: they are placeholders for answers to normatively significant questions that we confront. The second is the bold claim that we can derive the content of these answers—what Korsgaard, following John Rawls, calls normative conceptions—from the concepts themselves.4 To put that another way, we can arrive at the substance of normative answers just by reflection on the form of our normative questions. Korsgaard traces this intriguing idea to Kant’s practical philosophy. I will call this interpretation of the constructivist program Ambitious Constructivism. Here I’ll argue against Kyla Ebels-­Duggan, A Question of One’s Own: Concepts, Conceptions, and Moral Skepticisms In: Normativity and Agency: Themes from the Philosophy of Christine M. Korsgaard. Edited by: Tamar Schapiro, Kyla Ebels-­Duggan, and Sharon Street, Oxford University Press. © Kyla Ebels-­Duggan 2022. DOI: 10.1093/oso/9780198843726.003.0003

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

A Question of One ’ s Own  51 this second thesis, while advocating for both the truth and the significance of the first. Successful execution of Ambitious Constructivism with respect to central moral concepts would, indeed, answer the normative question on the skeptic’s demanding terms. Moreover, it would do so in a way that addresses some of the best-­known challenges to the authority of morality, including the problem standardly called Prichard’s Dilemma, along with long-­standing metaphysical and epistemological puzzles. But deriving a conception of some normative concept from the concept is a difficult matter. The best candidate attempts to do so are found in Kant’s Groundwork and Korsgaard’s own positive moral theory, but on reflection neither achieves the stated aims of Ambitious Constructivism. The good news is that there is an alternative. Modest Constructivism, a view that endorses the constructivist account of moral concepts, but gives up on the idea of deriving conceptions from these concepts, retains most of the advantages of Ambitious Constructivism. It, too, can address Prichard’s Dilemma and parry the metaphysical and epistemological challenges. It cannot answer all intelligible skeptical questions. But I argue that it can answer all those that are ours in Korsgaard’s important sense, and that we should be content with this.5 In the first section, I reconstruct Korsgaard’s critique of Moral Realism. I distinguish different aims that we might have in addressing skeptical challenges to morality, lay out Prichard’s Dilemma, and explain why Korsgaard holds that the Realist fails even to acknowledge, much less answer, the normative question. I also pose the metaphysical and epistemological challenges that she contends that the Realist cannot overcome. In Section 2, I defend the Constructivist view that the normative question is meaningful—that is, not a pseudo-­question—and, in Section 3, present Ambitious Constructivism as an attractive strategy for answering it. I turn to Kant’s defense of the Formula of Universal Law in the Groundwork as an example of how this strategy might be executed, and I show how successful execution of such a strategy would indeed dissolve Prichard’s Dilemma, address metaphysical and epistemological puzzles about morality, and answer the skeptic on his own terms. However, I then argue in Section 4 that neither this, nor later Kantian arguments, realize these constructivist ambitions. Finally, in Section 5, I assess the standing of the challenge voiced by the normative question and the prospects for a more Modest Constructivism. I argue that these prospects are pretty good: though Modest Constructivism cannot answer all intelligible skeptical challenges to morality, it can plausibly address all those that deserve acknowledgment as our own.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

52  Kyla Ebels-Duggan

1.  Addressing the Moral Skeptic When a skeptic presses the question Why should I be moral? he presents himself as doubting that there are compelling reasons to follow moral requirements, and asks us to convince him that there are. There are at least four things that we might expect a moral theory to do in response to this sort of challenge. I list them here in decreasing order of difficulty: Four conceptions of the aims of moral theory: (1) Convince and convert any skeptical challenger. (2) Give an argument to the skeptic that he can reject only on pain of irrationality. (3) Give an argument about the skeptic, or the skeptical position, that explains why his lack of conviction shouldn’t destabilize our own moral commitments. (4) Declare confidence in our moral commitments. We can dismiss (1) as a standard for a successful moral theory out of hand. No argument could convert those who won’t be moved by good reasoning, nor should an argument be criticized for failing to do so. Korsgaard advocates for (2), and indeed for a particularly demanding interpretation of (2). She seeks an argument that lays bare some internal inconsistency in the moral skeptic’s commitments.6 On this approach we need to show, not just that the skeptic is blind to certain reasons or values—blindness is, after all, not a form of irrationality—but that, like believing contradictory claims, maintaining the skeptical position requires affirming attitudes that are in rational conflict with one another. Korsgaard takes Kant’s idea that Reason seeks the unconditioned condition of its commitments to show that we have to meet this demanding standard: so long as you can question the justification of some commitment you hold—so long as a justificatory why question is intelligible—it needs an answer.7 Seeking and finding such answers is, she argues, an internal standard of the activity of Reason.8 Reason will be restless until it rests in the rationally unconditioned, some bedrock commitment that no one can intelligibly question further. To meet standard (2) is to produce a valid argument for the authority of morality that proceeds only from premises that the skeptic has to accept, that is to say from premises that anyone has to accept. Such an argument would address

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

A Question of One ’ s Own  53 every intelligible justificatory question and so meet the Kantian ideal of finding the unconditioned rational condition of moral conviction. Meeting this demanding standard would provide escape from H.A. Prichard’s famous dilemma for moral justification. Prichard holds that nothing could count as an adequate answer to the justificatory challenge, “why be moral?”9 He argues that if we justify the authority of morality by appeal to some non-­ moral—­that is, extra-­moral—­reason we would fail to justify a distinctively moral demand. If we demonstrate that acting morally conduces to our self-­interest, we succeed in explaining away distinctively moral considerations, rather than showing that they should play a special role in our practical thinking. However, it also seems impossible to justify morality with an internal appeal to moral reasons. It’s obviously question-­begging to cite the fact that it would be morally wrong not to follow moral directives in response to the skeptical query. But, the dilemma presses, these are the only two types of answers there could be, so nothing can satisfactorily answer the skeptical challenge. Demonstrating the internal inconsistency of any position that rejects the authority of moral demands seems to offer a third alternative. To do so would be to produce an argument to the conclusion that morality is authoritative, rather than presuppose its authority, escaping the second horn. So long as this argument doesn’t rely in its premises on an assertion of extra-­moral reasons, such as appeal to reasons of self-­interest, it would not fall afoul of the first horn. Of course, we might wonder what such an argument could be, but Ambitious Constructivism purports to have a way of generating one.10 Even so, given how very demanding standard (2) is, it may seem that we should set it aside and take (3) to be the best that we could hope to achieve. We would accomplish (3) by making the case that the skeptic, though he may have an internally consistent view, is unresponsive to reasons and values to which he should be sensitive. Unlike (2), (3) allows appeal to substantive normative or evaluative premises—assertions that certain considerations are reasons, or certain things or relationships are of value, or that certain ideals should be affirmed and pursued. These claims can be intelligibly doubted, but insofar as we are not ourselves skeptical, (3) might be enough. But Korsgaard urges us to regard this possibility as moot. Her more intuitive way of putting the Kantian claims about the characteristic activity of Reason is that, even if we are not currently skeptics, the normative question that the skeptic presses is, or could become, our own question—it could become your own question. She directs you to consider a scenario in which you are gripped by the doubts that this question expresses. She sketches some

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

54  Kyla Ebels-Duggan brief examples, scenes meant to illustrate that even a morally decent person can find the moral demand difficult and may reasonably wonder if it is really so categorical as it purports to be: And in ethics, the question can become urgent, for the day will come, for most of us, when what morality commands, obliges, or recommends is hard: that we share decisions with people whose intelligence or integrity don’t inspire our confidence; that we assume grave responsibilities to which we feel inadequate; that we sacrifice our lives, or voluntarily relinquish what makes them sweet.11

You may find yourself in such circumstances, she warns, and if so then you may need an answer to the normative question. She then imagines an advocate of a moral theory that meets only standard (3), a figure she calls the Substantive Realist, trying to address your doubt. The Substantive Realist is willing to rely on the sorts of contentful moral claims just mentioned.12 Korsgaard represents these assertions as the Realist simply insisting that you must do what morality requires because the purported obligations in question really are obligations.13 She rightly observes that this would not just fail to provide an adequate answer to the normative question you are asking, but would not even so much as acknowledge the question. Asserting the reality of the obligations under discussion is no help at all, certainly no help to the skeptic. Insofar as you are the skeptic, gripped by the doubts that the normative question expresses, it’s no help to you. In her rejection of strategy (3), Korsgaard reframes the challenge voiced in the normative question. She turns our attention from thinking about “the skeptic,” as an interlocutor who might confront and challenge us, to thinking about—or, better, thinking as—ourselves. She has us consider what questions we have, or might have, and what answers we could give to them. If this is the right way to approach matters, then perhaps a moral theory that merits our endorsement needs to have a (2)-type answer at the ready. To put this another way, perhaps any attempt to accomplish (3) without meeting the standard set by (2) will inevitably collapse into the table-­pounding response of (4) if and when we confront it with our own doubts. Saying you must do something because you really have an obligation has no more content than reaffirming, perhaps more stridently or in a louder voice, that you really must do it. That response seems like an embrace of the second of the two horns of Prichard’s Dilemma. If this is all that the Realist has to say then it seems fair to describe him as merely table-­pounding.14

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

A Question of One ’ s Own  55 In fact, Korsgaard thinks that matters are even worse for the Substantive Realist. She interprets the assertion of real obligations, not as completely empty, but rather as presuming and relying upon a certain account of normative concepts and the truth-­makers of normative claims. On this view, the function of normative concepts is to refer to and describe what she characterizes on the Realist’s behalf as “the normative part of reality.” What makes a normative claim correct on the Realist view is that the description in question accurately represents this part.15 Answering normative questions is then just like—indeed it is a subset of—answering descriptive questions, the kinds of questions that we ask in the empirical sciences. The only difference is that the properties or entities that concern us in the normative case are non-­material ones that don’t tell us what is, but what ought to be and what we ought to do. Once we start thinking of moral concepts in this way, metaphysical and epistemological puzzles arise. In order to vindicate the authority of morality, it appears that we now need to posit the existence of presumably non-­material, normatively laden entities or properties, and to suppose that human beings have some faculty by which we could detect these mysterious properties. Korsgaard plausibly contends that this amounts to descent into nonsense: we do not even understand what we could be talking about here, what it would be to be such a thing or have such a property.16 Insofar as it generates these sorts of worries, Realism, far from answering a skeptical challenger, appears to provide him further fuel for his skeptical flames. Korsgaard argues that these challenges are insurmountable. Moreover, she contends, even if we were willing to reconcile ourselves to living with these metaphysical and epistemological mysteries, the Realist view would still be unsatisfying. For it seems that we could, even then, ask why we should care about these mysterious entities, or on what basis we should accept that they have authority to direct our actions. But this is just to raise the normative question all over again.17 That this question can re-­arise shows that, construed as a part of reality that would figure in a complete description of the world, reasons and obligations could not do the normative work for which we need them. There is thus no point in positing them. We do not know what we are talking about when we try to refer to metaphysical entities of this kind, and would not be talking about the right sort of thing even if we did. So, Korsgaard argues, even if the Substantive Realist is willing to bear the metaphysical and epistemic costs of his position, this buys him nothing. So far, we have seen a range of challenges that a satisfying moral theory must seemingly meet. It must provide some response to the skeptic who questions the authority of moral claims. This answer must meet the test that

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

56  Kyla Ebels-Duggan Prichard’s Dilemma sets and it must avoid the metaphysical and epistemological pitfalls into which Korsgaard thinks that the Realist falls. Ambitious Constructivism commits to meeting standard (2), which would give us a way to meet all of these demanding requirements. Substantive Realism threatens to be limited to (4), and to meet none of them. In what follows I show how Modest Constructivism might land in the middle space designated by (3).

2.  Interpreting the Normative Question In his first dramatic scene, Korsgaard’s normative questioner confronts a purported moral obligation with two questions: must I do really this? And why must I do it?18 The first of these suggests doubt that what is asked of him is something that morality genuinely demands—a first-­order question about the content of moral obligations. Some of Korsgaard’s examples also lend themselves to this interpretation: we aren’t always required to share decisions with people whose characters don’t inspire confidence, or assume grave responsibilities to which we feel inadequate, or give up what makes our lives sweet. So in any particular situation we could meaningfully ask whether we’re required to do one of these things, there and then.19 But Korsgaard clearly says that this is not the question, or the sort of question, that she’s after. Her skeptic is not trying to have a discussion that is internal to morality, but to ask about the standing of morality as whole. Admittedly, this makes it a little hard to keep in view what, exactly, he is asking. It’s supposed to be a completely general question about the authority of moral demands, about why we should think that any requirement carries the kind of authority that moral demands purport to have. In Kant’s terms, it is the question of how a categorical imperative is possible.20 T.M. Scanlon denies that there is any meaningful question concerning why we must recognize the authority of moral reasons, any query that comes apart from first-­order questions about which reasons we have. He suggests that talk of the “authority of morality” misleads us into supposing that there is a question here, and that we should be careful to treat this talk as merely metaphorical.21 There does seem to be an important difference between the moral case and other cases in which the concept of authority is clearly in play. In the case of a purported divine authority or the authority of a government we have a clear grasp on what fixes requirements of a certain content. We can say that these are the laws of government in question, or those are the commands of the deity. Without reference to the particular content of these requirements,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

A Question of One ’ s Own  57 we can ask whether we have good reason to regard their source, the agent who makes the demands, as genuinely authoritative for us. But morality is different. It is not a source in the way that a god or a legislative institution might be. Morality is not a person, or other more complex agent, that issues commands. Rather, we might think that to accept that something is morally required of us just is to accept that it is an authoritative requirement, something we must do no matter how we feel about it or what our aims are, the sort of practical requirement that Kant would call cat­e­go­ri­ cal. If this were right, then we could still always intelligibly consider whether there is an authoritative demand to act in some particular way, say to always keep our promises. But if we grant that there is such a moral requirement, as Korsgaard says her skeptic does, we would have thereby settled the question about its authority and there would be no further normative question to ask.22 This argument looks compelling, and Korsgaard herself acknowledges that the normative question tends to “slip through our fingers.”23 But in fact there is a real question here distinct from the first-­order question about which moral obligations are genuine, and closer consideration of the constructivist strategy can help us get a grasp, or perhaps keep a grip, on what that question is. The constructivist account of normative concepts and the concept/conception distinction allows us to pin the normative question down securely. And, once we get the challenge in view, we can see the distinctive appeal of constructivist approaches to answering it. Recall that the constructivist regards normative concepts as placeholders for the answer to normatively significant questions.24 This is the first of the two theses of Ambitious Constructivism. Terms that function this way include reason, value, and justice. In answering the moral skeptic, Korsgaard focuses on the term obligation. Obligation is the place-­holder answer to the question: What, if anything, must I do, no matter what else I aim at or want? Having specified the question associated with a particular normative term we can start trying to formulate and defend an answer, a conception of the normative concept. For example, Kant’s conception of obligation is act only in accordance with that maxim through which you can at the same time will that it become a universal law.25 Scanlon also offers a conception of moral obligation, roughly, avoid acting in ways that would be permitted only by principles for the general regulation of behavior that someone also committed to identifying such principles could reasonably reject.26 While the concept is a place-­ holder for the answer to a normatively significant question, whatever this answer may be, a conception is a contentful, informative account of what the answer is.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

58  Kyla Ebels-Duggan With this background, we can understand Korsgaard’s skeptic as asking for a defense of his interlocutor’s conception of moral obligation. To answer the normative question about obligation is to defend a certain conception of obligation, to argue that that conception provides the true or best answer to the question about what we must do. Korsgaard confirms this interpretation of the normative question in saying, “Even those who are convinced that ‘it is right’ must be in itself a sufficient reason for action may request an account of rightness which this conviction will survive.”27 To put this another way, we are not asking, given some conception of obligation, whether a certain action falls under it. That would be to ask a first-­order normative question. Rather, we ask why we should accept that actions that fall under a given conception are such that we must do them.28 This is the question that Korsgaard accuses the Substantive Realist of failing to acknowledge. A view that simply asserts certain first-­order normative claims would be guilty of this charge, amounting to a mere list of purported obligations. The skeptic pressing the normative question is not asking about whether the particular action demanded of him appears on any such list. Rather, he wants a persuasive answer about what makes the particular list on offer one to which he should look to regulate his actions. To answer that question, one first needs some account of what it is that lands an action on that list, the standard or rule that the list is tracking, or the procedure that we are supposed to use to identify which actions belong on it. Then one needs to make the case that that standard or procedure is the best one for the purposes. Korsgaard has the Realist claim that the set of moral facts or truths, or real obligations, corresponds to and supports the lists. That is to say, the rule for formulating the list of obligations is: include all and only the things that really are obligations. But that standard is no help; it is just an obfuscating way of reiterating the list itself. So, Korsgaard’s Realist is not even forwarding, much less defending, a conception of obligation.29 That is why he cannot acknowledge the normative question. Nevertheless, the question is a genuine one.

3.  The Ambitious Constructivist Answer The second Ambitious Constructivist thesis is a claim about how to provide and defend an answer to the normative question, a bold position about how to demonstrate that a candidate conception of obligation is superior to competitors: we can determine the answer to a normative question just by

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

A Question of One ’ s Own  59 reflecting on the content of the question itself. Put another way, we can derive normative conceptions from their corresponding concepts. Korsgaard presents this as an interpretation of the Kantian idea that we can generate substance from form, the substance of a normative conception from the form of the concept, or the substance of the answer to a normative question from the form of the question itself. In Groundwork II, Kant appears to advance a thesis of this kind. He says that he will derive the content of the Categorial Imperative from the concept of a categorical imperative, or anyway that he will “. . . inquire whether the mere concept of a categorical imperative may not also provide its formula . . .,”30 an inquiry that he presents as yielding a positive resolution. Kant says that, whereas a hypothetical imperative requires us to do something on the condition that we want or aim to accomplish something else, a categorical imperative would require us to act in certain ways no matter what else we want or what ends we have. That is to say, its requirement has no condition, but is unconditional. This, he claims, is the very idea or concept of a cat­e­go­ri­ cal imperative. Reasoning just from this idea or concept, Kant arrives at a conclusion about what the Categorical Imperative is, or—as he puts it—what it contains, a conception of the Categorical Imperative, that is a conception of obligation.31 Such a principle commands unconditionally, as if it gives us a law. We have, as yet, no information about the content of the law in question. All that we can take from the concept of a categorical imperative is that it tells us to act on a law, that is, to act on a maxim that is fit to be a law.32 Kant argues that this is enough to yield the Universal Law Formula of the Categorical Imperative: act only in accordance with that maxim through which you can at the same time will that it become a universal law.33 In the examples that follow, he gives evidence that this principle has content and claims that violating it would involve you in a kind of contradiction.34 Here, it certainly appears that Kant purports to meet standard (2). He seems to be arguing that anyone must accept his conception of obligation on pain of affirming a contradiction, the paradigm case of narrow or formal irrationality. Some commentators have tried to make this out as a logical contradiction, in which the agent holds attitudes the conjunction of which violates laws of logic. But Korsgaard argues that the contradiction Kant has in mind here is, instead, distinctively practical.35 On her reading, the cases work like this: one formulates the maxim on which one proposes to act. The maxim states both the act one proposes to undertake and the aim with which one

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

60  Kyla Ebels-Duggan undertakes it. To see if the maxim can be willed as a law without contradiction, one must imagine a situation in which everyone with an aim like one’s own undertakes an action like one’s own in order to accomplish it. Then one asks: in this situation, the World of the Universalized Maxim, will my proposed action work? Will it effectively accomplish my end? If the answer is no, the contradiction is laid bare and the maxim is thereby revealed to be unfit to be a law. As is familiar to many, the lying promise provides the most straightforward illustration. The answer to Kant’s question about the effectiveness of this undertaking in the World of the Universalized Maxim is negative: if everyone tries to use lying promises to secure a loan, then everyone will laugh at such promises as vain pretense, and it will be impossible to use them to secure loans.36 The “contradiction” as Korsgaard understands it is between willing the maxim and willing the World of the Universalized Maxim, the situation in which anyone with your aim wills the maxim as a way of accomplishing it. The contradiction is practical, rather than theoretical, because the problem is not that one has contradictory beliefs, but rather that one wills contradictory aims: the maxim on the one hand, and its universalization on the other. Or perhaps better, in willing the maxim, one wills the aim that it specifies, but in willing the World of the Universalized Maxim, one wills to undermine this aim. Korsgaard looks to this part of Kant’s practical philosophy to provide the model of how an Ambitious Constructivist strategy would work, that is, how we might derive a contentful normative conception from the mere form of a normative concept. Suppose that we can indeed derive a conception of obligation from nothing other than the concept of obligation. The advantages would be impressive. First, this sort of defense of a normative conception would dissolve the threat posed by Prichard’s Dilemma. The derivation of conception from concept does not appeal to extra-­moral reasons. But neither is showing that some substantial answer follows from a complete understanding of the question it addresses mere table-­pounding. The derivation would show why what makes an action wrong, and so makes it fall under the conception, also gives us requiring reason not to do it, that is, meets the concept. And showing that whatever it is that makes acting in a certain way wrong also gives us requiring reason not to do it seems like the only way to address Prichard’s Dilemma. A successful Ambitious Constructivism would also provide a satisfying answer to metaphysical challenges. It furnishes an account of what it would be for an answer to a normative question to be true or correct that does not require positing any mysterious entities or faculties. Such an answer is correct

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

A Question of One ’ s Own  61 just in case, and just because, it can be derived from the question. Ambitious Constructivism thus yields a procedural account of normative truth: to be a true answer to a normative question is just to be the conclusion of good reasoning about how to answer the question. Assuming the success of the Ambitious Constructivist program, such reasoning delivers determinate answers, and these are the normative truths. By the same token, the constructivist can address the epistemological questions that often follow in the wake of the metaphysical ones. Even if we grant that there are answers to normative questions, normative truths or facts, it may seem puzzling how we could ever know anything about them.37 But, given a successful execution of Ambitious Constructivism, these answers are available to us—and our confidence in them is warranted—just because we understand the question and we have ordinary powers of reflection and reasoning. No additional faculty capable of tracking normative truths is required. The constructivist strategy successfully engages the skeptic and aims to answer him on his own terms. This is the power of Korsgaard’s simple observation that if you recognize a normative question as one that you face, and an answer to this question as the best one available, you thereby recognize and accept the authority of that answer.38 If you fail to guide yourself by this answer, you have attitudes that are in conflict with one another. Korsgaard’s skeptic is pressing the normative question, so it is clear that he already recognizes it as his own. Ambitious Constructivism would show that any such skeptic must accept the answer that it offers on pain of irrationality, meeting standard (2). Any skeptic should then see, by his own lights, that this answer is one he must accept.

4.  In Search of an Ambitious Constructivist All of this is very promising. But, on reflection, we can see that the Kantian argument above does not accomplish the Ambitious Constructivist aim, nor does Kant really present it as doing so. Though the argument accomplishes something interesting, it doesn’t answer the normative question. Suppose that Korsgaard’s procedure does turn up practical contradictions, not just in the lying promise case, but in a wider range of cases that plausibly track the content of our moral obligations. This would give us a method for sorting permissible from impermissible maxims of action.39 The Universal Law Formula, interpreted with the practical contradiction test, could tell us that it is wrong to act on certain maxims. This would already be significant,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

62  Kyla Ebels-Duggan but by itself would leave the Kantian in a position very much like that of the Realist who merely asserts some collection of first-­order normative claims. We can confront the list of purportedly impermissible maxims generated by the Practical Contradiction Test with the very questions that Korsgaard imagines the skeptic asking the Realist: why should we care about this particular list and avoid acting on the maxims that this test sorts as impermissible? Why accept the claim that these are the very things that we must not do? The Practical Contradiction Test goes at least one step beyond mere sorting: a mechanism or method the functioning of which was totally opaque could sort maxims. A guru might issue declarations concerning which maxims it would be impermissible to act on, without transparency about how she was making these determinations. Fairly or not, this is how Korsgaard represents the conclusions of her Realist interlocutor. By contrast, her Kantian Universal Law test transparently displays the feature in virtue of which it sorts between impermissible and permissible maxims. It thus provides, not just a list of obligations, but a conception of obligation: an impermissible maxim fails to have the form of law, where this means that one could not effectively accomplish the end described in the maxim by acting as the maxim directs in the World of the Universalized Maxim. Let us mark this by saying that the Practical Contradiction Test tells us not only that the maxim is permissible or impermissible but how—that is, in virtue of what—it is so. The Practical Contradiction Test thus defines a conception of obligation, but the normative question asks for not only a candidate conception of obligation but a defense of that conception. And this question persists in the face of the argument we have considered so far, because the practical contradiction that the Kantian test reveals is not necessarily a contradiction in the agent’s own attitudes. It is not like believing both p and ~p. This sets violations of the Categorical Imperative in contrast with violations of hypothetical imperatives, on Kant’s account. He argues that someone who violates a hypothetical imperative both wills to pursue an end and wills not to pursue it. As in the case of contradictory beliefs, if someone acknowledges that she is violating a hypothetical imperative, she can require no further account of why she must revise the set of attitudes that she holds. By contrast, the contradiction that the Practical Contradiction Test reveals appears not in what an agent actually wills, but rather between what she actually wills (her maxim) and the universalization of this maxim. An agent who is not willing the universalization is not involved in any inconsistency. But then it is open to her to ask why she should will the World of the Universalized Maxim, or why she should care about the fact that what she proposes to do here and now would not work

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

A Question of One ’ s Own  63 under those circumstances. Since avoidance of this sort of contradiction is the conception of obligation on offer, this just is a way of asking for a defense or justification of that conception. It is a way of pressing Korsgaard’s normative question against Kant’s Universal Law Formula. Kant himself would not be troubled by this, as he does not present the argument discussed above as answering Korsgaard’s normative question or as giving a defense of the conception of obligation on offer. While he claims that this argument can show that the Universal Law formulation is the only possible candidate for a categorical imperative—that is, the only candidate conception of obligation—he does not claim to vindicate the conception in this stretch of the argument. At the end of Groundwork II, he says that, for all that he has established so far, morality might yet be a chimera.40 That is, while he has forwarded a certain conception of moral requirement, he has yet to defend the claim that we should accept this—or any—conception, and so acknowledge its normative authority for us. For all he has said so far, the concept of obligation could lack application: it may be that there is nothing that we are required to do. It is only in Groundwork III that Kant turns to addressing the further challenge, so it is here, if anywhere, that he addresses Korsgaard’s normative question. She interprets the argument there as another application of the Ambitious Constructivist strategy. This time, Kant works from the concept of freedom, which he treats as a placeholder for the answer to the fundamental practical question, what to do?41 He argues both that we cannot escape this question, and that we can derive the Categorical Imperative, in its Universal Law formulation, from this concept as well.42 That is, we are faced with the task of acting freely or autonomously, and acting on the Categorical Imperative is the best way to carry out that task. He concludes that the Categorical Imperative has inescapable authority for us. Kant opens the Groundwork III argument by claiming that to act freely one must act in a principled way; random or arbitrary action is not freedom. From here he proceeds along a path familiar from the derivation of the Formula of Universal Law from the concept of obligation: to act freely, you must give yourself some law. But it seems that acknowledging any particular law would limit your freedom, and you would have no reason to accept such limits. Since you need a law, but cannot justify one of any particular content, you adopt a principle that merely directs you to act in a principled way: act on a maxim that can be willed as a law. But that just is the Universal Law Formula. We can apparently conclude that the best conception of freedom and the best conception of obligation are one and the same. That is to say, the answer to the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

64  Kyla Ebels-Duggan question of what to do is the same as the answer to the question what obligations do I have? You have most reason to do what you are morally obligated to do. But this argument succeeds only if the meaning of a maxim being willable as a law is the same here as in the argument above; otherwise, the identification between freedom and obligation will turn on an equivocation. The Practical Contradiction Test provided an interpretation of what it is for something to be willable as a law in the argument from the concept of obligation. But underlying this interpretation is a distinctively moral ideal, which Kant characterizes as a requirement not to “make an exception for ourselves”: no one should act in a way that everyone isn’t permitted to act.43 Appeal to a moral ideal is fine insofar as we are working out a conception of moral obligation, but would be illicit in the interpretation of the conclusion of the freedom argument. So when we ask what the Groundwork III argument means by saying that a maxim can be willed as law, we cannot just import the Practical Contradiction Interpretation. Korsgaard endorses this assessment when she says that the principle derived from the freedom argument does not yield determinate guidance because it leaves unspecified the domain over which universalization ranges. So, far from settling this domain, this argument permits it to range even just over the desires of the moment, yielding what she calls “the law of the wanton,” a principle of acting on my current strongest desire. But a principle that doesn’t rule that out surely isn’t ruling out much of anything.44 Only by specifying the domain of universalization as the Kingdom of Ends, or the community of all rational beings, can we take the final step along the path from the concept of freedom to a conception with substantive, moral content. But this specification is in no way entailed by the freedom argument or the practical question to which it responds. It does not follow just from the concept of freedom. The specification of this domain may indeed capture an appealing moral ideal, but, just as the skeptic can ask why he should will the World of the Universalized Maxim, he can ask why he should care about universalizing over that particular domain. So the conception of obligation remains undefended and the normative question remains without answer. We might wonder whether the Formula of Humanity can help. The Formula of Universal Law is the statement of the Categorical Imperative that Kant derives from the concept of a categorical imperative and the formula he purports to vindicate in Groundwork III, but the Formula of Humanity is widely regarded as a more promising statement of Kant’s fundamental moral principle. Kant’s argument that we must treat humanity as an end appears to be an

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

A Question of One ’ s Own  65 argument from elimination with a conditional conclusion: if there is an objective end, it must be humanity. Korsgaard develops a more promising approach, presenting the argument as a regress on the rational conditions of valuing anything at all.45 According to this argument, any given thing can be valuable only if humanity is valuable, so if something is valuable, then humanity must be valuable. Thus if you regard something as valuable, but deny the value of humanity, you hold contradictory attitudes just like someone who believes p and ~p. This argument has its own promise and problems, but in any case represents a shift away from the Ambitious Constructivist strategy defined by the two constructivist theses with which we began.46 Rather than starting with a bare normative concept, treated as a placeholder answer to some question, it begins with some substantive evaluative commitment that the skeptical questioner endorses: that x is valuable, for some particular x, and then works out the rational entailments of this commitment. Even if successful, this argumentative strategy leaves a route of escape from moral demand, albeit at the high cost of granting that there is nothing that is valuable, a position that Korsgaard calls complete practical normative skepticism.47 While undeniably a significant conclusion, this argument does not purport to be one that anyone would have to accept, or an answer to all possible skeptical challenges.48 The Universal Law Formula is plausibly derivable from the mere concept of obligation, but this argument provides only an interpretation of our obligations; it does not defend their authority. The argument from the concept of freedom may establish the inescapable authority of the conception that it yields. But—while it takes the same form as the Universal Law Formula—we lack grounds for thinking that the content of this conception is the same as that of the moral law. The regress argument for the Formula of Humanity holds promise, but not the promise that we can derive a conception of moral requirement from any concept. So we have yet to find a successful execution of Ambitious Constructivism.

5.  Prospects for a Modest Constructivism Above I argued that a successful execution of the Ambitious Constructivist program would have a range of significant advantages, culminating with the claim that it could answer the normative skeptic on his own terms by showing that maintaining his skeptical position would require affirming attitudes that are rationally incompatible with one another. But in what followed, I’ve

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

66  Kyla Ebels-Duggan argued that neither Kant’s own work, nor Korsgaard’s development of the Kantian view, fully executes this bold agenda, suggesting dim prospects for Ambitious Constructivism. In this last section I will argue that we can still secure most of the advantages of the constructivist approach, even if we abandon the idea of deriving authoritative conceptions from normative concepts, so long as we maintain the first of the two constructivist theses. Call a position that affirms this account of the nature of normative concepts, while jettisoning the project of deriving authoritative conceptions from them, Modest Constructivism. Modest Constructivism does not purport to demonstrate the internal inconsistency of moral skepticism as imagined in standard (2). But it need not collapse into the table-­pounding of (4). It can meet standard (3) and address Prichard’s Dilemma along with the metaphysical and epistemological challenges canvassed above. It cannot answer every intelligible skeptical question. But I will argue that it can answer those that deserve recognition as our own. In Section  1, I remarked on a shift in viewpoint in Korsgaard’s account from that of a skeptical interlocutor back to our own point of view, with the suggestion that the skeptic’s question could be one that we raise ourselves. I now want to return to this idea and consider whether this change in view may be more significant than Korsgaard acknowledges. Moving from the Kantian notion of seeking the unconditioned condition of our commitments to the idea that we might actually find ourselves in the grip of a certain question or doubt changes the task of a moral theory, and so what would count as success. So long as we can intelligibly question a commitment, we’ve not yet found the unconditioned condition. But not every intelligible question is, or could be, your own. A question that could be your own would be one that you don’t just understand, but are, or can at least imagine being, concerned with or troubled by. Moreover, and especially where the question is of moral import, it needs to be one that you can imagine seriously entertaining without becoming—to borrow a phrase from Korsgaard—a person you could no longer recognize as yourself.49 For, failing this, the person that you imagine asking the question would be one with whom you could not identify, and so it would be wrong to think that the possibility that you are imagining makes the question potentially yours. Recall that Korsgaard offers some examples of situations in which she claims you might find yourself in the grip of the normative question: you have to share decisions with those whose judgment you do not trust, or assume grave responsibilities, or give up your life. In these situations, she tells us, “the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

A Question of One ’ s Own  67 question—why?—will press, and rightly so. Why should I be moral?”50 What these examples have in common, and what allows them to do the work that they do, is that a person of good will can imagine being gripped by doubt in these situations without thereby abandoning her good-­willed stance or most fundamental values. The person here cast in the role of asking the normative question is thus one with whom we can sympathize; her need for an answer in the circumstances does not make us think ill of, or feel alienated from, her. Contrast the person who would like to sidestep moral demands simply for his own convenience or aggrandizement or just because he does not care about anyone but himself. Think of Hume’s knave, or his suggestion that it is not irrational to prefer the destruction of the whole world to a scratch on your little finger. Let’s grant that Hume is right that such a preference need not involve any inconsistency or contradiction. Even so, I submit that we cannot really imagine sharing it, and if we encountered someone who claimed to have it we would first regard him as insincere and, failing that, as utterly foreign to our own sensibilities. Why should I not prefer the destruction of the world to a scratch on my finger? is not a meaningless question. The words are not nonsense. But neither is it a question that we can take seriously as one that could truly be or become our own. I suspect that Korsgaard’s cases work as well as they do because when we imagine—or perhaps even remember—being in the kinds of situations she describes, we find plausible doubt about the first-­order question, the question of whether morality really demands the relevant action. We can imagine wondering whether we really must share power, take on responsibility, or sacrifice our lives. But we’ve already seen that such first-­order doubt is a distraction from what is supposed to be at issue: is the conception of obligation on offer defensible? If not, should some revised conception replace it? Or should we rather conclude that the concept of obligation is—in the end—an empty concept, that there is nothing at all that we have to do? But a decent person who can imagine entertaining the first-­order doubt may not sympathize with someone asking this second sort of question. In particular, she may resist taking the nihilist settlement of the question seriously, as Korsgaard’s skeptic sometimes does. That skeptic is much more like the one who seeks the destruction of the world to save his finger than Korsgaard’s characters at first appear to be. Return now to Kant’s argument. The conception of wrongness that Korsgaard attributes to Kant, the feature of an action that the practical contradiction test identifies, is that what you are doing would not succeed if everyone did it. It is

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

68  Kyla Ebels-Duggan prima facie plausible that this is a morally important feature of an action, one that might pick it out as impermissible. To think that it is fine to accomplish your ends in a way that would not work for everyone does seem to require making an exception for yourself, allowing yourself to do what others then cannot do. This, in turn, appears to rely on the idea that you have some special standing or authority that others lack, and is thus incompatible with an attractive ideal of moral equality. Rather than purporting to derive this conception from the very idea of impermissibility, the Modest Constructivist treats the claim that this is the best conception of the impermissible as a substantive normative assertion, an exercise of judgment about how best to understand the category of what we must not do. She recognizes that someone might, without inconsistency, ­render an incompatible judgment. But this need not undercut her own judgment, nor draw her to thinking that there are no standards for these judgments at all. A person of good will could genuinely wonder whether practical contradictions provide the best conception of the impermissible. Barbara Herman expresses and supports such doubts with purported counter-­ examples to Korsgaard’s reading of the Universal Law Formula. She offers the maxim: I will play tennis early on Sunday mornings, in order to avoid crowds on the court.51 Like the lying promiser, the person who acts on this maxim cannot accomplish his aim if everyone acts on it. But, unlike the lying promise, action on this maxim seems straightforwardly permissible. This case suggests that practical contradiction does not, in fact, capture the underlying moral ideal of equality we find attractive. Herman proposes that we can better understand the test for permissibility this way: consider whether, if it were a matter a public knowledge that you were acting on your maxim, you would still be able to accomplish your end. The tennis maxim appears to pass this test, while the lying promise fails, counting in favor of the revision. The point for our purposes concerns how we evaluate a challenge and proposal like Herman’s. We cannot choose between their conceptions of the impermissible by deriving one from their shared concept. Instead, we have to exercise judgment about which better captures the underlying moral ideal.52 Kant himself is clear that the application of any rule or principle requires an exercise of judgment, so we should not expect that the Formula of Universal Law will self-­interpret. We have standards on which we can rely in making these judgments and comparing available alternative conceptions. We can consider which seems to sort cases correctly, as we did above. But, more importantly, we can think about whether and why we have reason to care about the feature in virtue of which they are sorting, whether it responds to some recognizable value.53 To accept Herman’s proposal that we understand

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

A Question of One ’ s Own  69 permissibility in terms of publicity is to judge that it captures a moral ideal of relationships that we can recognize and endorse, and that it does so better than any of the available alternatives. To acknowledge this ineliminable role for substantive moral judgment in determining requirements on action is to abandon the second constructivist thesis. Yet we must still have some answers to certain normative questions, or at the very least treat some settlements as answers in practice. We must decide how to act, how to regard and treat one another, and what sorts of interpersonal relationships we should seek to realize. We can exercise judgment about these matters in a way that is responsive to the standards suggested above, without encountering any epistemological or metaphysical mysteries. We need not wait on some meta-­normative account in order to settle on answers, any more than scientists must wait on a parallel such account of empirical truth before making judgments about what conclusions their data supports. Nor, importantly, does the lack of such an account create a presumption in favor of any particular normative conceptions—including the nihilist conceptions that the skeptic may entertain. The normative nihilist claims that there is no reason to act one way rather than another, no values that guide our treatment of one another, and no ideal for our relationships worth taking seriously. We should assess those claims, like any others, on their merits, in accordance with the sorts of standards listed above. I do not myself find them very plausible. In addition to side-­stepping metaethical challenges, Modest Constructivism addresses Prichard’s Dilemma: defending a conception of obligation amounts to making the case that the features of action on which it focuses are worth caring about and that it makes sense to treat them as playing the distinctive role that we assign to moral reasons. Doing so is not an appeal to extra-­moral reasons to show that we should follow moral demands, nor is it just a louder reassertion of these demands. In fact, Korsgaard’s own view addresses Prichard in just this way. She advances ideals of the virtuous agent as unified and autonomous and as standing to other agents in the appealing relationship of respect and cooperation that she characterizes as being fellow members of the Kingdom of Ends.54 Similarly, in the Groundwork III argument, Kant relies on an attractive ideal of freedom as principled action, an ideal that Korsgaard also affirms.55 Throughout her work, Korsgaard makes the case for these moral ideals, sometimes by arguing that we have reason to affirm or care about them, but sometimes just by presenting them in a compelling light that makes their attraction salient. She begins with a concept of obligation: that which we must do, no matter what else we want, or what other aims we have. She offers a conception, or set of conceptions: that which makes us free, or unifies our agency, or could be a law in the Kingdom of Ends. And she

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

70  Kyla Ebels-Duggan defends these conceptions by appeal to the value of the relationships, with others and with ourselves, that they allow us to realize. These are ways of forwarding and defending normative conceptions without purporting to derive them from concepts, and they are among the most powerful passages in Korsgaard’s work.56 Modest Constructivism retains Korsgaard’s insight that insofar as we face a normative question, and regard some answer to this question as the true or best one, we thereby take that answer to have authority for us. But it lingers longer than Ambitious Constructivism does over whether a particular question is indeed one that we face or might face. If we understand the questions that grip the protagonists of Korsgaard’s examples as directed at a particular candidate conception of obligation, then we can recognize their questions as ones that we ourselves might raise. But we do not, thereby, introduce for real consideration the skeptical idea that the nihilist may be correct, and there is nothing that we are required to do. Our recognition of the sorts of doubts Korsgaard describes in her cases does not amount to a recognition of this thought as our own, or even potentially so. Nor do we need to rule out the nihilist conception as impossible; seeing that it would be coherent to take that position is not enough to show that we can identify with a person who takes it seriously. Genuinely doubting that others have the same kind of standing or moral authority that you have is already a way of going wrong morally.57 If the normative question is understood as expressing this doubt, or some attitude that entails it, then it is a question that the person of good will cannot seriously entertain. The formulation, interpretation, and application of any such conception requires the exercise of judgment, and to recognize this is to allow that someone could judge differently without internal inconsistency. If that’s right, then Korsgaard’s normative question can always be asked intelligibly— it can be asked of any conception—and in this sense can never be answered definitively. What we lose in abandoning Ambitious Constructivism is the purported ability to address every intelligible skeptical question. But it’s ok to let this go. Modest Constructivism’s avoidance of the metaphysical and epistemological puzzles and its response to Prichard show that in abandoning the second of the possible aims of moral theory, we need not settle for the fourth. We can find solid ground to stand on in the third. While it is always possible to keep asking justificatory questions, not all possible questions are our own questions. It is also possible to be satisfied with the answers that we have.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

A Question of One ’ s Own  71

Notes 1. Christine  M.  Korsgaard, The Sources of Normativity (Cambridge: Cambridge University Press, 1996). 2. G.A. Cohen raises this worry in his original response to the argument of The Sources of Normativity. “Reason, Humanity, and the Moral Law,” in Korsgaard, The Sources of Normativity, 167–88. Cf., T.M.  Scanlon, Being Realistic About Reasons (New York: Oxford University Press, 2014), ch. 1, 9–15 and William FitzPatrick, “The Practical Turn in Ethical Theory: Korsgaard's Constructivism, Realism, and the Nature of Normativity,” Ethics 115 (4): 651–91. 3. Korsgaard, The Sources of Normativity, 7–10. 4. Korsgaard advances the point most clearly in this form in “Realism and Constructivism in Twentieth-­ Century Moral Philosophy,” in The Constitution of Agency: Agency, Identity, and Integrity (New York, NY: Oxford University Press, 2009). Cf., Korsgaard, The Sources of Normativity, 113–19 and John Rawls, A Theory of Justice (Cambridge, MA: Harvard University Press, 1971), section 2, 7–11. 5. A common way of carving up the terrain of constructivist views contrasts restricted or local constructivism, on the one hand, with unrestricted or global constructivism, on the other. See Sharon Street, “Constructivism About Reasons,” in Russ Shafer-­Landau, ed., Oxford Studies in Metaethics, vol. 3 (Oxford: Oxford University Press, 2008), 207–45; Street, “What is Constructivism in Ethics and Metaethics?” Philosophy Compass 5 (2010): 363–84; Melissa Barry, “Constructivism,” in Tristam McPherson and David Plunkett, eds., The Routledge Handbook of Metaethics (New York: Routledge, 2017), 385–401; David Enoch, “Can there Be a Global, Interesting, Coherent Constructivism About Practical Reason?” Philosophical Explorations 12 (2009): 319–33. Views of the first type construct one normative conception in terms of another. Leading examples are Rawls’s conception of justice as fairness and Scanlon’s construction of a conception of wrongness from reasons. Rawls, A Theory of Justice; T.M.  Scanlon, What We Owe to Each Other (Cambridge, MA: The Belknap Press of Harvard University Press, 1998). My contrast between Ambitious and Modest Constructivism is distinct, but related. Ambitious Constructivism is a possible strategy for realizing global constructivism, and one might even argue that it is the only possible strategy for doing so. (But, notably, Korsgaard interprets Rawls’s constructivism as an Ambitious Constructivism. Christine  M.  Korsgaard, “Realism and Constructivism in Twentieth-­ Century Moral Philosophy,” in The Constitution of Agency, 302–26.) 6. See, e.g., Korsgaard, The Sources of Normativity, 33–40; Christine M. Korsgaard, “The Normativity of Instrumental Reason,” in The Constitution of Agency, 53–5; Korsgaard, “Realism and Constructivism in Twentieth-­ Century Moral Philosophy,” 316–17; Christine M. Korsgaard, Self-­Constitution: Agency, Identity, and Integrity (New York: Oxford University Press, 2009), 5–7, 32–3. 7. Korsgaard, The Sources of Normativity, 33, 61, 257–8. Cf., Kant, Critique of Pure Reason, e.g., A508/B536–A510/B538. 8. On internal standards, see Korsgaard, “The Normativity of Instrumental Reason,” and cf., “Self-­Constitution in the Ethics of Plato and Kant,” in The Constitution of Agency,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

72  Kyla Ebels-Duggan

9.

10.

11. 12.

13. 14.

15. 16. 17.

110–13. See also Christine M. Korsgaard, “The Activity of Reason,” Proceedings and Addresses of the American Philosophical Association 93 (2009): 23–43. H.A. Prichard, “Does Moral Philosophy Rest on a Mistake?” Mind 21 (1912): 21–37. Cf., the discussions of Prichard’s Dilemma in Scanlon, What We Owe to Each Other and Korsgaard, The Sources of Normativity. Note that Prichard himself does not draw a skeptical conclusion here, but rather rejects the question as a mere pseudo-­question, and draws a realist conclusion. Scanlon, What We Owe to Each Other, 149–51, contends that Ambitious Constructivism would not escape the dilemma even if it succeeded in meeting the high bar set by (2). He holds that this sort of view, which he calls a formal theory of moral motivation, would represent moral error as just a species of logical error, and that that has the same sort of problem as thinking of it as a species of prudential error: it is a way of replacing, rather than accounting for or vindicating, the idea of a distinctively moral error and so distinctively moral reasons. But this does not seem right. A successful formal theory uses logic or formal principles of reasoning in order to establish a contentful conclusion, for example the conclusion that we must treat each person as an end, never merely as a means. The advocate of the formal theory can hold that the distinctive nature of a moral wrong is to act against this conclusion, rather than understanding the wrong as just a matter of violating the laws of logic. Korsgaard, The Sources of Normativity, 9. Korsgaard, The Sources of Normativity, 9. Scanlon appears to concede this description of the realist, and to acknowledge that such insistence could not be thought likely to move a skeptical interlocutor. Scanlon, Being Realistic About Reasons, 9–10. Korsgaard, The Sources of Normativity, 38. Scanlon’s strategy for establishing what he calls the priority of morality arguably faces this sort of problem. See Scanlon, What We Owe to Each Other, 160ff. He begins by drawing attention to the value that he claims underlies moral demands, the value of standing to others in relationships of mutual justification. After urging us to recognize this relationship as valuable, he must further assert that this value is very great, in particular that it is great enough that our reasons to stand in these relationships override our other reasons in all or at least almost all cases. This, on his view, is what the priority of morality comes to. According to Korsgaard’s approach, this does not go far enough. At best, Scanlon shows that moral reasons are stronger or weightier than other reasons that we have. By contrast, Korsgaard believes that the categorical demand of morality differs from other reasons not just in degree, but in kind. If we establish that moral demands are cat­eg­ o­ri­cal in the way that they purport to be, we close the question about whether some other value could yet undermine or outweigh them. But Scanlon’s approach could not do this. Even if we are convinced of the great value of mutual justification, we need to compare this to other values one by one, and render contestable judgments about the relative import of the moral value in each case. Cf., Korsgaard, “Realism and Constructivism.” Korsgaard, “Realism and Constructivism.” Korsgaard, “Realism and Constructivism,” 315–17.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

A Question of One ’ s Own  73 18. 19. 20. 21.

22. 23. 24.

25. 26.

Korsgaard, The Sources of Normativity, 16. Korsgaard, The Sources of Normativity, 9. Kant, Groundwork, 4:453. In What We Owe to Each Other, 148, Scanlon says that the question cannot always be answered, but does not deny that there is such a question. He expresses more doubt in Being Realistic About Reasons, 9–14. Later, he discusses whether, once we have granted that some consideration is a reason, we can then meaningfully raise a further question about whether to acknowledge its authority. He argues that we cannot: “But this idea of an ‘authority’ is a misleading metaphor. For any standard, in the form of a set of principles or precepts, there is the question ‘Why do that?’ But when I arrive at a conclusion about the correctness of a normative judgment . . . there is no further such question. These conclusions carry their own normative authority, as it were. They do not need to derive it from some further source.” Being Realistic About Reasons, 68. Though he is here discussing a purported question about the authority of reasons more generally, it is clear how to apply the argument to the question about morality, or moral demands or reasons. We can ask of any such demand: “Why do that?” But this is only a first-­order question about what the reasons for doing some particular thing are, and whether they are really compelling. Though in Scanlon, “The Ideas of the Good,” he seems to acknowledge such a question, which he there calls the question of acceptance. Korsgaard, The Sources of Normativity, 42, 68. Compare Melissa Barry’s understanding of constructivism as hypothetical proceduralism: the constructivist holds that normative conceptions are the results of applying a given procedure. Barry, “Constructivism.” This follows Korsgaard’s contrast between procedural and substantive realism, in The Sources of Normativity, 35, but raises the question: what is the normative status of the procedure (cf., Enoch, “Can there be?”)? The second thesis of Ambitious Constructivism avoids this problem by appealing only to “procedures” of valid reasoning. In “What is Constructivism in Ethics and Metaethics?” Street argues that constructivism is best understood as holding that a normative conception follows from a standpoint or point of view. Whether this differs from the idea that normative concepts are placeholder answers to questions depends on how we understand what is meant by a point of view. On one reading, a point of view is constituted by some substantive commitments. This seems to be what Street has in mind when she glosses a point of view as that of “an agent or set of agents who accept certain values (or whose situation embodies certain values, as in Rawls’s original position),” 366, and cf., Street, “Constructivism About Reasons,” 208. On another reading, a point of view is instead defined by a task or problem. Street’s characterization of “the practical point of view” as that of a creature who values something sounds more like this: the practical point of view is constituted by the task of valuing something, or the question of what to value. “Constructivism About Reasons,” 366. The former reading sets Street’s view in contrast to mine, while the latter makes them much more similar, and perhaps identical. Kant, Groundwork, 4:421. Scanlon, What We Owe to Each Other, though Scanlon puts his view in terms of the concept of wrongness.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

74  Kyla Ebels-Duggan 27. Korsgaard, The Sources of Normativity, 9 (emphasis added). 28. John Stuart Mill first argues that we are obligated to promote the greatest happiness overall. Then he asks, what is the sanction of morality, what gives us reason to follow it? He takes the connection between an action’s being wrong and being such that we have decisive reason to avoid it to be an external, contingent connection. John Stuart Mill, Utilitarianism, ed. John Cottingham (New York: Oxford University Press, 1998). By contrast, in the Groundwork, Kant seeks to establish an internal connection: the very thing that makes an action wrong is also the reason not to do it. Establishing this sort of connection would provide the answer to the normative question. 29. Korsgaard may be attributing to the realist the following conception of wrongness: what is wrong is what is ruled out by the moral facts. Such a conception would give rise to the objections in Korsgaard, “Realism and Constructivism.” But Korsgaard considers Scanlon a substantive realist, despite the fact that Scanlon advances the more contentful conception of wrongness mentioned above. Whether her case against realism applies to this sort of view is somewhat less straightforward. 30. Kant, Groundwork, 4:420. 31. Kant, Groundwork, 4:420–1. 32. Christine  M.  Korsgaard, “Kant’s Formula of Humanity,” in Creating the Kingdom of Ends; Korsgaard, Sources, 90–100; Christine  M.  Korsgaard, Fellow Creatures: Our Obligations to the Other Animals, 118–22. 33. Kant, Groundwork, 4:421. 34. Kant, Groundwork, 4:422–4. 35. Korsgaard, “Kant’s Formula of Universal Law.” Cf., John Rawls, Lectures on the History of Moral Philosophy (Cambridge, MA: Harvard University Press, 2000), 162–80. 36. Kant, Groundwork, 4:422. 37. In Kant’s own thinking this shows up in his denial, in the Critique of Pure Reason, that we can know the nature of things in themselves. 38. Korsgaard, “Realism and Constructivism,” 322. 39. The test does not run afoul of Scanlon’s criticism, described in note 10, for what is wrong with the wrong action is not just that it is based on poor reasoning on the part of its agent. What is wrong is more contentful and specific: it is that the proposed action would be ineffective in the world of the universalized maxim. That, on this interpretation, is what it means for a maxim to fail to have the form of law. 40. Kant, Groundwork, 4:445. 41. Christine M. Korsgaard, “Morality as Freedom,” in Creating the Kingdom of Ends. 42. The inescapability of the question is what Korsgaard, in “The Activity of Reason,” describes as our plight. Cf., Korsgaard, Sources, 95. 43. Cf., Barbara Herman, “Moral Deliberation and the Derivation of Duties,” in The Practice of Moral Judgment (Cambridge, MA: Harvard University Press, 1993), chapter 7 and cf., Karen Stohr, Choosing Freedom (Oxford: Oxford University Press, 2022), chapter 5. 44. In The Sources of Normativity, she distinguishes between the Universal Law formulation, which she calls the Categorical Imperative, and “the moral law,” which she identifies with the Formula of Humanity. The latter she views as fixing a domain that yields determinate and recognizably moral content. See 92–125. 45. For alternative readings of the argument, see Adrienne M. Martin, “How to Argue for the Value of Humanity,” Pacific Philosophical Quarterly 87 (2006): 96–125; Japa

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

A Question of One ’ s Own  75 Pallikkathayil, “Deriving Morality from Politics: Rethinking the Formula of Humanity,” Ethics: An International Journal of Social, Political, and Legal Philosophy 121 (2010): 116–47; and Nandi Theunnissen, “Might We Be Just Plain Good: On Regress Arguments for the Value of Humanity,” Ethics 128 (2018): 346–72. 46. Street treats this argument as representative of Korsgaard’s constructivism, but argues that the formal constraints imposed by valuing something yield no content. Street, “What Is Constructivism in Ethics and Metaethics?” 47. Korsgaard, The Sources of Normativity, 163. For this argument to yield a conception of obligation it must also take the further step of specifying what, exactly, valuing humanity requires of us. In the terms used above, it must tell us that some actions are wrong or required. In “Kant’s Formula of Humanity,” Korsgaard offers a reading of the Formula of Humanity designed to meet this challenge: we fail to treat someone as an end when we treat her in ways to which she could not consent. The promise of this interpretation is that it will generate first-­order moral conclusions without relying on moral premises. Cf., Onora O’Neill, “Between Consenting Adults,” in Constructions of Reason (New York: Cambridge University Press, 1989), 105–25. For an argument against the possible consent interpretation, see Pallikkathayil, “Deriving Morality from Politics.” 48. In The Sources of Normatvity, Korsgaard appears satisfied with this conclusion. But in Self-­Constitution she returns to the more demanding standard of demonstrating that the authority of the moral demand is absolutely unconditional, and to the hope that the best conception of freedom and the best conception of obligation are one and the same. 49. Obviously this echoes Korsgaard’s notion of practical identity in The Sources of Normativity. 50. Korsgaard, The Sources of Normativity, 9. 51. Herman, “Moral Deliberation,” 138–40. 52. Cf., Elizabeth Anderson, Value in Ethics and Economics (Cambridge, MA: Harvard University Press, 1993), 93–8 on the need for substantive judgment about ideals. 53. It seems to me that the best answers here will rely on ideals of relationships with others. That such relationships are of value, and also that others are valuable in ways that the relationships can both acknowledge and respond to, are both attractive normative claims. Here also cf., Anderson, Value in Ethics and Economics, 93, for a list of constraints that govern our deliberations about objective values, and Rawls, A Theory of Justice, 48–51 for discussion of comparing conceptions. 54. See, for example, Korsgaard, “Morality as Freedom,” especially 171–83; Korsgaard, “Creating the Kingdom of Ends”; Korsgaard, “Self-­Constitution in the Ethics of Plato and Kant.” 55. See Korsgaard, “Morality as Freedom,” where she argues that we have reason to identify with our noumenal selves. What sort of reason is this? It seems to arise from the fact that the freedom characteristic of the noumenal self is an ideal worth striving to realize. 56. In What We Owe to Each Other, 160–8, Scanlon follows the same model in his address to Prichard. Developing an analogy to friendship, he argues that guiding ourselves by his conception of obligation allows us to stand in a relationship that he calls mutual recognition. This is a relationship that we have reason to value and so try to realize with one another.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

76  Kyla Ebels-Duggan 57. Cf., Kyla Ebels-­Duggan, “Learning from Love: Reasoning, Respect, and the Value of a Person,” in Sara Buss and Nandi Theunissen, eds., Reconsidering the Value of Humanity (Oxford: Oxford University Press, forthcoming).

Bibliography Anderson, Elizabeth. Value in Ethics and Economics. Cambridge, MA: Harvard University Press, 1993. Barry, Melissa. “Constructivism.” In Tristam McPherson and David Plunkett, eds., The Routledge Handbook of Metaethics, 385–401. New York: Routledge, 2017. Cohen, G.A. “Reason, Humanity and the Moral Law.” In Christine M. Korsgaard, The Sources of Normativity, 167–87. Cambridge: Cambridge University Press, 1996. Ebels-Duggan, Kyla. “Learning from Love: Reasoning, Respect, and the Value of a Person.” In Sara Buss and Nandi Theunissen, eds., Reconsidering the Value of Humanity. Oxford: Oxford University Press, forthcoming. Enoch, David. “Can there Be a Global, Interesting, Coherent Constructivism About Practical Reason?” Philosophical Explorations 12 (2009): 319–39. FitzPatrick, William. “The Practical Turn in Ethical Theory: Korsgaard’s Constructivism, Realism, and the Nature of Normativity.” Ethics 115 (2005): 651–91. Herman, Barbara. “Moral Deliberation and the Derivation of Duties.” In The Practice of Moral Judgment, chapter 7. Cambridge, MA: Harvard University Press, 1993. Kant, Immanuel. Critique of Pure Reason. Translated by Paul Guyer and Allen W. Wood. New York: Cambridge University Press, 1998. Kant, Immanuel. Groundwork of the Metaphysics of Morals. In Practical Philosophy, translated and edited by Mary  J.  Gregor. Cambridge: Cambridge University Press, 1999. Korsgaard, Christine  M. “Creating the Kingdom of Ends.” In Creating the Kingdom of Ends, 188–222. New York: Cambridge University Press, 1996. Korsgaard, Christine M. “Kant’s Formula of Humanity.” In Creating the Kingdom of Ends, 106–32. New York: Cambridge University Press, 1996. Korsgaard, Christine  M. “Kant’s Formula of Universal Law.” In Creating the Kingdom of Ends, 77–105. New York: Cambridge University Press, 1996. Korsgaard, Christine M. “Morality as Freedom.” In Creating the Kingdom of Ends, 159–87. New York: Cambridge University Press, 1996. Korsgaard, Christine  M. The Sources of Normativity. Cambridge: Cambridge University Press, 1996.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

A Question of One ’ s Own  77 Korsgaard, Christine  M. “The Normativity of Instrumental Reason.” In The Constitution of Agency: Essays on Practical Reason and Moral Psychology, 27–68. New York: Oxford University Press, 2008. Korsgaard, Christine  M. “Realism and Constructivism in Twentieth-Century Moral Philosophy.” In The Constitution of Agency: Essays on Practical Reason and Moral Psychology, 302–26. New York: Oxford University Press, 2008. Korsgaard, Christine  M. “Self-Constitution in the Ethics of Plato and Kant.” In The Constitution of Agency: Essays on Practical Reason and Moral Psychology, 100–28. New York: Oxford University Press, 2008. Korsgaard, Christine  M. “The Activity of Reason.” Proceedings and Addresses of the American Philosophical Association 83 (2009): 23–43. Korsgaard, Christine  M. Self-Constitution: Agency, Identity, and Integrity. New York: Oxford University Press, 2009. Korsgaard, Christine M. Fellow Creatures: Our Obligations to the Other Animals. Oxford: Oxford University Press, 2018. Martin, Adrienne  M. “How to Argue for the Value of Humanity.” Pacific Philosophical Quarterly 87 (2006): 96–125. Mill, John Stuart. Utilitarianism, edited by John Cottingham. New York: Oxford University Press, 1998. O’Neill, Onora. “Between Consenting Adults.” In Constructions of Reason, 105–25. New York: Cambridge University Press, 1989. Pallikkathayil, Japa. “Deriving Morality from Politics: Rethinking the Formula of Humanity.” Ethics: An International Journal of Social, Political, and Legal Philosophy 121 (2010): 116–47. Prichard, H.A. “Does Moral Philosophy Rest on a Mistake?” Mind 21 (1912): 21–37. Rawls, John. A Theory of Justice. Cambridge, MA: Harvard University Press, 1971. Rawls, John. Lectures on the History of Moral Philosophy. Cambridge, MA: Harvard University Press, 2000. Scanlon, T.M. What We Owe to Each Other. Cambridge, MA: The Belknap Press of Harvard University Press, 1998. Scanlon, T.M. Being Realistic About Reasons. New York, NY: Oxford University Press, 2014. Stohr, Karen. Choosing Freedom. Oxford: Oxford University Press, 2022. Street, Sharon. “Constructivism About Reasons.” In Russ Shafer-Landau, ed., Oxford Studies in Metaethics, vol. 3, 207–45. Oxford: Oxford University Press, 2008. Street, Sharon. “What is Constructivism in Ethics and Metaethics?” Philosophy Compass 5 (2010): 363–84. Theunnissen, Nandi. “Might We Be Just Plain Good: On Regress Arguments for the Value of Humanity.” Ethics 128 (2018): 346–72.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

4 The Two Normativities J. David Velleman

1 Two kinds of normativity have been identified by moral philosophers. One is the normativity of action- or attitude-­guiding language; the other is the normativity of reasons for actions or attitudes. The normativity of language and the normativity of reasons have much less to do with each other than is generally supposed. Hume’s argument about “is” and “ought,” which is an argument about language, has very little to do with his argument about reason and the passions, which is an argument about reasons. “Ought” is a verb lexically designated as forming injunctions, recommendations, and similar speech acts; no such use is lexically designated for “is.” Of course, statements with “is” can have the illocutionary force of a recommendation or injunction, but not in virtue of a use designated for the verb in the English lexicon; such illocutionary force must be provided by the context, which may include gestures, tone of voice, presuppositions as to the speaker’s or hearer’s attitudes, and so on. One argument for the impossibility of deriving “ought” from “is” is purely linguistic. Any conclusion having context-­independent force as a recommendation or injunction must in that respect be ampliative in relation to premises that lack such force, hence not derivable from them via any non-­ampliative procedure of derivation. Under another interpretation, Hume’s argument is also meant to rule out ampliative inferences from “is” to “ought,” in which case it would not be a purely linguistic argument; but neither would it be half as persuasive as it is when taken linguistically. The statements that express reasons for acting are not normative in this linguistic sense. Reasons for acting are facts, often naturalistic facts, that can be asserted without recommending or injunctive force. They can also be asserted with such force, but it must derive in part from contextual factors over and above the lexically designated use of the words. When only their J. David Velleman, The Two Normativities In: Normativity and Agency: Themes from the Philosophy of Christine M. Korsgaard. Edited by: Tamar Schapiro, Kyla Ebels-­Duggan, and Sharon Street, Oxford University Press. © J. David Velleman 2022. DOI: 10.1093/oso/9780198843726.003.0004

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Two Normativities  79 lexically designated use is considered, the statement expressing a reason has only the force of an assertion. Hume’s argument about reason and the passions is not linguistic; it is psychological. He argued that action can be in or out of accord with reason only insofar as it can be governed by reason; whereas the role of reason is to govern belief, which doesn’t govern action except in combination with desire or some other conative attitude. Although conation can be transferred from object to object by a belief in the latter as means to or part of the former, it has to start with an object to which it hasn’t been transferred from elsewhere, hence independently of belief and reason. These two arguments have defined the problematic of metaethics since Hume. The first, linguistic argument has morphed into the question whether normative statements have truth-­conditions or (as the jargon has it) whether they are truth-­apt. The second argument, about the faculty of reason—­reason in the singular—has morphed into a question about reasons in the plural, that is, considerations that count for or against an action. The question in this case is whether considerations can favor an action if they cannot move one to perform it, or (as the jargon has it) whether there can be external ­reasons for acting.

2 The distinction between internal and external reasons has become blurred since its famous treatment by the foremost contemporary Humean, Bernard Williams.1 The definition I’ve just proposed is that external reasons are considerations that favor an action but cannot move one to perform it. Williams simply assumed that reasons cannot be external in this sense, because reasons for one to act must have the potential to become the reasons for which or on the basis of which one does act. But Williams didn’t use the term “external reasons” in this sense; he used it for considerations that favor an action but cannot appeal to conative attitudes one already has or could arrive at by way of reflection—or, as he put it, by taking a “sound deliberative route” from one’s “subjective motivational set.” I’ll mark these two senses of the distinction between internal and external reasons by calling the former sense “practical” and the latter “psychological.” Williams just assumed that reasons must be internal in the practical sense, that is, capable of moving one to act and thus becoming the reasons for which one acts. That was his premise. His conclusion was that reasons must be internal in the psychological sense as well, that

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

80  J. David Velleman is, capable of engaging one’s subjective motivational set, at least as it would be after sound reflection. Williams noted that even if a consideration could not engage one’s subjective motivational set, it might perhaps be capable of leading to action if one had a rational (hence, presumably, not “subjective”) desire to act in ac­cord­ ance with reasons as such. He then ruled out that hypothesis on the grounds that considerations could not engage that desire unless they were first recognized as reasons; and how could they be recognized as reasons, he asked, unless they were first recognized as a basis on which one might act, hence as internal in the practical sense? Considerations would have to be recognized as practically internal in order to be recognized as reasons; but under the present hypothesis, they wouldn’t be internal in that sense unless they were already recognized as reasons. Caught in this circle, one could never get started acting on the basis of psychologically external considerations. Williams thus rejected the possibility of psychologically external reasons, on the grounds that they could never become practically internal, as all reasons must be. Derek Parfit claimed that reasons for acting can be external in both senses and, more importantly, that they can be practically internal without being internal psychologically.2 That is, considerations can be reasons, and can move one to act, without appealing to conations that are in, or are rationally accessible from, one’s subjective motivational set. Parfit’s main argument for his externalist theory of reasons was that it yields more intuitively satisfying results than internalism in various cases. But because Parfit thought that the nature of reasons is unanalyzable, and he didn’t reveal how reasons can be recognized, we simply have to take his word for the results he claimed to obtain in these cases, like the gamblers who have to trust the high-­roller Big Julie to remember where the spots formerly were on his all-­white dice.3 I suspect that Parfit wasn’t worried about getting counterintuitive results from his own theory because his epistemology of reasons was intuitionist. For if reasons are detected by intuition, then of course the reasons in a given case will turn out to be what we intuitively think they are. Unfortunately, reasons for acting are unlike other matters considered accessible to intuition, such as the facts of arithmetic, because they are a matter of deep disagreement. Is a person’s wrongdoing any reason in itself to make him suffer? Is social inequality a reason to favor forcible redistribution? Is there ever a reason to kill an innocent person? Is greater biodiversity a reason to accept slower economic growth? Is there any reason to pity a person who dies young? Intuitions differ. According to Parfit, one side of each question is wrong, but there is nothing that makes it wrong—its wrongness is a brute fact—and we just have to see it.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Two Normativities  81 This theory leaves the interesting questions unresolved on the grounds that they are conveniently unresolvable. That said, I do not accept Williams’s argument against externalism about reasons, nor do I accept his version of internalism. On the contrary, I think he was on the right track when he began to speculate about how reasons that are psychologically external could satisfy the requirement that they be practically internal—that is, how reasons that cannot engage our pre-­existing conations, even after reflection, can nevertheless become the reasons for which we act. For I believe that reasons owe their behavioral influence to something like the conation that Williams briefly considered but dismissed, namely, the desire to act in accordance with reasons.4

3 Another Humean who argued from practical to psychological internalism was Donald Davidson. Davidson pointed to the possibility of having several reasons for an action but performing it on the basis of only one, and he asked what differentiates the reason on which one acts from the reasons one has but doesn’t act on.5 His answer was that the reason on which one acts is the reason that causes one’s action, which it must do via the psychological process of motivation. A reason for acting must therefore be a potential motive, consisting of a belief-­desire pair, in which the belief represents the action as a means to the object of the desire. Davidson thought that when a belief-­desire pair motivates action, it also, in that very process, serves as the agent’s reason for acting. Acting for a reason, according to Davidson, is simply a matter of being moved to act by a means-­end related pair of belief and desire. A serious—in my view, fatal—flaw in Davidson’s theory is that it fails to explain the relation between a reason’s justificatory force and its motivational force. According to Davidson, a belief-­desire pair can justify action, at least in the agent’s eyes, by representing it as conducive to something desirable. But even if desiring something can be described colloquially as holding it to be desirable,6 it does not fit that description in the way required to justify action, not even to justify it subjectively, in the eyes of the agent. Davidson’s view of desire as half of a subjective justification depended on the age-­old assumption that we desire things “under the guise of the good,” and as I have argued in the past,7 this assumption is false, at least in the sense required for subjective justification. Unfortunately for my present purposes, I cannot repeat that argument here.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

82  J. David Velleman In any case, the influence of belief-­desire pairs in their capacity as reasons is distinct from their influence as motives. An agent who acts for reasons consisting in a belief-­desire pair is normally influenced by those attitudes twice: once by their motivational force and also, independently, by their force as justifications. A rational agent can perceive whether he is justified in his pursuit of a desired end, and he can rein in or ride the desire accordingly, under the influence of that perception. Because an agent can be encouraged to pursue a desired end by the sense of doing so with justification, or inhibited from pursuing it by the sense of lacking justification, acting with justification must be the object of some conative attitude on his part. And acting with justification is a second-­order end, that is, an end with respect to the manner in which first-­order ends are pursued. Even if we begin by thinking that reasons for acting are relative to the agent’s desires, we have to acknowledge that the agent can also be encouraged by an awareness of desiring the ends and the sense of being thereby justified in adopting means to them; or inhibited by a lack of that awareness or of the accompanying sense of justification. And then we should ask ourselves whether the action-­guiding force of reasons consists in the first-­order motivational force of the desire or rather in the force of the awareness of it as justifying action, together with a second-­order conation toward acting with justification. The clearest evidence for such a conation may be that people sometimes base their actions on desires they don’t have, pursuing ends that they merely think they want, only to discover that they never actually wanted them. When people are influenced to pursue something by merely thinking they want it, they are acting on that basis as their (albeit unsound) reason, and given that their motive cannot be the merely imagined desire, I suggest we consider that it’s the reason they imagined it to provide, plus a conation toward acting with justification. Admittedly, such a conation is not the only possible explanation for pursuing an end falsely believed to be wanted. Maybe people have an egoistic second-­order desire to get what they want. That desire would lead them to see a reason for pursuing what they think they want, as a means to the desired end of desire-­satisfaction, even if they didn’t have the desire they took themselves to be satisfying. Yet this competing explanation is undermined by a different case, in which an agent is moved by a desire that he isn’t aware of having. In that case, his action is likely to be impulsive, manifesting an urge that seems to have pre-­ empted practical reasoning, even though his motives include the knowledge

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Two Normativities  83 that the action is conducive to an (unwittingly) desired end. His action may even amount to symptomatic behavior of the sort that places his very agency in question, given that he was driven to it unconsciously. The reflexive “dig” that betrays unacknowledged resentment, the home-­ stretch stumble that manifests fear of success, the conveniently missed appointment—such behaviors exemplify a pathology of human agency despite being known means to desired ends. Why would the agent’s ignorance of his motive alter the quality of his behavior so radically if it merely blocked the influence of an ancillary, second-­ order motive for his action? Why is the resulting behavior so unlike ordinary cases of acting on the basis of a reason, given that it is motivated by the agent’s desire for something to which he believes it conducive? The most plausible answer, to my mind, is that what becomes disengaged when the agent is unaware of his first-­order motive is not just any additional motive but rather a motive that’s crucial to the normal functioning of human agency. And a conation toward acting with justification is a more plausible candidate for that role than egoism. I therefore favor the view that even if reasons are beliefs about means to desired ends (or, as Davidson thought, the belief and desire together) their normative force is distinct from the motivational force they exert. The normative force of reasons is rather the motivational force of whatever conation it is that inclines the agent toward actions for which he sees justification and against actions for which he sees none; and such a conation is precisely what Williams posited in his reductio of externalism.

4 Where Williams went wrong, I suggest, was in his claim that positing this desire would lead to absurdity. For there are two different ways of understanding the hypothesis. We can understand the hypothesis as characterizing the relevant conation de dicto, that is, as a conation toward actions under the description that they have some justification; or we can understand it as characterizing the conation de re, as a conation toward actions with some characteristic that does in fact justify them, though not toward them under that description. Understood de dicto, the conation cannot motivate the agent to act for a reason until he recognizes it as a consideration that justifies acting, hence until he recognizes it as a reason; and then the question arises how he can recognize anything as a

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

84  J. David Velleman reason without recognizing it as the sort of thing that can motivate him— which Williams thought it cannot do until so recognized. That was Williams’s argument, and it relied on the de dicto reading of the rational-­ motive hypothesis.8 Suppose, then, that we understand the hypothesis de re, as positing a conation toward actions with a particular characteristic or characteristics that, as a matter of fact, are justifying. Williams’s argument would not touch the hypothesis so understood, and that’s how I propose to understand it.

5 For a well-­known version of this hypothesis, I turn to Aristotle, who suggests two different strategies for blocking the question “What for?” The more general of these strategies is to divide human action into poiesis (“making”) and praxis (“activity”). The end of poiesis is a product that is distinct from the action itself; the end of praxis is always eupraxia, doing well, which is inseparable from the praxis.9 As I interpret Aristotle, these categories of action are not mutually exclusive; on the contrary, a poeisis can also be viewed from a different perspective as a praxis. Teaching philosophy can be a poiesis insofar as it aims to produce understanding in the minds of the students; it can at the same time be a praxis insofar as it aims at being done well. Under the former aspect, its success is judged by the students’ performance on the final exam; under the latter, it is judged by whether it manifests the teacher’s pedagogical virtues of expertise, eloquence, wit, and so on. What interests me about this statement of Aristotle’s is that eupraxia can be the end of every praxis only if it constitutes the activity, in the sense that what one is doing is determined by what one is aiming to do well. Otherwise, there will always be the possibility of doing something without aiming to do it well, because one is aiming to do it indifferently or even badly. The only way to rule out that possibility is to say that in aiming to do something indifferently or badly, one is not in fact engaged in that praxis but is rather engaged in a different one instead: not eating dinner but grazing or snacking or picking at one’s food; not playing checkers but humoring a child by pretending to lose; not sailing but messing about in boats.10 Humoring a child by losing a game of checkers is something that can itself be done well, not by playing checkers well, of course, but by putting on a convincing pretense of trying to. If the pretense fails, the child might protest, “You’re not playing the game!”—and rightly so.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Two Normativities  85 Whether Aristotle thought that there is a constitutive aim of all human action depends on whether he thought that every action can be view as a praxis, even if it is also a case of poiesis—a question on which I have no opinion. In any case, he certainly thought that living a human life is a praxis and that we are necessarily aiming at eupraxia in it, which he says is one and the same as eudaimonia. He defines eudaimonia as eu zēn kai eu prattein— living well and doing well, where the Greek for “doing well,” eu prattein, is the verbal form of the noun eupraxia. That’s why he thinks we necessarily aim at eudaimonia: it’s what we are aiming at simply in virtue of living human lives. Not to aim at it would leave one messing about among the materials of life without ever getting underway as a human being. And because it is necessarily the end of living a human life, it is necessarily justifying for us.

6 Within this general constitutivism about human agency, Aristotle offers a more specific constitutivism applying to prohairesis, or choice. He says that there are just three objects of choice: the useful, the pleasant, and the noble or the “fine.”11 I take it that Aristotle meant these terms as what Elizabeth Anscombe called desirability characterizations—that is, characterizations that forestall the question “What do you want that for?”12 A person can be asked what he wants flaming torches for; and if he answers “To juggle with,” he can be asked what he wants to juggle flaming torches for; but if he answers, “Juggling flaming torches is fun,” he has reached a desirability characterization, because the desire for fun cannot be challenged with the question “What do you want that for?” (The result would have been the same, according to Aristotle, if the answer had been “For a circus performer like me, juggling flaming torches is a fine thing to do.”) Desirability characterizations leave no open question “What for?,” and so they bring the regress of justifications to a close. To say that the objects of choice must fall into one of three categories—the useful, the pleasant, and the noble—is to say that any psychological phenomenon whose object falls outside these categories cannot be a choice. It is therefore a version of constitutivism, and it has the notable feature that at least two of the categories are not explicitly normative. For Aristotle, it seems, our desire for pleasure is partly constitutive of our agency.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

86  J. David Velleman

7 This digression into Aristotle’s Ethics is intended merely to sketch the general form of the view I favor—the view that human agency is constituted by conation toward behavior having particular characteristics that are, in fact, justifying. Those characteristics, whatever they are, must be the aim of any creature who functions as an agent, hence an aim constitutive of action. Not to aim at them is not to act. If the relevant aim is not one of those cited by Aristotle, what is it? As I suggested a moment ago, it is an aim that makes the difference between impulsive, unwitting, or even neurotic behavior—the reflexive “dig,” the missed appointment, the home-­stretch stumble—and well-­formed human action. The conation constitutive of action must therefore be the conation that plays that role. That conation will be the second-­order motive that adds its force to beliefs and desires when someone acts not just out of them, as motives, but on them, as reasons. And their normative force as reasons will turn out to be their potential to engage that conation via his awareness of them (or, as the case may be, his false belief that an action would have them). This method of philosophical discovery is functionalist: it interprets the terms “normativity” and “normative force” as denoting something whose role in human behavior can be described in non-­normative terms; and it identifies normativity with whatever actually plays that role. The role played by normative force is that its influence transforms what would otherwise be impulsive, unwitting, or neurotic behavior into human action.

8 What plays this role in human psychology, I believe, is a drive toward intelligibility. I believe that various features by which we recognize behavior as an exercise of human agency can be explained in terms of this drive. To wit. When behavior falls under the influence of this drive, there is a sense in which it is caused by the agent, not just by his impulses, and a sense in which it is undetermined by prior events from the agent’s point of view. The drive toward intelligibility accounts for the agent’s non-­observational knowledge of what he is doing and will do. It accounts for various instances of normativity, from the axioms of formal decision theory to personal ideals. It accounts for the multiplicity of values; for some moral emotions, such as shame; and for the rationality of co-­operative agreements in prisoners’ dilemmas. And its proper

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Two Normativities  87 function is that whose failure accounts for some generally recognized ­pathologies, such as the experimentally demonstrated phenomena of cognitive dissonance and the attribution effect. These are bald assertions, but this is not the place to defend them.13 My purpose here is to locate the resulting constitutivist theory on the broad map of metaethics. I suspect that the foremost metaethical constitutivist, honored in this volume, will find the locale uncongenial—for one thing, it is in the same general neighborhood as expressivism—and yet I believe it is not very far from the neighborhood that she has explored over the course of her distinguished career. Expressivists bypass head-­on analysis of a term like “ought” in order to analyze what is going on when someone uses the term: the speaker is expressing a psychological state that can be described and studied naturalistically, by the role it plays in conversation and social interaction. This methodology is sometimes described as a “sideways-­ on” view of normativity.14 Like the expressivist looking sideways-­on at “ought,” I bypass a head-­on analysis of reasons and look instead at what is going on when someone acts for a reason: the agent’s behavior is being influenced by his perception of a justification. That too must be the manifestation of a psychological state that can be studied naturalistically—in this case, by the role it plays in giving human action its characteristic features, such as the features I enumerated in the previous paragraph but one. That psychological state must constitute the agent’s susceptibility to the normative force of reasons, and so their normative force must consist in what it makes them susceptible to. This methodology also resembles expressivism in that it severs metaethics from normative discourse and hence from normative ethics. Some ­philosophers say that “the normative” is an autonomous realm that cannot be studied from the outside, like a society that can be studied only by ­participant observers.15 For them, metaethics is necessarily a normative discipline. I disagree. Functionalism about normativity presupposes that metaethics is purely descriptive. It has implications with respect to what people have ­reason to do, but so can a descriptive theory of anything; and it is descriptive only in the sense that it does not contain, and does not by itself imply, any ­linguistically normative statements: it does not entail any “oughts.” Although metaethical theories neither contain nor imply any linguistically normative statements, they can have implications with respect to what one has reason to do; for implications about what one has reason to do need not be linguistically normative. Of course, telling someone that he has reason to do something can be a way of recommending or even demanding that he do

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

88  J. David Velleman it: the speech act can have the illocutionary force of a recommendation or injunction. But that force must rely on additional features of the context: it is not inherent in the meaning of the words, as we can tell from cases in which the same statement is neither an endorsement nor an injunction. Some people have reason to believe outlandish falsehoods because their prior beliefs are so screwy that no sound epistemic route would lead them to the truth: only non-­rational procedures such as conversion or indoctrination or therapy would help. Conceding that they have reason to believe these things is just a way of saying that reasoning will not lead them aright, not that they ought to have those screwy beliefs. Indeed, it is a way describing just how screwy their prior beliefs are—so screwy as to lie beyond the help of reasoning. Similarly, some people’s subjective motivational set is so screwy that without some non-­ rational intervention they will have reason to do outrageous things, and saying that they have reason to do such things need not suggest that they ought to.16 Although saying that someone has reason to do something need not offer practical guidance for it, showing him is a different matter. What someone has reason to do depends on what would make the most sense for him to do. There is no sideways-­on way to show someone that an action would make sense: the only way is to show him the sense that it would make. And according to my view, that showing will give a justification of the action, by giving reasons, because reasons for an action just are considerations in light of which it would make sense. Thus, a reason for acting is action-­guiding even though neither the assertion of the reason nor the assertion that it is a reason is linguistically normative. The assertion of the reason itself is usually a purely descriptive statement whose normativity is rational, not linguistic. It is action-­guiding via the agent’s aim of acting in a way that has a particular property, a property that is, in fact, justifying—namely, intelligibility. Asserting the reason guides the agent by adducing a fact in light of which some action would make sense. And the assertion that it is a reason need not be action-­guiding at all, ra­tion­ ally or linguistically: it is a descriptive statement of metaethics, which is not a normative discipline in the sense that would raise questions about its unity with other naturalistic disciplines.

9 I don’t mean to deny that metaethical theories can have normative implications. They can indeed generate reasons to act—again, without either

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Two Normativities  89 containing or implying any linguistically normative statements. Like most reasons, the reasons generated by metaethics are just descriptive facts. A metaethical theory can give us reasons, according to my view, because it gives us a particular conception of how practical reasoning actually functions, and that conception makes some deliberative moves highly intelligible and others less so. By “deliberative moves” I mean actions such as turning our attention in particular directions, imagining alternative possibilities, totting up pros and cons, arguing with ourselves on the devil’s behalf, and so forth. When engaged in such acts, we can be guided by reasons for deliberating in one way rather than another. A psychological conception of how practical reasoning actually works will enable us to find some ways of deliberating more intelligible than others, thereby giving us reasons for deliberating in those ways. Thus, consequentialists tend to adopt consequentialist methods of deliberation when deliberating explicitly. On other occasions they entertain all sorts of non-­consequentialist reasons, such as the emotions they take themselves to be expressing, the daily routines they are following, or the core traits of character they are manifesting, all adduced without reference to consequences. When they are happy they smile, sometimes only reflexively but sometimes deliberately, on the grounds that they are happy; when sad, they not only break out in tears but go on to indulge deliberately in a good cry; they do some things every morning on the grounds that, well, that’s what they do every morning; there are some especially significant actions they take because, as they put it, “That’s just who I am.” All of these considerations are reasons for acting, if my metaethical theory is correct, because they render the associated actions intelligible, but of course they aren’t reasons according to consequentialism. When challenged, consequentialists claim that their mundane justifications have consequentialist reasoning behind them. And because of the endless plasticity of the descriptions available or devisable for their situations and options, they usually manage to come up with consequentialist reasons. But those consequentialist reasons were not their reasons, not the reasons for which they smiled or cried or brushed their teeth or stood up for a cause. The reasons for which they acted are the considerations that actually influenced them via practical reasoning, and those are the non-­consequentialist reasons they gave before being reminded that they are consequentialists. Their subsequent, post facto rationalizations are the considerations that, as self-­styled consequentialists, they can make sense of having taken into account—another manifestation of the drive toward intelligibility.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

90  J. David Velleman

10 To summarize. Reasons are not normative statements; they state facts in light of which an action would promote not only its first-­order ends but a second-­ order aim that is constitutive of action. Metaethical theory identifies that aim by explaining how it accounts for the characteristic features of human action—the kind of behavior that is based on reasons. This explanation is a theoretical exercise that does not depend on any normative statements as premises, though it may generate reasons as consequences, in the form of factual statements in light of which some deliberative steps but not others promote the aim it identifies as constitutive of action. I admit that this theory seems to leave the normative force out of normativity. It purports to explain how considerations justify action, yet it does so in psychological terms that are not themselves justifying. It explains the justificatory force of reasons without displaying it. In this respect, however, the theory is no different from expressivism, which explains but does not display the normativity of recommendations and injunctions. That’s why expressivism is called a sideways-­on view of normativity. The problem with expressivism is not that its view of normativity is sideways-­on; the problem is rather that it encompasses only linguistic normativity, which should not be the primary focus of moral philosophy. To quote Professor Korsgaard, expressivism is boring.17 Once we distinguish linguistic normativity from the normativity of reasons, we can focus our attention on what really matters to moral philosophy.

11 Let me conclude by commenting on a well-­known objection to functionalist theories of practical reason, the “shmagency” objection of David Enoch.18 Enoch argues that in order to take seriously considerations identified as reasons by the functional role they play in agency, one would need a reason for functioning like that rather than like creatures who function differently (call them shmagents). The first answer to this objection is that if one already is an agent, then one doesn’t need a reason for becoming one; the most one may need is a reason for continuing to be one, and reasons of that sort are easy enough for an agent to find. If one is already a shmagent, of course, then one has no use for reasons for becoming an agent instead; one needs shmeasons, in which case shmagency

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Two Normativities  91 will turn out to be an unstable condition. If, alternatively, the shmeasons come down on the side of shmagency, then there will be two independently stable conditions. So what? We happen to be agents, and we’re happy to be agents; why shouldn’t we be satisfied with modes of reasoning suited to creatures such as we happen and are happy to be? Enoch believes that an agent should mistrust reasons for being an agent that rely for their normative force on the constitution of agency itself. Without a more deeply anchored justification for being an agent, he thinks, an agent should mistrust any reasons whose grip on him depends on his being one. Enoch compares this mistrust of agent-­style reasons in support of agency to the epistemological skeptic’s mistrust of reasons delivered by our cognitive faculties in support of those faculties themselves.19 The epistemological skeptic is entitled to doubt the reliability of human cognition altogether, and Enoch claims the corresponding entitlement to doubt the reliability of human practical reasoning as functionally characterized by a theory such as mine. This analogy penetrates to the core of our disagreement. Epistemological skepticism presupposes an objective criterion of correctness for the output of cognition. What the skeptic doubts is whether our cognitive faculties can be relied on to deliver the truth. Enoch, a normative realist, believes that there is an analogous, objective criterion of correctness for the output of practical reason. If there is such a criterion, then of course any mode of practical reasoning characterized functionally may fail to meet it and is therefore subject to radical doubt. As a normative realist, then, Enoch is fully entitled to his skepticism about practical reasoning as the functionalist describes it, though he will go on using that mode of reasoning, just as he continues to use the faculties doubted by the epistemological skeptic. Such skepticism can no more undermine morality than epistemological skepticism can undermine science.

Notes 1. Bernard Williams, “Internal and External Reasons,” in Moral Luck: Philosophical Papers 1973–1980 (Cambridge: Cambridge University Press, 1981), 101–13; “Internal Reasons and the Obscurity of Blame,” reprinted in Making Sense of Humanity and Other Philosophical Papers (Cambridge: Cambridge University Press, 1995), 35–45. 2. Derek Parfit, On What Matters, 2 vols. (Oxford: Oxford University Press, 2011), ch. 1. 3. Guys and Dolls, Abe Burrows and Jo Swerling (Frank Music Corp., 1953): Big Julie: I’m rolling a thousand, and to change my luck, I will use my own dice. Nathan Detroit:  Your own dice!

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

92  J. David Velleman Big Julie:  I had them made for me especially in Chicago. Nathan Detroit:  They’re blank. Big Julie:  I had the spots taken off for luck. But I remember where the spots formerly were. 4. Many philosophers draw a distinction among reasons for acting between “normative” and “motivating” reasons. This distinction is better avoided. When someone acts for a reason, he is being influenced by that reason, and in that sense the reason is motivating him. But acting for a reason requires being influenced in a particular way. A reason justifies an action, or helps to justify it, and someone acts for a reason only insofar as it exerts its influence on him by justifying the action in his eyes. Sometimes an agent is influenced in this way by a consideration that is false, or by his belief in a fact that doesn’t genuinely justify his action. In that case, the reason for which he acts is said to be merely motivating but not normative, because it does not really justify. Fine. But distinguishing between normative and motivating reasons not only obscures the fact that reasons are normative in virtue of their justificatory force, which is at least potentially motivating; it also obscures the fact that even when motivating ­reasons lack that force, they motivate as if they had it. That is, a false or rationally irrelevant belief contains the reason for which someone acts only if he is influenced by it as he would be influenced by a rationally relevant, true belief. So normativity is in the wings, as it were, when we speak of reasons that are merely motivating, and motivation is in the wings when we speak of normative reasons. 5. Donald Davidson, “Actions, Reasons, and Causes,” in Essays on Actions and Events (Oxford: Oxford University Press, 1980), 3–21. 6. “Holding” is Davidson’s term. “Intending,” in Essays on Actions and Events, 97, n. 7. 7. See my “The Guise of the Good,” in The Possibility of Practical Reason, 2nd edn. (MPublishing, 2015). 8. I am not suggesting that the de dicto reading saves Williams’s argument; on the contrary, even under that reading, it was still fallacious. Williams seems to have assumed that the capacity of a consideration to motivate the agent was not just a necessary condition of its being a reason but a necessary pre-condition that must be prior in the orders of constitution and discovery. But the procedural nature of reasons does not require their motivational force to be antecedent to their status as reasons. They can satisfy the motivational condition in consequence of being reasons and being recognized as such. A conation toward acting with justification, even de dicto, can therefore satisfy the motivational condition. What such a conation cannot do is to provide practical reasoning with a criterion of correctness, which it requires in order to give application to the concept of justification. In order to justify an action, reasons must contribute to showing that it meets some standard, and the standard cannot simply be support by the strongest reasons, on pain of vacuity. For an elaboration of this argument, see my “The Possibility of Practical Reason,” in The Possibility of Practical Reason. 9. Nicomachean Ethics, VI.ii.5, 1139b: “Whoever makes something does so for some purpose, and what is made is not an end in itself but of and for something. But what is done [is an end], for eupraxia [is] the end, and desire is for that”; VI.iv.4–6, 1140b: “Making [poiēsis] has some further end, but not praxis; for eupraxia itself is the end.”

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Two Normativities  93 10. The point made in this paragraph is similar to that made by Christine Korsgaard in ­Self-­Constitution: Agency, Identity, and Integrity (Oxford: Oxford University Press), 31–2. 11. Nicomachean Ethics, II.iii.30, 1104b. 12. G. E. M. Anscombe, Intention (Cambridge, MA: Harvard University Press, 2000), §§37 ff. 13. For these explanations, see the papers in The Possibility of Practical Reason and Self to Self: Selected Essays, 2nd edn. (Ann Arbor: Michigan Publishing, 2020), chs 3, 11, and 12. 14. I believe the term was introduced by John McDowell, Mind and World (Cambridge, MA: Harvard University Press, 1996), index s.v. It is widely thought that Christine Korsgaard’s constitutivism is not naturalistic, because it says that agency is constituted by the Kantian Categorical Imperative. As I read Korsgaard, however, she thinks that agency is constituted by a commitment to the Categorial Imperative, which is a psychological state that can be studied naturalistically. 15. See, e.g., T.  M.  Scanlon, Being Realistic About Reasons (Oxford: Oxford University Press, 2016). 16. Actually, I can afford to concede the possibility that “reason” is a lexically directive term. For even if is lexically directive, it is what we call a “thick” normative term, because it has non-­normative descriptive content, unlike “ought.” And that non-­ normative content enables us to identify the constitutive aim in virtue of which it earns its normative force, by identifying the conation at the center of human agency. 17. Christine  M.  Korsgaard, “Realism and Constructivism in Twentieth-­Century Moral Philosophy,” Journal of Philosophical Research: Philosophy in America at the Turn of the Century 28 (2003): 122, n. 49. 18. David Enoch, “Agency, Shmagency: Why Normativity Won’t Come From What Is Constitutive of Action,” The Philosophical Review 115 (2006): 169–98. 19. David Enoch, “Shmagency Revisited,” in Michael Brady, ed., New Waves in Metaethics (Basingstoke: Palgrave Macmillan, 2011), 208–33.

Bibliography Anscombe, G. E. M. Intention. Cambridge, MA: Harvard University Press, 2000. Aristotle, Nicomachean Ethics. Translated by David Ross. Oxford World’s Classics. Oxford: Oxford University Press, 2009. Davidson, Donald. “Actions, Reasons, and Causes.” In Essays on Actions and Events, 3–21. Oxford: Oxford University Press, 1980. Davidson, Donald. “Intending.” In Essays on Actions and Events, 83–102. Oxford: Oxford University Press, 1980. Enoch, David. “Agency, Shmagency: Why Normativity Won’t Come From What Is Constitutive of Action.” The Philosophical Review 115 (2006): 169–98. Enoch, David. “Shmagency Revisited.” In Michael Brady, ed., New Waves in Metaethics, 208–33. Basingstoke: Palgrave Macmillan, 2011.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

94  J. David Velleman Korsgaard, Christine  M. “Realism and Constructivism in Twentieth-Century Moral Philosophy.” Journal of Philosophical Research: Philosophy in America at the Turn of the Century 28 (2003): 99–122. Korsgaard, Christine M. Self-Constitution: Agency, Identity, and Integrity. Oxford: Oxford University Press, 2009. McDowell, John. Mind and World. Cambridge, MA: Harvard University Press, 1996. Parfit, Derek. On What Matters, 2 vols. Oxford: Oxford University Press, 2011. Scanlon, T.  M. Being Realistic About Reasons. Oxford: Oxford University Press, 2016. Swerling, Jo and Abe Burrows. Guys and Dolls, 2nd edn. Frank Music Corp., 1953. Velleman, J.  David. The Possibility of Practical Reason, 2nd edn. Ann Arbor: Michigan Publishing, 2015. Velleman, J.  David. Self to Self; Selected Essays, 2nd edn. Ann Arbor: Michigan Publishing, 2020. Williams, Bernard. “Internal and External Reasons.” In Moral Luck: Philosophical Papers 1973–1980, 101–13. Cambridge: Cambridge University Press, 1981. Williams, Bernard. “Internal Reasons and the Obscurity of Blame” (1989). Reprinted in Making Sense of Humanity and Other Philosophical Papers, 35–45. Cambridge: Cambridge University Press, 1995.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

5 Self-­Consciousness and Self-­Division in Moral Psychology Richard Moran

The nature of first-­person thought first became of interest to me through the study of later work of Wittgenstein, where he takes our psychological concepts to be characterized by a basic asymmetry between the first-­person and the third-­person points of view. From here I became occupied with the idea of logical or conceptual limits to how much our thought about ourselves (our actions and attitudes) can be modeled on the corresponding thoughts we have about the lives of others. Self-­consciousness, in particular, seems to be a distinctive form of thought, reserved for thought about oneself as such, and not simply a kind of awareness that might in principle apply to oneself or to others. Versions of this idea are explored in philosophy of mind, in logic, and in philosophy of language, but it was Christine Korsgaard’s work that brought the first-­person point of view to the center of moral psychology and the understanding of practical normativity, and I have drawn inspiration from it ever since.1 As is well known, Korsgaard takes the human capacity for self-­ consciousness to be central to the fact that we do and must act for reasons and take ourselves to be subject to practical norms. Equally well known is that she conceives of this capacity in terms of ideas of distancing and “stepping back” from our inclinations and other states of mind. In this paper, I want to explore to what extent the metaphor of self-­division is necessary to the idea of self-­ consciousness and the role it plays in her account, and in particular whether the idea of self-­division is compatible with what I see as the fundamental differences between first-­person thought and thought about others. For our purposes we can think of self-­consciousness as the human capacity to think about oneself in a peculiarly immediate way, different from the way one thinks about other people or the objects in one’s environment, and typically expressed in the first person. I can think about myself in various guises, by name, by description, by perception (in a mirror or otherwise). And when Richard Moran, Self-­Consciousness and Self-­Division in Moral Psychology In: Normativity and Agency: Themes from the Philosophy of Christine M. Korsgaard. Edited by: Tamar Schapiro, Kyla Ebels-­Duggan, and Sharon Street, Oxford University Press. © Richard Moran 2022. DOI: 10.1093/oso/9780198843726.003.0005

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

96  Richard Moran I do think about myself in these different ways, it is possible for me to do so unknowingly, not realizing that the person I am thinking about is in fact myself.2 Oedipus is famously in this situation when he is seeking the killer of Laius his father, and before his moment of tragic realization. And when he does realize that he himself is that person, it seems that the form of his realization can only be expressed in the first person. “I am the killer of Laius. I myself have done it.” One thing this suggests is that the use of the first person is in place when there is no possibility of the speaker failing to realize the identity between himself as speaker or thinker and the person he is talking or thinking about. The use of the first person thus expresses a form of awareness, the awareness Oedipus comes to, and that he didn’t have before when he identified the killer of Laius (who is himself) in other different ways.3 The other ways of referring to himself don’t necessarily involve this awareness of the identity of thinker and object of thought. But it seems that use of the word “I” does. This awareness of the identity of thinker and object of thought seems built in to the logic of the first person. We can call this awareness “self-­consciousness,” to distinguish it from the other ways a person may be thinking of himself or talking about himself, but unknowingly. The self-­ consciousness that Oedipus arrives at has dramatic consequences for his subsequent thought and action; it provides him with a completely different set of reasons. And yet these very same features of first-­person thought have been found by some philosophers to cast doubt on the idea that thoughts expressed by the first person can be said to express knowledge or awareness of a particular person at all. Compare how Descartes arrives at his realization, the basis for his cogito, through a process of methodological doubt. That is, he takes the first-­ person thought “I seem to be sitting before the fire” to be available to him even when he doesn’t take himself to know his own name, or where he is, or whether he even has a body. He has not identified himself in any way when he says “I think, therefore I am.” And in a sense he doesn’t take himself to know anything about who this person might be, other than “the thinker of this very thought.” Given the apparent independence of this thought of any identifying information about its object, we might begin to wonder what sort of knowledge or awareness can be in question here, when it can co-­exist with total ignorance of the facts about the person it is supposed to be knowledge or awareness of. This feature of the first person applies not just to the cogito or to the realization of Oedipus, but to much ordinary first-­person thought. When I think or say “I have a headache” or “I see a tree outside” I do not come to this thought by first identifying a particular person and then saying of him that he

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Self-Consciousness and Self-Division  97 has a headache or sees a tree outdoors. I can have such thoughts without knowing anything about the particular person I am. So again we might worry about the supposed content of this thought. How is it a thought about a particular person at all? What is this thing called “self-­consciousness”? Another way to suggest a problem with the content of such thoughts is to recall that the first person is in place when the speaker could not fail to know the identity between himself and the person he is speaking about. In speaking in the first person it is impossible for me to be mistaken as to which person I am speaking about. When I say “I see a tree” or “I have a headache,” it could not be the case that I am right that someone sees a tree or has a headache but am mistaken in thinking that it is myself. It is sometimes said that first-­person statements of this sort are “immune to error through misidentification.”4 We might ask whether there is some confusion between logic and epistemology when we say that an error of identification is being “logically excluded” by certain uses of the first person. Or, granting that, we might question whether the putative reference to a particular person is really empty in such a context. If there is no logical possibility of getting the wrong person here, then there is no content to the idea of getting the right person.5 In which case we may lose grip on the idea of knowledge here, and with that the notion of self-­ consciousness as a form of awareness may be challenged as well. And yet, against all this skepticism, there remains the fact the Oedipus does come to a new and shattering awareness, and one that could not be expressed in any terms that don’t include the first person, with precisely the logical features we have just described.

1 In moral psychology, self-­consciousness is often associated with the capacity for critical reflection, the ability to appraise one’s own conduct, and to integrate that appraisal in one’s practical reason. While other creatures pursue their goals and act intelligently on the world around them, human beings have the ability to reflect on their goals themselves, to reflect on their own pursuits as such and hold them up to critical evaluation. Such reflection will be first-­personal in the sense described, since such reflection is not carried out under any names or descriptions by which one could fail to realize that the conduct being judged is one’s own, but is instead a form of reflection that is expressed using the word “I,” just as with Oedipus. It is sometimes argued that it is this capacity for self-­conscious reflection that makes our thought and

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

98  Richard Moran behavior subject to norms in the first place. And yet it is striking in this context how often the form of self-­consciousness in question is described in terms of self-­division, separation, and distance, when the notion we began with here defines itself by ruling out various possibilities of distance or misrecognition. The language in the following passages from Arthur Lovejoy and Adam Smith is typical in this regard. For by virtue of his self-­consciousness, man is divided within and against himself . . . . He has a complex mass of instincts, appetites, drives, desires, potentialities, feelings, emotions, attitudes, which make up what, at any given moment, he primarily is; but he is also capable, at least at other moments, of taking the attitude of external observer towards all of these, of looker-­on at the very process of his own experience, and at himself as the experiencer or the doer of it all. He is in short, both actor and spectator, both performer and commentator on or critic of the performance.6 When I endeavor to examine my own conduct, when I endeavor to pass sentence upon it, and either to approve or condemn it, it is evident that, in all such cases, I divide myself, as it were, into two persons; and that I, the examiner and judge, represent a different character from that other I, the person whose conduct is examined into and judged of.  The first is the spectator, whose sentiments with regard to my own conduct I endeavor to enter into, by placing myself in his situation, and by considering how it would appear to me, when seen from that particular point of view. The second is the agent, the person whom I properly call myself, and of whose conduct, under the character of a spectator, I was endeavoring to form some opinion. The first is the judge; the second the person judged of.  But that the judge should, in every respect, be the same with the person judged of, is as impossible, as that the cause should, in every respect, be the same with the effect.7

In Lovejoy and Smith we have two representative expressions of human deliberation or critical reflection as requiring a picture of self-­division, and as involving a relationship between two distinct “selves.” The passage from Lovejoy is explicit in attributing this self-­division to the human capacity for self-­consciousness, and the idea of self-­consciousness invoked here immediately suggests the duality of “actor and spectator.” The special human capacity for critical reflection is identified with the human capacity for self-­ consciousness, and the form of consciousness in question is understood as a kind of observer’s awareness, “the looker on at the very process of his own experience” as Lovejoy puts it. In both cases the duality introduced here is

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Self-Consciousness and Self-Division  99 described in terms of the contrast between the person as agent or actor, and the person as observer or spectator of that agent, and both writers find it natural to describe the situation in terms of the interaction of distinct “selves” or “persons.” This talk of distinct “selves” is acknowledged by Smith to be figurative in some sense, when he says that “I divide myself, as it were, into two persons,” but the duality introduced here is still meant to be a central, and perhaps ineliminable, feature of the phenomenon being described. The very capacity to self-­consciously pass judgment on one’s own conduct is thought to require a form of self-­division, a distinction of roles, if not something s­ tronger, as in Lovejoy’s depiction of the self-­conscious person as “divided within and against himself.” Here self-­consciousness is pictured not only as involving a distinction of roles but as essentially a form of self-conflict.8 What sort of division between the person judging and the person judged is thought to be necessary for self-­assessment or critical reflection? By the terms of the case itself, this is not supposed to be the same sort of division there is between the judge and the person judged when one of these people is standing in the dock and the other person is sitting on the judge’s bench. Rather, we are meant to be discussing the case of one person judging or appraising himself. So while we may speak of the person as judge and also as person judged, at the end of the day we mean to be describing the thinking of a single person. And importantly, we mean to be describing the thinking of a person who recognizes in his very thought that he is identical with the person he is judging. That is, Smith is not considering a different possible case, one in which I see someone in a mirror and think to myself “What a shabby pedagogue” (as in the anecdote from Ernst Mach).9 In that sort of situation, I am well and truly passing judgment on my own appearance, but I do not recognize myself as identical with the person I see in the mirror. Rather, Smith means to be describing the situation of a person who is self-­consciously appraising his own conduct as such. So this qualification raises our initial question in a more pressing form: If the form of thought Smith means to be describing is not only in fact a thought about the person who is himself, but is also self-­ consciously conceived of by the person as thought about himself, then what is the meaning of the insistence that, in some other sense, we are dealing with different “selves” here? What is the need to invoke self-­division of any kind here? It is not as though other ordinary reflexive activities encourage us to invoke self-­division. When I scratch myself or (better) when I defend myself, there is no temptation to think that the description of such a situation requires a division between the self as scratcher and the self as scratched, or a distinction between the person defending and the person being defended. And, as

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

100  Richard Moran just mentioned, in the case of reflexive relations in thought, such as self-­ appraisal in Smith’s case, there would seem additional reason to avoid talk of division because, unlike the case of scratching myself (which might be done under circumstances of mistaken identity), self-­conscious attitudes involve the recognition of the identity of thinker and object of thought. In The Sources of Normativity and in later work Christine Korsgaard understands self-­consciousness as playing an important role in understanding (moral) normativity, and our authority over ourselves. She understands the relevant notion of self-­consciousness in terms of “reflective distance” and metaphors of “stepping back” from an inclination or perception, a shift from attention directed outward to attention directed within. The role this notion plays in Korsgaard’s moral psychology has led critics like Pamela Hieronymi to see the invocation of self-­consciousness on the model of the requirement of awareness for something that one seeks to control (and hence a bad model for ordinary agency). I believe that Hieronymi’s criticism does not capture the sense of self-­consciousness that is really Korsgaard’s concern (and that of the Kantian tradition she is drawing on), but that seeing this will require some revision of Korsgaard’s own stated view. For I will argue that the relevant notion of self-­consciousness, the one that is expressed by thoughts using “I,” should not be understood in terms of an assertion of distance or difference between thought and the object of thought. It should not be seen as related to the requirement of perceptual awareness for control of objects in one’s environment. A notion of “self-­consciousness” of that form would not distinguish between the awareness one may have of other people and the particular form of self-­consciousness of “I” thoughts, and hence could not ground any particular normativity attached to one’s own thought and action. Further, once the relevant notion of self-­consciousness has been understood in terms of “reflective distance” it is inevitable that we begin the story of practical reason with a picture of self-­division, a division which will in turn have to be overcome somehow when an inclination of mine (something within me, different and distinct from the person reflecting on the problem) becomes something that I will. But the self-­consciousness expressed in “I will” cannot be of the same sort with which we began, cannot be an assertion of reflective distance, but must on the contrary be an expression of unity, of self-­identity, a form of awareness not related to the awareness of objects either within or without. The same notion of self-­consciousness in terms of reflective distance and self-­ division cannot be employed to unite the self, to assert identity and not difference. In rejecting this notion, I do not mean to reject the connection between the capacity for self-­consciousness and the capacity for self-­criticism and

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Self-Consciousness and Self-Division  101 correction, but I argue that this will have to be explained in terms of the non-­object-­oriented notion of self-­consciousness expressed in “I” thoughts. It will also need to be explained in what way “I” thoughts express a form of self-­ awareness, while not being transitive or object-­directed. Here is a much-­cited expression of Korsgaard’s view of the human capacity for self-­consciousness and the difference it makes in our lives. But the human mind is self-­conscious in the sense that it is essentially reflective. I’m not talking about being thoughtful, which of course is an individual property, but about the structure of our minds that makes thoughtfulness possible. A lower animal’s attention is fixed on the world. Its perceptions are its beliefs and its desires are its will. It is engaged in conscious activities, but it is not conscious of them. That is, they are not the objects of its attention. But we human animals turn our attention on to our perceptions and desires themselves, on to our own mental activities, and we are conscious of them. That is why we can think about them. And this sets us a problem no other animal has. It is the problem of the normative. For our capacity to turn our attention on to our own mental activities is also a capacity to distance ourselves from them, and to call them into question. I perceive, and I find myself with a powerful impulse to believe. But I back up and bring that impulse into view and then I have a certain distance. Now the impulse doesn’t dominate me and now I have a problem. Shall I believe? Is this perception really a reason to believe? I desire and I find myself with a powerful impulse to act. But I back up and bring that impulse into view and then I have a certain distance. Now the impulse doesn’t dominate me and now I have a problem. Shall I act? Is this desire really a reason to act? The reflective mind cannot settle for perception and desire, not just as such. It needs a reason. Otherwise, at least as long as it reflects, it cannot commit itself or go forward.10

This picture of the capacity for self-­consciousness that distinguishes human life from that of other creatures conceives of it as a turn from attention directed outward to attention directed inward, and sees it as a capacity to separate or distance ourselves from what we discover within, to give or withhold endorsement from the impulse or perception. It is the difference between apprehending an inclination of mine, which I hold at a distance and evaluate, and the assertion of what I will to do, where the “I” expressed here is essentially not another element within me I might distance myself from but rather the very self that I am, the unity beyond the manifold of internal psychological elements.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

102  Richard Moran The picture of self-­consciousness here is distinctive in a number of ways. The human capacity for critical reflection on one’s own actions is equated with the ability to take one’s own mental states (“our perceptions and desires themselves”) as objects of thought. Elsewhere this is described in terms of “mov[ing] th[e] attributive operation indoors.”11 The contrast is between a creature whose attention is focused outward, on the world of attractions and obstacles, and a creature which also attends to its perceptions and desires themselves, and calls them into question before it can act.12 In addition, the human capacity for reflection is described in terms of separation and distance from its object. What is reflected upon is conceived of as distinct from or separate from the consciousness that is engaged in reflection, the consciousness that has called a certain impulse into question and is deciding whether or not it shall count as a reason for action. I draw attention to these two features because they frame so much of the subsequent discussion and it can seem that any invocation of self-­consciousness as part of human practical reason is committed to both the move from a gaze directed outward to a gaze directed inward, and the idea of distance or separation between consciousness and the object of consciousness. It is these assumptions that I mean to question, along with the picture of self-­division that we see in Lovejoy and Smith. To anticipate, we should distinguish two different invocations of the “first-­person perspective.” On a Cartesian view, the first-­person perspective is an internal view of my psychological states purely as states of my consciousness. In the context of methodological doubt, I bracket any claims about, for example, the fire before me and restrict myself to claims about the appearance of a fire before me, a qualification of my consciousness rather than a claim about what is actually before me. This is a different thought from the idea of the primacy of agent’s perspective in thinking about normativity. For Korsgaard this practical orientation is methodologically fundamental for philosophical thinking about the demands of morality. The normative question is a first-­person question that arises for the moral agent who must actually do what morality says. When you want to know what a philosopher’s theory of normativity is, you must place yourself in the position of an agent on whom morality is making a difficult claim. You then ask the philosopher: must I really do this? Why must I do it?13

The first-­person perspective of deliberation is that of the agent determining what to do, in contrast to a third-­person perspective that might be concerned with what some agent will do, or concerned with what might motivate an

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Self-Consciousness and Self-Division  103 agent to do something, or with explaining what moved an agent to do what he did. The questions that arise from that agent’s perspective are not answered by a prediction or an explanation of the agent’s behavior, but rather by a decision or a commitment of that agent. When we lay it out like this, we can see that it remains an open question whether the first-­person character of the practical agent’s perspective involves anything like a Cartesian perspective on one’s mental states as such. The contrast between the agent’s perspective of deliberation and an external perspective of prediction or explanation is a different contrast from that between attention directed outward and attention directed inward. So far anyway, it seems that we can characterize what is first-­personal in the agent’s point of view in terms of a distinctive set of practical considerations directed upon the world with its opportunities and obstacles. We will consider later whether there are considerations that force us to describe the deliberative agent as adopting something like the inward-­looking Cartesian perspective on himself as well. The other distinctive feature of this picture of self-­consciousness is the metaphor of distance or separation. On this view, when I turn my attention to my perceptions and desires themselves as mental states, then I hold them at a certain distance, as I do other objects of my attention. I “back up” from them and bring them into view, and in this way I can relate to them as “other” to myself, the consciousness which has backed up and brought them into view. This consciousness of separation from my desires and perceptions means that I can now hold open the question of how I shall relate myself to them, specifically whether they should count for me as good reasons for action or belief. As with Lovejoy and Smith, self-­consciousness in the life of a deliberative agent is characterized in terms of a kind of self-­division, if not self-­conflict. Here the picture of self-­division is seen to be necessary to capture the sense that the practical agent is not simply driven by impulses and appearances but has some say in what is to count as a good reason for action or belief. As the first-­person perspective of the deliberative agent is in principle distinct from a Cartesian “inward glance,” so this perspective is in principle distinct from any form of self-­division, and it remains to be seen what consideration may require us to characterize the agent’s point of view in terms of any kind of internal division or relating to one’s inclinations as “other” to oneself. This characterization of self-­consciousness is intuitive in its own way and is shared by both philosophers in Korsgaard’s Kantian framework and by philosophers who take these features of self-­consciousness to be reasons to doubt that it is really an aspect of the agent’s deliberative perspective at all or plays a role in characterizing what is distinctive in human agency. In a series of papers,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

104  Richard Moran Pamela Hieronymi takes on board the characterization of self-­consciousness in terms of the ideas of separation and attention redirected inward, as an occasion to question why self-­consciousness, understood in the first sense of reflection on one’s internal states, should be thought to have any special relation to freedom, responsibility, and being subject to reasons. In her paper “Reflection and Responsibility” she writes: A common line of thought claims that we are responsible for ourselves and our actions, while less sophisticated creatures are not, because we are, and they are not, self-­aware. Our self-­awareness is thought to provide us with a kind of control over ourselves that they lack: we can reflect upon ourselves, upon our thoughts and actions, and so ensure that they are as we would have them to be. Thus, our capacity for reflection provides us with the control over ourselves that grounds our responsibility. I will argue that this thought is subtly, but badly, confused. It uses, as its model for the control that grounds our responsibility, the kind of control we exercise over ordinary objects and over our own voluntary actions: we represent to ourselves what to do or how to change things, and then we bring about that which we represent. But, I argue, we cannot use this model to explain our responsibility for ourselves and our actions: if there is a question about why or how we are responsible for ourselves and our actions, it cannot be answered by appeal to a sophisticated, self-­directed action. There must be some more fundamental account of how or why we are responsible. I will replace the usual account with a novel but natural view: responsible mental activity can be modeled not as an ordinary action, but as the settling of a question. This shift will require abandoning the tempting but troublesome thought that responsible activity involves discretion and awareness—which, I argue, we must abandon in any case.14

If self-­consciousness is seen as the awareness of inner states, and if the normative point of invoking self-­consciousness were to secure the conditions for responsibility and the demand for reasons, then it may well seem as though this required a severely alienated picture of the ordinary responsibility a person has with respect to his own actions and attitudes. On the model Hieronymi criticizes, we are responsible for our actions and attitudes, and for the reasons we act upon, because we exercise a kind of “managerial control” with respect to them, and in order for us to exercise such control we must have a particular kind of awareness of them, just as we do with respect to the objects in our environment.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Self-Consciousness and Self-Division  105 While this criticism has a certain force, I doubt that Korsgaard and others in this Kantian tradition would accept the diagnosis that Hieronymi offers. They would deny the analogy between rational control of one’s inner states and the managerial control of objects in one’s environment. And they would deny that their invocation of “self-­consciousness” commits them to the alienated picture of rational agency that she sees as stemming from this. For one, in the passage quoted above, Korsgaard insists that self-­consciousness is fundamental to our authority with respect to our actions and attitudes and not to their control, and Hieronymi’s diagnosis does not directly address this distinction.15 This does not remove the force of the criticism, however, since it will indeed be difficult to re-­constitute the notion of “authority” over oneself as distinct from “control” over one’s states once we begin with the picture of self-­ division (as we will see later). Hieronymi’s criticism is one that Korsgaard’s account leaves itself open to, insofar as it is committed to seeing the relevant self-­consciousness in terms of distance, division, and the metaphor of “stepping back.” Hieronymi herself appears to share this picture of the nature of self-­consciousness as separation but draws from this the conclusion that self-­ consciousness cannot then play anything like the role in rational agency and responsibility that it has in the Kantian picture. I mean to question this shared assumption and see what remains of the story of self-­consciousness and ra­tional agency once the relevant notion of self-­consciousness has been distinguished from any form of transitive or object-­consciousness and purged of the ideas of separation and self-­division. Hieronymi’s criticism focuses on the parallel between awareness of objects in one’s environment and awareness of one’s internal states, and sheds im­por­ tant light on the essentially different forms of agency one exercises in manipulating objects and the agency one exercises in making up one’s mind, endorsing a plan, or settling a question. When I act on something other than myself I will typically need to have it in view in order to monitor the progress of my affecting it. I can act on myself as an other as well, when I scratch myself or when a doctor treats his own injury.16 In cases like these the identity between the agent and patient is accidental, and the treatment or the scratching are actions which the person could perform on himself or on another, and which another could perform on her.17 In cases like these Hieronymi notes that a person can perform the action for any number of different reasons. I could scratch myself to relieve an itch or simply to demonstrate that I can do so or because someone asked me to. I have here what she calls “discretionary control” over my action and the reasons for it. As she points out, one does not typically have such discretion with respect to determining what to believe in a given situation or whether to endorse some desire as a reason for action. Here

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

106  Richard Moran reasons of a certain kind are determinative of the kind of question to be settled. The attempt to form the intention to X for reasons disconnected with doing X could only succeed in the formation of a different intention than the intention to X, something like the intention to produce a certain state in oneself.18 In the ordinary case of forming an intention, or settling one’s belief, or endorsing a desire as reason for action, we are active but we are not acting on ourselves.19 In effect, Hieronymi asks: If “self-­consciousness” is a form of awareness like the awareness of objects in one’s environment, only directed “inwards” at one’s own mental states, then what could be the importance self-­ consciousness to rational agency unless we saw it as an aid to the manipulative control of our inner states, with all the paradoxes that entails?20 What I hope to suggest, however, is that just as forming an intention or endorsing a desire is a way of being active but not a form of “acting upon oneself,” so the relevant form of self-­consciousness is a form of awareness but not analogous to awareness of objects, either inner or outer. In self-­consciousness the identity between the thinker and what is thought about is essential and not accidental, and the form of awareness does not take a dyadic form.

2 On Korsgaard’s account the human capacity for self-­consciousness sets us a problem that the other creatures do not have. Our impulses do not translate directly into action but are first set at a distance from us and must pass the test of reflection before we can accept them as reasons for action. But it is one and the same capacity which sets the human problem and within which the practical solution is found. Self-­consciousness is what divides us from ourselves, but it is in the exercise of that same capacity that we can put the pieces back together again and achieve a kind of unity. As she puts it in Sources, “If the problem springs from reflection then the solution must do so as well. If the problem is that our perceptions and desires might not withstand reflective scrutiny, then the solution is that they might.”21 In a particularly searching discussion, William Bristow raises a question about how one and the same capacity of self-­consciousness can play both these roles: On the one hand, self-­consciousness is consciousness of a mental state (e.g., a desire) as one’s own: “[W]e human animals [in contrast to other animals] turn our attention on to our own perceptions and desires themselves, on to our mental activities, and we are conscious of them” (Sources 93).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Self-Consciousness and Self-Division  107 This is object-­directed consciousness that has as its object some particular mental content of the subject of consciousness. On the other hand, practical self-­consciousness is consciousness of that “which is you, and which chooses which desire to act on” (Sources 100). It is essential that this second sort of self-­consciousness not be understood on the model of the first, because, whereas the first exactly implies the practical problem (the problem of whether this psychological content on which I reflect is a sufficient practical reason), the second is supposed to provide the resources for solving the practical problem. The first sort of self-­consciousness implies a reflective distance between the reflecting subject and the object of self-­reflection (my desire, say). The practical problem implied by this self-­ consciousness, according to Korsgaard, is whether to identify oneself with the object of reflection or not. The second sort of self-­consciousness can be the ground for self-­determination, thus solving the problem, only if there is no reflective distance in this self-­consciousness. Thus the self-­consciousness the principle of which serves as the ultimate norm cannot have the same structure as other-­directed consciousness; it cannot be consciousness of another, where the other is oneself. It must be immediate, unreflective self-­consciousness.22

Here Bristow brings out a fundamental tension within the conception of self-­ consciousness at issue. On the one hand, it is the name for a form of reflective distance from what it is directed upon. In what Bristow calls “object-­directed consciousness” I relate to some aspect of myself or some element of my inner life as other, as something I assert my separation and distance from. But of course, in this very stepping back, in this very assertion of my separation from some part of “me,” I affirm who I essentially am as independent of that psychological element. And the “I” who speaks here is essentially not an object to me of any kind, essentially not something (or someone) that I do or even could assert my separation from. This is the “self ” about which Korsgaard says: “When you deliberate, it is as if there were something over and above all of your desires, something which is you, and which chooses which desire to act on. This means that the principle or law by which you determine your actions is one that you regard as being expressive of yourself.”23 The assertion of “I,” when I raise the question for myself of whether to endorse some desire or not is itself a form of self-­consciousness, is the very capacity for reflection itself, the source of authority for my will as distinct from any candidate inclinations of mine. This consciousness cannot be described in terms of the kind of separation that characterizes consciousness of something other than oneself, not even when I relate to some inclination of my own as alien to me.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

108  Richard Moran Rather, if the “I” of reflection is to represent itself as solving the practical problem, as speaking with the authority of the person as such, it must express a form of thought that affirms identity and not separation, otherwise the practical problem remains open, the division not overcome. As Bristow puts it, this self-­consciousness “cannot have the same structure as other-­directed consciousness; it cannot be consciousness of another, where the other is oneself. It must be immediate, unreflective self-­consciousness.” Bristow writes of “two sorts of self-­consciousness,” one of which he calls “object-­directed consciousness” of a particular mental content. There are two assumptions contained here that I mean to question. One is whether the first-­ person perspective of practical reason should be characterized in terms of attention directed toward a mental content in the first place. As characterized earlier, the standpoint of the practical agent is in the first instance contrasted with an external, theoretical, or explanatory perspective, and we have not seen a reason yet to identify that with the contrast between attention directed outward and attention directed inward. The metaphor of “stepping back” does not itself require this. As I will describe later, the practical agent has the ability to step back from and call in to question his projects and his situation themselves, and this is the expression of a genuinely first-­person capacity, but not one that involves the re-­direction of attention from outward to inward. Critical reflection can be a form of self-­consciousness without being directed toward one’s internal states. Both Bristow and Korsgaard begin by describing the position of the practical agent in terms of reflection on states of mind (impulses, inclinations) and how they are to count in one’s proposed action, but the more natural starting point of deliberation is not reflection on one’s motivations as such but reflection on the situation before one, one’s goals, opportunities, and obstacles. We can capture what is right in the metaphor of “stepping back,” and what is first-­personal in the thinking of the practical agent, without departing from this focus. And secondly, when we do mean to describe the form of awareness a person has of what he thinks, feels, or wants, we can resist seeing this awareness as akin to “object-­directed consciousness,” where that is understood to mean something like “consciousness of another, where the other is oneself,” in Bristow’s terms. This second assumption would mean that the ordinary self-­ consciousness of what one thinks, feels, or wants is per se a form of alienated consciousness, a form of relating to one’s mental states like the mental states of another (even if that “other” is somehow “within” me). But the ordinary immediate awareness a person has of what he is doing at the moment, or what he wants, is not in the mode of separation or distance. And more importantly,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Self-Consciousness and Self-Division  109 even the cases of reflection which Korsgaard and Bristow begin with, where the person is indeed reflecting on an inclination or impulse as such, holding it at a distance and determining whether to act on it, these forms of reflection are not to be modeled on the consciousness of another, even if that “other” is within oneself. Even in the case of a thoroughly alienated impulse, the person’s relation to it is “immediate” and first-­personal in the sense described at the beginning of the paper. When Sartre’s gambler is alienated from his temptation to return to the gaming tables, he recognizes this as his own desire and no one else’s.24 This is the apprehension of a desire which he may disown, but it is not a judgment made possible through the identification of a particular person and the attribution of a desire to him. (Sartre’s gambler might be amnesiac at his moment of weakness and self-­struggle, with no information about the particular person he is.) Even the apprehension of one’s desires and impulses in these familiar alienated ways is immediate, and of a different logical form from subject-­object consciousness, including the consciousness of an other. Now of course it is important to Korsgaard’s account25 that there is room for a question of the form “But is this inclination really ‘me’ or not?” or “Do I ‘identify’ with this impulse (of mine) or not?” Here the temptation will be strong to reify different “selves,” as though there could be different “selves” within one and the same person. This temptation should be resisted not only because of its dubious coherence but because postulating another “self ” here only serves to obscure what is distinctively first-­personal in this form of consciousness and in the agent’s relation to the alienated or rejected inclination. When I confront an impulse like the desire to return to the gaming tables and “disown” it, I withhold my endorsement from it as a reason to act (even though, of course, it may in fact get the better of me). In distancing myself and seeking to assert my authority with respect to it, there is still no question for me whether this impulse belongs to me or to someone else. There is no other person in question, and as mentioned, my original apprehension of the impulse was not made on the identification of a particular person or “self ” in the first place. Despite Bristow’s deliberately paradoxical formulation (“consciousness of another, where the other is myself ”) my distancing myself from a rejected desire does not mean that I relate to it as the desire of another person in any sense. For that to be possible I would need to deploy some way of determining which person, among possible people, I am thinking about. I would have to be able to entertain thoughts of the form “This inclination or desire, does it belong to RM or to some other person, LW perhaps?” “Failure or refusal to identify” with a desire is still an essentially first-­personal relation to it, and not akin to consciousness of an other, even an internal

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

110  Richard Moran other. When I reject the desire and withhold endorsement from it, the desire loses its authority over my action, but it does not thereby belong to anyone else. It is either mine or no one’s (as when the loss of authority means the loss of its actual power, and perhaps its going out of existence altogether, hence no one’s desire). The very gesture of withdrawing authority from the rejected desire only makes sense as a relation to a desire recognized as one’s own. One does not relate to the desires of other people in this way. One can oppose such desires or try to dissuade the other person from acting on them or giving them any importance, but one does not withhold authority from the other person’s desire. With respect to some other person’s impulse or desire I am not in a position to delegitimize or to “silence” it in their practical reason.26 The metaphors of “separation” and “distance” are misleading in this domain for the sense of “separation” from a desire of mine I reject or am alienated from is not the same sense of “separation” or “distance” I have with respect to the desires of other people. In the one case I relate to the desire as one I exert a certain authority over, in that I am the one who is responsible for integrating my desires in my practical reason and my action. In the other case, I relate to the desire as one that I may approve or disapprove of, and I may act on that approval or disapproval, but I do not have authority over the integration of that desire in the other person’s practical reasoning. And in any case, insofar as I am relating to the desires of a genuine other person, I would have to identify a particular person, distinguishing them from among others, to ascribe some desire to them. I do not do so in my own case. Even in the case of an alienated desire, such as with Sartre’s gambler, my apprehension of this (rejected) desire is by the “outward directed” attention to the attractions of the gaming tables, not by identifying a particular person and thinking about that person’s desire. In rejecting his desire to gamble, Sartre’s character is thinking about himself, and self-­consciously so, but not in virtue of identifying a particular person (a genuine other), but in virtue of being the person who vividly apprehends the attractions of the gaming tables, and who equally vividly recalls the disastrous consequences. His awareness of his desire is the purely reflexive awareness expressed in “I” thoughts, where the thinker could not fail to be aware of the identity between the thinker and the object of thought. The invocation of another “self ” at this point would merely raise the question of how this person is in a position to say “I” here, to speak for this “self ” that is “within him” but separate from him. What could give him that authority? And if he couldn’t say “I” here, with respect to this rejected desire, then he couldn’t “step back” from it either; he could not raise the question of endorsing it or not. Here is another way in which the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Self-Consciousness and Self-Division  111 metaphors of distance and separation are misleading, for I cannot open up the space of questioning whether to identify with some desire or not except insofar as I recognize it as my own desire and no one else’s. Mine are the only desires with respect to which I stand in a relation of possible endorsement or delegitimation. The very gesture of “stepping back” and calling into question precludes the kind of separation or distance there is between me and the desires or mental states of another person.

3 Bristow refers to a “pure” or “formal” conception of self-­consciousness in Kant. In the First Critique Kant famously says, “The ‘I think’ must be able to accompany all my representations, for otherwise something would be represented in me that could not be thought at all, which is as much as to say that the representation would either be impossible or at least would be nothing for me.”27 For our purposes here let’s understand the meaning of adding “I think” to a representation of mine as the ability to convert a thought of the form “There’s a hawk” to a thought of the form “I see a hawk,” or to convert a thought of the form “Snow is white” to a thought of the form “I believe that snow is white.” On this interpretation, how should we understand Kant’s thought that I “must” “be able” to make such a transition in thought? Much will depend on how we understand the relation between the “necessity” invoked and the “possibility” in question. It is not necessary that I in fact add “I think” to any representation of mine, but it is necessary that this be possible for me, and possible for any “representation” of mine. Whatever necessity this may be, it does not apply to the possibility of adding to a representation of mine the thought that “RM thinks . . . ” or “the son of WTM thinks . . ., ” or “the man there in the mirror thinks . . ..” That is, it is essential to whatever necessity is invoked here that it express itself in the first person, and is not a judgment based on the identification of any particular person at all. The Kantian “I think” is available to me when I have no idea which person I am.28 And the fact that in such thought I do not identify myself in any way, or distinguish myself from any other thinker, is surely fundamental to the thought that this form of self-­consciousness is not a form of “object consciousness” at all, and thus is incompatible with any picture of self-­consciousness as separation or “making other.” This same “immediacy” and independence of any identifying information in such judgments is the reason for the peculiar “emptiness” of the first person, the sense that the very reason for the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

112  Richard Moran “necessary availability” of “I think” for any representation is because the “I think” adds no additional content to the original thought about the hawk or the snow. Self-­consciousness is an expression of identity and not difference, and is “always available” just because it is in an important sense not a transition to a logically distinct thought content at all (for such a transition would require some basis for inference, logical or empirical, and that is out of the question here). One conclusion to draw would then be a deflationary one, that this very idea of “self-­consciousness” is in fact empty, and a fortiori has no role to play in accounting for what is distinctive about human rational agency, but I don’t think that conclusion would be justified. From the fact that there is a sense in which the “I think” adds no additional content to my thought, it does not follow that the capacity for self-­conscious thought of the sort expressed in “I”-thoughts does not make a decisive difference in human rational agency. Nor does it follow that the “formal” character of the self-­consciousness expressed in “I”-thoughts is in conflict with the idea that, on an individual level, the person changes his thinking when he goes from thinking about the “shabby pedagogue” in the mirror to thinking about himself as such, and this change makes a decisive difference in his subsequent thought and action. However logically “formal” or “empty” we understand the first person to be, we have at the same time to describe the actual difference in thinking, the difference in reasons made available to a person when he comes to self-­consciousness. Let’s look a bit more closely at Kant’s “I think” as an expression of self-­consciousness. I am out in the countryside and see something in the sky and exclaim: (a)  “There’s a hawk.” Here there is nothing in my consciousness but the hawk.29 My awareness is totally absorbed in it. But in virtue of (a) it is also open to me to say (or think): (b)  “I see a hawk.” I have this thought “available” to me just by “reflecting,” or in being prompted by another person.30 The thought expressed in (b) is somehow already contained in the original thought about the hawk. But if all that is presented in experience is the hawk itself, one might question the apparent reference to a person in (b). One might conclude that this appearance is illusory,31 or conclude that, contrary to our understanding of Kant’s dictum, any transition

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Self-Consciousness and Self-Division  113 from (a) to (b) must be illegitimate. One could insist, that is, that despite appearances, the thought expressed in “I see a hawk” is as fully “impersonal” as the original statement “There’s a hawk.” This does not seem satisfying, however. For one, the thought expressed in (b) “I see a hawk” is certainly different from another possible thought, not about oneself: (c)  “A hawk is being seen by someone.” In contrast, the person who says (b) “I see a hawk” is not being non-­committal as to who is seeing the hawk (even while, of course, that person is not deploying any identifying information about himself). He is saying that he, the person speaking, is the person who sees the hawk. In addition, as we’ve been noting throughout, the first-­person thought (b) “I see a hawk” expresses a peculiarly immediate form of self-­consciousness and we need to be able to describe the change in mind such a person undergoes when he engages in first-­personal thought. John Perry’s shopper comes to an embarrassing realization about himself, and Oedipus comes to a shattering realization about himself, when they first come to think in the first person about what they have done.32 This new form of awareness makes for important differences in their further thought and action. Less dramatically, the person looking with absorption at the hawk comes to a new form of thought when his companion asks him why he suddenly stopped on the path and he replies, “I see a hawk.” If the change made by self-­consciousness is real, then what of the “availability” of Kant’s “I think,” the fact that the person looking at a hawk has available to him the thought “I see a hawk,” even though there is nothing in his field of experience but the hawk itself? Let’s contrast the transition from (a)  “There’s a hawk” to (b)  “I see a hawk” with the transition from (a)  “There’s a hawk” to (d)  “The man over there sees a hawk” or (e)  “RM sees a hawk.” These latter two thoughts are not immediately available to someone thinking the thought in (a). Any transition from (a) to a thought like one of these would involve a change of topic, with the logical and epistemic burdens that

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

114  Richard Moran would imply. But while the transition from (a) to (b) represents a change of consciousness in the person it is not a mere change from one topic to a different one, like the transition from a thought about that man over there to a thought about RM over here. Thoughts of this form [(d) and (e)] are thoughts with places for two distinctly conceived objects, RM (or the man over there) and the hawk that he sees. By contrast, the thought in (b) “I see a hawk” has a place for only one distinctly conceived object, the hawk, and that is the same object that is the only object singled out in the first statement (a) “There’s a hawk.” That is one way in which the two original thoughts in (a) and (b) are more internally related to each other than either of them is to the latter two (d) and (e). The fact that there is only one and the same object being singled out in both (a) and (b) (i.e., the hawk) is a reason in favor of saying that they are both modes of thinking about the same thing, that (b) can be characterized as self-­consciously seeing a hawk. In the person’s contemplation of the hawk, he has such thoughts available to him as that the hawk is before him, above him, and that it is by seeing it rather than by some other means that he is aware of its presence. Here again, however, the “him” in question is first-­ personal and reflexive. The person need not have any way of describing or identifying the person in front of whom there is a hawk, but his perception of the hawk is nonetheless situated in these ways and he has available to him the first-­person thoughts which register this situatedness. The formal difference between this type of thought and thoughts about another person’s sighting of a hawk, which involves the specification of two independently identified objects in relation to each other, is crucial for understanding the “availability” of the Kantian “I think,” the thought that it “must be possible” for it to accompany any of my representations of the world. And this formal difference between thoughts like (b) and thoughts about some other person tells us something about the “internal relation” to each other of (a) “There’s a hawk” and (b) “I see a hawk.” What I have in mind is that we cannot understand a statement (or a thought) of the form “There’s a hawk” without taking there to be answers to questions like, for example, “Where do you see it?”, and taking it that the speaker or thinker of (a) will be in a special position to answer such questions, without observing himself or distinguishing himself from among others. It would be difficult to make sense of a thought or statement like “There’s a hawk” where the thinker/speaker himself does not take what is claimed in that thought itself to be committed to answers to questions about whether the hawk was seen or heard, in what part of his field of vision, and so on. If the thinker/speaker rejected such questions by mis-­invoking Wittgenstein or Sartre and saying “My consciousness contains nothing but the hawk itself,”

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Self-Consciousness and Self-Division  115 I maintain we could not understand his original statement. At the same time, mastery of the first person is coordinated with the corresponding third-­ person forms. If the thought that RM sees a hawk is a thought that is available to other people (as it surely is), and if I am RM and thus a creature capable of reflection and the “I”-thoughts to which it gives expression, then I must be capable, by reflection, of thinking/saying “I see a hawk,” and mean by that to be talking about the same thing, the same person, that the other people are speaking/thinking about when they say “RM sees a hawk.”

4 The question of how a person’s actions or perceptions can be available to thought in first-­person reflection, even while the person does not thereby “posit” or empirically identify himself in any way, is thematic in Jean-­Paul Sartre’s early reflections on consciousness. He seeks an account of the form of thought by which someone engaged in action or absorbed in looking at something nonetheless has “available” to him a thought expressed in the first-­person. When I run after a tram, when I look at the time, when I become absorbed in the contemplation of a portrait, there is no I. There is a consciousness of the tram-­needing-­to-­be-­caught, etc., and a non-­positional consciousness of consciousness. In fact, I am then plunged into the world of objects, it is they which constitute the unity of my consciousnesses, which present themselves with values, attractive and repulsive values, but as for me, I have disappeared, I have annihilated myself. There is no place for me at this level, and this is not the result of some chance, some momentary failure of attention: it stems from the very structure of consciousness.33 The immediate consciousness that I have of perceiving [. . .] is not knowledge of my perception, it does not posit it: all that is intentional within my current [act of] consciousness is directed outwards, towards the world. On the other hand, this spontaneous consciousness that I have of my perception is constitutive of my perceptual consciousness. In other words, any positional consciousness of an object is at the same time a non-­positional consciousness of itself. If I count the cigarettes that are in this case, my impression is that of disclosing an objective property of this group of cigarettes: they are twelve. This property appears to my consciousness as a property existing in the world. I may well have no positional consciousness at all of counting them. I do not “know myself as counting”. [. . .] And yet, at the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

116  Richard Moran moment when these cigarettes disclose themselves to me as “twelve”, I am non-­thetically conscious of my adding activity. Indeed, if I am questioned, if someone asks me: “What are you doing now?”, I will reply immediately “I am counting” [. . .].34

Sartre’s counter of cigarettes, when he responds to the question “What are you doing?” replies spontaneously “I am counting.” Prior to that he was absorbed in his activity, but he was not in the position of someone acting in a trance. He was all along acting with awareness of what he is doing, even if his conscious thoughts were elsewhere as he counted away. Sartre calls this “non-­ positional” consciousness, and part of what “non-­positional” means here is that when absorbed the activity of counting is not an object of thought for the person but is itself a self-­conscious activity. Sartre adopts this terminology because he believes it enables us to say four things that seem true but which can seem in tension with each other. One is the more or less phenomenological fact that while absorbed in counting the person may not be thinking about anything to do with his counting. He may be humming a tune and thinking about last night, or thinking about nothing at all. Secondly, this fact does not imply that the person is acting in a trance or purely mechanically. Rather he is engaged in an ordinary purpose-­driven activity whose structure, goals, and progress he understands and is engaged with. Thirdly, the fact that he can answer spontaneously when he is asked what he is doing shows that in some sense he was aware all along of what he was doing. He draws on this awareness in providing the answer to the question asked. But, fourthly, when he does respond to the question, he is now aware of his activity in a different way, or in a different mode. His answer “I am counting” expresses an explicit self-­ consciousness of what he is doing. This is still a form of first-­person consciousness of his action, in that he does not report on his activity of counting by observing the motion of his hands with the cigarettes and concluding that he must be counting them. Rather, his explicit self-­consciousness of what he is doing is “immediate” and first-­personal. For our purposes the important thing to note at this point is that we can understand this form of explicit self-­ consciousness of his action as a form of “stepping back,” as making possible a kind of “critical distance” on his activity. He can now raise questions for himself about what he is doing; he can respond to the request for reasons from the person who asked him what he is doing. We can say that in saying “I,” when he says, “I am counting,” he asserts his agent relation to, his responsibility for, this activity and his reasons for it. This counts as a form of self-­consciousness in virtue of the form of thought expressed, its first-­person character and the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Self-Consciousness and Self-Division  117 fact that his transition from absorbed attention to the activity of counting to his explicit reflection on it is not made by his consulting any evidence about himself. When he answers the question about what he is doing his attention remains on the activity of counting itself, it does not shift to attention to his mental states (e.g., his “intention to be counting”). But the question has now brought him to a different and more reflective relation to that activity, one that he can express in first-­person thoughts about it. Here we see the possibility and the importance of distinguishing explicit self-­consciousness of one’s activity from attention to one’s mental states or “moving the operation indoors.” The attention of Sartre’s character is still directed outward at his activity of counting the cigarettes, but he is now reflectively aware of it, a form of awareness expressed in first-­person thoughts. We could say he is “stepping back” from his absorbed activity, in that he is now in a different position to raise or respond to various questions about it. But his “stepping back” is neither an alienated relation to his activity nor a shift of attention from his activity itself to his internal states. We can say similar things about the case of seeing a hawk. The person who sees the hawk has his attention fixed on the hawk. There is nothing else in his field of view. His exclamation “There’s a hawk!” is about the hawk in the sky. But as we distinguished the case of counting cigarettes from acting mechanically or under a trance, we can distinguish the ordinary case of seeing a hawk from any kind of “blindsight,” where the person has no awareness of seeing anything at all. We could agree with Aristotle’s thought in De Anima that it is by seeing itself that one is aware of seeing, that seeing an object contains a consciousness of itself as seeing, even when the explicit object of attention is nothing but the object of sight itself.35 When the person exclaims “There’s a hawk!” he may be challenged by another person who says “Really? Where?” and like Sartre’s counter of cigarettes he can spontaneously answer by saying “I’m looking right at it.” His answer here is in the first person and expresses self-­consciousness of his seeing, different and more reflective than his earlier absorbed attention to the hawk. But, again, we don’t have to describe this more reflective relation to his seeing the hawk as a shift of attention to his mental states as such. When he says “I’m looking right at it,” his attention is still fixed on the hawk, keeping it in view, but his first-­person statement is now not simply about the existence of a hawk but also expresses his self-­ consciousness of his own position vis-­à-­vis the hawk, that it is in front of him now, and that it is by seeing it rather than in some other way that he is aware of it. Again, in parallel with Sartre’s story, we can say that this reflective awareness is spontaneous in the sense of not being based on observation of himself

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

118  Richard Moran or inference from anything beyond the appearance of the hawk itself, and that the explanation of this reflective awareness must presume a non-­positional or pre-­reflective self-­awareness that belongs to seeing itself. There is a parallel to this in the practical situation as well. The practical point of view is a first-­personal point of view on the question of what to do, and in this it is fundamentally different from an external, explanatory perspective on what one will do or what forces are acting upon one. In the practical point of view I may be said to “step back” from the “givens” of my situation and reassess my relation to them, distance myself from what in the situation may be prompting me to react heedlessly. The situation before me may be painful, confusing, or tempting in various ways and when I seek to gain perspective on it, I may assert the first-­person perspective of deliberation and ask “How am I to relate myself to the situation? What am I to make of all this?” Nothing at this point, however, obliges me to turn my gaze inward. As with the person with his attention on the hawk, who is in a position to self-­ consciously say to his companion “There’s a hawk” because, as he puts it, “I’m looking right at it,” so the practical agent asserts his first-­person relation to the situation before him in attending to what is painful, confusing, or tempting in the situation itself. Before reflection he may have been simply absorbed like Sartre’s counter of cigarettes, and buffeted about by the conflicting currents of the situation, but he now reflects and relates to that same situation self-­consciously, calling into question his framing of it. In some of her formulations, Korsgaard moves imperceptibly from the assertion of the perspective of practical deliberation, as against an explanatory or third-­person perspective, to an inward-­looking concern with one’s desires and inclinations as such: From a third-­person point of view, outside of the deliberative standpoint, it may look as if what happens when someone makes a choice is that the strong­est of his conflicting desires wins. But that isn’t the way it is for you when you deliberate. When you deliberate, it is as if there were something over and above all of your desires, something which is you, and which chooses which desire to act on. This means that the principle or law by which you determine your actions is one that you regard as being expressive of yourself.36

In this passage we move from the rejection of the “external” or “battlefield” metaphor of practical deliberation to a depiction of the agent’s own relation to his own desires as such. But we can reject the explanatory point of view here

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Self-Consciousness and Self-Division  119 in favor of the perspective of the practical agent himself without thereby seeing the practical problem as a matter of which desires to act on. That “interior” framing of the issue, regarding one’s internal states, does not follow from what is first-­personal in the agent’s point of view on the question before him. A similar eliding of the difference between a Cartesian sense of the first-­person (where the contrast is between outer and inner) and the standpoint of the practical agent (where the contrast is between a theoretical or explanatory question and a practical question about what to do), occurs in her discussion of how the attractions of an activity like dancing (as incentive) take on the character of inclinations (as state of the person): At the moment, so to speak, just prior to choice, you have reflective distance from the incentive, and therefore do not take it for granted that it provides a reason. The pleasure of the dancing is what is calling out to you, but you do not, just then, take it that that pleasure provides a reason to dance: that is what you have to decide about, whether in this case you are going to take it for a reason. So for now, viewing the incentive with the questioning eye of reason, you regard the incentive’s effect on you as a mere psychological state, something normatively undecided—you regard it, that is, as just an inclination. If you do decide to act as the incentive bids, you are endorsing the incentive as providing a reason, and then it becomes more natural to mention the incentive in its character as a feature of the object rather than in its character as the source of your own psychological states.37

But here again it seems we can resist this appeal to psychological states and still respect what is first-­personal in the situation of deliberation. When I view the pleasure of dancing with the questioning eye of reason I am focused on the prospect of dancing but holding it open whether this is really the time or place for dancing. I am aware of its attractions, but they are normatively undecided for me at the moment, appearing less important to me than other competing attractions or obligations. Throughout this deliberation my attention remains on the dancing and not on any “mere psychological state.”38 And even in those situations where I am reflecting on my own desires and raising the question of whether to identify with one of them or not, who is the “I” here and how does the identity of that person concern me in my deliberations? My concern is not whether RM endorses this desire, even when I am in no doubt that I am RM. I am concerned with whether this desire is good and sufficient reason for action. From within first-­person reflection when I step back from some desire of mine and question it, my concern is not to identify

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

120  Richard Moran a particular person and ask whether he endorses this desire, but rather to ask myself whether it is to be endorsed or not. I am here self-­consciously reflecting on my desire but my concern is with the course of action to be performed and not with the identity of the person performing it. Even here, the self-­ consciousness in question is a form of thought about the world and my situation in it, like Sartre’s character running after the tram, or the counter of cigarettes when he is asked what he is doing or why and which he can answer spontaneously. He can step back from his action and question his relation to it, but in doing so he does not divide himself, as in the descriptions of self-­ consciousness in Lovejoy and Smith. And we may understand Bristow’s challenge in the following way. The authority to call into question a course of action or an appearance, and the authority to answer that question, have to be expressions of the same capacity, expressed with the same “I.” That’s the fundamental problem for any account of the self-­consciousness of critical reflection as a form of separation or self-­division. The idea of “division” first suggests itself to capture the sense of the authority to call into question, to demand reasons, for one’s actions and attitudes. But whatever authority is invoked to call into question must be the same authority to settle the question, to decisively assert identity and not difference, to give one’s answer to the question at hand, and hence cross whatever gulf was opened up when I began to reflect and demand reasons of my own actions and attitudes. Whatever critical reflection is, and the basis for holding oneself responsible, it must be one and the same capacity to raise these questions and to answer them.39 Any conception of self-­consciousness in terms of a form of self-­division will face the challenge of how the raising and the settling of the practical question can be expressions of the same capacity. One of Christine Korsgaard’s great contributions to philosophy is her placing the standpoint of the practical agent at the center of the understanding of practical normativity, and insisting on what is essentially first-­personal in the very question of normative obligation and the forms of its possible vindication. My hope here is to have shown how the importance of this re-­orientation of practical philosophy is best served by distinguishing the relevant notion of self-­consciousness from the ideas of self-­ division or the shift of attention from outward to inward.40

Notes 1. Aside from Kant himself, an important precursor here is Bernard Williams’s claim in Ethics and the Limits of Philosophy (Cambridge, MA: Harvard University Press, 1985), 67, that “practical deliberation is first-­personal, radically so,” which he deploys to very

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Self-Consciousness and Self-Division  121 different ends. In some ways my discussion in this paper is consonant with his in­sist­ ence that we not equate reflection and detachment, but I do not pursue those connections here: Ethics and the Limits of Philosophy, 69. 2. See David Kaplan, “Demonstratives,” in Joseph Almog, John Perry, and Howard Wettstein, eds., Themes from Kaplan (Oxford: Oxford University Press, 1989), 481–564 and his example of someone thinking “That man’s pants are on fire” when it is himself. 3. See Hector-­Neri Casteñeda, “‘He’: A Study in the Logic of Self-­Consciousness,” Ratio 8 (1966): 130–57; G.  E.  M.  Anscombe, “The First Person,” in Collected Papers, Vol. 2: Metaphysics and the Philosophy of Mind (Minneapolis: University of Minnesota Press, 1981), 21–36; Gareth Evans, The Varieties of Reference (Oxford: Oxford University Press, 1982); and John Perry, The Problem of the Essential Indexical and Other Essays (Oxford: Oxford University Press, 1993). To integrate my discussion with the others I discuss, and to avoid confusion, I will use masculine pronouns throughout. 4. Sydney S. Shoemaker, “Self-­Reference and Self-­Awareness,” The Journal of Philosophy 65 (1968): 555–67. 5. Anscombe, “The First Person,” raises this as a reason to doubt that such thoughts refer to anyone at all. 6. Arthur O. Lovejoy, Reflections on Human Nature (Baltimore, MD: The Johns Hopkins University Press, 1961), 106. 7. Adam Smith, The Theory of Moral Sentiments (Oxford: Oxford University Press, 1976), PART III, “Of the Foundation of our Judgments concerning our own Sentiments and Conduct, and of the Sense of Duty,” ch. 1, “Of the Principle of Self-­ approbation and of Self-­disapprobation,” 109. 8. Two further questions about Smith’s formulation: When “I divide myself ” into two persons, am I identical with either of the two remaining persons? And what do we make of the thought that it is one of the two, “the agent,” who is “the person I should properly call myself ”? Why is one of these more properly myself than the other? How are such distinctions to be made? 9. Ernst Mach, The Analysis of Sensations (Mineola, NY: Dover, 1959), p. 4, n. 1. 10. Christine  M.  Korsgaard, The Sources of Normativity (Cambridge: Cambridge University Press, 1996), 92–3. 11. Christine  M.  Korsgaard, Self-­Constitution: Agency, Identity, and Integrity (Oxford: Oxford University Press, 2009), 121. 12. “Reasons arise within the space of reflective distance; to that extent an inward glance is essential to generating them”: Korsgaard, Self-­Constitution, 124. 13. “[T]he position from which the normative question arises, the first-­person position of the agent who demands a justification of the claims which morality makes upon him”: Korsgaard, The Sources of Normativity, 16. 14. Pamela Hieronymi, “Reflection and Responsibility,” Philosophy and Public Affairs 42 (2014): 3–4. 15. “It will help to put the point in Joseph Butler’s terms, the distinction between power and authority. We do not always do what upon reflection we would do or even what upon reflection we have already decided to do. Reflection does not have irresistible power over us. But when we do reflect we cannot but think that we ought to do what on reflection we conclude we have reason to do”: Korsgaard, The Sources of Normativity, 104.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

122  Richard Moran 16. Aristotle, Metaphysics, V, 12, 1019a15–25, and Physics, Book II, 1, 192b24, in The Complete Works of Aristotle: The Revised Oxford Translation, ed. Jonathan Barnes (Princeton, NJ: Princeton University Press, 1984). 17. Vincent Descombes, Le complément de sujet: Enquete sur le fait d’agir sur soi-­meme (Paris: Gallimard, 2004), ch. XII. The entire book is an immensely rich resource for thinking about these issues. 18. See Pamela Hieronymi on the “Toxin Puzzle” in “Controlling Attitudes,” Pacific Philosophical Quarterly 87 (2006): 45–74. 19. Cf., Matthew Boyle, “‘Making Up Your Mind’ and the Activity of Reason,” Philosopher’s Imprint 11 (2011): 1–24 for a particularly helpful discussion of the different forms of activity in question. 20. “Our take on the world is not a take for which we are responsible unless we can also step back from, reflect upon, and change that take. The question we must ask is: Why should this be so? Why is responsibility subject to such a requirement? I believe that the usual thought again concerns control: with the capacity to reflect comes the possibility of controlling or having controlled one's own mind in a way that includes discretion and awareness”: Hieronymi, “Reflection and Responsibility,” 25 (emphasis mine). 21. Korsgaard, Sources of Normativity, 103. 22. William F. Bristow, “Self-­Consciousness, Normativity and Abysmal Freedom,” Inquiry 49 (2006): 509. 23. Korsgaard, Sources of Normativity, 100. 24. Jean-­Paul Sartre, Being and Nothingness, trans. Sarah Richmond (Abingdon: Routledge, 2018), ch. 1. 25. As well as that of Harry Frankfurt, “Freedom of the Will and the Concept of a Person,” The Journal of Philosophy 68 (1971): 5–20, and other “second-­order” theorists. 26. John McDowell, “Virtue and Reason,” The Monist 62 (1979): 331–50. 27. Immanuel Kant, Critique of Pure Reason, trans. N.  K.  Smith (London: Macmillan, 1933), §I6. On the original-­synthetic unity of apperception, see B132. 28. Ludwig Wittgenstein, The Blue and Brown Books (New York: Harper and Row, 1958); Anscombe, “The First Person”; and Evans, The Varieties of Reference. 29. Cf., Ludwig Wittgenstein, Tractatus Logico-­Philosophicus, trans. D.  F.  Pears and B.  F.  McGuinness (Routledge, 1961), 5.633; and Fred Dretske, “How Do You Know You are Not a Zombie?” in Brie Gertler, ed., Privileged Access: Philosophical Accounts of Self-­Knowledge (Aldershot: Ashgate, 2003), 1–13. 30. See Matthew Boyle, “Transparency and Reflection,” Canadian Journal of Philosophy 49 (2019): 1012–39. 31. Anscombe, “The First Person,” famously argues that in such contexts “I” does not serve to refer to a person at all. For defense of Anscombe’s view, see Vincent Descombes, Le parler de Soi (Paris: Gallimard, 2014); Jean-­Philippe Narboux, “Is Self-­ Consciousness Consciousness of One’s Self?” in Oskari Kuusela, Mihai Ometita, and Timur Uçan, eds., Wittgenstein and Phenomenology (New York: Routledge, 2018), 197–247; and James Doyle, No Morality, No Self: Anscombe’s Radical Skepticism (Cambridge, MA: Harvard University Press, 2018). For dissent, see Shoemaker, “Self-­ Reference and Self-­ Awareness”; Evans, The Varieties of Reference; and Béatrice Longuenesse, I, Me, Mine: Back to Kant, and Back Again (Oxford: Oxford University Press, 2017).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Self-Consciousness and Self-Division  123 32. Of course, the self-­consciousness attained by Perry’s shopper and by Oedipus is not available to them by mere reflection but requires actual investigation. My point is that reflexive first-­personal thought is not purely “formal” if that means that it does not involve a genuine change of mind for the person. 33. Jean-­Paul Sartre, The Transcendence of the Ego, trans. Andrew Brown (Abingdon: Routledge, 2004), 8. 34. Sartre, Being and Nothingness, 19. 35. “Since it is through sense that we are aware that we are seeing or hearing, it must be either by sight that we are aware of seeing, or by some sense other than sight. But the sense that gives us this new sensation must perceive both sight and its object, viz. colour: so that either there will be two senses both percipient of the same sensible object, or the sense must be percipient of itself. Further, even if the sense which perceives sight were different from sight, we must either fall into an infinite regress, or we must somewhere assume a sense which is aware of itself. If so, we ought to do this in the first case”: Aristotle, De Anima, Book 3, ch. 2  425b11–425b17, in Barnes, The Complete Works of Aristotle. 36. Korsgaard, The Sources of Normativity, 100. 37. Korsgaard, Self-­Constitution, 123–4. 38. “Reasons arise within the space of reflective distance; to that extent an inward glance is essential to generating them”: Korsgaard, Self-­Constitution, 124. It is precisely the identification of “reflective distance” with an “inward glance” that I mean to be questioning here. In the continuation of this passage Korsgaard elaborates her view with respect to the metaphysics of value and defends a “subject related” view of value, but I don’t believe that affects the point here. 39. And this is surely the thought behind Hieronymi’s critique and recommendation of an account of practical reflection in terms of the posing and answering of a question, in her “Controlling Attitudes” and elsewhere. 40. In a quite general way my thinking on these issues has benefitted from conversations with and the writing of Matt Boyle. I am grateful to him for comments on this paper, as well as to Vincent Descombes, Sandy Diehl, Jeremy Fix, Pamela Hieronymi, Jed Lewinsohn, Béatrice Longuenesse, Berislav Marusic, Jean-­Philippe Narboux, Tim Scanlon, Tamar Schapiro, and Martin Stone.

Bibliography Anscombe, G.  E.  M. “The First Person.” In Collected Papers, Vol. 2: Metaphysics and the Philosophy of Mind, 21–36. Minneapolis: University of Minnesota Press, 1981. Aristotle, De Anima, in The Complete Works of Aristotle: The Revised Oxford Translation, edited by Jonathan Barnes. Princeton, NJ: Princeton University Press, 1984. Aristotle, Metaphysics, in The Complete Works of Aristotle: The Revised Oxford Translation, edited by Jonathan Barnes. Princeton, NJ: Princeton University Press, 1984.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

124  Richard Moran Aristotle, Physics, in The Complete Works of Aristotle: The Revised Oxford Translation, edited by Jonathan Barnes. Princeton, NJ: Princeton University Press, 1984. Boyle, Matthew. “‘Making Up Your Mind’ and the Activity of Reason.” Philosopher’s Imprint 11 (2011): 1–24. Boyle, Matthew. “Transparency and Reflection.” Canadian Journal of Philosophy 49 (2019): 1012–39. Bristow, William  F. “Self-Consciousness, Normativity and Abysmal Freedom.” Inquiry 49 (2006): 498–523. Casteñeda, Hector-Neri. “‘He’: A Study in the Logic of Self-Consciousness.” Ratio 8 (1966): 130–57. Descombes, Vincent. Le complément de sujet: Enquete sur le fait d’agir sur soimeme. Paris: Gallimard, 2004. Descombes, Vincent. Le parler de Soi. Paris: Gallimard, 2014. Doyle, James. No Morality, No Self: Anscombe’s Radical Skepticism. Cambridge, MA: Harvard University Press, 2018. Dretske, Fred. “How Do You Know You are Not a Zombie?” In Brie Gertler, ed., Privileged Access: Philosophical Accounts of Self-Knowledge, 1–13. Aldershot: Ashgate, 2003. Evans, Gareth. The Varieties of Reference. Oxford: Oxford University Press, 1982. Frankfurt, Harry. “Freedom of the Will and the Concept of a Person.” The Journal of Philosophy 68 (1971): 5–20. Reprinted in Frankfurt, The Importance of What We Care About (Cambridge: Cambridge University Press, 1988), 11–25. Gertler, Brie. “Self-Knowledge and the Transparency of Belief.” In Anthony Hatzimoysis, ed., Self-Knowledge, 125–45. Oxford: Oxford University Press, 2011. Hieronymi, Pamela. “Controlling Attitudes.” Pacific Philosophical Quarterly 87 (2006): 45–74. Hieronymi, Pamela. “Two Kinds of Agency.” In Lucy O’Brien and Matthew Soteriou, eds., Mental Actions, 138–62. Oxford: Oxford University Press, 2009. Hieronymi, Pamela. “Reflection and Responsibility.” Philosophy and Public Affairs 42 (2014): 3–41. Kant, Immanuel. Critique of Pure Reason, translated by N.  K.  Smith. London: Macmillan, 1933 (1787). Kaplan, David. “Demonstratives.” In Joseph Almog, John Perry, and Howard Wettstein, eds., Themes from Kaplan, 481–564. Oxford: Oxford University Press, 1989. Korsgaard, Christine  M. The Sources of Normativity. Cambridge: Cambridge University Press, 1996.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Self-Consciousness and Self-Division  125 Korsgaard, Christine M. Self-Constitution: Agency, Identity, and Integrity. Oxford: Oxford University Press, 2009. Kosman, L. A. “Perceiving that We Perceive: On the Soul III, 2.” The Philosophical Review 84 (1975): 499–519. Longuenesse, Béatrice. I, Me, Mine: Back to Kant, and Back Again. Oxford: Oxford University Press, 2017. Lovejoy, Arthur  O. Reflections on Human Nature. Baltimore, MD: The Johns Hopkins University Press, 1961. McDowell, John. “Virtue and Reason.” The Monist 62 (1979): 331–50. Reprinted in McDowell, Mind, Value, and Reality, 50–73 (Cambridge, MA: Harvard University Press, 1998). Mach, Ernst. The Analysis of Sensations. Mineola, NY: Dover, 1959. Moran, Richard. Authority and Estrangement. Princeton, NJ: Princeton University Press, 2001. Moran, Richard. “Self-Knowledge, ‘Transparency,’ and the Forms of Activity.” In Declan Smithies and Daniel Stoljar, eds., Introspection and Consciousness, 211–36. Oxford: Oxford University Press, 2012. Narboux, Jean-Philippe. “Is Self-Consciousness Consciousness of One’s Self?” In Oskari Kuusela, Mihai Ometita, and Timur Uçan, eds., Wittgenstein and Phenomenology, 197–247. New York: Routledge, 2018. Perry, John. The Problem of the Essential Indexical and Other Essays. Oxford: Oxford University Press, 1993. Sartre, Jean-Paul. The Transcendence of the Ego, translated by Andrew Brown. Abingdon: Routledge, 2004. Sartre, Jean-Paul. Being and Nothingness, translated by Sarah Richmond. Abingdon: Routledge, 2018. Shoemaker, Sidney  S. “Self-Reference and Self-Awareness.” The Journal of Philosophy 65 (1968): 555–67. Smith, Adam. The Theory of Moral Sentiments. Oxford: Oxford University Press, 1976. Williams, Bernard. Ethics and the Limits of Philosophy. Cambridge, MA: Harvard University Press, 1985. Wittgenstein, Ludwig. Philosophical Investigations, translated by G. E. M. Anscombe. Oxford: Basil Blackwell, 1956. Wittgenstein, Ludwig. The Blue and the Brown Books. New York: Harper and Row, 1958. Wittgenstein, Ludwig. Tractatus Logico-Philosophicus, translated by D.  F.  Pears and B. F. McGuinness. New York: Routledge, 1961.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

6 What Makes Weak-­Willed Action Weak? Tamar Schapiro

1.  The Question Weak-­willed action is often characterized as action against one’s own better judgment. You know you shouldn’t drink another cocktail, but you drink it anyway. You know you shouldn’t put off fixing your car, but you put it off anyway. You know you shouldn’t spend so much money, but you spend it anyway. Acting in this way amounts to a kind of failure, by your own lights. You are doing something wrong or bad or, at least, deficient, according to your own standards. But the characterization of these actions as “weak” suggests something further. It suggests that in violating your own standards, you are also giving in to some kind of pressure. Strength and weakness are notions that we define in relation to some kind of force. We use muscle power to oppose the force of gravity. What sort of force does will power oppose? What does the strength and weakness of a will consist in?1 The standard characterization of weak-­willed action, as action against one’s better judgment, tends to make it hard to see this question clearly. Consider the example I just gave. You know you shouldn’t drink another cocktail, but you drink it anyway. Notice that in picturing this scenario, we tacitly assume the situation is one in which you want to drink another cocktail, in the sense of feeling like it, or being inclined to do so. Suppose it were otherwise. You know you shouldn’t drink another cocktail, and you have no inclination to do so, but you drink it anyway. How odd. Described in this way, the scenario is hardly a paradigmatic case of weak-­willed action. It is paradigmatic if we assume that in acting against your better judgment, you are giving in to your inclination. And yet philosophers standardly leave this last clause out of the description of the explanandum. This makes it hard to even raise the question I am interested in. What sort of pressure are you giving in to, when you go along with your inclination, against your better judgment? What sort of counterforce is required to oppose that pressure? Tamar Schapiro, What Makes Weak-­Willed Action Weak? In: Normativity and Agency: Themes from the Philosophy of Christine M. Korsgaard. Edited by: Tamar Schapiro, Kyla Ebels-­Duggan, and Sharon Street, Oxford University Press. © Tamar Schapiro 2022. DOI: 10.1093/oso/9780198843726.003.0006

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

What Makes Weak-Willed Action Weak?  127 In characterizing such action as weak, it is important to avoid embracing a picture according to which one is literally overpowered by a physical force. I am assuming that when you drink another cocktail, against your better judgment, your behavior is not simply compelled. It is not the equivalent of a hiccup or a seizure. You are not physically forced to bring the glass to your lips, against your will. Granted, there may well be such thing as responsibility-­ undermining dependence on substances like alcohol, nicotine, and opioids, as well as responsibility-­undermining compulsive conditions, like Tourette’s syndrome. Just how to conceive of these conditions is a separate philosophical question. My interest in weak-­willed action is an interest in an ordinary, nonpathological type of failure, for which we rightly, and routinely, hold ourselves and others responsible. My point here is that in thinking about this ordinary type of failure, it is unhelpful to conceive of ourselves as being literally overpowered by a brute force. That picture may seem to capture the sense in which weak-­willed action is weak, but it does so by sacrificing the sense in which such action is willed. The standard alternative is to characterize weak-­willed action as a type of deliberative error—a violation of one’s own standards of practical thinking. The accounts I have in mind have the following structure: they specify a conception of proper deliberation, and then identify the specific way in which the weak-­willed agent diverges from this procedure. According to Aristotle, the weak-­willed agent fails to properly grasp the practical syllogism.2 On Donald Davidson’s account, the agent violates the principle of continence, which tells her to act on her all-­things-­considered judgment.3 More generically, the agent is characterized as acting on the weaker rather than the stronger reason, or on the partial rather than the comprehensive reason, or on the apparent rather than the actual reason. All of these accounts avoid describing the agent as being overpowered by a brute physical force. Instead they preserve the sense in which weak-­willed action is willed, by characterizing the agent as actively engaged in practical thinking. But then in what sense is such action weak? Granted, the agent is violating her own deliberative standard, and in that sense she is acting against her own better judgment. But in making this mistake, how is she giving in? To see this worry in its most acute form, it is helpful to understand how it arises within Kant’s theory of action. A central feature of Kant’s theory of motivation is his insistence that the will is free, in the sense that it is always in principle possible for us to choose to refrain from satisfying our inclinations.4 What is sometimes overlooked is a further aspect of this freedom, commonly referred to as Kant’s “Incorporation Thesis.”5 Kant writes:

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

128 Tamar Schapiro . . . freedom of the power of choice has the characteristic . . . that it cannot be determined to action through any incentive except so far as the human being has incorporated it into his maxim (has made it into a universal rule for himself . . .); only in this way can an incentive, whatever it may be, coexist with the absolute spontaneity of the power of choice (of freedom).6

Without getting too far into the terminological weeds, the basic idea is that insofar as you have an inclination to φ (insofar as you “feel like φ-ing”), you are not yet φ-ing.7 Your inclination, simply as such, cannot determine you to act. If you are to act on your inclination, you must determine yourself to do so. You must freely accept, or endorse, or choose to act on your inclination. In Kant’s terminology, you must “incorporate” your “incentive” into your “maxim.” Later I will say more about how I think these specific terms should be interpreted, but for now the basic idea is what matters. Kantian freedom is not just the freedom to choose to refrain from acting on the inclinations you happen to have. It is also the burden of not being able to simply allow your inclinations to do the choosing for you. The Incorporation Thesis implies that weak-­willed action always involves some sort of free choice. That makes it particularly challenging to explain how weak-­willed action is weak. Like the standard theorists I discussed earlier, Kant seems to characterize weak-­willed action as a kind of deliberative mistake. While there is much room for controversy about exactly what the mistake is, it involves some sort of failure to give the moral law priority over self-­love.8 Kant holds that all human agents are necessarily committed to these two basic practical principles. Proper deliberation involves making the moral law the unconditioned condition of self-­love, in the sense that you commit to satisfying your inclinations only on the condition that you can do so without violating the moral law. The weak-­willed agent at some level reverses this ordering. She satisfies her inclination regardless of whether doing so is consistent with the moral law, and so gives self-­love priority. This is a mistake by her own lights, because as Kant sees it, she is at the same time necessarily—if only implicitly—committed to the supremacy of the moral law. Suppose we can give a coherent account of this deliberative error. Such an account might preserve the sense in which weak-­willed action is willed. But can it explain how weak-­willed action is weak? If the agent is free in the sense spelled out by Kant’s Incorporation Thesis, then even a strong inclination cannot, in principle, determine her to reverse the proper order of the two principles. And if that is so, then what could it mean to say, even in principle, that her inclination puts pressure on her will?

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

What Makes Weak-Willed Action Weak?  129 What, in principle, could pressure a free will?

2.  Rational Pressure Kant himself does not give us a developed theory of the relation between inclination and will.9 Indeed he even seems to talk about the influence of inclination in various incompatible ways. Sometimes he uses a gravitational metaphor. He describes our inclinations as providing the “counterweight” to the commands of duty.10 And he writes that because human nature is affected by inclinations, “virtue can never settle down in peace and quiet with its maxims adopted once and for all but, if it is not rising, is unavoidably sinking.”11 The gravitational metaphor is clearly in tension with Kant’s Incorporation Thesis, because we can in principle be helpless to overpower gravity, while we cannot in principle be helpless to refrain from acting on the inclinations we have. That said, Kant does not rely solely on the gravitational metaphor. Sometimes he describes inclinations as “claims” that one part of ourselves addresses to the rest of ourselves.12 Could this idea help us to understand weak-­willed action? Suppose that, in being inclined to drink another cocktail, you aren’t just being pushed around. Your inclination does not influence you as a brute force. Instead it operates as a kind of lens through which you “see” certain considerations as reasons. When you are inclined to drink another cocktail, you are, let’s say, “seeing” the pleasant effects of drinking another cocktail as a reason to drink one.13 The idea is that the motivational pressure of your inclination is not the pressure of a brute, physical force, but rather the normative pressure of a reason, as seen through the lens of your inclination. What is the normative pressure of a reason? Consider the pressure you feel when you understand the premises of a valid argument, but you have yet to draw the conclusion: All human beings are mortal. Your beloved is a human being. Insofar as you recognize the soundness of the premises, your mind is subject to pressure to draw the conclusion: your beloved is mortal. Call this rational pressure. Perhaps the form of pressure in relation to which the will can be strong or weak is rational pressure, in this sense. Does that help us understand the sense in which weak-­willed action is weak? The idea would be that

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

130 Tamar Schapiro in drinking another cocktail, you are seeing the situation as warranting this action. Your awareness of salient features of the situation is analogous to your recognition of the soundness of the premises, and your action is like your drawing of the conclusion. In giving in, you are giving in to rational pressure. Does this picture stand up to scrutiny? We can imagine it in two different ways. In the first scenario, you are not aware of the insufficiency of the reason you are acting on. You are aware of the consideration that drinking another cocktail would be pleasant, and you are taking that consideration as a reason to drink it, but you are not aware of any stronger reason to refrain. In that case, you are not making a mistake, by your own lights. You are just acting on the reasons, as you see them. You might be wrong, but you are not violating your own deliberative standard. You are acting on a reason that you take to be sufficient. If there is any sense in which you are “giving in,” you are giving in to the pressure of the strongest reason you are aware of. That is hardly a notion of weakness, in the pejorative sense. Alternatively, we can imagine that you are aware of the insufficiency of this reason, even as you act on it. You are acting on what you take to be the weaker reason. If that is what is going on, then you are making a mistake by your own lights. But in doing so, what are you giving in to? The answer cannot be that you are giving in to the normative pressure of a reason, as you see it. The stronger reason, by your own lights, supports forgoing the cocktail. In acting on the weaker reason, you must be giving in to some other kind of pressure. One could argue, perhaps, that in giving in to the weaker reason, you are giving in to something analogous to an optical illusion. Your inclination is not just a lens, but a distorting influence, through which weaker reasons appear to you as if they were stronger, just as a straight stick in the water appears bent. When you are inclined, against your better judgment, to drink another cocktail, you are acting on the illusion that you have stronger reason to drink.14 Does this appeal to perceptual illusion help? Again, we can imagine the scenario in two different ways. Either you are unaware of the situation as involving any distorting influence, or you are aware of it as such. In the first version, you are seeing the situation as ra­tionally warranting you drinking another cocktail, and you have no awareness of any distorting influence at work in this perception. If that is the situation, then in drinking another cocktail, you are not making a mistake by your own lights. And you are not giving in to any force other than the force of the ­stronger reason as you see it. There is nothing weak about that. On the other interpretation, you are aware of the illusion. You are “seeing” the situation as giving you reason to drink another cocktail, but you are at the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

What Makes Weak-Willed Action Weak?  131 same time aware of this perception as the product of a distorting influence. You are aware of it as mere appearance. If you nevertheless give in to the force of this appearance, what are you giving in to? Just as you are not forced to conclude, on the basis of what you regard as mere appearance, that the stick in the water is actually bent, so you are not forced to drink another cocktail on the basis of what you regard as a mere appearance that the situation calls for doing so. If you nevertheless give in, then you acted on the weaker reason. And we still do not have an account of what you are giving in to when you do that.

3.  Having an Inclination If the relevant form of pressure is neither brute force nor rational pressure, then what is it? I am going to suggest an answer that I think Kant would embrace. The only thing that can pressure a free will is the burden of freedom itself. Inclinations, as such, cannot put pressure on us, in our role as free choosers of our actions. The experience of frustrating an inclination is, no doubt, painful. But as free choosers, it is up to us to decide which pains are worth enduring. A given inclination, simply as such, cannot give us reason to judge that the pain of frustrating it is, or is not, worth enduring. But inclinations, simply as such, do provide us with the opportunity to flee the burden of our freedom. Let me explain. I will start by giving an account of what it is to have an inclination, prior to determining yourself to act on it. This is something I call “the moment of drama,” because in this moment, you are free to act on your inclination or not. You are in the moment of drama when you are having an inclination to drink another cocktail, but you have not yet determined yourself to do so. In this moment, you are free. The sense of freedom I am invoking here is not the contra-­causal freedom imagined by the incompatibilist. It is the freedom to choose among the actions that depend on you, the actions you could do if you chose to do them. A full defense of this conception of freedom would show that it is not threatened by causal determinism, or by mechanism more broadly. It would also show that attributing this sense of freedom is not arbitrary. It is both necessary and legitimate, but only insofar as we have to engage in practical reasoning.15 I will not try to offer such a defense here. The point for my purposes is that freedom in this sense is something we have to attribute to ourselves insofar as we find ourselves faced with the task of deciding what to do. Insofar as

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

132 Tamar Schapiro you find yourself having an inclination to drink another cocktail, you are faced with that task.16 To understand just what this task involves, we need a clearer picture of what you are faced with when you are inclined, but not yet determined, to do something. We need a clearer account of what I am calling the moment of drama, or the moment of “having an inclination” to do something. On the theory of inclination that I have developed elsewhere, to have an inclination to φ is “to be drawn out of yourself.”17 A part of you is seeing and responding to the world in a distinctive way, while the rest of you is aware of that activity. You are not in conflict with yourself, but you are aware of a division within yourself, such that there is something you call an “inclination” that you are “having.” To illustrate: when you are having an inclination to drink another cocktail, prior to determining yourself to act, part of your conscious mind is spontaneously focusing on the cocktail, real or imagined, in a distinctive way (I will say more about this in a moment). Moreover, this part of your mind exerts a guiding influence over your nonvoluntary agential capacities. You may start salivating, and some of your muscles may contract, as if initiating the pursuit of the cocktail. These conscious and physiological responses, though nonvoluntary, are purposively structured. They are oriented toward drinking another cocktail. In this respect, they are unlike sneezes and hiccups. When you are aware of yourself having an inclination, I claim, you are aware of a form of activity of that is purposively guided by a “someone” within, a part of your mind that is not your free, choosing self.18 The reason I am claiming that this “someone” is not your free, choosing self, is that your inclination is not attributable to your will. You can choose to drink another cocktail, and thereby drink another cocktail. But you cannot choose to feel like drinking another cocktail, and thereby feel like drinking another cocktail. What we choose to do is up to us. What we feel like doing is not up to us, in the same sense. This is a fundamental feature of our human predicament as agents. We can strive to cultivate our inclinations over time and, with luck, we can end up feeling like doing the things we choose to do. But we cannot simply feel like doing things at will. We “have” inclinations that are not the same as our actions because, for better and for worse, part of us has a practical mind of its own. This subagential part of us can initiate activity independent of our choosing selves. But it does not have a life of its own to lead. In this respect, it is not analogous to a separate individual who might happen to be located within the boundaries of your natural organism. Rather, it is more closely analogous to an organic system, like your digestive system. Your digestive system can

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

What Makes Weak-Willed Action Weak?  133 initiate activity independent of your choosing self. But it does not have a life of its own to lead. Its function is to contribute to your life. Now, your inclinations are consciously guided, in a way that your digestive activity is not. This is why it makes more sense to think of the source of your inclinations as a kind of “someone.” But that someone does not have any claim to a life of its own. Its function is to contribute to your life. I call this someone your “inner animal,” because I think it makes sense to understand it as having the structure of what Christine Korsgaard calls an “instinctive mind.”19 As Korsgaard characterizes it, the mind of a creature of instinct relates itself to a “teleological” world, a world that is infused with purposiveness. More specifically, this mind sees a world that is infused with its own purposiveness. It is a world structured in terms of that creature’s purposes. In Korsgaard’s words, the instinctive mind sees objects as “to-­be-­fled,” or “to-­be-­eaten.” In this respect, she claims, this mind differs from the “ra­ tional mind,” which does not see the world teleologically. Exactly how to fill this out requires interpretation, but what Korsgaard is getting at is the idea that the instinctive mind neither can, nor has to, raise a certain kind of question about how it should respond to the world. Its responses are built into its nature, and are projected onto its way of seeing its world. The rational mind, by contrast, does not have its responses built into its nature in the same sense, nor does it project its responses onto the world in the same sense. Because of this, the rational mind is forced to ask a certain kind of question that the instinctive mind cannot ask, namely, a question about how it should respond to the world. To dig a bit further into the details: as I mentioned, Korsgaard characterizes the instinctive mind as seeing objects as “to-­be-­fled,” “to-­be-­eaten,” and so on. But I believe what Korsgaard intends, and needs for her argument, is a characterization that is even a bit more alien than this. In saying that the animal sees an object as “to-­be-­fled,” she risks suggesting that the animal has the capacity to separate the object, conceived non-­teleologically, from the property, conceived teleologically. She risks suggesting that the antelope can see, for example, that approaching lion as simply a thing (viewed ­non-­teleologically), to which the (teleological) property of to-­be-­fledness happens to attach. And this in turn might suggest that the antelope could, at least in principle, raise the question, “is this lion (viewed non-­teleologically) really to-­be-­fled?” To fully appreciate the idea that the nonhuman animal sees the world teleologically, I think it is better to characterize the animal as seeing, for example, “lion-­y-­to-­be-­fledness.” The structure of the thought is not that of non-­teleological object plus teleological predicate. Rather, the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

134 Tamar Schapiro package as a whole is thoroughly teleological.20 This will become important later on. Turning back now to the theory of inclination, my claim is that when you are inclined, say, to drink another cocktail, part of your mind is seeing and responding to another cocktail in this instinctive mode. Your nonvoluntary agential capacities, conscious and physiological, are purposively guided by a thought of the form, “cocktail-­y-­to-­be-­drunkenness.” The effects of this activity constitute “what it feels like” to be inclined to drink another cocktail. Your inner animal is active, while the rest of you, the one who is “having” the inclination, is as yet undetermined. You are, in this sense, drawn out of yourself. This is not yet an account of weak-­willed action. It is an account of what it is to have an inclination to do something. I have not assumed in this description that your inclination runs contrary to your better judgment, nor have I said much about what it takes to go from having the inclination to acting on it. But just this much of the theory might suggest a familiar picture of the human agent, according to which reason is the rider, and passion is the horse.21 Is that the model? That model would be problematic, given what I have already argued. The rider/horse picture implies that our inclinations influence us externally, as brute forces that wholly bypass the will. Moreover, horses have lives of their own to lead, whereas your inner animal does not. So what picture is more accurate? I think it is a significant philosophical challenge to lay out a conception of the relation between your inclinations and the part of you that “has” them. It is tempting to try to do this by assimilating that relation to something more familiar, but I believe that our relation to our inclinations is, at bottom, sui generis. Analogies to more familiar relations cannot capture what is distinctive about being a rational animal.22 Still, I think I can say more about what it takes to go from having an inclination to acting on it, and what I say about that will help us answer the question I have been raising about what makes weak-­willed action weak.

4. Incorporation Given the inner animal view as I have laid it out thus far, how are we to conceive of the task we face when we are in the moment of drama? When you find yourself having an inclination to drink another cocktail, your inner animal is guided by its instinctive awareness of “cocktail-­y-­to-­be-­drunkenness.” I am going to use the term “incentive” to refer to this thought. Your incentive

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

What Makes Weak-Willed Action Weak?  135 is the instinctive thought that is guiding your inner animal when you are having an inclination. Now recall that Kant’s Incorporation Thesis holds that your incentive, simply as such, cannot determine your will. Why not? Kant does not spell out why not. This is because he does not say clearly enough what an incentive is, or how it is related to your will. Now, I have just given a theory of what an incentive is, at least insofar as we have in mind an incentive of inclination.23 An incentive of inclination is the instinctive thought that guides the activity that constitutes your inclination, such as “cocktail-­y-­to-­be-­drunkenness.” I want to claim that although “cocktail-­y-­to-­ be-­drunkenness” is fit to guide instinctive thinking, it is not fit to guide the thinking of your deciding self. This is because there is a mismatch between the form of the incentive, and what you need to function as a free agent. As a free agent, you have to choose your actions. To choose your actions, you have to first represent them to yourself, as possible objects of choice. My claim is that your inner animal’s thought, “cocktail-­y-­to-­be-­drunkenness,” is not a representation of a possible action. It is not fit to play the role of a possible object of practical choice. To see this, consider the fact that it makes no sense to ask, “should I cocktail-­y-­to-­be-­drunkenness?” “Cocktail-­y-­to-­be-­ drunkenness” does not describe a possible action as a kind of thing—an object of reflection, distinct from the situation you face—that can be looked at, critically evaluated, and chosen or rejected. In this sense, there is a mismatch between the form instinctive thinking takes, and the form of thinking required to engage in practical deliberation.24 I believe this is why something Kant calls “incorporation” is necessary. When you are having an inclination to drink another cocktail, your incentive is “cocktail-­y-­to-­be-­drunkenness.” If you are to make the transition from having that inclination to acting on it, you have to transform “cocktail-­y-­to-­be-­ drunkenness” into something that you can deliberate about. You have to make it into a representation that can serve as an object of your choice. I believe this is the role of what Kant calls your “maxim.” Your maxim, as I understand it, is a description of your action that can serve as a possible object of your choice. “Cocktail-­y-­to-­be-­drunkenness” cannot play this role, but something like “drink another cocktail” looks like it can. It at least makes grammatical sense to ask, “should I drink another cocktail?” So as a first pass, we can say that in order to go from having an inclination to acting on it, you have to transform your incentive, “cocktail-­y-­to-­be-­drunkenness,” into a maxim, where that is something like, “drink another cocktail.” Now, this isn’t quite right, because “drink another cocktail” is not yet a maxim in Kant’s sense. A maxim, in Kant’s sense, includes something about

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

136 Tamar Schapiro the agent’s reasons for choosing to do what she does. It is a description of your action, under which you find it worthy of your choice. So construed, the notion of a maxim is the correlate of Korsgaard’s notion of a “practical identity.”25 A practical identity is a description of who you are, under which you value yourself. A maxim, I want to say, is a description of what you are doing, under which you value it. And just as your practical identity sets the standard of your integrity as a person, your maxim sets the standard of the integrity of your action. It is with reference to your maxim that you judge whether or not you are successful in doing what you have chosen to do. If that is the right way to understand the notion of a maxim, then “drink another cocktail” is not yet fit to be one. “Drink another cocktail” is simply an outward description of an action. It says nothing about what you might find choiceworthy in drinking another cocktail. So if you are to incorporate your incentive into your maxim, you have to do more than transform “cocktail-­y-­ to-­be-­drunkenness” into an outward description, like “drink another cocktail.” You have to transform it into something like, “drink another cocktail, in this way, at this time, for this reason, and so on.” You have to represent your action to yourself as the specific form of cocktail-­drinking that you could find choiceworthy, given your situation. Your maxim is that more specific description. Let me construct a scenario to make this more concrete. Suppose the context in which you are inclined to drink another cocktail is your child’s wedding reception. Even though you know you will likely incur a headache in the morning, you consider that it might be worth it to have another drink. You have been very anxious in anticipation of this event, worrying about all the logistics, and now you just want to be able to relax a bit and enjoy each moment. In considering this choice, you are representing the possible action to yourself in a certain way. You are thinking about drinking another cocktail, given that this is a special occasion, for the purpose of celebrating the moment more deeply, on the condition that you are not likely to become objectionably inebriated, in a way that might embarrass yourself or anyone else, and on the condition that you are not going to get behind the wheel of a car. That is your maxim.26 It is much more detailed than a generic outward description of the action. Indeed, it describes the specific version of cocktail-­drinking that you can claim as the object of your possible choice, given your situation. To summarize, then, on my interpretation of Kant’s Incorporation Thesis, your incentive, “cocktail-­y-­to-­be-­drunkenness,” is unfit to serve as the object of your choice, because it is not a representation of an action. This tells us something about the structure of the task you face, when you are inclined, but not yet determined, to drink another cocktail. In order to act on your

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

What Makes Weak-Willed Action Weak?  137 incentive, you have to make it into a representation of an action that could be the object of your choice. You can do this by taking the instinctive thought, “cocktail-­y-­to-­be-­drunkenness” as motivational raw material out of which to construct your maxim. I call this “humanizing” your incentive. You are transforming a guiding thought that is fit for an instinctive mind into a guiding thought that is fit for the mind of a free agent. To “incorporate” your incentive into your maxim is to humanize it, in this sense.

5.  Giving In I have claimed that when you are having an inclination to do something, you are faced with the following task: can you transform your incentive into a representation of a possible object of your free choice? This will not always be possible. Suppose you are inclined to drink another cocktail at your child’s wedding, but you are committed to driving family members home at the end of the evening. In that case, there may be no version of drinking another cocktail that you can choose, consistent with your values. You are inclined to drink another cocktail, but given your commitment to being a conscientious driver, you simply have to forgo it. Fair enough. But notice what Kant’s Incorporation Thesis seems to imply in that situation. It seems to imply not only that you ought not to drink another cocktail but that you cannot. Kant’s claim is that your incentive cannot determine your will unless you incorporate it into your maxim. We are assuming that in this situation, it is impossible for you to construct a version of drinking another cocktail that you can find choiceworthy. In that sense, it is impossible for you to incorporate your incentive into your maxim. It follows, then, that you cannot act on that incentive. But of course you can act on it. You can certainly drink another cocktail, against your better judgment, even though you have committed to drive family members home. What are you doing when you are doing that? I think Kant should have said more here, but he didn’t.27 I want to show that the inner animal view provides an answer. When you are having an inclination to drink another cocktail, you are, as Kant says, “at a crossroads.”28 I want to characterize the crossroads in this way: you can accept the burden of your freedom, or you can flee that burden. If you accept it, you take “cocktail-­ y-­to-­be-­drunkenness” as motivational raw material out of which to try to construct your maxim. That is, you take your incentive as a kind of resource, out of which to actively construct a worthy object of your choice. You take it upon yourself to humanize your impulse as a condition of acting on it. In this sense, you affirm your position as governor of yourself, leader of your life.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

138 Tamar Schapiro But there is an alternative path open to you.29 You can flee the burden of your freedom, by fleeing into your animality. Instead of taking it upon yourself to humanize your impulse, you dehumanize yourself. How might this be possible? You cannot choose to become a nonhuman animal, and thereby become a nonhuman animal. You can fall into a condition that is something like that of a nonhuman animal, due to either injury or disease, or as a response to a situation that your organism registers as an emergency. But you cannot put yourself into this condition at will. What I am trying to describe in weak-­willed action is a way of freely putting yourself into the position of a nonhuman animal, while all the while not being one. When you take this low road, you inhabit the instinctive part of your mind, as if it were your whole self, in such a way as to avoid taking responsibility for what you do. Your inner animal is already guided by cocktail-­ y-­ to-­ be-­ drunkenness, and you, the one “having” the inclination, allow yourself to be so guided. You act as if the instinctive part of your mind were your whole mind. You act as if it were simply your nature to be a drinker of cocktail-­y-­to-­ be-­drunkenness. In this sense, you act as if you have no choice in the matter of what to do with your incentive. To see how this way of inhabiting your inner animal’s mind counts as fleeing your freedom, I want to contrast it with a different way of inhabiting your inner animal’s mind. The former I will call “giving in,” while the latter I’ll call “loosening up.” There is nothing wrong with loosening up. Suppose you are dancing at your child’s wedding. You find yourself becoming self-­ conscious, thinking too much, and failing to enjoy the moment. So you decide to try to loosen up. You try focusing more directly on the music, tuning out all the other people in the room, so as to allow your body to move more spontaneously. As you are doing this, you are aware of limitations. You are not going to move in such a way as to step on someone else’s toes, or make a complete spectacle of yourself. But you do make an effort to quiet your deciding mind, so as to allow your instinctive mind to have greater latitude in guiding you. With any luck, you will stop overthinking and start “going with the flow.” In loosening up, you are trying to give your instinctive mind a greater role to play in shaping what you do. But notice that this is entirely consistent with taking responsibility for what you do. You are choosing to loosen up, as a good way of doing something you value, namely enjoying yourself at your child’s wedding. Loosening up is thus something you do at the executive, rather than the legislative stage of action. It is something you do, as a way of more effectively doing what you have chosen to do.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

What Makes Weak-Willed Action Weak?  139 Giving in is different. It is not something you do at the executive stage, as a way of more effectively doing what you have chosen to do. It is, rather, something you do instead of choosing to do anything. When you give in, you are not enlisting your inner animal as a resource, something you can use to enhance your effectiveness in leading your life. Instead you are abandoning your post as the leader of your life, by escaping into your instinctive mind. Your inclination does not force you to do this. But it creates the opportunity for you to do this, by making you aware of an instinctive mind to flee into. Isn’t this explanation too fancy? Why not just say that in giving in to your inclination, you are choosing to avoid the pain of frustrating your inclination? The idea that it is painful to thwart an inclination, simply as such, makes sense on my view. Recall that in having an inclination, your non-­voluntary capacities are already purposively engaged, and you experience this physiologically and consciously. Once non-­voluntary activity is underway, there is a sense in which thwarting that activity before it has reached completion is painful. So why not just say that in giving in to your inclination, you are choosing to avoid this pain? As a free agent, it is up to you to decide which pains are worth avoiding, and which pleasures are worth pursuing. Sometimes it can make sense to act on an inclination, in part because enduring the pain of resisting it is simply not worth it to you. When that is the case, your maxim will reflect this consideration. But when you simply give in to your inclination, you are acting as if it is your nature to be an avoider of pain, simply as such. In this sense, you are fleeing your freedom. When you give in, are you acting without a maxim? On my view, the answer is that you are acting on a defective maxim. You are taking the instinctive thought, “cocktail-­y-­to-­be-­drunkenness,” and forcing it into the role of your maxim, even though it is unfit to play this role. As a consequence, your action is defective, qua action. It is essentially reactive, rather than active. Granted, it is not like a hiccup or a sneeze. But there is a sense in which you are acting without really choosing your action, and without really knowing what you are doing. You are reacting to the situation, instead of deciding what to do, given the situation. As I have emphasized, “cocktail-­y-­to-­be-­drunkenness” is not a description of an action, and it cannot really serve as an object of choice. But insofar as you can inhabit your instinctive mind, pushing aside the awareness of your capacity for free decision, you can still press this raw incentive into the role of a maxim. When you do this, you are coming as close as is humanly possible to acting without choosing, and to acting without knowing what you are doing. But since you are not externally coerced or

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

140 Tamar Schapiro deceived—since the defect in your action is your own doing—you are fully responsible. I just said that when you give in, you push aside the awareness of your capacity for free decision. What I mean to note is that in giving in, you are necessarily engaged in a kind of self-­deception.30 You are not simply deceiving yourself about which consideration is the stronger reason.31 To put it this way is to make weak-­willed action sound like a mistake within deliberation. On my view, weak-­willed action is a flight from deliberation. To the extent that you are deceiving yourself, you are deceiving yourself about your nature. You are acting as if you are not burdened with the freedom of having to construct and choose your actions. Rationalization, in the pejorative sense, is made possible by this more fundamental form of self-­deception. What is it to rationalize your weak-­willed action? I want to resist the idea that a rationalization is just an inaccurate account of the deliberation that led you to act as you did. When you offer a rationalization, to yourself or to others, you are not just being insincere about which reasons you acted on. More fundamentally, you are being insincere about having acted on reasons at all. Instead of deliberating, you simply gave in, and then you made up a story about having had reasons for what you did. This may make it sound as if, on my view, weak-­willed action is something essentially impulsive, something that cannot be calculated and instrumentally sophisticated. That is actually not implied by my view. Suppose you give in to your inclination to drink another cocktail. If there are no cocktails close to hand, you will have to find one. To do this, you might have to strategically make your way through a crowd of wedding guests, politely disengaging from any lengthy conversations, and targeting the bartender who can serve you most efficiently. To this extent, you have to know how to master objects in your environment. You have to approach those objects as resources for you to use, and you have to use them effectively. But mastery over your environment should not be confused with mastery over yourself. You can be fleeing your freedom internally, while making effective technical use of the means at your disposal.32

6.  Another Kind of Weakness: Going Along Inclinations do not pressure you, qua free will. Instead they give you the opportunity to flee into your animality. To do so is to flee into something Sartre called your “facticity.” It is to act as if your nature is something fixed

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

What Makes Weak-Willed Action Weak?  141 and given, something that determines your choices. Fleeing into your animality is one way of fleeing into your facticity. But it is not the only way. Another way to dehumanize yourself is to flee into automaticity. Our social environment provides us with scripts to follow. Our local cultures and subcultures are shot through with norms that tell us how to interpret the situations we are in, and how to respond to those situations, so interpreted. They tell us what to say and what to do, independent of what we may happen to feel like doing. Once we have internalized these norms and scripts, they exert a motivational pull on us, a pull that can, in principle, be at odds with what we happen to feel like doing. My claim is that these socially scripted incentives, like incentives of inclination, put us at a crossroads. Insofar as we accept our freedom, we regard these scripts as raw material out of which to construct our actions. We choose to make specific features of our cultures and subcultures our own, and perhaps to reject others, according to our values. But when we flee our freedom, we act as if we were simply built to do and say what we are expected to do and say. We allow ourselves to act as if we were cogs in the social machine. Let me explain in more detail. Socially scripted incentives are not products of an instinctive mind. They do not have the form, “cocktail-­ y-­ to-­ be-­ drunkenness.” Rather, they generally specify actions as outward performances. “Smile!” “Act like a lady!” “Dress professionally!” But as I argued earlier, to describe an action as an outward performance is not yet to construct a maxim. When you take the high road with respect to an incentive like “act like a lady!” you make a decision about, say, whether and in what respects you value acting like a lady. If you decide to act like a lady, you decide to do so only in this way, for this purpose, under these conditions, given these qualifications, and so on. In doing so, you find a way of acting like a lady that counts as your own. Or, if there is no version of acting like a lady that you can value, you dispense with that incentive entirely, and find some other incentive to use as your starting point. But when you take the low road, you flee into automaticity. You act as if it is simply your function to act like a lady, and as if you have no choice in whether and how to do so. This is a kind of weak-­willed action, even if it does not fit the paradigm of indulging your inclinations against your better judgment. It is the kind of weakness that Nietzsche and Emerson were centrally concerned about, the weakness that consists is blindly following the herd or the crowd. And it is arguably the kind of weakness that Hannah Arendt and Stanley Milgram were concerned about as they tried to make sense of human nature in the wake of the Holocaust. The forms of evil they describe are not mistakes

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

142 Tamar Schapiro within deliberation. They are flights from deliberation, and from our shared humanity. This point is actually broader than it appears. Customs, habits, and routines are sources of scripts that can function in the same way. Those can be shared or idiosyncratic. Donald Davidson describes someone who has no inclination to get up and brush his teeth, but who does so anyway, against his better judgment (because what he really needs at the moment is sleep).33 Davidson uses this example to argue that weakness of will need not involve giving in to one’s inclinations. I agree. As I read it, this person has allowed himself to flee into automaticity. He has allowed himself to become a cog in a machine of his own making. Routines can be helpful when employed at the executive stage of action. They can allow us to “go on autopilot,” for the sake of being more effective at doing what we have chosen to do. But they can also provide us with the opportunity to avoid deciding what to do in the first place. Autopilot is wonderful as a tool for an autonomous person to use. It is not a substitute for autonomy.

7. Conclusion One aim of this paper has been to highlight a constraint that any theory of weak-­willed action has to meet. It has to show not only the sense in which weak-­willed action is willed but also the sense in which it is weak. Most theories that do the former characterize weak-­willed action as a kind of mistake within deliberation. But they fail to show any sense in which the agent making this mistake is weak. My suggestion is that weak-­willed action is weak because it is a flight from deliberation. Whether we are fleeing into our animality (“giving in to feel good”) or into automaticity (“going along to get along”) what we are doing, more fundamentally, is fleeing the burden of having to lead human lives.

Notes 1. This paper is an attempt to state more clearly the view I put forth in Feeling Like It: A Theory of Inclination and Will (Oxford: Oxford University Press, 2021), ch. 6. 2. Here I am trying to sidestep more fine-­grained interpretive debates of Aristotle’s treatment of akrasia in Nicomachean Ethics, Book VII. 3. Donald Davidson, “How is Weakness of the Will Possible?” in Essays on Action and Events: Philosophical Essays, vol. 1, essay 2 (Oxford: Oxford University Press, 2001).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

What Makes Weak-Willed Action Weak?  143 4. “Suppose someone asserts of his lustful inclination that, when the desired object and the opportunity are present, it is quite irresistible to him; ask him whether, if a gallows were erected in front of the house where he finds this opportunity and he would be hanged on it immediately after gratifying his lust, he would not then control his inclination. One need not conjecture very long what he would reply. But ask him whether, if his prince demanded, on pain of the same immediate execution, that he give false testimony against an honorable man whom the prince would like to destroy under a plausible pretext, he would consider it possible to overcome his love of life, however great it may be. He would perhaps not venture to assert whether he would do it or not, but he must admit without hesitation that it would be possible for him”: Immanuel Kant, Critique of Practical Reason, in Practical Philosophy, trans. and ed. Mary J. Gregor, (Cambridge: Cambridge University Press, 1996), 5:30. 5. The name comes from Henry Allison, Kant’s Theory of Freedom (Cambridge: Cambridge University Press, 1990), I.2.II. 6. Immanuel Kant, “Religion Within the Boundaries of Mere Reason,” in Religion and Rational Theology, trans. and ed. Allen W. Wood and George Di Giovanni (Cambridge: Cambridge University Press, 2001), 6:24. 7. For a detailed account of how I am carving out the concept of inclination, along with an account of the philosophical method I am employing, see my Feeling Like It, ch. 1. 8. The literature on how to reconcile Kant’s Incorporation Thesis with his account of volitional weakness includes Marcia Baron, “Freedom, Frailty, and Impurity,” Inquiry 36 (1993): 431–41; Robert N. Johnson, “Weakness Incorporated,” History of Philosophy Quarterly 15 (1998): 349–67; Iain Morrisson, “On Kantian Maxims: A Reconciliation of the Incorporation Thesis and Weakness of the Will,” History of Philosophy Quarterly 22 (2005): 73–89; Pablo Muchnik, Kant’s Theory of Evil: An Essay on the Dangers of Self-­love and the Aprioricity of History (Lanham, MD: Lexington Books, 2009); and Marijana Vujošević, “Kant’s Account of Moral Weakness,” European Journal of Philosophy 27 (2019): 40–54. 9. Christine Korsgaard mentions this as a gap in his theory in her “From Duty and For the Sake of the Noble,” in The Constitution of Agency: Essays on Practical Reason and Moral Psychology (Oxford: Oxford University Press, 2014). There she does not try to fill the gap, but in later work she seems to endorse a certain view, at least as a feature of her brand of Kantianism, if not as a reading of Kant. The view is that inclinations are perceptions of reasons, in some non-­ realist sense: Korsgaard, Self-­Constitution: Agency, Identity, and Integrity (Oxford: Oxford University Press, 2009), 6.1.4. Later, I will explain why I don’t think this view solves the problem I am addressing here. 10. Immanuel Kant, Groundwork of the Metaphysics of Morals, in Gregor, Practical Philosophy, 4:405. 11. Immanuel Kant, Metaphysics of Morals, in Gregor, Practical Philosophy, 6:409. 12. Kant, Critique of Practical Reason, 5:74. 13. Proponents of this type of view include Dennis Stampe, T. M. Scanlon, Talbot Brewer, Sergio Tenenbaum, and others. See my discussion of the perceptual analogy in Feeling Like It, ch. 3. 14. Jessica Moss uses this idea to interpret Aristotle on Akrasia: “Akrasia and Perceptual Illusion,” Archiv für Geschichte der Philosophie 91 (2009): 119–56 and Aristotle on the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

144 Tamar Schapiro Apparent Good: Perception, Phantasia, Thought, and Desire (Oxford: Oxford University Press, 2012). I am making no claim about how to read Aristotle. 15. See Hilary Bok’s detailed defense of practical standpoint compatibilism in her book, Freedom and Responsibility (Princeton, NJ: Princeton University Press, 1998). 16. Putting it this way is a little ambiguous. Does the condition of having an inclination generate the task in the first place, or do we face the task independently, and then inclinations suggest possible ways of meeting it? I have not tried to argue the point, but I believe I am committed to the former. I am grateful to Kyla Ebels-­Duggan for pressing me on this. 17. Schapiro, Feeling Like It, ch. 4. 18. Kant comes closest to saying this when he refers to the “pathological self ” striving to make claims as if it were your whole self: Critique of Practical Reason 5:74. But Kant is not consistent in the way he writes about having an inclination. 19. Christine  M.  Korsgaard, Fellow Creatures: Our Obligations to the Other Animals (Oxford: Oxford University Press, 2018), 3.2. 20. I am eliding many qualifications and details here. I do want to take on Korsgaard’s view that the instinctive mind is intelligent and capable of learning through experience. I do not want to take on her specific way of spelling out the difference between the instinctive mind and the rational mind, as laid out in Fellow Creatures, 3.2.4. I think the way to arrive at an account of the rational mind is to start by trying to characterize our distinctive relation to the instinctive part of our own minds, which is what I try to do in Feeling Like It. 21. Or the elephant, as Jonathan Haidt argues in The Happiness Hypothesis: Finding Modern Truth in Ancient Wisdom (New York: Basic Books, 2006). 22. Korsgaard relies on the Platonic analogy between the parts of the soul and the parts of a constitutionally ordered polity: Self-­Constitution, ch. 7. I explain the shortcomings of this analogy in Feeling Like It, ch. 5. 23. The moral law is a special case of an incentive, one that does not conform to the model of an incentive of inclination. Yet another sort of incentive is one I will identify in Section 6. 24. This is a point that I should have developed more clearly in Feeling Like It, chs. 5 and 6. 25. Christine  M.  Korsgaard, Sources of Normativity (Cambridge: Cambridge University Press, 1996), “The Authority of Reflection,” and elsewhere. 26. I am making no claims that this representation has to be “conscious,” in the sense that it is readily available to you as an explicit thought. That said, the theory as a whole is addressed to you, insofar as you are trying to figure out what it takes to act well in relation to your inclinations: Schapiro, Feeling Like It, ch. 1. As such, it is addressed to an agent who is interested in governing herself properly. To spell out what that takes, I have to individuate steps in a procedure that need not always be forefront in our minds. Self-­ governed action is necessarily self-­ conscious, in that it necessarily involves some awareness of our own agency. But that does not imply that we are necessarily paying attention to that awareness. 27. See the references in note 8 for discussions of the relevant texts. 28. Kant, Groundwork, 4:400. 29. See also Schapiro, Feeling Like It, ch. 6.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

What Makes Weak-Willed Action Weak?  145 30. This is arguably another point of agreement with Kant. 31. Actually, on my view, weakness does not need to involve acting against your better judgment, or the stronger reason. The core of weakness is dealing badly with your inclinations. You could escape into your animality and still in some sense end up acting in a way that might be described as being in accordance with your better judgment. But the identification of what you are doing with what you would be doing if you acted on your better judgment would be strained, since what you are doing would be guided by a defective maxim, whereas what you would be doing would be guided by a proper maxim. 32. On Korsgaard’s mature view, efficacy presupposes autonomy: Korsgaard, Self-­ Constitution. I might not be denying that here. Although you may be efficacious in the pursuit of another cocktail, you are also fleeing your nature at this moment, and in that sense, there is a real question as to who is efficacious in that pursuit. 33. Davidson, “How is Weakness of the Will Possible?”

Bibliography Allison, Henry. Kant’s Theory of Freedom. Cambridge: Cambridge University Press, 1990. Baron, Marcia. “Freedom, Frailty, and Impurity.” Inquiry 36, no. 4 (1993): 431–41. Bok, Hilary. Freedom and Responsibility. Princeton, NJ: Princeton University Press, 1998. Davidson, Donald. “How is Weakness of the Will Possible?” In Essays on Action and Events: Philosophical Essays, vol. 1., essay 2. Oxford: Oxford University Press, 2001. Haidt, Jonathan. The Happiness Hypothesis: Finding Modern Truth in Ancient Wisdom. New York: Basic Books, 2006. Johnson, Robert N. “Weakness Incorporated.” History of Philosophy Quarterly 15, no. 3 (1998): 349–67. Kant, Immanuel. Critique of Practical Reason (1788). In Practical Philosophy, translated and edited by Mary  J.  Gregor. Cambridge: Cambridge University Press, 1999. Kant, Immanuel. Groundwork of the Metaphysics of Morals (1785). In Practical Philosophy, translated and edited by Mary  J.  Gregor. Cambridge: Cambridge University Press, 1999. Kant, Immanuel. The Metaphysics of Morals (1797). In Practical Philosophy, translated and edited by Mary J. Gregor. Cambridge: Cambridge University Press, 1999. Kant, Immanuel. Religion Within the Boundaries of Mere Reason (1793). In Religion and Rational Theology, translated and edited by Allen  W.  Wood and George Di Giovanni. Cambridge: Cambridge University Press, 2001.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

146 Tamar Schapiro Korsgaard, Christine  M. The Sources of Normativity. Cambridge: Cambridge University Press, 1996. Korsgaard, Christine M. Self-Constitution: Agency, Identity, and Integrity. Oxford: Oxford University Press, 2009. Korsgaard, Christine  M. “From Duty and For the Sake of the Noble: Kant and Aristotle on Morally Good Action.” In The Constitution of Agency: Essays on Practical Reason and Moral Psychology, ch. 6. Oxford: Oxford University Press, 2014. Korsgaard, Christine M. Fellow Creatures: Our Obligations to the Other Animals. Oxford: Oxford University Press, 2018. Morrisson, Iain. “On Kantian Maxims: A Reconciliation of the Incorporation Thesis and Weakness of the Will.”  History of Philosophy Quarterly  22, no. 1 (2005): 73–89. Moss, Jessica. “Akrasia and Perceptual Illusion.” Archiv für Geschichte der Philosophie 91 (2009): 119–56. Moss, Jessica. Aristotle on the Apparent Good: Perception, Phantasia, Thought, and Desire. Oxford: Oxford University Press, 2012. Muchnik, Pablo. Kant’s Theory of Evil: An Essay on the Dangers of Self-Love and the Aprioricity of History (Lanham, MD: Lexington Books, 2009). Schapiro, Tamar. Feeling Like It: A Theory of Inclination and Will. Oxford: Oxford University Press, 2021. Vujošević, Marijana. “Kant’s Account of Moral Weakness.”  European Journal of Philosophy 27, no. 1 (2019): 40–54.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

7 Integrity, Truth, and Value Sigrún Svavarsdóttir

1. Preliminaries When someone trusted tells me that a new colleague is a person of integrity, I come to expect that I am dealing with a principled person. Moreover, I expect her principle to be rooted in firm convictions and commitments as well as in a strong disposition to stand by them in word and deed. I will know where she stands, whether we will agree or not. These expectations are hardly based on an idiosyncratic understanding of “integrity.” The stereotype of the person of integrity as someone who stands by her convictions and commitments runs like a unifying theme through the philosophical literature on integrity. I rely on this stereotype in picking out the character trait that is my primary subject in this paper. It develops the position that personal integrity comes with a cluster of mental dispositions that warrants the praise of being open-­ minded, thoughtful, and honest when figuring out what to believe and to what to commit. In other words, integrity is not only a matter of standing by one’s convictions and commitments in word and deed but also a matter of forming and regulating these convictions and commitments in an open-­ minded, thoughtful, and honest manner. I will refer to this as the thoughtfulness condition on integrity. Satisfying the condition manifests a certain kind of concern for truth and value, even if there is no guarantee that the concern will result in convictions and commitments fastened on to truths and values. I argue that insofar as there is a requirement that links integrity to truth and value, it is satisfied by meeting the thoughtfulness condition. The mental dispositions in question are characteristic of reflective agents: agents who have awareness of themselves and their own states of mind, who have attitudes toward their own attitudes, and who exercise some regulative control over their attitudes. These are features of agency emphasized in Christine Korsgaard’s influential work on moral psychology, including her work on integrity. However, I suspect that Korsgaard will think my approach Sigrún Svavarsdóttir, Integrity, Truth, and Value In: Normativity and Agency: Themes from the Philosophy of Christine M. Korsgaard. Edited by: Tamar Schapiro, Kyla Ebels-­Duggan, and Sharon Street, Oxford University Press. © Sigrún Svavarsdóttir 2022. DOI: 10.1093/oso/9780198843726.003.0007

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

148  Sigrún Svavarsdóttir to integrity fundamentally misguided because I do not accept that the integrity of a person and the integration of the self are one and the same phenomenon. Korsgaard ingeniously draws a connection between the stereotype of a person of integrity as someone who lives by her commitments and the understanding of integrity as essentially a matter of being whole or undivided. Apropos, in The Sources of Normativity, she introduces integrity into her discussion as follows: Etymologically, integrity is oneness, integration is what makes something one. To be a thing, one thing, a unity, an entity; to be anything at all: in the metaphysical sense, that is what it means to have integrity. But we use the term for someone who lives up to his own standards. And that is because we think that living up to them is what makes him one, and so what makes him a person at all.1

True enough, a person or a self is a complex entity, and it is an interesting metaphysical question what its various parts are and what unifies them into a single whole, both at a time and across time. Since “integrity” means (in some of its uses) something like the wholeness of a complex entity given by a unifying structure, this metaphysical question may be worded as concerning “the integrity of a person.” However, Korsgaard is not making this mere terminological point but rather expressing the substantive idea that being a person of integrity, in the sense used at the opening of this paper, is one and the same phenomenon as being a person with a well-­integrated self. As intriguing as this idea is, I am skeptical of it. In earlier work, I have disputed what I have called “the integration view of integrity”: the thesis that personal integrity—the character trait picked out by the stereotype of standing by one’s commitments—and the integration of the self are one and the same phenomenon.2 I have argued that even if there may be an interesting relation between these two phenomena, it is not that of identity. It is possible to hold on to one’s integrity when one’s self is badly integrated. Indeed, integrity can be superbly displayed in handling problems that arise due to a poor integration of the self. I will not revisit these arguments in this paper, but I proceed from my earlier conclusion that the integration view of integrity is false. So, I do not use “integrity” interchangeably with “integration.” I use it exclusively to refer to the character trait picked out by the stereotype of a person who has firm convictions and commitments and a strong disposition to stand by them in word and deed, though I argue that we need to add to this stereotype the characteristic of being apt to form and regulate

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Integrity, Truth, and Value  149 these convictions and commitments in an open-­minded, thoughtful, and honest manner. A few more preliminary points are in order. First, as I use “conviction,” it designates the doxastic attitude of believing a proposition. I assume that such an intentional mental state grounds not only a disposition to assent to the proposition in question but, also, behavioral dispositions conditioned on the presence of appropriately content-­related conative attitudes. I do not make any additional assumptions about the nature of belief unless specially noted. Second, I use “commitment” for a subspecies of conative attitude.3 As all conative attitudes, these are intentional states that combine with beliefs in motivating action: that is, the person acts in a certain way because she believes such and such and is committed to thus and so, when the “because” signals that (cases of over-­determination aside) this person would not have acted so unless she had had these attitudes and, moreover, that these attitudes, due to their content, render the action intelligible from this person’s intentional perspective. Now, I am of the persuasion that it is important for a theory of agency to distinguish between various subspecies of conative attitudes. The mental attitudes that I call “commitment,” whether long term or short term, are among those conative attitudes which fix the central parameters of how the agent frames her practical deliberations. Here, I will not make any additional assumptions about the nature of commitment unless specially noted. My final preliminary point concerns the idea that the commitments of the person of integrity are “identity-­conferring.”4 True enough, since a person of integrity has strong commitments and lives by them in word and deed, integrity goes hand in hand with a strong personality and, hence, a strong identity in one sense of that ambiguous term. However, those who emphasize the connection between integrity and identity often portray the person of integrity as committed to a specific self-­conception. This is something that I reject, which marks a second point of disagreement with Korsgaard. She draws a close connection not only between integrity and the integration of the self but, also, between integrity and practical identity, when practical identity is given by “a description under which you value yourself, a description under which you find your life to be worth living and your actions to be worth undertaking.”5 You integrate yourself and, thus, achieve integrity by living up to your own standards, that is, by realizing the role-­descriptions under which you value yourself and, thus, by establishing your practical identities. Although Korsgaard makes clear that having a practical identity need not be routinely manifested in explicit thoughts about oneself as satisfying or aspiring to satisfy the relevant role-­ description,6 I understand her as ascribing

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

150  Sigrún Svavarsdóttir commitments to the person of integrity whose intentional objects essentially involve a reference to oneself. I reject that the commitments of the person of integrity need to include such a self-­reference or amount to anything like valuing oneself under one or more role-­descriptions.7 Moving beyond preliminaries, Sections 2 and 3 develop the main thesis of this paper: namely, personal integrity comes with a cluster of mental dispositions that warrants the praise of being open-­minded, thoughtful, and honest when figuring out what to believe and to what to commit. Section 4 makes the case that such mental dispositions are rightly seen as manifesting a concern for truth and value, although this concern is not to be construed as a desire or a passion for truth and value. This section includes a discussion of the role of the concept of value in the examination of commitments as well as of what it takes for such an examination to manifest a de dicto concern for value. Section 5 considers objections to my claim that insofar as there is a requirement to the effect that the person of integrity is, in some sense, oriented toward truth and value, that requirement is fully satisfied with the kind of concern for truth and value that is characteristic of those who meet the thoughtfulness condition. In this context, the thoughtfulness condition is invoked to explain the difference between integrity and fanaticism, and virtue theoretic considerations in favor of the thesis that integrity requires de re commitments to things of value are addressed but rejected. I will bring this introductory section to a close with a few remarks on the philosophical approach taken in this paper. It is an examination of a character trait that is usually considered a virtue. Although I end with some tentative remarks about what kind of virtue integrity is, the paper focuses for the most part on the nature of the character trait of integrity quite apart from the questions of whether and why it is a virtue. This is a character trait of quite sophisticated agents both with respect to the nature of their commitments, their ability to stand by these commitments in word and deed, and their capacity to be aware of and to re-­evaluate their commitments. So, a study of this ­character trait takes us into theory of agency. However, I am not offering an account of integrity intended to play a pivotal role in a systematic theory of the nature of reflective agency—let alone of unified agency, personal identity, or normativity as in Korsgaard’s seminal work on these issues. My musings on integrity are but philosophical reflections that aim to clarify which character trait is ascribed to an agent when she is described as a person of integrity. I am merely doing the groundwork of elucidating one of the phenomena to which a systematic theory of agency as well as a theory of virtue need to be alert.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Integrity, Truth, and Value  151 In my reflections, I pay attention to how we talk about integrity, and I rely enough on tailor-­made examples to invite the understanding that I am running thought experiments to test hypotheses about the concept that “integrity” expresses in ordinary usage. I have no objections to being thus understood so long as it is not mistaken for an ambition to provide a classical analysis of the ordinary meaning of the term “integrity” by precisely spelling out necessary and sufficient conditions for its applicability. First of all, I am skeptical that this can be done because I suspect that imprecision or vagueness is, for a good reason, a pervasive characteristic of thick evaluative concepts like the one expressed by “integrity” and that this characteristic should be preserved. Second, when I speak of conditions for integrity, I have in mind not necessary or sufficient conditions for the applicability of “integrity” but, rather, dimensions along which a person can be assessed as having more or less integrity. Third, although I find questions about the nature of our current (and past) concepts fascinating, I am not solely interested in our ordinary concept of integrity but also in the quality of mind or features of agency that ordinary thoughts about integrity at least attempt to capture, whether our current concepts are optimal for that task or not. In any case, some degree of conceptual clarification is needed when elucidating what we are talking about when ascribing integrity to a person. I will start with observations about the relationship between integrity and honesty.

2.  Integrity, Honesty, and Concern for Truth The person of integrity is often described as a person who has the courage of her convictions. In some cases, courage may be needed because living by one’s convictions requires an incredible physical feat or psychological stamina. However, the idiom “having the courage of one’s convictions” is more commonly used in cases where, due to social circumstances, speaking one’s mind would pose a risk to someone or something dear, not the least one’s own life, liberty, security, or reputation. There are many and diverse factors that may call for courage in living by or expressing one’s convictions, especially if these are religious, moral, or political views not widely shared or at least not shared by the powerful. However, the convictions at stake need not be religious, moral, or political. Sometimes it is risky to voice convictions about empirical matters. The danger of expressing an acceptance of heliocentrism in seventeenth-­century Europe is a case in point. Like in this case, it can be risky to express empirical views seen as religiously, morally, or politically

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

152  Sigrún Svavarsdóttir subversive, but the risk may also be of an entirely different kind. It may be a risk to one’s reputation as a scholar or a scientist. This last observation broaches the subject of intellectual integrity: integrity distinctively displayed in theoretical inquiry rather than in practical affairs.8 The stereotype of a person of integrity as someone who stands by her convictions captures an important ingredient in intellectual integrity. A person may display great intellectual integrity by firmly and publicly sticking to her convictions rather than cowing to the intellectual establishment or pandering to intellectual fashion. However, this hardly captures all that there is to intellectual integrity. An inquirer who stubbornly holds and advocates a view, while unreasonably ignoring available evidence against it or forming an unreasonable assessment of the evidence, does not display intellectual integrity. Of course, a person of integrity can make honest mistakes both by overlooking and misconstruing evidence, but by “unreasonable” in this context, I mean to single out oversights that call the honesty, rather than the intellectual aptitude, of the inquirer into question: be these oversights due to wishful thinking, self-­deception, or more willful acts of dishonesty. There are even more unsavory cases of lack of intellectual integrity than such unreasonable oversights: cases of calculated fabrication of evidence and misrepresentation of ideas as one’s own. This draws attention to the connection between integrity and honesty. In my opening remarks, I noted that if I learn that a new colleague is a person of integrity, I come to expect that I will know where she stands, whether we will agree or not. With integrity comes a great degree of transparency of commitments and convictions both because they can be inferred from how the person behaves and because she is apt to own up to them.9 Concealing one’s commitments and convictions, let alone being deceptive about them, is not exactly a sign of integrity, even if integrity is compatible with reserve and some degree of calculation about when it is most opportune or tactful to disclose one’s convictions and commitments. Furthermore, the above observations about intellectual integrity draw attention to the fact that the manner in which a person of integrity responds to evidence in the course of inquiry manifests honesty. It is also worth noting that an investigator may have shown much integrity when forming her views yet later displayed a lack of integrity by sticking to her guns. Imagine someone who has seemingly solved a great intellectual puzzle. By a stroke of genius, she advanced a theory that reconciled apparently inconsistent evidence, plugged explanatory gaps, and made sense of all the extant data. It became the received view. At that point, she basically closed

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Integrity, Truth, and Value  153 shop and reveled in the glory. She never bothered to push the theoretical questions further or design novel experiments to test the theory more extensively, not because in her best assessment it was pointless or because she was old and ready to retire but, rather, because she did not dare to take the risk of being proven wrong. Of course, she knew that future generations might find the theory wanting, but she was not going to be the one  to  hasten that discovery. If this scholar ever had integrity, success ­corrupted her. Indeed, how a person is disposed to arrive at her beliefs is probably less important for integrity than how she is disposed to rethink or revise her beliefs. Sometimes we come to believe something merely based on a hunch. This way of forming a belief will not enhance one’s intellectual integrity, but neither will it compromise it, so long as the person in question is honest to herself and others about the lack of support for her belief and is disposed to seek confirmation of her hunch in an open-­minded and responsible fashion.10 Moreover, a person of integrity will be alive to the possibility that even well-­ supported beliefs of hers may be mistaken, and she will be disposed to seek further confirmation of them in due course. Add to this that a person of integrity will not always adjust her beliefs to eliminate an inconsistency, lack of evidential support, or an explanatory gap in her system of beliefs. She will find such lack of coherence in her beliefs troublesome and attempt to work out where she has gone wrong. However, she may deem that the only way to restore coherence is ad hoc and that it is better to wait for more results before deciding which beliefs to give up. She knows that something is not quite as she thinks it is, but she is not sure where the problem lies. In this way, the person of integrity shows more concern for truth than for having a coherent story to tell. This brings us to the connection between integrity and concern for truth. If my thoughts on intellectual integrity are on target, it is integral to integrity to show concern for truth in the way that an honest inquirer displays concern for truth. This consists in being responsive to considerations that bear on the truth of one’s beliefs (i.e., evidence for and against one’s beliefs), being disposed to seek further evidence for one’s beliefs, and keeping an open mind for the possibility that further research may call for revision or even rejection of them.11 A person of integrity need not burn with, or even calmly harbor, a passion for uncovering every possible truth. Nor need she have such a passion with respect to some restricted domain of inquiry. Curiosity is hardly an essential characteristic of a person of integrity. Rather, integrity with respect to one’s convictions is a matter of being honest and using honest methods in

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

154  Sigrún Svavarsdóttir whatever inquiry one may undertake. Only in that sense is it distinctive of a person of integrity to have a concern for truth.

3.  The Thoughtfulness Condition on Integrity Intellectual integrity has very much to do with honesty in inquiry. Not surprisingly this has to do with how the inquirer is disposed to regulate her beliefs. It may not be as obvious that leading life with integrity requires a disposition to regulate one’s practical commitments in a thoughtful, open-­ minded, and honest way that may involve a critical examination of extant commitments. Is it not possible to be a person of integrity while remaining so fiercely loyal to friends and family as to have absolutely no predilection to reconsider one’s commitment to them? Is it not possible to be a person of integrity while being unquestionably committed to a cause or a project? Indeed, isn’t a disposition to submit one’s commitments to critical scrutiny antithetical to the firmness of commitment and steadfastness of purpose characteristic of the person of integrity? True enough, having firm commitments is necessary for having the seriousness of purpose and principle characteristic of a person of integrity. Sometimes it is even claimed that the commitments of the person of integrity are unconditional, and one way for a commitment to be unconditional is that the subject would retain it no matter what she learns about its intentional object.12 However, I see no reason for thinking that it is a requirement on integrity that some (let alone all) of one’s commitments be unconditional in this way. What seems crucial is that the person of integrity does not treat everything as a matter of mere taste or convenience. Integrity requires commitments that are not conditioned on them or their objects benefitting oneself in any way. The person of integrity is disposed to act on and acknowledge her commitments, even when it does not bring her enjoyment, the approval of others, or some other incidental benefits and, indeed, even when it brings her pain, the disapproval of others, or some other personal costs. This is the kind of seriousness of purpose and principle that is characteristic of the person of integrity, and it is hardly at odds with having mental dispositions that warrant the praise of being open-­minded and thoughtful in forming and, especially, revising one’s commitments.13 Indeed, integrity does not sit well with close-­mindedness and rigidity in commitments. Take a commitment to a cause or a project and say that questions have been raised about whether that cause or project is quite what one

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Integrity, Truth, and Value  155 thought it to be. Pursuing that query may lead to a loss of that commitment. Would it be a sign of integrity to refuse to engage the issue? Quite the contrary, I would think. Or take a commitment to a friend and say that questions have been raised about the loyalty of that friend or about the ethical record of that friend. Would it be a sign of integrity to refuse to investigate whether there is something about this person that would prevent you from remaining committed to him as a friend? Such refusals may be signs of integrity if you have reasons to think that the person attempting to sow doubts is up to no good, but if you have no reason to be suspicious of this person, it seems that integrity requires that you examine whether you can stand by your commitment upon increased knowledge and understanding of its object. I am suggesting that the person of integrity has a cluster of mental dispositions in virtue of which she is rightfully seen as open-­minded and thoughtful about her commitments. They include a disposition to pay attention to the nature of the things to which she commits and to make sure that she knows and appreciates to what she commits. The suggestion is not that a person of great integrity must have been extremely thoughtful when originally forming each and every one of her commitments. Some commitments are formed before the age of reason. Others grow on an adult when, over an extended period, she interacts with the same people, engages in the same kinds of activities, or is exposed to the same kinds of experiences. Yet other commitments are formed spontaneously, as when a person instantaneously falls for someone or something. I do not mean to suggest that a person of integrity cannot have commitments formed in these ways. However, she has to have a fairly strong disposition to pay attention to what she is committing and the ramifications of being thus committed. Most importantly, she has to be open to reconsidering her commitments when something indicates that things have gone awry, and she has thoughtlessly committed to something that does not bear scrutiny. In other words, a person of integrity may have commitments that were formed unreflectively or spontaneously, but she must hold onto these commitments non-­dogmatically. This is of a piece with the observations about intellectual integrity in the previous section. A person of integrity is disposed not only to stand by her convictions and commitments in word and deed but also to reform or revise these convictions and commitments in an open-­minded, thoughtful, and honest manner. She is responsible about her convictions and commitments in that she is disposed to examine whether she can stand by them upon increased knowledge and understanding of relevant matters.14 This is what I call the thoughtfulness condition on integrity. A precise specification of the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

156  Sigrún Svavarsdóttir dispositions necessary and sufficient for meeting this condition will not be attempted here.15 However, they may be circumscribed as the kind of mental dispositions that warrant the praise of being open-­minded, thoughtful, intellectually responsible, and honest when figuring out what to believe and to what to commit. The thoughtfulness condition may seem to yield a conception of integrity that requires too much reflectiveness of the person of integrity. One way that I have heard people get at this worry is to ask whether it is an implication of my view that no one of integrity is to be found within extremely traditional communities: a putative implication that they find problematic. The thought seems to be that people within such communities have completely unreflective and uncritical allegiance to the traditions within which they are raised and which have shaped their commitments, yet they are persons of great integrity.16 I stand by my claim that people with rigid mindsets, who find it absolutely unthinkable to entertain any critical questions concerning their commitments, are too dogmatic and close-­minded to count as persons of integrity, even if they are upright members of their community who steadfastly stand by their commitments in word and deed. However, it misrepresents traditional communities to portray everyone within them as having such rigid mindsets, let alone as being completely unreflective. True, people raised within traditional communities might not conduct their critical reflections in the same manner as contemporary secular liberal humanists do, but that does not mean that they all have dogmatic, rigid, and closed minds. The exact content and structure of such reflections will depend to a large extent on the content and strength of the convictions, commitments, and other concerns of the reflecting agent as well as on the challenges that he can imagine to these convictions, commitments, and concerns. All of this will be deeply affected by his historical and social milieu, including the traditions and values within which he has been raised. When re-­examining his commitment to a wayward friend, a person of integrity in eighteenth-­century China or, for that matter, in eighteenth-­century America will not think of the same questions and considerations as a person of integrity in twenty-­first-­century America. The same can be said about two contemporaries with radically different experiences and outlooks on the world. There is no fixed set of considerations that any open-­minded and thoughtful agent will bring into a re-­examination of his commitments. It will depend on the conceptual tools that each individual agent has, the power of his intellect, the experiences he can draw on, his background views, the fertility of his imagination, and so on.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Integrity, Truth, and Value  157 It would, indeed, be dubious to spell out the thoughtfulness condition to require the “inward-­looking” self-­examination symptomatic of our age, initiated with questions like “Who am I?” or “Who do I want to be?” I trust that it has been clear from how I introduced the thoughtfulness condition that self-­ scrutiny is not what I have in mind. The kind of reflections that the person of integrity is disposed to undertake focuses on the objects of her commitments: that is, to what she is committed. This may but need not have anything to do with herself. She may be committed to specific people, to specific moral or political ideals, to a certain kind of intellectual endeavor, and so on. An examination of the object of a commitment will put the commitment to a test: the test of whether it survives increased knowledge of and critical reflection on its object. Although I refer to such an examination as “critical reflections on one’s commitments,” I do not mean to suggest that the gaze of the reflecting agent is “inwardly” focused on her attitudes or on herself rather than “outwardly” focused on the objects of her attitudes. Nevertheless, it is an implication of my view that integrity requires considerable intellectual sophistication and developed reflective capacities that involve a measure of awareness of one’s own attitudes and agency. Although the gaze of the reflecting agent is “outwardly” focused on the object of commitment that may have nothing to do with the agent herself; the purpose of the reflections is to figure out one’s stand on the matter: to make up one’s mind about whether to be or remain thus committed.17 I readily embrace this implication. Integrity is a virtue of agents who are capable of an awareness of where they stand on various matters, even if they need not be preoccupied with the kind of self-­scrutiny that is symptomatic of our age.

4.  Integrity, Thoughtfulness, and Concern for Value (as Well as for Truth) At the end of Section 2, I noted that it is integral to integrity to show concern for truth in the way that an honest inquirer displays concern for truth: integrity with respect to one’s convictions is (partly) a matter of being honest and using honest methods in whatever inquiry one may undertake. It (partly) consists in being responsive to considerations that bear on the truth of one’s beliefs, being disposed to seek further evidence for one’s beliefs, and keeping an open mind for the possibility that further research may call for revision or even rejection of them. There is, I contend, something similar to be said of the connection between integrity and concern for value. Concern for value is

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

158  Sigrún Svavarsdóttir distinctive of a person of integrity. It is manifested in an orientation toward value that comes with the kind of dispositions required for meeting the thoughtfulness condition on integrity: dispositions to be thoughtful and honest regarding the objects of one’s commitments, and to be open to reconsidering a commitment if further knowledge of its object provides a good case for doing so. This section examines what kind of concern for value (as well as for truth) is betrayed by those who meet the thoughtfulness condition on integrity. According to the thoughtfulness condition, a person of integrity is responsible about her convictions and commitments in that she is disposed to examine in an open-­minded, thoughtful, and honest manner whether she can stand by them upon increased knowledge and understanding of relevant matters. Notably, the concept of value seems tailor-­made for framing questions that initiate and drive such critical scrutiny of commitments and other kinds of practical attitude. We may ask whether what we desire really has any value, whether what we value or hold dear is really all that valuable, or whether a project to which we have committed has sufficient value. The question concerns whether the emotional or motivational attitude at stake is well directed: directed at an object that is worth it. What has value is worth such an attitude, what lacks value is not. I am not going to make a strong assertion here regarding the centrality of this function of the concept of value but only draw attention to it. There is an analogous function for the concept of truth. It can be used for framing questions that initiate and drive critical scrutiny of beliefs or other kinds of doxastic attitudes. We may ask whether what we believe is really true or whether what we doubt is perhaps true. The question concerns, again, whether the attitude is well directed: in this case, directed at a proposition that is worth the kind of doxastic attitude in question. Only true propositions are worth believing in the relevant sense of being worth an attitude. Of course, it can pay for an agent to commit to a worthless object or believe a false proposition. These attitudes may bring him fortune. The worth-­relations at stake pertain to the suitability18 of the object for the kind of attitude at stake and not the benefit, or even the intrinsic value, of having such an attitude toward the object in question. So, a person of integrity may open a re-­examination of her commitments and convictions by raising questions about the value or truth of their intentional objects. Presumably, the ensuing examination will take into account considerations that the agent takes to bear on their value or truth. Such a person would be displaying, in a very robust sense, concern for value in her commitments and truth in her beliefs. However, this is not the only way that a

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Integrity, Truth, and Value  159 person of integrity can initiate and carry on an open-­minded and responsible re-­examination of her commitments and convictions. Focusing on the critical scrutiny of commitments for now, consider Earnest, who has been convinced by careful philosophical arguments to accept skepticism about value. He did not reach his conclusion lightly but through arduous and conscientious critical examination of the case for and against. No wishful or sloppy thinking was involved. Quite the contrary, Earnest is the reluctant value skeptic who followed the argument to its logical conclusion and swallowed the bitter pill. We need not assume that the arguments in question are sound but only that it is currently beyond the cognitive powers of Earnest to see their faults, if they have any. Now, instead of leaving his skepticism in his study, he has earnestly sought to abandon the concept of value in his deliberations about how to live, for he considers this concept no more respectable than the concept of witchcraft. Nevertheless, Earnest is exceedingly thoughtful about his practical commitments. He rarely commits to anything without first examining closely the nature of the phenomenon in question and its relation to other things, how a commitment to it will affect his ability to stand by his other commitments, what sort of means he would have to take and how feasible it is for him to stand by that commitment, and how it would affect him, others, and the environment. Moreover, he has a strong disposition to undertake such a re-­ examination of his extant commitments from time to time, or at least when his epistemic position seems to have improved or his circumstances changed in important respects. These are not idle reflections on his commitments. Rather, they are emotionally and motivationally engaged. To the best of his abilities, Earnest tries to make sure that the considerations brought into his purview engage all his extant concerns, even if he also stands ready to submit any of these concerns to similar critical scrutiny. (I will describe such examination of commitments as “widely emotionally engaged” from now on.) Since such a thorough and fully engaged examination may overwhelm, he undertakes the task piecemeal over time, trying his best to gauge whether this is a commitment he could stand by upon informed consideration, while widely emotionally engaged. I see no reason to doubt that Earnest meets the thoughtfulness condition on integrity. Provided that he also has firm commitments which he lives by in word and deed, it would be hard to deny that Earnest is a person of integrity. Surely, he re-­examines his commitments sufficiently open-­mindedly, thoughtfully, and honestly, albeit having no explicit thoughts of or concern for their value. So, does this example show that it is possible to meet the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

160  Sigrún Svavarsdóttir thoughtfulness condition without having any concern for value? I doubt that. As noted, questions about the value of the object of commitment concern whether the mental attitude at stake is well directed: directed at an object that is worth it. Even if Earnest has sought to abandon the concept of value in his thinking, isn’t he still examining whether the intentional objects of his commitments are worth them? Isn’t he concerned with and seeking value in his commitments, even if he does not raise explicit questions of value, does not articulate his concern as a quest for value, and does not bring explicit judgments of value to bear on whether to retain his commitments? It may seem that the distinction between de re and de dicto reading of mental attitude ascriptions would help in articulating my questions. I may be taken to intimate that although ascribing concern for value to Earnest is false on a de dicto reading, it may be true on a de re reading.19 In other words, there may be a phenomenon that Earnest actually investigates and that phenomenon may be the value of the things to which he has committed, even if Earnest does not conceive of the subject of his inquiry in that way. Accordingly, the orientation toward value that I claim to be characteristic of those who meet the thoughtfulness condition on integrity may be thought to be a de re rather than a de dicto concern for value.20 However, that is not my intent. Indeed, I doubt this is the lesson to be drawn from the Saga of Earnest. It is not clear that Earnest has succeeded in abandoning the concept of value in his deliberations about how to live, given that he still examines the objects of his commitments in the way that he does. The following second chapter to the Saga of Earnest serves to illuminate the issue that I am raising. Imagine that not only the concept of value but also the concept of truth is a casualty of Earnest’s skeptical attitude. He is convinced on the basis of the Liar Paradox that the concept of truth is hopelessly incoherent and has sworn to abandon it in his thinking. Principled as he is, he has managed to banish the truth predicate from his active vocabulary and never articulates his questions and judgments explicitly in terms of truth, cumbersome as that may be. (Whether or not the deflationists offer an adequate theory of truth, it is hard to dispute their claim that everything expressed in terms of the truth-­ predicate is in principle expressible without the use of the truth-­predicate because it is axiomatic that, for any sentence “s,” “s” is true if and only if s.) However, Earnest continues to reason as carefully as before, holds firmly onto the distinction between appearance and reality, engages in meticulous inquiry, and undertakes a re-­examination of his convictions from time to time. Sure, he never opens his re-­examination of his belief that q with an explicit question about its truth nor makes explicit judgments about truth in the course of

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Integrity, Truth, and Value  161 his inquiry, but he initiates his re-­examination with explicit questions like “May I have missed something when arguing for q?” and concludes his inquiry by noting, for example, that since p and if p, then q, he can rest assured that q. It is questionable to conclude that Earnest has stopped operating with the concept of truth and, thus, cannot have de dicto concern for truth in his beliefs, even if he has given up on the truth predicate and avoids any explicit thoughts of truth. Granted, a common way of identifying a general concept is with reference to a linguistic predicate: the concept of F is whatever the predicate “F” expresses. However, this does not imply that possessing and applying the concept of F consists in the predicate “F” being in one’s active vocabulary. The parrot that has picked up the truth predicate and crows “That’s true! That’s true!” has not acquired the concept of truth, and there are reasons to believe that children have at least a partial grasp of the concept of truth once they distinguish between appearance and reality (around the age of three) regardless of their facility with the truth predicate in their native language. Needless to say, difficult issues concerning concept possession and concept application in general, as well as issues concerning the possession and application of the concepts of truth and value in particular, will not be resolved here. However, it seems safe to assume that an individual has and operates with the concept of truth, if he holds firmly onto the distinction between appearance and reality, engages in meticulous inquiry, reasons well, corrects his reasoning when he realizes that his conclusion does not follow from his premises or acquires evidence against the premises on which he has relied, and, to boot, undertakes re-­examination of his convictions in a way that demonstrates an awareness of his own fallibility. Similarly, we should not hastily conclude from the first chapter in the Saga of Earnest that he has stopped operating with the concept of value and, thus, cannot have any de dicto concern for value in his commitments. Consider how Earnest would conduct his inquiry were he a value realist rather than a value skeptic. Presumably, he would be apt to raise explicit questions of value and bring explicit value judgments to bear on whether to retain his commitments. In some cases, he might establish the value of that to which he is committed against the background of some general value assumptions. He might, for example, proceed on the assumption, say, that eradicating poverty is of value and, then, rest content once he saw that a certain project, to which he is committed, provides one small step towards eradicating poverty. Or, he might satisfy himself that he can stand by his commitment to the study of some arcane historical matter by reminding himself that any kind of intellectual

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

162  Sigrún Svavarsdóttir activity is of value. However, what if the value realist kept pushing the inquiry ever further, leaving no stone unturned, re-­examining even his most basic assumptions about value? How would he struggle, for example, with the question whether he is right that any kind of intellectual activity is of value, even the study of the arcane historical matter to which he is committed? Is there any way for him to address that question but to proceed in something like the way that the original Earnest, the value skeptic, does: that is, try his best to gauge whether it is something to which he could remain committed upon well-­informed and emotionally engaged consideration? I am inclined to think that inquiry into the most basic questions of value takes something like this form. Speaking as an evaluator rather than a meta-­ ethicist, I do not know how ultimately to ascertain whether something has value except by examining it (non-­evaluatively) the best I can and seeing whether it resonates with me emotionally and motivationally such that I can make it central to how to lead my life, or at least such that I can empathetically understand those who do.21 If I am right on this score, Earnest would proceed in much the same way regardless of whether he were a value realist, who understood himself as addressing basic questions regarding the value of the objects of his commitments, or a value skeptic who, while remaining emotionally and motivationally engaged, understood himself as examining whether he can stand by his commitments upon gaining increased (non-­ evaluative) information about their objects. This raises the question whether our original Earnest, the value skeptic, has broken as radically with thinking in terms of value as he takes himself to have done. It is not clear that he has succeeded in abandoning the concept of value in his deliberations about how to live, if he still examines the objects of his commitments much as he would, were he inquiring into the most basic questions of their value. This is true, even if he does not keep track of his results with explicit value judgments that, in different contexts, he can bring to bear on what to do or to what to commit. Of course, proceeding as if one were doing something does not entail that one is doing that. Appearances may mislead. However, if Earnest’s disavowal of the concept of value has fairly superficial effects on his thought processes, it is not clear that he has indeed abandoned thinking in terms of this concept (i.e., abandoned evaluative thinking) rather than merely eliminated the value predicate from his vocabulary and exorcised explicit thoughts of value. The distinction between a de re and a de dicto reading of intentional mental state ascriptions has been introduced to handle a cognitive disconnect between two ways in which an individual thinks of one and the same thing

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Integrity, Truth, and Value  163 that explains how he proceeds in thought and action. To use a well-­worn example, consider Lois Lane at a time when she has not yet come to suspect that Clark Kent is Superman and would indeed find the idea ridiculous. Since Lois Lane avidly pursues an inquiry into Superman’s secrets, she is examining Clark Kent’s secrets, even if that is not her understanding of what she is doing. It makes a considerable difference as to how Lois Lane conducts her inquiry whether or not she conceives of the subject of her investigation as Clark Kent’s secrets. If she did, she would be paying much closer attention to her fellow journalist, sitting at the next desk, than she is. At least, she would be doing so if she were rational. In other words, due to ignorance, there is a cognitive disconnect between two ways in which Louis Lane thinks of one and the same man and that, rather than irrationality, explains how she proceeds in her actual circumstances. By contrast, it is not clear that, in the case of Earnest, there is a cognitive disconnect between two ways of thinking of one and the same phenomenon so that the distinction between a de re and a de dicto reading of intentional mental state ascriptions can be similarly evoked in an explanatorily fruitful way. Earnest’s avowed skeptical attitude toward value is manifested in a disposition to shy away from articulating any questions or thoughts, either to himself or to others, by using the value predicate just as his avowed skeptical attitude toward truth is manifested in a disposition to shy away from articulating any questions or thoughts, either to himself or to others, by using the truth predicate. Otherwise, he proceeds much as he would if he were pursuing questions of value in his commitments and truth in his convictions (at least, if I am right about the form that an inquiry into basic questions of value takes). It is not clear that the effect of his avowed skeptical attitude is profound enough to demonstrate that he has no de dicto concern for truth in his convictions and value in his commitments. However, there may be a cognitive disconnect between his reflective conception of what he is investigating and the conceptual tools he actually deploys when pursuing his inquiry. Even if I have warned against concluding from the Saga of Earnest that the person of integrity need not have de dicto concern for value, I do not mean to suggest that a de dicto desire for value (or some other kind of conative attitude with the same content) is required for integrity. The concern for value distinctive of a person of integrity is an orientation toward value that comes with the kind of dispositions required for meeting the thoughtfulness condition on integrity: dispositions to regulate one’s commitments in thoughtful, open-­ minded, and honest manner. As noted at the beginning of this section, it is comparable to the concern for truth that is distinctive of a person of integrity.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

164  Sigrún Svavarsdóttir It is not clear that such dispositions are best seen as grounded in a de dicto desire for truth and value. For sure, integrity need not come with a thirst for truth, a passion for uncovering every possible truth, not even truths within some restricted domain of inquiry. Similarly, I see no reason to think that the concern for value distinctive of a person of integrity is a love of or an appetite for all things valuable, or even for the valuable within some limited area of the practical domain. The mindset of the person of integrity need not be anything like that of the aesthete. Instead, the kind of concern she has for value is manifested in bothering to examine thoughtfully and honestly the objects of her commitments, while remaining open to revising them if there is a good case for doing so. Not just anything goes as far as the person of integrity is concerned. She treats it as a matter of importance to what she is committed: what exactly it is and what the ramifications of being thus committed are. Regardless of whether she has explicit thoughts of value, she is disposed to do everything that one could, ultimately, do to ascertain the value of that to which one has committed, were one inclined to raise explicit questions of value. This is what is required for integrity, and Earnest satisfies this condition in spite of his avowed skepticism about value and his banishment of the value predicate from his active vocabulary. This is what I have in mind when I talk about a kind of concern for value that is characteristic of those who meet the thoughtfulness condition.

5.  Is There a Stronger Connection between Integrity and Value? I am inclined to think that insofar as integrity is related to truth and value, the connection is forged via the kind of concern for truth and value that was circumscribed in the previous section: a concern manifested in a disposition to examine thoughtfully and honestly the object of one’s convictions and commitments, while remaining open to revising them if there is a good case for doing so. However, before closing this discussion, let’s consider whether the connection between integrity and value is even stronger, approaching this issue by taking a look at the difference between integrity and fanaticism.22 A fanatic may be as committed and as unwavering in acting on his commitments as a person of great integrity is, and he may be just as forthcoming about them. It may even be questioned whether these are two distinct character traits. Could the difference be merely terminological? Could the terms “integrity” and “fanaticism” simply have different flavors, the former being appropriately used only if one approves of the commitments of the person in

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Integrity, Truth, and Value  165 question,23 whereas the latter is appropriately used only if one disapproves of her commitments. A competing view asserts that these are distinct character traits, while identifying the difference between them in how the agent is, as a matter of fact, related to value: the person of integrity lives by de re commitments to things of value, whereas the fanatic lives by de re commitments to something bad, even evil. This view draws a strong connection between integrity and value—much stronger than the one introduced in last section. The person of integrity is not portrayed as someone who forms and regulates her commitments in a manner that shows concern for value but as someone who has de re committed to things of value. I do not find it plausible that integrity requires a de re commitment to things of value. On this issue, I agree with Cheshire Calhoun: When Dan Quayle stood on his pro-­life principle, refusing to sanction an abortion for his own young daughter and even for a twelve-­year old raped by her father, one might have thought him hopelessly misguided; but that thought alone would not have been reason to think he lacked integrity.24

Although Calhoun stops short of ascribing integrity to Dan Quayle, her remark articulates the thought that a person may have integrity even if he has captured neither truth nor value in his convictions and commitments: integrity neither requires nor guarantees that one has gotten things right. Indeed, one may have strayed quite far from truth and value. Likewise, I do not find it plausible that the fanatic must have a de re commitment to something bad, even evil. Despite the frequent use of the term “fanatic” to designate a ruthless and violent supporter of odious causes, it seems an entirely familiar thought that someone is a fanatic about a good cause. I understand this to be an individual who is dogmatically committed to a good cause, someone who is unwilling to entertain any critical question about that cause. This is a person who would not only eschew critical discussions of the relevant issues in specific social settings but would, under virtually any condition, resist a critical reflection on the matter as if it were a heresy to subject the object of his commitment to scrutiny. Indeed, I take it to be a consideration in favor of the thoughtfulness condition that it pinpoints nicely the crucial difference between the fanatic and the person of integrity. Like the person of integrity, the fanatic stands by his convictions and commitments in word and, when feasible, in deed. Thus, if he happens to have latched onto something of value, he can be a powerful force for the good. However, unlike the person of integrity, he lacks the cluster of

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

166  Sigrún Svavarsdóttir mental dispositions that is required by the thoughtfulness condition: dispositions to be thoughtful and honest regarding the objects of one’s commitments and to be open to reconsidering a commitment if further knowledge of its object provides a good case for doing so. Fanaticism is manifested in a rigidity of mind that goes well beyond the firmness of commitment and steadfastness of purpose characteristic of the person of integrity. It is manifested in close-­ mindedness that translates into unwillingness to consider whether one has thoughtlessly committed to something that does not bear scrutiny. These are mental dispositions that manifest lack of concern for value quite analogous to the lack of concern for truth displayed by a close-­minded and dogmatic thinker. Now, this diagnosis of the crucial difference between fanaticism and integrity can be accepted without embracing my position that satisfying the thoughtfulness condition suffices for whatever requirement there is to the effect that a person of integrity be appropriately oriented toward value. An additional condition may be proposed requiring a stronger connection to value. Indeed, the following case may be thought to show the need for such a condition: imagine a person who is committed to the ethnic cleansing of a territory, acts on it and is not ashamed of publicly acknowledging it. Furthermore, imagine that this person sees his commitment as justified by the historical injuries that the targeted ethnic group has inflicted on his own. He has thought through the issues again and again as hard as he possibly can, paying attention to any consideration that seems relevant to him. Moreover, he can be easily engaged in conversations about the matter in which he earnestly struggles with all the objections that his interlocutors offer him. Indeed, he seeks out such conversations rather than shies away from them. Besides being forthcoming about his commitment, he is eager to submit it to well-­ informed and widely emotionally engaged examination. However, in spite of his thoughtfulness, the agent remains full of hatred and sticks to his commitment to the ethnic cleansing of the territory in question. This extremely reflective ethnic cleanser may seem to satisfy the thoughtfulness condition, while having a mindset that is incompatible with integrity. True, he does not have the mental rigidity of a close-­minded fanatic, but it may be questioned whether he has the mindset of a person of integrity either. Assuming that the thoughtfulness condition was adequately motivated in Sections 2 and 3, this example is best seen as putting pressure on my claim that, insofar as there is a requirement to the effect that the person of integrity is concerned about value, it is fully satisfied by anyone who meets the thoughtfulness condition. It may be thought that an additional condition is needed,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Integrity, Truth, and Value  167 requiring at least that the person of integrity has not strayed in his commitments this far from value. A tempting response to this challenge is to dispute that the extremely reflective ethnic cleanser satisfies the thoughtfulness condition on integrity. An open-­minded, thoughtful, and honest examination of a commitment to ethnic cleansing would hardly ignore the predictable suffering that ethnic cleansing inflicts on the targeted group. Were this man truly open to changing his ways if he uncovered a good case for doing so, he would abandon his commitment to ethnic cleansing upon dwelling on the victims’ suffering. Hence, if his commitment remains intact, there are grounds for challenging the assumption that he has met the thoughtfulness condition on integrity. The main weakness of this argument lies in the second premise. True enough, hatred and cruelty often depend on culpable ignorance, self-­ deception, or selective attention to the relevant facts, but that does not imply that the problem can always be cured by hard and honest thought, even if widely emotionally engaged. Hatred and anger, not the least moralized anger, can make one indifferent or even unfriendly to the suffering of those who are the target of that hatred so that the consideration that one’s cause will visit suffering upon these people will not serve to weaken one’s commitment to that cause—it may even strengthen the commitment. The claim that someone deserves to suffer is all too familiar. Thus, hate may make it psychologically impossible for the ethnic cleanser to see the suffering of the targeted group as a reason to abandon his commitment to the ethnic cleansing of the relevant territory. However, if this were true, would the extremely reflective ethnic cleanser be sufficiently open-­minded to count as a person of integrity? Does the hate blind him and close his mind in such a way that his integrity fails along the dimension specified by the thoughtfulness condition? These questions connect back to the issue discussed at the beginning of Section 3 concerning how exactly to delineate the kind of open-­mindedness that the condition requires of the person of integrity. I noted that it should not be specified such as to rule out the mental dispositions that are characteristic of people with firm commitments, while now the issue raised is whether it should be specified such as to rule out mental dispositions characteristic of people whose hate has rendered them insensitive to the plight of others, at least some others. Ex hypothesi, the hate that blinds the reflective ethnic cleanser cannot be uprooted by widely emotionally engaged, hard, honest thought alone. He has a mindset that might have some tendency to counter hatred but cannot always uproot it or weed out commitments to bad causes that grow out of it, especially when

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

168  Sigrún Svavarsdóttir there are circumstantial factors that lend themselves easily to the justification of anger. The hate limits the impact of his reflections. The question before us is whether the thoughtfulness condition should be specified such as to require only the kind of openness of mind that ensures widely emotionally engaged, hard, honest thought about the object of commitment or also the kind of openness to the suffering of others that comes with the absence of hate.25 The second option would, if successfully defended, undermine the counterexample of the extremely reflective ethnic cleanser insofar as it is put forward as a case of someone who meets the thoughtfulness condition, yet is not oriented toward value in the way required for integrity. Nevertheless, I incline toward the first option: namely, to develop the thoughtfulness condition such as to require only the kind of openness of mind that ensures hard honest thought and wide emotional engagement, treating it as a contingent matter whether such efforts can, in any given case, uproot hatred and other psychological factors that may make one amenable to committing to bad, even evil, projects. Ex hypothesi, the extremely reflective ethnic cleanser will meet the thoughtfulness condition thus developed. So, I accept the case as a challenge to my thesis that insofar as there is a requirement to the effect that the person of integrity is concerned about value, it is satisfied by anyone who meets the thoughtfulness condition. However, this is a bullet that I am willing to bite. I accept that, in spite of his de re commitment to an evil cause, the extremely reflective ethnic cleanser has the orientation toward truth and value that is characteristic of the person of integrity: a disposition to examine thoughtfully and honestly the object of his convictions and commitments, while remaining open to revising them if he finds a good case for doing so. This reflects my agreement with Calhoun that integrity neither requires nor guarantees that one has gotten truth or value right. Although this is hardly an iconoclastic position, there are those who find it implausible that a person of integrity may have as radically misguided commitments as does the extremely reflective ethnic cleanser.26 Those people may seem to have virtue-­theoretic considerations on their side, for my view of integrity may appear to have the undesirable upshot that integrity is not a virtue. This would, at least, be implied on an account of virtue which claims that what makes a character trait a virtue is the way the trait emotionally and motivationally disposes the agent (positively) toward what is de re good and (negatively) toward what is de re bad.27 This quite compelling approach to virtue can thus be marshaled against the position that I have taken on the connection between integrity and value.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Integrity, Truth, and Value  169 One possible response to the virtue-­theoretic objection consists in embracing Bernard Williams’s suggestion that integrity is not a virtue: . . . one should perhaps say that integrity is not a virtue at all. In saying that, I do not mean that there is not all that much to be said for it, as one might say that humility was not a virtue. I mean that while it is an admirable human property, it is not related to motivation as the virtues are. It is not a disposition which itself yields motivations, as generosity and benevolence do; nor is it a virtue of that type, sometimes called ‘executive’ virtues, which do not themselves yield a characteristic motive, but are necessary for that relation to oneself and the world which enables one to act from desirable motives in desirable ways—the type that includes courage and self-­control. It is rather that one who displays integrity acts from those dispositions and motives which are most deeply his, and has also the virtues that enable him to do that. Integrity does not enable him to do it, nor is it what he acts from when he does so.28

I will not borrow this response. True, integrity is not like benevolence, generosity, and kindness in having a characteristic motive. However, it seems less clear that it does not count among the “executive virtues”—virtues like courage and self-­control.29 More importantly, I believe we need to recognize a third category of virtue: virtues that orient the agent toward truth and value not through a de dicto or de re intentional state directed at truth and value but, rather, through dispositions of mind that make the agent attend to her mental attitudes—doxastic and conative—in ways that regulate for truth and value, even if there is no guarantee that mental attitudes thus regulated actually succeed in fastening on to truths and values. Virtues like honesty, intellectual responsibility, rationality (theoretical and practical), and integrity belong in that category, or so I would argue. These are important virtues of limited agents, like us, who need to cope with epistemic as well as emotional constraints in figuring out what to believe and to what to commit: agents who cannot always overcome these constraints with the intellectual and emotional resources at their disposal, however hard they try. In particular, it seems a ­virtue to be not only disposed to stand by one’s convictions and commitments in word and deed but, also, disposed to make the effort to think hard and unflinchingly about that to which one is committed, so that the commitments one stands by in word and deed are apt to be those that one can stand by upon reflection.30

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

170  Sigrún Svavarsdóttir

Notes 1. Christine  M.  Korsgaard, The Sources of Normativity (Cambridge: Cambridge University Press, 1996), 102. This idea is also at work in Christine M. Korsgaard, Self-­ Constitution: Agency, Identity, and Integrity (Oxford: Oxford University Press, 2009). 2. Sigrún Svavarsdóttir, “Coherence of Attitudes, Integration of the Self, and Personal Integrity,” in David Shoemaker, ed., Oxford Studies in Agency and Responsibility, vol. 3, 62–84 (Oxford: Oxford University Press, 2015). 3. The word “commitment” is used differently in the following sentences: “You are committed to such and such an implication of your assertion, even if you would rather retract your claim than accept the implication,” and “He has incurred a commitment to financially support his illegitimate child regardless of what his attitude toward doing so may be.” The sense of “commitment” in these sentences is clearly normative. It speaks to something that one ought to believe or do given that one believes or has done something else. This is not the notion of commitment in play when the person of integrity is characterized as having firm commitments. To keep these two notions distinct, I propose the locutions “normative commitment” for the former and “attitudinal commitment” or merely “commitment” for the latter. 4. I borrow this term from Lynne McFall, who in turn credits it to John Kekes. See Lynne McFall, “Integrity,” Ethics 98 (1987): 13. The idea that there is an intimate connection between integrity and practical identity is traceable to the influence of Bernard Williams’s remarks on integrity in the context of his “integrity challenge” to utilitarianism. See Bernard Williams, “A Critique of Utilitarianism,” in J.  J.  C.  Smart and Bernard Williams, Utilitarianism: For and Against, 77–150 (Cambridge: Cambridge University Press, 1973), and “Utilitarianism and Moral Self-­indulgence,” in Moral Luck: Philosophical Papers 1973–1980, 40–53 (Cambridge: Cambridge University Press, 1981). 5. Korsgaard, The Sources of Normativity, 101. These are role-­descriptions: her examples are “someone’s friend or lover, or as a member of a family or ethnic group or a nation,” “a Citizen of the Kingdom of Ends,” “the steward of her own interest,” or “the slave of her passions.” See also Korsgaard, Self-­Constitution, especially 19–26. 6. Korsgaard, Self-­Constitution, 21. 7. See my remarks in the final paragraphs of Section 3. 8. I believe it is best to view ascriptions of specific kinds of integrity—say, intellectual, artistic, or moral integrity—as proceeding on the assumption that the person in question has a specific kind of commitment. In the case of intellectual integrity, the assumption is that the person has a commitment that drives her to one kind of inquiry or another, whereas ascriptions of artistic integrity presuppose a commitment to some art on the part of the one appraised, and ascriptions of moral integrity presuppose a commitment to morality on the part of the one appraised. Of course, there is an important difference between the two latter cases. Whereas we do not think that an artistic commitment is normatively required of all people, many of us presume that a moral commitment is universally required, so that everyone ought to be appraisable in terms of moral integrity. The main subject matter of this paper is the character trait ascribed without adding any modifier to “integrity” (although it is sometimes called “personal integrity”).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Integrity, Truth, and Value  171 9. So, it is incompatible with integrity to have a commitment to deceive. That is a commitment that one cannot be disposed to stand by in word and deed as public acknowledgment of this commitment is an act of non-­deception. This suffices to explain why a highly committed conman is not a person of integrity. 10. Does intellectual integrity require that one suspend belief in such cases and merely accept the proposition as a hypothesis on which one proceeds? Is it even possible to retain belief while acknowledging that it is merely based on a hunch? Here is not the space to make the case, but I accept that a theory of belief should make room for the possibility that, at times, one cannot help but believe something even if it is only based on a hunch and one is fully aware of that. In such cases, one’s epistemic responsibility and theoretical rationality is in question. However, if one is honest about the grounds for one’s belief and undertakes an honest inquiry to confirm or disconfirm the belief, I cannot see that believing on a hunch compromises one’s intellectual integrity. Thanks are due to Declan Smithies for pressing me on these points. 11. I am not suggesting that intellectual integrity requires that one be constantly reconsidering one’s beliefs. Such constant scrutiny is more likely to be rooted in lack of confidence or lack of firm convictions than the kind of honesty that is characteristic of a person of integrity. However, firm convictions are compatible with seeking further evidence for them and keeping an open mind for the possibility that more extensive research may call for revision or even rejection of them. This is what the person of integrity is disposed to do, although the strength of that disposition will be inversely related to the strength of the conviction and the disposition will not be manifested in every possible circumstance. 12. On the unconditionality of the commitments of the person of integrity, see McFall, “Integrity,” 11–13. Here is not the space to discuss McFall’s understanding of the relevant sense of unconditionality, but I believe it is too demanding. 13. A second caveat: in making a commitment, one is settling questions as to what to devote oneself. So, in some ways, one is less open-­minded after than before one becomes committed. However, it is implausible that a commitment, even a firm commitment, forecloses the kind of openness to reconsidering one’s commitments that grows out of an awareness of one’s fallibility and rashness. It is incumbent on a theory of commitment to make room for the metaphysical possibility of having a firm commitment yet, aware of one’s fallibility or rashness, submitting its intentional object to an open-­minded scrutiny, even if foreseeably this may lead to a change in attitude. (This is parallel to the need for a theory of conviction to make room for the possibility of having a firm conviction—that is, being quite settled on what to believe—yet, aware of one’s fallibility, conducting further research, even if foreseeably this may lead to belief revision; cf., note 11.) None of this is to deny that constant scrutiny of one’s commitments may indicate lack of confidence in them and even a lack of firm commitments rather than open-­mindedness and thoughtfulness. 14. The examination does not require that the relevant convictions or commitments be suspended. Indeed, if they are firm, it will be psychologically hard if not impossible to suspend them prior to uncovering any strong reasons for doing so. Moreover, it will make no sense, at least from the perspective of the subject, to do so. That, however, does not mean that it is impossible or makes little sense to examine challenges to one’s convictions or commitments, nor does it mean that it makes little sense for creatures,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

172  Sigrún Svavarsdóttir aware of their own fallibility, ever to put their convictions or commitments to the test provided by increased knowledge and understanding of their subject matter. 15. Indeed, there is a question concerning how much precision is feasible, even desirable, when specifying the thoughtfulness condition as well as the other dimensions of assessment. As I said in Section 1, I suspect that it is for a good reason that imprecision or vagueness is a pervasive characteristic of evaluative concepts and that this characteristic should be preserved. 16. It may be thought that this points to a connection between integrity and authenticity, which I am ignoring. Alas, I do not have the space to address this thorny issue adequately. I find the notion of authenticity, especially when applied to a person, extremely elusive. I have a sense of what it means to accuse someone of lacking cultural authenticity, which may or may not speak to her integrity, depending on whether she is committed to be of the culture referenced. However, I don’t really understand what it is to accuse someone of being inauthentic without any further (implicit or explicit) qualification. If it simply means that she is not true to her own commitments and convictions, then I see the connection to integrity, but it does not add anything new to what has been said on the subject. 17. Pace Korsgaard, I do not take this to be the same as figuring out a role-­description under which one values oneself. 18. By “suitability” here, I do not mean the mere possibility of functioning as the intentional object of such an attitude. For an examination of the worth-­relation and the role of value judgments in the critical scrutiny of attitudes, see Sigrún Svavarsdóttir, “Having Value and Being Worth Valuing,” The Journal of Philosophy 111 (2014): 84–109. 19. On a de dicto but not de re reading of a mental attitude ascription, the verb that designates the attitude introduces an opaque context: that is, within that context, sentences that have the same truth-­conditions are not substitutable without potentially affecting the truth of the attitude ascription. A slightly more theoretically loaded and hence more controversial exposition of the distinction is as follows: on the de dicto reading, the sentential complement of the verb (the sentence that embeds in contexts like “desire that” and “believes that”) does not merely state the truth-­condition of the belief ascribed or the satisfiability-­condition of the desire ascribed, it also expresses the agent’s way of conceiving (or her mode of presenting) what she believes or desires. 20. Talk about a de dicto mental attitude should be understood as shorthand for talking about a particular mental attitude ascription to an agent being true on a de dicto reading. 21. I give a partial defense of a view of value inquiry along these lines in Svavarsdóttir, ‘Detecting Value with Motivational Responses,” in Gunnar Björnsson, Caj Strandberg, Ragnar Francén Olinder, John Eriksson, and Fredrik Björklund, eds., Motivational Internalism, ch. 11 (Oxford: Oxford University Press, 2015). (By “empathetic understanding,” I have in mind an emotionally and motivationally engaged understanding: a vicarious experience of valuing.) 22. I am not the first one to consider how to distinguish integrity from fanaticism. See Mark  S.  Halfon, Integrity: A Philosophical Inquiry (Philadelphia, PA: Temple

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

Integrity, Truth, and Value  173 University Press, 1989), 63–70; Mark R. Hébert, “Integrity, Identity, and Fanaticism,” Contemporary Philosophy 24 (2002): 25–9; and Greg Scherkoske, Integrity and the Virtues of Reason: Leading a Convincing Life (Cambridge: Cambridge University Press, 2013), 7. 23. Or as McFall puts the suggestion (crediting Robert Koons): “Integrity is what you attribute to a fanatic with whom you agree”; see Lynne McFall, “Review of Integrity: A Philosophical Inquiry by Mark S. Halfon,” The Philosophical Review 101 (1992): 464. 24. Cheshire Calhoun, “Standing for Something,” The Journal of Philosophy 92 (1995): 248, n. 19. 25. Why not develop the thoughtfulness condition such as to require emotionally detached, hard, honest thought about the object of commitment? Here is not the space to give a proper answer to this question but, in a nutshell, I am skeptical that a fully emotionally detached reflection can lead to a change in commitment that amounts to the agent having revised her commitment. 26. For a defense of the position that integrity requires morally good principles or at least good, though not infallible, moral judgment, see Jody  L.  Graham, “Does Integrity Require Moral Goodness?” Ratio 14 (2001): 234–51. 27. Thomas Hurka’s recursive account of virtue is a particularly interesting example of this kind of approach. On his account, a character trait is a virtue insofar as it involves attitudes that amount to loving what is good and hating what is evil. As I understand Hurka, the agent need not love what is good under the guise of the good or hate what is evil under the guise of the bad. Instead, she must love what is good and hate what is bad under guises that appear on the list of values and disvalues in the base clause of the recursive account. See Thomas Hurka, Virtue, Vice, and Value (Oxford: Oxford University Press, 2001), esp. 11–20. Not surprisingly, this informs the account Hurka, along with Daniel Y. Eldstein, has given of what it is to act with integrity: “An initial analysis of ‘x is an act of integrity’ . . .runs something like ‘x is good, and x involves an agent’s sticking to a significantly good goal despite distractions and temptations, where this property makes any act that has it good,’ . . . ”: Daneil Y. Eldstein and Thomas Hurka, “From Thick to Thin: Two Moral Reduction Plans,” Canadian Journal of Philosophy 39 (2009): 526. 28. Williams, “Utilitarianism and Moral Self-­Indulgence,” 49. 29. In Robert Adams’s terminology, it may be a structural virtue rather than a motivational virtue as Adams himself suggests. See Robert Merrihew Adams, A Theory of Virtue: Excellence in Being for the Good (Oxford: Oxford University Press, 2006), 37 and 195. 30. Earlier versions of this paper were presented at Syracuse University, the NIH’s Department of Bioethics, Tufts University, Boston University, and the Harvard Philosophy and Law Colloquium. A more recent version was presented at University of Michigan and Stockholm University. I thank all these audiences for helpful discussions. Special thanks are due to my commentator at the University of Michigan, Mercy Corredor, as well as Declan Smithies, Selim Berker, and the editors of this volume for helpful written comments. Finally, I thank Justin D’Arms, Nalini Bhushan, Erin Kelly, and the late Amelie Rorty for stimulating conversations about integrity.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 20/05/22, SPi

174  Sigrún Svavarsdóttir

Bibliography Adams, Robert Merrihew. A Theory of Virtue: Excellence in Being for the Good. Oxford: Oxford University Press, 2006. Calhoun, Cheshire. “Standing for Something.” The Journal of Philosophy 92 (1995): 235–60. Eldstein, Daniel  Y. and Thomas Hurka. “From Thick to Thin: Two Moral Reduction Plans.” Canadian Journal of Philosophy 39 (2009): 515–36. Graham, Jody  L. “Does Integrity Require Moral Goodness?” Ratio 14 (2001): 234–51. Halfon, Mark  S. Integrity: A Philosophical Inquiry. Philadelphia, PA: Temple University Press, 1989. Hébert, Mark  R. “Integrity, Identity, and Fanaticism.” Contemporary Philosophy 24 (2002): 25–9. Hurka, Thomas. Virtue, Vice, and Value. Oxford: Oxford University Press, 2001. Korsgaard, Christine  M. The Sources of Normativity. Cambridge: Cambridge University Press, 1996. Korsgaard, Christine M. Self-Constitution: Agency, Identity, and Integrity. Oxford: Oxford University Press, 2009. McFall, Lynne. “Integrity.” Ethics 98 (1987): 5–20. McFall, Lynne. “Review of Integrity: A Philosophical Inquiry by Mark S. Halfon.” The Philosophical Review 101 (1992): 463–5. Scherkoske, Greg. Integrity and the Virtues of Reason: Leading a Convincing Life. Cambridge: Cambridge University Press, 2013. Svavarsdóttir, Sigrún. “Having Value and Being Worth Valuing.” The Journal of Philosophy 111 (2014): 84–109. Svavarsdóttir, Sigrún. “Coherence of Attitudes, Integration of the Self, and Personal Integrity.” In David Shoemaker, ed., Oxford Studies in Agency and Responsibility, vol. 3, 62–84. Oxford: Oxford University Press, 2015. Svavarsdóttir, Sigrún. “Detecting Value with Motivational Responses.” In Gunnar Björnsson, Caj Strandberg, Ragnar Francén Olinder, John Eriksson, and Fredrik Björklund, eds., Motivational Internalism, 213–36. Oxford: Oxford University Press, 2015. Williams, Bernard. “A Critique of Utilitarianism.” In J. J. C. Smart and Bernard Williams, Utilitarianism: For and Against, 77–150. Cambridge: Cambridge University Press, 1973. Williams, Bernard. “Utilitarianism and Moral Self-indulgence.” In Moral Luck: Philosophical Papers 1973–1980, 40–53. Cambridge: Cambridge University Press, 1981.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

8 Shadows of the Self Reflections on the Authority of Advance Directives Japa Pallikkathayil

People are commonly taken to have the authority to issue advance directives governing their medical care should they cease to be competent. My aim in this paper is to challenge the circumstances and extent to which we should regard individuals as having authority over treatment decisions made for a future state of incompetence. I focus on the loss of competence due to dementia as an important case study. Around fifty million people worldwide are currently living with dementia, and that number is expected to rise to 152 million by 2050.1 In the course of this discussion, I consider the implications of my account for other circumstances of incompetence. An adequate account of decision-­making for those with dementia should be able to do justice to two related aspects of that condition. First, notice the way in which, for practical purposes, sufferers of dementia both are and are not the same people they were before. Consider reflections of both caregivers and sufferers: I find myself writing in the past tense, although Mom is still living. I suppose I do so because of Mom’s condition now and how different she is. But although very different, mentally and physically, from what I describe above, she is still Mom. This is simply another phase of her life. However, I do miss her and it does tend to feel a little like bereavement.2 When shunned for her behaviour, Minnie would say that it was not her but the other woman! (It is true, from her perspective, that it was acted out by the disease not by her per se.)3 Mother has for some time referred to herself in the third person.4

Japa Pallikkathayil, Shadows of the Self: Reflections on the Authority of Advance Directives In: Normativity and Agency: Themes from the Philosophy of Christine M. Korsgaard. Edited by: Tamar Schapiro, Kyla Ebels-­Duggan, and Sharon Street, Oxford University Press. © Japa Pallikkathayil 2022. DOI: 10.1093/oso/9780198843726.003.0008

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

176  Japa Pallikkathayil My wife gets frustrated with me . . . and she is right to be frustrated. She asks me to put a can in the recycling . . . and I don’t do it. She says, “I know this is because of your illness, that this is not you.”5

Second, notice the way in which the “otherness” of the demented self reflects an unraveling rather than a becoming. The demented self is not a new self but rather the same old self in the process of disintegrating. I am going to begin by exploring this phenomenon. I will suggest that Christine Korsgaard’s treatment of personal identity provides a helpful framework for understanding this phenomenon. I will then give an account of decision-­making for dementia sufferers that is appropriately responsive to this phenomenon.

1.  Identity and the Broken Self It will be helpful to begin by considering a related but distinct challenge to the authority of advance directives in cases of dementia. Some worry that in cases of dementia the psychological continuity necessary for the persistence of a person is disrupted, leaving another person in her place. In these cases, an advance directive issued by a person who has ceased to exist has no authority over the treatment of the new individual. Let us call this the Other Person Problem.6 As I suggested above, I do not think this is the right way to conceptualize the otherness of the demented self. But examining responses to this worry will provide some important guidance about how to think about personal identity. I am going to consider two ways of responding to the Other Person Problem. First, one might claim that the kind of psychological continuity necessary for personal identity is not actually disrupted by dementia. Second, one might claim that psychological continuity is not actually necessary for personal identity after all. Let us consider each of these strategies in turn. Allen Buchanan takes up the first strategy, arguing that we ought to take the degree of psychological continuity necessary for persistence to be so low that we should take cases of progressive dementia to be ones in which the same person persists until there is no person at all.7 What remains after the capacities for personhood are lost may still be a living being with the capacity for pleasure and pain. In this case, the interests of this non-­person may need to be weighed against the surviving interests of the person who has ceased to exist, interests, for example, in what happens to her living remains. But this will not be a case in which respecting an advance directive would involve

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Shadows of the Self  177 subjecting a person to another person’s treatment decisions. Furthermore, in many cases, respecting an advance directive would not involve violating any obligations to the remaining non-­person given the radically truncated nature of that being’s interests. The painless termination of life support, for example, would not be in tension with such a being’s interest in the nature of her experiential states.8 This response to the Other Person Problem invites two questions. First, given that psychological continuity comes in degrees, why take it that there should be a threshold above which identity claims are taken as settled? Second, supposing there should be some threshold, why take it to be as low as in Buchanan’s argument? Buchanan’s responses to both questions invoke the practical implications of a conception of personal identity. In response to the first question, Buchanan notes epistemological problems about how to distinguish degrees of psychological continuity, and coordination problems that would be created by expecting people to follow rules nuanced enough to be sensitive to degrees of continuity.9 In response to the second question about why the threshold should be low, Buchanan argues: Some of our most important social practices and institutions—those dealing with contracts, promises, civil and criminal liability, and the assignment of moral praise and blame—apparently presuppose a view of personal identity according to which a person can survive quite radical psychological changes and hence a high degree of psychological discontinuity. If this is so, then given the value of these practices and institutions, we would have to have extraordinarily weighty reasons for giving up the view of personal identity upon which they are founded.10

Buchanan argues that there are no such weighty reasons. Moreover, although we might be able to reconstruct some of our practices in light of a high threshold, we would face the daunting challenge of responding to a vast new problem of intergenerational justice in light of “the ‘births’ of large numbers of ‘new persons’ who would as it were spring full-­blown into the world and who would not, strictly speaking, be the sons, daughters, husbands, wives, or friends of anyone.”11 And Buchanan suggests that nothing about the view that psychological continuity is necessary for identity forces on us such a radical revision in our thinking. There are, then, two steps in Buchanan’s defense of a low threshold. First, many of our practices presuppose that a person can survive radical

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

178  Japa Pallikkathayil psychological changes. And second, if we reconstructed these practices in light of requiring a high degree of psychological continuity for personal identity, we would face a serious practical problem about what the “new people” these new practices acknowledge are owed. I suggest, however, that both steps in this reply reveal a methodological tension in Buchanan’s view. In the first step, Buchanan relies on the presuppositions of our practices to support a low threshold. But there are at least some ways in which our practices also seem to presuppose that a person survives even through the late stages of dementia, the stages in which Buchanan holds that the person is gone. For example, we typically take familial obligations to the demented to last even through the late stages of that condition. Thus the presuppositions of at least some of our practices seem to indict requiring even a low degree of psychological continuity. In the second step of his argument, Buchanan points to a practical problem that revised practices employing a high threshold would face, namely, what to do about all the “new people.” But Buchanan’s own view faces a version of the same problem since it leaves the severely demented untethered in networks of social relationships. Although the severely demented are not persons on Buchanan’s view, we still face questions about who is responsible for their care, questions that have no straightforward answers if we take them to be “new beings.” The methodological tension in Buchanan’s view thus lies in arguing against a high threshold by pointing to practical presuppositions and problems while allowing some practical revisions and accepting some practical problems for the sake of a low threshold. I suspect this tension arises because Buchanan begins his paper by assuming for the sake of argument that some degree of psychological continuity is necessary for personal identity. This assumption is thus not subjected to the same practical constraints as the competing proposals about how to understand this requirement. To be clear, I do not intend this criticism of Buchanan’s view to be a criticism of a psychological continuity requirement on personal identity. We will consider that requirement more closely towards the end of this section. Rather, what I take this discussion to show is that reflection on the relationship between a conception of personal identity and our practices must begin at an earlier stage than Buchanan’s. Considering the second of the possible responses to the Other Person Problem will take us a step closer to doing that. Ronald Dworkin takes up this strategy, denying that psychological continuity is necessary for personal identity. Dworkin takes what he calls the “assimilationist” approach to personal identity. This approach begins by working

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Shadows of the Self  179 out the conception of personal identity that is implicit in our moral and social practices. When we encounter novel cases in which claims about identity are difficult to adjudicate, we should try to extend or adapt our conception of identity “with an eye to the efficiency and fairness of the public system of prudential concern and temporal co-­responsibility.”12 For example, if “travel” by replicating one’s body and destroying the original became commonplace, Dworkin argues that we would take this procedure to preserve personal identity since doing so would be least disruptive to our practices. Dworkin doubts that any such extension of our ordinary, implicit understanding of personal identity is needed in cases of dementia. He takes the judgment that the life of a single person may have a demented stage to be already firmly rooted in our public understanding of personal identity, an understanding that reflects the following rule of thumb: “if stages of human experience are connected by either of the two main continuities—physical or psychological—then we should assume continued personal identity, for the purposes of the twin assumptions of prudential concern and temporal co-­ responsibility, unless this would be irrational because prudence could take no hold or unjust because co-­responsibility is unfair.”13 Although psychological continuity may be disrupted in cases of dementia, physical continuity is maintained. So, Dworkin sets out to show why neither of the excepting conditions obtains in this case. Let us begin by considering whether prudential concern is appropriate in such cases. Dworkin seems to take prudential concern to be concern about “the character and value of [one’s] life as a whole.”14 And he offers as an example of when prudential concern would be “unnatural and unstructured” the following: “someone told he must be prudent for the combination of himself and someone else, whose fate he cannot connect with the value of his own life, has no way to be prudent as distinct from altruistic.”15 In contrast, he argues that the concern that people have for what happens during the demented stages of what he proposes to consider their later lives is easily construed as concern for the character and value of their lives as a whole. This concern has the same character as many other familiar prudential concerns, for example, concerns about the manner of one’s death and concerns about what happens after one’s death, including the success or failure of one’s projects and how one is remembered. Dworkin argues that we could not regard prudential concern for a demented stage of one’s life as irrational without calling these other prudential concerns into question. Notice two problems with this argument, both stemming from the conception of prudential concern on which Dworkin relies. First, since Dworkin

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

180  Japa Pallikkathayil takes prudential concern to be appropriately directed at events after one’s death, noticing that some event is an appropriate object of prudential concern is consistent with denying that this event is an episode in the life of a single, persisting person. One might, for example, have prudential concern in Dworkin’s sense for what happens to one’s children after one’s death. But given where Dworkin places the burden of proof in his argument this might not be very problematic for his purposes—prudential concern gives us no reason for doubting the identity of the demented self and the competent self even if it also does not give us a positive reason for affirming that identity. The deeper and more pressing problem for Dworkin’s argument lies in the passive understanding of prudential concern. Dworkin is, at least here, primarily focused on your concern about what happens to you. But that is in many ways not the primary focus of prudential thought. Of course it matters to you how factors beyond your control shape the character and value of your life as a whole. But much of your prudential thinking is not focused on those factors but instead on how you shape the character and value of your life as a whole. You do this by making choices that unfold in time and thereby draw together your earlier and later selves. But this is precisely the thing you can no longer do once dementia looms in the future. You might now make plans for what will happen to your demented self, much like you might make plans for the disposition of your estate after your death. But you cannot make plans for your demented self to carry out. Your thinking will not be integrated with hers in the right way. And that constitutes a profound break with the kind of prudential thinking most of us are engaged in most of the time. I will return to this point shortly. But first let us consider the second of the two excepting conditions that Dworkin takes it would tell against identity in cases of dementia: the unfairness of “co-­responsibility.” By this, I take it he means the potential unfairness of holding someone accountable for earlier choices. In particular, we normally take people’s current opportunities to depend in certain ways on their past choices. Dworkin suggests the following as a case in which that would be unfair. Suppose that my body divided into two identical bodies with the same memories and mental life. Dworkin suggests that if I knew this would happen I might coherently have prudential concern for both of the lives that would follow my division, and I might even take measures to provide for them both. But requiring this “might well be thought unjust, because I would then have either to cheat my life before I was divided, in order to make available enough for two lives later, or these lives would have to share what would normally do for one.”16 Dworkin asserts that there is no comparable injustice in making one’s provisions as a demented

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Shadows of the Self  181 person depend on the prior choices of one’s competent self, though he does little to argue for this claim.17 Since this aspect of responsibility over time is rather removed from my main focus, let us grant Dworkin for the sake of argument that at least this aspect of co-­responsibility is not undermined by dementia. Nevertheless, I suggest that this is not the only important aspect of responsibility through time that dementia might threaten. Consider whether it would be fair to hold your demented self accountable for keeping promises you have made, promises that you can no longer remember or no longer have the capacities needed to fulfill. This seems like a paradigmatic case in which you would be excused from keeping promises. The judgment of excuse rather than irrelevance suggests a single, persisting person. But notice that as soon as you know this excusing condition looms in the future, your ability to make promises is compromised. This is the interpersonal analogue of the problem for prudential reasoning noted above. Dementia threatens your ability to make both selfand other-­directed commitments. And this marks a profound break with your future self—that is, a self who you cannot obligate. When dementia looms in the future, your choices do not have the significance for your future self that they otherwise would. The same is true in the other direction. The choices of your future demented self also do not have the same significance for you that they would otherwise. Contrast two cases. Suppose that, for most of his life, Edward has worked to support equality. But after a series of misfortunes later in life, Edward becomes a vocal racist. Suppose next that Melanie has also worked to support equality for most of her life. But as Melanie slips into dementia, she begins to refer to individuals of certain racial groups pejoratively. She even begins referring to them with slurs. In these cases, Edward responds poorly to the ups and downs of life whereas Melanie’s mind is degenerating. For this reason, we might wholeheartedly say, “What a shame Edward became such a racist in later life,” while we may hesitate to say the same of Melanie. We might, as in some of the quotations in the introduction, suggest that Melanie’s racism is “the disease talking.” We must be careful with this point. There is a straightforward sense in which Melanie is doing the talking and foreseeing that one might behave in these ways is one of the horrors of anticipating dementia. But Melanie may not be blameworthy for her racism in the way that Edward is. Perhaps her new views are the product of increasing paranoia and confusion. Or perhaps the views are not so new after all but instead reflect a revealing of her true self brought about by the loss of inhibitions. Especially in this latter case, Melanie’s actions may not be wholly excused by her dementia. But even if she has really

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

182  Japa Pallikkathayil always been a racist deep down, the way in which dementia loosens her tongue mitigates her responsibility for her actions. And in this way, Melanie’s actions may not reflect poorly on the person she was before in the way that Edward’s do. Let us take stock. Dworkin highlights two important dimensions of our lives in which judgments about identity figure: our prudential reasoning and our responsibility for our choices. But in neither case are the judgments of identity as firm in cases of dementia as Dworkin seems to think. There is a noteworthy way in which your prudential concern for your future demented self has something of the character of estate planning, a kind of planning that is predicated on the assumption of your absence. And judgments about responsibility are likewise a mixed bag. You cannot obligate your future demented self and her actions may not impugn your character. For all that, I do not think this is good reason for taking the demented self to be a new person. As I noted at the outset, I think that the otherness of the demented self is best conceptualized as a different kind of disruption of identity. Rather than the Other Person Problem, we should consider what I will call the Broken Person Problem: the demented self is a self in the process of falling apart. And I suggest that Christine Korsgaard’s work on personal identity can help us appreciate the nature of this problem. Korsgaard’s work will enable us to overcome three shortcomings in Buchanan’s and Dworkin’s accounts. First, I take Korsgaard’s view to provide a response to a worry about the practically focused methodology both Buchanan and Dworkin employ, namely that it is overly conservative with respect to our practices. Second, Korsgaard’s view will enable us to appreciate the kind of psychological continuity that is relevant to personal identity, an important nuance that Buchanan’s view misses. Finally, Korsgaard’s view will enable us to begin to understand how and why bodily continuity is important for identity, a question which Dworkin’s disjunctive criteria for personal identity leaves unanswered. Korsgaard is focused on the way in which agency requires us to think of ourselves as unified. Action requires resolving conflicts between one’s motivational states.18 The deliberative standpoint from which you do this requires that you identify yourself as something other than any of these desires. Your actions then take you forward through time: [T]he choice of any action, no matter how trivial, takes you some way into the future. And to the extent that you regulate your choices by identifying yourself as the one who is implementing something like a particular plan of life, you need to identify with your future in order to be what you are even

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Shadows of the Self  183 now. When the person is viewed as an agent, no clear content can be given to the idea of a merely present self.19

Action thus requires one to think of oneself as unified both synchronically and diachronically. Many of the practices Buchanan identifies as involving a conception of personal identity are ones that we might coherently consider revising. Dworkin’s focus on prudential concern and co-­responsibility brings us to practices that are more deeply embedded in our thinking. But these ways of thinking might nonetheless be called into question, as Derek Parfit, for example, does.20 The kind of identity judgments Korsgaard’s practical reflections suggest, however, have something more like an inescapable character. Acting requires thinking of oneself as persisting through time. And it is not clear how we could coherently cease doing that. As Korsgaard puts it: “There is a necessary connection between agency and unity which requires no metaphysical support.”21 In this way, the practically focused methodology Korsgaard employs is not subject to a worry that the practical concerns she relies on reflect an undermotivated conservativism about our practical thinking. Consider, then, the kind of psychological continuity that is important for identity on this view: “You are not a different person just because you are very different. Authorial psychological connectedness is consistent with drastic changes, provided these changes are the result of actions by the person herself or reactions for which she is responsible.”22 With this in mind, return to Buchanan’s observation that many of our practices presuppose the possibility of persisting through radical psychological changes. But our practices are sensitive to the etiology of those changes. Consider again our responsibility for our actions. Edward’s racism is attributable to him because it reflects the development of his thinking. Melanie’s racism may not be attributable to her because it may reflect a malfunction in her thinking. In this way, the agential account of identity Korsgaard proposes is able to account for nuances in our downstream practices, suggesting that these practices are in these respects well-­motivated. But as Dworkin noted, there are other aspects of our practices that treat Melanie as persisting even in the late stages of dementia. Is this an aspect of our practices that can be vindicated on Korsgaard’s account? It may initially appear as though not. Melanie is already at a stage in which diachronic unity is a problem for her. Her thinking is not properly integrated with the thinking of her past and future selves. And her progression through dementia may eventually problematize even synchronic unity for her. She may be reduced to

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

184  Japa Pallikkathayil competing impulses with nothing left to adjudicate between them. It may seem, then, that this account pushes us to conclude that Melanie becomes a series of new persons until she ceases to be a person at all. Korsgaard’s take on the role the body plays in identity over time makes this issue even more pressing. She takes a physical criterion of identity to be contingently correct for us because, as things stand, the body is the basic kind of agent. But things might be different if technology were different. Suppose, for example, that technological advances allowed us to replicate our bodies, destroying the original, once every year after age thirty in order to preserve our youth.23 Korsgaard suggests that this series of bodies would be a single, persisting person because she “would be able to carry out unified plans and projects, and have ongoing relations with other persons.”24 I imagine she would say the same about transportation by replication. But if we are identified with our bodies only so long as they are the basic site of agency, it is especially unclear why we would take identity to hold through the breakdown of agency in cases like Melanie’s. There is something right about this line of thought, but I believe it overlooks an important aspect of the relationship between the body and agency. Agency is an element of a properly functioning mature human body.25 Our bodies set us the task of coordinating our impulses and sensations in a way that results in action. Even when a body cannot sustain agency, these impulses and sensations—the raw materials of agency—may remain. And when they do, what we have is not a new person or a new being. It is something broken. Something may be more or less broken. Compare a drinking glass that has crack allowing water to leak out with a glass that has shattered. Likewise, agents may be more or less broken. A person in the early and middle stages of dementia may still be able to act. But diachronic unity is a problem for such a person, and so her ability to complete drawn-­out or complex actions may be dwindling. And as I noted above, in the later stages of dementia, even synchronic unity may become a problem. I do not think it is impossible to imagine psychological breaks in a single body so sharp, complete, and permanent that they are best understood as bringing about the end of one agent and the beginning of another. So, just as we may in principle outlast our bodies, our bodies may in principle outlast us, and not simply in vegetative states but as new agents. But this is not a circumstance we presently encounter in real life. Elements of agency persist even through radical neurological impairment. Those experiencing retrograde amnesia, for example, may retain procedural memory.26 And even if in the late stages of dementia all that remains is a wash of experiences, these

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Shadows of the Self  185 experiences are the product of the same agential capacities for responding to one’s environment that have been present all along. So, in real life we do not encounter new agents inhabiting old bodies. We encounter broken agents. With this description of the phenomenon in hand, we can turn to considering how we ought to respond to it.

2.  The Authority of Advance Directives As I noted at the outset, people are commonly taken to have the authority to issue advance directives governing their medical care should they cease to be competent. It will be helpful to begin by considering Dworkin’s very influential defense of this position. He begins by asking why we should ever respect the decisions people make when we believe that these decisions are not in their best interests. According to what Dworkin calls the evidentiary view, “we should respect the decisions people make for themselves, even when we regard these decisions as imprudent, because each person generally knows what is in his own best interests better than anyone else.”27 But as Dworkin points out, this view will have trouble explaining why we would defer to people’s akratic or altruistic decisions. So, Dworkin instead proposes what he calls the integrity view, according to which the value of autonomy “derives from the capacity it protects: the capacity to express one’s own character—values, commitments, convictions, and critical as well as experiential interests— in the life one leads. Recognizing an individual right of autonomy makes self-­creation possible.”28 Dworkin holds that the seriously demented may lack this capacity, and so have no right to autonomy. But the precedent autonomy of such individuals should still be respected: “A competent person making a living will providing for his treatment if he becomes demented is making exactly the kind of judgment that autonomy, on the integrity view, most respects: a judgment about the overall shape of the kind of life he wants to have led.”29 This deference to one’s past, competent decisions extends even to situations in which those decisions contradict one’s present demented desires. Dworkin uses a case of temporary derangement to help motivate this view. Suppose a Jehovah’s Witness has an advance directive forbidding blood transfusions. If he ends up needing such a transfusion he may countermand his previous decision so long as he remains competent, even if such a reversal would reflect weakness of will. But in a state of derangement, he will not have the authority to change his mind: “Suppose we were confident that the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

186  Japa Pallikkathayil deranged Witness, were he to receive the transfusion and live, would become competent again and be appalled at having had a treatment he believed worse for him than dying. In those circumstances, I believe, we would violate his autonomy by giving him the transfusion.”30 Seana Shiffrin argues, however, that here Dworkin is overlooking an important difference between cases of temporary and permanent derangement: It is one thing to refuse to listen to those who have temporarily lost their senses . . . and will have to live with our action. We might think metaphorically about their deranged behavior in terms of an alien self who has temporarily commandeered their bodies. We decline to obey them because a very different personality with a different form of judgment will have to live with the effects of their temporary yet powerful reign. It is both that they will regret the decision and that the decision was made under conditions that were not true to themselves. In the case of the permanently demented, however, the “real” self will not return and be forced to live with the consequences of a temporary period of insanity.31

As I will go on to argue, I think there is an important difference between many cases of temporary and permanent impairment. But Shiffrin focuses on the wrong difference. This can be brought out by thinking about another case. Suppose you decide to go through a serious surgery because you very much want to live. You have been made aware of what the surgery will involve, including that you will be intubated and will need to remain intubated for some time after you wake up. You also know the medications you will be on when you wake up will confuse your thinking. After the surgery, you do indeed wake up confused and you begin trying to rip out your breathing tube. If we accede to your deranged wishes, your competent self will not return and have to live with the consequences in Shiffrin’s sense since you will be dead. Yet this seems like a case in which it would be entirely reasonable to defer to the decisions of your competent self. Dworkin’s own objection to Shiffrin’s argument emphasizes a somewhat different problem. He maintains that Shiffrin fails to take seriously the identity of the competent and demented individual.32 Even in cases of permanent dementia, the earlier competent self has to live with the consequences of what happens to her when she is demented. Those choices affect the character and value of her life as a whole, the very same life as the demented person’s. I take this objection to Shiffrin to have some force. It is striking that she takes respect for autonomy to provide no reason to implement advance directives in cases of permanent dementia. And it is hard to see how that

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Shadows of the Self  187 could be consistent with taking dementia to be the last stage of the competent person’s own life.33 But it also seems as though Dworkin’s view might be charged with a similar mistake. On Dworkin’s view, there seems to be no significant difference in the force of an advance directive in cases in which the person is in a permanent vegetative state and in cases in which the person is permanently demented. To be sure, in the latter case, the person may still have interests in the quality of her experiences. But the value of those experiences is taken to have already been accounted for by her previously competent self. Whether some outcome is or is not worth suffering for is a judgment that has already been rendered. Thus, on Dworkin’s view, even when considering what is required by beneficence rather than autonomy, the point of view of the demented individual is practically inert. In this way, despite his in­sist­ ence that the same person persists through dementia, Dworkin treats a person as gone as soon as competence is permanently lost. Shiffrin and Dworkin both fail to appreciate the sense in which a demented agent is an agent in the process of falling apart. Shiffrin fails to take a competent person’s decisions to have any import for how demented individuals should be treated. Dworkin fails to take the exercises of agency of which the demented self is still capable to be worthy of any respect. In the next section, I offer a view that does better on both of these fronts.

3.  Rethinking Advance Directives To begin, let us consider a different kind of temporary impairment than the one that Dworkin and Shiffrin have in mind. Suppose Greg is a devout Jehovah’s Witness who has had a traumatic brain injury. Greg did not need a transfusion in the course of his treatment for this injury. His treatment has proceeded in accordance with the instructions on the advance directive card he carried with him. As Greg’s recovery begins, the extent of his injuries become apparent. Greg struggles with some memory loss and some difficulty forming new memories. He has lost facility with many concepts and must relearn many basic skills. Doctors believe that he may eventually regain enough of his mental faculties to be competent to make his own medical decisions, but he is not presently so and it may be many months, and maybe even a year or two before he improves to that extent. Greg has a generally positive attitude toward his present circumstances. Although he experiences some understandable frustrations during his therapy, he takes pride in his accomplishments and looks forward to conquering further challenges. He  also enjoys the time he spends talking with his family and friends.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

188  Japa Pallikkathayil On the other hand, although Greg remembers some aspects of his religious beliefs, they now strike him as confusing and somewhat frightening. And he expresses a preference not to participate in the religious practices that were once very important to him. Now suppose that Greg has an accident resulting in massive blood loss and a transfusion is needed to save his life. Should his caregivers still defer to the decisions of his earlier, competent self? This seems like a much harder question to answer than when we imagined a Jehovah’s Witness in a fit of derangement, as in the case Dworkin and Shiffrin considered. I suspect that is because Greg is living a life that includes exercises of agency that are worthy of respect. Greg understands the basic goals of therapy and embraces them. Greg participates in his relationships with others. These are not the ravings of an unhinged person. At the same time, the religious convictions that motivated his earlier competent decisions have no place in his present point of view. His previous treatment instructions are thus at odds with the life he is now living. One might be tempted to explain hesitation about Greg’s case in a different way. Although Greg’s competence is expected to return, he may very well have different values by then. Greg’s case suggests the possibility of two rather different competent points of view. But that possibility was present even in the case of the deranged Witness. Suppose we knew, as is not implausible, that near-­death experiences unsettle people’s values in some non-­trivial number of cases. We might then think it is an open question whether the Witness would indeed by appalled if his competence returned after a transfusion is performed in defiance of his earlier decisions. But I do not think that the openness of that question weakens the case for respecting the Witness’s advance directive in that case. The right to live one’s life as one sees fit is supposed to protect even choices that may very well be regretted. Let us return, then, to the thought that the key difference between the original case of the deranged Witness and Greg’s case is that the latter but not the former is living a life that includes exercises of agency that are worthy of respect. Greg thus has a point of view that may suggest a course of treatment even if he does not understand his treatment options well enough to be able to decide between them. And this difference between the cases makes no reference to Greg’s prospects for recovering competence. I therefore take that feature of Greg’s case to be inessential. And I take this to reveal that the important difference is not between cases of temporary and permanent impairment, but between cases in which respect-­worthy exercises of agency remain and those in which they do not. So far, this has just been a diagnosis of why we might hesitate to abide by Greg’s earlier competent decisions in his present circumstances. Now I turn to

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Shadows of the Self  189 arguing that this hesitation is well-­founded. This will involve gesturing toward a somewhat different reason for deferring to people’s choices than Dworkin’s.34 Dworkin emphasizes the value of expressions of character that reflect one’s values and commitments, especially those that are oriented toward enacting a view of one’s life as a whole. Although I do not doubt that there is value in such expressions, I doubt that this is the value that is operative in our deference to people’s choices. Rather, that deference is primarily oriented toward realizing a certain kind of relationship—a relationship of mutual in­de­pend­ ence. I defer to your choices because I recognize that you and I have the same standing to determine how we each live our lives. And this is true even if we very often fail to live up to our own values and commitments, and rarely act with a view to our lives as a whole.35 Rather, our claim to stand in relationships of mutual independence is grounded in the much more basic capacity to act in ways that are responsive to the reasons we take to bear on our choices. I take this view of the grounds of deference to provide an attractive explanation of the expanding scope of the deference we afford to children as they grow into maturity. One ground for deferring to the choices of children even if we believe them to be misguided is educative. But I do not take this consideration to be exhaustive.36 Instead, notice that it may be appropriate to begin treating children as independent in certain limited domains before it is appropriate to treat them as independent full stop.37 I suggest doing so reflects acknowledgment that their capacity to reason about the relevant domain is adequate to the task. And this can be so even if they are still not capable of envisioning their lives as whole and aiming to bring about lives with a certain character. We may have judgments of the same kind about Greg. There are limited domains within which it is appropriate to defer to Greg’s choices. We have reason, for example, to defer to his choices about when to receive visitors because there is nothing importantly lacking in his ability to think through this matter for himself. He may sometimes regret spending too much or not enough time with those who want to see him. But he has enough sense of what seeing visitors involves to defer to his judgment about this aspect of his life. I take this line of thought to confirm that there really is a sense in which Greg continues to live his life even in his impaired state. He makes decisions that are worthy of our deference. And the point of view that emerges in the course of living his life is one that speaks in favor of saving his life even if that requires transfusion. How then are we to adjudicate between this point of view and his earlier one? Simply deferring to the earlier point of view overlooks the way in which Greg is still living a life. But Greg is not in a position to integrate his current

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

190  Japa Pallikkathayil point of view with his previous one. He has ceased to have the religious beliefs that motivated his advance directive, but not because he rejects them—he does not understand them well enough to do that. This might incline us to think that the task of integration is best left up to Greg’s earlier competent self. But that is also not a self that can complete the task of integration. Greg’s earlier self cannot choose to remain steadfast in his religious convictions even through impairment. Remaining steadfast is something that only a later self can accomplish. Greg never occupies a point of view that is adequate to the task of integrating his earlier and later selves. This is the way in which his condition reflects a breakdown in agency. And this is the reason that independence with respect to this task is inappropriate. Integrating Greg’s earlier and later points of view is something that we have to do for him. The order in which our lives are lived generally privileges our later selves with respect to the task of integration. To be sure, as we saw above, one begins unifying one’s present and future selves by making plans and having commitments. But as the future unfolds, one is in a position to respond to one’s past plans and commitments in a variety of ways, including by rejecting them. One plausible way of integrating Greg’s earlier and later selves involves taking the viewpoint that emerges from the respect-­worthy exercises of agency Greg continues to engage in to constitute a rejection of his previous commitments. This would suggest deferring to his later point of view when making treatment decisions for him. Notice, however, that this is not the only way the task of integration might be completed. Greg’s loss of religious conviction is bound up with confusion about the meaning of religious claims. For this reason, one might see his reluctance to affirm those claims as merely “the brain damage talking.” On this view, although Greg engages in discrete exercises of agency that are worthy of deference, the point of view that emerges in the course of those exercises of agency is not itself worthy of deference. Hence the task of integration does not after all involve reconciling two points of view. Whether we see Greg’s loss of religious conviction as part of a new point of view that is worthy of deference or as a malfunction in his thinking thus has tremendous significance for the treatment decisions that it would be appropriate to make on his behalf. This suggests the following conclusions about the authority of advance directives. The authority of advance directives is limited when one continues to live a life that includes respect-­worthy exercises of agency. In those cases, there is an important question about whether those exercises of agency add

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Shadows of the Self  191 up to a new point of view that is worthy of deference. There may be cases in which the answer to this question is relatively clear. But I take it that answering this question will very often require an exercise of judgment on the part of decision-­makers. And this suggests that in cases like Greg’s the real power the past self has over the future self lies in the choice of a surrogate decision-­ maker who is tasked with interpreting and integrating one’s past and present points of view. I initially presented Greg’s case as one of prolonged but temporary impairment. But I took what was relevant to Greg’s case to be orthogonal to the temporary aspect of his condition. And for that reason, the above reflections suggest that there may be cases in which deference to the point of view of someone suffering from dementia is appropriate even when it conflicts with an earlier competent decision. This may be especially so for certain kinds of dementia. Alzheimer’s disease, for example, is thought to display a pattern of cognitive decline that mirrors cognitive development in children.38 So, just as increasing domains of deference are appropriate for children as they develop, decreasing domains of deference may be appropriate for Alzheimer’s patients as they decline. But that leaves room for the possibility that such a patient may be living a life even when they are no longer competent, a life that may involve a point of view that is at odds with their earlier competent judgments. Suppose Christina thinks that a life in which she is unable to care for herself in many important respects would not be worth living. For that reason, she makes an advance directive indicating that she should not be given life-­ saving treatments if she were permanently in such a condition. She is diagnosed with dementia and as her disease progresses she is able to do less and less on her own. For example, she can no longer bathe or cook for herself. But she expresses minimal frustration with respect to these losses. And she enjoys seeing her family, listening to her favorite music, and eating her favorite meals. She also expresses fear at the prospect of dying. The way Christina lives her life in the limited ways in which she is able to suggests a point of view that treats her life as worth living.39 A decision-­maker might reasonably take this as a ground for deferring to Christina’s later point of view. Though as we saw above, this involves an important interpretative choice. Consider also a case with the opposite structure. Suppose John’s religious convictions lead him to regard life as sacred in a way that is in tension with withholding life-­saving treatments. His advance directive reflects these convictions and requests that such treatments be administered. But during the course of his decline, his wife dies. After that, he ceases to show much interest in the activities of which he is still capable and begins to remark that he would

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

192  Japa Pallikkathayil welcome death. When asked about his religious convictions, he responds, “I don’t know about all that. I just know that I want to die.”40 John here evinces a point of view that is in tension with his advance directive. And the same considerations that suggested deference to later points of view in the previous cases are applicable here too. But not all cases will be like this. The activities of a demented individual may not suggest a stable, coherent point of view to which to defer. And here it will be important to recall that there are not two lives under consideration— one pre- and one post-­dementia. There is just one life. And where there are no competing viewpoints, there is no task of integration. All that remains for us to do in these cases is to respect the agent’s final point of view. Thus, this view continues to countenance the authority of advance directives in an important class of cases.

4. Conclusion I have argued that dementia involves a quite literal falling apart. It is no surprise then that so many people fear and dread such a condition. And a natural response to this fear is to try to take control of the future now—to bind everyone else to ensure that you live the kind of life you think you ought to. But I have tried to show that there are limits to the extent that one can do so. Insofar as you go on living, you cannot freeze yourself in place. The point of view you may one day have may be strikingly different from the point of view you have now. And that point of view may emerge in circumstances in which you are not fully in charge of yourself. The most you can do to prepare for such a future is to entrust your care to people who can try to hold together the parts of a self that is breaking.41

Notes 1. “Dementia,” World Health Organization, September 2, 2021, https://www.who.int/ news-­room/fact-­sheets/detail/dementia. 2. Jennifer Davies, “We Don’t Know What is Going Through her Mind,” in Lucy Whitman, ed., Telling Tales About Dementia: Experiences of Caring (London: Jessica Kingsley Publishers, 2010), 35–6. 3. U Hla Htay, “We Learn to Enter her World,” in Whitman, Telling Tales About Dementia, 60. 4. David Shenk, The Forgetting: Understanding Alzheimer’s; A Biography of a Disease (London: HarperCollins, 2002), 191.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Shadows of the Self  193 5. Shenk, The Forgetting, 32. 6. See Rebecca Dresser, “Life, Death, and Incompetent Patients: Conceptual Infirmities and Hidden Values in the Law,” Arizona Law Review 28 (1986): 373–405, for an early articulation of this problem. 7. Allen Buchanan, “Advance Directives and the Personal Identity Problem,” Philosophy & Public Affairs 17 (1988): 277–302. 8. Buchanan curiously does not consider an advance directive subjecting the remaining non-­person to painful continued existence. In light of this kind of case, it is not clear that there is as little room for conflict between an advance directive and the interests of a remaining non-­person as Buchanan supposes. 9. Note, though, that Buchanan’s final position allows for the consideration of degrees of continuity below the threshold needed for personal identity. That is, above the threshold identity claims are an all-­or-­nothing matter, but below the threshold, we can acknowledge the practical import of degrees of continuity while denying that identity is preserved. Given that this is so, I am not sure how deep the epistemological and coordination problems associated with tracking degrees of continuity are supposed to be. 10. Buchanan, “Advance Directives and the Personal Identity Problem,” 292. 11. Buchanan, “Advance Directives and the Personal Identity Problem,” 293. 12. Ronald Dworkin, “Philosophical Issues Concerning the Rights of Patients Suffering Serious Permanent Dementia,” written for the United States Congress Office of Technology Assessment (1986), 103. 13. Dworkin, “Philosophical Issues,” 104. 14. Dworkin, “Philosophical Issues,” 105. 15. Dworkin, “Philosophical Issues,” 104. 16. Dworkin, “Philosophical Issues,” 107. 17. Some care is needed to understand this claim. Dworkin’s argument is situated in the context of an argument about how the provisions for demented people should depend on the results of a hypothetical insurance scheme in which the preferences of their earlier competent selves figure. So the relationship between the choices of the earlier self and the provisions for the later self is actually rather complicated. But these complications are not important for my purposes. 18. Mere desires may not suffice for action even when they are not in conflict. If so, then there is an even richer role for the deliberative standpoint to play than I have indicated in the main text. Thanks to Tamar Schapiro for bringing this to my attention. 19. Christine  M.  Korsgaard, “Personal Identity and the Unity of Agency: A Kantian Response to Parfit,” in Creating the Kingdom of Ends (Cambridge: Cambridge University Press, 1996), 372. 20. See Derek Parfit, Reasons and Persons (Oxford: Clarendon Press, 1984), chs. 14 and 15. 21. Korsgaard, “Personal Identity and the Unity of Agency,” 374. 22. Korsgaard, “Personal Identity and the Unity of Agency,” 379–80. 23. Korsgaard borrows this thought experiment from Thomas Nagel. 24. Korsgaard, “Personal Identity and the Unity of Agency,” 374. 25. Although not a point Korsgaard draws on in her discussion of personal identity, I take this point to be congenial to her larger project. See, for example, Christine M. Korsgaard, Sources of Normativity (Cambridge: Cambridge University Press, 1996), 149–50.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

194  Japa Pallikkathayil 26. For a helpful discussion of how the fictional depictions of amnesia and related disorders compare with their real-­life counterparts, see Sebastian Dieguez and Jean-­Marie Annoni, “Stranger than Fiction: Literary and Clinical Amnesia,” in J.  Bogousslavsky and S. Dieguez, eds., Literary Medicine: Brain Disease and Doctors in Novels, Theater, and Film (Basel: Karger, 2013), 137–68. 27. Ronald Dworkin, Life’s Dominion: An Argument About Abortion, Euthanasia, and Individual Freedom (New York: Knopf, 1993), 223. 28. Dworkin, Life’s Dominion, 224. 29. Dworkin, Life’s Dominion, 226. 30. Dworkin, Life’s Dominion, 227. 31. Seana Shiffrin, “Autonomy, Beneficence, and the Permanently Demented,” in Justine Burley, ed., Dworkin and His Critics: With Replies by Dworkin (Malden, MA: Blackwell Publishing, 2004), 202. 32. Ronald Dworkin, “Reply to Critics,” in Burley, Dworkin and His Critics, 367–8. 33. For Shiffrin’s discussion of this issue, see “Autonomy, Beneficence, and the Permanently Demented,” 207–8. 34. Both Shiffrin and Agnieszka Jaworksa also argue for somewhat different conceptions of autonomy in response to Dworkin. Jaworska argues that demented individuals may retain the capacity to value even after they have lost the ability to implement their values. And she argues that this capacity to value is worthy of respect in a way that Dworkin’s emphasis on shaping one’s life as a whole overlooks. See Agnieszka Jaworksa, “Respecting the Margins of Agency: Alzheimer’s Patients and the Capacity to Value,” Philosophy & Public Affairs 28 (1999): 105–38. Shiffrin, on the other hand, focuses on the importance of simply having control over one’s experiences even if that does not involve imbuing one’s experiences with meaning. This suggests that respect for autonomy may be appropriate even after the capacity to value has been lost. See Shiffrin, “Autonomy, Beneficence, and the Permanently Demented.” I suspect the account of autonomy I give in the main text is more akin to Jaworska’s than Shiffrin’s, though I will not attempt a detailed comparison here. But neither Jaworska nor Shiffrin has a satisfying account of how to reconcile respect for the autonomy of a competent person with respect for the kinds of autonomy still possible for people with dementia. As we have seen, Shiffrin problematically rejects the question by denying that respect for the autonomy of the competent person is still relevant to the treatment of the demented person. Jaworska is less committal on this point but much of her discussion points in the same direction. 35. Shiffrin also notices that Dworkin’s integrity account seems to do less well protecting akratic actions than it purports to. She suggests on his behalf that perhaps the reason to defer even to these choices is that “without the freedom to deviate from one’s values, the significance of one’s accomplishment in creating and sustaining a life guided by certain values is lessened. Perhaps also all self-­propelled actions contribute to a self-­fashioned life character and life, although the sum of these actions may produce a different product than what one hopes or intends”: Shiffrin, “Autonomy, Beneficence, and the Permanently Demented,” 201. These may well be ways of enabling the integrity view to get the right answers about deference to akratic or unprincipled actions. But I nonetheless think that it focuses on the wrong reasons for these answers, and I try to motivate an alternative in the text.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Shadows of the Self  195 36. As Shiffrin helpfully points out, the educative reason for deferring to children’s choices will not explain our deference to the choices of children who are terminally ill. But I take the conception of autonomy Shiffrin proposes to make sense of these cases to be too minimal to provide plausible grounds for our deference to people’s choices. See Shiffrin, “Autonomy, Beneficence, and the Permanently Demented,” 205–6. 37. For a helpful discussion of these “domains of discretion,” see Tamar Schapiro, “What Is a Child?” Thinking: The Journal of Philosophy of Children 15 (2001): 4–15. 38. Shenk, The Forgetting, ch. 8. 39. This case is drawn loosely from Dworkin’s presentation of the real-­life case of Margo. See Dworkin, Life’s Dominion, 220–1. 40. This case is drawn loosely from Jaworska’s presentation of the real-­life case of Mr. O’Conner. See Jaworksa, “Respecting the Margins of Agency,” 107. 41. I am indebted to Ryan Davis, Kyla Ebels-­Duggan, Tamar Schapiro, and the participants at the Workshop in Ethics and Political Philosophy at Brigham Young University for helpful comments on an earlier draft of this paper. I have also benefitted tremendously from many illuminating discussions with the participants in my graduate seminar on personal identity: Anwar ul Haq, Stephen Mackereth, Caleb Reidy, Rebecca Thomas, and Daniel Webber.

Bibliography Buchanan, Allen. “Advance Directives and the Personal Identity Problem.” Philosophy & Public Affairs 17 (1988): 277–302. Burley, Justine, ed. Dworkin and His Critics: With Replies by Dworkin. Malden, MA: Blackwell Publishing, 2004. Davies, Jennifer. “We Don’t Know What Is Going Through Her Mind.” In Lucy Whitman, ed., Telling Tales About Dementia: Experiences of Caring, 35–9. London: Jessica Kingsley Publishers, 2010. Dieguez, Sebastian and Jean-Marie Annoni. “Stranger than Fiction: Literary and Clinical Amnesia.” In  J.  Bogousslavsky and S.  Dieguez, eds., Literary Medicine: Brain Disease and Doctors in Novels, Theater, and Film, 137–68. Basel: Karger, 2013. Dresser, Rebecca. “Life, Death, and Incompetent Patients: Conceptual Infirmities and Hidden Values in the Law.” Arizona Law Review 28 (1986): 373–405. Dworkin, Ronald. “Philosophical Issues Concerning the Rights of Patients Suffering Serious Permanent Dementia.” Written for the United States Congress Office of Technology Assessment (1986). https://digital.library.unt.edu/ark:/67531/ metadc97370/. Dworkin, Ronald. Life’s Dominion: An Argument About Abortion, Euthanasia, and Individual Freedom. New York: Knopf, 1993. Dworkin, Ronald. “Replies to Critics.” In Justine Burley, ed., Dworkin and His Critics: With Replies by Dworkin, 339–95. Oxford: Blackwell Publishing, 2004.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

196  Japa Pallikkathayil Htay, U Hla. “We Learn to Enter Her World.” In Lucy Whitman, ed., Telling Tales About Dementia: Experiences of Caring, 58–64. London: Jessica Kingsley Publishers, 2010. Jaworska, Agnieszka. “Respecting the Margins of Agency: Alzheimer’s Patients and the Capacity to Value.” Philosophy & Public Affairs 28 (1999): 105–38. Korsgaard, Christine M. “Personal Identity and the Unity of Agency: A Kantian Response to Parfit.” In Creating the Kingdom of Ends, 363–98. Cambridge: Cambridge University Press, 1996. Korsgaard, Christine  M. The Sources of Normativity. Cambridge: Cambridge University Press, 1996. Parfit, Derek. Reasons and Persons. Oxford: Clarendon Press, 1984. Schapiro, Tamar. “What Is a Child?” Thinking: The Journal of Philosophy of Children 1 (2001): 4–15. Shenk, David. The Forgetting: Understanding Alzheimer’s; A Biography of a Disease. London: HarperCollins, 2002. Shiffrin, Seana. “Autonomy, Beneficence, and the Permanently Demented.” In Justine Burley, ed., Dworkin and His Critics: With Replies by Dworkin, 195–217. Oxford: Blackwell Publishing, 2004. Whitman, Lucy, ed. Telling Tales About Dementia: Experiences of Caring. London: Jessica Kingsley Publishers, 2010. World Health Organization. “Dementia.” September 2, 2021. https://www.who. int/news-room/fact-sheets/detail/dementia.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

9 Korsgaard on Responsibility T. M. Scanlon

1.  Introduction: Holding Responsible and Being Responsible In “Creating the Kingdom of Ends: Reciprocity and Responsibility in Personal Relations,” Chris Korsgaard writes, “To hold someone responsible is to adopt an attitude towards him rather than to have a belief about him or about the conditions under which he acts.”1 It is certainly true that holding a person responsible can involve “doing something” in the broad sense that includes adopting an attitude. Holding a person responsible can involve a variety of actions and attitudes, and the appropriateness of doing any of these things depends, as she says, on many factors, such as the significance of what the person has done, the way one has been affected by it, and one’s prior relationship with that person. It seems, however, that there is such a thing as a person’s being responsible in a sense that does not depend on whether we do any of these things. Rather, being responsible for something in this sense seems to be fact about the person that is a precondition for the appropriateness of doing these other things. For example, if I failed to meet you because I have been kidnapped and held prisoner, then I was not responsible for my failure to show up to meet you at the appointed time, and it would therefore not be appropriate for you to hold me responsible for my failure. In addition, it might seem that a person can be responsible for something even though, for some reason of the kind Korsgaard mentions, it would not be appropriate for anyone to hold that person responsible for it. For example, some things we do are so trivial in their effects that it would be ridiculous for anyone to say that they are holding us responsible for them, but also odd to deny that we are responsible for doing them. Holding a person responsible for something can mean a number of different things. As Angela Smith points out, “holding a person to be responsible for X” might mean (1) taking the person to be open, in principle, to some form of moral appraisal for X, (2) taking the person to be culpable for X, or T. M. Scanlon, Korsgaard on Responsibility In: Normativity and Agency: Themes from the Philosophy of Christine M. Korsgaard. Edited by: Tamar Schapiro, Kyla Ebels-­Duggan, and Sharon Street, Oxford University Press. © T. M. Scanlon 2022. DOI: 10.1093/oso/9780198843726.003.0009

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

198  T. M. Scanlon (3) actively blaming the person in some way for X.2 Holding a person responsible in any of these three senses is something we do in a broad sense of the term. But I find it very plausible to say that when we hold someone responsible in either of the first two senses, what we “do” is just to judge something to be the case. So even if to hold someone responsible in either of these senses is to do something—namely to form a judgment—there is also a fact about whether this judgment is correct. This is a fact about whether the person is responsible for something in sense (1) or about whether the person is culpable. By contrast, holding someone responsible in sense (3) is not just a matter of judging something to be the case, but of adopting some attitude or course of action in response to what one takes to be these facts (what Smith refers to as “active blaming”). So Korsgaard’s thesis is controversial only if she is understood as denying that there is such a thing as a fact about whether a person is responsible in the first of the senses that Smith distinguishes, which I will call the fundamental sense, since it is a precondition for the appropriateness of holding a person responsible in either of the other senses. As I have said, it seems to me that there is such a fact. In order to determine the extent to which Korsgaard and I disagree, I need first to explain how, in my view, facts about responsibility in this sense should be understood.

2.  Conditions of Responsibility To say that a person is responsible for something is to say that it is the kind of fact about what that person is like or has done that can in principle have certain moral consequences. These consequences can be of two different kinds. Facts about what a person is like or has done can make appropriate what P. F. Strawson called reactive attitudes, such as resentment, blame, praise, or gratitude.3 But facts about what a person has done can also change the person’s moral situation in other ways, such as by creating an obligation or giving permission. “Being responsible” is a necessary condition for something about the person to have consequences of either of these kinds: people should not be blamed for things that they are not responsible for, and they are not obligated by promises that they were not responsible for making. The question of what kinds of things a person is responsible for in the fundamental sense is just the moral question of what kinds of facts about the person can have these moral consequences. It follows that in my view an inquiry into the conditions of responsibility should begin by identifying the kind of moral consequences

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Korsgaard on Responsibility  199 in question and then ask the moral question of what kind of facts about a person can have moral consequences of this kind.4 One possible view is that the conditions of responsibility for moral consequences of these two kinds are the same. Individuals are responsible for just those things that are under their voluntary control: they can be blamed only for things that are under their control, and they are bound only by agreements that that they enter into voluntarily. I believe, contrary to this, that the conditions of responsibility are different for moral consequences of these two kinds.5 In this paper, however, I will be concerned only with responsibility of the first kind, which I call moral reaction responsibility—that is to say, with the kinds of facts about what a person is like or has done that can make moral reactive attitudes such as blame and gratitude appropriate. It is widely held, and often just assumed, that individuals are responsible in this sense only for things that are under their voluntary control.6 But it is worth asking why this should seem to be so. There seem to me to be two possible reasons.7 One is that moral reactions such as blame are seen as a kind of sanction, an undesirable form of treatment that it is unfair to subject people to unless they have had the opportunity of avoiding this by choosing differently.8 This is true of most forms of criminal punishment, but not of moral reactions of the kind we are discussing. Unlike criminal punishment, justifiable forms of blame, such as disapproval and withdrawal of trust and affection, do not deprive individuals of things that they normally have unconditional claim to. We do not owe it to people, for example, to trust them, or treat them as friends, regardless of their attitudes toward us. A second possible reason for thinking that voluntariness is a condition for moral reaction responsibility is that only facts about individuals that are under their control have the kind of significance that makes moral reactions appropriate. Our friends can be blameworthy for revealing things we told them in confidence if they do this in lighthearted conversation for the amusement of others, but they would not be blameworthy for revealing these things when threatened with torture and death. Such examples do not, however, support the view that voluntary control is a condition for moral reaction responsibility. First, they presuppose that what makes moral reactions such as blame appropriate are facts about what a person treats as reasons, and they do nothing to show that these facts have this kind of significance only if they are under a person’s voluntary control. Second, they do not show that a lack of voluntariness has moral consequences by undermining responsibility. A bank teller who hands over money to an armed robber does so involuntarily. But it would be a mistake to say that the teller is not responsible for what he does.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

200  T. M. Scanlon The teller should not be blamed for handing over the money, but he may deserve commendation, or perhaps a bonus, for handling a dangerous situation so calmly. What the robber’s credible threat does is not to undermine the teller’s responsibility for what he does, but rather to change what it is that he is responsible for doing, making it something that is not blameworthy, and even commendable. In my view, the things a person is responsible for in the most basic sense relevant to moral reactive attitudes are facts about what the person takes to be reasons (for actions or other attitudes) and facts about how the person governs him or herself in the light of these reasons. A person is responsible for other things, such as voluntary actions, insofar as these reflect facts about that person of this fundamental kind. Here I should acknowledge that I am using the expression “responsible for” in a distinctively broad way that departs in one respect from common usage. Perhaps because of the influence of the idea that individuals are responsible only for things that are under their voluntary control, to say that a person was “responsible for X” often means that that person had some role in bringing it about that X obtains. In my view this need not be the case. To say that a person is responsible for X is merely to say that X is the kind of fact about what that person is like or has done that can, in principle, have certain moral consequences, among which is the appropriateness of holding the person responsible in various ways. Such a claim has two aspects. The first is a purely descriptive matter—that X is a fact about the person’s attitudes toward reasons and the ways that the person responded to these reasons. The second is the normative claim that this descriptive fact has certain moral implications. Both of these claims concern facts about the person that hold independent of anything others may do. But it is important to see that two kinds of facts are involved, descriptive, and normative. I will return to this point later in comparing my view with Korsgaard’s. The range of things a person is responsible for in this fundamental sense includes many things that are not under that person’s voluntary control. First, it includes facts about a person’s attitudes toward reasons—about what she sees as important or unimportant—that are not voluntary. Second, it includes not only judgments about reasons that the person is conscious of making, and endorses on reflection, but also attitudes that occur to the person spontaneously even if the person rejects them once they occur. These are things a person is responsible for even if the person cannot stop them from recurring. This may seem harsh. We all find ourselves with thoughts that we are embarrassed by and reject. But if such thoughts are not attributable to us in the way that responsibility requires, why should we be embarrassed or ashamed by

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Korsgaard on Responsibility  201 them and eager to be rid of them? It is more plausible to say that these are things that a person is responsible for in this basic sense, but that it would be overly harsh to judge a person severely for every such thought. As Korsgaard says, “even the best of us can just slip.”9 A person can also be responsible in this sense for attitudes toward reasons that are not consciously held but are nonetheless manifested in the person’s behavior. Some cases of forgetting have significance of this kind.10 Insofar as one regards something as important, one will tend to remember it, and to think of it when this is appropriate. A failure to remember something, such as a friend’s birthday, can therefore indicate that one does not really care that much about it, and this brings such cases of forgetting within the class of things one is responsible for on the account I am presenting. The things a person is responsible for in my view include not only attitudes toward reasons but also “facts about how the person, as a rational agent, governs herself in the light of what she regards as reasons.”11 This covers actions that are done out of weakness of will, contrary to the person’s own conscious judgment about what he has reason to do. But it also covers things such as cases of forgetting that do not reflect any deficiency in what a person cares about. A person can forget things simply out of constitutional absent-­ mindedness, even though there is ample evidence that he cares deeply about them. Even when failures of this kind reflect no lack of concern, they are among the things that we are responsible for. They are aspects of what we are like that our friends, and others as well, have reason to care about and take into account in deciding how to relate to us. That is why they are things that we are embarrassed by and feel the need to apologize for.12

3.  Relationships, Reasons, and Responsibility Strawson emphasizes the significance that these facts about a person have for us. He speaks of “the very great importance that we attach to the attitudes and intentions towards us of other human beings, and the great extent to which our personal feelings and reactions depend on, or involve, our beliefs about these attitudes and intentions.”13 As he says, it matters greatly to us “whether the actions of other people—and particularly some other people—reflect attitudes towards us of goodwill, affection, or esteem on the one hand or of contempt, indifference, or malevolence on the other.”14 This concern with the attitudes of others, and our reactions to these attitudes are parts, he says, of the various relationships we have with each other.15

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

202  T. M. Scanlon Korsgaard also sees our reasons for holding others responsible as parts of important relationships. She writes: . . . it is not possible for us to avoid holding one another responsible. For holding one another responsible is the distinctive element in the relation of adult human beings. To hold someone responsible is to regard her as a person—that is to say, as a free and equal person, capable of acting both ra­tion­ ally and morally . . . When you hold someone responsible, . . . you are prepared to accept promises, offer confidences, exchange vows, cooperate on a project, enter a social contract, have a conversation, make love, be friends, or get married. You are willing to deal with her on the basis of the expectation that each of you will act from a certain view of the other: that you each have your reasons which are to be respected, and your ends which are to be valued.16

What does this appeal to relationships add to the bare thesis that we have reason to care about what other people see as reasons applying to their behavior toward us and about the way they govern themselves in the light of these reasons? The answer, I think, is that it provides a context that explains the significance of these particular facts about a person and explains why they are the basis for the particular responses that moral reactions such as blame involve. To make this clear I need first to say more about the nature of these relationships. Relationships in the relevant sense are defined by shared intentions and expectations about the attitudes of the parties in the relationship and the ways they will behave toward one another. Friendship, for example, centrally involves special kinds of affection and concern. Friends see themselves as having special reason to care about what happens to each other, to be pleased when good things happen to the friend and sad when things do not go well, and to take pleasure in each other’s company. These reasons support mutual intentions and corresponding expectations, such as to “keep in touch,” with one another, to help one another when needed, and to confide in one another and respect the confidentiality of what the other discloses to them. To be a friend involves having these expectations about the other and, reciprocally, intending to live up to the other person’s expectations about one’s own behavior. Friendship involves special kinds of concern, and special kinds of spontaneous responses, such as feelings of affection, feelings of pleasure when things go well for a friend, and regret and sadness when they do not. If a person whom one has regarded as a friend does not act in these ways, or fails to show these feelings, one has reason to respond, by complaining or

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Korsgaard on Responsibility  203 protesting and even, when these things are persistent, by ceasing to regard the person as a friend and changing one’s intentions and expectations accordingly. Friends have reasons of several kinds to engage in these forms of active blaming. One reason is to call a friend’s attention to the significance of what he is doing by complaining about it, in the hope of correcting his behavior and preserving the friendship. But protesting in this way is not always or merely a kind of sanction, intended to influence the other’s behavior. A friend has reasons for complaining and protesting in such a case even if there is no hope of changing the other friend’s behavior. To accept this behavior without complaint, and to continue to treat him as a friend despite its continuance, would be to accept an understanding of the relationship that permits this. Insofar as the value of a friendship and one’s reasons for remaining in it depends on its reciprocal and equal character, these facts about the other party can be reasons to complain and even to end the relationship. This is not to say that every relationship worth having must be one of perfect equality, but only that lack of reciprocity can be a good reason for reacting in these ways. Even feeling affection for a person does not have the same meaning when these feelings are not reciprocated. The facts about a friend that make these reactions appropriate fall into the class that I described: facts about what the person sees as reasons and facts about how the person, as a rational agent, governs him- or herself in the light of these reasons. Some of the things that we have reason to react to in this way are actions and failures to act that are under the friend’s voluntary control, but we have the same reasons to react to many facts about a person that are not under his or her control. A good friend is one who cares about you, but also can be counted on to act in the ways that caring involves. It is not unfair to react in the ways I have described to failures that are not under a friend’s voluntary control. We do not owe it to a person to have the special feeling and concern that friendship involves when these are not reciprocated, or to confide in them and rely on them if they are unreliable. I have identified the reasons we have for having certain reactive attitudes as reasons we have insofar as we stand in the particular relationship of being friends. But it is important to see that what is basic in the order of explanation here is not, as Strawson may be read to suggest, simply the facts about what “our practices” and the relationships that these practices involve happen to be.17 What is truly basic is not the relationships we have, whatever they may be, but rather our reasons for having relationships of the right kinds, and for objecting when it seems that they are not. For example, being friends involves having the reasons I have described only insofar as our friendship is a

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

204  T. M. Scanlon relationship worth having. The expectations defining a very unequal relationship might license complaint when a party in the inferior role is insufficiently deferential. But this does not mean that the person in the superior role has good reason to protest in such a case, or that the other party has reason to accept such a protest. Rather, it means that the relationship is not one that there is good reason to be a party to. It may seem implausible to draw conclusions about moral responsibility from personal relationships such as friendship and the forms of responsibility and blame that they involve. Moral duties apply between strangers who, by definition, need have no special affection for one another. Unlike friendship, the requirements of morality are not optional. It is crucial to the meaning and value of friendship that one can choose one’s friends, but moral duties are owed to all. Moreover, unlike the obligations of friendship, which must be reciprocal, at least most moral obligations are owed even to those who flout them. These differences between the relationship between friends and the “moral relationship” that I am claiming holds even between strangers raise questions about the relevance of the former relationship as a basis for drawing conclusions about the latter. The first question is, given the differences in the reasons we have for being concerned with the intentions and expectations involved in friendships and those with which morality is concerned, how much similarity is there between the kinds of facts about a person that can be the basis for reactive attitudes in the two cases? The second question concerns differences in the relevant reactive attitudes. In the case of friendship, the reactions made appropriate by the failures of one party generally include modifying or even rejecting these expectations. Given the unconditional and non-­optional character of moral requirements, how much room is there for reactive attitudes that go beyond mere assessment? The answer to the first question is that even though morality is not, like friendship, based on special ties of affection, it is concerned with the reasons that individuals see as governing their interactions. Even in the case of strangers, we have reason to care about whether their attitudes toward us are, in Strawson’s terms, “attitudes towards us of goodwill, affection, or esteem on the one hand or of contempt, indifference, or malevolence on the other.” So, even though friendship and morality have to do with different kinds of concern, what they have in common is being relationships that specify attitudes that individuals conceived as rational self-­governing agents are to have toward one another.18 In both cases, therefore, the deficiencies that can make reactive attitudes appropriate are facts about what the person sees as reasons and facts about how the person, as a rational agent, governs him or herself in the light of these reasons. Facts about a person that are not of this kind, such as height,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Korsgaard on Responsibility  205 are things that the person is “not responsible for.” Facts that are of this kind are in principle relevant bases for moral reactive attitudes, but whether they are relevant, and the significance that they have, depends on the content of substantive moral requirements. This is why conditions of responsibility of the kind we are discussing are necessary but not sufficient conditions for moral reactions. The second question concerns the range of reactions that can be appropriate when the expectations of the “moral relationship” are not met. When others, even strangers, behave in ways that manifest the attitude that our interests are of no significance, we have reason not to accept this characterization of our relationship, and reason to complain and to protest it, even if this will not change the person’s attitude. Reacting in this way already involves more than a mere judgment of culpability, but even in the moral case there is room for further responses that go beyond moral assessment. Moral requirements not to harm others are not generally conditional on their abiding by these limits. But there are other aspects of normal moral relations that we can be justified in suspending. We are not required to trust or enter into cooperative relations with individuals who have shown themselves to be untrustworthy, and we are not required to regard such people as candidates for friendship or other forms of personal relationship. Reactions of these kinds are, I believe, among the most important forms of moral blame.19 They deprive those toward whom they are directed of things that they have reason to want. But they are not properly understood as sanctions, since they are not done because they are bad for these individuals, and need not be done only, or even mainly, with the aim of affecting their behavior. These reactions can be made appropriate simply by facts about individuals as self-­governing rational agents—about what they regard as reasons and about the ways in which they respond to these reasons—whether or not these facts are under their voluntary control.20 The things that these reactions deprive these individuals of are not things that we owe them unconditionally. It is more than can reasonably be asked of us to go on trusting them and treating them in other normal ways no matter how they behave.

4.  Korsgaard on the Theoretical Conception of Responsibility Korsgaard also rejects the idea of voluntariness as the condition of responsibility.21 But she takes this idea to be an instance of a wider view, which she also rejects. This is what she calls the theoretical conception of responsibility, according to which

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

206  T. M. Scanlon it is a fact about a person that she is responsible for a particular action, or that there is some fact about her condition either at the time of action or during the events which led up to it which fully determines whether it is correct to hold her responsible. It is a fact, say, that she could have done otherwise, or that she could have avoided the condition which made it impossible for her to do otherwise. Similar although somewhat more complicated claims would be made about the person’s reactions and attitudes: facts about the person settle the question whether she is accountable for them.22

By contrast, on the practical conception of responsibility that she favors, “While . . . facts about the agent and about her condition at the time of action guide your decision whether to hold her responsible, they do not fully determine it.”23 “[T]o hold someone responsible,” she writes, “is to adopt an attitude towards him rather than to have a belief about him or about the conditions under which he acts.”24 The view I have been advocating certainly counts as a “theoretical conception of responsibility.” When we hold a person responsible for some action in the fundamental senses we are, in my view, claiming something to be the case about the person at the time of action, namely that the action reflected facts about what the person was inclined to treat as reasons and the way the person governed him- or herself in the light of these.25 To claim that these facts are, ultimately, what a person is responsible for is on my view simply to make the normative claim that these facts about the person have the relevant kind of moral significance. In later work, Korsgaard seems to offer a different, perhaps deeper explanation of why certain facts about a person have this significance. We are, she says, responsible for the kind of person that we are because we have made ourselves that kind of person. She writes: It is because our actions are expressive of principles we ourselves have chosen, principles we have adopted as the laws of our own causality, that it makes sense for us to hold one another answerable in this way: to demand one another’s reasons, and to take it, as we say, personally, when we hear what they are.26

This emphasis on principles we have chosen might be read as endorsing something like the voluntarist account that Korsgaard earlier rejected. I believe that this is not the proper reading, however. As she goes on to say:

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Korsgaard on Responsibility  207 But the sense in which we must create ourselves in order to be responsible is not that we must literally bring ourselves into being. Rather, we are responsible because we have a form of identity that is constituted by our chosen actions. We are responsible for our actions not because they are our products but because they are us, because we are what we do.27

But even if it is not a voluntarist account, this passage makes it sound as if a person’s being responsible in the fundamental sense is nonetheless a fact about that person, namely a fact about how the person has constituted him or herself. The question is why this is not a “theoretical” conception of responsibility, and how it is consistent with Korsgaard’s view that taking a person to be responsible in this fundamental sense is something that we do. I believe that these claims can be understood in a way that is not in­con­sist­ ent. In saying that we constitute ourselves Korsgaard does not mean that we “literally bring ourselves into being.” The idea is not one of control over our attitudes, but rather a way of understanding the content of these attitudes themselves. Attitudes are “ours” only because they have what might be called an element of endorsement. It is part of having an attitude of the kind in question that we “take it to be ours” and in this way “take responsibility” for it. This explains the apparent tension between what Korsgaard says in “Creating the Kingdom of Ends” and what she says later in Self-­Constitution. When she rejects voluntariness as a condition for responsibility in “Creating the Kingdom of Ends,” Korsgaard is both rejecting the idea that judgments of responsibility can be understood as purely theoretical claims, and rejecting the substantive moral thesis that all of the conditions that excuse agents from blame are a matter of the lack of voluntariness. In Self-­Constitution on the other hand, Korsgaard is making a claim about what it is to see oneself, or another, as a self-­governing agent—as a person. Her claim is that to do this is to see oneself as having a life that is unified in the way that being a person requires, unified by one’s being the author of one’s actions and attitudes. This understanding of what it is to be responsible for an attitude may seem inconsistent with my claim that the things individuals are responsible for includes attitudes that they reject and are even appalled by when they become aware of them. Perhaps Korsgaard would deny that people can be responsible for such attitudes, but she need not take this view. If an attitude that I am appalled by did not have this element of endorsement—if it did not seem to me, when it occurs, to be something that I think, then I would not be so appalled and embarrassed by it. On the other hand, insofar as an attitude has

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

208  T. M. Scanlon this character, it would be bad faith on the person’s part to just say, “That’s not me.” This same character also makes an attitude something that others have reason to regard as significant for their relations with me, and thus to “hold me responsible” for. This would not, however, apply to all of the things that a person is responsible for on the account I have been defending. It would seem to leave out responsibility for cases of forgetting that merely imply an attitude toward reasons and also cases such as constitutional absent-­mindedness. Failure to bear something in mind when it would be relevant need not involve any endorsement of the view that it does not matter. So the range of things a person is responsible for on Korsgaard’s view may be narrower than the class I have described. In the light of all of this, what should we conclude about Korsgaard’s claim that holding a person responsible in the fundamental sense is something we do, and that there is no fact, independent of what we do, about whether some other person is responsible for something in this sense? I  have argued that a person is responsible just for facts about what that person takes to be reasons (for actions or other attitudes) and facts about how that person governs him- or herself in the light of these reasons. It would seem that a fact about a person either is or is not a fact of this kind, independent of anything that we do. Such a claim about what a person is like or has done is a purely descriptive claim. But, as I pointed out earlier, the claim that such facts are things that a person is responsible for is a normative claim, namely a claim that these facts give us reasons to react to the person in certain ways. As a normative realist, I believe that if facts about what the person is like do give us such reasons this is just a further, normative fact about the person. But things are different for Korsgaard. In her view, that we have certain reasons, including reasons to adopt reactive attitudes, is not a normative fact about the world independent of us, but depends on what we will, or could will.28 We can understand this as follows. As I have argued above, a person’s attitudes and actions give us reasons to react in these ways because of the significance these attitudes have within our relationships with that person: because, for example they are incompatible with the expectations that define this relationship. For Korsgaard, standing in such a relationship with a ­person, including what I called “the moral relationship,” is something we choose to do. To see a person as a self-­governing agent who is responsible for what she does is therefore not just to recognize an independent psychological fact or normative fact about her, but rather to (choose to) view her as a rational agent

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Korsgaard on Responsibility  209 with whom one stands in a certain relationship, within which certain facts about her have meaning and give us reasons. As Korsgaard says, “The reason we must view another as a fellow rational person rather than as a psycho-­ social phenomenon is not that he is in fact one of these things rather than the other. In fact, he is both. That another is responsible is what Kant calls a postulate of practical reason: a belief or attitude that can be formulated theoretically, but is practical and moral in its basis.”29

5. Conclusion My aim in this paper has been to understand Chris Korsgaard’s view that to take a person to be responsible for something is not just to judge something to be true of that person but rather to do something—to take a certain attitude toward the person. As a contrast to her account, I described my own view, according to which a person’s being responsible for something in the most fundamental sense is a fact about that person and about the reasons we have for responding to him or her in certain ways, reasons that depend on our relationship with the person. I then discussed Korsgaard’s claim that people are responsible for what they are like because they have made themselves that kind of person, and explained why this is not what she calls a theoretical conception of responsibility and is consistent with her view that holding a person responsible is something that we do.30

Notes 1. Christine  M.  Korsgaard, “Creating the Kingdom of Ends: Reciprocity and Responsibility in Personal Relations,” in Creating the Kingdom of Ends (Cambridge: Cambridge University Press, 1996), 188. 2. Angela  M.  Smith, “On Being Responsible and Holding Responsible,” The Journal of Ethics 11 (2007): 465–84. 3. P. F. Strawson, “Freedom and Resentment,” in Gary Watson, ed., Free Will, 2nd edn., 72–93 (Oxford: Oxford University Press, 2003). 4. This may make the account of responsibility that I am defending what R. Jay Wallace calls a “normative” account, although not necessarily one that is based on an idea of fairness to the person being held responsible. See R.  Jay Wallace, Responsibility and the Moral Sentiments (Cambridge, MA: Harvard University Press, 1994), 15. 5. As I have argued in a number of places, most recently in “Forms and Conditions of Responsibility,” in Randolph Clarke, Michael McKenna, and Angela  M.  Smith, eds., The Nature of Moral Responsibility: New Essays, 90–113 (Oxford: Oxford University Press, 2015).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

210  T. M. Scanlon 6. See, for example, Galen Strawson, “The Impossibility of Moral Responsibility,” Philosophical Studies 75 (1987): 5–24. 7. I discuss the following points more fully in Moral Dimensions (Cambridge, MA: Harvard University Press, 2009), ch. 4. 8. Galen Strawson says, for example, that he is concerned with the kind of responsibility that would make ideas of heaven and hell at least make sense. See “The Impossibility of Moral Responsibility,” 9. 9. Korsgaard, “Creating the Kingdom of Ends,” 211. 10. Here I follow Angela M. Smith, “Responsibility for Attitudes: Activity and Passivity in Mental Life,” Ethics 115 (2005): 236 and passim. 11. See T. M. Scanlon, What We Owe to Each Other (Cambridge, MA: The Belknap Press of Harvard University Press, 1998), 268ff. 12. I am grateful to Noel Dominguez for pressing on me the importance of cases of this kind. Counting such failures as things a person can be responsible for is something of a departure from the “rational relations” view that Smith and I have defended. But even though these failures may not reflect anything about what the person sees as reasons, they do reflect the way he governs himself: the degree to which what he sees as reasons have the effects that they should, rationally, have on what he does. So the “rational relation” between reasons and remembering is relevant to a person’s responsibility in such cases, albeit in a different way than Smith describes. 13. Strawson, “Freedom and Resentment,” 75. 14. Strawson, “Freedom and Resentment,” 76. 15. Strawson, “Freedom and Resentment,” 76. 16. Korsgaard, “Creating the Kingdom of Ends,” 189–90. 17. To take our practices, whatever they may be, as basic would lead to an unacceptable form of relativism, as Patrick Todd emphasizes in “Strawson, Moral Responsibility and the ‘Order of Explanation’: An Intervention,” Ethics 127 (2016): 208–40. 18. As Korsgaard observes, even our relationship with our pets may be something like this, but not the same. See Fellow Creatures: Our Obligations to the Other Animals (Oxford: Oxford University Press, 2018), 107–8. 19. As I argue in chapter 4 of Moral Dimensions. 20. They are deserved, simply in virtue of these facts about what the individuals are like, as I argue in “Giving Desert Its Due,” Philosophical Explorations 16 (2013): 1–16. 21. Korsgaard, “Creating the Kingdom of Ends,” 198–9. Smith rejects it as well. 22. Korsgaard, “Creating the Kingdom of Ends,” 197. 23. Korsgaard, “Creating the Kingdom of Ends,” 198 (emphasis in original). 24. Korsgaard, “Creating the Kingdom of Ends,” 188. 25. Believing a person to be responsible for something in this sense “shapes” our decisions about whether to hold the person responsible in other senses because these further responses are appropriate only if the person is responsible in this fundamental sense. The person’s responsibility does not, however, “fully determine” this decision. As I have said, individuals can be responsible for acts that are not culpable but praiseworthy, and even for ones that are of no moral significance whatever. And the appropriateness of any of the actions and attitudes that can be involved in actively blaming can depend on many factors beyond mere culpability, such as the significance of what

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Korsgaard on Responsibility  211 the person has done, the way one has been affected by it, and one’s prior relationship with that person. 26. Christine  M.  Korsgaard, Self-­Constitution: Agency, Identity, and Integrity (Oxford: Oxford University Press, 2009), 131. 27. Korsgaard, Self-­Constitution, 130; see also 18–20. 28. See Christine  M.  Korsgaard, The Sources of Normativity (Cambridge: Cambridge University Press, 1996), 19 and passim. 29. Korsgaard, “Creating the Kingdom of Ends,” 208. 30. I am delighted to contribute to this volume honoring Christine Korsgaard, to have this opportunity to express my admiration for her work and my gratitude to her for almost thirty years of collegial partnership. I am grateful to Tamar Schapiro and Angela Smith for helpful comments on earlier versions of this paper, and to Noel Dominguez for research assistance and for fruitful discussion of questions of responsibility.

Bibliography Korsgaard, Christine  M. “Creating the Kingdom of Ends: Reciprocity and Responsibility in Personal Relations.” In Creating the Kingdom of Ends, 188–221. Cambridge: Cambridge University Press, 1996. Korsgaard, Christine  M. The Sources of Normativity. Cambridge: Cambridge University Press, 1996. Korsgaard, Christine M. Self-Constitution: Agency, Identity, and Integrity. Oxford: Oxford University Press, 2009. Korsgaard, Christine M. Fellow Creatures: Our Obligations to the Other Animals. Oxford: Oxford University Press, 2018. Scanlon, T. M. What We Owe to Each Other. Cambridge, MA: The Belknap Press of Harvard University Press, 1998. Scanlon, T.  M. Moral Dimensions. Cambridge, MA: Harvard University Press, 2009. Scanlon, T. M. “Giving Desert Its Due.” Philosophical Explorations 16 (2013): 1–16. Scanlon, T.  M. “Forms and Conditions of Responsibility.” In Randolph Clarke, Michael McKenna, and Angela  M.  Smith, eds., The Nature of Moral Responsibility: New Essays, 90–113. Oxford: Oxford University Press, 2015. Smith, Angela. “Responsibility for Attitudes: Activity and Passivity in Mental Life.” Ethics 115 (2005): 236–71. Smith, Angela. “On Being Responsible and Holding Responsible.” The Journal of Ethics 11 (2007): 465–84. Strawson, Galen. “The Impossibility of Moral Responsibility.” Philosophical Studies 75 (1987): 5–24.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

212  T. M. Scanlon Strawson, P.  F. “Freedom and Resentment.” In Gary Watson, ed., Free Will, 2nd edn., 72–93. Oxford: Oxford University Press, 2003. Todd, Patrick. “Strawson, Moral Responsibility and the ‘Order of Explanation’: An Intervention.” Ethics 127 (2016): 208–40. Wallace, R. Jay. Responsibility and the Moral Sentiments. Cambridge, MA: Harvard University Press, 1994.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

10 Animal Value and Right Stephen Darwall

In Fellow Creatures (FC), Christine Korsgaard seeks to show how a Kantian moral theory can provide a more satisfying account of our obligations toward other animals than Kant’s own theory affords.1 Notoriously, Kant holds that any moral constraints on how we treat our fellow animals come only from indirect effects on our treatment of rational agents. Korsgaard argues, however, that what grounds the injunction to treat rational agents as ends in themselves in Kant’s own moral theorizing can warrant an extension to fellow animals—if you like, that the reasoning leading to the Formula of Humanity can be extended to a Formula of Animality. All animals, not just rational ones, she argues, should be treated as ends in themselves. Korsgaard is best known for an interpretation of Kant’s moral philosophy that is grounded in reflections on the nature of action and in the distinctive questions and commitments that rational agents have or must make in deliberating about what to do.2 A central move has been to argue that rational agents must value their ends and that they can justifiably do this only if they value themselves as rational agents. Moreover, they can consistently value themselves as rational, only if they treat both themselves and other rational agents as ends.3 This commits any deliberating agent to the Formula of Humanity (FH): So act that you use Humanity, whether in your own person or in the person of any other, always at the same time as an end, never merely as a means.4

FC continues this line of argument for what Korsgaard here calls the “active sense” of treatment as an end itself: I must regard you as an end in itself in the active sense if I regard you as capable of legislating for me, and so as capable of placing me under an obligation both to respect your choices, and to limit my own choices to things compatible with your value as an end itself.5 Stephen Darwall, Animal Value and Right In: Normativity and Agency: Themes from the Philosophy of Christine M. Korsgaard. Edited by: Tamar Schapiro, Kyla Ebels-­Duggan, and Sharon Street, Oxford University Press. © Stephen Darwall 2022. DOI: 10.1093/oso/9780198843726.003.0010

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

214  Stephen Darwall All, and only, rational beings, having the capacity to reflect on their own motivations and act on self-­given laws (that is, having “autonomy” in Kant’s sense), are reciprocally obligated to treat one another as ends in themselves.6 To FH, interpreted in this “active” sense, FC adds a “passive” sense of treatment as an end in itself: “I must regard you as an end in itself in the passive sense if I am obligated to treat your ends, or at least the things that are good-­ for you, as good absolutely.”7 Although nonhuman animals are unlike rational ones in not coming within the scope of FH interpreted in the active sense, that does not mean that we are not obligated to treat other animals as ends in themselves in the passive sense. FC argues that we are indeed so obligated. In effect, FC argues for a Formula of Animality (FA): So act that you use Animality, whether in yourself or in any other animal, always at the same time as an end, never merely as a means.8

We are obligated, FC argues, to comply with FA, that is, to treat any and all animals’ ends, or what is good-­for them, as good absolutely, as giving anyone reasons, whether the animal is rational or not.

1.  Two Senses of “Good” and “Good-­For” Korsgaard distinguishes two different uses of “good,” each having its own associated sense of “good-­for.”9 There is an omnibus evaluative use of “good,” “our most general term of positive evaluation,” which we use to evaluate “any kind of thing for which we have any use, or with which we interact.” Evaluative uses of “good” always involve some implicit standard or “function.” With artifacts and instruments, these come from the uses to which we put these things. Good knives are well-­ suited to cutting, for example. Natural organisms, plants, and animals, on the other hand, have natures that determine what it is for them to function well or flourish. So Korsgaard also calls the evaluative sense the “functional” sense of “good.” Things that contribute to something’s being a good thing of its kind in this evaluative or functional sense are good-­ for that thing in this sense; they enable it to function well as that kind of thing. Korsgaard’s second, or “final,” sense of “good” designates goodness as a goal or end. “We call something ‘good’ in the final sense when we consider it worth having, realizing, or bringing about for its own sake.”10 When we want something and are inclined to pursue it as an end, we see it as good in this final sense; we see it as worth pursuing or bringing about. Nonrational animals are

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Animal Value and Right  215 no less susceptible to final-­good-­representing motivational states; indeed, their susceptibility and consequent activity is part of what it is for them to function well. “An animal,” Korsgaard says, “is an organism that functions, at least in part, by representing her environment to herself, through her senses, and then by acting in light of those representations.”11 The representational states to which animals are susceptible have a “‘valenced’ character.” As the things the animal encounters “strike her as attractive or aversive,” she represents them as “to-­be-­ sought” or “to-­be-­avoided”; she “perceive[s] the world evaluatively” in this final sense—as worth pursuing or avoiding.12 “The final good,” Korsgaard says, “came into the world with animals, for an animal is, pretty much by definition, the kind of thing that has a final good.”13 Final and functional good are tied together for animals. When animals function well, they represent as final goods, and then pursue, things that are essential to their good functioning. This means that for animals, “the kind of thing that has a final good,” “final goods are just functional goods when taken as the ends of action.”14 Things are good-­for animals in this final sense when they constitute or contribute to their final goods. Now, we are animals too. So we also are subject to the valenced final-­good-­ representing states that are essential to good animal functioning. Korsgaard seems to lack a canonical term for these motivational states, but we can call them, as she sometimes does, desires. When we desire to have something, or not to have it in the case of aversion, we represent it as something to have or to avoid. We see it as finally good in the sense of something to pursue or worth pursuing (or to avoid, in the case of aversion). Whether or not other animals are able to, however, we rational animals can distinguish between whether something actually is worth pursuing (that is, is finally good in the fully normative sense of being desirable) or only seems to be, as it does in the valenced state of desire. Presumably, then, whether something really is a final good (normatively) for us, consists not just in whether we do desire and pursue it as an end, but in whether we should desire it, whether it is desirable for (and/or by) us or worth pursuing and making our end. We are rational agents, and it has always been, in Korsgaard’s view, this aspect of our agency that brings normativity before us, that leads us to ask and to seek to answer normative questions.15 Unlike nonrational agents, we can reflect on our motivations, the “grounds” of our actions, as Korsgaard calls them in FC, and on whether the desires that move us, presenting options before us as ends to be pursued or worth pursuing, really do provide us with normative reasons for acting. We can ask whether what we desire really is desirable.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

216  Stephen Darwall In FC, however, Korsgaard’s focus is on the fact that we are rational animals. Because we are, we are subject to the same final-­good-­representing motivational states as part of our human nature, as any other animal is. Therefore, when we choose ends on this basis, even if we do so as rational in a way that nonrational animals cannot, she argues, we are in effect valuing our animal nature in a way that commits us to valuing similarly that of any animal. We are treating our animality as an end in itself; so we cannot without inconsistency treat other animals simply as means to our ends. We are committed to treating them as ends in themselves in the passive sense of treating their ends and what is good-­for them as “good absolutely.” We are, that is, committed to FA. We could also say that we are committed to bringing other animals within the scope of FH, if we take “humanity” literally as referring to our human (and not just our rational) nature and interpret FH in the “passive” sense.

2.  Moral Obligation: Directed and “Pure and Simple” In what follows, I want to evaluate this line of thought as an argument for the thesis that we are morally obligated to treat other animals as ends in themselves in this passive sense. Korsgaard generally adds that our obligations toward other animals are owed to them. This would mean that these obligations are directed (alternatively, relational or bipolar) obligations.16 What hangs on this? Directed obligations are generally thought to have distinctive conceptual entailments that are not part of the idea of moral obligation pure and simple. In particular, following Hohfeld, obligations we owe to others are often thought to entail correlative claim rights that the obligee has against the obligor (the person obligated to them).17 If A is obligated to B to do X, then it follows that B has a claim right (against A) to A’s doing X, which gives B the standing to demand X of A.18 This typically brings a bundle of normative powers in its train. Often, at least, B may release A from the obligation, seek apology and/or compensation from A and/or perhaps ­forgive A if it is violated, and so on. Unlike the undirected concept of moral obligation (pure and simple), directed obligations conceptually entail ­obligees’ distinctive individual standing or authority, as the very individual to whom the obligation is owed, to make (or forbear making) claims and demands of the obligor.19 Others don’t generally have standing to release or forgive.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Animal Value and Right  217 So far as I can see, Korsgaard’s arguments in FC do not include anything that brings these distinctive directed (relational or bipolar) elements into play. There is nothing in her arguments that would show that any obligations we have toward other animals entail any rights other animals have to the obligatory treatment. By the same token, however, I doubt that anything Korsgaard says gives her much investment in this stronger directed claim. She may well be happy to disavow any ambition to argue that we have obligations to other animals in the distinctive directed sense that entails other animals having correlative rights and individual authorities to make claims and demands of us. She might, of course, still hold that these obligations are owed to animals in some different sense. A more serious issue, I think, is that Korsgaard’s arguments do not provide the conceptual materials even to conclude that we have moral obligations toward other animals in the undirected, nonrelational sense, that is, moral obligations toward them pure and simple. I shall argue that the arguments of FC are impotent, without supplementation, to show not just that we have good reason to treat other animals as ends in themselves, but that we are morally obligated to do so in the distinctive deontic sense that failing so to treat other animals is a culpable moral wrong (if we fail so to treat them without excuse). The reason this is so, as I have argued in the past, is that deontic moral concepts like moral obligation, right, and wrong, are conceptually connected to accountability (and second-­personal attitudes like moral blame) in ways that the concept of rational or reasons-­responsive action is not, and nothing in Korsgaard’s arguments even purports to establish the requisite connection to accountability.20 Despite this, I shall argue that Korsgaard’s arguments may admit of being appropriately supplemented to take account of the fundamentally second-­personal character of moral accountability in ways that enable us to draw the conclusion that it is indeed morally wrong, at least pro tanto, not to treat other animals as ends in themselves in Korsgaard’s passive sense. I shall argue, indeed, that once we accept these arguments, there is no conceptual bar to holding that we also have directed obligations to other animals so to treat them. A second-­personal conceptual framework simply tells us what we would have to accept in order to reach this conclusion. We would have to hold that other animals have a distinctive authority to demand the obligatory treatment from us. They might have this authority without being able to exercise it, in which case we would be committed to having a view about what it would be for such an authority to be exercised on their behalf. In the past, I have argued that appreciating the second-­personal character of deontic moral concepts commits us to a set of theses about morality and

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

218  Stephen Darwall moral agency where the latter includes “second-­ personal competence.”21 Moral agency, in the sense of being an apt subject of moral obligations and, as well, a member of the moral community to whom we are accountable for complying with them, requires the capacity to take up an impartial second-­ person perspective on oneself and others and have Strawsonian “reactive” (better: “holding accountable”) attitudes toward oneself and others from that point of view. So far this is a conceptual thesis. The very idea of moral obligation commits us to the proposition that moral obligations entail accountability to the moral community or representative persons and, therefore, that beings with second-­personal competence have a representative authority to hold themselves and others accountable for compliance. Accountability is always a reciprocal second-­personal matter; we can intelligibly hold accountable only those we assume can hold themselves accountable also. Indeed, part of what holding someone accountable involves is calling for them to hold themselves accountable. In addition to this conceptual thesis, I have argued for the further normative theses that we have moral obligations toward, and indeed, to any and all second-­ personally competent agents.22 This means that second-­ personal competence is not only sufficient for the representative authority to hold agents accountable, as a representative person, for compliance with the moral obligations to which they are subject. It also means that there are moral obligations toward second-­personally competent beings that are grounded in their interests as second-­personally competent. And it means further that, because at least some of these are owed to such beings, second-­person competence is sufficient also for the individual authority to hold obligors accountable to them as the very individuals to whom they are obligated. This may encourage the thought that the second-­personal framework I have defended is committed to the claim that we only have moral obligations toward such second-­personally competent agents. This by no means follows. My arguments do attempt to show that second-­personal competence is sufficient for representative authority, individual authority, and the kind of moral considerability entailed in being the object of moral obligations. But my arguments neither attempt to show, nor do they actually show, that second-­ personal competence is necessary for moral considerability of this kind or, indeed, for the individual authority of right holders or, equivalently, of obligees to whom directed obligations are held. I believe my arguments do show that second-­personal competence is both sufficient and necessary for the kind of mutual accountability of members of the moral community, understood to

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Animal Value and Right  219 include both moral agents, those subject to moral obligations, and those (representative persons) to whom they are accountable in being thus subject. In what follows, I seek to show that although FC’s arguments, taken on their own, are insufficient to show that we have moral obligations either to or toward nonrational animals, a second-­personal framework can make use of them to provide plausible arguments for both of these positions. Most plausibly, they can help to show that we do indeed have obligations toward nonra­ tional animals that are grounded, not indirectly, like Kant’s, in effects of nonrational animal abuse on our treatment of rational, second-­personally-­ competent agents, but rather in interests that humans and nonrational beings share as animals. And they can be brought to bear on more speculative arguments that we owe obligations to nonrational animals as well.

3.  Value and Good Before we consider FC’s arguments for FA, we should note some remarks Korsgaard makes about the nature of good and value. In the course of criticizing the idea that human beings have a kind of superior importance that could warrant relative disregard for other animals’ interests, Korsgaard advances what she calls the “Tethered Values” thesis. “Nothing,” she claims, “can be important without being important to someone—to some creature, some person or animal.”23 This takes the wind out of the sails of any attempt to claim that human beings stand higher in some absolute scale of importance or mattering. There is no such scale. Things matter only to beings, and they matter to nonrational beings no less than they do to rational ones. “All importance is tethered . . . to the creature to whom the thing in question is important.”24 Korsgaard points out that her Tethered Values thesis does not rule out a kind of absolute importance, since absolute importance can be understood in terms of universal importance-­to, that is, importance “to everyone [i.e., every being] for whom things can be important.”25 In particular, there is, she says, “logical space” for: The Absolute Importance of Importance-­ To: It is absolutely important, important to us all, that every sentient creature get the good things it is important-­to her to get and avoid the bad things that it is important-­to her to avoid.26

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

220  Stephen Darwall This absolute importance claim, she then says, can be equivalently formulated in the language of “good” and “good-­for”: The Absolute Goodness of Goodness-­For: It is absolutely good, good-­for us all, that every sentient creature get the things that are good-­for her, and avoid the things that are bad-­for her.27

Below we shall consider how the idea of absolute goodness figures in Korsgaard’s arguments for FH, and, by extension, for FA. But notice first some important preliminary conceptual points. First, “us” in both The Absolute Goodness of Goodness-­For and The Absolute Importance of Importance-­To presumably refers to all us animals (i.e., “everyone for whom things can be important”). Second, although these remarks do not quite claim that value and importance-­to are the very same thing as goodness-­for, it is hard to read Korsgaard differently. The notion of good-­for that seems to be in play here is the one associated with “good” in the final rather than the evaluative or functional sense. For something to be good for a being in this sense is for it to be something that being should desire and take as goal or end. Third, although the Tethered Values thesis is initially put in terms of importance-­to, this should presumably not be read in a non-­normative sense as what a being in fact cares about, at least when that being is a rational agent. Korsgaard emphasizes that rationality essentially includes the capacity to step back from desires and consider the normative question of whether they provide normative reasons for acting, whether what they present as desirable really is so. As we shall see, Korsgaard importantly distinguishes between desiring something in the final-­good-­representing way we share with other animals and valuing it as we do when we choose it as an end as only rational beings can. The valuing that is distinctive of rational agency undertakes rational commitments to continued pursuit and instrumental rationality; it necessarily involves, Korsgaard argues, giving oneself a law that purports to bind both one and other rational agents alike. More on this below. At this point, I just want to note how this affects how Korsgaard’s claims about absolute importance and goodness should be interpreted. When Korsgaard makes her arguments for FH and FA, she repeats the Tethered Values thesis, contrasting it with the view (which in The Sources of Normativity she called “substantive moral realism”) that “certain things just have value . . . and we are doing our valuing correctly when we value those

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Animal Value and Right  221 things that actually have value.”28 By contrast, the Tethered Values thesis holds that “value depends on valuing—it is the act of valuing that has to come first.”29 “[C]orrect valuing,” she adds, “confers value on its object.”30 Note: correct valuing. Value is not magically conferred by just any valuing; the valuing must be “correct.” Obviously not substantively correct in a sense that would entail substantive moral realism. Rather, we might say, procedurally correct, in the sense of Sources’ “proceduralist moral realism.”31 Value is constituted by the successful execution of a procedure involved in rationally choosing something as an end. When a rational agent reflects on their desires and comes, as rational, to choose something as an end, they necessarily take themselves to have reason to do so, and this involves, Korsgaard will argue, their taking their end to be “good absolutely” in the sense of there being “reason for any rational being to promote it.”32 To be absolutely good in this sense, she says, quoting Kant, is to be “‘an object of the faculty of desire in the judgment of every reasonable human being.’”33 If absolute goodness is understood in terms of rational choice, as what there is reason to choose to bring about or promote, and if there are no other normative notions in play (at least for rational agents), we face a normative conceptual landscape that is simply too spare to derive the moral conclusions Korsgaard wants. We have the notions of final good, or good for, that is, what is desirable and choiceworthy for individual beings, and absolute good, what is desirable and choiceworthy for all. And that’s it. We lack normative notions necessary even to formulate deontic questions of obligation and right. And there are yet other important value concepts that we lack. Korsgaard herself observes that “[v]alue is not a univocal notion—different things are valued in different ways, even as ends.”34 “The kind of value that Kant thinks attaches to persons,” she then notes, “is one in which we must respect their choices.”35 We inherit from Kant the idea that all (rational) moral agents have a “dignity” that makes them apt objects of respect.36 Respect values its object in a different way than desire and choice do. Dignity is not a good, but a value of a fundamentally different kind. To take seriously the difference between different valuing attitudes we should follow Elizabeth Anderson’s idea that these mark out different value concepts and species of value—not just different kinds of one species, like different kinds of goodness, but wholly different forms of value.37 Final goodness in Korsgaard’s sense is analytically connected to desire and choice, hence to desirability and choiceworthiness, to what we should or have reason to desire or choose. The value that moral recognition respect recognizes, on the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

222  Stephen Darwall other hand, is of a fundamentally different kind.38 Dignity’s conceptual connections are to authority, standing, and rightful treatment. It is analytically connected to the deontic in ways that Korsgaard’s notions of final and absolute good are not. Understanding different value species in relation to differing valuing attitudes suggests a more general account of normative concepts and kinds, both other moral concepts and kinds (like deontic moral concepts) and normative concepts in other areas also, like that of credibility in epistemology. All normative concepts are “fraught with ought,” in Sellars’s famous phrase.39 They all contain a common basic normative concept, to refer to which different philosophers may use different terms (“ought,” “-worthy,” “-able,” “justified,” “reason-­supported,” “fitting,” and so on) which is then combined with some attitude or other that is analytically distinctive of a more specific normative concept. Sidgwick says that ethical judgments all contain “the fundamental notion represented by the word ought.”40 Ewing tends to prefer the term “fitting,” and Gibbard generally uses “rational,” “makes sense,” or “warrants.” “Reasons firsters,” like Parfit and Scanlon, express roughly the same idea in terms of normative reasons, that is, reasons in favor of or to have some attitude or other.41 Presently, I shall argue that deontic moral concepts, like obligation, right, and wrong, are analytically related to culpability and the attitude of moral blame, and that they cannot therefore be reduced to or understood in terms of choiceworthiness, what it is good to do, or, indeed to any species of good. This will mean that even if it could be shown, for example, that the most rational choice, or what there is most reason to do, is always to treat rational beings and other animals as ends in themselves in the active and passive sense, respectively, that could not show that we have any moral obligation, directed or undirected, to do so. Good and right are fundamentally different normative kinds; neither can be reduced to the other. Even the “good” that Kant says is “the only object of practical reason,” good or choiceworthy choice or action, that is, an action’s being best supported by practical reason, is of a fundamentally different normative kind than the deontic kinds of obligatory right and wrong.42 Before I make this argument, however, it is worth noting that there is another evaluative concept and kind that is tied to another form of valuing, namely love, care, or sympathetic concern, hence another kind of value, that other animals clearly instantiate. We do not have a canonical, or even very good, term to express the concept of being worthy of, or a fitting object of, love or care, but we clearly have the concept. Being “lovable” or “precious” are

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Animal Value and Right  223 probably as good terms as any. Clearly, (many kinds, at least, of) animals are precious in the sense of being worthy of care and concern for their own sake. When we care for a being for their sake, we see them as valuable in a distinctive way and want what is good for them for their sake, that is, because they are valuable themselves. We see them as instantiating this distinctive kind of value—as being a fitting object of the distinctive valuing attitude of love or care. Care and respect are different attitudes to which different kinds of value correspond. A caricaturing example from Lawrence Blum’s Friendship, Altruism, and Morality makes the point.43 Blum imagines a scenario in which an astronaut, Joan, is circling the moon in a space module when something goes wrong with her life support system. In Mission Control, two NASA scientists, Manny and Dave (so named to conjure up a respect-­ centered Kant and a sympathy-­based Hume), do everything they can to fix the problem and save Joan’s life. We can stipulate that they both take the very same actions, with the same seriousness of purpose, Manny being moved by respect for Joan’s dignity and legitimate claims and Dave being concerned for her and her well-­being. When it becomes clear that nothing can be done, however, they respond differently. Manny turns his energies and attention to other matters that make a claim on his agency—“What other duties do I have? How can I be most useful?” Dave, for his part, is stricken with anguish on Joan’s behalf, and then with grief. “I hope she didn’t suffer too horribly. I hope she found some consoling thoughts.” Respect and care or love are different ways of valuing, and they respond to different kinds or dimensions of value. Respect responds to its object’s dignity, care to its object’s preciousness, its being worthy of care.44 Because it essentially concerns treatment, respect is an agential attitude that is manifested in our deliberations about how to act toward its object.45 Care and love, by contrast, are focused on their object’s welfare or well-­being (what we also call “their good” in this distinctive sense), and are concerned with action and treatment only if and because they affect well-­being.46 Respect is agent-­ centered, focused on one’s treatment of others. Care is agent-­neutral, focused on others’ well-­being and how they fare quite generally, whether as a result of one’s agency or not. Although love and care are agent-­neutral, they are nonetheless, like respect, focused on their object’s value as a particular individual. As Iris Murdoch emphasizes, “loving attention” to individuals in their full particularity enables us to appreciate their value and counteract discounting self-­serving preconceptions and fantasies.47 Korsgaard herself makes a similar point in FC when she quotes a passage from George Elliot’s Middlemarch that, she says,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

224  Stephen Darwall “reminds us how hard it is to keep in view—not just tell yourself, but to feel with the ‘directness of sense’ what other people’s lives” are like from their point of view.48 “How much harder, then, must it be,” she adds, “to wrap our minds around the ways in which creatures of a different species . . . might experience their own fates and their own existence.”49 When, however, we do attend carefully to the lives of other animals, it seems clear that they matter in the sense of being worthy of our care and concern in themselves. If we accept that other animals and their lives are precious in this sense, it follows directly, as I see it, that we have weighty reasons to promote their well-­being (their good in the welfare sense), since this just is what there is reason to want for them on the assumption that they are worthy of care.50 Valuing an animal, whether rational or nonrational, through care or love commits us to their mattering in a way that entails (agent-­neutral) reasons to promote their welfare. These reasons are not, in themselves, however, deontic. No moral obligations are entailed by them. We may of course be morally obligated to promote or not set back such value, but that is not conceptually guaranteed. We can admit that things matter or are important, in this sense, and nonetheless deny that they give us moral obligations without conceptual confusion.

4.  Value, Obligation, and Right Korsgaard’s aim is to argue that other animals matter in a way that necessarily does make us morally obligated to treat them as ends in themselves in the passive sense, that is, to promote their final good. Her ambition thus goes beyond simply claiming that they matter in the way I have just suggested they do. That is sufficient to give us reason to promote their good, but a moral obligation to do so does not necessarily follow. Let us turn then to Korsgaard’s argument, both for the moral obligation to comply with FH in the active sense regarding rational animals and to comply with it in the passive sense with regard to all animals, that is, to comply with FA. Central to the argument is a contrast Korsgaard makes between desiring something, on the one hand, and valuing or “setting a value” on it, on the other.51 The former is the “valenced” motivational state rational animals share with nonrational animals through which we represent the state’s objects as finally good, as something to desire and choose. What makes us rational animals is that we can think about these states—the “grounds” of our end-­ seeking—consider whether they really do present us with normative reasons

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Animal Value and Right  225 for our ends, and choose whether or not to adopt ends on that basis. When we do adopt an end, we set a value on it. We are committed to it as supported by normative reasons, and we desire to pursue it, and to continue to desire it, for these reasons. We have a reflective, second-­order endorsement or approval of acting on the desire that itself involves a first-­order approval in its valenced content. What has driven Korsgaard’s arguments in the past are what she claims are Kantian presuppositions of our so valuing our ends: “autonomy” and the value of rational persons as ends in ourselves. These elements are in FC also. When we value and adopt an end, she says, we impose a “law” on ourselves. Finding myself with a desire to grow vegetables, if I take the further step of adopting the end of having a garden and thereby set a value on it, I make a commitment to that end that “legislate[s] a kind of categorical imperative for myself.”52 Autonomy, therefore, is a necessary presupposition of rational endand value-­setting. And so is taking oneself to be an end in oneself. “[T]here are some things we must value if we are to value anything at all,” Korsgaard says, endorsing what she takes to be Kant’s view: “we have to value ourselves as ends in ourselves.”53 However, we cannot sensibly presuppose either our own autonomy alone or that just we are ends in ourselves. Any law capable of binding me would have also to bind all other rational agents. Neither could such law be legislatable only by me. It would have to be “co-­legislated” by any and all rational beings.54 Similarly, in valuing myself as an end in myself, I am committed similarly to valuing all other rational beings, and likewise, they to valuing me. Setting a value on ends, Korsgaard argues, commits any rational being to a “Kingdom of Ends [as] a republic of all rational beings, whose laws are the product of reciprocal legislation.”55 Now there are deontic senses of both autonomy and rational persons as ends in themselves—equivalently, the dignity of persons—that bring the ideas of obligation, right, and wrong along with them as conceptual entailments. If it can be shown that persons are ends in themselves in this deontic sense, it simply follows that we are obligated so to treat them. The normative concept of the dignity we ordinarily have in mind—being an end in itself in this deontic sense—is analytically tied to the attitude of (moral recognition) respect. To have this attitude toward someone just is to see the requisite treatment of them not simply as supported by good or conclusive normative reasons, but as morally obligatory, such that failing so to treat them would violate morality’s demands (the moral law). Indeed, the obligations the dignity of persons analytically brings into play are directed obligations owed to each and every individual person, entailing claim rights held against their obligors. Failing to

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

226  Stephen Darwall treat rational beings as ends in themselves not only is a wrongful violation of the moral law but also wrongs and violates persons themselves. This is plainly the sense of dignity (Würde) in the German Grundgesetz: “Human dignity shall be inviolable . . . . The German people therefore acknowledge inviolable and inalienable human rights.”56 Similarly, the ideas we normally take to be expressed by the Kantian terms “autonomy,” “self-­legislation,” and “reciprocal co-­legislation” are deontic in their nature. As we ordinarily understand the concepts we take these terms to express, it is not as if, should it be established that all rational agents are bound by laws that are the product of “reciprocal legislation” in a Kingdom of Ends, we would have to take a further argumentative step to show that these laws obligate us and that failing to comply with them is morally wrong. As we normally understand these notions, they are already implicitly deontic. It is clear, moreover, that Korsgaard herself regards this deontic aspect as already built in. Assume, as she does, that the only way one can decide whether a ground is a good reason for acting is by Kant’s universalizability test: Here is one (very informal) way to see why you might think that a universalizability test is the correct standard for evaluating a person’s reasons for action. If you do something wrong, you have wronged someone; and that means someone has a legitimate complaint against you.57

The reason why, she says, autonomy requires “reciprocal legislation,” is that the requisite laws “are supposed to be laws on the basis of which we can make claims on one another, claims that are supposed to give rise to reasons.”58 Similarly, Korsgaard understands treatment as an end in itself to require its object’s actual or possible consent.59 But the very idea of consent assumes a deontic background of entitlements. Consenting just is allowing someone to do something they would otherwise not be permitted to do. Now these aspects are indeed central to the idea of deontic morality, to the notion that persons stand under an obligatory moral law failure to comply with which is, lacking excuse, a culpable wrong. It is part of the very ideas of moral obligation, right, and wrong, that failure to do what we are, all things considered, morally obligated to do, is blameworthy, that is, fittingly subject to the distinctive attitude of moral blame. Deontic moral notions, I argue, have a conceptual connection to accountability, and to the attitudes through which we hold ourselves and others accountable—resentment, blame, and guilt.60 A conceptual connection to the attitudes through which we hold accountable is essential to deontic moral concepts’ normative profile. In the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Animal Value and Right  227 terms of the general outline of normative concepts laid out earlier, blameworthiness (culpability) is what is justifiably or fittingly blamed, and deontic moral concepts are analytically connected to that.61 Mill may have been the first to remark on this conceptual connection, although it is implicit in many writers before him and after: We do not call anything wrong, unless we mean to imply that a person ought to be punished in some way or other for doing it; if not by law, by the opinion of his fellow creatures; if not by opinion, by the reproaches of his own conscience.62

When I say that second-­personal competence is necessary for moral agency, I have this point in mind. Conscience is the capacity to take up an impartial second-­person standpoint on oneself and acknowledge wrongdoing—blame oneself—from that “representative” point of view. Morality, conceived as constituted by what we are accountable to the moral community (if you like, the Kingdom of Ends) for doing—what everyone, as a representative person, can justifiably expect of everyone—seems part and parcel of the notions of autonomy and persons being ends in themselves in the active sense as Korsgaard is thinking of them in passages of the kind to which I have just been calling our attention. What, however, in the argumentative materials she provides, can ground deontic moral claims of this kind, including a moral obligation to comply with FH and so FA? The fundamental problem is that the concepts of good in terms of which Korsgaard’s arguments are formulated—final goodness, goodness-­for, and absolute goodness—and the deontic concept of moral obligation (directed and undirected) are mutually irreducibly concepts. Premises expressed entirely in terms of the one cannot establish conclusions expressed in terms of the other. To get a deontic moral conclusion, we need a deontic moral premise. As Bernard Williams put it, “obligation out, obligation in.”63 Consider the argument for FH that a rational agent must value its ends and that doing so involves both self-­legislation and the agent’s valuing themselves as ends in themselves. Korsgaard argues that valuing an end involves a ra­ tional commitment to it as absolutely good, which means that it is good-­for all rational agents in the sense of being what they have reason to choose, or at least not interfere with, as well: Importantly, Kant takes the judgment that the end is good to imply that there is reason for any rational being to promote it—that it is good absolutely. As he says in the Critique of Practical Reason: “What we are to call

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

228  Stephen Darwall good must be an object of the faculty of desire in the judgment of every reasonable human being . . .”64

Assume that this is correct. Assume even that because it is, a rational being should not (that is, cannot have normative reason to) adopt and value an end unless it meets the test that every rational agent has decisive reason to promote (the first agent’s) pursuing that end, or at least not to interfere with that. Nothing having the distinctive normative profile of deontic morality follows. From claims about the absolute goodness of an end, as Korsgaard understands it, no moral obligations follow. Nothing has yet been said about Korsgaard’s claim that valuing ends ra­tion­ ally requires valuing oneself, seeing oneself as an end. Although, as I’ve argued, the Kantian idea of dignity is normally understood deontically as entailing entitlements to respect and as including what I would call a fundamental second-­personal authority to makes claims and demands and hold others accountable, nothing like this can follow from Korsgaard’s arguments.65 How then is it to be understood in her framework? Obviously, the idea that persons are ends in themselves cannot mean that they themselves are good in any of the senses Korsgaard canvasses—their value cannot be expressed in terms of the concepts of final or absolute goodness, or of the concept of good-­for that is related to these. So, what can it mean? It can mean only that, of every rational agent, it is true that other rational agents’ actions toward them can be justified (supportable by normative reasons) only if it is something that rational agent has reason to choose also. That is just what Korsgaard calls “The Absolute Goodness of Goodness-­For” interpreted with a scope restricted to rational beings. But again, nothing deontic can follow from this. Nothing about an obligation to treat rational beings as ends in themselves is entailed by it.

5.  Animal Right Suppose, however, that we could somehow establish that any rational agent is morally obligated to comply with FH with respect to any and all rational agents. Might the arguments Korsgaard provides for extending FH, read passively, to all animals, rational and nonrational, have force nonetheless? Here things are more promising. Interpreted actively, FH applies to rational agents as autonomous end-­setters; it does not “look behind,” as it were, their setting of ends to consider what more specifically might justify that. It treats that as a black box. It simply says that, whatever grounds it, their end-­setting has to be

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Animal Value and Right  229 consistent with the end-­setting of other rational agents in the right way and thus all end-­settings are mutually constrained by others as end-­setters. We rational animals, however, take our desires, the final good-­representing motivational states we share with other animals, as pro tanto reason giving. When we take desires like those for “food, sex, comfort, freedom from pain and fear,” and the like as giving us pro tanto normative reasons to adopt ends, then, as Korsgaard points out, we are taking our end-­setting to be justified by “things that are good for us insofar as we are animals.”66 “[T]here is no reason,” she adds, to think that because it is only autonomous rational beings who must make the normative presupposition that we are ends in ourselves, the normative presupposition is only  about  autonomous rational beings. And in fact it seems arbitrary, because of course we also value ourselves as animate beings.67 

This seems right, at least insofar as our taking our desires and feelings (that is, both taking seriously the evaluative representations these motivational states provide and the fact of the desires and feelings themselves) as giving us pro tanto reason for end- and value-­setting is a way of valuing ourselves as animate beings. If we rational animals value ourselves as animals in this sense and expect our fellow rational animals to respect our so valuing our animal nature, then why should we not be expected to respect the way in which nonrational animals value their animal nature also? This seems a promising line of argument for a moral obligation to comply with FA. In this final section, I want, albeit briefly, to sketch how this Korsgaardian point can be placed within the second-­personal account of moral obligation I defend to reach a similar conclusion. Start with the assumption that moral obligations fall on any and all second-­ personally competent agents and involve accountability to such agents as representative persons or members of the moral community. Might there be some way of arguing for moral obligations constraining our conduct towards other animals, or perhaps even as owed to them, on this basis? I believe there might. First, take second-­personal competence itself. Although we rational animals may be the only beings we know to have the kind of moral agency necessary to be subject to moral obligations and held accountable for compliance with them, many other animals seem capable of forms of second-­personal engagement and of holding accountable response. Justice Holmes famously remarked that dogs distinguish between being kicked intentionally and

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

230  Stephen Darwall accidental “stumbling,” and object to the former and not to the latter.68 Adam Smith and John Stuart Mill both discuss what they call “animal resentment.”69 Of course, part of their point is to distinguish a more “primitive” retaliatory response from a moralized form of resentment that has the impartially regulable structure of a Strawsonian attitude. Thus, Smith remarks that what resentment (of this latter kind) “is chiefly intent upon” is not retaliation or exacting retribution but to make its object “sensible, that the person whom he injured did not deserve to be treated in that manner.”70 Resentment like this is more a respectful demand for respect. But the responses of objecting dogs can seem a version of this also. When appropriately reciprocally acknowledged by their injurer, an injured dog can return to something like mutually respectful relations. Still, we do not hold nonrational animals morally responsible in the way we do rational animals (at least, second-­personally competent ones).71 Even if, however, only rational animals have the moral capacities necessary to be full members of a mutually accountable moral community—that is, a community of those subject to moral obligations but also to whom they are simultaneously accountable for compliance—that does not mean that moral obligations do not include substantial constraints on our treatment of other animals. It does not mean that there is not a moral obligation to comply with FA. If we assume, as I argue, that moral agents are all obligated to comply with FH in their conduct toward all second-­personally competent beings, it does not follow that they are not also morally obligated to comply with FA in their conduct toward all animals. And this is where Korsgaard’s argument comes in. Being a full member of the moral community in the sense I have just mentioned—being both subject to moral obligations and among those to whom those subject are simultaneously accountable—does not yet say anything about the content of moral obligations. It concerns moral obligation’s “form” rather than its “substance.”72 But it does not leave everything up for grabs. Since, I argue, persons can be held accountable only for their free choices through their own autonomous practical reasoning, the very idea of mutual accountability commits one to a robust sphere of freedom, with persons having an autonomy right to lead their own lives and make their own choices.73 Similarly, I argue that the metaethical idea of morality as mutual accountability provides a rationale for a contractualist approach to normative moral theory.74 Any contractualism—Rawlsian, Scanlonian, or otherwise—must assume some theory of the good (cf., Rawls’s “thin theory of the good”), and I have claimed that the appropriate theory concerns the good of agents conceived as second-­ personally competent.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Animal Value and Right  231 The fundamental idea is Rawls’s that to be a person is to be “self-­originating source of valid claims.”75 Put in my terms, this means that second-­personal competency gives one a shared basic second-­personal authority to make claims on other second-­personally competent agents. This is not simply the formal idea that only second-­personally competent beings are subject to moral obligations and among those to whom they are simultaneously accountable for compliance. It is the substantive idea that they are sources of claims in a way that affects the substance or content of moral obligations. More specifically, it grounds the contractualist theory that moral obligations’ content is subject to a test of hypothetical impartial choice (behind a veil of ignorance, say), or of not being reasonably rejectable, in light of interests that second-­personally competent persons share in autonomously leading their own lives on terms of mutual accountability.76 But so far, like FH, this does not look behind people’s autonomous choices and projects to consider their reasons for adopting them. It considers effects on their interests as autonomous and mutually accountable choosers regardless of what motivates or justifies their choices. It treats the latter as a “black box.” Taken in themselves, such matters as physical and emotional pain and suffering, and the quality of lived experience—for example, whether it consists of fear and torment or calm and enjoyment—lack intrinsic relevance. A purely second-­personally based contractualism would take account of them only for their instrumental effects on autonomous, mutual accountable living. But this seems too spare. Korsgaard’s point is that if we take such matters of lived human experience to matter intrinsically, then insofar as it concerns our animal nature, it is unjustifiably arbitrary not to give it weight with respect to other animals also. Suppose, then, that we want to hold that we are morally obligated not to cause torturous suffering to human beings because of what it would be like for them (given their animal nature) and not just because avoiding this pain is something they happen to choose or because of its downstream effects on autonomous, mutually accountable living. If we think that the desire to avoid torturous suffering does not simply motivate adopting the end of avoiding it but justifies it, then we will think that the apparent reasons such a desire represents actually are normative reasons. Moreover, we are likely to think that the normative reasons in this case warrant a claim, that avoiding torturous pain normally is sufficiently important in itself to human beings, is an “urgent interest” in Scanlon’s sense, and not just something they happen to prefer and choose.77

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

232  Stephen Darwall On reflection, there are many such interests that are simply impossible adequately to appreciate if we abstract from the embodied animal nature of human existence. Consider, for example, the interest in not being subjected to rape. No doubt, the evil of rape essentially involves a form of disrespect of rational personality and will. But it seems true also that it is impossible to appreciate its distinctive horror as the trauma of a deeply personal bodily violation and defilement if we abstract from our embodied animal nature. Urgent human interests that a plausible contractualism will incorporate will thus include urgent animal interests, interests we have in virtue of being not just rational (second-­personally competent) beings, but rational animals. If we take animal interests to ground claims of rational animals, it seems arbitrary not to take comparable, in some cases even identical, interests, to ground claims of nonrational animals also. I doubt that any strict in­con­sist­ ency is involved, but how might it be justified? It is true that nonrational animals are not in a position to make claims with the full second-­personality of a reciprocally recognizing transaction.78 But this is true also of many human beings, for example, of children, or even the temporarily mentally impaired, where we want to hold that they nonetheless have an authority they cannot exercise and that competent moral agents are obligated to organize themselves so that the authority can be exercised on their behalf. A contractualist normative theory, grounded in the second-­person standpoint, can thus recognize moral obligations not just toward human beings who lack second-­personal competence, but also to them. In this way it can account for directed obligations that give obligees’ claim rights that they themselves lack the competence to exercise. But if it can, there is no reason why such an approach cannot, in principle, also ground directed obligations to nonhuman animals, when interests of the very same kind are involved. Of course, we do not have to think we have obligations to other animals, with their having correlative rights against us, to think that we have moral obligations toward them. Moreover, the practical issues involved in recognizing directed obligations to animals do not arise in the latter case. We do not have attribute the individual authority to hold us accountable to animals to believe that we are morally obligated, pure and simple, to treat them in ­certain ways, for example, to comply with FA. This would make us accountable for such treatment in the way we are with any such obligation, namely, to the moral community or representative persons. And we could extend the argument further, as discussed earlier, to conclude that we are indeed obligated to other animals in a way that entails their having claim rights against us.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Animal Value and Right  233 When placed within a second-­personal framework, Korsgaard’s argument shows us what can justify these beliefs. When we take aspects of our animal nature and lived bodily experience to justify claims we make on one another, we should also think that these provide no less justification for comparable claims concerning our fellow animals.

Notes 1. Christine  M.  Korsgaard, Fellow Creatures: Our Obligations to the Other Animals (Oxford: Oxford University Press, 2018). 2. Christine  M.  Korsgaard, The Sources of Normativity (Cambridge: Cambridge University Press, 1996) and Self-­Constitution: Agency, Identity, and Integrity (Oxford: Oxford University Press, 1996). 3. Korsgaard, The Sources of Normativity, 121. 4. Immanuel Kant, Groundwork of the Metaphysics of Morals, in Practical Philosophy, trans. and ed. Mary J. Gregor (Cambridge: Cambridge University Press, 1996), 4:429. 5. Korsgaard, Fellow Creatures, 141. 6. Autonomy, Kant says, is “the property of the will by which it is a law to itself independently of any property of the objects of volition”: Kant, Groundwork of the Metaphysics of Morals, 4:441. 7. Korsgaard, Fellow Creatures, 141. 8. Not to be confused, of course, with what Kant scholars more usually call FA, i.e., the Formula of Autonomy: “the idea of the will of every rational being as a will giving universal law”: Kant, Groundwork of the Metaphysics of Morals, 4:430. 9. Korsgaard, Fellow Creatures, 17. 10. Korsgaard, Fellow Creatures, 17. 11. Korsgaard, Fellow Creatures, 20. 12. Korsgaard, Fellow Creatures, 20. 13. Korsgaard, Fellow Creatures, 21. 14. Korsgaard, Fellow Creatures, 28. 15. Korsgaard, The Sources of Normativity. 16. See e.g., Michael Thompson, “What Is It to Wrong Someone? A Puzzle about Justice, in R. Jay Wallace, Philip Pettit, Samuel Scheffler, and Michael Smith, eds., Reason and Value: Themes from the Philosophy of Joseph Raz (Oxford: Oxford University Press, 2004), 333–84; Stephen Darwall, “Bipolar Obligation,” in Russ Shafer-­Landau, ed., Oxford Studies in Metaethics, vol. 7 (Oxford: Oxford University Press, 2012), 333–58; and R. Jay Wallace, The Moral Nexus (Princeton, NJ: Princeton University Press, 2019). 17. Wesley Newcomb Hohfeld, Fundamental Legal Conceptions, ed. Walter Wheeler Cook (New Haven, CT: Yale University Press, 1923). 18. Margaret Gilbert, Rights and Demands: A Foundational Inquiry (Oxford: Oxford University Press, 2018). 19. Darwall, “Bipolar Obligation.”

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

234  Stephen Darwall 20. Stephen Darwall, The Second-­Person Standpoint: Morality, Respect, and Accountability (Cambridge, MA: Harvard University Press, 2006); Morality, Authority, and Law: Essays in Second-­Personal Ethics I (Oxford: Oxford University Press, 2013); and Honor, History, and Relationship: Essays in Second Personal Ethics II (Oxford: Oxford University Press, 2013). 21. Darwall, The Second-­Person Standpoint; Morality, Authority, and Law; and Honor, History, and Relationship. 22. Stephen Darwall, “Moral Obligation: Form and Substance,” Proceedings of the Aristotelian Society 110 (2010): 31–46 and “Bipolar Obligation.” 23. Korsgaard, Fellow Creatures, 9. 24. Korsgaard, Fellow Creatures, 10. 25. Korsgaard, Fellow Creatures, 10. 26. Korsgaard, Fellow Creatures, 10. 27. Korsgaard, Fellow Creatures, 10. 28. Korsgaard, The Sources of Normativity, 34; Fellow Creatures, 135. 29. Korsgaard, Fellow Creatures, 135. 30. Korsgaard, Fellow Creatures, 135. 31. Korsgaard, Fellow Creatures, 135. 32. Korsgaard, Fellow Creatures, 138. 33. Korsgaard, Fellow Creatures, 138; Immanuel Kant, Critique of Practical Reason, in Gregor, Practical Philosophy, 5:61. 34. Korsgaard, Fellow Creatures, 132. 35. Korsgaard, Fellow Creatures, 132, emphasis added. 36. Kant, Groundwork of the Metaphysics of Morals, 4:439–40. 37. Elizabeth Anderson, Value in Ethics and Economics (Cambridge, MA: Harvard University Press, 1993). 38. Stephen Darwall, “Two Kinds of Respect,” Ethics 88 (1977): 36–49. 39. Wilfrid Sellars, “Truth and ‘Correspondence,’” Journal of Philosophy 59 (1962): 44; Allan Gibbard, Thinking How to Live (Cambridge, MA: Harvard University Press, 2003), 21. 40. Henry Sidgwick, The Methods of Ethics, 7th edn. (London: Macmillan, 1967), 25. 41. Derek Parfit, On What Matters, 2 vols. (Oxford: Oxford University Press, 2011); T. M. Scanlon, What We Owe to Each Other (Cambridge, MA: The Belknap Press of Harvard University Press, 1998); Being Realistic About Reasons (Oxford: Oxford University Press, 2016). In Principia Ethica, rev. edn., ed. Thomas Baldwin (Cambridge: Cambridge University Press, 1993), G.  E.  Moore famously argued that good is the basic ethical concept. For arguments that Sidgwick rather than Moore was right on this fundamental point, see William Frankena, “Obligation and Value in the Ethics of G. E. Moore,” in P. A. Schilpp, ed., The Philosophy of G. E. Moore (LaSalle, IL: Open Court), 93–110; and Stephen Darwall, “Moore, Normativity, and Intrinsic Value,” Ethics 113 (2003): 468–89. One might read Frankena and Sidgwick as asserting a kind of “tethered” view against Moore’s “untethered” view of good. So viewed, the general approach to value and the normative here suggested is a “tethered” approach. 42. Kant, Critique of Practical Reason, 5:58.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Animal Value and Right  235 43. Lawrence Blum, Friendship, Altruism, and Morality (London: Routledge and Kegan Paul, 1980), 146–7. 44. Judith Butler’s concept of the “grievable” seems similar, since a disposition to feelings on another’s behalf, like grief, seem essential to love and care: Judith Butler, Frames of War: When Is Life Grievable? (London: Verso, 2009). 45. Darwall, “Two Kinds of Respect.” 46. In Welfare and Rational Care (Princeton, NJ: Princeton University Press, 2002), I argue that the concept of welfare or well-­being is analytically tied to care. What is for someone’s good in the welfare or well-­being sense is what there is reason to want for someone from the perspective of caring for that person. Equivalently, it is what, if there is reason to care for someone, there is reason to want for them. Also equivalently, it is what there is reason to want for someone for their sake. Kant uses “das Wohl” to refer to well-­being and “das Gute” to refer to what it is rational to desire and choose. 47. Iris Murdoch, The Sovereignty of Good, 2nd edn. (New York: Routledge Classics, 2001). 48. Korsgaard, Fellow Creatures, 14. 49. Korsgaard, Fellow Creatures, 14. I am indebted here to Austen Hall. 50. Darwall, Welfare and Rational Care. 51. Korsgaard, Fellow Creatures, 38–44, 138–44. 52. Korsgaard, Fellow Creatures, 143. 53. Korsgaard, Fellow Creatures, 135. 54. Korsgaard, Fellow Creatures, 150. 55. Korsgaard, Fellow Creatures, 153. 56. The Basic Law (Grundgesetz) 2016: The Constitution of the Federal Republic of Germany (May 23, 1949), trans. Axel Tschentscher. 57. Korsgaard, Fellow Creatures, 122. 58. Korsgaard, Fellow Creatures, 123. 59. Korsgaard, Fellow Creatures, 77, 176–9. 60. P.  F.  Strawson, “Freedom and Resentment,” in Gary Watson, ed., Free Will, 2nd edn. (Oxford: Oxford University Press, 2003), 72–93; Darwall, The Second-­Person Standpoint; Morality, Authority, and Law; and Honor, History, and Relationship. 61. Stephen Darwall, “Making the ‘Hard’ Problem of Normativity Easier,” in Errol Lord and Barry Maguire, eds., Weighing Reasons (Oxford: Oxford University Press, 2016), 257–78; “What Are Moral Reasons?” The Amherst Lecture in Philosophy 12 (2017): 1–24. 62. John Stuart Mill, Utilitarianism, ed. George Sher (Indianapolis, IN: Hackett Publishing Co., 2002), 48. 63. Bernard Williams, Ethics and the Limits of Philosophy (Cambridge, MA: Harvard University Press, 1985), 181. 64. Korsgaard, Fellow Creatures, 138. 65. Although I cannot make the case properly here, it is worth noting that the same problem affects Kant’s own discussion. In The Critique of Practical Reason, indeed, Kant says that the moral law applies only to finite rational agents. Perfect rational agents stand under the Fundamental Law of Pure Practical Reason (“So act that the maxim of your will could always hold at the same time as a principle in a giving of universal law” (5:57, 22, 30)). Practical reason gives only “(to the human being) a universal law

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

236  Stephen Darwall which we call the moral law” (5:31). The difference, Kant says, is that obligation, oughts, and imperatives apply only to rational agents having arational “finite” motivational sources other than pure practical reason. Obligation, Kant holds, “involves not just practical necessity (such as a law in general asserts), but also necessitation” (Kant, The Metaphysics of Morals, 6:222–3). “Necessitation” is a technical term for Kant that refers to the psychological process in finite rational agents, involving the feeling of respect, through which they are able to comply with practical (and for them, moral) law (Critique of Practical Reason, 5:20). Through respect we feel imperatives or oughts. “An imperative, that is, a rule indicated by an ‘ought,’” Kant writes, “expresses objective necessitation to the action and signifies that if reason completely determined the will the action would take place in accordance with this rule” (Critique of Practical Reason, 5:20). “They [imperatives] say that to do or to omit something would be good, but they say it to a will that does not always do something just because it is represented to it that it would be good to do that thing” (Groundwork of the Metaphysics of Morals, 4:413). This means that Kant does not even recognize moral obligation and the right as something that could make an independent normative claim on the will. The morally obligatory is simply the choiceworthy (the good to do), as it appears to a less than fully rational agent. 66. Korsgaard, Fellow Creatures, 144. 67. Korsgaard, Fellow Creatures, 144. 68. Oliver Wendell Holmes, Jr., The Common Law, ed. Stephen  L.  Carter (Chicago, IL: ABA Publishing, 2009), 2. 69. Adam Smith, The Theory of Moral Sentiments, ed. D.  D.  Raphael and A.  L.  MacFie (Indianapolis, IN: Liberty Classics, 1982), 104; Mill, Utilitarianism, 51. For discussion of Mill, see Stephen Darwall, “Justice and Retaliation,” in Honor, History, and Relationship. 70. Smith, The Theory of Moral Sentiments, 96. 71. See Michael Tomasello, A Natural History of Human Thinking (Cambridge, MA: Harvard University Press, 2014) and A Natural History of Human Morality (Cambridge, MA: Harvard University Press, 2016), for an extended argument that the second-­personal competence necessary for moral agency is not shared by other animals and for a speculative account of its evolutionary basis in human beings. 72. Darwall, “Moral Obligation: Form and Substance.” 73. Darwall, “Moral Obligation: Form and Substance.” 74. Darwall, The Second-­Person Standpoint, 300–22. 75. John Rawls, “Kantian Constructivism in Moral Theory,” The Journal of Philosophy 77 (1980): 544. 76. In “Kantian Constructivism in Moral Theory,” 525, Rawls stipulates that the parties to the original position have “highest-­order interests” in exercising their moral personality (living on terms of mutual respect) and in rationally choosing and revising their own conceptions of the good and a “higher-­order interest” in “protecting and advancing their conception of the good.” 77. T. M. Scanlon, “Preference and Urgency,” The Journal of Philosophy 72 (1975): 655–69. 78. Stephen Darwall, “Demystifying Promises,” in Hanoch Scheinman, ed., Promises and Agreements: Philosophical Essays (Oxford: Oxford University Press, 2011), 255–76.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Animal Value and Right  237

Bibliography Anderson, Elizabeth. Value in Ethics and Economics. Cambridge, MA: Harvard University Press, 1993. Blum, Lawrence. Friendship, Altruism and Morality. London: Routledge and Kegan Paul, 1980. Butler, Judith. Frames of War: When is Life Grievable? London: Verso, 2009. Darwall, Stephen. “Two Kinds of Respect,” Ethics 88 (1977): 36–49. Darwall, Stephen. Welfare and Rational Care. Princeton, NJ: Princeton University Press, 2002. Darwall, Stephen. “Moore, Normativity, and Intrinsic Value,”  Ethics 113 (2003): 468–89. Darwall, Stephen. The Second-Person Standpoint: Morality, Respect, and Accountability. Cambridge, MA: Harvard University Press, 2006. Darwall, Stephen. “Moral Obligation: Form and Substance,” Proceedings of the Aristotelian Society 110 (2010): 31–46. [Reprinted in Darwall, Morality, Authority, and Law: Essays in Second-Personal Ethics I. Oxford: Oxford University Press, 2013.] Darwall, Stephen. “Demystifying Promises,” in Hanoch Sheinman, ed., Promises and Agreements: Philosophical Essays, 255–76. Oxford: Oxford University Press, 2011. [Reprinted in Darwall, Honor, History, and Relationship: Essays in SecondPersonal Ethics II. Oxford: Oxford University Press.] Darwall, Stephen. “Bipolar Obligation,” in Russ Shafer-Landau, ed., Oxford Studies in Metaethics, vol. 7, 333–58. Oxford: Oxford University Press, 2012. [Reprinted in Darwall, Morality, Authority, and Law: Essays in Second-Personal Ethics I. Oxford: Oxford University Press, 2013.] Darwall, Stephen. Honor, History, and Relationship: Essays in Second-Personal Ethics II. Oxford: Oxford University Press, 2013. Darwall, Stephen. Morality, Authority, and Law: Essays in Second-Personal Ethics I. Oxford: Oxford University Press, 2013. Darwall, Stephen. “Making the ‘Hard’ Problem of Moral Normativity,” in Errol Lord and Barry Maguire, eds., Weighing Reasons, 257–78. Oxford: Oxford University Press, 2016. Darwall, Stephen. “What Are Moral Reasons?” The Amherst Lecture in Philosophy 12 (2017): 1–24. http://www.amherstlecture.org/darwall2017/index.html. Darwall, Stephen. “Justice and Retaliation,” in Honor, History, and Relationship: Essays in Second-Personal Ethics II, 50–71. Oxford: Oxford University Press, 2013. Ewing, A.  C. “A Suggested Non-Naturalistic Analysis of Good,” Mind 48 (1939): 1–22.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

238  Stephen Darwall Frankena, William. “Obligation and Value in the Ethics of G.  E.  Moore,” in P.  A.  Schilpp, ed., The Philosophy of G.  E.  Moore, 93–110. LaSalle, IL: Open Court, 1942. Gibbard, Allan. Thinking How to Live. Cambridge, MA: Harvard University Press, 2003. Gilbert, Margaret. Rights and Demands: A Foundational Inquiry. Oxford: Oxford University Press, 2018. Hohfeld, Wesley Newcomb. Fundamental Legal Conceptions, edited by Walter Wheeler Cook. New Haven, CT: Yale University Press, 1923. Holmes, Oliver Wendell, Jr. The Common Law, edited by Stephen  L.  Carter. Chicago, IL: ABA Publishing, 2009. Kant, Immanuel. The Metaphysics of Morals, in The Metaphysics of Morals, translated and edited by Mary Gregor. Cambridge: Cambridge University Press, 1991. [References are to page numbers of the Preussische Akademie edition.] Kant, Immanuel. Groundwork of the Metaphysics of Morals, in Practical Philosophy, translated and edited by Mary  J.  Gregor. Cambridge: Cambridge University Press, 1996. [References are to page numbers of the Preussische Akademie edition.] Kant, Immanuel. Critique of Practical Reason, in Practical Philosophy, translated and edited by Mary J. Gregor. Cambridge: Cambridge University Press, 1996. [References are to page numbers of the Preussische Akademie edition.] Korsgaard, Christine  M. The Sources of Normativity. Cambridge: Cambridge University Press, 1996. Korsgaard, Christine M. Self-Constitution: Agency, Identity, and Integrity. Oxford: Oxford University Press, 2009. Korsgaard, Christine M. Fellow Creatures: Our Obligations to the Other Animals. Oxford: Oxford University Press, 2018. Mill, John Stuart. Utilitarianism, edited by George Sher. Indianapolis, IN: Hackett Publishing Co., 2002. Moore, G. E. Principia Ethica, revised edition with the preface to the (projected) 2nd edition, and other papers, edited by Thomas Baldwin. Cambridge: Cambridge University Press, 1993. Murdoch, Iris. The Sovereignty of the Good, 2nd edn. New York: Routledge Classics, 2001. Parfit, Derek. On What Matters, 2 vols. Oxford: Oxford University Press, 2011. Rawls, John. A Theory of Justice. Cambridge, MA: The Belknap Press of Harvard University Press, 1971. Rawls, John. “Kantian Constructivism in Moral Theory,” The Journal of Philosophy 77 (1980): 515–72.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Animal Value and Right  239 Scanlon, T.  M. “Preference and Urgency,” The Journal of Philosophy 72 (1975): 655–69. Scanlon, T. M. What We Owe to Each Other. Cambridge, MA: The Belknap Press of Harvard University Press, 1998. Scanlon, T.  M. Being Realistic About Reasons. Oxford: Oxford University Press, 2016. Sellars, Wilfrid. “Truth and ‘Correspondence,’” Journal of Philosophy 59 (1962): 29–56. Sidgwick, Henry. The Methods of Ethics, 7th edn. London: Macmillan, 1967. Smith, Adam. The Theory of Moral Sentiments, edited by D.  D.  Raphael and A. L. MacFie. Indianapolis, IN: Liberty Classics, 1982. Strawson, P.  F. “Freedom and Resentment,” in Gary Watson, ed., Free Will, 2nd edn., 72–93. Oxford: Oxford University Press, 2003. Thompson, Michael. “What Is It to Wrong Someone? A Puzzle about Justice,” in R. Jay Wallace, Philip Pettit, Samuel Scheffler, and Michael Smith, eds., Reason and Value: Themes from the Philosophy of Joseph Raz, 333–84. Oxford: Oxford University Press, 2004. Tomasello, Michael. A Natural History of Human Thinking. Cambridge, MA: Harvard University Press, 2014. Tomasello, Michael. A Natural History of Human Morality. Cambridge, MA: Harvard University Press, 2016. Tschentscher, Axel, trans. The Basic Law (Grundgesetz) 2016: The Constitution of the Federal Republic of Germany (May 23, 1949). http://papers.ssrn.com/sol3/ papers.cfm?abstract_id=1501131. Wallace, R. Jay. The Moral Nexus. Princeton, NJ: Princeton University Press, 2019. Williams, Bernard. Ethics and the Limits of Philosophy. Cambridge, MA: Harvard University Press, 1985.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

11 Juridical Personality and the Role of Juridical Obligation Barbara Herman

This paper is about the conjoined moral ideas of juridical personality and juridical obligation as set out in Kant’s major work of political philosophy, The Doctrine of Right, the first half of his Metaphysics of Morals.1 Across interpretations, much attention has been paid to Kant’s seemingly free-­standing idea of moral personality: to the actualization of autonomy, to the distinctive account of moral motivation, and to the moral terms of relation between persons as ends in themselves. That this idea might be incomplete, that there might be, at least for human beings, an equally important moral notion of juridical personality, has not been appreciated. Partly this is because juridical law-­abiding action does not carry any heightened moral value (even a “race of devils” can do it), and partly, I suspect, because the wide variation in possible juridical institutions robs this normative realm of the universality that Kantian morality calls for. This paper argues that, to the contrary, articulated juridical personality, the self-­conception of a person in a well-­formed state, is a condition of effective moral personality: in important regions of human action and interaction, juridical duties are the possibility condition of our acting well. There might be a kind of rational being that can realize its moral personality in action on its own. We can do this only in a certain kind of community with others.

1 When I started seriously studying Kant’s moral philosophy, it was near commonplace to dismiss the Doctrine of Right as late work, sloppy or editorially compromised. Even when, thanks to Mary Gregor’s translation in the 1990s, the whole of the Metaphysics of Morals became available to English readers, the Doctrine of Right still played a secondary role in most discussions of Kant’s Barbara Herman, Juridical Personality and the Role of Juridical Obligation In: Normativity and Agency: Themes from the Philosophy of Christine M. Korsgaard. Edited by: Tamar Schapiro, Kyla Ebels-­Duggan, and Sharon Street, Oxford University Press. © Barbara Herman 2022. DOI: 10.1093/oso/9780198843726.003.0011

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Role of Juridical Obligation  241 ethical theory, though a more prominent one as a stage in the social contract tradition. As a result, its central question has been taken to be about the legitimacy and limits, perhaps the moral legitimacy and limits, of public law and coercion. A standing point of contention concerns the status of the Doctrine of Right’s “Universal Principle of Right” (UPR)—whether it instances the moral law, is a fourth (or fifth?) variant of the categorical imperative, or is a principle of a different order, say, of free embodied rational persons interacting in space and time.2 As I’ll explain, I have come to regard the Doctrine of Right, along with its partner, the Doctrine of Virtue, as providing the first principles of a unified system of morals, of duties, for beings like us—with our natural and rational capacities, situated in (roughly) our material conditions. They are the necessary practical complement of the foundational arguments in the Groundwork and Critique of Practical Reason. Why necessary? Because, as Kant plainly insists, the foundational arguments have as their target the wills of finite rational beings as such, aiming to answer the question whether and how such a rational nature alone supports their being under obligation. The materials of the foundational arguments lack the resources to generate our duties. Not as a flaw but as a feature. Generations of Kant’s readers have been drawn to think otherwise through Kant’s Groundwork examples, by the moral compass he finds in the universalization test of the formula of universal law, and by the intuitively powerful idea of treating people as ends in themselves in the Formula of Humanity. Yet, as critics delight to point out, the examples resist providing a principle for generating duties. And as any scout or mountaineer will tell you, a compass without a map, or some other means of orientation, is of limited use. It can tell you whether you are headed toward the magnetic north pole, but not whether 27º NNW will get you to your destination. Nor can the Formula of Humanity provide the independent criterion many hope for: in application it depends on the Formula of Universal Law. You fail to treat another as an end in themselves when your maxim of action (that includes their proposed treatment as an end) is one they cannot share: that is, will as a universal law without contradiction (the validity, as opposed to the usefulness, of a maxim cannot depend on whose maxim it is).3 So how did this interpretive confusion come about? When the examples are introduced in the Groundwork, Kant has just finished arguing that if there is a principle of good willing, it can be so only because of its form, not its matter, and if it is to obligate, its form must be the form of a law, and so be universal. It follows that any maxim whose form is

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

242 Barbara Herman inconsistent with the form of law cannot be a maxim of a good will. What remains to be shown is that this analysis of good willing tracks what counts as moral in a non-­accidental way. If clear examples of actions contrary to duty have maxims that are inconsistent with universal form (cannot be willed a universal law without contradiction), then the analysis is on track. A proof of concept. But note, what we get is this. If we deploy concepts of property or promise in our maxims, then a maxim that in effect asserts that the concept holds for you but not for me cannot be willed a universal law and is disallowed or invalid. The concepts of property and promise are simply assumed. As they have to be for the proof of concept. A validity test does not require assessment of the truth of premises to reject a faulty inference. This gives us a positive reading of the same data that fuels the infamous “emptiness” objection: that any employment of the categorical imperative assumes some moral duty to derive its result. The emptiness objection is right about the data and wrong about the conclusion it draws. The objection arises from the conflation of two very different philosophical tasks that Kant kept separate: one is a radical approach to a metaphysical question (what would it take for a principle of morality to be possible for dependent rational beings?); the other is a generative normative project (given what we are like—human beings ­ ­capable of moral agency—what are our duties?).4 The Groundwork has powerful things to say about the former, and despite appearances, almost nothing about the latter.

2 The task of determining the content of our duties is delegated to the Metaphysics of Morals. Although both parts of the Metaphysics of Morals ­contain principles, examples, and elaborations of duties—of right and of ethics5—the presented array is not meant to be exhaustive. Here the arguments are generative, supplying something like the categories or first principles for duties and exhibiting some stretch of what they look like in concreto, as Kant likes to say. Just as the categories of understanding supply a priori concepts that make our ordered experience of nature (and so science) possible, the moral categories of right and ethics introduce a priori concepts for a rational, law-­governed system for human conduct and relations. They give us the basic terms of the moral project (a practical science, Kant might call it).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Role of Juridical Obligation  243 To derive duties, the Metaphysics of Morals uses two distinct generative principles, one about free action (roughly: an action counts as free only if it is compatible with like freedom for all), one about necessary or obligatory ends (which are only two: self-­perfection and the happiness of others). While they might appear to establish two independent spheres of duty, I favor a reading that has the two principles together providing the elements of a unified and dynamic system of moral duties for human persons. The more controversial claim I want to make is that the anchor of the system is on the juridical side. And this is not just to claim that juridical duties have some kind of priority in the system, but that it’s on the juridical side that the central Kantian idea of the equal moral status of persons is secured. Now one might think that moral status in the Kantian system is already determined by the species marker “rational being.” However, if we think of moral status as what duties respect and preserve, the species marker is not sufficient. We need to know what about us as human beings enters the space of moral concern. Surely something about our bodies, something about the kinds of relationships we have, something about our projects, private and collective, something about the fact that we live in structured social groups. Knowing that we are all rational beings doesn’t tell us nearly enough. In the Doctrine of Right, the status of human persons as a kind owed obligations is framed in the terms of the four elements of “innate right” (MS 6:237f.).6 They comprise our independence (from other persons), our equality (with other persons), our separateness (as responsible agents), and our autonomy as a thinker (reasoner, deliberator . . . ). This is our standing as free persons: we are not to be (in principle) constrained by others’ choices, not bound (as property) to another, not guilty for another’s deeds, not prevented from speaking our own mind. How this status is to be realized for all persons is articulated and secured by a system of juridical institutions and attendant duties. This is the work of a doctrine of right. Some features of right are general. No system of duties that respects innate right can tolerate caste or slavery or any form of indentured servitude. There must be juridical protections for life and limb. These are not legal forms of moral duties or individuals’ rights in a state of nature then drawn into public law for instrumental reasons. Having the enforceable right not to suffer assault on one’s life or body is necessary for the independence and integrity of the person: to be free from the private authority of others. The law that distinguishes assault and battery from innocent contact thus creates the boundary-­ space of the human body. Its formal point is not to protect persons from injury as such, but to define the space of free action, and from that, determine what will count as wrongful interference or injury.7

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

244 Barbara Herman Other features of right are formally necessary, though particular in content—for example, resolving how we can, consistent with innate right, withdraw material means for our own extended purposes, given that each is equally entitled to set ends and so take means, and using material means excludes others. What would entitle us to do that, consistent with their equal moral status? Kant’s answer is that a system of property rights (and whatever institutions are needed to support the system) is necessary to secure both productive use and equality. Whether a particular system of property succeeds from the perspective of right is a function of the freedom from dependency it secures.8 In general, the problems solved by duties of right come from the fact that persons will do things for production and reproduction that, considered as natural actions, leave them morally vulnerable from the point of view of innate right (the way land use excludes others; that domestic life inevitably creates relations of dependence that threaten subordination). If it’s not reasonable to avoid affecting one another in these ways (and how could it be?), and we have no a priori norm for how things ought to go, then there is a threat, a kind of moral impossibility, in the bare facts of human existence. None of this is about our being bad or selfish or weak: the human life form itself introduces fundamental moral problems. Kant’s argument is that we require some kind of juridical order, one that includes coercive enforcement, to make these naturally problematic relations morally possible by giving persons delimited powers to act that they would otherwise lack, consistent with the freedom of all. No exemplary action of one person (or of many acting on their own or in voluntary concert) can give another the status of being “free from interference.” It’s not that things would necessarily go badly; people might treat each other well. It’s that dependence on the good will of others is not equivalent to recognition of a person’s equal status. This goes some way to explaining why the system of duties cannot begin with principles of good willing, but has as its first concern enforced norms for external actions and their effects. Setting aside misplaced Kantian orthodoxy, this makes intuitive moral sense. We should want a moral theory to register the significance of what we do in the world as well as what we will—whether it matters that the unintended effects of well-­intentioned actions cause harm, for example. We need to know what kinds of things we are and are not free to do, what effects we can be responsible for, before we can set ends or assess our maxims. If the only path to my health is through the medicine in your

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Role of Juridical Obligation  245 cabinet, there are questions about possession and need that can’t be answered from the perspective of rational volition alone. The priority of right insures two additional things: shared concepts as we move into ethics (about, e.g., family, work relations, promise), and deliberative orientation for imperfect duties. Consider beneficence or gratitude, duties that depend on what is ours to hold or give. The laws that make property possible, by way of rules of possession, loss, transfer, taxation, and so on, implicitly give an account of what property is morally for. That is: if a system of property is for the support of private purposes, then its rules will reflect that, and property-­burdening duties and taxes will be seen as impingements on freedom—to be resisted and limited. If, on the other hand, the rules point to a system that enables common use of the earth’s resources, one might view the limits on use that come with duties and taxes as more about sharing and common work than about loss, so a freedom-­supporting good. It may not be easy to read purpose from law, and what it is can be widely contested, but that doesn’t defeat so much as reinforce the importance of the connection. Although I won’t be able to go fill out the claim here, ethical duties, duties of private or interpersonal morality, descend from the basic framework of right.9 Think of the variety of duties to self and others about managing wealth, duties about preserving and developing one’s mental and physical abilities, duties that concern others’ physical and practical well-­being, duties of truthfulness in speech and relationships, duties of respect—all connect to the essential values of innate right: of maintaining and preserving in­de­pend­ ence, equality, separateness, and sound deliberative reasoning. Although moral status and the conditions of agency are initially set in the terms of public morality, what they amount to in the relations between persons is not clear until the full system of duties, public and private, is laid out. But the root idea is an old one: where public institutions are crude, or corrupt, or absent, where freedom to act and speak is wrongfully curtailed, what individuals can do for the sake of their ethical ends is gravely compromised.10 The obvious obstacle to pursuing this view of Kant’s moral project lies in the combining of the different sources of obligation of the two spheres of duty. It’s not just that ethical duties are about ends and juridical duties about external actions. The principles are profoundly different in what they require of agents. Ethical duties require that we act out of recognition of the value content of duties—that we act for the sake of our own perfection and the happiness of others (in effect, that we act “from duty”). Juridical duties are indifferent to incentives and ends: it’s all about external conformity, not value

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

246 Barbara Herman conformity; a nudge or a sanction will do. How then could ethical duties depend on juridical ones? Indeed, how could Kant’s idea of right even be, as he says it is, a moral concept? Or juridical obligation any species of moral obligation? A host of issues cluster here. Whether we’ve understood what makes a principle a moral principle or a duty a moral duty for Kant. Whether there is an essential connection between any moral duty and a (formula of) a cat­e­go­ ri­cal imperative. Could the UPR be a formula of the categorical imperative (I  think it pretty clearly is not)? And what about the importance of “acting from duty”—did we mistake its importance for Kant?11 On the other hand, might we be missing something about the intersection of juridical duties and moral character? What if moral character in us is not just the having of a stable good motive, but is instead a norm- or principle-­constituted power, partly ours by nature, partly ours to create together in a juridical order? After all, the juridical is about freedom of action, and the moral is about “freedom of will”; it would be more than odd if the two ways that persons are free didn’t amount to something together.

3 This brings us to the idea of “juridical personality”: the effect on what we’re like as persons of standing in juridical relations. One might hope that if moral personality is “nothing other than the freedom of a rational being under moral laws” (MS 6:233), juridical personality would be a species of moral personality, typed by a kind or application of moral law. But the laws that make up the juridical condition aren’t exactly moral laws: they are human made, and they do not obligate us directly (no “inner obligation”), but constrain by way of a permission granted to others to coerce violations (in anticipation and in response). Moreover, the freedom at issue isn’t so  much the freedom of the individual as the freedom of many persons united in accordance with a (universal) law (MS 6:230). There’s nothing problematic about the idea of collective reciprocal freedom; it would, ­however, call for some rethinking of the genus of which juridical personality is a species. A different approach might try to add juridicality to the line-­up of human predispositions—animality, humanity, morality: our powers for the good. Juridicality might fit between humanity and morality. What makes this a plausible conjecture is the fact that the juridical condition involves the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Role of Juridical Obligation  247 introduction of new rational powers—kinds of things we can do if we stand together in certain ways, none of which are available to us as agents acting separately, regardless of our good wills. Any amplification of our rationally determined powers alters what we are like as persons, whether or not it makes us better persons. Such a view looks promising. Curiously, the story of juridical powers begins not in the Doctrine of Right itself but in the Introduction to the whole of the Metaphysics of Morals. It’s an important thread to follow. The Introduction starts with an account of the kind of will that can be under obligation followed by a survey of the general forms obligation can take (the obligatory, the forbidden, etc.). The piece we need comes next, where, in just a few paragraphs, Kant lays out how we should look at human action given the view of will and obligation. Rather than seeing action as a matter of changes in the natural world caused by a rational volition, the action itself is to be regarded as a moral event in its own kind of order. That is, just as the necessary connections between natural events are explained by natural laws in terms of which we understand both normal and deviant effects, so in the moral case, there is moral causality, with its law, that explains moral connections between actions under obligation and effects. This allows for the asking and answering of questions about the activity issuing from the rational will that lack traction in the language of natural causality. Suppose, concerned with questions of responsibility, we need to know where an action ends. In natural terms, I think there is no clear answer. Is it when an action’s causal influence peters out? Does it ever? Is it when it no longer plays a leading role in explaining how things go on? Or perhaps when the counterfactuals play out both ways? But why take such a model of explanation to be what matters? After all, even if event E would have happened had I not A’d, that I did A could still matter morally in the history of E. For Kant, principles of moral causality (that is, of obligation) connect effects to choice and give completion conditions for action (when, for example, culpability for consequences outstrips intention). Imputation is the doctrine that explains how moral effects are assigned to actions under obligatory law. The details of this are important. Imputation begins with the idea of the deed. Kant says: “An action is called a deed insofar as it comes under obligatory laws and hence insofar as the subject, in doing it, is considered in terms of the freedom of his choice” (MS 6:223). It follows that “[b]y such an action the agent is regarded as the author of its effect.” That is, through the idea of the deed, both action and effect can be imputed to the agent. Indeed, being open to imputation is what

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

248 Barbara Herman distinguishes a “person” from a “thing.”12 Imputation, Kant says, marks out our moral personality (in contrast to “psychological personality”) as: “the freedom of a rational being under moral laws.” But that doesn’t quite mean what one might think, for as Kant elaborates: “a person is subject to no other laws than those he gives to himself (either alone or at least along with others)” (emphasis added). The disjunction matters. Although we are talking about the person as a subject with freedom of choice under laws of obligation, the laws referred to are not in the register of the Groundwork’s law of autonomous self-­legislation but are the laws of ethics and right. Kant’s idea of a deed can be seen as part of a general approach to action and activity in living things. Activity is to be understood through its own causality, identified as a kind according to a specific governing law. So, at one extreme, if the activity source is the autonomic nervous system, the activity or movement—one’s heartbeat or breathing—is regarded as falling under a biological law of the organism. Breathing is then assessed by its functional role— supplying adequate oxygen—not, say, in terms of its adding CO2 to its environment. If at the other extreme the activity source is the free choice of a subject, the relevant law is a law of obligation. So if the context is one in which there is the potential for harmful interaction (an overheated argument, say), the subject’s activity—her deed—is to be described according to the relevant duty (no disrespectful exchange), whether or not her choice of words is taken with an eye toward the obligation that applies. The moral perspective of the deed is third personal, in this sense: the assessment does not refer back to the subject’s principle of willing, or to the action’s positive (or negative) effects on others, but looks to whether the deed is in conformity with a duty or not (MS 6:224). So if, suddenly thinking about something else, a hot-­headed duty-­ignoring arguer leaves the scene without saying anything, her deed conforms with duty (the “matter” of the obligation: no disrespectful exchange); there is then no imputable failure, whatever flaws of character may be present. Kant continues: A deed that is contrary to duty is called a transgression (MS 6:224).13 If the transgression is unintentional, it is a mere fault; if it is intentional (where the deed is “accompanied by consciousness of it being a transgression”), it is called a crime. Activity that does not have the status of a deed cannot be a transgression in either sense. Because it is not sourced in free choice, it is not under obligatory law. Now, when we think in terms of volition, we think of action-­as-­willed either for its own sake or as a means to an end adopted by the agent. By contrast, the deed is the external side of free choice; that puts it in the space of

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Role of Juridical Obligation  249 effects, not ends. Whether the effect of an action counts as a moral effect depends on the law of obligation governing the deed. If I do what I am required to do, if my conduct is in keeping with what is (morally) owed, it has no imputable effect at all: neither the good nor the bad effects belong to my deed (MS 6:228). As an event in the natural world, my action has all sorts of effects. As a deed, a required action has none—no moral effects, that is. Suppose what is required is that I do X for A (repay a debt, say), and doing X will help B. Suppose my aim is to benefit B by Xing, but as it happens my Xing has a detrimental effect on B.  I may be disappointed, but my deed is not a transgression and the detrimental effect doesn’t belong to it. And this is true even if I am obligated to benefit B. It is not up to me which law of obligation applies or has priority.14 If, however, someone aims beyond what is owed, doing “more in the way of duty than he can be constrained by law to do” (MS 6:227), her action is meritorious and the good results are imputed to her. I am obligated to provide some help and I do more, or, in order to respect your right I make a sacrifice beyond what the duty could require of me (I have discretion about taking extra steps, but they are steps that belong to the duty’s grounding value). By doing more in the way of duty I am the moral author of the good results that follows. But if I do less than what is owed, less than what the law requires, the action is morally culpable, and any bad results of acting this way can be imputed to me as their author.15 The idea seems to be this. Conformity of action to obligatory law is a kind of safe harbor. If you are there, the moral causality of your action is finished, regardless of further natural effects, good or bad, expected or not. One might say that it is the duty, not you, that owns the effects. If you exceed the requirements of obligatory law and do “more in the way of duty,” it is your doing that takes the duty’s concern (its value) further than was required, so the good effects are imputed to the deed and to you as its author. Actions that are contrary to duty are still deeds (an obligatory law applies). If there is a good effect, it is, morally speaking, an accident (there is no principle of moral causality that connects the action with the good effect). But if there are bad effects, because the deed is a transgression there is no safe harbor; the subject acted on her own authority. Even if the transgression is unintentional (the subject fails to see that a known duty applies or makes a mistake about the grounds of obligation), not being where she ought to be, she lacks the protection of the safe harbor: objectively, she’s in unprotected waters. How things are subjectively can still make a difference. There can be a question about the degree of imputation: the extent of authorship of the deed and its effects may be treated

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

250 Barbara Herman differently depending on how easy it was for the subject to have avoided the error (MS 6:228).16 Now, among the unexpected effects caught in the net cast by imputation are punishment and reward. So when Kant says that “[t]he rightful effect [‘der rechtliche Effekt’] of what is culpable is punishment” (MS 6:227), he is not talking about punishment as the correct response to wrongdoing; punishment is rather a necessary and imputable effect of wrongdoing. And if, under imputation, punishment is a “rightful effect” of a deed, the acting subject is its author.17 This follows from the moral metaphysics of action and causality. Formally, reward and punishment are not contingent effects of a deed; they belong to the complete (moral) causal path that is a function of the obligatory law that applies. And if in the domain of right what is wrongful has punishment as its imputable necessary effect, there is no need in the theory for any of the usual justifications for punishment (and, as is obvious, Kant doesn’t offer any). A deed contrary to right is to-­be-­punished. The specifics of hard treatment for any type of wrongdoing are matters to be decided by the state—within the moral terms of the justificatory account in the doctrine of imputation.18 That leaves the question why right is the sort of law of obligation that has this structure of imputation. To find the answer we need to take a close look at the concept of right as Kant introduces it.

4 The Doctrine of Right begins with the concept of right as a moral concept (i.e., not as a juridical concept): “the sum of the conditions under which the choice of one can be united with the choice of another in accordance with a universal law of freedom” (MS 6:230). It concerns external relations and actions as a product of free choice insofar as they can affect others’ freedom. According to the UPR: “Any action is right if it can coexist with everyone’s freedom in accordance with a universal law, or if on its maxim the freedom of choice of each can coexist with everyone’s freedom in accordance with a universal law” (MS 6:230).19 Kant says two things follow. First, that a hindrance to a rightful action is wrong (“it cannot coexist with freedom according to a universal law”), and second, that there is no requirement that UPR be anyone’s maxim, since one either does or does not impair external freedom by one’s external actions, regardless of one’s maxims. As restated: agents are not

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Role of Juridical Obligation  251 morally required to limit their own freedom; reason, as a postulate, instead says that freedom is limited (by the conditions of UPR) and “may be actively limited by others” (MS 6:231). Kant then argues: because interference with right is wrongful, it is not wrongful to hinder it. That is, since a wrongful action is not a valid exercise of free choice, it has no moral standing to resist coercion that hinders or counteracts it. In this way right entitles one to the completion of one’s action. Not as what one wants, but as what right provides. The first thing to notice is that right (or having a right) is essentially relational. To say that my acting a certain way is rightful is to say that others are obligated not to hinder it. Others’ interference with what I have a right to do is wrongful. If there were no interacting others, there would be no domain of right. The second thing is that so far there is no content to right. It is a strictly formal notion: the bare concept of right does not tell us which acts are right or who is authorized to hinder wrongful action. The third thing is about the juridical incentive. If we are not required to make UPR our principle in acting, and so not required to act out of respect for others’ freedom, it can seem reasonable to conclude that the point of UPR is to justify state coercion, and that right is not in any essential way a part of the motivational–conceptual–agential repertoire of the moral agent. While Kant does say that one can act as right requires from an ethical incentive (MS 6:220), the point of the remark in context is about a back-­up in case of the unavailability of external constraint. One can and ought to be ethically attached to the moral concept of right: the sum of juridical law; but that’s not the incentive our acting juridically requires. Still one might ask: wouldn’t it be better if one did act as right requires out of the ethical incentive? That is, even if the state turns out to be necessary even for the good-­willed, and even if general conformity to law can only be secured through the apparatus of coercion, shouldn’t those of us who can bring law-­abiding into our lives ethically do so? And if not, wouldn’t that impeach juridical obligation as a well-­ formed part of morality and juridical personality as marking a moral power? In thinking about the two incentives, it is natural to focus on the ways they bring about required action: targeted natural causality versus acting from moral principle. In that light, the juridical incentive—whatever coercible pathological interests can be engaged—looks second class, even regrettable, from a moral point of view. The question is whether this is correct. Might the juridical incentive nevertheless be a kind of moral cause?

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

252 Barbara Herman The ethical incentive is an agential response to the moral value of a duty, involving judgment, deliberation, and arousal to action-­oriented volition. There’s nothing automatic in the response to recognition that one is in a duty-­ organized space. A lot can happen between finding certain considerations salient given the value of a duty that applies and coming to act in a suitable way. Response can be routine: having promised I then perform. Or it can be freighted: having promised I come to see that my performance will cause harm. Now there’s a deliberative task: to sort the grounds of duty (their respective value) and only do what I promised if that ground prevails, and also, if need be, prepare a remedy for a duty set aside. The responsibility for getting this right is for the most part on the agent, even if she calls on others’ experience and established scripts for familiar forms of conflict. Nothing like this is present to the agent of a juridical duty. The reason her incentive can be whatever is sufficient for conformity to right is that she is not to be independently engaged with the value or ground of a requirement. That work is done by the juridical institution that generates the duty: a legal system of property or contract, the juridical form of the relationship between parents and children, and so on. The juridical duty’s value is expressed through laws that set boundaries and priorities, even penalties, either directly or through linked rules of remedy. It would be inappropriate for the agent of a juridical duty to regard herself as deliberatively independent in determining what, in her circumstances, she ought to do. It’s not her moral job, even if the value-­ tracking juridical institutions aren’t doing the job all that well. The indifference to the others’ freedom that Kant speaks of is not a mark of disengagement from the moral project. Even if I cared about the freedom of others, in acting under right I cannot make others’ freedom my object. That would be a kind of category mistake.20 The point of right—from innate right, to property, and the rest—is to introduce uniform terms for rightful relations in regions of interaction that persons (generically) cannot avoid being in and which, as I’ve argued, on their own unavoidably threaten independence and freedom. To secure rightful relations between persons in these regions, the uniformity of action-­kinds that coercive juridical institutions provide is a necessary feature. Still, we appear to lose something important. A defining aspect of the ethical incentive is that in acting out of respect for the moral law we realize a moral relation to self and other as a practical reasoner with needs and vulnerabilities of both mind and body that are of moral concern to us. Conforming to right by way of the juridical incentive looks like it couldn’t lead us to relate to each other in a distinctively moral or even juridical way: I am moved to

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Role of Juridical Obligation  253 treat you as right requires not out of respect for your status but to avoid coercive sanctions. There wouldn’t seem to be much moral value in the self-­ conception of a person with a well-­formed juridical personality if it works like that. What more can we say? UPR says that each of us acting rightfully puts others with whom we might interact under obligation. This fact of reciprocal obligation is prior to its terms: one might say that it calls for terms. The call is answered by the matter of civil society—its laws and institutions—that alter the relations between persons. In an important sense, when we are in a rightful condition, we are not strangers. Standing that we have as a matter of innate right, articulated in law, becomes the fabric of rightful relations to others within the juridical state. It’s a form of membership, pervasive in its effect. It is also morally necessary membership: we are not free not to be a member and so not free not to stand in the relationships created by right. But the moral necessity doesn’t make it a moral relation. For we are not just all obligated to act as law requires; we can only have this kind of obligation if everyone is equally constrained to conform, not just equally obligated. Some pervasive apparatus of state coercion is not merely permissible, it is necessary: it makes it true that all of our actions lie within the bounds of right. That looks more like military than moral authority. To resolve the problem, we need to pick up the thread about imputation and the conditions for action completion. If violations of right are wrongful hindrances of free action, they have no standing as free actions, even though they arise, as much as rightful actions do, from the choices of rational agents. A rightful action is one entitled to go forward. It’s like a valid move in a game. That I can physically move the rook on a diagonal doesn’t make it a chess move; and so a wrongful move has to be taken back. Coercion, thought of abstractly as a hindrance, is an element of moral causality, a taking back or canceling of a wrongful move in the domain of right. While an action in space and time can’t be taken back or canceled, as an action in the space of right, it can be. If as a matter of right you are made to give back what you wrongfully took (which may include the value lost to me while you had what is mine), it is as if there had been no wrongful action.21 rightful relations are maintained. So if right is “the sum of the conditions under which the choice of one can be united with the choice of another in accordance with a universal law of freedom,” coercion is the system of counter-­action, the condition without which the community of free action isn’t possible. That is why Kant has to reject the idea that right is a two-­element notion: an obligation in accordance with a law

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

254 Barbara Herman and an authorization to use coercion to enforce its fulfillment (MS 6:232). In the language of imputation, coercion is the condition of completion of rightful action. It makes it the case that in the domain of right, wrongful action is not possible.22 This is not to say that no one acts contrary to right. It is to say that no one violates right. A violation of right would be a kind of success, itself a real resistance to its authority. By building coercion into the very idea of right, it is not a response to wrongful choice but a formal moral barrier to its realization. Formally, the barrier—coercion and its partner, the pathological incentive—has to be present wherever wrongful action might be. In a formal sense, we must all be coerced. Now, to be prevented from doing what one wants to do by pathological means, one will be subject to pain, or the threat of pain, or face some other brute negative cost sufficient to make one want to avoid rather than gain from the forbidden performance. That sure looks like a low road, one that reinforces the sense of juridical personality as instrumental, even demeaning or subservient, sharing none of the sublimity of the moral incentive. This would be a damning conclusion if the distinctively moral work that the ethical incentive provides was absent from the juridical realm. But when we look at the whole of right, it’s clear that it’s not absent. The value at issue is reciprocal freedom of choice; the subject is the person with innate right of independence from domination by others; the rule of right that secures freedom is positive law and the institutions of the juridical state. In action done “from duty,” the value that belongs to a duty enters the agent’s maxim through the work of the moral motive. In rightful action, the value-­tracking role of motive is performed by the institutions of right, not by any rational attitude or motive of the agent. If an actual law fails to secure external freedom (say it gets the principle of adverse possession wrong, or requires consideration in contract that is too demanding), the subject of the law has no authority to act in terms of what she regards as a better law (even if she is correct in her judgment). She can, as a citizen, act to get the law changed—but that’s a very different kind of project, and one which might well call upon a value-­tracking motive, though still not in the ethical manner. Change in law has to be a change that is possible for the people of a state to decide for themselves (the shadow effect of what Kant calls “omnilateral willing”—a necessary condition of public law). There is a hint in this that juridical personality might be a common possession. But even if the coercion story doesn’t completely tarnish the moral glow of the juridical condition, the deployment of a pathological incentive drawn from what Kant calls “the aversions” may remain troubling.23 We regard the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Role of Juridical Obligation  255 aversions as “low” since they are about fear and avoidance of pain, primitive responses, self-­concerned, without moral import. From the point of view of the incentive, their relation to right looks to be instrumental and prudential: avoiding the juridical sanctions. But must we think of all the aversions this way? Perhaps there is room in the theory for “higher quality aversions.” We can get a different perspective on the aversions by looking at the early development of a healthy child. Her increasing powers and emerging sense of self are created partly by her own forays and partly by the limits and constraints introduced by her parents. If all goes well, the constraints respect the child: arising from the love and obligation of her parents, they have as their aim her developing autonomy. This happens even though the constraints operate via the pathological determining grounds of the child’s will: responses to scowls and no-­sayings, physical constraint and withholding of affection. It’s not pretty. And it’s not trivial that in the constraining the aim can simply be compliance—No peanut butter in your ear! The parent-­constrainer likely won’t see herself in the moment of constraining as having high aims. But none of this is “low” in any sense that makes it demeaning or a sign of defect or being crudely instrumental, even if it’s a bit crude in the doing. If things go well, what is done to avoid sanction comes to be done to sustain connection and then as part of what makes a person who she is—generous and able to share, less destructive with the peanut butter. The upshot is a training of the aversions (abetted by the pleasures) that help to establish the boundaries of a personality. They are a source of integrity that aligns the pathological determining grounds with love and respect.24 So likewise for the healthy person, the aversions belong not just to the history of her moral personality, they become part of its support system. The training of the aversions does not turn them into moral incentives. It couldn’t. What it can do is reconfigure their object, and in so doing make them more social. Think of our aspirations for the adolescents in our lives—that they might come to see our disapprovals as part of a system they can rely on. The charge of the state is to secure the external freedom of citizens through its institutions. The state could carry out its mission with a race of devils because the prospect of pain is for them sufficient as an incentive. But we are not devils. Just as we needn’t regard the development of a child’s personality as determined by the material or psychological whip of parental discipline, so we needn’t see the securing of juridical relations of right as just an expression of the materials of fear or the prospect of pain, as such. The aversions are not simply about the brute avoidance of pain, though pain is the telltale of them as an incentive. Aversion’s cast of characters includes such things as

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

256 Barbara Herman social shaming; loss of attachment, or connection, or trust; negative evaluation; exclusion from benefits—both material and cooperative—as well as standard sources of fear and credible threat. Alongside the desire not to be interfered with, shaped into a rightful claim not be wrongfully interfered with, arises a need to sustain cooperative relations without domination. This suggests a possible trajectory of the aversions in providing support for social or civic personality in conditions of right. So not so low after all.25

5 There’s a case to be made that such a trajectory is reflected in the ascent or unfolding of rightful social relations, starting in innate right. We’ve already noted the importance of having an enforceable right not to suffer assault on one’s life or body and the concurrent juridical determination of what is assault and battery and what is incidental contact. That this gives us juridically defined bodies, from what counts as hands and feet (but also prosthetics) to the boundaries of availability and use. We might even be thought to gain a juridical analog to proprioception: we come to feel some kinds of encroachment as an offense. While the initial specifications of innate right assert normative in­de­pend­ ence, the development of the conditions of right through property, contract, and domestic right, develop and extend our juridical relations to other persons. We could say: each stage of right transforms us. And it does so through juridical institutions and the aversions they organize. We acquire new moral powers that can bring classes of natural action and relation under obligatory laws. To be able to act freely on the material stuff of the world, to have ordinary human purposes, we need possession at a distance and over time, which calls for the exclusion of others. To render this prima facie wrongful act rightful requires an alteration of the relations between persons so that exclusion is a gain for the free action of all, an alteration individuals cannot make on their own authority, but which they can have through civil and juridical rule. In Kant’s way of speaking: we gain the power to obligate each other with respect to things. Signposts and fences can then arouse feelings that encourage us to change direction (or head into court if the claim they represent looks to us to be impossible). The coercive apparatus of the state can be in a symbol as well as a gun. To be able to form and execute performance agreements—a general good— we need to be able to obligate others into the future. But to constrain a future

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Role of Juridical Obligation  257 performance in a way consistent with another’s freedom is not a power we have by nature. The juridical regime of contract gives us that power, and so allows us to meet this need consistent with our standing as equals (which is then also a moral test for any particular regime of contract). It makes it the case that my agreement to perform doesn’t make me your servant, and yet it is not empty or dependent on my continuing interest in the enterprise. Setting aside Kant’s views about sexual relations and marriage, the central issue raised by domestic union concerns the mingling of what is ours with what is someone else’s: so sharing space and wealth, authority, and, to some extent, voice (who speaks for “us” when our interests are indivisible). The conventional members of a domestic union are adult partners and children, but also as Kant saw it, domestic servants (nannies and live-­in housekeepers still fit the function if no longer the label). Because of the intermingling of lives, if not bodies, there is, for Kant, something not manageable through contract (for delimited goods and services), and so a distinct question about securing the equal status of all members while in the union (which includes determining what counts as dissolving the union, as distinct from fleeing or abandoning it). It’s not that one can’t find moral terms to describe various kinds of domestic abuse without appeal to juridical status. The juridical is about creating terms of standing—of role, responsibility, complaint, and redress. The point is not to add a further dimension of moral scorekeeping, but to have juridically defined and protected roles that give individuals determinate rights and powers in a region of human living that in its natural forms threatens innate right. These are called domestic rights. It allows saying: absconding with a domestic union’s wealth is not just wrong, it is to be juridically undone. Abandoning a domestic union’s dependents may harm them in many ways that are morally objectionable, but only juridically do they have determinate avenues of redress. Though Kant was surely too focused on matters of paternal authority and obedience, it’s not implausible to see in the aversion-­ backed sensitivities for the juridical boundaries of (relatively) enduring domestic relations a means of sustaining a safe space for persons of varying natural powers and shifting roles to be and be seen as equals. Or not, if the facts betray their rationale.26 It’s of more than passing interest that Kant did not view the family as private. This view of the trajectory of right makes one of the hard moments of the Doctrine of Right a bit clearer. Given the need for an increase in our moral powers, creating the conditions for rightful relations is necessary moral work for what Kant calls an “omnilateral willing” that can move us from a state of nature to civil society that has the form of a state (a political order

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

258 Barbara Herman under the rule of law). That is, it is a moral duty of each, and of all together, to make it that all who must associate with one another are under enforceable reciprocal obligation with respect to relations it would be contrary to reason to avoid and which, without our together being in juridical relations, would be inconsistent with innate right. In short: the moral existence conditions for human persons in a finite shared material word are that they together make themselves into citizens. The juridical state is then a moral space in which, through a particular regime of right, we know a host of things about each other, about our powers and our vulnerabilities. I know that along the rails of right that connect us you can send things, sell things, loan them, jointly use them; we can make things separately and together; form lasting complex relationships; rightfully bring dependent beings into existence. None of this is because of who we are (in a personal sense), or because we are good. We can do this because the positive laws that give specific expression to right make such activities possible for equal and independent persons. Although these are powers we each have only on the contingent condition that others have them as well, they are rational powers that alter what it means to be a person under obligatory law. As persons with juridical personality we can enter into a variety of relationships that involve uncertainty and risk without having to investigate the other’s deep psychology to get security about his or her attitude toward rightful performance. The juridical obligations that express juridical personality are sufficient. Formally, at least, we have security through enforceable rights: actions that are contrary to right can be hindered, through either lawful prevention or a kind of juridical annulment (of repair or punishment). As noted, Kantian juridical duties hold for citizens even if legal institutions do their work badly. Making the actual institutions that do this formal work less messy and more just is a permanent moral, not a juridical, task. As we are vulnerable to stepmotherly nature for the success of our best intentions, we are vulnerable to the quality of our institutions for our independence.27 A subprime mortgage industry that targets economically vulnerable citizens could act with impunity when a sub-­domain of right was defectively instantiated in law. It could be an open question whether regulation or liability is the appropriate corrective. What is not open is whether the moral cause was insufficiently virtuous bankers. It wasn’t. The failure was collective (however it is represented) in not securing institutional arrangements that protected the independence of vulnerable borrowers. A kind of state-­sanctioned piracy. A different kind of response is called for when action is contrary to law (and so contrary to right). We stand in juridical relation to one another not

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Role of Juridical Obligation  259 as voluntary members of a group but as juridical persons. A principle of external action that would self-­exempt the agent from the law is juridically impossible.28 Since the action is empirically possible, the assertion of the primacy of juridical authority is in hindering the completion of the deed (either directly or by cutting off the fruits of wrongful action through punishment). Such post factum hindering is the imputed rightful effect of wrongdoing, a necessary collective resistance to an assault on our common juridical authority.

6 With this we have assembled, albeit in rough form, the pieces of an account of juridical obligation and juridical personality as essential elements of Kant’s account of moral agency and the system of duties. The starting point is the principle of practical reason—the moral law—that governs both our volitions and our external actions. It gives rise to principles of obligation based in reason whose authority does not depend on our personal ends or purposes. On the ethical side, reason takes on the project of the human will, subject to both rational and sensible incentives, requiring of us that we conform our willings to its principle. On the juridical side, reason’s target is a plurality of embodied persons interacting in a shared physical environment. It requires that we together create the conditions for external rational or rightful relations. On the ethical side our rational nature gives us obligatory ends—things we must attend to in order to maintain and develop our moral powers. On the juridical side, there are things that it is rational for us to do—reason rejects our not being able to do them—that we cannot do on our own or through voluntary cooperation or good willing consistent with innate right. There is therefore a duty that we have together to alter the terms of our interactions to make it possible for us to relate to each other in freedom- or independence-­preserving ways. It is a moral story through and through. To speak of juridical personality is to speak of a kind of power we have as rational beings in relations with other persons in a shared world. Juridical obligations arise from the laws of a state whose rationale looks to innate right and external freedom. The worry about juridical obligation as a kind of moral obligation stemmed from its reliance on a pathological incentive, the partner of coercion. I’ve argued that if the reason to prize the ethical incentive or motive of duty is that it keeps the agent oriented to what’s of value in required action, in the juridical context that work has to be done by institutions of law, not individual willings. Juridical

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

260 Barbara Herman structures are deliberation-­sensitive, juridical obligations are not. In ethical life, respect transforms how we regard self and other; a parallel transformation comes in juridical life through the rightful conditions for our interactions. Juridical personality thus adds to the moral conception of the person, but with a piece possible only in community with others.29

Notes 1. I am pleased to offer this paper as a much delayed reflection on some themes in Christine Korsgaard’s bold 1997 essay, “Taking the Law into Our Own Hands: Kant on the Right to Revolution.” It is testimony to the power of Chris’s ideas that they always direct one’s attention to first principles. My aim here is not to debate the ethics of the revolutionary, but to give an account of the stakes on the other side of the issue: the extent to which the ethical life we prize depends on our taking care of the institutions of our public life. I expect I share with Chris the periodic surprise about how much (still) lies waiting in Kant’s thought. 2. Arthur Ripstein’s Force and Freedom: Kant’s Legal and Political Philosophy (Cambridge, MA: Harvard University Press, 2009) elegantly makes this case. For a different take on the principle, see Paul Guyer, “Kant’s Deduction of the Principles of Right,” in Mark Timmons, ed., Kant’s Metaphysics of Morals (Oxford: Oxford University Press, 2002), 23–64. 3. I give an extended argument for this conclusion in “The Difference that Ends Make,” in Julian Wuerth, ed., Perfecting Virtue: Kantian Ethics and Virtue Ethics (Cambridge: Cambridge University Press, 2011), 92–115. 4. That is, the possibility of morality depends on there being a source of action essentially connected to our autonomy as rational beings; the matter or content of our duties and obligations is about us as human beings. There is a further question of how a dependent rational being like us—whose will is subject to inclination—can take an interest in the moral law in an autonomy-­preserving way. Kant takes this up in the second half of Groundwork III and in Part Three of the Critique of Practical Reason. 5. To help with tracking terms, I use right for Kant’s recht when it describes the system (also rightful for rechtliche), but the ordinary “right” for e.g., the right to property. I’ll also use ethics for tugend; the usual translation of tugend as “virtue” takes the contemporary ear in an overly Aristotelian direction. 6. References to Kant’s work are to volume and page number of the Prussian Academy Edition of Kant’s Complete Works. Abbreviations: MS for the Metaphysics of Morals; G for the Groundwork of the Metaphysics of Morals. 7. This is an appropriate place to acknowledge Japa Pallikkathayil’s insightful “Deriving Morality from Politics: Rethinking the Formula of Humanity,” Ethics 121 (2010): 116–47. Pallikkathayil sees how right might give (some) content to the Groundwork idea of humanity and so of treating persons merely as a means. The argument I’m making is more radical: it’s not just external actions that depend on right for their content; our ethical duties arise in and are shaped by juridical duties and institutions.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Role of Juridical Obligation  261 The formula of universal law would be deliberatively ineffectual on its own, either begging the content of duties question, or caught in a web of socially generated false negatives (e.g., about hierarchy or competition or telling tall tales). 8. So if one group or class of persons controls the means of production, making others dependent in their productive lives on imposed arrangements, that makes those who are dependent unfree in a way that violates the conditions of a justified property system, whether or not they are benefitted by the arrangement. 9. I argue for this in detail in The Moral Habitat (Oxford: Oxford University Press, 2021). 10. This is the taking-­off point for Korsgaard’s “Taking the Law into Our own Hands: Kant and the Right to Revolution,” in Andrews Reath, Barbara Herman, Christine M. Korsgaard, eds., Reclaiming the History of Ethics: Essays for John Rawls (New York: Cambridge University Press, 1997), 297–328. 11. That we can “act from duty” is essential evidence that our practical rationality is not exhausted by instrumental principles. But it doesn’t follow that whenever there is a requirement whose authority comes from the moral law that what it requires is acting from respect for it as an incentive. 12. For those agitated by Kant’s Formula of Humanity remark about animals as things, this definition matters. The remark is not a claim that there is no morally significant difference between animals and rocks, it’s that the actions of animals are not under obligatory law and so not imputable to them. If, looking ahead, imputation is the anchor of Kant’s version of moral responsibility, this is not a controversial claim about animals. 13. Unlike a volition, a deed is identified by the law that applies. So the same deed can issue from different maxims. But then also, acting from a maxim of duty does not guarantee that the correct deed was done. 14. I think this is why the discussion of conflicts of duty follows the preliminary articulation of deed and imputation. Since a deed cannot have two opposing moral valences, only one of the conflicting grounds of obligation can actually obligate. 15. Suppose “doing more in the way of duty” makes things worse. Had I not made what turned out to be a futile extra effort, someone else would have stepped in and done better. Neither perfect nor imperfect duties tell us exactly what to do on their behalf: there is always some contingency in how we act “in the way of duty.” Failure from “doing more in the way of duty,” so long as the way of acting is sensible, may be just another instance of the world limiting our successes. 16. Even a subject’s “state of mind” can make a difference (“whether he committed the deed in a state of agitation or with cool deliberation” (MS 6:228)). 17. It is in this sense, and at this point in the Doctrine of Right, that we have the explanation for the claim that wrongdoers will or are the authors of their own punishment. In a formal sense. 18. Including suspended or minimal sentences. It’s not surprising, given the origin story for punishment in coercion’s hindering a hindrance, that Kant’s view about proper punishment is talionic. Though it’s worth noting that apart from his views about murder, his application of the talionic idea tends to be more expressive than literal. So something that strikes down the pride of a high-­born thug, or work without access to property (wages) for a thief (MS 6:332–3).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

262 Barbara Herman 19. And why the disjunction in the UPR? A maxim can fail to be a possible law of free choice without reference to its deliberative justificatory structure. A question about right can thus be asked about action understood under juridical principles for property or contract or status—as the doing of a parent or a banker or a policeman. 20. But if that’s so, how could the ethical incentive ever even play a back-­up role? Suppose we are miners in the Gold Rush, aiming to own and mine potentially profitable land. The law of California asserts some kind of provisional right to a staked claim until it’s filed, but there is no effective enforcement. While you travel down the mountain to the claims office, what incentive do I have not to take ore from your holding if I safely can? The ethical incentive can’t latch on to the provisional juridical duty—it’s a relatively arbitrary bit of regulation (or worse, a rule made to favor the banks and large landholders). The nearest possible object of the ethical incentive is something like lawfulness as such. Since claim-­jumping challenges the authority of law to make secure possession of land possible, we can have an ethical duty not to undermine lawfulness serving as a back-­up for an unenforced juridical duty, but only indirectly. (Kant’s own example at MS 6:220 is of a contract that cannot be enforced.) 21. This condition is not realized if you just voluntarily give back what you wrongfully took. 22. That’s why, for Kant, punishment has to flow out of the juridical meaning of the wrongful action. 23. The determining grounds of coerced choice are drawn from the aversions, because, as Kant says, “it is a lawgiving, which constrains, not an allurement, which invites” (MS 6:219). Should we then, on Kantian grounds, reject the idea that a nudge is as good as a shove? 24. In the terms of Kant’s moral psychology, the trained aversions function like the analog moral emotions of sympathy and honor (see the remarks at G 4:398): amiable, attractive, even reliable, but unable to confer moral worth on the actions they support. One can “take delight” in something for its own sake without valuing it as good in itself. 25. Of course, things don’t always go so well. The child of a parent who consistently belittles him or who offers approval only on proof of his domination over others is not going to find joy in a community of equals. His future law-­abiding behavior will likely be mired in the low road, and is to be regretted. 26. Friedrich Engels’s work on the centrality of the family in the history of oppression is a historical-­materialist descendent of this argument. 27. There might even be room here for civil disobedience as a strategy of juridical re­sist­ ance. Kant does argue that it is imperative that the state support publicity, transparency, and access so that its citizens can engage together in the task of bringing the state closer to the condition of right. (Cf., various remarks in MS 6:313–28, and especially 6:329.) 28. Ethical error is inappropriately making an exception for yourself; juridical error is about exempting oneself from the authority of law. 29. But weren’t we already in community with others as co-­legislating members of a kingdom of ends? This would be to treat a formal notion as a material one. The kingdom of ends is a concept—of a systematic union under law of ends willing ends—that we are led to from the idea of each as legislating universal law (G 4:433). It is an ideal. At

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Role of Juridical Obligation  263 most that tells us that the concept is consistent with the moral law. And, Kant thinks, it is a “fruitful” way to regard ourselves. However, the possibility of a moral community, and the consistency with the concept of a rational being as a member (and not its head), is not for all that being in community with others.

Bibliography Guyer, Paul. “Kant’s Deduction of the Principles of Right.” In Mark Timmons, ed., Kant’s Metaphysics of Morals, 23–64. Oxford: Oxford University Press, 2002. Herman, Barbara. “The Difference that Ends Make.” In Julian Wuerth, ed., Perfecting Virtue: Kantian Ethics and Virtue Ethics, 92–115. Cambridge: Cambridge University Press, 2011. [Reprinted in Barbara Herman, Kantian Commitments. Oxford: Oxford University Press, 2022.] Herman, Barbara. The Moral Habitat. Oxford: Oxford University Press, 2021. Kant, Immanuel. The Metaphysics of Morals. In Practical Philosophy, translated and edited by Mary J. Gregor. Cambridge: Cambridge University Press, 1996. [References are to page numbers of the Prussian Academy edition.] Kant, Immanuel. Groundwork of the Metaphysics of Morals. In Practical Philosophy, translated and edited by Mary  J.  Gregor. Cambridge: Cambridge University Press, 1996. [References are to page numbers of the Prussian Academy edition.] Korsgaard, Christine  M. “Taking the Law into Our Own Hands: Kant on the Right to Revolution.” In Andrews Reath, Barbara Herman, and Christine M. Korsgaard, eds., Reclaiming the History of Ethics: Essays for John Rawls, 297–328. New York: Cambridge University Press, 1997. [Reprinted in Christine M. Korsgaard, The Constitution of Agency: Essays on Practical Reason and Moral Psychology. Oxford: Oxford University Press, 2008.] Pallikkathayil, Japa. “Deriving Morality from Politics: Rethinking the Formula of Humanity.” Ethics 121 (2010): 116–47. Ripstein, Arthur. Force and Freedom: Kant’s Legal and Political Philosophy. Cambridge, MA: Harvard University Press, 2009.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

12 The Social Conditions for Autonomy Kant on Politics and Religion Faviola Rivera-­Castro

In her highly influential Self-­ Constitution: Agency, Identity and Integrity, Christine Korsgaard explains the essential features of “becoming a unified agent,” of being a person.1 Though constituting oneself into a particular person is something that all human beings do under any conditions, she maintains that there are better and worse ways of carrying out this unavoidable task. Following Kant, Korsgaard maintains that the categorical imperative is the highest standard of self-­constitution. Thus, on her account, being internally well-­constituted is being autonomous or, what is the same for Kant and Korsgaard, being morally good. People who are immoral or evil have constituted themselves badly. However, Korsgaard, unlike Kant, has never explicitly discussed the social conditions that would be most adequate for successful self-­constitution. In this paper, I argue that despite her Kantian commitments Korsgaard appears to suggest that the social conditions for autonomy are importantly different from those that Kant affirms.2 While Korsgaard seems to think that a liberal society would be the best social setting for constituting oneself successfully, Kant explicitly argues for a specifically religious kind of ­ socialization into the morality of autonomy. By a “liberal society” I mean one in which there is a plurality of conceptions of the individual good and individuals are free to affirm the conception of their choice. It is a society that secures the protection of basic individual rights and liberties and also welcomes the pluralism of conceptions of the good that results from the exercise of human reflective capacities under conditions of freedom. A diversity of religious conceptions has usually been part of a pluralistic ­liberal society. While Kant affirms that citizens should be granted basic rights and freedoms, he also maintains that there are moral and political reasons to promote Faviola Rivera-­Castro, The Social Conditions for Autonomy: Kant on Politics and Religion In: Normativity and Agency: Themes from the Philosophy of Christine M. Korsgaard. Edited by: Tamar Schapiro, Kyla Ebels-­Duggan, and Sharon Street, Oxford University Press. © Faviola Rivera-­Castro 2022. DOI: 10.1093/oso/9780198843726.003.0012

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Social Conditions for Autonomy  265 citizens’ endorsement of the true religion of reason and its true church. On his account, the stability of both individual virtue and political society calls for membership in an ethical community that takes the form of a church. Kant’s arguments turn on the role that the right kind of religious sociability plays for instilling and reinforcing citizens’ respect for law in both morality and politics. By contrast, the wrong kind of religious sociability—which, on his view, exists under the condition of “sectarianism”—undermines the social acceptance of both true morality and the rule of law. I argue that this explains why Kant, unlike contemporary liberal political philosophers, does not recommend reconciliation to the proliferation of religious conceptions and ecclesiastical faiths that are consistent with justice. Instead, he favors collective enlightenment on religious matters, which, on his view, importantly entails the criticism of existing religious practices. Religious enlightenment, on his view, leads to the recognition of moral religion as the true religion of reason and to the public affirmation of the church. Most of the paper explores the question why Kant maintains that a specifically religious kind of sociability is necessary for both individual autonomy and the rule of law. At various points I indicate where Kant departs from contemporary liberal philosophy and its vision of a liberal society. In Section 1, I briefly explain what I take to be Korsgaard’s position regarding the kind of society in which individuals could best constitute themselves successfully. Since she has never explicitly addressed this issue, my discussion is somewhat speculative. I suggest that Korsgaard’s account of self-­constitution calls for a liberal society as the best social setting in which individuals could constitute themselves successfully. In Section 2, I present Kant’s argument for an ethical community as a necessary condition for the stability of virtue. Here I explain why, on his account, a church based on the moral religion of reason provides the right kind of socialization for the social acceptance of the morality of autonomy. In Section 3, I examine his arguments against “sectarianism” and argue that what he calls the “ethical state of nature” can plausibly be interpreted to refer to this condition. Here I suggest that the ethical community is Kant’s response to the public undermining of true morality in the condition of sectarianism. Finally, Section 4 develops the claim that in a Kantian state the ethical community plays the role of securing the public ac­cept­ ance of the supremacy of law. In sum, my central claim is that, on Kant’s account, a church based on the moral religion of reason plays the dual public role of socializing individuals into the morality of autonomy and the rule of law.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

266 Faviola Rivera-Castro

1.  Korsgaard’s Liberalism Korsgaard’s account of self-­constitution appears to call for a liberal society as the best social context in which individuals could constitute themselves successfully. To constitute oneself, she tells us, “involves finding some roles and fulfilling them with integrity and dedication,” as well as “integrating those roles into a single identity, into a coherent life” (Self-­Constitution 20, 25). This is the task of “carving out a personal identity,” of “making something of ourselves,” of “creating ourselves” (24, xiii, 120–30). Though constituting oneself is something that all humans do in all societies under all sorts of circumstances, she argues that there are better and worse ways of carrying out this unavoidable task. People constitute themselves badly when they fail to integrate themselves into a single coherent identity. We do this when we fail to live up to the principles that are internal to the social roles with which we identify, but also when we fail to live up to the principles that are internal to the work of self-­ constitution and necessarily govern it. Following Kant, Korsgaard maintains that the categorical imperative is the highest standard of self-­constitution. Thus, on her account, to be internally well-­constituted is being autonomous or, what is the same for Kant, being morally good. People who are immoral or evil have constituted themselves badly. Though Korsgaard does not elaborate on the social conditions that are most favorable for the task of self-­constitution, it seems reasonable to suppose that on her view such conditions must allow for the revision and possible rejection of practical identities. A practical identity is a role to which we commit ourselves, such as being a parent, a citizen, a wife, a daughter. Most of the practical identities we affirm are “contingent” in the sense that “we can walk away from them” (23). It is “up to us,” she tells us, whether we “treat them as [] source[s] of reasons and obligations” (23). This does not mean, however, that we cannot treat some of our practical identities as ­necessary, as “the sources of absolute inviolable laws” (23). We do this all the time in constituting ourselves as the particular persons that we are. The point I wish to highlight is that if we are to constitute ourselves well and thus become autonomous, it would be best if social conditions do not obstruct the possibility of walking away from those practical identities that we think we should not endorse. Otherwise, it may become hard or even impossible to achieve integrity, to lead a coherent life. Societies that mandate individuals to practice a particular religion or prohibit women to get an education and decide for ­themselves who they wish to become obstruct the possibility of successful self-­constitution.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Social Conditions for Autonomy  267 The task of constituting oneself, as Korsgaard describes it, can best be carried out under social conditions that grant individuals the freedom to choose their own practical identities. Such conditions include the protection of basic liberties such as freedom of conscience, of expression, and of association. These social conditions also allow and welcome the creation of new forms of practical identities that are not immoral or unjust. Arguably, such a society would also seek to eradicate forms of domination that work through the social imposition of practical identities on individuals, such as racial and gender stereotyping. In sum, it seems reasonable to suppose that Korsgaard’s account of self-­constitution calls for a liberal society as the best social setting in which individuals could constitute themselves successfully. Such a society protects basic individual rights and liberties and also welcomes the pluralism of ways of life that results from the exercise of individual reflective capacities under conditions of freedom. One the few explicit remarks that link her moral theory to liberalism is her explanation of the sense in which human beings have a life in terms of the notion of a “conception of the good.” She writes: “a human being has, and is capable of choosing, what we sometimes call a ‘way of life’ or, following Rawls, a ‘conception of the good’” (128). According to Rawls, the latter is a conception of “what is of value in human life” that includes “a more or less determinate scheme of final ends” as well as “attachments to other persons and loyalties to various groups and associations.”3 The familiar liberal thought is that individuals have the capacity to revise their conceptions of the good and to abandon them in favor of other alternatives. This is the capacity for self-­ determination, which, on a liberal view, must be institutionally protected. Though liberals agree on this important point, they disagree as to whether the liberal state should also promote an ideal of individual autonomy understood as fully governing oneself, in all aspects of one’s life, on the basis of values and principles freely affirmed upon reflection.4 It is unclear where Korsgaard stands regarding this disagreement, since her account of self-­constitution is compatible with either position.5 The point I wish to stress is that, on a liberal view, the institutional protection of the capacity for self-­determination produces, in the long run, a growing pluralism of conceptions of the good. The notion of pluralism is built into the notion of a liberal society and is difficult to exaggerate its importance within liberal theorizing. Liberal authors take pluralism to be, in Rawls’s well-­known formulation, the long-­term result of the exercise of human reason under free institutions.6 A liberal society is committed to welcoming the pluralism of ways of life insofar as the latter is  taken to be the ongoing result of liberal institutions themselves.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

268 Faviola Rivera-Castro Pluralism, which for historical reasons has importantly included a diversity of religious conceptions, is an integral part of a liberal society as the expression of the protection of individual liberties.7 As we will see in the following sections, Kant held a notoriously different view of the relation between moral autonomy and pluralism in matters of faith.8 On his view, the social acceptance of moral autonomy calls for a particular kind of religious unanimity. Though he agrees that the proliferation of ecclesiastical faiths is a sign of the official protection of freedom of conscience, he does not draw the normative implication according to which religious diversity is expressive of the exercise of reflective capacities under conditions of freedom. By contrast with Korsgaard, he does provide an account of the social conditions that are necessary for the successful struggle after individual autonomy. The required kind of sociability, on his view, is specifically religious and involves the endorsement of a particular conception of religion as well as membership in a particular type of church. As we will see, Kant considers sectarianism—the proliferation of ecclesiastical faiths—to be morally objectionable and politically undesirable.

2.  Kant’s Argument for an Ethical Community Kant holds the extraordinary view (to contemporary liberal sensibilities at least) that we all have a duty to join a church. The ground for this duty is the moral demand to do everything in our power to establish an “ethical community” on earth, the visible form of which, he maintains, is a church (R 115, 6:105).9 By an “ethical community” he means “an association of human beings merely under the laws of virtue” (R 106, 6:94). A church, in turn, is “[a]n ethical community under divine moral legislation” (R 111, 6:101).10 Kant’s argument for the need of an earthly ethical community turns on his claim that the individual commitment to virtue can never be stable as long as social vice is pervasive in social relations. In a society ridden with vice, virtuous individuals are in “incessant danger” of relapsing into evil (R 106, 6:94). In light of this, the ethical community “has for its end the prevention of this evil and the promotion of the good in the human being” (R 104, 6:94). Though the demand to strive after moral goodness is addressed to each person qua ­individual, the struggle against evil requires a concerted collective effort. The struggle to become morally good is as much a collective as an individual one. By “evil” Kant means the natural propensity, rooted in reason, to subordinate moral requirements to the claims of self-­love, which procures whatever

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Social Conditions for Autonomy  269 is agreeable or convenient and may be contrary to morality. In Religion he maintains that temptations to evil do not originate in natural inclinations, which are not reprehensible in themselves (R 78, 6:58). Temptations to evil, on his view, have their source in the vices that plague social interaction. Echoing Rousseau, Kant maintains that, as soon as they are together, human beings “mutually corrupt each other’s moral disposition and make one another evil” (R 105, 6:94). His point is not that evil human beings may drag others down along with them, though this may also be the case, but that social interaction by itself harbors social vices. He mentions envy, ingratitude, joy in other’s misfortunes, addiction to power, and avarice (R 51, 6:27; R 105, 6:93–4). These social vices originate in the natural tendency to compare oneself to others, which, in turn, gives rise to an inclination “to gain worth in the opinion of others.” This inclination is good as long as the individual seeks to be regarded as an equal (“not allowing anyone superiority over oneself ”) but may also give rise to the “unjust desire to acquire superiority for oneself over others” and to the consequent attitudes of jealousy and rivalry that pave the way to the “vices of culture” (R 51, 6:27). “Virtue,” as he characterizes it, is “the moral strength of man’s will in fulfilling his duty.”11 Such a personal firm determination, however, is not enough for the “triumph of goodness” since, as already mentioned, virtuous individuals are in danger of falling back into evil in a society ridden with social vice. So conceived, the triumph of goodness is not within the reach of each virtuous individual alone but depends on the collective effort to relate to one another through laws of virtue. On this picture, morality calls for a profound transformation of social relations that, Kant maintains, would result if all human beings were to join an ethical community in which the laws of virtue have become public laws.12 Kant presents this social change as a transition from an “ethical state of nature” to an ethical community (R 108, 6:97). Drawing an explicit contrast with the juridical state of nature, he characterizes the ethical state of nature as one in which each individual prescribes the law to herself—there is no common judge, and no effective public authority. To readers of his moral philosophy, it may come as a surprise that Kant conceives of the ethical community through such a close analogy with political society. What he now considers an “ethical state of nature” captures exactly what the morality of autonomy maintains: that autonomous individuals prescribe laws to themselves while there is no common judge and no public authority capable of applying and executing such laws. In order to apply the laws of virtue, the relevant judge would require the capacity to determine whether individuals have indeed adopted

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

270 Faviola Rivera-Castro such laws as their maxims of action. As Kant repeatedly maintains, however, we lack such a capacity both in our own case and in judging other people’s conduct. In order to execute the laws of virtue, on the other hand, a public authority must have the capacity to give each individual what she deserves according to her virtue. As is well known, Kant claims that individuals deserve happiness in exact proportion to their virtue. Nevertheless, if it is to be an association in which human beings see themselves as being collectively under the laws of virtue, the ethical community must have a superior authority: a common legislator, judge, and public authority capable of applying the laws of virtue (R 109–10, 6:98–100). The problem, however, is that no human public authority can be up to this task. A human public authority could not possibly legislate, apply, and enforce the laws of virtue. The latter are “internal” in the sense that they govern the choice of maxims. A public human authority can only be “external” insofar as it can only legislate, apply, and enforce laws for the legality of actions (i.e., their outward conformity to law). Since the supreme common legislator of laws of virtue cannot be human, Kant’s solution is to conceive of the ethical community as “a people of God” (R 109, 6:98).13 In his capacity as a supreme legislator, God is conceived as a being “with respect to whom all true duties, hence also the ethical, must be represented as at the same time his commands” (R 110, 6:99). Consequently, God must also be represented as capable of discerning our innermost motives and of distributing happiness in exact proportion to virtue. To this end, he is also conceived as the author of the laws of nature. Thus, the possibility of the ethical community, which is necessary for the triumph of goodness, leads to religion. The latter, on Kant’s account, is the recognition of all our duties as divine commands (R 153, 6:153–4).14 As soon as it must be conceived of as a people of God, the ethical community takes the form of a church. Kant makes clear that the impossibility of establishing an ethical community through human forces does not mean that all the work should be left to God. On the contrary, he maintains that each must “so conduct himself as if everything depended on him” (R 111, 6:101). On his view, individuals in an ethical state of nature fulfill their duty to do everything in their power to enter into an ethical community by joining an existing church that is susceptible to transformation according to the idea of “an ethical community under divine moral legislation” (R 111, 6:101). Here there is a clear parallel with the duty to leave the juridical state of nature, which, likewise, is not to be discharged by founding a new state but by submitting to the authority of an existing one (usually the state one is already living under). There are, however, three important asymmetries.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Social Conditions for Autonomy  271 First, membership in the ethical community is entirely voluntary. Nobody can be coerced to join a church, though individuals are expected, from an ethical point of view, to do so. Second, the ethical community aims to include all human beings since “duties of virtue concern the entire human race” (R 107, 6:96). Kant envisions the process of building an ethical community as a gradual progress toward religious unanimity that extends across political boundaries. Third, whereas individuals usually have no choice but to submit to the authority of the state they live under, it is possible to choose among various churches provided that the state protects freedom of belief. However, Kant maintains that not all existing churches are equally good. Individuals remain in an ethical state of nature before they join a church that bears the characteristics that make it a “true” church.15 As the visible form of an ethical community that seeks to include the entire human race, a true church aspires to universality and has as its primary aim the moral improvement of its members. As already mentioned, the pure religion of reason says that all God requires from his subjects is to lead a morally good life. According to this, churches that do not aspire to universality or that reject the pure religion of reason do not qualify as the visible form of an ethical community. Such churches include those that condition membership to some characteristic that not all human beings share (such as race, nationality, or gender) or that make salvation conditional on anything other than moral improvement (such as fulfilling certain outward acts and rituals). On Kant’s view, the transition to the ethical community is underway with the appearance of a historical faith that affirms the pure religion of reason. As is well known, he takes Christianity to be such a faith on the grounds that “of all the public religions so far known” only this faith is a religion of good life conduct (R 71, 6:51–2). Ideally, the true church should be founded on pure rational faith alone since this is the only faith that can be “convincingly communicated to everyone.” At the same time, however, Kant acknowledges that a true church stands in need of a historical (i.e., ecclesiastical) faith that contains statutory laws accessible through revelation despite the fact that a historical faith, based on facts, “is incapable of a transmission that commands conviction universally” (R 118, 6:109). He grants that “no doctrine exclusively based on reason would seem to the people to make an unalterable norm” and that they demand a “historical authentication” of the authority of pure rational faith in the form of a divine revelation (R 120, 6:112). Historical faith thus serves as a “vehicle” for pure rational faith.16 The former is a mere instrument that may be gradually discarded when it turns into a “fetter” as the process of enlightenment advances and the people no longer need the support of a historical faith to

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

272 Faviola Rivera-Castro affirm the true religion of reason (R 127, 6:121). He regards the historical component of ecclesiastical faith as non-­essential, dispensable, and even perilous insofar as it can be taken, as often happens, as being essential and unconditional. On his account, ecclesiastical authorities would become gradually dispensable as enlightenment progresses and “equality springs from true freedom” (R 127, 6:122). The distinction between the true and universal moral religion grounded on reason, on the one hand, and the historical component of ecclesiastical faith, on the other, is central to Kant’s account of sectarianism. While only the ­former is essential and unconditional, the latter is a mere instrument or vehicle for making moral religion accessible. As we will see in the following section, sectarianism emerges, on his account, when moral religion is subordinated to the historical element, which, in turn, is taken to be unconditional.

3.  Sectarianism as an Ethical State of Nature In this section I discuss, from a different perspective, the question why Kant thinks that the adequate kind of sociability for fostering individual autonomy is specifically religious.17 In the previous section I considered his argument for the need of a divine legislator if the laws of virtue are to be public laws. In this section I address the prior question why, on his view, the laws of virtue must become public in the first place. It is reasonable to ask why the laws of virtue might not remain the laws that each individual gives to herself, as proponents of autonomy usually maintain. Kant’s response, as we saw in the previous section, appeals to the morally corrupting nature of social relations. In this section I suggest that the ethical community can also be thought of as his solution to a far more specific social problem: the morally corrupting nature of most existing religious practices. On his view, I argue, such practices pose the greatest obstacle to the social acceptance of the morality of autonomy. Churches, by making claims to unconditional obedience to God, come into direct conflict with the morality of autonomy as soon as the former reject practical reason’s claim to be the only source of unconditional (i.e., moral) demands. This conflict undermines the social acceptance of the morality of autonomy given the powerful influence that churches often exercise on individual consciences. The ethical community plays the public function of correcting this pernicious social tendency by providing an alternative way of socialization in which true unconditional demands become public laws.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Social Conditions for Autonomy  273 The ethical community, as a church in which God’s unconditional demands are those of true morality, is Kant’s response to this social challenge. Kant’s criticism of ecclesiastical practices is primarily contained in his objections to “sectarianism.” His approach to the latter presupposes his view that moral religion is the true religion of reason which, as we have seen, is grounded in universal morality. The true religion of reason holds that the only unconditional demand religion makes on us is moral improvement. From this point of view, the historical and statutory elements of ecclesiastical faith are instrumental for moral religion. Such elements are secondary and even dispensable. With this view in place, sectarianism turns out to be morally objectionable insofar as “it arises from a mistake on the part of ecclesiastical faith: the mistake of regarding its statutes (even if they are divine revelations) for essential parts of religion” (CF 273, 7:50). Kant refers to this mistake as “dogma” which, he says, “is a prolific source of innumerable sects in matters of faith” (CF 273, 7:50). He also calls this mistake “counterfeit” serv­ice in which “revealed faith is to come ahead of religion” and “what is mere means is unconditionally commanded” (R 162, 6:165). According to his diagnosis, disagreements in matters of faith concern those elements of an ecclesiastical faith that are not moral but historical and statutory. He remarks that “the so-­ called religious struggles, which have so often shaken the world and spattered it with blood, have never been anything but squabbles over ecclesiastical faiths” (R 117, 6:108). Sectarianism is thus morally objectionable insofar as it is motivated by the mistake of subordinating the essential element of religion (moral improvement) to the historical and statutory components, which are in turn taken to be unconditional. Once this inversion takes place, ecclesiastical faiths that differ in their historical and statutory components present themselves as offering alternative paths to making oneself agreeable to God with the consequent conflicts and violent confrontation that such claims might provoke. Kant grants that the existence of a diversity of historical faiths is desirable insofar as it indicates that there exists freedom of belief,18 but he remarks that “in itself, such a public state of affairs in religion is not a good thing” unless all ecclesiastical faiths can agree on the essentials of the true religion of reason (CF 275, 7:52). Kant’s position on the problems posed by sectarianism, as well as on the appropriate solution to them, radically differs from the views that have figured most prominently since the outbreak of the European wars of religion. As is well known, the main problems have usually been taken to lie in the conflicts and violence that threaten both individual integrity and social stability. This is clear in Hobbes’s description of the state of nature and in Locke’s views on

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

274 Faviola Rivera-Castro toleration. The usual solutions, accordingly, have been directed at making peace possible. On Kant’s account, by contrast, the fundamental problem posed by the proliferation of ecclesiastical faiths is not the fact of violence, actual or possible, among them. From his point of view, even if sectarianism did not lead to violence, the problem would remain that sectarianism is morally objectionable. His position on this point parallels his view on the problem posed by the juridical state of nature. While that problem has ­traditionally been taken to lie in the conflicts and violence that the absence of a recognized common authority brings with it, Kant takes the fundamental problem to lie in the absence of a rightful condition, which makes the juridical state of nature morally objectionable (DR 124, 6:312).19 On his account, even if the latter were not a state of war, it would still be morally wrong to remain in it. Likewise, in his approach to sectarianism, the fundamental problem does not lie in the conflicts and violence that such a state of affairs brings with it. Even if sectarian violence were to cease, the proliferation of ecclesiastical faiths that subordinate moral demands to ecclesiastical but morally indifferent requirements would still be morally objectionable. In both cases, the juridical and the religious one, conflict and violence turn out to be derivative of an underlying moral deficiency. Violent resolutions of conflicts of rights remain a possibility in the absence of a rightful condition just as sectarian conflict and violence may break out as long as people make the moral mistake of taking the non-­essential parts of religion as unconditional. Kant’s solution in both cases is addressed at correcting the underlying moral problem. In the case of sectarianism, his solution is collective religious enlightenment, which, he maintains, makes people realize that the only unconditional demand religion makes on us is moral improvement and that the historical and statutory components of religion are secondary and even dispensable. According to this solution, sectarianism is to be gradually replaced with the universal expansion of the moral religion of reason and its true church, which alone recognizes moral improvement as religion’s only unconditional demand. By correcting the moral deficiency in sectarianism, the expansion of a universal true church also dissolves the possibility of sectarian conflict and violence. The close analogy between Kant’s approach to the juridical state of nature and to sectarianism strongly suggests that the ethical state of nature refers to the proliferation of ecclesiastical faiths offering competing moral conceptions.20 This is a situation in which there is a multiplicity of sectarian views on what is unconditionally required—differing sects compete among themselves for adherents, such competition leads to conflicts and violence, and there is

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Social Conditions for Autonomy  275 no common authority to adjudicate conflicting claims on what God truly requires. The only common authority could be God himself as the legislator of moral requirements, exactly what Kant claims is lacking in the ethical state of nature. In such a situation there is widespread confusion about what religion truly requires, which, at the same time, is also confusion about what morality requires. By subordinating moral religion’s unconditional demand for moral improvement to compliance with morally indifferent ecclesiastical demands, sectarian views wrongly declare that the latter is what morality truly requires. In this condition, individuals are misled on moral and religious matters. As a consequence, they subordinate the task of becoming morally better to the performance of morally indifferent but ecclesiastically required actions. Such individuals can hardly commit themselves fully to the morality of autonomy. By trying to be good through the ecclesiastical faiths on offer, well-­meaning individuals end up being evil. In such a situation, as Kant claims, even virtuous individuals are in continuous danger of relapsing into evil. The pervasiveness of evil, on this reading, specifically results from the corrupting influence of ecclesiastical faiths that subordinate unconditional moral duty to morally indifferent ecclesiastical demands. If we think of sectarianism as an ethical state of nature, it becomes clearer why Kant maintains that the solution to the latter requires a common divine legislator who transforms the laws of virtue into public laws. If the ethical state of nature refers to the proliferation of ecclesiastical sects and their conflicting views on what God demands, the solution to this conflict can only be a common legislator who cannot be other than God himself. Since, as Kant also claims, the only access we have to what God commands is pure practical reason the true religion can only be one that requires as unconditional those duties that practical reason itself recognizes as such. The solution to the problem in sectarianism is thus to represent practical reason’s laws as divine commands. This is the only way in which such laws can become public. Consequently, individuals leave the ethical state of nature when they join a church that can be reformed according to the moral religion of reason. This means that only pure practical reason can put an end to sectarian conflict in the right way by establishing what God truly demands. However, if this is to be a solution at all, the representation of moral duties as divine commands must not constitute a threat to moral autonomy. Though Kant is not explicit on this point, I will close this section by briefly indicating why there is no such a threat. First, as a common legislator, God’s authority is not incompatible with autonomy insofar as the religion of reason is grounded in morality. To be a

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

276 Faviola Rivera-Castro religious person, on this view, is identical to being a virtuous person who subjectively regards moral duties as divinely commanded (R 153, 6:153–4).21 Belief in God as the ethical community’s superior authority presupposes that morality is grounded in the autonomy of reason. God does not legislate any commands other than those we legislate autonomously. To see moral requirements as divine commands is only a subjective way of conceiving them.22 Second, as a common judge, God’s capacity to access our innermost motives does not threaten our capacity to act from respect for the moral law. The conception of God as a being capable of knowing us better than we can know ourselves may even provide encouragement for honest self-­ examination. Third, God’s power to distribute happiness in accordance with morality would threaten autonomy only if the promise of happiness were presented as an incentive to morality.23 But this way of conceiving God’s superior power would be incoherent. The attempt to act morally with a view to happiness as a reward would be self-­defeating since morality requires that we act from respect for law alone. Finally, and despite his superior powers, God is no threat to autonomy insofar as he cannot possibly coerce us to adopt morally good maxims. Only individuals themselves have the capacity to constrain themselves to act morally. Since God cannot coherently be represented as offering natural incentives to act morally; cannot directly force us to do so (as an external authority can force us in relation to juridical laws); and ­legislates only those duties that practical reason recognizes as unconditional, belief in a divine legislator for the sake of the possibility of the ethical ­community is not incompatible with moral autonomy.

4.  Political Society and the Ethical Community Kant is explicit that the establishment of an ethical community presupposes a political society (R 106, 6:94). As we saw in the previous sections, a true church as the vehicle of an ethical community comes into being through the reformation of existing churches according to the moral religion of reason. This transition necessarily presupposes a rightful condition, since only the latter can grant individuals the rights to join the church of their choice, to communicate freely their thoughts to others on religious matters, and to subject to criticism existing ecclesiastical organizations in light of the true religion of reason. If individuals are to be able to join a true church, they must enjoy freedom of belief and of association. If they are to participate in the reformation of their own churches according to the idea of the moral religion

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Social Conditions for Autonomy  277 of reason, they must enjoy freedom of thought and of expression. In sum, if they are to engage collectively in a social process of religious enlightenment leading to the collective affirmation of a true church individuals must enjoy rights that only political society can grant. In this final section I argue that in addition to providing necessary conditions for the transition to an ethical community, political society also has an interest of its own in this transition. There are good political reasons, on Kant’s account, to favor the coming into being of a true church as the institution in which the citizenry is socialized into free submission to the rule of law. In matters of faith the primary role of the Kantian state is to guarantee the conditions of external freedom that are necessary for both individual practice and collective enlightenment. On Kant’s account, the political authority must protect freedom of thought and of belief, reject religious establishment, and abstain from interfering in the internal organization of churches. In his terminology, what we today call the “right to religious freedom” is the right to practice the ecclesiastical faith of one’s choice. As “an inner disposition lying wholly beyond the civil power’s sphere of influence,” religion can neither be prohibited nor demanded by the state (DR 137, 6:327–8). “Ecclesiastical faith,” by contrast, which is a “belief in the dogmas of a church and in the power of priests” (DR 173, 6:368), can be subject to state’s regulation. In his juridical and political writings, Kant does not explicitly mention the individual right to practice freely the ecclesiastical faith of one’s choice. Arguably, however, this right can be derived from the “innate right to freedom,” which is a right to “independence from being constrained by another’s choice” (DR 63, 6:237).24 A person is independent in this sense when she sets her own ends herself as opposed to having her ends set by others (including the state). Persons have a right to independence under the condition that their freedom to choose their own ends can coexist with everyone’s right to do the same in accordance with universal laws. According to the innate right to freedom, which, Kant maintains, is “the only original right belonging to every man by virtue of his humanity” (DR 63, 6:237), individuals are free to pursue their own ends subject to universal laws that grant the same freedom to everyone. Since practicing an ecclesiastical faith or joining a church is a particular end that an individual may choose, a system of equal freedom grants individuals the right to make that choice freely, subject to the formal condition of its being consistent with universal laws. The right to practice freely the ecclesiastical faith of one’s choice places strict limits on what the state can be authorized to do in matters of faith. In particular, it follows from this right that the state must not attempt to force the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

278 Faviola Rivera-Castro people into a particular ecclesiastical faith (by establishing one as official, say) or to forbid a particular faith or faith as such (DR 173, 6:368). To have a particular faith or no faith at all is a choice that the state must leave to individuals to make.25 On Kant’s account, churches are independent from the state in the sense that the latter does not have the right to legislate their internal constitution “or to organize them in accordance with its own views” (DR 137, 6:327).26 He especially emphasizes the state’s duty not to hinder the progress of scholarly debates on matters of faith and to guarantee freedom of thought.27 Kant’s arguments for what we today call a regime of separation between churches and the state are purely conceptual and aim to hold regardless of historical circumstances. In particular, such arguments are not addressed to the factual historical problem of conflicts among ecclesiastical faiths. According to a familiar view, the main motivation against religious establishment is its failure to treat equally people who affirm a diversity of religious conceptions. On this view, church–state separation is conditional on the ex­ist­ence of religious diversity. This is not, however, Kant’s position. Far from invoking social and political conflicts over matters of faith in his argument for separation, his account implies that in the absence of a diversity of ecclesiastical faiths the grounds for separation would remain firmly in place. Even if the social process of collective religious enlightenment were to advance and all citizens came gradually to affirm the true religion of reason, the state must not attempt to establish an official faith or to interfere in the internal organization of churches. The reason, as we have seen, is that a church, no matter how dominant it may be in society, is an institution through which individuals pursue collectively their own particular ends concerning faith. Since the transition from the ethical state of nature to the ethical community is voluntary, it may appear that a Kantian state has no role to play in favoring it. After all, the crucial asymmetry between exiting the political and the ethical states of nature is that we may be properly coerced to enter a rightful condition, while joining the universal true church is entirely voluntary.28 Nevertheless, it is Kant’s view that there are good political reasons for the state to favor the transition to an ethical community. If, as I argued in the previous section, we think of the ethical state of nature as the condition of sectarianism, the claim is that there are good political reasons to exit a condition in which churches teach people to subordinate true morality to morally indifferent ecclesiastical requirements. Such reasons concern the state’s interest in having loyal subjects who affirm the rule of law. On Kant’s account, churches play an important social role in either supporting or undermining citizens’ respect for the rule of law. Those churches that make something other than

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Social Conditions for Autonomy  279 the moral law unconditionally required not only fail to support the supremacy of law but also tend to undermine citizens’ allegiance to it. By contrast, a church that affirms the true religion of reason teaches people respect for law as the core element of both moral and religious practice and, to this extent, supports respect for law in political society. Kant reminds us that the ways of thinking that churches encourage cannot but have effects in the conduct of the citizenry. The political authority cannot be oblivious to this fact.29 Kant’s main concern with religion in the political context is not with the protection of individual liberty but with the social influence of ecclesiastical organizations. By contrast with contemporary liberalism that has traditionally focused on the protection of individual religious freedom, Kant directs his attention to churches as social powers that exert a profound influence on the citizenry. He explicitly warns against the power of churches to instill habits of obedience to despotic rule, which is rule by particular persons (be them one, many, or all) as opposed to the rule of law. On his account, those churches that do not affirm the moral religion of reason have a despotic constitution and, for this reason, exert a corrupting influence on the citizenry. In churches of this kind the clergy becomes “the single authoritative guardian and interpreter of the will of the invisible lawgiver” (R 174, 6:180), and thus rules despotically. Despotic ecclesiastical rule, Kant warns, fosters compulsory service and the habit of hypocrisy that comes with it. The deleterious effect for political society is that “the habit of hypocrisy undermines, unnoticed, the integrity and loyalty of the subjects; sharpens them in the simulation of service also in civil duties, and, like all wrongly accepted principles, brings about exactly the opposite of what was intended” (R 174, 6:180). Kant is concerned here with a danger to political society that persists even in the absence of sectarian conflict and violence. It is no mystery why Kant focuses on churches as posing a special kind of danger to the supremacy of law. Insofar as churches make claims to unconditional obedience, they can come into direct conflict with the state’s claim to supreme authority. Kant’s solution to this conflict not only differs from Hobbes’s defense of state control over religion and from Rousseau’s recommendation of an official civil religion. Kant’s position also differs from liberal views that favor a combination of church–state separation with a permanent policy of toleration toward the continuous proliferation of ecclesiastical faiths that are consistent with justice. By contrast with these approaches, his proposal is to favor the development and expansion of an ecclesiastical faith, independent from official control and grounded in the moral religion of reason, that provides the adequate kind of sociability for the collective ­

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

280 Faviola Rivera-Castro affirmation of the morality of autonomy and the rule of law. The true visible church plays the role of a social institution in which people are socialized into free submission to the supremacy of both moral and juridical laws. This is not to deny that all sorts of religious practices that are not in­con­sist­ ent with justice must be tolerated. On his view, however, an official policy of toleration of those faiths that comply with justice cannot be the definitive solution to the problems posed by sectarianism. Religious toleration, though required by freedom of belief is, on his view, a merely provisional measure to be superseded, through collective enlightenment, by the replacement of sectarianism with a religious unity centered on the moral religion of reason. Tolerance of other ecclesiastical faiths can only be provisional before everyone consents to “an ecclesiastical form commensurate to the dignity of a moral religion, viz. a free faith” (R 128, 6:123n). A Kantian political society admits of a certain degree of ecclesiastical pluralism provided that all ecclesiastical faiths affirm the true religion of reason and thus meet the condition of not including anything in their constitutions “which contradicts the duty of its members as citizens of the state” (R 107, 6:96). By contrast, tolerance without enlightenment leaves the door open to all sorts of religious superstition and enthusiasm (subjection of reason to feeling) that undermine the morality of autonomy and the rule of law. Kant maintains that the best political means to favor the collective affirmation of the true religion of reason is the full protection of freedom of thinking which, on his view, brings collective religious enlightenment with it. He argues that, contrary to what is usually thought, this protection does not cause civil instability. He explicitly claims that “in freedom there is not the least cause for anxiety about public concord and the unity of the commonwealth” (WE 21, 8:41). Though the freedom to communicate one’s thoughts to others is a right that the state must protect, his point is that unrestricted freedom of thinking is also in the government’s own interest. Kant’s conception of freedom of thinking, however, differs from today’s constitutionally protected freedom of thought. While the latter is only subject to the external constraints of justice, the former is also subject to an internal standard of correctness. According to him, freedom to think not only is opposed to compulsion both civil and over conscience but also signifies the subjection of reason to only those laws that it gives to itself (WO 12, 8:144–5). Civil compulsion in this case would be the government’s taking away citizens’ freedom to publicly communicate their thoughts to others. This freedom, Kant maintains, is indispensable in order to be able to think correctly. By “compulsion over conscience” he means the power that church authorities exert through their influence on people’s minds in obstructing the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Social Conditions for Autonomy  281 examination of reason in religious matters. The rejection of the authority of reason in thinking is not freedom, according to Kant, but lawlessness that manifests itself in multiple ways: as enthusiasm, superstition, or libertinism (the principle of recognizing no duty) (WO 12, 8:145–6). He warns that rejection of the authority of reason results in the self-­destruction of freedom of thinking (WO 12, 8:146). Lawlessness undermines morality, true religion, and the rule of law. The full exercise of freedom of thinking brings collective religious enlightenment with it insofar as the former necessarily entails subjection to the laws of reason. In religious matters this means the public subjection of ecclesiastical tenets and practices to criticism in light of the moral laws of practical reason. This rational criticism, according to Kant, shows that some, perhaps most, existing ecclesiastical practices go wrong in the demands they make on the faithful and that the moral religion of reason is the only true religion. Thus, on his account, the public philosophical critique of morally corrupting social practices is crucial for bringing into being the ethical community. He assigns philosophers the task of publicly submitting to rational criticism those social practices that undermine the morality of autonomy and citizens’ respect for the rule of law.30 Kant’s emphasis on the importance of this kind of socio-­philosophical critique for bringing into existence the social conditions that most favor the collective affirmation of the morality of autonomy and the rule of law leaves no doubt that political conditions alone do not suffice. A rightful condition that grants basic rights and liberties is a necessary condition for the possibility of an ethical community but hardly sufficient. This brings us back to the question with which we began about the most adequate social conditions for successful self-­ constitution. I suggested that, according to Korsgaard, such conditions take the form of a liberal society that grants basic rights and liberties and also welcomes the pluralism of ways of life that ensues with the exercise of reflective capacities under conditions of freedom. Kant, however, is emphatic that unless such political conditions are accompanied with collective enlightenment, pluralism can go terribly wrong. This conclusion follows from the premise that Korsgaard shares with Kant according to which individual autonomy is subject to an internal standard of correctness. Successful self-­constitution entails endorsement of the right principles, not just any values whatsoever. From this normative perspective, the pluralism of ways of life that a liberal society enables is not, by itself, a good thing, as Kant’s objections to sectarianism make clear. On a Kantian view, only a pluralism of ways of life that affirm the morality of autonomy can be welcomed as an expression of human freedom.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

282 Faviola Rivera-Castro To express human freedom, however, the right kind of pluralism cannot be promoted through coercive means, political or otherwise. Kant’s proposal is to further collective enlightenment through the socio-­philosophical critique of morally corrupting social practices. Those who disagree with his claim that an ethico-­religious community provides the right kind of socialization into the morality of autonomy and respect for the rule of law should still agree on the need to account for the social conditions that are most favorable for becoming autonomous individuals and law abiding ­citizens. Korsgaard’s view of self-­constitution, I believe, also calls for such an account.31

Notes 1. Christine  M.  Korsgaard, Self-­Constitution: Agency, Identity, and Integrity (Oxford: Oxford University Press, 2009), 7, 20–6. 2. I thank Kyla Ebels-­Duggan for helpful comments and important objections to a previous version of this paper. I also thank Fernando Rudy-­Hiller and all the participants at the practical philosophy workshop at Instituto de Investigaciones Filosóficas (National Autonomous University of Mexico) for their comments and suggestions. 3. John Rawls, Political Liberalism (New York: Columbia University Press, 1993), 19. 4. In A Theory of Justice, rev. edn. (Cambridge, MA: Harvard University Press, 1999), Rawls maintains that a just society actively promotes such an ideal (“moral education is education for autonomy,” 452) but takes this claim back in Political Liberalism (New York: Columbia University Press, 1993) (“justice as fairness . . . leaves the weight of ethical autonomy to be decided by citizens severally in light of their comprehensive doctrines,” 78). 5. Though Korsgaard (Self-­Constitution, 108) employs the term “autonomy” for both the capacity for self-­determination and its ideal exercise, the distinction between the two is clear enough: the former is a mere capacity that grounds the possibility of its ideal exercise. 6. Rawls, Political Liberalism, xxiv. 7. As some authors maintain, the historical origins of liberalism are to be found in the controversies over religious toleration during the European wars of religion. See Rawls, Political Liberalism, xxiv. 8. Frederick C. Beiser, Enlightenment, Revolution, and Romanticism: Genesis of Modern German Political Thought, 1790–1800 (Cambridge, MA: Harvard University Press, 1992), chs. 1–2 provides a view of Kant’s liberalism in its historical context. 9. Immanuel Kant, Religion Within the Boundaries of Mere Reason, in Religion Within the Boundaries of Mere Reason and Other Writings, trans. and ed. Allen  W.  Wood and George di Giovanni (Cambridge: Cambridge University Press, 1998). Hereafter I refer to this work as R followed, according to standard usage, by the page number of the English edition and the page number and volume of Kants Gesammelte Schriften. The duty to establish an ethical community is, Kant maintains, a duty of the entire human

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Social Conditions for Autonomy  283 race toward itself (R 108, 6:97). This is a duty “sui generis” that “differs from all others in kind and in principle” insofar as the idea of “a universal republic based on the laws of virtue differs entirely from all moral laws”: while the latter “concern what we know to reside within our power,” we cannot know whether the whole of a universal republic is in our power (R 109, 6:98). 10. Kant distinguishes between the church “invisible” and “visible”: The former is “the mere idea of the union of all upright human beings under direct yet moral divine world-­governance”; the latter is “the actual union of human beings into a whole that accords with this ideal” (R 111, 6:101). 11. Immanuel Kant, Metaphysical First Principles of the Doctrine of Virtue, in The Metaphysics of Morals, trans. and ed. Mary Gregor (Cambridge: Cambridge University Press, 1991), 206 (6:405). Hereafter I refer to this work as DV. 12. In the following section I explore further why Kant maintains that moral laws must become public laws. 13. Ebels-­Duggan provides an alternative interpretation of the ethical community that does away with a divine legislator in her “Moral Community: Escaping the Ethical State of Nature,” Philosopher’s Imprint 9 (2009): 1–19. For an explicit objection to the religious character of the ethical community, see Paul Guyer, “Kantian Communities,” in Charlton Payne and Lucas Thorpe, eds., Kant and the Concept of Community (Rochester, NY: University of Rochester Press, 2011), 113. For a broader questioning of the religious implications of Kant’s moral philosophy, see Onora O’Neill, “Kant on Reason and Religion,” in Grethe  B.  Peterson, ed., The Tanner Lectures on Human Values (Salt Lake City, UT: University of Utah Press, 1997), 269–308. 14. This argument for the transition from morality to religion is independent from the desideratum to realize the highest good (a state of affairs in which individuals are virtuous and each one of them is happy in exact proportion to their virtue), though Kant also maintains that the ethical community has the latter as its end. For a detailed discussion of the relation between the ethical community and the highest good, see Lawrence R. Pasternack, Routledge Philosophy Guidebook to Kant on Religion within the Boundaries of Mere Reason (Abingdon, UK: Routledge, 2014). For criticism of the idea of the highest good, see David Sussman, “The Highest Good: Who Needs It?” in Joachim Aufderheide and Ralf M. Bader, eds., The Highest Good in Aristotle and Kant (Oxford: Oxford University Press, 2015), 215–29. 15. Kant is not committed to the claim that there can be only one true church. Instead, he observes “in the various churches divided from one another because of the difference in their kinds of faith, one and the same true religion can nevertheless be met with” (R 116, 6:108). According to this, there can be a diversity of “true” churches as long as all of them affirm the true religion of reason. 16. He also maintains that “in the molding of human beings into an ethical community, ecclesiastical faith naturally precedes pure religious faith,” though, from a moral point of view, it ought to happen the other way around (R 115, 6:106). 17. This question has been largely ignored in contemporary studies of Kant’s Religion. 18. Immanuel Kant, The Conflict of the Faculties, in Religion and Rational Theology, trans. and ed. Allen W. Wood and George di Giovanni (Cambridge: Cambridge University Press, 1996), 275, 7:52. Hereafter CF.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

284 Faviola Rivera-Castro 19. Immanuel Kant, Metaphysical First Principles of the Doctrine of Right, in Gregor, The Metaphysics of Morals, 124, 6:312. Hereafter DR. 20. The possibility of understanding the ethical state of nature as the condition of sectarianism has been overlooked in recent studies of Kant’s Religion. An advantage of this understanding is that it provides an answer to those who question the religious nature of the ethical community. 21. Kant writes that in the true church “each indeed obeys the law (not the statutory one) which he has prescribed for himself, yet must regard it at the same time as the will of the world ruler as revealed to him through reason” (R 127, 6:122). 22. As Allen Wood emphasizes in “Religion, Ethical Community, and the Struggle against Evil,” in Charlton Payne and Lucas Thorpe, eds., Kant and the Concept of Community: Volume 9 NAKS Studies in Philosophy (Rochester, NY: University of Rochester Press, 2011), 122–4, “subjective” here does not mean “private” or “optional,” and he warns us against what he calls an “extreme liberal interpretation.” Instead, “subjective” means that belief in God is not grounded on anything objective as a proof of His existence, which according to Kant is impossible anyway. As Wood indicates, “a religious person, who regards her duties as divine commands, need not be certain that her duties are in fact commanded by God . . . Religion requires that I have duties, that I have a concept of God (as a possible supreme cause of things) and that my awareness of duty is subjectively enlivened by the thought that if there is a God, then my duties are God’s commands to me” (124). “Religiousness” on this reading “is solely a matter of a person’s subjective attitude toward the moral life” (121). 23. Kant writes: “The teacher of the Gospel speaks of a recompense in the world to come . . . [not as] an incentive of actions but only (as an uplifting representation of the consummation of divine goodness and wisdom in the guidance of the human race) as an object of the purest admiration and greatest moral approval for a reason which passes judgment upon human destiny as a whole” (R 159, 6:162). 24. Arthur Ripstein, Force and Freedom: Kant’s Legal and Political Philosophy (Cambridge, MA: Harvard University Press, 2009), 31. As he indicates, “[m]any individual rights are grounded in the ‘authorizations’ that are ‘already contained’ in the innate right to freedom” (209). Such rights include freedom of expression and of association, both of which are crucial for the transition to an ethical community. 25. Kant actually holds that atheism would be dangerous to the state. He maintains that, for this reason, “it is never advisable to destroy a popular faith” (R 111, 6:111). 26. Churches must also be independent from the state financially. As Kant explicitly maintains, the costs of maintaining them must be charged to the congregation (DR 137, 6:328). 27. Immanuel Kant, “An Answer to the Question: What Is Enlightenment?” in Practical Philosophy, trans. and ed. Mary Gregor (Cambridge: Cambridge University Press, 1996), hereafter WE; “What Does It Mean to Orient Oneself In Thinking?” in Wood and Giovanni, Religion and Rational Theology, 12–14, 8:144–6, hereafter WO; and Religion 122, 6:114. 28. He writes: “woe to the legislator who would want to bring about through coercion a polity directed to ethical ends!” (R 107, 6:96). 29. He explicitly claims that morality is “the mainstay on which the government must be able to count if it wants to trust the people” (CF 281, 7:60). Mark Lilla examines the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

The Social Conditions for Autonomy  285 political implications of what he calls Kant’s conception of Homo religious in “Kant’s Theological-­Political Revolution,” The Review of Metaphysics 52 (1998): 397–434. 30. In The Conflict of the Faculties Kant advises the government on how to regulate the teachings of the Theology and Philosophy Faculties with a view to favoring religious enlightenment and thus having loyal subjects. For reasons of space I cannot go into this discussion here. 31. I thank the editors for their kind invitation. It is a great honor for me to contribute to this volume. Christine Korsgaard was my mentor at Harvard and I take this opportunity to express my gratitude and affection.

Bibliography Beiser, Frederick C. Enlightenment, Revolution, and Romanticism: The Genesis of Modern German Political Thought, 1790–1800. Cambridge, MA: Harvard University Press, 1992. Ebels-Duggan, Kyla. “Moral Community: Escaping the Ethical State of Nature,” Philosopher’s Imprint 9 (2009): 1–19. Guyer, Paul. “Kantian Foundations for Liberalism.” In Kant on Freedom, Law, and Happiness, 235–61. Cambridge: Cambridge University Press, 2000. Guyer, Paul. “Kantian Communities.” In Charlton Payne and Lucas Thorpe, eds., Kant and the Concept of Community: Volume 9 NAKS Studies in Philosophy, 88–120. Rochester, NY: University of Rochester Press, 2011. Kant, Immanuel. Metaphysical First Principles of the Doctrine of Right. In The Metaphysics of Morals, translated and edited by Mary Gregor, 35–177. Cambridge: Cambridge University Press, 1991. Kant, Immanuel. Metaphysical First Principles of the Doctrine of Virtue. In The Metaphysics of Morals, translated and edited by Mary Gregor, 181–279. Cambridge: Cambridge University Press, 1991. Kant, Immanuel. “An Answer to the Question: What Is Enlightenment?” In Practical Philosophy, translated and edited by Mary Gregor, 15–22. Cambridge: Cambridge University Press, 1996. Kant, Immanuel. The Conflict of the Faculties. In Religion and Rational Theology, translated and edited by Allen  W.  Wood and George di Giovanni, 239–328. Cambridge: Cambridge University Press, 1996. Kant, Immanuel. “What Does It Mean to Orient Oneself In Thinking?” In Religion and Rational Theology, translated and edited by Allen W. Wood and George di Giovanni, 1–18. Cambridge: Cambridge University Press, 1996. Kant, Immanuel. Religion Within the Boundaries of Mere Reason. In Religion Within the Boundaries of Mere Reason and Other Writings, translated and edited by Allen  W.  Wood and George di Giovanni, 33–191. Cambridge: Cambridge University Press, 1998.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

286 Faviola Rivera-Castro Korsgaard, Christine M. Self-Constitution: Agency, Identity, and Integrity. Oxford: Oxford University Press, 2009. Lilla, Mark. “Kant’s Theological-Political Revolution.” The Review of Metaphysics 52 (1998): 397–434. O’Neill, Onora. “Kant on Reason and Religion.” In Grethe  B. Peterson, ed., The Tanner Lectures on Human Values, 269–308. Salt Lake City, UT: University of Utah Press, 1997. Pasternack, Lawrence  R. Routledge Philosophy Guidebook to Kant on Religion within the Boundaries of Mere Reason. Abingdon, UK: Routledge, 2014. Patten, Alan. “Liberal Neutrality: A Reinterpretation and Defense.” The Journal of Political Philosophy 20 (2012): 249–72. Pogge, Thomas W. “Is Kant’s Rechtslehere a ‘Comprehensive Liberalism’?” In Mark Timmons, ed., Kant’s Metaphysics of Morals: Interpretative Essays, 133–58. Oxford: Oxford University Press, 2002. Rawls, John. Political Liberalism. New York: Columbia University Press, 1993. Rawls, John. A Theory of Justice, rev. edn. Cambridge, MA: Harvard University Press, 1999. Ripstein, Arthur. Force and Freedom: Kant’s Legal and Political Philosophy. Cambridge, MA: Harvard University Press, 2009. Sussman, David. “The Highest Good: Who Needs It?” In Joachim Aufderheide and Ralf  M.  Bader, eds., The Highest Good in Aristotle and Kant, 215–29. Oxford: Oxford University Press, 2015. Wood, Allen W. “Religion, Ethical Community, and the Struggle against Evil.” In Charlton Payne and Lucas Thorpe, eds., Kant and the Concept of Community: Volume 9 NAKS Studies in Philosophy, 121–37. Rochester, NY: University of Rochester Press, 2011.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Index For the benefit of digital users, indexed terms that span two pages (e.g., 52–53) may, on occasion, appear on only one of those pages. action: and unity of the self  182–3, see also agency, and unity of the self as a functional kind  13–15 belief–desire model of see belief–desire model of action constitutive aims or principles of  85–7, 90; see also agency, constitutive aims or principles of; constitutivism see also agency; deed agency: and self–consciousness  100–1, 103–6, 112; see also self–consciousness and the body  184–5 and unconscious motivation  82–3; see also self–deception and unity of the self  182–5; see also action, and unity of the self as the source of normative standards  1–10, 13–16; see also action, constitutive aims or principles of; constitutivism; normativity constitutive aims or principles of  x, 2–4, 7, 13–14, 85–7; see also action, constitutive aims or principles of; constitutivism respect-worthy exercises of, see autonomy, respect for akrasia, see weakness of the will Anderson, Elizabeth  221–2 animals: and self–preservation  5–6 defined 215 human beings as  215–16, 229, 231–2 inner see desire minds of  133–4 our obligations to  x–xi, 213–14, 216–17, 219, 224–5, 228–33 what is good–for, see good

Anscombe, Elizabeth  85 Aristotle  viii, 11, 84–6, 117–18, 127, 143n.14 autonomy: and reciprocal legislation  226 as a capacity  214, 225–6, 260n.4; see also self–consciousness formula of  233n.8 respect for  185–8, 190–2, 194n.34 social conditions for  264–8, 275–6, 281–2 see also agency; freedom; self-constitution belief–desire model of action  81–2; see also desire Blum, Lawrence  223 Bristow, William  106–9, 111–12, 119–20 Buchanan, Allen  176–8, 182–3 Buddhism 36–7 Categorical Imperative  viii, 23, 59–60, 62–5, 74n.44, 93n.14, 240–2, 246, 264, 266; see also autonomy, formula of; formula of universal law; humanity, formula of; kingdom of ends; moral law consequentialism  ix, 89 constitutive aims see action, constitutive aims or principles of; agency, constitutive aims or principles of; constitutivism constitutive principles see action, constitutive aims or principles of; agency, constitutive aims or principles of; constitutivism constitutivism  x, 85–7, 93n.14; see also action, constitutive aims or principles of; agency, constitutive aims or principles of; constructivism

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

288 Index constructivism: and the normative question  50–1, 57–65 modest vs. ambitious  65–6, 68–70 the general strategy of  23–7, 43n.6, 73n.24 see also constitutivism contractualism 230–2 control: discretionary or voluntary  105–6, 199–201, 203, 205–7 managerial or manipulative  103–6 see also responsibility Davidson, Donald  81, 127, 142 deed 247–50; see also action dementia 175–96 Descartes, Rene  96–7 desire  79–83, 100–3, 105–11, 118–20, 126–35, 137, 139–41, 182–3, 215, 220, 224–5, 260n.4, 268–9; see also incentive dignity  221–2, 225–6, 228, see also end–in–itself; humanity, formula of duty: motive of  18n.4, 245–6, 254, 259–60; see also incentive, ethical vs. juridical juridical, see obligation, juridical Dworkin, Ronald  178–89 end-in-itself  213–14, 216, 222, 224–6, 228–9; see also dignity; humanity, formula of Enoch, David  90–1; see also realism ethical community  264–5, 268–73, 275–82; see also Kant, on ethical community ethical life  2, 4, 12–13, 17 expressivism  87, 90 fanaticism 164–7 first–person, see self–consciousness, and the first person formula of autonomy, see autonomy, formula of formula of humanity, see humanity, formula of formula of universal law  59–65, 68–9, 226, 241–2, 260n.7; see also categorical imperative freedom: and determinism  131–2

as a burden  131, 137–8, 140; see also desire of belief/conscience/thought  267–8, 270–1, 273, 276–8, 280–2 of the will  63–4, 131–2, 246; see also autonomy external (independence)  188–9, 243–6, 250–7, 259–60, 277 friendship  154–5, 202–5 good: contrasted with right/obligation  222, 227–8 evaluative or functional vs. final  214–15, 220 for animals  215–16 guise of the, see guise of the good pluralism of conceptions of the  264, 267–8, 281 tethered vs. absolute  216, 219–21 guise of the good  81, see also incorporation thesis Herman, Barbara  68–9, 240–63 Hieronymi, Pamela  100–1, 103–6 humanity: contrasted with personality  2 formula of  64–5, 74n.44, 213–14, 216, 224–9, 241, 260n.7, 261n.12; see also categorical imperative; dignity; end–in–itself; moral law Hume, David  67, 78–9 identity, see integrity; personal identity; practical identity imputation, see deed; responsibility, and imputability incentive: of inclination vs. socially scripted  141 juridical vs. ethical  251–6 role in theory of motivation  118–19, 127–8, 134–40 see also desire; duty, motive of; incorporation thesis inclination, see desire incorporation thesis  127–9, 134–7; see also desire; incentive independence, see freedom integrity 147–74 and agency  13–15, 147–50

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

Index  289 and commitments/convictions  147–50, 152, 154–69, 170n.8 and honesty  147, 151–3, 157–8 and practical identity  135–6, 149–50; see also practical identity and reflection  156–7, 166–8 and thoughtfulness  147, 154–60, 163–8 and truth  147, 153–4, 157–61, 163–4, 169 as a virtue  150, 168–9 internal standards see action, constitutive aims or principles of; agency, constitutive aims or principles of; constitutivism intuitionism 80–1 Jaworska, Agnieszka  194n.34 Kant, Immanuel:  ix, 240–86 on ethical community  268–72, 276–82; see also ethical community on obligations to non-human animals  x–xi, 213, 219, 261n.12; see also animals on religion  40–1, 45n.15, 264–5, 268, 270–82; see also ethical community; religion; state of nature, ethical on self-consciousness  111–15; see also self-consciousness on the formula of humanity  64–5, 213, 221–3, 227–8; see also humanity, formula of on the Categorical Imperative  viii, 1–3, 59–60, 63–4, 67–8, 235n.65 his theory of motivation  127–9, 134–7; see also desire; incentive; incorporation thesis kingdom of ends  10, 12, 64, 69–70, 225–7, 262n.29; see also end-in-itself; ethical community; humanity, formula of; moral law law see formula of universal law; moral law liberalism  264–8, 279–81 love  7–8, 27–8, 31–2, 128, 222–4, 255; see also valuing (attitude) Lovejoy, Arthur  97–9, 102–3, 119–20 maxim  59–64, 68, 135–7, 139–41, 241–2, 250–1, 269–70, 275–6, see also formula of universal law; incorporation thesis

Mill, John Stuart  74n.28, 227, 229–30 Murdoch, Iris  223–4 moral law  1–4, 13–14, 17–18, 65, 74n.44, 128, 144n.23, 225–7, 235n.65, 240–1, 246–50, 252–3, 258–9, 260n.4; see also categorical imperative; humanity, formula of; kingdom of ends; respect, for the moral law normative question, the  5–7, 9, 16, 50–1, 53–5, 56–8, 62–4, 66–7, 70, 102–3; see also normativity, skepticism normativity  x, 2–6, 78–9, 85–90, 95, 100–3, 119–20, 215; see also action, agency, constitutivism obligation: and accountability  217–19, 226–7, 229–32 and identity  5–6 and reactive attitudes (e.g., blame)  226–7; see also reactive attitudes and rights  216–17, 225–6, 232 concepts and conceptions of  57–60, 62–5, 67, 69–70 directed (relational, bipolar)  216–17, 225–6, 232 juridical  240, 243, 245–6, 252–3, 257–60; see also incentive, juridical vs. ethical obligatory law see moral law purported vs. real  54, 56 second–personal character of  217–19, 227, 229–33 to other animals, see animals, our obligations to see also normative question, the pain and pleasure  6, 38–40, 131, 139, 176–7, 231, 254–6 Parfit, Derek  80–1, 183, 220 personal identity: and physical continuity  179, 182, 184–5 and psychological continuity  176–9, 182–3 see also practical identity personality: contrasted with humanity  2 juridical  240, 246–7, 251, 254, 258–60; see also obligation, juridical moral  240, 246–8, 251, 255

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 19/05/22, SPi

290 Index practical identity  ix, 4–9, 12–13, 135–6, 149–50, 266–7; see also personal identity Prichard’s Dilemma  51, 53–4, 60, 69–70 principle in acting/of action, see maxim promises  60, 68, 181, 242, 252 prudential concern  26, 53, 178–82 racism 181–3 rational agency see agency Rawls, John  vii–viii, 50–1, 230–1, 267–8 reactive attitudes (e.g., blame)  12–13, 198–200, 203–5, 217–18, 222, 226–7, 229–30; see also obligation, and reactive attitudes; obligation, and accountability; responsibility; Strawson, P.F. realism  22, 50, 54–6, 58, 61–2, 91, 161–2, 208, 220–1 reason (the faculty)  52–3, 78–9; see also agency, autonomy, self–consciousness reasons (for actions or attitudes): as essentially public  10–11 external 79–81 internal 79–81 motivating vs. normative  92n.4 religion  21–2, 32–3, 40–1, 45n.15, 187–92, 264–5, 268, 270–82 respect: as a normative attitude  221, 223–6, 229–30, 232 for the moral law  252–3, 275–6 see also autonomy, respect for; end-in-itself responsibility 197–212 and control  103–4, 199–201, 203, 205–7; see also control and imputability  247–50, 253–4; see also deed and personal identity  178–83; see also personal identity conditions that undermine  127 theoretical vs. practical conceptions of 205–9 right 240 universal principle of  240–1, 246, 250–1, 253 innate  243–5, 253–4, 256–60, 277 see also obligation; good

Sartre, Jean–Paul  108–11, 115–20, 140–1 Scanlon, T.M.  56–7, 72nn.10,12,14, 75n.56, 197–212, 220, 231 self-consciousness 95–125 and Kant’s “I think”, see Kant, on self-consciousness and normativity  6, 95, 100–3, 106, 119–20, see also normativity and reflection/reflective distance  4, 97–111, 116–20, 147–8 and self-division  95, 97–103, 105–11, 119–20 and the first-person  95–8, 102, 108–20 self-constitution  6, 13–15, 206–7, 264–8, 281–2; see also agency; autonomy; practical identity self–deception  16–17, 140, 152, 167 Shiffrin, Seana  186–7, 194nn.34–36 Sidgwick, Henry  222 skepticism  17–18, 50–6, 65–7, 69–70, 159–64; see also normative question, the Smith, Adam  97–100, 102–3, 119–20 Smith, Angela  197–8 social practices  10–12, 15–17, 177–9, 182–4, 203–4, 264–5, 272–3, 280–1; see also ethical community state of nature: ethical  265, 269–75, 277 juridical or political  243, 257–8, 269–70, 273–5, 277 Strawson, P.F.  198–9, 201, 203–5 value and goodness, see good depends on valuing  22, 220–2 is not univocal  221–4 of animality  214, 216, 220–1, 229 of humanity  64–5, 213–14, 216, 221, 229; see also humanity, formula of and integrity, see integrity, and value see also good, valuing valuing (attitude)  38–40, 44n.14, 220–4; see also love; value weakness of the will  126–46, 185–6, 194n.35, 201 Williams, Bernard  17, 19n.11, 79–81, 83–4, 120n.1, 169, 227 Wittgenstein, Ludwig  10, 95, 113–15