Reason, Value, and Respect: Kantian Themes from the Philosophy of Thomas E. Hill, Jr. [Hardcover ed.] 0199699577, 9780199699575

Citation preview

Reason, Value, and Respect

Reason, Value, and Respect Kantian Themes from the Philosophy of Thomas E. Hill, Jr.

Edited by

Mark Timmons and Robert N. Johnson

1

1 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 2015 The moral rights of the authors‌have been asserted First Edition published in 2015 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: 2014951244 ISBN 978–0–19–969957–5 Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY 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.

For Robin and the many graduate students, friends, and colleagues who have enriched my work and my life.

Contents Acknowledgements Contributors Introduction Mark Timmons and Robert N. Johnson

ix xi 1

PART I.  Respect and Self-Respect 1. Servility and Self-Respect: An African-American and Feminist Critique Bernard Boxill and Jan Boxill

19

2. Humility, Arrogance, and Self-Respect in Kant and Hill Robin S. Dillon

42

3. Respect as Honor and as Accountability Stephen Darwall

70

PART II.  Practical Reason 4. Hypothetical Imperatives: Scope and Jurisdiction Mark Schroeder

89

5. More Right than Wrong Jonathan Dancy

101

6. Autonomy and Public Reason in Kant Onora O’Neill

119

PART III.  Social, Political, and Legal Philosophy 7. Private and Public Conscience (Or, Is the Sanctity of Conscience a Liberal Commitment or an Anarchical Fallacy?) Gerald Gaus

135

8. Kant on Three Defenses in the Law of Homicide Jeffrie G. Murphy

157

9. Virtue, Repugnance, and Deontology Matt Zwolinski and David Schmidtz

178

10. But What About the Animals? Cheshire Calhoun

194

viii contents

PART IV.  Kant’s Ethics 11. The Supererogatory and Kant’s Imperfect Duties Marcia Baron

215

12. Did Kant Hold that Rational Volition is Sub Ratione Boni? Andrews Reath

232

13. Kantian Complicity Julia Driver

256

PART V.  Conclusion 14. Looking Back: Main Themes and Appreciation Thomas E. Hill, Jr.

269

Published Works by Thomas E. Hill, Jr. Index

297 303

Acknowledgements Thanks to Tom Hill for his good advice in the planning stage that led to the present volume. Tom wishes to thank Jordan MacKenzie and Adam Cureton for their assistance with his contribution and to Jordan for the bibliography. A special thanks to Oliver Sensen, who served as an OUP referee for this volume. And finally, we are grateful to our OUP editor, Peter Momtchiloff, for his encouragement and advice.

Contributors Marcia Baron is Professor of Moral Philosophy at the University of St Andrews and Rudy Professor of Philosophy at Indiana University. Bernard Boxill is Pardue Distinguished Professor at the University of North Carolina, Chapel Hill. Jan Boxill is Senior Lecturer in Philosophy and Director of the Parr Center for Ethics at the University of North Carolina, Chapel Hill. Cheshire Calhoun is Professor of Philosophy at Arizona State University. Jonathan Dancy is Professor of Philosophy at the University of Texas at Austin and Senior Research Professor at the University of Reading. Stephen Darwall is Andrew Downey Orrick Professor of Philosophy at Yale University. Robin S.  Dillon is William Wilson Selfridge Professor of Philosophy at Lehigh University. Julia Driver is Professor of Philosophy at Washington University in St Louis. Gerald Gaus is James E. Rogers Professor of Philosophy at the University of Arizona. Thomas E.  Hill, Jr. is Kenan Professor of Philosophy at the University of North Caroline, Chapel Hill. Jeffrie G. Murphy is Regents’ Professor of Law, Philosophy, and Religious Studies in the Sandra Day O’Connor School of Law at Arizona State University. Onora O’Neill is Hon. Fellow at Newnham College, Cambridge. Andrews Reath is Professor of Philosophy at the University of California, Riverside. David Schmidtz is Kendrick Professor of Philosophy and joint Professor of Economics at the University of Arizona. Mark Schroeder is Professor of Philosophy at the University of Southern California. Matt Zwolinski is Associate Professor of Philosophy at the University of San Diego.

Introduction Mark Timmons and Robert N. Johnson

Thomas English Hill, Jr., was born in Atlanta, Georgia in 1937. He earned a BA degree in philosophy from Harvard University in 1959, completed the Bachelor of Philosophy degree at Oxford under Gilbert Ryle in 1966, and went back again to Harvard for the Ph.D. under John Rawls in 1966. His first academic job as assistant professor was a short-lived appointment at Johns Hopkins followed by two years at Pomona College. He then became a member of the UCLA department from 1968 to 1984. Since 1984 he has been Professor of Philosophy at University of North Carolina at Chapel Hill, and Kenan Professor since 1996. In 1994 he gave the Tanner Lectures on Human Values and in 2003 he was elected fellow of the American Academy of Arts and Sciences. The chapters in this volume are a testament to the range and influence of Hill’s work. From his early work taking up Kant’s views on supererogation and practical reason, through his treatments of ethical and political topics such as self-respect, affirmative action, punishment and terrorism, to his work on theoretical topics such as constructivism in normative theory, Hill has consistently brought a thoughtful and commonsensical approach to questions at every level of practical philosophy and its history. We have divided the contributions into four general topics that cover the range of Hill’s work: respect and self-respect; practical reason; social, political, and legal philosophy; and Kant’s ethics. Hill then has added some reflections on how he came to hold the views in each of these areas and provides a “big picture” of the themes of his work.

I  Respect and Self-Respect Hill’s work on respect and self-respect, beginning with one of his most widely read and influential essays, “Servility and Self-Respect,” has shown that how one treats oneself can raise ethical questions that are more serious than one might have assumed. Indeed, a theme in much of his work is that we can fully understand what is morally objectionable or honorable about many types of behavior and attitudes only when we understand that they essentially involve objectionable or honorable behavior and attitudes toward oneself. However, in Hill’s view, while respect is a foundational element of the Kantian normative perspective, it is not always well understood what normative role

2  MARK TIMMONS AND ROBERT N. JOHNSON it plays in that perspective. On one picture, Kantian ethics tells us to take up a certain sort of valuing attitude toward something that is present in each human being, their “humanity.” Hill’s picture is considerably different from this. The substantive principle that we respect a person’s humanity doesn’t yield enough, or at least enough that is precise and non-controversial, in the way of action-guidance. A view that Hill thinks is preferable is one in which respect for humanity is not a substantive principle from a distinctive “Kantian” value system, but a characteristic of a general procedure for deliberating rationally in the midst of competing claims about what one ought to do. Oversimplifying a bit, to “respect persons” is to adopt and abide by principles that anyone, were they reasonable, would agree to live by. In, “Servility and Self-Respect” Hill considers three individuals—Uncle Tom, the Self-Deprecator, and the Deferential Wife—whose behavior is described as being servile and for that reason, morally defective. The defect in question, as characterized by Hill, is either a failure to recognize one’s moral rights or a failure to appreciate the true worth of those rights. In their “Servility and Self-Respect: An African-American and Feminist Critique,” Bernard and Jan Boxill challenge Hill’s suggestion that the behavior of the three characters as described by Hill is genuinely servile. As they point out, the dictionary definition of the term “servile” refers to one’s “having or showing an excessive willingness to serve or please others,” but Hill’s depiction of the Self-Deprecator, according to the Boxills, is more aptly classified as apathetic rather than servile, which may be a troubling condition, but not necessarily a moral defect. Moreover, the Boxills also challenge Hill’s claim that the Self-Deprecator’s behavior is caused by a failure to understand his rights. They raise similar worries about Hill’s Deferential Wife, who is described as being overly willing to obey her husband. But if this is to count as servility, she must display an excessive willingness to serve and please her husband, which on the Boxill’s understanding of the example is not in evidence. They speculate that one reason why Hill’s depiction of the Deferential Wife strikes many as a compelling portrait of servility is because readers are likely to be influenced by the unstated assumption that this character is dominated and therefore servile, even though this is not made explicit in Hill’s story. (Similar remarks apply to the Uncle Tom character.) Of course, Hill might re-describe the case slightly to make clear that the husband in his example does indeed dominate his wife. And then the argument might be that domination corrupts both dominators and the dominated, tending to make the latter servile. But the Boxills resist this line of argument by examining the idea that domination breeds servility in the dominated. They argue that domination is likely to lead the dominated to pretend servility in order to deceive, which in turn leads to associated vices, including dishonesty, hypocrisy, insincerity, and many others, and so to a kind of corruption that need not result in genuine servility. Rather, those who are dominated likely resent being dominated, an indication that such individuals, far from not recognizing or properly valuing their moral rights, should be understood as affirming their moral equality.

Introduction  3 In another highly influential article, “Ideals of Human Excellence and Preserving Natural Environments,” Hill argues that persons who unnecessarily destroy the environment exhibit vices that hinder the development of various virtues including “proper humility” which, as he describes this trait, involves both a correct sense of one’s relative importance and a kind of self-acceptance that recognizes the nature and limits of the kind of creature we are. Robin S. Dillon’s, “Humility, Arrogance, and Self-Respect in Kant and Hill” examines the claim that humility is a genuine virtue. On one understanding of humility (which follows dictionary definitions of the term), this attitude involves a lowly estimation of oneself, which one would naturally associate with a kind of objectionable submissiveness and self-degradation, and so hardly a companion to such virtues as wisdom, courage, benevolence, and justice. However, according to Dillon, those who champion what they call “humility” associate it with respect for persons, and tend to use the term for what would more aptly be described as wisdom, honesty, or integrity. Dillon finds this puzzling and mistaken, and suggests that one source of this mistake is that humility is often thought to be the opposite of arrogance; that is, one might suppose that insofar as exaggerated self-importance is characteristic of arrogance, individuals who lack this vice are properly thought of as humble, whereas these individuals are better understood as having a proper sense of interpersonal recognition respect for others as moral equals. Indeed, the lowly self-estimation characteristic of the dictionary definition of humility is typically manifested in the vice of servility, as Kant characterizes this vice. Kant distinguishes “false humility” from “true humility,” which serves as the basis for Dillon’s inquiry into whether there is a form of humility that can play an important role in a moral life. Kant held that true humility results from an honest comparison of one’s conduct and character with the moral law, or, more precisely, with the high standard of moral perfection that the law requires us to strive toward. On this understanding, according to Dillon, humility “is the knowledge of one’s limitations and deficiencies judged in comparison with a self-given law.” This sort of knowledge, then, plays a crucial role in cultivating the virtue that Dillon calls “agentive recognition self-respect,” a form of respect that involves committing oneself to the high standards of conduct and character set by the self-given moral law. So, while not a virtue, true humility provides an agent with information that figures importantly in a morally committed agent striving for self-perfection. One dominant theme in Hill’s work is his interpretation of Kantian respect for persons which, in his Tanner Lecture “Respect for Humanity,” he contrasts with an aristocratic view according to which individuals merit differential treatment owing to heredity and social rank. Stephen Darwall’s contribution, “Respect as Honor and as Accountability,” explores this contrast, arguing that the sort of respect characteristic of the Kantian notion grounded in one’s dignity and the sort of respect characteristic of honor codes involve different conceptions of personhood and, correspondingly, different conceptions of moral and social order. According to Darwall, both types of respect are forms of recognition respect (in contrast to appraisal respect) and so

4  MARK TIMMONS AND ROBERT N. JOHNSON concern how individuals are to regulate their behavior toward one another. According to Darwall’s view of the Kantian notion, respect for persons essentially involves what he calls second-personal address and acknowledgment. In holding another person morally responsible for some moral offense, one addresses that person as an individual with dignity, as someone who is subject to one’s legitimate demands and can be held accountable for that offense, and as someone who has authority to hold oneself (the addressor) accountable for any moral offenses. Furthermore, one’s full recognition of such mutual accountability requires that one acknowledge others’ second-personal authority, in the sense that one not merely comply with the legitimate demands of others, but that one relate to those others as persons with equal dignity and thus acknowledge their standing to hold accountable. This conception of second-personal respect, then, involves conceiving of persons as having equal standing, and morality as equal accountability. Honor respect, according to Darwall, involves conceiving of persons as having a particular hierarchical status that is socially constituted in the sense that the “status just consists in the common knowledge that it has been publically recognized by the relevant people.” One significant point of contrast between second-personal respect and honor respect concerns differences in appropriate responses to violations of these forms of respect. Responses to violations of the former typically involve indignation and resentment on the part of those offended and guilt on the part of the offender, which are appropriate because of their conceptual tie to second-personal address and blame. By contrast, contempt is the natural attitude to violations of honor respect, the response to which on the part of the offender is shame. Of particular note is that while contempt in response to honor disrespect involves treating the offender as an inferior, the negative reactions associated with second-personal respect implicitly affirm the equal standing of those being held morally accountable. Differences in these responses to violations bring into sharp relief the contrasting conceptions of social order and personhood associated with these two forms of recognition respect.

II  Practical Reason The dominant conception of practical reason and rationality among philosophers is and long has been Humean. On that conception, as Hume himself said, reason is the slave of desire and so makes no pronouncements about any action beyond whether it is the best way of achieving some goal set by what we want. Hill’s work on practical reason has been to argue that reason has significantly more to say than this. Hill’s approach has been to delineate and develop the Kantian conception of practical rationality by emphasizing a particular advantage it has over the Humean conception. That advantage is that it is addressed to the perspective of a deliberating agent, not to the perspective of an observer concerned to produce an accurate description of practical reasoning. That is, in approaching the topic of practical reason (and all other topics

Introduction  5 in practical philosophy), Hill has consistently emphasized the Kantian aim of finding a suitable and rationally defensible answer to the question, What ought I to do? The advantage of addressing this perspective is that it must take seriously the possibility that some things must not be done or must be done. This is so in the case of instrumental as well as moral deliberation. It is, however, not at all obvious how such answers can be arrived at, beginning with descriptions of what human beings in fact do. As Mark Schroeder remarks in his “Hypothetical Imperatives:  Scope and Jurisdiction,” Hill’s 1973 article, “The Hypothetical Imperative” continues to be one of the best entrees into ongoing discussions of instrumental rationality, and arguably represents the standard contemporary understanding of Kant’s doctrine of hypothetical imperatives. One contemporary debate over hypothetical imperatives (instrumental rationality) is over matters of the scope of “ought” and thus over the proper logical structure of such imperatives. Roughly speaking, according to so-called wide scope interpretations, if one believes that one’s doing something A is necessary for bringing about some intended end E, then one rationally ought to either do A or give up E. The problem with this interpretation—one that Hill adopts—is that if one understands the “ought” involved as “x won’t do everything he or she is rationally required to do unless” the resulting wide interpretation is far too weak to capture the intuitive force of requirements of instrumental rationality; indeed, as Schroeder explains, it makes the wide interpretation uncontroversially true—a logical implication of supposedly competing narrow and intermediate scope interpretations (which are explained in his chapter). One question this raises is, Which stronger interpretation of “ought” should figure in wide scope interpretations that would make debates between narrow and wide scope substantial? Schroeder’s proposal for understanding what is at stake in disputes over scope is to focus on the notion of jurisdiction in the law. New York driving laws, to use Schroeder’s example, have jurisdiction only over those driving in New York and allows for a meaningful distinction between complying with law and merely avoiding it by, say, not being in New York and hence under the jurisdiction of that state’s traffic laws. As Schroeder notes, if one interprets “ought” in this example as “New York traffic regulations require one to,” then a wide interpretation of this requirement (which is met automatically by those not living or traveling in New York) is implausible, for on this interpretation those outside of New York are in compliance with this state’s traffic laws. The important upshot for understanding scope debates over law-like norms of instrumental rationality is that relating matters of scope to jurisdiction help in properly framing meaningful debates among narrow, intermediate, and wide interpretations of such norms. Of course, principles of instrumental rationality such as Kant’s hypothetical imperatives on the wide interpretation are supposed to have universal jurisdiction over all rational agents. In light of his remarks about jurisdiction and scope, Schroeder asks how it can be that on the wide interpretation of requirements of instrumental rationality, one can make sense of their purported universal jurisdiction. Moreover, referring to Kant’s own view, he asks how the notion of rational agency can suffice to explain such

6  MARK TIMMONS AND ROBERT N. JOHNSON universal jurisdiction. These questions lead Schroeder to briefly explore Kant’s notion of autonomy and the idea that one has authoritative jurisdiction over oneself. He proceeds to explain why narrow and intermediate accounts of instrumental rationality are attractive, given the particular picture of Kantian autonomy, and he proposes what he calls a Kantian Intermediate interpretation of hypothetical imperatives. Jonathan Dancy’s contribution, “More Right than Wrong,” takes up the claim, made by some critics of Kant, that while Kant has trouble making sense of the notion of a moral reason, W. D. Ross’s view does not. In “The Importance of Moral Rules and Principles,” Hill addressed this challenge, arguing that for Kantians, particular facts are reasons because they are “especially salient features” that figure in the fuller possible rationale one might give for or against a particular course of action. As Dancy notes at the outset of his chapter, unless more is said about what makes some fact an especially salient feature of one’s circumstances, one will lose the distinction between roles played by reasons and roles played by other aspects of the “fuller rationale.” But even if more can be said in addressing this concern, Dancy’s main claim is that Ross’s moral theory enjoys far less advantage compared to standard readings of Kant, because Ross fails to make good sense of the relation between his conception of moral reasons as prima facie duties and the rightness of an action. After explaining the troubles for Ross’s view, Dancy goes on to consider how one might improve upon Ross. As Dancy explicates Ross, what Ross’s notion of prima facie duties was meant to do is specify morally relevant features of actions (moral reasons) whose presence contributes to the overall rightness (or wrongness) of actions as a matter of degree, in the sense that, in a particular context, some of the features present in a situation can “weigh more” than others in determining the overall rightness of an act. The problem, as Dancy sees it, is that Ross’s account of prima facie duty (as presented in Ross’s 1930 The Right and the Good) fails to make sense of the normativity of the contributory because it makes no sense of the idea of a prima facie duty being a matter of degree, which is essential to the very idea of some feature contributing some normative weight toward the overall deontic status of an action. And so Ross’s attempt to explain the normativity of the overall in terms of the contributory fails. In exploring this failure, Dancy discusses H. A. Prichard’s objections to Ross’s notion of prima facie duty which leads to a discussion of the advantages of a buck-passing account of rightness, a view that Dancy has defended in the past. Dancy concludes by raising a host of worries for deontic buck-passing and then offers a suggestion about how to answer them. Onora O’Neill points out in “Autonomy and Public Reason in Kant,” that one of Hill’s most striking contributions to our understanding of Kant’s conception of autonomy is that it should not simply be equated with contemporary conceptions of autonomy as identical to some form of individual independence. Rather, on Hill’s reading, Kant’s conception of autonomy or positive freedom essentially involves select principles of action that can be communicated and shared with others, which represents a distinctively public form of reasoning. O’Neill’s contribution builds on Hill’s reading by contrasting what she refers to as Kant’s modal conception of public reason with the conceptions of

Introduction  7 public reason we find in the writings of John Rawls and Jürgen Habermas. According to O’Neill, although both Rawls and Habermas in their political philosophies stress the role of public reason, they are primarily concerned with the qualifications for actual inclusive participation in public reasoning, which they argue leads to a form of deliberative democracy. However, O’Neill notes that in addition to real world problems confronting the proposed deliberative forms of democracy, Rawls and Habermas devote little time to the question of what sorts of norm ought to guide such public reasoning. Despite these two political theorists drawing on Kant’s work, O’Neill argues that Kant’s own conception of public reason differs significantly from theirs. As is well known, Kant distinguishes private from public uses of reason. Private uses proceed from ungrounded assumptions such as church edicts or state proclamations from which such reasoning derives its (conditional) authority. By contrast, public uses of reason are not achieved merely by universal participation in deliberation and decision making. As O’Neill explains, the focus is “not on the context and conditions of actual discourse, but on the normative conditions required if discourse is to communicate with unrestricted audiences.” Kant’s approach to public reason, then, concerns those conditions governing proper public reason-giving rather than requirements for participation. Among the conditions in question are the demands that principles governing public reasoning be law-like and so be followable by others, but also that they have universal scope in that they could be adopted by all. Only if principles governing public reasoning are both law-like and universal in scope, and so freed from various forms of dogma and ungrounded assumptions, can public reasoning have the kind of authority that expresses the genuine autonomy of reason. After developing this conception of Kantian autonomy and public reason, O’Neill concludes with some reflections on whether this modal conception of public reason can be relevant to contemporary public life.

III  Social, Political, and Legal Philosophy For a philosopher who admires Kantian values, it can be distressingly difficult to square that admiration with many of Kant’s own words. How can the lofty Kantian ideal of respect for human dignity be married to his seemingly harsh retributivist statements about punishment, or to the view that no revolution is justified, no matter how unjust the government? Hill’s work acknowledges these tensions, but consistently tries to separate what is essential to the Kantian perspective from what is not. From Hill one learns that Kant may have mistakenly thought that his views had implications that they do not in fact have (the categorical imperative, for instance, need not imply that one must tell the truth to a murderer at the door), or the context can show that something Kant may at first seem to overvalue, such as rational consent, is in fact valuable in certain well-defined theoretical arguments depending on idealizations of rational agents deliberating together about the acceptable grounds of social cooperation.

8  MARK TIMMONS AND ROBERT N. JOHNSON In two of his essays, “Four Conceptions of Conscience” and “Conscience and Authority” Hill develops a conception of conscience, based on remarks by Kant, that contrasts with certain religious and relativist conceptions of this concept. The guiding metaphor of Kant’s view is that conscience is an inner judge whose principal purpose is to judge whether one’s behavior lives up to one’s own moral standards. The verdicts of one’s conscience are in one sense fallible regarding one’s objective duty. As Kant says, “I can indeed be mistaken at times in my objective judgment as to whether something is a duty or not [however] I cannot be mistaken in my subjective judgment as to whether I have submitted my practical reason (here in its role as judge) for such a judgment . . .” (MS 6: 401). As Hill develops this Kantian view, in addition to the general task of judging whether one’s conduct lives up to one’s internalized moral standards, conscience’s special task involves one’s duty of taking “due care” in passing judgment on whether one’s conduct is consistent with one’s internalized moral standards. One question Hill raises about conscience is how one should respond when the deliverances of one’s conscience conflict with the demands of authorities—conflicts that can arise dramatically in relation to war, business affairs, as well as in more mundane affairs. Should one always follow one’s conscience in such cases of conflict? In responding to this question, Hill embraces what he describes as a modest position, namely, that neither conscience nor authority always trumps the other in cases of conflict; rather, as he writes, “both should be respected, but neither is an infallible moral guide; and if we cannot satisfy both, there is need, time permitting, to look for a resolution in a process of moral reasoning” (Hill 2000: 261). Hill’s modest “resolution” of conflicts of conscience and authority is the topic of Gerald Gaus’s “Private and Public Conscience.” Gaus notes that Hill’s Kantian conception of conscience admits of two interpretations. According to what he calls the personal interpretation, one’s conscience requires that one reflect on one’s own internalized moral convictions, and while in doing so one may take account of claims of public authority, one need not. By contrast, according to the inclusive interpretation, the duty of “due care” requires that in passing judgment on one’s conduct, one consider and take seriously the reasons offered in defense of a public authority’s verdicts. As Gaus argues, it is the inclusive interpretation that best fits Hill’s Kantian conception. However, an implication of this interpretation is that, unlike the personal interpretation, supposed cases of conflict between conscience and authority are illusory because, as Gaus says, on the inclusive interpretation one has factored into one’s conscientious judgment the considerations favoring the verdicts of public authorities and so one should, contrary to Hill’s official line, follow one’s conscience. So, to preserve the potential conflict between conscience and authority, Hill’s Kantian conception requires the personal interpretation. However (for reasons Gaus explains) the inclusive conception is far superior, which, as just noted, leaves no room for the conflict in question. Furthermore, as Gaus points out, even on the more plausible inclusive interpretation, there is the problem of whether, given the depth of moral disagreement in which one finds interpersonal conflict in the demands of people’s inclusive consciences, those

Introduction  9 representing public authority are justified in imposing their views on those whose inclusive consciences are at odds with what a public authority demands. In the final part of his chapter, Gaus explores a contractualist understanding of cases of interpersonal conflicts of inclusive conscience, suggested in one place by Hill, that involves a commitment to public justification—a perspective from which one takes “up the public view and seeks accommodation with the consciences of others.” Among the topics Hill discusses in “Kant on Punishment:  A  Coherent Mix of Deterrence and Retribution?” is the question: Who should be punished through criminal law? Hill briefly comments that Kant apparently allows for cases in which there are “moral grounds for limiting the legal right to punish in special circumstances.” The cases in question concern the alleged right of necessity in certain circumstances to knowingly kill an innocent person; a mother killing her illegitimate child, and a soldier killing in a duel. Jeffrie Murphy’s contribution, “Kant on Three Defenses in the Law of Homicide,” agrees with Hill’s interpretation of Kant that these cases are not ones where there is a legal duty to punish but a moral prohibition against doing so, but rather cases of morally justified legal defenses. Murphy’s chapter examines in some detail Kant’s reasons for holding that such cases are morally justified cases of homicide. With regard to the cases of infanticide and killing in duels—both cases of so-called honor killings—Kant appeals to such considerations as, for example, the illegitimate child has the status of (or a status analogous to) contraband merchandise, the likely shame experienced by the mother resulting from public attitudes toward her, and the alleged “state of nature” both mother and duelist find themselves confronting. Murphy is critical of such defenses, though he hypothesizes that Kant’s treatment of these cases might be viewed somewhat sympathetically as striving toward the contemporary distinction in law between excuse defenses and justification defenses. Turning to the right of necessity cases, Kant’s view was that there is no right of necessity in, for example, life-threatening shipwreck cases, but nevertheless because the threat of the death penalty cannot effectively deter killing in such cases where one’s death seems imminent, the law has no authority to make such killing a criminal offense. Murphy finds Kant’s deterrence argument for this position unconvincing, but notes that Kant seems to have some appreciation for the distinction between what is relevant in defining a crime and what may be relevant in the sentencing phase of those guilty of a crime. Murphy then goes on to ask whether Kant was correct in holding that there is no legitimate doctrine of a right of necessity. Murphy organizes his discussion around the much discussed case of Regina v Dudley and Stephens. After rehearsing the essentials of that case, Murphy argues that cases of necessity do not justify killing an innocent person (thus agreeing with Kant), but that in light of the justification/excuse distinction, homicide law might include excuses for such cases. The focus of “Virtue, Repugnance, and Deontology” by Matt Zwolinski and David Schmidtz is Parfit’s much discussed “repugnant conclusion” that presumably makes trouble for utilitarianism. The problematic conclusion is easily illustrated by considering a version of act utilitarianism that emphasizes the maximization of total happiness

10  MARK TIMMONS AND ROBERT N. JOHNSON and its implications for large populations. Compare a state of affairs of, say, one billion people, all of whom are living a high quality of life, with a state of affairs involving a much larger population with people whose lives are barely worth living. Judged in terms of total utility, the much larger population is hypothesized to realize a greater total amount of utility than that realized by the smaller group, and thus counts as a better state of affairs of the two, according to utilitarianism. But surely the claim that a world with a large population in which people’s lives are barely worth living really is preferable to one in which the individuals living in a large but somewhat smaller population are well off is morally repugnant. After explaining in some detail Parfit’s case for the repugnant conclusion, Zwolinski and Schmidtz argue that the problem generalizes not only to other versions of utilitarianism, but to all standard act-centered theories of right conduct, including standard deontological theories. Inspired by Hill’s influential article on environmental ethics mentioned earlier (which has us shift attention from questions about actions that involve rights violations or involve interfering with the interests of sentient creatures to questions of character), our authors appropriate Hill’s character-based approach as a way of addressing the repugnant conclusion. In short, from a character-based assessment of someone who would prefer the much larger population filled with people whose lives are barely worth living to a smaller population of people well off, such a person does not possess the kind of humility that would view persons as playing a limited role in the biotic community. And in general, as Zwolinski and Schmidtz point out, on questions of contributing to overpopulation, those who contribute to an unsustainable growth in population by having large families are in effect guilty of free riding on the restraint of others and can be evaluated in virtue-theoretic terms as “short sighted, irresponsible, ill-willed, selfish, and so on.” The conclusion these authors draw from reflection on a character-based approach about the repugnant conclusion and related issues of overpopulation is not that act-centered theories are completely misguided and should be abandoned in favor of character-based theories. Rather, their claim, again, inspired by Tom Hill’s work, is that when it comes to certain moral problems such as a person’s relation to insentient living things such as trees and, in general, insentient features of our environment, Hill’s question about character is the proper question to ask. As have many other contemporary readers of Kant, Hill finds Kant’s doctrine on the treatment of non-human animals a “morally repugnant doctrine.” In Kant’s view, human beings have no “direct” duties to non-human animals but only “indirect” duties with regard to them. The sole rationale for this view is twofold. First, non-human animals are not rational agents (and so they are owed nothing) but second, since mistreating animals will likely impede a person’s duty to cultivate positive dispositions such as compassion that are of moral significance both in striving to perfect one’s own nature and in one’s dealings with other human beings, humans are morally obligated to not mistreat non-human animals. This essentially causal rationale is why our duties to non-human animals are only indirect.

Introduction  11 In her “But What About the Animals?” Cheshire Calhoun, inspired by certain themes in Hill’s work on such topics as self-respect and snobbery, argues that there are good Kantian grounds (even if Kant did not advance them) for holding that there are three realms of concern that human beings should have with respect to non-human animals: a positive regard for their interests, a non-instrumentalist valuing attitude toward them, and refraining from subjecting them to mockery, forms of injustice, ingratitude, and other such attitudes. As Calhoun notes, some of Kant’s remarks about animals emphasizes certain analogies between humans and certain non-human animals. Attending closely to these analogies (as a Kantian might reconstruct them) is the basis, according to Calhoun, for developing a plausible Kantian account of the three areas of concern lately mentioned. The challenge for the Kantian, of course, is that despite any analogies that do hold between the human species and select species of non-human animals, the fact that humans have a dignity in virtue of their rational natures that non-humans lack represents a morally significant difference of such import that humans only have indirect duties toward non-human animals. As Calhoun explains with respect to animal interests, we “need a positive account of why the morally significant differences between animals and humans do not warrant discounting (perhaps severely discounting) animal interests.” With regard to reasons for having a direct positive regard for animal interests and taking a non-instrumentalist attitude toward their value, Calhoun notes that although both humans and non-human animals are natural species, animal behavior, conforming as it does to deterministic laws, is necessitated to avoid such vices as drunkenness and gluttony. By contrast, humans, in misusing their freedom of choice, may succumb to “lawlessness” and thus debase themselves as below beasts. Reflection on such facts is a basis for taking a suitably humble attitude toward our own species and avoiding a kind of species snobbery that discounts if not ignores animal interests and thinks of their worth in a purely instrumentalist fashion. As for the third ethical concern about not treating non-human animals unjustly, not mocking them, and showing gratitude for faithful service, a puzzle remains. For example, animals (unlike humans) are not able to obligate one to be grateful since they cannot act on the freely adopted end of beneficence, and moreover they have no concept of gratitude needed to appreciate the reciprocity that grateful feelings and corresponding actions express. Looking to Hill’s view of Kantian ethics as an ethics of attitude as much as of action, Calhoun argues that even if refraining from ingratitude, contempt, and indifference to fairness toward non-human animals is not rationally required, nevertheless adopting such attitudes is evidence of what she calls being morally “small-minded”—something Kant himself mentions in his Lectures on Ethics—a moral defect whose avoidance is grounds for having and expressing such attitudes as gratitude toward non-human animals. After developing in some detail the points just mentioned, Calhoun’s conclusion is that Kantian ethics does have the resources—superior to those found in utilitarian approaches to the ethical treatment of animals—to account for our what intuitively seem to be proper moral responses to select non-human animals.

12  MARK TIMMONS AND ROBERT N. JOHNSON

IV  Kant’s Ethics At the heart of Hill’s work is, of course, his commitment to the core ideas of Kantian ethics. None of these is more important than the idea of autonomy. A striking feature of Hill’s work on autonomy (as Onora O’Neill points out in her contribution) is that it emphasizes that autonomy is best understood as a property of principles, not a metaphysical property of a noumenal human will. Again oversimplifying, an autonomous practical principle would be one that does not embody the achievement or realization of some value as an impetus to comply with it. The latter makes a principle unsuitable for, for instance, justifying one’s actions publicly to any reasonable agent (i.e. unsuitable for contexts in which there are competing values among a plurality of agents). This idea again brings to the forefront the deliberative perspective of one asking, What ought I to do? Indeed, most of the topics and issues that Hill discusses, such as practical questions concerning the importance of personal values and projects, make apparent the importance of this perspective, the deliberative question, and the search for answers embodied in autonomous principles. Can Kant’s ethics (or more generally, Kantian ethics) recognize the category of the supererogatory, understood roughly as actions that go “beyond” the demands of duty? In his “Kant on Imperfect Duties and Supererogation,” Hill argues that Kant’s ethics can recognize this moral category as a sub-class of actions that fulfill imperfect duties, and provides a list of characteristics a supererogatory action would likely have if such acts have a place in Kant’s ethics. Marcia Baron in her contribution, “The Supererogatory and Kant’s Imperfect Duties,” argues that although Kant’s ethics has the resources to do justice to most of the characteristics philosophers typically associate with supererogatory actions, she disagrees with Hill’s claim about the best way to characterize supererogation within the Kantian framework, and she disagrees that Kant’s ethics has a place for the category of the supererogatory recognized as such. According to Baron, the problem with Hill’s characterization is the requirement that for Kant a supererogatory action be “motivated by a sense of duty (or, perhaps, respect for moral reasons).” She argues that this requirement is only plausible if one equates moral worth with moral merit in Kant’s ethics and so holds that because a supererogatory act on Kant’s scheme would be categorized as an imperfect duty and thus count as meritorious, it would thereby have moral worth which, on Kant’s view, requires being motivated by duty. But, according to Baron, there are good textual reasons for not making this equation. To accommodate this point, Hill could drop this particular requirement, and thereby slightly modify his conception of Kantian supererogation. The resulting proposal would be roughly this: if one has adopted the maxim of beneficence and if one has continually acted on this maxim and thereby done enough to promote the happiness of others, then one may, but is not required, to do even more so long as one is not in violation of perfect duty. However, there remains an obvious tension between the claim that a supererogatory act is to be located within Kant’s category of imperfect duty and the claim that such

Introduction  13 actions go beyond duty. Hill, of course, recognizes this, and suggests that Kant’s adherence to the restrictive notion of duty as sufficient to cover all of morality did not allow him to simply distinguish between what is morally obligatory and what it is good, but not obligatory, to do. Again, Baron disagrees that Kant’s deontic terminology is too restrictive, and questions whether it would be an advantage of Kant’s ethics to think in terms of having done enough—what is minimally required by the imperfect duty of beneficence—in order to make room for the supererogatory. But even if the category of the supererogatory is not at home in Kant’s ethics, one may still ask whether Kant can more or less do justice to the sorts of phenomena typically associated with the acts of saints and heroes that are held up as examples of the supererogatory without recognizing the category itself. In the final section of her chapter, Baron argues that indeed Kant’s ethics can mostly accommodate the phenomena in question. In his “Personal Values and Setting Oneself Ends,” Hill casts doubt on the interpretive claim that Kant held a so-called strong thesis about the relation between one’s personal ends (and related maxims) and an agent’s commitment to their value, according to which the non-moral personal ends one adopts and pursues involve a commitment to “the objective moral goodness of our ends and the acts based on our maxims” (2002: 263). As Andrews Reath points out in his contribution, “Did Kant Hold that Rational Volition is Sub Ratione Boni?” the specific issue about the adoption and pursuit of ends that Hill’s essay raises extends to broader questions about how Kant’s conception of rational volition is related to the moral law, the topic of Reath’s essay. In particular, Reath is concerned with the issue of whether Kant’s conception of rational volition, properly interpreted, is committed to a strong, rationalist thesis about the relation between rational volition and the moral law, according to which, as Reath puts it, “rational volition is based on practical reasoning aimed at judgments of goodness that make a tacit claim to universality.” As Reath notes, this thesis amounts to the controversial claim that in Kant’s view all rational volition is guided by the Universal Law formulation of the categorical imperative, or at least something like this formula, as its formal constitutive norm. Reath’s aim is to examine the textual and doctrinal support for what he calls “the rationalist thesis about the will” and its implications for Kant’s practical philosophy. Reath’s case for interpreting Kant as holding the rationalist thesis in question involves a particular conception of rational volition consistent with and perhaps suggested by select passages in the Groundwork concerning a rational being’s capacity to act according to principles, remarks in the Metaphysics of Morals about the will as faculty of desire in accordance with concepts, and by a particular reading of Kant’s positive conception of freedom. Very roughly, the conception in question (very roughly, because Reath develops this conception in much greater detail) is as follows. The structure of rational volition involves adopting ends (principles or values) and reasoning from these “initiating” elements of deliberation to conclusions about which action (or series of actions) under one’s specific circumstances would achieve the end (or be called for by the principle of value), and on the basis of such reasoning forming the intention (adopt

14  MARK TIMMONS AND ROBERT N. JOHNSON a maxim) to perform the action (or series of actions) that such reasoning supports. With this structure in mind, the picture of rational volition that Reath claims is at least consistent with, if not suggested by, the passages in question is one according to which the agent takes the end (principle or value) to be good “on grounds that are valid for every rational being as such” (G 4: 413) and furthermore takes the conclusion of practical deliberation (and thus the maxim and associated action) to be rationally supported and thus as good. On this picture, then, both initiating conditions (ends, principles, values) and the conclusions about specific maxims and actions derived through practical reason are taken by the agent to have universal validity. In this way, the picture entails the rationalist thesis that all rational volition (even bad volition), because constitutively committed to the idea of universal validity, is tacitly guided by the Universal Law formulation of the categorical imperative. Reath further supports the rationalist thesis interpretation of Kant’s theory of rational volition by examining Kant’s doctrine of the autonomy of the will, and Kant’s claim that a free will and a will under moral laws are “one and the same” (G 4: 447). One putative challenge to Reath’s interpretation concerns immoral action. If rational volition—acting for what one takes to be good reasons—is constitutively tied to the (at least tacit) assumption that ends and means that figure in one’s practical reasoning are universally valid, what account can be given of an agent’s knowingly acting contrary to morality? Reath concludes with suggestions for how this worry can be addressed by a defender of the rationalist thesis. In his “Moral Responsibilities of Bystanders,” Hill considers so-called tolerance complicity—a person’s moral complicity and thus blameworthiness in cases where he or she, either though indifference or culpable ignorance, does nothing and is thus negligent in at least not standing up and confronting the wrongdoing of others. The cases Hill discusses involve instances of social oppression and those bystanders who, for example, stood by while Nazis killed Jews in the Second World War, or who did little or nothing while women were discriminated against in their own countries. Hill’s particular focus in this essay is forward-looking: exploring preparatory responsibilities of bystanders for avoiding such complicity. He emphasizes the cultivation of dispositions that would combat the sort of ignorance and indifference involved in tolerance complicity, which can be grounded in a Kantian conception of respect for the rational nature of oneself and those being oppressed. As Julia Driver explains in “Kantian Complicity,” accounting for and explaining the wrongness of such forms of complicity is an advantage of a Kantian approach to the general issue of moral complicity over direct consequentialist approaches. As she explains, consequentialists are able to account for the wrongness of cases of participation complicity—complicity involving one’s being causally efficacious in another’s wrongdoing, as when one is the wrongdoer’s accomplice. But, there are other cases in which one intentionally and knowingly participates in another’s wrongdoing with the awareness that one is not causally contributing to the wrongdoing. Such cases have led some to propose causally neutral accounts of complicity, but as Driver explains, such accounts are not able to adequately explain the kinds of bystander cases of complicity Hill

Introduction  15 considers—cases in which those complicit do not participate in, for example, wrongful oppression of targeted groups. In exploring the advantages of a Kantian approach to tolerance complicity beyond those highlighted by Hill’s discussion, Driver stresses the manner in which engaging in such complicity poses a problem for personal integrity. As she explains, in addition to the Kantian requirement that one express in attitude and action respect for the humanity of self and others, one important wrong-making feature of tolerance complicity concerns positive self-esteem which, distinct from respect, is not owed to everyone. Driver explains that this conception of self-esteem possesses a kind of integrity that would compel one to stand up for one’s values. Furthermore, as she points out, a failure of integrity in passively standing by while others engage in forms of oppressive wrongdoing may be viewed from a Kantian perspective as allowing oneself to be used—used in the sense that one’s silence is properly interpreted as a sign that one agrees with the values that fuel oppressive practices. Driver’s contribution thus extends Hill’s Kantian approach in explaining the rich resources Kantians have in plausibly explaining the ways in which forms of complicity are morally wrong.

V Conclusion In his concluding reflections, Hill traces the development of his views which begins, unsurprisingly, with the early influence of his father, the philosopher Thomas E. Hill, Sr. Philosophers at Harvard and Oxford further molded those views, but it was likely the encouragement of Ryle and Rawls to study and develop the views of Kant that had the most impact. Although unlike his father, Hill rejected Moore’s ideal consequentialism, it is clear that, beyond Kant and Rawls, many of the concerns that motivated Moore’s views have provided an important touchstone for Hill’s work. Although Hill does not try to address or reply to the contributions, he tries to put them together with his own thoughts about the contours of his work overall and the concerns that resulted in his work having the shape and content that it has.

Abbreviations G

Groundwork of the Metaphysics of Morals, in Kant, Practical Philosophy, trans. and ed. Mary J. Gregor (Cambridge: Cambridge University Press, 1996)

MS The Metaphysics of Morals, in Practical Philosophy

References Hill, Thomas E., Jr. 2000. Respect, Pluralism, and Justice: Kantian Perspectives. Oxford: Oxford University Press. Hill, Thomas E., Jr. 2002. Human Welfare and Moral Worth: Kantian Perspectives. Oxford: Oxford University Press.

PA RT I

Respect and Self-Respect

1 Servility and Self-Respect An African-American and Feminist Critique Bernard Boxill and Jan Boxill

Thomas E. Hill Jr.’s “Servility and Self-Respect” inspired a generation of moral philosophers to think seriously about servility and self-respect, moving feminists and students of race to ponder the consequences of sexism and racism for their victims. It launched an impressive literature on the virtue of self-respect and immediately on its publication its marvelously suggestive examples of the Uncle Tom, the Self-Deprecator, and the Deferential Wife became widely and familiarly acknowledged in the literature as exemplars of servility. Unsurprisingly although Hill cautioned that he was providing an account of only one kind of self-respect and two kinds of servility, the sheer breadth and power of his arguments soon created a near consensus that he had provided an account of the only kind of self-respect and servility worth our attention. We are among the feminists and students of race who have been and continue to be inspired and stimulated by the profoundly moving arguments in Hill’s great essay. But we have become wary of the consensus that it has created. We believe that its arguments have subtly coaxed his readers into supposing that his Kantian view of self-respect as appropriately understanding and valuing one’s rights is more salient and significant than it really is. So we do not challenge Hill’s claim to have provided an account of one kind of self-respect and two kinds of servility. We believe, however, that there are other kinds of self-respect and servility that are more widespread, more vexing, and perhaps more interesting that demand our closer and more serious attention.

Hill’s Theory Here are the broad outlines of Hill’s argument for his account of self-respect and servility: To help provide a “preliminary idea” of what he means by servility Hill presents us with his famous examples of the Uncle Tom, the Self-Deprecator, and the Deferential Wife carefully and painstakingly describing their behavior (Hill 1995: 77, 78). Assuming that his readers will agree that their “attitudes,” presumably reflected in their behavior,

20  Bernard Boxill and Jan Boxill are indeed servile, and that servility as he has described it is morally objectionable, Hill then asks for the grounds for supposing that these attitudes are “morally objectionable” (Hill 1995: 79, 79). After considering and dismissing several possible grounds for the supposition Hill “suggests” that the attitudes are morally objectionable because of a “moral defect in each case,” namely that the Uncle Tom, the Self-Deprecator, and the Deferential Wife, fail “to understand and acknowledge” their “own moral rights” (Hill 1995: 82). However conceding that these individuals may continue to display similarly servile and objectionable attitudes even after being successfully instructed about their rights Hill then argues that there must be a second possible ground for a servile attitude which he identifies as “placing a comparatively low value” on one’s rights. Summing up he concludes that “there are at least two types of servility: one resulting from ignorance and confusion about one’s rights and the other from placing a comparatively low value on them” (Hill 1995: 85). At this point Hill reminds his readers that his suggestion that the “objectionable feature of the servile person is his tendency to disavow his own moral rights either because he misunderstands them or because he cares little for them” is in fact so far only a suggestion. Now he looks for arguments to support that suggestion. To that end he develops a Kantian argument that every person has a moral duty as far as possible to “respect the moral law.” According to Hill this means that every person has a duty both to “do what is morally required and refrain from what is morally wrong,” and also to hold “the system in esteem,” and to “treat all the provisions of morality as valuable—worth preserving and prizing as well as obeying” (Hill 1995: 86). Hill is confident that persons who misunderstand their moral rights or understand them but care too little for them cannot possibly fulfill these duties. Since it is morally objectionable to fail to do one’s duty to respect the moral law we now have sound philosophical grounds for supporting the judgment that the attitudes and behavior of the Uncle Tom, the Self-Deprecator, and the Deferential Wife are morally objectionable.

Preliminary Critique We will not challenge Hill’s argument that people who fail to understand and value their moral rights cannot possibly fulfill the duty to respect the moral law. We also will not challenge his descriptions of the behavior of the Uncle Tom, the Self-Deprecator, and the Deferential Wife. Indeed we cannot challenge these descriptions since they are constitutive parts of Hill’s examples which he can describe however he pleases. However we can and will challenge his stipulation or suggestion that the behavior of his three characters as he has stipulated it to be is servile. Whether such behavior is servile or not depends on what the word “servile” means and he cannot just stipulate its meaning. The examples are supposed to show how Hill’s theoretical claims about the failure to do one’s duty to respect the moral law fit into our intuitions about behavior

Servility and Self-Respect  21 we recognize as servile. If he presents us with behavior we do not recognize as servile theory and intuition may not fit. We should also be wary of Hill’s stipulation or suggestion that the behavior of the characters is morally defective or that it is morally defective because of its servility. If their behavior is not necessarily servile, it may not be morally defective because it is servile. Finally we find Hill’s claim that the alleged servile behavior in question is morally defective because the Uncle Tom, the Self-Deprecator, and the Deferential Wife fail to understand and value their rights to be especially worrisome. His exact words are: “The moral defect in each case, I suggest, is a failure to understand and avow one’s own moral rights” (Hill 1995: 82). Since it would make no sense to suppose that the behavior of the three characters is morally defective if they just happened also to fail to understand and value their rights, we will assume that their failure is supposed to either explain or cause their allegedly servile and morally defective behavior. But Hill does not make it clear whether we should take this assumption as a suggestion or as a stipulation. The word “suggest” in the sentence seems to imply that we should take it as a suggestion, which we are entitled therefore to appraise for its plausibility. At times, however, Hill writes as if we should take the assumption as a stipulation. To make these and other concerns clearer let us consider the examples separately and in greater detail.

The Self-Deprecator The Self-Deprecator “is reluctant to make demands,” “says nothing when others take unfair advantage of him,” and “when asked for his preferences or opinions, he tends to shrink away as if what he said should make no difference” (Hill 1995: 77, 78). Hill stipulates that this behavior displays a kind of servility. Then he claims that it is morally defective because it stems from or displays the Self-Deprecator’s failure to appropriately understand and value his basic moral rights. Finally he relates certain facts about the Self-Deprecator; namely that he is a failure, knows that he is a failure, believes falsely that basic moral rights have to be earned, and consequently believes also falsely that he has no moral rights and little moral status (Hill 1995: 81). We have no difficulty believing that a person who has very few merits, and who believes he is a failure, will likely behave like the Self-Deprecator. And we also have no difficulty believing that such behavior is flawed. But we do have difficulty in believing that it is servile. This simply does not ring true. Our dictionary tells us that “servile” means “having or showing an excessive willingness to serve or please others.” We can see how it has come to have that meaning when we note that its origin is the Latin “servus” meaning slave (Oxford American Dictionary). The slave must be eager to please, eager to obey. He must behave as if his master’s will that he do something is reason that he do it at once and without hesitation. Frederick Douglass cited Hugh Auld his master as unfolding the “true philosophy of slavery” as follows: The slave “should know nothing but the will of his master, and learn to obey it” (Douglass 1987: 92).

22  Bernard Boxill and Jan Boxill Presumably the good slave must also search out ways to please his master. But we do not see either an excessive willingness to obey or an excessive willingness to please in the stipulated description of the behavior of the Self-Deprecator. We do not see that it even says that he acknowledges that he has a master and consequently we cannot see how he can know only the will of his master and be eager to obey it. We also think that he lacks the confidence to think that he can successfully search out ways to please anyone. He does not protest when others take advantage of him, and is reluctant to express his preferences or opinions. But this is not the same as displaying an excessive eagerness to obey or to please them. Rather than showing servility his tendency to refuse to speak up when invited to do so suggests intransigence inconsistent with servility. A slave who failed to speak up when his master invited him to do so would be in for a good hiding for failing to show an adequate willingness to obey. Further it is unwarranted to suppose that the Self-Deprecator’s failure to protest and complain at interference is motivated by an eagerness to please and obey. His failures and inadequacies naturally incline him to believe that protesting interferences would be pointless. His failures and inadequacies are also quite enough to make him reluctant to express his opinions even after being invited to do so. In his experience accepting such invitations has repeatedly exposed him to be a fool and he is tired of making a fool of himself. We are not trying to change Hill’s examples. They are his examples and he can describe them however he chooses to. He can describe the Self-Deprecator’s behavior. And he can describe the Self-Deprecator as servile. But he cannot just stipulate that the Self-Deprecator’s behavior as he has stipulated it to be is servile. Whether that behavior is servile depends on the meaning of the word “servile” and no one can just stipulate what that word means—unless of course he wants to change the subject. We believe that the behavior Hill ascribes to the Self-Deprecator is properly described as apathetic rather than servile. Consider how Rawls describes the behavior of a person who he says lacks self-respect. According to Rawls such a person fails to have a “sense of his own value, his secure conviction that his conception of his good, his plan of life is worth carrying out,” or fails to have a “confidence” in his “ability,” so far as it is within his power to, to fulfill his intentions (Rawls 1971: 440). These attitudes and beliefs seem very similar to those Hill’s Self-Deprecator is likely to have given his painful awareness of his failures. A person who knows that he is a failure is very likely not to have confidence in his ability so far as it is within his power to fulfill his intentions. Further such a person is very likely to behave like the Self-Deprecator. Nevertheless Rawls does not describe him as servile—correctly we believe. He describes him as plagued by “self-doubt” as lacking a will to strive for his ends, and as sinking into “apathy and cynicism” (Rawls 1971: 440). These traits may be faults but they are not servility. Someone may be plagued by apathy, self-doubt, and cynicism and yet not be at all too eager or too willing to please or obey others. Indeed an excessive willingness to please and obey seems like a commitment that the apathetic person is unable to make. Hill is likely to protest that the Self-Deprecator’s behavior is meant to display only a kind of servility that is morally defective. We are wary of such disclaimers. Saying that

Servility and Self-Respect  23 one is only describing one kind of servility may seem modest, but it may be a Trojan Horse that conceals something that one is bringing into the argument that has no business being there. Thanks to David Sachs we have a famous example of the practice and its consequences in Plato’s stipulation (in the Republic) that one kind of justice is sticking to your own job. We believe that we have given many reasons to doubt that the Self-Deprecator’s behavior displays any kind of servility. Consequently we cannot agree that it displays a kind of servility that is morally defective. It may be objected that it does not matter that the Self-Deprecator’s behavior displays apathy rather than servility because a lack of self-respect may show itself as servility or as apathy. We may be reminded that Rawls suggests that the person he describes as apathetic fails to have self-respect because he is apathetic. But Rawls’s suggestion led him to a view of self-respect very different from Hill’s. If Hill were to allow that the Self-Deprecator’s behavior is apathetic rather than servile he might be led to endorse Rawls’s view of self-respect rather than the view he famously endorsed. Let us move to Hill’s suggestion or stipulation that the Self-Deprecator’s behavior displays a kind of servility that is morally defective because he does not believe in or value his basic moral rights. Since we do not agree that the Self-Deprecator’s behavior is servile we cannot agree that it is an example of a kind of servility that is morally defective. We have agreed that it evinces apathy but we do not see why it must therefore be morally defective. Apathy is troubling and perhaps self-destructive, but not obviously or necessarily defective morally. But suppose that Hill says that he stipulates that the Self-Deprecator’s behavior stems from his failure to understand and value his basic moral rights. We are not sure that he can do that. He can stipulate whatever he wants to about the Self-Deprecator’s behavior. It is his example. For the same reason he can stipulate that the Self-Deprecator does not understand and value his rights. But he cannot just stipulate that the Self-Deprecator behaves as he does because he fails to understand or value his basic moral rights. That one thing happens because of another seems to be a factual claim that no one can just stipulate to be the case. Of course philosophers often concoct examples with highly unlikely situations. Think of some of the many variations of the “trolley car” examples. This may encourage the complaint that Hill can stipulate that the Self-Deprecator behaves as he does because he fails to believe in and value his rights—even if we do not know whether this stipulation is empirically credible. But the improbable stipulations in the trolley car examples are attempts to isolate certain facts in order to test our intuitions about how we would feel it justifiable or permissible or excusable to behave if we were in circumstances that instantiated those isolated facts precisely as they are stated to be. But Hill’s stipulation (if it is a stipulation) that the Self-Deprecator’s behavior is servile and morally defective because he fails to understand and value his moral rights does not test any intuitions about how we should feel it justifiable to act if the stipulation turned out to be true. It certainly does not help us decide whether it is possible that the Self-Deprecator’s behavior can be servile and morally defective just because he fails to understand and value his rights. On the contrary it compounds two claims in such

24  Bernard Boxill and Jan Boxill a way as to defy any testing of our intuitions about that possibility. Specifically it combines the claim that the Self-Deprecator fails to believe in and value his rights with the claim that he is acutely aware of his failures. Since the second claim is fully adequate to explain his behavior the stipulation cannot test an intuition that the first claim explains his behavior. But suppose that Hill protests that he was not stipulating a causal relation between a failure to understand and value one’s basic rights and the Self-Deprecator’s behavior. Suppose he says that he meant that the failure in question explained the behavior in question, in the sense that it provides the reason why the Self-Deprecator behaved as he did. We agree that behavior tainted in this way by a failure to understand and value one’s basic rights would probably be morally defective; but we do not see that the Self-Deprecator’s behavior is necessarily tainted in this way. We have already shown that the Self-Deprecator has sufficient reason to behave as he does without appealing to his beliefs about his rights.

The Deferential Wife The Deferential Wife seems very willing to please and obey her husband. Hill describes her as “wearing clothes” that her husband “prefers,” as inviting guests “he wants to entertain,” as making love “whenever he is in the mood,” and as “moving to a new city in order for him to have a more attractive job” (Hill 1995: 78). But servility is an excessive willingness to please and obey. Is her willingness to please and obey her husband excessive? Is she therefore servile? Hill stipulates or suggests that she is. We argue however that his readers have easily acquiesced to his suggestion (if he is making one), only because he has led them to assume that the Deferential Wife is a dominated person. Since dominated people, especially dominated wives, have a reputation for being servile, his readers therefore easily agree that the Deferential Wife is servile. Since Hill never actually says that the Deferential Wife is a dominated person his readers may not be obliged to acquiesce to his suggestion that she is servile. Although Hill does not actually say that the Deferential Wife is a dominated person, he does very carefully describe her as a wife, refraining from giving her a name like “Mary” or “Jane” which his readers could use to think of her as anything else but a wife. The feminist and republican literature has long argued wives were dominated by their husbands—at least until very recently. Those arguments are sound. Until very recently in America all the forces of law, tradition, respectability, and custom, not to mention sheer brute strength and the capacity to inflict bodily harm and injury to women without fear of serious retaliation combined to give husbands overwhelming power over their wives, and consequently made them their wives’ dominators, whether they wanted to occupy those roles or not. These considerations created a consensus that husbands dominate their wives. Relying on that consensus and subtly constraining his readers to think of the Deferential Wife as most saliently a wife, Hill leads his readers to

Servility and Self-Respect  25 think of her as dominated by her husband. They cannot help thinking that way even if they do not know that they are. It does not matter that he does not say that her husband abuses her or threatens to abuse her. A dominator does not have to abuse or threaten to abuse to be a dominator. The benevolent dictator is still a dictator. The benevolent dominator, and therefore the benevolent husband is also still a dominator. The feminist and republican literature has also long argued that domination tends to make the dominated servile and in particular that upon the slightest suspicion, wives are servile to their husbands because they are dominated by their husbands (Skinner 2008: 92–4). Many of the liberal and enlightened (which includes many if not most philosophers) have been persuaded by its arguments and a near consensus has been reached at least in those classes that many wives at least during the period of their greatest domination were servile and also that wives more generally upon the slightest suspicion are servile. Hill did not invent that near consensus but he makes good use of it. Having powerfully but quietly implied that the Deferential Wife is dominated, he uses the consensus that wives upon the slightest suspicion are servile to suggest that she in particular is servile, though he takes care to strengthen the implication by including the word “Deferential” in her name. We argue that the unstated assumption that the Deferential Wife is dominated by her husband and therefore probably servile does most of the work in leading Hill’s readers to agree with him that she is servile. Her actual behavior as Hill stipulates it to be does none of that work. It only confirms the tacit and antecedently entrenched assumption that her husband dominates her and that she is therefore probably servile. If Hill’s readers were denied that assumption and forced to rely only on her behavior as he stipulates it to be they would be little inclined to agree with him that she was servile. Consider an example of the behavior that allegedly makes her servile: she moves to another town so that he can pursue his career. Such self-sacrificing behavior does not make her servile! Sports fans do not cry “servility” when they hear familiar tales of doting parents sacrificing everything and moving to new cities just so that their little skating or tennis or gymnastics prodigies have a better chance to fulfill their dreams of winning Wimbledon championships or Olympic medals. Or consider Sam. We stipulate that Sam is a white male, intelligent, educated, profitably employed and able bodied and that he indulges his wife in exactly the way the Deferential Wife indulges her husband. Do we leap to the conclusion that Sam is servile? Certainly not. Most people would probably say that he was head over heels in love with his wife. But Sam differs from the Deferential Wife only in there being no implication that his wife dominates him while there is implication galore that her husband dominates her. Having persuaded his readers of the Deferential Wife’s servility (illicitly we believe), Hill stipulates that her servility is morally defective. Then he suggests or stipulates that her servile behavior is morally defective because it stems from or is explained by her failure to understand and value her rights. Suppose that he suggests this. The only ground for taking such a suggestion seriously is the assumption of her morally

26  Bernard Boxill and Jan Boxill defective servility. Given that assumption, when Hill offers us an explanation of her behavior that would present it as faulty—that she fails to understand and value her rights—we are ready to take it. But we have already shown that we have little reason to believe that she is servile, aside from unstated assumptions that we may not even be aware that we have. Consequently we cannot use her supposed servility to help us conclude that her behavior is morally defective and stems from or is only adequately explained by her failure to understand or value her rights. We must base that conclusion only on Hill’s description of her behavior. According to that description she often fails to insist on her rights in order to please him and help him succeed. Is such behavior morally defective or does it strongly suggest that she fails to understand and value her rights? Only the unstated and covert implications of her servility allow Hill to suggest that it is, and escape the retort that he has slandered human nature by implying that only deep moral flaw can account for or explain supererogatory behavior. When a soldier throws herself on a live grenade to save her comrades no one leaps to the conclusion that she did so because she fails to understand and value her rights. A person does not fail to understand and value her rights just because she fails to insist on them at every turn or fails to insist on them at many turns. Paradigmatically having rights means being free to choose to exercise them or not. If they obligated us to exercise them they would be just another kind of duty. Of course we may be obliged to do what we have a right to do, but never simply because we have a right to do it. Consequently setting aside the powerfully motivating though never stated and perhaps unconscious implications of the Deferential Wife’s domination and servility, we see no reason to doubt that she prizes and values her rights. Indeed she would seem to have good reason to prize and value them since they assure her that she can justly defy rules that might forbid her from acting on her free choices to please her husband. There would be some ground for suspicion that she did not prize her rights if she never asserted them in order to prevent her husband from doing what she had a right he not do. But we have no grounds for such suspicions. Hill does not say that she never draws the line to stop her husband from doing something a sane or moral person would not allow him to do. He does not say for example that she defers to her husband’s proposal to commit a crime. Further as far as we know she may punctiliously press her rights against everyone other than her husband. If the above discussion is sound Hill very much needs his readers to assume that the Deferential Wife’s husband dominates her; for if they were deprived of that assumption they could reasonably doubt his suggestion that she is servile and fails to understand and value her rights. But such doubt would be fatal to his attempt to present them with a believable case of servile and morally defective behavior caused by a failure to understand and value one’s rights. The mystery is Hill’s reticence to say explicitly that the Deferential Wife’s husband dominates her. If he needs the assumption that her husband dominates her why does he not state it explicitly? To answer this question we need to say a little more about domination.

Servility and Self-Respect  27 One person, say W, is dominated by another, say M, if the following conditions hold: M has the power to interfere with some of W’s more important interests at his pleasure including those interests protected by her rights, and can do so without significant cost to himself; W may perhaps be able to escape M’s domination, but may perceive the costs of doing so to be high or very high. As a result of that perception W may choose to remain dominated by M. If she makes that choice she will then have two main options for dealing with his domination: First, she may conduct her life making no concessions to the fact that he dominates her. Second, she may try to devise strategies to reduce the frequency or severity of his interferences. Suppose she chooses the first option. In that case if she is lucky and has a kind dominator her life may be tolerable except for the fear that there is nothing to prevent him from becoming a monster. If she is unlucky and has a monster as dominator to begin with, which is the more likely possibility, her life will be miserable. Suppose she chooses the second option and tries to devise strategies to reduce the frequency or severity of her dominator’s interferences. One such strategy would be to simply keep a low profile and keep out of his way. Unfortunately this strategy is apt to be unsuccessful especially if as is likely her dominator’s power has given him a taste for abusing his underlings. If so she may decide to try the strategy of pleasing the dominator. She knows that like most everyone else he naturally tends to look kindly or less unkindly on those who please him and to treat them less harshly that he otherwise would. She knows too that he is likely to be especially pleased by those who seem to recognize and esteem him for his admirable qualities. Of course she is unlikely to see any such qualities in him, but she will know that he will fancy that he has some of them and would be well pleased to have that fancy credited by the recognition and esteem of others. Consequently she may strive to persuade him that she esteems and admires him for the reasons he thinks he deserves to be esteemed and admired. Of course he may suspect that she is flattering him. But she will learn from her mistakes, perfect and polish her performance, and eventually often convince him of her sincerity. Nor will she be able easily to end the charade once she has started it. She will know that he will want it to continue since it feeds his pride and that if he discovers that she has been deceiving him that his outrage at her betrayal—as he will see it—will put her in extraordinary danger. Consequently once started their intertwined maneuverings will likely continue and end up with her appearing to be excessively willing to please and obey him. In other words she will come to appear to be servile. The feminist literature has argued that at least until recently in the US married women have chosen to deal with their domination by trying to please their husbands. That argument assumes the reasonableness of women and is consequently probably sound. Until recently the costs to a woman of defying or getting free of her husband were very high. In fact they often still are. But it is reasonable and perhaps sometimes even obligatory to please a dominator in order to mitigate the costs of his domination at least if these costs are very high. Without considerable qualification Patrick Henry’s maxim “Give me liberty or give me death” does not entail an obligation to die

28  Bernard Boxill and Jan Boxill to avoid enslavement. Slaves are not obliged to commit suicide to escape their enslavement. Neither are they obliged to disobey their dominator’s every order and be beaten to death for their temerity. For the same reason, a married woman is not obliged to commit suicide if this extraordinary measure is the necessary cost of escaping her husband’s domination. She is also not obliged not to try to please him if only pleasing him will prevent his murdering her children. In other words pleasing one’s dominator does not necessarily mean that one is either servile or fails to understand or value one’s rights. These considerations may help explain the mystery of Hill’s reticence to admit explicitly that the Deferential Wife’s husband dominates her. For suppose that Hill admitted this explicitly; then it would enter his readers’ minds that the Deferential Wife’s behavior might stem from her domination. Since in that case her behavior would not be evidence of either her servility or of a failure to understand and value her rights, those readers might then start wondering whether Hill has given them a believable case of morally defective servility caused or explained by a failure to understand and value one’s rights. Of course domination is in various ways very much a matter of degree. The costs of escaping it may be large or small; and dominators often cannot do terrible things to the dominated with impunity. Thus while the dominated may sometimes justifiably choose a pretense of servility to deal with their domination, it does not follow that they are always justified in doing so. If the worst the boss can do to an employee with impunity is to shorten her break time, lavish flattery to avoid his unwanted sexual advances would be servile and morally objectionable. This possibility may seem to imply that Hill’s readers reasonably acquiesce to his suggestion that the Deferential Wife is servile and fails to understand and value her rights. He implies that her husband dominates her, and therefore leads them to suppose that she is servile since most dominated people are; but since he never tells them that her husband can beat her with impunity, he also manages to intimate that her deference to him is undertaken to evade some petty inconvenience, and consequently is servile and morally defective. But they should be wary of these tortuous maneuverings. Not everything dominated people do they do because they are dominated. The more we lower the power of the dominator to do serious harm with impunity, the more we raise the possibility that those he dominates behave as they do for innocent reasons. Fear probably motivates the deference of those dominated by a very powerful dominator even if he is kindly and unlikely to use his power to harm them. It less probably motivates the behavior of those dominated by a less powerful but also less kindly dominator. Especially if his power is very limited and he is also kindly the dominated may please him because they like him, not because they fear him. Consequently if the only advantage the Deferential Wife seems to gain from pleasing her husband is avoiding some petty discomfort, there is no compelling reason to believe that she pleases him to avoid that petty discomfort. If we think that there is, we must agree with Hobbes that only fear and the prospect of advantage motivate people.

Servility and Self-Respect  29 It may be objected that the above argument trivializes the well-known tendencies of domination to corrupt both the dominator and the dominated. For if the dominated are corrupted, then upon the slightest suspicion anything they do that happens to please the dominator must be viewed as tainted by corruption. But the corruption peculiarly associated with domination is servility. Consequently we are justified in suspecting servility whenever the dominated please their dominators. We deny the charge that we trivialize the dangers of domination. We agree with the maxim that power tends to corrupt. Although that maxim is typically announced as a reminder that power corrupts the dominators, we agree that power also tends to corrupt the dominated. We deny however that it necessarily makes them servile. Our position is somewhat controversial. As we have noted the republican literature has created a near consensus that domination breeds servility in the dominated. Republican thinkers from John Milton and Algernon Sidney in the seventeenth century to Quentin Skinner and Philip Pettit in the twentieth and twenty-first centuries maintain that domination makes the dominated servile. Their strongly avowed position is part of the more general republican contention that domination is rife, evil, and should be abolished or at least greatly reduced. Naturally republicans have therefore belabored the idea that domination is not only generally corrupting, but corrupting in the specific sense that it breeds servility in the dominated. They perceive correctly that this contention constitutes perhaps the greatest indictment of domination they could possibly contrive; certainly servility is destructive of the very agency of those it afflicts and consequently cripples them more thoroughly than any other vice. Unfortunately they do not for the most part present elaborate arguments to support their contention. Apparently they take it to be obvious and typically rely on their observations of dominated people who they report to be almost invariably servile. We do not challenge their reliance on observation. The contention that domination breeds servility is a fundamentally factual claim about cause and effect and observation must therefore play a considerable role in establishing its truth. But we believe that republicans have interpreted their observations in a way that unfairly supports the contention that they are so zealous to demonstrate. They seem to forget that the dominated are under pressure to conceal their real thoughts, beliefs, and attitudes, and to behave and speak as if they were abjectly servile. In short the dominated are under pressure to put on a performance of servility. But a performance of servility cannot be morally defective just insofar as it is a performance. To be morally defective it must somehow reflect badly on the thoughts, beliefs, and attitudes of the performers. Consequently the performance of servility of the dominated if it is to be morally defective cannot be taken at face value. But the republicans take it at face value and therefore beg the question. The following remarks should therefore not be read as trying to demonstrate that domination does not always or often make the dominated servile. They are meant to suggest that the republican observers of domination may have been duped by the performances of servility put on by the dominated. Perhaps the dominated intended to dupe them. The republican observers of domination are often members

30  Bernard Boxill and Jan Boxill of the privileged classes who the dominated might suspect to be informants for their dominators. Republicans may take it to be obvious that domination breeds servility in the dominated because as we have already admitted the dominated often seem excessively willing to please and obey. Since servility is an excessive willingness to please and obey it may therefore seem to follow that domination makes the dominated servile. In fact what follows is that domination often makes the dominated seem servile. It may be argued that the continual pretense of servility of the dominated will inevitably lead them to become servile. We do not see the force in this argument. Continually doing something will probably develop into a habit. But since we have admitted only that the dominated continually pretend to be servile, the most that can be inferred from our concession is that the dominated will probably develop a habit of pretending to be servile. Indeed the dominated probably intend to develop such a habit. It might secure the pretense on which their well-being depends. The republican observers may complain that when the appearance of servility is as constant and unrelieved as it is in the dominated, then one is entitled to conclude that the appearance is the reality. We doubt first of all that the appearance of servility of the dominated is as unrelieved as this complaint supposes, and will later present evidence that it is not. But the republicans cannot conclude that the appearance is the reality even if the appearance of servility in the dominated is as constant as they say it is. It may be constant because their well-being or their lives depend on it being constant. Spies known as moles are constant in their pretense of being the enemy. Do they therefore become the enemy? One republican writer has at least gestured at arguments that might seem to support the view that domination tends to make the dominated servile. According to Philip Pettit, “Domination is generally going to involve the awareness of control on the part of the powerful, the awareness of vulnerability on the part of the powerless, and the mutual awareness—indeed the awareness among all the parties to the relationship—of this consciousness on each side. The powerless are not going to be able to look the powerful in the eye, conscious as each will be—and conscious as each will be of the other’s consciousness of this asymmetry.” And he concludes that as a result “The master–slave scenario will materialize, and the asymmetry between the two sides will be a communicative as well as an objective reality” (Pettit 1997: 60, 61). Pettit does not actually say at this point that as a result domination breeds servility in the dominated, but in the next paragraph he claims that “conscious” of the problem he just listed, John Milton “deplored the ‘perpetual bowings and cringing of an abject people’ and Algernon Sidney observed that ‘slavery doth naturally produce meanness of spirit, with its worse effect, flattery’ ” (Pettit 1997: 60, 61). If these considerations are supposed to constitute an argument that domination breeds servility in the dominated we fail to see its force. Perhaps its force is supposed to lie in the fact that the dominated will see in every face he looks into that he is dominated. We admit this fact of course but cannot see how it is supposed to imply that the dominated will therefore be servile.

Servility and Self-Respect  31 Perhaps the argument relies on the idea that we gain an estimation of ourselves from others’ estimation of us that we see revealed in their faces. As Adam Smith indicated others’ faces are like mirrors in which we look and gain an estimation of ourselves (Smith 1982: 109–12). But even if this idea is to be credited we fail to see how it constitutes an argument that domination makes the dominated servile. At best it suggests that domination will lead the dominated to see that he is dominated because everyone sees him as a dominated person. But we doubt that the dominated will need anyone’s testimony in order to know that he is dominated. But perhaps Pettit intends a subtler argument. He may not be saying that the dominated will discover that she is dominated because most others see that she is dominated. He may be pointing to the fact that the dominated will discover that everyone knows that she is dominated. The fact that everyone knows that one is dominated is not strictly a part of being dominated; as Pettit himself admits secret domination is possible though rare. Consequently the fact that everyone knows that he is dominated may be news to the dominated. We admit that this bit of news is likely to make domination considerably harder for him. Probably, it will make him ashamed of his condition, if he was not already ashamed of it. But although his shame will be a hardship that has to be added to his already onerous hardships we do not see why it must make him servile. On the contrary his shame for being dominated may move him to action. After all, the vicious and weak are often shamed in the hope that shame will lead them to better their conditions or themselves. Another argument Pettit may have in mind focuses not on the shame of the dominated for being dominated, but on her shame for resorting to deception for dealing with her predicament, if indeed she has resorted to that strategy. This is an interesting possibility. As everyone knows exposed wrongdoing is more likely to cause shame than secret wrongdoing. Of course as we have established the deception of the dominated may be no wrongdoing. And such justified deception should therefore cause no shame even if it is exposed. But there is the following complication: for the usual reasons the dominated will be at least tempted to resort to unjustified deception; further like most human beings they will want to believe that they behave appropriately. Consequently they will be tempted to deceive themselves into believing that their deception is justified even when it is not. If they give in to the temptation and believe that no one is watching, perhaps they can succeed in their self-deception and live contentedly. Their lives will be much harder if they become aware that their deception is public. Information that they need to forget they cannot as easily forget with the awareness that witnesses will point to it. As a result, the publicly dominated, worriedly scrutinizing her reasons for deception, and aware that the world is looking over her shoulder ready to expose attempted deception, may refrain from resorting even to justified deception—and for that she will suffer. Even when she does justifiably deceive she will worry far more than the privately dominated that she is not justifiably deceiving—and for that too she will suffer. But this argument only shows that the publicity of domination adds to the burdens of domination. It does not show why the dominated must be

32  Bernard Boxill and Jan Boxill servile. Further we may have exaggerated the burdens that the publicity of domination adds to the burdens of the dominated. Those most likely to see through her deception will be other dominated people and these will be likely to applaud her deception and to encourage her to further deception. The others, her dominators, may be only too ready to allow themselves to be deceived into believing that she regards them as demigods. These considerations intercept a final argument that can be teased out of Pettit’s remarks. In the course of making these remarks he refers to Helmer Thorvald, the husband in Ibsen’s play A Doll’s House, who Pettit says is “clearly aware of dominating Nora his wife,” but is “absolutely blind to the fact that this domination could come to seem irksome and demeaning to Nora herself ” (Pettit 1997: 60). Concluding he notes that the “lesson” in the story is that “even where domination exists and is recognized, it may not be seen for what it is when the dominated parties cannot speak for themselves” (Pettit 1997: 60). Thorvald’s imperturbable assurance that his behavior is faultless, although he knows that he dominates his wife, suggests that he believes that she is his true subordinate and that her deference to him shows that she seconds that belief. Pettit’s generalization of Thorvald’s attitude, that on the condition that “the dominated parties cannot speak for themselves,” domination even when recognized “may not be seen for what it is,” suggests that when that condition is satisfied even those aware of domination will suppose that the dominated are the dominator’s subordinates and that their deference to him shows that they accept their subordination. This generalization and the view that we gain our estimation of ourselves from the images of ourselves that we see revealed in others’ faces suggest that the dominated would conclude that they were their dominators’ true subordinates and that they accepted their subordination. Since having such beliefs constitutes servility, then indeed the dominated would be servile, and domination would indeed breed servility in the dominated. But the generalization in question is inconsistent with the claim that it supposedly generalizes. Presumably Nora cannot speak for herself, but we are led to believe that she comes to find her domination to be irksome and demeaning and consequently that she cannot believe that she is her husband’s subordinate. In other words though Nora cannot speak for herself, she still manages to see domination for what it is. Actually the claim that Nora cannot speak for herself is not strictly true, or at least is somewhat ambiguous. True, Nora cannot speak for herself in the sense that she cannot speak back to her husband and tell him to his face that his behavior is arrogant and blind. But she can and probably does speak for herself to those sharing her fate, namely the other members of the community of wives. To counter Pettit’s literary example we propose another literary example which shows a better understanding of the resources of the dominated. In her novel Adam Bede George Eliot tells of one Mrs Poyser, a tenant of and therefore dominated by the Squire Donnithorne, who silently enduring his unjustified interferences for years finally and strongly expresses her resentment of his behavior openly and to his face (Scott 1990: 6). This in itself is significant since it shows that apparent submission to domination need not imply any concession to its righteousness. But the more important point of the story for our purposes is that

Servility and Self-Respect  33 her outburst was a summation of conversations that for years she had been having with other tenants about their common condition. We are led to believe, plausibly, that these prior interactions with sympathetic others played a key role in energizing her actual outburst. Common experience tells us that people who have their feelings reflected back to them from sympathetic listeners are more likely to keep those feelings vigorous and strong. And Hume’s deployment of his theories of sympathy plausibly explains and accounts for the phenomenon (Hume 2000: 206–11). If Hume and the lesson of common experience are right, the dominated will be able to keep their resentments alive and strong if they can constitute themselves into communities of listeners. Confirming this conjecture is the fact that dominators think it important to constantly try to disrupt any community among the dominated. Slave masters bought and sold slaves and separated families partly for this reason; and they also sent both spies and the suspicion of spies into communities of slaves in order to dampen and discourage the exchange and consequent strengthening of the rebellious passions that these communities tended to nurture. Insofar as communities of dominated exist it also seems that the dominated will find faces in such communities that will provide them with images of themselves as dominated people who resent their domination and hate to be forced to pretend servility. If so the republicans are mistaken to imply that the dominated see only their dominators’ faces, and therefore, always seeing themselves as servile must come to believe that they are servile. The republicans may object that if the dominated see other faces that provide them with more flattering images of themselves they nevertheless place greatest trust in the images of themselves they see in their dominators’ faces. But this objection implies that the dominated place greatest trust in their tormentors! Does it not therefore confirm our claim that the republicans beg the question? We are not idealizing the dominated. They are often disagreeable people. Domination may not make them servile, but it does tend to corrupt them because of the pretense and deception it encourages them to engage in. The pretense alone could be harmless. Despite Plato it is not necessarily corrupting; on the contrary it could conceivably be morally educative since to pretend effectively one must learn to see things from others’ perspectives. Thus even the pretense of servility could be morally educative. But the dominated do not only pretend servility. They pretend servility in order to deceive. Therefore their domination tends to corrupt them. Someone may wonder why. We have admitted that the deception of the dominated may sometimes be justified when they resort to it to ease some of the heavier burdens of their condition. Assuming that justified actions cannot be corrupting it may therefore seem that domination need not always be corrupting even when it moves the dominated to deceive. But we believe that deception is always corrupting even when it is justified and the deceiver is neither blameworthy nor even at fault. This claim is not paradoxical. We may be fully justified in taking a drug on some occasion though we know full well that doing so could possibly lead us to become addicted to it. In the same way deception used even justifiably may pave the way for the dominated to use

34  Bernard Boxill and Jan Boxill it unjustifiably. Their pretense of justified servility will teach them to see things from others’ perspectives; this will tend to teach them how to make their deceptiveness more effective; and if their improved deceptions succeed, this will make them confident that they can successfully engage in additional deceptions. Combined with the constant temptation to deceive unjustifiably, which domination always encourages, it is no surprise that the dominated often become skilled, habitual, and unjustified deceivers. Their unjustified deceptiveness is likely to lead to further associated vices such as dishonesty, hypocrisy, insincerity, trickiness, slyness, cowardice, disloyalty, and cunning. We also concede that domination also probably directly breeds undesirable traits such as self-doubt, short-sightedness, foolhardiness, a craving for immediate gratification, and tendencies to be insecure, paranoid, superstitious, and fickle. If these traits and the vices already listed are indeed likely products of domination then we can take it as established that domination is corrupting. But it does not follow that domination is corrupting in the special sense that it breeds servility in the dominated. Dishonesty, hypocrisy, insincerity, trickiness, slyness, and cunning, are vices, but they do not add up to servility even when they are combined and compounded with short-sightedness, insecurity, foolhardiness, and paranoia. Nevertheless the whole unruly heap, conjoined with the habit of a pretended servility, must closely resemble genuine servility and invite the mistaken judgment that it is servility. So far our arguments have tried to disarm the republicans’ contention that domination breeds servility in the dominated. Sometimes we tried to explain why the republicans seem to suppose that their contention is obviously true; and sometimes we suggested that their arguments begged the question. But we may be accused of begging the question ourselves. Do we not assume that the dominated repudiate their supposed appropriate subordination to their dominators? Is not that unproved assumption the basis of our claim that the apparent servility of the dominated is only a pretense? What if any are the grounds for that assumption? Pettit thinks that the dominated parties should be allowed to speak for themselves, and with this we agree. So instead of beginning with the reports of people like Milton and Sydney we begin with the report of Frederick Douglass. Douglass gives vivid, detailed, and firsthand accounts (for he was himself a slave for twenty years) of the consequences of domination in the lives of the slaves of the American South. The consequence that he mentions most often and most insistently is fear. The slaves lived in constant fear. They feared the lash; they feared sudden and unprepared separation from their families; and they feared the constant scrutiny of their masters which they believed could succeed in penetrating to their inmost and inevitably rebellious and vengeful thoughts. And they got no respite because of good behavior since some masters maintained that slaves must be whipped to be kept good as well as to be made good, and consistently adopted a policy of using the lash in advance of bad behavior. As a consequence of living in such fear many slaves behaved in a servile manner. But Douglass never says that such slaves were ever fully persuaded that they were properly subordinated to their masters and not his moral equal. He does admit that their

Servility and Self-Respect  35 fears often distracted them from their love of liberty. And he did say that the extraordinary regimen that they endured sometimes induced him to “forget to love freedom” (Douglass 1987: 254). But being distracted from one’s love of liberty is not the same as not loving liberty. And forgetting to love freedom is not the same as permanently forgetting to love freedom. Indeed once the pressure on the slave is relaxed Douglass maintains that “dreams of freedom intrude;” the slave “wishes to become his own master,” and the “clear conception of rights rises up to life and power and leads him on” (Douglass 1987: 161, 162). We anticipate the objection that we tendentiously take Douglass’s claims to be serious attempts at telling the truth about slavery and human nature when in fact they are only rhetoric from a man who understandably wanted to defend human nature, especially the human nature of the slaves. But republican declamations about the servility of the dominated may also be rhetoric aimed at arousing detestation of domination. We therefore cannot prefer the republicans’ position over Douglass’s on the ground that his is rhetoric. However we think that there are both moral and epistemological considerations that favor at least starting off with what the dominated have to say about themselves and their situation rather than to disparage them as servile and morally defective victims who fail to understand and value their rights. Is there anything to be said for the claim of the dominated that they rarely if ever accept their subordination? Useful considerations begin with David Hume’s observations that when people interact with each other their sentiments and opinions tend to flow easily, naturally, and automatically between them. Hume postulated a theory and a mechanism that he called sympathy to account for this flow of sentiments and opinions between human minds. Very briefly according to the theory we become aware of others’ gestures, words, and the expressions on their faces, infer automatically the thoughts and feelings that they reveal, and then via sympathy come to share these thoughts and feelings (Hume 2000: 206–11). If this theory is sound, and it is plausible to some considerable extent, human beings effortlessly and irresistibly have considerable access to and familiarity with the contents of each others’ minds. As a result we all know that all of us sometimes feel love, hate, envy, jealously, and so on through the other feelings, and moreover that these feelings interact with each other in basically the same way in all of us. Further and for our purposes more importantly since human beings can make comparisons and see resemblances and differences they must therefore also be drawn irresistibly to think that their minds work in basically the same ways and contain basically the same sentiments though some of these sentiments occur more frequently in some people’s minds than in others. We emphasize that Hume’s theory suggests that this thought grows without anyone’s conscious effort. It is not as if the thought arises because people try successfully to read each others’ minds. It is rather as if, as someone aptly put it, sympathy ensures that feelings and thoughts flow among people as if by contagion. As a result the idea of his sameness with others arises in everyone’s mind.

36  Bernard Boxill and Jan Boxill So far we have suggested that all normal human beings will have an idea of their fundamental sameness. We now further suggest that all normal human beings are strongly disposed to believe that idea to be true. This suggestion stems from what they all do when the easy and effortless contagion of thoughts among them lags and some of them want to know what the others are thinking and feeling. Suppose for example that one of them tries to control his expressions, gestures, and words, in order not to share his thoughts and feelings with his fellows. The others who want to know what he is thinking will unhesitatingly start by studying his face. What sympathy accomplished easily and automatically they now self-consciously duplicate, though with deliberate and often considerable effort. They assume that he will be like them in having facial expressions that reveal his feelings and in finding it difficult to control those expressions. Macbeth was trying to look the loyal subject and kinsman of Duncan, but his wife could tell him, “Your face my thane is like a book where men may read strange matters.” Accomplished liars may learn to control their facial expressions making their faces into masks that express feelings and thoughts that they do not have or that try to express or reveal nothing. The latter effort may not be as effective as the first since everyone knows that everyone naturally makes facial expressions that reveal their thoughts, and consequently that a poker face reveals that the agent is trying not to reveal his thoughts. Sometimes as in poker this makes no difference. Often however a poker face is a reason for concern. Jefferson was allowing his distaste for blacks to compromise his good sense when he complained that the black skin on their faces was a veil that concealed their emotions. Their emotions were concealed, but not by the black skin on their faces. They were concealed because the slaves deliberately wiped off every expression of emotion from their faces in order to prevent him from knowing how much they resented him. They knew that if they allowed their faces to express their feelings that he would know how they felt. And they knew that he would know how they felt because they knew or believed that he believed that white expressions of resentment were the same as black expressions of resentment and consequently to that extent that whites and blacks were basically the same. The idea that human beings believe in the essential sameness of the contents and operation of their minds is even more compelling when we consider what we all do when we want to know what another who is not present is thinking or feeling. We cannot study his face, but we immediately try to figure out what we would think or feel if we were in his place, reasonably confident that what we think or feel will be what he is thinking and feeling; thus betraying our confidence that his mind and our minds work in basically the same way. In greater detail here is what we do: First, we try to gather together what we know of the other’s thoughts and feelings; second, we try to imagine that we have those thoughts and feelings while also trying to imagine that we do not have any of our own thoughts and feelings that we think conflict with his; third, we put ourselves in his place; fourth, we see what thoughts and feelings we come to have; fifth, we take those thoughts and feelings

Servility and Self-Respect  37 as a kind of working hypothesis of what he is thinking and feeling. This method of mind reading is not guaranteed to give correct results. What we find impressive is our confidence that it usually works and our conviction that when it fails it is because we do not have the other’s thoughts and feelings just right or because we are overlooking something about his situation which he sees and we do not see. Rarely do we ever give up on the method because we think that his mind simply does not work like ours. This second method of mind reading is probably more reliable than the first method which consists in reading faces. Jefferson may have been stymied by his slaves’ poker faces, but he had the good sense to correctly divine their rebelliousness relying on his belief that their thoughts and feelings worked just like his did, and consequently that since he would be resentful if he was enslaved, they too would be resentful since they were enslaved. So confident was he of his diagnosis of their rebelliousness that he predicted race war if they were freed and allowed to remain in their native land and consequently recommended that they be deported, and if that proved impossible, that they be kept forever enslaved. Often the two methods can be usefully combined. Douglass relates how both slave and slave master combined the two methods, both thereby conceding their convictions of their sameness. The masters he tells us, “Conscious of the injustice and wrong they are every hour perpetrating, and knowing what they themselves would do if made the victims of such wrongs,” are on the lookout for “the first signs of the dread retribution of justice” and “watch therefore with skilled and practiced eyes, and have learned to read, with great accuracy, the state of mind and heart of the slave, through his sable face.” The slave for his part is just as conscious of the vigilance of the master, and of the reasons for that confidence, and especially when he has hatched a scheme of escape, behaves with great “caution and reserve” fearing that his “sable face might prove altogether too transparent for the safe concealment” of his scheme and is no “proof against the daily, searching glances” of his master (Douglass 1987: 169). The moves and countermoves of dominators and dominated trying to read each others’ minds have often been described. What we find particularly impressive is that these moves and countermoves would hardly be attempted and would make no sense unless both dominators and dominated assume the basic similarities of their minds. It may be objected that these considerations suggest at most that dominators and dominated are aware of their basic factual sameness as human beings. But what needs to be established is that domination does not make the dominated believe that he is not the moral equal of the dominator. But we maintain that what we have suggested does help to establish what needs to be established. We are not trying to derive an “ought” from an “is.” Our point is that the powerful indeed apparently irresistible belief of human beings that they are basically the same strongly disposes them to also believe that they are morally equal.

38  Bernard Boxill and Jan Boxill John Locke made the point centuries ago. Human beings, he wrote, are in a “state of equality, wherein all the Power and Jurisdiction is reciprocal, no one having more than another: there being nothing more evident, than that Creatures of the same species and rank promiscuously born to all the same advantages of nature and the use of the same faculties, should also be equal one amongst another without Subordination or Subjection, unless the Lord and Master of them all, should by any manifest Declaration of his Will set one above another, and confer on him by an evident and clear appointment and undoubted Right to Dominion and Sovereignty” (Locke 1988: 269). We do not take Locke to be committing a Naturalistic Fallacy saying that the sameness of our faculties entails our moral equality. We take him to be saying that the sameness of our faculties disposes us to suppose that no one rightly and naturally dominates another. This reading of Locke’s passage assumes that we naturally suppose that there must be a reason for moral inequality. Therefore it is a tribute to our human nature. Although we endorse this reading of Locke’s passage and think that he is right, we admit that it faces many difficulties. We list only four of the most obvious. First, as Locke himself pointedly allows, God could subordinate the dominated to their dominators, wives to their husbands, and slaves to their masters, although the dominated and the dominators, wives and husbands, slaves and masters have the same faculties. Second, we have the same faculties, but some of these faculties may be better in some of us than in others. For example some may be more intelligent than others. Conceivably some may argue that the more intelligent rightly and naturally dominate the less intelligent. Third, we may not have the same faculties. For obvious reasons we dub this the Aristotelian objection. Fourth, Hume’s theory of the sympathetic sharing of our feelings and emotions is wrong. Some people do not share their feelings and emotions in the way he described. Moreover they do not try to read each others’ minds in the ways we described. Consequently being unaware of the sameness of their faculties they lack any strong disposition to assume their moral equality with others. For less obvious reasons perhaps we dub this the Rousseauian objection. We cannot try to respond to these serious and interesting objections to the case we have suggested for a strong human disposition to suppose that no one is naturally entitled to dominate others. Our chapter is already too long. We hope it will suffice to say that we believe that there are ways to cut them down to size. If we are right philosophers studying domination should resist the common compulsion to assume that the dominated almost immediately capitulate to the dominator. They must instead begin with the presumption that the dominated stubbornly if silently resent being dominated. We may be surprised how this presumption may change how we interpret the behavior of the dominated. Behavior that seemed servile and evidence of a failure of the dominated to understand and value their rights may turn out to be evidence that they do understand and value their rights. Appearing not to want what one does want may be a strategy to secure it. Brer Rabbit begged not to be thrown into the briar patch precisely in order to be thrown there.

Servility and Self-Respect  39

Conclusion If our suggestions are correct that there is an insistent human disposition to believe in the moral equality of all individuals, then actual cases of Hill’s kind of servility must be rare. He may protest that he never said that his account of servility was of a particularly common kind of servility, and that he would be satisfied if it picked out at least one important kind of servility. But his choice of examples, especially the examples of the Deferential Wife and the Uncle Tom, tell otherwise. They advise us to doubt that Hill did not intend to give an account of servility that would cover a large number of cases. Who are the likeliest candidates for servility in America? Why of course blacks, women, and to a lesser extent, acknowledged failures. So Hill naturally uses a black, a woman, and an acknowledged failure, as bases for the construction of his paradigm examples of servility. Naturally his readers readily agree with him that they are servile. Everyone knows that blacks and women are dominated, and most everyone is inclined to believe that dominated people tend to be or to appear to be servile. And Hill names his examples so as to plant the causes of their servility firmly in his readers’ minds; they are the Uncle Tom, the Deferential Wife, and the Self-Deprecator. But then Hill switches the reasons for thinking them servile. His readers think they are servile (at least the Uncle Tom and the Deferential Wife) because they are dominated people. He has made them think they are dominated. But then he substitutes his own reasons for their servility which have nothing to do with the reasons he has planted in their minds. He tells them that the Uncle Tom and the Deferential Wife are servile because they do not understand and value their rights. And his readers do not notice the switch because he has not explicitly told them that the Uncle Tom and the Deferential Wife are dominated. Nevertheless by disguising his account of servility as the standard kind caused by domination he has succeeded in persuading his readers that the standard, paradigm, and most common cases of servility stem from a failure of the persons involved to understand and value their rights. Of course we do not mean that Hill deliberately misled his readers. Knowing and admiring him as we do we dismiss the possibility. We think that he made mistakes because of his eagerness to give Kant a place in the analysis of the most common and vexing cases of servility. Without those mistakes his Kantian account of servility would be easily recognized as a rarity, at least if we are right that easy capitulation to domination is a rarity. Hill may object that he never says that the Deferential Wife or the Uncle Tom are dominated persons, or that they are subject to the fear that typically moves dominated persons to behave in a servile manner. Indeed he is liable to remind us that he has gone out of his way to stipulate that neither of these persons behave as they do out of any fear of the repercussions of behaving otherwise. Our answer to this objection is that such stipulations cannot erase the connotations and emotive force of the category of persons he uses to give his examples, and the names he chooses to give his examples. These must remain, and though unnoticed they still move his readers to agree that his

40  Bernard Boxill and Jan Boxill suggestions of the servility of his arguments are plausible when in fact they are not. Put the point in the form of a question: Why, though Hill erases all fear of repercussions for self-respecting behavior from the Deferential Wife and the Uncle Tom, does he still persist in presenting them as the Deferential Wife and the Uncle Tom—unless he wants to retain somewhere in his readers’ minds the idea that those individuals are dominated and consequently may behave as they do from fear? His persistence in using those emotive names while striving to erase all traces of fear from their motivation strikes us as an effort to have his cake and eat it too; namely, to keep the thought that his characters are motivated by fear in the back of his readers’ minds while telling them that they are not. By thus compelling his readers to draw the conclusion he wants them to draw while simultaneously denying them the reasons why they drew that conclusion, he further compels them, in order to save their own rationality, to accept his suggestion of the reasons for drawing the conclusion. We reject the reasons that Hill gives to account for the moral defectiveness of the servility of the dominated. He says that their servility is morally defective because they fail to understand and value their rights. We say that their servility is morally defective because of the long list of vices and weaknesses we cited as the results of the dishonesty that domination encourages or compels the dominated to engage in. That list we hope is more substantial and interesting than the abstract and perhaps almost unintelligible failure “to understand and value one’s rights.” Note that Hill never hints at what might be done to correct that failure. He tells us that people can be instructed in the nature and value of their rights but warns that they may persist in their servility even after being thus instructed. He describes and indeed names his two characters—the Deferential Wife and the Uncle Tom—so as to draw his readers irresistibly to conclude that domination has something to do with the failure of those characters to understand and value their rights; but he also warns his readers that he makes no speculations about the preconditions of his characters’ failure to understand and value their rights, leaving one to wonder why he so carefully leads his readers to assume that his characters are dominated. We submit that our account of the moral defectiveness of the kind of servility we wish to emphasize not only deals with the most widespread cases of servility and pretenses of servility, but also gives us some guidance about how to correct it. Right off it suggests that the dominated suffer from no deep mysterious moral flaw like failing to understand and value their rights, and do not need to be instructed to value their rights. As Douglass insisted during one of the darkest periods of slavery, “The Negro is all right.” The main thing the Negro—and the woman—need is to be relieved of the burdens of their domination.

References Douglass, Frederick. 1987. My Bondage and My Freedom, ed. William L. Andrews. Urbana and Chicago: University of Illinois Press. Hill, Thomas E., Jr. 1995. “Servility and Self-Respect,” in Robin Dillon, ed., Dignity, Character, and Self-Respect. New York: Routledge.

Servility and Self-Respect  41 Hume, David. 2000. A Treatise of Human Nature, ed. David Fate Norton and Mary J. Norton. Oxford: Oxford University Press. Locke, John. 1988. Two Treatises of Government, ed. Peter Laslett. Cambridge:  Cambridge University Press. Pettit, Philip. 1997. Republicanism. Oxford: Oxford University Press. Rawls, John. 1971. A Theory of Justice. Cambridge, MA: Harvard University Press. Scott, James C. 1990. Domination and the Arts of Resistance. New Haven and London:  Yale University Press. Skinner, Quentin. 2008. “Freedom as the Absence of Arbitrary Power,” c­ hapter 4 in Cecile Laborde and John Maynor, eds., Republicanism and Political Theory. Oxford: Blackwell Publishing, pp. 83–101. Smith, Adam. 1982. The Theory of Moral Sentiments. Indianapolis: Liberty Classics.

2 Humility, Arrogance, and Self-Respect in Kant and Hill Robin S. Dillon

Although Kant’s ethics is widely regarded as chiefly concerned with issues of moral duty and constraints on action, one of the distinctive aspects of Tom Hill’s work in Kantian ethics is his focus on questions of character and moral attitudes. Hill’s essays often treat both specific issues, such as the justifiability of suicide or the moral responsibilities of bystanders, and more theoretical issues, such as the justification of moral judgments and standards, through consideration of the kinds of objectionable attitudes and defects of character that might give rise to wrongdoing or failures of duty, as well as the virtues and appropriate attitudes towards the self and others that would characterize a morally ideal person or reflect the deepest rational commitments of moral agents. Indeed, Hill’s approach to Kant’s ethics is to treat it as “an ethics of attitude,” for he takes one of its central ideas to be that “to be a moral agent is to be a person (with a ‘will’) committed to regarding oneself and others in certain ways” (Hill 2000a: 21). To be a morally good person is thus not merely to have a good will but also to have and be strongly motivated by the proper kinds of attitudinal commitments (Hill 2008). For Hill, as for Kant, the attitudes and virtues of self-respect and respect for other persons are morally central and enormously powerful both theoretically and motivationally. Thus much of Hill’s work and Kant’s later ethical writings are concerned to a great extent to explicate the nature, value, and moral implications of self-respect and respect for others. My aim in this chapter is to contribute to our understanding of self-respect by focusing on some questions about its possible relation to other stances toward the self. In particular, having looked at connections between self-respect and arrogance, I want now to consider a quality typically regarded as the opposite of arrogance: humility. Arrogance is often thought to involve disrespect for other people; I have argued elsewhere that it is also and in the first place a failure to respect oneself (Dillon 2003, 2007). If this is correct, then how, if at all, is humility related to self-respect (and respect for others)? Just as arrogance is usually treated as a serious vice, so humility is widely regarded as an important virtue. Indeed, it is supposed to be the virtue that opposes

Humility, Arrogance, and Self-Respect  43 arrogance, keeping it in check or preventing it from developing in the first place. But while it is obvious why disrespectful arrogance is a vice, humility is not unproblematic from a Kantian perspective. For humility, which is often thought to involve having a low opinion of oneself or valuing oneself less than others, seems to be in tension with, if not opposed to, the respect one is morally required to have for oneself as a person equal in dignity and moral standing with all other persons. So how are we to understand humility? Is it indeed a virtue? That is, is it a quality that we should admire from a moral point of view, encourage in other people, and develop in ourselves? Or is it instead a characteristic that might be useful at times but is of no real moral value? Or, worse, is it, like arrogance, morally objectionable because it is incompatible with respect for ourselves as free and rational persons?1 If it is a virtue or at least a morally good trait, might at least a part of its value derive from connections to self-respect? Now, while Kant and Hill have much to say about self-respect, they have much less to say about humility, so it might seem odd to look to these philosophers for help in understanding it or its possible relevance to self-respect. Nevertheless, what they do say helps to shed light on both humility and arrogance as well as on a certain kind of self-respect. What I conclude with their help is, first, that humility is not the virtue opposing arrogance; rather, self-respect is. Second, humility is at best an ancillary and instrumental virtue and the servant of self-respect; but at worst, it is as serious a vice as arrogance, indeed, an aspect of it.

1 Let me say a bit more in a preliminary way about arrogance and humility. Arrogance is widely regarded today as a serious moral vice and a cause of social, economic, political, and environmental ills.2 It has long been regarded as such. The ancient Greeks called it hubris and regarded it as the worst of evils: a civic crime that threatens social order and leads to political disaster, and a vice that undermines good judgment, destroys virtue, and brings ruin for self and others (Fisher 1992). Hubris was the bane of wax-winged   In posing these questions in this way, I am adapting Hill’s queries regarding conscience (Hill 2002: 234).   To cite just a few of the recent manifestations of this view: In their last debate before the May 6, 2012 election in France, the most cutting criticisms Nicholas Sarkozy could make of rival François Hollande was to call him arrogant (New York Times, May 12, 2012, and ), while Sarkozy’s own arrogance was cited as a cause of his downfall ; Evangelos Venizelos of Greece’s socialist Pasok party blamed “arrogance, petty party politics and opportunism” for the collapse of efforts to form a coalition government after the recent elections (The Guardian, May 15, 2012  ); and arrogance has been identified as a chief cause of large-scale corporate debacles in the U.S. like the 20BP oil spill in the Gulf of Mexico last year (Joseph Romm, “The Three Causes of BP’s Oil Disaster,” ) and the 2008 collapse Wall street collapse (“Lehman Docs Show Wall Street Arrogance Led to Financial Collapse,” Huffington Post ). 1

2

44  Robin S. Dillon Icarus, the real Achilles’ heel of both that greatest of warriors and his internecine rival Agamemnon, and the doom of the Persian empire. Contemporary discussions cite it as a chief character flaw of politicians and business leaders3 and a cause of calamities such as the 2009 collapse of Wall Street,4 the 2011 Japanese nuclear catastrophe,5 and failures in the U.S. wars in Iraq and Afghanistan.6 The medievals called it superbia, the sin of pride, and ranked it as the deadliest of the Seven Deadly Sins: deadly not only because it is the source of all other sins and wrongdoings, but also because it corrupts and destroys the very constitution that makes us what, most fundamentally, we are.7 Arrogant pride was the downfall of Milton’s Lucifer and the first sin of Adam and Eve (Augustine 1996: 14.13, 335); less dramatically, it is the central flaw distorting the characters and judgment of the principals in Austen’s Pride and Prejudice. Hill and Kant agree with this general assessment of arrogance. Although it does not figure significantly in his writings, Hill does call it a vice and notes the incompatibility of arrogance both with respect for the moral law and the respectful recognition of the humanity of each person as an end in itself; he also identifies arrogance as an obstacle to respectful relations with people from cultures quite different from one’s own (Hill 1991a: 4; 2002: 252; 2000b: 84). The Arroganz that Kant also calls “Hochmut,” “Übermut,” and “Eigendunkel” (self-conceit) is the most significant of the vices in his ethics, for it is the severest violation of the core moral duties to respect the moral law, other persons, and ourselves, and the chief impediment in moral life, inasmuch as it corrupts our abilities to value, to make rational judgments, and to act autonomously, abilities that are constitutive of personhood. Arrogance, that is, corrupts the person-defining powers of rational moral agency (Dillon 2007). Humility is typically treated in both everyday and philosophical discussions as the opposite of arrogance and the virtue that prevents or combats it. Regarded as the proper orientation to self, God, and other persons that makes right living and other virtues possible, it is a preeminent virtue in the Deadly Sins tradition and in contemporary Christianity, Islam, and Judaism, according to which only the humble individual can live in right relation to God and one’s fellow human beings. Moses, Christ, the Prophet Muhammad, as well as the Buddha and Confucius—all are regarded as paragons of humility whose model we are exhorted to follow. But even in non-religious

3   E.g. Burbach and Tarbell 2004; Kershaw 1993; Heidi Moore, “Facebook’s IPO Debacle: Greed, Hubris, Incompetence,” The Guardian, May 23, 2012  ; Susan Brooks Thistlethwaite, “JPMorgan and the problem of hubris” . 4   E.g. Cohan (2009). 5   E.g. Christian Parenti, “Nuclear Hubris: Could Japan’s Disaster Happen Here?” The Nation, March 13, 2011 . 6   E.g. Isikoff and Corn (2006); Perry White, “Afghanistan: American’s hubris and folly,” . 7   The two clearest expressions of this view are in Gregory the Great (1850) and Thomas Aquinas (1921: I-II, q. 84). I follow Taylor (1994: 145) in treating the deadliness of the Deadly Sins as constitutional corruption.

Humility, Arrogance, and Self-Respect  45 contexts humility is widely prescribed as the cure for personal, economic, political, and environmental arrogance. Hill doesn’t go so far as to call humility a preeminent virtue, but he does regard it both as what is needed to think morally well and engage with others on terms of mutual respect rather than arrogance and as a morally important and admirable character trait, one of the human excellences that, in particular, we who face environmental problems would do well to cultivate (Hill 2000b, 1991c). Interestingly, however, Kant bucks the trend with regard to the moral value of humility. For although he says some positive things about humility, he writes neither as much nor as strongly as one might expect about the reputed principal foe of the principal moral failing. Nor does he regard it as a virtue, although he does give it a role to play in the moral life.8 Most of what Kant has to say about humility concerns a morally pernicious form of it: servility. I admit to finding the characterization of humility as a great virtue or admirable human excellence puzzling. For what the OED defines as a matter of having “a lowly opinion of oneself ” or “a low estimate of one’s importance, worthiness, or merits”— highlighting its etymological roots (from L. humilis: low, lowly; from humus: the ground)—seems odd company for traditional virtues like wisdom, courage, benevolence, and justice: not noble or inspiring, too meek and self-effacing to be a strength of character. It is odder still that the cognates of an esteemed trait—“humble,” “humiliate,” “humiliation”—refer to states and activities that are far from admirable: lowliness, submissiveness, degradation of position or value, abasement. Humility is certainly not a trait that Aristotle could call a virtue. Nor is it a virtue from a Humean perspective. Hume regards it as a “monkish” virtue (Hume 1972: 270), since it is neither useful to oneself (a common view is that what is needed to get ahead in business, politics, sports, and other competitive arenas is not humility but arrogance, cockiness, aggressive self-confidence) nor pleasurable to oneself (being pleased about one’s humility is a sure sign that one lacks it), nor pleasurable to others (at least, it is perverse to delight in another’s lowly self-estimate), nor useful to others (indeed, to laud humility for its usefulness to others borders on sinister, given the long history of casting it as a virtue of subordinated peoples—how much easier to dominate those who believe that submissiveness makes them good). Moreover, its reputation as the foe of arrogance is puzzling, since humility doesn’t preclude arrogance. It makes perfect sense to speak of arrogant humility, as Proust does in mocking the Princesse de Parme (Proust 1934: 1023).9 Perhaps my difficulty in seeing it as a virtue arises from not really understanding what it is, dictionary definitions notwithstanding. But if so, I seem to be in good company, for there is significant disagreement about how to define it. Putative definitions 8   Here I disagree with Grenberg (2005), who argues that humility is a core virtue for Kant. I develop this disagreement in sections 3 and 4. 9   The Princesse, Proust writes, had had “instilled in her from her earliest childhood the arrogantly humble precepts [precepts orgueilleusement humbles] of an evangelical snobbery: Remember that if God has caused you to be born on the steps of a throne you ought not make that a reason for looking down upon

46  Robin S. Dillon seem to be of two basic types. In the Deadly Sins tradition and contemporary religious perspectives, humility is identified as the absolute submission to God that requires viewing oneself as nothing or worthless, as annihilation of or obliviousness to or unconcern for the self, as restraining one’s ambition and desire for excellence, and as valuing others humans higher than oneself.10 This view, however, strikes many contemporary non-religious thinkers as objectionable: self-abnegation cannot be what makes humility a virtue. And, as we’ll see, Kant would agree, since (a) it is false that the self is nothing or worthless, (b) concern for self-worth is of the greatest moral importance, (c) self-abjection or annihilation is the essence of the vices opposed to the fundamental moral duty to respect oneself as a being with dignity, (d) all persons are equal in dignity so that none should be valued more highly than others, and (e) we have a moral duty to make our own moral perfection our end and so strive for moral excellence. Since contemporary theorists of humility accept (usually without question) that it is a virtue, they define it in ways that omit disrespectful self-abnegation as, for example, underestimating or not overestimating or having an accurate sense of one’s abilities and achievements,11 being unconcerned or unimpressed with them,12 regarding self-worth as unimportant,13 sharing credit for one’s accomplishments,14 having a realistic view of one’s limitations,15 being unconcerned with others’ opinions of one’s worth,16 and self-transcendence.17 But these definitions are also problematic. Hill’s discussion of humility as a human excellence is of the second type and illustrates the main problems with this approach. In “Ideals of Human Excellence and Preserving Natural Environments” Hill (1991c) argues that at least part of what makes destruction of the environment morally objectionable is that the kind of person who would do things such as replace his lovely garden with asphalt, “strip mine a wooded mountain, or level an irreplaceable redwood grove” (108–9) exhibits vices or defective attitudes that hinder the development of certain virtues, most prominently the virtue Hill calls “proper humility.” While Hill does not claim to be defining or fully explicating proper humility, he does identify two aspects of it: a proper sense of one’s relative importance and self-acceptance. The first aspect refers to a more-than-intellectual appreciation of one’s place among persons and in the natural order that requires regarding things other than oneself as valuable or important in their own right. The major obstacle to this aspect of humility is self-importance, which is “a tendency to measure the significance of everything by its relation to oneself and those with whom one identifies” (113). The second aspect, self-acceptance, itself “long those to whom Divine Providence has willed . . . that you must be superior by birth and fortune . . . Therefore never seem, in your speech, to be recalling these great privileges . . . because it is useless to point out that you are better than other people . . . since everyone knows these facts already.” 10   See, for example, Aquinas (1921: II-II); Benedict (1875); Bernard (1929); Eckhart (1981); Weil (2002); Carlson (1944); Warren (2002). 11   See, for example, Driver (2001); Flanagan (1990); Richards (1992); Peterson and Seligman (2004). 12 13 14 15   Garcia (2006).   Roberts (2009).   Nuyen (1998).   Snow (1995). 16 17   Roberts and Wood (2003).   Taylor (2006); Peterson and Seligman (2004).

Humility, Arrogance, and Self-Respect  47 considered to be a human excellence,” involves “acknowledging, in a more than merely intellectual way, that we are the sort of creatures that we are,” which is a matter of “understanding, facing squarely, and responding appropriately to who and what one is, i.e., one’s powers and limitations, one’s affinities with other beings and differences from them, one’s unalterable nature and one’s freedom to change.” The denial or disowning of such aspects, the pretense that one is something other than one is, is “incompatible with proper humility, because it is seeing oneself as better than one is” (114). While I  would agree that self-importance, self-denial, and pretense are morally problematic and incompatible with a proper attitude toward the self, I am puzzled about why the attitudes Hill describes should count as humility rather than something else. Regarding other persons as having value in themselves is, a Kantian would say, valuing them appropriately as ends in themselves, i.e., respecting them as persons, while appreciating one’s real relative value and place as a person among equal persons is respecting oneself. Many theorists also treat regarding the natural environment as valuable in itself as a kind of respect. These attitudes of respect are not the same as, nor do they require, having a lowly opinion of one’s relative worth or being submissive to other persons or things. Likewise, it is odd to regard not pretending that one is other than one is and understanding and responding appropriately to one’s powers, freedom, and similarities or differences vis-à-vis others as aspects of humility rather than, say, wisdom, honesty, or integrity. Where is the lowliness? This same problem of identifying as humility what is really something else also infects other contemporary accounts. For example, why call having a realistic view of one’s good qualities and accomplishments or sharing credit for them with others humility rather than honesty? When humility is lauded as morally admirable or a virtue, I often find that I can’t see any relation between what is lauded and the lowly estimate of oneself that every dictionary I consult ascribes to humility; it seems to me that what is being lauded is something else. But when humility is so defined, then I find it difficult to see much moral merit in it. Having an accurate sense of one’s flaws, limitations, or dependence could be understood in terms of a low self-estimate, but why think it has moral value over and above the moral value of honesty? (Interestingly, many contemporary accounts that include such an aspect hasten to add that the humble self-appraisal includes one’s merits as well as flaws; but then we’re back to the first problem.) Why think that a standing attitude toward the self that regards one’s merits or worth as unimpressive or unimportant or that views the self as worthless or less valuable than others or as something to be ignored or transcended is anything virtuous, anything we should, morally, develop on ourselves or encourage or seek to instill in others?18 I suspect that one source of the mistake in these accounts is the common view that humility is the, i.e., the only, opposite of arrogance. For example, if self-importance or self-enhancing self-denial is a form of arrogance (I think both are), then one might 18   Of course, in certain religious contexts, forms of self-abnegation may be quite appropriate—compared to God or absent one’s absolute dependence on God, one really is (next to) nothing. But contemporary

48  Robin S. Dillon think that its absence must be humility, or at least, that the prevention or cure for it must be humility. The tendency, I suspect, is to ignore the core sense of humility and to give the concept whatever content makes humility both a virtue and the opposite of arrogance however the latter is being understood: the absence of that inappropriate attitude toward the self just is what humility is. But this move ignores the possibility that the appropriate and non-arrogant stance toward the self or the inhibitor or cure for arrogant self-importance or self-enhancing self-denial is something else, such as self-respect or respect for others, and that developing a certain kind of humility is but a means (perhaps only one way for some but not all people in some but not all circumstances) to be self-respecting. That, at any rate, is what Kant held. Or so I’ll argue.

2 Kant has relatively little to say about “proper humility,” but he does discuss some improper forms of it. Chief among these is servility, something that Hill has also famously and very helpfully discussed. In The Metaphysics of Morals Kant identifies servility, which he calls false or lying humility, as one of the vices opposed to the duties all persons have to themselves, contrasting it with both arrogance and self-respect (MS 6: 434–7). His discussion begins by emphasizing the dignity that persons have simply as persons, i.e., as rationally autonomous moral agents—beings with the capacity to choose freely to act on purely rational motives. Each person has dignity and so can “exact respect for himself from all other rational beings in the world” (MS 6: 435). Each of us also has a duty to respect ourselves, i.e., a duty to not disavow our dignity but to act always with consciousness of our supreme worth and status as equal persons. Servility is deliberate self-abasement: “the disavowal of all claim to any moral worth in oneself ” that is “contrary to one’s duty to oneself since it degrades one’s personality” (MS 6: 36). Hill helps us understand the nature and moral badness of servility through three now-familiar examples: an Uncle Tom, an extremely deferential black who regards himself as having a lower status than whites and treats his own values, aspirations, and interests as less important than those of whites; an overly deferential wife who subordinates herself to her husband, valuing him and his needs, desires, interests, and ideals far above, even instead of, herself; and an extremely self-deprecating person, who makes no demands on others and allows himself to be ill-treated or used by them because he thinks his shameful inadequacies and failures make him unworthy of decent treatment (Hill 1991a). Hill explains that what makes their servility morally objectionable is their failure to appreciate their moral status and their equal basic rights as human beings. Their actions and attitudes convey the idea that they are accounts typically steer clear of theistically grounded evaluations of humility. To be sure, various of those accounts do seek to explain in other ways why these “humble” attitudes are virtuous, and in another context I would argue that the explanations are unpersuasive.

Humility, Arrogance, and Self-Respect  49 lesser sorts of beings—less important, less worthy of decent treatment, having fewer or lesser rights than other people. That is, they degrade themselves, ranking themselves lower in a moral hierarchy than whites or husbands or more successful or morally better people. Hill identifies two possible reasons for this self-degrading stance. Servile individuals might let others violate even their morally inviolable rights or not exercise basic rights that they ought to exercise either because they don’t understand that they are the moral equals of other persons with the same basic rights that every person has just because they are persons or because they don’t value their moral status and rights as they should. But in both cases, servile individuals, Hill holds, lack a certain kind of self-respect, one that involves proper regard for the basic rights one has as just a person. This kind of self-respect is what I call “interpersonal recognition self-respect,” the reflexive form of the respect that each person is owed and has a right to claim from every person.19 That respect, interpersonal recognition respect, is the practical acknowledgement of the moral worth and status that all and only persons have, the basic rights that come with that status, and what Darwall (2006) calls the “second-personal authority” of each person to exact from all persons respect as equals in moral worth and status. Interpersonal recognition respect is a categorical moral duty: one formulation of the Categorical Imperative, the supreme principle of morality, declares that our fundamental moral obligation is to respect persons.20 As rationally autonomous moral agents, persons are ends in themselves with the “absolute inner worth” that Kant calls “dignity” [Würde] (e.g. MS 6: 435). Dignity is not the kind of worth that others can bestow or strip one of or that one has to earn or could forfeit or even waive or give up. Persons have it unconditionally, regardless of personal qualities, social status, and accomplishments and failures, including our individual records as a moral agent. Dignity is intrinsic worth: we all have dignity simply in virtue of having been born with the rational capacities that make us persons.21 It is also supreme worth in the sense that it trumps all other values in considerations of how to act. Thus, no matter what else might be gained by doing so, the moral law categorically forbids treating persons as if they were things, i.e., treating beings of unconditional intrinsic worth as if they had value only relative to someone’s feelings or desires, treating ends in themselves as if they were at best merely means to be used to get something else that someone happens to want or need. Persons are all equal in dignity, and this gives each person both the moral status of an equal person among persons and an equal set of basic rights, at least some of which are inviolable and inalienable. The dignity and basic rights of each person must always be acknowledged by every person, in   I borrow the term “recognition respect” from Darwall (1977).   I follow Wood (1999) in reading the second formulation of the Categorical Imperative—“Act in such a way that you treat humanity, whether in your own person or the person of any other, never simply as a means but always at the same time as an end”—as saying that our fundamental moral obligation is to respect persons. 21   For rather different accounts of Kantian dignity, see Sensen (2009) and Hay (2013). 19

20

50  Robin S. Dillon attitude as well as action. Interpersonal recognition respect is the practical acknowledgment of the dignity of persons and the moral implications of their dignity. More specifically, interpersonal recognition respect for a person is, as Hill puts it, the “full recognition” of someone as a person with the same basic moral worth as any other and the status of “co-membership in the community whose members share the authority to determine how things ought to be” (Hill 2000b: 63). Recognition respect is more than a mere intellectual grasping of a fact about someone; it involves acknowledging and valuing her as a being with dignity, appreciating her basic rights as well as other moral constraints on all moral agents to which the dignity of persons give rise, having the attitudes that such appreciation involves, and acting in regard to her only in morally appropriate ways out of this valuing of her as a person. Interpersonal recognition self-respect, which Kant sometimes calls “real self-esteem [Selbstschätzung]” and “pride [Stolz] in the dignity of humanity in one’s own person” (MS 6: 459), involves understanding and valuing oneself as a being with dignity and an equal person among persons with an absolute right to—and the moral authority to demand from all others—the same practical acknowledgment of one’s moral status and worth that one owes all others, as well as a categorical duty to acknowledge one’s own dignity in actions and attitudes, and living in light of this self-understanding and self-valuing. Individuals with interpersonal recognition self-respect regard certain forms of attitude and treatment from others as their due as a person and other forms as degrading and beneath the dignity of persons; and, other things equal, they are not willing to be regarded or treated by others in ways that mark them as less than a person. They claim the right to be shown “respect for a human being as a moral being” out of “a concern to yield nothing of one’s human dignity in comparison to another” (MS 6: 465). It is this complex and powerful, indeed, life-shaping, standing attitude of moral appreciation of what one is and of the implications of one’s moral nature that the servile individual lacks. The low estimate of self-worth and status that servility involves makes it clear that it is a kind of humility. Of the two forms of servility that Hill identifies, one that involves misunderstanding one’s moral status, rights, and worth and another that involves not caring about them, Kant is more concerned with the latter. Both are morally objectionable, but whereas the former could be merely a lack of virtue for which the individual may not be to blame, the servility that involves the deliberate discounting of one’s moral status and basic rights is a vice. Two features make the latter form of servility deeply bad. First, it is “false humility” (MS 6: 436), in that its invitation to others to regard one as a being of a lesser sort reflects a false view not of one’s own worth and status but of the worth and status of all persons. In its denial of one’s status as an member of the moral community with the same dignity and basic rights as every other member, servility conveys the view that the moral community is not a relation of equals but a hierarchy of two moral castes, one composed of beings with higher fundamental worth and the other, to which one belongs, of beings with lower fundamental worth who are entitled to much less in the way of consideration and respect.

Humility, Arrogance, and Self-Respect  51 Second, servility is not just false but also “lying humility.” That is, as Kant makes clear, servility is not simply a matter of the possibly blameless misunderstanding of one’s basic rights and status as a person that might, for example, plague someone who was raised in a racist or sexist society to believe she was a lesser sort of being than others.22 It is, rather, motivated self-abasement: the deliberate and so culpable devaluing of oneself. Servility lies about one’s worth, status, and rights; and the lies are motivated, as lies typically are, by desires for something else. The servile person disavows his true moral worth not because he doesn’t care at all about worth, but because he wants to be valued in some other way. He trades in his dignity in order to “acquire a borrowed worth” or “as a means to acquiring the favor of another” (MS 6: 435–6). Rather than claiming from others the respect as an equal that is his due as a person, or even demanding that others value him more highly than they value themselves, the servile individual seeks to “borrow” worth from the conditional value that others ascribe to him for his submissiveness, usefulness, flattering dependence, self-castigation for flaws and failures that both makes others look good by comparison and seems to show something like honesty about oneself, and so on. Perhaps he thinks trading in his basic rights is an easier path to self-valuing, perhaps he is expressing feelings of inadequacy or inferiority, perhaps pretending to a lesser status is a calculated ploy to gain some advantage over others. But whatever the specifics, the servile person makes himself “a plaything of the mere inclinations and hence a thing” (MS 6: 420). What, at bottom, makes servility wrong, Kant explains, is that it is a violation of the “prohibition against depriving [oneself] of the prerogative of a moral being, that of acting in accordance with principles, that is, inner freedom” through the willingness to let one’s choices be determined by the inclinations rather than by reason (MS 6: 420). Servility degrades that which, on Kant’s view, is most truly one’s self: the rationality that makes one a person. Clearly, this kind of humility is a deeply serious moral vice. In explaining how servility violates the duty to accord oneself (interpersonal recognition) respect, Kant also contrasts it with both arrogance and “true humility.” I’ll focus on arrogance now and return to the latter later in the chapter. Let me first distinguish two kinds of arrogance.23 What I call “interpersonal arrogance” involves an inordinate sense of superiority of worth and status that is marked by disdain, contempt, or disregard for those one regards as inferiors. What I call “unwarranted claims arrogance” involves a sense of entitlement and a disposition to arrogate, i.e., to lay claim without warrant to, things associated with high worth and status, such as rights, authority, status, knowledge, ability, etc. While they are different traits,   Hill stresses that not every servile person is to be blamed for misvaluing their rights, and two of his examples—the Deferential Wife and the Uncle Tom—are good examples of people whose self-conception and abilities to value themselves have been distorted by oppression. However, since Kant thinks that persons are always aware of their dignity and status as persons, even if some individuals sometimes ignore it or act as if they weren’t aware, it is not unreasonable to think that for Kant all servility involves lying, to others and perhaps to oneself. 23   I’ve explicated this distinction in more detail in Dillon (2003, 2007, 2012). 22

52  Robin S. Dillon comprising different beliefs, attitudes, objects, and behaviors, the core of each is an orientation to one’s self-worth that is fundamentally flawed: they either express inordinate self-valuing or seek inordinate worth for self.24 In contrast with servility, which waives all claims to moral worth in violation of the moral law and invites others to deny one the respect one is owed, Kant describes arrogance as an unjustified belief in one’s superior worth (MS 6: 436) that “demands from others a respect it denies them” (MS 6: 465). As he explains, “arrogance [Hochmut] (superbia and, as the word expresses it, the inclination to be always on top) is a kind of ambition [Ehrbegierde] (ambitio) in which we demand that others think little of themselves in comparison with us . . . Arrogance is, as it were, a solicitation on the part of one seeking honor for followers, whom he thinks he is entitled to treat with contempt” (MS 6: 465). This presumptuous “lack of modesty in one’s claims to be respected by others,” which Kant also calls “self-conceit [Eigendünkel] (arrogantia)” (MS 6: 462), is “a vice opposed to the respect that every human being can lawfully claim” (MS 6: 465). Just as servility disrespects oneself by denying one’s own dignity and moral status, so interpersonal arrogance disrespects other persons by denying that they have dignity and the status of equality with oneself. It involves viewing other persons, and demanding that they view themselves, as beings who belong to a lower order, who have less worth than oneself, fewer or no basic rights, and no second-personal authority to make claims of respect on oneself. Now, since servility and arrogance seem like opposites, it might also seem that as servility is the vice of deficiency with regard to which a certain kind of self-respect is the virtue, so interpersonal arrogance is the vice of excess. But this is mistaken.25 Arrogance is not too much of a good thing, but rather its complete absence. For while the arrogant individual does makes an unwarranted claim to superior worth, it isn’t dignity he claims. The worth he seeks is essentially comparative and competitive, not incomparable and unconditional. Interpersonal arrogance, Kant says, is the inclination to think highly of oneself, but it asks “not what one is worth, but how much more one is worth than another”; the arrogant person “already believes in his own worth, but he esteems it solely by the lesser status of other people” (L 27: 41). The arrogant person falsely believes that the only worth all persons have is comparative and thus regards neither himself nor others as beings with dignity unconditionally deserving respect. Indeed, Kant says, the arrogant person is always “mean [niederträchtig] in the depths of his soul. For he would not demand that others think little of themselves in comparison with him unless he knew that, were his fortune suddenly to change, he himself would not find it hard to grovel and to waive any claim to respect from others” (MS 6: 466).

24   By “inordinate” I mean not merely “too high” but disordered, inappropriate in any number of ways. It is the seeking as well as the worth sought that is inordinate. 25   When writing about self-respect many years ago and being too much under the influence of Aristotle, I made just this mistake. I have come to think, however, that the continuum model is not a good one for understanding any of these traits.

Humility, Arrogance, and Self-Respect  53 Here, things get rather interesting. For the “meanness” that is prepared to waive all rights to respect is the same self-abasement that characterizes servility. The arrogant, that is, are servile at heart. What is more, just as pernicious servility is not simply a matter of misunderstanding moral worth and status but is motivated by a desire to value the self, so interpersonal arrogance springs from a desire for self-value. Indeed, they both arise out of what contemporary psychologists call the “self-esteem motive,” the powerful desire to value oneself ever more highly, especially through being highly valued by others. The self-esteem that is sought is “thinking well of oneself ” or “feeling good about oneself,” having a positive, high, and self-approving attitude about oneself.26 It differs from interpersonal recognition self-respect. For whereas self-respect involves the well-grounded consciousness of one’s intrinsic dignity as a person, selfesteem need not be grounded in anything about oneself that is morally significant, and it typically derives from external sources such as how one is regarded by others. Such self-esteem involves, as Kant says, the “delusion” that “consists in valuing the mere opinions of others regarding the worth of things as equal to their real worth” (A 7: 270). Both servility and arrogance essentially involve the comparative valuing of persons: one has either more or less worth, a higher or lower status, more or fewer rights than others. They thus also share the same hierarchical view of the moral community, differing only in where they locate themselves (at least for now). Both are false and lying stances toward the self; the view of relative moral status and the comparative valuing of persons that they share lies about the incomparable, unconditional, and equal dignity of all persons. They both make one “a plaything of the mere inclinations and hence a thing.” However, Kant makes it clear that servility is not the only false form of humility. All humility that involves comparing oneself to other human beings and valuing oneself less is false (MS 6: 435, L 27: 349), since the only true interpersonal worth is non-comparative dignity. It is important to recognize that Kant’s objection applies to 26   I use the term here as it is used by contemporary American psychologists and social psychologists. See e.g. Coopersmith (1967), Rosenberg (1965), Baumeister (1994), Mruk (2006), where self-esteem is defined as “a positive attitude toward the self,” “feeling good about oneself,” or “self-approval,” which may or may not (depending on the theorist) be based on one’s assessment of one’s abilities and achievements, but which is always socially comparative in that it (a) partially consists in reflected appraisals of others or involves one’s view of one’s acceptability to others, (b) involves self-assessment by social standards, and/or (c) is typically caused and sustained by acceptance and praise from significant others. Most researchers accept it as an axiom that the desire to maintain and enhance self-esteem is an enormously powerful, ubiquitous, even universal motivation (albeit one whose operation is shaped by social and cultural context); some consider self-esteem to be the master motive in personal and interpersonal relations. Mruk (2006) contains a comprehensive review of the relevant psychological literature, as do the essays in Kernis (2006). On the self-esteem motive, see especially Baumeister (1994); Gecas (2001) (and other essays in that volume); Kaplan (1975). Let me note that Kant uses the terms “self-esteem” [Selbstschätzung], “self-respect” [Achtung für sich selbst] and “reverentia” interchangeably to refer to interpersonal recognition self-respect (consciousness of one’s dignity as a person); indeed, he uses the first far more often than the second term in this connection. But he also uses the term “self-esteem” [Selbstschätzung] to refer to what I have elsewhere called “evaluative self-respect” (appreciation of one’s moral merit or “inner moral worth”) and to refer to self-valuation that is not grounded in one’s dignity as a person. I use the term exclusively for the latter.

54  Robin S. Dillon comparisons of not just the kind of worth related to moral status and rights, but any worth, including the value of one’s moral merits, accomplishments, personal qualities, social position, likeability, etc. Anyone who entertains a low opinion of his worth in comparison to others does not display a virtue, but only something “unnatural,” a mere analogue of the real thing that Kant, echoing Hume, calls a “monkish virtue” (L 27: 349). Thus, any account of virtuous humility that identifies it as involving beliefs about one’s worth, significance, or importance relative to other humans (or things) misidentifies it. Whatever “true humility” is, it involves no interpersonal comparison whatsoever. Not only is interpersonal humility false, but Kant warns that it is also morally dangerous because it inevitably becomes competitive. For “when a human being values his own worth according to others, he seeks either to raise himself above others or to diminish the value of others” (P 9: 491). Interpersonal humility is just another arena for competition, as one tries to “equal or surpass others in . . . [being humble], believing that in this way one will gain even greater inner worth” (MS 6: 435). Interpersonal comparison always gives rise to competition, and competitors always seek to win, to come out on top: ambitio et superbia. Even in the humility arena, as it turns out, arrogance is at work: the individual who takes others as the measure of his own worth and holds a low opinion of himself “is actually proud thereby” (L 27: 349). In the beliefs, attitudes, desires, and aims they share, then, servility and arrogance are opposed to self-respect but not to each other. Arrogance is just humility in waiting; interpersonal humility is just arrogance in disguise.27 Thus, interpersonal humility not only does not correct interpersonal arrogance, but it leads to or expresses it and involves the same fundamentally false view of the worth of persons. They are both the same kind of violation of one’s categorical moral duty of interpersonal recognition self-respect. Interpersonal humility is not a virtue; it is as serious a vice as interpersonal arrogance.

3 What, then, of true humility? Because humility involves a low estimation of one’s worth, it necessarily involves a comparative judgment. Interpersonal humility and 27   Servility and other forms of interpersonal humility are not the only guises arrogance can wear, nor is direct violation of the duty of self-respect the only moral problem with interpersonal humility, as Kant notes repeatedly. For example, while beneficence and sympathy are duties of love for others, Kant warns about arrogant beneficence and sympathy: pitying others or, like Lady Catherine de Burgh in Austen’s Pride and Prejudice, helping others in ways that casts oneself as the superior master over inferior dependents, humbling or humiliating them and undermining their self-respect (MS 6: 568–75, L 27: 697–8). The comparative/ competitive self-valuing of interpersonal humility also leads to the vices of hatred: envy, ingratitude, and malice (Schadenfreude) (MS 6: 576–8, L 27: 694–8, P 9: 491). By contrast, because interpersonal recognition self-respect is at the same time recognition respect for all other persons, it neither disguises arrogance nor poses other moral dangers.

Humility, Arrogance, and Self-Respect  55 interpersonal arrogance take the wrong standard for comparison and wrongly traffic in status and worth. True humility, by contrast, concerns the merits and demerits of the self as determined by the only true standard of worth, the moral law. In discussing servility, Kant also discusses true humility, defining it in contrast with arrogance: The consciousness and feeling of the insignificance of one’s moral worth in comparison with the law is humility [Demüt] (humilitas moralis). A conviction of the greatness of one’s moral worth, but only from failure to compare it with the law, can be called moral arrogance [Tugendstolz] (arrogantia moralis). (MS 6: 558)

The moral worth to which Kant refers here is not dignity (Würde), the absolute and unconditional worth that all persons have equally simply as beings capable of freely choosing to act through a purely rational motive, but, rather, the Werth that he calls “inner worth” or “moral worth” and that I’ll call “moral merit,” which is a conditional, scalar worth grounded in the use individuals make of their rational capacities. An individual earns moral merit for herself, her actions, and her character through goodness of the will, but fails to earn or earns little moral merit through acting wrongly, acting rightly for the wrong reasons, or having a weak or bad character. Since something can have “no worth other than that determined for it by the law” (G 4: 436), moral merit or lack thereof must be judged solely in comparison with the individual’s own self-given moral law or idea of virtue. True humility, which “follows unavoidably from our sincere and exact comparison of ourselves with the moral law” (MS 6: 436), “presupposes a correct estimation of self ” (L 27: 39). For when we honestly compare our moral performance to the standard of perfection set by the moral law, even the morally best of us fall far short, so that humility—having a low opinion of our moral merit—is appropriate. By contrast, arrogance [Tugendstolz, literally, “virtue pride”] involves claiming great moral merit and doing so by failing to assess one’s worth properly. Moral merit is grounds for a kind of self-respect different from that grounded on dignity, which I  call “evaluative self-respect.”28 Evaluative self-respect involves the judgment that one’s character and conduct at least come up to scratch in the absence of significant moral demerit; one feels it as “inner self-contentment” (C 5: 88, 117; MS 6: 387, 391). It is evaluative self-respect we mean when we say “I couldn’t respect myself again if I . . .” or “I’d lose my self-respect if I . . .” Of course, self-judgment is always liable to warping by inattention, blindness to one’s flaws, bias, self-deception, and so on, but warranted evaluative self-respect involves judging correctly that one is genuinely trying to live morally appropriately while also acknowledging that one still falls far short in comparison with what is morally required. Kant discusses something like evaluative self-respect in connection with “the first command of all duties to oneself,” which is “know (scrutinize, fathom) yourself . . . in terms of your moral perfections in relation to your duty,” to examine not only our 28   I borrow this term from Hudson (1980). Darwall’s term for it is “appraisal self-respect.” See Dillon (2004).

56  Robin S. Dillon actions but also our motives and patterns of motivation in light of the moral law (MS 6: 441). From this command follows the duties of “impartiality in appraising oneself in comparison with the law, and sincerity in acknowledging to oneself one’s inner moral worth or lack of worth” (MS 6: 442), which Kant identifies as “love of honor [Ehrliebe] (honestas interna, justum sui aestimium)” (MS 6: 420). Although true humility is part of honest evaluative self-respect, as Kant makes clear in the Lectures (L 27: 40, 349), here, as elsewhere, Kant also makes clear that humility is not necessarily the whole content of the judgment of inner worth derived from comparison of self with the moral law, a point to which we’ll return. Hill also discusses evaluative self-respect as what is at stake in trying to justify to oneself the choices that one faces or makes (Hill 1991d). If we engage in what Hill calls “deep deliberation” about a course of action, we look for reasons that would justify a decision to ourselves, not in terms of cost/benefits to self or others or in terms of anticipated pains and pleasures, but with a concern for whether we can be “reflectively satisfied with what the choice reveals (or makes) of  ” ourselves, for whether we can, when we reflectively and thoughtfully scrutinize ourselves, avoid self-condemnation and self-disapproval (175–6). What a deep deliberator seeks through reflective self-scrutiny is to have good reason to believe that, as the author of the person she will become as the result of choosing this course over another, she has been appropriately responsible both for and to herself. What is at stake, then, in making justifiable choices is not maintaining a pleasing self-image or good feelings about ourselves, but maintaining evaluative self-respect over time.29 Evaluative self-respect is not a favorable attitude toward oneself,30 nor is it pride or self-esteem, even “moral self-esteem.”31 Putting it in such terms does not do justice to what really matters to the self-respecting person, since it makes the focus of the self-respecting person’s concern her subjective state and so ultimately her comfort and happiness; but her real concern is the reality of her worthiness and the appropriateness of her self-valuing to that reality. Moreover, the worthiness with which it is concerned has to do not with the highly prized accomplishments or spiffy qualities that pride and esteem concern but rather, with congruence with and expression of one’s dignity

29   In a lovely scene from the classic western film Ride the High Country (MGM, 1962), Steve Judd, an aging former lawman whose frayed shirt cuffs and holey boots signal how far down he has come in a world that no longer values men like him, defends his choice to not help his old friend and current partner Gil Westrum steal gold they’ve been hired to guard, in terms of the value of maintaining self-respect. To Westrum’s taunt, “What’s that worth on the open market?,” Judd replies, “Not much to some people; a great deal to me. I lost it, but now I’m getting back a little respect for myself, and I intend to keep it.” Westrum’s scorn continues: “Know what’s on the back of a poor man when he dies? The clothes of pride. And they’re not a bit warmer to him dead than when he was alive. Is that what you want?” Judd answers, “What I want is to enter my house justified.” 30   For example, Telfer (1968) analyzes what she calls “estimative self-respect” as a favorable attitude toward oneself grounded in the belief that one’s character and conduct “come up to scratch.” 31   Darwall (1977) identifies what he calls “appraisal self-respect” as a species of self-esteem. See also Thomas (1995).

Humility, Arrogance, and Self-Respect  57 as a person. The self-respecting person’s overriding concern is not to regard herself as admirable or excellent, but rather, as Hill’s discussion makes clear, to know herself to be “justified.” Evaluative self-respect can be motivational, like self-esteem. But whereas those seeking self-esteem are motivated to achieve or to best others or to gain others’ esteem, those seeking to maintain or regain evaluative self-respect are motivated to forbear, to not engage in unworthy activity, and to live in a morally decent way. The moral value of evaluative self-respect lies in the importance to moral motivation, moral self-development, and reflective self-government of the disposition to appraise one’s conduct and character and to regulate oneself in light of one’s findings. Evaluating oneself and caring enough about the results of the survey to stake one’s sense of self on them is essential to keeping oneself on track, morally speaking, and to getting oneself back on track when one strays off. I will come back to the relation between true humility and self-respect. But it is important to say a bit more about the arrogance with which Kant contrasts true humility. Recall how he defines it: “a conviction of the greatness of one’s moral worth, but only from failure to compare it with the law” (my emphasis). The italicized bit indicates that arrogance is not only a matter of unjustifiable claims to superiority over other persons but also of ascribing any worth to oneself independently of the moral law. Since the moral law is the only determination of the worth of anything, moral arrogance is morally objectionable even apart from its denial of respect to other persons. The arrogance that is involved here is not interpersonal arrogance but unwarranted claims arrogance, an arrogation of moral merit which is the source of a false moral self-esteem that is not to be confused with evaluative self-respect. Kant sees unwarranted claims arrogance about moral merit as a serious problem, to judge from the frequency of his condemnations of it. It is the “egotistical self-esteem [eigenliebegen Selbstschätzung] which takes mere wishes—wishes that, however ardent, always remain empty of deeds—for proof of a good heart” (MS 6: 441); the “arrogantia [or] pride [Stolz], when we presume to a value that we do not possess” (L 27: 458); the self-regard that is self-satisfaction [Wohlgefallens an sich selbst] (arrogantia) or self-conceit [Eigendünkel], in which one’s “claims of self-esteem . . . precede conformity to the moral law” (C 5: 74); the self-conceit [Eigendünkel] in which a person “build[s]‌ too much on his own powers . . . [being] such a fool as to think he can fulfill [the moral law in its full purity] quite purely by his own efforts” (L 27: 351); the moral self-conceit [moralische Eigendünkel] of believing oneself “to be perfect in comparison to the law,” which is far worse than believing oneself superior to other people (L 27: 349); the “arrogance, or moral self-conceit [Arroganz, moralischen Eigendünkel]” that “makes an unwarranted pretension to merit” and is a “far more damaging defect” than philautia or moral self-love [Eigenliebe], which is “an inclination to be well-content with oneself ” and “devoid of self-reproach” (L 27: 357); the “moral self-conceit [moralischen Eigendünkel] of thinking [oneself] morally good, and having a favorable opinion of [one]self,” which is a “dream-like condition” that is one of the “tendencies to evil” in

58  Robin S. Dillon us that we must “constantly contend against” (L 27: 464); the complacent self-love that rests on the failure to assess our true moral worth and is “the cause of great harm” (L 27: 621–3); and a “source of all evil” (R 6: 46). A number of questions arise from Kant’s discussion of unwarranted claims arrogance, true humility, and moral merit. For example, why is a low opinion of moral worth morally valuable (independently of the value of honesty)? And what is so bad about a too-high opinion of one’s inner worth? In particular, why is arrogating moral merit not merely foolish but rather a “very damaging defect” that is far worse than disrespect of others, a “cause of great harm,” and a “tendency to evil,” indeed, a “source of all evil”? If we attend again to the way Kant defines arrogance, we can identify a form of unwarranted claims arrogance that is a much more serious vice than misjudging one’s inner worth, a form that is linked to a third kind of self-respect.

4 Given, on the one hand, the ineluctable imperfection and inevitable failings of every human and, on the other, the high standards for conduct and character set by the moral law, it is reasonable to suppose that an accurate assessment of one’s inner worth in comparison with the law would yield humility. Thus any high opinion of moral merit would be unjustifiably high. Nevertheless, moral arrogance lies, Kant says, not in overestimating one’s moral worth but in doing so “only from failure to compare it with the law.” This suggests that any claim to merit, even if it happens to be accurate, is unwarranted if it doesn’t come from comparing oneself against the standards of moral excellence contained in the law. The arrogance of claiming great moral merit is thus not merely the conceit that results from poor self-assessment skills, but rather the elevation of one’s desire to think well of oneself to the level of the standard of moral worth. Moral arrogance, the arrogation by desire of the authority to determine worth, is the refusal to acknowledge the moral law as the supreme condition of all worth of persons and hence the only standard for self-evaluation. The unwarranted claims arrogance we have here is not simply the absence of evaluative self-respect but also a failure of what I call “agentic recognition self-respect.” True humility plays its important moral role in relation to this kind of self-respect. My earlier discussion of recognition self-respect emphasized its interpersonal nature: because each rational being is a person among equal persons, interpersonal recognition respect and self-respect ought to structure all of our relations with other persons. But self-respect is not only about how one engages with others in light of the fact that one is a person; it is also, and more centrally, about how one engages with oneself as a person. The self-respecting person understands herself not only as the object of other people’s attitudes and actions but also as the subject of her own, and insofar as she understands herself to have control over her conduct and character, she can regard the dignity she has as a person as both demanding and constraining the exercise of her

Humility, Arrogance, and Self-Respect  59 agency in living her life, as giving rise to a responsibility to shape herself and direct her living so that they are congruent with and honor her dignity as a person. She has, that is, due regard for herself as a moral agent. The differences between the interpersonal and agentic forms of recognition selfrespect reverberate in different ways throughout the individual and her life, and it is possible for an individual to respect herself in one way and not the other. Consider again Hill’s example of an Uncle Tom. The Uncle Tom, says Hill, “always steps aside for white men; he does not complain when less qualified whites take over his job; he gratefully accepts whatever benefits his all-white government and employers allot him, and he would not think of protesting its insufficiency . . . he accepts without question the idea that, as a black, he is owed less than whites” (Hill 1991a: 5). Such an individual clearly does not think of himself as an equal person among persons; he is servile and lacks interpersonal recognition self-respect. Nevertheless, we can imagine that Uncle Tom need not think of himself as lacking all worth as a person. We can easily imagine, for example, that he regards character as something of great importance, that he thinks of himself as a moral agent with duties to others and even to himself, that he regards the moral agency that he knows he has as demanding certain things from him with which he tries to live in accord. So, he may hold himself to high moral standards regarding honesty, integrity, perseverance, temperance, self-reliance, dependability, courage. He may regard it as a matter of self-respect that he put forth his best effort in whatever he does, that he gives an honest day’s work for an honest day’s pay, that he be a good provider for his family not only materially but also emotionally, morally, and spiritually, that he shun prurient forms of entertainment that less worthy men enjoy. And he may be successful in living in accord with his understanding of a worthy life, conducting himself, we might say, with dignity, living as he believes self-respecting people live. Because he is (mostly) right about this, he has agentic recognition self-respect.32 While interpersonal recognition self-respect is the reflexive form of interpersonal respect for other persons, agentic recognition self-respect does not correspond to a kind of other-directed respect; it is uniquely self-respect. Hill describes something very like it when he “reconsiders” his analysis of servility and self-respect (Hill 1991b). Reflecting on a talented artist who alters what he knows to be a work of originality to make it more appealing to “a tasteless public,” an actress who supports herself as a prostitute because the pay is better than waitressing, and a respectable professor whose infatuation with a burlesque dancer leads him to humiliate himself as a cabaret clown, Hill argues that they lack a kind of self-respect that is not a matter of understanding and properly valuing their basic rights, but that requires instead “that one develop and live by a set of personal standards by which one is prepared to judge oneself ” (Hill 1991b: 22). These standards would be “an important part of oneself,” such that violating 32   Of course, he is wrong about something very important, namely, his moral status and equal dignity, and that is a failing as regards agentic recognition self-respect. I leave aside the question of whether the latter is, unlike interpersonal recognition self-respect, a matter of degree.

60  Robin S. Dillon them would be a self-betrayal that understandably would give rise to attitudes of negative self-judgment such as the shame, self-disgust, self-contempt, even self-hatred that is experienced by the individuals Hill considers. Such attitudes reflect the absence of evaluative self-respect; they express the judgment that one has failed even to come up to scratch by the light of the standards one holds oneself to live in accord with. Now, while Hill stresses that these standards need not be (narrowly) moral ones nor ones that the individual regards as objective and would hold others to, we might reasonably agree with Kant that there are some moral standards that every moral agent ought to incorporate into their personal set, by which each of us ought to live and evaluate ourselves morally. But even where the standards are applicable to all persons, the commitment to hold oneself to them and judge oneself by their light makes them one’s own, a part of oneself. Committing oneself to live in light of standards that are, to a great extent, self-defining, reflects what I’m calling agentic recognition self-respect. This kind of self-respect involves an appreciation of oneself as a moral agent, a being with the ability and responsibility to act autonomously and value appropriately. An individual with agentic recognition self-respect values herself appropriately by committing herself to live, and then striving mightily to live, in accord with norms that configure the kind of life she regards as appropriate to her as a person, norms that define the kinds of character traits, attitudes, and concerns she should and shouldn’t have, the kinds of projects and activities that would be worthy and unworthy of her pursuit, the talents and abilities she should and should not cultivate and exercise, the emotions and motivations she should endorse or ignore, the modes of conduct that are appropriate and inappropriate. She regards certain forms of acting, thinking, desiring, and feeling as befitting her as a person and other forms as self-debasing or shameful, and she expects herself to adhere to the former and avoid the latter. Indeed, she understands the latter as what she must not and cannot do or be, no matter what she might gain from it, and she restrains herself from what she regards as unworthy of herself as a person.33 In these ways she invests herself in herself, and if her will and the world cooperate, she creates a worthy self. The person with agentic self-respect regards the development of one’s character and the shaping of one’s life as vitally important tasks that a person must take up and do well. Her commitment to being a worthy person living a worthy life means that it matters very much to her that the person she is or is becoming and the life she is leading are genuinely worthy, genuinely congruent with and expressive of her dignity as a person.

33   Some of the norms for conduct and character to which she is committed she regards as such that other people, maybe all others, ought also to live in accord with them; others she regards as pertaining especially or only to herself. In a sociocultural environment in which individuality is highly valued and regarded as an important expression of the dignity of persons, the fully self-respecting individual will regard it as important to exercise her agency in becoming her own person and in living a life that is particularly hers. So, she will commit herself to living in accord with norms that define and befit her as the particular person she is and should be, and she will expect herself (but no one else) to adhere to certain personal standards, the disappointment of which she would regard both as self-betrayal and as degrading to her dignity as a person.

Humility, Arrogance, and Self-Respect  61 And so it matters to her that she know whether her efforts to shape herself and her life are on track, whether she is worthy or at least headed in the right direction. Thus she has reason to reflect on and evaluate herself to determine how she stands. This is where evaluative self-respect comes in. The agentically self-respecting person respects herself by living in a certain way; the person with evaluative self-respect respects herself to the extent that, as she believes, she does live that way. Someone who respects herself as a moral agent takes her responsibilities seriously, especially her responsibilities to honor her dignity as a person, to govern herself fittingly, and to make of herself and her life something she believes to be good and worthy. For Kant, the most vital of the responsibilities that each moral agent has is the responsibility to realize one’s personality, one’s capacity for moral valuing and autonomous agency, by choosing to act through rational motives, i.e., from respect for the moral law, acknowledging the absolute authority of the moral law, which is to say, the dictates of one’s own rationality unimpeded by the importuning of the inclinations. When one acts on maxims that derive their motivating power from reason rather than from inclination, one acts like the being one is, honoring the dignity one has as a person, i.e., a being with the capacity to act on reason. When one acts otherwise, one betrays one’s dignity and fails to respect oneself as a moral agent. The unwarranted claims arrogance with which Kant is concerned involves just this kind of failure to respect oneself. What makes it morally objectionable is the source of the unjustified claiming of great moral worth: the arrogation by inclination of the authority to determine worth and so the refusal to acknowledge the moral law as the supreme condition of all worth. As with interpersonal arrogance, unwarranted claims arrogance about moral merit is motivated by a particular desire: it sacrifices honest self-assessment in exchange for a more easily obtained enhancement of self-esteem. The arrogant individual wants to think well of himself, but rather than striving to earn moral worth, he arranges his judgments so as to declare himself to be morally admirable. In ignoring the supremacy of the moral law in the adjudication of all matters of worth, the arrogant individual subordinates his rationality to his inclination and so makes himself a “plaything of the inclinations,” debasing his dignity as a rational being by failing to respect his own moral agency. Kant’s discussion of arrogance in the second Critique reinforces this assessment of it. In his explanation of how the moral law becomes an incentive, that is, how it can directly determine our choice of action independently of our inclinations, Kant contrasts self-love [Selbstliebe, Eigenliebe] with what he calls self-conceit [Eigendünkel] or arrogance [Arrogantia] (C 5: 73) in this way: “This propensity to make the subjective determining grounds of one’s choice into an objective determining ground of the will in general is called self-love; when it makes itself legislative and an unconditional practical principle, it can be called self-conceit . . . [S]‌elf-conceit . . . decrees the subjective conditions of self-love as laws” (CB 5: 74). As Beck puts it, arrogance is “the inclination to take one’s own subjective maxims and interests as having the authority of law” (Beck 1960: 291).

62  Robin S. Dillon Unwarranted claims arrogance is the most serious threat to morality, and its corrupting effect on motivation and judgment explains why.34 The arrogant person desires to heighten or maintain his self-esteem; but an agent can’t think well of himself morally without acknowledging moral standards and taking his conduct, motives, and character to conform to them. Someone of modest merit, great demerit, and a compelling desire to think well of himself arrives at a high opinion of himself not by ignoring the moral law or thumbing his nose at it—which Kant regards as impossible for a rational being (R 6: 35)—but rather by taking himself to meet the demands of the law which he “flatters himself that he inwardly reveres” (MS 6: 430). This he does by casting the merely contingent features of his personal psychology as objectively justified principles, giving the desire for self-esteem the kind of authority that is possessed only by the moral law generated by his own practical reason. The desire motivates him to deceive himself not only about his own actions and motives, but also about the law, as he “tinkers with” it (L 27:465), adjusting the law’s standards to his actions so that he can think well of himself as doing his moral duty, no matter what he does. The desire for self-esteem makes self-love a “legislative and an unconditional practical principle,” usurping for itself the authority that is rightly possessed only by the moral law generated by pure practical reason. Unwarranted claims arrogance is thus what Kant identifies as the deepest source of evil in human nature, in which “the mind’s attitude is corrupted at its root” (R 6: 30), for it involves not merely the frailty of wanting to do right but being too weak to resist temptations, nor the impurity of needing to be pushed by our inclinations to do what we know we ought to do, but the “depravity” or “perversity” of subordinating the incentives of the moral law to those of the inclinations (R 6: 29–30), making “the incentives of self-love and their inclinations the condition of the compliance with the moral law” (R 6: 36). The unwarranted claims arrogance of self-conceit must therefore be eliminated to make possible the practical acknowledgment of the authority of the moral law. And here is where true humility comes in. True humility, recall, is the emotionally experienced recognition of the insignificance of our moral worth that arises inexorably whenever we sincerely and accurately compare our moral performance with the exacting standards of the moral law.35 True humility curbs the high assessment of inner worth (L 27: 350) to which we are enticed or pushed by the self-esteem motive, “confining self-esteem in its legitimate bounds” (L 27: 636). We might well have a longer list of accomplishments, a better moral track record, spiffier personal qualities, or a more admirable character than someone else (especially if we choose the someone else cleverly). But when we honestly hold ourselves up to the standard of the moral law, we realize any opinion of our moral worth

34   I am indebted to Bernard Reginster for calling my attention to and helping me to understand this process. 35   Kant’s understanding of true humility thus parallels the view we find in the Deadly Sins tradition, where the exalted standard is God’s power and majesty: compared with that, all human worth is insignificant.

Humility, Arrogance, and Self-Respect  63 other than a low one is unjustified. When we realize this, our pretentious self-conceit is, as Kant says in the second Critique, “struck down” as the moral law “unavoidably humiliates” [demütigt] our unjustified claims of moral worthiness, reducing them “to nothing” (C 5: 73–4). This humiliation is part of the subjective experience that constitutes the incentive of pure practical reason, i.e., the effect that the moral law has on our motivational system as soon as we recognize it. The humiliating recognition of unworthiness sweeps away the self-importance that otherwise blocks us from directing the abilities and efforts of our agency toward making morally appropriate choices, willing properly, acting rightly, and improving ourselves morally, rather than wasting them on making morally poor choices, willing badly, acting wrongly, and shoring-up our illusions of decency through rationalization, self-deception, and willful self-ignorance. But while true humility has this useful, perhaps essential, role in making it possible for us to exercise our moral agency well, Kant warns that it also poses a danger. For if humility is all we are left with upon self-evaluation, we are quite liable to lose all moral motivation, sinking into despair, “despondency” [Kleinmütigkeit], and “timorousness” [Mutlosigkeit] (L 27: 350). When we focus on the inevitably long list of our flaws, deficiencies, inadequacies, and failures, we may come to believe that our actions “never comply with the moral law,” that our characters lack any morally redeeming dimensions, that we will never amount to anything, morally speaking, that can never hope to be the good persons that we believe we should be; and then “inertia arises [and we] venture to do nothing at all” (L 27: 350). There is a real possibility, that is, that unconstrained humility will undermine moral agency. The danger posed by true humility is different from that posed by the false humility and arrogance produced by the desire for self-esteem, but it is every bit as serious: the threat of distorted agency is cause for deep concern no matter what its source. Thus even true humility is as much in need of constraint as is the self-esteem motive. The peril to moral agency from unconstrained humility is averted, however, because humility is not all that results from comparing oneself sincerely to the moral law. For at the same time as I recognize the extent of my failure to live up to the moral law, I recognize three other important things about myself. First, I understand myself to be the author of the law. The standards I don’t live up to are my own standards, the expression of my own ideas about how it is necessary and appropriate for a being like myself to live and act. The judgment of deficiency is my own judgment, the verdict of my own conscience, the judge in the internal courtroom in which I put myself on trial;36 and the punishment of humiliation is one I inflict on myself and willingly endure as a matter of moral integrity, an expression of my commitment to the moral law and to myself as an agent committed to living worthily. Second, I am conscious of my “sublime predisposition” for moral self-governance and virtue, and I understand that it is “only through the noble predisposition to the good in us” that I can and do judge myself to lack moral

  See Hill’s (2002) explanation of the structure and workings of conscience.

36

64  Robin S. Dillon worth (MS 6: 441). Only someone who sincerely cared about and was committed to being morally good and acting morally worthily would bother to evaluate herself honestly, would face and accept the sad and humbling truth about herself, would think that the assessment could and does matter for how she goes on from here. Then third, as a result, I recognize myself for what I am most essentially: a person; and I feel an awe and “an exaltation of the highest self-valuation” that is respect for my inalienable dignity as a person (MS 6: 436). In this way, arrogating self-conceit is replaced by recognition self-respect and the self-esteem motive is demoted to a subordinate place in the motivational structure. As Kant’s discussion in the second Critique makes clear, humility is thus only the first stage in a complex response to ourselves and our worth; and Kant’s emphasis in discussing this response throughout the ethical works is not on lowly humility but on the elevated respect for ourselves that assessment of worth inevitably produces.37 What is more, the judgment of insignificant inner worth is not the whole of the verdict of self-evaluation, at least for all but the most depraved of us (who wouldn’t make these judgments anyway). For, like the “moderately honorable man” who abstains at some cost to himself from wrongdoing because he knows it is wrong, we, too, can be conscious of “maintaining humanity in its proper dignity in our person and honoring it” so that we needn’t “dread the inward view of self-examination” (C 5: 88). Indeed, everyone, Kant believes, “will feel in himself that, once at least, he has performed a good action from good dispositions and is capable of doing yet more of more” (L 27: 351). Thus in an honest self-evaluation, it is possible for an individual to be conscious of actual goodness and virtue as well as deficiency.

5 As Kant explains it, the attitude of true humility and the experience of felt humiliation are inevitably evoked whenever we are aware of the moral law. Since we can have no duty to experience something, let alone something that we can’t help but experience, humility is not something we have, or could have, a duty to cultivate. Nor is humility a Kantian virtue. A virtue in Kant’s view is not just any morally good quality. It is, rather, a matter of will, of a strong and effective considered resolve to master and restrain inclinations “in accordance with a principle of inner freedom” and “in relation to an end one ought to make one’s own” whenever inclinations “rebel against the law” (MS 6: 381, 383, 395, 405).38 Humility is a matter of judgment, not a matter of will. Thus, true humility is not a virtue, let alone a preeminent virtue. But it can be part of a virtue. Humility involves the knowledge of one’s limitations and deficiencies judged 37   In reversing the significance of these two moments of humility and self-respect, Grenberg (2005) misrepresents Kant’s views of both the moral value of true humility and its place in the moral life. In this assessment I agree to a certain extent with Louden (2007). 38   See Hill (2008) for a valuable discussion of Kant’s account of virtue.

Humility, Arrogance, and Self-Respect  65 in comparison with self-given moral law; the virtue lies in dealing properly with this knowledge,39 which is what the virtue of agentic recognition self-respect involves. The moral value of true humility derives from its role as a servant of this virtue. Like interpersonal recognition self-respect, agentic recognition self-respect is a duty we have to ourselves. But it becomes a virtue when appreciation of one’s dignity becomes the motivational core of one’s self-conception, when one freely commits oneself to the self-valuing that duty requires and when one’s self-defining considered resolve is to live in ways that honor one’s dignity as a moral agent and a person among persons. Because one is aware always of the “continuing propensity to transgression” (C 5: 128), the commitment to honor one’s dignity as a person involves both the commitment to live up to the standards of the moral law and to restrain and try to master the importune urges of the self-esteem motive, and thus also a commitment to moral self-improvement. These commitments—to valuing oneself as an end in oneself and valuing the self one is in light of the self that one knows oneself to be fully capable of continually approaching—make agentic recognition self-respect a virtue and distinguish it from self-esteem. It is important to understand that what the self-respecting person wants is not to esteem herself, not to feel good about herself or think well of herself, but to know whether she is living up to her commitments to be worthy of her dignity. Since she needs to know this about herself, she is also rationally committed to engage in honest moral self-assessment. Both evaluative self-respect’s judgment of coming up to scratch (if one does) or being morally worthy (if one is) and humility’s low assessment provide an agent with information that is of crucial importance to the core task of living up to these commitments, guiding her continuing efforts to improve herself morally. Nevertheless, the main purpose of comparing ourselves with the law and accepting its judgment is not to show us how unworthy we are, thereby humiliating self-conceit, although that is its foreseeable effect. The central purpose is to remind ourselves of the work yet to be done and in this way to motivate renewed striving to be good and do right. Because the self-respecting person does not place a premium on thinking well of herself or feeling good about herself, she is both strongly motivated and well equipped to resist the siren call of the self-esteem motive. Thus she can avoid both the Scylla whose twin heads of servility and arrogance seek to devour moral agency and personhood and the Charybdis of enervating hopelessness that threatens to drown agency.40 For the agentically self-respecting person, true humility yields neither despondency and timorousness nor attempts to boost the moral worth score by self-deception and tinkering with the law, nor even efforts to compensate for a low score in the moral domain by seeking high

39   Christine Swanton (2011) usefully distinguishes between humility as knowledge and as the disposition to deal appropriately with it. 40   Kant writes that arrogance and timorousness are “the two rocks a man runs in to, if he departs, in one direction or the other, from the moral law” (L 27: 350).

66  Robin S. Dillon self-esteem in other domains of life. The self-respecting person puts humility to work in “a firm determination” to the “tenacious pursuit” of her principles in “dutiful obedience to the law” (L 27: 610).41 The content of the consciousness that is humility thus has an important role to play, but only in service to the commitment to respect oneself as a moral agent. For the self-respecting person, self-worth matters tremendously. Dignity is the hallmark of one’s nature as a person, concern for moral merit is the sign of one’s commitment to honor and manifest that dignity, assessment of moral merit is an exercise of one’s agency, the degree of moral merit one earns or loses is both the fruit of agency well or poorly exercised and the signal to persevere in one’s striving to improve one’s actions, one’s character, and the shape and course of one’s life. To think that virtuous humility is a matter of being oblivious to or unconcerned about or unimpressed with self-worth or of underestimating one’s real worth, or a matter of restraining one’s ambition and desire for moral excellence, to think that its value lies chiefly in freeing one to pay more and better attention to more important things or in preventing or curing arrogance, is not only to misunderstand what humility is and how it matters morally. It is to deeply misunderstand its relation to arrogance and self-respect as well as the significance of self-respect in properly valuing the moral dimensions of human life (which is to say, all dimensions of human life) and in effectively motivating morally appropriate living. Humility not anchored in, constrained by, and at the service of agentic recognition self-respect is liable to the agency-undermining excess and interpersonal comparisons of self-worth that misvalue the self and give rise to arrogance. Self-valuing not structured by the commitment to value oneself appropriately generates self-corrupting unwarranted claims arrogance and dignity-violating, self- and other-disrespecting interpersonal arrogance. The virtue of agentic self-respect combats the tendencies to arrogance, makes proper use of the right kind of humility, motivates compliance with the duties of interpersonal recognition respect for all persons including oneself, preserves sound moral judgment and effective moral agency, and ensures that one strives diligently and effectively to “be worthy of the humanity that dwells within” (MS 6: 387).

Abbreviations Citations to Kant’s writings are to the Akademie edition. I have used the following abbreviations for Kant’s texts in translation: G

Groundwork of the Metaphysics of Morals, in Kant, Practical Philosophy, trans. and ed. Mary J. Gregor (Cambridge: Cambridge University Press, 1996)

41   For the self-respecting person, negative assessment is valuable in other ways: it manifests to oneself (as well as others) the value one places on being a certain kind of person, reaffirms the virtues and norms one has violated, and assures oneself of one’s continuing commitment to the moral law and the dignity of persons, thus strengthening the resolve to resist inclinations that tempt one to abandon or subordinate these commitments.

Humility, Arrogance, and Self-Respect  67 C

Critique of Practical Reason, in Practical Philosophy

CB

Critique of Practical Reason, trans. Lewis White Beck (Indianapolis: Bobbs-Merrill Company, 1956)

MS The Metaphysics of Morals, in Practical Philosophy L

Lectures on Ethics, ed. Peter Heath and J.  B. Schneewind, trans. Peter Heath (Cambridge: Cambridge University Press, 1997)

R

Religion Within the Boundaries of Mere Reason, in Kant, Religion and Rational Theology, trans. and ed. Allen W.  Wood and George DiGiovanni (Cambridge: Cambridge University Press, 1996)

P

Lectures on Pedagogy, trans. Robert B. Louden, in Kant, Anthropology, History, and Education, ed. Gunter Zoller and Robert B. Louden (Cambridge: Cambridge University Press, 2007)

A

Anthropology from a Pragmatic Point of View, ed. Robert B.  Louden (Cambridge: Cambridge University Press, 2006)

References Aquinas, Thomas. 1921. Summa Theologica I-II, literally translated by the Fathers of the English Dominican Province. London: Burns Oates and Washborn. Augustine. 1996. The City of God Against the Pagans, trans. Phillip Levin. Loeb Classical Library, London: William Heineman and Cambridge, MA: Harvard University Press. Baumeister, Roy F. 1994. “Self-Esteem,” in Encyclopedia of Human Behavior, Volume 4, ed. V. S. Ramachandran. San Diego, CA: Academic Press. Beck, Lewis White. 1960. A Commentary on Kant’s Critique of Practical Reason. Chicago: University of Chicago Press. Benedict. 1875. The Rule of our Most Holy Father St. Benedict, ed. Fathers of St. Michaels. London: Washbourne. Bernard of Clairvaux. 1929. The Twelve Degrees of Humility and Pride, trans. Barton R.V. Mills. London: MacMillan. Burbach, Roger and Jim Tarbell. 2004. Imperial Overstretch: George W. Bush and the Hubris of Empire. London: Zed Books. Carlson, Sebastian. 1944. “The Virtue of Humility.” The Thomist, 7, no. 2, 3: 135–78, 363–414. Cohan, William D. 2009. House of Cards: A Tale of Hubris and Wretched Excess on Wall Street. New York: Doubleday. Coopersmith, Stanley. 1967. The Antecedents of Self-Esteem. Palo Alto, CA:  Consulting Psychologists Press. Darwall, Stephen. 1977. “Two Kinds of Respect.” Ethics, 88: 34–49. Darwall, Stephen. 2006. The Second Person Standpoint: Morality, Respect, and Accountability. Cambridge, MA: Harvard University Press. Dillon, Robin S. 2003. “Kant of Arrogance and Self-Respect,” in Cheshire Calhoun, ed., Setting the Moral Compass: Essays by Women Philosophers. Oxford: Oxford University Press, pp. 191–216.

68  Robin S. Dillon Dillon, Robin S. 2004. “ ‘What’s a Woman Worth? What’s Life Worth? Without Self-Respect?’: On the Value of Evaluative Self-Respect,” in Margaret Walker and Peggy DesAutels, eds., Minds, Hearts, and Morality:  Feminist Essays in Moral Psychology. Lanham, MD:  Rowman & Littlefield, pp. 47–68 Dillon, Robin S. 2007. “Arrogance, Self-Respect, and Personhood.” Journal of Consciousness Studies, 14: 100–26. Dillon, Robin S. 2012. “Arrogance,” in Hugh LaFollette, ed., International Encyclopedia of Ethics. New York: Wiley-Blackwell. Driver, Julia. 2001. “The Virtues of Ignorance,” in Driver, Uneasy Virtue. London: Cambridge University Press. Eckhart, Meister. 1981. “On Humility” and “On Detachment,” in The Essential Sermons, Commentaries, Treatises, and Defense, trans. Edmund Colledge and Bernard McGinn. New York: Paulist Press. Fisher, N.  R. E. 1992. Hybris:  A  Study of the Value of Shame and Honor in Ancient Greece. Warminster: Aris and Phillips. Flanagan, Owen. 1990. “Virtue and Ignorance.” Journal of Philosophy, 87: 420–8. Garcia, J. L. A. 2006. “Being Unimpressed with Ourselves: Reconceiving Humility.” Philosophia, 34: 417–35. Gecas, Viktor 2001. “The Self as Social Force,” in Timothy Owens, Sheldon Stryker, and Norman Goodman, eds., Extending Self-Esteem Theory and Research: Sociological and Psychological Currents. Cambridge: Cambridge University Press, pp. 85–100. Gregory the Great. 1850. Morals on the Book of Job, vol. 4. Oxford:  John Henry Parker, and London: F. and J. Rivington. Grenberg, Jeanine. 2005. Kant and the Ethics of Humility: A Story of Dependence, Corruption, and Virtue. Cambridge: Cambridge University Press. Hay, Carol. 2013. Kantianism, Liberalism, and Feminism. New York: Palgrave Macmillan. Hill, Thomas E., Jr. 1991a. “Servility and Self-Respect,” in Hill, Autonomy and Self-Respect. Cambridge: Cambridge University Press, pp. 4–18. Hill, Thomas E., Jr. 1991b. “Self-Respect Reconsidered,” in Hill, Autonomy and Self-Respect. Cambridge: Cambridge University Press, pp. 19–24. Hill, Thomas E., Jr. 1991c. “Ideals of Human Excellence and Preserving Natural Environments,” in Hill, Autonomy and Self-Respect. Cambridge: Cambridge University Press, pp. 104–17. Hill, Thomas E., Jr. 1991d. “Pains and Projects: Justifying to Oneself,” in Hill, Autonomy and Self-Respect. Cambridge: Cambridge University Press, pp. 173–88. Hill, Thomas E., Jr. 2000a. “Kantian Pluralism,” in Hill, Respect, Pluralism, and Justice: Kantian Perspectives. New York: Oxford University Press, pp. 11–32. Hill, Thomas E., Jr. 2000b. “Basic Respect and Cultural Diversity,” in Hill, Respect, Pluralism, and Justice. New York: Oxford University Press, pp. 59–86. Hill, Thomas E., Jr. 2002. “Punishment, Conscience, Moral Worth,” in Mark Timmons, ed., Kant’s Metaphysics of Morals: Interpretative Essays. New York: Oxford University Press, pp. 233–54. Hill, Thomas E., Jr. 2008. “Kantian Virtue and ‘Virtue Ethics’ ” in Monika Betzler, ed., Kant’s Ethics of Virtue. Berlin: Walter de Gruyter, pp. 29–60. Hudson, Stephen D. 1980. “The Nature of Respect,” Social Theory and Practice, 6: 69–90.

Humility, Arrogance, and Self-Respect  69 Hume, David. 1972. Enquiries Concerning the Human Understanding and Concerning the Principles of Morals, 2nd edn., ed. L. A. Selby-Bigge. Oxford: Oxford University Press. Isikoff, Michael and David Corn. 2006. Hubris: The Inside Story of Spin, Scandal, and the Selling of the Iraq War. New York: Crown. Kaplan, Howard B. 1975. “Prevalence of the Self-Esteem Motive,” in Self-Attitudes and Deviant Behavior, ed. Howard B. Kaplan. Pacific Palisades, CA: Goodyear. Kernis, Michael H. ed., 2006. Self-Esteem Issues and Answers:  A  Sourcebook of Current Perspectives. New York: Psychology Press. Kershaw, Ian. 1993. Hitler, 1889–1939: Hubris. New York: Penguin Books. Louden, Robert B. 2007. “Kantian Moral Humility: Between Aristotle and Paul.” Philosophy and Phenomenological Research, 75: 632–9. Mruk, Christopher J. 2006. Self-Esteem Research, Theory and Practice, 3rd edn. New York: Springer Publishing Company. Nuyen, A. T. 1998. “Just Modesty.” American Philosophical Quarterly, 35: 101–9. Peterson, Christopher and Martin E. P. Seligman. 2004. “Humility and Modesty,” in Character Strengths and Virtues: A Handbook and Classification. New York: Oxford University Press, pp. 461–76. Proust, Michel. 1934. Remembrance of Things Past, III. The Guermantes Way, vol. 1, trans. C. K. Scott Moncrieff. New York: Random House, 1934. Richards, Norvin. 1992. Humility. Philadelphia: Temple University Press. Roberts, Robert C. 2009. “The Vice of Pride.” Faith and Philosophy, 26: 119–33. Roberts, Robert C. and Jay Wood. 2003. “Humility and Epistemic Goods,” in Intellectual Virtue, ed. Michael DePaul and Linda Zagzebski. Oxford: Clarendon Press, pp. 257–79. Rosenberg, Morris. 1965. Society and the Adolescent Self-Image. Princeton: Princeton University Press. Sensen, Oliver. 2009. “Kant’s Conception of Human Dignity.” Kant-Studien 100: 309–31. Snow, Nancy. 1995. “Humility.” Journal of Value Inquiry, 29: 203–16. Swanton, Christine. 2011. “Kant’s Impartial Virtues of Love,” in Lawrence Jost and Julian Wuerth, eds., Perfecting Virtue. Cambridge: Cambridge University Press, pp. 241–59. Taylor, Gabriele. 1994. “Vices and the Self,” in Philosophy, Psychology, and Psychiatry. Royal Institute of Philosophy Supplement 37, ed. A. Phillip Griffiths. Cambridge:  Cambridge University Press, pp. 145–57. Taylor, Gabriele. 2006. Deadly Vices. London: Oxford University Press. Telfer, Elizabeth. 1968. “Self-Respect.” The Philosophical Quarterly, 18: 114–21. Thomas, Laurence. 1995. “Self-Respect:  Theory and Practice,” in Dignity, Character, and Self-Respect, ed. Robin S. Dillon. New York: Routledge, pp. 251–70. Warren, Rick. 2002. The Purpose Driven Life. Grand Rapids, MI: Zondervan. Weil, Simone. 2002. Gravity and Grace, trans. Emma Craufurd. London: Routledge. Wood, Allen. 1999. Kant’s Ethical Thought. New York: Cambridge University Press.

3 Respect as Honor and as Accountability Stephen Darwall

In his Tanner Lecture, “Respect for Humanity,” Thomas Hill opposes a Kantian “ethics of respect for persons” to an earlier “aristocratic doctrine,” according to which people merit differential treatment in accordance with “heredity and social rank” (Hill 1997: 9). Hill’s contrast is both important and correct. In what follows, however, I want to investigate how both the ancien régime and its modern Kantian replacement can be seen as interpretations of the idea of recognition or respect for persons, only in fundamentally different senses of “person” corresponding to different conceptions of (recognition) respect. For an aristocratic outlook, “person” has the sense of “persona”—social presentation or “face”—and respect for persons in this sense is an honor that helps constitute differences of status or rank as social facts. For its Kantian alternative, on the other hand, a person is a moral subject—a being with the capacity to be morally obligated. And respect for persons is recognition of an inviolable dignity that all moral agents share equally by virtue of sharing this capacity. To explore this difference, we may begin by contemplating Bernardo Strozzi’s wonderful painting, The Tribute Money, which depicts the young Jesus among Pharisees sent, as Mark tells us, “to catch him in his words.”1 Herod and his followers want to “lay hold” of Jesus, but fear popular reaction, so they seek to make Jesus incriminate or dishonor himself. Here are Mark’s words in the King James Version: And they sought to lay hold on him, but feared the people . . . And they send unto him certain of the Pharisees and of the Herodians, to catch him in his words. And when they were come, they say unto him, Master, we know that thou art true, and carest for no man: for thou regardest not the person of men, but teachest the way of God in truth: Is it lawful to give tribute to Caesar, or not? Shall we give, or shall we not give? But he, knowing their hypocrisy, said unto them, why tempt ye me? Bring me a penny, that I may see it. And they brought it. And he saith unto them, whose is this image and superscription? And they said unto him, Caesar’s. And Jesus answering 1   Bernardo Strozzi, The Tribute Money, 1630–5, Museum of Fine Arts, Budapest. Autograph versions exist, for example, in the Uffizi Gallery in Florence.

Respect as Honor and as Accountability  71 said unto them, Render to Caesar the things that are Caesar’s, and to God the things that are God’s. And they marvelled at him. (Mark 12: 13–17)

There are several elements of this story to which I shall wish to return, not least, first, the Pharisees’ comment that Jesus “regardest not the person of men,” or as it is sometimes said also of the law, that he is not a “respecter of men,” and second, that Jesus “teachest the way of God in truth.”2 Before that, however, notice Jesus’ “regard,” and imagine that Strozzi is depicting the moment when Jesus confronts the Pharisees with the charge: “Why tempt ye me?”3 Jesus looks the Pharisee to his left in the eye and asks him, in effect, to answer for himself. He implicitly holds the Pharisee accountable and attempts thereby to get the Pharisee’s recognition of his authority to do so. But where does the Pharisee look? Does he reciprocate Jesus’ regard? Imagine, at least, that he does not, that he is looking at the Pharisee on Jesus’ right.4 On this imaginative interpretation, the painting’s dramatic tension partly depends on the fact that the Pharisee to whom Jesus looks refuses to look back at him, looking steadfastly in the direction of the other Pharisee, who returns his colleague’s regard. This creates a tension between two axes, one aspirational, between Jesus and the Pharisee on his left, and one actual, between the two Pharisees. In The Second-Person Standpoint, I referred to Strozzi’s painting as an illustration of an aspect of what I there call “Pufendorf ’s Point,” namely, that even God must gain our recognition to hold us accountable (Darwall 2006: 142n.–143n.). The thought was that holding someone responsible involves commitment to a form of mutual respect, since it commits the holder to the idea not just that he has an authority to hold the other responsible, but also that the other can hold himself and others responsible as well. A major aim of the book was to argue that the second-personal address involved in holding people responsible—“Why tempt ye me?”—commits addresser and (an acknowledged) addressee alike to their sharing a second-personal competence and authority. It commits them, that is, to their reciprocal or mutual accountability—their having a shared standing to make claims and demands of each another, and hold one another answerable, as equal, second-personally-competent members of the moral community. As I see it, it is this shared second-personal authority that Jesus implicitly demands recognition of when he puts the question “Why tempt ye me?” to the Pharisees. The question implies that Jesus has standing to demand that the Pharisees not seek to entrap him, as he charges them with doing. In holding them answerable, he demands recognition (respect) of the authority to make this demand.

  Emphasis added.   This will require an act of imagination, since, evidently enough, the moment depicted must be later since Jesus asks for the coin only after his charge, and here the coin is already on the scene. The moment Strozzi seems actually to be portraying is the Pharisees’ response to Jesus’ “marvelous” reply. 4   Or perhaps, as William Lycan and Gerald Postema have suggested to me, Strozzi means it to be ambiguous whether he is or not. 2 3

72  Stephen Darwall This was, of course, a radical idea, most certainly at the beginning of the Christian era, but also for Strozzi’s seventeenth century. It is, I believe, a thought that deeply underlies the contemporary idea of a moral law enshrining universal human rights, as that notion was developed in the doctrine of natural liberty and natural law in Grotius, Locke, and other early modern natural lawyers, Kant’s idea of the equal dignity of persons, and the American and French revolutions, which made the “rights of man” part of the moral and political rhetoric of the West.5 So understood, morality affirms an equal second-personal authority of all persons in both its content and its form. It obligates us to treat one another with mutual respect and gives us all an authority to demand this treatment.6 On this conception—morality as equal accountability, as I call it—moral obligations are what we are accountable to one another, as representatives of the moral community, for doing. And moral rights are what we right holders have the individual authority to demand of others’ treatment of us in particular—“Why tempt ye me?” For these reasons, I  argue in The Second-Person Standpoint that (recognition) respect for persons is fundamentally second personal. It is recognition inter alia of a second-personal authority that persons have as such to claim and demand certain treatment from one another. So far, this claim does not significantly revise the view of respect for persons that I had advanced earlier, when I first distinguished recognition respect from appraisal respect (Darwall 1977). Then I argued that whereas appraisal respect is a form of esteem, recognition respect is something we realize in our treatment of others by regulating our conduct towards them by giving adequate weight to the fact that they are persons.7 Once we have the idea that second-personal authority is part of the dignity of persons, that can simply be slotted into the earlier account of recognition respect: giving appropriate weight to their second-personal authority in our treatment of persons. Nevertheless, I argue in The Second-Person Standpoint that this does not yet adequately capture the full second-personality of recognition respect for persons, as we can see in fact in Strozzi’s painting. As I imagine it, Jesus is not simply calling the Pharisee to take account of his authority to demand that they not entrap him, say, by ceasing to do so for that reason. He is summoning the Pharisee to acknowledge that authority by reciprocating his address—for example, by looking him back in the eye and recognizing it through recognizing him, which the Pharisee can do only through reciprocating Jesus’ second-personal regard of him.8 As I am asking you to imagine   On the last, see Hunt (2007).   There is an important distinction, which we can mostly ignore for our purposes, between the authority we have as individual right holders to demand certain treatment of ourselves in particular (and demand compensation, forgive offenses, and so on) and the authority we have just as members of the moral community to whom all are accountable for respecting rights and complying with moral obligations in general. 7   Hill takes a similar view,“the disposition to give appropriate weight in one’s deliberations to the fact that someone is a person (whether meritorious or not)” (Hill 1998: 284). 8   John Bollard suggests to me that we might also interpret the child at the bottom left corner of the painting in second-personal terms as similarly summoning us, the viewers. 5

6

Respect as Honor and as Accountability  73 the painting, the Pharisee refuses to do this. He continues to look steadfastly at his colleague, leaving the dramatic tension of the painting unresolved. He refuses to hold himself accountable to Jesus. For these reasons, I argued that recognition respect for persons is irreducibly second personal. It cannot be realized in private practical reasoning through adequate weighing of any fact, including the fact of second-personal authority. Rather we fully respect persons only through relating to them second personally, as we might imagine the Pharisee Jesus addresses would do were he to look Jesus back in the eye and say something like, “You’re right. I was trying to catch you out. I apologize,” acknowledging his violation and Jesus’ authority to hold him accountable and thereby holding himself accountable. We respect one another, in other words, by making ourselves mutually accountable. This brings me to the topic of the current chapter. I want to suggest that the tension between the two axes I identified in Strozzi’s painting, the “aspirational” axis between Jesus and the Pharisee to his left and the “actual” axis between the two Pharisees, reflects a tension between two very different kinds of recognition respect for persons. I have already mentioned the kind that I take to be reflected along the aspirational axis. Jesus is attempting to establish reciprocal second-personal respect of one another as having a common second-personal authority. But on our imaginative interpretation, a quite different kind of recognition respect is instantiated in the regard between the two Pharisees, which regard they refuse to bestow on Jesus. This is, significantly, the very kind of respect to which Mark refers in his Gospel when he has the Pharisees say to Jesus that he “regardest not the person of men” and “carest for no man.” Clearly, Mark’s Pharisees do not mean that Jesus is disrespectful or that he is not benevolently inclined toward others. They mean what we mean when we say that the law (or God) is no “respecter of persons”—that is, that Jesus treats people equally without regard to their status or social place.9 In this very different sense (of both “respect” and “person”), one person respects another by recognizing or honoring him as having some specific social role, status, or place that, in principle, not every can have. “Person” here is used in its original sense; “its etymon,” as the OED puts it is “the classical Latin persōna,” “a mask used by a player, character in a play, dramatic role, the part played by a person in life, character, role, [or] position.” Respecting someone in this sense is, roughly, allowing him to play the role he is attempting to play by playing along with him or by bestowing on him a higher status or role, that is, an honor. There are important connections between honor, “face,” and shame. Face, in the relevant sense, is the persona (mask) we present socially. As Erving Goffman brilliantly analyzed, the “presentation of self in everyday life,” is through “face-to-face” “interaction rituals” (Goffman 1959, 1982). When others see and honor the face we wish to present, this enhances our sense of social place, confidence, and pride. To lose face, on   And, of course, without regard to personal interest or personal relations.

9

74  Stephen Darwall the other hand, is to lose one’s social place. When others will not see us as we wish to be seen, or worse, when they show contempt or “look us out of countenance,” this is an occasion for shame, which we naturally express by hiding, by removing our face from public view.10 I will call respect of this second kind honor respect. Honor respect is a species of recognition respect; it is a form of treatment, something we do in recognizing someone in some distinctive way. We “do someone an honor” through this recognition, which honor the person thereby comes to have. Alternatively, someone can be “done an injury” by a failure of such recognition. Honor contempt dishonors its object and accords it low status. The injury is, quite precisely, to the person’s honor, which just is the status that the contempt lowers. Someone can dishonor himself also, as in fact the Pharisees hope they can make Jesus do if they can catch him out, ultimately also in those of “the people.” In all these cases, it is important that honor or dishonor is socially constituted. The status just consists in the common knowledge that it has been publicly recognized by the relevant people. This is an important difference between honor and the dignity of persons as I understand it, namely as entailing second-personal authority (Darwall 2006). So understood, dignity is an irreducibly normative rather than a socially constituted status. Failing to recognize someone’s dignity by refusing to accord her second-personal respect may injure her in some way or other, but it cannot injury her dignity, at least not directly.11 Suppose that the Pharisees had refused to bring Jesus the penny, had not been drawn into his elenchus (“Whose is this image and superscription? And they said unto him, Caesar’s”), and so had not made themselves vulnerable to his “marvelous” reply (“Render to Caesar the things that are Caesar’s, and to God the things that are God’s”). As the situation actually transpires, Jesus gets the better of the Pharisees. But suppose he had not. Suppose that he had been struck speechless by the Pharisees’ temerity or drawn into some self-dishonoring response. Even if his honor had been thereby injured, his dignity would not. His authority to hold the Pharisees’ responsible for attempting to entrap him would still be intact even if went unrecognized. In what follows, I want to explore the differences between these two very different kinds of recognition respect for persons. I shall argue that they underlie and mediate two very different conceptions of social or moral order. Second-personal respect mediates and underlies morality as equal accountability. We respect people in this sense when we enter into relations of reciprocity and mutual accountability with them and regard and treat them as having the same basic second-personal competence and authority that we have. Honor respect, on the other hand, mediates and underlies an

10   In this connection, it is significant that the Greek term prosopon, which Mark translates as “person” in “thou regardest not the person of men,” has the same sense as the Latin persona, that is a mask or face worn by actors in a drama. I am indebted here to John Bollard. 11   So to what, then, is the injury? Perhaps to the person’s sense of his own dignity or his ability to rely on its being recognized, or something similar.

Respect as Honor and as Accountability  75 essentially hierarchical social order of honor. It recognizes someone as having a specific social status that not just anyone can have, and it helps constitute that status in a way that is very different from the relation between the second-personal respect involved in mutual accountability and equal authority or dignity. Montesquieu said that honor is “the principle” of monarchy (Montesquieu 1989: 55). Honor respect from those who enjoy it, most prominently the king himself, constitutes the privileges and entitlements that characterize any order of honor, whether monarchical, military, or academic. Honor respect brings these into existence; it realizes the social relations in which the relevant privileges and entitlements consist. Having honor in this sense is simply for the status to be recognized, that is, to have honor respect from those who have the honor of bestowing it. It is important that honor respect is a kind of recognition rather than appraisal respect. It consists neither in any form of esteem nor in any view of someone as worthy of esteem, nor, for that matter, as worthy of honor. As with recognition respect of any kind, we honor someone by how we conduct ourselves toward her, by regulating our conduct in certain ways. It is something we broadly do. Honor respect is thus unlike any attitude, like esteem, which we may simply have toward a person.12 And it is, in addition, like second-personal respect (and unlike both appraisal respect and other forms of recognition respect) in being essentially social. It calls for, and may often require, a certain uptake from others.13 Nevertheless, though honor respect differs from appraisal respect or esteem in these ways, what they share is that both honor and appraisal respect respond to difference. Just as it is part of the very idea of appraisal respect that different persons deserve differential appraisal respect by virtue of how they conduct themselves, so also is it part of the very idea of honor respect that different persons warrant it differently. Both are tied to what Nietzsche called “rank-ordering, rank-defining value judgments” (Nietzsche 2006: 11).14 I speculate, in fact, that as cultures of honor give way historically to cultures of mutual accountability (equal reciprocity), that is, as the kind of recognition respect for persons that most fundamentally defines a society is as sharing the same basic authority rather than as having fundamentally different statuses, then, where honor respect is neither marginalized nor relegated to certain spheres (the military, the academy, the sports field, and so on), it tends to get psychologized as esteem, and reconceived as responding to different forms of merit. The idea that emerges is that the 12   More carefully, recognition respect can also refer to the attitude of being disposed to regulate ourselves in the appropriate ways. Appraisal respect can of course show itself in our conduct also, but it is not, in the first instance, a disposition to do anything in particular. 13   We should bear in mind though that although honor and second-personal respect are social in this way, it is true only of honor respect and not of second-personal respect that what it recognizes is essentially social, status, as opposed to dignity. 14   Nietzsche makes this remark about the aristocratic ethos (“good”/“bad”) that he contrasts with the morality of the herd (“good”/“evil”). As I suggest below, it is useful to interpret Nietzsche’s contrast between forms of responsibility involved in these different ethical frameworks as a contrast between honor respect and second-personal respect.

76  Stephen Darwall only kind of warranted differential respect for persons as such is differential appraisal respect or esteem for their differential merit, for how people conduct themselves, that is, how they exercise their shared moral competence.15

Recognition Versus Appraisal Respect It will help to mark more carefully the distinction between recognition and appraisal respect.16 Appraisal respect is the attitude we refer to when we say that we have more or less respect for someone, either as a person or in some specific capacity (as a philosopher, say) or to have gained or lost respect for someone. In these contexts, “respect” refers to a kind of esteem. Of course, esteem can take forms we would never call “respect.” Esteem can be based on any qualities we admire, look up to, envy, or wish to acquire or emulate, and many such excellences cannot intelligibly be objects of respect, at least, not without some long and unobvious story. The esteem we call “respect”—appraisal respect—is an assessment of someone’s conduct or character or of something that somehow involves these. Appraisal respect for someone as a person is moral esteem: approbation for her as a moral agent. We can have appraisal respect for people in more specific capacities also, but though this differs from moral esteem, it still concerns conduct in the relevant capacity. Thus, although appraisal respect for someone as a tennis player differs from respect for her as a person, neither is it the same thing as esteem for her tennis-playing skills or achievements or for her as a tennis player. Appraisal respect is esteem that is merited or earned by conduct or character. By contrast, recognition respect is no form of esteem at all. When we think that even scoundrels have a dignity that entitles them to respectful forms of treatment (in holding them accountable, for example), we clearly have something other than esteem in mind. The idea is not that personhood is somehow an admirable quality: What is in play is recognition rather than appraisal. The object of recognition respect is not excellence or merit; it is dignity or authority, in the case of second-personal respect, or status, in the case of honor respect. Recognition respect concerns not how something is to be evaluated or appraised, but how our relations to it are to be regulated or governed. Broadly speaking, we respect something in the recognition sense when we give it standing in our relations to it. As I noted above, since the authority that persons have as such is fundamentally second personal, respect for it must be second personal also; it must involve acknowledgment. There are forms of recognition respect, however, that need not be second personal, whenever, in fact, the relevant authority or status the respect recognizes is not second personal. In these cases, the authority or status can be recognized without any form of

  Thus “moralizing” respect in general in the way to which Nietzsche objects.   In what follows, I draw on my discussion in Darwall 1977 and 2006.

15

16

Respect as Honor and as Accountability  77 acknowledgment. In many of these instances, moreover, the authority can be merited or earned. But even when that is so, respect for the authority differs from esteem for any form of merit. Consider epistemic authority, for example. One can respect the knowledge or wisdom one overhears in another person’s solitary musings, for example, and regulate one’s private reasoning by them without even imaginative acknowledgment. But though epistemic authority must be earned, recognition respect for it differs from esteem for epistemic virtues. The latter shows itself in and partly just is a positive appraisal of someone as a cognizer or of her contributions to collaborative inquiry. The former, however, is realized in our epistemic conduct in relation to him, for example, by giving his views weight or authority in deciding what to believe ourselves.

Second-Personal Respect In this section, we shall consider in more detail why respect for the dignity of persons is second personal in two distinct, but related, senses: the authority it recognizes is second personal, and appropriate recognition of that authority comes from the second-person standpoint we take up when we relate to someone. Consider the former aspect first. Notice that someone might accept first-order norms that we take to structure the dignity of persons and regulate himself scrupulously by these standards without accepting anyone’s authority to demand that he do so. If such a person were to refuse to rape, pillage, and steal out of, for example, the conviction that such conduct was beneath him, it should be clear that his forbearance would not constitute respect for the persons. He would be seeing the conduct as beneath his dignity, not as required by theirs. Another way of seeing the point is to note the close connection between recognition respect for persons and “respect for their rights” (Thomson 1990:  210–11). As Feinberg argued, someone can have claim rights only if she has the standing or authority to claim them, and “it is claiming that gives rights their special moral significance” (Feinberg 1980: 151). To respect the dignity of persons, therefore, we must recognize their rights not just in the sense of avoiding their violation, but by being disposed to avoidance because this is their right, that is, because of their authority to claim or demand this. This is the first way in which respect is second personal:  it recognizes a second-personal authority to address claims or demands to others. But authority of this kind is appropriately recognized second personally also, through relating to others in a way that acknowledges their second-personal authority. Not only is the authority second-personal respect recognizes second personal, but that authority’s distinctive form of recognition is second personal also. We realize second-person respect by relating to one another in ways that acknowledge each other’s standing to demand, remonstrate, resist, charge, blame, resent, feel indignant, excuse, forgive, and so on. Since accountability is second personal in both senses, respect for persons is as well. Since, as

78  Stephen Darwall Feinberg says, “having rights enables us to ‘stand up like men’ [and] look others in the eye,” we recognize this standing when we look others in the eye in return and recognize our common dignity reciprocally (Feinberg 1980: 151). Now it is possible to respect the fact that someone has second-personal authority— for example, that she can claim certain conduct as her right—without yet relating to her in the second-personal way that genuine respect for her as a person involves.17 To see this, consider, first, an analogous case in theoretical reasoning. Suppose you are disinclined to trust someone’s judgment, but that, on reflection, you reject your distrust and think you should take his testimony as evidence. When he affirms p, this does not increase your inclination to believe p until you recall your considered view, which corrects your distrust. Here it seems natural to say that although you respect the fact of his epistemic authority, you do not, yet anyway, treat him as an authority and so respect him in this sense. Consider now a fully second-personal case. Imagine a parent who is not yet fully disposed to regard his college-age daughter as an independent person with second-personal authority. We can well imagine that he believes, on reflection, that he should so regard her, indeed, that she has the same claim to respect that he does. But parental habits die hard. When she says that she doesn’t want to do something within her discretion, which he believes to be for her greater good, he is disinclined to defer in his own deliberations until he reminds himself of his considered conviction. When she calls him on his paternalism, he is not initially disposed to hold himself responsible, but defends himself until he reminds himself that she is right; she really is no less entitled to respect than he is. Here again, it seems that although the parent respects the fact that his daughter has the same dignity he does, he does not yet fully respect her in the sense of relating to her on terms of equal respect. He does not yet accord her second-personal authority in his relations with her. Second-personal respect is thus an irreducibly second-personal way of relating to someone. To respect someone as a person is therefore not just to regulate one’s conduct by the fact that one is accountable to him, or even just to acknowledge the truth of this fact to him; it is also to make oneself or be accountable to him, and this is impossible outside of a second-personal relation. This, I believe, is the deepest resonance of respect’s root, respicĕre (to look back), at least when it comes to respect for persons. To look back at someone and return his addressing regard just is to establish second-personal relationship and acknowledge the other’s second-personal authority. To look someone in the eye is to make oneself and one’s eyes, the “windows” of one’s soul, vulnerable to him, and in both directions. One gives the other a window on one’s responses to him but also makes oneself vulnerable to his attitudes and responses by empathy. When address is an attempt to hold someone accountable, returning the

  In what follows, I am indebted to Amanda Roth.

17

Respect as Honor and as Accountability  79 other’s address, being open to his claims, including by giving him this window, is itself part of holding oneself accountable to him.

Second-Personal Respect and Answerability When we hold someone accountable, we implicitly address a claim or demand on his conduct. We hold him to the conduct, either explicitly, through legal and social practices, or implicitly, with the states of mind that Strawson dubbed “reactive attitudes,” such as indignation, resentment, blame, and guilt (Strawson 1968). We take ourselves to be in a position, not just to point to good reasons why someone should act or should have acted, in the way an adviser might, but to claim or demand the conduct.18 But neither do we view this as mere force; we do not think our demands are naked demands. We assume them to be legitimate, that they are backed by an authority that the person we hold responsible can and should freely accept and guide himself by. Moreover, when we hold someone morally responsible, we commit ourselves also to the idea that he has the very same competence and authority to hold himself (and us) responsible also (Darwall 2006). What is at issue in moral responsibility is culpability, that is, whether a person did something wrong (violated his moral obligation) and, if so, whether he is to blame for doing so (had no excuse). I follow Mill, Strawson, and a number of more recent writers, in holding that the ideas of moral obligation and wrong are related to that of moral responsibility analytically. It is a conceptual truth that an act is morally obligatory if, and only if, it is one that the agent would warrantedly be held responsible and blamed for not doing unless he had a valid excuse.19 This builds answerability and “justification to” into the very ideas of moral responsibility and moral obligation. If someone is able to justify his conduct to us by showing that he had good and sufficient reasons for it and that it would be unreasonable to expect someone in his position to act otherwise, then he shows that his conduct was outside the scope of the morally obligatory. Alternatively, he might show that although his action violated a moral demand, it should nonetheless be excused or his responsibility mitigated for some reason, say, ignorance or duress. If, however, someone cannot answer for himself in either of these ways, then what he did is culpable and blame is warranted. According to morality as equal accountability, moral responsibilities, obligations, and rights all involve what Rawls might have termed a “model conception” of the moral community as a group of mutually accountable persons with the authority to

  Cf. Hobbes’s distinction between “command” and “counsel” (Hobbes 1983: XIV.1).   Mill is the classic source for this view: “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.” Unless we think “blame” is warranted, Mill continues, we think, “it is not a case of moral obligation” (Mill 1998: ch. 5, ¶14). See also Gibbard (1990: 42); Brandt (1979); Skorupski (1999: 29, 142). 18

19

80  Stephen Darwall hold themselves to moral demands (Rawls 1980, 1993). Moral obligations, on this conception, are what we have standing as such members to demand of one another and ourselves. Accordingly, when people violate moral obligations and rights, they fail adequately to respect this authority. Reciprocal, second-personal respect thus mediates moral responsibility. In holding someone morally responsible, we address demands in the name of an authority that our addressee and we share; we acknowledge this shared standing in holding them morally responsible and call on them to acknowledge it also. Morality as equal accountability differs fundamentally in these regards from ways of holding responsible and demanding honor respect within an honor culture, as we shall see in more detail presently. When someone’s honor has been injured or insulted, retaliation of some kind is called for; unavenged dishonor stands. And even when, say, norms of gentlemanly conduct are observed, a challenge or demand for satisfaction, that is, for renewed honor respect, does not itself necessarily honor its object as an equal; it is not implicitly reciprocal in the way that holding someone morally answerable is.20 Holding someone morally responsible returns second-personal respect for second-personal disrespect, whereas avenging dishonor returns honor disrespect for honor disrespect.

Honor Respect and Honor Cultures What makes reactive attitudes like indignation, resentment, and guilt distinctively appropriate for moral responsibility is their conceptual tie to second-personal address and to blame. To feel indignation, for example, is to feel as if someone is to blame for wrongful conduct and therefore appropriately held answerable, even if only by being subject to reactive attitudes guilt from himself (guilt) and others (moral blame). The point is not that it seems as if a sanction of some kind is desirable or that that would make for a more valuable or fitting whole (poetic justice). Indignation invariably includes some sense of authoritative demand that may be absent from feelings of desirability or fittingness. Whereas the natural response to contempt (honor disrespect) is shame (the experience of loss of face), the appropriate response to a warranted charge of second-personal disrespect lodged through blame or resentment is guilt, a felt acknowledgment of having violated some authoritative demand. To feel guilty is to feel as if one is appropriately blamed (to blame) and held responsible for something one has done. It is an implicit acknowledgment of one’s blameworthiness that recognizes both the grounds of blame as well as the authority to level it (even if only “to God”). To feel guilt, consequently, is to feel as if one has the requisite capacity and standing to be addressed as responsible. 20   It may nonetheless be the case that in the early modern period, practices of gentlemanly honor formed something of a transition to a conception of equal dignity and morality as equal accountability. On this point, see LaVaque-Manty (2006). See also Anderson (2008). I am indebted here to discussion with Colin Bird.

Respect as Honor and as Accountability  81 And guilt’s natural expressions are themselves second personal—confession, apology, making amends, and self-addressed reproach. It is an important point about the nature of second-personal respect, and an important contrast with honor respect, that second-personal respect lacks a straightforwardly contrary attitude of the kind honor respect has. The contrary of honor respect is contempt or disdain, treating someone as inferior, looking down on him. Honor respect and contempt play contrary roles in recognizing and constituting evaluative hierarchies of honor; they are “rank-ordering, rank-defining,” helping to establish a shared sense of the noble and base. But shared by whom? Unlike reactive attitudes, contempt and disdain do not implicitly address their objects or call upon them for reciprocation. Contempt may have an addressee, of course, but it need not, and even when it does, its addressee is very frequently not its object. We more naturally express contempt to peers we hold competent to judge the contemptible object as we do. A reactive attitude, however, registers a charge with its object and holds him responsible in way that, as Strawson puts it, “continu[es] to view him as a member of the moral community” (Strawson 1968: 93). And it calls for acknowledgment; it comes with an implicit RSVP. Consider the difference between a disdainful rolling of the eyes and a look that expresses resentment or indignation (“looking daggers”). The latter addresses its object (with an RSVP) in a way that the former need not. For example, consider the “staredown” that the Italian ice dancer, Barbara Fusar Poli, gave her partner Maurizio Margaglio after he dropped her during the Ice Dancing competition at the 2006 Winter Olympics in Torino. According to one account, “When the music stopped, the two stood facing each other for nearly a minute at center ice, Fusar Poli steaming as her eyes burned through Margaglio” (Macur 2006). Or as one caption had it: “Barbara Fusar Poli to partner Maurizio Margaglio: Look me in the eye and tell how you dropped me.”21 Suppose, however, that Fusar Poli had disdainfully rolled her eyes after the fall and skated off in disgust. Any expressive address in that instance would most likely have been to the cognoscenti off ice: “Can you believe what he just did?” Resentment and indignation are not opposites of second-personal respect. To the contrary, they implicitly involve respect by recognizing their objects’ authority as equal members of the moral community and issuing a charge within an assumed second-personal space of reciprocity. But isn’t moral contempt the opposite of second-personal respect? Not really, in my view. To be sure, contempt can function as a moral emotion, but when it does, it seeks not so much to exclude its object from the moral community altogether as from morally optional relations like partner, friend, colleague, and so on. And this can itself be a way of holding someone accountable as a member of the moral community (Mason 2003). Viewing someone 21   In the Internet edition of the San Diego Union-Tribune: .

82  Stephen Darwall as lacking the competence to be held morally responsible at all, like a very young child or an advanced Alzheimer’s patient, is something else altogether, not moral contempt. Again, as guilt is the emotion that responds to (apparently) warranted indignation, resentment, and blame, so shame is the emotion that responds to (apparently) warranted disdain or contempt. It follows from what we said in the last paragraph, however, that though shame is the response as if to warranted to honor disrespect, guilt is not the response as if to warranted second-personal disrespect. It is the response to an apparently warranted respectful charge that one has disrespected the other. Unlike honor respect, there is no such thing as warranted second-personal disrespect. Just as the feeling of being addressable (so a suitable object of second-personal respect) is part of the experience of guilt, so also does the feeling of shame clue us to what is at issue in honor respect and contempt, namely, presentability or face—having an honorable persona in a social space. Honor respect consists in regarding and treating someone as having such a status, acting toward her in a way that honors her self-presentation, whether grandly, by doing or giving her an honor of some sort, or just by “supporting” her in her role, as we say also of (professional) actors. The currency of contempt, by contrast, is the put down, taking someone down a peg. Consider, in this vein, Samuel Johnson’s catty remarks about Colley Cibber, playwright, actor, and Poet Laureate in eighteenth-century England: “Colley Cibber, Sir, was by no means a blockhead; but by arrogating to himself too much, he was in danger of losing that degree of estimation to which he was entitled. His friends gave out that he intended his birth-day Odes should be bad: but that was not the case, Sir; for he kept them many months by him, and a few years before he died he shewed me one of them, with great solicitude to render it as perfect as might be, and I made some corrections, to which he was not very willing to submit. I remember the following couplet in allusion to the King and himself: ‘Perch’d on the eagle’s soaring wing, The lowly linnet loves to sing.’ Sir, he had heard something of the fabulous tale of the wren sitting upon the eagle’s wing, and he had applied it to a linnet. . . . Grand nonsense is insupportable.” (Boswell 1998: 285)

Perhaps the most honored dispenser of literary honors and brickbats of the period, Johnson laces his remarks with the language of honor and contempt. By “arrogating to himself too much” in Johnson’s view, Cibber sets himself up for Johnson’s “grand nonsense is insupportable.” Johnson has a measure of esteem for Cibber, but will not support or honor the measure that Cibber presents himself as meriting. Already in eighteenth-century England, elements of a pure honor culture had given way to psychologized versions:  honor has been replaced, at least partly, by esteem and dishonor by disesteem. It is arrogating to himself “a measure of estimation” beyond his due that makes Cibber subject to Johnson’s scorn. These remarks are almost snugly within a culture of honor, but not quite. Johnson does say not just

Respect as Honor and as Accountability  83 that Cibber is worthy of less esteem than he himself thinks, but that he “arrogates to himself ” too much, that is, that he presents himself as meriting more; he presumes on others’ esteem. And Johnson speaks of Cibber’s relatively unwilling submission to his correction. But even so, it is esteem that Johnson accuses Cibber of arrogating, not honor.22 In my view, we can see in Johnson’s remarks (in 1762) effects of the emerging idea that all persons are entitled to equal respect of their equal dignity.23 The idea that different orders of men have inherently different privileges warranting differential honor respect was in the process of becoming “insupportable” itself, at least in certain circles.24 In closing, I  want to note some important contrasts between the ways in which honor respect and its correlative forms of responsibility mediate honor cultures, on the one hand, and the role of second-personal respect and answerability in morality as equal accountability, on the other. I suggest that these differences at least partly concern different attitudes that these different conceptions of social order take toward relevant truth in their respective domains; that is, whether they regard that truth to be publicly constituted, as in honor cultures, or rather as independent and subject to public inquiry, as in morality as equal accountability. Honor facts are made true by actual social relations, by who honors or dishonors whom, by how people respond to challenges to their honor, and so on. An unavenged insult or dishonor, or a challenge unsuccessfully repelled, thereby makes it the case that its object has lower status or honor. It puts forward no putative truth claim, say, that its object is worthy of dishonor, that might occasion an inquiry. Its object is given no charge to which he may or must answer by providing evidence. A challenge rather requires some action that can remove or annul it; it may be no help whatsoever to provide evidence that it was unwarranted.25 In his insightful history of honor cultures, James Bowman gives an especially vivid example from Mallory’s Morte d’Arthur (Bowman 2007. Virtually everyone in Camelot knows that Launcelot is violating his oath of fealty to Arthur by having an affair with 22  See Lovejoy 1968 for the importance of “approbativeness” (the desire for others’ approval) in eighteenth-century Europe and, of course, Rousseau’s second Discourse for the related idea of amour proper (Rousseau 1997). 23   In Darwall 2006, I suggest that Adam Smith, whose Theory of Moral Sentiments was published in 1759, might fairly be regarded as the first “second-personal philosopher.” Lynn Hunt notes that Rousseau first used the phrase “the rights of man” in 1762 in Émile (Hunt 2007: 70). 24   On the ways in which honor cultures have been undermined by internal critique within an honor framework (for example, by ridiculing what was formerly deemed honorable, see Appiah 2010. Appiah 2010 and Waldron 2012 propose that the idea of equal dignity be understood within broadly honor terms as elevating all to an equally high status of peers. An implication of the current chapter, in my view, is that their proposal lacks the second-personal element that is essential to the idea of the dignity of persons. Roughly, mutually respecting and accountable persons do not merely look upon one another, third personally, as peers, they look to one another by holding themselves answerable to one another. They do not merely defer to others in matters of right, but take others to have the standing to claim their rights and hold them answerable for respecting them. 25   Imagine, for example, responding to the challenge that “you ain’t got nuttin’ ” by producing a resume of previous conquests.

84  Stephen Darwall Guenevere. But no one dares to speak of the liaison publicly to Launcelot’s face, since that would “invite Launcelot, whose fighting prowess makes him the most honorable of all knights, to call him a liar,” and “the charge of lying against any knight would in turn have obliged that knight to challenge Launcelot to a single combat to the death, or else to be forever dishonored himself as one who has allowed himself to be ‘given the lie’ . . . without a fight” (Bowman 2006: 42). As Bowman observes, “Malory portrays a system of honor in which what is known privately by everyone nevertheless does not matter or even exist, in some important sense, so long as it is not spoken of publicly” (Bowman 2006: 42).26 One might think that traditional honor codes must actually value truth and truth telling, since they generally make lying the most serious offense. But the fact is actually the reverse, as in Mallory’s Camelot. There intimidation is not only an obstacle to inquiry; it actually determines socially relevant honor “truths” in a way that is impervious to the actual facts. Were a knight to respond to a challenge from Launcelot as though it were a charge he could answer with evidence, he would dishonor himself. In the Introduction to his Treatise of Human Nature, David Hume prepares the way for his “attempt to introduce the experimental method of reasoning into moral subjects.” He notes that the experimental method had only begun to be systematically applied to “natural subjects” (that is, to “natural philosophy” or science) with Bacon and that in the century since, “some late philosophers in England . . . have begun to put the science of man [‘moral subjects’] on a new footing,” so that it is reasonable to expect that a Newton of the moral sciences might now appear (an honorific Hume clearly hopes his readers might bestow on him). The point I want to note in closing, however, is Hume’s subsequent remark that “these improvements in reason and philosophy can only be owing to a land of toleration and liberty” (Hume 2000: 5). A liberal culture in which people are answerable to one another for their conduct is, perforce, one that values public inquiry, getting at the truth behind social appearances and “speaking truth to power.” We might thus think of morality as equal accountability as a secular version of Mark’s characterization of Jesus’ teaching as “the way of God in truth.” When we respect “not the person of men” in the sense of honoring their persona or social appearance, but rather respect all equally in the second-personal sense, we commit ourselves to a mutual accountability that implicitly honors fact over appearance.

26   Cf. Montaigne: “There are two sets of laws, the law of honour and the law of justice which are strongly opposed in many matters (the first condemns an unavenged accusation of lying; the other condemns the revenge; a gentleman who puts up with an insult is, by the laws of arms, stripped of his rank and nobility: one who avenges it incurs capital punishment; if he goes to law to redress an offence against his honour, he is dishonoured; if he acts independently he is chastised and punished by the Law . . .” (quoted in Bowman 2006: 67).

Respect as Honor and as Accountability  85

References Anderson, Elizabeth. 2008. “Emotions in Kant’s Later Philosophy: Honor and the Phenomenology of Value,” in Monika Betzler, ed., Kant’s Virtue Ethics. New  York and Berlin:  de Gruyter. Appiah, Kwame Anthony. 2010. The Honor Code. Princeton: Princeton University Press. Boswell, James. 1998. Life of Johnson. Oxford: Oxford University Press. Bowman, James. 2007. Honor: A History. New York: Encounter Books. Brandt, Richard. 1979. A Theory of the Good and the Right. Oxford: Oxford University Press. Darwall, Stephen. 1977. “Two Kinds of Respect.” Ethics, 88: 36–49. Darwall, Stephen. 2006. The Second-Person Standpoint: Morality, Respect, and Accountability. Cambridge, MA: Harvard University Press. Feinberg, Joel. 1980. “The Nature and Value of Rights,” in Rights, Justice, and the Bounds of Liberty. Princeton: Princeton University Press. Gibbard, Allan. 1990. Wise Choices, Apt Feelings. Cambridge, MA: Harvard University Press. Goffman, Erving. 1959. The Presentation of the Self in Everyday Life. New York: Anchor Books. Goffman, Erving. 1982. Interaction Ritual: Essays on Face-to-Face Behavior. New York: Pantheon Books. Hill, Thomas E., Jr. 1997. “Basic Respect and Cultural Diversity,” in Grethe B. Peterson, ed., The Tanner Lectures on Human Values 18. Salt Lake City: University of Utah Press, pp. 3–37. Hill, Thomas E., Jr. 1998. “Respect for Persons,” in Edward Craig, ed., Routledge Encyclopedia of Philosophy, vol. 8. London: Routledge. Hobbes, Thomas. 1983. De Cive, the English Version, Entitled in the First Edition, Philosophicall Rudiments Concerning Government and Society, ed. and trans. Howard Warrender. Oxford: Clarendon Press. Hume, David. 2000. A Treatise of Human Nature, ed. David Fate Norton and Mary J. Norton. Oxford: Oxford University Press. Hunt, Lynn. 2007. Inventing Human Rights: A History. New York: W. W. Norton & Company. LaVaque-Manty, Mika. 2006. “Dueling for Equality: Masculine Honor and the Modern Politics of Dignity.” Political Theory, 34: 715–40. Lovejoy, Arthur O. 1968. Reflections on Human Nature. Baltimore: Johns Hopkins University Press. Macur, Juliet. 2006. “After Staredown, a Détente for an Italian Pair.” New York Times, February 21. Mason, Michelle. 2003. “Contempt as a Moral Attitude.” Ethics, 113: 234–72. Mill, John Stuart. 1998. Utilitarianism, ed. Roger Crisp. Oxford: Oxford University Press. Montesquieu, Baron de. 1989. The Spirit of the Laws. Cambridge: Cambridge University Press. Nietzsche, Friedrich. 2006. The Genealogy of Morals, ed. Keith Ansell-Pearson, trans. Carol Diethe. Cambridge: Cambridge University Press. Rawls, John. 1980. “Kantian Constructivism in Moral Theory.” The Journal of Philosophy, 77: 515–72. Rawls, John. 1993. Political Liberalism. New York: Columbia University Press. Rousseau, Jean-Jacques. 1997. Discourses on the Origin and Foundation of Inequality Among Men. In The Discourses and Other Early Political Writings. Cambridge: Cambridge University Press. Skorupski, John. 1999. Ethical Explorations. Oxford: Oxford University Press.

86  Stephen Darwall Strawson, P. F. 1968. “Freedom and Resentment,” in P. F. Strawson, ed., Studies in the Philosophy of Thought and Action. London: Oxford University Press, pp. 71–97. Thomson, Judith. 1990. The Realm of Rights. Cambridge, MA: Harvard University Press. Waldron, Jeremy. 2009. Dignity, Rank, and Rights. Tanner Lectures at University of California, Berkeley. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1461220. Waldron, Jeremy. 2012. Dignity, Rank, and Rights (Berkeley Tanner Lectures), ed. Meir Dan-Cohen. Oxford: Oxford University Press.

PA RT I I

Practical Reason

4 Hypothetical Imperatives Scope and Jurisdiction Mark Schroeder

1  Hypothetical Imperatives vs. the Hypothetical Imperative The last few decades have given rise to the study of practical reason as a legitimate subfield of philosophy in its own right, concerned with the nature of practical rationality, its relationship to theoretical rationality, and the explanatory relationship between reasons, rationality, and agency in general. Among the most central of the topics whose blossoming study has shaped this field, is the nature and structure of instrumental rationality, the topic to which Kant has to date made perhaps the largest contribution, under the heading of his treatment of hypothetical imperatives. After forty years, Tom Hill’s 1973 article, “The Hypothetical Imperative,” remains one of the best entrees into the issues surrounding instrumental rationality, as well as the main voice of probably the most mainstream interpretation of Kant’s own view of hypothetical imperatives.1 Hill’s article offered not only an interpretation of Kant’s theory of hypothetical imperatives, but a general account of why, in Kant’s view, demonstrating the possibility of a categorical imperative requires a synthetic argument of the kind that he proceeds to give in section 3 of the Groundwork and in the Critique of Practical Reason—showing that understanding Kant’s view of hypothetical imperatives is essential for understanding the structural issues in his practical philosophy more generally. The main theme of Hill’s article is that just as Kant’s moral philosophy relies on an important distinction between categorical imperatives and the Categorical Imperative, 1   My own engagement with Hill’s article actually led to both of my first two publications in philosophy, including both my not-so-creatively titled dissent, “The Hypothetical Imperative?,” and “The Scope of Instrumental Reason,” whose origin was as a splinter off the other paper.

90  Mark Schroeder Kant’s broader practical philosophy also relies on an equivalent distinction between hypothetical imperatives and the Hypothetical Imperative (hence the title).2 For Hill’s view it is important not only that there is something called “the Hypothetical Imperative,” but that it is itself an imperative. Whereas individual hypothetical imperatives, such as “if you want to lose weight, then count your calories” and “if your aim is to graduate, then you ought to attend your classes” enjoin particular, concrete, actions to agents with particular ends, according to Hill the Hypothetical Imperative says simply: “take the necessary means to your ends.” Rather than enjoining particular concrete ends, it enjoins simply taking the necessary means to your ends, whatever those are, and rather than addressing only the people with one end or another, it addresses every rational agent as such.3 It will be helpful to compare imperatives, which after all for Kant are simply the expression of laws to imperfect wills which do not necessarily obey them, to more mundane laws. One important feature of laws is that they have jurisdictions. For example, in the state of New York, it is illegal to turn right at a red light. The jurisdiction of that law is drivers in New York, and what it prohibits is turning right on red. In general, anyone who is simultaneously a driver in New York and is turning right on red is in violation of this law, but it is important to appreciate that being a driver in New York and turning right on red make different contributions to this fact. If you are a driver in New York and you don’t turn right on red, then you are complying with the law, whereas if you are a pedestrian in New York or a driver in Buenos Aires or Cairo, the law simply doesn’t apply to you. The reason why drivers in Cairo who turn right on red aren’t in violation of New York traffic laws is that the New York state legislature doesn’t have jurisdiction over drivers in Cairo—not that it does have jurisdiction, but they are in compliance. According to Hill, the significance of the Hypothetical Imperative (in the singular, with capitals) is that its jurisdiction is just as universal as the Categorical Imperative— it has jurisdiction over absolutely all rational agents, no matter what they are like. The reason this shapes Hill’s broader interpretation of Kant is that this makes the Hypothetical Imperative sound an awful lot like Kant’s description of categorical imperatives—which are universal laws, with jurisdiction over every rational agent, no matter what she is like. For on one highly eligible interpretation of Kant it is precisely the universal jurisdiction of categorical imperatives which makes them so hard to argue for and explain, but Hill cannot accept this interpretation, for if it were right,

2   Curiously, Kant actually never unambiguously refers to “the” hypothetical imperative, in the singular, as he does to the Categorical Imperative. Hill claims otherwise, twice citing a single passage in which the term “the hypothetical imperative” appears; unfortunately, the passage is taken out of context, and comes from the second half of a long German sentence whose first half refers alternately to “the imperative of prudence” and “the technical imperative,” and in which the most natural reading of “the hypothetical imperative” in the second half is as a bound reading, so that it refers alternately to either the imperative of prudence or the technical imperative. Nothing much should hang on this issue. 3   Compare also pp. 51–7 of Hill and Zweig (2002), where this view is reiterated.

Hypothetical Imperatives  91 then hypothetical imperatives would be just as hard to argue for and explain, since they derive from the Hypothetical Imperative, which shares with categorical imperatives the feature of having universal jurisdiction.4 In earlier work I argued that Hill’s textual interpretation of Kant’s account of hypothetical imperatives is under-motivated and fails to account for a number of interesting facts about the text, including the details of Kant’s analytic argument for hypothetical imperatives, the significance of his distinction between problematic and assertoric hypothetical imperatives, and changes in his treatment over time, particularly leading up to the publication of the Critique of Judgment.5 My goal here is not to rehash those arguments, but to explain why it is so important—not only for Kant, but for contemporary efforts to explain requirements either of rationality or of morality—what we take the underlying jurisdiction of those requirements to be. My main focus will be on the constraints on any theory that are placed by the answer we give to this question, rather than on interpretive questions about Kant, but I will try to indicate along the way why it is fruitful to understand Kant as having been motivated by precisely the issues that I will be trying to make vivid, here. The remainder of the chapter consists of five sections; in section 2 I’ll introduce a mostly familiar dialectic framed in contemporary terms about the structure of hypothetical imperatives, framed in terms of their scope, and try to get clearer about how they are related to one another. One of the most important points to be made in that section is that in order for questions about scope to be interesting, it helps a great deal to impose certain constraints on the proper interpretation of the normative concept that we use to formulate them. In section 3 I’ll argue that the law gives us a model for the kind of normative concept that allows the interesting scope question to be asked, and that this is because it makes room for non-vacuous distinctions in jurisdiction. Then in section 4 I’ll discuss the relationship of scope to jurisdiction, and lay out two broadly Kantian concerns one might have—and which I think we should have—about how to explain wide scope rational requirements with universal jurisdiction. In section 5 I’ll offer a positive, Kant-inspired but not necessarily strictly Kantian picture which, by giving one interpretation of what makes narrow-scope

4   Compare Hampton (1998: 165–6), who argues on just these grounds that Kant’s view that hypothetical imperatives are easier to explain than categorical imperatives is incoherent (boldface added for emphasis):

Kant’s position on the nature of hypothetical imperatives must be construed (contra his explicit wishes) such that understanding the bindingness of a hypothetical imperative is no easier than understanding the bindingness of a categorical imperative. My interpretation cannot save Kant’s belief that the former is more straightforward than the latter; indeed, my argument is that Kant’s belief is wrong. The only way to analyze Kant’s analyticity claim is to do so in a way that locates in hypothetical imperatives the same mysterious objectivity that attends the categorical imperative. Even more strikingly, I have argued that the force of hypothetical imperatives is dependent on, and is at least in part constituted by, the force of some antecedent categorical imperative that is in part definitive of instrumental rationality. 5

  Schroeder (2005).

92  Mark Schroeder statements of conditional requirements of rationality true, gives us a rich and satisfying alternative to the concerns raised in section 4. Finally, in section 6 I’ll close by revisiting my original disagreement with Hill, and suggesting that in some (particularly important!) respects, our readings of Kant may not actually be so far apart.

2  Instrumental Rationality: Some Possible Views The topic of instrumental rationality explores how what we ought to do depends on our ends. It is actually plausible that there are in fact several different closely related topics which can loosely be described in these terms and have sometimes been confused,6 but glossing over complications, we may take the basic datum to be that at least in paradigmatic cases, there is something going wrong with someone who intends some end, believes that some means are necessary for that end, and has no intention whatsoever for the means. To simplify things by avoiding quantifying over ends and means, I will assume that e and m are arbitrary actions, that Ex is the proposition that x intends to do e, Mx the proposition that x intends to do m, and Bx the proposition that x believes that m is a necessary means to e. Also for simplicity, and glossing over several complications, I will follow a common convention in deontic logic, and assume that we can represent the claim that x ought to do a as OUGHTx(x does a), treating “OUGHTx” as a propositional operator.7 These simplifications make it possible to distinguish clearly between four important kinds of view from the literature about instrumental rationality, distinguished from one another by the scope of the “ought”: Narrow Intermediate Wide Myth

∀x (Ex&Bx → OUGHTx(Mx))

∀x (Bx → OUGHTx(~Ex ∨ Mx))

∀x (OUGHTx(~Ex ∨ ~Bx ∨ Mx))

∀x (OUGHTx(~Ex) ∨ OUGHTx(~Bx) ∨ OUGHTx(Mx))

Controlling for differences of opinion about exactly how “OUGHTx” is to be understood, I’ve advocated a view with the structure of Narrow in Schroeder (2009), Jonathan Way has advocated an attractive view with the structure of Intermediate in Way (2010), Joseph Raz (2005) and Niko Kolodny (2008a, 2008b) have advocated views something along the lines of Myth, and Wide has been widely endorsed, including by Hill (1973), Darwall (1983), and Broome (1999), among other prominent proponents. Proponents of Wide sometimes say that only Wide is uncontroversial, but that depends on how we interpret “OUGHTx.” If we interpret “OUGHTx” as meaning “x won’t do everything that x rationally ought to do unless,” then Wide is actually entailed by   Including by me; see Schroeder (2004).   I don’t believe this simplification affects anything important in this chapter, but for some of the reasons why it is arguably an oversimplification, see Schroeder (2011). 6 7

Hypothetical Imperatives  93 each of the other views. After all, if x won’t do everything that x rationally ought to do unless ~Ex, it is clear that x won’t do everything that x rationally ought to do unless ~Ex ∨ ~Bx ∨ Mx. So on this reading, the first disjunct of Myth entails Wide, and similar reasoning goes for the other disjuncts. Similarly, if the truth of Ex&Bx guarantees that x won’t do everything that x ought unless Mx, x certainly won’t do everything that x ought unless either Mx or it is not the case that Ex&Bx—so on this reading, Narrow also entails Wide—and similar reasoning goes for Intermediate. So if we interpret “OUGHTx” in this way, Wide isn’t controversial; it is just a description of the data that everyone wants to be able to explain.8 Now, the problem is not that we cannot formulate some dispute between Wide and Narrow scope views about instrumental rationality, by use of a normative concept like “x will not do everything that x rationally ought to do unless.” Using such a concept, as I’ve noted, all sides agree that Wide is true. But there can still be a very real debate about whether Wide is the only true thing to be said, at this level of generality. However, prominent objectors to the Wide scope view in the literature—including Niko Kolodny, Joseph Raz, and myself—have not presented their view as granting that Wide is true but insisting that one of the other theses is also true. On the contrary, many remarks of these critics suggest that they have been trying to argue directly against Wide. So it seems that charity requires taking them to interpret “OUGHTx” as meaning something else. And indeed, though Wide is uncontroversial when we interpret “OUGHTx” in this way, what is controversial is what makes this true. On this interpretation of Wide, for all that it says, the explanation of why x will not be doing everything that x ought unless ~Ex ∨ ~Bx ∨ Mx works the same way as the explanation of why someone who ought to post a letter will not be doing everything that she ought unless she either posts it or burns it.9 But if someone who has promised to post a letter (and hence ought to post it) burns it instead, she is either posting it or burning it, but there is nothing intuitive about the claim that she is satisfying one of her obligations.10 Moreover, if she does post the letter, there is intuitively nothing extra that she does right, in virtue of the fact that she thereby either posts it or burns it. So when proponents of Wide say things like that you can satisfy the requirements of instrumental rationality just as well by giving up your ends as by going on to intend the means,11 or that there is something distinctive going wrong with someone who fails to intend the believed means to her intended end,12 it does not make sense to interpret them as meaning simply that x will not do everything

8   The data is actually more complicated; I’m simplifying by ignoring what happens when, for example, you believe that the means is necessary for the end but that you will do it without intending to do so. See Setiya (2007) and Kolodny (2008a) for more discussion of these sorts of important details, over which I will proceed to gloss. 9   This is Ross’s Paradox, originally raised in the context of imperatives. See Ross (1941). 10   Although compare Wedgwood (2007: ch. 4) on how easy it may be to satisfy some obligation versus how hard it is to satisfy all obligations. 11 12   Compare Hill (1973), Broome (1999).   Compare Wallace (2001).

94  Mark Schroeder that x rationally ought unless ~Ex ∨ ~Bx ∨ Mx—they must mean something stronger.13 It is therefore this stronger thing that detractors of Wide Scope views, including Joseph Raz, Niko Kolody, and myself have meant to deny.14

3  Scope in the Law This raises the question of just what this stronger interpretation of “OUGHTx” is supposed to be. And for the answer to that, it is helpful to look back to the example of the law. Just as someone who intends e, believes that m is a necessary means to e, and does not intend m is not entirely as she rationally ought to be, someone who is a driver in the state of New York and turns right on red violates the New York state traffic regulations. If we let Ex be the proposition that x is driving, Bx be the proposition that x is in the state of New York, and Mx be the proposition that x does not turn right on red, we can now interpret Narrow, Intermediate, Wide, and Myth as accounts of what is going on in this case. Again, if we interpret “OUGHTx” as meaning “x will be in violation of New York state traffic regulations unless,” there is nothing controversial about Wide. But again, since it is also true that you will be in violation of New York state traffic regulations unless either you are not in New York or you are not a driver or you do not turn right on red or you drive a convertible, this does not seem like a particularly interesting thing to say. On the other hand, if we interpret “OUGHTx” as “New York state traffic regulations require x to,” then Wide is not obviously true after all. On the contrary, Narrow and Intermediate are much more natural views. According to Narrow, New York state traffic regulations have jurisdiction over drivers in New York, and require them not to turn right on red. According to Intermediate, New York state traffic regulations have jurisdiction over everyone in New York, driver or not, and require them to not turn right on red while driving. Both of these views are plausible, because it is (relatively) easy to understand why the New York legislature has jurisdiction over people in the state of New York—particularly if they are driving. But on this interpretation, Wide says that New York state traffic regulations have jurisdiction over drivers in Buenos Aires and Cairo, and require them to either not turn right on red or else not be in New York. This claim is particularly implausible, because it is very hard to see how the New York state legislature could have gotten jurisdiction over drivers in Cairo or Buenos Aires. The much more natural way of understanding what is going on in this case is therefore that Wide is not, in fact, true on this interpretation. Another way of seeing the same thing, I think, is to observe that drivers in Cairo and Buenos Aires are not complying with New York state traffic regulations, simply because 13   For this point in connection with Wide, see van Roojen (unpublished). Similar issues were originally introduced in the context of understanding conditional “oughts” as oughts with material conditional complements within Standard Deontic Logic by Chisholm (1963). 14   See particularly Raz (2005), Kolodny (2008a, 2008b), and Schroeder (2004, 2009).

Hypothetical Imperatives  95 they are not in New York. In contrast, drivers in New York who don’t turn right on red are complying with New York state traffic regulations. It is true that there are two ways to avoid violating New York traffic regulations—you can refrain from turning right on red, or you can leave the state. But these are not two ways of complying with the regulations. One is compliance, and the other is escape. The distinction between compliance and escape tracks the regulations’ jurisdiction, because you can comply with a regulation only if you fall under its jurisdiction, and leaving the jurisdiction of the regulation is sufficient to avoid violating it. What these remarks illustrate, I believe, is that the concept of what is required by New York traffic regulations is the right kind of concept to allow for a meaningful scope debate precisely because it allows for a non-vacuous distinction between who falls inside and outside the jurisdiction of the law. Any concept that allows, at least in principle, for a non-vacuous jurisdiction distinction enables us to have a real scope debate, because the concept of jurisdiction guarantees that no one who falls outside the jurisdiction is bound by any of its requirements. And that means that all requirements are conditional on falling under the relevant jurisdiction. So if proponents of the Wide scope view are willing to use such a law-like concept, and hold that something—a kind of means-end consistency—is required of all rational agents, then they are committed to a view about the jurisdiction of this requirement or its source—that it has jurisdiction over all rational agents. What I’ve just been arguing, is that the dispute between advocates of Wide and its detractors is not best understood as a dispute about whether Wide is true, if interpreted in such a weak way that it follows from Narrow, Intermediate, or Myth, and moreover that understanding this dispute requires drawing an important distinction between compliance and avoidance, which tracks the important concept of jurisdiction. The distinction between compliance and avoidance is precisely what makes the dispute among proponents of Narrow, Intermediate, Wide, and Myth an interesting one.

4  The Impact of Jurisdiction on Explanatory Resources So far I’ve been arguing that insofar as there is an intelligible debate about the scope of instrumental rationality, it is a debate about the scope of a concept that is law-like, in that it distinguishes between violation and non-compliance, and that this distinction is closely related to the concept of jurisdiction. The fact that law-like normative concepts are precisely of the right kind to allow for these important distinctions should encourage us, I believe, about whether these distinctions are important for Kant’s own theory of instrumental rationality. For Kant himself employed a rich terminology deeply indebted to thinking in terms of the concept of law. Imperatives are defined outright in the Groundwork as the expression of laws to imperfectly rational wills.

96  Mark Schroeder The reason why scope is closely related to jurisdiction is that a law only needs to have jurisdiction over those who satisfy its condition. Since Narrow and Intermediate only postulate laws given some substantive condition, the laws they postulate need only have narrow jurisdiction. But since Wide postulates an unconditional law, that law must have universal jurisdiction. This is precisely what makes the Wide interpretation of New York state traffic laws so implausible—for it is implausible that the New York state legislature has the authority to legislate laws with jurisdiction over drivers in Cairo or Buenos Aires. So what about the Wide interpretation of the requirement of instrumental rationality? Is it plausible that it has universal jurisdiction over all rational agents? I have to confess that I find this idea as puzzling as I do the idea that the New York state legislature has jurisdiction over drivers in Buenos Aires. What is the source of this universal requirement supposed to be, and how does it acquire jurisdiction over every rational agent? I find it difficult to even get my head around this question. Moreover, I think that this sort of puzzlement is distinctly Kantian. For according to Kant, rational agents are autonomous, in the sense that they act according to laws that they set for themselves: The will is therefore not merely subject to the law, but subject in such a way that it must be considered as also giving the law to itself and only for this reason as first of all subject to the law (of which it can regard itself as the author).15

But if hypothetical imperatives derive from a master Hypothetical Imperative with universal jurisdiction over every rational agent, then their authority seems to come from outside the agent—for it comes from whatever authority has jurisdiction over all rational agents. So although it is possible that this is based on a misconception on my part, such an external source of rational requirements sounds prima facie exactly like heteronomy of the will. Autonomy, in contrast, would be each agent being bound only by laws she sets for herself—that is, the idea that each agent falls only under her own jurisdiction. A second, related but also important question, is how bare rational agency suffices to explain why someone falls under this particular requirement. Kant explains in the Groundwork that the hypotheticality of hypothetical imperatives is exactly what makes them unpuzzling, and the kind of thing whose possibility can be established by analytic means. The “objective necessity” which they present, he says, is only “based on a presupposition”—a presupposition that is rich enough for us to know that someone who satisfies it is indeed bound by that imperative. By contrast, “How is the imperative of morality possible?” is beyond all doubt the one question in need of a solution. For the moral imperative is in no way hypothetical, and consequently the objective necessity, which it affirms, cannot be supported by any presupposition, as was the case with hypothetical imperatives.16 15

  Kant (2002: 232 [4: 431]). Boldface added for emphasis; italics in original.   Kant (2002: 220 [4: 419]).

16

Hypothetical Imperatives  97 I believe that the insight behind this passage is that the stronger the condition on which an imperative applies, the richer the explanatory materials that we will have, in order to explain why and how it is, that the imperative applies. What makes categorical imperatives puzzling and in need of explanation, on this view, is precisely that they are unconditioned, and so we have nothing more to work with than rational agency as such, in seeking to explain them. So Hill’s Hypothetical Imperative, which unconditionally requires means-end coherence of every rational agent, should be puzzling for exactly the same reason, and I think that it is. So I believe that there are two, distinct but closely related, puzzles about unconditional requirements like that postulated by the Wide scope view of instrumental rationality, as I’ve been interpreting it in this chapter. The first is the puzzle of what could have jurisdiction over every rational agent, and the second is how the bare idea of rational agency as such could suffice to explain not only why an agent falls under this jurisdiction, but why the requirement is in force for her. Since I believe that these puzzles are closely related, however, I don’t think it should be surprising if a single move addresses both.

5  Autonomy of the Will Whereas it is puzzling, I think, how something could come to have jurisdiction over every rational agent, I don’t think it is similarly puzzling how a rational agent could come to have jurisdiction over herself. This is not to say that there are no philosophical puzzles about the latter—on the contrary, if this is Kant’s view, that I have authority over myself is something that I may only be able to establish through a transcendental argument—but only that the same puzzles do not arise. If an agent has jurisdiction over herself, then she can create rules or laws for herself. In the helpful terminology of Sam Shpall (2013, 2014), she can rationally commit herself to acting in one way or another.17 Because the New York state legislature has jurisdiction over drivers in the state of New York, it has the authority to require things of them. But in order to exercise this authority, it must act. If it passes legislation which says, “drivers may not turn right on red,” then drivers in New York become required not to turn right on red. And if it later passes legislation which says, “drivers may turn right on red,” then this former requirement is relaxed. Similarly, if a rational agent has jurisdiction over herself, then she has the authority to create rules for herself. But in order to exercise that authority, she must do something. So in order for an agent to have meaningful such authority over herself, there must be something that she can do to exercise it. 17   Shpall (2013) argues convincingly for the importance of our intuitive notion of commitment in considering cases like instrumental rationality and enkrasia, and in (2014), he develops a rich characterization of the distinctive and important features of this sense of commitment.

98  Mark Schroeder The virtue of the Narrow and Intermediate accounts of instrumental rationality is that they tell us exactly what a rational agent must do, in order to exert this authority over herself. On the Narrow view, what she must do to commit herself to intending m, is to intend e and believe that m is necessary for e. On the Intermediate view, what she must do to commit herself to not both intending e and not intending m, is to believe that m is necessary for e. Because each of these views makes the requirement governing an agent not only conditional, but conditional on something that agent does (in a very expansive sense of “does” which includes belief and intention), they are very naturally construed as simply telling us what an agent must do, in order to commit herself. This picture does away with the idea that there is some peculiar source of rational requirements which somehow has jurisdiction over every rational agent, and replaces it with the idea that each rational agent has jurisdiction over herself. In that way, it addresses the first problem which puzzled me in section 4. And it makes good on Kant’s observation that was at the heart of the second problem from that section. Kant’s observation, I suggested, was that conditional requirements give us richer explanatory resources in order to explain requirements, and on the picture being developed in this section, we utilize those explanatory resources by conceiving of them as the things an agent must do, in order to exercise her authority over herself. I don’t know that Kant’s own view is anything like this, but if it is, I suspect that it has neither the shape of Narrow nor that of the Intermediate view I’ve described, but rather that of an alternative, Kantian Intermediate view: Kantian Intermediate

∀x (Ex → OUGHTx(~Bx ∨ Mx))

After all, if rational agents give themselves laws, on Kant’s picture it will not be their beliefs which do so, but the exercises of their will. In willing an end, I commit myself to taking the means I believe to be necessary to it. In ceasing to will this end, I relieve myself of that commitment. So long as I am under the commitment and fail to intend some means I believe to be necessary, I am failing to do everything to which I am committed, and hence am irrational. But that is not because there is any more general rule of rationality which explains why this is so. It is just because through my will, I have the power to rationally commit myself, and rationality is nothing more than living up to my own commitments—that is, being successfully governed by the laws that I set for myself.18

6 Kantsequences In this chapter I’ve been trying to argue that it helps to use a law-like normative concept—one which allows for meaningful distinctions about jurisdiction—in order to get 18   This description matches the interpretive claims about Kant advanced in Schroeder (2005). Compare especially the discussion of autonomy in Hill (1985), and in Hill (1989: esp. 140–1).

Hypothetical Imperatives  99 to the heart of what has been at stake in the so-called scope dispute over instrumental rationality. And if this is right, then we should not be surprised, given Kant’s own reliance on the concept of law, if this dispute is important for his purposes. And I’ve promoted two Kantian ideas as having real import for this dispute: first, that rational agents are autonomous, in the literal sense that they give themselves their own laws, and second, that the very thing that makes hypothetical imperatives easier to explain is their conditionality, since that provides us with greater explanatory materials. When we put these two Kantian ideas together, we get the view that instrumental rationality is not something that is required of us, as rational agents, but rather, simply a reflection of our capacity to require things of ourselves. It is not right—or at least not illuminating—to think of instrumental rationality, on this picture, as backed up by a law-like Hypothetical Imperative. I have not intended the main claims of this chapter to amount to interpretive claims about Kant; merely to emphasize what I  take to be Kantian themes that I think are independently important in this domain. But if Kant really did accept something like the Kantian Intermediate view, holding that the will is our capacity to require things of ourselves, he certainly did not think that it gives us the authority to require just anything of ourselves. Just as the New York state legislature has the authority to require drivers not to turn right on red but lacks the authority to require voters to pay a poll tax, rational agents, though they have the authority to require some things of themselves, lack the authority to require themselves to lie, for example, or to neglect their self-development. Famously, for Kant all limits on what agents have the authority to require of themselves must come from the bare condition that all requirements must have the form of a law. So whereas Kant’s account of hypothetical imperatives, on this view, reflects the authority that rational agents have over themselves, his account of categorical imperatives reflects the limits of that authority. And this means, I think, that there is something worth calling the Hypothetical Imperative after all—even though I don’t think it is right to think of it as an imperative. After all, at the end of “The Hypothetical Imperative,” Hill emphasizes that what is most important about the claims that he makes in the article is that it shows how the Hypothetical Imperative and the Categorical Imperative are simply two different reflections of a single capacity for practical reason, which can be “paradoxically” summarized with the edict, “Do what you will.” But this is exactly what I have just been arguing follows, if anything like the picture of this chapter is accurate about Kant’s own views. If practical reason is auto-nomous in the way I’ve described, then the Hypothetical Imperative is a reflection of the auto, and the Categorical Imperative is a reflection of the nomous. If that is true, it is an important truth, and no paradox.19 19   Special thanks to Tom Hill, Robert Johnson, Mark Timmons, Errol Lord, Andrew Sepielli, and an anonymous referee.

100  Mark Schroeder

References Broome, John. 1999. “Rational Requirements.” Ratio, 12(4): 398–419. Chisholm, Roderick. 1963. “Contrary-to-Duty Imperatives and Deontic Logic.” Analysis, 24(1): 33–6. Darwall, Stephen. 1983. Impartial Reason. Ithaca, NY: Cornell University Press. Hampton, Jean. 1998. The Authority of Reason. Cambridge: Cambridge University Press. Hill, Thomas E., Jr. 1973. “The Hypothetical Imperative.” Philosophical Review, 82(4): 429–50. Hill, Thomas E., Jr. 1985. “Kant’s Argument for the Rationality of Moral Conduct.” Reprinted in Hill 1992: 97–122. Hill, Thomas E., Jr. 1989. “Kant’s Theory of Practical Reason.” Reprinted in Hill 1992: 123–46. Hill, Thomas E., Jr. 1992. Dignity and Practical Reason in Kant’s Moral Theory. Ithaca, NY: Cornell University Press. Hill, Thomas E., Jr. and Arnulf Zweig. 2002. “Editors’ Introduction.” In Kant 2002: 19–108. Kant, Immanuel. 2002. Groundwork for the Metaphysics of Morals, trans. Arnulf Zweig, ed. Arnulf Zweig and Thomas E. Hill, Jr. Oxford: Oxford University Press. Kolodny, Niko. 2008a. “Why Be Disposed to be Coherent?” Ethics, 118(3): 437–63. Kolodny, Niko. 2008b. “The Myth of Practical Consistency.” European Journal of Philosophy, 16(3): 366–402. Raz, Joseph. 2005. “The Myth of Instrumental Rationality.” Journal of Ethics and Social Philosophy, . Ross, Alf. 1941. “Imperatives and Logic.” Theoria, 7(1): 53–71. Schroeder, Mark. 2004. “The Scope of Instrumental Rationality.” Philosophical Perspectives, 18 (Ethics): 337–64. Schroeder, Mark. 2005. “The Hypothetical Imperative?” Australasian Journal of Philosophy, 83(3): 357–72. Schroeder, Mark. 2009. “Means-End Coherence, Stringency, and Subjective Reasons.” Philosophical Studies, 143(2): 223–48. Schroeder, Mark. 2011. “Ought, Agents, and Actions.” Philosophical Review, 120(1): 1–41. Setiya, Kieran. 2007. “Cognitivism about Practical Reason.” Ethics, 117(4): 649–73. Shpall, Sam. 2013. “Wide and Narrow Scope.” Philosophical Studies, 163(3): 717–36. Shpall, Sam. 2014. “Moral and Rational Commitment.” Philosophy and Phenomenological Research, 88(1): 146–72. van Roojen, Mark. Unpublished. “Consequents of True Practical Conditionals Detach.” Unpublished paper. Wallace, Jay. 2001. “Normativity, Commitment, and Instrumental Reason.” Philosophers’ Imprint, 1(4), . Way, Jonathan. 2010. “Defending the Wide-Scope Approach to Instrumental Reason.” Philosophical Studies, 147(2): 213–33. Wedgwood, Ralph. 2007. The Nature of Normativity. Oxford: Oxford University Press.

5 More Right than Wrong Jonathan Dancy

I claim some credit for having been instrumental in W.  D. Ross’s return to philosophical favour after thirty years of oblivion at the hands of R. M. Hare. I was struck, therefore, when Tom Hill told me that he had taught Ross every year since his original appointment at Chapel Hill. This is characteristic of that admirable intellectual independence which we also see in his relation to Kant. In the present chapter I consider one aspect of the difference between Kant and Ross. It is often said that Kant has great difficulty in making any sense of the notion of a moral reason, and that in this respect Ross’s notion of a prima facie duty puts him at a great advantage—an advantage inherited by those who think that the basic notion in all this is the notion of favouring, understood as a specific relation that holds directly between certain features of the agent’s situation and certain forms of response (mainly action). In his interesting paper ‘The Importance of Moral Rules and Principles’, Hill suggests in response that particular facts are reasons because they are ‘especially salient features of a possible fuller rationale that we might give’. This is an example of a Kant-inspired account which diverts our attention away from the supposed relation of favouring, and seeks to explain reasons as parts of a rational structure, dominated by some grand Kantian Principle, Categorical or Hypothetical, that generates an overall ought, or duty. The difficulty for this manoeuvre lies in the phrase ‘especially salient’. Reasons, it might be said, are not just particularly eye-catching elements in a structure. Unless more can be said than this remark about special salience, we lose entirely the distinction between the roles played by reasons and the roles played by other aspects of the supposed ‘fuller rationale’. In what follows I suggest that those inspired by Ross have much less advantage on this point than they tend to suppose, for Ross fails to make any sense of the relation between moral reasons as he understands them (that is, as prima facie duties) and the rightness of the action that is made right by those reasons. He can give no account of what it is that his moral reasons do. I then consider the best way of trying to improve things.

102  Jonathan Dancy

1  Ross on the Relation between Duty Proper and Prima Facie Duty Ross was the first person to see that the rightness of an action is built out of certain features that go to make it right, which he called prima facie duties. Ewing rightly said of this that it was ‘one of the most important discoveries of the century in moral philosophy’ (1959: 126). This was true in 1959, and it remained true forty years later. What is extraordinary, and very hard to believe nowadays, is that before Ross nobody had any inkling of this idea. Everyone thought in terms of principles of absolute duty. If a principle applied at all, it determined the answer to the question what one’s duty was. Evidence of this fact, if evidence is needed, comes from the structure of the argument in the final chapter of Ewing’s early The Morality of Punishment (1929). Ewing argues that all known principles of absolute duty (that is, the Kantian and the utilitarian versions) are false, and that therefore there are no moral principles. The result of this reasoning was that Ewing was a particularist at the beginning of his long career, which consisted of a series of lurches from one meta-ethical position to another. (His Second Thoughts in Moral Philosophy (1959) might have been better called Yet Another Change of Mind.1) Ross offered in The Right and the Good (1930) a way of escaping the need to choose between Kantianism, Utilitarianism, and particularism: there are principles, but they are all principles of prima facie duty. Principles of this form do not determine the answer to the question what one’s duty is in a particular case; on that matter they are silent, leaving the issue to judgement. They play their role at a preliminary stage, and several of them may apply to the same case at the same time. The role of judgement is to choose between them, and all that the principles of prima facie duty do is to tell the judger what she is choosing between. The system (if it can be so called) serves up a list of directly relevant features, but it does nothing to tell us how significant the various features are in a particular case, and so does nothing to tell us how to act in response to the different prima facie duties that we know ourselves to have here. This is no doubt disappointing, considering the ambitions of traditional moral theory; but Ross held that no more could be expected or achieved. Many people now think Ross right about this, though there are still some who continue to write as if Ross had never existed. Derek Parfit’s recent book is an impressive example of this; the general impression (on this topic) is that if it is not in Sidgwick or in Kant, it can be confidently ignored. But this prejudice is dangerous, for absolute moral principles are much harder to find and to defend than are prima facie versions of the same thing. Compare these two versions of a principle of consent: (1) if a person cannot consent to the way you are treating him, what you are doing is wrong, and (2) if a person cannot consent to the way you are treating him, what you are doing is prima facie wrong. The nub of the matter is that counter-examples to the second of these are 1   This reminds me of Broad’s remark that Ross’s second book, Foundations of Ethics, could well have been called ‘The Righter and the Better’ (1940: 239).

More Right than Wrong  103 much harder to produce than are counter-examples to the first. So one would think that wary theorists would do well to restrict their pronouncements to the prima facie. Parfit is a contemporary example of someone who is not wary in this way. So Ross was the first person to see the need for a distinction between what I call the contributory and the overall levels. And this was indeed an intellectual triumph. But sadly, this is as far as the triumph went. Pretty well everything else in Ross’s meta-ethics is wrong—indeed, hopelessly wrong. In saying this, I am not referring to his intuitionism (by which I  mean his pluralism and the associated epistemology), nor to the list of particular duties that he provides (which is not a bad first shot, I would say, if such a thing is needed). I am referring to the details of the account he gives of the contributory, and of its relation to the overall. None of this can be got to work. So what Ross shows is that we need an account of the contributory, and we need an account of the relation between what goes on at that level and the level at which facts about how we ought in fact to act emerge. But Ross’s own versions of these things are way off the mark. What Ross rightly wants is the idea of something normative at the contributory level that is a matter of degree. This is what he is getting at when he talks of one prima facie duty being more incumbent or more pressing than another. And he tries to express this idea by talking of prima facie duties. But he immediately announces that these prima facie duties ‘are not duties at all, but something related in a special way to duty’ (1930: 20). We need, therefore, to understand the way in which one of these things can be more pressing than another, when greater pressingness is not a matter of greater duty. And Ross’s official account of a prima facie duty just prevents us from doing this. He writes that being a prima facie duty is ‘the characteristic (quite distinct from that of being a duty proper) which an act has, in virtue of being of a certain kind, . . . of being an act which would be a duty proper if it were not at the same time of another kind which is morally significant’ (1930: 19). But the first and most obvious thing about this account is that prima facie duty, as characterized, is not a matter of degree. One cannot have more of being an act which would be a duty proper if it were not at the same time of another kind which is morally significant; the idea of such a thing doesn’t make any sense. Ross has left us no room to say that one of these things can be more pressing than another. The normativity of the contributory has vanished from view entirely. At best, what Ross offers is a consequence of the thing he is really trying to get at. The same can be said of the use of the term ‘prima facie justification’ in contemporary epistemology. Those who use this term acknowledge their debt to Ross. But the unfortunate fact is that they also think of a prima facie justification as having some normative force—as generating a little bit of justification, to be put in the balance and compared with the prima facie justifiedness of alternative hypotheses. And this idea is one that Ross’s views completely undermine. If we take what he says seriously, a prima facie justifier is a feature such that, if it were the only relevant feature, it would make the belief at issue justified. But this isolation test is not the way in which the notion of prima facie justification is used in epistemology, and rightly so.

104  Jonathan Dancy I said at the outset that Ross was attempting to make sense of the way in which particular features of an action can, as I put it, go to make that action right. This notion of ‘making right’ is worth further thought. For Ross denies that any individual feature can ever succeed in making an action right. An action is never made a duty proper by one feature alone. A single feature can only make the action a prima facie duty. If it is a duty proper, it is, he says, made so by all of its features, not just one, or even some, of them. The equality of the two angles [of an isosceles triangle] is a parti-resultant attribute. And the same is true of all mathematical attributes. It is true, I may add, of prima facie rightness. But no act is ever, in virtue of falling under some general description, necessarily actually right; its rightness depends on its whole nature and not on any element in it. The reason is that no mathematical object (no figure, for instance, or angle) ever has two characteristics that tend to give it opposing resultant characteristics, while moral act often (as everyone knows) and indeed always (as on reflection we must admit) have different characteristics that tend to make them at the same time prima facie right and prima facie wrong. (1930: 33–4)

This passage rewards attention. The first thing to say is that the last sentence introduces a confusion. Later in c­ hapter 2 of The Right and the Good, Ross suggests an understanding of prima facie duty that is different from the one I have already mentioned, and is run in terms of tendencies. But these tendencies are tendencies to make actions actually right or actually wrong: they tend to make them duties proper, not to make them prima facie duties. Otherwise the appeal to tendencies would do nothing to explicate the notion of a prima facie duty. (I return to this suggestion about tendencies in section 2.) But it is in the third sentence that the real mistake occurs. Why is the (supposed) fact that no act is ever, in virtue of falling under some general description, necessarily actually right, relevant to the question whether an action can be made right by its possession of a limited number of characteristics? It is only relevant if one assumes, as Ross did but I do not, that if an action is made right by having certain features, any other action that has those features must also be right, no matter how it may differ in other respects. If one makes this assumption, the idea that an action can be made right by the good it does, or by any other particular character it may have, is indeed a non-starter. So Ross, making this assumption, denies the possibility of what I have been calling ‘right-making features’, and I view this as another abject failure. The sort of contribution that is made by a particular feature must be compatible with the idea that such a feature can on occasion be what makes an action the one we ought to do. Of course this feature will only be able to do it if no other feature succeeds in making some alternative act a duty proper instead. But this does nothing to show that when that condition is satisfied, this feature is not what makes the action a duty. What we should be hoping to provide is an understanding of the thought that this action is a duty proper because you promised you would do it; your promise makes it a duty proper and not just a prima facie duty. But Ross fails to provide any such understanding; indeed, he asserts

More Right than Wrong  105 firmly that duty proper is, as he puts it (rather exaggeratedly) ‘toti-resultant’: that is, that what makes the action a duty proper is all its properties at once. Now we know that this is false. If an act is a duty despite being hard-hearted, it is not a duty partly because it is hard-hearted. Ross has left no room for the idea of a ‘despite’. And Ross knew that it is false too, for he admits in a footnote (1930: 33–4) that his account is exaggerated, because many features of an action will be irrelevant to whether it is a duty. But this is not enough; he still insists that no act is made a duty by any one consideration, and this is something we should reject in advance, even though we of course allow that many duties are made duties by several features acting together, pooling their normative resources.

2  The Tendency Account Ross writes (1931: 28) that We have to distinguish from the characteristic of being our duty that of tending to be our duty. Any act that we do contains various elements in virtue of which it falls under various categories. In virtue of being the breaking of a promise, it tends to be wrong; in virtue of being an instance of relieving distress it tends to be right. Tendency to be one’s duty may be called a parti-resultant attribute, i.e. one which belongs to an act in virtue of some one component in its nature. Being one’s duty is a toti-resultant attribute, one which belongs to an act in virtue of its whole nature and of nothing less than this.2

What are we to make of this?3 One possible reply is that tendencies do not belong to particular actions, but to actions of a type. A particular action does not tend to be dangerous; either it is dangerous or it is not. But perhaps Ross thinks otherwise. If he does, what is offered here does indeed look like a sort of normative force coming from below, as it were, which has the great advantage that there is a clear account on offer of the relation between the lower and the upper strata—being the right-maker and the rightness made. Nonetheless I do not think that this account can be got to work.4 Certainly, tendencies to be right can be stronger or weaker, and so a tendency to be right is a matter of degree, which helps; but a tendency of this sort is not itself normative. The normativity of the tendency comes entirely from the normativity of the rightness that acts of this sort tend to have. It is just not clear that an unfulfilled tendency is the right sort of thing to generate residual duties, and it should be clear if an unfulfilled tendency had its own normativity (see section 3 for more on this topic).   Though Ross adds in a footnote that this last claim, about toti-resultance, is somewhat exaggerated.   For a discussion of this suggestion, and a reply to some early complaints that I made about it, see Gay (1985). 4   There is a full-scale discussion of the tendency account of prima facie duties in ch. 6 of my Moral Reasons (Dancy 1993). Here I focus only on the use of that account to solve my worries about the relation between overall and contributory normativity. 2 3

106  Jonathan Dancy It is worth comparing what Ross says about prima facie rightness with what he says on occasion about prima facie goodness, since he clearly thinks that as far as the term ‘prima facie’ goes, these two are on a par: If it were self-evident that the right coincides with the optimific, it should be self-evident that what is prima facie right is prima facie optimific. But whereas we are certain that keeping a promise is prima facie right, we are not certain that it is prima facie optimific (though we are perhaps certain that it is prima facie bonific).5

Here Ross does not seem to feel any strain in applying his conception of the prima facie to good-making rather than to right-making. The prima facie optimific is not something that tends to be best, but something that is already good-making in a certain respect. Being already good in a certain respect is not the same as tending to be best (nor is it the same as being something that if it were alone would be best). Similarly, the prima facie bonific should be something that is actually bonific in a respect, not something that tends to be so overall. And this is the sort of thing that Ross should be saying about prima facie rightness—except that he doesn’t.

3  Further Difficulties While I am at it, I might as well mention a further failure of Ross’s account of the prima facie, that is, of the contributory. One attractive aspect of his approach is that it seems to make room for a duty that we have, and that we recognize that we have, but which we think of as defeated by another duty which is stronger. Now I have already said that Ross in fact offers no workable understanding of how one prima facie duty can be stronger than another. But at least we might have hoped that he can give some sense to the idea of a defeated but still real duty (unlike the friends of absolute duties, who have to say that no real duty can be defeated). Unfortunately it turns out that this is not so. For his account of a prima facie duty as a feature that would decide the issue if it were the only relevant one fails entirely to make sense of the idea that if a prima facie duty loses out, it can still generate catch-up duties (such as the duty to apologize). This is because the account is officially silent about what happens to a prima facie duty when it is not the only relevant feature. This is exactly the problem with isolation tests. And for that reason it can say nothing about how a defeated duty behaves, either in general, or in particular as a generator of catch-up duties. All these problems stem from the one central failure of Ross’s account, which is that it fails entirely to make any sense of the normativity of the contributory. But the idea of the contributory, and the distinction between the contributory and the overall, are things that we very much want to retain. One feature of the contributory will occupy us for much of this chapter, which is the fact that it comes in degrees. Rightness, for Ross, is a matter of degree, but this applies only to prima facie rightness, not to rightness proper. Duty proper is not a matter of degree, since no duty is more pressing than any other duty. But some prima facie duties are more pressing than others.   Ross (1930: 36).

5

More Right than Wrong  107 I may say, at this point, that I don’t find it at all helpful to think of Ross’s prima facie duties in terms of Foot’s distinction between the verdictive and the evidential.6 The verdictive side of this is not too bad. To say of an act that it is the right one in the circumstances can indeed be understood as the casting of a verdict in favour of that act. In fact, this verdictive conception is compatible with both views about the nature of rightness, for one can cast a verdict in favour of an act equally well by saying that it is the most right act and by saying that it, and it alone, is right (in the absolute sense). So I will continue to use what is effectively the notion of the verdictive in what follows. It is the evidential conception of the prima facie that seems mistaken. A prima facie duty is not to be understood as evidence that the relevant action is a duty proper. And a defeated prima facie duty is not like misleading evidence, that is, evidence that leads one away from the true answer. In my own terms, a reason not to do the action remains in play as that reason even when defeated by stronger reasons on the other side. It is not shown to be misleading, or substandard in some way by this defeat. The notion of evidence just seems out of place here. After all these criticisms of Ross, it is worth pointing out that in his later Foundations of Ethics he completely changed his view of the relation between the contributory and the overall. The earlier view, which we have been examining, was that the contributory (the prima facie, that is) is to be understood in terms of its relation to duty proper. The later view reverses this, understanding the contributory in terms of a relation of fittingness or suitability in a respect, and the right action as that action that has ‘the greatest amount of such suitability possible in the circumstances’ (1939: 53; but see also p. 84, where it is far from clear that Ross has simply abandoned his earlier views). This view is much closer to the one that Prichard pressed against Ross in the years between the publication of Ross’s two books, as we will now see. It is also much more clearly opposed to the structure we see in Kantian accounts. The metaphorical distinction between top-down and bottom-up needs careful handling in the present context. Kant’s position is top-down, in the sense that the role played by the distinct elements in generating a duty is explained by their relation to the duty generated at the top. Ross’s first position is top-down in this sense as well; it is his second view that reverses the direction of explanation and is therefore bottom-up. The first position does, however, seek to explain duty proper as generated by normative forces coming up from below (expressed in such terms as ‘stringent’ and ‘pressing’); and in this second sense of bottom-upness both of Ross’s views are bottom-up. Kant’s position thus emerges as distinctive in being doubly top-down.

4  Prichard’s Complaint It is always interesting, when thinking about Ross, to ask oneself about the relation between his views and those of Harold Prichard. In many ways Prichard was Ross’s mentor. Ross was enormously influenced by Prichard’s argument that motive is not   Foot (1978: 182) and Stratton-Lake (2000: 14).

6

108  Jonathan Dancy choosable, and that therefore that which is our duty cannot include any motive. It cannot, therefore, be our duty to act from a certain motive. And Ross followed Prichard’s misguided lurch from objectivism about duty to subjectivism.7 And we might be under the impression that Prichard agreed with Ross about the distinction between prima facie duty and duty proper. For Prichard did write, in his famous ‘Does Moral Philosophy Rest on a Mistake?’ (1912) that: . . . obligation admits of degrees, and . . . where obligations conflict, the decision of what we ought to do turns not on the question ‘Which of the alternative courses of action will originate the greater good?’ but on the question which is the greater obligation. (2002: 14n.)

And it is easy to see in these remarks a position similar to the one that Ross eventually articulated in his book. The same applies to Prichard’s alternative suggestion to Ross’s terminology of prima facie duties, namely that we should think of such things as claims. Ross decided not to accept this idea on the grounds that if there is a claim, there should be a claimant, but in the case of some duties there is no claimant to be found (1930: 20). Still, there is something of a hint here that Prichard was sympathetic to Ross’s position, the only question between them being one of terminology. But there is compelling evidence that Prichard was in fact opposed to Ross’s substantive position—though not to his stress on the contributory. This evidence consists mainly of a letter which Prichard wrote to Ross in 1932. For the publication of this letter we are indebted to Jim McAdam’s recent edition of Prichard’s works, which contains a fair amount of previously unpublished material (2002: 286–7). The letter reads as follows: My dear Ross, I hope you will forgive my taking a sort of pot shot at something in your book. But I have got it on my chest. On p. 27 you reduce the 6 or rather 7 prima facie duties enumerated on pp. 21 to 4, viz. your original (1)a, (1)b & (2)8 & the prima facie duty of producing as much good as possible . . . And I have for long wondered that you should feel comfortable in coordinating the last with (1)a, (1)b and (2), e.g. the prima facie duty of keeping a promise. But I have never till now seen clearly what has been puzzling me about this procedure— though I have always felt it was queer. The queerness I think lies in this—that whereas e.g. to describe an act as one of keeping a promise or as one of making reparation is to describe it in respect of a character it has in itself, to describe [an act] as producing as much good as possible, is only to do this verbally; it is really to describe it as having a character which it possesses only in relation to all the other acts the man can do—the character of producing good in a degree greater than that produced by any of the others. This difference seems to me vital. The last does not seem to me the character of an action in the way in which the first 3 are. And the difference seems to me one which is paralleled in your distinction between some prima facie duty of a man and his duty sans phrase, and also to be out   I have documented these shifts of view in my ‘Prichard on Duty and Ignorance’.   These are the duties of (1)a: promising; (1)b: reparation; (2): gratitude.

7

8

More Right than Wrong  109 of place in a list of prima facie duties. Whether the phrase ‘prima facie duty’ is appropriate for the thing you refer to in your doctrine or not, the thing referred to is some character which an action of a certain kind possesses in itself i.e. as an instance of a certain kind and apart from its relatedness to the actions of other kinds possible to a man—a character e.g. due to an act being one of keeping a promise. And I take your view about duty to be that in a given situation the action which it is my duty to do is that out of all the actions which I can, there is the greatest prima facie duty to do—so that where some action is what you call ‘my duty’—or as I would rather put it the act which I am bound to do, or what I ought to do, it is so in virtue of having more prima facie obligatoriness than any of the other acts I can do—and so in virtue of a character which the act possesses only in relation to all the others. (cf. Pickard-Cambridge, note 3, p. 150, Mind, April, 1932.) Hence while the basis of a prima facie duty is a character which the action has in itself, i.e. apart from its relation to the others, the basis of a particular act’s being ‘my duty’ is not. Hence, also, it seems to me, given your distinction between a prima facie duty of mine and my duty, the so-called character of producing as much good as possible, i.e. really the character of producing more good than any other possible action, while it might possibly be maintained to be the basis of some action’s being my duty, can’t be held to be the basis of an action being a prima facie duty. But, further, strictly speaking I don’t see how it can be maintained even to be the basis of the thing called an action being my duty,—i.e. your ‘an action being my duty sans phrase’. For to produce more good than any other possible action is to have a certain character in a greater degree than certain other actions, and this cannot give rise to anything but a greater degree of something else; yet on your view to say of some act that it is a prima facie duty and to say of it that it is my duty are to make qualitatively different statements. The truth is that the more I consider it the less I can make sense out of ‘the act which I am bound to do’—as distinct from ‘an act which I ought to do’—and the more I get to think that the only fact corresponding to the phrase is ‘the act9 which I ought to do more than I ought to do any other’, and that your ‘prima facie’ duty is really a duty, your ‘my duty sans phrase’ is really that of a man’s duties which he most ought to do, i.e. that so far as these phrases can be made to stand for facts these must be the facts. . . . yrs ever HAP

This letter can be seen as having three sections. In the first Prichard distinguishes between the character which an act possesses as being of a certain kind, and a character which an act possesses in relation to all other acts available to the agent. (This distinction is different from, but analogous to, Ross’s own distinction between parti-resultant and toti-resultant.) And Prichard maintains that Ross’s prima facie duties are all officially of the first sort, and his duties proper are of the second sort. He then points out that the character of being optimific, which Ross treats as the 9   The text as MacAdam gives it reads here ‘corresponding to the phrase “is the act . . .”’, which renders the thought unintelligible.

110  Jonathan Dancy basis of a prima facie duty, is in fact of the second sort when officially it ought to be of the first sort. This by itself would not be damaging, since it only concerns the way in which Ross has formulated his duties of beneficence, non-maleficence, and justice. Ross has offered versions of these in explicitly maximizing terms, and one could imagine non-maximizing versions of those duties, which would perhaps make more sense.10 The idea would be that what makes the charitable act a duty is not that it does more good than any other act available, but just the good that it does. Its doing that good is a feature that it has as being of a certain kind, not in relation to all other acts available to the agent. So this particular point might be able to be cleared up. We would still need to find some role for the fact that the act produces more good than any alternative. If an act is right because of its production of good, is it right because of the amount of good produced, or because that amount is greater than any alternative would have produced—or both? Perhaps the answer is both, and that we are dealing with two different ‘becauses’ here. The ground for the rightness is the good produced but what enables that amount of good to make this action right is that no alternative action would have produced more. The comparative fact is relevant, but not in the same way as is the amount of good produced. This pattern may seem rather contrived, but it is in fact perfectly general. Every act that is right is so partly because there was no alternative action that would have taken precedence, or been more of a duty. But this does not mean that all right actions are made right partly by one and the same fact. The role of the absence of alternative more pressing duties is that of an enabler, not that of a ground. With this, I  return to Prichard, who has not finished yet. He points out that, though the act that is one’s duty proper is the act that is the most pressing prima facie duty, these characteristics are, for Ross, very different—qualitatively different, as he puts it. One is a matter of degree and the other is not. And Prichard doesn’t see how this can work. If we are to understand the relation between thoughts about prima facie duty and duty proper, we have to abandon the thought that ‘is my duty’ expresses something that is not a matter of degree, and to allow that all it can mean is ‘is the most pressing of my prima facie duties’. Far from its being the case that ‘duty’ in ‘prima facie duty’ means something completely different from the ‘duty’ in ‘duty proper’, so that prima facie duties are not duties at all, but something related in a special way to duties, for Prichard there are no prima facie duties at all. What we call our duty proper is simply our most pressing duty proper, and there is no room for a distinction between the ground for the duty proper and the ground for the prima facie duty that it also is. As we might put it, the ‘right’ in ‘right-making’ is not prima facie rightness but the only sort of rightness there is, real rightness. And the same applies to ‘ought’, and to ‘duty’.

  As Wiggins suggests (1998).

10

More Right than Wrong  111 Ross simply assumed that ‘ought to be done’ has an uniqueness implication, as in the phrase ‘the act that one ought to do’.11 But Prichard finds in this the same severance between the supposed matter of degree, as ground, and something that is not a matter of degree as conclusion. Really all we can make sense of is the idea of an act which to some extent we ought to do, or which is to some extent our duty. This notion enables us to ask what we ought to do, understanding that question as the question which of the acts which to some extent we ought to do is the one (if there is one) that we most ought to do. But there is no available notion of duty other than this, and so no conceptual distinction between duty proper and most pressing prima facie duty. I end this section with a diagram of Ross’s position as Prichard understands it: Duty proper

most pressing prima facie duty PF1

PF2

PF3

PF4

Ground 1

Ground 2

Ground 3

Ground 4

This diagram is intended to bring out the atomistic idea that degree of pressingness of a prima facie duty is an intrinsic matter, not comparative, since the pressingness of one prima facie duty will not affect the pressingness of another. The relation between those independent pressingnesses determines which of the prima facie duties is the most pressing. Comparison comes in only where we move up from level 2 to level 3. But to be a duty proper is (for Ross) to have a characteristic that is toto caelo different from that of being the most pressing prima facie duty, even though every act that has one of these characteristics also has the other.

5  Assessment of Prichard’s Objection Prichard needs to offer some defence of the main hinge principle in his letter, which is this: For to produce more good than any other possible action is to have a certain character in a greater degree than certain other actions, and this cannot give rise to anything but a greater degree of something else.

This remark seems to be intended as entirely general. It is that having some character in a greater degree than other actions can only give rise to something that is itself 11   For this assumption, see Ross (1930: 3–4), where he argues that though there can be two alternative actions both of which are right, only one of them can be what we ought to do, that is, our duty.

112  Jonathan Dancy a matter of degree. This looks like a ban on threshold properties with a comparative base of a certain sort. And we need to understand why this ban is supposed to be justified. With the ban in place, the situation seems to be like this. That this action did a certain amount of good could be the ground for a duty proper (if there were such a notion), since doing that amount of good is not having a certain character in a greater degree than certain other actions do. But that the amount of good done is greater than what any other action would have achieved is only capable of ‘giving rise’ to a greater degree of something else, such as a ‘more ought’ or a greater prima facie duty, and could not be the ground of any such thing as a duty proper. In interpreting this ban, we might wonder whether to produce more good is the same thing as to have a greater degree of good-producingness (or of ‘beneficence’)? What are we to make of Prichard’s ‘having a certain character in a greater degree than certain other actions’? There seems to be a distinction between being more productive of good and being productive of more good. And if there is, which of these is the ban concerned with? It looks as if it should be the former, if either, since the latter is not obviously a matter of degree. And there is certainly a distinction between being more impassioned about music and being impassioned about more music, and only the former seems to fit Prichard’s phrase. On the other hand, it seems to be the latter (being productive of more good) that is at issue when we think about which of two beneficent acts we have the greater prima facie duty to do. If so, Prichard has no complaint against the way in which Ross treats the duty of beneficence. The property of being productive of more good is not itself a matter of degree. To confirm intuitions, suppose that there is only one prize, and it is given to the person who makes the most cakes. Bernie makes the most cakes, and so deserves the prize. Are we to suppose that the other cake-makers also deserved the prize, but deserved it less, so that Bernie is the most prize-deserving, but not the only prize-deserver? (This seems to be the way that Prichard would have to read the case.) Does Lucio, who made the fewest cakes, deserve the prize to some extent, only not so much as Bernie? This seems most improbable. Only Bernie deserves the prize. But does this case really come under Prichard’s ban on threshold properties that are not matters of degree but which are grounded in something that is a matter of degree? No, since making most cakes is not a matter of degree. There is the now familiar distinction between making most cakes and doing the most making of cakes. Bernie is not more of a cake-maker than Lucio, even though he makes more cakes. My conclusion at this point is that Prichard’s reasoning is not very persuasive. And in addition, I think that it is at least open to Ross to abandon the claim that no duty proper is more of a duty proper than is any other. Perhaps he could allow that of two actions, both of which are duties proper (and which, of course, are therefore not competing), it would be more wrong not to do one than it would be not to do the other. If so, the threshold property will end up being a matter of degree in just the sort of way that Prichard demands.

More Right than Wrong  113

6  Another Shot So far, then, it has not been shown incoherent to add a conception of absolute duty to our conception of something that is a matter of degree. As far as that goes, the structure of Ross’s position remains intact. Elsewhere, however, also in material that has only recently come to light,12 Prichard writes: To maintain that we ought to do that of two actions which we ought to do more is to be involved in a contradiction, because it is to imply both that while the one act is a duty the other is not, and that the latter is a duty though to a lesser degree than the former. Putting this otherwise, if we think that of two acts there is a greater obligation to do one of them, we cannot go on to think that we ought to do that action without implying that there is no obligation to do the other in any degree whatever. Consequently we gain nothing by maintaining the existence of degrees of obligation in addition to absolute obligation. (2002: 79)

In this we see Prichard prefiguring the complaint I made in section 1, that Ross ends up trying to say both that prima facie duties are duties and that they are not. Ross’s official view, of course, is that prima facie duties are not duties at all. He is not trying to say that they are and are not duties. But there is a sense in which Prichard is right. As I explained in section 2, Ross is trying to hold on to the idea that there is something normative to be found at the prima facie level, and he does this by calling what he finds there ‘prima facie duties’. But this is specious. He might as well have called them ‘prima facie cuties’ for all that he achieves by it. This is a different complaint from the general complaint that it is incoherent to add something that is not a matter of degree to something that is. So there is something new in what Prichard says here. If we think in terms of degrees of duty, meaning now by this something that Ross did not succeed in substantiating, namely that there is something normative that we are thinking of as a matter of degree, we gain nothing by adding to that a conception of duty that is not a matter of degree. We gain nothing because in terms of that second notion, only one action is a duty and all the others are not duties at all. And we would have to say of those actions that they are to some degree a duty and to no degree a duty, and this is hardly a gain. Prichard could reasonably claim that, if anything, it makes matters worse than before, because by adding a conception of duty that is not a matter of degree, we undermine our sense that what is a matter of degree is a matter of duty (so more or less pressing as a duty) rather than just something related in a special way to duty. This addition is not merely something whose relation to what we already had on the table (duty in degrees) is hard to make sense of—which would have to be admitted anyway; the one positively interferes with the other.13 12   This is from a set of student’s notes on a lecture of Prichard’s given in 1928, two years before the publication of The Right and the Good. These notes read as if Prichard had dictated them. 13   I should admit that, although the passage I have quoted, and am here discussing, does seem to me to contain a powerful argument against Ross’s position, that is not Prichard’s purpose in his lecture, and he

114  Jonathan Dancy This returns us to the point made in section 2, that Ross fails to locate something both normative and contributory at the prima facie level. But it goes beyond that in claiming that this is no surprise, because every attempt to achieve that, while insisting on a distinctive non-contributory level of normativity, undermines its own starting point.

7  A Distinctive Argument for Deontic Buck-Passing Prichard’s view, as I understand it, is a form of deontic buck-passing. We are more familiar nowadays with evaluative buck-passing, which is the idea that for something to be of value is for it to have features that give us reasons to respond to it in certain ways. Deontic buck-passing is the idea that for an action to be right is for the reasons that favour it to be stronger than the reasons favouring any alternative. The right act is the one that one has most reason to do. This formulation in terms of reasons is distinct from anything that Prichard actually says, and I suspect that he would say that we need more than reasons at the base. The idea would be that the notion of a reason does not have the right sort of strength to make duty; we need something more peremptory. And anyway the characterization of deontic buck-passing in terms of reasons needs a certain amount of refining, because if it is moral rightness that we are dealing with, we should probably restrict ourselves to moral reasons, so that for an action to be right is for the moral reasons in favour of it to outweigh any moral reasons for alternative actions. And if we take this line, we face the difficulty that nobody knows the difference between moral and non-moral reasons; there seems to be no list of considerations that are somehow moral, and no prior distinction between reasons of a moral sort and reasons of a non-moral sort. Luckily, Derek Parfit has shown us how to do the necessary refining.14 We can say that what it is for an action to be morally right is for those considerations that are morally relevant in the case to come down in favour of doing it rather than any alternative. Doing things this way, which avoids explicit appeal to reasons, is probably the most cautious approach15 and so the wisest. And perhaps it would be does not present things in that way. His purpose is to make sense of the idea of a conflict of duties, and the suggestion that he is considering at this exact point is whether the way to do that is to add a notion of obligation that comes in degrees to a Rossian notion of duty proper (which is what Prichard means when he speaks of what we ought to do). And his conclusion is that this cannot work: that since we need a notion of what we ought to do, we cannot add to that a notion of what we most ought to do, or what we ought to some extent to do. I think he should have realized that the same thing holds in reverse: that we cannot add a notion of what we ought (simpliciter) to do to a notion of what we ought to some extent to do, without undermining the latter notion. (I am indebted in these comments to Tom Hurka.) 14   Parfit (2011: 166–7). Parfit’s actual distinction is between decisive moral reasons and morally decisive reasons; he prefers the latter expression, for the reason that I give in the text. I have adapted his distinction for my own purposes. 15   Another advantage of this approach is that it allows that not every relevant consideration is a reason.

More Right than Wrong  115 enough to satisfy any Prichardian sense that our original base properties were too weak to make a duty. There remains the worry that the natural way to understand moral relevance is as relevance to how it is right or wrong to behave, which threatens to take us back to non-scalar conceptions of the right and the wrong. If we deny ourselves such a conception of moral relevance, is any other available? The arguments for evaluative buck-passing are in fact pretty weak, as I think is now agreed. One argument was that the badness of pain does not add to the reasons to avoid it that are given us by the way pain feels. What is bad is always made bad by something, and it is that something that gives us reasons to avoid the bad, not the badness. And this fact can be explained if we think of the badness itself as the presence of reasons that give us reasons to avoid the thing that is bad. Now this is not itself a very powerful argument, even if its starting point were secure.16 And there are also doubts about the starting point; it might be, for example, that the badness of our behaviour gives us reason to regret doing what we did. As far as this goes, the situation is the same for deontic buck-passing. It might be that the wrongness of what we have done gives us reason to be ashamed of ourselves. But there is another and much better reason to be a deontic buck-passer, which is that we thereby avoid trying to add something that is not a matter of degree to something that is, which, if Prichard is right, leads us eventually to incoherence. Everything is done by the weight of morally relevant considerations, and there is no independent quality of rightness to add to that. This reason does not apply to evaluative buck-passing, where everything is a matter of degree, since we are understanding degrees of value in terms of degrees of (strength of) reason. So deontic buck-passing is in a noticeably stronger position than is evaluative buck-passing.

8  Five Worries about Deontic Buck-Passing I committed myself to deontic buck-passing in an early paper on the topic (Dancy 2000), having also argued against evaluative buck-passing. And that is still my position. But I have ever since been worried about various things that seem to be at odds with the deontic buck-passing view that all we need is something that is a matter of degree—whether we talk of degrees of duty, or more simply in terms of stronger and weaker moral reasons or morally relevant considerations. There are three such things to worry about—or perhaps three ways of expressing a sense that we need something that is not a matter of degree. The first derives from a distinction between commendation and recommendation to which I was first alerted by the work of R. M. Hare.17 Hare held that the OED rightly   This was one of the main points of my (2000).   The only relevant remark I can find in Hare’s work is this: ‘According to the O.E.D. “commend” is sometimes used with the sense “recommend”; but this use is not common, and it is not in this sense that the word 16 17

116  Jonathan Dancy says of ‘good’ that it is the most general term of commendation. And he always maintained that to say that a sort of action is right is to prescribe acting in that way. But as I see it, prescription is a form of recommendation, not a form of commendation, no matter in how high a degree. Now of course one can recommend something by saying that it is the best available. But still, to say that it is the best is no more than to commend it more highly than one commends other things. One then recommends it for that reason. It is not as if the options that are less highly commended are also less strongly recommended. Only one of them is at all recommended. So commendation is a matter of degree, and recommendation is not. And we need both. I think the answer to this worry is to point out that the structure of the evaluative is relevantly distinct from that of the deontic, since to say that an action is good in some respect is a sort of partial commendation, but to say that there is some reason to do the action is not to commend it at all. All that we need for deontic purposes is something that is a matter of degree, with recommendation coming in only when the balance of duty has been determined. The basic point perhaps is that commending and recommending are just different speech acts. Nothing that emerges directly from the differences between these two sorts of speech act could serve to require a distinction between two sorts of concept, one to be in operation in acts of one sort and the other in those of the other sort. But there is another way of making the basic point. (Call this the second worry.) Consider the phrase ‘You ought to do what you most ought to do’. Even if it is a truism, this phrase just does not sound like a tautology. If I think I am required to do the act that is most right, and required not to do any of the others, I seem to have moved away from any tautology to something substantial. It is not that I am required to do each of the others severally, only less stringently so as we go down the list, but most required to do the one that comes out top. So even if we allow degrees of duty, we need to add an independent concept of something that is not a matter of degree. Even if the action that is our most stringent duty is what we are required to do, requirements don’t seem to come in degrees. They select one action as the one required. The third thing that has always worried me, as a deontic buck-passer, is the suspicion that there is a notion of the wrong as that which is simply ruled out, and that nothing that is ruled out in this sort of way is more ruled out than anything else that is ruled out. Derek Parfit has expressed this recently with his talk of a notion of what ‘mustn’t-bedone’ (2011: 165–6). This notion is supposedly distinct from any notion of what we have decisive reason not to do,18 and Parfit thinks of it as unanalysable. It therefore occurs in The Language of Morals. We normally use “recommend” when a particular choice is in question, but “commend” when a thing is being mentioned as in general “worthy of acceptance or approval” ’ (Hare 1957: 103n.). This distinction is of course quite different from the one I outline in my text. 18   The notion of a decisive reason, which is certainly available to buck-passers, deserves more attention than it has so far received. Parfit (2011: 32) understands any set of winning reasons as decisive; but this is something that I neither recommend nor commend.

More Right than Wrong  117 appears to be something that resists buck-passing treatment, at least by those forms of buck-passing that are offered as conceptual analysis. The opposite of the ruled-out is that which is required—a requirement. And Maggie Little and Coleen Macnamara have recently raised doubts about whether a requirement can be built out of contributory reasons (Little and Macnamara forthcoming). Some oughts can be so built; Little calls these ‘prescriptive oughts’. But these are merely the acts which one has most reason (of a certain sort) to do. A requirement, as she conceives it, is made in a different way. It is something that no matter what the reasons against, I just have to do. To fix intuitions, consider an order given by a competent authority. If my commanding officer tells me to stand on guard, I do not suppose that his command is just a stronger reason than any reason I had not do to do something else, reasons, perhaps, to curl up in front of the fire and watch the tennis, so that I have to work out whether these reasons are stronger than those. The command somehow obliterates those reasons. And Little and Macnamara conceive of moral requirements as functioning in much the same sort of way. The fourth worry is that there may be oughts without reasons. I tackle this problem in my ‘Reasons and Rationality’, and will not pursue it further here. The fifth point is that we may lose the possibility of tragic dilemmas, which seem to require something more than two equally strong sets of reasons.

9  Attempt at an Answer Some of these difficulties derive from Prichard’s talk about what we ‘most ought’ to do. Thinking of the issue in terms of reasons, we ought to do what we have most reason to do. But this formulation seems to suggest that the only thing we can do with reasons is to add them up. And though Ross and Prichard do seem to have thought in this atomistic sort of way about the nature of moral reasoning, others have not. If we suppose that other relations are possible than the simple one of combining to make a greater or lesser total, we might come to think that the supposed tautology ‘you ought to do what you most ought to do’, far from being a tautology, is not even always true. And if it means that adding up independent normative contributions will always give us the answer, it is not true. Similar thoughts may explain Little and Macnamara’s attitude to what they call requirements. The reasons for fulfilling a requirement are not reasons that simply turn out to be weightier, in combination, than the reasons to do something else instead. Requirements are built in a certain way. And on the same lines we might try to accommodate Parfit’s notion of what mustn’t-be-done. Taking this line of defence leads to a form of deontology that is not exactly scalar, but which still does not seem to introduce a separate concept of duty or obligation in the sort of way that made trouble for Ross. Everything is understood in terms of the interactions between contributory considerations. The fact that we have

118  Jonathan Dancy complicated our story about what goes on at that level does not amount to adding a new upper level of a distinct metaphysical type, which was the source of the original difficulties.

References Broad, C. D. 1940. ‘Critical Notice of W. D. Ross, Foundations of Ethics, Oxford, 1939’. Mind, 49: 228–39. Dancy, J. 1993. Moral Reasons. Oxford: Blackwell. Dancy, J. 2000. ‘Should we Pass the Buck?’, in O’Hear 2000: 59–73. Dancy, J. 2002. ‘Prichard on Duty and Ignorance’, in Stratton-Lake 2002: 229–47. Dancy, J. 2004. Ethics without Principles. Oxford: Clarendon Press. Ewing, A. C. 1929. The Morality of Punishment. London: Kegan Paul, Trench and Trubner. Ewing, A. C. 1959. Second Thoughts in Moral Philosophy. London: Routledge and Kegan Paul. Foot, P. 1978. Virtues and Vices. Oxford: Blackwell. Gay, R. 1985. ‘Ethical Pluralism: A Reply to Dancy’. Mind, 94: 250–62. Hare, R. M. 1957. ‘Geach: Good and Evil’. Analysis, 17(5): 103–11. Little, M. and C. Macnamara. Forthcoming. ‘Between the Optional and the Obligatory: Toward a Deontic Pluralism’. O’Hear, A. ed. 2000. The Good, the True and The Beautiful. Cambridge: Cambridge University Press. Parfit, D. 2011. On What Matters. Oxford: Clarendon Press. Prichard, H. A. 2002. Moral Writings, ed. J. MacAdam. Oxford: Clarendon Press. Ross, Sir W. D. 1930. The Right and the Good. Oxford: Clarendon Press. Ross, Sir W. D. 1939. Foundations of Ethics. Oxford: Clarendon Press. Stratton-Lake, P. J. 2000. Kant, Duty, and Moral Worth. London: Routledge. Stratton-Lake, P. J. ed. 2002. Ethical Intuitionism: Re-evaluations. Oxford: Clarendon Press. Wiggins, D. 1998. ‘The Right and the Good and W. D. Ross’s Criticism of Consequentialism’. Utilitas, 10(3): 261–81.

6 Autonomy and Public Reason in Kant Onora O’Neill

In a striking and convincing comment on Kant’s view of autonomy, Thomas Hill observed some twenty-five years ago that ‘Autonomy is a central concept in contemporary moral debates as well as in the discussion of Kant; but the only thing that seems completely clear about autonomy in these contexts is that it means different things to different writers.’1 Until I read this remark, I had assumed—in the face of considerable textual evidence to the contrary!—that Kant’s discussions of autonomy deal with some variant of the topics discussed in contemporary writing on autonomy. Contemporary views of autonomy see it as some form of individual independence, but disagree about which sort of independence it requires. Views range from existentialist or quasi-existentialist conceptions of autonomy, that identify it with mere, sheer independence in choosing and acting, to a wide range of positions that see autonomy as independence in choosing and acting that meets the demands of one or another conception of rationality. Although Kant too assumes both that agents are capable of independence (they can choose and refuse freely) and that they are rational, his claims about autonomy introduce very different considerations. The textual evidence for the negative claim that Kant does not identify autonomy with any version of individual independence is abundant. In What is Enlightenment?— possibly his most popular discussion of autonomy—Kant identifies it with a distinctive conception of the public use of reason.2 In Groundwork of the Metaphysics of Morals he introduces the Formula of Autonomy as a version of the Categorical Imperative that articulates ‘the idea of the will of every rational being as a will giving universal law’, and claims ‘that the above principle of autonomy is the sole principle of morals can well be   Hill (1992: 76).   Kant (1784: 8: 33–42). References to Kant’s works cite the Cambridge translations, using the standard Prussian Academy volume and page numbers; where short titles are conventional, I use them. 1

2

120  Onora O’Neill shown’.3 In the Critique of Practical Reason he asserts that ‘the moral law expresses nothing other than the autonomy of pure practical reason.’4 In the Conflict of the Faculties he states that ‘the power to judge autonomously—that is freely (according to principles of thought in general)—is called reason’.5 In these and many other passages Kant links autonomy with reason rather than with independence. He neither construes autonomy merely as a form of individual independence, nor speaks of autonomous persons or selves. The conception of principled autonomy that he proposes and defends requires the adoption of principles that combine law-like form and universal scope, so are fit to be offered as reasons to all others.6 Kant identifies autonomy with a specific conception of public reason, rather than with a specific conception of individual independence. Yet this has been obscured both by the very different accounts of autonomy that became popular in the late twentieth century, and by the quite different articulations of public reason proposed by political philosophers in the same period. Here I shall approach Kant’s account of autonomy by a roundabout path, first contrasting his account of public reason with contemporary conceptions of public reason, and then commenting on the considerations that led Kant to link his conceptions of public reason and of autonomy.

1  Public Reason and Deliberative Democracy: Habermas and Rawls Conceptions of public reason are among the big, bold legacies of the European Enlightenment that regained a certain lustre in the political thought of the late twentieth century. Despite this renewed prominence, there is still little agreement about conceptions of public reason, or about its importance. So I shall begin with a sketch of two well-known contemporary ways of thinking about public reason, and then look at some of their implications and limitations, and gesture to the conceptions of democracy to which they are linked. Jürgen Habermas and John Rawls advanced the two best-known contemporary discussions of public reason.7 Their approaches focus on the public among whom reasoning is to take place, but say remarkably little about reason. Public reason is seen simply as debate and discussion, discourse and deliberation that take place between and among the members of the public, variously conceived, hence in the public sphere. In The Structural Transformation of the Public Sphere8 Habermas offered an account of the emergence during the European enlightenment of a public sphere no longer dominated, if still harassed, by the power of Church and state. In this world, communication could take place with a measure of freedom in salons and coffee houses, in

4 5   Kant (1785: 4: 440).   Kant (1788: 5: 33).   Kant (1798: 7: 27).   See O’Neill (2004) for greater textual detail. 7 8   For an overview and comparisons see McCarthy (1994).   Habermas (1962). 3

6

Autonomy and Public Reason in Kant  121 conversation and correspondence, in newspapers and periodicals. This public sphere was not democratic, since participation was open only to those with education and means, but it was politically significant because its swirling discourses were not tightly subordinated to Church, state, or other authorities. Questions of more sorts could be raised and debated by more people in more ways. Habermas saw this Enlightenment world as precursor to a world that supports participation in debate by all competent persons within (perhaps beyond) states, so as precursor to an increasingly democratic (perhaps global) conception of public reasoning that is free from coercion. In later writings Habermas elaborates a dialogical account of reasoning, which he sees as taking place in inclusive, non-coercive discourse among free and equal participants.9 He holds that such dialogue counts as reasoned, provided that certain minimal conditions are met. Each competent subject must be free both internally and externally to take part in discourse, to introduce their own assertions and question others’ assertions, and to express their attitudes, desires, and needs. This account of the conditions of participation in dialogue says little about norms of reasoning: the requirements on which Habermas focuses are simply requirements for participating. This focus on conditions for participation underpins the close links he draws between public reason and deliberative conceptions of democracy, in which, as he sees it, citizens exchange views and seek agreement (however, agreement is neither the only aim of democratic deliberation, nor its invariable result). While any agreement reached by citizens in such deliberation can presumably count as democratically legitimated, it is unclear why, given the lack of any account of norms of reasoning, it should count as justified (a consensus may, after all, be iniquitous, or false). Habermas says nothing about norms of reasoning, nothing about the connections between reason and autonomy of the sort Kant explored in his account of public reason, let alone about contemporary conceptions of autonomy. In his later work John Rawls also put forward an explicitly political conception of public reason. His account embodies certain substantive political norms, but he too is silent about norms of reasoning. In Political Liberalism he wrote: Public reason, then, is public in three ways: as the reason of citizens as such, it is the reason of the public; its subject is the good of the public and matters of fundamental justice; and its nature and content is public, being given by the ideals and principles expressed by society’s conception of political justice, and conducted open to view.10

Rawls sees public reason as taking place among the citizens of a bounded, liberal and democratic political society, who ‘enter by birth and leave by death’ and are willing to accept constraints, provided others too will accept and abide by them.11 Public reasoning as he conceives it is not internal to a specific social or religious group. It addresses

10   Habermas (1995).   Rawls (1993: 213); see also Rawls (1999: 573–615, esp. 574 ff.).   Rawls denied that this bounded society is a state. However, the fact that he depicts it as having and defending geographical boundaries puts pressure on this claim. See O’Neill (1998: 411–28). 9

11

122  Onora O’Neill ‘the public world of others’,12 and can be deployed by fellow citizens with varying comprehensive ethical and political views. The substantive norms presupposed in this account of public reason are commitments to a bounded society and (the search for) its good, to liberalism, and to democracy.13 Rawls too says nothing about norms of reasoning, about connections between reason and autonomy of the sort Kant explored in his account of public reason, let alone about contemporary conceptions of autonomy. Rawls does not see his account of public reason as a generic account of reason, and notes that other stretches of reasoning (conceived of in quite general and conventional terms) may have more restricted contexts and assumptions, so cannot count as public reasoning: Not all reasons are public reasons, as there are the non-public reasons of Churches and universities and of many other associations in society.14

Interestingly Rawls does not see non-public reasoning as defective reasoning: it is not unreasoned. It cannot, however, offer reasons to the public at large, since many will (at best) see non-public reasoning as conditional on assumptions they reject, indeed may not understand. In his later works, Rawls increasingly emphasized both the geographical bounds and the political context of public reason, and followed Habermas in linking public reason to democratic legitimation. His late essay The Idea of Public Reason Revisited begins with the forthright assertion that: The idea of public reason, as I understand it, belongs to a conception of a well ordered constitutional democratic society. The form and content of this reason . . . are part of the idea of democracy itself.15

These prominent accounts of public reason say almost nothing about norms of reasoning, as opposed to requirements for participating in reasoning. Each conceives of reason generically and conventionally, and focuses on necessary conditions for individuals, thought of as citizens, to participate. Each approach has been influential in contemporary discussions of democracy, and in particular of deliberative democracy. It is a small step from these accounts of public reason to accounts of deliberative democracy, in which policy and decisions about public affairs are to be shaped by inclusive deliberation among citizens. However, inclusive deliberation about public affairs is difficult, if not impossible. Citizens have limited time and energy, and many demands on both. Constant deliberation about ‘the good of the public and matters of fundamental justice’ demands a lot.   Rawls (1993: 53).   Nor is this a limitation to be overcome at some future time, when processes of public reasoning have led all to a shared outlook. Rawls sees persisting pluralism of outlooks and ideologies within societies as the natural outcome of the free public use of reason, and sets aside aspirations to justify a comprehensive moral outlook. That is why he characterizes his liberalism specifically as political liberalism, and why his conception of public reason offers not only a basis for legitimation but (arguably) a coherentist justification of principles of justice that is relativized to the political sphere in which it takes place. See Rawls (1993: xv; 1999: 578). 14 15   Rawls (1993: 220).   Rawls (1999: 573). 12 13

Autonomy and Public Reason in Kant  123 One trouble with deliberative democracy, it has been said, is that it takes too many evenings,16 even if deployed only for limited or local decision making. Deliberation and dialogue may offer an attractive, even compelling, model for organizing the affairs of small associations, but are not feasible at great or even at medium scale.17 Perhaps deliberative forms of democracy can be achieved in the smallest polities, or for a few fundamental decisions, but aspiring to deliberative democracy in large polities looks like fruitless nostalgia for small communities and times past. Real world democracies cannot rest much on inclusive public reasoning. Or so it seems.

2  Technological Fixes and Quasi Communication Some optimists think that new communications technologies may overcome these problems and make deliberative democracy possible for larger associations and polities. This is implausible, for many reasons. First, the rapid expansion of access to these technologies in the richer parts of the world may not be matched on a global scale: a digital divide persists. Second, the dearth of evenings, and the many other uses people have for them, are unlikely to end. Third, even those who are free to spend many hours communicating with many others across great distances often prefer to spend that time on social, cultural, educational, professional, or business communication with selected others, rather than on political communication with all others. Some of these problems might be overcome in part, but the deeper difficulties with supposed technological solutions to the limitations of deliberation among large publics arise because they secure only some aspects of communication. Communication technologies, old and new, indeed make it possible to transmit content to many others. This has been possible since the invention of printing, and greatly extended by broadcasting and the internet. But transmitting content need not and often does not secure communication: it is necessary but not sufficient for communication. The new technologies cannot secure global communication or dialogue, because they cannot ensure that inclusive (or even less-than-inclusive!) audiences grasp or even notice what others seek to communicate, let alone participate in discussions of public affairs. Communicative speech remains, as it always has been, vulnerable to failure: it works only when speech acts are not merely accessible to intended audiences, but intelligible to and assessable by them. And where dialogue is the aim, those audiences must be able to respond and be heard, and must actually do so. Indeed, since effective communication nearly always has to be differentiated for differing audiences, neither dialogue nor deliberation can hope to be fully inclusive, whatever the technologies they use. Where standards for effective communication are ignored, the dissemination of speech content amounts only to quasi communication.18 These points may 16   The quip, originally about socialism rather than deliberative democracy, has been variously ascribed to Oscar Wilde, George Bernard Shaw, and George Orwell. 17   Kuper (2006). 18   On the other hand, non-inclusive communication and deliberation among selected members of the public is feasible. Activities such as deliberative polling, focus groups, and citizens’ juries can secure limited

124  Onora O’Neill seem so obvious that they should not need making. And yet they do. We live in a public culture in which we all too often speak of communicating, even when referring to speech that is not grasped by intended, let alone unrestricted, audiences, indeed may not be comprehensible to them, so does not and cannot communicate. Thinking that ignores the needs of audiences is likely to overlook the necessary conditions for effective communication, deliberation, or dialogue, including the necessary conditions for offering others anything that can count as a reason. A selective focus on the production of speech acts that ignores the conditions for their reception, and overlooks the demands of genuine interaction and communication between participants, is apparent in many contemporary discussions of the ethics of speech and communication. Some discussion, and quite a lot of legislation and regulation, focuses on standards that can be satisfied by speech acts that need not communicate with anybody, such as self-expression, dissemination, or disclosure; or on norms that do not require successful communication, such as norms of transparency or openness.19 Other discussions do not focus on speech acts at all, but only on requirements for handling or processing speech content, or for data protection. In such discussions, norms that must be observed if communication with audiences is to be effective, among them norms of reasoning, are easily ignored and marginalized.

3  The Necessary Conditions of Public Reason Superficially Habermas and Rawls appear to draw on Kant’s account of public reason. Their central demand that ‘the public use of one’s reason must always be free’ echoes, indeed quotes, Kant’s very words.20 However, Kant’s account of public reason is very different from those offered by Habermas and Rawls. Rather than equating public reason with actual inclusive participation in communication or dialogue, Kant sees it as a commitment not to preclude the possibility of reasoning with others. Public reasoning, on Kant’s account, may not rely on principles that others cannot follow. His conception of public reasoning is normative and modal, and is put forward as a basis for ethical justification rather than for democratic legitimation (which requires actual deliberation and agreement).

and focused communication and engagement among small groups. These exercises typically ask a limited sample of the public to take a limited amount of time to address a limited range of questions in the light of a limited amount of information provided and formatted by the organizers. The results can be informative and useful, both for those who organize and for those who take part. Such exercises are probably most useful in addressing fairly specific questions, for example about consumer preferences, or about local environmental or recreational priorities. They exclude most citizens, usually pass unnoticed by those excluded, and cannot be scaled up to achieve deliberative democratic governance. 19   Transparency limits secrecy, but often does not secure communication. See Fung, Graham, and Weil (2007); Heald and Hood (2006). 20   Kant (1784: 8: 37).

Autonomy and Public Reason in Kant  125 Kant articulated these points by proposing a surprising way of distinguishing between public and private uses of reason. He characterizes uses of reason that rely on ungrounded assumptions, such as the edicts of Church or state, or theological or ideological dogma, not as public but as private. In What is Enlightenment? he (disconcertingly) classifies the reasoning of those who hold public office (military officers, pastors of the established Church, civil servants) as private uses of reason. As he sees it, these public officials derive their authority from their office, and when their communication assumes that authority, it is not (fully) reasoned so cannot count as public reasoning.21 He contrasts such private uses of reason with ‘the public use of one’s own reason . . . which someone makes of it as a scholar before the entire public of the world of readers . . ., who by his writings speaks to a public in the proper sense of the word’.22 The image of scholarly communication as a paradigm of fully reasoned communication may be self-flattering, but provides a suggestive metaphor for the idea of communicating in ways that do not preclude the possibility of reaching an unrestricted audience. Kant’s distinction between public and private uses of reason is a distinction between private uses of reason that assume but do not justify authority, so may seem pointless or unreasoned to some audiences, and public uses of reason that do not assume authority without justification, which could in principle be followed by any audience.23 So Kant does not see public reason as achieved by universal participation, but as a matter of respecting norms that enable claims to be intelligible and assessable to all. His focus on what is possible, rather than on what is actually done, has profound implications. It supports an account of public reason that focuses not on the context and conditions of actual discourse, but on the normative conditions required if discourse is to communicate with unrestricted audiences. His view is that reasoning among pluralities of agents can work only if participants can offer others reasons, and those to whom they are offered can accept, challenge, or refuse what they are offered, and can offer reasons in their turn. Public reasoning fails if the conditions for giving and receiving, accepting and rejecting others’ claims are ignored, damaged, or flouted. Reasoned discourse of all sorts must therefore be structured so that others can follow it in thought, and practical reasoning must be structured so that others can adopt its recommendations and act on them in suitable circumstances—although, of course, they may decide not to act on reasons that they can follow in thought, or find no occasion to act on principles of action that they have adopted. It is a curious feature of the accounts of public reason put forward by Habermas and Rawls that they say so little to show why freely entered into dialogue and engagement among citizens should count as reasoned. By contrast, Kant’s approach to public reason 21   ‘What I call the private use of reason is that which one may make of it in a certain civil post or office with which he is entrusted.’ Kant (1784: 8: 37). 22   Kant (1784: 8: 37). 23   A corollary of this view is that Kant would see contemporary exercises in deliberative democracy as private uses of reason, because they assume the authority of civilly and socially constituted institutions and practices.

126  Onora O’Neill articulates essential normative requirements not for participation, but for reason giving. If we follow Habermas and Rawls, we have to regard all unfettered speech acts that are made publicly available, however they are constructed or structured, as reasoned. Kant evidently does not accept this line of thought. He insists that public reason demands more than freedom to participate in dialogue or discourse, and depicts those who make public use of their reason not merely as free from constraint, but as deploying cognitive and normative disciplines that permit their speech to reach an unrestricted audience. Consequently Kant can distinguish between acceptable and unacceptable uses of freedom to reason. He does so in many writings, using many different images. One interesting account can be found in What Does it Mean to Orient Oneself in Thinking?, published soon after What is Enlightenment? While the earlier essay argued that rulers should accord freedom to those whose reasoning is fit to be public in Kant’s distinctive sense of the term, the later one focuses on standards or norms that must be respected if discourse is to count as reasoned. The first of these standards is the demand that reasoning be law-like, and so in principle followable by others:  discourse that lacks the structure and discipline necessary for others to follow it does not even offer reasons to others. Kant challenges sceptics about reason, who imagine that freedom from constraint is enough for discourse to count as reasoning, asking ‘have you thought about  .  .  . where your attacks on reason will lead?’24 He argues against those who think that reason requires no more than free, unconstrained discourse, and claims that they will find that the sort of freedom they claim to prize actually stands in the way of communicating with others: we cannot communicate our thoughts or offer one another reasons for belief or action without imposing sufficient structure or discipline on our discourse to enable intended audiences to follow our communication. Lawless thinking simply ignores necessary conditions for reaching possible audiences. Even a private use of reason must enable intended audiences to follow it, so must be law-like. Although a demand that communication be based on law-like principles is quite minimal, and leaves many possibilities open, its justification is not simple. Any direct argument for this or other norms of reason would descend into circularity, and in What is Orientation? Kant takes a more indirect approach, using fiercely sarcastic rhetoric25 to deride those who suggest that reasoning can do without structure or discipline, or that a ‘lawless’ absence of structure could be liberating. He had clear contemporary targets in mind, including purveyors of religious enthusiasm, superstition, and exaggerated views of the powers of genius. Today his targets might include post-modernists, sceptics, and deconstructionists. Such sceptics about reason, he   Kant (1786: 8: 144).   A reasoned rebuttal of deniers of reason is not possible—as those who tried to engage reasonably with post-modernists often discovered. 24 25

Autonomy and Public Reason in Kant  127 suggests, simply fail to see where unstructured, lawless ‘liberation’ of thought and action by itself will lead: . . . if reason will not subject itself to the laws it gives itself, it has to bow under the yoke of laws given by another; for without law, nothing—not even nonsense—can play its game for long. Thus the unavoidable consequence of declared lawlessness in thinking (of liberation from the limitations of reason) is that the freedom to think will ultimately be forfeited and—because it is not misfortune but arrogance which is to blame for it— . . . will be trifled away. . .26

The hubris that expresses itself in such ‘lawless’ thinking leads, Kant suggests, not only to confusion but to cognitive disasters: First genius is very pleased with its bold flight, since it has cast off the thread by which reason used to steer it. Soon it enchants others with its triumphant pronouncements and great expectations and now seems to have set itself on a throne which was so badly graced by slow and ponderous reason. Then its maxim is that reason’s superior law-giving is invalid—WE common human beings call this enthusiasm, while those favoured by beneficent nature call it illumination. Since reason alone can command validly for everyone, a confusion of language must soon arise among them; each one now follows his own inspiration . . .27

Such anarchic, ‘lawless’ discourse lacks resources for reasoned responses to the claims of dogma, rabble rousing and public relations, spin and superstition, and risks confusion and anarchy in thought and action. However, if reason required only reliance on law-like principles, its capacity must be quite limited. A requirement to avoid ‘lawlessness’ can be met by many sorts of reasoning, among them reasoning that defers to the categories and norms of particular political, social, cultural, or religious institutions and practices. Private reasoning too is reasoning. Although reasoning that relies on local categories and norms will inevitably be heteronomous reasoning, and can have no more than local or conditional authority and normative force, it is nevertheless reasoned up to a point. However, on Kant’s account, heteronomous reasoning does not count as a fully public use of reason.

4  Heteronomy and Autonomy: The Form and Scope of Public Reason As Kant sees it law-likeness is not enough for fully public reasoning. Fully public reasoning cannot assume the authority of some arbitrary (‘alien’) norm, standard or power. It must both be law-like and meet further standards. Here, it seems Kant must face difficulties: how are any norms of reasoning to be identified, if they are not to be or defer to the norms promulgated by one or another political, social, religious, or other ‘authority’? Kant addresses this issue by arguing that fully public reasoning requires agents to think, act, and communicate on principles that are not only law-like in form but have 26

  Kant (1786: 8: 145).

  Kant (1786: 8: 145) and see also Kant (1781: A707/B735).

27

128  Onora O’Neill universal scope. Failure to adopt principles with law-like form undermines all reasoned communication, including that which rests on heteronomous considerations; failure to adopt principles that could be principles for all indeed offers reasons (it is after all a private use of reason)—but only to the like-minded: If the will seeks the law that is to determine it anywhere else than in the fitness of its maxims for its own giving of universal law . . . heteronomy always results.28

Those who seek to make public use of their reason must meet a more exacting standard: To make use of one’s own reason means no more than to ask oneself, whenever one is supposed to assume something, whether one could find it feasible to make the ground or the rule on which one assumes it into a universal principle for the use of reason.29

Public reasoning, as Kant conceives it, indeed cannot rely on ‘lawless’ freedom in thinking, doing, or communicating. But it also cannot rely on law-like principles that cannot be principles for all, since these can support heteronomous but not autonomous reasoning. Thinking, discourse, and action are fully reasoned, so count as public reasoning, only if they are based on principles that could be followed in thought or adopted for action by all. As is well known, Kant formulates this combination of demands on the form and scope of reasoning—with more compression than would be ideal!—in the strictest formulation of the Categorical Imperative: act only on that principle through which you can at the same time will that it be a universal law.30 By approaching the nature and authority of reason in this way Kant takes a specific view of the normative requirements for reasoned thought, action, and discourse. Anything that is reasoned must offer reasons to others—which they may or may not consider or accept. Lawless discourse fails to offer others reasons. Law-like but heteronomous discourse offers reasons that are directed to and can be grasped by restricted audiences, so at least achieves a private use of reason. Only law-like and autonomous discourse can offer reasons to ‘the world at large’ and provide a basis for fully public uses of reason. Kant’s conception of autonomous reasoning is a conception of the conditions for fully public reasoning, not for real-time, actual discourse. His arguments are about the necessary conditions for anything to count as reason giving, and his most basic claim about reasoning might be put quite crudely as the thought that if we offer others only considerations that they cannot follow, we simply fail to reason with them. On this reading, Kant’s account of public reason is neither individualistic nor anchored in a philosophy of consciousness, and articulates minimal conditions for reasoned thought, action and discourse. It may still seem puzzling that Kant chose the term Selbstgesetzgebung—literally self-legislation, and by etymological affinity and convention autonomy—for the idea of ensuring that what are offered as reasons can be reasons for all, and are not contingent on specific or variable features of agents. Where is the legislating self in   Kant (1785: 4: 441).

28

  Kant (1786: 8: 146n.).

29

  Kant (1785: 4: 421).

30

Autonomy and Public Reason in Kant  129 this picture? I would suggest that this question reflects a problem in reading and translating Kant more than it reflects his thinking. As I read the texts, Kant cannot be referring to a self who (miraculously?) legislates: if that were his thought, those who adopt heteronomous principles would also be self-legislators, since they too adopt law-like principles. Indeed, it might seem that they would have a stronger claim to be called self-legislators, since the principles they adopt or propose may be self-serving or incorporate self-affecting considerations. But that is not Kant’s thought. The point of the reflexive element selbst in Selbstgesetzgebung is not to suggest that there is some self that legislates, but to indicate that a ‘law’ or principle that gives reasons for all cannot be conditional on or dependent on features found in some but not in others. It is therefore a law to or for itself: not the law of some individual self.31 The image of ‘self-legislation’ suggests not that selves can legislate, but that legislation or principles can be non-derivative: they can be independent of features that are found in some but not in others—i.e. can have unrestricted scope. Self-legislation means not legislation by a self, but (as Kant often puts it) legislation that is for itself, that is a possible ‘law for itself ’ that combines law like form and universal scope.32

5  Public Reason Today Kant’s modal conception of public reason is ostensibly less directly political than its contemporary successors. Those successors link public reason closely with the public and political processes of democracies, yet the conceptions of public reason that they advocate are not feasible at great, or even at moderate scale. As is well known, these approaches have not proved helpful for thinking about justice at a global scale, and offer an insufficient basis for reasoning that could be received by an open ended, universal audience. Could a modal conception of public reason—a more direct descendant of Kant’s line of thought—be relevant to public life today? To test this thought we need to ask how communication might secure the possibility of reasoning with all and any others. Broadly speaking, I think that today this would require the shaping of communicative practices to provide unrestricted audiences the possibility of grasping and testing what others, including institutions, claim and the commitments that others, including institutions, make. This agenda is not wholly absent in current aspirations for public communication put forward by governments, companies, and other institutions, but its patchy implementation reflects, among other things, a failure to consider what reason-giving for an unrestricted audience would demand. 31   See ‘Autonomy of the will is a property of the will by which it is a law to itself (independently of the objects of volition)’, Kant (1785: 4: 440). 32  These points and the textual considerations that lie behind are set out more fully in O’Neill (2004: 13–26).

130  Onora O’Neill Too often it is said that we need more and better communication in some area of life, yet the remedies proposed are not adequate for communication even with restricted audiences. Simplified, let alone dumbed-down, streams of information formatted for wide public ‘consumption’ may not offer an unrestricted audience the possibility of understanding what is communicated, and will not provide genuine opportunities for response and interaction, for check or challenge, or adequate ways of judging truthfulness or trustworthiness. This is one reason why so many claims ‘to communicate with the public’ simply fail. Sometimes they offer considerations that go over the heads of many members of the public; on other occasions they oversimplify and distort; sometimes their glossy presentation raises suspicions that realities and prospects are not as depicted, without offering adequate means to check or challenge what is said; sometimes they invoke more to the demands of public relations than to the needs of genuine communication.33 But there are also examples of effective communication that genuinely seek to be followable and assessable for unrestricted audiences. Such attempts at public reasoning may include: plain speech, openness about sources, straightforward reporting of the limitations of what is communicated, provision of ready ways of replying and of dealing with communicative failures, ensuring that responses reach the appropriate destinations, rather than ‘communications officers’ or recorded messages, and a gamut of other possibilities. Incompletely reasoned communication and quasi communication may dominate public discourse, but this is not inevitable and we could secure greater respect for the epistemic and ethical norms that matter for reasoned communication that is fit to reach an unrestricted audience.34

References Fung, Archon, Mary Graham, and David Weil. 2007. Full Disclosure: The Perils and Promise of Transparency. New York: Cambridge University Press. Habermas, Jürgen. 1962. The Structural Transformation of the Public Sphere: An Inquiry into a Category of Bourgeois Society, trans. Thomas Burger. Cambridge: Polity Press, 1989. Habermas, Jürgen. 1995. ‘Reconciliation through the Public Use of Reason: Remarks on John Rawls’s Political Liberalism’. Journal of Philosophy, 92: 109–31. Heald, David and Christopher Hood, eds. 2006. Transparency: The Key to Better Governance? Proceedings of the British Academy 135. Oxford: Oxford University Press. Hill, Thomas E., Jr. 1992. ‘The Kantian Conception of Autonomy’, reprinted in his Dignity and Practical Reason in Kant’s Moral Theory. Ithaca, NY: Cornell University Press, pp. 76–96. Kant, Immanuel. 1781. Critique of Pure Reason, trans. and ed. Paul Guyer. Cambridge: Cambridge University Press, 1998. 33   Problematic types of quasi communication are often deployed in the name of securing better communication, but are unfit to do so. This is often apparent in company reports, promotional literature, press releases, as well as the ‘creative’ use of small print, league tables, opinion polling, and much more. 34   Earlier versions of parts of this chapter were presented at conferences in Edinburgh and in Oslo. I am grateful for helpful comments on both occasions.

Autonomy and Public Reason in Kant  131 Kant, Immanuel. 1784. An Answer to the Question: What is Enlightenment?, in Immanuel Kant, Practical Philosophy, trans. and ed. Mary J.  Gregor. Cambridge:  Cambridge University Press, 1996. Kant, Immanuel. 1785. Groundwork of the Metaphysic of Morals, in Immanuel Kant, Practical Philosophy, trans. and ed. Mary J. Gregor. Cambridge: Cambridge University Press, 1996. Kant, Immanuel. 1786. What Does it Mean to Orient Oneself in Thinking?, in Immanuel Kant, Religion and Rational Theology, trans. Allen W.  Wood and George di Giovanni. Cambridge: Cambridge University Press, 2001. Kant, Immanuel. 1788. Critique of Practical Reason, in Immanuel Kant, Practical Philosophy, trans. and ed. Mary J. Gregor. Cambridge: Cambridge University Press, 1996. Kant, Immanuel. 1798. Conflict of the Faculties, in Immanuel Kant, Religion and Rational Theology, trans. Allen W. Wood and George di Giovanni. Cambridge: Cambridge University Press, 2001. Kuper, Andrew. 2006. Democracy Beyond Borders:  Justice and Representation in Global Institutions. Oxford: Oxford University Press. McCarthy, Thomas. 1994. ‘Kantian Constructivism and Reconstructivism: Rawls and Habermas in Dialogue’. Ethics, 105: 44–63. O’Neill, Onora. 1998. ‘Political Liberalism and Public Reason: A Critical Notice of John Rawls, Political Liberalism’. The Philosophical Review, 106: 411–28. O’Neill, Onora. 2004. ‘Self-Legislation, Autonomy and the Form of Law’, in Herta Nagl-Docekal and Rudolf Langthaler, eds., Recht, Geschichte, Religion:  Die Bedeutung Kants für die Gegenwart, Sonderband der Deutschen Zeitschrift für Philosophie. Berlin: Akademie Verlag, pp. 13–26. Rawls, John. 1993. Political Liberalism. New York: Columbia University Press. Rawls, John. 1999. ‘The Idea of Public Reason Revisited’, in his Collected Papers, ed. Samuel Freeman. Cambridge, MA: Harvard University Press, pp. 573–615.

PA RT I I I

Social, Political, and Legal Philosophy

7 Private and Public Conscience (Or, Is the Sanctity of Conscience a Liberal Commitment or an Anarchical Fallacy?) Gerald Gaus

1  Conscience and Public Authority:  Hill’s Hobbesian Kantianism In two insightful essays Thomas Hill grapples with the persistent problem of the relation of conscience to the dictates of public authority.1 A traditional view is that one should always follow one’s conscience. “In favor of conscience,” Hill notes, “some argue that we lack moral integrity if we violate our conscientious convictions just because someone told us to” (CA: 260). But if this is so, then it seems that a person of moral integrity cannot acknowledge the authority of others to make decisions for her—at least not on moral matters. She must always act as she thinks best. Hobbes saw this as a poisonous and seditious doctrine, corrosive to the commonwealth. For a man’s conscience and his judgment is the same thing; and as the judgment, so also the conscience may be erroneous. Therefore, though he that is subject to no civil law sinneth in all he does against his conscience, because he has no other rule to follow but his own reason, yet it is not so with him that lives in a Commonwealth, because the law is the public conscience by which he hath already undertaken to be guided. Otherwise in such diversity as there is of private consciences, which are but private opinions, the Commonwealth must needs be distracted, and no man dare to obey the sovereign power farther than it shall seem good in his own eyes.2

1   Thomas Hill, “Conscience and Authority,” in his Respect, Pluralism, and Justice: Kantian Perspectives (New  York:  Oxford University Press, 2000), 260–74, (hereafter referred to as “CA.”), and “Four Conceptions of Conscience” in Ian Shapiro and Robert Adams, eds., NOMOS XL: Integrity and Conscience (New York: New York University Press, 1998), 13–52 (hereafter referred to as “FCC”). 2   Thomas Hobbes, Leviathan, ed. Edwin Curley (Indianapolis: Hackett, 1994), 212 (ch. 29, ¶7). Emphasis added.

136  Gerald Gaus Because there is moral disagreement—our moral judgments, and so the dictates of our consciences differ—those who insist on always acting on their “private consciences” are enemies of social order. We should be guided by the “public conscience” of the law. In these two essays Hill develops a Kantian conception of conscience that concurs with three of Hobbes’s core claims: (i) that conscience is intimately tied to judgment, and (ii) that because our judgment can be mistaken, conscience may be erroneous (CA: 272). And, importantly, Hill concurs with Hobbes’s core claim (iii) that the doctrine that one should always follow one’s “private” conscience undermines public authority, and leads to disasters. After the horrors of the twentieth century, however, we cannot be as sanguine as Hobbes about eliminating reliance on “private conscience.” 3 Hill seems to offer a way out: both conscience and public authority are to be respected but both are fallible. Sometimes we should follow one, sometimes the other (CA: 272–3). Thus Hill seeks to avoid both extremes, the deification of private conscience and the Hobbesian supplanting of private conscience with public authority. In this chapter I examine Hill’s analysis of conscience, and embrace his thesis that the Kantian conception is to be preferred to the competitors he considers. On my reading, however, his account of Kantian conscience is open to two interpretations, the personal and the inclusive. Hill’s ultimate position on the relation of conscience and public authority, I maintain, presupposes the personal interpretation, but the inclusive interpretation is far superior. The inclusive interpretation, though, leads to precisely the doctrine that both Hobbes and Hill oppose—that a moral agent should always follow her conscience. In the final sections I sketch an analysis that seeks to show why, pace Hobbes and Hill, this understanding of the sanctity of conscience is not inimical to public authority, even public authority about the dictates of morality.

2  Three Conceptions of Conscience 2.1  The Religious Conception Hill’s method in these essays is to contrast different conceptions of conscience,4 showing how the competitors to the Kantian view go astray. The first is what Hill deems a “popular religious conception” according to which conscience is a “God-given instinctual access to moral truth” (FCC: 17). Hill focuses on an extreme version characterized by six theses (FCC: 17–18). 3   Hitler is said to have explicitly attacked moral conscience: “Conscience is a Jewish invention. It is a blemish, like circumcision. Providence has ordained that I should be the greatest liberator of humanity. I am freeing men from the dirty and degrading self-mortification of a chimera of conscience and morality.” As quoted by Herman Rauschning and reported in Robert E. Conot, Justice at Nuremberg (New York: Basic Books, 1983), 10. 4   In CA Hill considers three accounts; in the much longer FCC, he deals in some detail with a fourth, Joseph Butler’s. Although Butler’s analysis is of interest in its own right, its consideration is not necessary to appreciate the attractiveness of Hill’s Kantian proposal; I shall thus follow CA’s identification of the alternatives here.

Private and Public Conscience  137 (i) Every human is born with a capacity for conscience, which is the ability to identify morally permissible and impermissible motives, intentions, acts, and aims. (ii) Certain acts are objectively morally wrong, independently of our consciences; if our conscience persuades us they are wrong and we nevertheless perform the act, we have done the additional wrong of acting against our conscience. (iii) Conscience speaks to us authoritatively: we do not view its dictates as advice. Violation incurs guilt and painful feelings. (iv) Conscience is a gift from God that makes moral truth accessible to us, and its authority stems from the fact that its dictates are part of God’s will. (v) It is not primarily a matter of rational judgment as much as an “instinct governed internal ‘voice’ ” (FCC: 18). (vi) “Once we have correctly identified and heard its ‘voice,’ conscience is a reliable source of knowledge of our moral responsibilities” (FCC: 18). On some views conscience is infallible, though we still might make mistakes in interpreting it. The popular view apparently grounds the sanctity of private conscience; it is a Godgiven gift that allows access to the moral truth, and we commit a moral wrong to disregard it. However, Hill suggests that the wide gulf it posits between rational judgment and conscience undermines the apparent infallible authority of conscience. Even if I am clearly “told” to perform some act, unless I form a sound judgment based on moral reasons why it is to be done, I cannot be certain that the voice I am hearing is really that of conscience, as opposed to some less reputable instinctive urging such as self-interest, masquerading as conscience (CA: 265). Thus insofar as it truly is an instinctual voice divorced from reason and reflection, the authority of our inner voice is undermined, for apart from judgment we cannot be sure that the voice is really that of conscience. Moreover, when another confronts us with a claim that her conscience demands that she must perform (or refrain from) some act, unless our inner voice confirms her claim, the soundness of her report to act on conscience must remain opaque to us. She is reporting an “instinctual” reaction; if we do not share it we have no grounds for supposing that it is properly her conscience that is instructing her rather than, say, self-interest or simply a dulled sense of right and wrong (FCC: 20). Thus the ground for respecting the differing consciences of others is undermined, and thus also the social basis for recognizing the sanctity of conscience.

2.2  Cultural Relativism An important aspect of Hill’s analysis is that he confronts what he calls the “cultural relativistic” analysis of conscience. The extreme version of this view advances two theses: “(1) a widely accepted causal explanation of the genesis and social function of the feelings ascribed to ‘conscience’ and (2) the controversial philosophic thesis that what is called conscience is not, even in the best case, a mode of access to moral truth, knowledge, or objectively justifiable beliefs” (FCC: 22, emphasis in original). I am skeptical that these theses imply, or are even consistent with, cultural relativism

138  Gerald Gaus understood as an account of morality; there is certainly nothing in cultural relativism that must deny that conscience involves access to moral truth, or refers to moral judgments with truth conditions, so long as this truth is indexed to contextual (i.e. cultural) facts.5 As I understand him, Hill is actually concerned with a certain sort of social scientific debunking view, according to which after we have done our moral psychology and understood the role of conscience in upholding social norms, there is nothing left to say. Hill quite rightly objects that we cannot fully understand the idea of acting from conscience divorced from some notion of what morality requires. He considers the case of Himmler. In the secret Posen speeches in 1943 in which Himmler laid out the agenda for the Holocaust to Nazi leaders, he proclaimed the “sentence: ‘The Jews must be exterminated.’ ” Himmler adds “what the sentence demands of the man who must execute it is the hardest and toughest in existence.” 6 This hardness required that the executioners steel themselves against pity and sympathy in carrying out their extermination of an entire race, including women and children.7 Even if one concluded that this was in accord with the norms of the Nazi Germany,8 we cannot say that “Himmler’s conscience” instructed him to ignore feelings of pity and sympathy to carry out this “hard” task. “If, as I suspect,” Hill writes, “Himmler’s norms were fundamentally vicious, self-serving, and subversive of morality, then any bad feelings he may have had when thinking about violating them do not deserve to be called pangs of conscience in the usual (partially laudatory) sense” (FCC: 26, emphasis in original). This is not to say, Hill adds, that it is incoherent for me to say of another that his conscience instructs him to do what I believe is morally wrong. We need to distinguish the immoral from the “grossly immoral” (FCC: 27). If I can see your action as moved by an intelligible or reasonable conception of morality, I can employ the laudatory sense of conscience even though I believe your action is immoral. However, at some point what a person calls “moral” is beyond the pale, and then I might say “it was false (corrupt, not genuine) conscience that told him to do that” (FCC: 27). Note that Hill’s argument is directed against the second claim of the “cultural relativist” conception, i.e. that “what is called conscience is not, even in the best case, a mode of access to moral truth, knowledge, or objectively justifiable beliefs.” However, Hill insists, Kantians need not reject all the elements of this empirical conception (FCC: 5   I argue for this claim more fully in “On Seeking the Truth (whatever that is) Through Democracy:  Estlund’s Case for the Qualified Epistemic Claim,” Ethics, 121 (Jan. 2011), 270–300. 6   Quoted in Conot, Justice at Nuremberg, 256. 7   Which Himmler apparently thought was especially “hard.” Conot, Justice at Nuremberg, 256. 8   Which, I think, is doubtful. After instructing them to fulfill this “hard” task, Himmler immediately adds “I ask you only to hear and never talk about what I tell you in this circle.” This indicates why we should be very careful in how we employ examples from the Nazi regime in philosophy; it is all too often supposed that Nazis simply had a different, alien, morality, which they could assert publicly without any inkling that what they were doing was unspeakable. Apparently the Nazi audience went out of the meeting reeling; Albert Speer was to insist that he never heard the speech. Conot, Justice at Nuremberg, 256.

Private and Public Conscience  139 24). “Contemporary Kantians who reject certain aspects of Kant’s metaphysics,” Hill writes, “should expect that the development of conscience can be explained empirically, and, in my opinion, there is no need to deny that conscience requires certain cultural contexts in which to develop” (FCC: 44). As I read him, Hill accepts that an adequate conception of conscience must accord with empirical findings in moral psychology and social theory, including the ways that conscience helps “secure a measure of conformity to local standards without entirely relying on external rewards and punishments” (CA: 266). This yields a desideratum: an adequate conception must account for conscience’s role in supporting social norms.

2.3  The Kantian Conception Hill describes Kant’s conception of conscience as “judicial self-appraisal” (FCC: 31). On the Kantian conception conscience is distinct from moral judgment—it is not an instinct or faculty for determining what is right and wrong, but comes “into play only after one has made, or accepted, a moral judgment” (FCC: 31). Conscience is an involuntary response to the recognition that what we have done, are doing, or are about to do is contrary to the moral judgment that we have made (by applying the moral law to different types of circumstances). Prominent among the many moral judgments that a person of conscience will have made is that they have the special, second-order duty to submit their acts to the “inner court” of conscience, scrutinizing them diligently, impartially, and sincerely. Once they submit their acts to appraisal, conscience gives its verdict and [as Kant says] “passes sentence” automatically, for this is just a metaphor for the painful awareness of wrongdoing that such sincere appraisal causes in a person with the basic disposition of “practical reason.” . . . [W]‌e can say that conscience can acquit or condemn with regard to accusations of both violations of first-order duties (e.g., truth telling) and failures to fulfill second-order duties of due care in scrutinizing and appraising our acts diligently . . . (FCC: 34–5).

Conscience, then, is not a type of instinct or intuition, but a monitoring of our first-order judgments, a determination as to whether we have fulfilled our second-order duty of due care and whether our actions conform to our considered judgments. Consequently, Hill claims that “[a]‌clear conscience is no guarantee that we have acted in an objectively right way, and so it is no ground for self-righteous pride or presumption that our moral judgment is superior to those who conscientiously disagree ” (FCC: 35, emphasis added). A clear conscience is an indication that one has done one’s best—or at least all that can be expected of one—in forming judgments and acting on them. Consequently, “[c]onformity to conscience is necessary and sufficient for morally blameless conduct, in Kant’s view, even though it cannot establish correctness” (FCC: 36). On the Kantian conception “conscience is not a non-verbal signal, like a flashing light. The metaphor represents it as speaking to us, accusing, examining, and passing sentence, in a familiar moral vocabulary” (CA: 270). This inner court determines when we fail to act on our own internalized moral judgments (CA: 270). The “warning

140  Gerald Gaus and the pangs of conscience” indicate to us that we could have done better in our deliberations and our actions—they “reflect our recognition that our conduct falls short of our own standards, they are a reliable sign that we are not doing our best” (CA: 271). Like any court, conscience can be overly harsh and convict us of malfeasance even we have adequately fulfilled our duties. Hill’s insistence that the pangs of conscience are not an immediate reaction like a flashing light indicates that they are more akin to judicial sentences or at least warnings: in our inner conversation we have concluded that we have failed to meet our own standards. If our conscience clears an action, it is a reliable, though by no means infallible, indication that we have lived up to our own moral standards (CA: 271); whether or not we have actually done so is a matter of fact, not a matter of our beliefs about that fact. To act with a clear conscience reflects an inner judgment that we have discharged our duty of due care in our moral deliberation; we have adjudged ourselves to have faultlessly arrived at our decision. This is not to say that we are correct, but that we have been acquitted by our inner court of faulty deliberations. Although, in contrast to what Hill described as the traditional religious view, conscience is not an infallible voice, if a judgment is, as it were, cleared by one’s conscience, it is nevertheless as “reliable a subjective guide as we [one] can get” (FCC: 36). It is, we might say, our best imperfect procedure for arriving at moral decisions, though of course we may be wrong.9 Recognizing that the conscience of each is fallible, Hill argues that the Kantian conception insists that conscience must be “checked by public, reason-governed, critical discussion of the standards that our consciences habitually rely upon” (CA: 272). The upshot is that conscience does not determine what it is objectively right to do. Under the best interpretation it must be respected, for its judgments are reliable, within their limits. But given this view, conscience is never sufficient by itself: only engaging in explicit moral reasoning, with others, enables us to live with a reasonable hope that our moral beliefs are justified. Ironically, assuming the “duty of care,” we cannot even have a clear conscience unless we are willing to check the opinions that our consciences rely upon by engaging in this process of moral reasoning (CA: 272).

There is, as I read it, a fundamental ambiguity in this passage between a personal (or private) and an inclusive (or public) interpretation of the demands of conscience. On the personal interpretation, conscience is akin to a rationalized, judicial, internal voice or faculty that concerns only whether one’s personal reasoning, given one’s own internalized norms and convictions, is adequate. “The general function of conscience is to alert us when we are not doing our best to live up to our own moral standards” (CA: 269, emphasis added). It makes perfect sense to say that I check this “personal conscience” by engaging in a sort of public reason—comparing it to others’ private 9   See John Rawls, A Theory of Justice, revised edn. (Cambridge, MA:  Harvard University Press, 1999), 74–5.

Private and Public Conscience  141 judgments to determine whether I have made a deliberative error or based my reasoning on flawed norms and values. However, insofar as conscience is an internal judge that determines whether one has exercised “due care” in forming one’s judgment, conscience itself would appear to demand that one takes an inclusive view of the relevant evidence and considerations, and so the very formation of conscience requires public engagement. The exercise of due care drives us from merely personal deliberation about our own norms and values to public engagement. Conscience itself demands that one’s personal judgment be publicly checked. After all, how could conscience give a verdict that one has exercised due care when one has not checked one’s deliberations in this way? But if so, then it is only one’s initial, personal, moral judgment, not one’s conscience, that is being checked by public engagement. It is perplexing that Hill thinks that it is “ironic” that we cannot even have a clear conscience unless we check our opinions through public moral reasoning; this would only be ironic if conscience was essentially a faculty or a judgment based only on personal norms, and that such a conscience needed to be checked by others’ reasoning. But there is nothing remotely ironic about an internal judge, insisting on due care, requiring that a conscientiously formed opinion must take an inclusive view of the relevant considerations and so include public reasoning. Indeed, this is essentially a Millian response to fallibility. As Mill pointed out in On Liberty, given the fallibility of one’s judgment, in order to arrive at adequately grounded beliefs one must engage those who disagree.10 On Mill’s view this is part of a good epistemic practice that is required for adequate justification of one’s beliefs; on the Kantian interpretation it would seem part of the due care that conscience itself demands. Unless conscience requires this wider “checking” it would simply be false that it is as “reliable a subjective guide as we can get” (FCC: 36). Rather than saying that “conscience must be checked by public reasoning and exchange,” it seems far more accurate to say “the Kantian conscience demands that initial personal moral judgments be so checked.” It thus seems that the inclusive or public interpretation of the demands of conscience is favored by the “due care” reading of the Kantian conception.

3  The Conflict between Conscience and Authority: The View of the Dissenter 3.1  The Conflict and Hill’s Response Hill asks: “What if conscience conflicts with the direct commands of those who have authority over us?” (CA: 260). To make the question more precise, suppose that Alf ’s 10  Mill, On Liberty, in The Collected Works of John Stuart Mill, ed J. M. Robson (Toronto: University of Toronto Press, 1963), vol. 18, ch. 2. I have called this Mill’s master epistemic claim in my “Controversial Values and State Neutrality in On Liberty” in C. L.  Ten, ed., Mill’s On Liberty:  A Critical Guide (Cambridge: Cambridge University Press, 2008), 83–104.

142  Gerald Gaus country is prosecuting a war that he deems immoral; the duly constituted majority, employing the apparatus of the state, has decided in favor of the war, and calls on Alf to support it, either through taxation or more specific measures, such as being a combatant. Alf believes that the war is wrong, and his conscience verifies this. What is he to do? Hill invokes the Kantian account of conscience: even after Alf has concluded with due care that the war is immoral, he must recognize that he could be wrong. In contrast to what Hill called the religious conception, Alf cannot claim that his conscience is infallible. Moreover, Hill insists that it is not just some “other people,” but a constituted political authority, that is commanding Alf. To acknowledge that they are authorities is to recognize that there are good reasons for them, rather than us, to have the right to make certain decisions. To ignore this crucial point can be disastrous, especially in times of crisis that call for immediate action. Even in the absence of crisis, when we have ample time to reflect, the need for authoritative decisions to coordinate group activities is a vitally important factor that our deliberations, and even consciences, should take into account (CA: 260, emphasis added).

One might think, then, that given the possibly “disastrous” outcomes of ignoring political authority, we should take a Hobbesian route and always defer to public authority. Hill refuses to take this radical route. Recall (section 2.2) that conscience has an ineliminable moral dimension and is not simply a tool for social order; there must, Hill reasons, be times when one cannot defy one’s personal moral judgments—“for example, where to obey would be plainly a crime against humanity” (CA: 261). Hill’s example here is extreme, citing the Nuremberg trials (CA: 261), suggesting perhaps that in a very wide range of normal cases deference to authority is appropriate. In any event, Hill concludes: Regarding the conflict between conscience and authority, my theme has been a modest one: both should be respected, but neither is an infallible moral guide; and if we cannot satisfy both, there is a need, time permitting, to look for a resolution in a process of moral reasoning. In this process we survey the facts of the case, crucially examine relevant arguments, and listen to diverse opinions, considering all of this from a moral point of view (CA: 272–3. emphasis added).

3.2  The Inclusive View and the Mirage of Conflict Because Hill’s account of conscience is open to two interpretations (section 2.3), so is his doctrine about the relation of conscience and authority. Turning first to the inclusive conscience interpretation, it would seem to imply that the individual conscience should always take into account the arguments for, and claims of, authority; Hill is clear that this is part of the relevant features of the act from the “moral point of view” (CA: 273). Due care requires that Alf take up the moral case for doing as the authority dictates; as Hill says, “the need for authoritative decisions to coordinate group activities is a vitally important factor that our deliberations, and even consciences, should

Private and Public Conscience  143 take into account” (CA: 260). If, after Alf takes up the more inclusive view, his conscience still does not approve (that is, his initial personal moral judgment stands after due care has been exercised), then he should conclude that the directives of the political authority provide him with insufficient reason to obey. On this interpretation, once Alf considers the entire problem inclusively from the moral point of view, his conscience will announce which is the correct judgment arrived at with due care—whether to follow his initial personal moral deliberations about the justice of the war or do what the public authority directs. He then should follow this judgment. This interpretation reconciles Hill’s claims (i) that one’s conscience should consider the importance of public decision-making authority and (ii) that one’s conscience is the best subjective guide to moral right and wrong one has. And, after all, one can only act on the basis of the best reasons as one sees them. However, if this is the proper interpretation, the conflict between the claims of conscience and public authority is ultimately illusory, for the original personal moral judgment was only, as it were, a judgment of the first look, before Alf had fulfilled the requirements of due care. After due care, when Alf considers the problem inclusively from the moral point of view and engages with the opposing views of others and the case for political decision making, he cannot conclude that there is an ultimate conflict between the pull of conscience and moral reasons to follow authority. Either, given all the inclusive considerations relevant from the moral point of view (including the moral goods of collective action, and taking into account the fallibility of personal judgment), Alf ’s conscience approves of the political directives, or else he concludes that the authority’s directives are defeated as reasons for him. In either case, at the end of the day the conflict is resolved in favor of one or the other and his conscience approves of the result arrived at with due care. So the inclusive view of conscience in no way denies that one should always follow one’s conscience. To be sure, it insists on a more inclusive view of what should be considered in political contexts: this should include the good of collective action conjoined with a lively recognition that one’s own deliberations might be wrong, so one should give weight to the collective conclusions. As Hill acknowledges, all of these are relevant considerations from the moral point of view (CA: 273). It is important, I think, that appreciating the fallibility of conscience does not undermine this conclusion. Mill provides a case that recognizing one’s fallibility should lead one to adopt certain attitudes towards inclusive engagement with others so as to arrive at the best justified beliefs. And one should embrace a continuing engagement as one recognizes the revisability of one’s judgments in light of further deliberation and evidence. However, having accepted all of this, having done one’s best to exercise due care in forming one’s beliefs and taking account of the claims of political authority, then one has no reason to further defer to the authority of the public on the grounds that one might be wrong. One has already, as it were, factored all that into one’s conscientious judgment through the duty of due care, which is what conscience demands. It thus seems that on the inclusive interpretation of the Kantian conception one should, pace

144  Gerald Gaus Hill, always follow one’s conscience. From the view of the dissenter, there really is no deep conflict: he should always follow his conscience.11

3.3  The Personal Conscience Interpretation and the Problem of Social Norms On the personal conscience interpretation, Alf ’s conscience requires reflection on his personal, internalized, moral convictions; while sometimes it may take account of the claims of public authority, conscience’s due care requirement does not demand such an inclusive view of the problem. Under this conception, a person might genuinely check his conscience by consulting public deliberation or the directives of public authority. Here it seems that we must distinguish what Alf ’s conscience directs from what he has most moral reason to do (i.e. what is ultimately approved of from the moral point of view). Recognizing that his personal moral reasoning on the basis of internalized rules is fallible, and so even after he has fulfilled his conscience’s demand of due care that he may be wrong, Alf sometimes will see that he should defer to the public authority rather than his conscience. This interpretation accords with Hill’s “modest” conclusion that both conscience and public authority “should be respected, but neither is an infallible moral guide” (CA: 272–3): it thus allows for a genuine conflict within Alf ’s moral consciousness between the demands of a public legitimate authority and his conscience, so that sometimes as a good citizen he will act against his conscience. On this view, as a citizen Alf should see both that he has strong moral reasons to act on his conscientious moral judgments while also appreciating that he has strong moral reasons (reasons from “the moral point of view”) to obey the state. But even if he concludes that he should obey the state in this case, this still would go against his conscience because, thinking about things as deeply as he can, he has concluded the state’s command is wrong. Alf is in the somewhat paradoxical situation of recognizing moral reasons to do what he conscientiously believes to be wrong. This interpretation seems more appropriate for a view that sees conscience as a personal faculty of moral judgment—as an automatic “flashing light” (section 2.1)— rather than, as in Hill’s Kantian view, a monitoring second-order judge that determines whether we have exercised due care in making moral judgments. At least on the Kantian view, the idea that there could be considerations from the moral point of view that are outside the purview of conscience seems puzzling.

11   Although Hill does not invoke an exclusionary reasons account of authority, such an analysis would not undermine this conclusion. On an exclusionary account such as Joseph Raz’s, when an authority provides Alf with a reason to go to war despite the fact that his first-order moral reasons indicate that it would be wrong, the authority does not provide countervailing first-order reasons that, if Alf accepted, would defeat his other first-order reasons. Rather, the authority is said to provide second-order exclusionary reasons— reasons not to consider his first-order moral convictions in deciding what to do. But again, if after due care Alf believes that the moral point of view endorses following the exclusionary reasons, then on the inclusive view his conscience would confirm that the morally appropriate thing to do is to obey the authority. On exclusionary reasons, see Joseph Raz, Practical Reason and Norms (London: Hutchinson, 1975), 35ff.

Private and Public Conscience  145 A less obvious problem also confronts this interpretation. As we saw in section 2.2, Hill embraces conscience’s role in supporting social norms—that conscience, by internalizing social norms, helps ensure compliance via guilt and other reactive attitudes. And Hill certainly seems right to embrace this. Good evidence and compelling social scientific models indicate that social norms that are not moralized—that are not internalized and induce feelings of guilt for noncompliance and resentment and indignation when their violation by others is observed—are ineffective.12 Now social norms are best viewed as a type of public authority. To be sure, they are not typically based on a political authority that issues explicit directives, but they are social mechanisms that instruct individuals to act in a certain uniform way even if, on the grounds of their personal deliberations, this is not the best way to act. A social norm confronts Alf much as does a political directive: it says “Perform action ϕ in circumstances C (regardless of whether you think that is the best thing to do).” Even if on Alf ’s view there is a better norm of, say, personal privacy, if he endorses the existing social norm he will typically feel guilt for violating it. For Alf to experience guilt at the violation of a social norm, he must internalize it,13 thus functioning social norms require that Alf internalize the dictates of public authority—he must drawn them within the ambit of his conscience. For example, suppose that inspired by John Stuart Mill Alf does not think it is best to consider a person’s voting behavior as private; he thinks it would be better if people were publicly challenged about their votes and required to defend them. Nevertheless, if Alf accepts the current norm he will not challenge others in this way or criticize them for failing to divulge their votes, and he will feel rude for pressing hard to get them to reveal how they voted. If, however, conscience is really simply about one’s personal commitments and not an inclusive evaluation of the act and the internalization of the necessary public authoritative directives, then we cannot expect such social normative guidance to come under the evaluation of his conscience. But if that is so, then we cannot expect conscience to be the enforcer of such norms—the voice of public authority—and without that we cannot explain conscience’s role in supporting social norms.14 The same sort of inclusive conscience extends to the internalization of legitimate legal directives. Given the great goods secured by a political framework for social order, the Kantian maintains that we have a duty to embrace the norm of fidelity to legitimate law, and so its violation can induce guilt. Just as social norms without moral internalization are ineffective, so too with the legal order. To the extent that the moral

12   I  consider some of the evidence for this in The Order of Public Reason (Cambridge:  Cambridge University Press, 2011), chs. III and IV. 13   For evidence of this, see my The Order of Public Reason, 202–5. 14   Perhaps, one might think, it is not important that conscience considers compliance with social norms so long as the norms are enforced through feelings of guilt. Like any other emotion, however, guilt can be irrational; if one does not believe that one has done wrong, guilt is inappropriate. And conscience is the inner court which determines whether we have done wrong. If there is room in a Kantian view for the idea that social norms are to be upheld by guilt, it must be a rational guilt. See my The Order of Public Reason, 204–5.

146  Gerald Gaus point of view requires acceptance of legal authority over some matters, the moral point of view itself recommends the internalization of the authority’s directives, and so their inclusion in the jurisdiction of conscience.

3.4  Avoiding Fanaticism: Abandoning the Personal Optimizing View of Morality It is hard not to conclude that the Kantian conception of conscience must endorse the inclusive, public, view. Not only does the overall idea of conscience as due care in making moral judgment clearly incline in that direction, but it also seems necessary to account for the role of conscience in upholding social norms and the law. Given that the inclusive conscience takes account of all that is relevant from the moral point of view, it follows that, after all, one should always obey one’s conscience, a claim that Hill set out to deny (CA: 261). Like Hobbes, Hill appears to see this insistence on the sanctity of conscience as threatening fanaticism: based on my inner voice, I refuse to abide by public authority. But the only way for Hill to deny that a moral person always follows her conscience is to accept the personal interpretation of conscience, which I have tried to show is problematic on several grounds. On the inclusive, public, view of conscience, one’s internal judge only will approve of one’s moral judgment as exercising due care if one (i) takes account of one’s own personal deliberations about what moral standards are best (let us call this the personal optimizing perspective), and (ii) considers the conflicting views of others, the moral importance of coordinated action, and common social norms and law. Now some philosophers may resist the latter requirement, identifying moral thinking with an optimizing perspective: thinking through the issue on his own from the moral point of view, Alf decides what is the best rule, or best principle, and that is what morality requires, and so what his conscience dictates.15 But this seems far too narrow. On the inclusive view, Alf must recognize that he is fallible, and that there are reasonable others who often have come to different conclusions. Hill is entirely correct that a plausible Kantian view must take reasonable moral disagreement seriously. Thus due care requires that Alf engage them in public discussion. More than that, however, acknowledging the case from the moral point of view for common action and shared norms, he has moral reasons to deny that his personal view about the optimal rules or principles is determinative of his final conclusion about what is demanded by morality; insisting on the conclusions of his personal optimizing perspective often will prevent Alf from embracing common moral norms, and the requirements of common political action, as moral requirements. Thus in a world of reasonable moral disagreement, the inclusive conscience requires that Alf often modify the conclusions of personal 15   “Where interests conflict, there are many possible regulations dealing with the conflict. The directive embodying the regulation would not be properly moral (as opposed to being legal or conventional) unless it purported to be the best possible way of regulating such a conflict.” Kurt Baier, “Moral Obligation,” American Philosophical Quarterly, 3 (July 1966), 210–26, at 226. Emphasis in original.

Private and Public Conscience  147 optimization deliberation and accept as genuinely moral many common rules that, while falling short of what he judges to be optimal, are acceptable to him as well as reasonable others. In these cases, though, the authority of the public is verified by the authority of Alf ’s conscience: his inclusive conscience transcends his personal optimizing perspective to require that he conform to the dictates of public authority. Of course at some point the moral case for common rules and action will not outweigh Alf ’s judgment that the rules or political action on offer are seriously wrong, and so then his conscience will instruct him to dissent from or otherwise oppose the directives of public authority. Recall here Hill’s comments on crimes against humanity (section 3.1), though there is no reason to think that the inclusive conscience need, as it were, to wait anything like that long before rejecting the claims of public authority. In any event, when Alf comes to this conclusion, the authority of his conscience has determined that the “public authority” is without moral authority. Thus it does seem, as Hill quotes Kant as saying, that one should never obey a command of the state to do what one thinks is wrong (FCC: 42).

4  The Conflict between Conscience and Authority: The View of the Imposer 4.1  The Ideal of Conscientious Engagement Let us suppose that each of us has taken the proper inclusive, public, view, and each person’s conclusions have been vindicated by the court of his conscience. Most moral philosophers, including Kant, have been much too sanguine that competent reasoners will converge in their moral beliefs. Hill concurs: Kant, I think, had too much confidence that all who take up the moral perspectives would reach agreement on moral principles. But in the face of disagreement about matters of vital moral importance, it is clear that his theory implies that the best each of us can do is, first, to make our own moral judgments about what we can sincerely recommend as reasonable to others who will take up the moral legislative point of view and, then, after duly consulting with others and giving due weight to their concerns, to act according to these judgments faithfully but with humility. (FCC: 37)

Consequently, Hill acknowledges, though this Kantian conception implies a certain respect for conscientious disagreement, it by no means bans “coercing someone against their conscience” (FCC: 37). The respect for those who disagree with us in matters of conscience implied by this Kantian conception is surprisingly close to the ideal of “conscientious engagement” advocated by the religious philosopher, Christopher J. Eberle. “[E]‌ach citizen,” on Eberle’s view, “ought to arrive at conclusions of conscience in a conscientious manner. A citizen who respects her compatriots won’t be satisfied with supporting coercive laws that she merely takes to be morally appropriate. She’ll do her level best to ensure

148  Gerald Gaus that the coercive laws she supports really are morally appropriate.”16 This is essentially Kant’s notion of conscience as requiring “due care” in making one’s moral judgments. After ensuring this due care, Eberle holds that a citizen who respects her compatriots (i) will withhold support from any coercive proposal that cannot be conscientiously defended in this way. (ii) She will communicate with her fellows, and seek to explain to them her reasons for coercing them, as well as listen to their evaluations of her reasons, and seek to learn from their moral objections to her proposed coercive policies. (iii) She will seek, as far as she can, to provide them with reasons that they can accept for being coerced. “She will pursue public justification for her favored coercive policies.” Finally, (iv) she will not endorse any policy that, on her view, rests on a rationale that violates or denies the dignity of her compatriots.17 Of course Eberle’s account is not the religious view of conscience as described by Hill (section 2.1); it insists that a sound claim to conscience requires that one be conscientious, and this implies that a claim to conscience involves genuine judgment and “rational justification.”18 But this, I think, shows how Hill’s contrast between the religious view and the Kantian account depended on a specific, and somewhat radical, version of the religious view, which divorced appeal to conscience from rational reflection. A sophisticated religious analysis such as Eberle’s relates private conscience and the imposition of coercive policies by the public in almost precisely the same way as does Kant’s. Once we reject the picture of the religious conscience as ruled by infallible voices that cannot be rationally reflected upon, we see that both Kant’s and a religious-friendly analysis such as Eberle’s allow that, after respectful engagement, the public may rightfully impose a line of conduct based on a view of what is morally correct that violates the conscience of some citizens. Indeed, Eberle’s ideal of conscientious engagement seems more respectful than the account Hill finds in Kant, since a necessary step (i.e. iii) in Eberle’s account is the attempt to show that those being coerced have their own moral reasons to endorse the measure. On Eberle’s analysis we should at least try to find such a justification, though failing to do so after a good faith (due diligence) effort does not make it wrong to go ahead with the policy. Certainly Kant’s view as described by Hill is not more respectful of conscientious difference than the account developed by Eberle.

4.2  Fallibility Contrasted with Reasonable Disagreement In my view, both Eberle’s and Hill’s analyses inadequately respond to the fact of deep reasonable moral disagreement, which is far stronger supposition than fallibility. In holding any belief with which a significant number disagree, the supposition of fallibility is appropriate. After all, infallibility—that we could not possibly be wrong—is an 16   Christopher J.  Eberle, Religious Convictions in Liberal Politics (Cambridge:  Cambridge University Press, 2002), 105. Emphasis in original. 17  Eberle, Religious Convictions in Liberal Politics, 104–5. 18  Eberle, Religious Convictions in Liberal Politics, 104.

Private and Public Conscience  149 extraordinarily strong supposition. A reasonable scientist who advocates a Darwinian account of evolution should suppose she is fallible; before pressing, say, a school curriculum based on evolution, she should monitor her beliefs to ensure that she has taken due care in forming them, and has considered objections and so on. That, however, is hardly an insuperable—or even a very strong—barrier to her proceeding. In contrast, what Rawls calls the “burdens of judgment” applies to a wide range of our moral convictions.19 Because moral questions are so complex, because there are multiple values that must be weighed and there is no clear way to weigh them, because the empirical evidence is uncertain and what we take as evidence is often dependent on our previous beliefs and values, we disagree about the proper answers. And even when we agree on a general principle, our reason leads us to disagree on its proper application or interpretation. In contrast to the evolutionary scientist, the question here is not simply about fallibility, but whether, given the limits of human reason, one can have enough confidence in these judgments to ground what Hill calls “the presumption that our moral judgment is superior to those who conscientiously disagree” (FCC: 35)—a presumption that Hill himself denies. If, though, Betty is not justified in presuming that her moral judgment is superior to the conscientious moral judgments of those who disagree, then on what moral grounds can she justify imposing her view of morality—requiring that Alf act on morality as she sees it rather than as his own conscience dictates? She cannot say that his conscience is too narrow a view, only taking account of his personal optimizing perspective and not the need for public coordination, for she admits that he has taken the inclusive view, and so has exercised due care in considering these wider demands. If she cannot say that her inclusive conscience is superior to Alf ’s (including on the need for coordinated action on this matter), then on what grounds does she claim moral authority over Alf such that he should follow hers, and not his own, conscience? Most of us are not Kantians, so even the most thoughtful derivation that an act is required by the categorical imperative will not be endorsed as justificatory by the conscientious evaluation of other reasonable people, just as Eberle’s conscientious moral judgments based on religious convictions will not be endorsed by many fully competent members of the moral community. If this is the world in which we find ourselves, the question arises as to what warrants Betty in insisting that the conclusions of her public conscience have authority that overrides the authority of Alf ’s own public conscience about what he is required to do. Should Betty (as I think she should) concede that she would indeed be asserting an unwarranted claim to superiority, this would not be demonstrating that she lacks the courage to act on her fully justified though fallible beliefs; she would be refusing to elevate her merely reasonable conscientious convictions into the required public standard for all. Hill is quite right: she ought not presume her public conscience is superior to Alf ’s. But if she cannot presume that, what

19

 Rawls, Political Liberalism (New York: Columbia University Press, 1996), p. xlix.

150  Gerald Gaus warrants her in making Alf conform to it? On the other hand, if she is not warranted in imposing her view, doesn’t this undermine the very possibility of political authority?

4.3  Hobbes’s Choice It may seem that we are being led along the path blazed by Robert Paul Wolff: in the end those claiming authority over others must run roughshod over some people’s conscience. If a Kantian really takes seriously the importance of moral autonomy—a person following her own conscience and making her own moral decisions—then she must reject all political authority. She must be an anarchist.20 On the personal interpretation of conscience, recognizing the sanctity of conscience certainly leads to anarchism. The personal optimizing stance leads us to disagree; if Betty, as the public authority, sees the personal optimizing stance as sacred—as deserving the highest degree of respect—then she will not claim authority over Alf to direct him against his own conscience. As Hobbes saw, it is very difficult to see how any system of political authority is consistent with such a view. It might seem that the inclusive, public, conscience avoids Hobbes’s choice: if Alf ’s conscience takes account of the demands of our common public existence, the conscience of each can endorse public authority. As we have just seen, however, the inclusive view is not on its own sufficient to avoid Hobbes’s choice: given the depth of reasonable pluralism about the right, the demands of each person’s inclusive conscience differ. If Alf and Betty’s public consciences differ, and if Betty represents public authority, she still seems to confront Hobbes’s choice: acknowledge the sanctity of Alf ’s public conscience and so draw back from authoritative imposition on him, or insist that he defer to hers.

5  Reconciliation through Public Justification 5.1  An Uncomfortable Conclusion and Unappealing Alternatives The analyses of the last two sections have led to a rather uncomfortable conclusion. From the perspective of the moral agent, Alf, he should always act on his inclusive conscience—the final court over him as to what is required by the moral point of view (section 3). But Betty, if she is to uphold genuine public authority, seems committed to imposing her inner court’s judgment on Alf (section 4). Or, to be more democratic, Betty and her fellow members of the majority seem committed to imposing the sentence of their concurring inner courts (their public moral consciences) on Alf, even though they also admit that Alf ’s moral integrity requires that he always act on his public conscience, the best subjective guidance he can have. To continue the judicial metaphor, their inner courts are simultaneously claiming jurisdiction over Alf ’s actions while acknowledging that Alf should be guided by no judgment

  Robert Paul Wolff, In Defense of Anarchism (New York: Harper & Row, 1970).

20

Private and Public Conscience  151 but his own. How, though, can any court claim jurisdiction over someone while insisting that he should always follow the directives of another, competing, court? Thus far it seems that the only way to avoid this incoherent view of moral jurisdiction is to either say that sometimes a person ought—morally ought—not follow his own conscience (as the personal conscience interpretation allows) or else embrace anarchism regarding both political and moral authority, and hold that each public person’s conscience has supreme and uncontested jurisdictional authority over him, and so no one may claim authority to direct the morally relevant actions of another. However, both the personal interpretation of conscience and anarchism are deeply objectionable. Can we avoid the uncomfortable conclusion as well as these unappealing ways to escape it?

5.2  The Inclusive Conscience and Public Justification Hill suggests a way out. “Universal agreement” among reasonable persons who disagree, he says, could be “a regulative ideal, perhaps constituting ‘correctness’ about what ‘objectively’ is right” (FCC: 37).21 A more contractual understanding of Kantianism might accommodate disagreement about the right. Rather than seeing moral legislation as an individual decision procedure, or even as each person coming to conclusions based on her own public conscience, we might understand the regulative ideal as one of agreement among diverse perspectives. Of course our judgments are still fallible, but the fallible judgment is now about what is the object of consensus among diverse perspectives (or, we might say, diverse consciences), not what each individual thinks is best from the inclusive moral point of view. To justify a moral judgment one must have a reasonable belief (which is fallible, and which one holds with humility) that all reasonable public consciences in some way endorse it. Suppose Betty takes the inclusive view: she considers not only the various standards supporting and telling against various rules and their prohibited, required, or permitted actions, but also considers the rules (and the interpretation of rules) that reasonable others see as acceptable and the case for coordination on some rule and reasonable interpretations of it. And given the requirement of due care, she engages in public deliberation with others so as to check her conclusions. Because she has rejected the optimizing stance, and recognizes the importance of coordination, her conscience is alert to the necessity that she not insist on what she sees as the best rule or interpretation as the one and only possible moral rule or interpretation (section 3.4). To again employ the judicial metaphor, if Betty’s personal deliberations are evaluated by the court of the first instance, this still must pass the verdict of a review court that insists that she take seriously public disagreement and the requirements of public coordination, and seek a public moral and political order that respects the consciences of all. She takes up the public perspective and appreciates reasonable disagreement in each 21   Yet, he immediately adds “in practice this would only be an aim and a hope.” My aim (or at least my hope) is to show that it can be a real, practical, requirement.

152  Gerald Gaus person’s public conscience about what is the best, or optimal, public authority. Indeed, she acknowledges that the very possibility of a public authority that can be endorsed by everyone’s conscience requires that she renounce the optimizing stance about the best system of public rules, or the best system of public authority. Reasonable pluralism about the right implies that conscientious moral agents, exercising their rational powers as well as can be expected, will not agree about the best or optimal moral rule or principle. 22 Suppose then that Betty orders the set of proposed public moral rules, or legal regulations, over some matter X; call this set {R1, . . . , Rn}. The rule that she ranks first is that which she would choose if she were the Kantian legislator for all. But she is not, and we face deep disagreement, so Betty must consider alternative rules. Because she has accepted the moral case for accommodating the inclusive consciences of others (section 3.4), she will not insist on her first, optimizing, choice, but will be willing to live under other rules too. To be sure, at some point she may simply be unable to endorse some candidate rule Ri as recognizably moral, and so something that she can accept with good conscience; even once Betty has rejected the stance of the lone Kantian, optimizing, legislator for the perspective of a community seeking agreement, there will be some proposals that she will see as beyond the pale (section 2.2).23 Suppose, then, she identified an eligible set {R1, . . ., Rk}. What might happen when she confronts the verdict of Alf ’s final inner court, his all-things-considered conscientious judgment about his eligible set? One possibility is that Alf too has identified a set of acceptable rules, and the sets overlap on just one rule, Rk. If so, their inclusive consciences endorse a common rule as acceptable: they both can act on this rule in good conscience, and obtain the benefits of a common public moral (and by extension, political) life. When Betty demands that Alf ϕ on the grounds that Rk requires it, Betty is employing the public perspective. She can properly demand that Alf perform ϕ rather than an alternative act favored by his personal optimizing stance (or his personal conscience) because his inclusive conscience endorses ϕ action. Consequently, even though it is not best by his personal lights, it is appropriate for him to feel guilty if he fails to ϕ; to fail would be, as it were, to act on his lower court’s judgment (not-ϕ because of his personal optimizing stance) which has been overturned on review. Thus the moral reactive attitudes would be engaged by his failure to ϕ24 so acknowledge the authority of Rk as the public moral rule. Another possibility is that there is overlap between Alf and Betty’s eligible sets, but the overlap is not a singleton, but a subset with multiple elements. They are then confronted with an additional problem: how to select from multiple possibilities such that 22   Cf. the aim of the parties in Rawls’s original position is to agree on the best alternative from a set of options. Rawls, A Theory of Justice, §21. 23   I consider these matters in more depth in The Order of Public Reason, 310–21. 24   See P. F. Strawson, “Freedom and Resentment,” Proceedings of the British Academy, 48 (1962), 187–211. I consider the conditions under which these reactions to the violation of public moral rules are appropriate in The Order of Public Reason, ch. IV. We can extend this analysis to legal rules.

Private and Public Conscience  153 both have reason to endorse the selected rule (principle, etc.). Here the case for a constituted public authority to select from the eligible is strong.25 There are many interesting things to say about how they might go about coordinating on one member of this common eligible set,26 but the crucial point is that public authority endorsed by their inclusive consciences is not a will-o’-the-wisp, but a live possibility. So let us consider the worrisome case, where Betty discovers that Alf ’s set of acceptable alternatives does not overlap with hers. What might they do? First, each might go back and reconsider his or her judgments. Having found that they cannot live together under common public authority without violating the consciences of at least one of them is a troubling conclusion. Thus one or both may decide that that he or she has been too insistent on the personal perspective, and has not taken sufficient account of the great goods of a common public life. It is easy to underestimate the importance of this reason to reconsider. If, as Hill indicates, it can be “disastrous” to reject the claims of public authority (CA: 260), this disaster is morally relevant and gives one powerful reason to think again about what common rules and principles are acceptable. Suppose, then, that Betty has reconsidered. She again examines Alf ’s eligible set, finds no overlap, and concludes that none of Alf ’s eligible rules are acceptable. Her final court of appeal convicts all of them as immoral. In good conscience she cannot live under any, even when giving all due weight to the great value of public authority. What might she conclude about Alf? She may, and in all probability will, conclude that he has failed to arrive at a judgment that adequately discharges the duty of due care. Just because Alf reports that his judgment has been arrived at by due care, it does not follow that Betty must accept this verdict. On the Hill-Kantian account, a report of the demands of conscience is not a report of an infallible inner voice, but a claim to conscientious judgment. But to claim that one has conscientiously arrived at a judgment, or even to sincerely believe that one has arrived at such a judgment, is distinct from actually having done so. A court may proclaim that it has arrived at a verdict with all due care, but an appeals court may strike it down because it did not properly consider some evidence before it, was biased, or rushed to judgment. When Betty disagrees with Alf she must satisfy herself that his claim to have exercised due care is warranted. If she finds it faulty, then she will not take his depiction of his eligible set as a bona fide demand of his conscience. Again, conscience requires that his deliberations live up to standards of due care. In this case Betty may form a warranted belief that, while Alf claims that Rk is not in his eligible set, it really is, and if he exercised due care and diligence he would have come to that conclusion. In this case, Betty’s insisting that he conform to Rk is not a violation of the sanctity of Alf ’s conscience. Some may think this rather high-handed of Betty. One might ask, “Who is she to appoint herself the review court of Alf ’s deliberations?” Yet the cognitive nature of a claim to conscience on the Hill-Kantian view invites others to evaluate your claim that   I make this case in Justificatory Liberalism (New York: Oxford University Press, 1996), parts II and III.   Some of which I consider in The Order of Public Reason, ch. VII.

25

26

154  Gerald Gaus you have exercised due care in canvassing the relevant moral considerations, relevant information, the judgments of others, and the great values of public authority. As with any review court, simply that Alf has arrived at different conclusions is not an adequate ground for reversal, but Betty’s conclusion that due care has not been fulfilled in his deliberations may well ground it. We cannot help but serve as review courts for each other’s claims to conscience, at least once we have rejected conscience as an inner voice that is simply reported by its hearer (section 2.1). If this is true, then, we must modify Hill’s claim that Alf ’s conclusion that he has fulfilled his duty of due care—the dictates of his conscience—shows that he is morally blameless (section 2.3). Although from his perspective he has done all that can be expected of him, Betty disagrees, as does any court of review that determines that a lower court has failed in due diligence. While it is nevertheless the case that any court can only do what it sees as best, it also can be true that it can make an error about whether it has adequately discharged its responsibilities. All this is distinct from Betty’s judgment as to whether Alf ’s views are objectively correct. To reconcile the sanctity of conscience with public authority, the latter must be justified to all those reasonable moral agents who have fulfilled their duty of due care in considering the moral issue at stake. This is more than a hope or an aspiration. As one who calls on the authority of public norms and laws, if Betty is to avoid the uncomfortable conclusion and the unappealing ways of avoiding it (section 5.1), she must have good grounds for thinking that these norms do not violate the conscience of any of those subjected to them. This reconciliation is only possible once we recognize that abandoning the optimizing viewpoint is itself required by due care. Those who insist on their sectarian conclusions—who insist that the only acceptable basis for public moral authority is libertarian, egalitarian, utilitarian, virtue, welfare, or whatever—believe that conformity to their personal public ideals is required by their inclusive conscience, but they have failed to exercise due care and have given insufficient weight to the great good of public authority that can be endorsed by the consciences of all.

5.3  Deep Moral Conflict The last possibility to be considered is that Betty’s court of review may find that Alf has exercised all due care, and yet his eligible set does not have any elements in common with hers, and after careful reconsideration of her own deliberations, she cannot in good conscience expand her eligible set to include elements of his. We now confront two questions: (i) whether Betty may impose her moral views in the sense of requiring Alf to ϕ because her inclusive conscience requires ϕ-ing though Alf ’s does not and (ii) whether she can morally exercise coercion against Alf, to stop him from γ-ing because her inclusive conscience dictates that γ-ing is wrong though Alf ’s does not.

Private and Public Conscience  155 Regarding (i), Betty must accept that any claim to authority she exercises over Alf to ϕ would violate his conscience. Yet she also believes that a moral person should always follow his conscientious opinions, so Betty would again confront the uncomfortable conclusion (section 5.1) that she has final authority over Alf such that she may direct him to ϕ but he has final authority to determine for himself whether to ϕ. Moreover, she would have to acknowledge that as he sees it he is under no duty to ϕ, and so he should not feel guilty for his failure to ϕ, for he has followed the best guide he has as to the moral thing to do—his inclusive conscience. In light of all of this, for Betty to insist that Alf follows the dictates of her public conscience rather than his own must be simply to assert the interpersonal superiority of her own public conscience; but this is just the sort of presumption of moral superiority that reasonable pluralism precludes, and that Hill wisely repudiates. Betty, then, should not insist that Alf ϕs because it is required by morality as she understands it. In many cases this grounds a recognition that Alf is a conscientious objector; Betty acknowledges that he objects to the requirement to ϕ on conscientious grounds, and he is following his own conscience.27 (ii) The second question raises different issues; it is not about a claim to moral authority over another, but a moral permission to act against him. Betty may see that Alf is about to γ, and on her view this is a very serious moral matter, and so she contemplates employing force, or the threat of force, to stop him. Now this need not violate Alf ’s conscience, for if Betty is employing a public authority his conscience deems legitimate, his conscience may approve of the use of force against him to prevent him from γ-ing, not because he thinks γ-ing is reasonably held to be wrong from the public point of view but because he recognizes the great good of a legitimate authority, and he does not wish his dissent on this matter to signal a global repudiation of the authority Betty represents.28 Thus in this case, coercion against him is approved of by his inclusive conscience; Alf is willing to acknowledge the legitimacy of this use of coercion against him. The deepest moral problems occur when Alf ’s conscience does not approve of the use of coercion against him; as he sees it he has a moral liberty to γ, and it is wrong for Betty to employ force to stop him. In contemplating these difficult cases I do not think we should focus on an Alf-qua-Himmler, who is engaging in crimes against humanity, a person who delights in killing or some such maniacal agent. We must remember that, ex hypothesi, Betty has concluded that Alf really has exercised due care in taking account of all the considerations that a reasonable moral agent would see as relevant; he does not merely claim that his conscience permits γ-ing, but Betty agrees that his inclusive conscience, having exercised all due care, leads to this conclusion. The question as to whether Betty can live on 27   These remarks only scratch the surface of exceedingly complex matters. See Rawls, A Theory of Justice, §§56, 58. 28   See Rawls, A Theory of Justice, §§57, 59.

156  Gerald Gaus terms that her conscience requires without violating the conscience of Alf, turns on the deeper question of whether the most outrageous moral wrongs—those that in conscience Betty simply cannot permit to occur—may be approved of by the conscience of Alf who (in Betty’s own view) has exercised all due care in forming his judgments. Hill, we have seen, maintains that one simply cannot conclude this of those who are engaging in grossly immoral acts; to say that another person’s conscience recommends an act is already to see that act as within the range of the morally intelligible (section 2.2). If Hill is right, then the scope of conscientious moral conflict is greatly constrained, and it becomes more dubious that one’s conscience dictates that one has no choice but to employ coercion against the consciences of others. For one must remember that that they are employing their best moral guides and that one agrees that they have done so with all due care with a morally intelligible result, and one’s own views of the demands of morality are not, from the public perspective, superior to theirs.

6  Conclusion: The Possibility of Liberal Order Chandran Kukathas has written that “among the worse fates that a person might have to endure is that he be unable to avoid acting against his conscience—that he be unable to do what he thinks is right.”29 To Kukathas this insight is the heart of liberalism. To Hobbes it is the enemy of social order. Hill aims to balance the liberal and the Hobbesian views, seeking to show that conscience is an important but fallible guide, and so sometimes we should follow not it, deferring instead to the directives of public authority, and we may indeed be coerced in ways that are condemned by our conscience. A reconstruction of Hill’s analysis that focuses on (i) the inclusive conscience and (ii) a commitment to public justification and its rejection of the optimizing stance, I believe, converges with Kukathas’s position. To be sure, as Hill shows us, conscience is not an infallible inner voice, and people can be wrong about what their consciences demand (because they may be wrong about having exercised due care in moral judgment). Yet recognizing the sanctity of the individual conscience is not, as Hobbes thought, inimical to public moral authority, for the inclusive conscience recognizes that its personal optimizing stance cannot dictate what is morally required in certain public contexts. The public conscience repudiates the sectarian, private conscience that so worried Hobbes. This repudiation, however, does not take the form of a public authority that overrides or supplants the individual conscience: the inclusive conscience transcends its personal optimizing stance to take up the public view and seeks accommodation with the consciences of others. 29   Chandran Kukathas, The Liberal Archipelago (Oxford: Oxford University Press, 2003), 55. Emphasis in original.

8 Kant on Three Defenses in the Law of Homicide Jeffrie G. Murphy

There are a number of cases where Kant seems to be admitting that, although in some sense a defendant ought to be punished (even with the death penalty), special circumstances undermine even the legal right and duty to punish (or to apply the death penalty), e.g., shipwreck “necessity” cases, and, in defense of “honor”, a soldier killing in a duel and a mother killing her illegitimate newborn infant. Apparently, these are not bizarre cases of a legal duty of officials to punish despite a moral prohibition against doing so, but rather cases of moral grounds for limiting the legal right to punish in special circumstances. Thomas E. Hill, Jr.1

Introduction Tom Hill is a man whose friendship I value and whose work always stimulates my own thinking, and thus I dedicate this chapter to him with affection and esteem. I have decided to use the above passage from his essay on Kant on punishment as a springboard for my own chapter for three reasons. First, the fact that I have devoted so much of my own philosophical career to the topic of criminal punishment will, I hope, give me some chance that I might have something both informed and useful to say about the three criminal defenses in the law of homicide that Kant notes in his Doctrine of Right (Rechtslehre) with some approval: killing as a necessity and two shame-avoiding “honor” killings: killing in a military duel and bastard infanticide.2 Second, although   Hill (2000: 177–8).   Kant generally does not offer the defenses in these honor killings as complete defenses—ones that should result in complete acquittal—but rather as mitigating defenses that should qualify the convicted defendant for a sentence less severe than the sentence (death) normally and properly given for murder. In the 1

2

158  Jeffrie G. Murphy I some years ago published an essay of my own on Kant and punishment,3 it has been a very long time since I—generally moving away from Kant studies—could claim anything like Hill’s scholarly expertise on the Kantian texts.4 I will here try to hide my scholarly inadequacies, however, by focusing on a quite specific issue: three defenses in the law of homicide, topics on which Kant provides a highly compressed but also highly puzzling and sometimes even disturbing discussion. I select this issue both because of its intrinsic and continuing interest in criminal law theory and because much of what Kant says about it strikes me as quite unhelpful for a variety of reasons. Third, I have always admired Hill’s ability to make sense of and render plausible many Kantian claims that, on first inspection, seem confused or otherwise unacceptable, and it is my hope that I can provoke him into engaging that ability by addressing the issue of Kant on these selected criminal defenses. What Hill says about them in the essay is found in the quoted passage—a passage that offers little commentary, pro or con, about Kant’s discussion of the three defenses noted. Does Hill find what Kant says about these defenses intrinsically acceptable and consistent with the larger theoretical claims in Kant’s moral and legal philosophy, or does he agree with me that they have significant—and perhaps fatal—shortcomings? I am anxious to hear what he might have to say about this and expect that, as usual, I will find what he says illuminating.

Honor, Shame, Bastards, and Duels Before exploring the complex issues involved in claiming a “right of necessity” as a legal defense to a charge of murder, I would like briefly to discuss the two other Kantian examples mentioned by Hill: a soldier killing a fellow soldier in a duel and a woman killing her illegitimate newborn infant. Some of Kant’s arguments on these two issues strike me as instances of unexampled feebleness and in no sense worthy of perhaps the greatest philosopher of the eighteenth century. I believe that Hill is quite correct in interpreting Kant to be saying, not that these are cases where there is a legal duty to punish despite a moral prohibition against doing necessity cases Kant seems to favor total acquittal. The present chapter is an expanded version of the essay “Kant on the ‘Right of Necessity’ and other Defenses in the Law of Homicide” that originally appeared in my essay collection Punishment and the Moral Emotions—Essays in Law, Morality, and Religion (Oxford: Oxford University Press, 2011). 3   See Murphy (1987). What I said in that essay about Kant on “honor” killings now strikes me as quite superficial and even wrong, and I hope that my expanded discussion of such killings in the present chapter is a significant improvement. What I say in the present chapter about Kant on the necessity defense overlaps some of what I said in my earlier essay, but involves some non-trivial modifications—indeed a complete change of mind on my part with respect to some central claims. 4   Indeed, I am not sure that I was ever in a position to claim comparable expertise. My teacher, Lewis White Beck, once told me—after reading one of my early Rawls inspired essays on Kant’s moral and legal philosophy—that he thought that I was destined to become America’s leading authority on what Kant should have said. This was definitely not a compliment.

Kant on Three Defenses in the Law of Homicide  159 so, but rather cases of moral grounds for limiting the legal right to punish in certain circumstances—in short, morally justified legal defenses. But are Kant’s arguments persuasive? For the most part—absent some creative rational reconstruction—I think that they are not. Let me consider primarily the bastard infanticide case, since I think that what I will say about that will hold, mutatis mutandis, for the case of a soldier killing in a duel. If there is a moral reason for limiting the legal right to punish in this case, what is it? One possible reason, of course, would be to claim that the child—being illegitimate— is not a human person at all in any morally relevant sense and thus totally lacks the rights and dignity generally attributed to human persons and protected by just systems of law. But this is surely a vile view and one cannot imagine Kant, of all people, holding such a view. But, alas, consider the following passage: A child that comes into the world apart from marriage is born outside the law (for the law is marriage) and therefore outside the protection of the law. It has, as it were, stolen into the commonwealth (like contraband merchandise), so that the commonwealth can ignore its existence (since it was not ‘right’ that it should have come into existence in this way), and can therefore also ignore its annihilation. (Kant 1996: 477)

As a first and rather obvious objection to this line of reasoning, consider this: Since the infant presumably loses its status as a relevant object of legal protection because it was not legally right that it came into existence in this way, how would a bastard ever earn a right to legal protection since the origin of its existence will remain the same throughout its entire life? And why is the right to kill limited to the mother? Should we not all have a legal right to kill this individual, because of its illegal origin, whether it is 3 days old or 30 years old? And what should our attitude to adult illegal immigrants be, given that they have intentionally stolen illegally into our society? This would seem to support an even stronger case that killing them should not be regarded as murder— or at least not as the most severe degree of murder, murder to be punished by death. Finally, if it is really the case that, with respect to an illegitimate child, we may “ignore its annihilation,” then why punish such infanticide at all instead of merely reducing the sentence to a degree of homicide that does not require the death penalty? We do not punish people at all simply for throwing out their garbage, and Kant’s discussion— whatever its actual intention—invites us to think of the illegitimate infant in an analogous way. How could Kant have been drawn to such a vile view? Although he at least has the decency to add the phrase “as it were,” Kant tries out the following analogical argument in defense of his view: the illegitimate child is, as it were, like a smuggler who sneaks contraband (in this case himself) into society—who has, in Kant’s words, “stolen into the commonwealth.” But this of course will not work since the use of the word “stolen”—suggesting intention or what the law calls mens rea—seems to attribute responsibility to an infant, and that is absurd. Kant quickly abandons this absurdity by adopting a view that is, at least morally, even worse—namely, he conceptualizes the

160  Jeffrie G. Murphy infant as itself contraband—in short, as a thing or object rather than a person who is an end in itself. This would seem to be exactly what the second formulation of the categorical imperative in the Groundwork (Grundlegung) commands us not to do (Kant 1996: 78–85).5 I am reluctant, of course, in haste to convict Kant of the kind of moral imbecility I have so far described. So, to interpret him more sympathetically, we might see in these tortured passages Kant groping toward a distinction not clearly analyzed until much later in the development of criminal law: the distinction between excuse defenses and justification defenses. To justify is to argue that what the defendant has done was right. To excuse is to argue, not that what the defendant has done was right, but rather that the defendant was not responsible—at least not fully—for the wrong that he has done.6 All just legal systems will regard some conduct normally wrong (intentionally killing, for example) as in some circumstances justified (intentionally killing in self-defense, for example) and will sometimes not punish even for wrong conduct if the individual who engaged in the conduct was not a free and responsible agent (as in the insanity defense, for example). How might this distinction bear on Kant’s discussion? Consider his introduction of the concept of shame with respect to the woman who kills her illegitimate infant: “No decree can remove the mother’s shame when it becomes known that she gave birth without being married” (Kant 1996: 477). The fact that a woman, given prevailing social conventions, will feel deeply ashamed in giving birth to an illegitimate child certainly cannot justify the intentional killing of that child—particularly if the conventions are themselves, to use Kant’s language, “barbarous and undeveloped” (Kant 1996: 477). But perhaps the psychological pressure of such shame might at least mitigate to some degree the appropriate legal punishment for such killing—making the killing of a less serious grade, perhaps, than killing done out of cruel malice. Kant speaks of “being constrained by public opinion” in the case of the soldier who kills in a duel, and surely a similar kind of “constraint” could apply in the case of bastard infanticide as well. If one acts from allegiance to conventions that are irrational and immoral, one surely cannot claim that one’s conduct was (all things considered) justified; but the oppressive nature of social conventions just might be regarded by some as mitigating—arguably reducing the crime from capital murder to a lesser degree of homicide carrying a lesser punishment—life imprisonment, perhaps. I have no idea if Kant was trying to grope toward some idea of excuse, but I can see no other reason why he would introduce the concept of shame at all.7 Also, Kant’s claim that legal punishment is justified in such 5   Given the heavy emphasis Kant places on capacity for rational choice as what makes human beings ends in themselves, one might think that he would deny significant moral status to all infants—not just illegitimate ones. There is no reason to suppose he holds such a view, however. 6   Some defenses—e.g. provocation—seem to straddle the distinction since the provocation that reduces a homicide from first to second degree murder or manslaughter must be a provocation that would have aroused a reasonable person. 7   My colleague Mary Sigler has suggested that fear of social ostracism may be just one motivating factor. Another might be the realization that one has done a shameful thing and killing as a desire to distance oneself from that by obliterating its object.

Kant on Three Defenses in the Law of Homicide  161 cases, but that the death penalty may not be, is best explained in terms of a mitigation excuse.8 There are, of course, grounds for being skeptical that the kind of shame endured from acting against a powerful social convention is adequate to trigger a legitimate excuse, even a mitigating excuse.9 Down that road perhaps lies excusing the gang banger who murders because of a fear that, if he does not, he will—given the social conventions operating in certain gangs—be shamed in the eyes of his fellow gang members and perhaps dropped from the gang. Or is it only the shame occasioned by violation of the social conventions of “respectable” middle class people that is to count in the realm of excuses for homicide? That does not seem very decent. Consider a recent case in Phoenix, Arizona. Faleh Hassan Almeleki, a 48-year-old conservative Muslim immigrant from Iraq, was convicted of second degree murder and sentenced to thirty-four years in prison for running down his daughter with his car and killing her. Why did he do this? Because, according to most reports, his “honor” demanded it—demanded it because his daughter had become so “Westernized” that she was exhibiting a level of sin that, in his view, shamed the entire family. (The United Nations Population Fund estimates that more than 5,000 women a year die in “honor” killings for such “crimes” as speaking to unrelated men or being raped.) It is sad that there are still places in the world that condone such mindless barbarism, and surely no civilized society would regard such actions as anything but culpable criminal homicides.10 But are they murders of the very worst sort—murders 8   Fearing shame and social ostracism are typically not such powerful emotions as to render a person delusional or powerless. They are not a kind of insanity. They might, however, make it more difficult to comply with the relevant criminal law than would be the case for persons who lack these fears, and thus fairness might point to a lesser punishment for killings motivated by these fears than for killings motivated in other ways— particularly if the motives in question are thought to be less evil than, say, malicious cruelty. Also, it might be argued that a person who kills in compliance with powerful social conventions can to some degree blame those conventions and argue that society shares responsibility for what has been done. I would not, however, be too hasty in going down this excusing road—particularly if the excuse is regarded as a total defense and not merely a mitigating defense. It is, of course, to ask a lot of a person when one asks that person to go against common social conventions or practices. But it is also true, it should be remembered, that criminal law must target people at the point when they are most strongly tempted to act contrary to what the law requires. Because of this, one must be cautious about recognizing difficulty of compliance as a criminal defense. The American Model Penal Code 2013(1)(b) provides for a defense of “extreme emotional disturbance.” The defense reduces murder to manslaughter if the emotional distress in question has “a reasonable explanation”—i.e. is a kind and level of distress that we think would have a tendency to overwhelm a reasonable person. Though similar to provocation in some ways (see note 6), this defense does not require a provoking act. 9   I am interested in whether Kant’s discussion can cast any light on controversies in contemporary criminal law theory. In fairness to Kant, however, I must admit that I have no knowledge of the degree of shame occasioned in his day and culture by the violation of the two social conventions he discusses. Dueling has the interesting feature that both parties willingly participate in the duel, but in contemporary criminal law consent is generally not a defense to homicide. When I later discuss the issue of the right of necessity, however, I will discuss the possibility (only to be rejected) that certain kinds of consent—hypothetical rational consent or what is sometimes called actual tacit consent—might render some intentional killings justified. 10   For a brief summary of this case see and other sites to which you will be referred from this one. For initial local newspaper coverage of the case, see The Arizona Republic, November 8, 2009, page B14—Leonard Pitts “Viewpoints” essay “Father, daughter, caught in clash of cultures in Ariz.” A month later (December 18, 2009, A2) The Arizona Republic

162  Jeffrie G. Murphy that require the death penalty? Kant thought not, and if he was thinking of mitigating excuses (leading to conviction of a lesser degree of homicide) rather than justifications (leading to total acquittal) the essence of his view may not be so feeble after all. With this background, let me now return to the case of bastard infanticide. There are, even in contemporary America, instances of killing an illegitimate child that might reasonably be regarded as open to a mitigating excuse—classification, perhaps, as a degree of homicide that does not demand the most severe penalties set down for first degree murder. However this generally will not be regarded as appropriate simply because the defendant faces shame. Rather, it will be because the defendant is, say, very young (in high school, perhaps), is poor, is very unsophisticated, and faces—not just shame—but perhaps brutal punishment at home or even being thrown out of the house and made homeless on the streets. (As Cheshire Calhoun has pointed out to me in conversation, the situation of a girl pregnant out of wedlock may take place in an environment of sexual predation by relatives and employers and thus the shame visited on them the result of coerced sex—problems bad enough in our day but no doubt even worse in Kant’s day since even the inadequate legal remedies available to women today were not available in his.) These worries impose pressures far more grave than the mere fear that one will endure the shame that comes from violating certain social conventions. Imagine, however, a woman who does not labor under the other fears that I have noted but who simply fears shame before, let us suppose, her Country Club circle of white friends who would shun her if she revealed herself as the parent of an illegitimate black baby. I see no place for a mitigating excuse of any kind in this case. Indeed, I am rather inclined to see killing an infant for this reason to be an aggravating factor. One final point worth noting in Kant’s highly compressed discussion is his introduction of the concept of the state of nature: Legislation cannot remove the disgrace of an illegitimate birth any more than it can wipe away the stain of suspicion of cowardice from a subordinate officer who fails to respond to a humiliating affront with a force of his own rising above the fear of death. So it seems that in these two cases people find themselves in the state of nature, and that these acts of killing (homocidium), which would then not have to be called murder (homocidium dolosum) are certainly punishable but cannot be punished with death by the supreme power. (Kant 1996: 476)

This strikes me as astoundingly weak as an argument that acts of bastard infanticide and killing in (some) duels are “crimes deserving of death, with regard to which it still remains doubtful whether legislation is also authorized to impose the death penalty.”11 reported a case in which a London judge sentenced to life imprisonment a Kurdish father for murdering his 15-year-old daughter because she fell in love with a follower of a different branch of Islam. Although the killing in the Arizona case was almost certainly premeditated, the prosecuting attorneys decided not to seek the death penalty—perhaps because they thought a jury might contain some aggrieved parents inclined to be somewhat sympathetic to a strict father driven to despair by the disobedient behavior of a near teenage child. 11   Normally one argues that a mitigating excuse should reduce the penalty because of a belief that, morally, the individual having that excuse does not deserve the death penalty. So what could it possibly mean

Kant on Three Defenses in the Law of Homicide  163 First, it is by no means obvious that “legislation cannot remove the disgrace of an illegitimate birth any more than it can wipe away the suspicion of cowardice.” Legislation alone cannot perhaps totally and immediately remove such disgrace, but it can surely reduce it—and perhaps in the long run assist in its total removal. Strictly enforced laws against dueling would surely have some effect in undermining cowardice as the only reason why a law-abiding and law-respecting person would refuse a duel, and strictly enforced laws against bastard infanticide might ultimately lead to social conventions that make killing an illegitimate infant more shameful than giving birth to it. We now generally punish as murder killing in duels and bastard infanticide, and we would no longer take shame over such matters as either an excuse or justification for such killing.12 Our community norms have changed, and the law has probably played some role in this normative shift—just as United States Supreme Court decisions declaring unconstitutional laws criminalizing homosexual sodomy and laws mandating racially segregated schools have probably played some non-trivial role in increasing social tolerance in areas of sexual orientation and race. In short, the law does not always have to follow. It can sometimes lead.13 (A legitimate concern here, of course, is that the law in its attempt to lead to a better future world may be using those in psychological bondage to the prejudices of the current world as means in Kant’s sense.) Second, even if legislation happens to be powerless to remove certain social conventions that stigmatize when violated, it hardly follows from this—as Kant seems to suggest—that, with respect to actions in accord with those conventions, individuals “find themselves in a state of nature.” If there are powerful social conventions that are motivationally operative, then this is certainly not a state of nature of the kind Hobbes described—a condition in which people are motivated to act solely to avoid pain and violent death. But then what kind of state of nature is it? Kant seems to be claiming that legal punishment for the killings in question would be legitimate, but not the death penalty. But what kind of state of nature is it that provides for any legal punishments at all? If there is legitimate punishment, then there is a legitimate civil society and thus no state of nature in any sense that I can understand. If one of the primary purposes that would to say—as Kant does—that the individual deserves to be executed but that legally the penalty should not be imposed? 12   As I have discussed earlier, if we add some complexities to the case and find ourselves dealing not merely with shame but with such things as youth or mental instability (diminished capacity), then we may have grounds for a mitigating excuse. 13   “Since it is the task of the law to form and project, as well as mirror and reflect, we should not, as judges, merely recite the expectations [of privacy] and risks [of loss of privacy],without examining the desirability of saddling them upon society” (Justice Harlan, dissenting, in United States v White, 401 US 745 (1971)). For a stimulating discussion of how the law and the punishments provided by law can play a role in the formation of beliefs in wrongfulness, see Kahan (2000). Kahan argues that when the goal is to change a deeply entrenched norm (e.g. to replace the norm that tends to trivialize date rape with a norm that condemns it as “real” rape) the best strategy is to employ a relatively mild sanction expressing condemnation of the old norm since juries will often not convict if they think the punishment is too severe. This would be a non-Kantian reason for thinking that Kant may be right when he suggests that the criminal punishment for honor crimes should not be the most severe in our punitive arsenal.

164  Jeffrie G. Murphy prompt rational beings to form civil society in the first place is not being realized for certain groups of people (minorities not getting police protection to prevent their being murdered, criminals being put in prisons in effect run by the Aryan Brotherhood, or battered women who cannot get adequate police protection, to use three examples) then I can understand why one might say that those people have been in effect thrown back into the state of nature. I am not, however, tempted to say this for people whose only complaint is that one of their familiar social conventions points one way and, if the law points in another way, they will be put in an uncomfortable bind: face criminal punishment or face the shame of social stigma. This is a tough situation for one to face, no doubt, but to call it the state of nature strikes me as nothing more than grandiose and misleading rhetoric But perhaps the above is a bit too quick. Even if it cannot be literally true that those who kill from honor find themselves in a state of nature, perhaps talk about the state of nature could be a way—a metaphorical way—of highlighting a worrisome problem that could have legitimate impact on determinations concerning severity of punishment—or at least condemnation. If one believes that a legal system gets its legitimacy in part by enforcing shared social or community norms—by encouraging compliance with those norms rather than punishing for loyalty to them14—then might one not be at least tempted to feel outside the scope of a legitimate rule of law if one is punished for doing what most people in the community think is the right thing to do—even if some, being more enlightened perhaps, would not agree? Southern white segregationists in the early days of the civil rights movement often expressed opposition to integration as a “challenge to our very way of life.” Justice required, of course, that force be used to bring them into line on this matter, but one can still—I think—have a certain limited sympathy for their view that their expectations concerning the rule of law were being undermined by the federal government. And this sympathy could to some degree mitigate the unambiguous harshness with which we might otherwise condemn them for peacefully expressing the view (which they had a constitutional right to express) that integration was wrong and should not be forced on them. This might at least make us see such people as something more than simply evil and ignorant racist thugs. They certainly deserve some condemnation (many white Southerners did, after all, see the evil of racial segregation) but surely not the same level of condemnation appropriate to apply to Sheriff “Bull” Connor (and those who applauded him) when he unleashed attack dogs on peaceful civil rights demonstrators or those jeering whites who shouted abuse and spat on the children being escorted by the National Guard to recently integrated schools. If one is inclined to mitigate at least to some small degree the condemnation that one addresses to those who opposed segregation because they saw it as an attack on their way of life, then perhaps one is—at least in a metaphorical 14   Lord Patrick Devlin’s belief that one of the law’s primary purposes is to uphold the shared moral norms of society came to be called “legal moralism.” For elaboration and discussion of this view, see Devlin (1965) and Murphy (2006).

Kant on Three Defenses in the Law of Homicide  165 sense—seeing them as in some sense in a state of nature. This, at any rate, is the best I can do in trying to make at least a little bit of sense of Kant’s claims on the matter.15 So much for honor, shame, bastards, and duels. Let me now move to a more complex issue: Kant on the “right of necessity.”

Shipwrecks and Killing to Preserve One’s Own Life In a passage that matches very closely what Kant says in his other writings about the so-called “right of necessity,” he says this in The Doctrine of Right (Rechtslehre): . . . This imagined right is supposed to give me permission to take the life of another person when my own life is in danger, even if he has done me no harm. It is quite obvious that this conception implies a self-contradiction within jurisprudence, since the point in question here has nothing to do with an 15   That this is the best I can do does not, of course, mean that it is the best that can be done. Tom Hill called my attention to a splendid essay by David Sussman (Sussman 2008). Mustering far more Kant scholarship than I currently possess, Sussman argues (p. 299) that “Kant takes agents to have a moral right to defend their honor. Unlike other rights, however, this right of honor can only be defended personally, so that individuals remain in a ‘state of nature’ with regard to any such rights, regardless of their political situation. According to Kant we should be lenient in these cases because the malefactors are caught between two kinds of authentic normative demand, at a point when the moral authority of the state collides with a certain authority which individuals must claim for themselves.” Also, Justin Tosi, responding to a version of this chapter that I presented at a conference, suggested the following in an email to me: “The state of nature . . . is a relational concept. Persons are in the state of nature relative to each other, concerning certain disputes, if there is no sovereign power in place to resolve those disputes. So, before civil society, Smith and Jones are in the state of nature relative to each other about every dispute they might have. After they join a sovereign commonwealth, they are in the state of nature with each other only with respect to disputes that [the commonwealth] does not regulate. . . . Kant might mean that two officers, Smith and Jones, are in the state of nature relative to one another concerning some shameful incident—[Smith insulting Jones in certain ways, for example]. . . .The law does enjoin citizens from dueling. But it does not forbid all actions that bestow shame. So in response Jones may rightly insult Smith’s honor in some way. But in killing him, he takes more than is owed. And killing is an action that the law is concerned with, so it may of course sanction Jones. But it may also take into account Jones’ right to reclaim some of his honor from Smith. So if Smith’s (private) offense was serious enough, the state could have reason to withhold imposing the death penalty [the most severe penalty for homicide] on Jones.” And Guyora Binder emailed me the following comment after reading a draft of my chapter on a flight back to Detroit after presenting a very stimulating paper at Arizona State University: “I think I understand what Kant means by characterizing the two honor killings as taking place within a state of nature. A society according status on the basis of honor (rather than viewing all moral agents as equals) is not governed by the moral law, and—insofar as status must be maintained through violence—rejects the rule of law. Just as punishment is impossible in the international arena (Binder 2009) it is impossible in a society where security of life depends on a willingness to defend one’s honor. Think of Jack Abbott’s portrayal of life in a maximum security prison, or Maxine Hong Kingston’s portrayal of Chinese village life in The Woman Warrior—which opens with a description of a young woman giving birth out of wedlock forced not only to kill the baby but also to commit suicide.” Finally, Carol Steiker in correspondence made the following relevant observation, an observation that I think is related to the point I was trying to make in my segregationist example: “The woman who gave birth and then tried to raise an illegitimate child would be forever defined by her failings in chastity. The social norms that she would defy if she failed to kill the illegitimate child would have led to a kind of informal punishment, likely for the rest of her life. This informal sanction derives from exactly the same wellspring as the law’s formal sanction—deeply inscribed social norms of morality. When these norms are diametrically opposed in the formal and informal spheres, as they appear to be in this case, there seems to be more than the usual force to the argument for moderating punishment for violation of the formal norm, because of the weight of the informal norm.”

166  Jeffrie G. Murphy unjust assailant on my own life, which I defend by taking his life . . . The question under discussion is whether I am permitted to use violence against someone who himself has not used it against me. It is clear that this allegation [of a right based on necessity] is not to be understood objectively, according to what a law might prescribe, but merely subjectively, as the sentence that might be pronounced in a court of law. There could be no penal law assigning the death penalty to a man who has been shipwrecked and finds himself struggling with another man—both in equal danger of losing their lives—and who, in order to save his own life, pushes the other man off the plank on which he had saved himself. For the first man, no punishment threatened by the law could be greater than losing his life. A penal law applying to such a situation could never have the effect intended, for the threat of an evil that is still uncertain (being condemned to death by a judge) cannot outweigh the fear of an evil that is certain (being drowned). Hence, we must judge that, although an act of self-preservation through violence is not inculpable . . ., it still is unpunishable, and this subjective immunity from punishment, through a strange confusion among jurists, is identified with an objective (legal) immunity from punishment. (Kant 1996: 391–2)

I do not know enough about German legal history to know if what Kant says here captures the law of his day. My interest here, however, is not historical. Rather, I want to see if what Kant says can provide illumination for contemporary criminal law theory—particularly illumination of defenses to homicide. I found very little illumination in exploring his mitigating defenses for bastard infanticide and killing in duels, and I now want to see if his discussion of the right of necessity fares any better. The example that Kant uses to frame his discussion of the right of necessity—Cicero’s (originally Hecaton’s) famous example of two men in danger of drowning contending for a floating plank after a shipwreck—is not an ideal case around which to build a discussion of a right of necessity with respect to homicide. The fact that Kant describes the man who pushes the other off the plank as “taking the life” of that other man suggests—along with the rest of Kant’s discussion—that this is to be understood as a case of murder. But murder is typically understood as the intentional and wrongful killing of another human being, and it is by no means obvious that this is the proper way to understand Cicero’s case. First, it should be noted that this case involves two men contending (each “struggling with another man” in Kant’s words) for a plank in the hope that this plank will serve as a life preserver. This is a plank to which neither man has a right (unless the fact that one man at the moment has it in his control constitutes a kind of “squatter’s right”) and thus the fact that one man ultimately wins out in this competitive struggle does not obviously constitute wrongdoing on the part of that man. It would be extremely nice—perhaps even heroic—for one of these men to withdraw from the competition so that the other might live, but surely there is no wrongdoing in a refusal to withdraw and criminal law does not punish people for failing to be extremely nice. If this is going to be understood as wrongdoing, then an argument for this is surely needed.16

16   In a rich essay (Finkelstein 2001) Claire Finkelstein outlines the history of the plank example from its origin in Hecaton through Cicero, Grotius, Pufendorf, Bacon, Blackstone, and, finally, Kant. In her own

Kant on Three Defenses in the Law of Homicide  167 Second, and more important, murder typically is understood as requiring a mens rea of intention or purpose—i.e. the actor, to be guilty of murder, must have as his purpose (what the American Model Penal Code calls his “conscious object”) the death of the person whose life is lost. It seems to me, however, highly unlikely that the man in this particular shipwreck case has as his purpose the death of the person who loses the struggle. The victor, surely, would be perfectly happy if the man over whom he has been victorious found another plank and saved his life as well.17 The law will sometimes treat knowledge that one will bring about death as equivalent to intention if the knowledge is certain or of extremely high probability, but we have no evidence that that kind of knowledge was present in the case before us. Perhaps there were other planks in swimming distance and the victor over the one plank (perhaps viewed as a particularly splendid plank) hoped that his rival would find another plank that was at least adequate and would survive. At any rate, no civilized legal system would—absent more information—regard the death of the sailor who lost the struggle for the plank as a case of murder—reckless or negligent homicide perhaps (although I suspect not even this) but nothing that would engage even a discussion of the death penalty. Suppose, however, we consider for a moment the case as Kant considers it—namely, as a case of murder that, as a matter of law, should be punished with death but for which the defendant should receive a sentence less than death or even no sentence at all.18 Kant here, in his contrast between the objective and the subjective view of the matter, seems to be noting an important distinction between what might be relevant at a trial to establish guilt and what might be relevant at a later stage at which a judge sets the sentence after conviction—a stage at which certain subjective features of the defendant, perhaps not thought possible to specify in advance and build into the statutory definition of the crime, can play a role.19 discussion of the example, she argues that each of the contending men has an equal right to the plank. It does not follow, of course, that because each man has an equal right to the plank that each man may do whatever is in his power to secure the plank. The idea of two men contending for the plank tacitly assumes, I think, that the competition between them is approximately equal—or at least not grotesquely unequal. Suppose one of the men has a pistol and shoots the other between the eyes in order to keep him from having a chance at the plank. Here I think we would have wrongdoing unless one wants to subscribe to the (to me) utterly implausible principle “If X has a right to Y, then X may do anything at all to secure Y.” 17   The absence of relevant mens rea in this case is noted by Suzanne Uniacke in a valuable article (Uniacke 1996). 18   There seems to be a lack of parallel here with the shame/honor cases. With respect to those, I have suggested that Kant may have been groping toward the articulation of a mitigating excuse rather than a justification—an excuse that reduces the crime from one requiring the death penalty to one that carries a lesser punishment. (However, when Kant says that the law may “ignore the annihilation” of the illegitimate infant, that one isolated remark could be taken as a justification for recommending no punishment at all.) A mitigation excuse, as Austin once said, gets the defendant out of the fire and into the frying pan—but a frying pan that is still in a fire. With respect to the shipwreck case, however, Kant does not say that the punishment should be less than death but rather says that the conduct in question is “unpunishable” and enjoys “immunity from punishment.” So total acquittal and no punishment at all seems to be what Kant is suggesting. 19   For this and other reasons, it might be better to consider such matters at the pardon or clemency stage. For a defense of this view, see Murphy (2007). Also, for a brief discussion of the dangers of pardon and clemency, see Kant (1996: 477–8).

168  Jeffrie G. Murphy What might these be? Somewhat surprisingly, at least for those who want to hold on to the view that Kant is primarily a retributivist, Kant offers only one argument that the sentence for this case of murder should be less than death—namely, that such a sentence can have no deterrence value. There are possible retributivist arguments that could be given—for example, the argument that one who kills under the extreme psychological pressure of fear of death does not deserve the same punishment as one who kills out of malice—but Kant seems to have no interest in arguments of this nature. Even his deterrence argument leaves much to be desired. Kant claims, as though it is an obvious fact, that one facing certain immediate death cannot be affected by a criminal statute that threatens as punishment only a possible future death. I believe, however, that this is by no means obvious—at least not when one sees the full nature of the way in which the law can control behavior. As I noted in my earlier discussion of the shame cases, the law can sometimes be understood as an expression of community norms and can even have some effect in molding those norms. So one might respect and obey such a law, not merely because one fears the threatened punishment, but because one sees the law as a particularly emphatic reminder (by imposing the death penalty) of an important community norm that one should obey out of duty and not just fear. I  believe that philosopher C.  D. Broad, in speaking of a situation of lifeboat survival, once said that he hoped that, if he were ever faced with such a situation, he would not engage in some unseemly fight over the small bit of remaining food but would rather, like a gentleman, go to his death in good order. Even if we consider only the simple motive of fear of death, however, is it possible that Kant might have been tempted by the fairly common view that the fear of immediate death is so strong that it will trump all other motivational considerations? This view is clearly mistaken, of course, as we know from the many cases in which, for example, sailors will—at great risk to themselves—save passengers before themselves on a sinking ship and ship captains, at least in the old days, felt a duty to go down with their sinking ship or, at least, not to leave the ship—again at great risk to themselves—until they were certain that all passengers and crew had been evacuated. These people—like those who sacrifice for comrades in battle—we regard as heroes, although they probably think of themselves as simply doing their duty or what is honorable. At any rate, the cases—although perhaps not the norm—certainly show that it is not beyond human possibility for a person to regard some non-self-interested considerations as of sufficient importance to justify his facing immediate death. These heroic sailors and captains surely feared death, but they feared something else even more: a shameful death, one in which they failed to do their duty and thereby failed to live up to what Freud called their ego-ideal. And their sense of duty may have been formed, at least in part, by the fact that the law provides severe penalties for a failure to suppress what Kant called “the dear self ”

Kant on Three Defenses in the Law of Homicide  169 (which would include the fear of immediate death) and, in some circumstances, requires the willing sacrifice of one’s own life.20 Of course, if one accepts Kant’s fear-based version of a deterrence rationale, his claim that there should be no punishment at all in the shipwreck case he discusses makes perfect sense. Since he rules out the death penalty because of his belief that it will not deter (the fear of immediate certain death always being stronger than the fear of a possible future death), then a fortiori any punishment less than death must be ruled out as well. However, if one adopts (as I have suggested here) a richer interpretation of the deterrence rationale—or employs a different (perhaps retributivist) rationale—one might find a basis for a mitigating excuse that would reject the death penalty but still allow some other punishment. In spite of these noted shortcomings, Kant’s discussion does show some appreciation of two important matters: (1) the distinction between what is relevant in defining a crime (of primary relevance at trial and conviction) and what may be relevant at the sentencing stage and (2) the related distinction between regarding conduct as “objectively” justified as a matter of right and seeing that, given certain “subjective” features present in the defendant, the conduct of that defendant might be at least partially excused for doing something he has no objective right to do. As noted in my earlier discussion of shame defenses to homicide, Kant sometimes seems to be groping toward a distinction in law and morality that is now quite familiar: the distinction, not always sharp of course, between justification (defendant had a right to do it) and excuse (defendant cannot be fully blamed for doing something he had no right to do). But is Kant correct in thinking that, other than in cases of legitimate defense against an unlawful aggressor, intentionally killing to save one’s own life or the lives of others, though perhaps sometimes excusable, is never a matter of objective right—that there is no legitimate doctrine of a “right of necessity”? Is the very concept of such a right, as Kant says, “a self-contradiction within jurisprudence”? This is the issue to which I shall now pass.

Another Shipwreck: The Case of Regina v Dudley and Stephens Like many other teachers of criminal law, I  often start my class with this famous case of murder and cannibalism on the high seas—a much better illustration of a shipwreck necessity case than Cicero’s.21 In 1884 the yacht Migonette went down 20   Recall that Kant, in discussing a soldier fighting a duel with another soldier, accepts that motivational considerations of honor can represent “a force . . . rising above the fear of death.” And sometimes, as noted by Suzanne Uniacke (Uniacke 1996: 120), “people can refrain from wrongful conduct because they believe it to be wrongful.” 21   [1884] 14 QBD 273 DC. This is a better illustration because it is a clear case of intentionally killing another human being—seeking the death of another human being as one’s conscious object. This claim will not be defeated by someone saying that their conscious object was to save their lives. For full details of the situation faced by Dudley and Stephens and of the case itself, see Simpson (1984) and Hanson (2000).

170  Jeffrie G. Murphy very far away from land. The four crewman Dudley (the captain), Stephens, Brooks, and Parker (a young boy at sea for the first time) managed to get into a small lifeboat—a boat with very little food or water packed on board—in the hope that they might find land or be rescued. They drifted for twenty days and two cans of turnips in the life boat and one small turtle, which they managed to catch, was their only food. They ran out of water and began what they reasonably believed was their rapid progress to certain death from starvation and dehydration. Parker, with no experience at sea, made the mistake of drinking seawater and became so ill and delirious that it was reasonably believed by the others that he was fading fast— that his death, given the poisonous nature of seawater and his illness and delirium, was inevitable. On the nineteenth day, Dudley, Stephens, and Brooks discussed killing Parker in order that they might eat his flesh and drink his blood and thereby increase their chances, by living longer, that they would survive long enough to be rescued. Dudley proposed the drawing of lots to determine who would be sacrificed—often thought of as the “custom of the sea” in such cases. Brooks said he would have no part in any of this—saying that he would rather die than murder or be murdered—and lots were not drawn.22 Dudley and Stephens then agreed to kill Parker—selecting him both because they believed he had no chance of survival anyway and because, unlike Dudley and Stephens, he had no family responsibilities. Dudley then approached Parker, said a prayer asking forgiveness “in case they should perform a rash act,” told Parker his time had come, and then killed him by slitting his throat. Dudley, Stephens, and Brooks (willing to eat the flesh but not to kill in order to do so) all fed on the body of Parker and drank his blood. They regained some of their strength, certainly enough to live a few more days, and four days later they were rescued.23 When they arrived at their home port, they were warmly welcomed—most of those in this community of mariners apparently believing that they had, of course, done the right thing in killing one so that they could survive—particularly since it was believed that Parker could not have survived no matter what.24   One might, of course, be tempted by the cynical view that Brooks was simply taking advantage of the fact that he believed that Parker would be killed (and provide Brooks with life sustaining sustenance) because Dudley and Stephens would cast two votes for the killing and Brooks would cast only one against— thereby giving Brooks the best of both worlds: retaining his moral (and legal) purity while still getting something to eat. Even if this was true of Brooks (and we have no way of knowing) it would in my view be overly cynical to assume that any person who expressed the view expressed by Brooks would simply be engaged in strategic behavior. 23   Why did Dudley say a prayer asking forgiveness if he thought his action justified? He actually was said to offer a prayer “asking forgiveness of them all if either of them should be tempted to commit a rash act, and that their souls might be saved.” Did the reference to the possibility of a rash act perhaps involve a recognition of human fallibility—a realization that even reasonable beliefs are sometimes not true beliefs? 24   I think that this is a much less controversial justification for selecting Parker than the argument that Dudley and Stephens had families and Parker did not. Having family responsibilities is a morally relevant factor, of course, but so is the loss of any opportunity to have a family or participate in other aspects of adult life. 22

Kant on Three Defenses in the Law of Homicide  171 Their warm welcome did not last very long, however, since those administering the English criminal law system—a group not dominated by individuals socialized as mariners—took a dim view of intentionally killing “a weak and unoffending boy,” not in self-defense against unlawful aggression, but simply to prolong one’s own life. They saw this as the intentional killing of an innocent and thus a non-controversial case of murder—the kind of premeditated killing that would require the punishment of death. (English law had long been tolerant of the “custom of the sea” and had not prosecuted for murder in cases such as this one, so it appears the government wanted to use this case to stop the law from seeming to give tacit approval to the custom—thereby ignoring the reasonable expectations of innocence that Dudley and Stephens had, given previous prosecutorial indifference, probably formed.) Dudley and Stephens and many of their friends were quite surprised by the decision to prosecute them since they thought they had a defense that would certainly earn them acquittal in any criminal trial: necessity (what is now sometimes called a choice of evils or a lesser evil defense). The intuitive idea was this: It is better that one person die rather than that all die—what they reasonably believed would happen if they did not kill and eat one of the crew members. In short, it was reasonably believed that such killing (clearly an evil) was necessary in order to avoid the greater evil of all dying. Also, it was reasonable to believe that Parker should be selected primarily because of his already terminal condition. But was such a defense to be recognized in English law? The trial court was unwilling to rule on this matter. They simply agreed on the facts (as sketched above) and then sent the case to the House of Lords (the institution in England that is closest in function to the United States Supreme Court in that it has the last word as a matter of existing law) for a “special verdict.” The special verdict would function to answer this one question: If the facts were indeed as presented (mainly facts about what was reasonable for Dudley and Stephens to believe) did this constitute capital murder or was there rather a defense, often called “the right of necessity,” that would merit acquittal or at least a reduced sentence? The Law Lords refused to recognize such a defense. They noted that the available legal authorities (with the possible exception of Lord Bacon) recognized as a defense only the necessity of self-defense involved in repelling an unlawful aggression—certainly not the issue in the present case. Given this absence of legal precedents, and no statutory authority either, the Lords were unwilling to preempt the legislative function and carve out, on their own, a new defense to murder. So, given existing law, Dudley and Stephens were found guilty of murder and sentenced to death. The Lords did not merely state existing law, however, but they also spoke with some eloquence (and some purple prose) in defense of their view that the existing law was absolutely right and that it would be unjust, unwise, unflattering to human nature, and even unchristian for anyone to introduce such a defense of necessity into English law. It would be unjust because this was a case of intentionally killing the innocent (a crime that surely deserves the death penalty), unwise because such a defense could so easily

172  Jeffrie G. Murphy be abused (letting a man judge necessity when he is likely to profit from such a judgment25), unflattering to human nature since it failed to take note of the recorded heroic deeds of sailors sacrificing themselves so that others might live (thereby proving that human nature is not as cowardly as those favoring a necessity defense seem to think), and unchristian because the example set by “the Great Example whom we profess to follow” set for Christians the standard of sacrificing ourselves for others, not a standard of sacrificing others for ourselves. In short: The Law Lords reached exactly the same conclusion as Kant on the supposed right or defense of necessity: there is no such legal right.26 Interestingly, they also hinted that they, like Kant, held the view that no punishment or at least a reduced sentence would be appropriate in a case like this. The Lords suggested that mercy should probably be shown to individuals such as Dudley and Stephens—suggesting that this might be a good case for the sovereign to exercise the power of pardon or at least reduce the sentence.27 Although they agreed with Kant on conclusion, however, Kant used none of the Lords’ rationales to get to that conclusion and the Lords did not use the only one that Kant offered: the argument that no deterrence purpose would be served by a criminal law threatening the possibility of a future death to someone believing that he now in fact faces the certainty of immediate death. When I first wrote on the necessity defense, I thought that both Kant and the Law Lords were mistaken about this matter and that, at least as a matter of principle, a justification defense of necessity should be recognized in cases such as Dudley and Stephens (Murphy 1987). The argument I gave I took to be Kantian in spirit since it drew on some of the hypothetical contractual or bargaining models that one occasionally finds in Kant and that are present (apologies to the ghost of Lewis White Beck28) in the greatest Kantian in twentieth-century moral philosophy: John Rawls.29 Simply put, I suggested the following: Imagine a group of rational sailors coming together to agree on rules to govern lifeboat survival situations of the kind faced by Dudley, Stephens, Brooks, and Parker. It seemed to me that they would surely select 25   This is not strictly correct, of course, since a jury would typically have to agree that the killing was indeed necessary (or at least reasonably believed to be so) before an acquittal on grounds of necessity. 26   I am not, of course, suggesting that the Law Lords had any knowledge of Kant’s views on the matter. 27   As reported in Simpson (1984) and Hanson (2000), there was a behind the scenes deal struck between the Law Lords and the Crown—perhaps explaining why the possibility of a manslaughter conviction was not considered. Dudley and Stephens were to be convicted of murder and sentenced to death, and the Crown was to pardon them—a way of being sympathetic to the great stress under which Dudley and Stephens operated that would not involve recognizing a new defense and thereby changing the definition of murder. Sovereign pardon does not recognize a new defense or change the definition of murder since it sets no precedent. In spite of what may have been agreed to behind the scenes, Dudley and Stephens were not pardoned. Their conviction stood but their sentence was reduced to six months. 28   See note 4 in which I talk about Beck’s less than enthusiastic response to my Rawlsian interpretations of Kant. 29   As an example of the use by Kant of agreement models to illustrate rational and moral choice, see his “On the Common Saying: That May be Correct in Theory but is of No Use in Practice” (Kant 1996: 277–306) and, for an analysis of that essay, Murphy (1998). The Rawlsian use of hypothetical consent models is to be found in Rawls (1971).

Kant on Three Defenses in the Law of Homicide  173 a rule of the form “kill one to save the many” since each sailor would stand a better chance of survival given this rule than if they adopted the rule “never kill one even if that means that all will die.” If Sailor Bill found himself in a grim situation comparable to that faced by Dudley, Stephens, Brooks, and Parker, he would, of course, not be at all pleased if he became the one selected to be executed and served as food and drink to his shipmates. But neither would he be pleased to die along with the rest of his shipmates if nobody is sacrificed. So, as a rational being seeking to maximize his chances of survival in lifeboat situations, Sailor Bill would surely agree (as would all other rational sailors) to a rule allowing sacrifice of one to save many in circumstances of the kind here described.30 I have now, under pressure from some very astute critics, come to the view that my earlier view simply will not work and only seems to by begging some very important questions.31 Some sailor (take Brooks as an example) might believe that it is always wrong to kill the innocent—no matter what—and I now see no good reason for thinking that someone who subscribes to such an absolute principle (a version of the venerable principle “Do no evil even if good will come of it”) is in any sense irrational—even if he subscribes to this principle out of religious conviction that does not rest on what Rawls calls “public reason” (Rawls 1993). Of course, the sailor who feels this way could always volunteer to be the one sacrificed—unless, of course, he regards that as suicide and regards suicide as absolutely wrong, another view that does not clearly establish him as irrational. These problems show some of the limitations of hypothetical consent models—particularly the danger of getting the result one wants (and tacitly begging the question at issue) by placing arbitrary constraints on what can count as rationality. Of course, a rational sailor will not agree to a principle that one may be sacrificed for all unless there is a procedure of selection to determine who that one will be—a 30   In a famous essay (Posner 1985) Richard Posner discusses the Dudley and Stephens case and argues, in defense of recognizing a right of necessity in the case, that “at some point the sacrifice of one person so that others will live must increase social welfare.” He then discusses the principle of drawing lots and the principle of sacrificing the weakest and finds advantages and disadvantages in both. Finally, he briefly notes in passing that hypothetical antecedent agreements might be relevant if they are indeed ways of establishing economic efficiency since he believes that we should “always [assume] that economic efficiency is to be the guidepost for criminal law doctrine.” Posner offers what are in effect utilitarian arguments (arguments of wealth maximization or economic efficiency) as the foundation for his conclusion, and it’s thus not surprising that a defense of necessity emerges as justified on such a theory. Posner, like all utilitarians, would no doubt brush aside Kantian or religious scruples here as a kind of sentimental irrationality. For a critique of the employment of Posner’s economic analysis to issues in moral, legal, and political philosophy, see Murphy (1986). 31   I am grateful to Guyora Binder, Cheshire Calhoun, Richard Dagger, Joshua Dressler, Antony Duff, Carissa Hessick, Sandra Marshall, Jerome Neu, Mary Sigler, Carol Steiker, Justin Tosi, and Michael White for very helpful comments on this and other issues discussed in the present chapter. My thoughts on necessity have been greatly influenced by Marshall (2005)—a critical response to Ripstein (2005). Ripstein’s rich and stimulating essay provides me with another embarrassing reminder of how out of touch I now am with Kant scholarship. His non-moralistic approach to criminal law doctrine is quite different from my moralistic approach. Unlike Ripstein’s, my approach—perhaps overly influenced by Kant’s sole reference to targeting “inner viciousness” (inneren Bösartigeit) in his doctrine of criminal liability—seeks to retain some elements of moral retributivism in Kant’s account of punishment.

174  Jeffrie G. Murphy procedure that will guarantee that he is no more likely than others to be the one selected for sacrifice. Thus the principle of selection must guarantee to each sailor that the person sacrificed does not, for reasons that are unreasonable or unfair, make him more likely to be selected than relevantly similar others. So I believe that rational sailors would select as the principle of selection the drawing of lots—the standard “custom of the sea.” This gives each sailor an equal chance of avoiding his own sacrifice. (An exception to the principle of drawing lots might be made in a case where the one to be sacrificed is, like Parker, reasonably believed to already be on the verge of certain death.) In short: If one is to be sacrificed that the rest might live, then drawing lots is generally the rational and fair principle of selection. This is a big “if,” however, and does nothing whatever to establish that it is ever proper to sacrifice anyone. If hypothetical consent will not work, how about actual consent? What I have in mind is a kind of actual tacit consent. Suppose (as was probably true for nineteenth-century sailors) that there is a socially evolved principle, known to all sailors as part of the established “custom of the sea,” that in extreme situations of the kind here discussed lots will be drawn to determine a person to be sacrificed so that others might live. If this is the case, it might be plausible to argue that any sailor, simply by agreeing to ship on, tacitly agrees to this custom. Even if there is tacit consent in such cases, however, this will not necessarily make any resulting murders justified (and thus, though homicides, not really murders). Why not? Because in criminal law, at least in the Anglo-American legal tradition, we do not generally accept victim consent as a defense to homicide, and several plausible rationales have been given for this: 1. The right to life is generally regarded as an inalienable right, and this simply means a right that cannot be waived or forfeited. 2. Because the right to life is inalienable, suicide and assisting suicide has long been regarded as criminal. If one does not have a legal right to kill oneself, then a fortiori one does not have a legal right to allow someone else to do the job. 3. Crimes are generally and properly regarded as offenses, not just against individuals, but against the entire community—something revealed in the fact that criminal cases, unlike cases in private or civil law, carry the name “State v X” or “Regina v X” and not “X v Y”. Thus victim consent cannot eliminate the state’s legitimate interest in prosecution and punishment—unless, of course, we are dealing with a crime such as rape where absence of consent is part of the very definition of the crime itself. 4. A context in which one might be called upon to consent to his own death is likely to be one in which the individual will feel great psychological pressure from others to consent. Thus we might have reasonable doubts that any consent given in such a context will be free enough to count as valid consent. The problem of Ulysses and the Sirens might also arise here. What should be counted as the real consent—consent given when the matter is being conceived in the

Kant on Three Defenses in the Law of Homicide  175 abstract with no strong emotional involvement or consent given when actually faced with the terrifying prospect of being killed? Of course, not everyone will accept the above four reasons as sufficient to undercut the kind of tacit consent that might be assumed when a sailor embraces the “custom of the sea.” Some will see the arguments—particularly 1 and 2—as depending on controversial religious or philosophical beliefs—e.g. Kant’s strange argument in the Groundwork against the moral permissibility of suicide or Peter Geach’s argument that it is not rational to accept any absolute moral principles unless they are grounded in a belief in God and the doctrine of divine providence.32 And others might point out that the practical worries raised in 4 might not apply in the context of signing on to go to sea as a sailor—a context that may not involve the kind of psychological pressure likely to be present in an actual lifeboat situation where others might pressure you into thinking that your going to your death in order that they might live is your duty. The Ulysses and the Sirens problem still remains, however. Even though these arguments against consent as a defense to homicide may not be conclusive, I think that they do have some power—at least enough to give us rational grounds to proceed with caution before using “the custom of the sea” as a basis for assuming a kind of valid tacit consent that keeps a so-called “right of necessity” from remaining highly problematic and dangerous—both morally and legally—if regarded as a justification. I think perhaps the best argument that one might give for a right of necessity in the Dudley and Stephens situation would be an argument that they are in a lawless state of nature—an argument that might have some plausibility in a case of starving men killing to save their lives in survival situations on the high seas. Kant, however, does not use the state of nature argument in necessity cases—only in the less persuasive cases of honor killings. Of course, this argument would at most address the legality (as a kind of jurisdictional matter) of what Dudley and Stephens did and not the morality of what they did. Even as a legal matter, however, the state of nature argument would face an uphill fight. The Law Lords did consider and reject an absence of jurisdiction claim— rejected because the Migonette was a registered British vessel and all crew members were British subjects. You may be inclined to dismiss the Law Lord’s argument as little more than a case of rigid legal formalism, but before doing this consider this case: Suppose that Dudley and Stephens were malicious sociopaths who killed Parker simply for the fun of it or suppose they decided to repeatedly rape him through acts of forced sodomy. Would you want to say that since they were in a state of nature that no law applied to them and thus they could not legitimately be punished by an English court?   Kant (1996: 73–4 and 80) and Geach (1969).

32

176  Jeffrie G. Murphy

Conclusion My purpose in this chapter has been to evaluate three claims made by Kant with respect to criminal defenses to homicide: the claim that we should not punish as a murderer the soldier who kills in a duel, the claim that we should not punish as a murderer the woman who kills her illegitimate baby, and the claim that there is no just and rational right of necessity in shipwreck survival situations (but that it would, for deterrence reasons, be pointless to punish such injustice as murder). I think that he is generally quite wrong with respect to the first two—at least to the degree that he is thinking in terms of justification—but that in his discussion of shame he may have partially laid the foundations for at least considering a mitigating excuse in some cases of this nature. On the third issue, necessity as a justification, I am now inclined to think that he (and the Law Lords in Dudley and Stephens) may have reached the right conclusion—namely, that the law should not recognize a justifying right of necessity in cases that involve the murder of a person innocent of any wrongdoing. In reaching this conclusion, however, the Law Lords and Kant part company. Kant relies on a very unpersuasive deterrence argument (an argument of a kind never even considered by the Law Lords) and he, unlike the Law Lords, does not recommend sovereign mercy on compassionate grounds for those in situations of the kind faced by Dudley and Stephens—a recommendation that strikes me as showing a proper understanding of the pressures under which Dudley and Stephens made their decision without setting a precedent and thus changing the law so that an improper justification would be legally recognized. If Kant and the Law Lords had been able to think more clearly about the justification/excuse distinction—a distinction Kant sometimes seems to flirt with in his discussion of the two honor/shame cases—they might even have been able to build an appropriate excuse into the prevailing doctrines of criminal law. Absent that, sovereign mercy strikes me as better than nothing. I do not, however, want to end on such a negative note about the great Kant—whose moral vision has been an inspiration to me throughout much of my philosophical career.33 When I find myself tempted to reject a right of necessity in cases such as Dudley and Stephens, I find that my temptation has a Kantian foundation—not a foundation in what he actually says about necessity cases but a foundation in the very essence of his moral theory. As I see this theory it makes respect for the inherent dignity of each person foundational—a value commitment that forbids that any person be used merely as a means rather than as an end in himself, as a mere resource to be exploited for the benefit of others. Killing an innocent person that others might live strikes me as a possible example of the very thing that Kant’s moral theory requires us to reject (far more powerful than a bogus deterrence argument) or a least to make us engage in a great deal more serious moral reflection before we endorse the recognition of a justifying 33   Often as that vision has been articulated by Tom Hill. My work on forgiveness, for example, was greatly influenced by his two essays on servility and self-respect (Hill 1991: 4–24).

Kant on Three Defenses in the Law of Homicide  177 right of necessity for homicide. We will thereby take seriously, as I think we should, the question that Sonja puts to Raskolnikov in Dostoevsky’s Crime and Punishment: “Who put me here to judge who is to live and who is not to live?”

References Binder, Guyora. 2009. “States of War: Defensive Force among Nations.” Ohio State Journal of Criminal Law, 7: 439–61. Devlin, Lord Patrick. 1965. The Enforcement of Morals. Oxford: Oxford University Press. Finkelstein, Claire. 2001. “Two Men on a Plank.” Legal Theory, 7: 279–306. Geach, Peter. 1969. “The Moral Law and the Law of God,” in God and the Soul. London: Routledge and Kegan Paul. Hanson, Neil. 2000. The Custom of the Sea. New York: Wiley. Hill, Thomas E., Jr. 1991. Autonomy and Self-Respect. Cambridge: Cambridge University Press. Hill, Thomas E., Jr. 2000. Respect, Pluralism, and Justice—Kantian Perspectives. Oxford: Oxford University Press. Kahan, Dan M. 2000. “Gentle Nudges vs. Hard Shoves: Solving the Sticky Norms Problem.” University of Chicago Law Review, 67: 607–45. Kant, Immanuel. 1996. Practical Philosophy, trans. and ed. Mary J. Gregor. Cambridge: Cambridge University Press. Marshall, Sandra. 2005. “Life and Death on a Plank—Ripstein and Kant.” Ohio State Journal of Criminal Law, 2: 435. Murphy, Jeffrie. 1986. “The Justice of Economics.” Philosophical Topics, 14(2): 195–210. Murphy, Jeffrie. 1987. “Does Kant Have a Theory of Punishment?” Columbia Law Review, 87(3): 509–32. Murphy, Jeffrie. 1998. “Kant on Theory and Practice.” Character, Liberty and Law—Kantian Essays in Theory and Practice. Dordrecht: Kluwer. Murphy, Jeffrie. 2006. “Legal Moralism and Retribution Revisited.” Proceedings and Addresses of the American Philosophical Association, 80(2): 45–62. Also published in Criminal Law and Philosophy, 1(1) (2007): 5–20. Murphy, Jeffrie. 2007. “Remorse, Apology and Mercy.” Ohio State Journal of Criminal Law, 4: 423–53. Posner, Richard. 1985. “An Economic Theory of Criminal Law.” Columbia Law Review, 85: 1229–30. Rawls, John. 1971. A Theory of Justice. Cambridge, MA: Harvard University Press. Rawls, John. 1993. Political Liberalism. New York: Columbia University Press. Ripstein, Arthur. 2005. “In Extremis.” Ohio State Journal of Criminal Law, 2: 415–34. Sussman, David. 2008. “Shame and Punishment in Kant’s Doctrine of Right.” Philosophical Quarterly, 58(231): 299–317. Simpson, Brian A. W. 1984. Cannibalism and the Common Law. Chicago: University of Chicago Press. Uniacke, Suzanne. 1996. “The Limits of Criminality: Kant on the Plank,” in Henry Tam, ed., Punishment, Excuses, and Moral Development. Aldershot: Avebury, pp. 113–26.

9 Virtue, Repugnance, and Deontology Matt Zwolinski and David Schmidtz

Introduction In classrooms and textbooks, even in scholarly work, philosophers sometimes speak and write as though the difference between deontology and virtue ethics were simple and dramatic.1 Deontology is about hard-and-fast rules that we are supposed to be able to simply follow. Virtue ethics offers rough-and-ready principles that guide rather than replace wisdom and common sense. Deontology offers us a clear cut decision procedure to ensure the morality of our behavior; virtue ethics can only counsel us to develop and exercise our practical reason. And, of course, deontology focuses on evaluating acts, while virtue ethics focuses on evaluating character.2 There are kernels of truth in these claims: useful simplifications that help to convey differences in emphasis and focus. A deontological system can, however, embrace questions, values, and insights more commonly associated with virtue ethics. Specifically, it can treat the assessment of an agent’s character as central to the project of moral evaluation. Thomas Hill has stressed the importance of character in thinking about certain problems of environmental ethics. In this chapter, we embrace and extend Hill’s insight. We proceed via an argument that Derek Parfit’s famous Repugnant Conclusion is a problem, not just for utilitarians, but for any act-centered theory. The solution, we 1   This chapter incorporates material from Zwolinski and Schmidtz (2005), and also from Schmidtz (2001). Schmidtz’s work on this chapter was supported by a grant from the John Templeton Foundation. The opinions expressed here are his and do not necessarily reflect the views of the John Templeton Foundation. 2   Ryan Hanley, a scholar very much alive to the subtleties of the various theories, nevertheless sets up his discussion with the standard line. “In this century, virtue ethics has emerged principally as a late-stage rival to both utilitarian or consequentialist approaches on the one hand and deontological or neo-Kantian approaches on the other. In contrast to these systems, which evaluate actions on the grounds of their capacity to maximize good effects (as consequentialism does) or on the grounds of their adherence to universally valid and rationally derived moral rules (as deontology does), virtue ethicists focus on evaluating characters and the specific virtues and vices that contribute to the composition of good and bad characters” (Hanley 2009: 53).

Virtue, Repugnance, and Deontology  179 suggest, is a vision of deontology unlike what we sometimes teach our undergraduates, but a vision that we believe Thomas Hill would endorse.

Maxims and Character In a typical introductory version of Kantian deontology, a relatively straightforward procedure determines the permissibility or impermissibility of an action. First, we identify the maxim, or “subjective principle of volition,” of the action. In our textbooks (though not in Kant, as such), maxims have something like the form “In circumstances C, I will perform action A in order to achieve purpose P. ” Second, we run the maxim through one of the formulations of the Categorical Imperative. In the case of the Formula of Universal Law, this involves formulating a universalized version of our maxim (“In circumstances C, everyone will perform action A in order to achieve purpose P”), then seeing if we can consistently will that universalized version. We can fail this test in two ways—either by being unable to conceive of a universalized version of our maxim, or by being unable to will it. If we fail in either way, then we cannot act on this maxim without violating our moral duty. Even students reading Kant for the first time are quick to see that the process of detecting contradictions in universalized maxims is more complicated than it first appears. Simply specifying the maxim turns out to be anything but simple. It is not at all clear that there is a uniquely correct way to specify the maxim of an action. The same act, for instance, can be described as “Bringing Mark Peters a cup of decaf coffee,” “Bringing my guest something to drink,” “Making my guest feel comfortable,” and so on. Such open-endedness in many cases seems to leave the permissibility/impermissibility of an action depending more than it should on how exactly its maxim is described. For example, consider Kant’s case of the false promise. Kant famously argued that a maxim of acquiring a loan by falsely promising to repay it is not universalizable, because a world in which all persons made such false promises would be a world in which promises as such ceased to exist. No one would believe, or make, promises if all promises were empty. But to borrow an example from Allen Wood, suppose that we describe the maxim not as merely “making a false promise” but as “making a false promise on Tuesday, August 21, to a person named Hildreth Milton Flitcraft” (Wood 2008: 102)? Surely, an agent can both conceive and will a universalized version of that maxim without contradiction. In this case, then, the added specificity appears to generate a “false positive”—a mistaken identification by the universalization test of an impermissible maxim as permissible. Too much detail can sometimes yield false negatives too—cases where permissible maxims are mistakenly identified as impermissible. Few things could seem more innocent, for instance, than playing tennis on a Sunday morning. But suppose that the agent’s maxim is to play tennis on Sunday in order to avoid crowded courts? In this

180  Matt Zwolinski and David Schmidtz case, the universalized version of the maxim is self-defeating and fails the test of the Formula of Universal Law. Nor is the problem always one of too much specificity. Sometimes, the problem is what is left out of the agent’s maxim. “I will do my job well in order to provide for my family” is a fine maxim for a construction worker, but not for a concentration camp guard.3 We bring up the problem of maxim specification not because it poses a fatal difficulty for Kant, or for Kantians. It does pose a fatal difficulty for a certain naïve view about how the Kantian universalization test should function in moral reasoning. But we believe that Kantians should (and most do) reject this view anyway. The Formula of Universal Law can play a useful role in moral thought, but only to the extent that it is supplemented by a careful examination of the issues of character that are more commonly associated with the virtue ethics approach. Onora O’Neill’s article, “Consistency in Action,” provides an illustrative example. For O’Neill, problems arise with universalization tests when we understand maxims as “surface level intentions.” Such an understanding not only runs into the problems of false-positives and false-negatives described above; it runs counter to various Kantian claims about the opaqueness of maxims—the way in which it sometimes (perhaps always) is hard to truly know which maxims we are acting on at any given time. The problem, as O’Neill sees it, is that while our intentions can be focused on incidental and unimportant features of our situation—which particular cup to grab in order to fill our visitor’s coffee, which path we walk to bring it to him—the point of focusing on maxims is to focus on the principles of our action. Maxims, then, are “those underlying principles or intentions by which we guide and control our more specific intentions” (O’Neill 1989: 84). Your choice of a cup for your visitor’s coffee is inessential to the point of your action. Your underlying aim is to make your visitor feel comfortable, and to do so by bringing him coffee requires that you choose some cup. But that choice is controlled and guided by your underlying principle, or maxim. Underlying principles can be psychologically opaque. We donate some money to help a stranger. Did we do it with a purely good will, from the motive of beneficence? Did we do it to convince others of our kindness? Or to convince our self? We cannot always know through simple introspection.4 In many cases, the best we can do is to imagine carefully constructed counterfactual situations, and to think about 3   For more on the difficulties of maxim specification, and an overview of various attempts to cope with these difficulties, see Rob Gressis (2010a and 2010b). 4   Kant recognizes the difficulty, perhaps even the impossibility, of empirically determining the maxims on which one’s own past actions (let alone the actions of others) were based (Groundwork 4: 407–8; see Kant 1996: 61–2). The point of thinking about maxims is thus not primarily to help us assess the morality of actions that have already been performed in the past. It is, rather, to help us shape our will for actions that we are thinking about performing in the future. But even this task, we think, requires us to go through the difficult psychological task of uncovering the various maxims on which we might choose to act, and thus (as we shall argue below) to engage with questions that are fundamentally about our moral character. We thank an anonymous referee for this intriguing suggestion.

Virtue, Repugnance, and Deontology  181 how our actions might have differed in them. Would I still have given the money if no one was watching? If it somehow made me feel worse, rather than better about myself? Answers to such questions are impossible to come by with any certainty. But notice that these questions are almost exactly the same ones we ask when we assess an agent’s moral character. In asking how an agent would likely act in a variety of counterfactual situations, we are attempting to understand his or her basic dispositions. But one’s moral character, of course, can itself be thought of as a disposition of the most general sort. Like maxims on O’Neill’s view, one’s character controls and guides the specific intentions one forms. And like maxims, one’s true character can often be opaque even to one’s own self. But if this is right, then none of the naive distinctions we initially made between deontology and virtue ethics will hold up. Deontology—as most sophisticated deontologists admit—has just as difficult a time specifying hard-and-fast rules (as opposed to rough-and-ready principles) for the specification of right and wrong conduct.5 Whatever specification deontology offers will involve the use of practical reason and judgment no less than does virtue ethics. And, as we have seen, deontology’s focus on the moral evaluation of acts is not only consistent with a focus on moral character; it arguably demands it. In all these ways (and, we suspect, more), the traditionally sharp distinction between virtue ethics and deontology is overstated.

Hill on Environmental Ethics and the Inadequacy of Act-Centered Theories Reflection on the problem of maxim specification provides one reason for thinking that considerations of character must play an important role in moral thinking. As Thomas Hill showed in a now-classic article, reflections on certain environmental issues can lead to a similar conclusion (Hill 2005). When Hill’s neighbor cut down a beautiful old avocado tree and covered his yard with asphalt, Hill was indignant, but paused to wonder whether he had any theoretical justification for his indignation. The problem was not that trees have rights, or that Hill’s neighbor had no right to cut down the tree. Hill’s neighbor was depriving others of the enjoyment of the tree, but although this was consistent with Hill’s indignation, it did not really account for it. Ultimately, Hill concluded, the core question was not what was wrong with the act, but what was wrong with the person. “What sort of person,” Hill asked, “would do a thing like that?” Hill’s answer was that an admirable person would not 5   See, for instance, O’Neill’s discussion in “Consistency in Action,” where she argues that the concept of duty is fundamentally about moral worth and only derivatively about the permissibility of external actions, and that indeed drawing any firm conclusions about the latter on Kantian grounds will be tremendously difficult.

182  Matt Zwolinski and David Schmidtz do it. A person would have to be insensitive, and lacking in humility. Interestingly, the humility about which Hill was talking was recognizably an ecological humility. It is what Aldo Leopold (1966) meant when he spoke of our need to acknowledge and ultimately cherish our proper status as citizens (not conquerors) of the biotic community. Like Tom Hill, many of us who are disturbed by the callous treatment of our natural environment would feel uncomfortable arguing that trees, brooks, or sand dollars have a right to be left alone. Also like Hill, though, neither is our discomfort grounded in a calculation of aggregate utility. The problem is that both deontological and consequentialist theories locate the source of our discomfort in the wrong sorts of considerations. As a result, they can give us false guidance. One particularly notorious example of this is Derek Parfit’s Repugnant Conclusion. Parfit (1984) conceived his Repugnant Conclusion as a problem for one version of a utilitarian value theory, but in the following sections, we generalize Parfit’s argument. We suggest, first, that all utilitarian theories of value face analogous problems. Second, we cannot solve the problem (although we might be able to obscure it) by adopting some other act-centered theory. Ultimately, Repugnant Conclusions suggest (we do not claim they entail) problems for the whole idea that moral theorizing should culminate in a simple formula for right action. We need a different sort of theory. We need not merely a better formula, but a better objective, such as is hinted at in Tom Hill’s environmental virtue ethics.

The Repugnant Conclusion Standard versions of the principle of utility say something like this: An act is right if and only if it maximizes happiness. What does it mean to maximize happiness? First, the principle is referring to an aggregate: that is, the sum of everyone’s happiness.6 Second, the principle on its face is quantitative, referring to the kind of thing that can be maximized. Third, the quantity is most naturally thought of as a total sum, as opposed to an average. As Parfit notes, however, there is a problem with this seemingly innocuous third point. Suppose we are deciding whether to have one or two children, and have no reason to doubt that the two children would each be about as happy as the one. If the two children would each be as happy as the one, then we conclude there is more total utility, indeed about twice as much, in having two children than in having one.

6   It must seem obvious we will have trouble contriving a number that reliably represents aggregate happiness, but utilitarians have little choice but to insist it can be done. Or, at least, to insist that the ideal of maximization can serve as a useful guide to moral deliberation even in the absence of reliable numbers. We accept the assumption for argument’s sake. But see Schmidtz (1992).

Virtue, Repugnance, and Deontology  183

A

B

C

Figure 9.1 

On its face, we seem to have utilitarian grounds for having two children rather than one. However, as Parfit points out, when we endorse the principle that the right act maximizes total happiness, we commit ourselves to the Repugnant Conclusion that For any population of at least ten billion people, all with a very high quality of life, there must be some much larger imaginable population whose existence, if other things are equal, would be better, even though its members have lives that are barely worth living. (Parfit 1984: 388)

Here is how Parfit reaches that conclusion. Figure 9.1 illustrates three possible populations. The width of the blocks represents the number of people living; the height represents their average quality of life. In population A, people’s lives are, on average, well worth living. Population B differs from population A only in that, in addition to the population of A, it contains an added group. These people’s lives are worth living, too, though less so than the lives of the persons in population A. Let us stipulate that their existence does not affect the persons in population A at all—members of population A do not even know of this new group’s existence. The question, then, is this: do we make a situation worse by moving from A to B—that is, by the mere addition of persons whose lives are worth living? It’s hard to imagine condemning such an addition. By hypothesis, no one’s rights are violated, total utility is higher, and each new person is happy to be there.7 At the least, B seems no worse than A. How does B compare with C? In C, the better-off group from B has been made worse off. But their loss is smaller than the gain achieved by the worse-off group, so total utility has risen. The resulting distribution is superior on egalitarian grounds as well. On various grounds, then, C is at least as good as B.8 The problem, of course, arises in the repetition of this sequence, as illustrated in Figure 9.2. For if the move from A to C is justified, why not the move from C to D, and 7   The average utility of the entire population has, of course, decreased. But the average utility of the original group remains unchanged, and there are reasons to think this should be the relevant consideration. If we judged the morality of an action based on the average utility of the population which exists ex post, rather than ex ante, then average utility would seem to condone (secretly) killing off anybody who was less happy than average. Surely this is not the way to make the world a happier place. 8   This is not to deny that there are impermissible ways of moving from B to C, or even that a state of affairs which would otherwise be desirable can be rendered condemnable by being brought about in an impermissible way. So long as we assume this is not the case, C looks preferable to B.

184  Matt Zwolinski and David Schmidtz

A

C

D

Z

Figure 9.2 

so on, all the way to Z? In other words, if a state of affairs is made better (or at least, not worse) than another by doubling the population while decreasing average utility by less than 50 percent, why not continue until we are left with an enormous number of people whose lives are, on average, only barely worth living?9 This is the repugnant conclusion. To people seeing this puzzle for the first time, the conclusion that Z is morally desirable, or even that it is merely no worse than A, sounds absurd. How could a society be superior to another simply in virtue of being so much larger, when average members are so much worse off that one more traffic jam, stubbed toe, or malfunctioning toilet will drive them to conclude that life is no longer worth living?

Generalizing: A Problem for Utilitarianism Here is an easy response. Total utilitarianism works well enough in a world of fixed population, but in a world where population size is one of our choice variables, we need a more sophisticated metric. In this more complex world, the intuitive attraction of utilitarianism is better captured by average utilitarianism: the theory that an act is right if and only if it maximizes average utility. As Parfit was aware, though, average utilitarianism has its own version of the problem. The Other Repugnant Conclusion: For any population of people with a quality of life more or less like ours, say, we can imagine a smaller population (in the limit, a single, god-like Utility 9   What does it mean to say that a life is only barely worth living? On its face, this description is compatible with a life’s being pretty miserable. This, at any rate, is the interpretation we operate with for the purposes of this chapter. On reflection, however, the judgment that a life is worth living might be one to which it is difficult to assign any concrete interpretation in the absence of a particular perspective. A person who is already living that life might be more disposed to sincerely believe that their life is worth living than a person who is observing the world from the outside, deciding whether or not it would be worth it to live that life (rather than remain in their current state of oblivion). Perhaps there are psychological forces which impel the person living their life to convince themselves that it is worthwhile. Or perhaps the fact that they are situated in a particular point in time is relevant (costs in the past are sunk, and there’s always hope for the future). Are these reasons to discount the preference expressed by actual individuals? What’s the alternative?

Virtue, Repugnance, and Deontology  185 Monster) that is on average so much happier that it would be better if our population were replaced by that smaller one.

Some environmentalists will not find this repugnant, but this is because they have other reasons to oppose overpopulation. To them, there are things in the world that matter more, maybe a lot more, than the happiness of persons. In other words, they are not utilitarians. To utilitarians, though, the Other Repugnant Conclusion is almost as big a problem as the original. If average utility can be increased by (secretly) killing off people who are somewhat less-than-averagely happy, then the maximization of average utility appears not to be a very attractive ideal. To summarize, the problem in its general form is that we have two kinds of mechanical measures of aggregate happiness: total or average. In practice it may be massively difficult, indeed impossible, truly to arrive at any such measure. But the problem suggested by our repugnant conclusions goes deeper: that is, we would not be able to trust a number even if it were easy to obtain, and even if its accuracy were indubitable. The fact would remain that neither version of a utilitarian number—that is, neither total nor average happiness—reliably tracks the intuitions that led us to find utilitarianism plausible in the first place.10

Generalizing Further: A Problem for All Act-Centered Theories The problem is more than a problem for utilitarianism. It is a problem for all act-centered theories: that is, moral theories that treat specifying action-guiding rules as their primary task. From an act-centered perspective, it is hard to explain why the repugnant conclusion is repugnant. We have already seen that utilitarian theories, which base their evaluation of actions on their tendency to produce desirable states of affairs, are left with little means of stopping once they accept the desirability of the initial moves from A to B, and B to C. Deontological theories might appear to be in better shape. After all, they do not base their evaluation of an act’s morality on its tendency to produce a certain state of affairs, so they can reject the move from A to B even if they judge B to be a better state of affairs. Admittedly, it is awkward to bite the bullet and say (as the hardest forms of deontology do) that consequences should play no role in our evaluations of states of affairs. But that consequences should play any role is all it takes to get repugnant conclusions off the ground. For if C is superior to A on some morally relevant grounds,

  Michael Huemer (2008) has likewise argued that the Repugnant Conclusion is inescapable within any utilitarian framework, although he treats this as a reason for accepting the Repugnant Conclusion rather than a reason for leaving utilitarianism behind. 10

186  Matt Zwolinski and David Schmidtz then the burden of proof is on the deontologist to show that there are countervailing moral considerations that override the moral case for moving from A to C. And this seems unlikely. Who is being treated as a mere means in the course of moving from A to C? Where in the chain is the step that cannot be universalized? For that matter, what if anything is wrong with Z from a deontological perspective? Whose rights are being violated? More profoundly, whose rights are violated by the bare admission that Z is a better state of affairs? A theory might try to save face by insisting that moving from A to Z is for some reason impermissible, but quibbling about the move’s propriety would be too little too late. If a theory admits that Z is better, it has already embarrassed itself, regardless of whether it prohibits moving to Z.11 Deontological theories are more apt for considering how to treat currently existing people than for considering whether it would be good for an additional population to come into existence. Which is to say, deontological theories are less apt for a world where moral problems increasingly are taking on ecological dimensions. Somehow, the initial utilitarian rationale for the move from A to C is spurious. In some way, a total utilitarian has the wrong idea, not about whether consequences are relevant, but about the way in which consequences are relevant.12 An average utilitarian at best does only a little better. We may hope for some escape from the problem once we factor in environmental considerations, but it is built into the problem that life in the larger population is, after all, worth living. So whatever loss of environmental amenities people face, average people nevertheless are eking out lives worth living, even if only barely. An anthropocentric deontology likewise would seem not to solve the problem, as there is no particular reason to suppose members of this larger population are failing to treat each other as ends in themselves. A deontology expanded to embrace animal rights merely treats one symptom of a larger problem, because the problem goes beyond our treatment of other sentient creatures to the wanton destruction of trees and such. The larger concern is not animal rights; it is more fundamentally an ecological concern, perhaps with aesthetic overtones.13 11   In the case of what we call the Other Repugnant Conclusion, the deontological prohibition of the crimes that would be involved in making most of humanity disappear is straightforward. Here too, however, we see a problem with using the language of universalizability or respect for persons to explain what is wrong with the kind of environmentalist who thinks the world would be better if virtually all of humanity were to disappear. 12   One other possibility would be to handle consequentialist issues in contractarian fashion, that is, by saying that the utilities in question are not fungible. It is true that for any given person, more utility is better for that person, but it is not true that anyone’s utility can be traded for anyone else’s in such a way as to produce a higher aggregate. Such a theory would reject what otherwise seems to be a utilitarian case for moving from A to Z, but at a cost of refusing to entertain comparisons at all, and thus also refusing to acknowledge that A is preferable to Z in some agent-neutral way. For discussion of non-aggregative forms of utilitarianism, see Coons (2012). 13   For related arguments that recast the question as a question about the right thing to do rather than reasons for preferring, see Narveson (1967). See also Booin-Vail (1996).

Virtue, Repugnance, and Deontology  187

Is There Another Way of Doing Moral Theory? There would seem to be a more direct way of going to the heart of the problem, but it would involve giving up on standard act-centered moral theories. The heart of the task is at least in part a task of defining ideals of human excellence—defining a conception of the good life for a person that makes sense as an ideal, with no presumption that the ideal must be adopted as a goal to be promoted. We have other ways of responding to ideals; simply respecting them comes to mind, an attitude perhaps more in keeping with intuitions that motivate deontology. Part of an ideal of human excellence could, in turn, incorporate forms of environmental sensitivity. Again, though, there is no assumption that environmental sensitivity must translate into an activist agenda—simply appreciating the beauty of nature is among the more admirable ways of being sensitive to it.14 Thomas Hill suggests that our discomfort with environmental exploitation is not wholly a product of our belief that environmental goods are being put to inefficient use, nor that those who exploit such goods are violating any rights in doing so. It is a mistake, he thinks, to suppose that “all moral questions are exclusively concerned with whether acts are right or wrong, and that this, in turn, is determined entirely by how the acts impinge on the rights and interests of those directly affected” (Hill 2005: 48). Instead, Hill suggests, we ought to ask “What sort of person would destroy the natural environment?” Approaching the issue from this perspective allows us to see that “even if there is no convincing way to show that the destructive acts are wrong (independently of human and animal use and enjoyment), we may find that the willingness to indulge in them reflects the absence of human traits that we admire and regard as morally important” (Hill 2005: 50). People who carve their initials in hundred-year-old Saguaros might not be violating any rights, and the satisfaction they get might well outweigh the suffering caused to other sentient beings, but the fact remains that there is some defect in such people’s characters. In Hill’s language, they lack a kind of humility—an ability to appreciate their place in the natural order or to “see things as important apart from themselves and the limited groups they associate with” (Hill 2005: 51). Hill’s approach offers a natural, straight-to-the-point way of thinking about what Richard Routley called the “Last Man Argument.” Routley’s thought experiment presents you with a situation something like this: You are the last human being. You shall soon die. When you are gone, the only life remaining will be plants, microbes, invertebrates. For some reason, the following thought runs through your head: “Before I die, it sure would be nice to destroy the last remaining Redwood. Just for fun.” What, if anything, would be wrong with destroying that Redwood? Destroying it won’t hurt anyone, so what’s the problem? The problem is, what kind of person would destroy that 14   Mathews (2001) argues that militant environmentalists make the same mistake as everyone else who seeks to conquer the biotic community: failing to embrace a truly ecological ethos of “letting it be.”

188  Matt Zwolinski and David Schmidtz last Redwood? What kind of person would enjoy such wanton destruction of such a beautiful, majestic, living thing? Hill’s question seems like exactly the right question.15 Indeed, for a number of philosophers Hill’s question promises to transform the way we think about environmental ethics. The relatively recent field of environmental virtue ethics has developed in order to study the norms of character that ought to govern human interaction with the environment, and scholars such as Philip Cafaro, Jason Kawall, and Ronald Sandler are developing this field in a way that is yielding fruitful theoretical and practical results (Cafaro 2004; Kawall 2003; Sandler 2007). These scholars seek to characterize attitudes and dispositions that are constitutive of environmental virtue, and to explain the proper role of an ethic of character within a broader environmental ethic. And, practically, scholars have focused both on the implications of and prerequisites of particular virtues such as benevolence or temperance, and also on the implications of a virtue ethics approach for specific practical problems such as consumerism and genetically modified crops. Hill’s approach also offers a way of dealing with repugnant conclusions. Even if we cannot provide a definitive account of the wrongness of preferring society Z, as described in the repugnant conclusion, there remains something wrong with being the kind of person who would prefer Z. That something could be hard to articulate, but no less real for that. In any case, Hill does provide pertinent articulation. The sort of person who would prefer Z is the sort who does not possess the humility that would lead a more virtuous person to see value in human society playing an appropriately limited role in the biotic community, for non-anthropocentric as well as anthropocentric reasons. It is difficult to make accurate judgments of character without setting the context in some detail. And it is difficult to imagine a context in which a person’s expressed preference for Z is worthy of being taken seriously enough to merit a moral evaluation. In the next section, we will attempt to describe a situation in which it makes sense to speak of a person “choosing” to move from A to Z. For now, though, suppose that someone you know were offered the option of snapping their fingers and thereby popping into existence a population either like A or like Z in some far-off and (otherwise) causally isolated universe. Presumably, since no other values are at stake, a disposition to choose population Z in such circumstances is simply part of what it means to believe that Z is a superior state of affairs to A. Given this fact, then, what are we to say about the character of a person who prefers Z to A? Bear in mind that Z is, in the end, a fairly miserable place. People’s lives are, it is true, still worth living. But only barely so. Think of how much misery a person can endure while still believing that life is worth living. Now think of a whole world—a very crowded world—filled with such people. What kind of person would bring that kind of universe into existence when they could just as easily have produced a universe 15   This paragraph borrows from the editors’ introduction, “The Last Man and the Search for Objective Value,” in Schmidtz and Willott (2012).

Virtue, Repugnance, and Deontology  189 with a smaller number of very happy people? What would be the point? Intuitively, there is no point, contra total utilitarianism. The most natural explanation for such a disposition seems to be a sort of obsessive focus on happiness as the sole consideration of moral relevance. It’s normal, of course, to think that the happiness of particular other people is important. It’s normal to generalize from this and think that happiness itself is important. It’s maniacal to think that this abstraction translates into a reason to prefer, over A, the concrete misery of world Z. Like cases of obsession in general, what seems to have gone wrong here is an extreme inability to grasp the larger context. In this respect, it is not unlike a person who originally pursues cleanliness for the sake of health (and health for the sake of a long, enjoyable life), but ends up cleaning compulsively. What started off as a reasonable principle—be as clean as you can—has been transformed into a manifestly unreasonable rule that trumps all other aspects of a worthwhile life. The compulsive pursuit of cleanliness thus ultimately undermines the very value that led them to pursue cleanliness to begin with. There is something similarly wrong in being a person who would think of maximizing a happiness number, as if, from an anthropocentric perspective, what matters most were happiness rather than persons themselves (or something about persons other than their happiness, such as whether they achieve excellence).

Note on Contributing to Overpopulation In the real world, of course, populations are never the product of any individual’s choice. Individuals and families do not choose populations; they choose whether to have children. Populations emerge only as a (typically unforeseen and unintended) consequence of the combination of many such decisions. The fact that we can raise questions about the character of one who holds a sincere preference for population Z might thus seem to have few implications for deciding what to say about people and policies which actually move us toward Z. Do character-centered ethics have anything to say about these more practical population-related questions? Insofar as they are not the product of any single individual’s choice, undesirable populations can be regarded as an externality along the lines of air pollution. No individual family’s decision to limit the number of children they have significantly affects the size or well-being of the overall population, just as no commuter significantly affects the amount of smog in the air by choosing to ride their bicycle to work one day. In either case, however, the aggregate result of many individuals failing to make such decisions is tragic. Suppose for argument’s sake that adding large numbers to the existing population will contribute to an unsustainable (or undesirable) population growth. In that case, raising a large family would amount to a failure to contribute to a sort of public good.16 16   To be clear, we do not accept the premise, but it nicely illustrates how indignation over minute harms can often be better explained in terms of universalizability than in terms of utility, and even better, in terms of what such behavior says about the characters of people who free ride at other people’s expense.

190  Matt Zwolinski and David Schmidtz If that were the case, achieving a desirable population would require a general policy of restraint on the part of most families, and having a large family in that situation would be to free ride on the restraint of others. The irresponsibility here needn’t be calculated—many families simply won’t know (and won’t bother to find out) the consequences of having more children than they can afford. The point, however, is that where act-centered theories fall short of explaining the indignation we sometimes feel (or explain the indignation in an ad hoc, roundabout, or otherwise unsatisfying way) when confronted with free riding that does only minute harm, a character-centered theory provides a rich vocabulary for criticizing those who contribute to repugnant conclusions as short-sighted, irresponsible, weak-willed, selfish, and so on. Does virtue ethics help to provide a specifically environmental perspective on what’s wrong with people who contribute to overpopulation? (1)  A  utilitarian can say an act has bad environmental consequences, which counts for something. (2)  A  deontologist can say an act that has sufficiently bad environmental consequences will as a result be non-universalizable, and may also fail to treat other would-be users of the same environmental amenities as ends in themselves. (3) A virtue theorist can say an actor is a bad person qua member of the biotic community, which is something else. A virtue theorist can acknowledge that the third conclusion is true partly because the first one is, but can go on to say there is more to being a good person than to act in a way that has good consequences. A good person is considerate, and therefore cares about consequences. A good person is humble, in the sense of seeing himself as a locus of value in a world where there are many loci of value, and seeing also that it is not only humans who can be worthy of appreciation.

Intuition and Theory One false ideal for moral theory is the idea that the right theory will be simple in the sense of being able to substitute for the wisdom of experience. Part of the point of the Repugnant Conclusion is that wise persons realize that the intuitions leading them to find utilitarianism plausible are not in fact captured by the simple formulas utilitarians sometimes claim to offer.17 It is not as if theory is a radical alternative to intuition. A theory is an attempt to capture our intuitions with a simple formula. How could we expect to do that without losing some of morality’s intuitive nuance? Of course theories will have counter-examples! It is in our veins as philosophers to continue to test our theories against the intuitions 17   One of the many features that made Mill a great moral philosopher was his refusal to place more weight on the simple formula of utilitarianism than it was meant to bear. Readers of Utilitarianism will find nothing like an algorithm at work in Mill’s thoughtful and nuanced moral analysis.

Virtue, Repugnance, and Deontology  191 we intend to be articulating with our theories, and of course those articulations will be an imperfect match. Theories try to systematize our intuitions, but that is like trying to launch a ballistic missile in a direction such that its simple trajectory will track the more complex trajectory of a guided missile. Counter-examples take the form of showing where the ballistic missile deviates from the guided missile’s more convoluted path. This is not to express skepticism about the whole project of moral theorizing, so much as about the more particular assumption that act-centered moral theorizing is the way to go. Act-centered theories are one way of trying to articulate. There is no reason to assume they are the best way. Nothing like that is guaranteed. What is more or less guaranteed—we see no counter-examples on the horizon!—is that act-centered theories will provide imperfect guidance. We are not presenting this as a knock-down argument against act-centered theory. Our conclusion is that act-centered theory has a certain kind of value, not that it has no value. Virtue ethics reminds us that providing moral agents with a decision procedure covering all possible situations is not the main purpose of moral theories (if it is even a purpose at all). The people for whom moral theories are intended are people already in the midst of living their lives. They come to philosophy hoping it can help them reflect on their lives. A moral theory is successful if it provides that assistance; unsuccessful if it does not. Cases like the repugnant conclusion show us that an act-centered theory is not useful as a universal decision procedure. The proper lesson is not that act-centered theories are useless, though, but rather that we are better off treating act-centered theory as the sort of thing from which wise persons can gain insight that is useful, even if limited.

Summary There are times, as Tom Hill says, when the question is not what is wrong with the act, but what kind of person would do it. The Repugnant Conclusion seems to show that there are cases where the moral problem, even from an act-utilitarian perspective, is not straightforwardly a problem of how to maximize total utility. Our “Other Repugnant Conclusion” seems to show that average utilitarianism does not solve the problem; therefore, even by its own lights, act-utilitarianism, the simplest, most mechanical of all moral decision procedures, is not reliable as a mechanical procedure even in principle. Moral decisions require wisdom, not mere computational power, and there is no simple recipe for wisdom. It is mere appearance, and misleading appearance, that act-centered theories are better than agent-centered ones at converting moral decision making from art into science. Prevailing act-centered theories incorporate theories of value that specify some of the considerations to which a wise moral agent will be sensitive. That is their

192  Matt Zwolinski and David Schmidtz contribution to moral wisdom. It is a significant contribution, but they have not done better than that, and probably never will. Human rights matter, as do animal rights, whenever they are at stake. Interests matter, when they are at stake. Treating persons as ends in themselves matters, when persons and their ends are at stake. Perhaps universalizability matters in some independent way, but if it matters in some independent way, it probably matters in virtue of what it says about an agent’s character. The idea is that to act in a way that you could will to be universal law arguably is the essence of acting with integrity. That is, when we do that, we are acting from motives that we would not hesitate to make transparent, for all the world to see. If Kant was right, acting in accordance with what one could will to be universal law was the essence of good will, which (although good will is a notoriously technical notion in Kantian scholarship) appears to be a state of character. A virtue theory might not agree with Kant that good will is the only thing good in itself, but might readily agree that good will is basic, and that unless one gets one’s character in order, the other good things in life become ashes. Finally, talk of rights, interests, and treating persons as ends seems especially apt when we are talking about how to treat persons, or perhaps other sentient beings, and that is what we were talking about in the previous paragraph. But what if the issue concerns a person’s relation to an insentient creature such as a Redwood, or to the biotic community as such? Intuitively, Tom Hill’s question is the question.18

References Annas, J. 2004. “Being Virtuous and Doing the Right Thing.” Proceedings and Addresses of the American Philosophical Association, 78: 61–75. Annas, J. 2006. “Virtue Ethics,” in D. Copp, ed., The Oxford Companion to Ethical Theory. Oxford: Oxford University Press. Booin-Vail, D. 1996. “Don’t Stop Thinking about Tomorrow: Two Paradoxes about Duties to Future Generations.” Philosophy and Public Affairs, 25: 267–307. Cafaro, P. 2004. Thoreau’s Living Ethics: Walden and the Pursuit of Virtue. Athens, GA: University of Georgia Press. Coons, C. 2012. “The Best Expression of Welfarism.” Oxford Studies in Normative Ethics, 2: 206–37. Gressis, R. 2010a. “Recent Work on Kantian Maxims I: Established Approaches.” Philosophy Compass, 5(3): 216–27. Gressis, R. 2010b. “Recent Work on Kantian Maxims II.” Philosophy Compass, 5(3): 228–39. Hanley, R. P. 2009. Adam Smith and the Character of Virtue. Cambridge: Cambridge University Press. 18   Schmidtz (2001b) argues that we can respect nature without being species egalitarians. Indeed, the view that potatoes and chimpanzees have equal moral standing is incompatible with genuine respect for nature. Genuine respect acknowledges what living things have in common, but also acknowledges differences.

Virtue, Repugnance, and Deontology  193 Hill Jr., T. 2005. “Ideals of Human Excellence and Preserving Natural Environments,” in R. Sandler and P. Cafaro, eds., Environmental Virtue Ethics. Lanham, MD:  Rowman & Littlefield. Huemer, M. 2008. “In Defence of Repugnance.” Mind, 117: 899–933. Hursthouse, R. 1999. On Virtue Ethics. Oxford: Oxford University Press. Kant, I. 1996. Groundwork of the Metaphysics of Morals, in M. Gregor, ed., Practical Philosophy. New York: Cambridge University Press. Kawall, J. 2003. “Reverence for Life as a Viable Environmental Virtue.” Environmental Ethics, 25: 339–58. Leopold, A. 1966. A Sand County Almanac. Oxford: Oxford University Press. Mathews, F. 2001. “The World Grow Old: An Ethos of Countermodernity,” in D. Schmidtz and E. Willott, eds., Environmental Ethics: What Really Matters, What Really Works. Oxford: Oxford University Press. Narveson, J. 1967. “Utilitarianism and New Generations.” Mind 76: 62–72. O’Neill, O. 1989. “Consistency in Action,” in O. O’Neill, Constructions of Reason: Explorations of Kant’s Practical Philosophy. Cambridge: Cambridge University Press, pp. 81–104. Parfit, D. 1984. Reasons and Persons. Oxford: Oxford University Press. Sandler, R. 2007. Character and Environment: A Virtue-Based Approach to Environmental Ethics. New York: Columbia University Press. Schmidtz, D. 1992. “Rationality within Reason.” Journal of Philosophy, 89: 445–66. Schmidtz, D. 2001. “The Language of Ethics,” in C. Davis, ed., Ethical Dilemmas in the Water Industry. Denver, CO: American Water Works Association. Schmidtz, D. 2012. “Are All Species Equal?,” in D. Schmidtz and E. Willott, eds., Environmental Ethics: What Really Matters, What Really Works, 2nd edn. Oxford: Oxford University Press. Schmidtz, D. and E. Willott, eds. 2012. Environmental Ethics: What Really Matters, What Really Works, 2nd edn. Oxford: Oxford University Press. Wood, A. 2008. Kantian Ethics. Cambridge: Cambridge University Press. Zwolinski, M. and D. Schmidtz. 2005. “Virtue Ethics and Repugnant Conclusions,” in P. Cafaro and R. Sandler, eds., Environmental Virtue Ethics. Lanham, MD: Rowman & Littlefield.

10 But What About the Animals? Cheshire Calhoun

Rather than argue directly with destroyers of the environment who say, “Show me why what I am doing is immoral,” I want to ask “What sort of person would want to do what they propose?” (Hill 1991a: 108)

Tom Hill’s larger Kantian project is, in his words, “to distinguish Kant’s basic moral theory from unwarranted particular conclusions, to show its appeal so far as possible, to call attention to its shortcomings as I see them, and to suggest modifications to make Kantian ethics more plausible at least on some issues” (Hill 2002: 165). That project has moved Hill to take up topics neglected in both Kant commentaries and in ethics generally: self-respect, snobbery, promises to oneself, multicultural education, symbolic protest, and “respect” for the environment. Were one to make a list of those aspects of Kant’s ethical thought standing most in need of attention to their shortcoming, modifications of the view, and heavy work to demonstrate the appeal of Kant’s thinking, Kant’s moral views regarding animals might easily top the list. Hill himself describes Kant’s view that we owe nothing directly to animals as a “repugnant moral doctrine” (Hill 1994: 58). Yet Hill’s own remarks about the treatment of animals are, if kinder, even briefer than Kant’s (Hill 1994: 57–8). So one might reasonably wonder, “But what about the animals?” Kant offers what looks like an entirely unsatisfactory answer to that question. We have a moral duty to be kind rather than cruel to animals, but only because we have a duty to ourselves to cultivate dispositions, like compassion, that are morally useful in our dealings with humans. Our duties to animals are thus indirect ones. Some Kantians have argued that Kant’s indirect duty account is not as repugnant as it seems (Denis 2000). Others have reconstructed Kantian arguments for our having a direct duty to take animal interests into account (Wood 1998; Korsgaard 2004). In both cases, the ultimate aim is to show that Kantianism can do as well as utilitarianism on the animal question.

But What About the Animals?  195

1  The Scope of Our Ethical Concerns Aiming to do as well as utilitarianism on the animal question is, I think, the wrong aim. Humans make animals suffer in so many ways and sometimes with such extraordinary viciousness that it may easily seem that animal welfare is the only moral issue. But everyday moral thinking about animals is extremely complex. We do not, for example, eat our pets when they die (Diamond 1978). Animal trainers think about when it would be fair or unfair to punish misbehavior. And those who are rescued by service animals sometimes feel deep gratitude. Our question—“But what about the animals?”—thus asks more than just “What about animal welfare?” It also asks, “What about respect, fairness, and gratitude to animals?” It is just here that I think Kantian ethical thought is more promising than is utilitarianism . . . or so I plan to argue. Kant’s apparently repugnant view of animals follows directly from his highly attractive account of the nature of humans. Humans have, in addition to their animal nature, a rational nature. They are capable of willing freely on the basis of reasons, of setting ends for themselves and assessing means, of reflectively ordering their ends into a conception of their own happiness, and of regulating their conduct by principles, including moral laws. Were there no beings with a rational nature, Kant says, “the world would have no value” and no morality since both value and moral obligation originate in persons’ ability to ask what would constitute a reason for choice (LE 27: 344). This capacity for introducing value and morality into the world is the feature of persons that must be prized in ourselves and in others as something of inestimable value. As the source of value, including moral law, persons are ends-in-themselves and have a dignity beyond price. Through the fulfillment of our duties to persons, we honor their status as ends-in-themselves. We have no evidence that nonhuman animals have a rational nature. Many animals have a complex psychology, the capacity to learn and sometimes to use tools, to feel bodily pleasure and pain as well as emotions, and to guide their own actions in light of how the world appears to them. Kant acknowledges that animals make choices on the basis of representations and thus have a will, though not a free one. A cat, for example, represents a mouse as to-be-chased and chooses to chase the mouse.1 The choosing and the willing, however, are not free, because the cat cannot reflect on whether the representation of the mouse as to-be-chased is a reason for chasing the mouse. Instead, animal willing is “necessitated by incentives and stimuli. Their actions contain bruta necessitas” (LE 27: 344). Although he thought instinct largely determines what animals choose, Kant also recognizes that animals learn—he observed, for example, that birds teach their young their species-specific song (LP 9: 443). So although animals do not will freely, they are not, in Kant’s view, mere automata in the Cartesian sense of lacking any conscious mental states; nor are their actions entirely governed by instinctive impulses.2   See Christine Korsgaard’s discussion of animal incentives (Korsgaard 2004: 6–8).   Some of Descartes’s followers were alleged to have defended the vivisection of dogs on the grounds that the dogs were not really experiencing pain, but were only producing pain-like sounds via their 1

2

196  Cheshire Calhoun But sentience, emotional responsiveness, the capacities to learn, to represent objects in the world, and to choose to act on those representations, however estimable the occurrence of these abilities in nature are, differ in kind from the capacity to value, to set ends, and to legislate with others a set of binding moral laws. Given this difference, animals cannot, in Kant’s view, be the direct objects of moral obligations. A person is, he says, “a being altogether different in rank and dignity from things, such as irrational animals, with which one may deal and dispose at one’s discretion” (APV 7: 127; see also LE 27: 372–3). Whatever obligations we have that involve how we treat animals are in fact obligations to ourselves to preserve and enhance our own moral perfection. Critics have repeatedly located the repugnancy of Kant’s view in three places. First, Kant seems to get the object of moral obligation wrong. Surely we owe it first and foremost to, say, the starving stray dog to feed it, rather than owing it to ourselves to strengthen our compassionate dispositions. Second, Kant seems to get the ground of the obligation wrong. One might reasonably think that it is some fact about the value of the animal and of its welfare that is the reason why we should treat animals in this way but not in that. Third, Kant seems to get the stringency of requirements concerning animals wrong. On his view, all moral requirements regarding animals appear to be entirely conditional—dependent on contingent empirical facts about whether mistreatment of, or failures of kindness to, animals is likely to undermine the compassionate dispositions that aid us in fulfilling our obligations to fellow humans. Were the causal connection between mistreating animals and damaging our sympathetic and compassionate dispositions different than it typically is, we might have no obligation to avoid cruelty to animals and might even be obligated to be cruel if doing so caused us to be more compassionate to persons.3 Utilitarianism, by contrast, avoids such repugnancies. Because it takes the capacity to feel pleasure and pain and the having of interests to be morally significant facts about both human and nonhuman animals, it can easily deliver a non-repugnant account of the ethics of interacting with animals. Animals are the direct object of our moral duties, facts about animal sentience and animal interests ground those duties, and the primary reasons for treating animals well are not the causal effects that our treatment of animals might have on our treatment of humans. Given the virtues of a utilitarian approach, it looks as though a Kantian reconstruction needs to answer the specifically utilitarian challenge, “But what about the animals?,” by delivering an account of why cruelty and the wanton disregard of animal interests is wrong that does as well as the utilitarian account does. It is certainly true that a minimally adequate Kantian account needs to do at least as well as utilitarianism

body-machines. Kant is well aware of the tortuous pain suffered by animals undergoing vivisection. He denies the Cartesian view of animals as automata in CPJ 5: 464n.   For critiques of the causal claim see Wood (1998), Timmermann (2005), Skidmore (2001), Denis (2000).

3

But What About the Animals?  197 with respect to prohibiting the torturous treatment of animals. However, not all of our ethical thinking about animals can be accounted for by focusing on animal pains, pleasures, interests, and preferences. Moreover, some of Kant’s most interesting ethical observations about animals have nothing to do with animal pains, pleasures, interests, and preferences. In order fairly to assess the Kantian resources for addressing our ethical concerns about animals, we would do well to return to the question, “But what about the animals?,” and look more carefully at what exactly it is about our interactions with animals that we want an adequate view to address. To that end, consider the case of military dogs. In the year 2000, Representative Roscoe Bartlett introduced HR 5314 into Congress. Titled “A bill to amend title 10, United States Code, to facilitate the adoption of retired military working dogs by law enforcement agencies, former handlers of these dogs, and other persons capable of caring for these dogs.” HR 5314 was aimed at revising the Department of Defense policy concerning the disposition of military dogs who were no longer able to serve their military function or were no longer needed by the DOD. The poster child was Robby, an 11-year-old war dog and three-time Pentagon champion who was, at the time Representative Bartlett met him, arthritic and missing teeth. Although Robby’s handler wanted to adopt him, DOD policy required Robby to be returned to Lackland Air Force base where older military dogs would either be used to train new handlers or would be caged for up to a year and then euthanized. That DOD policy was a legacy of the 1949 Federal Property and Administrative Services Act, enacted after the Second World War, which reclassified military dogs as equipment. Under it, the nearly 4,000 dogs who served during the Vietnam war were treated as surplus equipment at the end of the war and either euthanized or given to the South Vietnamese army. Signed into law on November 7, 2000, HR 5314 made possible the adoption of retired military dogs. The treatment of military dogs as equipment that might become inoperable or surplus raises at least three distinct ethical concerns. One is the impact of this classification on the dogs’ animal interests when, for instance, they are deprived of ongoing human companionship, caged for long periods of time, and then euthanized. Separable from the issue of the animals’ interests is the classification of the dogs as mere equipment. Valuable equipment may be extremely well cared for; and service and competition animals regarded as expensive and not easily replaced pieces of equipment may receive much better care during their service life than many pets regarded as companions. But regardless of how well valuable animal equipment is treated, taking an entirely instrumentalist valuing attitude toward beings who are in fact more like us than like assault rifles may strike us as ethically deficient. From a utilitarian viewpoint, attitudes are ethically worrisome only insofar as they tend to result in ethically condemnable action. Kantian ethical thought takes our attitudes themselves, apart from the consequences of holding those attitudes, to be assessable.

198  Cheshire Calhoun Finally, there is the ethical concern that is really at the heart of HR 5314—the failure to reward the dogs’ long, faithful, and important service. “It was obvious to me,” says Bartlett, “that Robby is a dog who has faithfully served his country . . . It was also obvious that Robby has a special bond with his handler. Understandably so, as the two spent several years working side by side” (HR 5314, Cong. Rec. 2000, 21991). The ethical concern here, like the previous one, is not with animal interests or welfare. It arises from what seems like an entirely apt description of Robby in terms borrowed from the realm of human action and human moral appraisal: Robby was a co-worker who faithfully served his handler, and in so doing, the US. And that invites—to use Cora Diamond’s words—“the extension to animals of modes of thinking characteristic of our responses to human beings” (Diamond 1978: 474). In Robby’s case, we extend to a dog our thinking about what would constitute a grateful and fair reward. Examples of this mode of thinking abound. Consider, for example, Barbara Smuts’s ethical reflections on whether to wrest toys from the mouth of her playful dog, Sufi: “Since the toys belong to her, and since she never substitutes objects like my new shoes, it seems fair that she decides when to keep the toy and when to share it with me” (Coetzee 1999: 117). Extending to Sufi the characteristically human features of having property and respecting property is the basis for responding to her as one would to a human— by adopting a fair policy. One might, of course think that it’s simply a mistake to extend to animals “modes of thinking characteristic of our responses to human beings.” There’s no genuine ethical concern here to be accounted for. Perhaps that’s right. But our own cultural practices increasingly militate against the confinement of our ethical considerations concerning animals to the sheer animality of animals as dogs come to serve in ever-expanding service capacities and pet ownership practices increasingly treat pets like family members.

2  The Apparent Warrant for (Severely) Discounting Animal Interests We are now quite far removed from an ethical concern that utilitarianism, or any ethical view that takes utilitarianism as the gold standard, could address. But we are not so far afield of Kant’s ethical concerns about animals. Kant remarks that “if a dog, for example, has served his master long and faithfully, that is an analogue of merit; hence I must reward it, and once the dog can serve no longer, must look after him to the end, for I thereby cultivate my duty to humanity, as I am called upon to do” (LE 27: 459). Significantly, Kant extends to animals modes of thinking characteristic of our responses to human loyal service in the midst of articulating his “repugnant” doctrine of indirect duties to animals: “. . . gratitude for the long service of an old horse or dog (just as if they were members of a household) belongs indirectly to man’s duty with

But What About the Animals?  199 regard to these animals; considered as a direct duty, however, it is always only a duty to himself ” (MS 6: 443). In short, it is the very same view that both seems deficient by comparison to utilitarianism and seems to improve upon utilitarianism by addressing a distinctive ethical concern about the treatment of service animals. Kant’s indirect duty argument regarding animals is part of a wider set of comments about how we should think of our relation to animals that emphasize multiple analogies between humans and animals. Attending to those analogies, both as Kant states them and as a Kantian might reconstruct them, will enable us to see how a Kantian might address our three ethical concerns—with animal welfare, with appropriate inner valuing attitudes toward animals, and with extending to animals moral responses characteristic of our responses humans. In presenting the indirect duty argument, Kant claims that “Since animals are an analogue of humanity, we observe duties to mankind when we observe them as analogues to this, and thus cultivate our duties to humanity” (LE 27: 459).4 Spelled out, the indirect duty argument goes something like this: We have dispositions that aid us in the performance of our duties to other persons. One important disposition is the natural predisposition to feel sympathetic joy and sorrow at others’ states of joy or sorrow (MS 6: 456). That predisposition is responsive to human joy and suffering as well as to the analogous joy and suffering of animals. We have a duty to cultivate feelings, including compassionate feelings, that aid us in our moral interactions with fellow humans and to refrain from actions that would tend to weaken or destroy those feelings. Thus, Kant claims, we have an indirect duty “not to avoid the places where the poor who lack the most basic necessities are to be found but rather to seek them out, and not to shun sick-rooms or debtors’ prisons and so forth in order to avoid sharing painful feelings one may not be able to resist” (MS 6: 457). We also, and for the same reason, have a duty to avoid cruel treatment of animals, since cruelty “dulls [our] shared feeling of their pain and so weakens and gradually uproots a natural predisposition that is very serviceable to morality in one’s relation to other men” (MS 6:443). Three points are worth noting about this argument. First, Kant uses the indirect duty argument not only to ground duties regarding animals, but also to ground duties regarding persons—such as the duty not to avoid contexts that expose us to human misery—where the duty is not owed to the miserable, but to ourselves whose moral performance with respect to suffering persons depends upon our cultivating our compassionate dispositions. Second, the indirect duty argument applies generally to any disposition or feeling that aids morality. Kant did not think that compassion is the only serviceable feeling

4   Louis Infield’s translation of this passage reads: “Animal nature has analogies to human nature, and by doing our duties to animals in respect of manifestations which correspond to manifestations of human nature, we indirectly do our duty toward humanity” (Immanuel Kant, Lectures on Ethics, trans. Louis Infield (Indianapolis: Hackett Publishing Co., 1980)).

200  Cheshire Calhoun that we have an indirect duty to cultivate. We also have, for example, a disposition to love beautiful things independently of their use value. That disposition promotes, or at least “prepares the way” for morality (MS 6: 443). Because wanton destruction of beautiful, inanimate natural things weakens or destroys that disposition, we have a duty not to engage in wanton destruction of nature. And the argument might be extended to dispositions and feelings that Kant does not mention in this context. Hill, for example, argues that duties regarding the natural environment might be grounded on considerations of the way that viewing and treating nature as a mere resource may weaken or impede development of an attitude of proper humility (Hill 1991a). Third, and following from the previous point, our duties regarding animals are not limited to actions that preserve or cultivate compassionate feelings toward human suffering. Kant makes the general claim that “if the acts of animals arise out of the same principium from which human actions spring, and the animal actions are analogues of this, we have duties to animals, in that we thereby promote the cause of humanity” (LE 27: 459). This claim is immediately preceded by the injunction to reward dogs’ service as analogous to human service, where, presumably, rewarding animal service helps cultivate our disposition to feel gratitude, thereby aiding us in appropriately responding to human benefactors. At the heart of the indirect duty argument is a causal claim: Some actions regarding both persons and non-persons have the causal effect of weakening, destroying, or interfering with the cultivation of dispositions that aid us in the performance of our moral duties to persons. So, for example, the person who has a dog that can no longer serve shot “damages the kindly and humane qualities in himself ” (LE 27:459). And, presumably, the well-off who avoid witnessing the miseries of poverty miss an opportunity to cultivate their compassionate disposition. But for our treatment of animals and the poor to have such widespread causal effects on our treatment of persons in general, we must first take the suffering of animals and the poor to be analogous to our own, not just as mental states but also in importance. In his statement of the indirect duty argument, Kant seems to just assume that our compassion is naturally calibrated to respond to both the suffering of different human kinds and the suffering of different animal kinds as though those sufferings were analogously important; as a result, hardening one’s heart to animal suffering or the suffering of the impoverished involves ignoring reasons for being kind and avoiding cruelty to all species and human kinds. But of course compassion and other morally serviceable dispositions are often not so calibrated. Which kinds of beings we are able to feel compassion, respect, gratitude, and other morally serviceable feelings towards is affected by how we think about the value of those kinds. People who take themselves to be members of a superior social kind—whether that be a class, a race, a nationality, or a sex—all too often engage in the systematic mistreatment of people they take to be members of an inferior social kind without it at all affecting how they treat their own kind. Nazi doctors, for example, went home after performing horrific experiments on Jews to loving relations with their family and

But What About the Animals?  201 respectful relations with their fellow Aryans. White researchers in the Tuskegee syphilis study withheld penicillin and information about its effectiveness in treating syphilis from the black participants in the study. Presumably, they would not have done this to whites. What prevents cruelty, ingratitude, contempt, and so on toward a supposedly inferior kind of person from causally affecting the dispositions that support beneficent, grateful, and respectful treatment of a supposedly superior kind of person are precisely the suppositions of inferior and superior worth. The racist directs mistreatment at Jews or blacks because he regards them with contempt as an inferior kind of human. Their suffering is not, to his mind and heart, analogously important even if it is qualitatively similar to the suffering of “superior” human kinds. All the more easily may mistreatment fail to have widespread causal effects when directed at animals who are not an inferior kind of human, but no human at all. Like the racist who feels no compassion for the miseries of inferior races, or the classist who views the miseries of the poor without compassion, so too the human, convinced of his greater human worth, may regard the suffering of animals without compassion precisely because, on his view, the sufferings are not analogous. One is the suffering of a mere animal. The other is the suffering of a valuable human being. For those who see no reason in the first place to accord humans and animals an analogous regard, the causal effects of animal cruelty on their treatment of fellow humans cannot be what gives them a reason to avoid animal cruelty, since there are, for them, no such causal effects. Instead, what we want to say is that they should see animals and humans as being more analogous than they do. Thus they should be so constituted that cruelty to animals has more widespread causal effects. The analogy must be more than just that both human and nonhuman animals are sentient and have analogous experiences of pain and pleasure. It must be one that justifies our taking animal and human suffering as analogously important. And here it may be difficult to imagine what a Kantian could say. The mistake that the racist or classist makes when she discounts the suffering of some human kinds is that she fails to recognize the moral equality of persons. The person who discounts the suffering of nonhuman animals does not make the same mistake. As Kant says, rational beings are “altogether different in rank and dignity from things, such as irrational animals” (APV 7: 127). And that would seem to warrant our discounting animal suffering, treating it as disanalogous in importance, however analogous animal and human suffering may be as felt experiences. If a Kantian view is to do at least as well as utilitarianism on animal welfare issues, we need some positive account of why, despite animals’ lack of a dignity beyond price, we should regard animal and human suffering as analogously important. Only then will we have an account of why animal and human responses that are described in the same terms should get analogous uptake in our emotional dispositions. And it’s not just Kant’s own indirect duty argument that depends on our being able to give such an account. Kantian reconstructions that attempt to ground a direct duty to animals in some feature shared by humans and animals must explain why the fact that animals

202  Cheshire Calhoun lack dignity beyond price is not a reason for discounting the moral significance of that shared feature when it is possessed by animals. Allan Wood and Christine Korsgaard both argue that there is some common feature of human and nonhuman animals that grounds a direct duty to animals to avoid cruelty and wanton disregard of their interests. Our duties are thus owed directly to animals in virtue of features of animals themselves and not in virtue of an assumed causal connection between cruelty and kindness to animals and cruelty and kindness to persons. Wood argues that respect is owed not to individual persons but to rational nature. Animals possess “fragments” (Wood 1998: 198, 200), or the “infrastructure” of (200), or “parts” of (197) that rational nature insofar as they make choices and act on the basis of representations. There is, thus, the same thing in nonhuman animals that is owed an attitude of respect and honoring in human animals, viz., the infrastructure of rational nature. “To frustrate an animal’s desires or to cause it pain maliciously or wantonly is to treat with contempt that part of rational nature which animals share with human beings” (Wood 1998: 200). In humans, of course, the infrastructure of rational nature is connected to capacities for using concepts, for making judgments, and for rational autonomy—features that are essential to our being able to place each other under moral obligation, to have moral rights, and to give ourselves the moral law. But given this moral difference that having a rational nature versus having only a fragment makes, we cannot just appeal to what animals and humans share in order to ground obligations concerning animal welfare. We also need a positive account of why the morally significant differences between animals and humans do not warrant discounting (perhaps severely discounting) animal interests.5 Korsgaard similarly argues that nonhuman animals and human animals share a morally significant feature. In her view, it is not a fragment of rational nature but rather our animal nature itself. All animals, human and not, are creatures for whom things can be naturally good or bad. When we demand to be treated as ends in ourselves by not being made to suffer, we confer normative significance on the fact that we have an animal nature. “We legislate that the things that are good or bad for beings for whom things can be good or bad—that is animals—should be treated as good or bad objectively and normatively. In other words, we legislate that animals are to be treated as ends in themselves” (Korsgaard 2011: 109). In her view, we cannot consistently place value on our own animal nature while refusing similarly to value nonhuman animals’ animal nature (Korsgaard 2011: 117 n. 51). But of course, when we value our animal nature, we do so within a wider set of evaluative commitments, including most fundamentally, valuing our own rational nature—a nature animals do not share. It would thus seem that we still need a positive account of why the morally significant differences

5   Skidmore (2001) raises a set of worries concerning the significant difference between beings with a rational nature and beings with fragments of rational nature.

But What About the Animals?  203 between animals and humans do not warrant discounting (perhaps severely discounting) animal interests. It is to that positive account that I now turn.

3  Valuing Animals Kant argues that we have duties to ourselves as animal beings. Those include duties not to commit suicide, not to mutilate oneself, not to defile oneself with lust, and to avoid the vices of drunkenness and gluttony (MS 6: 420, 421–7). We also have duties to harden our bodies against discomforts and toward frugal needs (LE 27: 379–80), as well as duties to promote the body’s “vigour, activity, strength, and courage” (LE 27: 380). And if we violate those duties, how are we to think of ourselves? Kant sometimes says that we thereby debase ourselves to the level of animals. The glutton, for example, indulges in a cow-like enjoyment (MS 6: 427). But more frequently, Kant makes the curious (for him) observation that when humans violate their duties to themselves as animals, humans debase themselves beneath the beasts (MS 6: 425, 6: 426; LE 27: 373, 27: 380). “[W]‌e deny the dignity of humanity,” he says, “when we, for example, take to drinking, commit unnatural sins, practice all kinds of immoderation, and so forth, all of which degrade the human being far below the animals” (LP 9: 489). But how, on Kant’s view, could humans ever degrade themselves below the level of beasts? Even the most self-degrading behavior does not alter the fact that human persons have a dignity beyond price and animals do not. To see how we might debase ourselves beneath beasts, we need to turn from the analogies rooted in our shared animal nature (such as our analogous capacity to suffer) to the analogy that obtains between the very things that make us different, namely, how our actions are produced. Human action is not a product of the operation of natural, deterministic laws but the product of free choice to conform one’s actions to law. Animal behavior, by contrast, is a product of deterministic laws. Despite that substantial difference, instinct, or more generally, deterministic psychological principles, play a role in the lives of animals analogous to the role that reflectively adopted principles play in our lives: both impose constraints on action and both render action orderly. The orderliness of human action, however, is not inevitable. The orderliness of animal behavior is. Animal instincts, for example, generally limit sexual behavior to behavior that serves reproduction. Human sexual impulses are not similarly constrained by natural instincts. Instead, rationally chosen principles of sexual restraint play, for us, the role of limiting sexual impulses. Animal instincts for self-preservation prevent animals from engaging in self-destruction. Humans, however, have the ability to conceive of incentives to commit suicide (from escaping a miserable future to preserving one’s honor). For us, a rationally chosen principle forbidding suicide plays the role of excluding self-destruction that instinct plays in animals. Animals don’t have to work at not mutilating themselves, and avoiding drunkenness and gluttony, while humans must adopt principles prohibiting

204  Cheshire Calhoun self-mutilation, drunkenness, and gluttony. Moreover cleanliness comes naturally to animals (Kant mentions the cleanliness of baby birds and prey animals with approval; LP 9: 441, 9: 495–6); and animals need no discipline as human children do, to develop their physical powers, put up with discomfort, and learn to have frugal needs. In short, animals instinctively do what humans do by imposing duties on themselves. A human “has no instinct and must work out the plan of his conduct for himself ” (LP 9: 441). While our ability to work out a plan for ourselves gives us dignity, it is also true, Kant says, that freedom that is not restrained under certain rules of conditioned employment . . . is the most terrible thing there could ever be. All animal acts are regular, for they take place according to rules that are subjectively necessitated. In the whole of non-free nature we find an inner, subjectively necessitated principium, whereby all actions in that sphere take place according to a rule. But if we now take freedom among men, we find there is no subjectively necessitating principium for the regularity of actions . . . If freedom is not restricted by objective rules, the result is much savage disorder. For it is uncertain whether man will not use his powers to destroy himself, and others, and the whole of nature. (LE 27: 344)

The same rational nature that gives us a dignity beyond price, elevating us above the rest of the animal world, also comes with the potential for a lawlessness that reduces us beneath the level of nonhuman animals. “Now if man freely follows his inclinations,” Kant explains, “he is lower even than the animals, for in that case there arises in him a lawlessness that does not exist among them” (LE 27: 345). And that disorder may be savage indeed. Animal rescue organizations feature horrifying stories of animal abuse in their solicitation letters. To mention just one: a dealer bought an already starved horse at auction. She refused to load into his trailer. In a rage, the dealer wrapped barbed wire around her head, tied her to the back of the trailer, and began to drive away. When she fell, he unhooked her from the trailer and backed the trailer over her head. (Amazingly, she survived.)6 The story’s donation-eliciting effectiveness lies in part in the way it invites compassionate animal lovers to feel terrible for the horse. But its effectiveness also lies in its inviting us to think not just “What kind of person would do such a thing?” but “What kind of species would do such a thing?” and thus, out of shame for being a fellow human, to make amends to animals on behalf of humans. That animals have on us not just orderliness and non-destructiveness but better performance with respect to what, for us, are duties to ourselves as animal beings does not, of course, make animals ends-in-themselves or warrant our regarding animals with the respect owed free beings. But these reflections do undercut a kind of species-conceit or snobbery that often goes along with the self-justifying claim, “It’s just an animal.” Those who discount animal interests with the thought, “They are just animals, while we are humans,” seem overly puffed up about belonging to the human   Habitat for Horses solicitation letter.

6

But What About the Animals?  205 species. Compared to fully rational beings, we humans have a great deal to be humble about. Compared to possible species of rational beings that Kant imagined existed on other planets, the human species also has reason for humility. Even without such comparisons, a clear-eyed survey of the human species might well lead us to think that the correct response to the query (posed by Kant) as to whether our species is good or bad is Kant’s own: “there is not much to boast about in it” (APV 7: 331). While snobbery towards humans involves a failure to appreciate humans’ equal dignity (Hill 1991b), species snobbery involves a failure to appreciate the place of humans in relation to nonhuman animal species whose orderly, non-destructive conduct both accords with what are, for us, duties with respect to one’s animal nature as well as provides a model of what, for humans, can only be achieved through a disciplined will, and even then, is not consistently achieved across the human species. Proper humility, Hill argues, involves self-acceptance: facing squarely one’s powers and limits and one’s affinities with and differences from other beings (Hill 1991a: 114). While correctly appreciating the importance of humans’ rational nature, the species snob fails to face squarely the fact that human dignity goes hand in hand with the capacity to degrade oneself beneath the beasts, and that many of our fellow humans do. The human who views animals as inferiors and uses nonhumanness to discount the interests of fellow animals does not make the same mistake that the racist and classist do when they view members of different social kinds as inferiors and use that difference to discount the interests of some fellow humans. The racist and classist fail to recognize the equal humanity of persons; and animals lack humanity. But the species snob does make a mistake. In asking how he should value the human species in relation to nonhuman animals, and thus whether and how much he should discount animal suffering as the suffering of an inferior being, he stacks the deck in his favor by only describing himself as a rational being who has the capacity to give himself the law. He neglects to note that he is also a natural species of animal. A proper appreciation of the analogies between what animal and human nature can achieve when their behavior conforms, respectively, to deterministic and to moral laws, and a proper humility about the destructive potential of human freedom give us reason to adopt a more humble attitude toward ourselves and a less arrogant attitude toward animals. What is wrong with disregarding the interests of “mere animals” and with regarding animals merely as equipment for human use are the objectionable attitudes of arrogance about humans and contempt for animals that it expresses. Instead, our valuing attitudes toward animals should be calibrated to reflect not only the fact of human dignity but also the facts that animals’ lawful care for themselves is an analogue of our own lawful fulfillment of duties to ourselves, and that their non-destructiveness and non-self-debasement is an analogue to what we aim to achieve. These kinds of reflections address our first two ethical concerns by explaining what is wrong with discounting animals’ welfare interests as well as what is morally defective about adopting an instrumentalist valuing attitude towards animals and regarding them as serviceable, inoperable, or surplus equipment. These reflections do not,

206  Cheshire Calhoun however, address our third ethical concern, since they do not explain why failure to gratefully reward service animals is morally criticizable. Indeed, it may be hard to see how it could be morally criticizable. From the service dog’s or horse’s point of view, what matters are a set of natural goods—food, shelter, companionship, absence of fear and pain, space to move about, and so on. That these are given as a reward—a gesture of reciprocity—and out of grateful appreciation could hardly matter to them. Nor is the service animal, when it serves, adopting our ends as its own. Although the bomb-sniffing military dog does us a service in the sense that we benefit from his actions; his is not the service of a benefactor who respects our end-setting capacity.

4  The Extension to Animals of Modes of Thought Characteristic of our Responses to Humans The puzzling view that many of us, including Kant, have that a grateful return is owed to service dogs is only one instance of a general pattern of introducing moral considerations into our thinking about what we owe to animals that are difficult to make sense of as anything more than anthropomorphism. The general pattern goes like this: treatment X is wrong because it is unkind, cruel, neglectful of animals’ interests and the wrongness is compounded by a moral failure that has nothing to do with the animal’s interests. So for example, caging service dogs for long periods of time and then euthanizing them is wrong because it neglects the animals’ interests. That wrong is compounded by ingratitude. The cruelty in Kant’s day of requiring carthorses to work long hours pulling heavy loads for starvation rations of hay is compounded by the unfairness of expecting them to do so much work for so little reward. The cruelty that baboons were subjected to in Gennarelli’s infamous head injury study was compounded by researchers’ mockery of brain-damaged apes. The neglect of animals’ interests involved in hunting them for sport is compounded by not giving them a fair or sporting chance. The cruelty of tying a firecracker to the tail of the family pet is compounded by taking advantage of her trust in order to do so. And so on. If moral wrongs to animals really can be compounded in this way, then two things would need to be true. First, what animals do is sometimes appropriately described in human terms. Second, we ought to respond to animals so described the same way we would to humans by, for example, not mocking their undignified behavior and by giving fair reward for service. Kant apparently did think that animals’ behavior is appropriately described in terms borrowed from descriptions of human action, because in his view animals provide an analogon rationis.7 For example, he says of dogs: 7   “Animals are wholly lacking in consciousness, their conduct occurs according to laws of imagination, which nature has laid within them—by analogy. That principle, which guides the animal as analogon rationis, is called instinct, the faculty to carry out actions without consciousness, for which humans require consciousness . . .” (From Dohna’s notes on Kant’s lectures on metaphysics, quoted in Naragon 1990: 20).

But What About the Animals?  207 [Dogs] seem to be the most perfect animal, and to manifest most strongly the analogon rationis . . . they carefully look after their responsibilities, remain with their master, if they’ve done something wicked they become disturbed, and if they see their master angry try to win him over with a submissive posture.8

And less flatteringly of monkeys: “Although monkeys have an analogon rationis, no analogon moralitatis will be found in them, as they are always wicked, spiteful and obstinate, and everywhere they go, they wreck havoc.”9 One might wish for more in the way of justification for using seemingly anthropomorphizing descriptions. I propose this: In some cases, the best available description of what animals are up to is in human terms. Since we do not know what the inner lives of animals are like, we have no basis for developing a distinct descriptive language for animal behavior that captures the connection between what animals do and animals’ actual inner lives. So we have a choice between using behaviorist stimulus-response descriptions that attempt to stay clear of any speculations about the content of animals’ inner lives; or we can employ behavioral descriptions designed for beings whose inner lives we do know about (us), such as “looking after their responsibilities” or “being spiteful.” Applying to animals descriptions borrowed from the world of human action often has a great deal more explanatory and predictive capacity. We can, of course, ask when we may attribute the sorts of analogies that would warrant analogues to our moral responses to humans. It seems to me we have the strongest warrant for doing so when we can see animals not just as having an analogon rationis, but as fellow social participants. One of the things we do with animals—and increasingly so—is incorporate them into animal-human social worlds. By ‘animal-human social world’ I mean a set of human practices involving animals that are specifically designed to incorporate animals as fellow participants in some shared human practice or project. Pet ownership practices increasingly incorporate animals into the family lives of humans. Pets share a human residence, go along on vacations, are played with, become jogging partners, are sent off to daycare or provided with sitters, share couches and beds, and are taken to school and to pet therapists to learn how to be more cooperative participants in the animal-human social world. Animals who are pets learn to fit into the routines of humans’ lives and learn to communicate their needs to their human owners—to ask, demand, beg, refuse, and protest. Perhaps as a result of the expanding practices of pet keeping, animal training has moved away from a focus on “breaking” toward training that focuses on communication, respect, and trust (witness the Also, “in comparing the artistic actions of animals with those of human beings, we conceive the ground of the former, which we do not know, through the ground of similar effects in humans (reason), which we do know and thus as an analogue of reason, and by that we mean also to indicate that the ground of the artistic capacity in animals, designated as instinct, is in fact specifically different from reason, but yet has a similar relation to the effect (comparing, say, construction by beavers with that by humans)” (CPJ 5: 464n). 8   Translated by and quoted in Kain (2010: 13) from Kaehler and Messina’s notes on Kant’s lectures on physical geography. 9   Kain (2010: 12).

208  Cheshire Calhoun many horse and dog whisperers). Pets and performance animals are taught to respect humans—to behave well because they understand their place in a hierarchy of authority within a social world populated by both humans and other animals. As we ask animals to cross species lines and participate in an animal-human world, we give animals a second nature analogous to the second nature humans give their children through education, discipline, and enculturation. While animals’ species vocation is not, as ours is, “to emerge from our crude state of nature as animals” (LE 9: 492), one principal way in which we “dispose” of animals is precisely to redirect their desires and instill new behavioral regularities, fitting them for life with us. There is nothing fanciful in Kant’s description of dogs carefully looking after their responsibilities, being disturbed by their own wicked (rule-breaking) deeds, or trying to win over their master. Having given dogs an analogue to our own second nature, their motives and behavior are appropriately read as an analogue of our own. Even so, one might doubt that animal analogues of our own second nature provide us with moral reasons. Why ought we to respond with gratitude to a mere analogue of loyal service? One thought might be that animal analogues do not in fact supply us with moral reasons. There are nevertheless reasons having to do with rational consistency to respond analogously to analogous behavior. Given that animals have an inner life and act from analogous principles, we have good reasons to, for example, regard service dogs as doing an analogue of human service work, rather than as doing an analogue of what useful pieces of equipment, such as metal detectors and heart monitors, do. We have good reasons to fairly reward human service work. It would thus be rationally inconsistent not to reward animal behavior that is best described as “service work,” even if we don’t think there is a moral obligation to reward animal service. On this view, DOD policy was objectionable because it prevented dog handlers (among others) from behaving in a rationally consistent way toward their canine co-workers. This analysis does not capture what we might have hoped to capture, namely the thought that there is something morally defective about the absence of gratitude toward an animal that renders loyal service. Moreover, one might find the appeal to rational consistency misplaced. Animal and human service, along with the principles from which they spring, are merely analogous. Things that are merely analogous also differ in important ways. Animal benefactors do not freely adopt our ends as their own (though what they do may be analogous). Their beneficence cannot be a form of respect for persons as end-setters. It thus cannot obligate us to gratitude. In addition, ingratitude toward animals cannot have the morally undesirable effect that it sometimes has on humans of undermining the incentive to fulfill one’s duty of beneficence, since (among other reasons) animals have no such duties. Finally, since animals presumably have no concept of gratitude, we can’t even have prudential reasons for making a grateful return. Whatever prompts animals to further service for us is not our past expression of gratitude (past praise is a different matter). In short, the analogies

But What About the Animals?  209 between animal and human beneficence are not of the right sort to make limiting one’s gratitude to humans a form of rational inconsistency. Notice that the preceding argument does not depend, as arguments for discounting animal suffering do, on one’s being puffed up about the human species in relation to inferior nonhuman species. It depends only on thinking that animal service isn’t the same as human service and that the analogies aren’t of the right sort to support a moral obligation to gratitude. So is there nothing morally defective about not gratefully rewarding service animals? I think there is. I suggest we return to Hill’s question, “What kind of person would do something like that?” Central to Hill’s Kantian reconstructions has been his emphasis on an ethics of attitudes. Acts that do not violate moral duties might nevertheless be criticizable for the attitudes they display. The snob’s inwardly looking down his nose at those he thinks inferiors is a moral defect even if outwardly he takes care to do his moral duty (Hill 1991b: 158); and snobbish behavior is wrong because of the attitude it displays. While attitudes that involve a failure to properly respect persons clearly involve a moral defect, defect may also be involved where the attitudes are not directed toward persons. Hill suggests that part of what is wrong with servility is a morally defective attitude toward morality. Someone who properly respected morality would be disposed to learn his place in the morality system and affirm it proudly (Hill 1991c: 14). Treatments of animals that display ingratitude, contempt, indifference to fairness, and so on cannot be criticized simply because of the attitudes they display. Animals are not persons, and thus we are not rationally required to adopt and display toward animals the attitudes we are required to adopt toward persons. However, not opting to extend those attitudes toward animals seems to involve a criticizable small-mindedness about morality. Kant suggestively remarks that “If a master turns out his ass or his dog, because it can no longer earn its keep, this always shows a very small mind in the master” (LE 27: 460). Kant does not explain what that small-mindedness amounts to. But we might say this: One may view the opportunities to honor morality in one’s life in either a small-minded or a generous way. The small-minded person, while committed to satisfying the requirements of duty, does not think about how her life might honor morality in ways that are not strictly required. The small-minded person limits her moral concern to her actions, doing what the law requires and doing it out of duty, but not being additionally concerned with her inner attitudes; she may inwardly mock or look down on others or regard them with indifference or contempt. She also takes her imperfect duties of beneficence and of cultivating her talents as permissions to ignore others’ needs and to indulge in laziness just so long as she satisfies the letter of the law. So, recognizing that she is not strictly required to respond to others’ needs on any particular occasion, even when she might do so easily and with virtually no cost to herself, she often avoids being kind or generous. While the small-minded person wrongs no one, her view of morality is of a yoke she must bear rather than something to be prized and

210  Cheshire Calhoun honored by living in accord with the spirit, not just the letter, of the law. Instead, she wants to be as unconstrained by morality as she can justifiably be, and she uses the letter of the law to accomplish just that. Small-mindedness may also manifest itself in one’s moral appreciation of others’ behavior. Someone who is generously, large-minded about morality would be disposed to notice and appreciate behavior that accords with duty even where the behavior may have no moral worth. For the most part, we do not know what in fact motivates particular acts of beneficence, respect for property, honesty, justice, and so on. Some of that morally noteworthy behavior likely springs from principles that don’t involve self-regulation, for example, habit, fear, unthinking conformity to social norms, or a naturally benevolent disposition. Someone who insisted on reserving her gratitude only for benefactors she could be confident had adopted her ends as their own would display an extraordinary small-mindedness about morality. So too would someone who reserved her moral appreciation only for those she was confident told the truth or respected her property out of respect for the moral law. This account of small-mindedness suggests several ways that those who refuse to extend their moral responses to animals are small-minded. The small-minded person insists that service animals are not really benefactors, and thus are not owed gratitude; the family dog does not really respect one’s ownership of one’s shoes, and so is not owed a return respect of its ownership of its toys; the performance horse does not really respect behavioral rules, and so punishment can be neither fair nor unfair. What seems small-minded about this is, first, the lack of appreciation of action that accords with duty. The stinginess of not extending to animals a positive moral description of what they do is of a piece with the stinginess of not describing human action in positive moral terms except when one is confident the action has moral worth. Second, what seems small-minded is the effort to limit one’s moral response to what is strictly required, even when that moral response might be easy and relatively cost free. Such a small-minded person might see the analogies between human and animal action—might see both as serving, or respecting property, or respecting rules—but she limits her gratitude, or reciprocal respect for property, or concern with fair punishment to those to whom she must morally respond—fellow humans. Finally, there is a kind of small-mindedness involved in not practicing one’s moral responsiveness where one reasonably can. Animals present us with reasonable analogues of positive moral behavior and thereby occasions for practicing gratitude, respect for property, and fairness. Perhaps this will improve our moral performance towards fellow humans, as Kant thought it would. But even if it does not, such practice is a way of generously honoring the spirit of morality.

But What About the Animals?  211

Abbreviations for Kant’s Works MS Metaphysics of Morals LE Lectures on Ethics APV Anthropology from a Pragmatic Point of View LP

Lectures on Pedagogy

CPJ Critique of the Power of Judgment

References Coetzee, J. M. 1999. The Lives of Animals. Princeton: Princeton University Press. Diamond, Cora. 1978. “Eating Meat and Eating People.” Philosophy, 53(206): 465–79. Denis, Lara. 2000. “Kant’s Conception of Duties Regarding Animals:  Reconstruction and Reconsideration.” History of Philosophy Quarterly, 17(4): 405–23. Hill, Thomas E. 1991a. “Ideals of Human Excellence and Preserving Natural Environments,” in Autonomy and Self-Respect. New York: Cambridge University Press, 104–17. Hill, Thomas E. 1991b. “Social Snobbery and Human Dignity,” in Autonomy and Self-Respect. New York: Cambridge University Press, 155–72. Hill, Thomas E. 1991c. “Servility and Self-Respect,” in Autonomy and Self-Respect. New York: Cambridge University Press, 4–18. Hill, Thomas E. 1994. “Respect for Humanity.” Tanner Lectures on Human Values, April 26 and 28, 1994. . Hill, Thomas E. 2000. Respect, Pluralism, and Justice: Kantian Perspectives. New York: Oxford University Press. Hill, Thomas E. 2002. Human Welfare and Moral Worth: Kantian Perspectives. New York: Oxford University Press. HR 5314 of the 106th Congress. 2000. “To amend title 10, United States Code, to facilitate the adoption of retired military working dogs by law enforcement agencies, former handlers of these dogs, and other persons capable of caring for these dogs.” 146 Congressional Record-House Tuesday, Oct. 10, 21990–21992. Kain, Patrick. 2010. “Duties Regarding Animals,” in Lara Denis, ed., Kant’s Metaphysics of Morals: A Critical Guide. Cambridge: Cambridge University Press. Kant, Immanuel. 1960. Education. Ann Arbor: University of Michigan Press. Kant, Immanuel. 1997. Lectures on Ethics, trans. Peter Heath, ed. Peter Heath and J. B. Schneewind. Cambridge: Cambridge University Press. Kant, Immanuel. 1991. Metaphysics of Morals, trans. Mary McGregor. Cambridge: Cambridge University Press. Kant, Immanuel. 2011. Anthropology, History, and Education, trans. Robert B.  Louden, ed. Gunter Zoller. Cambridge: Cambridge University Press. Kant, Immanuel. 2006. Anthropology from a Pragmatic Point of View, trans. and ed. Robert B. Louden. Cambridge: Cambridge University Press.

212  Cheshire Calhoun Korsgaard, Christine M. 2004. “Fellow Creatures:  Kantian Ethics and Our Duties to Animals.” Tanner Lectures on Human Values, February 6. . Korsgaard, Christine M. 2011. “Interacting with Animals:  A  Kantian Account,” in Tom L. Beauchamp and R. G. Frey, eds., The Oxford Handbook of Animal Ethics. New York: Oxford University Press. Naragon, Steve. 1990. “Kant on Descartes and the Brutes.” Kant-Studien, 81(1): 1–23. Skidmore, J. 2001. “Duties to Animals: The Failure of Kant’s Moral Theory.” The Journal of Value Inquiry, 35(4): 541–59. Timmermann, Jens. 2005. “When the Tail Wags the Dog: Animal Welfare and Indirect Duty in Kantian Ethics.” Kantian Review, 10(1): 128–49. Wood, A. W. 1998. “Kant on Duties Regarding Nonrational Nature: Allen W. Wood.” Aristotelian Society Supplementary, 72: 189–210.

PA RT I V

Kant’s Ethics

11 The Supererogatory and Kant’s Imperfect Duties Marcia Baron

1 It is an honor to contribute to a volume on Tom Hill’s work. But what to write on? This decision was not an easy one, given how many stimulating and important papers Tom has written, several of which have played a leading role in putting a topic on the philosophical map. I’ve settled on my topic because his work in this area had a particularly significant impact on my own. At the time that I first read “Kant on Imperfect Duty and Supererogation,”1 I had recently decided to work intensively on Kant’s ethics (thanks in part to the outstanding 1983 NEH Institute on Kantian Ethics and in particular the lectures by Tom and by Onora O’Neill) and was mulling over the claim that the scope of duty is too broad in Kant’s ethics, leaving no room for anything “beyond” duty. Tom’s groundbreaking essay helped me improve my grasp of Kant’s taxonomy of duties and then develop my own reply to the claim that Kant does not leave room for the category of the supererogatory. I agreed (and agree) with Tom in rejecting the position that Kant cannot do justice to the facts to which J. O. Urmson drew our attention in “Saints and Heroes,” inter alia, that there are saints and heroes, and saintly and heroic acts.2 Urmson’s claim that for Kant “every duty is equally and utterly binding on all men”3 is not incorrect, exactly, but it occludes the fact that duties come in various shapes (as well as differing levels of abstraction). Duties differ in the sorts of things they bind us to do; moreover, what they 1   Thomas E.  Hill, Jr., “Kant on Imperfect Duty and Supererogation,” Kant-Studien, 62 (1971), 55–76, reprinted in Hill, Dignity and Practical Reason (Ithaca, NY: Cornell University Press, 1992), 147–75. Unless otherwise indicated, all subsequent quotations from Hill will be to this work, using the pagination from Dignity and Practical Reason, and page references will be inserted parenthetically into the text. 2   J. O. Urmson, “Saints and Heroes,” in Essays in Moral Philosophy, ed. A. I. Melden (Seattle: University of Washington Press, 1958), 198–216. 3   Urmson, “Saints and Heroes,” 206.

216  Marcia Baron bind us to is not always something aptly characterized as a doing. Some Kantian duties are not duties to do this or that action but to embrace an end, and they do not entail duties to (say) help people in need at every opportunity. Saintly and heroic acts are not morally required, except where the alternative is a violation of a narrow (perfect) duty. (Refusing to sign a false deposition that will send an innocent man to his death is heroic if one knows that by refusing one is in effect signing one’s own death warrant, but it is nonetheless morally required.) So far Tom and I are in agreement. But exactly how can Kant do justice to saintly and heroic acts (and more jejune good acts of a sort we hesitate to classify as morally required yet also do not want to view as merely permissible)? How can he avoid treating them as simply permissible? Tom’s way of working it out seemed to me to involve something of a stretch, or perhaps a forcing of a round peg into an octagonal hole. As I sought to pinpoint the apparent failure of fit, I became more convinced that Kant offered something distinctive, different from and arguably superior to the standard approach of recognizing a category of supererogatory acts. Hill and I agree that Kant can (mostly) do justice to the facts that there are saintly and heroic acts, but differ on how he does them justice. In sections 2, 3, and 5, I present and probe Hill’s explanation and indicate points where I disagree with him, focusing in these sections on his “Kant on Imperfect Duty and Supererogation” (hereafter, KIDS); in section 4, I consider another argument one might give for thinking that Kant has room for supererogatory acts, viewed as such, and reject it. In section 6, I briefly consider whether Hill’s more recent work on beneficence and Kant’s imperfect duties reflects a change in his position and note that although in discussing KIDS in “Meeting Needs and Doing Favors”4 he seems generally not to take his view to have changed,5 the fact that he lays emphasis on the more cautious statements in KIDS, not on the claims with which I take issue below may perhaps indicate a (subtle) change in his position. In sections 7 and 8, I expand on my reply to the claim that Kant cannot do justice to the category of the supererogatory.

2 I turn now to the details of Hill’s position. Pointing out that Kant’s imperfect, or wide, duties differ in their “width”—i.e. in their latitude—Tom suggests that if “Kant’s scheme has a place for supererogatory acts . . . they will be found as a subclass of acts which fulfill principles of wider imperfect duty” (168). Only a subclass: “we cannot equate supererogatory acts with those that fulfill principles of wider imperfect duty, such as beneficence,” for “in some circumstances promoting the happiness of others will be obligatory; e.g. if one’s only alternatives on that occasion are contrary to perfect 4   Thomas E.  Hill, Jr., “Meeting Needs and Doing Favors,” in his Human Welfare and Human Worth (Oxford: Oxford University Press, 2002), 201–43. 5   One exception occurs in note 38 of “Meeting Needs.” See my note 18 and accompanying text.

The Supererogatory and Kant’s Imperfect Duties  217 duty or if one who has continually neglected to help others is faced with his last opportunity” (168).6 The “best candidate for a supererogatory act” is an act that (a) is of a sort commended by a principle of wider imperfect duty, (b) is motivated by a sense of duty (or, perhaps, respect for moral reasons), (c) is neither forbidden nor required by another, more stringent duty . . ., (d) is in a context where no alternative is required by more stringent duty and there is at least one alternative that is neither forbidden by more stringent duty nor commended by other principles of wide duty, and (e) is done by an agent who has adopted the relevant principle of wider imperfect duty and has often and continually acted on that principle. (168–9)

Hill is claiming (1) if Kant has a place for supererogatory acts, it is here. He also holds (2) Kant indeed does have a place for them. This is clear from his opening paragraph, where he explains his “general aim” and his “special aim,” the latter being “to show . . . that Kant has a place in his moral scheme for supererogatory actions” (147).7 I have misgivings about (1) and will explain them in section 3, but my main concern is with (2). I disagree, assuming that Hill means that Kant has a place for supererogatory acts as such, i.e. viewed as supererogatory acts.8 I agree that Kant has a way to do justice to most of the phenomena that contemporary moral philosophers have tried to capture by adding the category of supererogatory acts to the categories of morally required acts, morally permissible acts, and morally impermissible acts. But as I see it, Kant’s way of doing justice to the phenomena is significantly different from their approach. The effort to fit into Kant’s ethics the category of supererogatory acts is, in my view, misguided. It is unnecessary and has a slightly distorting effect. Is my assumption that Hill holds that Kant has a place for supererogatory-acts-assuch correct? Or is his view simply that Kant’s ethics does justice to the phenomena that are often so classified, not also that he does justice to them by classifying them as supererogatory acts? It is clear that his aim in KIDS is to show the former, i.e. that Kant has 6   I am uneasy about the second disjunct in this explanation of why promoting the happiness of others will in some circumstances be obligatory. First, it is hard to wrap one’s mind around the idea of someone having a “last opportunity” to help others. Opportunities to help others abound (though tendencies to, inter alia, shield our gaze so as not to see need or to see it but not see a way we might be able to help, and to treat something as an occasion for helping only if it is right there before us, may dim our awareness of the opportunities). So unless he is on his deathbed, or about to be executed (and has a last chance, say, to offer some kind words that will comfort his family), it is hard to imagine how it could be, and how the agent could know it to be, a last opportunity to help others. But second, and more important, if after continually neglecting to help others he now does help someone, and does so in the recognition that he had better seize this “last opportunity,” would his action “[fulfill] a principle of wider imperfect duty”? Only if he has in effect just had a conversion experience, and so helps now not because it is his last opportunity but because he sees the moral importance of others’ happiness. An action can fulfill a principle of imperfect duty only if the agent has adopted the relevant obligatory end. That he had not done so is evident from the fact that up to now he continually neglected to help others. 7   Later in his paper he is more circumspect; see my section 6. But given the aims stated in his opening paragraph, I take it that he does aim to show that Kant’s ethical scheme has a place for supererogatory acts. 8   I emphasize this because in his “Meeting Needs and Doing Favors,” Hill says that I deny that “there is a significant place in Kant’s ethics for supererogatory acts” (203). While I deny that there is a (significant) place in Kant’s ethics for supererogatory acts viewed as such, I do not deny that there is one for what are typically referred to as supererogatory acts.

218  Marcia Baron a place for supererogatory acts as such. That this is his aim is evident in his discussion of possible sources of dissatisfaction with Kant’s position. The first source of dissatisfaction concerns (b). The worry is that Kant’s position, as Hill has presented it, implies that “the only acts that are supererogatory (‘beyond duty’) for Kant are acts motivated by a sense of duty” (171–2). Hill is sympathetic to this concern, and I take his sympathetic reaction to reflect a hope to show that Kant’s treatment of acts generally thought of as supererogatory is in accord with the approach of contemporary (or recent) ethical theorists who regard the category of supererogatory acts as of great importance to ethical theory (and perhaps also to informal, everyday thought about moral matters). To be sure, those who so regard the category of the supererogatory—I’ll refer to them as “supererogationists”—often differ as to what motivation is required for an action to count as supererogatory.9 One might therefore question my speaking of “the” approach, as if supererogationists all speak in one voice. Differences there are. But none of them hold that to be supererogatory, an action must be done from duty. Hill knows this, and is concerned to show that Kant’s treatment is not (at least not significantly) out of synch with their conceptions of supererogatory actions. He wants to show that Kant has a place for supererogatory acts as such, understood in a way that is not at odds with the approach of contemporary and recent ethical theorists who emphasize the importance of the category of the supererogatory to ethical theory. To address the concern about (b), Hill proposes a modification of Kant’s view, designed to show that Kant has a place for supererogatory acts viewed as supererogatory, not only for those acts that go beyond duty but which the agent mistakenly thinks are morally required.10 Hill writes that “it is not an unreasonable extension of Kant’s position to say that what is required for moral worth is not a motive to do one’s duty but a motive to do what is demanded or encouraged by moral considerations” (172, italics added). With the extension, Hill can hold that acts the agent sees to be encouraged by moral considerations and opts to do for that reason11 can have moral worth, on Kant’s view, and if they meet the other criteria listed above (a, c, d, and e) can count as supererogatory. 9   For discussion of the differences, see my Kantian Ethics Almost without Apology (Ithaca, NY: Cornell University Press, 1995), chs. 1–2. 10   One might claim that this would not be a problem since, as is often remarked, many who perform heroic deeds say, when asked why they took so huge a risk, something along the lines of “I was only doing my duty, only doing what any decent person in my circumstances would do.” (See inter alia Nechama Tec, When Light Pierced the Darkness: Christian Rescue of Jews in Nazi-Occupied Poland (New York: Oxford University Press, 1986). For discussions of this phenomenon, see Lawrence A. Blum, Moral Perception and Particularity (Cambridge: Cambridge University Press, 1994) and my “Imperfect Duties and Supererogatory Acts,” Annual Review of Law and Ethics, 6 (1998), 57–71, esp. 65–8.) Those in favor of utilizing the classification “supererogatory acts” will (if they do not call into question the sincerity of such claims or suggest that heroic acts motivated by a sense of duty should not count as supererogatory at all) generally not want to limit supererogatory acts to these, however. Many will want also to count as supererogatory those acts which the actor sees as simply “a nice thing to do;” some might find this excessively liberal but would count those which the actor sees as “a nice thing to do, morally speaking.” 11   This requirement for supererogatory acts would not sit well with some supererogationists. But it is at least closer to an account that most would find acceptable than is a view according to which an act can count as supererogatory only if regarded by the agent as morally required (and done for that reason).

The Supererogatory and Kant’s Imperfect Duties  219 That Hill is concerned to show not merely that Kant has some way to do justice to the phenomena but that he has a way to recognize supererogatory acts as such, i.e. as beyond duty, is also evident in his mention of a second source of dissatisfaction. The second source of dissatisfaction is that locating supererogatory acts within Kant’s scheme of imperfect duties is in tension with regarding them as beyond duty. “How, one may wonder, can I be going ‘beyond duty’ if I am guided by a principle of duty?” (172). The fact that Hill raises and takes seriously this question leaves no doubt that he wants to show that Kant recognizes, or can recognize, supererogatory acts as supererogatory. Otherwise he could have dismissed the question by replying, “In fact it isn’t possible. But that isn’t a problem. For Kant can do justice to the phenomena often classified as supererogatory without regarding them as ‘beyond duty’.”

3 As noted above, my main concern is with Hill’s claim that Kant has a place in his ethical theory for supererogatory acts (with (2), rather than with (1)). But in this section I’ll focus on (1), that is, on Hill’s claim that the best candidate for a supererogatory act would be an act with the features (a)–(e) mentioned above. The feature that will concern me is (b).12 I question his claim that if Kant can recognize supererogatory acts, the acts he would recognize as supererogatory would have, in addition to (a), (c), (d), and (e), the feature that they are “motivated by a sense of duty (or, perhaps, respect for moral reasons)” (168). The claim is plausible if we equate merit with moral worth. But if merit and moral worth are not identical, there is no reason to expect that the best candidate for a supererogatory act would have the feature that it was done from duty.13 That Hill treats merit and moral worth as identical is evident in KIDS, especially in sect. III. Referring to a passage from Metaphysics of Morals 6: 227, he writes: “This tells us that an action is meritorious (of positive moral worth) if it ‘does more in the way of duty’ than juridical duty” (163).14 Also noteworthy is that although the passage to which he refers makes no mention of moral worth, he introduces it with the following sentence: “The pattern by which moral worth is attributed to actions is suggested by 12   In discussion in a mini-seminar I taught with Allen Wood at Stanford University (June 2012), doubt was cast on (e) as well, the thought being that as long as one has adopted the relevant principle of wider imperfect duty, the same sort of action that would arguably qualify as supererogatory if done after one has acted repeatedly on that principle should also count as supererogatory if it is one of the first things one does after adopting the principle (supposing that we can peg the adoption of a principle to a particular time). I will not take this up here. 13   In thinking about the relation between moral worth and merit, I’ve been helped by Allen Wood, “Moral Worth, Moral Merit, and Acting from Duty” in his The Free Development of Each:  Studies on Freedom, Right, and Ethics in Classical German Philosophy (Oxford: Oxford University Press, 2014), and by Robert N. Johnson, “Kant’s Conception of Merit,” Pacific Philosophical Quarterly, 77 (1996), 310–334. 14   Hereafter this work will be cited as ‘MS’. The first page number refers to the volume of the Akademie edition (Kants gesammelte Schriften, herausgegeben von der Deutschen Akademie der Wissenschaften, Berlin: Walter de Gruyter and predecessors, 1902), and the second page number refers to the page number within that volume. Quotations from MS are taken from Immanuel Kant, Practical Philosophy, trans. and ed. Mary

220  Marcia Baron the following passage.” Here is the passage in question: “If someone does more in the way of duty than the law can compel him to do, his action is meritorious (meritum). If he does only the exact thing required by the law, he does what is due (debitum). Finally, if he does less than the law requires, this is moral guilt (demeritum)” (MS 6: 227).15 I have hinted that moral worth and merit are not identical, but how certain is it that they are not? In fact there is some textual support for equating them; see MS 6: 390.16 On the whole, the evidence is against it, but there is room for debate. This much, however, is clear, and is all that matters for my purposes: if there is any sense of “moral worth” in which “moral worth” and “merit” can be treated as interchangeable, it is not the one used in the Groundwork. I will not try to sort out whether in the Metaphysics of Morals Kant has a different notion of moral worth, according to which an action’s having moral worth is equivalent to its being meritorious, for my concern here is only to show that the notion of the moral worth of actions, as explicated in the Groundwork, is not the same as that of merit, as explicated in the Metaphysics of Morals. That they are different is evident from the following: An action can have moral worth only by virtue of being done from duty, but can be meritorious on other grounds, among them the subjective difficulty of the action. See for example MS 6: 228: “The greater the obstacles (of sensibility) and the less the moral obstacle (of duty), so much the more merit has to be accounted for a good deed, as when, for example, at considerable self-sacrifice, I rescue a complete stranger from great distress.” That moral worth (moralische Wert) as understood in Groundwork and merit (Verdienst) as understood in The Metaphysics of Morals are different is relevant to (b) for the following reason. Unless moral worth and merit are equivalent, there is no reason to think that actions that stand the best chance of counting as supererogatory in Kant’s ethics would be done from duty. Therefore (b) should be dropped. Modifying Hill’s characterization of the best candidate for a supererogatory act helps Hill’s case, for it removes the obstacle explained above in section 2 (as the “first source of dissatisfaction). The problem generated by (b) does not arise if (b) is dropped. His account can thus be somewhat closer to the standard supererogationist conceptions of supererogatory acts.

4 I drew attention to the fact that merit and moral worth are not the same in order to question (b) and suggest that it be dropped from the list of features we should expect an act J. Gregor (Cambridge: Cambridge University Press, 1996). Exceptions are quotations from Kant that Hill quotes; see n. 15. Apart from these exceptions, all quotations from Kant are from the Cambridge Edition of the Works of Immanuel Kant. 15   Hill is quoting from Mary Gregor’s 1964 translation. Kant, The Doctrine of Virtue, trans. Mary J. Gregor (Philadelphia: Pennsylvania University Press, 1964). 16   “Imperfect duties alone are . . . duties of virtue. Fulfillment of them is merit (meritum) = +a; but failure to fulfill them is not itself culpability (demeritum) = –a) but rather mere deficiency in moral worth = 0, unless the subject should make it his principle not to comply with such duties.”

The Supererogatory and Kant’s Imperfect Duties  221 to have if it is going to count for Kant as supererogatory. But the passages in MS where Kant speaks of meritorious actions call for discussion for another reason as well: on first reading, they may seem to lend strong support to the claim that Kant can—in fact, does—recognize supererogatory actions. Supererogatory actions, it might appear, are what Kant refers to as “meritorious actions.” (To clarify: this is not something Hill says. But it is something readers might expect, and for that reason needs to be addressed.) A passage quoted above initially seems to support understanding what Kant calls ‘meritorious actions’ as supererogatory actions: “If someone does more in the way of duty than the law can compel him to do, his action is meritorious . . .” (MS 6: 227). (Or, to use the translation in the 1996 Cambridge edition: “If someone does more in the way of duty than he can be constrained by law to do, what he does is meritorious . . .”) More in the way of ethical duty? Or more in the way of juridical duty? And if more by way of ethical duty, wide or narrow ethical duty? Only if it is more in the way of wide ethical duty than the agent can be constrained by law to do would it make sense to understand meritorious actions to be supererogatory actions. A survey of passages where Kant uses the word “meritorious” (verdienstlich) reveals that he uses “meritorious” to refer to actions not required by juridical duty, and sometimes more narrowly to refer to actions required by wide but not by narrow ethical duty. At 390–1 he writes: Although there is nothing meritorious in the conformity of one’s actions with right (in being an honest human being), the conformity with right of one’s maxims of such actions, as duties, that is, respect for right, is meritorious. For one thereby makes the right of humanity, or also the right of human beings, one’s end and in so doing widens one’s concept of duty beyond the concept of what is due (officium debiti), since another can indeed by his right require of me actions in accordance with the law, but not that the law be also my incentive to such actions.17

Here Kant uses “meritorious” to indicate that the action is not something that another can demand of one as his or her right. It would of course be a mistake to suggest that Kant holds that everything meritorious, so understood, is supererogatory, or even that he holds that everything required by wide but not by narrow obligation is supererogatory. Hill is well aware that this would be a mistake. Asking where in Kant’s scheme we should look for supererogatory acts, he writes, “Clearly they cannot be identified with acts that fulfill imperfect duties, for we have seen that duties of respect for others are imperfect” (168). Only if “meritorious” is understood more broadly than Kant uses the term do the passages containing the word “meritorious” lend support to the view that there is a place in Kant’s scheme for supererogatory acts. Hill notes this in “Meeting Needs and Doing Favors.” Correcting his earlier reading, he observes: Kant appears to be concerned in this context with what can be legally demanded and enforced rather than, as I earlier thought, with moral law. If so, by saying that someone’s meritorious act is “more in the way of duty than he can by law be constrained to do” Kant does not affirm or deny 17

  See also the rest of the same paragraph.

222  Marcia Baron that extraordinary efforts to make others happy can be “more in the way of duty than moral law demands.”18

So Kant’s use of “meritorious” (verdienstlich) lends far less support than one might initially think to the view that he recognizes supererogatory actions. Is it nonetheless in the spirit of Kant’s discussion to recognize supererogatory acts? Hill thinks so. Right after the sentence just quoted, he writes: “Such efforts, however, are a kind of thing that having an obligatory end requires of us (namely, working towards others’ happiness), and some of these efforts in particular contexts are more of that kind of thing than the imperfect duty of beneficence requires because one could have been sincerely committed to the end without doing quite so much.”19 In sections 7 and 8 I’ll speak to the question of whether, from a Kantian standpoint, it is valuable to think in terms of just how much is required by sincere commitment to the end, and whether it is apt to distinguish the efforts that go beyond what is required from those that are only what is required. My suggestion will be that only against the background of assumptions that are no part of Kant’s ethics would it be apt to do so. For now, my point is that although the distinction between merit and moral worth is helpful to the position that Kant can recognize supererogatory acts as such, it is helpful only in that it removes one obstacle: it is not necessary to suppose that because supererogatory acts would count as meritorious, they would have to have moral worth and therefore would have to be done from duty. So that obstacle is removed. But Kant’s use of the epithet “meritorious” does not show—indeed lends virtually no support to the claim—that he recognizes supererogatory acts.20

5 As I explained in section 2, Hill argues both (1) if Kant has a place for supererogatory acts, it will be here (see above), and (2) he does have a place for supererogatory acts.21   Hill, “Meeting Needs and Doing Favors,” 215–16, n. 38.   Hill, “Meeting Needs and Doing Favors,” n. 38. 20   There is, however, textual evidence in Kant’s lectures that he was willing, at least at the time of those particular lectures, to classify some actions as supererogatory. In lectures transcribed in 1798 by Johann Friedrich Vigilantius, Kant reportedly said, in a discussion of actions “worthy of honour,” “Towards men we realize more morality than is incumbent on us, and in this, therefore, lies at the same time the merit of our actions. This honestas can be attained by practice, no less in conduct towards our own humanity than in serving the ends of other men; for example, when the manual worker cultivates his talents beyond the call of duty, so as thereby to become more useful to others, and is guided by the dutiful conviction that he is bound to cultivate his talents to the best of his ability” (27: 668). Two pages later, in a discussion of duties of parents to their children, a distinction is drawn between the “coercive duty to nourish their children,” a “mere debitum,” and “the rearing of children up to the point of self-sufficiency, i.e. an education so ordered that the children thereby obtain contentment with their lot, and pleasure in their existence,” which is classified as an “opus supererogationis” (27: 670). The latter passage is less noteworthy for my purposes since the contrast is not between an ethical duty and a supererogatory act but a juridical duty and a supererogatory act, but the first passage merits more attention than I am able to give it in this chapter. Thanks to Mark Timmons for drawing my attention to these passages. 21   It should be borne in mind that I take issue with (2) only if is understood to mean “. . . supererogatory acts viewed as such” and that as I argued in section 2, it is clear that Hill is concerned to support the view that Kant has a place for supererogatory acts viewed as such—not only the view that Kant has a place for what are classified as supererogatory acts. 18

19

The Supererogatory and Kant’s Imperfect Duties  223 In section 3, I called (1) into question, taking issue with one feature Hill said that supererogatory acts, in Kant’s scheme, would have; in section 4, I noted that Kant’s use of the word “meritorious” might seem to lend support to (2), but in fact does not. The rest of my chapter focuses on (2). One reason for misgivings about (2) was articulated by Hill in KIDS. “How, one may wonder, can I be going ‘beyond duty’ if I am guided by a principle of duty?” (172). More generally, locating supererogatory acts within Kant’s scheme of imperfect duties is in tension with regarding them as beyond duty. Whereas the first source of dissatisfaction that he mentioned, the one concerning feature (b), is easily addressed because (b) should be dropped, this problem is much more serious. Supererogatory acts, whatever else they are, are supposed to be acts that go beyond duty. Locating them within Kant’s category of imperfect duties, even in the widest of imperfect duties, is at odds with this crucial feature. Hill responds to the concern as follows: Kant’s terminology of “imperfect” and “perfect” duty does confuse the issues. It is as if Kant started to work out a moral theory on the model of legal-like strict duties, and then, discovering that there is more to morality than duty, still retained the old labels for types of duty rather than spoil the symmetry of his theory by changing to more natural expressions. For example, what Kant is concerned to say about beneficence is (i) it is a duty to adopt a maxim of beneficence, and therefore (ii) it is a duty to promote the happiness of others sometimes, but also (iii) when one has satisfied these minimum and rather indefinite requirements, one may promote their happiness or not, as one pleases, but to do so with the proper motive will always be of positive moral worth. Kant tried to say all of this with his restrictive terminology of duty when it could be put more simply by making an early distinction between what is obligatory and what is merely good to do. (172)

Hill recognizes that there is a problem, but he believes it is due to restrictive terminology, terminology which he thinks does not reflect Kant’s real views. The problem is thus superficial. Kant was trying to say something which his terminology of duty made it difficult to say. I disagree. Kant’s terminology is not inapt; he did not mean to distinguish between what is obligatory and what is merely good to do. Nor would he have done better to do so. I also disagree with Hill’s claim above regarding “what Kant is concerned to say about beneficence.” I doubt that Kant would endorse (iii) and definitely do not think that he is concerned to say (iii). He certainly is concerned to affirm (i); I do not think he is concerned to affirm (ii). He would probably assent to (ii), but it is not something he is concerned to show. His focus is less action oriented than (ii) suggests.

6 Before elaborating on this disagreement with Hill, I  want to note briefly, without resolving the matter, the possibility that his more recent work on beneficence and

224  Marcia Baron Kant’s imperfect duties reflects a change in his position concerning (2)—his position that Kant does indeed have a place for supererogatory acts, viewed as such. Hill’s recent work on the topic is “Meeting Needs and Doing Favors” in his Human Welfare and Moral Worth. It is less clear in this work than it was in Dignity and Practical Reason that Hill is concerned to show not only that Kant has some way to do justice to the phenomena commonly classified as supererogatory actions but also that his scheme of imperfect duties gives him a place for supererogatory actions understood as such. Even in KIDS there are sentences that leave some room for doubt as to whether Hill takes the stand I explained in section 2. At one point in KIDS he takes a more circumspect tone, saying not that Kant does have room for supererogatory acts but only that it “is obvious . . . when we compare Kant’s position with what recent philosophers call ‘supererogatory’ that it is far less misleading to grant that Kant allows for supererogatory acts than to deny it” (169). In “Meeting Needs,” Hill summarizes his position in KIDS in a way that picks up on the cautious tone reflected in the sentence just quoted from KIDS: “My initial account of Kant’s moral categories proposed that, although Kant does not acknowledge a category of supererogation, his scheme for classifying actions leaves room for meritorious acts that have the main features of what some philosophers at the time were calling ‘supererogatory’.”22 “Main features” allows for the possibility that the acts lack some of the features of what some philosophers at the time were calling “supererogatory;” “some” in front of “philosophers” adds a further qualification. Although Hill does not retract anything he said in KIDS, apart from correcting his reading of MS 6: 390–1,23 what he draws on from the earlier article does not include some of the parts that I have highlighted in this chapter. In particular, he does not suggest in “Meeting Needs and Doing Favors” that Kant would have done better to distinguish between “what is obligatory and what is merely good to do” rather than adhere to his “restrictive terminology of duty” (74, quoted more fully above). My main disagreement with Tom concerns precisely this point: I see the terminology of duty to work fine, the only difficulty being that Kant’s distinctions are not as tidy and crisp as one might wish.24 I do not think Kant would have done better to distinguish between what is merely good to do and what is obligatory. In section 7, I take a closer look at the distinction and consider what there is to be said for it. It might seem

22  Thomas E.  Hill, Jr., “Meeting Needs and Doing Favors,” in Human Welfare and Human Worth (Oxford: Oxford University Press, 2002), 234. See also 237, where instead of saying that Kant has room for supererogatory acts he speaks in terms of an analogue: “In Kant’s scheme the closest analogue to the acts that others label supererogatory . . .” 23   See text accompanying n. 18, above. 24   But although some of the untidiness might be rectifiable, to a considerable extent the fuzziness of the boundaries and the difficulty in classifying a particular duty under one label or another simply reflect the way the world is.

The Supererogatory and Kant’s Imperfect Duties  225 obvious that Kant’s ethics would be enhanced by utilization of the distinction. But I will suggest that the value of the distinction hinges on certain assumptions made by many ethicists, but not by Kant, and that without those assumptions, the distinction is not particularly helpful. I want to emphasize that I am not suggesting that Hill accepts any of these assumptions nor that he thinks Kant does;25 I suspect that he simply did not question the standard view that an ethical theory needs a category of supererogatory acts.

7 Against the background of what assumptions is it important to draw the distinction between what is morally obligatory and what is, morally speaking, merely good to do? If one holds that, as Urmson puts it, it is “part of the notion of a duty that we have a right to demand compliance from others,”26 it makes perfect sense to regard the distinction as very important. After all, we need to know when others have the right to demand x of us; we also want to be clear on when we have a right to demand certain conduct of others. Relatedly, there needs to be a limit on what can legitimately be demanded of us by others (and of them by us), and putting a “cap” on duty provides that limit. In an ethical theory that assumes that duties—all duties—entail a corresponding right to demand compliance, the distinction makes a great deal of sense, in fact might well be essential. Kant, of course, rejects that assumption, holding that only some duties entail such a right.27 One might favor a narrow notion of duty, according to which duties do entail a right to demand compliance from others, but it will not do to say that because every duty entails a corresponding right to demand compliance, Kant is worse off for not distinguishing between what is morally obligatory and what is merely good to do.28 25   I emphasize this in part because I can see that in Kantian Ethics Almost without Apology I failed to indicate as clearly as I should have when I was taking issue with the specifics of Hill’s interpretation of Kant, and when I was questioning the assumption that every ethical theory needs to have room to recognize supererogatory acts and trying to unearth some of the background assumptions motivating that assumption. In his “Meeting Needs and Doing Favors” Hill sometimes takes me to be challenging his claims at points in my discussion when in fact I was not doing so. 26   Urmson claims, for instance, that it is “part of the notion of a duty that we have a right to demand compliance from others” (“Saints and Heroes,” 213). There is, however, some tension in his view, as I note in my Kantian Ethics Almost without Apology, 65 n. 11. 27   For more on the differences between Kant’s conception of duty and those of Urmson and other supererogationists, see Kantian Ethics Almost without Apology, chs. 1–2. 28   Another assumption that motivates the distinction is that if x is morally obligatory, it is appropriate, in the absence of excusing conditions, to blame anyone who fails to do x. (Exactly for whom it is appropriate is not always clarified, but the idea seems to be that it is generally appropriate for bystanders to do so.) There is no notion in Kantian ethics that blame is generally appropriate, and I think it best not to understand the concepts of duty, moral rightness, or obligatoriness as entailing the appropriateness, or even the overridable presumption of appropriateness, of blaming another for failure. Whether it is appropriate to blame another hinges not only on matters of responsibility but also on what the wrong in question was. A failure to treat others with respect warrants blame more than does a failure to seek to develop one’s talents.

226  Marcia Baron Likewise, if one thinks of morality on the model of a code, as Urmson does in “Saints and Heroes,”29 it makes sense to suppose that we need to know what is obligatory and what is merely good to do. Just as we need to know whether we are required by law to (say) shovel the snow on the sidewalk in front of our home or whether it is merely good to do so, it is important, if morality is understood as a code, to know just what the rules are. If we reject the supposition that duties entail a right to demand compliance, and recognize a distinction between juridical and ethical duties and between wide (or imperfect) duties and narrow (or perfect) duties, the importance to moral philosophy of the distinction between what is obligatory and what is merely good to do diminishes. There is room for debate, however, as to whether it diminishes enough that Kant can do perfectly well without the distinction. Those who think that Kant would be much better off with the distinction might press the point as follows: Even if the assumptions noted above are rejected, it still is important to know (a) what I really must do (and although a little less important, it is helpful to have a sense of what others are required to do), and (b) what it is permissible to exact from each other. The juridical duties cover (b); but what about (a)? The challenge has some merit. Kant’s ethics offers a little less by way of an answer to (a) than does a contemporary approach such as Urmson’s that tries to distinguish as sharply as possible between what is morally obligatory and what is good to do but not morally required. The juridical duties tell me only what duties others have a right to compel me to do, not what I really must do (beyond what others may compel me to do). The perfect duties provide much of the answer to (a) but not the whole story; the imperfect duties after all are real duties, too. So if I want to know what I really must do by way of helping others (and improving myself) what does Kant’s ethics tell me? (I focus in my answer on helping others, since that is Hill’s focus, and a general answer that encompasses all imperfect duties will not do since the latitude attached to the imperfect duties varies depending on the imperfect duty in question.30) Hill offers a friendly amendment to Kant’s ethics which if accepted provides a partial answer. His suggestion is that as long as one has adopted the maxim of beneficence and promoted the happiness of others sometimes, one may promote others’ happiness or not, as one pleases, provided that the action in question is not required by perfect duty. But this provides a more definite answer than is warranted. (The answer is misleading, too: that I have done a lot in the past in no way frees me from a duty to continue to help

29   Without commenting on his use of the word “code,” much less defending a conception of morality as a code, Urmson argues that “our code should distinguish between basic rules, summarily set forth in simple rules and binding on all, and the higher flights of morality of which saintliness and heroism are outstanding examples” (“Saints and Heroes,” 211). 30   See Hill, “Kant on Imperfect Duty and Supererogation,” as well as Baron, Kantian Ethics Almost without Apology, ch. 3, and Marcia Baron and Melissa Fahmy, “Beneficence and Other Duties of Love in The Metaphysics of Morals,” in The Blackwell Guide to Kant’s Ethics, ed. Thomas E.  Hill, Jr. (Malden, MA: Wiley-Blackwell, 2009), 211–28.

The Supererogatory and Kant’s Imperfect Duties  227 others. It is different from, say, a duty to do my share in my household to keep the house neat. See below.) The answer should be, “Kant’s ethics doesn’t tell you what you really must do with respect to the imperfect duties. In fact, there is no determinate answer to that question. You can certainly go wrong here, but we can’t spell out for you how to avoid that. If you really have embraced the obligatory end of others’ happiness—have adopted the maxim of beneficence—you will do okay, assuming that in addition you don’t have vices that keep you from seeing what you should see, using good judgment, and doing what you see you should do.” We might add here that one is less likely to have such vices, at least to have them in anything but a very mild form, if one takes seriously (and has taken seriously throughout one’s adult life) duties to cultivate one’s character.31 Striving to be better, including striving for self-knowledge,32 reduces the risk that one’s conduct will be seriously marred by complacency and rationalizations. The challenge is best met by questioning whether it is better for a moral theory to tell us more, rather than less, about just what it is that we are supposed to do, and raising a concern about efforts to infer from what Kant says about the duty to adopt the end of others’ happiness claims about how much, or how often, we should act accordingly. Efforts to translate what Kant says about obligatory ends into statements to the effect that (as Hill puts it) “it is a duty to promote the happiness of others sometimes” very subtly mislead us. Naturally, if we are going to try to infer such statements, this seems to be a safe one; who could deny that we have a duty to do that? The problem is that the very choice of the word “sometimes” suggests that we do not have a duty to do it very often. One could of course quickly add “but whether a great deal more than that is required is not clear” but Hill does not. He adds that “when one has satisfied these minimum and rather indefinite requirements, one may promote their happiness or not, as one pleases . . .” But it is not just as one pleases; it is a matter of judgment. And thinking in terms of satisfying minimum requirements also is not quite right (though the addition of “indefinite” does help); the idea of “minimum requirements” is not quite at home in Kant’s ethics. I see it to be an advantage of Kant’s approach that it does not invite us to reason, “I’ve done a lot, so although others in my situation would have a duty to help, I don’t.” More generally, it does not invite us to think of morality on the model of a code of rules, or on an institutional model. Within institutions and clubs, it is indeed useful to have a sense of how much one should do in order to “do one’s part” and thus at what point one is going “beyond duty.” This is because one can be, in effect, swallowed up by the numerous committees on which one could serve, the events one could help sponsor, the ceremonies and celebrations and receptions and meetings and protests 31   On the cultivation of character, see my “Kantian Moral Maturity and the Cultivation of Character,” in Harvey Siegel, ed., The Oxford Handbook of Philosophy of Education (Oxford: Oxford University Press, 2009), 227–44. 32   The “First Command of All Duties to Oneself,” Kant writes, is “know (scrutinize, fathom) yourself” (MS 6: 441; all capitalization and emphasis taken from the 1996 Gregor trans.; the part with the capitalization is a heading).

228  Marcia Baron one could attend. Unless we hold—as Kant of course does not33—that morality too threatens to swallow one up if one takes it too seriously or doesn’t draw a line and say “No more,” the distinction between what one must do and what it is good to do is not helpful for morality the way it is in the context of institutions and clubs, and the (apparent) demands of one’s job. Treating the distinction as important for morality risks encouraging us to view morality as something that needs to be prevented from swallowing us up. Duties to help others are not just duties to do one’s share. In this way they differ from (many) institutional duties or duties we have as members of certain kinds of clubs; they also differ from duties we have as members of a household to do our share to keep the house tidy. If others do not do their share, then, unless there are vulnerable people in the house—e.g. someone with limited vision and poor balance who is endangered if items are left here and there on the floor—it is perfectly appropriate for me to decide that given how much I have been doing to keep the house tidy, I am now going to do a lot less. Duties to help those in need are different. If someone is in need, the fact that I have been doing a great deal to help others in the past (much more than others have) does not entail that I am now free to help or not, as I choose. To be sure, many factors enter into whether it is compatible with my having adopted the end of others’ happiness that I now do not take up (or seek out) an opportunity to help someone. It is not that my past activities have no bearing at all. But it is hopeless to try to enumerate the factors and say when I am free to opt not to help. And with the possible exception of such factors as “This person asking for my help tried to murder my sister,” no one factor suffices to settle that I am in this situation free to help or not, as I please. I have been suggesting that the assumptions that motivate the distinction between the morally obligatory and the merely good-to-do are not part of Kant’s ethics, and that Kant would not be better off limiting the scope of duty so as to leave room for the distinction. What Hill calls Kant’s “restrictive terminology of duty” is not a liability.

8 I said at the outset that Hill and I agree that Kant can mostly do justice to the phenomena that there are saints and heroes, and saintly and heroic acts, but that I think

33   It is sometimes suggested that the “fantastic virtue” passage shows that Kant thinks it possible to take morality too seriously. Kant writes that “the human being can be called fantastically virtuous who allows nothing to be morally indifferent (adiaphora) and strews all his steps with duties, as with mantraps; it is not indifferent to him whether I eat meat or fish, drink beer or wine, supposing that both agree with me. Fantastic virtue is a concern with petty details which, were it admitted into the doctrine of virtue, would turn the government of virtue into tyranny” (MS 6: 409; italics as in Gregor edition). There is nothing in the passage to suggest that one can take morality too seriously. Fantastic virtue does not involve taking too seriously what really is morally significant but rather taking something to be morally significant that is of no moral importance whatsoever.

The Supererogatory and Kant’s Imperfect Duties  229 Kant can do so without recognizing supererogatory actions. How? Part of the answer is already in view: saintly and heroic actions will typically be among those actions that fall under a principle of wide imperfect duty. They are not severally required, but they are not beyond duty. They are of a type—actions promoting others’ happiness—that are, as a class, morally required. That doesn’t take us very far, however. We want to be able to distinguish, among actions promoting others’ happiness, those that show remarkable generosity from those that show only a little. But we can, in the language I just used: we say that the action shows remarkable generosity. I don’t see that it helps—except in an institutional context—to say that it goes beyond duty. It does help if “duty” is understood in ways quite different from the way Kant understands it; if it is understood as foisted on us, as arising only from societal need, as exclusively other-directed, or as entailing corresponding rights to exact compliance. My claim is not that the category is never helpful, but that it is not helpful in the context of Kantian ethics. In general, much of what supererogationists want to capture by adding to our classifications of actions can be accomplished very well—I think better—by use of such terms as “generous,” “uncommonly thoughtful,” “unbelievably courageous,” and by assertions about the person’s character, as reflected by the action. So I believe Kantian ethics is not really limited by the fact that it cannot recognize the category of supererogatory actions. And to try to insert into Kant’s ethics among the wide imperfect duties a classification of supererogatory actions, even one that would (as Hill proposes) be understood contextually, both reflects and encourages an approach to life and to the place of morality in a life well lived that is at odds with Kant’s ethics. Morality is not to be thought of as a check list of items I need to do (even do regularly). Nor is it helpful to think in terms of overall volume; the question, “Have I done enough to fulfill my imperfect duties?” (or even, “Have I done enough today to fulfill my imperfect duties, for the time being?”) is not an apt one. We should indeed be engaging in moral self-scrutiny, but not in a way that involves separating out what we have to do from what it is good but not required. Earlier I said that Kant can mostly do justice to the phenomena for which a category of the supererogatory is thought to be necessary. Why the qualification? The main reason34 I added “mostly” is that there may be—some people think there obviously are—actions that are “out of character” and which nonetheless deserve 34   Another reason for the qualification is that one may think that we need the category of the supererogatory as part of a practice of singling out the great heroes, doing so by reference to their heroic actions. As I argued in Kantian Ethics Almost without Apology, Kant’s ethics does not encourage or support a singling out of great heroes or saints. Recall Critique of Practical Reason 5: 155, where after suggesting that educators of young people would do well to “[search] through the biographies of ancient and modern times in order to have at hand instances for the duties presented” and thereby “activate their pupils’ appraisal in marking the lesser or greater moral import of such actions,” Kant writes: “But I do wish that educators would spare their pupils examples of so-called noble (supermeritorious) actions, with which our sentimental writings so abound, and would expose them all only to duty and to the worth that a human being can and must give himself in his own eyes by consciousness of not having transgressed it; for, whatever runs up into empty wishes and longings for inaccessible perfection produces mere heroes of romance who, while they pride

230  Marcia Baron praise if they go beyond duty. If by “out of character” we mean that the excellence of the action in no way reflects any excellence in the person, and if we think that it is important to praise the action not only for its results but in some other way, then we may well think it a shortcoming of Kant’s ethics that it does not leave room for saying that it goes beyond duty (beyond ethical duty, that is). Whether this is a problem depends on what is meant by “out of character.” On one reading of “out of character” the act in fact does reflect well on the person, at least as the person was in that time period; on another, I do not know how to make sense of the concept. Oskar Schindler’s widow—and not one of his greatest admirers—observed that it was only from 1939 to 1945 that he showed extraordinary heroism.35 Let’s assume that what she said was true. She may have claimed—I don’t recall if she used this terminology—that the actions he took to save the lives of some 1,200 Jews during the Holocaust were therefore out of character. If we count this as out of character, well, it is nonetheless the case that in those years his altruism was steady, and surely reflected some excellences of character (even if it also reflected qualities that are not excellences, but which facilitated his work, for example, a penchant for risk-taking and, as Thomas Keneally put it, a “capacity to inveigle and charm officials and businessmen”).36 So if “out of character” means only that it is not characteristic of the person as he was throughout his life, only of a period of five or six years, an action’s being out of character is compatible with its reflecting something significant about the person—the person at that time. And so there is no problem capturing what is morally remarkable about the action without referring to it as supererogatory; we can say the action was heroic, remarkably generous, etc. If, however, we can imagine an action that is, as some say, a complete fluke, e.g., an action of rescuing someone where the action was motivated by a genuine concern for the other person (rather than, say, bravado and a wish to be honored as a hero) but which in no way at all reflects anything about the agent, and if we think it important to be able to celebrate the action yet cannot celebrate it for what it shows about the agent, this may be a phenomenon to which Kantian ethics cannot do justice. I do not understand what it could be for an action to be a complete fluke. I understand being unable to explain why someone acted as he did, as it does not comport with our sense of who or what he is, but in those circumstances we should be moved to wonder if we really know the person, and have not misjudged him. If we aren’t—if it is clear he really is the complete jerk we thought he was, yet nonetheless acted very themselves on their feeling for extravagant greatness, release themselves in return from the observance of common and everyday obligation, which then seems to them insignificant and petty” (Critique of Practical Reason 5: 154–5; see also the footnote Kant appended to this text). One might of course disagree with Kant on this and claim that (a) a practice of singling out the great heroes is important, and (b) to this end the category of the supererogatory is important; but both (a) and (b) need to be argued for. Neither is at all obvious. 35   This is reported by Thomas Keneally, referring to an interview on German television in 1973. Keneally, Schindler’s List (New York: Simon and Schuster, 1982), 396–7. 36  Keneally, Schindler’s List, 396.

The Supererogatory and Kant’s Imperfect Duties  231 generously just now—I don’t know how to make sense of that. But if it does somehow make sense, I wonder how clear it is that we need to be able to celebrate the act in a way other than by celebrating the consequences (supposing that they were good) and also saying (if this makes sense) that this was a very generous act, though done by a very ungenerous person. I am not convinced that we need to be able to say more than that, and in particular, I am not convinced that it is important to be able to praise the act as an act that goes beyond duty.37

37   I am grateful to Allen Wood for extensive discussion of an earlier draft of this chapter and to Mark Timmons and an anonymous reviewer for written comments on an almost-final draft. In addition, I would like to thank discussants at the Stanford University mini-seminar of June 2012 and at the June 2012 University of St Andrews Philosophy Reflectorium for helpful comments on material I presented to them from this chapter.

12 Did Kant Hold that Rational Volition is Sub Ratione Boni? Andrews Reath

1 Introduction In an essay entitled “Personal Values and Setting Oneself Ends,” Tom Hill casts doubt on what he rightly characterizes as a “strong thesis” about rational volition that several commentators have attributed to Kant. The thesis is “that anyone who has values at all is thereby implicitly committed to objective moral values” or that “adopting ends and maxims carries commitment to objective moral values that could ground moral claims.”1 The specific version of this thesis on which Hill focuses concerns the value judgments that “we implicitly make when we set ourselves personal ends and adopt maxims about how to achieve them,” and claims that when one adopts a rationally optional end through practical reason, one takes one’s end to have “objective value, in a robust sense.”2 This claim appears to initiate an argument to the effect that commitment to moral principle is built into rational choice per se, and that agents who violate moral principle act irrationally in the narrow sense of acting contrary to principles or commitments that they actually accept.3 Both Christine Korsgaard and Stephen Engstrom have defended versions of this thesis about the value commitments presupposed in rational choice of ends (including the end of happiness)—both as interpretation of Kant and on philosophical grounds. In her early well-known essay “Kant’s Formula of Humanity,” Christine Korsgaard attributes to Kant the view “when we act we take ourselves to be acting reasonably and so we suppose that our end is, in his sense, objectively good,” that is, “provides reasons for action that apply to all rational beings.”4 I take it that the “objective goodness” that the agent supposes her end to have can be unpacked through the thoughts that 2   Hill (2002: 244, 262). See also Hill (2002: 251, 258, 263, 266–7).   Hill (2002: 246, 258).   On the narrow sense of irrationality (practical irrationality is acting contrary to one’s actual judgments about applicable principles or reasons), see Scanlon (1998: 25–30). 4   Korsgaard (1996a: 116, 115). 1

3

Did Kant Hold that Rational Volition is Sub Ratione Boni?  233 pursuit of the end is supported by reasoning that justifies it to any rational agent, that it is a good thing from anyone’s point of view that the agent achieve her end, and that others in certain relations to the agent have objective reasons to support her pursuit of it. As I understand it, the key idea (both to her reading of Kant and her own view) is that it is a necessary feature of rational choice of an end that the agent takes the end to be objectively good in this sense. This evaluative attitude is, as it were, a component of the self-consciousness of rational choice, part of the rational agent’s own understanding of what goes into rational choice of an end. And she assumes that from this feature of rational self-consciousness one can uncover a commitment to morality that is likewise implicit in all rational choice of an end. Korsgaard does not deny that for Kant personal ends are “relative” or “subjective” ends that are of value to agents because of our interests or their fit with our nature. She argues that adoption of a relative or subjective end by an agent who is an end in itself confers objective value on the end in the above sense. The objective goodness that we suppose our relative ends to have presupposes that we understand rational choice to have a value-conferring capacity, and that in turn presupposes a valuing of rational nature as an end-in-itself that is implicit in all rational choice of ends.5 The latter value commitment is, of course, fundamental to morality. In the Form of Practical Knowledge, Engstrom develops what he terms a “practical-cognitivist conception of the will,” one main element of which is that rational volition is based on practical judgments of goodness made under “the presupposition of universality.” The cognitive aspect of volition is a judgment about good that is taken to be sharable by any rational subject—that is, it takes itself to satisfy a condition of universal validity.6 This is a fact about the nature of rational volition—an essential feature of volition that is given to agents through practical self-consciousness, or the understanding of rational volition from the perspective of an agent engaged in it. Since human agents necessarily will own happiness as an end, we take our own 5   In this respect, in Korsgaard’s conception the valuing of the self and its rational capacities that is implicit in all choice is the framing assumption of practical reasoning and judgment, and is needed to back up the objective goodness that an agent takes her ends to have. In her later work, she continues both to ascribe this view to Kant and to defend it herself. Her main points are that we do take our ends to be valuable, that doing so supposes that rational choice is value conferring and that the capacity for rational choice has unconditional value. See e.g. Korsgaard (1996b: ch. 3, esp. p. 122 (§3.4.8)), and Korsgaard (2009: 122–3 (§6.3)). Her view in (1996b) is that valuing one’s human identity is a condition of having reasons for action (because practical reasons are based in practical identities and our practical identities get their normative grip on us through our valuing our human identity), and that valuing one’s human identity is implicit in all rational choice or finding an end to be valuable or worth pursuing. In (2009: 123) she writes: “As a Kantian, I believe that it is our own choices that ultimately confer value on objects, even though our choices are responsive to features of those objects. In choosing objects, in conferring value on things that answer to our nature in welcome ways, an agent is affirming her own value. She takes what matters to her to matter absolutely and so to be worthy of choice.” 6  Engstrom (2009:  124–7). Barbara Herman has argued for a similar conception of the will as a “norm-constituted power,” where the principle that defines the power (as I understand her view) is the moral law. See “The Will and its Objects” in Herman (2007). The conception of volition that I explore in this chapter is close to (and has drawn on) Herman’s and Engstrom’s views.

234  Andrews Reath happiness to be objectively good (ceteris paribus) and will it under the presupposition of universality. On Engstrom’s reading of Kant, the Formula of Universal Law is the internal norm of rational volition and we are committed to following it by the self-consciousness that is part of volition. Hill expresses skepticism both about this thesis and the argumentative use to which it has been put. He has the good sense to be wary of the idea that ordinary choice of subjective ends carries any commitment to their objective value—this thesis appears to overreach in the value commitments that it attributes to ordinary agents—and he does not think that there is clear textual support for saddling Kant with such a view. He allows that for Kant choice of an end carries some limited value judgments—e.g. that an action advancing such an end is good as a means, or good insofar as it contributes to one’s happiness and in that sense good for an agent. But he thinks that Kant’s attitude towards subjective ends is closer to Hobbes than many recent commentators believe, and argues that “[n]‌one of these commitments in adopting ends and maxims attributes any impersonal, nonrelative, or moral goodness to our personal ends as individuals.” Furthermore, it seems obvious that agents often willingly and knowingly (intentionally) make morally unjustifiable choices. All told Hill thinks that it is a “very thin reed” on which to base an argument that it is irrational to act contrary to moral principle.7 The strong thesis about the value commitments implicit in rational choice of an end readily extends to broader issues about Kant’s understanding of the nature of rational volition—in particular did he think that all rational volition proceeds sub ratione boni, and if so in what sense? In this chapter I would like to explore what can be said on behalf of ascribing to Kant a conception of rational volition that I would state as follows: that rational volition constitutively understands itself to satisfy a condition of universal validity. That is, rational volition is based on practical reasoning aimed at judgments of goodness that make a tacit claim to universality. As one might say, it is part of the self-consciousness of rational volition that it proceeds on maxims that are taken to satisfy a condition of universal validity. This amounts to the admittedly controversial claim that all rational volition is tacitly guided by something like the Universal Law version of the categorical imperative as its formal or internal constitutive norm. I shall refer to this conception as the “rationalist thesis” about the will. It is a claim about the nature of rational volition (not the moral psychology of choice), and since volition is constituted by its own self-understanding, a claim about the nature of volition as given in a rational agent’s self-understanding of what it is to exercise the will. It does bear on the rational authority of morality, since the authority of morality would follow if the fundamental principle of morality is indeed the internal principle of rational volition, the principle that, as it were, we impose on ourselves through our understanding of what it is to exercise the will. However I am more interested in exploring the contours of the rationalist thesis than in using it to argue that bad conduct is contrary to

7

  Hill (2002: 261, 267).

Did Kant Hold that Rational Volition is Sub Ratione Boni?  235 principles and commitments that an agent actually accepts and is therefore irrational in a narrow sense. Not having left all of Tom Hill’s good sense behind, I have some misgivings about ascribing the rationalist thesis to Kant, and I don’t claim to have resolved the interpretive issue in this chapter. While I believe that there is strong support for this reading, there are also clearly passages where Kant seems to allow that we can freely and intentionally choose actions and ends that we see are not universally justifiable. The texts are not decisive. My aim will be to see what support for the rationalist thesis as a reading of Kant can be drawn from the texts and to explore what it would do for his moral conception. In section 2, I canvas various discussions of the will for possible textual support, and in section 3 I argue that Kant’s well-known theses about autonomy and free agency provide stronger philosophical support for this thesis. In section 4, I’ll consider how this thesis can accommodate certain forms of bad willing and will address some counter-intuitive implications that it might be thought to have.

2  Does Volition Understand Itself to Satisfy a Condition of Universal Validity?—Initial Textual Support As I’ve indicated, the rationalist thesis about the will that I wish to explore is the idea that rational volition constitutively understands itself as part of its self-consciousness to satisfy a condition of universal validity. If so, all rational volition, including morally bad choice, is tacitly guided by the Universal Law version of the categorical imperative in some form. (As I will sometimes put it: the Formula of Universal Law is the formal principle of rational volition.8) What might this conception of volition involve and is there textual evidence that Kant accepted it? In this section, I’ll lay out some initial ideas. In various well-known passages, Kant appears to identify the will with practical reason. In the Groundwork, Kant writes: “Only a rational being has the capacity to act according to the representation of laws, i.e., according to principles, or [has] a will. Since reason is required for deriving actions from laws [die Ableitung der Handlungen von Gesetzten], the will is nothing other than practical reason” (G 4: 4129). In a perfectly rational agent whose will is unfailingly determined by objective reason, “the will is a   For this use of “formal principle” see Reath (2010: esp. section III).   Citations to Kant are to the volume and page in the Berlin Academy edition of Kant’s Gesammelte Schriften (Berlin: de Gruyter, 1900– ), using the translations (with some modifications) in the Cambridge Edition of the Works of Immanuel Kant. I use the following abbreviations: 8

9

G KpV KU MS Rel

Groundwork of the Metaphysics of Morals (in Kant 1996a). Critique of Practical Reason (in Kant 1996a). Critique of the Power of Judgment (in Kant 2000). The Metaphysics of Morals (in Kant 1996a). Religion within the Boundaries of Mere Reason (in Kant 1996b).

236  Andrews Reath capacity to choose only that which reason, independently of inclination, recognizes as practically necessary, i.e., as good.” In imperfectly rational agents such as ourselves, the will is a capacity to choose what reason recognizes as good, but whose exercise, experience tells us, is not unfailingly directed at objective good, and the determination of the will by objective principles is “necessitation” [Nöthigung]. In our case practical reason constrains volition through imperatives. Even so imperatives represent actions as good and, moreover, determine volition by representing actions as “practically good . . . on grounds that are valid for every rational being as such” (G 4: 413).10 In this discussion, the identification of the will with practical reason is based on the idea that volition is a capacity for “deriving actions from laws” (or principles), thereby representing an action as good in some respect. Presumably the will is a complex capacity to move from rational principle to action through an agent’s representational capacities, by representing or judging an action to be good. Likewise, in the Metaphysics of Morals, Kant defines the will as “the faculty of desire in accordance with concepts” whose “inner determining ground, hence even what pleases it, lies within the subject’s reason” (MS 6: 213). Call this a “rational faculty of desire.” This passage contains the important distinction between Wille (will) and Willkür (choice), which are the legislative (or rational) and executive (or causal) aspects of the rational faculty of desire, or will [Wille] in a broad sense.11 The power of choice is the rational faculty of desire “insofar as it is joined with one’s consciousness of the ability to bring about its object by action.” It is the causal aspect of volition, involving the ability to realize the object or action that one judges to be supported by rational grounds. Wille in a narrower sense is the rational faculty of desire considered “in relation to the ground determining choice to action, and has itself properly no determining ground, but is, insofar as it can determine the power of choice, practical reason itself ” (MS 6: 213). Kant’s identification of the will in a narrow sense with practical reason indicates that volition has a rational or cognitive aspect, while its identification with practical reason references the fact that it is a kind of faculty of desire—the capacity in a living being “to be, by means of its representations, the cause of the objects of these representations” (MS 6: 213; cf. KpV 5: 9n. and KU 5: 220). Will is a faculty of desire in which the representations through which the living being guides its causal activity are based in reason. For example, the representations that guide the being to realize their objects are rational principles, or judgments that represent an action or end as good by “deriving” it from a principle. The practical dimension

10  Cf. KpV 5: 61: “What we are to call good must be an object of desire in the judgment of every reasonable human being, and evil an object of aversion in the eyes of everyone; hence for the appraisal of action, reason is needed in addition to sense.” 11   Beck (1960: 199–202) and Allison (1990: 129) have suggested that Wille and Willkür distinguish the legislative and executive functions of volition. Another important discussion on which I have drawn is Engstrom (2010). Engstrom has argued persuasively that volition has a cognitive dimension and argues that the distinction between Wille and Willkür tracks the distinction between the cognitive and the causal moments of volition. For further discussion see Reath (2013: 41–8).

Did Kant Hold that Rational Volition is Sub Ratione Boni?  237 of volition is not in dispute, and I am concerned to explore the implications of its rational or cognitive dimension. That volition involves “deriving action from laws” (or principles) suggests the following picture. Volition is initiated by ends, principles, or values taken to be good (and which presumably are genuinely good in some circumstances or subject to certain limitations).12 Practical reasoning and judgment takes into account relevant features of one’s circumstances and available alternatives, and it moves from its starting point through these features to a representation of an action that will achieve the end or is called for by the principle or value under the circumstances (perhaps through one or more subsidiary ends or principles for which means are required, perhaps directly). Since it is a species of reasoning, it understands itself to move correctly or in a warranted way from initiating end through circumstances to choice, thus concluding in a representation of the action (or action kind) as rationally supported. Obviously the practical reasoning may be quite complex when multiple or competing ends and values that are assigned differing degrees of priority are on the table, some of which may set limiting conditions on the satisfaction of others. Kantian maxims represent actions (or action kinds) as intelligible occurrences with rational support and are intended to capture the practical reasoning and judgment that guides volition. A maxim is a representation of an action or action kind as rationally supported by facts about an agent’s ends, principles, circumstances, and so on, and thus represents the action as good. (Since the reasoning is general and applies to all relevantly similar cases, the maxim itself is a principle to the effect that given a certain end or value and a certain set of circumstances, one is to φ, or φ-ing is good.) And it is through such a representation of the action as rationally supported or good that the agent guides the deployment of his causal powers. That is, the maxim is a representation of an action or end, through which an agent directs his causal powers, and it is efficacious in realizing the represented action through the agent’s self-understanding as moving correctly from an initiating end or value to choice, that is, by representing the action as (judging it to be) rationally supported or good—viz., rational agents act according to the representation of laws or principles.13 Though it addresses a different issue about volition, the so-called “Incorporation Thesis”—that “the power of choice . . . cannot be determined to action through any incentive except so far as the human being has incorporated it into his maxim (has made it a general rule for himself, according to which he wills to conduct himself)” (Rel 6: 23–4)—readily fits into the conception of will as the capacity to derive actions   The end or principle that initiates practical reasoning is not its starting point in time, but rather the first premise in its rational reconstruction (which admittedly is often artificial). 13   For another discussion of Kant’s conception of practical reasoning, see Herman (2006: 44–61). As the causal aspect of volition, Willkür is the ability to guide one’s causal powers to realize the action represented as good, through that very representation. It is volition moving all the way to action. And as a rational faculty, it likewise understands itself to move in a warranted way from that representation, viz. from a maxim to action. 12

238  Andrews Reath from principles. The point of the Incorporation Thesis is to clarify the freedom of the power of choice, denying that incentives influence choice causally by holding that they do so only through an act of spontaneity on the part of an agent. The act of spontaneity is here characterized as “incorporating” or taking up (aufnehmen) an incentive into a maxim, or “making it a general rule” for one’s conduct. What the thesis claims is that an incentive of any sort can influence choice only by an agent taking the incentive up into practical reasoning and judgment that can be expressed as a maxim, or general rule for one’s conduct. The “incorporation of an incentive into a maxim” could consist of one’s taking the incentive up as an end (making its object an end) that is to initiate practical reasoning to action, or of taking the incentive up into practical reasoning from some pre-existing end (happiness, for example, or some more specific end) as a way of specifying or furthering that end. The “general rule” that the agent adopts is a function of the generality of the relevant practical reasoning. As I see it, the central claim here is that an incentive influences choice, not causally, but by being taken up into practical reasoning that leads to a judgment of an action as good or rationally supported, on which the agent acts. This is a thesis about freedom of choice because it addresses the role of incentives in practical reasoning carried all the way to action (the maxim on which an agent acts).14 The rationalist thesis would hold that the practical reasoning that guides volition constitutively understands itself as part of its self-consciousness to satisfy a condition of universal validity—e.g. to derive or move toward action with sufficient rational support. That is, the thesis is that practical reasoning and judgment aimed at affirming a detachable conclusion about action begins from ends, principles, or values that are taken to be “good . . . in the judgment of every reasonable human being,” and that it takes itself to move correctly from there to a representation of an action as good in just that sense. Moreover, the thesis is that it is part of the self-understanding of rational volition that it is initiated by an end or principle that satisfies a condition of universal validity and that it moves correctly or in a warranted way from that starting point to a practical conclusion (a judgment) about the objective goodness of an action. Assuming that the reasoning is captured in the maxim of action, that is to say that it is part of the self-consciousness of rational volition that agents take their maxims to satisfy a condition of universal validity. In this sense volition would tacitly be guided by and would understand itself to be guided by something like the Universal Law version of the categorical imperative. (This is not to say that all rational volition does satisfy a condition of universal validity, but only that the agent tacitly understands it to.)   On one common understanding of the Incorporation Thesis, the act of spontaneity that is essential to the freedom of choice resides in the adoption of a maxim (as a principle representing an action as good), that is, in the decision to act on the maxim. My suggestion here is that the spontaneity referred to is present throughout the derivation of an action from a principle. It is a feature of all moments of volition, including the taking up of the incentive into practical reasoning as well as the movement from judging an action to be good to action. The Incorporation Thesis is a thesis about Willkür because it concerns volition that moves all the way to action. 14

Did Kant Hold that Rational Volition is Sub Ratione Boni?  239 The example of the deposit in the second Critique is a case of volition that illustrates features of this structure. The agent has made it his maxim “to increase my wealth by every safe means.” The end of increasing one’s wealth is good through its contribution to happiness, which is an objectively good end when its pursuit is properly constrained. The fact that the agent has in his hands a deposit whose owner has died without leaving a record presents “a case” for his maxim. Applying the maxim of increasing one’s wealth by every safe means to this case leads the agent to ask whether “he can give through his maxim such a law as this: that everyone may deny a deposit which no one can prove has been made” (KpV 5: 27). I take it that the agent here applies the initial end to his circumstances (prompted by the owner’s death and realization that he can keep the deposit without detection) to formulate what appears to be a second maxim of denying the existence of the deposit as a means to safely increasing his wealth—a maxim that expresses a specific intention to act. That maxim represents the action of denying the deposit as good by representing it as rationally supported by the initial maxim (the end of safely increasing one’s wealth by all safe means) in the agent’s circumstances. 15 Now of course it is not good, and if the agent acts under, in Engstrom’s phrase,16 the presupposition of universality, the not terribly taxing realization that the secondary maxim fails to satisfy this condition should lead the agent to change his initial judgment and to abandon this intention.17 The rationalist thesis about the nature of volition is certainly consistent with Kant’s remarks about the will and with the conception of practical reasoning suggested by some of his examples. But how will it deal with a kind of case that is clearly possible (since it commonly occurs) in which an agent decides to keep the deposit even after realizing that the relevant maxim fails of universal validity. Bad choice is one obvious difficulty for this conception of volition.18 The thesis must hold that bad willing is due 15   This discussion illustrates an ambiguity in Kant’s understanding of maxims: are they general principles or ends that are sometimes called “Lebensregeln,” or are they more specific principles with an action-end-circumstances schema that can express an intention to act? What Kant initially calls the agent’s “maxim” in this passage is a general end or principle (“to increase my wealth by every safe means”), and its application to the case leads to a maxim of the second kind (to deny the existence of the deposit when one can do so without detection as a means to increasing one’s wealth), which is a representation of the practical reasoning that leads to a specific action under the circumstances. We need not resolve this question about the nature of maxims, but there is no deep problem here. Presumably principles of the first kind are adopted as having rational support, and to articulate that support, they can be stated in the second form as means to happiness (or some other general end) given standing facts about human life. Happiness is my end and something I take to be good; and given various general facts about life in certain social settings along with facts about what I need to lead a satisfactory life, adopting the principle of increasing my wealth by all safe means is a means to happiness—a valid “counsel of prudence.” 16   See Engstrom (2009: 124–7). 17   If the initial maxim has the form of or qualifies as a practical law, it is a principle that all can agree that all are to act from and will apply consistently to all cases that fall under it, including this one. But since applying the initial maxim to this case leads to a contradiction in conception, it does not have the form of a law, nor does the subsidiary principle that results from its application to this case. 18   It is important to distinguish genuine vice, which involves endorsing and acting from bad principles (bad willing), and weakness, which involves knowing failure to live up to principles that one endorses. For an important discussion (aimed at showing how genuine weakness is consistent with the Incorporation Thesis), see Johnson (1988: esp. 358–62). A defense of the strong thesis (both philosophical and as Kant

240  Andrews Reath to either of two kinds of defects in rational commitment or reasoning (or to a combination of both). One is erroneous or untenable assignment of value to the initiating end—for example, the agent takes the initiating end to be objectively good or good without restriction when it is not, or assigns it a priority in the circumstances that cannot be sustained. Or perhaps the agent mistakenly takes his happiness to have greater objective importance than that of others (thereby licensing privileges not extended to other agents). The second kind of error is bad reasoning from the initiating end to action—e.g. the agent might begin from a just assessment of the standing of the end and take the reasoning that it initiates to satisfy a condition of universal validity, when it does not. In the first case, the agent who keeps the deposit might take own happiness to be good without restriction or to have a value that overrides applicable common sense moral norms. Or—the second kind of case—the agent might rightly acknowledge the conditional value of own happiness but “make an exception for herself ” in this case and mistakenly judge that the maxim of denying the deposit as a means of furthering her own happiness is endorsable by anyone. (That sounds pretty bloodless, but what precludes defective value commitments or errors in practical reasoning from being “willful” or “motivated”?) Is this a plausible reading of what goes on in bad choice? Perhaps the following “alternative picture,” is more compelling. Practical reasoning (Wille in the narrow sense) reveals that denying the deposit fails to meet a condition of universal validity and thus would be wrong. But even an agent who sees this clearly still needs to decide whether or not to comply with moral principle (i.e. to abide by the conclusions of Wille), and in this case elects not to, fully aware that the action is Gesetzwidrig. Rational volition does require practical reasoning and judgment about what is good or what one has reason to do (the province of Wille in the narrow sense). But action also requires choice (Willkür), which is the further decision about whether or not to follow the conclusions of practical reason. The conclusions of (objective) practical reason ideally guide choice, but because “the will is not in itself completely in conformity with reason,” they necessitate or constrain.19 This “alternative picture” of bad choice—one that does not ascribe the rationalist thesis to Kant—is committed to a certain understanding of the distinction between Wille and Willkür. To give it a name, I’ll call it the “elective conception” of the will. I think that there are problems with the picture of bad choice that results from the elective conception of the will, and with this conception of the will more generally, and these problems push one towards that rationalist thesis. interpretation) needs to accommodate both. What I say in these paragraphs (about “bad willing”) applies to what Kant calls “true vice” (MS 6: 408), and I make some remarks about weakness at the end of section 4. 19   G 4: 413. One is tempted to take “the will” [der Wille] in this remark—what is not in itself in complete conformity with reason and is thus subject to rational necessitation—to refer to the causal moment of volition, to what Kant in later works specifies as Willkür. But it should refer to all moments of volition. In finite rational agents, any moment of volition—our practical reasoning as well as our implementation of our practical judgments—can deviate from objective norms and thus is subject to rational necessitation.

Did Kant Hold that Rational Volition is Sub Ratione Boni?  241 What are these problems? I’ve suggested that Kant understands volition in the broad sense to be a complex capacity to move from rational principle to action through one’s own representational activity that has both a cognitive dimension involving reasoning and judgment (Wille in the narrow sense) and a causal or executive dimension (Willkür) that are linked. Practical reasoning concludes in a representation of an action as good, and the causal dimension of willing is the ability to actualize the action thus represented. But does the cognitive dimension of volition do any real work in the alternative picture of bad choice—and if not, is it genuine volition? Assume here that practical reasoning leads the agent to see that there are compelling reasons not to deny the deposit, then (perversely) ignores this conclusion. The judgment of Wille then drops out of the picture and does not figure in what he does. Choice is then divorced from practical reasoning and judgment and appears arbitrary. If so, in what sense is bad choice a case of volition? Perhaps the proper analysis is that the agent reasons instrumentally from the end of happiness, which is mistakenly taken to be good in an unrestricted sense, or decides on further reflection that an exception to ordinary moral norms is (after all) warranted in this case. But that story belongs to the rationalist thesis, since it takes the choice to result from defective principle, reasoning, or judgment of Wille. It is volition that is initiated by an unsustainable assessment of the standing of the initial end, or leads to choice through unsound practical reasoning. A general problem with the alternative elective conception of the will is that it detaches choice from practical reasoning—as though once one sees what practical reason judges to be worth doing, one still needs to decide whether or not to do it. (And how would one go about deciding that?) I am inclined to think that the distinction between Wille and Wilkür at MS 6: 213 suggests a different conception of volition that is more coherent. Wille in the narrow sense is the cognitive aspect of volition that it is constitutively aimed at determining what there is good and sufficient reason to do, and it takes its exercises to satisfy a condition of universal validity. Choice is the causal aspect of volition whose function is to execute the conclusion of practical reason, e.g. a representation of a specific action as good through which the agent deploys his causal powers. (Until Wille has put an action or end on the table by representing it as good, choice has nothing to do.20) One might grant that Wille in the narrow sense as the cognitive aspect of volition is constitutively guided by a standard of universal validity and that its fundamental principle is indeed the categorical imperative.21 But surely, one might object, Willkür 20   I develop this reading further (which is indebted to Stephen Engstrom) in Reath (2013: 41–9). Among other things, I suggest there that the internal norm of Willkür, understood as the causal or executive dimension of volition, is the hypothetical imperative. 21   Hence Kant can say that “Laws proceed from the will . . . the will, which is directed to nothing beyond the law itself, cannot be called free or unfree . . . since it is directed immediately to giving laws for the maxims of actions (and is therefore practical reason itself)” (MS 6: 226). Presumably this passage implies that objective practical reason (practical reason correctly applied) follows the categorical imperative. In that case one should hold that individual uses of practical reason, while admittedly fallible, understand themselves to satisfy a condition of universal validity.

242  Andrews Reath or choice need not have any such constitutive aim beyond choosing what to do. It has no internal norm of its own, but is a purely elective capacity to follow the conclusions of objective practical reason that an agent does not always exercise. Sometimes it does, but sometimes it does not follow objective practical reason. While it is not disconnected from practical reasoning, it represents a further step beyond simply concluding that an action is supported by good and sufficient reasons. I find this conception of volition unsatisfactory. If the conclusion of practical reasoning still leaves a further question to be decided—whether to follow that conclusion or not—then choice is disconnected from practical reason in a problematic way. Moreover, in Kant’s philosophical system, we should expect a rational capacity to have its own internal norm that makes it what it is, and (I would argue) that tacitly guides all exercises of the faculty—or better that the faculty is self-consciously guided by its own awareness of its internal norms. (Therein lies the spontaneity of its operation.) This would be the positive conception of the faculty, what it is a power to do. The objection grants that Wille in the narrow sense is constitutively guided by a standard of universal validity, but denies that choice has any such internal norm. However while choice need not have the constitutive aim of satisfying a condition of universal validity, if it is the causal or executive dimension of volition in the broad sense, it will be charged with and will understand itself to be carrying out the conclusions of practical reason (which is constitutively guided by a principle of universal validity). So its constitutive aim ties it to the internal norms of Wille. Kant’s claim that choice cannot be defined as a capacity to choose for or against the moral law is relevant here. Kant tells us that the positive conception of free choice (die Freiheit der Willkür) is “the ability of pure reason to be of itself practical . . . through subjecting the maxim of every action to the condition of its qualifying as a universal law” (MS 6: 213–14). And although only choice (as opposed to will in the narrow sense) is properly free, “freedom of choice cannot be defined. . . as the ability to make a choice for or against the law (libertas indifferentiae), even though choice as a phenomenon provides frequent examples of this in experience . . . [F]‌reedom can never be located in a rational subject’s being able to choose in opposition to his (lawgiving) reason, even though experience proves often enough that this happens.” Our experience of how choice is exercised cannot give us “the expository principle [Erklärungsprinzip] (of the concept of free choice) and the universal feature for distinguishing it (from arbitrio bruto s. servo)” (MS 6: 226). One point to make about this passage is that the positive conception of freedom of choice should be the positive conception of the causal dimension of practical reason. (That is, he is not concerned with the positive conception of Wille in the narrow sense.) The claim is that the positive conception of free choice is the ability of pure reason to be practical—that is, the ability of pure reason to realize, through its representations (of actions and ends as good), the objects of those representations independently of conditions of sensibility, or the ability to act from pure practical reasoning. But then, freedom of choice does have its own internal norm or constitutive aim,

Did Kant Hold that Rational Volition is Sub Ratione Boni?  243 that of realizing the conclusions of (pure) practical reason, and if that is its positive principle, its exercise does not float free of Wille in the narrow sense. A second point concerns how to understand the positive conception of a faculty. I’ll suggest in section 3 that the positive conception of a faculty—what Kant here terms its “expository principle” (Erklärungsprinzip)—is the internal norm consciousness of which tacitly guides all exercises of the faculty. If so, then all practical reasoning understands itself to be guided by the condition of universal validity (as sketched above) and all rational choice understands itself to be implementing the conclusions or judgment of practical reasoning, or doing what practical reason judges good in this sense. This passage from the Metaphysics of Morals does not require this reading of the “positive conception,” but it does offer a way of filling out what this claim amounts to. This section of the chapter has shown that Kant’s key discussions of the will are consistent with and point toward the rationalist thesis about the nature of rational volition. However, I don’t think that these texts require this interpretation of volition. First, the passages from the Groundwork and related passages from the second Critique (KpV 5: 57–62 ff.) make it clear that Kant thinks that rational volition is guided by representations of actions as good, but such representations could include judgments that an action is good as a means, or that some action or end is good conditionally or on prudential grounds. It is not obvious that such practical judgments make a claim to their own universal validity in the sense needed for the rationalist thesis (in fact, many philosophers think that it is obvious that they do not . . .), and Kant never explicitly makes this claim. Second, the claim that an imperfectly rational will “does not always do something just because it is represented to it that it would be good to do” (G 4: 413) and similar passages certainly lend support to the conception of freedom of choice as an elective capacity that operates independently of practical reason in the sense that it often ignores or acts contrary to the conclusions of objective practical reason. This tends to be the default reading of free choice. My aim in this section has been to suggest alternative readings of such passages that are (at least) sufficiently compelling to supplant the elective reading as the default reading. I’ll discuss some of these passages in the final section. Finally, the rationalist thesis relies on a certain take of what goes into the “positive conception” of free choice. I think that it is philosophically compelling and will take the point up in section 3, but I grant that it is not the only way to read the texts. In sum, the rationalist thesis is consistent with and supported by, but not required by key texts about the will.22 22   Chapter 2 of the Analytic of the second Critique is one place where one might expect to find support for Kant’s acceptance of a strong guise of the good thesis, but this chapter does not seem to me to address the issue clearly one way or the other. First, Kant does say that good and evil are the “only objects of a practical reason” and that they are “necessary objects” of the faculty of desire or aversion “both however in accordance with a principle of reason.” The necessity in question appears to be that what is good “must be an object of the faculty of desire in the judgment of very reasonable human being, and evil an object of aversion in the eyes of everyone”—what is good or evil is what everyone can rationally judge to be so (KpV 5: 58, 61). But Kant does not clearly say that volition is necessarily directed at what one judges to be good or that the good is the formal object of rational volition. Second, Kant finds an ambiguity in the scholastic formula nihil appetimus,

244  Andrews Reath

3  Philosophical Support In this section I  will explore what the rationalist thesis that volition constitutively understands itself to satisfy a condition of universal validity would do for Kant philosophically. I will be focusing on two well-known ideas from the Groundwork and second Critique—the thesis of autonomy of the will and the claim that “a free will and a will under moral laws are one and the same” (G 4: 447). I will suggest, first, that the thesis about rational volition is required by Kant’s conception of autonomy, and second, that it gives the best account of Kant’s claim that “a free will and a will under moral laws are one and the same” (G 4: 447)—or as we might say, that the moral law is the basic principle of free rational agency. Furthermore, it offers a satisfying account of how bad willing can be genuine free volition, given the close connection that Kant draws between free agency and the moral law. To begin with the first, Kant characterizes autonomy of the will as “the property of the will by which it is a law to itself (independently of any property of the objects of volition)” and formulates that law (that the will is to itself) as the Universal Law version of the categorical imperative (G 4: 440, 444, 447). The best reading of the thesis that the will is a law to itself “independently of any property of the objects of volition” is that the nature of rational volition by itself is the source of its own fundamental norm.23 But what does that mean? First, rational volition must have a “nature” that is sufficiently rich to generate a fundamental practical principle that can guide volition, without the need to turn to any object or interest given to the will from outside. To say that a rational faculty has a nature is to say that it has a formal aim or constitutive principle that can guide its own exercise, and moreover that does tacitly guide all exercises of the faculty. We know that the moral law (categorical imperative) is the law given by its nature (“not to choose in any other way than that the maxims of one’s choice are also comprised as universal law in the same willing” (G 4: 440)). So unpacking Kant’s thesis of autonomy in this way leads directly to the thesis that rational volition constitutively understands itself to satisfy a condition of universal validity, viz. that the Universal Law version of the categorical imperative is the formal principle of all rational volition.24 Must this norm tacitly guide all instances of rational volition? Late in Groundwork II Kant claims:

nisi sub ratione boni . . ., depending on whether bonum refers to das Gute or das Wohl. He thinks the formula “indubitably certain” if bonum refers to das Gute, in which case it can be rendered: “we will nothing under the direction of reason [wir wollen nach Anweisung der Vernunft nichts] except insofar as we hold it to be good or evil” (KpV 5: 60). When reason properly directs volition, it is of course sub ratione boni. But does this phrasing allow that we can sometimes will though not under the direction of reason, or that choice can ignore reason’s judgments? That would be the alternative elective conception of the will. I need Kant to hold that all volition involves the exercise of reason and takes itself to conform to reason’s standard, but it seems to me that this passage could be read either way.   I defend the points in this paragraph and the next in more detail in Reath (2013: section III).   Here note also G 4: 433, that the will is a “will giving universal law according to its natural end.”

23

24

Did Kant Hold that Rational Volition is Sub Ratione Boni?  245 An absolutely good will, whose principle must be a categorical imperative, will therefore, indeterminate with regard to all objects, contain merely the form of willing as such [bloß die Form des Willens überhaupt], and indeed as autonomy; i.e. the fitness of the maxim of every good will to make itself into a universal law is the sole law that the will of every rational being imposes on itself, without underpinning it with any incentive or interest as its foundation. (G 4: 444)

That the principle of a good will “contains merely the form of willing as such” is easily read as the claim that the Formula of Universal Law expresses a formal feature of any exercise of the will. That would make it the formal principle that tacitly guides all rational volition. Furthermore, the role that the thesis of autonomy plays in Kant’s foundational argument is to support the unconditional authority of the moral law. Kant does that by demonstrating that the moral law is the formal principle of rational volition to which one is committed simply in exercising the will. Turning now to the second cluster of ideas, early in Groundwork III, Kant argues that from the negative explication of free agency as independence from determination by alien causes there flows a positive conception—a positive specification of what free agency qua capacity is that includes a statement of the principle according to which free agency operates.25 Freedom of the will can only be “autonomy, i.e. the property of the will of being a law to itself,” and Groundwork II has argued that the Universal Law version of the categorical imperative is the law that the will, as practical reason, gives to itself. More precisely: “the proposition: the will is in all actions a law to itself, designates only the principle of acting on no maxim other that that which can also have itself as object as a universal law.” Thus “a free will and a will under moral laws are one and the same” (G 4: 447)—and by that I take it Kant means that the basic principle of free agency is that of acting from maxims that can be willed as universal law or that satisfy a condition of universal validity. Likewise in the second Critique Kant argues that a will for which “the mere law-giving form of maxims alone is the sufficient determining ground” is a free will (KpV 5: 28–9).26 Setting aside the details of the argument, how should we understand the claim that “a free will and a will under moral laws are one and the same”? A standard reading of this claim finds two general points in it. The first is a conception of free agency: free agency is the capacity to determine choice by moral principles, and free and imputable actions are those chosen by a free agent in whom this capacity is unimpeded or undiminished, whether or not it is fully exercised. The second general point is to unpack the claim that 25   Here compare also KpV 5: 105–6, where Kant holds that the reality of free agency is established by identifying the principle of its operation. This is “an objective principle of causality . . . in which reason . . . already itself contains this determining ground by that principle, and in which it is as pure reason itself practical.” Thus freedom is “not merely thought indeterminately and problematically . . . but is even determined with respect to the law of its causality and cognized assertorically . . . [O]‌ur reason itself, by means of the supreme and unconditional law cognizes itself . . . and indeed even with a determination of the way in which, as such, it can be active.” These passages are naturally read as implying that all exercises of free agency are in some sense guided by the moral law (or in our case the categorical imperative). 26   I assume that the will in these passages is the faculty of rational volition in the broad sense (including both Wille in the narrow sense and Willkür).

246  Andrews Reath the moral law is the basic principle of free agency in terms of the basic commitments of free rational agents.27 The Universal Law version of the categorical imperative is reason’s own principle, and action from that principle is free. And since it is a necessary component of the self-consciousness of rational agency that one acts under the idea of freedom and identifies with this capacity (the capacity that makes one an agent), one is rationally committed to acting on the principle by which it is defined. When one does, one fully realizes the capacity, and that is a good for the agent. (It is the good that the opening of Groundwork I presents as unconditional and without limitation, now understood in a way that shows how it can be a good for a finite rational agent.) According to this conception, a free agent has the capacity to judge what is required by and to act from universally valid principles, and is committed to acting on such principles by one’s self-conception as free. When one does act from maxims that satisfy this standard, one realizes one’s free agency and the good that it represents. But sometimes free agents act badly—on maxims or practical judgments that fail to satisfy this normative standard, indeed sometimes knowing full well that one’s maxim falls short of accepted moral standards. In such cases, though practical reason shows that one’s maxim is not universally valid, the agent makes the elective choice to act on it all the same. Here the agent fails to exercise the capacity to act from universally valid principles, and in this respect bad choice represents a failure to exercise one’s free agency. But

  Hill (1992) has developed one of the more interesting reconstructions of the argument that a free agent is committed to the fundamental moral principle, which has not been given sufficient recognition in the literature. Hill takes the argument to show that if one grants (plausibly) that the negative conception of freedom includes a capacity for desire-independent motivation, any free rational agent is committed to “some principle or principles which are rational and yet not hypothetical imperatives,” and thus must acknowledge some rational principle beyond those recognized by instrumentalists such as Hobbes and Hume. Practical reason in free rational agents is not simply instrumental (112). His reconstruction of Kant’s argument from negative freedom to autonomy goes roughly as follows (111–12): 27

A will with autonomy is committed to at least one practical principle that does not simply prescribe the means to the agent’s desired ends and is “one’s own” in the sense that it is deeply rooted in one’s rational nature. A negatively free will can act for reasons that are not based on desires or hypothetical imperatives. Since a lawless will is absurd, such agents must act from some principle (when they act on non-desire-based reasons), thus must be committed to or accept some rational principle that is not a hypothetical imperative. The previous step rules out various substantive principles that are clearly hypothetical imperatives (such as desire-based principles, principles adopted because accepted by some external authority or convention, etc.). It also appears to rule out the principle of assigning prima facie weight as reasons to informed preferences that have survived critical reflection. (115) The only remaining candidate principles are those “which reflect some necessary features of rational agency independently of its special contexts.” (112) Since such principles satisfy the definition of autonomy in Step (1), an agent who is negatively free has autonomy. This argument is “non-constructive” in the sense that it does not clearly identify the principle(s) to which free agents are rationally committed, but Hill thinks that it points towards the principle that any rational agent is committed to valuing his or her practical rational capacities, which is the basis of the Formula of Humanity.

Did Kant Hold that Rational Volition is Sub Ratione Boni?  247 the action is free and imputable because chosen by a free agent in whom the capacity is undiminished. While there is nothing decisively unsatisfactory with this conception, here are two worries that one might raise. The first, due to its reliance on the elective conception of the will mentioned in section 2, is whether it can count bad choice as a genuine exercise of volition, thus as genuinely bad willing. The issue here is whether a failure to exercise the capacity for volition counts as genuinely bad willing. This conception appears to treat bad choice as an elective choice that is divorced from, or at least sets aside, practical reasoning and judgment, and that makes it seem not just perverse or weak, but arbitrary. In good willing, choice follows the judgment of practical reason, but what happens in bad willing? Presumably the agent chooses an end that initiates action, but not through any representation of it as good. (And if not, what grounds the choice of the end?) Since bad willing so conceived does not engage the capacity to make judgments of good, it represents a volitional failure—a failure to exercise the capacity for rational volition. But then bad choice is not really volition—though it is still imputable because it is the act of an agent with a capacity for rational volition that he has failed to exercise. Second, it is part of this conception that when one does act from maxims that genuinely satisfy the normative standard of universal validity, one “realizes one’s nature” as free and rational. It is certainly an asset of a moral conception to be able to show that acting from moral principles secures a “higher good” of self-realization for the agent. But the Groundwork, at any rate, connects freedom and morality to establish the unconditional authority of morality. This conception appears to locate the reasons to give deliberative priority to the moral in the good of the complete realization of one’s nature. But can the good of self-realization ground the conclusion that morality is required of us? Fully realizing one’s nature is a good, but must one care about it above all other goods?28 I am proposing a different way to understand the claim that the moral law is the basic principle of free agency. I read it as the claim that the Universal Law version of the categorical imperative is the formal principle of free rational agency, to be unpacked through the rationalist thesis that rational volition constitutively understands itself to satisfy a condition of universal validity. As we’ve seen, this thesis holds that volition constitutively understands itself to move correctly from ends or principles taken to be good to judgment of an action as good—that is, it has the formal aim of acting on maxims that satisfy a condition of universal validity. To say that rational volition constitutively understands itself to satisfy a condition of universal validity is to say that this self-consciousness is a necessary feature that guides all volition; expressed as a principle, it is the Formula of Universal Law (FUL). The will “imposes” this principle on itself 28   Several years ago Henry Allison pointed out to me that many interpretations of Kant’s attempt to ground the authority of morality in free agency make it, wrongly in his view, an ethic of self-realization or perfection. See Allison (1996: 117–18).

248  Andrews Reath through the self-consciousness that guides volition, thus is normatively committed to complying with this principle by its own self-understanding.29 The proposed conception makes volition an act of the (rational) faculty of desire that understands itself to satisfy a condition of universal validity. This mark applies as much to bad as to good willing, and that makes bad willing free, a genuine exercise of volition on all fours with good willing. What makes it bad is that it fails to satisfy the condition of universal validity that is the internal standard of volition because of some defect in the rational commitments or practical reasoning that guides volition. (For example, the agent takes a bad end or principle to be universally valid, or assigns unrestricted value to a conditionally good end, or reasons badly from the initiating end or principle, and so on, through perversity or weakness, and so on.) The thesis that the FUL is the formal principle of free agency (that volition constitutively understands itself to satisfy a condition of universal validity) needs some qualification. The claim is not that all rational volition does satisfy a condition of universal validity, but rather that it understands itself to satisfy this condition and proceeds under this self-conception. Thus as I frame the thesis, it allows the formal norm of universal validity to come apart from genuine moral principles in an agent’s practice— that is, it is not claiming that agents necessarily take their maxims to have moral justification or to accord with commonly accepted moral principles. An agent who denies the authority or universal validity of ordinary moral norms (e.g. by taking himself to be exempt from them) can still take his maxims to satisfy an indeterminate condition of universal validity. Thus, the formal principle can be part of the self-consciousness of an agent with a very attenuated moral sense. In this respect, my reading departs from Kant’s wording at Groundwork 4: 447: the thesis is not that the moral law, properly speaking, but that the principle of universal validity is the formal principle of free volition (where an agent’s understanding of universal validity might fall short of moral justifiability). Finally, by holding that the mark of free volition is that it understands itself to satisfy a condition of universal validity, one avoids the objection that bad willing does not really count as willing. If, to count as volition, an act of the faculty of desire must be guided by in the sense of satisfying the condition of universal validity, action on maxims that are not fully universalizable would not represent genuine free volition, thus would not count as bad willing. But the idea that volition constitutively understands itself to satisfy a condition of universal validity implies that the essential mark of volition is not that it does, but that it takes itself to satisfy a condition of universal validity. What makes an act of the faculty of desire an act of rational volition is this self-conception. But an act could be guided by this self-consciousness and fail to satisfy this normative standard. In that case, it is still volition—genuine free, though bad, willing.

  For elaboration of this last point, see Reath (2013: 47–9).

29

Did Kant Hold that Rational Volition is Sub Ratione Boni?  249 What one wants here is a sense of being guided by a principle that allows for genuine violations, or failures to conform the principle. I take that form of normative guidance to be supplied by the self-understanding or self-consciousness of a kind of cognitive or rational activity. The thought is that what makes an act the kind of rational activity it is—in this case what makes an act of the faculty of desire an act of rational volition—is that it understands itself to have a certain formal aim. This self-consciousness normatively guides the activity and gives the norms internal to that activity a grip on the agent: through the self-consciousness that one is engaging in a certain kind of rational activity, one takes oneself to be following and correctly applying the relevant internal norms (in this case, that one is acting from maxims that satisfy a standard of universal validity). But one can take oneself to be guided by a set of norms and still violate or fail to conform to them. As a final note, this reading of the idea that the FUL is the formal principle of free agency includes the conception set aside earlier in this section and does full justice to Kant’s claim that a free will is “subject to moral laws” (unter sittlichen Gesetzen). A faculty of desire that is constitutively guided by the condition of universal validity certainly has the capacity to act from genuine moral principles and fully realizes its nature when it does. Further, it is subject to genuine moral principles in the sense that it is committed to accepting the authority of morality by its own self-consciousness. In willing it understands itself to be acting from maxims that satisfy a condition of universal validity, it imposes this requirement on itself through this self-conception, and it is thus in self-contradiction when its maxims fail to satisfy the standard of universal validity. I’ll conclude this section by contrasting the main features of my own proposal with the elective conception of volition that I want to set aside. According to the elective conception, Willkür is a capacity for choice that is independent of and floats free of Wille (in the narrow sense) in that it involves the further decision whether or not to follow the judgment of practical reason. Willkür is the capacity to act from universalizable maxims, but (though it ought to follow pure practical reason) it has no constitutive aim or formal principle and thus does not necessarily operate sub ratione boni. Its freedom is its capacity to act from genuine moral principles, which an agent sometimes, but does not always exercise. According to the elective conception, the essential elements of Kant’s idea that the moral law is the basic principle of free agency are that free agency is the capacity to act from moral principles, that free agents are rationally committed to acting from the moral law, and perhaps that doing so secures the good of self-realization. One problem with this conception is that it is not clear that it can count bad choice as genuine willing, rather than simply a failure to exercise the capacity for rational volition. (Still it is free choice and imputable because it is an act of an agent with the capacity for free choice.) The contrasting conception that I  wish to defend takes Willkür to be the causal moment of a complex faculty of volition whose constitutive aim—to realize what practical reason represents as good—links it to the constitutive aim of Wille. Wille (the

250  Andrews Reath cognitive aspect of volition) understands itself to move correctly from principles taken to satisfy a conception of universal validity to representations of actions as good, and Willkür is constitutively aimed at realizing what practical reason represents as good. Rational volition as a whole thus proceeds sub ratione boni in the sense that it constitutively understands itself to satisfy a condition of universal validity. I read Kant’s claim that the moral law is the basic principle of free agency as the claim that the Universal Law version of the categorical imperative is the formal principle of volition that tacitly guides all willing, as I’ve explained that thesis above. I’ve argued that this conception is required by Kant’s doctrine that the will is a law to itself (the central plank of which is that the nature of the will is the source of its own formal principle). Further, the claim that FUL is the formal principle that tacitly guides all volition provides a natural reading of Kant’s idea that the moral law is the basic principle of free agency. One clear strength of this reading is that it makes bad willing genuine free volition—free volition because it understands itself to satisfy a condition of universal validity, but bad because of defects in fundamental principles and value commitments and in reasoning. That is to say that bad willing is not just a failure to exercise the capacity distinctive of volition, but a misuse of that capacity. That the rationalist thesis connects in these ways with these central Kantian ideas provides powerful (even if not decisive) reasons to think that Kant is committed to it. Kant thinks that it is a necessary feature of practical self-consciousness that rational agents act under the Idea of Freedom. If the Universal Law version of the categorical imperative is indeed the formal principle of rational agency, then acting sub ratione boni in this sense is just a different aspect of this same Idea.

4  Bad Willing? In this section, I consider briefly how the rationalist thesis that FUL is the formal principle of rational volition fits the phenomenon of bad willing. Does the thesis fit what Kant says about bad willing? And in making a case for ascribing this thesis to Kant, am I saddling him with a highly implausible conception of volition and action? Kant clearly thinks that we can knowingly and intentionally act contrary to moral principle. To cite a few instances:30 a finite will “does not always do something just because it is represented to it as something that it would be good to do,” which presumably implies that such agents sometimes do things that they know to be bad or contrary to reason (G 4: 413). The scoundrel is conscious of the good will in him that “constitutes the law, the repute [or authority] of which he recognizes as he transgresses it [dessen Ansehen er kennet, indem er es übertritt]” (G 4: 455). Human beings are “unholy enough 30   Thanks to Jens Timmermann for these passages and for discussion of this issue generally. The first three passages suggest cases of weakness, but I think that all can be read either as weakness or as “perversity” or “true vice” (bad willing). Bad (rather than weak) willing is easier for my account to handle, but I’ll comment briefly on weakness at the end of this section.

Did Kant Hold that Rational Volition is Sub Ratione Boni?  251 that pleasure can induce them to break the moral law, even though they recognize its authority” (MS 6: 379). “Now through experience we can indeed notice unlawful actions, and also notice (at least within ourselves) that they are consciously contrary to law [das sie mit Bewußtein gesetztwidrig sind]” (Rel 6: 20). The rationalist thesis about the will does not preclude the possibility of action that is knowingly contrary to moral principle. As I said in section 3, it does not claim that rational volition necessarily takes itself to satisfy standards of moral justification. In general, the thesis traces bad willing back to some part of the cognitive aspect of volition, and here we’ve mentioned two general possibilities. One is that practical reasoning and judgment are initiated by defective or unsustainable principles or value commitments that are taken to satisfy a condition of universal validity. For example, an agent might reason from an unrestricted principle of own happiness that is (wrongly) taken to be universally valid. Here the agent reasons from a non- or counter-moral principle and does not accept the deliberative priority of morality. A second is bad or defective reasoning from acceptable principles that is incorrectly taken to lead to a universally valid conclusion about action—e.g. as when an agent who professes acceptance of the conditional value of happiness (wrongly) ignores those conditions in his reasoning to action. While it is perhaps clearest in the first kind of case, an agent who consciously rejects the deliberative priority of morality can take his or her practical reasoning or maxim to satisfy a condition of universal validity, while knowing that it does not conform to accepted moral principles. What this shows is that the condition of universal validity can be understood by an agent in such a way that it falls short of genuine moral justifiability, so that the formal principle of universal validity can come apart from morality in the agent’s reasoning. This allows for genuine volition that is consciously contrary to morality—‘mit Bewußtsein gesetzwidrig’. Kant’s discussion of self-love and self-conceit in the second Critique provides a model for tracing bad willing to defective practical principles that are taken to satisfy a condition of universal validity. Indeed, “self-conceit” appears to elevate a counter-moral principle to the status of “law.” To cite a well-known passage: Now, however, we find our nature as sensible beings so constituted that the matter of the faculty of desire . . . first forces itself upon us, and we find our pathologically determinable self, even though it is unfit to give universal laws through its maxims, nevertheless striving to make its claims primary and originally valid, just as if it constituted our entire self. This propensity to make oneself, on subjective determining grounds of choice [nach subjectiven Bestimmungsgründen seiner Willkür] into the objective determining ground of the will [des Willen] in general can be called self-love; and if self-love makes itself lawgiving and the unconditional practical principle, it can be called self-conceit. (KpV 5: 74)

I take the “pathologically determinable self ” to be the (rational) self as susceptible to the influence of sensible inclination. It experiences sensible inclinations as incentives (including the natural inclination to distinguish oneself in comparison with and to be well regarded by others; cf. MS 6: 465). But as a rational subject it experiences these

252  Andrews Reath incentives by taking them up into practical self-consciousness in a way that makes them available for use in practical reasoning and judgment. Desires that originate independently of one’s rational capacities are taken up in rational form, as potential ends or principles from which one can reason and that can be the basis of “claims.” Simplifying, in self-love an agent (wrongly) “claims” an unrestricted value for own happiness so that it is the basis of sufficient reasons for action, while in self-conceit an agent (wrongly) claims for one’s own person an unconditional standing not reciprocally accorded to others. This passage makes it clear, first, that these forms of bad willing rest on defective fundamental principles and value commitments. What is corrupted, as it were, are the basic premises of practical reasoning, in this case due to a defective conception of the self: the “pathologically determinable self ” strives to make the claims of self-love and self-conceit valid “just as if it constituted our entire self.” Further, the claims of self-love and self-conceit are treated as universally valid. Though the pathologically determined self is “unfit to give universal laws through its maxims,” it puts its claims forward as “primary and originally valid,” as “objective determining grounds of the will” or as “law-giving.” These remarks suggest that it is the nature of the rational self, even as sensibly affected, to operate through principles that are understood to make a claim to universal validity. Finally, if contrary to fact the sensibly affected self were the entire self—if the self were exhausted by incentives of self-concern (Selbstsucht)—then by default the principles underlying self-love and self-conceit would be universally valid principles. But this conception of universal validity would support only a weak notion of justification that falls short of moral justifiability, because these principles cannot ground genuinely sharable or reciprocally recognized normative claims.31 The thesis that rational volition constitutively understands itself to satisfy a condition of universal validity does not locate bad willing (willing on bad principles) in an act of choice (Willkür) that ignores or proceeds independently of the judgments of practical reasoning. Rather, it traces it back either to defective rational principles and value commitments or to bad reasoning from (possibly innocent) principles, that, as exercises of practical reason, understand themselves to satisfy a condition of universal validity. This conception has strong interpretive support from Kant’s discussion of self-love and self-conceit (as well as the discussion of “impurity” and “depravity” or “perversity” in the Religion (Rel 6: 30, 36)). Further, the thesis can accommodate bad willing that is knowingly contrary to morality because volition can be guided by nonor counter-moral principles adopted on the supposition of their universal validity, in full awareness that they violate common moral principles. (As I have said, the formal standard of universal validity can be understood in a way that falls short of moral justifiability.)

  For further discussion, see the Appendix of ch. 1 of Reath (2006).

31

Did Kant Hold that Rational Volition is Sub Ratione Boni?  253 A complete account must also show that this conception of volition can accommodate weakness—volition that fails to conform to principles that the agent actually accepts. Indeed weakness may pose the greater challenge, and I don’t have the space to address it adequately here. But the preferred strategy will be to locate weakness in the cognitive dimension of volition—in the fundamental principles or reasoning and judgment that lead to action, rather than in a simple failure of choice (Willkür) to follow practical reason (Wille). One possibility is to trace weakness to unclarity in the content of an agent’s principles—for example, if an agent is unclear about what a principle demands in some situation, or understands it so as to permit exceptions in its application that it does not strictly license.32 Another is to trace weakness to defective reasoning from or application of principles that are otherwise sound. Such cases would be genuine but weak willing. A back-up possibility for handling certain kinds of weakness is to treat them as intentional action that is not the result of volition. In this respect, the thesis is less strong than it might initially appear since it need not imply that all intentional behavior understands itself to satisfy a condition of universal validity. Such choices would be intentional in the sense that they issue from a faculty of desire—a capacity by means of one’s representations to realize the objects of these representations. But they would not be rationally willed since they do not engage the distinctive capacity for rational volition, which initiates action from a principle understood to satisfy a condition of universal validity.33 (Perhaps the action in this case would be the work of a sub-agential system in an agent of which the agent is conscious, and which the agent has the capacity but fails to control.) Instances of weakness understood in this way would not, strictly speaking, be “weak willing,” but volitional failure—failures to exercise the will. But the resulting action can still be regarded as free and imputable, since it is performed by an agent who possesses but fails to exercise the relevant rational capacities. One advantage of the rationalist conception of volition that I have been trying to articulate on Kant’s behalf is that it counts bad or perverse willing as genuine volition—not just a failure to exercise the capacity for volition, but a misuse of it. Above I suggested that the alternative elective conception has to view bad (and weak) willing across the board as volitional failure—as based on an arbitrary act of choice that, because it does not engage the capacity for practical judgment, is not genuine volition but a failure to exercise the will. Bad action would still be imputable because it is the choice of an agent with the capacity for rational volition that he has failed to exercise. In resorting to this model of imputability for some instances of weakness, the advantage of my account runs out at this point. But since my conception uses this model only 32   Here see Hill (2012: section 5). Hill suggests a Kantian account of weakness that locates it either in the content of a person’s principle or in how resolutely a person wills a principle. Both are ways of locating weakness in a person’s Wille rather than Willkür. 33   Engstrom seems to allow for the possibility of intention that falls short of volition—for instance in practical (efficacious) thought whose efficacy is not based on a judgment of good. See Engstrom (2009: 44 ff.).

254  Andrews Reath in certain cases of weakness, rather than using it across the board for all bad and weak willing, the advantage is retained. Obviously more needs to be said about both perverse and weak willing than I have room for here, and I do not underestimate the challenges that these phenomena pose for the conception of volition that I have tried to articulate. What I hope to have done in this chapter is to lay out a case for ascribing this conception to Kant that focuses on his central doctrines of autonomy and the connection between morality and free agency.

References Allison, Henry A. 1990. Kant’s Theory of Freedom. Cambridge/New York: Cambridge University Press. Allison, Henry A. 1996. “Kant on Freedom: A Reply to my Critics,” in Allison, Idealism and Freedom. Cambridge/New York: Cambridge University Press. Beck, Lewis White. 1960. A Commentary on Kant’s Critique of Practical Reason. Chicago: University of Chicago Press. Engstrom, Stephen. 2009. The Form of Practical Knowledge. Cambridge, MA:  Harvard University Press. Engstrom, Stephen. 2010. “Reason, Desire and the Will,” in Lara Denis, ed., Kant’s Metaphysics of Morals: A Critical Guide. Cambridge/New York: Cambridge University Press. Herman, Barbara. 2006. “Reasoning to Obligation.” Inquiry, 49(1): 44–61. Herman, Barbara. 2007. Moral Literacy. Cambridge, MA: Harvard University Press. Hill, Thomas E., Jr. 1992. “Kant’s Argument for the Rationality of Moral Conduct,” in Hill, Dignity and Practical Reason. Ithaca, NY: Cornell University Press. Hill, Thomas E., Jr. 2002. “Personal Values and Setting Oneself Ends,” in Hill, Human Welfare and Moral Worth. Oxford/New York: Oxford University Press. Hill, Thomas E., Jr. 2012. “Kant on Weakness of Will,” in Hill, Virtue, Rules, and Justice: Kantian Aspirations. Oxford/New York: Oxford University Press. Johnson, Robert. 1988. “Weakness Incorporated.” History of Philosophy Quarterly, 15(3): 349–67. Kant, Immanuel. 1996a. Kant: Practical Philosophy, trans. and ed. Mary J. Gregor. Cambridge/ New York: Cambridge University Press. Kant, Immanuel. 1996b. Religion and Rational Theology, trans. Allen W. Wood and George di Giovanni. Cambridge/New York: Cambridge University Press. Kant, Immanuel. 2000. Critique of the Power of Judgment, trans. Paul Guyer and Eric Matthews. Cambridge/New York: Cambridge University Press. Korsgaard, Christine M. 1996a. Creating the Kingdom of Ends. Cambridge/New York: Cambridge University Press. Korsgaard, Christine M. 1996b. The Sources of Normativity. Cambridge/New York: Cambridge University Press. Korsgaard, Christine M. 2009. Self-Constitution:  Agency, Identity and Integrity. Oxford/ New York: Oxford University Press, 2009. Reath, Andrews. 2006. Agency and Autonomy in Kant’s Moral Theory. Oxford/New York: Oxford University Press.

Did Kant Hold that Rational Volition is Sub Ratione Boni?  255 Reath, Andrews. 2010. “Formal Principles and the Form of a Law,” in Andrews Reath and Jens Timmermann, eds., Kant’s Critique of Practical Reason: A Critical Guide. Cambridge/ New York: Cambridge University Press. Reath, Andrews. 2013. “Kant’s Conception of Autonomy of the Will,” in Oliver Sensen, ed., Kant on Moral Autonomy. Cambridge/New York: Cambridge University Press. Scanlon, T. M. 1998. What We Owe to Each Other. Cambridge, MA: Harvard University Press.

13 Kantian Complicity Julia Driver

Kantian ethics has long been thought of as having a rather severe problem with complicity.1 Notoriously, if we are to take Kant at his word, responsible adherence to the Categorical Imperative (CI) seems to require complicity in wrongdoing. The infamous case in which Kant holds that one cannot lie, even for an altruistic motive, is used to illustrate the point.2 Kant quite clearly holds that “Truthfulness in statements that cannot be avoided is the formal duty of man to everyone, however great the disadvantage that may arise therefrom for him or for any other.”3 The verdict that one must tell the truth even if the truth will be used to kill an innocent person strikes many as implausible. It seems to conflict with an intuitively plausible principle, that of complicity avoidance: (AV) One ought to avoid complicity in wrongdoing (pro tanto), when so avoiding does not result in a greater harm. Contemporary philosophers who accept Kant’s system have either tried to show that Kant’s view is really not so implausible, when understood in the right way, or they have tried to show that Kant may have misapplied his own system in failing to recognize that the maxims of action need to be specified in “more subtle” and “context-sensitive” ways. One way to develop the first strategy is to insist that Kant’s system does not commit one to violating (AV). That’s because when a person, for example, tells the prospective murderer where his friend is hiding, he is not really complicit in his friend’s death. The idea is that assiduously conforming to the CI is incompatible with wrongdoing, and complicity constitutes a kind of wrongdoing. Therefore, if the CI requires truth   Thanks are owed to members of the St Louis Ethics Workshop (SLEW). I would also like to thank the audiences at Ghent University, the University of Amsterdam, the University of Nebraska at Lincoln, and the Fall 2012 Meeting of the Indiana Philosophical Association. 2   “On a Supposed Right to Lie because of Philanthropic Concerns,” in Grounding for the Metaphysics of Morals, trans. James Ellington (Indianapolis: Hackett Publishing Co., 1981), 63–7. Ellington notes it originally appeared in Berlinische Blätter (published by Biester) September 1799. 3   “On a Supposed Right to Lie because of Philanthropic Concerns,” 64. 1

Kantian Complicity  257 telling, even to someone about to commit murder, that truth telling is not complicity in wrongdoing. This line of reasoning is mistaken for a variety of reasons that will be explored in this chapter. However, the primary problem is that it is just not plausible. The second strategy seems much more plausible to me, though that strategy involves giving up at least some of Kant’s stated constraints on how to use the CI. But even the second strategy is incomplete in that it fails to take advantage of a feature of Kant’s ethics that distinguishes it from theories such as consequentialism. On Kant’s theory there is no need to tie wrong making to production of the bad. Instead, wrong-making features will be understood quite differently, for example, involving the intrinsic disvalue of certain attitudes expressed in our actions. And, in spite of the initial problem that Kant’s system seems to have with complicity, I will argue that there may be a way in which a Kantian can work out an account of complicity that is in some respects more plausible than alternatives. The aim is to develop an account of complicity that does not make it necessary for complicity that the complicit agent has made a causal difference to an outcome. This is often where Utilitarian accounts of complicity are taken to fall short. A Kantian account, however, can identify at least one wrong-making feature of complicity in the failure of self-esteem—in the failure to maintain conditions of responsible moral agency.

What is Complicity? To describe someone as ‘complicit’ has negative connotations. One is complicit in wrongdoing, one is not complicit in doing good. It doesn’t follow from this that complicity in wrongdoing is always wrong, or criticizable. Suppose that Melissa lives in a borderline oppressive society and has the following options: She can vote for candidate A, she can vote for candidate B, or she may not vote at all. Let’s suppose that candidate A is clearly corrupt and would implement terribly unjust policies; candidate B is much better, though would engage in some unjust policy making. Let’s also stipulate that voting for A is worse than voting for B, and doing nothing at all is worse than voting for B, since it makes A’s election more likely. Here, Melissa cannot avoid complicity in some wrongdoing. What should be guiding her is the intention to minimize wrongdoing. Opting for bystander status does not in itself rid one of complicity. If one has acted to minimize wrongdoing, however, even if one is complicit in some wrongdoing, one is not criticizable. Complicity should be distinguished from other forms of wrongdoing. There are two kinds of complicity, broadly speaking:  participation complicity and tolerance complicity. When one participates in wrongdoing with others one engages in participation complicity; when one tolerates the wrongdoing of others one engages in tolerance complicity. Because ‘complicity’ involves some relation to the wrongdoing of others, it would be odd to describe a lone murderer as complicit in wrongdoing, since he directly and actively engaged in the wrongdoing by himself. We can thus note a distinction

258  Julia Driver between the actor who is primarily responsible for wrongdoing, and someone who helps or tolerates the actor’s wrongdoing. Often, those who are complicit in wrongdoing are either accomplices or bystanders. However, there is also a more attenuated notion of complicity in which members of groups or collectives are complicit in wrongdoing or in generating harm, even in the absence of a primary actor. Thus, we might argue, that everyone in the world who drives a gas-guzzling car is complicit in global warming. Most work on complicity has focused on legal complicity, and, as such, tends to tie complicity to lending aid in a causally efficacious way. We need to distinguish several cases to make this idea clear. One can be causally involved in wrongdoing without one’s participation being efficacious. This is typical of causal overdetermination cases. To be causally efficacious one’s actions or inactions must make a relevant difference to the outcome. Some writers, who are focused on legal issues, such as John Gardner, restrict complicity to the causally efficacious.4 However, writers such as Christopher Kutz have noted that one can be at least morally complicit, even when one’s aid is not causally efficacious.5 In this chapter we are considering moral complicity, and there is a good deal of intuitive support for the view that moral complicity need not be causally efficacious, nor need it even be the case that the complicit agent must believe or expect her participation to be causally efficacious. Indeed, in Kutz’s view this is a major problem for accounts of complicity that are Utilitarian, or, more broadly, consequentialist. This is because the consequentialist seems committed to holding that the only acts of complicity that are blameworthy must be actually harmful—that is, they must cause harm. Kutz believes that consequentialists, at least direct consequentialists, are committed to the Individual Difference Principle which holds: “I am accountable for a harm only if what I have done made a difference to that harm’s occurrence . . .”6 This runs counter to views about particular cases of participation in harms—cases that are clear cases of complicity in harm. One case Kutz discusses is that of a carpet bomber whose bombing makes no difference to an outcome, let’s say, because there are so many other bombers involved that the harm is overdetermined. Even supposing that he made no difference at all, he is still complicit. Kantian accounts of complicity hold the promise of providing a principled way of divorcing the blameworthiness of the action from the generation of harm—or even the mere prospect of a generation of harm. Thus, it seems that Kant’s framework might be better able to account for complicity, even if some of his specific verdicts seem to render complicity a problem.   “Complicity and Causality,” Criminal Law and Philosophy, 1 (2007), 127–41.   While Kutz ends up rejecting a Kantian account of complicity, he does so out of a strict interpretation of application of the Categorical Imperative, and we are considering abandoning this approach. A “Kantian” approach is still available, and at least one writer concerned with complicity in collective wrongs, David T. Schwartz, thinks a broadly Kantian view to have promise. See his Consuming Choices: Ethics in a Global Consumer Age, particularly ch. 4. 6   Complicity (New York: Cambridge University Press, 2000), 116. 4 5

Kantian Complicity  259 Consider the following case: Robert and his ex-wife, Maria, are involved in a very bitter custody dispute. Maria has recently decided to remarry, to Albert, and Robert is convinced that her increased financial stability will lead to her winning the custody dispute. He decides to kill Albert, thus eliminating Maria’s perceived advantage. He approaches his friend Mark and asks him for a ride to Albert’s, making it quite clear that he intends to kill Albert. Mark knows that if he doesn’t give a ride to Robert, it will end up making no difference to Albert since Robert will just walk a few blocks over to Albert’s and kill him anyway. This is like Kutz’s overdetermination cases in that there are many ways for Robert to easily get to Albert’s house. Mark’s presence, and his willingness to drive Robert, make no difference to the relevant outcome (i.e. that Albert is dead). Clearly, if Mark gives Robert a ride to the murder scene, he is complicit in the murder. Intuitively, this holds true even assuming that the ride made no difference to the outcome. Thus, accounts of complicity that limit it to either causally efficacious aid or aid that is thought to be causally efficacious, are too narrow to capture the intuitive notion. The question then arises, what makes the complicity wrong in these sorts of cases? Note that causal contribution of a sort is not at stake—there is clearly causal impact, just, Kutz claims, not of the relevant sort. On Kutz’s view, complicity is captured in the following way: I am accountable for what others do when I intentionally participate in the wrong they do or the harm they cause . . . I am accountable for the harm or wrong we do together, independently of the actual difference that I make.7

Kutz is primarily interested in an account of complicity that would handle individual contributions to collective wrongdoing—e.g. my driving my car more than I need does not in itself cause global warming, though it is part of a pattern of collective action that does. Thus, he seeks an account of complicity that is causally neutral. That is, on his view, one can be complicit in wrongdoing even if one’s participation made no causal difference whatsoever. The view Kutz is advocating takes intention to be key. However, it seems to me that not only must the complicit individual intentionally participate in the wrong, they do so under the description of the action or project, which renders the action, or project wrong. Thus, they intentionally and knowingly participate in the wrong. Consider the following case: Sam intentionally returns Robert’s hunting rifle to him, at which point Robert uses the hunting rifle to kill a rich relative in order to inherit the relative’s money. If Sam knows (or, more weakly, should know) that Robert would do this, he is complicit, otherwise not.

 Kutz, Complicity, 122.

7

260  Julia Driver We also need to distinguish providing a standard for complicity and accounting for its wrongness. Many may agree on Kutz’s standard for example, but strongly disagree on why complicity is wrong. Further, the account of its wrongness needn’t appeal to a single factor. It may be that one wrong-making feature is the production of bad effects, or the prospect of bad effects. Given one is rejecting consequentialism, there may be others, indeed, there may be very many other wrong-making features. One could plausibly argue that in the case of causally inefficacious complicity the badness has an expressive component. In failing to stand up to evil the evil actor may take the failure as a kind of endorsement, a kind of passive expression of endorsement. Further, Kutz’s account will have trouble with certain sorts of complicity relating to bystander ethics. These are cases where people are complicit in the oppression of others out of a studied indifference, or a negligent failure to notice the signs that something very wrong is happening.8 This is partly because Kutz views complicity as a kind of participation in wrongdoing; his account is geared towards participation complicity. Tolerance complicity does not involve participating in a wrong. Tom Hill considers this sort of complicity when he discusses bystander ethics. On his view we have responsibilities to cultivate dispositions that militate against this sort of tolerance complicity. For example, we have a responsibility to learn to notice things about our social environments that indicate oppressive social practices, such as noticing when police seem to be pulling over members of one ethnic group disproportionately. Moral insensitivity involves a failure to see things that are morally significant. Sometimes this failure is willful or self-deceptive, as when one is always ready with excusing explanations of the anomalous behavior. We have a healthy desire to avoid unpleasant confrontation with others, and studied indifference feeds into this desire. Unfortunately, it allows social predators latitude. For this sort of tolerance complicity, in which primary wrongdoers depend upon lack of awareness and indifference, Hill’s suggestions would be useful in avoiding complicity. So, there is a kind of complicity that involves moral negligence in failure to develop virtues of moral sensitivity. Consider the regular complicity problem we began with. A standard maneuver is to argue that Kantians need to reject absolutism, though, technically, it isn’t the absolutism so much as the constrained nature of Kant’s recommended maxims that really causes the problem. On this view, also suggested by Hill (amongst others), allowing for maxims that are context-sensitive would eliminate absurdity and still be in keeping with the formal nature of Kant’s approach. I think this is correct. However, I think that another strategy is available to the Kantian (and nothing rules out both strategies being operative). The second strategy exploits a feature of complicity that is often overlooked. 8   On pp. 155–65 of Complicity, Kutz notes that intentional participation is too strong a requirement, and that some forms of participation in wrongdoing involve agents who are not fully aware of the ramifications of what they are doing. However, the focus here is still on participation in wrongdoing, even if the sense of “participation” is more attenuated.

Kantian Complicity  261 Complicity is relational. An action or omission can be complicit with respect to one instance of wrongdoing while undermining (and thus failing to be complicit) in another. The absolutists who would avoid dirty hands really cannot avoid it on this view. To modify a case used by Thomas Nagel, if Anna refuses to encourage Hannah to twist the arm of a terrorist’s child, in order to frighten the terrorist into revealing the location of a bomb, then she is complicit in the terrorist’s wrongdoing, even if she strongly disagrees with the terrorist and disapproves of what the terrorist is doing. Complicity does not require agreement. However, she avoids being complicit in the wrongdoing of harming the child. Thus, in dealing with the complicity problem a Kantian could, at least in some cases, bite the bullet and hold that, for example, in lying to a prospective murderer Sally wrongs the murderer, but avoids the more serious offense of complicity in the prospective murderer’s wrongdoing. Further, imagine a mixed complicity case: Sally is standing next to Roger, who she knows is lying to the prospective murderer, and she does nothing to stop him, she does nothing in aid of the murderer. She may be complicit in the undermining of the prospective murderer, but thereby avoid complicity in harming an innocent person. This analysis has problems in that it fails to account for some of Kant’s views regarding the impossibility of genuinely mixed cases. But the Kantian has to give up something—the real issue has to do with what she gives up. Sally is a bystander whose inaction is causally efficacious. But often bystanders may court complicity even when there is nothing they can really do to make a material difference. Avoidance of complicity is important as an expression of respect and esteem for others; but also as an affirmation of one’s own value commitments (or value commitments one ought to have). Thus, an issue for Kutz’s account will be those cases of tolerance complicity that do not involve participation in wrongdoing. And these cases will be the best cases for a Kantian account. To account for the wrongness of complicity the Kantian needs to invoke an attitude that is at odds with the sorts of attitudes demanded of morality. That is, the complicit agent who tolerates wrongdoing is failing to have the right attitude—and attitude that is crucial to moral action, inaction, and simply being a morally good Kantian agent. This brings into play a consideration of the sorts of attitudes needed to effectively display respect for self and others. Kant does not tie this effective display necessarily to outcomes. This does not mean outcomes are always irrelevant; it just means that someone can be wronged even if not actually harmed, though the nature of the wrong will vary depending on whether or not there is harm. So, if Mark drives Robert to Albert’s house, and Albert is not home, then Mark still has wronged Albert even if Albert never suffers any ill-effects from Mark’s action; in the same way, Mark may wrong Albert even if—though Albert suffers harm—the harm is not dependent upon Mark’s acting a certain way. When Hill discusses bystander ethics, he discusses the sorts of attitudes that moral agents need to cultivate to avoid complicity. Hill writes in the context of articulating a duty to resist oppression—even bystanders have a duty, and this duty involves

262  Julia Driver cultivating the ability to see the signs of oppression.9 On Hill’s view, respect is of central value, and a concern for happiness is one way of manifesting respect. But “respect” is a notion that can be unpacked in ways that merely demand that we refrain from harming others ourselves, or in ways that add something more positive, including a duty to help others maintain the conditions of autonomous agency. Hill’s own view ties respect to a recognition of the rights of rational beings. Of course, rights in turn can be given more and less demanding treatments. On the more demanding conception of “respect” it isn’t enough to simply refrain from harming someone, or refrain from other forms of damaging behavior. On the more demanding view one fails to respect another rational being when one fails to help that being maintain the conditions that render her worthy of such respect. For a starving person, this means providing food or the means to procure food. Onora O’Neill had written convincingly on this topic.10 For the oppressed it means providing them the means to resist the oppression, and to even resist that oppression oneself, as a bystander. To fail in this is to be complicit in the oppression. Hill argues that not only do the oppressed have a duty to resist their oppression, but bystanders have a duty to resist said oppression as well. This translates into a duty to avoid complicity in oppression.

Bystanders and Complicity What is a bystander? Paradigm examples of bystanders are those who, for example, stood by while Nazis killed Jews during the Second World War; those who stood by during segregation and witnessed abuses to others and did nothing; those who stood by while women were systematically discriminated against, and did nothing. These are the cases Hill brings up in discussing the duties of bystanders to resist oppressive social practices. The characteristics he identifies are not necessary for bystander status. Hill believes, though, that they characterize most bystanders. He goes on to note the following features that (typical) bystanders have: (1) they do not intend to harm, exploit, or oppress anyone; (2) they are either unaware of the wrongful oppression or they think that they have no responsibility to do anything about it; (3) there is in fact serious oppression in the situation all around them; (4) they do have at least forward-looking (first-order) responsibilities to try to eliminate, oppose, diminish, and/or ameliorate the effects of the oppression; (5) they see themselves as basically people of good will, well-meaning people who would not tolerate atrocities, gross injustices, or systems of abuse in their communities, at least not if they had any legitimate and effective way to prevent or stop them; (6) for the most part, they are in fact fundamentally people of good will who would not knowingly and entirely willingly tolerate preventable atrocities, gross injustices, and systems of abuse in their communities; and (7) among the things that keep them from a more realistic,   “Moral Responsibilities of Bystanders,” Journal of Social Philosophy, 41 (2010), 28–39.  “The Moral Perplexities of Famine Relief,” in Tom Regan, ed., Matters of Life and Death (Philadelphia: Temple University Press, 1980), 260–98. 9

10

Kantian Complicity  263 painful, and effective awareness of their actual situation and first-order responsibilities are reluctance to disturb a familiar way of life, aversion to effort and risk, habits of making excuses and blaming others, a public culture that masks oppression, the influence of primary oppressors who intentionally put a favorable “spin” on its symptoms, the support of other self-justifying bystanders, and other social conditions that promote conformity and undermine confidence in one’s ability to think for oneself.11

Everyone agrees that what the oppressors themselves do is wrong and they have a duty not to oppress others. What Hill is concerned with is the more contentious issue of what, if anything, bystanders have a duty to do in these situations, and how that can be understood within a Kantian framework. Kant clearly holds that one ought not harm another, and Kant clearly holds that we have duties to the self. But, just as clearly, Kant seems to also hold that there is no duty to undermine the immoral agency of another. So, this makes a duty to resist oppression more problematic—it isn’t like just any duty to help others, the imperfect duty of benevolence. Giving money to the poor is not undermining anyone’s agency. But lying to an oppressor, or restricting the movements of an oppressor, or otherwise interfering with them does seem to undermine agency in ways that Kant seemed to disapprove of. Thus, if a Kantian, using some of Hill’s insights, can provide an alternative Kantian analysis of what bystanders do wrong when they don’t intervene, one that does not commit the Kantian to “negative responsibility,” that will be quite significant. Again, Hill’s remarks on Kant’s inflexibility and the duties we have as bystanders allow such an account. Hill notes that Kant’s theory does not support his absolutism and extreme position on lying and obedience to the law.12 He writes, “. . . a full and honest articulation of the maxim behind many conscientious lies would be more subtle and context-sensitive than those mentioned in Kant’s examples.”13 Further, when we consider the full range of Kant’s thinking on the duty to honor rational nature, Hill notes a very substantive account of positive duty in Kant, that is, putting a premium on this respect “should also lead us . . . to give positive support for institutions and practices that increase everyone’s opportunities to live as rational, free, persons.”14 These give the Kantian enough resources to argue for active engagement against oppression even on the part of mere bystanders (rather than oppressors and victims), and the resources to argue that failure to resist can make one complicit in wrongdoing. Consider a case from Emerson, the “withdrawing” citizen.15 When citizens in an oppressive state remain unreflective about the oppression around them, when they   “Moral Responsibilities of Bystanders,” Social Philosophy and Policy, 41 (Spring 2010), 30.   See his “Happiness and Human Flourishing in Kant’s Ethics,” Social Philosophy and Policy (1999), 152–4. 13   Hill, “Happiness and Human Flourishing in Kant’s Ethics,” 154. 14   Hill, “Happiness and Human Flourishing in Kant’s Ethics,” 155. 15   This case is discussed by Alex Zakaras in Individuality and Mass Democracy: Mill, Emerson, and the Burdens of Citizenship (New York: Oxford University Press, 2009). Zakaras explicitly ties being a good citizen of a democratic state to avoidance of complicity. The virtue of individuality involves citizens’ exercising 11

12

264  Julia Driver simply conform to the dictates of the oppressors, they are complicit in the oppression.16 The example Emerson used was the failure of many citizens to respond appropriately after the Fugitive Slave Law was passed in 1850. This law required the return of runaway slaves and stripped free African Americans of many legal rights (e.g. to contest, legally, a claim of ownership). Emerson was aghast at the response of his fellow citizens in Massachusetts. They should have revolted, but instead were largely silent, and thus complicit. In an address in 1851 he said: Every hour brings us from distant quarters of the Union the expression of mortification at the late events in Massachusetts, and at the behavior of Boston. The tameness was indeed shocking. Boston, of whose fame for spirit and character we have all been so proud; Boston, whose citizens, intelligent people in England told me they could always distinguish by their culture among Americans; the Boston of the American Revolution, which figures so proudly in John Adams’s Diary, which the whole country has been reading; Boston, spoiled by prosperity, must bow its ancient honor in the dust, and make us irretrievably ashamed. . . . The tameness is indeed complete.17

The factors that Hill identifies when delineating bystanders are telling, particularly the last factor regarding the promotion of conformity. In the rest of his address Emerson notes that people he had thought of as intelligent, reflective, and responsible seemed, after the passage of the act, to simply conform to the dictates of their political party. One of the deleterious effects of conformity is the resultant loss of habits in thinking for oneself. Thinking for oneself is important for integrity and self-respect. Kant regarded all rational persons as having the capacity to understand norms and conform to them—but actively. That is, for an agent to genuinely be acting morally the agent subjects herself to the norm. Agents who simply follow the lead of others on moral matters risk undermining habits that are important to moral agency. Indeed, this is what partly undergirds worries in general about moral deference. One thing crucial to freedom is the ability to think for oneself. Agents recognize the value of setting ends, and the rational pursuit of ends, not just in ourselves, but in others.18 We can justify a presumption that practices that interfere with agential freedom should be resisted—and should be resisted even if the resistance isn’t causally efficacious. But one of the enduring themes of Hill’s works is that, in Kantian terms, we should be mindful of what the dignity of this freedom demands of us when it comes to our

their critical and reflective capacities in such a way that they “work to ‘detach’ themselves from complicity by speaking and acting bluntly against the unjust uses of power” (97). 16   Alex Zakaras in “Complicity and Coercion: Toward an Ethics of Political Participation” (unpublished manuscript) also discusses the complicity issues that arose around the passage of the Fugitive Slave Act, focusing on the work of Henry David Thoreau. Zakaras notes that Thoreau viewed American citizens—in virtue of their citizenship and in virtue of regarding the American government as authoritative for them—as more than mere bystanders. 17   Ralph Waldo Emerson, Address to the Citizens of Concord, May 3, 1851. 18   Thomas Hill, “Humanity as an End in Itself,” in Dignity and Practical Reason in Kant’s Moral Theory (Ithaca, NY: Cornell University Press, 1992), 54.

Kantian Complicity  265 attitudes towards the self, not simply to others. “Servility, as is often revealed in groveling, flattery, simpering, and self-disparagement, is condemned because it symbolizes an attitude which does not place the dignity of one’s own humanity above all price.”19 Consider Emerson’s charge against the citizens of Boston again. Emerson’s shock was in part due to the fact that many had expressed—sincerely—a commitment to freedom and yet did not speak out when the Fugitive Slave Law was passed—and, not only passed but actively implemented by the state of Massachusetts, resulting in widespread harms against African Americans who had found what they thought was freedom in Massachusetts. The failure was one of a failure to stand up and speak out against gross injustice. It was demeaning to those who were complicit in this way; it was a kind of servility to the state that is unbecoming free citizens of a democratic state. This would be the case even if standing up and speaking out changed nothing, materially. This is one form of tolerance complicity. Of course, tolerance complicity also occurs in cases where the complicit agents actually share the values embodied in the activities that they are tolerating, but the wrong here isn’t as much of a puzzle. The wrong here has to do with shared ends that are intrinsically bad ends. One ought not value the subjugation and humiliation of other people. Tolerance complicity in which the ends are not shared, however, is still blameworthy for what it reveals about the agent’s attitudes and commitments to her own set of values. Thus, being complicit in wrongdoing does not mean that one approves of the wrongdoing, just that one is not going to challenge it. Often people—due to factors beyond their control, even—end up unavoidably involved in events, and then the relevant issue is how to be involved. The parable of the Good Samaritan illustrates how simply feeling that something is not one’s business doesn’t absolve one of involvement, it is just that one has made a choice to be involved in a way that does not require some kind of interference.

More on Tolerance Complicity The Kantian should make most of the intuition that tolerance complicity is, pro tanto, wrong even when causally inefficacious. Some reasons that can be cited are familiar from the anti-consequentialist literature. For example, tolerance complicity seems to pose problems for personal integrity, and to the extent that we think integrity is a virtue, we should find this sort of complicity morally problematic. A person who fails to stand up for her values is lacking in integrity. But this intuition needs much more unpacking. What is it about tolerance complicity that, within a Kantian framework, makes it problematic in this way? Many writers on Kant find the notion of self-respect crucial to understanding his moral system. Recall that on the Kantian view, self-respect is very minimal. In accounting for some of the peculiar wrong-making features of complicity

19

  Hill, “Humanity as an End in Itself,” 55.

266  Julia Driver we can go further than this and also argue that self-esteem, not just self-respect, is important. Self-esteem involves seeing a special value of some sort in oneself. It requires integrity. All persons deserve respect, but not all deserve esteem. Given that failure to stand up for one’s values involves loss of integrity, there will be a corresponding loss of self-esteem.20 But not just any failure to stand up for values leads to complicity. A very important factor has to do with one’s audience, and what the audience makes of silence, or what the agent can reasonably expect the audience to make of the silence. As noted earlier, some might view silence as a kind of assent. If that is the case, the agent is at risk of allowing herself to be turned into, and used, as an example of someone who agrees rather than disagrees with values she in fact does not hold. The consequentialist will view this as a bad-making feature of silence, because it can lead to a loss of collective inhibition against doing wrong.21 A different approach would focus on the issue of “allowing oneself to be used” to provide cover for those one disagrees with. And this is in keeping with a Kantian approach. When one is participating, and complicit as a participant in wrongdoing, the wrong itself is directly attributable to the participant. When one is silent, the wrong instead is dependent in part on how the audience takes the silence. Even when speaking out doesn’t change anything, if one is silent then there will be many situations where the silence allows the members of the audience to presume agreement, and to use the silence as a sign that their position is persuasive. Thus, in 1851, one reason Emerson was upset with those who remained silent (though I am sure he also felt that speaking out would have been causally efficacious) was that they were allowing others to presume their agreement, and thus be emboldened. It is the allowing oneself to be used that is objectionable—as in the case of Hill’s servile wife. But this is even more extreme—one is allowing oneself to be used to ends that one does not share.

20  For more on distinguishing self-respect and self-esteem, see David Sach’s “How to Distinguish Self-Respect from Self-Esteem,” Philosophy and Public Affairs, 4 (1981), 346–60. 21   See my “Caesar’s Wife: On the Moral Significance of Appearing Good,” Journal of Philosophy (1992).

PA RT V

Conclusion

14 Looking Back Main Themes and Appreciation Thomas E. Hill, Jr.

The contributors to this volume have written excellent original essays that deserve attention on their own, and I am honored and grateful that the authors engaged critically with many of the themes in my work and were willing to include their own essays in this volume. I am grateful too to the editors, Mark Timmons and Robert Johnson, for proposing and putting together this collection. Instead of giving a detailed response to each chapter, I want primarily to take this opportunity (as suggested by the editors) to review the background and main themes in my essays as they have developed over time. This may help to provide some context and continuity for what may otherwise seem a rather disconnected series of essays. By giving some autobiographical background I can at the same time disclose some of the early influences on my thinking and express my appreciation for those who were most helpful and inspiring. My continuing study of philosophy spans many decades and so my retrospective will inevitably reflect my view of some of the major changes in moral and political theory over that time. By reviewing themes in my work so far I can not only provide some context for the chapters in this volume but also call attention to issues that I addressed beyond those on which the contributors chose to focus. I comment only very briefly on the particular essays collected in this volume, acknowledging that this does not give them the full attention that they deserve. My main aims here are to highlight some central themes in my work and then to comment briefly on the contributors’ chapters, suggesting some points that we might consider in further discussions.

A Brief Autobiographical Sketch My father (also named Thomas E. Hill) was a professor of philosophy at Macalester College for most of his professional life. He wrote five books and taught a wide range of philosophy courses, but his primary interest was ethics. Not surprisingly, then, I was introduced to philosophy—and came to love it—when I was still in high school

270  Thomas E. Hill, Jr. (1951–5). At that time my father accepted G. E. Moore’s “ideal utilitarianism,” and, as he explained and used the theory, I initially found it persuasive. The basic idea that we ought to promote the most good possible for all concerned seemed a high-minded and flexible alternative to the social and religious norms that often seemed arbitrary and unduly restrictive. The theory was non-hedonistic and called for appreciation of “intrinsically valuable experiences” of all kinds. Moore’s principle of organic unities fit well with reasonable doubts that we can add and subtract bits of intrinsic value in order to determine what we ought to do. According to Moore’s principle, the intrinsic value of a whole is not necessarily the sum of the intrinsic value of its parts. Because of this, Moore concluded, what is objectively right depends ultimately on whether our acts will help to bring about the most intrinsically valuable universe. Critics often charge that utilitarianism would recommend lying, killing, and punishing the innocent in cases where this is counter-intuitive, but Moore’s theory seemed wonderfully flexible regarding particular cases. Apparently we could always judge that in the long run the universe will be intrinsically better if we respect the usual moral constraints. Empirical facts cannot prove otherwise. A  major problem, of course, is that Moore’s theory is too flexible because he takes intrinsic value to be a simple, indefinable, non-natural property that is known only by intuition—and intuitions can vary widely and reflect blindly accepted individual and social biases. Some biases can be tentatively identified but there is no standard beyond intuition to rely on. As an undergraduate at Harvard, after a brief flirtation with pre-medical study and despite a competing interest in history, I settled on philosophy as major. Certain philosophy professors were especially helpful and influential, and I remain grateful to them. I had tutorials with Burton Dreben, Maurice Mandelbaum, and Stephen Barker, and courses from Dreben, Raphael Demos, Roderick Firth, Rogers Albritton, Henry Aiken, D. C. Williams, and others. I particularly remember sophomore tutorials with Dreben, which consisted of fast-paced walks around the Yard during which he pressed me to explain how “Murder is wrong” is different from “You ought not to eat peas with your knife.” Also memorable were junior tutorials with Maurice Mandelbaum in which he walked me through his then-new Phenomenology of Moral Experience. At my father’s urging I wrote a senior thesis in epistemology, which focused on objections to sense data theories of perception. This probably contributed to my later skepticism about the alleged intuition of intrinsic value as presented by G. E. Moore and C. I. Lewis. During college years I participated in two varsity sports (football and wrestling) and this no doubt helped me to win a Rhodes Scholarship which enabled me, following graduation, to study three years at Oxford at a time of extraordinary intellectual excitement. J. L. Austin, Gilbert Ryle, J. O. Urmson, Geoffrey Warnock, and R. M. Hare encouraged the close study of ordinary language, partly for the sake of resolving or dissolving traditional philosophical problems. Moral intuitionism, represented by the works of G. E. Moore and W. D. Ross, seemed to be thoroughly rejected by all, for various reasons, many of which were subtly expressed in Peter Strawson’s classic paper on

Looking Back  271 the subject.1 R. M. Hare presented his universal prescriptivism in metaethics, sharply opposed by Philippa Foot and Elizabeth Anscombe. My special topics for B.Phil. study were ethics, theory of knowledge, and (as urged by Ryle, my supervisor) Immanuel Kant. My B.Phil. thesis was a critique of ideas of intrinsic value in C. I. Lewis and G. E. Moore. This represented a sharp break from the value theory that I had earlier taken for granted, and no doubt it left me more receptive to alternatives that I found in the works of Kant and (later) John Rawls. A persistent theme throughout my work has been that claims that something is valuable are essentially claims about attitudes that we have reason to adopt and act on and, further, that particular reasons of this sort ultimately depend on general principles of rational choice. After completing the B.Phil. I  spent a third year at Oxford working with J.  O. Urmson and William Kneale but primarily focusing on political philosophy, which included attending the marvelous lectures of Isaiah Berlin, John Plamenatz, and H. L. A. Hart. Advice from many sources, especially Urmson, recommended returning to the US for the Ph.D., and in particular trying to go wherever John Rawls was going to be teaching. That happened to be Harvard, and I was pleased that I was able to return there and work under Rawls’s supervision for two and a half years. The influence of his lectures, even in that early period (1962–4), was powerful and lasting, I think, but later when I tried to express gratitude Rawls brushed it off, insisting that we cannot know what ultimately shapes our thinking. He was not showing false modesty or (worse) a desire to disassociate from me, I assume, because he followed the remark by explaining that he felt sure that the tragic deaths of his brothers had influenced his work in some way but he could not say how. Like many others, I admired Rawls as a person and found his attitude towards philosophical inquiry inspiring. He had extraordinary respect for the history of moral and political philosophy and saw his and our task as continuing in its traditions. He studied and was inspired by systematic moral and political theories, as is evident in his lectures on Kant, Hegel, and Bradley in an era when these systematic thinkers were commonly dismissed by analytical philosophers. He could simplify ideas for methodological reasons and for clarity of exposition, but he would not compromise his theoretical project for popularity as a “public intellectual.” He had faith that serious engagement with deep issues in moral and political theory can ultimately make a real difference in the world, but he thought that readers should focus on an author’s ideas rather than on the author. A common theme was that moral theory is hard and still underdeveloped. All existing theories are imperfect, and what we need now is not more quick counter-examples and replies but serious efforts to construct better alternatives, taking cues from our forebearers. Rawls saw philosophy as a cooperative and constructive enterprise, and a major aim was to give a structure and defense to plausible commonsense ideas. 1  P. F.  Strawson, “Ethical Intuitionism,” Philosophy (1949), reprinted in Philosophical Writings (Oxford: Oxford University Press, 2011).

272  Thomas E. Hill, Jr. Rawls urged me to continue my work on Kant, in part perhaps as an antidote to the excessive concern with ordinary language that was prominent at Oxford. In a debriefing session after the defense of my Ph.D. thesis (1966), he compared doing systematic moral theory to painting a mural. One cannot make progress by painting small bits in great detail before first blocking out the whole in large strokes in charcoal, then revising and revising until the big picture becomes clearer. It was good advice, I thought, even though initially I found myself more suited to smaller scale projects. There were other distinguished teachers at Harvard then and an incredible collection of impressive graduate students who went on for careers in philosophy, including, for example, Sharon Bishop, Claudia Card, Ted Cohen, Alan Fuchs, Paul Eisenberg, Gilbert Harman, J.  Charles King, Edmund Leites, David Lewis, David Lyons, Tom Nagel, Onora O’Neill, T. M. Scanlon, Robert Shope, Michael Slote, Steven Smith, Lawrence Stern, Michael Stocker, Louis Werner, Roger Werthheimer, and Margaret Wilson. I did not know them all well but was inspired by their seriousness, intelligence, and philosophical spirit. Over many years of teaching I have learned from inspiring colleagues and friends, too many to acknowledge fully. At UCLA in moral and political philosophy these included at various times Robert Adams, Bernard Boxill, Philippa Foot, Jean Hampton, Arnold Kaufman, Gregory Kavka, Herbert Morris, Warren Quinn, John Taurek, and Richard Wasserstrom, as well as visitors Angela Davis, Norman Dahl, Robert Solomon, Jeffrie Murphy, and others. John Perry, Tyler Burge, and Herbert Morris were especially encouraging. Burge helped me to develop my understanding of Kant’s argument for the rationality of moral conduct. Kavka gave me excellent comments on all of my early papers, and Robert Adams offered new and different ways of looking at familiar questions. I found Morris’s exploratory style liberating, and I admired his work on guilt, shame, and punishment. He was a trusted advisor, a good friend, and (I can now admit) a superior tennis player. During my years at UNC Chapel Hill I have had excellent colleagues representing quite different perspectives in moral and political philosophy—Robert Adams, Simon Blackburn, Jan Boxill, Bernard Boxill, Geoff Brennan, Michael Corrado, David Falk, Douglas Long, Douglas MacLean, Geoffrey Sayre-McCord, Gerald Postema, Ryan Preston-Roedder, Gregory Trianosky, and Susan Wolf. Working on joint projects with Bernard Boxill, Arnulf Zweig, and Adam Cureton has been especially rewarding. These and other colleagues, students, and friends at other universities have inspired and enriched my life professionally and, even more, personally.

Respect and Self-Respect Self-respect and respect for others have been recurring themes in my work from the earliest years. These are obviously central concerns in Kant’s ethics, but I have tried to highlight their importance in connection to other topics, including personal

Looking Back  273 autonomy, symbolic protests, weakness of will, affirmative action, suicide, punishment, bystanders to oppression, and John Rawls’s theory of justice. “Servility and Self-Respect” is representative, illustrating how questions about respect for oneself and others are connected to other practical and theoretical issues. That early essay (1973) has had a surprisingly long shelf-life, and I have been pleased that other authors have taken up the topic, moving the discussion forward. The essay was prompted by a convergence of old and (then) new interests. My interest in broadly Kantian themes is evident, even though the essay was not an attempt to interpret Kant’s own account of servility. My idea of servility was based on a premise about the equality of moral rights whereas Kant treats servility in his Doctrine of Virtue, which is concerned with equal respect for humanity but not explicitly with equal rights. I was also intrigued by contemporary controversies about the idea of duties and obligations to oneself. Marcus Singer and others quickly dismissed this idea with arguments that seemed persuasive but too facile.2 Over time I came to see, as I argued later in “Promises to Oneself,” that the critics were right about certain conceptual points, but that they failed to capture the moral significance of serious claims expressed in terms of duties and obligations to oneself. Besides these more theoretical interests, I was moved and puzzled by the contempt that feminists often showed for the stereotypical “happy housewife” and that progressive African Americans often showed for those they regarded as “Uncle Toms.” In addition, John Rawls made prominent use of an idea of self-respect in A Theory of Justice (first published in 1971), but Rawls, in contrast to Kant, was concerned with the causal effects of social institutions on the basic self-esteem needed for happiness, rather than an ethical duty to act in a self-respecting way.3 These interests came together in my argument for the limited claim that, given certain specified definitions and assumptions, we have a duty to ourselves to be self-respecting and therefore to avoid servility, understood as failing to affirm one’s equal moral rights either from ignorance or from caring too little. “Servility and SelfRespect” and several of my others essays started from carefully designed examples that I hoped would stimulate interest but that were not intended to represent everyone who might be called by my labels, “Uncle Tom,” “Deferential Wife,” and “Self-Deprecator.”4 The aim there and elsewhere was not to give a psychological profile of personality types or to take a stance about how often real people approximated my paradigms. The cases were partly constructed as a response to Marcus Singer’s argument for rejecting “duties to oneself ”: if A has a duty to B, then B has a right against A; if one has a duty to oneself,   M. G. Singer, “Duties to Oneself,” Ethics, 69 (1959), 202–5 and Generalization in Ethics (New York: Alfred A. Knopf, 1961), 311–18. 3   “Stability, A Sense of Justice, and Self-Respect,” A Companion to Rawls, ed. Jon Mendel and David Reidy (New York: Wiley-Blackwell Publishing), ch. 11. 4   In other essays too I used exaggerated paradigm examples to illustrate various objectionable attitudes, for example, “Amiable Amy” (in “Weakness of Will and Character”), “class snobs,” “ talent snobs,” “ achievement snobs,” and “moral snobs” (in “Social Snobbery and Human Dignity”), “impulsive suicides,” “apathetic 2

274  Thomas E. Hill, Jr. then one must have a right against oneself; but to suppose one could have a right against oneself is absurd. The underlying conceptual point in Singer’s argument was that one can violate the rights of others but not one’s own rights. My cases were meant in part to make vivid that one’s attitudes and behavior can fail to properly respect one’s basic equal rights even though it is conceptually impossible to violate one’s own rights in the same ways that one can violate the rights of others. You cannot literally steal from yourself, but you can fail to affirm and take up your rightful place in a system of moral equals as you should unless you have strong reasons, such as justified fear of reprisals, not to. The examples were meant to focus attention on possible cases where individuals let others violate, deny, or disregard their basic rights, and where this is intuitively objectionable—not necessarily blameworthy, but morally undesirable. The intuitively objectionable behavior and attitudes in my particular paradigm cases, I thought, could reasonably be described as servile and not fully self-respecting, though these words may also have other meanings. Acknowledging that it is not morally objectionable to subordinate oneself (e.g. by behaving and talking like a servile person out of reasonable fear of oppressors), I pictured my paradigms as subordinating themselves not from fear but instead from ignorance or caring too little for their equal rights. The ultimate causes of their not understanding and caring for their moral equality may well be unjust and oppressive historical and cultural conditions. This background explanation is plausible, but would not make them less servile and lacking in self-respect in my senses. Their behavior and attitude remain morally undesirable even though the servility may not be their fault and it may be more difficult for them than for others to stand up for their rights. The bottom line is not that the servile person should be blamed but that those who can must try to rectify the unjust and oppressive conditions that encourage it. Let me explain my exploratory methodology in this and related essays, that is, ­chapters 1, 2, 3, 5, 7, and 8 in Autonomy and Self-Respect. The idea was to focus on particular examples designed to isolate moral values and principles that seem not to be well explained by familiar theories. The challenge for me then was to articulate these values and principles, to show how they would help us to understand the examples, and finally to propose some reasons for accepting or at least not dismissing them too readily. Practical issues (race, gender, lying, symbolic protests, suicide, and preserving natural environments) were raised not as “applied ethics” (presupposed theories applied to practice) but as a way of exploring and clarifying values and principles that might be worth incorporating into moral theory. As I expected, the values and principles that emerged corresponded to some extent with a reasonably modified Kantian perspective, as I understood this. These early investigations—which I enjoyed more than much of my subsequent work—illustrated my belief that, in normative ethics, theoretical and suicides,” and “self-abasing suicides” (in “Self-Regarding Suicide: A Modified Kantian view”), “bystanders,” “oppressors”, and “victims” (in “Moral Responsibilities of Bystanders”), “Bad Bubba,” “Weak Willie,” and “Good Gracie” (in “Two Conceptions of Virtue”).

Looking Back  275 practical issues should be considered together and, further, that exploratory thinking can be just as important as the more customary attempt to give definitive arguments to nail down a point in whatever debates happen to be current. Later papers, some of which were more explicitly concerned with Kant’s ethics, also emphasized the role of self-respect and respect for others. “Self-Respect Reconsidered” called attention to a conception of self-respect as having and living up to personal standards, which is different from the conceptions of self-respect in Kant and in my earlier essay. The Tanner Lectures—“Basic Respect and Cultural Diversity” and “Must Respect Be Earned?”—were primarily about basic (“recognition”) respect for persons in different cultures, in penal institutions, and in university curricula. In the more recent essays “Treating Criminals as Ends in Themselves” and “Kant and Humanitarian Intervention” I attempted to show some practical implications of the related Kantian idea of human dignity (as I develop it). In “Donagan’s Kant,” I tried to emphasize admirable aspects of Alan Donagan’s work, but I was also critical of his attempt to treat respect for rational agents as the comprehensive moral standard from which all other moral requirements must be derived. The essay “Moral Responsibility of Bystanders” focused mainly on Kant’s treatment of second-order duties to scrutinize one’s motives for passivity, to exercise due care in deliberation, and to develop moral strength of will, all of which can be seen as requirements of self-respect. In the section of this volume labeled “Respect and Self-Respect” the contributors, Jan and Bernard Boxill, Robin Dillon, and Stephen Darwall, raise important critical questions and carry the discussion forward in new and interesting ways. Here I will only add a few comments to explain my current understanding of some of the issues. The Boxills have inspired and helped to shape my thinking about respect in the context of racial and gender issues for many years. In particular, from early years at UCLA I have learned from Bernard Boxill’s work on reparations and protest, and my essay “The Message of Affirmative Action” began as a response to his powerful book, Blacks and Social Justice. He and I have jointly taught several courses on Rousseau, Kant, and Rawls, and in this context I have had the benefit of his good-spirited challenges to my views. Jan Boxill has helped to make me more keenly aware of what respect means in sports and in social policies towards women. In their chapter for this volume the Boxills argue that ignorance of moral equality is rare, that my “Self-Deprecator” is not servile but just depressed, and that I led readers to believe that my “Uncle Tom” and “Deferential Wife” lived under domination but then denied that they acted from the fear characteristic of those who live under such conditions. There are substantive as well as verbal disagreements here that I leave to others to assess. The Boxills’ main point, however, is that the primary problem with the people who in real life resemble my paradigms and are suggested by the labels “Uncle Tom” and “Deferential Wife” is not that they do not know or care enough about their place in a system of equal basic moral rights. The problem, rather, is that the actual people suggested by my labels are dominated and live in fear of those who dominate them. According to the Boxills, these people are not unaware of their moral equality, but rather they behave in a deferential,

276  Thomas E. Hill, Jr. self-effacing way out of fear. What they need is not “instruction” in the equality of moral rights but removal of the system of domination. Whether intended or not, they argue, my labels “Uncle Tom” and “Deferential Wife” led people to assume that my essay was about the vast numbers of blacks and women who live under domination. As a result, they argue, the influence of “Servility and Self-Respect” has been to deflect attention from the more important problem of oppression—living under domination and needing to lower oneself out of fear. Whether or not my essay has had the unintended effects that they claim, they now quite reasonably focus attention more squarely on how to understand and address today’s persistent and urgent problems of race and gender relations. When that is the focus, although I am less confident of their sociology and republican theory of justice than they are, I share their conclusion that our primary social task is not to preach equal rights to those who appear servile but to try to correct unjust social conditions and attitudes. That said, I still maintain that those who satisfy or approximate my model of servility owe it to themselves to try to overcome their condition, even if their being servile resulted from oppression and their overcoming servility is not the most pressing social problem. Robin Dillon analyzes a cluster of related moral attitudes with her usual subtlety and thoroughness, with special attention to the details of Kant’s texts. In particular, she argues that Kant rightly held that self-respect, not humility, is the virtue that opposes arrogance and, further, that “humility is at best, an ancillary, instrumental, contextual virtue and the servant of self-respect; but at worst, it is as serious a vice as arrogance, indeed, an aspect of it” (p. 43). Although Dillon’s analysis and careful review of Kant’s texts are illuminating, I still think that there is a kind of humility, which I would call “proper humility,” that is a virtue incompatible with arrogance, but her discussion rightly highlights the need to say more about what proper humility is, why it is a virtue, and how it relates to arrogance. In the essay from which Dillon draws my idea of humility (“Ideals of Human Excellence and Preserving Natural Environments”) I simply appealed to an intuitive idea of humility and suggested that certain experiences with nature can be humbling in a good way, constraining our tendencies to evaluate everything by its relation to ourselves. Curbing our self-importance and learning to accept ourselves as natural beings in an immense and complex natural world, I suggested, may be a step towards proper humility in our relations with fellow human beings. This, in my view, is not an obsequious or servile attitude of inferiority but, to the contrary, presupposes an appreciation of one’s basic moral equality with others, as one among many in a worldwide community. One problem here is that words like “humility” can be used in different ways, and there is no Platonic form of humility by which to judge diverse uses. In moral theory, as I see it, we want to make our descriptions of putative virtues and vices understandable in ordinary language, but we should aim to define them carefully so that we can see that the virtues so defined are really excellences of character and the vices the opposite or contrary traits. For this reason, I think, theories of virtue need to specify what is “proper pride,” “courage as a virtue,” “the moral vice of weakness of will,” and so on, to focus sharply

Looking Back  277 on the traits we mean to assess and distinguish them from other traits (similar in some ways) that might be labeled as “pride,” “courage,” and “weakness of will,” etc., in ordinary usage and in dictionaries. Dillon does an extraordinary job of explaining humility, arrogance, and self-respect in the senses with which she and Kant may be working, and she rightly calls for those of us who disagree to explain what we have in mind. Anyone should agree that humility as it is often conceived can be a vice and that if one calls a trait “proper humility” and identifies it as moral virtue then one must define it so that it is clearly not a matter of lowering oneself by denying the equal worth of one’s humanity. In “Respect as Honor and as Accountability” Stephen Darwall expands upon his influential distinction between two kinds of respect. Appraisal respect is a kind of esteem for a person’s merits whereas recognition respect is not esteem for merit but acknowledgment of a person’s dignity or status. Here Darwall distinguishes two kinds of recognition respect—second personal respect for the dignity and moral authority of all persons and honor respect for some (but not all) persons for their socially constituted rank or status. He traces the history of these concepts, seeing honor respect as prominent in the honor culture of the ancien régime, and as having been partly replaced by Kant’s moral respect for the dignity of all persons (which Darwall treats as a kind of second person respect). The distinction between honor cultures and morality is familiar but has not been sufficiently explored by philosophers. Darwall’s contribution here is to clarify and highlight some distinct concepts and to locate them in his own account of moral authority and responsibility as dependent on a second-person perspective. He develops his moral theory (“the second person perspective”) elsewhere.5 This is important systematic work, more than I can address here, but I welcome and commend Darwall’s new investigation of a kind of respect beyond the basic distinction (recognition vs. appraisal respect) that he famously articulated in “Two Kinds of Respect.”6

Practical Reason Practical reason has been a prominent topic in the history of moral and political philosophy as well as contemporary discussions. My writing in this area was primarily the result of my attempt to interpret and extend Kant’s work but was further stimulated by reviewing Stephen Darwall’s Impartial Reason7 and by finding myself dissatisfied with other contemporary accounts of practical reason. Here I summarize some of the main themes I have tried to articulate over the years and then acknowledge briefly the relevant contributions to this volume by Mark Schroeder, Jonathan Dancy, and Onora O’Neill. 5   Stephen Darwall, The Second-Person Standpoint:  Morality, Respect, and Accountability (Cambridge, MA: Harvard University Press, 2000). 6   Stephen Darwall, “Two Kinds of Respect,” Ethics, 88 (1977), 36–49. 7   “Darwall on Practical Reason,” Ethics, 96 (April 1986), 604–19.

278  Thomas E. Hill, Jr. The core of Kant’s conception of practical reason is his distinction between hypothetical and categorical imperatives. In the early essay “The Hypothetical Imperative” (1973) I  presented an interpretation of this distinction, which Kant introduced to explain moral and non-moral practical judgments that we commonly express by “I ought to do X,” especially when this is understood strictly as “I must do X.” Just as he did in other areas of philosophy, Kant asked how we can understand various puzzling claims about what is necessary for human beings—necessary for understanding, knowledge, rational choice, and reasonable hope. In moral philosophy the main question was what is practically necessary to choose and to do, taking into account the inevitably limited perspective, fallibility, and weakness of human beings. Kant’s answer lies in his extended discussion of hypothetical and categorical imperatives. Among the main features of Kant’s position are these: First, judgments that I ought to do X purport to express imperatives, which are true or warranted claims about what it is rationally necessary to do expressed in the imperatival form (“a command of reason”) appropriate to imperfectly rational beings who can conform but might not. They are not identifiable by grammatical form and are not identical to empirical propositions of the form “In order to get E, it is causally necessary to do X.” Second, to support a particular ought-claim a rationale is needed that goes beyond the relevant empirical facts to explain the reasoning why we would take the prescribed course of action if we were sufficiently rational. In other words, particular ought-judgments presuppose a general principle of practical reason that, together with relevant facts, supports that conclusion. Merely citing empirical facts surrounding the case and calling them “reasons” will not by itself explain the normative judgment “I must (or ought to) do it.” Third, particular instrumental non-moral ought-judgments (“hypothetical imperatives”) presuppose a general principle of reason (that I called “the Hypothetical Imperative”) which is itself unconditionally rational but in application always leaves us options. Loosely expressed, the principle says that it is rationally necessary (and so one ought), if one has willed an end for which there are causally necessary means available and within one’s power, either to take the means or abandon (or suspend) the end. Fourth, understanding the requirements of non-moral reason as conditional in this way, Kant can maintain that moral requirements are unconditional (categorical) requirements of reason but nevertheless hold that they are always compatible with non-moral requirements of reason. Although morality may conflict with the pursuit of happiness, hypothetical and categorical imperatives are always compatible and there can be no conflict between fully rational prudence and morality. Obviously these claims need further explanation and qualification, and reasonable critical questions have been raised both about their basis in Kant’s texts and their philosophical defensibility. My later essays most relevant to reasons and rational principles do not directly revisit these issues, but argue for some further points that I will mention but cannot elaborate here. For example, “Reasonable Self-Love” contrasts a Kantian account of self-regarding and other-regarding reasons with other prominent theories and shows how the Kantian account, perhaps surprisingly, more

Looking Back  279 closely fits with commonsense beliefs about reasons. The essay “Pains and Projects” makes a case for thinking that physical and psychological conditions, such as being in pain, are not intrinsically bad in a sense that implies reasons to avoid them independently of ends that we will or would will if rational. We understandably call severe pains “bad in themselves,” but a deeper explanation is needed in order to explain why they are reason-giving. Several other essays, most recently “Kantian Constructivism as Normative Ethical Theory,” endorse the idea that moral reasoning about specific principles and potential exceptions can be understood as having a constructivist structure. Here, setting aside contemporary controversies about constructivism as a meta-ethical alternative to realism and expressivism, I try to show how Kant used constructivist procedures in normative ethics, even in developing his system of ethical principles in his later work The Metaphysics of Morals. A theme that runs through several essays, especially those on happiness, personal ends, and helping others, is that both morality and rational prudence leave us a wide range of permissible options in normal circumstances.8 Morality places strict boundaries on our conduct and prescribes positive attitudes and ends that we must adopt and incorporate into our life plans, but, except in extraordinarily challenging circumstances, moral requirements leave us ample room to adopt and pursue our own personal projects, and the justification for this is not merely that pursuing personal projects makes us better able to fulfill our duties to others. In addition, in Kantian theory, rational prudence and self-love normally also leave us a wide range of optional choices. There are no non-moral ends, such as one’s pleasure or desire-satisfaction, that a rational person must try to maximize. Those who are fully rational will fulfill moral requirements and adopt and revise their set of personal ends in an informed, consistent, and coherent way, but neither morality nor non-moral rational requirements dictate every choice or imply that there is always something one ought to be doing. Finally, in “The Importance of Moral Rules and Principles” I sketched the structure of a Kantian rationale for a particular non-moral ought-judgment and contrasted this with the structure of particular moral ought-judgments. In both cases, what I called a (full) rationale lays out explicitly (as in everyday life we never do) both the facts and the principles (basic and derivative) that together make sense of claims that a person is rationally required to act in a certain way. The thesis that particular ought-judgments presuppose general principles of reason, illustrated by the rationales that I sketched, distinguish Kantian theory from the prevailing contemporary belief that empirical facts can be considered normative reasons in favor of acts without any further explanation of what makes those facts reasons. Whether successful or not, Kantian theory at least accepts the task of identifying and articulating general principles of rational choice—moral and non-moral. A more minor theme in the same essay was

  See Part II of Human Welfare and Moral Worth, 97–274.

8

280  Thomas E. Hill, Jr. my conjecture that in ordinary practice the particular facts that we pick out as “the reasons” for an act are relative to context, in particular relative to the features of a fuller rationale that needed to be emphasized to those asking for “the reason.” A full rationale, if we could ever articulate it, would not be relative to context in this way. What I suggested is only that the element of a rationale that we call “the reason” is relative to the conversational context. Turning back now to the chapters in this volume, Mark Schroeder takes a fresh look at interpretations of the Hypothetical Imperative in contemporary philosophy as well as in Kant. Using (as Kant did) terms drawn from the law, he distinguishes between the scope and the jurisdiction of principles of instrumental rationality in order to clarify ongoing disputes about their scope. He also questions the source of their alleged authority and gives an original argument that draws from the idea that rational agents are autonomous (i.e. have the power to rationally commit themselves) and concludes that “instrumental rationality is not something that is required of us, as rational agents, but rather, simply a reflection of our capacity to require things of ourselves” (p. 99). The Hypothetical Imperative, it turns out, is not actually an imperative. Schroeder’s idea of autonomy may not be Kant’s, but Schroeder, I suspect, means to be proposing an explanation that reflects broadly Kantian themes that go beyond Kant. The chapter defies brief summary but is challenging and well worth close study. Jonathan Dancy notes with some surprise that I have regularly included selections from W. D. Ross in my ethics courses. Dancy is surely right that Ross was historically important in offering a middle way between utilitarianism and theories that endorsed absolute moral principles that hold without exception, and it is to Dancy’s credit that interest in Ross has revived in recent times. Dancy’s chapter in this volume, however, goes beyond appreciation to confront confusions in Ross’s theory and to explain their source. His critique and diagnosis of the problems in the work of Ross and his critics are important for contemporary controversies about reasons and their relation to overall moral judgments, but rather than engage his larger project here I will simply respond to his brief comment about my essay “The Importance of Moral Rules and Principles.” Dancy mentions my dissatisfaction with contemporary accounts of moral reasons that give no explanation of how the facts that are taken to be reasons can explain an overall moral judgment, and I was surprised and pleased to see that he concludes that Ross, despite first appearances, also leaves the relation between overall judgments and the alleged reasons for them unexplained. Dancy also quickly dismisses my suggestion about how a Kantian theory might construe the relation between overall moral judgments and the particular facts we call “the reasons” for them. This is understandable because on this point my discussion was too brief and incomplete. But it also presupposed a background Kantian framework in which the particular question that Dancy raises is not an especially important one. That is, Kant was not concerned to identify a particular fact as the reason for a moral judgment and distinguishing this

Looking Back  281 fact from background facts that also belong to a full rationale, “enabling” the judgment and showing that it is not “defeated” by special circumstances. Instead, Kant set himself the task of finding comprehensive rational principles that ideally specify relevant factual circumstances and procedures of deliberation that we need to make overall moral judgments. What matters is the full justifying rationale for the judgments, not how we identify which particular fact in a full rationale constitutes “the reason.” Also, for Kant, but not for Dancy or Ross, the primary ethical question “What ought I to do?” calls for moral standards for choices faced by fallible human beings who are inevitably uncertain of potentially relevant facts. The question is not “What cluster of particular facts metaphysically determines that a certain act is right?” but rather “Given the available options, which, if any, are reasonable and morally justifiable for me to choose?” In this background, what matters is whether there is a rationale for one’s choice, not how we can pick out the elements of that rationale that we should (or do) call “the reason.” I presupposed this background when suggesting that the particular fact that we identify as the reason for a moral judgment is relative to context, but even so I acknowledge that there may be more subtle criteria than the ones I suggested by which normally we distinguish what we call “the reason” from other background facts and principles necessary for a full rationale. That aside, the larger issue between Kantians and particularists is still whether it is possible to articulate and defend the general principles of reason that can plausibly serve as the basis of a full rationale. In “Autonomy and Public Reason in Kant” Onora O’Neill distinguishes Kant’s idea of public reason from the ideas of public reason presented by Jürgen Habermas and John Rawls, connects Kant’s idea of public reason with his idea of autonomy, contrasts this with contemporary conceptions of autonomy, and draws practical implications about reforms needed to promote public reason in Kant’s sense. She kindly acknowledges and develops further my earlier attempts to distinguish Kant’s special idea of autonomy as a property of the rational will from more recent conceptions of autonomy as various kinds of independence that individuals may have.9 Contemporary confusion about Kantian autonomy, she explains, arises from attributing autonomy to individuals rather than to principles or “determinations of the will.” Although her basic point here is important, I suspect it would be misleading and unnecessary for Kantians to deny altogether that autonomy is a property of the wills of persons. Presumably, all persons have a will with the property of autonomy because, insofar as they are moral agents, all persons have practical reason which is, in one sense, the will (Wille) with the property of autonomy, i.e. the property of “being a law to itself (independently of 9  Thomas E.  Hill Jr., “The Kantian Conception of Autonomy”, in John Christman, ed., The Inner Citadel: Essays on Individual Autonomy (Oxford: Oxford University Press, 1989), 91–105 and reprinted in Thomas E. Hill, Jr., Dignity and Practical Reason in Kant’s Moral Theory (Ithaca, NY: Cornell University Press, 1992), 76–96. I commented on autonomy more recently in “Kantian Autonomy and Contemporary Ideas of Autonomy,” in Oliver Sensen, ed., Kant’s Conception of Autonomy (Cambridge University Press, 2013), 15–31.

282  Thomas E. Hill, Jr. any property of the objects of volition).”10 It is consistent with O’Neill’s main point, I think, to say that human beings have the capacity and rational disposition to follow the principle of autonomy (the Categorical Imperative) because they possess rational wills with autonomy. Nevertheless, because they also have a power of choice (a will in another sense, Willkür) that is not always determined by their rational will with autonomy, they can fail to follow the Categorical Imperative. In a sense, then, we human beings have autonomy, but, as O’Neill emphasizes, it is not a kind of independence that licenses amoral choices. O’Neill’s discussion is illuminating, especially in the ways that it highlights how far our societies today are from the Kantian ideal of public reason. As she points out, new technologies enable us to “transmit content to many others,” but genuine communication also requires making speech accessible and intelligible to the intended audiences. The Kantian ideal of public reason, as she says, includes norms of correct reasoning and not just criteria for acceptable participation in public debates. Drawing from her long-standing interpretation of Kant’s Categorical Imperative, she explains that for Kant “fully public reasoning requires agents to think, act and communicate on principles that are not only law-like in form but have universal scope” (p. 127). O’Neill argues that, in contrast to Habermas and Rawls who focus on actual agreement as a condition of full political legitimacy, Kant’s standard of public reasoning requires that arguments could in principle be communicated and followed by an unrestricted audience. Questions remain for me about the nature and adequacy of this standard, though O’Neill has written extensively on this elsewhere. Also, while her views about the role of consent in establishing political legitimacy significantly differ from the views of both Habermas and Rawls, it is worth emphasizing that (as she knows) the views of Habermas and Rawls are importantly different. In particular, a central tenet in all versions of Rawls’s theory of justice is that political principles are primarily justified by appeal to the hypothetical consent of parties in an ideal choice situation. In Rawls’s Political Liberalism full political legitimacy depends at least on the possibility of an overlapping consensus of a society’s reasonable comprehensive moral and religious doctrines, but it is unclear to what extent this requires the actual consent of those subject to the laws.11 O’Neill has often raised objections to the appeal to hypothetical consent to justify principles, arguing that instead the standard must be whether or not all can consent. I have tried to develop and defend a broadly Kantian perspective that, in contrast to 10   Immanuel Kant, Groundwork for the Metaphysics of Morals, ed. Thomas E. Hill, Jr. and Arnulf Zweig (Oxford: Oxford University Press, 2002), 240 (4: 440). 11   The liberal principle of legitimacy, Rawls writes, says that “our exercise of political power is fully proper only when it is exercised in accordance with a constitution the essentials of which all citizens as free and equal may be reasonably expected to endorse in the light of principles and ideals acceptable to their common human reason.” Political Liberalism (1996), 137 (see also 217). The crucial requirement, it seems, is that there are reasons sufficient to support (and so in a sense “reasonably expect” citizens to endorse) the constitution, not whether citizens actually consent to the constitution. In any case full legitimacy also requires that the principles would be endorsed by parties in an appropriately designed “original position.”

Looking Back  283 O’Neill’s position, gives a significant role to what principles (hypothetically) rational and free persons would endorse. Her view stems from a certain interpretation of Kant’s formula of universal law, whereas mine extends Kant’s idea of a kingdom of ends and draws from Rawls as well. O’Neill’s position has substantial textual support, but I remain optimistic about the value of hypothetical consent standards in moral theory.12 Although I note some possible in-house disagreements here, the important point is that O’Neill’s chapter lays out a distinct Kantian account of public reasoning that has challenging implications.

Social, Political, and Legal Philosophy My interest in this area has been long and persistent, but most of my writing on political and legal philosophy has been focused on Rawls and Kant. After working with and admiring Rawls I decided not to be either a disciple or a critic but to turn instead to Kant and moral questions about self-regard and personal relations with others. The practical questions on which I  later focused were typically moral questions about social policies and political institutions. Punishment, revolution, humanitarian interventions, suicide, affirmative action, symbolic protests, conscience vs. authority, and even attitudes towards the natural environment are topics where it is hard to separate moral, social, political, and legal considerations. For better or worse, I never tried to articulate a comprehensive theory on these matters but instead I tried to draw out what I thought were practical implications of Kantian theory and also to explore some issues independently and critically. Before turning to the contributor’s chapters, I will mention a few themes that I have tried to articulate and defend (beyond those already mentioned). My interest in Rawls’s political philosophy has persisted, especially as he shifted from A Theory of Justice to Political Liberalism. Despite my earlier resolution to concentrate on other matters, I have now written four papers about his work. In “Kantian Constructivism in Ethics” (1989) I explained the view that Rawls had presented as Kantian constructivism and contrasted it with my reconstruction of Kant’s kingdom of ends formulation of the Categorical Imperative.13 Later my contribution to an American Philosophical Association Author Meets Critics panel was reworked and published as “The Problem of Stability in Political Liberalism.”14 Here I explained 12   I review O’Neill’s objections to Rawls in “Varieties of Constructivism,” in David Archard, Monique Deveaux, Neil Manson, and Daniel Weinstock, eds., Reading Onora O’Neill (London:  Routledge, 2013), 37–54, and in “Kantian Constructivism as Normative Ethics,” Oxford Studies in Normative Ethics (Oxford: Oxford University Press, 2011), 26–50, reprinted in Virtue, Rules, and Justice: Kantian Aspirations (Oxford: Oxford University Press, 2012). 13   “Kantian Constructivism in Ethics”, Ethics, 99 (1989), 752–70. 14   “The Problem of Stability in Political Liberalism,” Pacific Philosophical Quarterly, 75 (1994), 333–52. Reprinted in Respect, Pluralism, and Justice, 237–59.

284  Thomas E. Hill, Jr. why Rawls moved to a “political conception” of justice and why certain objections missed the mark. The most important point, which Rawls acknowledged, was that the reason for arguing that an overlapping consensus of reasonable comprehensive doctrines is possible was not just to show that a well-ordered just society could endure, but also to show that a society structured by his political conception of justice would satisfy the liberal principle of legitimacy, whether or not in fact an overlapping consensus will actually develop. For this purpose, what especially matters is that there are sufficient reasons for all reasonable citizens to support this political conception regardless of their diverse (but reasonable) moral and religious doctrines. More recently, my “Varieties of Constructivism” (2013) contrasts the views of Rawls and Onora O’Neill and attempts to assess her objections to Rawls’s work. Then “Stability, A Sense of Justice, and Self-Respect” places Rawls’s discussions of a sense of justice and self-respect in context and considers what remains of value in the third part of A Theory of Justice despite Rawls’s shift to a different explanation of stability in Political Liberalism.15 My main attempts to interpret and criticize Kant’s political philosophy are the essays on punishment, humanitarian interventions, and revolution.16 In each case Kant advocated a hard, strict position with which I substantially disagree. But I have tried to interpret his position charitably, take his arguments seriously, and emphasize deeper aspects of his theory. Kant’s unjustifiably hard stances in my view included his insistence on capital punishment for murder and treason, his opposition to any forcible interventions in the governance of other states, and his claim that it is always wrong to attempt to overthrow the lawful rulers of one’s state. These claims in his theory of law and justice, I argued, are at odds with his ethical commitment to human dignity, and this tension calls into question his apparent attempt to make a sharp division between legal justice and ethical duties and ideals. With regard to punishment I argued that despite Kant’s reputation as the paradigm retributivist, his theory of judicial (as opposed to divine) punishment is not deeply retributivist in its structure. The law of retribution (lex talionis) was Kant’s proposed measure of the degree and kind of punishment for crimes, but his ground for the state’s authority to punish at all was the legitimacy and need for coercion “as a hindering of hindrances to freedom.”17 The law cannot know the ultimate motives or comparative inner moral worth of criminals, and so must concentrate on the wrongful “external acts” that can be imputed to them and the relative injuries to others that they cause or risk. When examined closely, many of 15   “Stability, A Sense of Justice, and Self-Respect,” in Jon Mandle and David Reidy, eds., A Companion to Rawls (New York: Wiley-Blackwell Publishing, 2014), ch. 11, 200–15. 16  Regarding revolution, see Respect, Pluralism, and Justice, ch. 8, 200–36, and Virtue, Rules, and Justice, ch. 12, 272–95. Regarding humanitarian interventions, see Virtue, Rules, and Justice, ch. 14, 320–42. Regarding punishment, see Respect, Pluralism, and Justice, ch. 7, 173–99, Human Welfare and Moral Worth, ch. 10, 310–39, and Virtue, Rules, and Justice, ch. 13, 296–319. 17   Immanuel Kant, The Metaphysics of Morals, trans. and ed. Mary J. Gregor (Cambridge: Cambridge University Press, 1996), 25 (6: 231). Bracket numbers refer to volume and page in the standard Prussian Academy edition.

Looking Back  285 the most deplorably harsh texts that seem to favor a deeply retributive interpretation turn out, I argue, to be compatible with my significantly qualified reading.18 Two other practical topics that I have addressed in more than one essay are suicide and preserving natural environments. Both focus attention on our attitudes— how we value our own lives and natural environments. My initial discussion of suicide, “Self-Regarding Suicide:  A  Modified Kantian View,” set aside important other-regarding considerations in order to focus on a person’s self-regarding reasons for considering suicide.19 Those who commit impulsive, self-abasing, apathetic, or hedonistic calculating suicides fail to value their own humanity as they should, I argued, but those who end their lives when terminally ill and in gross irremediable pain do not necessarily fail in this way because their human lives are effectively over. A recent sequel, “Killing Ourselves: Suicide and the Appreciation of Life,” contrasted religious, consequentialist, and libertarian perspectives on suicide with the previously sketched Kantian perspective and a further viewpoint that stresses the importance of appreciation of the good things within one’s life that are not reducible to pleasures and avoidance of pain.20 In “Human Excellence and Preserving Natural Environments” I  considered the attitudes that a virtuous person would have towards natural environments. My focus was not directly on practical questions of public policy but on the question of why we may find it morally objectionable for someone to value nature simply because of the rights and welfare of human beings. Failure to appreciate nature for its own sake, I suggested, may be evidence of a lack of natural attitudes that underlie virtues such as proper humility, gratitude, and appreciation of the good in others. The sequel, written much later, proposes that it may be a virtue to appreciate aspects of nature that are good in themselves in an ordinary sense that is distinct from Moore’s understanding of intrinsic value. We must distinguish desiring something for itself, valuing it for its own sake, and taking it to be valuable in itself.21 The essay also details different claims that may be meant in environmental ethics by “anthropocentric” so that we can distinguish between those claims that are objectionable and others with which they may be confused. Here it may seem that I have come full circle from my initial rejection of Moore’s theory of intrinsic value but I have actually only taken a few steps towards acknowledging insights in the Moorean view without accepting its metaphysical and epistemological baggage. 18   See especially “Kant on Wrong-doing, Desert, and Punishment,” in Law and Philosophy, 18(1) (1999), 407–41. Reprinted in Human Welfare and Moral Worth: Kantian Perspectives (Oxford: Oxford University Press, 2002), 310–39, and in Sharon Byrd and Joachim Hruschka, eds., Kant and Law (Aldershot: Ashgate Press, 2006), 337–68. 19   “Self-Regarding Suicide: A Modified Kantian View,” in Suicide and Life Threatening Behavior, 13(4) (1983), 254–75. 20   “Killing Ourselves: Suicide and the Appreciation of Life,” in Steven Luper, ed., Cambridge Companion to Life and Death (Cambridge: Cambridge University Press, 2014). 21   “Finding Value in Nature,” Environmental Value, 15(3) (2006), reprinted in Virtue, Rules, and Justice, 95–106.

286  Thomas E. Hill, Jr. Turning now to the contributor’s essays in this volume, in “Private and Public Conscience” Gerald Gaus calls attention to some important distinctions regarding conscience and authority, and he argues that my discussions of Kant’s conception of conscience contained ambiguities that need to be resolved. He distinguishes between what he calls a personal conception of conscience and an inclusive conception, and suggests, perhaps rightly, that at times my essays switch from one to the other.22 More importantly, Gaus develops his own account of inclusive conscience and public authority, which emphasizes the role of public justification in a pluralistic world. I cannot give his essay the full discussion here that it deserves, but it may be helpful if I try to explain again my main attempt to understand the Kantian conception of conscience.23 The background is Kant’s view of moral agents as having a mostly reliable but often unreflective grasp of what it is their duty to do and what is impermissible. They have the capacity (freedom of will) to do their duty despite contrary inclinations. Their practical reason enables them to think critically about their initial moral beliefs and to judge how general principles apply to particular cases. Nevertheless, they often fail to think critically and make good judgments. They may be ignorant, perhaps non-culpably, of facts that would be relevant in ideal moral decision making. In addition to other duties, they have a duty of due care to use their critical rational powers when judging what they, all things considered, ought to do about important matters. Conscience in Kant’s theory is not the same as practical reason or critical judgment, though both should play a role in its “inner court.” In the court of conscience one serves as one’s own accuser, one’s defender, and one’s judge. Conscience is, in the first instance, an involuntary warning and call to judge whether or not what one has done (and may be still doing) is contrary to one’s moral beliefs. In Religion Kant implies that conscience especially judges whether one is complying with a duty of due care in making moral judgments, but again a subjective judgment of conscience that one is complying is no guarantee that one is really doing what one ought to be doing or really exercising due care in deliberation. Understood in this narrow way, conscience is primarily concerned with relations among one’s moral beliefs, one’s judgment that these beliefs were formed with due care, and one’s understanding of what one was (or is) intentionally doing. Kant apparently thought (mistakenly) that conscience so conceived cannot err because what conscience judges (i.e. relations among one’s beliefs, judgments, and understanding) are inner psychological factors obvious to the agent. The voice of conscience, whether a warning call or a final verdict, expresses a judgment that one is or is not complying with what one takes one’s duties to be. Conscience, so conceived, is not an infallible judge about what is in fact

22   My main attempt to interpret Kant on conscience is in “Four Conceptions of Conscience,” Nomos, 40 (1998), 13–52. A briefer discussion was published earlier as “Conscience and Authority”, Joseph Reich Annual Distinguished Lecture on War, Morality, and the Military Profession (Colorado Springs: US Airforce Academy, 1997). Reprinted in Respect, Pluralism, and Justice, ch. 10, 260–74. 23   This is better expressed in “Four Conceptions of Conscience” than in “Conscience and Authority.”

Looking Back  287 one’s duty. In reaching a verdict in the court of conscience a person may or may not have considered the opinions of others and the demands of political authorities, but arguably on matters of public importance due care in deliberation would require us to take these into account. If the verdict of conscience is that one’s acts (as one understands them) conform to one’s moral beliefs and that one has exercised due care in forming those beliefs, and if (as Kant suggests) the verdict of conscience on these points cannot err, then (as Kant says) one would be blameless for so acting. Nevertheless, assuming that due care in forming moral beliefs is no guarantee that they are correct, it is still possible (as Kant also says) for one’s blameless acts to be objectively wrong. Gaus favors a broad inclusive conception of conscience that treats the judgment of one’s conscience as one’s best informed and duly reflective moral judgment, all things considered, about what one ought to do. On this conception, it makes sense to say that one always ought to follow one’s conscience, even contrary to authoritative public orders, because by hypothesis conscientious reflection has already taken into account to the best of one’s ability the rational and moral basis for recognition of public authorities. Understandably, following conscience in this sense leaves one morally blameless, though not infallibly correct or free from liability to punishment. Gaus and I agree that on serious public matters conscience should take into account the opinions and reasoning of others, but he develops this basic and relatively uncontroversial point in connection with his own significant but more controversial account of public justification. Even our basic point of agreement about the need for reasoning in consultation with others is not so explicit in Kant’s comments on conscience, probably because Kant had more confidence than we do in the capacity of individuals to reach reliable moral judgments through their own practical reasoning. Jeffrie Murphy takes up some remarkable passages in Kant’s legal theory that I have long avoided. Murphy is a friend whose integrity, philosophical seriousness, and moral sensibility I have long admired despite disagreements on particular points. His early work is what first led to my interest in Kant’s political philosophy.24 Despite his modest disclaimer, his chapter shows that he knows Kant’s texts well but is puzzled by them for good reasons. Never one to pull his punches, Murphy highlights three deplorable passages in Kant’s treatment of punishment, but then with some ingenuity and generosity finds Kant groping towards recognition of the important legal distinction between justification and excuse. In the three controversial passages Kant argues against applying the death penalty for murder in cases where (a) from shame a woman kills her “illegitimate” newborn infant, (b) from a sense of honor a soldier kills another in a duel, and (c) after a shipwreck a man pushes another man off a floating plank that each needed to survive. 24   Jeffrie Murphy, Kant: The Philosophy of Right (London: Macmillan Press, 1970; reprinted by Mercer University Press, 1994).

288  Thomas E. Hill, Jr. I agree with Murphy that Kant’s arguments on the first case are ill-considered and morally indefensible and that there are similar problems with his comments on the second case. In my view (contrary to Kant) capital punishment should not be the default position (“objectively” before reason), but at least (as Murphy says) Kant seemed to recognize possible mitigating excuses that could justify a lesser penalty (“subjectively” before a court of law). As Murphy rightly argues, wanting to avoid shame and dishonor is not in general either a justification or excuse for homicide. What is shameful and dishonorable according to existing social conventions might be relevant, however, if the society that authorizes the punishment has supported those social conventions and if there is something morally commendable about them. Kant evidently thought that the conventions of honor that motivated the mother in his first case satisfied these conditions, even though from our point of view we may agree with Murphy that the convention that makes it shameful for an unmarried woman to have to a child belongs to an indefensible “bourgeois morality.” Kant also apparently thought that the convention of dueling to defend one’s honor, even though flawed, was supported by a society that punishes and was at least an expression of something morally commendable. These points, if correct, help to make sense of Kant’s comments, though they fall short of justifying them. Turning to the third (shipwreck) case, Murphy is again critical of Kant but also gives an illuminating general discussion of the supposed “right of necessity” in the law, including comments on the famous case of Regina v Dudley and Stephens. He gives convincing reasons to reject Kant’s argument that making killing in these cases a punishable offence could have no deterrence value and so the law has no authority to make them criminal offenses. Modifying his early position, Murphy also argues that neither hypothetical consent nor implicit actual consent could justify killing a non-offending person to save the lives of others. He concludes with the thought that “killing an innocent person that others might live” may be rightly rejected on the Kantian ground that it treats a person as a mere means, whether or not the person consents or would consent if fully rational. Although in agreement with Murphy on most points, I remain more optimistic than he is about arguments from hypothetical consent, especially when there is also implicit actual agreement. The conditions for hypothetical consent and implicit consent need to be carefully specified, but my repeated proposals for a broadly Kantian legislative perspective (expanding on Kant’s idea of a “kingdom of ends”) are meant to be a step in that direction. Whether Murphy is right that a person could reasonably hold on to an absolute prohibition of killing in the problematic cases for me depends on how we assess the merits of this perspective and the particular arguments from that perspective for exceptions vs. absolute prohibitions. In my view, we are not treating persons merely as means unless we are failing to treat them as ends in themselves with human dignity. The issue is whether in the extraordinary hard cases such as those that arise when we consider a supposed “right of necessity,” a rule or law that allows an exception to a prevailing and extraordinarily important dignity-based norm for most cases necessarily fails to respect the human dignity of

Looking Back  289 each and every person.25 I would argue that in extraordinary cases where there are only horrible options it is not necessarily contrary to proper respect for the human dignity of each and every person to follow principles that allow killing a non-offending person. The key is whether there are compelling dignity-based reasons to think that every reasonable person would endorse these principles. That one or more other people will be saved is not, of course, sufficient by itself to justify sacrificing the life of an innocent person. A dignity-based justification cannot say that only the numbers of survivors count, but that does not mean that the numbers are never relevant. In their essay “Virtue, Repugnance, and Deontology” Matt Zwolinski and David Schmidtz highlight and extend themes from my essay “Ideals of Human Excellence and Preserving Natural Environments.” There I  argued that environmental ethics should look beyond questions of rights and utility to consider our attitudes towards natural environments. When we contemplate destroying or despoiling a forest, a mountain top, or other parts of the natural world, we should ask not only “who owns it?” and “what are the costs and benefits to human beings?” but also “what sort of person would do that?” The attitudes with which we do things, rather than their effects, sometimes explain why these acts are morally objectionable. Also, our attitudes may express or stem from virtues or vices of character that are important independently of whether we actually act on them. The implications, as the authors note, are not only that we should consider environmental ethics from the perspective of human virtues and vices of character, but also that the apparently neat lines often drawn between agent-assessment (“virtue ethics”) and act-assessment (“consequentialism” and “deontology”) in ethics are significantly blurred. Zwolinski and Schmidtz extend these ideas to respond to Parfit’s famous argument that total utilitarianism leads to a “repugnant conclusion” regarding issues about population control. As they note, average utilitarianism leads to a different but almost equally repugnant conclusion. Extending the argument further, they argue that there is a similar problem with “moral theories that treat specifying action-guiding rules as their primary task” (p. 185). The problem, they explain, is not just indeterminacy of application but that virtually any rules that precisely determine what to do will fail to track our intuitions. If we interpret Kant’s Categorical Imperative as a simple and definitive moral decision procedure, for example, it cannot take into account all the morally relevant considerations. Rules are no substitute for wisdom, judgment, and ideal moral attitudes. As the authors see and kindly acknowledge, I agree with the 25   These are extremely controversial issues, and I will no doubt be suspected of opening the door for many consequence-based killings that are morally repugnant. I discuss reasons for making room for very limited justifiable exceptions without embracing consequentialism especially in “Making Exceptions without Abandoning the Principle; or How a Kantian Might Think about Terrorism”, in Ray Frey and Christopher Morris, eds., Violence, Terrorism, and Justice (Cambridge: Cambridge University Press, 1991), 196–229, reprinted in my Dignity and Practical Reason in Kant’s Moral Theory (Ithaca, NY: Cornell University Press, 1992), ch. 10, 196–225. See also Virtue, Rules, and Justice, ch. 8, 185–202.

290  Thomas E. Hill, Jr. spirit and much of the letter of what they say. My response to the concerns they raise, however, is not to turn to a comprehensive “virtue ethics” as the default solution but rather to show that non-consequentialist theories, such as Kant’s, need to be understood as more subtle, complex, and responsive to context than commonly presented in textbooks. A tenable moral theory needs to include (as Kant’s does) an account of virtue and of when and why consequences matter.26 Although Kant at times suggests otherwise, a simple test to distinguish right from wrong is unlikely to have the depth and scope needed to explain even those intuitions that best survive critical reflection. Wisdom and judgment will always be needed, but in my view it is premature to give up altogether on the attempt to develop moral theories that have a place for rules, virtues, and consequences and that help us to understand how these are related. The broadly Kantian theory that I have sketched is just one proposal in this direction. The most common question that confronts me when I give a talk trying to explain admirable features of Kant’s ethics on any subject is “What about the animals?” This has been a disturbing question because, as Cheshire Calhoun notes, I saw Kant’s official position about the moral status of animals as repugnant despite attempts by Kant’s friends to defend or supplement it. Calhoun directly confronts this question and gives a subtle, complex, and morally sensitive response that helps to illuminate and supplement Kant’s position, drawing from a range of texts that are commonly overlooked. She shows that ordinary moral concern for animals is not exclusively about animal rights or animal welfare. Especially in the case of pets and service animals, we often see a need for fairness, respect, and gratitude in our attitudes towards animals. But can Kantians do better than utilitarians in making sense of these intuitions despite the fact that in Kant’s ethics animals lack the dignity and respect-worthy capacities of free and rational persons? Calhoun’s response focuses primarily on our attitudes about the animals with which we form special relationships and the analogies between our relations with animals and with human beings. Rather than addressing all the usual questions directly, she poses a question of the kind I raised in a different context, “What sort of person would treat an animal like that?” Her project, though limited, is original and illuminating. I doubt that it will completely satisfy Kant’s critics regarding the moral status of animals, but it reveals more significant resources in Kant’s thought than his critics—and friends—have identified. My own views in this area are not yet clear, but one point bears repeating. Kant held that only persons, by virtue of their practical reason and autonomy of the will, can determine what morality requires and only imperfectly rational persons (such as human beings) are subject to moral obligations. This does not necessarily mean, 26   My discussions of Kant’s account are in “Two Conceptions of Virtue,” Theory and Research in Education, 11(2) (July 2013), 167–86, “Kantian Virtue and ‘Virtue Ethics’,” in Monika Betzler, ed., Kant’s Ethics of Virtue (Berlin: Walter de Gruyter, 2008), 29–60 (reprinted in Virtue, Rules, and Justice, 129–59), and (with Adam Cureton) “Virtue and Self-Improvement in Kant’s Ethics,” in Nancy E. Snow, ed., Cultivating Virtue: Multiple Perspectives (forthcoming), ch. 5.

Looking Back  291 however, that only persons are morally considerable. Kantian arguments for the ethical treatment of animals, in my view, depend not only on the fundamental features of moral reasoning but also on facts about human nature, the nature of the animals, and how we are related in a common world. Indeed most, if not all, arguments for specific obligations presuppose background facts about the global and local context. The abstract moral law is supposed to be derivable from pure practical reason, independently of empirical facts about human nature and the world we inhabit, but even so the law must be interpreted and applied in the light of facts. On my reconstruction the basic moral law requires us to work out (and “legislate to ourselves”) a system of substantive principles relying both on fundamental Kantian values (e.g. implications of the humanity formula) and facts about ourselves and the world. From this point of view many, though not all, familiar reasons for treating animals well remain relevant. Kant’s argument that those who abuse animals are more likely to abuse human beings is grossly insufficient by itself, but this is surely not all (or the best) that Kantians can say. Kantian moral theory does not uncritically accept unsupported assertions that animals have rights, that it is wrong to favor our own species, or that the interests of all sentient creatures should count equally, but, as Calhoun has begun to show, moral arguments regarding how human beings should treat animals need not start from uncritical acceptance of these assumptions.

Kant’s Ethics Since I have been writing about Kant’s ethics for a long time, it may be helpful if I try to summarize here some of the main themes that have occupied my attention. I will start with themes beyond those the contributors to this volume have focused on, and then will turn more specifically to the main themes in the contributors’ chapters. First and most generally, my aim has been to understand Kant’s moral and political theory, to see if I could explain it without distortion in accessible non-specialist language, and to show its relevance to theoretical and practical questions of interest today. I distinguish between purely historical points, informed interpretation, and acknowledged extensions and modifications of Kant’s ideas, and I try to make it clear which I am proposing to do at different places. A working hypothesis has been that, although Kant uses a special vocabulary and places his ethics within his systematic critical philosophy, his views are often closer to common moral thinking than is usually supposed. He strongly emphasizes certain points in stark contrast to influential moral theories that he thought were mistaken, sometimes leaving an exaggerated and one-sided impression of ethics, but we can get a more balanced understanding by paying close attention to his later works, especially The Metaphysics of Morals, Religion within the Boundaries of Mere Reason, and Lectures on Ethics. Second, a persistent theme over many years has been that the kingdom of ends formulation of the Categorical Imperative incorporates ideas expressed in the earlier

292  Thomas E. Hill, Jr. formulations but avoids some of the problems that those earlier formulations raise when considered by themselves. To some extent I think that my idea of the relations among the formulations of the Categorical Imperative is a reasonable interpretation of Kant’s text in Groundwork II, but I also develop a reconstruction that goes beyond the letter of the text. The basic idea is that mid-level moral principles, such as those in Kant’s Doctrine of Virtue, are moral principles (“laws”) that would be adopted (“legislated”) by an inclusive moral community of rational persons with autonomy of the will who respect themselves and other members as ends in themselves. Each of these features requires interpretation and defense, of course, but the structure of the proposed model is partly analogous to Rawls’s idea of “the original position.” The universal law formula asks what personal maxims we can will consistently as universal law, but the kingdom of ends formula (as I construe it) asks what moral principles we would will as universal laws if we were adopting principles from an ideal position for doing so. In several essays over the years I have filled out this reconstruction, for example, clarifying how I understand autonomy, treating persons as ends in themselves (with “dignity”), “abstracting from individual differences,” etc. An attractive feature of the kingdom of ends formulation, as reconstructed, is that its use does not require us first to articulate the particular maxim that we want to act on. The kingdom of ends model, however, raises questions of its own, for example whether the principles suitable for the ideal kingdom of ends can be made relevant to our very imperfect world. In several of my later essays I address these questions of application, though no doubt insufficiently.27 The kingdom of ends model and even aspects of Kant’s Doctrine of Virtue, I suggest, can be seen as a kind of constructivism in normative ethics distinct from metaethical constructivism and the Kantian constructivisms described by Onora O’Neill and John Rawls.28 Third, like many others I have been impressed with Kant’s ideas of persons as ends in themselves and the dignity of humanity as opposed to unconstrained consequentialist thinking. These ideas are also an important part of the model of a kingdom of ends. Recurring themes in several of my essays are (i) that Kant uses both a more formal (“thin”) and a more substantive (“thick”) idea of the dignity of humanity, (ii) that “Never treat a person merely as a means” presupposes and gets its meaning in relation to the more general principle “Always treat a person as an end in himself,” (iii) that “humanity” in many passages is more than the capacity to set ends, (iv) that dignity is not an intuited metaphysical property, (v) that the ideas of human dignity and persons as ends in themselves are most reasonably considered as fundamental abstract values to be used to guide deliberation about a system of moral principles, taken together, rather than used to determine right and wrong intuitively case by case, and (vi) that 27   See Thomas E. Hill, Jr., Respect, Pluralism, and Justice, chs. 2, 3, and 8, Human Welfare and Moral Worth, ch. 3, and Virtue, Rules, and Justice, chs. 8, 13, and 14. 28   See especially Thomas E. Hill, Jr., Virtue, Rules, and Justice (Oxford: Oxford University Press, 2012), chs. 3 and 11.

Looking Back  293 although the humanity formula holds without exception (dignity is never to be violated), the more specific principles that follow for normal conditions (e.g. ‘Don’t kill human beings’) are subject to exceptions provided that those exceptions can be justified for the circumstances by the basic principle itself. Turning now to the contributors’ chapters in the section on Kant’s ethics, each one is a new and important discussion of topics in Kant’s ethics that I have addressed at some point. First, Marcia Baron returns to my early (1971) essay “Kant on Duty and Supererogation” and my later (2002) revisit to the topic (which was partly a response to her criticisms).29 The question is whether Kant’s ethics has (or could be reasonably extended to have) a place for acts that are good to do but beyond duty. Controversy over this question was prompted by J. O. Urmson’s classic paper “Saints and Heroes,” where he argued that Kant’s theory ruled out the possibility that there are such acts (often called “supererogatory”). Baron is admirably thorough and detailed in defending her view that Kant cannot and should not grant that any acts are supererogatory in the sense that they are “good to do but beyond duty.” Nevertheless she argues that, by using such terms as remarkable generosity, uncommonly thoughtful, and unbelievably courageous, “Kant can mostly do justice to the phenomena for which a category of the supererogatory is thought to be necessary.”30 As Baron acknowledges, we are in agreement on many of the main points in this controversy. I would not want to defend everything in my earliest paper on this subject. She is right, for example, to distinguish “morally worthy” and “meritorious” and to explain that “more in the way of duty” (6: 227) may merely imply a contrast with juridical duties or other perfect duties. The word “supererogatory” has many unfortunate connotations that are best to avoid when describing Kant’s ethics. In my later paper “Meeting Needs and Doing Favors” I tried to highlight certain central points and avoid some misleading ways that I expressed my view earlier. Even in the earlier paper, however, I  never proposed (as Baron suggests) “to translate what Kant says about obligatory ends into statements to the effect that . . . ‘it is a duty to promote the happiness of others sometimes.’ ” That would be a serious mistake, inviting the ridicule that David Cummiskey heaped on what he called my “anemic” principle of beneficence. Rather than offering a translation or full exposition, my point was just that, when considering the implications of obligatory ends for action, the basic duty to adopt the happiness of others as an end implies that it is a duty to promote the happiness of others sometimes but does not specify when, how, and to what extent. This general principle of beneficence, however, is far from a complete specification of our duties to help others in particular cases. In the end the main disagreement between Baron’s view and mine may be less about how to describe Kant’s ethics than about whether moral theory should acknowledge that some acts are morally good to do but not required. This is my 29   Thomas E. Hill, Jr., “Meeting Needs and Doing Favors” in Human Welfare and Moral Worth: Kantian Perspectives (Oxford: Oxford University Press, 2002), ch. 7, 201–43. 30   Eds. note: p. 17 of Baron’s chapter in this volume.

294  Thomas E. Hill, Jr. own view in moral theory, but I now suspect that it may have been a mistake to suggest that Kant’s theory could be easily extended to accommodate it. The changes needed might be quite substantial and, in any case, more needs to be said about the sense in which the acts are “good to do.” In “Did Kant Accept a Strong ‘Guise of the Good’ Thesis?” Andrews Reath addresses another long-standing controversy in Kant interpretation. The question is to what extent, if any, does Kant accept a version of the idea that we always choose (will) to act “under the guise of the good”?31 The issue is important because a strong version of this doctrine leads to the conclusion that no one ever knowingly chooses to act badly (do wrong) and so all wrongdoing must be due to error or self-deception. In other words, because we always see what we choose to do as objectively good, we never culpably choose to do what we know to be objectively bad (or wrong) from weakness of will or willful defiance of our moral standards. We are pulled towards this interpretation by the fact that Kant says in the Groundwork that “the will is nothing but practical reason” (4: 412) and that pure practical reason commands unconditional conformity to the moral law. It seems, then, that we always will to act well (i.e. to avoid wrongdoing) but we are sometimes mistaken about what specifically this requires. As Reath explains, I have several times expressed my skepticism about attributing this doctrine to Kant—at least in its strong form that implies that we always act for the sake of what we take to be objectively and universally good. As Reath says, there are many texts to consider and it may be that no unambiguous interpretation can explain all the passages. In Kant’s later works, especially Religion, he seems to acknowledge clearly that we can willfully make the “evil” choice to subordinate the moral law to self-interest. Much depends on how we understand Kant’s distinction in his later works between will-as-legislative-practical-reason (Wille) and will-as-the-ability-to-choose (Willkür) and also on how we understand his arguments that the moral law is the principle of every rational will with autonomy. Reath scrutinizes texts judiciously with extraordinary patience, constructs several possible interpretations, and argues that a moderately strong version of the guise of the good thesis coheres with main themes in Kant’s moral theory and makes sense of some puzzling texts. Reath’s chapter should set the agenda for any further discussion of this issue. For now I will add just a few comments to clarify my position. First, Kant clearly held that a person, as an end in himself, has the power to set permissible (universalizable) personal ends that he correctly conceives as giving reasons to others which might be expressed as “it is good that the person’s end be realized.” This is uncontroversial because it is an implication, Kant says, of the formula of humanity as an end in itself. What is controversial is whether, apart from a prior commitment to that moral standard, in setting personal ends one always and necessarily takes them to be permissible and good 31   Virtue, Rules, and Justice, Introduction, 3–5; Hill, “Legislating the Moral Law and Taking One’s Choices to be Good,” Philosophical Books, 4(2) (2008): 97–106; Human Welfare and Moral Worth, 251–68; “Practical Reason, the Moral Law, and Choice,” Analytic Philosophy 53 (1) (2012): 71–7.

Looking Back  295 in a sense that gives reasons to others. The issue is the order of argument. Only a taking-one’s-end-to-be-good that is not based on the moral law could be a premise for an argument for the moral law. Second, my alternative to Reath’s interpretation is not to treat Wille (i.e. practical reason) as purely cognitive and then to treat Willkür (power of choice) as somehow motivating us and putting into effect the rational prescription. Kant undeniably thought that pure practical reason is both cognitive and rationally motivating, even though imperfectly rational (human) beings are also inclined and able to act contrary to the full requirements of reason. Without being overpowered by the causal “force” of inclination, we sometimes use reason instrumentally in choosing an (impermissible) maxim that incorporates our inclination in a way that (as we may know) is not rationally defensible all-things-considered. Third, does this interpretation make the immoral choice arbitrary? We can, of course, explain what the agent wanted, what the purpose was, how it coheres with the agent’s other ends, why particular means were selected, and how reason was used to understand the factual circumstances. This is enough, I suppose, to say “what the agent’s reason was” in an ordinary sense, but the choice cannot be explained as fully justified by sufficient reasons because by hypothesis it was not fully in accord with reason. Fourth, is this a satisfactory interpretation of Kant? We can still say that for those who make immoral choices the moral law is an unconditionally authoritative standard even though it is not the principle that explains their immoral choice. Why do rationally competent persons sometimes choose evil over good? This is just where for Kant, I think, explanations run out. If empirically explicable (the agents were overpowered by sensuous forces), the choices could not be evil or good; and if fully explicable by reason (based on the moral law), the choices would have been good. As philosophers we may want an explanation, but did Kant have one—or think he had one? In “Kantian Complicity” Julia Driver offers a rich and subtle exploration of the resources in Kantian moral theory for explaining why it is wrong and morally complicit to be a passive bystander to oppression. She takes into account various writings of mine relevant to the project, but goes beyond them in ways that I had not considered. The initial problem that she identifies is that Kantian moral reasoning (from the Categorical Imperative) seems (implausibly) to require a person to be morally complicit in certain circumstances. This would be a serious objection to Kant, I think, because “complicity” (as I  understand it) implies wrongdoing. Driver and I  agree, however, that the initial suspicion that Kant’s ethics requires complicity in wrongdoing is not justified by deep features of Kant’s moral theory, but instead comes from Kant’s arbitrary exclusion of context-sensitive maxims and his rigid stance on particular issues. For example, Kant’s unfortunate claim that it would be wrong to tell a lie to a murderer to save an innocent person’s life reflects an inflexible opposition to lying arguably not warranted by the best reconstruction of his full theory, especially the later formulations of the Categorical Imperative. In my discussion of bystanders, I stipulated that they had certain features that may not be fully present in all real life examples and I bypassed the question that is practical (often urgent) in particular

296  Thomas E. Hill, Jr. contexts, viz. “What should we do to resist the oppression now?”32 I focused instead on second-order preparatory responsibilities of due care in deliberation, self-scrutiny, and the cultivation of a strong will. Beyond this Driver addresses the more pressing practical question, arguing that Kant has grounds for judging bystanders to be wrongfully complicit in oppression independently of whether they are causing harm or even potentially generating harm.33 I found especially significant her arguments that integrity and self-esteem, beyond basic self-respect, require active thinking and resistance to oppression and that to remain uncritically passive is to allow oneself to be used merely as a means. The chapter illustrates the explanatory power of Kantian moral reasons that are not simply a matter of avoiding harm to others. To conclude, I am grateful to the editors for the opportunity to call attention to main themes in my work over many years and I appreciate the excellent essays of the contributors and their willingness to publish their original work in this volume.

32   “Moral Responsibilities of Bystanders,” Journal of Social Philosophy, 41(1) (2010), 28–39. Reprinted in Virtue, Rules, and Justice, 343–57. 33   There is a brief discussion of moral complicity and disassociation from evil in “Symbolic Protest and Calculated Silence,” Philosophy and Public Affairs, 9(1) (1979), 83–102, reprinted in Autonomy and Self-Respect (Cambridge: Cambridge University Press, 1992), 52–66.

Published Works by Thomas E. Hill, Jr. Books (Published) 2012a Virtue, Rules, and Justice: Kantian Aspirations. Oxford: Oxford University Press. 2009a The Blackwell Guide to Kant’s Ethics. Malden, MA and Oxford: Wiley Blackwell edited with an introduction, pp. 1–16. 2002a Kant: Groundwork for the Metaphysics of Morals, co-edited with Arnulf Zweig, trans. by Arnulf Zweig. Oxford: Oxford University Press. 2002b Human Welfare and Moral Worth: Kantian Perspectives. Oxford: Oxford University Press. 2000 Respect, Pluralism, and Justice: Kantian Perspectives. Oxford: Oxford University Press. 1992a Dignity and Practical Reason in Kant’s Moral Theory. Ithaca, NY: Cornell University Press. 1991a Autonomy and Self-Respect. Cambridge: Cambridge University Press. Forthcoming Articles 2014a (forthcoming) “Virtue and Self-Improvement in Kant’s Ethics,” with Adam Cureton, in Nancy E. Snow, ed., Cultivating Virtue: Multiple Perspectives. New York: Oxford University Press, ch. 5. Published Articles and Book Chapters In addition to indicating those articles and chapters that are reprinted in Hill’s books, only a select set of other reprints are listed. Many of his articles have appeared elsewhere, some of them reprinted numerous times. 2014b “In Defense of Human Dignity:  Comments on Kant and Rosen,” in Christopher M. McCrudden, ed., Understanding Human Dignity, Proceedings of the British Academy Vol. 192. Oxford: Oxford University Press, ch. 18, pp. 315–27. 2014c “Rational Foundations of Human Dignity in Kantian Approaches,” in Marcus Düwell, Jens Braarvig, Roger Brownsword, and Dietmar Mieth, eds., The Cambridge Handbook of Human Dignity. Cambridge: Cambridge University Press, ch. 22, pp. 215–21. 2014d “Stability, A Sense of Justice, and Self-Respect,” in Jon Mandle and David Reidy, eds., A Companion to Rawls. Malden, MA:Wiley-Blackwell Publishing, pp. 200–15. 2014e “Killing Ourselves: Suicide and the Appreciation of Life,” in Steven Luper, ed., Cambridge Companion to Life and Death. Cambridge: Cambridge University Press, pp. 265–81. 2013a “Rüdiger Bittner on Autonomy.” Erkenntnis (October 2013), online. 2013b “Two Conceptions of Virtue.” Theory and Research in Education (July 2013): 11(2). 2013c “Varieties of Constructivism,” in David Archard, Monique Deveaux, Neil Manson, and Daniel Weinstock, eds., Reading Onora O’Neill. London: Routledge, pp. 37–54. 2013d “Kantian Autonomy and Contemporary Ideas of Autonomy,” in Oliver Sensen, ed., Kant’s Conception of Autonomy. Cambridge: Cambridge University Press, pp. 15–31.

298  Published Works by Thomas E. Hill, Jr. 2013e “Supererogation,” with Adam Cureton, in Hugh LaFollette, ed., International Encyclopedia of Ethics. Oxford: Wiley-Blackwell. 2013f “Kant on Imperfect Duties to Oneself,” in Andreas Trampota, Oliver Sensen, and Jens Timmermann, eds., Kant’s Tugendlehre. Berlin and New York: de Gruyter, pp. 293–309. 2012b “Scanlon on Moral Dimensions.” Philosophy and Phenomenological Research, 83(2): 482–9. 2012c “Practical Reason, the Moral Law, and Choice: Comments on Stephen Engstrom’s The Form of Practical Knowledge.” Analytical Philosophy, 53(1) (March), 71–7. 2011 “Kantian Constructivism as Normative Ethical Theory,” in Oxford Studies in Normative Ethics. Oxford: Oxford University Press, pp. 26–50. Reprinted in Hill 2012a. 2010a “Kant’s Tugendlehre as Normative Ethics,” in Lara Denis, ed., Kant’s ‘Metaphysics of Morals’: A Critical Guide. Cambridge: Cambridge University Press, pp. 12, 234–55. Reprinted in Hill 2012a. 2010b “Kant,” in John Skorupski, ed., The Routledge Companion to Ethics. London: Routledge Publishing Co., pp. 156–67. Reprinted in Hill 2012a. 2010c “Moral Responsibilities of Bystanders.” Journal of Social Philosophy, 41(1):  28–39. Reprinted in Hill 2012a. 2009b “Kant and Humanitarian Intervention.” Nous Supplement:  Philosophical Perspectives, 23: 221–40. Reprinted in Hill 2012a. 2008a “Kant on Weakness of Will,” in Tobias Hoffmann, ed., Weakness of Will from Plato to the Present. Washington, DC: Catholic University of America Press, pp. 210–30. Reprinted in Hill 2012a. 2008b “Legislating the Moral Law and Taking One’s Choices to be Good.” Philosophical Books, 49(2): 97–106. 2008c “Moral Construction as a Task:  Sources and Limits.” Social Philosophy and Policy, 25(1):  214–36. Reprinted in Hill 2012a and in Ellen Frankel Paul, Fred D.  Miller, Jr., and Jeffrey Paul, eds., Objectivity, Subjectivism, and Relativism in Ethics. Cambridge: Cambridge University Press, 2008, pp. 214–36. 2008e “Kantian Virtue and ‘Virtue Ethics’,” in Monika Betzler, ed., Kant’s Ethics of Virtue. Berlin: Walter de Gruyter, pp. 29–60. Reprinted in Hill 2012a. 2008f “Talking about Ethics: A Conversation with Thomas Hill, Jr. about Kantian Ethics,” in Lawrence Hinman, Ethics: A Pluralistic Approach to Moral Theory. Belmont, CA: Centgage Learning, Inc., Thomson-Wadsworth, pp. 182–4. 2007 “The Importance of Moral Rules and Principles.” The Annual Lindlay Lecture, University of Kansas Press. Reprinted in Hill 2012a. 2006a “Finding Value in Nature.” Environmental Value 15(3). Reprinted in Hill 2012a. 2006b “Kantian Normative Ethics”, in David Copp, ed., The Oxford Handbook of Ethical Theory. Oxford: Oxford University Press, pp. 480–514. Reprinted in Hill 2012a. 2005a “Assessing Moral Rules:  Utilitarian and Kantian Perspectives,” in Ernest Sosa and Enrique Villanueva, eds., Normativity, Philosophical Issues. A Supplement to Nous, 15: 158–78. Reprinted in Hill 2012a. 2005b Contributions to International Kant Interview. 100 Et”udov o Kant, Istoriko-Filosofsky Almannach, Vipusk 1. Moskva: Sovremennie Tetradi, SS 3-16, edited by Vadim Vasiley, pp. 46–7, 81–2, and 111. .

Published Works by Thomas E. Hill, Jr.  299 2003a “Die Würde der Person:  Kant, Probleme und ein Vorschlag”. “Human Dignity:  Kant, Problems, and a Proposal”, ed. Ralf Stoecker and trans. Joachim Schulte, in Ralf Stoecker, ed., Menschenwürde: Annäherung an einen Begriff, Vienna: Öbv & Hpt, pp. 153–73. English version published in Hill 2012a. 2003b “Treating Criminals as Ends in Themselves,” Jahrbuch für Recht und Ethik /Annual Review of Law and Ethics, vol. 11, pp. 17–36. Reprinted in Hill 2012a. 2002a “Punishment, Conscience, and Moral Worth,” in M. Timmons, ed., Kant’s Metaphysics of Morals: Interpretative Essays. Oxford: Oxford University Press, pp. 340–61. Reprinted in Hill 2002b. 2002c “Questions about Kant’s Opposition to Revolution,” Journal of Value Inquiry 36(2–3): 283– 98. Reprinted in Hill 2012a and in Arthur Ripstein, ed., Immanuel Kant. International Library of Essays in the History of Social and Political Thought. Aldershot: Ashgate Publishing, 2008, pp. 401–16. 2001a “Comments on Franz and Cafaro.” Philosophy in the Contemporary World, 8(2): 59–62. 2001b “Hypothetical Agreement in Kantian Constructivism.” Social Philosophy and Policy, 18(2): 300–29. Reprinted in Hill 2002b and in Ellen Frankel Paul, Fred D., Miller, Jr., and Jeffrey Paul, eds., Moral Knowledge. Cambridge: Cambridge University Press, 2001, pp. 300–29. 2001c “Kant and Race,” co-authored with Bernard Boxill, in Bernard Boxill, ed., Race and Racism. Oxford: Oxford University Press, pp. 448–71. 1999a “Kantianism,” in Hugh LaFollette, ed., The Blackwell Guide to Ethical Theory. Oxford: John Wiley and Sons Ltd Blackwell Publishers, ch. 12, pp. 227–47. Revised 2nd edition, ch. 14, pp. 311–31. Reprinted as “Kantian Analysis: From Duty to Autonomy” in Hill 2002b. 1999b “Kant on Wrong-doing, Desert, and Punishment” in Law and Philosophy, 18(1): 407–41. Reprinted in Hill 2002b and in Sharon Byrd and Joachim Hruschka, eds., Kant and Law. Aldershot: Ashgate Press, 2006: pp. 337–68. 1999c “Autonomy and Agency.” William and Mary Law Review, 40(3): 847–56. 1998a “Respect for Persons,” in Edward Craig, ed., Encyclopedia of Philosophy, vol. 8. New York and London: Routledge Publishing Co., pp. 283–7. 1998b “Kant on Punishment, Conscience, and Moral Worth.” The Southern Journal of Philosophy, 36, Supplement, pp. 51–71. Reprinted in Hill 2002b and in Mark Timmons, ed., Kant’s Metaphysics of Morals: Interpretative Essays. Oxford: Oxford University Press, 2002, pp. 233–53. 1998c “Four Conceptions of Conscience.” Nomos, 40: 13–52. Reprinted in Hill 2002b. 1998d “Happiness and Human Flourishing in Kant’s Ethics,” Social Philosophy and Policy, 143– 75. Reprinted in Hill 2002b and in Ellen Frankel Paul, Fred D. Miller, Jr., and Jeffrey Paul, eds., Human Flourishing. Cambridge: Cambridge University Press, 1999, 143–75. 1997a “Dignité et respect de soi,” in Monique Canto-Sperber, ed., Dictionnaire d’éthique et philosophie morale. Paris: Presses Universitaires de France, pp. 413–17. French translation. 1997b “Kant’s Theory of Punishment: A Coherent Mix of Deterrence and Retribution?” Jahrbuch für Recht und Ethik. Annual Review of Law and Ethics, vol. 5, pp. 291–314. Reprinted in Hill 2000. 1997c “Conscience and Authority”, Joseph Reich Annual Distinguished Lecture on War, Morality, and the Military Profession. Colorado Springs: US Airforce Academy. Reprinted in Hill 2000 and in J. Carl Ficarrotta, ed., The Leader’s Imperative:  Ethics, Integrity, and Responsibility. West Lafayette, IN: Purdue University Press, 2001, pp. 228–42.

300  Published Works by Thomas E. Hill, Jr. 1997d “A Kantian Perspective on Political Violence.” The Journal of Ethics, 1: 105–40. Reprinted in Hill 2000. 1997e “Basic Respect and Cultural Diversity,” in Grethe B. Peterson, ed., The Tanner Lectures on Human Values 18. Salt Lake City: University of Utah Press, pp. 3–37. Reprinted in Hill 2000. 1997f “Must Respect be Earned?,” in Grethe B. Peterson, ed., The Tanner Lectures on Human Values 19. Salt Lake City: University of Utah Press, pp. 37–76. Reprinted in Hill 2000. 1997g “Reasonable Self-Interest.” Philosophy and Social Policy, 14(1): 52–85. Reprinted in Hill 2002b. 1996a “Is a Good Will Overrated?” Midwest Studies in Philosophy: Moral Concepts, 20: 299–317. Reprinted in Hill 2002b. 1996b “Moral Dilemmas, Gaps, and Residues,” in H. E. Mason, ed., Moral Dilemmas and Moral Theory. New York and Oxford: Oxford University Press, pp. 167–98. Reprinted in Hill 2002b. 1995a “Rawls’ Legacy: an Ideal and a Project,” in Hoke Robinson, ed., Proceedings of the Eighth International Kant Congress, 1(3). Milwaukee: Marquette University Press, pp. 1157–63. 1994a “Kant on Responsibility for Consequences,” in B. Sharon Byrd, Joachim Hruschka, and Jan C. Joerden, eds., Jahrbuch für Recht und Ethik. Berlin: Duncker & Humbolt, pp. 159–76. Reprinted in Hill 2000. 1994b “The Problem of Stability in Political Liberalism,” Pacific Philosophical Quarterly, 75: 333– 52. Reprinted in Hill 2000 and in Henry Richardson and Paul Weithman, eds., The Philosophy of Rawls, vol. 4: Moral Psychology and Community. New York and London: Garland Publishing Inc., 1999, 167–86. 1993 “Donagan’s Kant.” Ethics, 104(1) (Oct.): 22–52. Reprinted in Hill 2000. 1992b “Self-Respect”, in Charlotte and Lawrence Becker, eds., Encyclopedia of Ethics. New York and London: Garland Press, vol. 2, pp. 1136–8, and in 2nd edition 2001, vol. 3, pp. 1563–5. 1992c “The Autonomy of Moral Agents,” in Charlotte and Lawrence Becker, eds., Encyclopedia of Ethics. New York and London: Garland Press, vol. 1, pp. 71–5, and in 2nd edition, 2001, vol. 1, pp. 111–15. 1992d “Kantian Pluralism.” Ethics, 102(4) (July): 743–62. Reprinted in Hill 2000. 1992e “A Kantian Perspective on Moral Rules.” Philosophical Perspectives, 6: 285–304. Reprinted in Hill 2000. 1992f “Beneficence and Self-Love: A Kantian Perspective.” Social Philosophy and Policy, 1–23. Reprinted in Hill 2002b and in Ellen Frankel Paul, Fred D. Miller, Jr., and Jeffrey Paul, eds., Altruism. Cambridge: Cambridge University Press, 1993, pp. 1–23. 1992g “Gibbard on Morality and Sentiment.” Philosophy Phenomenological Research, 52(4): 957–60. 1991b “Making Exceptions without Abandoning the Principle; or How a Kantian Might Think about Terrorism”, in Ray Frey and Christopher Morris, eds., Violence, Terrorism, and Justice. Cambridge: Cambridge University Press, pp. 196–229. Reprinted in Hill 1992a. 1991c “The Message of Affirmative Action.” Social Philosophy and Policy, 2 (Spring): 108–29. Reprinted in Hill 1991a and in Ellen Frankel Paul, Fred D. Miller, Jr., and Jeffrey Paul, eds., Reassessing Civil Rights. Oxford: Blackwell Publishers, 1991, 108–29. 1989a “The Kantian Conception of Autonomy,” in John Christman, ed., The Inner Citadel: Essays on Individual Autonomy. Oxford: Oxford University Press, pp. 91–105. Reprinted in Hill 1992a. 1989b “Kantian Constructivism in Ethics.” Ethics, 99: 752–70. Reprinted in Hill 1992a and in Henry Richardson and Paul Weithman, eds., The Philosophy of Rawls, vol. 3: Opponents and

Published Works by Thomas E. Hill, Jr.  301 Implications of A Theory of Justice. New York and London: Garland Publishing Co., 1999, 102–20. 1989c “Kant’s Theory of Practical Reason.” The Monist, 72(3):  363–83. Reprinted in Hill 1992a and in Lawrence Pasternak, ed., Kant’s Groundwork of the Metaphysics of Morals. London: Routledge, pp. 99–120. 1987 “The Importance of Autonomy,” in E. F. Kittay and D. T. Meyers, eds., Women and Moral Theory. Lanham, MD: Rowman and Littlefield, pp. 129–38. Reprinted in Hill 1991a and in German translation, “Die Bedeutung der Autonomie,” in Karl Ameriks and Dieter Sturma, eds., Kants Ethik: Beiträge der kontinentaleuropäischen und angloamerikanischen Philosophie. Paderborn: Mentis, 2004, pp. 178–89. 1986 “Weakness of Will and Character.” Philosophical Topics, 14(2) (Fall): 93–115. Reprinted in Hill 1991a and trans. Johannes Schulte in Thomas Spitzley, ed., Willenschwäche. Münster: Mentis Verlag, 2005, pp. 168–90. 1985a “Darwall on Practical Reason.” Ethics, 96: 604–19. 1985b “Kant’s Argument for the Rationality of Moral Conduct.” The Pacific Philosophical Quarterly, 66: 3–23. Reprinted in Hill 1992a and in Paul Guyer, ed., Critical Essays on Kant’s Groundwork of the Metaphysics of Morals. Rowman and Littlefield, 1998, pp. 249–72. 1984a “Autonomy and Benevolent Lies.” Journal of Value Inquiry, 18: 251–67. Reprinted in Hill 1991a. 1983a “Self-Regarding Suicide: A Modified Kantian View.” Suicide and Life Threatening Behavior, 13(4) (Winter): 254–75. Reprinted in Hill 1991a and in Margaret P. Battin and Ronald W. Maris, eds., Suicide and Ethics. New York: Human Sciences Press, 1983, pp. 38–59. 1983b “Ideals of Human Excellence and Preserving Natural Environments.” Environmental Ethics, 5 (Fall): 211–24. Reprinted in Hill 1991a and in Finnish translation in Theoria, 38(2) (June 1995): 1–156. 1983c “Self-Respect Reconsidered,” Tulane Studies, 31: 129–37. Reprinted in Hill 1991a and in Robin S. Dillon, ed., Dignity, Character, and Self-Respect. New York and London: Routledge, 1995, pp. 117–24. 1983d “Moral Purity and the Lesser Evil.” The Monist, 66(2) (Apr.): 213–32. 1980a “Humanity as an End in Itself.” Ethics, 91(Oct.): 84–99. Reprinted in Hill 1991a and in Heiner F. Klemme and Manfred Kuehn, eds., The History of Philosophy: Kant, vol. 2. Brookfield: Ashgate Publishing Company, 1998, pp. 101–16 and R. Arneson, ed., Liberalism, vol. 1. Cheltenham: Edward Elgar Publishers, 1992, pp. 493–508. 1979a “Value Assumptions in Clinical Judgment,” in H. T.  Engelhardt, Jr., S. F.  Spicker, and B. Towers, eds., Clinical Judgment:  A  Critical Appraisal. Dordrecht:  D. Reidel Publishing Company, pp. 254–8. 1979b “Symbolic Protest and Calculated Silence.” Philosophy and Public Affairs, 9(1) (Fall): 83– 102. Reprinted in Hill 2001a. 1978 “Kant’s Anti-Moralistic Strain.” Theoria, 44(3): 131–51. Reprinted in Hill 1992a. 1974a “Kant’s Utopianism,” in Gerhard Funke, ed., Akten des 4. Internationalen Kant-Kongresses, Mainz, Part II. Berlin and New York: Walter de Gruyter & Co., pp. 918–24. Reprinted in Hill 1992a. 1973a “The Hypothetical Imperative,” The Philosophical Review, 82(4): 429–50. Reprinted in Hill 1992a and in Heiner F. Klemme and Manfred Kuehn, eds., The History of Philosophy: Kant, vol. 2. Brookfield: Ashgate Publishing Company, 1998, pp. 117–38.

302  Published Works by Thomas E. Hill, Jr. 1973b “Servility and Self-Respect.” The Monist, 57(1) (Jan.):  87–104. Reprinted in Hill 1991a and in German translation in Henning Hahn, ed., Selbstachtung oder Anerkennung? Beitrag zur Begründung von Menschenwürde und Gerechtigkeit. Bauhaus Universität Weimar Universitätsverlag, 2005. 1972 “The Kingdom of Ends,” in Lewis White Beck, ed., Proceedings of the Third International Kant Congress. Dordrecht: D. Reidel Publishing Company, pp. 307–15. Reprinted in Hill 1992a. 1971 “Kant on Imperfect Duty and Supererogation.” Kant-Studien, 55–7. Reprinted in Hill 1992a and in Ruth Chadwick, ed., Immanuel Kant: Critical Assessments, vol. 3. London: Routledge, 1993. Book Reviews 2013g “Review of Jeffrie Murphy, Punishment and the Moral Emotions.” Faith and Philosophy, 30(4): 490–3. 2004 “Samuel Kerstein, Kant’s Search for the Supreme Moral Principle.” The Philosophical Review, 113 (2): 272–5. 2001d “John Rawls, Collected Papers.” The Journal of Philosophy, 98(5) (May): 269–72. 1995b “Barbara Herman, The Practice of Moral Judgment.” Journal of Philosophy, 92(1) (Jan.): 47–51. 1992h “Gerald Dworkin, The Theory and Practice of Autonomy.” Nous, 26(1): 99–100. 1991d “Rudiger Bittner, What Reason Demands.” Journal of Philosophy, 88(9): 497–501. 1990 “P.C. Lo, Treating Persons as Ends.” The Philosophical Review, 99(2): 278–80. 1989d “Hans Reiner, Duty and Inclination: The Fundamentals of Morality Discussed and Redefined with Special Regard to Kant and Schiller.” Kant-Studien, 80: 243–5. 1984b “Robert Elliot and Arron Gare, eds., Environmental Philosophy: A Collection of Readings.” Environmental Ethics, 6 (Winter): 367–71. 1984c “Richmond Campbell, Self Love and Self-Respect. Ottawa: Canadian Library of Philosophy, 1979.” The Philosophical Review, 92 (July): 470–3. 1980b “Robert J. Benton, The Problem of Transcendental Argument in Kant’s Second Critique.” The Journal of the History of Philosophy, 18 (July): 356–7. 1980c “Richard A. Wasserstrom, Philosophy and Social Issues. Notre Dame: University of Notre Dame Press, 1980.” UCLA Law Review, 28(1) (Oct.): 135–43. 1979c “Onora O’Neill, Acting on Principle. New York: Columbia University Press, 1975, x, 155 pp.” Ethics, 89(3) (Apr.): 306–11. 1974b “Jussi Tenkku ‘Are Single Moral Rules Absolute in Kant’s Ethics?’ Jyvaskyla Studies in Education, Psychology and Social Research, No. 14. 1967, 31 pp.” Theoria, 40(1): 57–61. 1973c “Robert Paul Wolff, The Autonomy of Reason. New York: Harper and Row, 1973, 228 pp.” Journal of Philosophy, 75(12) (Dec.): 743–7. 1970 “T. C. Williams, The Concept of the Categorical Imperative. Oxford: Clarendon, Press, 1968, 136 pp.” The Journal of the History of Philosophy, 8(2) (Apr.): 222–4. 1969a “Jan Narveson, Morality and Utility. Baltimore: The Johns Hopkins University Press, 1967, ix, 302 pp.” The Philosophical Review, 78(4) (Oct.): 547–9. 1969b “Richard T. DeGeorge (ed.), Ethics and Society: Original Essays on Contemporary Moral Problems. Garden City: Doubleday and Company, 1966, xi, 217 pp.” The Philosophical Review, 78(1) (Jan.): 102–4.

Index Note: “n.” after a page reference indicates the number of a note on that page Abbott, Jack   165 n.15 accountability, respect as  3–4, 70–84 acknowledgment  4, 76–7, 81 act-centered theories inadequacy of  181–2 overpopulation 190 virtue, repugnance, and deontology  178, 181–2, 185–6, 190, 191–2 Adams, Robert  272 address, second-personal  4, 70 Afghanistan War  44 Aiken, Henry  270 Albritton, Rogers  270 Allison, Henry  236 n. 11, 247 n. 28 Almeleki, Faleh Hassan  161 American revolution  72 anarchism  150, 151 animals  10–11, 194–210, 290–1 apparent warrant for (severely) discounting their interests  198–203 extension to modes of thought characteristic of our responses to humans  206–10 rights  186, 192 scope of ethical concerns  195–8 valuing 203–6 Anscombe, Elizabeth  271 answerability and second-personal respect  79– 80, 83 apathy  2, 22–3 Appiah, Kwame Anthony  83 n. 24 appraisal respect  3, 72, 75–7, 277 Aquinas, St Thomas  44 n. 7 Aristotle  45, 52 n. 25 arrogance 277 interpersonal  51–2, 55, 57 self-respect  3, 42–66 unwarranted claims  51–2, 57–8, 61–2 attitude 42 Auld, Hugh  21 Austen, Jane,  Pride and Prejudice 44, 54 n. 27 Austin, J. L.  167 n. 18, 270 authority conflict with conscience: dissenter’s view 141–7 conflict with conscience: imposer’s view 147–50 conflict with conscience: reconciliation through public justification  150–6

epistemic  77, 78 honor and accountability  76–8, 80 instrumental rationality  96, 97–8, 99 public  8–9, 135–6, 141–56 second-personal 49 autonomy  12, 281–2 conscience 150 hypothetical imperatives  6, 95 and public reason  6–7, 119–30 rational agents  96 of the will  14, 96, 97–8, 99, 244–5, 246 n. 27 Bacon, Francis  84, 171 bad willing  239–40, 244, 247, 284, 250–4 Baier, Kurt  146 n. 15 Barker, Stephen  270 Baron, Marcia  12–13, 293–4 Bartlett, Roscoe  197–8 bastard infanticide  9, 158–61, 162, 163, 176 Beck, Lewis White  158 n. 4, 172 n. 28 arrogance 61 homicide law  172 rational volition  236 n. 11 beneficence duty  110, 112 humility and arrogance  54 n. 27 supererogatory and imperfect duties  12, 13, 216, 222, 223, 226–7, 293 Berlin, Isaiah  271 Bible, honor and accountability  70–1, 73, 84 Binder, Guyora  165 n. 15 Bishop, Sharon  272 Blackburn, Simon  272 blame complicity 258 duties and  225 n. 28 honor disrespect  80, 82 Bollard, John  72 n. 8 Bowman, James  83–4 Boxill, Bernard  272 servility and self-respect  2, 275–6 Boxill, Jan  272 servility and self-respect  2, 275–6 BP  43 n. 2 Brennan, Geoff  272 Broad, C. D.  102 n. 1, 168 Broome, John  92 Burge, Tyler  272

304 Index Butler, Joseph  136 n. 4 bystanders and complicity  14–15, 257, 258, 260, 261–6, 295–6 Cafaro, Philip  188 Calhoun, Cheshire animals  11, 290–1 pregnancy out of wedlock  162 Card, Claudia  272 Categorical Imperative  278 autonomy 119 complicity  256–7, 258 n. 5, 295 conscience 149 homicide law  160 implications 7 instrumental rationality  89–90, 90 n. 2, 91, 97, 99 interpersonal recognition respect  49 kingdom of ends formulation  291–2 maxims 179 rational volition  13, 14, 234, 235, 238, 241, 244–5, 245 n. 25, 246–50 reason 128 Universal Law formulation  13, 14, 90, 119, 128, 179, 180, 192, 234, 235, 238, 244–5, 246–50, 292 virtue, repugnance, and deontology  289 character acting out of  229–30 cultivation of one’s  227 environmental ethics  178 virtue, repugnance, and deontology  178, 179–81, 189, 190 Chisholm, Roderick  94 n. 13 Cibber, Colley  82–3 Cicero  166, 169 classism  201, 205 code, morality understood as a  226 coercion 155–6 Cohen, Ted  272 commendation and recommendation, distinction between  115–16 communication public reason  123–4, 129–30 quasi  123–4, 130 communications technologies  123–4 compassion to animals  196, 200, 201, 204 to humans  196, 199–200 competition 54 compliance to duties  225, 226 hypothetical imperatives  94–5 complicity 295–6 bystanders and  262–5 Kantian  14–15, 256–66 nature of  257–62 participation  14, 257–8, 260, 261

tolerance  14–15, 257–8, 260, 265–6 conceit  52, 61–3, 64, 65, 251–2 conformity 264 Connor, Sheriff “Bull”  164 conscience  8, 286–7 conflict with authority: dissenter’s view  141–7 conflict with authority: imposer’s view 147–50 conflict with authority: reconciliation through public justification  150–6 cultural relativism  137–9 humility and arrogance  63 inclusive interpretation  8–9, 142–4, 146–7, 149, 151–5, 156 Kantian conception  139–41, 144, 146, 148, 153 liberal order  156 personal interpretation  8, 144–6, 150, 153, 156 private  8–9, 135–56 public  8–9, 135–56 public authority and  8–9, 135–6 religious conception  136–7, 140, 142, 148, 149 conscientious engagement  147–8, 149 consent 102–3 homicide law  288 hypothetical  173, 174 rational 7 tacit 174–5 consequentialism complicity  14–15, 257, 258, 260 environmental ethics  182 virtue ethics  178 n. 2 constructivism 292 contempt honor 74 honor disrespect  4, 80, 81–2 contributory, prima facie duty  106, 107 Corrado, Michael  272 corruption domination and  29, 33, 34 humility and arrogance  44, 62, 66 pride 44 cowardice 162–3 culpability 79 cultural relativistic analysis of conscience  137–9 Cummiskey, David  293 Cureton, Adam  272 custom of the sea  170, 171, 174, 175 Dahl, Norman  272 Dancy, Jonathan  6, 280–1 Darwall, Stephen appraisal self-respect  56 n. 31 instrumental rationality  92 practical reason  277 respect as honor and accountability  3–4, 277 second-personal authority  49 Davis, Angela  272

Index  305 Deadly Sins  44, 46, 62 n. 35 death penalty  157, 159, 161, 162, 163, 167, 169, 171 deception, and domination  31–2, 33–4 deep deliberation  56 Deferential Wife  2, 19–20, 24–38, 39–40, 48, 51 n. 22, 273, 275–6 deliberative democracy private reason  125 n. 23 public reason  120–3 Demos, Raphael  270 deontic buck-passing  6, 114–17 deontology  9–10, 178–92, 289–90 environmental ethics and inadequacy of act-centered theories  181–2 intuition and theory  190–1 maxims and character  179–81 moral theory  187–9 overpopulation 189–90 Repugnant Conclusion  182–6 see also duties Descartes, René  195–6 n.2 desire, rational faculty of  236, 248–9 deterrence, and homicide law  168, 169, 172, 176 Devlin, Lord Patrick  164 n. 14 Diamond, Cora  198 difference 75–6 dignity  204, 205, 275, 292–3 of animals  201, 202 complicity 264–5 honor and accountability  72, 74, 76, 77, 78, 83 humility and arrogance  46, 48, 49–50, 52, 53, 55, 58–9, 60 n. 33, 61, 64, 65, 66 as motivational core of one’s self-conception 65 servility  48, 51, 51 n. 22, 265 universal human rights  72 Dillon, Robin S.  3, 276–7 disdain  81, 82 domination  2, 24–5, 26–40 Dostoevsky, Fyodor,  Crime and Punishment 177 Douglass, Frederick  21, 34–5, 37 Dreben, Burton  270 Driver, Julia  14–15, 295–6 due care  139, 140, 141, 142, 143, 144, 146, 148, 149, 153–4, 155–6 duels, killing in  9, 158–9, 160, 161 n. 9, 162–3, 165 n. 15, 169 n. 20, 176 duties 101 to animals  194, 196, 198–9, 200, 201, 202, 209, 210 of bystanders  263 deontic buck-passing  6, 114–17 ethical 226 evaluative self-respect  55–6 imperfect  12–13, 215–31, 293 juridical 226 Kant’s terminology  223, 224, 228 moral  49, 62

to natural environment  200 to oneself  203, 263, 273–4 perfect 226 prima facie  6, 101, 102–14 proper  102–12, 114 n. 13 to resist oppression  263 rights versus  26 self-respect  20, 65 servility 20 tendency account  105–6 see also deontology; ought Eberle, Christopher J.  147–8, 149 ecological humility  182 ego-ideal 168 Eisenberg, Paul  272 Eliot, George,  Adam Bede 32–3 Ellington, James  256 n. 2 Emerson, Ralph Waldo  263–4, 265, 266 Engstrom, Stephen  233–4, 236 n. 11, 239, 241 n. 20, 253 n. 33 Enlightenment 120–1 enthusiasm 127 environmental ethics  289 humility  3, 45, 46, 47 virtue, repugnance, and deontology  10, 178, 181–2, 186, 187–8, 190 esteem honor and accountability  72, 75–7, 82–3 see also self-esteem ethical duties  226 ethics  42, 291–6 animals  11, 195–8 complicity  14–15, 256–66 environmental  see environmental ethics humility and arrogance  44 neglected topics  194 rational volition  13–14, 232–54 servility and self-respect  1–2 supererogatory and Kant’s imperfect duties  12–13, 215–31 virtue  see virtue see also morality evaluative buck-passing  114, 115 evidential, prima facie duty  107 Ewing, A. C.  102 excuse defenses  160, 169, 176 “face” (as persona)  73–4, 82 Falk, David  272 fallibility conscience  141, 142, 143, 144, 146, 148–50, 156 homicide  170 n. 23 reasonable disagreement contrasted with 148–50 false promises  179 fanaticism 146–7

306 Index favoring 101 fear of death  168, 169 servility and self-respect  27, 28, 34–5, 39–40 slavery 34–5 Feinberg, Joel  77, 78 feminism domination  25, 40 servility and self-respect  19, 25, 27, 39, 40 wives 24 Finkelstein, Claire  166–7 n.16 Firth, Roderick  270 Foot, Philippa  107, 271, 272 French revolution  72 Freud, Sigmund  168 Fuchs, Alan  272 Fugitive Slave Act  264, 265 Fusar Poli, Barbara  81 Gardner, John  258 Gaus, Gerald  8–9, 286–7 Geach, Peter  175 Goffman, Erving  73 goodness, prima facie  106 gratitude to animals  11, 206, 208–9 Gregory the Great  44 n. 7 Grenberg, Jeanine  45 n. 8, 64 n. 37 Grotius, Hugo  72 guilt honor disrespect  80–1, 82 violation of social norms  145, 145 n. 14 Habermas, Jürgen  7, 120–3, 124, 125–6, 281, 282 Hampton, Jean  91 n. 4, 272 Hanley, Ryan  178 n. 2 Hare, R. M.  270, 271 commendation and recommendation, distinction between  115–16 and Ross  101 Harman, Gilbert  272 Hart, H. L. A.  271 Hecaton 166 Herman, Barbara  233 n. 6 heroes and heroic acts  215–16, 228–9, 229–30 n.34, 230 heteronomy, public reason  127–9 Hill, Thomas E., Sr.  15, 269–70 Himmler, Heinrich  138, 155 Hitler, Adolf  136 n. 3 Hobbes, Thomas command versus counsel  79 n. 18 conscience  135–6, 142, 146, 150, 156 motivation 28 nature, state of  163 rational volition  234, 246 n. 27 Hollande, François  43 n. 2 homicide law  9, 157–77, 287–9

honor, shame, bastards, and duels  158–75 Regina v Dudley & Stephens 169–75 shipwrecks and killing to preserve one’s life 165–7 honesty, humility and  3, 47 honor contempt 74 cultures 80–4 killings  9, 161–5, 167 n. 18, 175, 176, 288 respect  74–5, 76, 81, 83, 277 respect as  3–4, 70–84 hubris 44 Huemer, Michael  185 n. 10 human rights universal 72 virtue, repugnance, and deontology  192 humanitarian interventions  284 humanity, respect for  2 arrogance 44 honor and accountability  3–4, 70 Hume, David humility  45, 54 morality 84 practical reason and rationality  4 rational volition  246 n. 27 sympathy  33, 35, 38 humility  205, 276–7 ecological 182 false  3, 50–1, 53, 63 interpersonal 54–5 natural environment  200 self-respect  3, 42–66 true  3, 51, 54–6, 57, 58, 62–3, 64–5 virtue, repugnance, and deontology  10, 182, 187, 188, 190 Hunt, Lynn  83 n. 23 hypothetical consent  173, 174 hypothetical imperatives  5–6, 89–99, 278, 280 assertoric 91 autonomy of the will  97–8 instrumental rationality  92–4 jurisdiction’s impact on explanatory resources 95–7 problematic 91 scope in the law  94–5 versus  the Hypothetical Imperative 89–92 Ibsen, Henrik,  A Doll’s House 32 ideal utilitarianism  15, 270 ideals of human excellence  187 illegitimate child, killing of  9, 158–61, 162–3, 176 illumination 127 imperfect duties  12–13, 215–31, 293 Incorporation Thesis  237–8 independence, individual  119, 120 indignation

Index  307 honor disrespect  4, 80 second-personal respect  81, 82 Individual Difference Principle  258 infanticide  9, 158–61, 162–3, 176 Infield, Louis  199 n. 4 ingratitude, refraining from subjecting animals to  11, 206, 208–9 instinct, animal  195, 203–4, 206 n. 7 instrumental rationality  5–6, 89, 92–9 integrity complicity  15, 264, 265–6, 296 humility and  3, 47 virtue, repugnance, and deontology  192 intention, and complicity  258, 260 n. 8 intrinsic value  270, 271, 285 intuition  190–1, 270 Iraq War  44 Japanese nuclear catastrophe  44 Jefferson, Thomas  36, 37 Johnson, Robert N.  219 n. 13 Johnson, Samuel  82–3 judgment burdens of  149 conscience  135, 136, 137, 139, 140–1, 142, 143, 144, 152, 153, 154, 156 juridical duties  226 jurisdiction hypothetical imperative and  5–6, 90–2, 94–7 moral 150–1 shipwrecks 175 justice duty 110 sticking to your own job as a kind of  23 justification homicide defenses  160, 169, 176 prima facie  103 Kahan, Dan M.  163 n. 13 Kaufman, Arnold  272 Kavka, Gregory  272 Kawall, Jason  188 Keneally, Thomas  230, 230 n. 35 King, J. Charles  272 Kingston, Maxine Hong,  The Woman Warrior 165 n. 15 Kneale, William  271 Kolodny, Niko  92, 93, 93 n. 8 Korsgaard, Christine  195 n. 1, 202, 232–3 Kukathas, Chandran  156 Kutz, Christopher  258, 259–60, 261 Last Man Argument  187–8 law conscience  135–6, 145–6 homicide  see homicide law instrumental rationality  94–6, 99

jurisdiction  see jurisdiction moral  see moral law roles 163 rule of  164, 165 n. 15 scope in the  94–5 law-like nature of reason  126, 127–8 learning 195 legal moralism  164 n. 14 Leites, Edmund  272 Leopold, Aldo  182 Lewis, C. I.  270, 271 Lewis, David  272 liberal order  156 Little, Maggie  117 Locke, John  38, 72 Long, Douglas  272 Louden, Robert B.  64 n. 37 Lovejoy, Arthur O.  83 n. 22 Lyons, David  272 McAdam, Jim  108 MacLean, Douglas  272 Macnamara, Coleen  117 Mallory, Sir Thomas,  Morte d’Arthur 83–4 Mandelbaum, Maurice  270 Margaglio, Maurizio  81 Marshall, Sandra  173 n. 31 Mathews, F.  187 n. 14 maxims ambiguity in Kant’s understanding of  239 n. 15 complicity 260 rational volition  237, 238–9, 240, 242, 245, 246, 247, 248, 249, 251, 252 virtue, repugnance, and deontology  179–81 merit moral 55 and moral worth, relation between  219–22 military dogs  197–200, 206, 208, 209, 210 Mill, John Stuart fallibility of judgment  141, 143 moral responsibility  79, 79 n. 19 social norms  145 utilitarianism  190 n. 17 Milton, John domination and servility  29, 30, 34 Paradise Lost 44 mind reading  36–7, 38 mockery, refraining from subjecting animals to  11, 206 monarchy, honor as the principle of  75 Montaigne, Michel de  84 n. 26 Montesquieu, Baron de  75 Moore, G. E. ideal utilitarianism  15, 270 intrinsic value  270, 271, 285 moral intuitionism  270

308 Index moral duty humility and arrogance  62 interpersonal recognition respect  49 intuitionism 270–1 jurisdiction 150–1 law honor and accountability  72 humility and arrogance  3, 44, 49, 55, 56, 57, 58, 61, 62–3, 64–6 rational volition  13, 242, 244, 245–6, 247, 249–51, 295 servility  20, 52 supererogatory and imperfect duties  221–2 merit 55 obligations 79–80 responsibilities  79–80, 82 rights honor and accountability  72, 79–80 humility and arrogance  49–50, 54 servility and self-respect  20, 21, 23–4, 25–6, 28, 35, 39, 40, 48, 49 theory 271–2 worth and merit, relation between  219–22 morality animals  194–6, 198–202, 205–10 attitude 42 complicity 258–66 conscience  137, 138, 139, 140–1, 142–5, 146–7, 149, 150–6 deep conflict  154–6 duty  114–15, 117 as equal accountability  72 homicide law  158–60, 169, 170 n. 24, 175, 176 honor and accountability  70, 72, 76, 83, 84 humility and arrogance  43–50, 53–4, 55, 57, 58–66 legal moralism  164 n. 14 personal optimizing view  146–7, 149, 150, 156 practical reason  278–9 rational volition  247, 248, 249, 251, 252 servility and self-respect  20, 21, 22–4, 25–6, 28, 29, 35, 37–8, 39–40, 48, 49, 51, 53 understood as a code  226 virtue, repugnance, and deontology  180, 181, 183 n. 7, 184, 186, 187–91 see also ethics Morris, Herbert  272 motivation dignity as  65 evaluative self-respect  57 Hobbes 28 humility and arrogance  63, 64, 65, 66 self-esteem 57 servility and self-respect  40 Murphy, Jeffrie  272 homicide law  9, 287–9

Nagel, Thomas  261, 272 natural environment  285 nature, state of  162, 163–4, 165 n. 15, 175 necessity, right of  9, 165–9, 171–3, 175, 176–7, 288–9 Nietzsche, Friedrich  75, 76 n. 15 non-maleficence 110 normative perspective, respect  1–2 Nuremberg trials  142 O’Neill, Onora  215, 272, 284 autonomy and public reason  6–7, 281–3 constructivism 292 deontology  181 n. 5 maxims  180, 181 respect 262 Orwell, George  123 n. 16 ought  278, 279 hypothetical imperatives  5, 92–4 see also duties overpopulation  10, 185, 189–90 Parfit, Derek duty  102, 103, 114, 116 Repugnant Conclusion  9–10, 178, 182–3, 184, 289–90 participation complicity  14, 257–8, 260, 261 perfect duties  226 perfection  3, 46, 55, 57–8 Perry, John  272 Pettit, Philip  29, 30–1, 32, 34 Plamenatz, John  271 Plato  23, 33 Posner, Richard  173 n. 30 Postema, Gerald  272 practical reason  4–7, 277–83 autonomy and public reason  6–7, 119–30 Hume 4 hypothetical imperative  5–6, 89–99 rational volition  237, 238, 239–43, 245–7, 251–3, 294–5 rightness and wrongness  6, 101–18 Preston-Roedder, Ryan  272 pretense 47 Prichard, Harold A.  6, 107–15, 117 prima facie duty  6, 101, 102–14 prima facie justification  103 private conscience  8–9, 135–56 private reason  7, 125, 126, 127, 128 promises, false  179 Proust, Marcel, Remembrance of Things Past  45 public authority  8–9, 135–6, 141–56 conscience  8–9, 135–56 justification and the inclusive conscience 151–4

Index  309 reason 281–3 autonomy and  6–7, 119–30 conscience 140–1 deliberative democracy  120–3 form and scope  127–9 necessary conditions  124–7 shipwrecks 173 technological fixes and quasi communication 123–4 today 129–30 sphere 120–1 punishment 284–5 quasi communication  123–4, 130 Quinn, Warren  272 race emotions and facial expression  36 servility and self-respect  19, 39, 40 slavery  36, 37, 40 racism  201, 205 rational consent 7 faculty of desire  236, 248–9 volition  13–14, 232–54, 294–5 bad willing  250–4 universal validity condition  235–50 rationality  204, 205 animals  195, 202 complicity 263 Hume 4 humility and arrogance  61, 62, 63 instrumental  5–6, 89, 92–9 servility 51 shipwrecks 173–4 will  13–14, 234–40, 241–3, 251 Rawls, John  1, 15, 271–2, 283–4 apathy  22, 23 burdens of judgment  149 conscience  152 n. 22 constructivism 292 homicide law  172 moral community  79 “original position”  292 public reason  7, 120–3, 124, 125–6, 173, 281, 282 respect and self-respect  273 Raz, Joseph  92, 93, 144 n. 11 reason law-like nature  126, 127–8 practical  see practical reason private  7, 125, 126, 127, 128 public  see public reason reasonable disagreement, fallibility contrasted with 148–50 Reath, Andrews  13–14, 294–5 recognition respect  3, 70, 72–3, 74, 75, 76–7, 277

recommendation and commendation, distinction between  115–16 Regina v Dudley and Stephens  9, 169–75, 176, 288 religion conscience  136–7, 140, 142, 148, 149 homicide 175 honor and accountability  70–3, 74, 84 humility and arrogance  44, 46, 47–8 n. 18 sacrifice 172 republican literature domination and servility  25, 29–30, 33, 34, 35 wives 24 Repugnant Conclusion  9–10, 178, 182–4, 188, 191–2, 289 intuition and theory  190–1 the Other  184–5, 186 n. 11, 191 overpopulation  10, 185, 189–90 requirements, deontic buck-passing  117 resentment honor disrespect  4, 80, 82 second-personal respect  81 respect  1–4, 272–7 appraisal  3, 72, 75–7, 277 and complicity  261, 262 as honor and as accountability  3–4, 70–84 humility and arrogance  47, 48, 49 interpersonal recognition  49, 50, 51, 58 recognition  3, 70, 72–3, 74, 75, 76–7, 277 second-personal  4, 72, 73, 74–5, 76, 77–81, 83, 277 see also self-respect revolution 284 Ride the High Country (MGM, 1962)  56 n. 29 right of necessity  9, 165–9, 171–3, 175, 176–7, 288–9 rightness, and duty  104, 106, 107, 110, 114, 115, 116 rights animal  186, 192 human  72, 192 moral  see moral rights Ripstein, Arthur  173 n. 31 Ross, W. D. duty  101–18, 280 Hill’s teaching of  101 moral intuitionism  270 moral reason  6, 101 Ross’s Paradox  93 n. 9 Rousseau, Jean-Jacques  83 n. 22, 83 n. 23 Routley, Richard  187 rule of law  164, 165 n. 15 Ryle, Gilbert  15, 270, 271 Sachs, David  23, 266 n. 20 saints and saintly acts  215–16, 228–9 Sandler, Ronald  188 Sarkozy, Nicolas  43 n. 2

310 Index Sayre-McCord, Geoffrey  272 Scanlon, T. M.  272 Schindler, Oskar  230 Schmidtz, David nature, respect for  192 n. 18 virtue, repugnance, and deontology  9–10, 289–90 Schroeder, Mark  5–6, 89–99, 280 Schwartz, David T.  258 n. 5 scope, hypothetical imperative  5, 91–9 sea, custom of the  170, 171, 174, 175 second-personal respect  4, 72, 73, 74–5, 76, 77–81, 83, 277 selfabasement  51, 53 abnegation  46, 47 n. 18 acceptance 46–7 conceit  52, 61–3, 64, 65, 251–2 defense, killing in  165–9 degradation  48, 49, 203 denial 47–8 Deprecator  2, 19–20, 21–4, 39, 48, 273, 275 esteem arrogance and servility  53 complicity  15, 266, 296 definition  53 n. 26 humility and arrogance  61–2, 63, 64, 65 motivation 57 importance  3, 46, 47–8 knowledge, striving for  227 legislation 128–9 love 251–2 respect  1–4, 272–7 agentive recognition  3, 58, 59–61, 65–6 appraisal  56 n. 31 complicity  264, 265–6 estimative  56 n. 30 evaluative  55–7, 58, 60, 61, 65 humility and arrogance  3, 42–66 interpersonal recognition  49, 50, 51, 53, 54 n. 27, 58, 59, 65 servility and  1–2, 17–40 service animals  197–200, 206, 208, 209, 210 servility  1–2, 17–40, 273–4, 275–6 complicity  265, 266 Deferential Wife  24–38 Hill’s theory, and preliminary critique  19–21 humility and  3, 45, 48–9, 50–1, 52, 53, 54, 55, 59 morally defective attitude toward morality 209 preliminary critique of Hill  20–1 Self-Deprecator 21–4 Setiya, Kieran  93 n. 8 Shakespeare, William,  Macbeth 36 Shall, Sam  97, 97 n. 17 shame animal cruelty  204

deontic buck-passing  115 domination 31 homicide law  9, 160–1, 162–3, 167 n. 18, 176 honor disrespect  4, 74, 80, 82 Shaw, George Bernard  123 n. 16 shipwrecks  9, 166–9, 176, 288 Regina v Dudley and Stephens  9, 169–75, 176, 288 Shope, Robert  272 Sidgwick, Henry  102 Sidney, Algernon  29, 30, 34 Sigler, Mary  160 n. 7 Singer, Marcus G.  273–4 Skidmore, J.  202 n. 5 Skinner, Quentin  29 slavery  21–2, 28, 30, 33, 34–5, 36, 37, 40 Fugitive Slave Act  264, 265 Slote, Michael  272 small-mindedness 209–10 Smith, Adam  31, 83 n. 23 Smith, Steven  272 Smuts, Barbara  198 social norms conscience  139, 145–6 homicide law  163, 164, 165 n. 15, 168 social ostracism  160 n. 7, 161 n. 8 Solomon, Robert  272 Speer, Albert  138 n. 8 spontaneity  238, 238 n. 14, 242 Steiker, Carol  165 n. 15 Stern, Lawrence  272 Stocker, Michael  272 Strawson, P. F. contempt 81 moral responsibility  79 reactive attitudes  79 moral intuitionism  270–1 Strozzi, Bernardo,  The Tribute Money 70–1, 72–3 suicide  173, 174, 203, 285 superbia 44 supererogatory  12–13, 215–31, 293 servility and self-respect  26 Sussman, David  165 n. 15 Swanton, Christine  65 n. 39 sympathy domination  33, 35–6, 38 humility and arrogance  54 n. 27 tacit consent  174–5 Taurek, John  272 technology, communications  123–4 Telfer, Elizabeth  56 n. 30 theory and intuition  190–1 Thoreau, Henry David  264 n. 16 tolerance complicity  14–15, 257–8, 260, 265–6 Tosi, Justin  165 n. 15

Index  311 Trianosky, Gregory  272 truth 83 Uncle Tom  2, 19–20, 39–40, 48, 51 n. 22, 59, 273, 275–6 Uniacke, Suzanne  167 n. 17, 169 n. 20 universal human rights  72 Urmson, J. O.  270, 271 duties  225, 226 morality understood as a code  226 saints and heroes  215, 293 utilitarianism animals  194, 195, 196–7, 198, 199, 201 complicity 258 ideal  15, 270 right of necessity  173 n. 30 virtue, repugnance, and deontology  9–10, 178, 178 n. 2, 182–5, 189, 190, 191 value of humans  195 valuing animals  11, 203–6 van Roojen, Mark  94 n. 13 Venizelos, Evangelos  43 n. 2 verdictive, prima facie duty  107 vice, true  239–40 n. 18 Vigilantius, Johann Friedrich  222 n. 20 virtue  9–10, 178–92, 289–90 environmental ethics and inadequacy of act-centered theories  181–2 fantastic  228 n. 33 imperfect duties  220 n. 16 intuition and theory  190–1 maxims and character  179–81 moral theory  187–9 overpopulation 189–90 Repugnant Conclusion  182–6 vivisection  195–6 n. 2 volition, rational  13–14, 232–54, 294–5 bad willing  250–4 universal validity condition  235–50

Waldron, Jeremy  83 n. 24 Wall Street collapse  43 n. 2, 44 Warnock, Geoffrey  270 Wasserstrom, Richard  272 Way, Jonathan  92 weakness  239–40 n. 18, 250 n. 30, 253–4 Wedgwood, Ralph  93 n. 10 Werner, Louis  272 Werthheimer, Roger  272 Wilde, Oscar  123 n. 16 will animals 195 autonomy of the  14, 96, 97–8, 99, 244–5, 246 n. 27 bad willing  250–4 elective conception of the  240–2, 243, 246–7, 249 free, and a will under moral laws  244, 245–6 imperatives 95 norm-constituted power conception  233 n. 6 practical-cognitivist conception  233 rationalist thesis about the  13–14, 234–40, 241–3, 251 Williams, D. C.  270 Wilson, Margaret  272 wisdom, humility and  3, 47 withdrawing citizen  263–4, 265, 266 Wolf, Susan  272 Wolff, Robert Paul  150 Wood, Allan animals 202 categorical imperative  49 n. 20 false promises  179 moral worth and merit, relation between  219 n. 13 wrongness, and duty  104, 115, 116 Zakaras, Alex  263–4 n. 15, 264 n. 16 Zweig, Arnulf  272 Zwolinski, Matt  9–10, 289–90