Aristotelian Ethics in Contemporary Perspective

1,020 59 1MB

English Pages 238 [245] Year 2013

Report DMCA / Copyright


Polecaj historie

Aristotelian Ethics in Contemporary Perspective

Table of contents :
Pt. I. Themes in (neo-)Aristotelian virtue ethics --
part II. Beyond (neo-)Aristotelian virtue ethics.

Citation preview

Aristotelian Ethics in Contemporary Perspective

By bringing together influential critics of neo-Aristotelian virtue ethics with some of the strongest defenders of an Aristotelian approach, this collection provides a fresh assessment of the strengths and weaknesses of Aristotelian virtue ethics and its contemporary interpretations. Contributors critically discuss and re-assess the neo-Aristotelian paradigm that has been predominant in the philosophical discourse on virtue for the past 30 years. Julia Peters is Lecturer in Philosophy at the University of Bonn, Germany.

Routledge Studies in Ethics and Moral Theory

1 The Contradictions of Modern Moral Philosophy Ethics after Wittgenstein Paul Johnston

12 Challenging Moral Particularism Edited by Mark Norris Lance, Matjaž Potrč, and Vojko Strahovnik

2 Kant, Duty and Moral Worth Philip Stratton-Lake

13 Rationality and Moral Theory How Intimacy Generates Reasons Diane Jeske

3 Justifying Emotions Pride and Jealousy Kristján Kristjánsson 4 Classical Utilitarianism from Hume to Mill Frederick Rosen 5 The Self, the Soul and the Psychology of Good and Evil Ilham Dilman 6 Moral Responsibility The Ways of Scepticism Carlos J. Moya 7 The Ethics of Confucius and Aristotle Mirrors of Virtue Jiyuan Yu 8 Caste Wars A Philosophy of Discrimination David Edmonds 9 Deprivation and Freedom A Philosophical Enquiry Richard J. Hull 10 Needs and Moral Necessity Soran Reader 11 Reasons, Patterns, and Cooperation Christopher Woodard

14 The Ethics of Forgiveness A Collection of Essays Christel Fricke 15 Moral Exemplars in the Analects The Good Person is That Amy Olberding 16 The Phenomenology of Moral Normativity William H. Smith 17 The Second-Person Perspective in Aquinas’s Ethics Virtues and Gifts Andrew Pinsent 18 Social Humanism A New Metaphysics Brian Ellis 19 Ethics Without Morals In Defence of Amorality Joel Marks 20 Evil and Moral Psychology Peter Brian Barry 21 Aristotelian Ethics in Contemporary Perspective Edited by Julia Peters

Aristotelian Ethics in Contemporary Perspective Edited by Julia Peters



First published 2013 by Routledge 711 Third Avenue, New York, NY 10017 Simultaneously published in the UK by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2013 Taylor & Francis The right of the editors to be identified as the author of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Aristotelian ethics in contemporary perspective / edited by Julia Peters. p. cm. — (Routledge studies in ethics and moral theory ; 21) Includes bibliographical references and index. 1. Aristotle. 2. Virtue. 3. Ethics. I. Peters, Julia, 1978– B491.E7A74 2012 171′.3—dc23 2012027696 ISBN: 978-0-415-62341-4 (hbk) ISBN: 978-0-203-07276-9 (ebk) Typeset in Sabon By Apex CoVantage, LLC


Introduction: Aristotelian Ethics in Contemporary Perspective



PART I Themes in (Neo-)Aristotelian Virtue Ethics 1 Aristotle on Virtue: Wrong, Wrong, and Wrong



2 Aristotle on Virtue: A Response to Hurka



3 The Benefit of Virtue



4 Well-Being and Eudaimonia: A Reply to Haybron



5 Virtue, Personal Good, and the Silencing of Reasons



6 Human Nature, Virtue, and Rationality JOHN HACKER-WRIGHT


7 Good (as) Human Beings





8 Attachment Theory, Character, and Naturalism



9 Notes Toward an Empirical Psychology of Virtue: Exploring the Personality Scaffolding of Virtue



10 Natural Virtue and Proper Upbringing



11 Kalou Heneka



PART II Beyond (Neo-)Aristotelian Virtue Ethics 12 A New Metaphysics for Virtue Ethics: Hume Meets Heidegger



13 A Kantian Plea for Virtues?



14 Toward a Humean Virtue Ethics



List of Contributors Index

225 229

Introduction Aristotelian Ethics in Contemporary Perspective Julia Peters

THE BACKGROUND DEBATE Since Elizabeth Anscombe’s programmatic call for a return to the notion of virtue in moral philosophy in 1958, the modern revival of virtue ethics has been predominantly associated with an Aristotelian approach.1 Most philosophers who followed Anscombe’s call turned to the works of Aristotle to find inspiration for a way of doing moral philosophy that, according to Anscombe’s intention, was not only to stand alongside the Kantian and Utilitarian paradigm, but also to be superior to them in several respects. As a result, the contemporary philosophical discussion of virtue has been primarily carried out within an Aristotelian framework—contemporary virtue ethics and neo-Aristotelian ethics came to be used almost synonymously. Today, more than 50 years of scholarly endeavor after Anscombe’s appeal, we are in the fortunate position of possessing a body of literature in which her programmatic suggestions have been transformed into a well-established branch of modern moral philosophy with many followers and defenders. The 1970s, ’80s, and ’90s were particularly productive decades for the development of neo-Aristotelian virtue ethics: they brought forth a number of seminal monographs and articles in which some of the central tenets of this line of moral philosophy received their most comprehensive and influential interpretation to date.2 Many of these central tenets can be traced back to Anscombe’s programmatic treatment of the subject in 1958. One is the idea that the notion of virtue is tied to that of human flourishing, in both its subjective and its objective meaning. Thus neo-Aristotelian virtue ethicists tend to be sympathetic to ethical naturalism and consider it one of the strengths of their theories that they seek to ground morality in human nature by associating moral excellence with human flourishing understood as natural human perfection. Regarding the subjective connotation of flourishing, they tend to share the ambition of demonstrating that the virtues benefit their possessor and enable him to flourish in the sense of leading a happy life. Even more essentially, neo-Aristotelian virtue ethicists obviously share the view that the notion of virtue, and the associated conception of acting virtuously or acting ‘from’ virtue, play a crucial if not fundamental role in moral philosophy. However, they also tend to interpret


Julia Peters

the notion of virtue in a specific fashion, inspired by Aristotle. That is to say they understand the virtues as morally excellent character traits that manifest themselves in their possessor’s actions and comprise two essential elements, harmoniously united: habit on the one hand, practical reason on the other. In various interpretational versions, these tenets are among the central assumptions that define the neo-Aristotelian virtue ethical school of thought today. However, after this ‘golden age’ of neo-Aristotelian virtue ethics, its discussion has changed significantly in the past 15 years or so. On the one hand, scholars have increasingly begun to raise severe criticism of its central tenets, or of the approach as a whole. On the other hand, alternative approaches to virtue have begun to be given increasing scholarly attention. As a result, Aristotelianism today can no longer claim the status of an unrivaled, undisputed contemporary approach to virtue. The criticisms raised against it come from various directions: they are philosophical as well as empirical. Thus, for instance, Aristotelian eudaimonism—as well as the self-centered conception of moral motivation supposedly associated with it—has been confronted with sharp philosophical objections.3 The same is true for neoAristotelian naturalism.4 Similarly, the Aristotelian conception of character has been seriously challenged, beginning in the late 1990s, by several waves of ‘situationist’ criticism raised against it by philosophers drawing on a body of literature from social and cognitive psychology.5 In this more critical climate, moreover, several monographs have emerged that present powerful alternatives to the neo-Aristotelian approach to virtue: for instance, Christine Swanton’s Virtue Ethics: A Pluralistic View; Thomas Hurka’s Virtue, Vice and Value; or Robert Adams’s A Theory of Virtue.6 The aim of this volume is to reflect and give voice to this newly emerged critical stance and thus to represent an up-to-date philosophical perspective on neo-Aristotelian virtue ethics. This critical perspective manifests itself in the volume in three different ways. Some chapters present criticisms of central tenets of the Aristotelian or neo-Aristotelian approach to virtue. Others offer defenses of Aristotelian views, in particular in light of recent criticisms that have been raised against the central tenets of neo-Aristotelian virtue theory. Finally, some chapters defend alternative accounts of virtue ethics to be set alongside the Aristotelian approach.7 In this way, by bringing together influential critics and defenders of Aristotelian ethics, while at the same time exploring alternative approaches to virtue, this collection seeks to provide a fresh basis for assessing the strengths and weaknesses of Aristotelian virtue ethics and its contemporary interpretations.

THE CHAPTERS Part I of the volume opens with a decidedly critical piece. In his chapter ‘Aristotle on Virtue: Wrong, Wrong, and Wrong’, Thomas Hurka advances penetrating criticisms of some of the core theses of the Aristotelian approach



to virtue. Hurka challenges the Aristotelian tendency to blur the distinction between the good and the right by making the virtues, which are constitutive of a person’s goodness, objects of praise or blame. He puts into question the Aristotelian doctrine of the mean and the idea that vice can always be explained in terms of either excess or deficiency. Most importantly, he challenges what he calls the foundational egoism of Aristotelian virtue theory, according to which a virtuous person’s ultimate reason for being virtuous is grounded in their concern for their own flourishing. Hurka contrasts his criticism with a sketch of his own recursive theory of virtue, which is opposed to the Aristotelian approach in crucial respects and thus suggests itself as an attractive alternative to it. Anthony Price’s reply to Hurka, ‘Aristotle on Virtue: A Reply to Hurka’—which can also be read as an independent piece reflecting on Aristotle’s conception of the motivation underlying virtuous action—argues that contrary to Hurka’s charge, the Aristotelian virtuous agent’s concern for his eudaimonia is not culpably egocentric. Rather, it is expressive of an essential aspect of morality: namely, the fact that each agent must be centrally concerned with, and take responsibility for, his own good or bad action. The exchange between Hurka and Price is followed by a block of chapters on eudaimonism. Christoph Halbig’s chapter ‘The Benefit of Virtue’ offers a critical examination of the thesis that the virtues benefit their possessor. Halbig looks at different versions of the claim represented in the contemporary neo-Aristotelian literature and reaches the conclusion that the most plausible and defensible version is much weaker than the one favored by most neo-Aristotelian virtue ethicists. Daniel Russell and Mark LeBar, in their chapter ‘Well-Being and Eudaimonia: A Reply to Haybron’, defend Aristotelian accounts of well-being against Daniel Haybron’s criticism.8 They argue that Aristotelian eudaimonism, which makes individual well-being at least in part dependent on the possession of features that are essential to our human nature, can be defended against Haybron’s objections and is more in tune with our intuitions about human well-being than Haybron’s alternative individualist account. In her chapter ‘Virtue, Personal Good, and the Silencing of Reasons’, Julia Peters suggests a way of reconciling McDowell’s claim that the virtuous never have to make genuine sacrifices when acting virtuously with the truism that even the virtuous agent (sometimes) has a reason for regret when forgoing a personal good for the sake of virtue. Three chapters take up the theme of naturalism. John Hacker-Wright, in ‘Human Nature, Virtue, and Rationality’, offers a defense of neo-Aristotelian naturalism by arguing that its appeal to human nature not merely establishes the fact that humans are essentially rational agents, but also can function as a basis for deriving substantive norms for human rational conduct. Philipp Brüllmann, in his chapter ‘Good (as) Human Beings’, argues that neo-Aristotelian naturalism—in particular in the version defended by Rosalind Hursthouse—implies a problematic tension because it puts a conceptual constraint on our conception of the moral good that is incompatible with the


Julia Peters

function it is supposed to play in testing the correctness of individual moral judgments, in particular, judgments about which character traits are virtues. Edward Harcourt’s chapter ‘Attachment Theory, Character, and Naturalism’ draws on attachment theory—a prominent empirical theory of child development—in order to suggest a ‘modest’ version of neo-Aristotelian ethical naturalism that conceives of human ethical life as continuous with our lives understood in psychobiological terms. This modest version, Harcourt argues, is to be preferred to more ambitious forms that seek to identify judgments about virtues and vices with judgments about natural perfections and defects. Nancy Snow, in ‘Notes Toward an Empirical Psychology of Virtue: Exploring the Personality Scaffolding of Virtue’, draws on empirical psychology in order to offer a response to the situationist challenge of the Aristotelian conception of (virtuous) character. Snow seeks to show that contrary to the situationist challenge, empirical psychology provides support for the kind of global character traits that, on Aristotelian conceptions, constitute the virtues. In her chapter ‘Natural Virtue and Proper Upbringing’, Candace Vogler offers a reading of the (neo-)Aristotelian view that proper upbringing is essential to the development of virtue and of the insight underlying corrective accounts of the virtues. Proper upbringing, Vogler argues, facilitates the development of natural virtue, which is a genuine, yet inchoate form of virtue, limited to constraining deviant natural impulses. Full virtue, in contrast, has to be combined with the possession of right practical reason. However, even full virtue maintains a certain corrective function since it is one mark of a virtuous character that its possessor is open to moral self-correction. Timothy Chappell, in his chapter ‘Kalou Heneka’, draws on Aristotle in order to defend the view that the category of the Noble or the Beautiful, which is often evoked in the explanation of virtuous action, constitutes a category sui generis of reasons for action that need not be reduced to the category of moral or prudential reasons in order to carry explanatory force. The second part of the volume contains three chapters devoted to exploring alternatives to the Aristotelian approach to virtue. Christine Swanton’s ‘A New Metaphysics for Virtue Ethics: Hume Meets Heidegger’ draws on Heidegger—in particular, his conceptions of aletheia and of ‘horizons of disclosure’—in order to develop the metaphysical scaffolding for a Humean, response-dependent type of virtue ethics. Erasmus Mayr, in his chapter ‘A Kantian Plea for Virtues?’, presents a Kantian framework for thinking about virtue, arguing that there is a distinctive and irreducible role to be played by the virtues—understood as character traits a human being needs in order to lead a good life—even within Kant’s ethics. In ‘Toward a Humean Virtue Ethics’, Lorenzo Greco gives an outline of a Humean virtue ethics that stresses the crucial differences between a Humean and an Aristotelian approach to virtue, but emphasizes at the same time that the peculiarities of Hume’s account have to be considered as potential strengths, rather than weaknesses of his virtue ethical position.



NOTES 1. Anscombe, ‘Modern Moral Philosophy’. 2. For instance Foot, Virtues and Vices; Annas, The Morality of Happiness; McDowell, ‘Virtue and Reason’; McDowell, ‘Two Sorts of Naturalism’; Hursthouse, On Virtue Ethics. 3. See for instance Hurka, Virtue, Vice and Value; Swanton, Virtue Ethics: A Pluralistic View. 4. See for instance Copp and Sobel, ‘Morality and Virtue: An Assessment of Recent Work in Virtue Ethics’. 5. See Harman, ‘Moral Philosophy Meets Social Psychology: Virtue Ethics and the Fundamental Attribution Error’; Doris, Lack of Character; more recently: Merritt, Doris, and Harman, ‘Character’. 6. While proposing theories of virtue, neither Hurka nor Adams, however, are virtue ethicists in the sense of considering virtue as a fundamental concept in moral philosophy. 7. It is of course not possible to neatly sort the chapters in the collection into these three categories because some of them pursue more than one of these aims. 8. Russell and LeBar are referring in their article to Daniel Haybron’s discussion in his much acclaimed book The Pursuit of Unhappiness.

REFERENCES Adams, Robert, A Theory of Virtue (Oxford: Oxford University Press, 2006). Annas, Julia, The Morality of Happiness (Oxford: Oxford University Press, 1993). Anscombe, G.E.M., ‘Modern Moral Philosophy’, Philosophy 33 (1958), 1–19. Copp, David, and David Sobel, ‘Morality and Virtue: An Assessment of Recent Work in Virtue Ethics’, Ethics 114 (2004), 514–554. Doris, John, Lack of Character (Cambridge: Cambridge University Press, 2005). Foot, Philippa, Virtues and Vices and Other Essays in Moral Philosophy (Berkeley: University of California Press, 1978). Harman, Gilbert, ‘Moral Philosophy Meets Social Psychology: Virtue Ethics and the Fundamental Attribution Error’, Proceedings of the Aristotelian Society 99 (1999), 315–331. Haybron, Daniel, The Pursuit of Unhappiness (Oxford: Oxford University Press, 2008). Hurka, Thomas, Virtue, Vice, and Value (New York: Oxford University Press, 2001). Hursthouse, Rosalind, On Virtue Ethics (Oxford: Oxford University Press, 2001). McDowell, John, ‘Two Sorts of Naturalism’, in Virtues and Reasons: Philippa Foot and Moral Theory, ed. R. Hursthouse, G. Lawrence, and W. Quinn (Oxford: Clarendon Press, 1995), 149–179. ———, ‘Virtue and Reason’, The Monist 62 (1979), 133–150. Merritt, Maria W., John M. Doris, and Gilbert Harman, ‘Character’, in The Moral Psychology Handbook, ed. John M. Doris and the Moral Psychology Research Group (New York: Oxford University Press, 2010), 355–401. Swanton, Christine, Virtue Ethics: A Pluralistic View (Oxford: Oxford University Press, 2003).

Part I

Themes in (Neo-)Aristotelian Virtue Ethics


Aristotle on Virtue Wrong, Wrong, and Wrong Thomas Hurka

Recent decades have seen a revival of philosophical interest in moral virtue. Prompted initially by an article of Elizabeth Anscombe’s,1 it has generated a school of thought called ‘virtue ethics’ that’s now often seen as a third main ‘method of ethics’ alongside consequentialism and deontology. While Mill and Kant are the classical exponents of these views, the classical exponent of virtue-based ethics is commonly taken to be Aristotle; the rise of virtue ethics has therefore been the rise of an Aristotelian approach to the subject. I agree that moral virtue is an important moral concept, but I think Aristotle is the wrong figure to look to for insight into it. Many of his central claims about virtue are mistaken, and present-day virtue-ethical theories that embrace them are therefore misguided. This chapter develops a critique of Aristotle’s account of virtue, but it first sketches a better account by contrast with which the flaws in his become evident.

VIRTUE AS A HIGHER-LEVEL GOOD This account was widely accepted in the late nineteenth and early twentieth centuries—in Britain by Hastings Rashdall, G. E. Moore, W. D. Ross, and others, in Europe by Franz Brentano and his followers.2 It treats virtue as a higher-level moral concept, involving a relation to items falling under other, independently applied moral concepts. More specifically, it sees the virtues as intrinsic goods that involve morally fitting attitudes to items with other moral properties, and the vices as evils involving unfitting attitudes. The account’s first proponents were consequentialists and therefore took all the virtues and vices to involve attitudes to items falling under the consequentialist concepts of good and evil. One of their claims was that if something is intrinsically good, then having a positive attitude toward it, that is, desiring, pursuing, or taking pleasure in it—in short, loving it—for itself is another intrinsic good and a form of virtue. Thus, if your pleasure is intrinsically good, my desiring, pursuing, or taking pleasure in it is also good and an instance of virtue, more specifically of benevolence. By contrast, if something is intrinsically evil, loving it for itself is another evil and vicious;


Thomas Hurka

thus, my desiring, pursuing, or taking pleasure in your pain for itself is evil and, more specifically, malicious. The fitting and therefore virtuous attitude to an evil such as your pain is negative, involving desire for or pursuit of its absence, or pain at its presence; this hating your pain for itself is good and involves the virtue of compassion. But hating something good, as when I enviously want your pleasure to end, is vicious and evil. Attitudes whose orientation matches the value of their object—positive to positive or negative to negative—are virtuous and good, while ones that oppose it are vicious. There can also be deontological virtues. If an act is right, my wanting to perform it because it’s right is fitting and therefore virtuous—it involves conscientiousness, or a Kantian good will. And it’s likewise virtuous to hate doing what’s wrong. But whether its object is good or right, a virtuous attitude need not care about it as good or right. If your pleasure is good, my wanting it because it’s good is virtuous, but so is my wanting it just because it’s a pleasure and independently of any thoughts about goodness. Likewise, my hating lying is virtuous not only when I think of lying as wrong but also when I just don’t like lying. An attitude to something good or right for the properties that make it so is virtuous even when it doesn’t think of them as good- or right-making. A complete higher-level account must also say how virtuous or vicious different attitudes are. Here it’s guided by an ideal of proportionality, which says it’s best to love objects in proportion to their degrees of goodness or evil. Thus, a fully virtuous person will be more pleased by another’s intense pleasure than by her mild pleasure, and by as much as the first pleasure is more intense; he’ll likewise be more anxious to relieve a worse pain. Something similar holds for deontological virtues. If some act’s being an instance of lying does more to make it wrong than its promoting pleasure does to make it right, he’ll be more averse to it as an instance of lying than drawn to it as a promoting of pleasure. However exactly it’s developed, the higher-level account treats the moral virtues as intrinsically good, so they have value not just instrumentally, or for the other goods they promote, but also in themselves. Being benevolent by itself makes your life better and being malicious makes it worse. But the account also makes virtue in several ways a secondary moral concept. First, as a response to items falling under other moral concepts, it can’t be the only or main such concept; unless other things are independently good or right, there’s nothing for it to care fittingly about. Second, as so understood virtue plays only a minor role in the evaluation of actions. Imagine that you can give either a large pleasure to one person or a small pleasure to another. Given the ideal of proportionality, it’s most virtuous to desire the larger pleasure more than the smaller and therefore to produce the larger pleasure. But the claim about virtue isn’t needed to establish that you ought to produce the larger pleasure. That already follows from the fact that it’s the greater good, or from that plus the claim that you ought to produce the most good you can. That in doing so you’ll also act from the most virtuous motive may

Aristotle on Virtue


be an additional reason to do the independently right act, but it can’t change what this act is; that already follows from the facts that make your motive best. Finally, and departing from many of the account’s proponents, I think virtue is a lesser intrinsic good in the sense that it always has less value than its intentional object. Compassion for another’s pain is good, but it isn’t more good than the pain is evil; it can’t be better for there to be pain and compassion for it than no pain and no compassion. Likewise for vice: a torturer’s malicious pleasure in his victim’s pain isn’t as evil as the victim’s pain. If you can eliminate only one of the two, you ought to eliminate the pain. This is a brief sketch of a ‘higher-level’ account of virtue, and when we turn to Aristotle’s account, we find several points of similarity. He too thinks moral virtue is good in itself, contributing to a desirable life not just instrumentally but in its own right. He also thinks virtue is a matter of your attitudes broadly conceived, of your desires, motives in acting, and pleasures and pains. An act’s virtuousness depends not on its effects or conformity to external moral rules but on inner states such as its motive and accompanying feelings. But on other central issues he’s mistaken.

PRAISE AND BLAME First a smaller point. In the Nicomachean Ethics Aristotle says the virtues and vices are traits for which we’re praised and blamed (1105b31–1106a1, 1106a7).3 Since he recognizes that praise and blame are appropriate only for things under our voluntary control (1109b30–33), he must hold that virtue and vice are voluntary, and he defends that view in NE III.5. But his arguments for it are unpersuasive. In one passage he seems to argue that it’s always in our power to act virtuously (1113b3–6), but a virtuous action must be done from a virtuous motive, and someone who’s vicious can’t now produce a virtuous motive in himself.4 He also argues that even if a vicious person can’t now act virtuously, he’s responsible for his vicious action because he could have avoided developing his bad character in the past: vicious people ‘are themselves by their slack lives responsible for becoming men of that kind’ (1114a3–5). But this claim is hard to square with his insistence on the importance for moral virtue of the right childhood training and education (1095b4–12, 1103a15–18, 1103b3–6, 24–25, 1104b11–13,1105a1–2, 1179b24–27). If you were raised badly by vicious parents, how could you start to develop virtuous desires, and if you couldn’t start, how can you be blamed for not having them now? The concepts of praise and blame, like those of right and wrong, presuppose voluntariness: you can’t have acted wrongly or be to blame unless you could have done otherwise. But no such requirement governs the concepts of good and evil. A serendipitous pleasure is good even if no one voluntarily produced it, and pain evil even when it’s no one’s fault. The higher-level account


Thomas Hurka

makes use of only these last two concepts. It says virtue is intrinsically good and vice evil, and they can be so regardless of how they came about. If Hume and Mill were right that we have innate tendencies to be pleased by others’ pleasure and pained by their pain, the account says we’re naturally virtuous and good, though we deserve no credit for this. If we innately delight in others’ pain, as a bleaker view has it, we’re naturally vicious but not blameably so. Aristotle’s claim that virtue is praised and vice blamed applies the wrong concepts to them, forcing him into implausible arguments about voluntary control. Those arguments aren’t needed if virtue and vice are instead said to be just good or evil.

DISPOSITIONS VS. OCCURRENT STATES Another issue concerns the primary locus of virtue. We make virtue ascriptions at two levels, one more global and one more local. Speaking globally, we may say someone has the character trait of generosity or is a generous person. More locally, we may say a particular act was generous or a particular feeling malicious. Is one of these two types of ascription primary? Do we first understand the virtues as traits of character and count individual acts or feelings as virtuous only when they issue from such traits? Or do we first identify individual motives and feelings as virtuous and understand a virtuous character as one that tends to produce them? The higher-level account takes the second view, ascribing virtue properties first to occurrent states such as individual desires, acts, and feelings and only then to dispositions. However, Aristotle takes the first view. He defines virtue as a state of character (hexis) (1105b20–1106a13) and says that to be done virtuously an act must issue from a ‘firm and unchangeable character’ (1105a33–34), otherwise it may be ‘in accordance with the virtues’ (1105a29) but it can’t be fully virtuous. Aristotle doesn’t think the mere possession of virtue is the highest good; that comes only in the active exercise of virtue, as in particular virtuous acts (1095b32–34,1098b33–1099a6). But they’re only done virtuously if they issue from a stable character. I think this view is both false to our everyday understanding of virtue and morally mistaken. If you see someone kick a dog just for pleasure, do you say ‘That was a vicious act, on condition that it issued from a stable disposition to perform similar acts on similar occasions’, or just ‘That was a vicious act’. Surely you say the latter. Your remark doesn’t concern only the kick’s physical properties; it turns essentially on the motive from which it was done. But it concerns only its motive at the time, independently of any longer-lasting trait. Or imagine that a friend who normally doesn’t do this gives $20 to a homeless person from concern at the time for his welfare. If you say ‘That was uncharacteristically generous of you’, you don’t contradict yourself. Or imagine that we’re a military committee deciding whether to give a medal for bravery to a soldier who threw

Aristotle on Virtue


himself on a hand grenade, knowing it would kill him and in order to save his comrades. If an Aristotelian says ‘This is a medal for bravery, and we can’t know whether his act was brave unless we know whether he would have acted similarly a week before or a week after’, we’ll throw him out of the room.5 Nor is the issue here just one of terminology. ‘Virtue’ is an evaluative term, in that to call something virtuous is to call it somehow good, and Aristotle’s claim that acts not expressing a virtuous character aren’t done virtuously implies that they aren’t fully good: since they don’t involve the ‘exercise of virtue’, they can’t make the same contribution to your good as ones that do. (Perhaps they make no contribution.) And that seems wrong. Considered just in itself and apart from the other things co-present with it in a life, an out-of-character act of generosity or courage seems every bit as good as one based in a stable disposition. The second act may be accompanied by more acts of similar value in the same life, and that life may be better as a whole, perhaps even in part because it contains enduring virtuous dispositions.6 But Aristotle’s claim that the in-character act is by itself better is unpersuasive. Both analytically and evaluatively, the primary locus of virtue is occurrent desires, actions, and feelings apart from any connection to more stable traits.

THE DOCTRINE OF THE MEAN A further issue concerns Aristotle’s differentia for the virtues among traits of character, his doctrine of the mean. It says that every virtue is a mean between two vices, and every vice an excess or deficiency with respect to the same feeling as concerns some virtue. Thus the virtue of temperance is a mean with respect to the desire for physical pleasure, a desire the excess of which is self-indulgence and the deficiency of which is insensibility. Courage is a mean with respect to fear, of which the excess is cowardice and the deficiency rashness. Many present-day Aristotelians distance themselves from the doctrine of the mean, but I think something like it can be part of an adequate account of virtue. It can’t be the whole, however, most clearly because of what it says about vice. By taking all the vices to involve excess or deficiency, the doctrine implies that there are no basic human impulses that are always evil: each is such that in a proper or medial form it’s virtuous and good. But this leaves out the worst forms of vice, such as malice and cruelty, which involve desire for or pleasure in another’s evil. No form of these feelings is good; all their instances are bad.7 The higher-level account makes traits like malice its central cases of vice because they involve the positively unfitting attitudes of loving an evil or hating a good. However, they’re excluded by the doctrine of the mean, and it’s therefore no surprise that they don’t figure in Aristotle’s main catalogue of vices in NE II–IV. These books discuss self-indulgence, cowardice, profligacy, and other vices but not the positive desire for harm


Thomas Hurka

to others that’s intuitively the worst vice of all. Aristotle does mention this desire in the Rhetoric (1382a1–16, 1386b33–1387a1, 1387b22–24), but that work doesn’t contain the doctrine of the mean, and when that doctrine appears in the NE, vices like malice don’t. Aristotle may seem to allow for these vices when he says that not all feelings admit of a mean since some such as spite and envy have names that already imply badness (1107a9–13). But his explanation is that if we attach a name to the excess or deficiency of some feeling, such as ‘gluttony’ to the excessive desire for food, there will be no mean with respect to it because there’s in general no ‘mean of excess and deficiency’ (1107a25). That’s precisely how he understands spite and envy, as the excess and deficiency of another feeling that can be virtuous (1108a35–b6). He continues to assume that our basic impulses all have medial forms and therefore continues to exclude the worst vices. I said the doctrine of the mean can figure in an account of virtue, and it can in particular express the ideal of proportionality. Thus, a desire can be ‘in a mean’ if it’s proportioned to its object’s value, wanting it neither more nor less than its degree of value compared to other objects makes appropriate. As so understood the doctrine can explain ‘vices of disproportion’ such as cowardice and selfishness. A coward is vicious because he cares much more about his comfort or safety than about some significantly greater good, such as the preservation of several people’s lives, that he could secure by risking it. By contrast, a rash person cares too little about his safety because he risks it for much smaller goods, and a selfish person wants his own pleasure much more than the greater goods of other people, which is again disproportionate. But this use of the doctrine of the mean isn’t available to Aristotle because it doesn’t fit the general structure of his ethical view. This leads to the most important objection to his account: that it gives the wrong explanation of what the virtues are, resulting in a wrong and even repellent picture of the virtuous person’s psychology.

EXPLANATORY EGOISM The general structure of Aristotle’s ethics is set out in NE I. In every act we aim at some good, and therefore, he argues, aim at a single chief good. This chief good is eudaimonia, and though he doesn’t say so explicitly, it seems clear that for each person the relevant good is just her own eudaimonia. (There are passages where Aristotle imagines an agent aiming at the eudaimonia of all, but the most common reading of his ethics gives it the egoistic structure I’ve described.8) Eudaimonia turns out to involve the active exercise of virtue, which consists in part in acts expressing moral virtues such as courage and liberality. Our ultimate reason to perform these acts is therefore that doing so is part of exercising virtue, which is what we must do to achieve the eudaimon or good life that’s our ultimate goal.

Aristotle on Virtue


This sketch of Aristotle’s ethics should be familiar, but it doesn’t allow the claim that states of other people such as their pleasure or knowledge are good in a way that by itself gives me sufficient reason to promote them. Any good playing that role must either be or contribute to a chief good that’s my own eudaimonia, and states of other people can’t do that: my life can’t be better or more eudaimon because of something true of you. Aristotle therefore can’t use claims of this kind to explain his doctrine of the mean. He can’t say courage is in a mean and virtuous because it cares proportionally about goods outside the self such as the preservation of another’s life, whereas cowardice and selfishness are vices because they care too little about others’ goods. He can’t value proportionality among goods that he can’t recognize in the first place. This isn’t to say he can’t include courage and liberality among his virtues and cowardice among his vices. He can assert that the former are good, in the sense of contributing our eudaimonia, and the latter bad. But these will be, and in the NE notoriously are, just assertions with no supporting rationale; he never makes a persuasive connection between his general claims about each person’s good in NE I and his list of specific virtues in II–IV. More specifically, he can’t say, as the higher-level account does, that the other-regarding virtues are virtues because they respond fittingly to independently good or bad states of others while the other-regarding vices respond unfittingly. He can’t say these things because he doesn’t think states of another have value from my point of view, or are relevant to my moral thought. And because he can’t say them, he can only assert what the higherlevel account explains. This points to the central flaw in Aristotle’s account of virtue: its underlying explanatory egoism. Imagine that you’re suffering pain and I can act to relieve your pain. Presumably I ought to do so, but what’s the ultimate explanation why? Aristotle’s explanation is that relieving your pain can make my life more desirable. If I do so from the right motive, my act will exercise virtue and so contribute to my eudaimonia; it will make my life better. But that’s surely not the right explanation, which is that relieving your pain will make your life better. My reason to aid you isn’t just superficially but fundamentally other-regarding, concerning your rather than my good. Aristotle’s conception of the good life isn’t hedonistic; he’s not saying I should relieve your pain as a means to something like pleasure for myself. His ideal is a eudaimonia of which virtuous action is an intrinsic constituent. Even so, my eudaimonia is necessarily a state of me and located in my life; it’s my eudaimonia rather than someone else’s. And that means his view grounds all my oughts or reasons in considerations about my good. That was the main criticism of his and other ancient ethical views by H. A. Prichard: that their egoism distorts duties concerning other people by making them really about oneself.9 And the criticism extends to those present-day virtue-ethical views that, like Anscombe’s, define the virtues as traits a person needs in order to flourish or live well.10 This definition relates the virtues not to external


Thomas Hurka

values such as others’ pleasure or pain but to my own good or flourishing; it therefore goes with the view that any reason I have to act virtuously likewise relates to my good. But that’s the wrong definition and the wrong reason. What makes something like benevolence a virtue isn’t its benefiting me but its caring properly about goods in other people’s lives.

EGOISTIC MOTIVATION The underlying egoism of Aristotle’s account seems to imply a similarly egoistic picture of the virtuous person’s motivation. If my ultimate goal is my own eudaimonia, shouldn’t I, while relieving your pain, have the desire for my eudaimonia as my ultimate motive? But isn’t helping you from concern for my good precisely not virtuous? Some present-day Aristotelians say that though my ultimate aim is my eudaimonia, this aim isn’t one I can achieve by trying to. Eudaimonia requires virtuous action, which is action motivated by concern for others, and I won’t have that if my primary desire is for a state of me.11 Though this is a possible move it makes the resulting ethical view ‘selfeffacing’ because it tells people not to believe or be guided by its own foundational claims.12 Rival views such as utilitarianism can also be self-effacing, but the Aristotelian one will be so in an especially troubling way. If utilitarianism tells people not to think in utilitarian terms, it’s because of the contingent psychological fact that their attempt to do so won’t succeed. But the proposed eudaimonist view tells them not to be guided by itself because that’s intrinsically objectionable or contrary to virtue, which is an odd thing for an ethical view to say. Whether or not this move is acceptable, Aristotle’s own view seems not to be self-effacing because his picture of the virtuous person is at many points precisely egoistic, involving a primary focus on his own virtuous action. This is clearest in his account of the proud person or megalopsychos in NE IV.3. The megalopsychos is said to have every virtue but also has an unattractively self-centered concern with his standing in virtue, especially compared to other people. He likes to give benefits but not receive them because ‘the one is the mark of a superior, the other of an inferior’ (1124b9–11). He’s also a person of few deeds, not doing ordinary acts of virtue but holding himself back for great and notable ones (1124b23–26). If you ask him to help you with a heavy package, he’ll say he doesn’t do trivial favors; he’ll only respond to something really serious like a threat to your life. Again, he’s less concerned with what an act will do for you than with what it means for his own status as exceptionally virtuous.13 It’s not that all concern for your virtue is objectionable. The higher-level account says that if your virtuous desire for another’s welfare is good, desiring or taking pleasure in it is also good. However, the account has two grounds for limiting these attitudes. One is its claim that a virtuous attitude has less

Aristotle on Virtue


value than its object, so your desire to relieve another’s pain is less good than the relief it aims at. The other is its ideal of proportionality, which says you should care less about lesser goods. Together they imply that you should care less about your virtuous desire for another’s relief from pain than you do about the relief, which is precisely what the megalopsychos doesn’t do. By caring more about his own virtue than about any benefits it can give others, he divides his concerns in a disproportionate and even vicious way. Defenders of the higher-level account have found the megalopsychos repellent. Rashdall noted ‘Aristotle’s revolting picture of the high-souled man’, while Ross said the description of the megalopsychos ‘betrays somewhat nakedly the self-absorption which is the bad side of Aristotle’s ethics’.14 That self-absorption appears often in the NE. In his discussion of courage Aristotle says ‘the end of every activity is conformity to the corresponding state of character’ (1115b20–21), as if a courageous person’s main aim is to express his own courageous disposition. He also says that the more virtuous a person is, the more he’ll be pained at the thought of his death because ‘life is best worth living for such a man, and he is knowingly losing the greatest goods’ (1117b9–12; also 1170a26–28). So does a courageous person think while on the battlefield about how virtuous he is? Or consider the account in IX.8 of the self-lover or philautos. Like the megalopsychos he cares that he more than anyone else should act justly and temperately (1168b25–26). He too prefers one great and noble act, such as dying for others, to many trivial ones (1169a25); mustn’t he then hope others’ lives will be threatened? He’ll sacrifice his wealth for a friend, but only because he thereby gains nobility and ‘assign[s] the greater good to himself’ (1169a28–30). He’ll also let his friend do virtuous deeds rather than do them himself, but his reason is that it may be ‘nobler to become the cause of his friend’s acting than to act himself’, so he again ‘assign[s] to himself the greater share in what is noble’ (1169a33–36). If his friend has the same competitive motive, they can engage in an Alphone-and-Gaston routine where each tries to get the other to do the virtuous deed so as to gain the ‘greater share’ of nobility for himself. Or the friend can say that while it’s nobler to let a friend do a virtuous deed, it’s even nobler to let a friend let you do it, leading to an infinite regress of nobler lettings. Even within friendship Aristotle imagines virtuous agents competing in virtue and more concerned with their comparative virtuousness than with any benefits they can give to others; E. F. Carritt rightly condemned ‘the egoistic self-righteousness of Aristotle’s philautos’.15

CHOOSING ACTS FOR THEIR OWN SAKES Nor is it only in his descriptions of particular characters that the egoism of Aristotle’s view comes out. Consider his well-known claim in NE II.4 that in order to act virtuously you must choose your acts ‘for their own sakes’


Thomas Hurka

(1105a33; also 1176b5–8). This is appropriate if you’re choosing to keep a promise, tell the truth, or do some other act required by a deontological duty, and even in non-deontological cases it’s better than choosing an act just as a means to your own wealth or pleasure. These last aren’t, however, the only or the best alternatives. Often a virtuous person will choose an act primarily as a means, but to a good state of some other person. If she virtuously relieves another’s pain, it will be mainly as a means to an outcome in which the other is free of pain. She may also choose the act for itself, for example, as one that’s virtuously motivated, but if her attitudes are properly proportioned this will be a secondary motive, with less importance in her psychology than the desire to do her act as a means. But Aristotle seems to make it the primary motive, as if virtuous agents always choose their acts above all for themselves, which makes virtue excessively self-concerned. It may be replied that this critique misreads Aristotle’s view. Any act that’s worth doing has properties that make it so, and to choose it for those properties is to choose it ‘for its own sake’. If an act is worth doing because it will free another from pain, someone who chooses it for that reason chooses it for its own sake.16 But this reply ascribes to Aristotle a view he never explicitly states, though he easily could. It also threatens to make his ‘for their own sakes’ condition vacuous. If choosing an act because it will result in another’s freedom from pain is consistent with choosing it for its own sake, why isn’t the same true of choosing an act because it will result in your having money or in a table’s being made? Shouldn’t all cases of choice on instrumental grounds be treated the same? The proposed reading therefore seems to imply that every act is chosen for its own sake, and that is not Aristotle’s view. He thinks it’s distinctive of ‘doing’, of which virtuous action is an instance, that it ‘itself is its end’ and is chosen for itself, whereas ‘making’ ‘has an end other than itself’ and is chosen as a means (1140b6–7). How on the proposed reading can there be any cases of making? In one passage Aristotle does, admittedly, take a different line. In NE X.7 he gives as one reason for the superiority of contemplation to moral virtue that, while the former has no end beyond itself, ‘from practical activities we gain more or less apart from the action’, so a statesman ‘aims at despotic power and honors, or at all events happiness, for him and his fellow citizens’ (1177b21–25). The reference to others’ happiness here suggests a more attractive view than in II.4, but now Aristotle denies that virtuous action is ‘loved’ or ‘desirable’ for its own sake, and in fact does so twice (1177b2, b18). This denial is puzzling since it contradicts the II.4 claim that virtuous agents do choose acts for their own sakes.17 But it confirms my reading of the earlier passage since it assumes that when you act as a means to an external goal you don’t choose your act for itself. It’s therefore hard to see how choosing an act ‘for its own sake’ is consistent with choosing it for how it will affect others. Even if it is consistent, however, there’s another objection to Aristotle’s view. If a truly virtuous per-

Aristotle on Virtue 19 son does what will free another from pain, her main concern is that the other be free from pain, and she desires her own act derivatively, as a means to that end. But then she’ll have various other attitudes concerning the other’s pain. If she can’t relieve that pain herself, she’ll hope it gets relieved in some other way. And if it is relieved, perhaps because someone else relieves it or because it goes away by itself, she’ll be pleased by that fact. She’ll care as or almost as much about goods of another that don’t result from her action as about ones that do. But nowhere in his main discussion of virtue in NE II–IV does Aristotle ever say that a virtuous person will have hopes or feel pleasures or pains about things that happen to other people independently of her own acts: it doesn’t occur to him to make this point. These attitudes are surely central to virtue; it’s surely a key part of being virtuous that you care about states of others just as states of them and apart from your role in producing them. But this kind of caring seems not to figure in Aristotle’s account. This isn’t because he thinks virtue involves only dispositions to act. He often says moral virtue is concerned with pleasures and pains and involves being pleased and pained by the right things (1104b4–28, 1105a4–16, 1106b18–22, 1121a3–4, 1152b1–6, 1172a21–23). His example of virtuous pleasure, however, is always pleasure in your own virtuous action: he says virtuous acts are pleasant to the lover of virtue (1099a10–20) and identifies virtuous people in part as ones who delight in acting temperately or courageously (1104b4–8; also 1110b12–13). This is especially evident in a passage that comes close to the higher-level account I have contrasted with his. In NE X.5 he says that if an activity is good, pleasure in it is also good, whereas if the activity is neutral, so is pleasure in it; likewise, if the activity is bad, pleasure in it is bad (1175b24–1176a3). He here recognizes that things can have value independently of our attitudes to them and that their values can make some attitudes to them good and others not. But the things he considers are only activities rather than states of a person such as her being free of pain, and they’re only your own activities rather than someone else’s; this is implied in his calling the pleasures ‘proper to’ the activities and so closely tied to them that it’s hard to tell the two apart (1175b30–33; also 1174b24–1175a2). While he recognizes that there are virtuous feelings, he again doesn’t include among them feelings about states of other people independent of your virtuous action. Aristotle does briefly discuss these feelings in NE VIII and IX, under the heading of ‘goodwill’ (1155b31–1156a10, 1159a5–12, 1166b30–1167a20), but he says that, except in relation to a close friend, goodwill is too weak an impulse to ever issue in action (1167a1–2, a7–9). And another discussion in these Books further highlights the egoism of his view. In IX.7 he says that just as craftsmen and poets care especially about what they themselves have created, so do virtuous benefactors. Since ‘that which they have treated well is their handiwork’ and even ‘is, in a sense, the producer in activity . . . to the benefactor that is noble which depends on his action, so that he delights in the object of his action’ (1167b34–1168a18). But this gets genuinely


Thomas Hurka

virtuous motivation precisely backward! A truly virtuous person cares first that another be free of pain and only secondarily about an act of hers that may produce that result. Aristotle’s benefactor cares first about her own virtuous action and only derivatively about its effect on others, as something she brought about. She may be pleased that another is free from pain, as making her own act of seeking that outcome successful and therefore a greater contributor to her eudaimonia.18 But she’s pleased by it only or mainly because it was produced by her. In fact Aristotle often prioritizes virtuous action over its effects. In NE VIII.1 he is arguing that friends are necessary for a good life and gives as one reason that rich people in particular need friends since money is useless ‘without the opportunity of beneficence, which is exercised chiefly and in its most laudable form towards friends’ (1155a6–9; also 14–16). This argument seems to value friends, like money, primarily as means to one’s own exercise of virtue. A notorious argument in the Politics criticizes the proposal for communal ownership in Plato’s Republic by saying it removes the opportunity for liberal uses of private property.19 Like similar arguments by present-day neo-conservatives against the welfare state, it assesses a scheme of property relations only by its effect on virtuous action by the rich and not at all by its implications for the condition of the poor.

CHOOSING ACTS AS KALON As well as saying virtuous agents choose acts for their own sakes, Aristotle says they act for the sake of the kalon, often translated the ‘noble’ or ‘fine’ but with aesthetic connotations of the beautiful.20 This raises some additional as well as some familiar issues. Because kalon is an evaluative concept, to choose an act as kalon is to be motivated by an explicitly evaluative thought, as you need not be if you choose an act for its own sake. If Aristotle thinks motivation by the kalon is necessary for virtuous action, his account excludes a kind of action allowed as virtuous by the higher-level account and on many views paradigmatically virtuous: where you choose an act for properties that make it right but without thinking of them as right-making, as when you relieve another’s pain just because you want it to end and without any thought of your act as required. If Aristotle denies that this kind of act is virtuous, his account is excessively intellectualist in the same way as Kant’s, which finds moral worth only in acts done from duty and not in ones that are simply compassionate.21 Another issue concerns the aesthetic connotations of kalon. Is choosing an act for its beauty not again choosing it for an inappropriately selfcentered reason, one focused on the aesthetic quality it can add to your life rather than on any benefits it will give others? Sidgwick read Aristotle this way, saying his virtuous agent makes ‘a deliberate choice of virtuous acts for the sake of their intrinsic moral beauty, and not for any end external to

Aristotle on Virtue


the act’, so ‘The limits of Aristotle’s Liberality are not determined by any consideration of its effect on the welfare of its recipients, but by an intuitive sense of the noble and graceful quality of expenditure that is free without being too lavish; and his Courageous warrior is not commended as devoting himself to his country, but as attaining for himself, even amid pains and death, the peculiar kalon of a courageous act’.22 The objection implied here is, however, too quick. Since being kalon is a supervenient property, any act that’s kalon has non-evaluative properties that make it so, and to choose it as kalon is to choose it believing it has those properties. What are they? Aristotle is characteristically disappointing on this topic, making only vague and even contradictory statements. Sometimes he suggests that an act is made kalon by properties it has just as an act and independently of its motive, as when he says a liberal person will ‘give for the sake of the noble, and rightly; for he will give to the right people, the right amounts, and at the right time’ (1120a24–6; also 1120b3–4, 1121b3–7, 1147a29–32, 1151b18–21, 1177b16–18). At other times whether an act is kalon seems to turn on its motive, as when he says the end of courageous action is conformity to a courageous state of character, which is noble (1115b20–22), or that it’s especially noble to act in the face of great danger, which you’re then not deterred by (1115a24–31, 1169a21–26), or to benefit another without a view to repayment (1162b36; also 1171b20–23, Rhetoric 1366b35–67a5). And of course an act could need both types of property to count as kalon, though Aristotle never explicitly says this. It’s surely central to an adequate account of virtue to specify clearly what non-evaluative properties a virtuous person chooses her acts for, but Aristotle’s discussions of the kalon do not do that. We can, however, consider the two main possibilities. One is that an act is made kalon by properties it has apart from its motive, which can include its being likely to benefit another person.23 Even if this is Aristotle’s view, however, it still faces the objection that the agent’s primary concern is his own giving of the benefit rather than the resulting state of the other, such as her being free of pain. (Recall that in X.7 the benefactor thinks the effect he produces is noble because it depends on his action.) And the view is hard to reconcile with the aesthetic aspect of the kalon since merely instrumental properties, though they can by themselves make an act worth choosing, don’t usually by themselves make it beautiful. (This may have been part of what motivated Sidgwick’s reading.) If I cut off your leg to save you from dying or upbraid you harshly because that’s the only way to improve your character, what’s remotely beautiful in what I do? There can be aesthetic quality in achieving an end in an especially elegant or efficient way, but not all instrumentally good acts do that. An act can also be beautiful if it’s ‘fitting’ to its situation, as an act of gratitude can be to a previous benefit; Ross suggested this reading for all ancient ethical uses of kalon.24 But as C. D. Broad argued, while the concept of the ‘fitting’ is appropriate for some


Thomas Hurka

moral considerations such as gratitude and promise-keeping, it isn’t appropriate to that of promoting good consequences, which involves the different concept of ‘utility’.25 To choose an act just because it will have good effects, as a virtuous person often does, isn’t to choose it for a property that can plausibly make it kalon. The other possibility is that acts are made kalon by their motive. This better fits the aesthetic side of the kalon since the motive of an action is intrinsic to it, and a good motive can be said, at least on the higher-level view, to ‘fit’ the value of its object. Moreover, several commentators have ascribed this kind of view to Aristotle.26 But as well as still not addressing the objection about valuing virtuous acts over their effects, the view makes virtuous motivation implausibly complex. A virtuous person, it holds, first has a base-level virtuous desire, for example, to relieve another’s pain. Then he sees that an act done from that motive will be kalon and forms a second, higher-level desire to do it because it will be kalon, or because it will have that initial virtuous motive. Must virtuous action always have this selfreflective, double motivation? Does it even often have it? And there’s again a question about self-centeredness. If the virtuous agent has two desires, one to relieve another’s pain and the other to do an act motivated by that desire and therefore kalon, which is his main or most strongly motivating desire? A parallel question can arise after he acts: what’s he most pleased by then, that he relieved another’s pain or that he acted from the virtuous desire to do so? Aristotle’s answer to both questions seems to be that the higher-level, self-reflective attitude is the stronger one. He says countless times that virtuous agents act for the sake of the kalon, which on the view now under consideration is to do an act because it will have another virtuous motive, and speaks much less often of agents’ doing acts because they’ll benefit others. Bernard Williams called an agent ‘morally self-indulgent’ if ‘what the agent cares about is not so much other people, as himself caring about other people’, or if he ‘focuses disproportionately upon the expression of his own disposition’.27 If Aristotle’s virtuous person chooses an act primarily as kalon, where that depends on its having another virtuous motive, he’s self-indulgent in Williams’s sense. It’s therefore not only Aristotle’s descriptions of characters such as the courageous person on the battlefield, the megalopsychos, the philautos, and the benefactor especially pleased by what he produced that give an unattractively self-centered picture of virtuous motivation. The same follows from some of his more general claims, such as that a virtuous person chooses his acts ‘for their own sakes’ or for having the quasi-aesthetic quality of being kalon. My main argument has been that this self-centeredness isn’t a lapse on Aristotle’s part but an expectable consequence of his overall ethical view. On that view, recall, all my acts are chosen as means to a chief good that’s my eudaimonia, so anything choiceworthy for me must contribute to my eudaimonia. But no state of another person, such as her being free from pain, can do that; my life can’t be better because of something true of someone else, and as

Aristotle on Virtue


a result no such state can be good in a way that by itself gives me a reason to act. What can give me a reason is only something true of me, such as that an act will be one of my relieving your pain, or one in which I act from a virtuous motive. It therefore can’t be surprising that those are the primary foci of Aristotle’s virtuous agent. He isn’t pleased or pained by states of others unconnected to his own agency because those states aren’t relevant to his good. And he doesn’t first want a good of another, such as her being free from pain, and only then want to do an act that will produce it; he first wants to do that act and will only value its result because it’s one he produced. The whole structure of Aristotle’s view pushes his virtuous agent to look mainly at his own acts and own motives in a way Ross said involves ‘self-absorption’. There’s a natural explanation for these facts. As C.C.W. Taylor has argued, Aristotle developed his account of virtue in a society still influenced by a Homeric conception of the good or admirable person as essentially competitive, wanting to be superior to others in aspects of life attended with honor, pleased with himself when he is superior, and therefore more selffocused than anyone we today could see as fully virtuous. Hence Aristotle’s jarring-to-us descriptions of ‘virtues’ like magnificence and megalopsychia, while foreign to our ethical outlook, fit that of his Greek society.28 I would extend Taylor’s point by saying the same influences led Aristotle to posit an ultimate goal for ethical life that’s similarly egoistic, involving for each person only features of his life and not giving ground-level importance to what happens to others. Like more specific features of his account of virtue, the underlying structure of Aristotle’s view reflects an agonistic Greek ethos that’s some distance from our moral thought today.

CONCLUSION I’ve argued that Aristotle wrongly thought virtue is praised and vice blamed, wrongly made the primary locus of virtue dispositions rather than occurrent mental states, and wrongly excluded, with his doctrine of the mean, the worst moral vices. But my main criticism has been that his account of virtue is objectionably egoistic, especially as compared to the higher-level account. This last contrast is worth making more abstractly. We can distinguish two general approaches to the concept of virtue, which can be called the outside-in and the inside-out. The outside-in approach takes there to be values or, more generally, normative factors outside a person’s motives and attitudes and holds that the virtues involve appropriate responses to those factors. What makes an attitude virtuous is its relation to something outside itself and often outside the agent, as when its object is another person’s pleasure or freedom from pain. This externally-based explanation of what makes the virtues virtues goes with a picture of virtuous motivation as likewise externally focused, so a virtuous person cares most about his virtues’ objects, such as another’s pleasure or pain, and only


Thomas Hurka

secondarily about his own virtuous motives in pursuing them. The insideout approach, by contrast, doesn’t relate the virtues to external values since it doesn’t recognize any. It just says the virtues are good states of the person, or intrinsic constituents of an overall good or eudaimon life for him. It therefore can’t explain why a given virtue such as benevolence is one; it can only assert that it is. And it goes with an internally-focused picture of virtuous motivation, where the virtuous person cares primarily about his own virtue and its expression and only secondarily about the states of others his acts can, if successful, bring about. The higher-level account illustrates the outside-in approach and Aristotle’s the inside-out, and I’ve tried to show that on several crucial points the former is more attractive. It gives better explanations of why the virtues are virtues and of why we should treat others in the way the other-regarding virtues would lead us to: the ultimate reason isn’t that this will make our lives better, but that it will make the others’ lives better. It also gives a better picture of the virtuous person’s motivation as externally rather than internally focused. For a long time the work of Rashdall, Moore, Ross, and other moral philosophers of their era was ignored and even denigrated. As a result their higher-level account of virtue was also ignored, and accounts modeled on Aristotle’s attracted the bulk of philosophers’ attention. But the higher-level account is by far the more illuminating of the two; in comparison, Aristotle’s is a dead end.

NOTES 1. Anscombe, ‘Modern Moral Philosophy’. 2. Rashdall, ‘Professor Sidgwick’s Utilitarianism’, and The Theory of Good and Evil; G. E. Moore, Principia Ethica; W. D. Ross, The Right and the Good; Franz Brentano, The Origin of Our Knowledge of Right and Wrong. I give a present-day elaboration and defense of the account in Virtue, Vice, and Value. 3. Aristotle, Nicomachean Ethics (NE). All references to the NE are to this translation. 4. Aristotle recognizes this at 1137a5–9, where he says acting ‘as a result of a certain state of character is neither easy nor in our power’. 5. I develop this argument more fully in ‘Virtuous Act, Virtuous Disposition’. 6. Ross held that what’s virtuous or morally good is not only occurrent ‘acts of will, desires, and emotions’ but also ‘relatively permanent modifications of character even when these are not being exercised’ (Foundations of Ethics, 292). However, he saw the value of the latter as only an addition to the value of occurrent virtuous attitudes, not something that increases their value when they’re present. 7. That the doctrine of the mean excludes vices like cruelty and malice is also noted by C. C. W. Taylor in Aristotle: Nicomachean Ethics Books II–IV, xix, 113. 8. For readings of Aristotle in which each person’s ultimate ethical goal includes the eudaimonia of others see Kraut, Aristotle on the Human Good, and McKerlie, ‘Friendship, Self-Love, and Concern for Others in Aristotle’s Ethics’. 9. Prichard, ‘Duty and Interest’, 21–49; for a similar criticism of Plato see Brown, ‘Glaucon’s Challenge, Rational Egoism and Ordinary Morality’, 42–60. 10. Anscombe, ‘Modern Moral Philosophy’, 18.

Aristotle on Virtue


11. Annas, The Morality of Happiness, 118, 127–128, 224; Whiting, ‘Eudaimonia, External Results, and Choosing Virtuous Actions for Themselves’, 286. Though I lack the space to elaborate, I don’t think Annas’s attempt to answer the egoism objection in Intelligent Virtue, 52–63, addresses the main points. 12. Parfit, Reasons and Persons, 24. 13. For criticism of this last feature of the megalopsychos see Sherman, ‘Common Sense and Uncommon Virtue’, 105–106. Sherman thinks the megalopsychos is unrepresentative of Aristotelian virtue; I think he’s all too representative. 14. Rashdall, The Theory of Good and Evil, 208. 15. Carritt, ‘An Ambiguity of the Word “Good” ’, 69. Taylor also notes the ‘selfreferentiality’ of Aristotle’s megalopsychos, philautos, and other virtuous agents in Aristotle: Nicomachean Ethics Books II–IV, 88–92. 16. Whiting, ‘Eudaimonia, External Results, and Choosing Virtuous Actions for Themselves’, 280. 17. Henry Sidgwick took Aristotle to be simply inconsistent on this point; see Outlines of the History of Ethics, 67–68. 18. Whiting, ‘Eudaimonia, External Results, and Choosing Virtuous Actions for Themselves’, 286–289. 19. Aristotle, Politics, 1163b11–14. 20. Richard Kraut argues that kalon has aesthetic connotations in ‘An Aesthetic Reading of Aristotle’s Ethics’ (forthcoming). 21. Taylor also makes this criticism of Aristotle on the kalon; see Aristotle: Nicomachean Ethics Books II–IV, 90–91. 22. Sidgwick, Outlines of the History of Ethics, 59; and ‘Hedonism and Ultimate Good’, 90. Another objection is that an aesthetic concept like kalon doesn’t line up perfectly with moral rightness since there can be beauty in wicked acts; Sidgwick made this point in the second edition of The Methods of Ethics, 100, as did Carritt in ‘Moral Positivism and Moral Aestheticism’, 141. 23. Kraut claims that for Aristotle it’s necessary for an act to be kalon; that it benefit either the agent or someone else (‘An Aesthetic Reading of Aristotle’s Ethics’, 15 in the typescript). But this is again something Aristotle doesn’t explicitly say; on the contrary, he contrasts the kalon with the beneficial as, alongside the pleasant, one of the three main objects of choice (1104b30–31). 24. Ross, Foundations of Ethics, 54. 25. Broad, Five Types of Ethical Theory, 218–220. 26. Korsgaard, ‘From Duty and for the Sake of the Noble: Kant and Aristotle on Morally Good Action’, 216–219; Price, Virtue and Reason in Plato and Aristotle, 74–76. 27. Williams, ‘Utilitarianism and Moral Self-Indulgence’, 45, 47. 28. Taylor, Nicomachean Ethics II–IV, xx–xxi; also 88–92.

REFERENCES Annas, Julia, Intelligent Virtue (Oxford: Oxford University Press, 2011). ———, The Morality of Happiness (New York: Oxford University Press, 1993). Anscombe, G.E.M., ‘Modern Moral Philosophy’, Philosophy 33 (1958), 1–19. Aristotle, Nicomachean Ethics, trans. Sir David Ross (Oxford: Oxford University Press, 1980). ———, Politics, in The Complete Works of Aristotle, trans. B. Jowett, ed. J. Barnes. 2 vols. (Princeton: Princeton University Press, 1984). Brentano, Franz, The Origin of Our Knowledge of Right and Wrong, trans. R. M. Chisholm and E. Schneewind (London: Routledge & Kegan Paul, 1969).


Thomas Hurka

Broad, C. D., Five Types of Ethical Theory (London: Routledge & Kegan Paul, 1930). Brown, Lesley, ‘Glaucon’s Challenge, Rational Egoism and Ordinary Morality’, in Pursuing the Good: Ethics and Metaphysics in Plato’s Republic, ed. D. Cairns, F. G. Hermann, and T. Penner (Edinburgh: Edinburgh University Press, 2007), 42–60. Carritt, E. F., ‘An Ambiguity of the Word “Good” ’, Proceedings of the British Academy 23 (1937), 51–80. ———, ‘Moral Positivism and Moral Aestheticism’, Philosophy 13 (1938), 131–147. Hurka, Thomas, Virtue, Vice, and Value (New York: Oxford University Press, 2001). ———, ‘Virtuous Act, Virtuous Disposition’, Analysis 66 (2006), 69–76. Korsgaard, Christine M., ‘From Duty and for the Sake of the Noble: Kant and Aristotle on Morally Good Action’, in Aristotle, Kant, and the Stoics: Rethinking Happiness and Duty, ed. S. Engstrom and J. Whiting (Cambridge: Cambridge University Press, 1996), 203–236. Kraut, Richard, ‘An Aesthetic Reading of Aristotle’s Ethics’, in Politeia in Greek and Roman Philosophy, ed. M. Lane and V. Harte (Cambridge: Cambridge University Press, forthcoming). ———, Aristotle on the Human Good (Princeton: Princeton University Press, 1989). McKerlie, Dennis, ‘Friendship, Self-Love, and Concern for Others in Aristotle’s Ethics’, Ancient Philosophy 11 (1991), 85–101. Moore, G. E., Principia Ethica (Cambridge: Cambridge University Press, 1903). Parfit, Derek, Reasons and Persons (Oxford: Clarendon Press, 1984). Price, Anthony, Virtue and Reason in Plato and Aristotle (Oxford: Clarendon Press, 2011). Prichard, H. A., ‘Duty and Interest’, in Moral Writings, ed. J. MacAdam (Oxford: Clarendon Press, 2002), 21–49. Rashdall, Hastings, ‘Professor Sidgwick’s Utilitarianism’, Mind o.s. 10 (1885), 200– 226. ———, The Theory of Good and Evil. 2 vols. (London: Oxford University Press, 1907). Ross, W. D., Foundations of Ethics (Oxford: Clarendon Press, 1939). ———, The Right and the Good (Oxford: Clarendon Press, 1930). Sherman, Nancy, ‘Common Sense and Uncommon Virtue’, in Midwest Studies in Philosophy, vol. 13, Ethical Theory: Character and Virtue, ed. P. A. French, T. E. Uehling, Jr., and H. K. Wettstein (Notre Dame, IN: University of Notre Dame Press, 1988), 97–114. Sidgwick, Henry, ‘Hedonism and Ultimate Good’, in Essays on Ethics and Method, ed. Marcus G. Singer (Oxford: Clarendon Press, 2000), 89–98. ———, The Methods of Ethics, 2nd ed. (London: Macmillan, 1877). ———, Outlines of the History of Ethics, 6th ed. (London: Macmillan, 1931). Taylor, C.C.W., Aristotle: Nicomachean Ethics Books II–IV (Oxford: Clarendon Press, 2006). Whiting, Jennifer, ‘Eudaimonia, External Results, and Choosing Virtuous Actions for Themselves’, Philosophy and Phenomenological Research 65 (2002), 270–290. Williams, Bernard, ‘Utilitarianism and Moral Self-Indulgence’, in Moral Luck (Cambridge: Cambridge University Press, 1981), 40–53.


Aristotle on Virtue A Response to Hurka Anthony Price

Thomas Hurka’s chapter is a splendid antidote to deferential treatments of Aristotle. If there are any Aristotelians who will read it without finding their understanding challenged and clarified, the present writer is not one of them. Hurka’s indictment comes ostensibly under seven heads. Yet, in fact, all but one of these (and that I shall neglect) relate to the same issue: the motivations that are proper to virtuous action. It is not simple to identify Aristotle’s view of these; it is hardly simpler to decide—as I shall not—what view to take of them ourselves.

I We may start by distinguishing external and internal goals characteristic of virtuous action. Let me take as an example an anecdote about Socrates’s courage told by Alcibiades in Plato’s Symposium: People were already scattered, and he was withdrawing, Laches with him . . . He was making his way along there just as he does here at home . . . observing people on our own side and on the enemy’s in the same calm way, and making it plain to anyone, even if they were some distance away, that if anyone laid a hand on him, they’d meet with some pretty stiff resistance. That’s why he actually got away safely, along with his companion. (221a2–b7, tr. Rowe) Socrates well instances Aristotle’s brave man who, in the presence of great danger, is only ‘gently’ or ‘mildly’ afraid (ērema, Eudemian Ethics [EE] III.1 1228b29, 38); though not without apprehension (which is why he keeps looking around), he is free of disturbance (he is atarachos, Nicomachean Ethics [NE] III.8 1117a19, III.9 1117a31). Like the megalopsychos or man of proper pride, he maintains an even step and a level voice (IV.3 1125a12–16). And all this not idly, but in order to save his own life and (as Alcibiades’s narrative emphasizes) that of his companion, Laches. A different agent might risk danger gratuitously in order to display his sang-froid; Socrates is a sensible


Anthony Price

man and not a swashbuckler. The danger is inescapable, and he faces it with composure in order to save two lives. That is his external goal. (In a happier situation, this would have been victory.) He may also have an internal goal: to show himself brave, and thereby—as Aristotle would have it—to achieve the fine or noble (to kalon). Practical deliberation starts, as I read Aristotle, from a concrete goal that an agent adopts in context. In the EE, the general statement ‘Those who have no target before them are not in a position to deliberate’ (II.10 1226b29–30) comes immediately after a specific example: ‘The carrying of goods is a cause of walking if it is for the sake of that that a man walks’ (b28–29). Socrates’s goal was to save his own life and that of Laches. Had his goal been different, say to kill a Bœotian, he would have behaved differently. However, saying this cannot suffice to show his action to be brave, for deliberation that is simply oriented toward achieving a certain concrete goal succeeds through a mere ‘cleverness’ (deinotēs) that ‘is such as to be able to do the things that tend towards the mark we have set before ourselves, and to hit it’ (NE VI.12 1144a24–26, tr. Ross). For courage as a virtue, closely linked to practical wisdom, the agent needs two further capacities: his initial selection of an end must manifest a good character sensitive to context, and in thinking through how to achieve that end he must exercise a good judgment about what is worth pursuing or enduring for the sake of what. (This may lead him to discard his initial end in favor of another.) Aristotle pairs these two when he writes, ‘Virtue makes the end right, practical wisdom the things towards this’ (1144a7–9). This somewhat cryptic remark is best understood by reference to EE II.11, where we read, ‘Does virtue make the goal or the things towards the goal? We say that it is the goal, for of this there is neither reasoning nor a logos’ (1227b23–25). The agent’s character reveals itself in what occurs to him as an external goal worth achieving in the situation in which he finds himself. He then calculates how he may achieve this, or—if there are alternative ways or means—best achieve this (NE III.3 1112b15–17). Obstacles of various kinds may then obtrude. It may turn out simply impossible to achieve the goal in the context: ‘If we come to an impossibility, we give up the search, e.g. if we need money and this cannot be got’ (1112b24–26). A different impossibility is if the goal can be achieved but only in a way that is ethically out of the question. Less dramatically, the cost of achieving this goal may be such that it is better to pursue another goal less desirable in itself but cheaper to achieve in context. Estimating this requires reasoning that is not purely instrumental, and yet also falls within practical wisdom. We read in the De Anima [DA], ‘Whether one shall do this thing or do that thing it is the work of reasoning to decide. And such reason necessarily implies the power of measurement by a single standard; for what one pursues is the greater good’ (III.11 434a7–9). What Aristotle means here by ‘measurement by a single standard’ (heni metrein) is not clear. It can hardly mean a universal standard that applies in every context; it might mean a

Aristotle on Virtue 29 single standard relevant to the present context, or (more realistically still) a single standard for each comparison that has to be made on the way toward a final arbitration. Such complications lie, I believe, behind a sentence that has been much debated: ‘If it is characteristic of the practically wise to have deliberated well, excellence in deliberation will be correctness with regard to what conduces the end of which practical wisdom is the true supposition’ (VI.9 1142b31–33). On the face of it, this contradicts statements that the end is provided not by practical wisdom, but by virtue. It may then be suggesting that practical wisdom has two distinct spheres of operation: it identifies first the best end (either in general, or in context), and then the best means (certainly in context). However, we do better justice at once to the other evidence I have cited, and to the realities of the case, if we notice that the sentence places wisdom’s role in the selection of an end within deliberation, and not in advance of it. I take Aristotle’s thought to be that truth in the selection of an end is only reliably achieved when an exploration of possible ways and means has identified which goal is best practicable in context. So understood, Aristotle is faithful to the familiar phenomenology of practical thinking. An agent doesn’t start his deliberations by (as it were) closing his eyes and determining an a priori starting point for deliberation. Nor does he achieve the feat, to be ruled as impossible by Hume, of inferring an ‘ought’ of decision from an ‘is’ of description. Rather, a reflective inspection of his situation prompts a selection of a provisional goal; this is followed by a thinking through of possible ways and means that may discard or revise the goal, or confirm that, in context, it is achievable in a way that is acceptable. Nowhere, here or elsewhere, is there is any indication of a general and effective decision procedure. (Anyone who reads DA III.11 in isolation as a gesture in the direction of one must find the gesture hollow.) Rather, there is a frequent emphasis upon the variability of circumstances and the absence of any universal rule (NE I.3 1094b11–27, V.10, IX.2 1164b27–1165a14). Aristotle envisages no superordinate end by reference to which subsidiary goals might be assessed. He displays no interest in innovation that could achieve neither theoretical truth (since it would not be a contribution to theory), nor practical truth (which would require its catching on and being realized in practice). To characterize what the agent is trying to achieve in general terms, we have to turn to internal goals. In a situation of danger, whatever other virtues may be operative, the well-intentioned agent aims to act courageously, and thence well, and nobly or finely (kalōs). This demands attention to whatever considerations may be relevant to action in context: ‘The man who faces and who fears the right things and from the right motive, in the right way and at the right time, and who feels confidence under the corresponding conditions, is brave; for the brave man feels and acts according to the merits of the case and in whatever way the logos directs’ (III.6 1115a17–20). What he achieves by doing justice to the situational variables


Anthony Price

can be identified from case to case by reference to external goals, but in general terms only by reference to internal goals. Thus Aristotle links the exercise of overall judgment to the agent’s achievement of an ethical end: ‘The end of every activity is conformity to the corresponding state of character. This is true, therefore, of the brave man as well as of others. But courage is noble. Therefore the end also is noble; for each thing is defined by its end. Therefore it is for a noble end that the brave man endures and acts as courage directs’ (III.7, 1115b20–24; cf. in general VI.5 1140a25–28). In Aristotle’s view, therefore, virtues are not to be construed instrumentally as dispositions that measurably maximize the good effects of action. Rather, they are dispositions to act well in ways that can be roughly classified within spheres of action or feeling (as courage involves responses to danger involving two passions, fear and confidence), but which all involve a sensitivity to whatever considerations merit attention (whence the unity of the virtues). The core of what the agent is choosing and doing (which is a ‘this’ for the sake of ‘that’, viz. a way or means for the sake of an end, EE II.10 1226a11–13) is specified by a piece of reasoning that Aristotle takes to be syllogistic in form (cf. NE VI.12 1144a31–2, De Motu Animalium [DMA] 7). Yet this is the tip of an iceberg of perceiving and thinking that involves the whole situation and has as its target nothing less than acting as, in context, is good and best. The agent thus displays a complex practical orientation that is all of a piece: it is only by pursuing the best external goal in context, with attention to other relevant costs and benefits, that he can aspire to realize his standing aspiration to live and act well.

II Action in Aristotle thus possesses a double teleology: for instance, A risks his own life in order to save B’s life—and all this (means and end taken together) for the sake of acting bravely, finely, and well. Here risking his own life is instrumental toward saving B’s; without any such worthwhile goal it would not be brave, but rash. And it is only if he risks his own life for another’s that we have a paradigm of acting finely. Aristotle finds it characteristic of the decent agent that ‘he does many things for his friends and country, and if necessary dies for them’ (NE IX.8 1169a18–20).1 We have to bring together two general claims. eudaimonia is always the final end of action (or more precisely, as becomes clear later, deliberate action): ‘It is a first principle; for it is for the sake of this that we all do all that we do’ (I.12 1102a2–3). Equally, acting virtuously involves choosing to do what is virtuous for its own sake (II.4 1105a28–32). Any apparent tension between these two statements can be allayed by noting that it is by and in doing what he does (say, risking his life in the right context for a good goal) that the agent succeeds in being eudaimōn in the sense of acting well. Within the agent’s motivation we can distinguish these two elements:

Aristotle on Virtue 31 instrumentally he risks his life as a means to achieving an external goal; constitutively he does all that as a way—and perhaps the only way open to him in context—of acting well. Aristotle at once confirms this, and apparently confuses it, when he is arguing in X.7 in support of his own opinion that even better than ethical action is the contemplation of scientific truth. One may half regret this chapter as a piece of higher salesmanship. Taken on its own (and in separation from X.8), it might be interpreted as taking back the NE’s valuation of ethical action as a great good in itself and replacing that by an eccentric claim that only intellectual contemplation is itself a way of being eudaimōn. In the course of pursuing his argument, Aristotle says the following: If among actions in accordance with the virtues political and military actions are distinguished by nobility and greatness, and these are unleisurely and aim at an end and are not desirable for their own sake, but the activity of reason, which is contemplative, seems both to be superior in serious worth and to aim at no end beyond itself . . . it follows that this will be perfect human eudaimonia. (1177b16–25) On the face of it, this is not only inconsistent, but incoherent: if, for example, military action can itself achieve nobility, how can it fail to be desirable for its own sake? (The kalon is not an instrumental value.) However, coherence can be restored if we recall the exact wording of II.4: acts that are in accordance with (kata) the virtues may fail to be done virtuously. For example, if I risk my life in battle in a way that might well benefit my comrades, but without the right motivations (I may not intend their benefit, or, even if I do, I find no intrinsic value in acting so, but view it as a fatal chore), I fail to act virtuously in the full sense and hence to act well. What accords with the virtues, but fails in itself to achieve an intrinsic value, is just what I do concretely (which I have been describing vaguely as ‘risking my life’). That has no value that is independent of its instrumental role. This differentiates it from contemplating whose value is only intrinsic and which can be appreciated as being incommensurable in value just so long as it is permissible in context. The evaluation of virtuous action is therefore more complex than that of contemplation: risking one’s life, taken as such, is only instrumentally valuable, that is, in serving a good external end; if it does pursue that extrinsic value in context, we have a more complex act—risking one’s life for a good end—that may well (in the absence of any overriding counter-consideration) possess the intrinsic value of acting well. We could put Aristotle’s point as follows: contemplating truth, taken as such, is closer to eudaimonia than is risking one’s life, taken as such; we have more to add in the second case than in the first if we are to identify a case of acting well and achieving eudaimonia. How far this justifies placing contemplation above virtuous action is debatable; yet, so read, it is intelligible—and confirms my understanding of Aristotle.


Anthony Price

The ethical structure that I have identified has two further features that we should note. First, it is not implied that A values B’s life less than he values his own acting well (as if, in a different situation, he would rather that B died than that he should, on a single occasion, fail to act well). It is saving B’s life that he pursues for the sake of acting well. Part, though not all, of the content of this is that, if the situation were such that saving B’s life would not be acting well, he would not pursue it (or only in error). Secondly, a question whether A acts well in order to save B’s life does not arise. Adding that A not only risks his life but acts well for B’s sake would reiterate the same motivation with nothing gained except confusion. If A were so motivated even in acting well, it should follow that he not only acts well, but also acts better than well since he is acting well out of a noble, because altruistic, motive. But then even acting better than well should be enhanced if it also is generously motivated—which would yield an infinite sequence of bootstrapping achievements. A safer rule is that motivation cannot thus be reduplicated one level up: the altruistic motive attaches to the act of risking his life whereby the brave man acts well, but not to acting well itself. Acting well is not the sort of thing to be enhanced in this way. It itself cannot be other-interested—which should absolve it of the charge of being self-interested. This already gives us an understanding of Aristotle’s claim in I.7 that, while many things are worth pursuing for their own sake, only eudaimonia is fully final or perfect (teleios) in being pursued for its own sake, and never for the sake of anything else (1097a33–34). Yet it becomes evident that, so read, this evaluation of eudaimonia is restricted: that its value is final in this sense does not do much to show that its value is great. Compare a connected concept: acting commendably. It is doubtless better to act commendably than to act otherwise, and acting commendably is a thing achieved by having the right motivations and not a thing that becomes valuable, or even more valuable, by being well motivated itself. And yet, as a value, it seems minor even if it is not epiphenomenal: meriting commendation is not a mere sideeffect (as perspiration may be of physical motion in battle) and may well be a goal of the agent; yet it is hardly a salient goal and one without which his attitude to his action would be significantly different. What makes it important that, by risking his life for a good cause, the brave man is thereby acting well? This must be some feature of his action that connects with another formal feature of eudaimonia, which is its selfsufficiency: ‘The self-sufficient we now define as that which when isolated makes life desirable and lacking in nothing; and such we think eudaimonia to be’ (I.7 1097b14–16).2 This should not be read as implying that, when an agent consciously acts well, he must suppose that, within the context of his present action, nothing could have gone better. That can’t be what Aristotle means when he later remarks that the good man ‘has nothing to regret’ (he is ametamelētos, IX.4 1166a29). There he may merely mean that the good man has never to regret how he acted, wishing that, even with the information he then possessed, he had decided and acted otherwise. I.7 must be

Aristotle on Virtue


claiming more than this, but not that nothing could have enhanced his life at the time. (Very trivially, his immediate experience might have been better still either without a pang of toothache, or with a praline in his mouth.) Exactly what more Aristotle has in mind must connect with his idea that the agent who acts well may also be acting finely or nobly. This is evidently in part a hedonic concept. There is a pleasure to be found even in dying as a hoplite: ‘The end which courage sets before itself would seem to be pleasant, but to be concealed by the attending circumstances, as happens also in athletic contests’ (III.8 1117a35–b3). One may compare a hope in I.10 that, even in great misfortunes, nobility may ‘shine through’ (1110b30–31). A more transparent pleasure accompanies the life and death of a hero: ‘He would prefer a short period of intense pleasure to a long one of mild enjoyment, a twelvemonth of noble life to many years of humdrum existence’ (IX.8 1169a22–24). However, more needs to be said to explain what this unique value is that can provide so rare and rich an enjoyment.3 Here, I feel, Aristotle rather lets us down: it remains unclear precisely what constitutes to kalon and how it supervenes intelligibly upon the conditions for acting virtuously laid down in II.4.4 There are thus at least two aspects to Aristotle’s conception of the agent who ultimately acts for the sake of acting well. In part, this requires that he is not putting on blinkers in pursuit of some limited goal but trying to do justice to however many, and varied, considerations arise within the context of his action. Additionally, it finds a special and intrinsic value in action that achieves this goal. It is within action (and certain other activities), and not in the results of action, that the best of life is to be found. In Alcibiades’s anecdote, it is Socrates, and not Laches, who had the better day. We may or may not find this credible. Yet it is just what Aristotle committed himself to when he argued that the human good is nothing other than ‘activity of soul in accordance with virtue’ (I.7 1098a16–17).

III In one respect, the egocentricity of Aristotle’s ethical eudaimonism enables him to capture an essential feature of morality. Any ethical agent must be centrally concerned, as he or she acts, to be acting well and not badly. This is what it is to take responsibility for one’s own actions—a responsibility that one cannot have for the free actions of others. Such a concern was manifested by Socrates on a famous occasion that he narrated to the Athenian jurors as follows: When the oligarchy was established, the Thirty summoned me to the Hall, along with four others, and ordered us to bring Leon from Salamis, so that he might be executed . . . Then I showed again, not in words but in action, that . . . death is something I couldn’t care less about, and that my whole concern is not to do anything unjust or impious. That


Anthony Price government, powerful as it was, did not frighten me into any wrongdoing. When we left the Hall, the other four went to Salamis and brought in Leon, but I went home. I might have been put to death for this, had not the government fallen shortly afterwards. (Apology 32c3–d8, tr. Grube)

He did not reason that, since in all likelihood an innocent man would die anyway (and why not then one rather than two?), he was free to save his own skin. Instead, he decided to act justly, where ‘acting justly’ denotes a way of acting open to him that he ranked incomparably above other ways. The thought that evil may come into the world, but must not through me, is not culpably egocentric. It places my agency, where it must be for each agent, at the center of my life. What it precisely does not do, which would indeed be egocentric, is to displace that concern by another that is focused upon personal pleasure or advantage. To love one’s neighbor as oneself is not to be indifferent as to whether I do wrong, or he does. Such an attitude would be not impartial, but irresponsible. Yet to take responsibility for one’s action is not to suppose, insanely, that it matters more from some non-agential point of view (say God’s, or—less intelligibly—Sidgwick’s point of view of the universe) if I act badly, than if someone else does. That is not the reason why one kicks oneself if one has acted badly oneself, and not if another has acted badly—which rests rather upon the logic of kicking oneself. Personal responsibility is inescapable, and inescapably first-personal. In refusing to collaborate in arresting Leon, even at the likely cost of losing his own life, Socrates was acting in accordance with what is known as ‘Democritus’ Maxim’ (B 45), roughly ‘It is worse to do wrong than to be wronged.’ This is most familiar to us from early Plato (e.g., Crito 49b, Gorgias 469c) but is also explicit in Aristotle: ‘Acting unjustly is the worse [of acting unjustly and being unjustly treated], for it involves vice and is blameworthy’ (NE V.11 1138a31–32). The thought is not that, for each agent, it is impersonally worse that he should wrong anyone else than that anyone else should wrong him; rather that, morally speaking, it is better for him, or from his point of view, to be a victim than a villain.5 Such an attitude respects the integrity of individual agency and is central and essential to morality as we know it. However, it may root this thought deeper in human nature if we view it as the ethical transformation of a basic truth about human responses. Richard Wollheim introduces the idea that self-concern is basic, and not derivative from other concerns, by imagining the following alternatives: I am told something like the following: Someone whom I know will to-night meet a friend whom he loves and misses. Someone whom I know will tomorrow morning wake up blind. Then I learn that this someone, the someone whom I know and to whom this will happen, is me. There is a characteristic . . . way in which I shall respond to such a lesson. This response I call ‘the tremor’.6

Aristotle on Virtue


The thought that will characteristically generate the tremor is just this: it is coming to me. It is a response that is spontaneous, involuntary, and unmediated. It does not arise through the following steps: I foresee something good or bad for a future self; I side with that self; I feel the tremor. It is enough that I learn that it will be me and am able to let that thought register and reverberate. The last thing it rests upon is some evaluation, from whatever reflective point of view, of my own welfare as a peculiarly proper object of my concern. I suggest that we can place Democritus’s Maxim as a moralized derivative of the same basic human attitude: agents come to care, spontaneously and imperatively, about how they act. In order to act well, they have at the very least to avoid acting wrongly. Aristotle accepts some general restrictions: one is never to commit adultery, theft, or murder (II.6 1107a8–17); and no doubt very many other ways of acting, though not universally wrong, are excluded ‘for the most part’. One may compare the side-constraints that Robert Nozick places upon how one may achieve one’s ends.7 Again, one needs to be careful in defining how this falls within Aristotle’s eudaimonism. It would be a distortion to include among the ends that one is not to achieve impermissibly acting well itself—as if, among the ways in which one might act well, there are wrong ways. Rather, no agent can count as acting well if he infringes such constraints; he thereby rather achieves the opposite of eudaimonia, which is kakodaimonia. Let us return to the case of courage. An agent may risk his life for the sake of his friends, and all this for the sake of acting well. Here, as I argued, he needs to be pursuing a good concrete end if he is to count as acting well. Now we have a further condition: he must not pursue that end impermissibly (say by targeting not the enemy, but some innocents who happen to be in the way) if he is to be acting well. The upshot is indeed a complexity of motivation. Is this an implausible complexity? It is hard to see that it is. Let me rehearse yet again the levels of justification: I pursue some concrete goal, aiming at the same time to respect other goals of mine, and all this in order that I can count as acting well all things considered; in so acting, since I have a special concern about my own agency, I take a special pleasure. We can well ask how much of this need be conscious, with an awareness of the danger of becoming self-conscious.8 Here we may lack the evidence to pin Aristotle down. He certainly expects deliberation to be conscious, but he also admits that at least obvious elements need not actually be rehearsed (DMA 7 701a25–28); so when he makes eudaimonia the end of ends (e.g., I.12 1102a2–3), he need not be supposing that ‘How am I to act well?’ is always a question that the agent consciously puts to himself. I am well aware that saying these things does not amount to a full rebuttal even of those of Hurka’s objections to which they are relevant. What I hope to have done is to provide a framework within which some distinctive features of Aristotle’s conception of virtuous action can justly be assessed.


Anthony Price

NOTES 1. If man endues death in order to escape greater pains, this is not courage (EE III.1 1229b12–14). Of course, different goals are appropriate to different virtues. The health that is a goal of temperance (NE III.11 1119a16) is doubtless the agent’s own; by contrast, the liberal are loved for being useful (IV.1 1120a22), evidently to others. 2. I do not here discuss the debated lines that follow (1097b16–20); on them, see my Virtue and Reason in Plato and Aristotle, 52–53. 3. It may appear a further element of egocentricity that this pleasure is taken in one’s own activity and not, say, in the benefit to a recipient. However, this is inevitable since the pleasure is a form of enjoyment, and I cannot enjoy another’s sense of relief in being saved from danger (though I might enjoy perceiving it). However, it would be wrong to infer that I lack the attitudes toward his well-being that one would expect of someone willing to risk his life on another’s behalf. Aristotle recognizes a sympathetic pleasure or pain (sunchairein, sunalgein, IX.4 1166a7–8, IX.10 1171a6–8, 29–30). He disparages men who resemble women in liking to receive sympathy (X.11 1171b10–11) but respects the attitude of a mother separated from her children for whom it is a sufficient consolation to see them flourish, even if she is unknown to them (VIII.8 1159a28–33). 4. I say a little in Price, op. cit., 68. 5. These things are illuminatingly discussed in Müller, ‘Radical Subjectivity: Morality versus Utilitarianism’, 115–132. 6. Wollheim, The Thread of Life, 237. 7. See his Anarchy, State, and Utopia (Oxford: Basil Blackwell, 1974), 28–35. 8. The congratulatory self-awareness of the megalopsychos (IV.3) is not something that modern Aristotelians would wish to defend.

REFERENCES Primary Plato

Apology Crito Gorgias Symposium

Aristotle De Anima (DA) De Motu Animalium (DMA) Eudemian Ethics (EE) Nicomachean Ethics (NE)

Secondary Müller, A. W., ‘Radical Subjectivity: Morality versus Utilitarianism’, Ratio 19 (1977), 115–132. Nozick, R., Anarchy, State, and Utopia (Oxford: Basil Blackwell, 1974). Price, A. W., Virtue and Reason in Plato and Aristotle (Oxford: Clarendon Press, 2011). Wollheim, R., The Thread of Life (Cambridge: Cambridge University Press, 1984).


The Benefit of Virtue Christoph Halbig

Bonum est multiplex—this time-honored scholastic slogan no doubt applies to human life: Its goodness comprises several dimensions. Two of the most important among them are the virtuous life and the happy life. The question of how these two dimensions are related to each other has been at the center of philosophical attention at least since Plato’s dialogues. Is leading a virtuous life the key to leading a happy life? Is it sufficient to be virtuous in order to be happy, as the Stoics held, or is it at least necessary, as Aristotle held? On the other hand, don’t the virtues demand sacrifices from their possessors even to the point of giving up one’s life and thus renouncing the pursuit of happiness once and for all? Nietzsche seems to have a point when he talks of the virtues victimizing (instead of benefitting) their possessors.1 The question of how the happy and the virtuous life are related to each other is of the utmost practical relevance: many of the most important decisions of our lives depend on the implicit or explicit answer we give to it (unless the very fact of having become a virtuous person settles the question for the better, as Aristotle would have thought, or the worse, as Nietzsche would have). But the question is also of utmost theoretical or philosophical interest: how much so is actually the subject of much debate in contemporary virtue theory. Many philosophers, especially in the neo-Aristotelian tradition, diagnose a conceptual link between virtues and happiness or flourishing (which they tend to prefer as a translation of the Greek eudaimonia). Virtues are defined as those character traits that human beings need in order to flourish.2 Its contribution to human happiness provides the key to the nature of virtue. Other philosophers from otherwise quite different traditions3 disagree with such an approach; they try to provide an independent account of the nature of virtue that might nonetheless allow them to provide an explanation for the fact that the virtues tend to be beneficial. The attempt to forge a conceptual link between happiness and virtue, however, seems to recommend itself for at least three reasons: First, it provides for theoretical unification4—instead of being settled with a dualism of virtue and happiness, the one is understood in terms of its contribution to the other. Second, it promises to pave the way for a naturalistic grounding of the virtues. At the very least, the category of happiness or flourishing seems


Christoph Halbig

to be more amenable to a naturalistic reading than that of virtues. Third, the old question of ‘why be moral?’ seems to find a ready answer: because it is necessary or sufficient for leading a happy life. This third reason, like the second one, relies on an asymmetry between the virtues on the one hand and human happiness on the other, this time not an ontological asymmetry, but a normative-cum-motivational one. It seems uncontroversial that everyone of us has a reason to seek his or her own happiness and a motive to do so—but do we really have a reason (and a corresponding motive) to act according to the demands of virtue? And, are there really such asymmetries, and do they really speak in favor of such a project as to forge a conceptual link between happiness and virtue? Ethical non-naturalists, rationalists about reasons, and moral internalists in the theory of motivation would hardly be convinced. If this question has to be answered in the negative, much of the motivation to account for the virtues in terms of their contribution to human happiness seems to evaporate. It even might look promising to turn the tables on those who seek to understand the virtues via their contribution to happiness and try to understand the good for of happiness in terms of the good of the virtues. Some basic prudential goods seem to be at least partially constituted by some of the virtues. Or something even stronger might hold: being overall virtuous might turn out to be a condition of basic prudential goods being goods in the first place (sadistic pleasure, for instance, might cease to be even a prudential good). We have gone full circle. If such approaches as the two last-mentioned are on the right track, we need an independent account of virtue in order to avoid a (more or less) vicious circularity: our hold on the concept of prudential value seems at least as much in need of our understanding of virtue as vice versa. Unfortunately, empirical research does not offer much help on these matters. Even if evidence came up that points to a considerable amount of covariance between virtue and happiness, that would still leave the question open as to which kind of connection exists between them (for instance, a causal one? a constitutive one?) and as to which way the connection actually goes. Is it that virtuous people tend to be happier just because they are virtuous, or is it rather the other way round: being happy, as ordinary wisdom has it, helps a lot in being a virtuous person. Whereas unhappiness, for instance, shuts one off from one’s fellow beings, happiness might open one’s eyes to the pains, needs, and expectations of others, thus facilitating the formation of some crucial other-regarding virtues. I A discussion of the connections between virtue and happiness obviously needs quite a bit of stage-setting for both the virtue and the happiness side of the problem. For lack of space, however, I have to be very brief here and confine myself to some rather dogmatic remarks. As to what it means to benefit someone, I will presuppose that there is a basic distinction between the good sans phrase and the good for: In order to show that an entity x benefits y it is not enough to show that x makes y a better

The Benefit of Virtue


person. It has to be shown that x makes for a better life for y. What counts as a benefit for x may depend solely on x’s individual makeup, or it may also depend on other factors, most prominent among them being his membership of a biological species. It might even be that the internal and the external aspects are interrelated—the virtues might be a case in point: Neo-Aristotelians try to show that they not only enable their possessors to live well but that they make them good human beings and that both features are inextricably intertwined.5 Unfortunately, I have to put this whole complex of the connections between the virtues, human goodness, and the good life for human beings to one side. I will also dismiss out of hand forms of strong objectivism about human happiness, that is, positions that hold that happiness is completely independent of subjective attitudes. How a person feels about her life, whether she enjoys herself, etc., is at least partially constitutive of her happiness. I again have to remain non-committal on how the subjective and the objective dimensions are related to each other—for instance, on how experiences of enjoying one’s life or being content with one’s life are related to ecological criteria like ‘the bright eye and the gleaming coat’6 on the one hand, and to the depth7 of such a life that brings into play the question of what a life is about, what it is focused on, on the other. The contribution the moral virtues make to the depth of our lives, however, will be addressed later. As to who it is who is supposed to benefit from the virtues, I will focus exclusively on their possessor. I will not consider other candidates for the role of the beneficiaries of virtue, for instance, (a) those who are in contact with virtuous persons, (b) the communities of which the virtuous persons form a part, or (c) the life-form characteristic of human beings. Let us now turn to the virtue side of our problem. Prima facie it seems relatively easy to make a case for a benefit that comes out of virtue for particular kinds of virtues, that is, the self-regarding or the structural ones, for instance, temperance or industriousness. In the following, however, I will focus instead on a discussion of the other-regarding, moral virtues, as these pose the most impressive challenge to the idea of a prudential value of virtue as such. As to the general theory of virtue, I will try to remain as non-committal as possible. Suffice it to say that I have of course to exclude any theory of virtue that, by its very structure, begs our central question, like for instance the aforementioned attempt (quite popular especially in the neo-Aristotelian tradition) to define the virtues through their very contribution to happiness. II After this much too brief look on the two relata of the connection of virtue and happiness, let us now proceed to the analysis of the connection itself. I would like to distinguish between five kinds of such a connection in decreasing order of strength: (i) Being overall virtuous is both necessary and sufficient for the happiness of the virtuous person.


Christoph Halbig (ii) Being overall virtuous is a necessary condition on other goods being of prudential value for their possessor and thus contributing to her happiness. (iii) Being overall virtuous is itself a basic kind of prudential value. (iv) Possessing some virtue or other is a necessary condition on other goods being of prudential value for their possessor. (v) Being overall virtuous is a contingent, personal project whose successful realization is of prudential value for its possessor.

Let me make some comments on each of them in turn. Both the strongest and the weakest kind of connection can be dismissed quite swiftly: The weakest one (v) would altogether trivialize the connection between virtue and happiness—virtue would come out as just one project among others that could as well be replaced with the pursuit of vice. Its prudential value would not lie in itself but in its being the object of a personal project whose successful realization would be rewarding as such and thus of prudential value quite independent of its content. The strongest one (i) is represented by the Stoic thesis that virtue is the only kind of value in the strict sense of the word—other kinds of value like health are at best so-called preferred indifferents. One might object at this point that virtue as the only bearer of genuine value falls into the category of the good, not into the category of the good for, and thus does not bear on the question of prudential value at all. This, however, would hardly be consistent with the Stoic claim that a life of virtue is a life that is ipso facto worth living—it is rewarding for the person leading it, even if that person is bereft of all preferred indifferents and is exposed to many of the dispreferred ones. Virtue would thus both be necessary and sufficient for happiness; there can be no sacrifice in the service of virtue because virtue is simply the only kind of intrinsic value in the game. The Stoic theory of value, however, needs no separate discussion here since even the somewhat weaker kind of connection (ii), which does grant genuine value to other entities beside virtue, will, as we shall see now, not stand critical scrutiny. The second but strongest (ii) kind of connection represents what in recent debates is often labeled as a moralized conception of happiness. According to it, no other goods (than virtue) like health, achievement, etc., are of prudential value to their possessor if they come at the expense of virtue. And no other evil (than vice) is of prudential disvalue to its possessor if it is suffered in following the requirements of virtue. That implies that sacrifices required by virtue cannot count as losses in happiness. In contemporary philosophy, such a position has been defended, for instance, by D. Z. Phillips, who holds that ‘death for the sake of justice is not a disaster’8 and that for the virtuous person who dies for the sake of justice, death itself even turns into ‘a good’9; John McDowell claims that the idea of happiness characteristic of the virtuous person equip[s] him to understand special employments of the typical notions of ‘prudential’ reasoning—the notions of benefit, advantage, harm, loss,

The Benefit of Virtue


and so forth—according to which (for instance) no payoff from flouting a requirement of excellence . . . can count as a genuine advantage; and, conversely, no sacrifice necessitated by the life of excellence, however desirable what one misses may be by other canons, can count as a genuine loss.10 Now, such a moralizing conception of happiness that makes the very idea of losses required by virtue unintelligible clashes with obvious counterexamples: If I am speaking up for a friend who is persecuted by a brutal regime for publicly drawing attention to its crimes, that might very well be what virtue demands, but it might also lead to my imprisonment and even physical and psychological torture. This might not only be intensely painful, but might even impair the exercising of some of my basic capacities as a human being forever. Can it be true that even in such a case I would have suffered—in McDowell’s words—‘no genuine loss’,11 or that even after undergoing the permanently debilitating torture I would have—in Philipps’s words—‘accomplished all’12 I could have asked for not just in moral, but also in prudential terms? Some distinctions are in order. First of all, it has to be clear at the outset that the defender of the no-loss-at-all thesis has to shoulder a heavy burden of proof: he has to show that the virtues function as conditions that not only are able to change the overall evaluative valence of a state of affairs but also are able to change their prudential evaluative valence. Let us take an example: My friendship with B is based on withholding from her some important piece of information that she has a right to learn but that would conceivably be incompatible with our entering into deeper personal relations. Now one might argue that insofar as my violation of B’s rights somewhat spoils our friendship, it is definitely not as good as it should be (and would be) if it were not built on such culpably shaky foundations. The friendship might be diminished in overall value or even spoilt to such an extent that it changes its evaluative valence—it would then turn into an overall bad thing. But is it diminished in prudential value for me, or does it even acquire a prudential disvalue for me? That seems at least a different question. Maybe the friendship works just as well for me, and I get all one could expect out of such a relationship. Such a friendship probably would have been impossible (or unlikely) if I had been a fully virtuous person in the first place, but that fact seems irrelevant for its eventual prudential payoff. At this point, one might follow a hang-tough strategy and argue that such a friendship might not even be a good thing for me in the first place. Enjoying the external benefits, I might simply not realize that personal relations based on suppression of facts, violation of rights, etc., actually fail to contribute to one’s own happiness the way genuine friendships do. At this point, however, we have, secondly, to take note of an important good/bad asymmetry: even if one might be tempted by such a hang-tough strategy in dealing with our example of the spoilt friendship,13 it is rather doubtful whether such a strategy proves as successful in the case of evils


Christoph Halbig

turned neutral—or even changed into goods—as it might prove to be in the case of goods that are spoilt and thus turned into something evil. As to the case of spoilt goods, Aristotle, for instance, remarks that the pleasure taken in vicious activities is not at all good, it is rather something bad.14 In the same vein, goods that presuppose viciousness might indeed lose even their prudential value. In the context of the no-loss-at-all thesis, however, what has to be shown is that something bad like pain that might have to be endured as the price of virtuous behavior, thereby (by its being required by virtue) loses its prudential disvalue and becomes prudentially neutral (as McDowell’s no-loss-at-all metaphor suggests), or even acquires a positive prudential valence (as Philipps’s has-accomplished-all metaphor suggests). This seems rather hard to believe. The pain that the virtuous martyr suffers on the rack for the freedom of press might be so excruciating that his steadiness of purpose in the service of the good cause will hardly even attenuate its disvalue, not to mention neutralizing it or even turning it into a positive prudential value. That the bad in such cases stubbornly remains just that, that is, something bad, seems not only to be attested by our intuitions but also to be required by the very logic of the way we evaluate situations of innocent suffering. What makes them so hard to bear is precisely that the virtuous person is indeed suffering and that the tortures succeed in doing him serious harm. On the other hand, even the self-respect of the very person suffering requires that she admits to herself that she has made a sacrifice in the service of some higher good15—someone recovering from such a torture and flippantly declaring that she has suffered no loss at all would hardly be considered a moral role model. McDowell seems to have deluded himself into the no-loss-at-all thesis by narrowly focusing on the deliberative perspective of the virtuous person: In thinking about what to do, it might well be that possible losses for herself are indeed silenced—but that just means that they are not admitted to the deliberative process, in the sense of providing reasons that have to be weighed against other kinds of reasons. It does not mean that the fact that such a loss is undergone itself loses its prudential disvalue; what it does lose is just its normative impact on the deliberating agent.16 The bystander will sadly notice the price that the virtuous agent had to pay and will hold it all the more against the perpetrators of evil. Thirdly, it might seem that even if one accepts the asymmetry thesis it is still possible to defend the idea that virtue never requires the agent’s sacrifice of happiness. Why so? Because a life in which prudential disvalue would have been avoided by acting, for instance, cowardly and not speaking out for one’s friend, would not be (a) a life with the chance of a virtuous action foregone, but the prudential value of, say, the life-long exercise of one’s basic capacities, and thus one’s happiness saved; it would rather be (b) a spoilt life incompatible with happiness because the virtuous agent would realize that he has compromised his ideals. In facing the decision between following the demands of virtue and putting one’s bodily or psychological

The Benefit of Virtue


integrity at risk on the one hand and simply ignoring them and saving one’s bodily or psychological integrity, it would thus be inaccurate to say that the agent is asked to sacrifice his happiness. He is already in the tragic situation in which happiness has become impossible for him whatever decision he takes.17 Facing such an alternative, there might even be a priori reason to decide in favor of the virtuous option: even if one loses one’s happiness in both cases, one loses it by one’s own choice in acting viciously (one is oneself responsible for the vicious action), whereas one loses it just by the disfavor of the circumstances in acting virtuously (the torturers who punish the virtuous behavior take the responsibility).18 Nonetheless, even in such a situation it would be highly misleading to claim that virtue does not demand the sacrifice of one’s happiness. The only reason for such a claim would be that happiness is not to be had anyway.19 But it is virtue itself that is doubly responsible for that: doing what virtue demands would imply sacrificing important prudential values necessary for happiness; not doing it would spoil the goods preserved by the vicious action (or omission) by changing their prudential evaluative valence from good to bad (or at least by neutralizing it). Let us take stock of the results reached so far: Virtue is not sufficient for happiness. Even a life in which virtue is realized might well be not worth living because crucial prudential goods might be lacking in that life. It does take, as Aristotle puts it, ‘external goods’ in addition to the practice of virtue to be happy.20 The practice of virtue might require the sacrifice of crucial prudential values; under unfavorable enough social and political circumstances, such sacrifices might even be the rule rather than the exception. What one is deprived of is not just an entity that has lost its positive prudential value by its very incompatibility with the requirements of virtue, it is the genuine prudential good itself, thus having full (and perhaps fatal) impact on one’s happiness. But is virtue at least necessary for happiness? Although it seems axiologically plausible to argue that incompatibility with virtue might indeed spoil the prudential value of some goods kept or acquired at the expense of the requirements of virtue (friendships based on the suppression of some important pieces of information, etc.), those cases are not sufficient to support a general requirement that would allow the attribution of prudential value to x only on the condition that (acquiring, keeping, etc.) x is consistent with the requirements of virtue.21 To borrow a metaphor from Jonathan Dancy, entities like pleasure, achievement, bodily, and mental functioning, etc., have a prudential default value:22 They bring their value to the situation in which their value might indeed be neutralized or reversed by the conflicting requirements of virtue. Unless that happens, however, their prudential value remains switched on. A non-virtuous person x, who happens to live in a well-ordered society that rewards behavior that does not conflict with the requirements of virtue, might, by sheer luck, enjoy a sufficient amount of goods whose prudential value simply remains switched on by the sheer


Christoph Halbig

lack of tension with the requirements of the virtues (vicious action is simply discouraged by the social framework; so the potentially ‘spoiling’ aspects of x’s character simply do not come to bear on his actions). Virtue is thus neither sufficient nor strictly necessary for happiness, although, as a matter of fact, it might be de facto necessary for most of us. The third kind of connection (iii) considers virtue as one fundamental category of prudential value among others. It allows that shortcomings in virtue might be compensated by an increase in other kinds of prudential value. Under fundamentally unjust political or social circumstances, parents caring for their children’s overall happiness might thus be tempted to inculcate only a moderate amount of virtue into their children so that they might fare better along the dimensions of, say, pleasure or social relations.23 Still, of two lives equally endowed with all the other prudential values, one of them being overall virtuous, the other vicious, we would, if the third kind of connection holds, consider the first of those lives not only as the better life sans phrase, but also as the better life for the person leading it. If we just focus on the question of prudential value and ignore questions of merit, etc., we would, it seems, indeed feel sorrier for the vicious person—his viciousness has deprived him of a crucial dimension of prudential value.24 But do we? At this point, I have to admit that it strikes me as highly artificial to take virtue as just another item on a list of prudential value and examine its status by keeping the amount of the other values in a life constant while varying it (by comparing the lives of the virtuous and the vicious). First, as we have already seen, virtue and vice interact in complex ways with those other values in which, for instance, achievement and knowledge as examples of other basic kinds of prudential value do not interact—up to the point of switching off their prudential default value (whereas, for instance, lack of achievement hardly spoils the prudential value of knowledge). These holistic relations stand in the way of keeping the cetera really paria in checking our intuitions on the vicious and the virtuous person otherwise equally endowed with prudential values. Things are further complicated by the fact that, second, our reactions of sympathy to matters of virtue and vice are strongly influenced by our moral attitudes that might confuse our prudential intuitions—the sympathy for the vicious person that we would otherwise feel might be superseded by our moral indignation at his self-depravation, etc. Third, virtues seem to contribute in complicated ways both to other fundamental dimensions of prudential value and to other dimensions that, though in themselves non-prudential, are nevertheless relevant for the happiness of a life. An example relevant in the context of our question is the dimension of depth:25 A life of virtue might be, if not the only, then at least a privileged way to give one’s life as a whole a certain depth that might enhance its quality along dimensions of prudential value like enjoyment—even great art, for instance, seems to resonate better with ‘deep’ characters than with shallow ones. A vicious life of indulging in petty, self-centered concerns might be successful in its own terms, but might prove incompatible with being a deep

The Benefit of Virtue


one. The reason why we feel sorrier for a vicious person than for a virtuous one, although none fares better along all other fundamental dimensions of prudential value except (possibly) virtue itself, might thus not be that he lacks an independent category of prudential value but—rather indirectly— that qua vicious he has condemned himself to leading a shallow life, a life that then might come even at a price in terms of basic prudential values. But, even if virtue forms no basic category of fundamental prudential value in itself, and quite apart from the role it plays in other dimensions of the good—if not prudentially good—life, some specific virtues or bundles of virtues—as opposed to virtuousness in general—might still play a constitutive role for basic prudential values. This is the possibility to which we will turn now. The fourth kind of connection (iv) drops both the requirement contained in (ii), namely that overall virtue might be a prerequisite for other goods being of prudential value for their possessor, and the fundamental idea of (iii), namely that overall virtue is itself a fundamental kind of prudential value. It approaches the question of the connection between virtue and happiness in a rather piecemeal fashion by trying to prove for individual kinds of prudential value26 that they presuppose some particular virtue or bundle of virtues. Achievement for instance, as Michael Slote has pointed out,27 unlike mere talent, seems to require the virtue of perseverance. Whereas sheer talent may come as windfall, genuine achievement requires a firm disposition to persist in the face of difficulties—which seems to be the core of the virtue of perseverance. Another important fundamental category of prudential value might be knowledge of important matters. Knowledge as a prudential value might be related to virtue in different ways. In the first instance, virtue itself implies a cognitive component; at least some crucial dimensions of moral knowledge seem to be the prerogative of the virtuous. If that holds true, an important part of knowledge, that is (a part of) moral knowledge, would require the virtues.28 In the second instance, knowledge quite independently of its content seems to require virtues such as courage—the pursuit of knowledge might put into question dearly held convictions, shatter long-cherished religious beliefs, etc. It takes the virtue of courage to avoid psychological escape mechanisms like wishful thinking in the face of these obstacles.29 Both examples—that of knowledge and that of achievement—however illustrate two structural difficulties of this fourth kind of approach to the connection between virtue and happiness. First, it is at least not obvious that some single virtues or a bundle of virtues are really necessary for the prudential good in question. For instance, it might be asked whether moral knowledge cannot be had without being virtuous at all; the self-possessed, for instance, might have full moral knowledge without being virtuous—he might just lack the necessary wit in doing what he knows he should do, etc. Second, it is questionable whether the virtues are really necessary for the prudential good as such: virtues such as courage might prove instrumentally useful in acquiring knowledge, but is the way it has been acquired essential to its status as prudential


Christoph Halbig

good for the one who possesses it? At least in the cases of achievement and of prudential goods that involve deep personal relations, such as friendship or life, it seems, however, not hopeless to try to identify such an essential connection. A fortune won in the lottery instead of having been amassed by relentless, perseverant, and courageous efforts of my own seems to enhance my life considerably less even in prudential terms. In the same vein, being generous and trusting toward others might not only prove instrumentally useful in building up social ties, it seems constitutive of the prudential value of genuine friendships or relationships of love. Having put myself at risk in unselfishly trusting another, it makes the resulting relationship a better thing for me.

III Does virtue benefit its possessor? We have discovered various respects in which it does. Some single virtues, especially self-directed ones, and possibly whole kinds of virtues (like the executive ones—the courageous robber is not a better person for being courageous—he is much better at his job than he would be if he were a coward—but it is arguably better for him to be courageous) are undoubtedly of great instrumental value, though their value very much depends on the contingencies of the social and political framework, conditions of scarcity, etc., in which the life of the virtuous person is led. Even apart from their instrumental value, some single virtues or bundles of virtues play a constitutive part in basic dimensions of prudential value, like deep personal relationships, achievement, or even knowledge. Being overall virtuous, however, does not form in itself a basic category of prudential value. It contributes nonetheless to other dimensions of a person’s life, which, although themselves likewise non-prudential (like depth), have at least important repercussions on its overall prudential value. Finally, although being overall virtuous does not prove sufficient for a good life, being overall vicious might in many cases spoil one’s happiness. Although the goods virtue requires us to forgo in many cases retain their prudential value, goods that can only be retained or acquired by flouting the requirements of virtue can indeed change their prudential evaluative valence and turn into something that, outward appearances notwithstanding, does not benefit us at all. Such a position implies rather the reverse of the Stoic optimism, namely that virtue can guarantee our happiness come what may. What is to be expected instead is that we are quite often exposed to situations of tragic choice in which we know in advance that our happiness is at risk or even lost, whatever option we choose. Doing what virtue requires might imply depriving us for good of crucial prudential goods; not doing it might spoil those goods and deprive them—even if retained—of their prudential value for us. But what about situations in which being virtuous does not benefit its possessor, where it does ask for sacrifices? Unfortunately, I won’t be able to go deeper into the intricate problem of how to weigh the claims of happiness

The Benefit of Virtue


against those of virtue here. Let me just conclude with some remarks on the question of who is going to raise that problem in the first place: Is it a problem that only comes up in the third-person perspective of an observer who tries to find out whether a loss in happiness undergone by x acting virtuously is overall justified? Or is it a problem that comes up in the deliberative perspective of x herself, who is about to decide what she is going to do? The question ‘Am I going to benefit from a virtuous action?’ is very unlikely to guide the first-person deliberation of the agent. What she will be concerned with, especially if she is virtuous, is to lead a good life simpliciter, not a life that is good to live for her own sake.30 There is no doubt that prudential values will figure in that perspective. Even the good life simpliciter is after all the life of the one who is leading it, and she is rightly concerned with her own well-being. In addition, attention to prudential values (or disvalues) is often a constitutive part in identifying the requirements of virtue. As we have seen earlier, a courageous person, unlike a merely rash one, will take the risks he is undergoing for himself into consideration, although if a good of others of sufficient weight is at stake, the normative force of these risks might be completely silenced. Even in cases where the requirements of virtue imply a loss in happiness and silence or at least outweigh this prudential loss, it is however implausible to suppose that the deliberating agent might be concerned with the amount of that loss: First, it seems incompatible with a virtuous character to explore the ‘price tag’ attached to saving the life of one’s neighbor, for instance. Second, it seems highly unpractical to compute the prudential gains and losses—as we have seen, the virtuous act might have some at least partly compensating prudential value. It would, however, be pointless for the deliberating agent to ponder over such a compensation; this would contribute nothing at all to the very point of deliberating, that is, to finding out what to do all things considered. Third, some aspects relevant to prudential value might necessarily elude the deliberative perspective. For instance, the pleasure taken in helping someone else might only be a true prudential reward if it comes as a mere byproduct of the (successfully realized) intention to help; if directly aimed at it would lose its value (helping someone as a means to one’s pleasure in it tends to defeat that very pleasure—one gets the genuine thing only when one is engaging in such an activity for its own sake). Doing so would thus be a self-undermining endeavor. In addition, the prudential value of an action might be augmented or diminished by sheer luck; causal chains exercise an influence that is beyond the deliberating agent’s control. But it is not just the benefit side that is unlikely to figure prominently in the deliberative perspective. Virtue-ethics is notoriously beset by the problem of self-indulgence. If it is, as virtue-ethics holds, the virtues that provide the


Christoph Halbig

sources of reasons for action, then it should be expected that in deliberating over what to do, we try to track those very sources of our reasons for action. This, however, seems morally repugnant. A compassionate person should genuinely care for the person in need, not for himself caring about such a person qua possessing the virtue of, say, compassion; it is the former and not the latter that should take center-stage in his deliberation.31 In order to avoid this problem of self-indulgence, virtue-ethics might even be compelled to become what Derek Parfit32 labels a self-effacing theory, thus positively discouraging its adherents from deliberating in its own terms.33 Without being able to go deeper into the problem of self-indulgence, it seems uncontroversial that the virtues, even if of crucial importance for the deliberative stance of the agent, should not be the focus of its content—thus turning the agent’s attention to his own character in a morally questionable way. The person who suffers a loss in the service of virtue should, in her own deliberating perspective, suffer that loss not for the sake of her own virtue, but rather for the sake of demands of the moral situation that the virtue in question is concerned with (in the case of compassion it is concerned with the misery of one’s neighbor). Both parts of the ‘does virtue benefit its possessor?’ question thus seem alien to the first-person perspective of the deliberating agent. In that perspective, prudential aspects are considered only in part and under the constraints mentioned previously, whereas virtue effaces itself by directing the agent’s attention away from herself towards the demands of the situation that call for her action.34 Looking back on her own past actions, the agent might of course take stock in a third-person perspective: If she comes to the conclusion that virtuous action regularly implies heavy losses in terms of her happiness, such an observation might encourage her to rethink the normative relevance of virtue on the one hand, happiness on the other, and the connection between both of these terms. Or she might adopt a critical stance toward the contingent social, political, or economical circumstances she finds herself in. In this chapter, I have tried to disentangle the various ways in which virtue does contribute to the benefit of its possessor. Losses and tragic dilemmas, however, remain possible. To what extent they become actual depends on those contingent but also (at least in part) malleable circumstances. So, notwithstanding the various levels pointed out in this chapter on which virtue and happiness are connected, and notwithstanding the complex ways in which these levels interact, it seems safe to state at least one thing: At each historical moment it seems to take much virtue and considerable sacrifices particularly from the virtuous to keep the virtuous life from becoming too demanding in terms of happiness.35 NOTES 1. Cf. Nietzsche, Die fröhliche Wissenschaft, § 21, ‘An die Lehrer der Selbstlosigkeit’. 2. Cf. Hursthouse, ‘Virtue Theory and Abortion’, 226.

The Benefit of Virtue


3. For instance Adams, A Theory of Virtue; Driver, Uneasy Virtue; Hurka, Virtue, Vice and Value. 4. Cf. Slote, ‘The Virtue in Self-Interest’, 264f. 5. Cf. Hursthouse, On Virtue Ethics, 526. For a critical discussion of such an approach cf. Copp and Sobel, ‘Morality and Virtue. An Assessment of Some Recent Work in Virtue Ethics’, 526ff., and Zagzebski, ‘The Admirable Life and the Desirable Life’, § 1. 6. Cf. Williams, Ethics and the Limits of Philosophy, 46. 7. On the dimension of depth cf. Foot, Natural Goodness, 86ff. 8. Phillips, ‘Does It Pay to be Good?’, 50. 9. Ibid., 51. See also ibid., 60, where Phillips claims that the ‘man who chooses justice’, even if he does not profit as the rogue might well do in terms of acquiring power, wealth, etc., ‘has accomplished all.’ 10. McDowell, ‘The Role of Eudaimonia in Aristotle’s Ethics’, 369. In ‘Some Issues in Aristotle’s Moral Psychology’, however, McDowell has added that eudaimonia/happiness marks out just ‘one dimension of practical worthwileness’ (122). So a serious injury, for instance, incurred in the pursuit of virtue might well turn out to be a genuine loss along some other dimension. Nevertheless, McDowell leaves no doubt that the dimension of eudaimonia/happiness is privileged with respects to the other dimensions—it marks out ‘excellence par excellence’ (123), which in turn implies that for someone who has learnt to appreciate that dimension ‘nothing else matters for the question what shape one’s life should take here and now, even if the upshot is a life that is less desirable along other dimensions.’ In a similar vein, Müller, speaking of a ‘secret’ contained in the virtuous person’s attitude toward her happiness, denies that the virtuous really believe that their goodness could actually deprive them of their happiness, cf. Müller, Was taugt die Tugend?, 191. 11. McDowell, ‘The Role of Eudaimonia in Aristotle’s Ethics’, 369. 12. Philipps, ‘Does It Pay to be Good?’, 60. 13. Slote (‘The Virtue of Self-Interest’, 274), who is not in the least tempted by the hang-tough strategy in the case of spoilt goods, at least implicitly subscribes to what I have labeled the asymmetry-thesis by noting that such a strategy would be ‘even more implausible’ in cases of prudential disvalues neutralized or turned into prudential values by being required by virtue. 14. Cf. Aristotle, Nicomachean Ethics (NE) X.5, 1175b26–28. 15. Cf. Haybron, ‘Well-Being and Virtue’, 10, for a more detailed argument that tries to prove the incompatibility of the no-loss-at-all thesis with the demands of self-respect. 16. And even the deliberating agent had better be aware that there is such a prudential disvalue involved, even if he immediately goes on to silence them in terms of their normative force—otherwise the distinction between, for instance, the courageous and the rash agent could hardly be made (the courageous agent unlike the rash one is aware of the serious risks he exposes herself to in acting virtuously). For McDowell’s theory of silencing see ‘Virtue and Reason’, § 3 and ‘The Role of Eudaimonia in Aristotle’s Ethics’, 370. 17. Cf. Foot, Natural Goodness, 97, who argues that in such a situation, the virtuous agent would not describe herself as sacrificing her happiness but as realizing ‘that a happy life had turned out not to be possible for him’. 18. This point is made by Zagzebski, ‘The Admirable Life and the Desirable Life’, 65. 19. As does Foot loc. cit. (n. 35). 20. Cf. Aristotle, NE, 1153b19. 21. Such a general requirement is at least hinted at in Foot, Natural Goodness, 96, who holds that ‘humanity’s good can be thought of as happiness, and


Christoph Halbig

22. 23.


25. 26.

27. 28.

29. 30.


32. 33. 34. 35.

yet in such a way that combining it with wickedness is a priori ruled out’. (Although some paragraphs earlier Foot allows the combination of a wicked character and a happy life as at least a conceptual possibility, ibid. 92.) Cf. Dancy, Ethics Without Principles, 184–187. For the possibility of such trade-offs between virtues and other prudential goods that might lead parents to recommend ‘a mixed virtuousness’ to their children see Copp and Sobel, ‘Morality and Virtue. An Assessment of Some Recent Work in Virtue Ethics’, 527ff. For an ‘argument from lack of sympathy’ (for the vicious person who has exactly the same amount of prudential values as the virtuous person, with the exception of virtue itself) to a rejection of virtue as a fundamental category of prudential value see Hooker, ‘Does Moral Virtue Constitute a Benefit to the Agent’, 149–155. For a discussion of the dimension of depth see Foot, Natural Goodness, 86ff. At the heart of Michael Slote’s so-called Platonic elevationism lies the thesis that such a connection can be shown not just for some elements, but indeed for ‘every element of human well-being’ [my emphasis, C.H.]. In case the theory of the unity of the virtues holds true, Slote’s Platonic elevationism would of course collapse into what he labels Aristotelian elevationism, that is, the thesis ‘that all elements of personal well-being must be compatible with virtue taken as a whole’ (Slote, ‘The Virtue in Self-Interest’, 274)—since the individual virtues required for some element of well-being would then in their turn presuppose all the others. Cf. Slote, ‘The Virtue in Self-Interest’, 276f. For a more detailed discussion of such a line of argument that tries to show that the virtues make a constitutive contribution to moral knowledge as a crucial dimension of the fundamental prudential value of knowledge, see Hooker, ‘Does Moral Virtue Constitute a Benefit to the Agent’, 146f. Cf. Slote, ‘The Virtue in Self-Interest’, 277f., who holds that ‘knowledge constitutes a distinctive form of personal good, and counts as wisdom, only when it takes courage to acquire it’. Ibid., 278. For this distinction see Haybron, ‘Well-Being and Virtue’, 20. Ibid., 19ff. Haybron makes some interesting observations on how Aristotle’s point of departure from a first-person, goal-setting perspective might have come in the way of developing an independent theory of well-being. This problem, however, does not seem to beset all the virtues in the same way; it seems that someone could act justly for the very reason of being just, whereas it seems unlikely that someone could act compassionately for the reason of being a compassionate person (he should definitely not care about himself ). Cf. Williams, Ethics and the Limits of Philosophy, 10f. Cf. Parfit, Reasons and Persons, 24. For a discussion of the self-indulgence problem and its possible consequences for virtue-ethics cf. Hurka, Virtue, Vice and Value, 246ff.; Cox, ‘Agent-based Theories of Right Action’; Keller, ‘Virtue Ethics is Self-Effacing’. The person and her character might of course be included in that situation. Be it only to keep the virtuous life attractive for those who are still on the way toward virtue and who might be deterred by circumstances that would make the practice of virtue ipso facto heroic (by, for instance, severely punishing acts of basic decency).

REFERENCES Adams, Robert M., A Theory of Virtue (Oxford: Oxford University Press, 2006). Aristotle, Nicomachean Ethics.

The Benefit of Virtue 51 Copp, David, and David Sobel, ‘Morality and Virtue. An Assessment of Some Recent Work in Virtue Ethics’, Ethics 114 (2004), 514–554. Cox, Damian, ‘Agent-based Theories of Right Action’, Ethical Theory and Moral Practice 9 (2006), 505–515. Dancy, Jonathan, Ethics Without Principles (Oxford: Oxford University Press, 2004). Driver, Julia, Uneasy Virtue (Cambridge: Cambridge University Press, 2001). Foot, Philippa, Natural Goodness (Oxford: Oxford University Press, 2001). Haybron, Daniel M., ‘Well-Being and Virtue’, Journal of ethics and social philosophy 2 (2007), 1–27. Hooker, Brad, ‘Does Moral Virtue Constitute a Benefit to the Agent’, in How Should One Live?, ed. R. Crisp (Oxford: Clarendon Press, 1996), 141–155. Hurka, Thomas, Virtue, Vice and Value (New York: Oxford University Press, 2001). Hursthouse, Rosalind, On Virtue Ethics (New York: Oxford University Press, 1999). ———, ‘Virtue Theory and Abortion’, Philosophy and Public Affairs 20 (1991), 223–246. Keller, Simon, ‘Virtue Ethics is Self-Effacing’, Australasian Journal of Philosophy 85 (2007), 221–231. McDowell, John, ‘The Role of Eudaimonia in Aristotle’s Ethics’, in Essays on Aristotle’s Ethics, ed. A. O. Rorty (Berkeley: University of California Press, 1981), 359–376. ———, ‘Some Issues in Aristotle’s Moral Psychology’, in Ethics (Companions to Ancient Thought, IV), ed. St. Everson (Cambridge: Cambridge University Press, 1998), 107–128. ———, ‘Virtue and Reason’, Monist 62 (1979), 331–350. Müller, Anselm Winfried, Was taugt die Tugend? (Stuttgart: W. Kohlhammer Verlag, 1998). Nietzsche, Friedrich, Die fröhliche Wissenschaft, in Kritische Studienausgabe Bd. 3, 1887. Parfit, Derek, Reasons and Persons (Oxford: Oxford University Press, 1984). Phillips, D. Z., ‘Does It Pay to be Good?’, Proceedings of the Aristotelian Society 65 (1964), 45–60. Slote, Michael, ‘The Virtue in Self-Interest’, Social Philosophy and Policy 14 (1997), 264–285. Williams, Bernard, Ethics and the Limits of Philosophy (London: Fontana 1985, 3ed.). Zagzebski, Linda T., ‘The Admirable Life and the Desirable Life’, in Values and Virtues, ed. T. Chappell (Oxford: Clarendon Press, 2006), 53–66.


Well-Being and Eudaimonia A Reply to Haybron Mark LeBar and Daniel Russell

Philosophers have been rightly enthusiastic about Daniel Haybron’s book The Pursuit of Unhappiness.1 Its central aim is to develop and defend a conception of happiness as a long-term, broadly positive emotional condition, a ‘stance of psychic affirmation’, which is ‘not merely a state of one’s consciousness’ but ‘more like a state of one’s being—not just a pleasant experience, or a good mood, but psychic affirmation or, in more pronounced forms, psychic flourishing’.2 But Haybron distinguishes between happiness as a matter of descriptive psychology and well-being as a normative notion; our concern here is with this normative notion.3 While we think that Haybron makes a significant contribution to an adequate account of well-being, we argue that even more can be gained by reconsidering Haybron’s reasons for distancing his theory from what he takes to be ‘Aristotelian’ alternatives. Haybron argues that a central component of well-being is the fulfillment of one’s ‘emotional nature’—fulfillment as a unique individual who is such as to find happiness in some things rather than others. For instance, someone who turns down a career that would make him happy, opting instead for a career doing something he thinks more important than his happiness, thereby makes himself worse off. Because he neglects his emotional fulfillment, his way of living fails to express who he truly is and, hence, is low in well-being.4 Accordingly, Haybron calls his account of well-being the ‘self-fulfillment’ view. The connections that Haybron draws between self-fulfillment and wellbeing seem to us insightful and just as important as Haybron thinks they are. However, we find the contrast he draws between his view and ‘Aristotelian’ views to be problematic in two ways. First, Haybron says that Aristotelian theories are ‘perfectionist’ theories, locating well-being in ‘the proper exercise of our distinctively human capacities’, understood as ‘being a good specimen of one’s kind, for instance, or fulfilling one’s capacities well’.5 Haybron’s worries about such a view are well taken, but we argue that such ‘perfectionism’ should be distinguished from Aristotelian eudaimonist theory. Our point is not historical. It is that without distinguishing Aristotelian eudaimonism from ‘perfectionism’, Haybron overlooks an account of well-being that capitalizes on the virtues of his own view without the defects of the perfectionist theories he rejects.

Well-Being and Eudaimonia


Second, Haybron’s self-fulfillment theory actually makes well-being wholly dependent on individual make-up, and we find that implausible. While Haybron agrees that well-being is ‘nature-fulfillment’, for Haybron this is the fulfillment of ‘the arbitrarily idiosyncratic make-up of the individual’.6 Haybron calls this thesis ‘welfare internalism’, on which ‘the constituents of an agent’s well-being are ultimately determined wholly by the particulars of the individual’s make-up qua individual (vs. qua group or class member)’.7 Haybron therefore rejects what he calls ‘welfare externalism’, on which well-being does not depend wholly on individual make-up. Although Haybron rightly rejects versions of welfare externalism that ignore individual nature, nonetheless an adequate account of well-being must be a form of welfare externalism. Not only does the Aristotelian alternative Haybron rejects escape his objections to perfectionism, but its welfare externalism also answers better to our considered convictions about well-being than does Haybron’s recommended alternative. In that case, Aristotelian eudaimonism warrants further consideration as an account of well-being and ought not to be dismissed as it is by Haybron. We begin by discussing perfectionism and its defects as a theory of wellbeing before mounting our defense of welfare externalism.

PERFECTIONISM AND WELL-BEING Haybron understands perfectionism ‘broadly enough to include any theory that takes well-being ultimately to consist at least partly in some kind of perfection’.8 More specifically: Perfection is commonly regarded as the perfection of one’s nature: being a good specimen of one’s kind, for instance, or fulfilling one’s capacities well. But I will understand perfectionism broadly enough to include any theory that takes well-being ultimately to consist at least partly in some kind of perfection, excellence, or virtue (or the exercise thereof).9 Haybron’s attack on this class of theories gets conflated with an attack on ‘Aristotelian theories’ at the outset because he takes Aristotelianism to be the ‘best-known example of a perfectionist theory’ of well-being.10 Let’s set aside (for the moment) questions about the subscription of Aristotle (or Aristotelians) to the view Haybron is targeting.11 We can simply use the term Perfectionism (with a capital P) to designate the family of views that Haybron targets, whoever may have held such views. We start with Haybron’s diagnosis of its problems.

Two Problems for Perfectionism The first problem for Perfectionism is that locating well-being in ‘perfection’ grossly under-specifies the nature of well-being. (Call this the under-specification


Mark LeBar and Daniel Russell

problem.) Haybron makes this objection with a thought experiment about the imaginary Frank, who assumes care of an orphaned autistic child; Frank thereby perfects himself by showing ‘greater virtue’ and making his life ‘more admirable’.12 However, Haybron says, Frank now has to spend more time caring for the child than developing his own interests and talents so that now his life ‘involves a lesser exercise of his human capacities: his functioning is sharply constrained and inhibited’.13 Frank’s situation, of course, is perfectly typical: perfection in one respect always costs opportunities for perfection in other respects. Indiscriminate perfection is zero-sum.14 The second problem arises because, as Haybron describes it, Perfectionism makes well-being consist in things like being a good specimen of one’s kind. So understood, Perfectionism locates well-being in a good that yields agentneutral, rather than agent-relative, reasons. (Call this the agent-neutrality problem.) Agent-relative reasons include an ineliminable reference to the agent whose reasons they are.15 Here, Haybron builds on the idea that wellbeing gives a special kind of reasons for the person whose well-being it is in a way that perfection does not. There is thus a mismatch between the kinds of reasons perfection and well-being generate. The kind of perfection involved in Perfectionism is, we might say, good simpliciter, not good for the person whose perfection it is. Agent-relative goods necessarily are good-for, while agent-neutral goods need not be. As Haybron notes, such perfection ‘bears no necessary connection to anything that can plausibly be viewed as an organism’s goals’.16 The agent-neutrality of the good of perfection is essential to Perfectionism, but even if being a good specimen is some sort of good, it has little if anything to do with wellbeing. Well-being must be a good for the agent in question, and being a good specimen just is not that kind of good. Haybron makes this point with the case of Angela, who must choose between retiring a few years early and staying in her career long enough to take another important assignment. It seems clear, Haybron says, that there would be a higher degree of human functioning, and thus greater perfection, involved in taking the assignment, but retiring early would make Angela better off.17 If the world is improved when one of its members becomes more perfect, then perfectionist value gives any agent whatsoever reason to sustain or increase it. By contrast, increases in well-being improve life for the person living that life. This sort of value gives special reason for action to the agent whose life it is: it is because this agent has this relationship with this life that this life is valuable to her in this way. Perhaps Angela’s being a better human specimen is some sort of good, but well-being in life is a good for the one living that life. Perfectionism as Haybron describes it does indeed have these two problems, and they are devastating. Whatever perfection may have to do with well-being, any account of well-being so indiscriminate as to assign just any form of perfection this kind of value must simply be mistaken. Likewise, we must neither conflate agent-relative goods and perfectionist goods nor

Well-Being and Eudaimonia


assume that one achieves the former by pursuing the latter.18 Since Perfectionism is a non-starter as a theory of well-being, we might do just as well to ignore it—except for the fact that Haybron takes Aristotelian eudaimonism to be a form of Perfectionism. But Aristotelian eudaimonism is an alternative both to Perfectionism and to Haybron’s theory of well-being, and by treating Aristotelian eudaimonism as a form of Perfectionism Haybron obscures this important third way.

Aristotelian Eudaimonism and the Under-Specification Problem Haybron argues that the under-specification problem especially plagues those theories that make virtue important for well-being, as Aristotelian eudaimonism does. It was no accident that Haybron set up his thought experiment about Frank so that it was greater perfection in virtuousness that cost him greater perfection in other areas of his life. Indeed, Haybron concludes from such considerations that ‘perfection, excellence, or virtue probably forms no fundamental part of well-being. . . . Or, alternatively, if perfection is fundamental to well-being, then it plays a smaller and very different role from that posited by Aristotelian accounts’.19 But what Frank’s case actually shows is a perfectly general problem: any view that takes some kind of self-development to matter for well-being must explain why that kind of self-development counts for well-being since all kinds of self-development have opportunity costs. Indeed, the same lesson applies to Haybron’s own view: emotional fulfillment as a unique individual is itself a kind of self-development, so he must explain why it should count for well-being in some special way among other kinds of self-development. It is this very issue, surely, that is pressed by the person who thinks that his emotional fulfillment is less important for his well-being than Haybron’s view takes it to be. So the under-specification problem is no special problem for Aristotelian eudaimonism (nor indeed for Perfectionism!), and that problem must be addressed by saying why that kind of self-development is important enough for well-being to warrant its opportunity costs. Perfectionism’s special problem is its failure to do so. Haybron’s objection to the Aristotelian account, then, must really be that the Aristotelian similarly fails to explain why virtue should have that kind of importance. Of course, the case that virtue is important for well-being depends (inter alia) on what we take virtue to be. How does Haybron think of virtue? Importantly, Haybron does not articulate, much less defend, the assumption that Frank’s decision to care for the orphan is the more virtuous one. The idea seems to be that we can understand ‘virtue’ as ‘admirability’, and admirability as a function of stereotypical ‘good deeds’. This is hardly an exotic way of thinking about virtuous activity, and if one thinks of virtue this way, then, as Haybron says, it is difficult to see why it should always warrant the sacrifice of opportunities for other kinds of self-development.


Mark LeBar and Daniel Russell

But that is not at all the sort of claim that Aristotle makes because Aristotle does not think of virtue as Haybron does. For Haybron, virtuous activity is a special type or class of activity—‘good deeds’, as opposed to other, more everyday sorts of deeds. By contrast, Aristotle thinks of virtuous activity as any kind of activity, insofar as it is done with practical intelligence and emotions that harmonize with practical intelligence—it is this activity, he says, that characterizes virtuous persons (Nicomachean Ethics [NE] I.13). A virtue is a state of character (II.5) concerned with actions and emotions (II.3), and thus with making good choices (II.6). Accordingly, Aristotle says that the virtuous person both cares about the right things and is intelligent in making decisions about those things (II.4, VI.12). Virtue therefore concerns all sorts of everyday affairs: handling things like money (IV.1–2) and recognition (IV.3–4), coping with desires (III.10–12) and emotions like anger (IV.5) and fear (III.6–9), being with friends and acquaintances (VIII–IX), even having a conversation (IV.8). Clearly, virtue involves a person’s handling of both worldly circumstances and features of his inner life that come into play in every area of life every day. This is a long way from virtue as a matter of ‘admirable good deeds’. Again, our point is not historical. It is that Haybron assumes a conception of virtue on which the under-specification problem immediately arises, but which Aristotelians patently reject. On the Aristotelian conception, activities involved in developing a talent (say) and virtuous activity are not two sets of activities. Rather, activities involved in developing talents are virtuous activities when they are done with appropriate forms of practical reasoning and feeling. So understood, virtue does have opportunity costs for other kinds of self-development—as a burglar, say, or a tyrant—but it is the kind of self-development that makes it possible for any other kind of selfdevelopment to be part of a good human life—including the kinds of selfdevelopment Haybron describes. Given its conception of virtue, Aristotelian eudaimonism has much more to say about the special importance of virtue for well-being than Haybron allows.

Aristotelian Eudaimonism and the Agent-Neutrality Problem Haybron’s second objection to Perfectionism is that, whereas well-being is an agent-relative good, Perfectionism identifies it with an agent-neutral good. Haybron also directs this objection specifically against Aristotelian theories: because of how Aristotle connects eudaimonia and virtue, Aristotle’s notion of eudaimonia most naturally is an account of the ‘good life’ as something distinct from well-being.20 However, we argue (1) that Aristotle understands eudaimonia as something we have agent-relative reason to seek and (2) that he understands the virtues as benefiting their possessor in an agent-relative way. Take the second point first: if Aristotle did not think that the virtues benefit their possessors in an agent-relative way, his argumentation would

Well-Being and Eudaimonia


make no sense. Indeed, much of the argument offered by Aristotle as well as Platonists and Stoics is intended to show that virtue is good for the virtuous agent and that a life of virtuous activity is good for the person whose life it is.21 There would hardly be a need to defend the place of virtue in eudaimonia if eudaimonia were seen as only agent-neutrally good.22 These arguments weigh in against the thought that virtue comes at a cost to wellbeing and that the life of virtue is not the best life one can live for one’s own sake. It is because the conception of eudaimonia at stake is thought to be something we have great agent-relative reason to seek that these arguments are needed in the first place. The greater point—that Aristotle takes eudaimonia to be a source of agent-relative reasons—is clear from his larger account of practical reasoning. Aristotle thinks that deliberation requires an ultimate end—an end we seek only for its own sake and for the sake of which we seek everything else—and that the good life or eudaimonia is what people already agree is that ultimate end (NE I.4, 1095a14–22). Aristotle takes himself to be entering a conversation about eudaimonia, and that is already a conversation about agent-relative goods. Aristotle observes that eudaimonia figures prominently in thought about goals, and so he begins by considering the popular views that it lies in pleasure, or wealth, or fame, or being a good person (I.5, 1095b14–1096a4). Aristotle goes on to reject all of these views, but the point is that they are all most naturally taken to be views about what is a good life for the person living it. The conversation that Aristotle joins, then, has all along been about something we have agentrelative rather than agent-neutral reasons to seek. This becomes even clearer if we look at Aristotle’s account of practical reasoning and especially the need for an ultimate end for deliberation. Aristotle starts by observing that we cannot desire everything we desire for the sake of something else since desire would then regress and thus be empty; so there must be at least one thing we desire that is such that we could desire everything for its sake and desire it for the sake of nothing further so that it can finally bring deliberation to a halt (NE I.2, 1094a18–22). Aristotle thinks there is exactly one such ultimate end (telos)—eudaimonia—which he thinks is what everyone already agrees is the ultimate end, despite disagreement about exactly what it amounts to (I.4, 1095a14–22). Aristotle then addresses that disagreement by discussing several formal constraints on anything that could properly serve as the ultimate end. For one, such an end must be something active since a life of inactivity would not be the sort of life we want (NE I.5, 1095b31–1096a2). For another, it must be something reasonably within one’s control, not a function of what other people are doing, or of how chance things turn out since a person’s ultimate end would then have too little to do with that person and his own actions (I.5, 1095b23–6). Moreover, the ultimate end must of course be something for the sake of which we could adopt all of our other ends and not itself adopted for the sake of any other end (I.7, 1097a15–b6). It also must not


Mark LeBar and Daniel Russell

leave out anything important for one’s central ends in life (I.7, 1097b6–21), and it must be a distinctly human form of life, and thus must exercise our practical rationality (I.7, 1097b22–1098a20). The formal constraints on the ultimate end have something in common: if any of them is unmet, then life lacks some important good for the person living it. For instance, notice the requirement that the good life must be desirable to such a degree that it may be said to be lacking in nothing with respect to one’s central ends. A life that is good in the sense of meeting this condition is Aristotle’s proposal for the good life that can serve as our ultimate end in deliberation, answering to the demands we place on such an end from our own first-personal perspectives as livers of such lives. As such, its value is deeply agent-relative. This explains the centrality to Aristotle’s outlook of practical wisdom and more generally the exercise of the virtues. For Aristotle, virtuous activity is an important good for a human, but not because without it one is an inferior specimen. It is important because anyone who thinks about his ultimate end without taking seriously his nature as an intelligent, emotional, and deliberating agent will live a life that is poorer for him, whether he realizes it or not. There are important questions about the existence of an ultimate end, what constraints there would be on such an end, and whether and how virtuous activity is supposed to meet those constraints. But those questions are not relevant to Haybron’s objections to the Aristotelian approach.23 Whatever we think of Aristotle’s account of the ultimate end, we cannot miss that virtuous activity enters that account as a good for the agent, someone with goals for his life. This is an important point but one that is often overlooked, as Haybron overlooks it. Aristotle’s eudaimonism is not merely a theory about what well-being is, as if that question arose in a vacuum. Rather, it is about what well-being is given the centrality that Aristotle thinks well-being actually has in practical reasoning. Haybron is correct to point out the chasm between Perfectionism and the agent’s own ends—but that is precisely why Perfectionism points us in the wrong direction for understanding what the good life consists of on a eudaimonist model. Return to the case of Angela. Perhaps continuing her work constitutes some sort of perfection that is of agent-neutral value, but for the Aristotelian eudaimonist that certainly does not settle the question of what Angela has agent-relative reason to do with her life. Perfection of that sort therefore cannot bring deliberation to a halt, so it cannot be what well-being is. Instead, the standard for wisdom here is finding an option that is congruent with the best life for Angela—best agent-relatively. Perhaps her life will be better off for retirement. Or perhaps retirement would be less wise because less good for her, if for instance it would involve a compromise on ends that Angela sees as giving her life meaning. And that is the point: in order for eudaimonia to play its role in practical reasoning, it has to have everything to do with the agent’s goals and the meaning she gives her life.24

Well-Being and Eudaimonia


Haybron is right that Perfectionism cannot explain either why one should seek perfectionist ‘virtue’ as opposed to other forms of self-development, or how such an agent-neutral good bears on one’s well-being. It does not follow, though, that there is no sense to be made of the idea that virtue is important for well-being. The Aristotelian eudaimonist conceives of virtue not as one kind of self-development among others, but the kind that makes it possible for any other kind of self-development to be part of a good life for creatures like us—creatures that live by practical reasoning and have emotions that can accord with practical rationality. Furthermore, Aristotelian eudaimonism holds that virtue is crucial for human life not for the sake of our being good specimens, but for the sake of our well-being of a distinctively human sort.

THE CASE AGAINST WELFARE INTERNALISM Haybron cannot dismiss Aristotelian eudaimonism as a competitor on the basis of his (sound) objections to Perfectionism, but Aristotelian eudaimonism runs headlong into Haybron’s objections to all forms of welfare externalism. This is our crucial point of disagreement with Haybron: we are welfare externalists but, pace Haybron, we think that that is good news. To see the motivation for Haybron’s welfare internalism, return again to Angela. What will make Angela well off depends crucially on what Angela is like—her ‘unique make-up’. But once we make that observation, it may be difficult to see how anything but her unique make-up could matter. We might then make a bold generalization: well-being is ‘determined wholly by the particulars of the individual’s make-up qua individual’ and not at all by the attributes one shares with others qua human being.25 That generalization is welfare internalism. Welfare externalism denies that claim, and to defend externalism, we offer five arguments for our claim that welfare externalism allows us to make normative judgments about well-being that we take to be strongly compelling, while welfare internalism does not.

Internalist Welfare Is Not the Relevant Kind of Welfare for People Consider an example from the documentary film Crumb about underground cartoonist Robert Crumb. At one very memorable point in the film Crumb confesses to a former girlfriend that he never loved her, and indeed that he has never loved anyone; in fact, he says, he doesn’t grasp the very idea of love. Likewise, his adult son says that although sometimes he would like to embrace his father—even by simply shaking his hand—his father ‘just can’t do that’. Suppose that all of this is true about Crumb and that his make-up is such that he is incapable of love. We can now ask whether Crumb would be better off with love in his life. This is not to ask about Crumb’s judgment


Mark LeBar and Daniel Russell

on whether he would be better off; as Haybron points out, there is no reason to think that the subject must have the last word on his well-being. Rather, we are setting aside what Crumb thinks and focusing on what Crumb is like: would love make that man better off? Our question is different from the question of whether, say, Crumb would be better off with a new car: Crumb would still be the same unique individual if we gave him a new car, but to have love in his life he would have to change deeply in his make-up. So we can ask whether a new car would be fulfilling for Crumb, given his unique make-up, but the question whether love would be fulfilling for Crumb is unlike that. After all, part of what makes Crumb the unique individual he is just is his incapacity to have love in his life. (That is why the scenes with his former girlfriend and his son are such important episodes.) Of course, one’s make-up changes over time, and perhaps Crumb could undergo such a massive change as to become capable of love; call Crumb so transformed Crumb* to mark the difference. Furthermore, perhaps Crumb* would be better off than Crumb for having a capacity for love that Crumb lacks. But that is neither here nor there: our question is whether love would make Crumb better off, not Crumb*. Evidently, the welfare internalist answers this question by saying that it is ill-formed: we cannot ask whether Crumb would be happier with love in his life because if he had love in his life then he would not have Crumb’s individual make-up. Surprisingly, it is not even a question Crumb can ask himself. Crumb’s well-being, on the internalist view, is his fulfillment given the make-up he has. And given the make-up he has, love is not part of his well-being. This is a serious problem for welfare internalism. When we ask whether Crumb is worse off for not having love in his life, one thing we want to know is whether Crumb is worse off for having a make-up that is incapable of love. This question is not what counts as well-being-for-Crumb, but whether Crumb is capable of all the well-being there is reason for him to want for his own sake—perhaps whether he is capable of something that really counts as well-being at all. To think about that sort of question, we think about what matters in human life—what things are so important for human beings that life without them is a much poorer thing. In this sense, to say that love is important for human well-being is to say something about what we humans are like. From this perspective, it is obvious that Crumb is significantly hampered in his potential for being well off. Of course, that likely gives us no license to interfere with his life. Moreover, Crumb himself may think we are crazy to think he is missing anything. But that would not show that his potential for well-being is not hampered; it would only show how deeply hampered it is. The grain of truth in welfare internalism is this: if Crumb is one sort of fellow and not another, then perhaps those who care for him should not force the issue. Crumb is a subject with a welfare—he can be better or worse off—and we can do better or worse in trying to improve his welfare. There

Well-Being and Eudaimonia 61 is such a thing as well-being-for-Crumb, after all, and for those who interact with Crumb—including Crumb himself—it is important to know what sort of welfare he is capable of and what sort he is not. If he is the sort of fellow whose welfare does not involve love, then it does not and that’s that. He is who he is: make him comfortable, and leave him alone. But notice that this is no more than what we say about the well-being of non-human animals. A dog has a make-up and can be better or worse off. Many dogs enjoy living outdoors, but if Max is not like that, then that’s that and it is no good making Max stay outside anyway. Furthermore, while many dogs are very affectionate, they are not capable of love; and we don’t pity dogs for lacking that capability. A dog is just a dog; make it comfortable, and leave it alone. So internalism does capture one way of thinking about well-being—wellbeing as a guide to ‘proper care and feeding’, so to speak. But the problem is that that falls too far short of capturing what is human about human wellbeing. To live a flourishing human life is not merely to be better off—dogs and for that matter ferns can be better or worse off, too. Surely those who care for Crumb would wish more for him than that he be made comfortable and left alone—that would be a tragic conclusion. The optimistic possibility is that he could become the sort of person who could know what fully being a person is like. While this possibility may give them no reason to interfere with him, it is clearly a wish that he could experience real human well-being, to have the distinctive kind of well-being of which humans are capable. To make sense of this further thought, we must attend to any potential Crumb might have for a better, more fulfilling make-up. This is a thought not only about Crumb’s individuality but also about the kind of being he is: his humanity. And that is the point of welfare externalism.

Internalism Can Yield No Reasons to Change One’s Make-up There is an analogy in Crumb’s case to someone who is deaf. There is no point in urging the beauty of listening to Bach on someone who cannot hear. But if there is some possibility that a deaf person might acquire the ability to hear, and he is wondering whether he should do so, then the matter is different, and the urging is in order. As in the case of impaired hearing, it matters what the story is for Crumb’s impaired capacity for love. If Crumb is not capable of becoming Crumb*, then his case is not relevant to those who can think about what will make them well off, given the normal range and panoply of capacities for human emotions and attachments. But suppose that he is capable of having Crumb*’s very different individual make-up. Is there a reason for him to pursue that possibility (by seeking the relevant therapy, say)? Unfortunately, for reasons we have seen, welfare internalism can countenance no such reasons relevant to Crumb’s well-being. For the internalist, Crumb’s well-being just is his well-being with respect to his actual makeup—the make-up of Crumb, not Crumb*. On that view, therefore, there


Mark LeBar and Daniel Russell

are no reasons for the sake of his well-being for Crumb to try to change. If Crumb can become Crumb*, this is not because of his individual make-up but only because of features he shares with other human beings, such as the capacity to love. It is his humanity, not his individuality, that opens this possibility. The internalist might reply that if Crumb is really capable of becoming Crumb* by becoming capable of love in this way, then that just is part of Crumb’s ‘individual make-up’, and hence the internalist can see that part of his make-up as imposing a normative demand on Crumb.26 But the problem for the internalist’s strategy here is, ironically, similar to the under-specification problem facing Perfectionism. The very plasticity of the nature that the internalist draws our attention to here allows for an indefinite number of ways in which we might retune our ’emotional natures’, without having the resources to distinguish between them normatively. Some ways of making ourselves over allow for increases in well-being, but some do not; internalism does not appear to have the resources to provide criteria to determine which is which. Externalism does, however. It can appeal to our shared nature as human beings, and to the process of shared reflection on those natures, and what makes us better off in light of our shared natures, to identify the lines of development that we find make us better off.27 This point generalizes, since often what we deliberate about just is what sort of make-up to have and what sorts of things to find our happiness in. Rationality does not, as Haybron suggests, contribute just by instructing us how to fulfill our commitments.28 It contributes by directing choices and actions that make substantive changes to our natures, happiness, and wellbeing. It is crucial to determining what commitments we ought to have. Curiously, the original impetus for internalism was the thought that one’s individuality is reason-giving whereas one’s humanity is not. But if there are reasons to change one’s individual make-up, then one’s individuality is precisely what cannot supply those reasons. One’s humanity must be reasongiving, too. That is the welfare externalist picture of reasons that are for the sake of well-being.29

Internalism Cannot Explain the Importance of Autonomy Haybron rightly maintains that self-fulfillment is important for welfare whether one thinks so or not. He is also correct to make autonomy necessary for self-fulfillment. Self-fulfillment, Haybron says, requires emotional fulfillment—happiness—but happiness that is based on manipulation or brainwashing, say, is not the requisite sort since in such cases ‘it isn’t really us responding to our lives: our happiness is not autonomous’.30 More specifically, Haybron says that self-fulfillment requires that ‘happiness not be based on values that are manipulated’, or on activities that one is forced to do (as in the case of slavery, for instance), or on pathological functioning (as in the case of someone perpetually high on feel-good drugs).31

Well-Being and Eudaimonia


Haybron’s endorsement of autonomy is sound, but his welfare internalism does not entitle him to make it. To see this, ask why autonomy in even the thin sense Haybron has in mind should ever be necessary for well-being in the first place. The answer is simply that it isn’t necessary for well-being full stop, but only for human well-being. Consider dogs again. Max has an individual make-up—some things will make him happy and others will not—and one can make Max better off by paying attention to that individual make-up. But it makes no sense to think that Max should be autonomous in his happiness; a dog is just not that kind of thing. (What would it mean to say that Max’s happiness should be based on values that are not the result of manipulation?) So we cannot explain the importance of autonomy for a human’s well-being in terms of his individual make-up. Rather, autonomy is important for a human’s well-being because a human is a creature of such a nature that autonomy is necessary for the well-being of creatures of that type. Human beings are capable of determining what they do on the basis of reasons in the first place; that feature of us is what opens up the possibility of manipulation or brainwashing in the first place, and that possibility is what gives autonomy its significance. The capacity for autonomy is therefore a property we share in virtue of the kind of being we are. So autonomy is important for a human only qua human, and for that reason, only the welfare externalist is entitled to make autonomy a necessary condition for human well-being. However, perhaps we could extend Haybron the courtesy of an even thinner conception of autonomy that applies to all individuals with a make-up, dogs as well as humans. Perhaps all the ‘autonomy’ that well-being requires is enough freedom from interference so that one can live in accordance with one’s individual make-up—a freedom to be oneself, if you like—whatever the provenance of that make-up, and even dogs may need to be free in that sense. A being that is just a brain in a vat ‘is liable to strike us as pathetic’, Haybron says, ‘failing badly to fulfill its nature’ because it is responding not to its life but to a ‘mirage’.32 Even Max is capable of living his own life, and a dog-brain-in-a-vat does indeed seem a pathetic thing for not living that way. Since even Max is better off with the freedom to be himself, perhaps that freedom is important for a creature’s well-being simply qua having an individual make-up. But while this thinner form of ‘autonomy’ suffices for dogs, it will not do for humans. Dogs and humans both need freedom to be themselves, but humans need the freedom to be themselves by exercising practical reasoning, by experiencing complex emotions, by making choices and living with their implications. That is, the human freedom to be oneself just is autonomy in Haybron’s more robust sense, and it is important for a human as a creature with a distinctly human make-up. Insofar as autonomy is normative— insofar as it gives one reason to live or choose one way or another—it is so only for creatures that can respond to norms and act on reasons. It is because individuals are of a kind that has that capacity—that is, because


Mark LeBar and Daniel Russell

they are human—that not being brainwashed or manipulated matters for their well-being. Moreover, autonomy of this sort is just as crucial to human selfhood. To have a self in the way that humans do is to choose and act and live through the exercise of practical reasoning. Because dogs are not autonomous, there is no ‘self’ that a dog has, even though a dog can have an individual makeup and be better or worse off. For a human, however, well-being requires the fulfillment not merely of whatever individual make-up one happens to have, but the fulfillment of a make-up that genuinely counts as a self of the sort that is characteristic of human beings. Indeed, Haybron takes the self in self-fulfillment to be a self by which one has an understanding of one’s identity and life.33 But to say that well-being requires the fulfillment of such a self is to say that human well-being requires the fulfillment of a distinctly human self. Other welfare subjects simply do not have a self in anything like that sense. Evidently, making autonomy and selfhood important for human wellbeing means that welfare internalism needs to smuggle in the centrality of one’s humanity to one’s well-being—the very idea that internalism rules out and that welfare externalism consists in.

Internalism Cannot Explain the Value of a Richer Life A similar point can be made about Haybron’s claim that well-being depends in part on the ‘richness’ of one’s life. Haybron says (plausibly) that happiness is of greater value with respect to well-being when (ceteris paribus) one’s happiness ‘is grounded in richer, more complex ways of living. For such ways of living more fully express one’s nature’.34 Given his welfare internalism, by ‘expressing one’s nature’ Haybron must have in mind ways of living that more fully express one’s individual make-up, not one’s humanity. So, for example, if Fred can find happiness in playing push-pin and in reading poetry, then on this criterion Fred would be better off devoting himself more to poetry than to push-pin (ceteris paribus) since (let’s suppose) devoting more time to poetry would yield a richer and more complex way of living for Fred. But whatever the merits of the richness criterion, it is not clear what greater complexity has to do with more fully expressing Fred’s nature as an individual. If Fred finds no more happiness in poetry than in push-pin, and if Fred’s well-being depends entirely on his individual make-up, then it is very difficult to see what difference the greater complexity of reading poetry could make with respect to Fred’s well-being. Surely it is because Fred shares the human property of being able to develop in complex and challenging ways that it makes sense to count richness as important for his well-being. Of course, Haybron may simply stipulate the richness criterion, but he cannot explain why there should be such a criterion in the first place without adverting to an aspect of our humanity of precisely the sort that internalism rules irrelevant.35

Well-Being and Eudaimonia


The welfare internalist must understand a ‘richer’ activity as more complex than other activities that also express one’s individual nature, not as an activity that more fully expresses one’s human nature. So understood, the richness condition requires that one not be a stunted version of oneself, but it still allows that one might be a stunted human being.

Internalism Must Treat All Emotional Fulfillments as Equal As we said at the beginning, Haybron is correct to say that emotional fulfillment is a central aspect of a person’s well-being. Self-fulfillment can be part of an externalist account of well-being, as well. Our final argument, though, is that we can make much better sense of the importance of emotional fulfillment for well-being on externalist grounds.36 The idea that emotional satisfaction is important for well-being has a couple of important implications: (1) well-being is diminished if certain forms of emotional experience are missing from one’s life, and (2) wellbeing is diminished if one’s emotional experience is of the wrong kind. We saw the first point in the case of Robert Crumb: because Crumb cannot love others, he cannot experience the emotions associated with love, either. That lack makes Crumb worse off, although that is a point the externalist, not the internalist, is in position to make. The second point is a new one. An adult with the emotional nature of a child would be seriously hindered in his capacity for human well-being, even if he were ‘fulfilled’ in that emotional nature. Consider Chance the gardener in the film Being There. Chance is a sunny but simple-minded, middle-aged man who would rather watch cartoons on TV than make love to a beautiful woman. Chance is about as emotionally fulfilled as he can be, and that is not trivial. Even so, Chance’s form of emotional fulfillment is a poor thing; he is a curiosity but not an object of envy. The reason is that Chance is incapable of a normal emotional life for the kind of creature he is, namely an adult human being. That, too, is a point that we can explain only on externalist grounds: Chance has an individual make-up and can be better or worse off, but to be capable of human well-being Chance’s emotional nature would need to be, well, more human. We can explain this further by taking up a point we believe Haybron has misunderstood in earlier writers: there is something significant about the kind of life we wish for our children.37 We do not wish for them a life of pleasant sensations or constant bliss. That would be unrealistic and unnatural for beings of our kind; they would not be human lives.38 Moreover, we would not wish for them the life of Genghis Khan, however successful or delighted a tyrant he might have been. Now, Haybron takes this point but dismisses its significance as not relevant to our concerns for our children’s well-being.39 This seems to us confused. It is because we care for our children as we do—that is, for their own sakes—that we want them to have lives that they experience as choiceworthy and good, as lives that it is good for them


Mark LeBar and Daniel Russell

to live. If it isn’t their well-being that is at stake, we don’t know what it is. And we see the contours of such lives as fixed by the kinds of being they are. We do not wish for them well-being as appropriate for a dog, or for Crumb, or for Chance, or for Genghis Khan. To be sure, there will be many reasons that we do not want the life of Genghis Khan for our children: such a life would be morally horrific, disgustingly violent and bloody, and so on. However, those are reasons not to wish that life for anyone, whereas by focusing on what we would or would not wish for our own children, we occupy the perspective of someone whose primary interest is in their welfare. From that perspective, it becomes clear that we want for them fulfillment of the sort possible for human beings, and we want it for them for their own sake. If emotional fulfillment has anything to do with human well-being, it must be understood along externalist lines. This form of externalism grants what is, we think, the most important of Haybron’s aspirations for welfare internalism, without internalism’s explanatory incapacities. CONCLUSION Aristotle said famously that any adequate conception of human well-being must always keep in view the mode of living that is distinctive of human beings (what he calls their ergon or ‘function’).40 This is not because Aristotle thinks that well-being is the perfection of human capabilities, however. It is because we can never hope to give an adequate account of human wellbeing unless we first say what is human about human well-being. That is what welfare externalism requires. We have argued that Aristotle is on to something important: in addition to escaping the problems for Perfectionism, the Aristotelian alternative better fits an array of our deep convictions about well-being than does welfare internalism. Aristotelian eudaimonism is interesting not because it may justly be ascribed to Aristotle, and not even because it is congruent with Haybron’s best insights into what is good for us. It is interesting because it is the most compelling conception of human well-being we know of.41

NOTES 1. Haybron, The Pursuit of Unhappiness. 2. Haybron, Pursuit, 182, italics in original. 3. Haybron discusses well-being in chapters 8–9. These chapters were previously published, respectively, as Haybron, ‘Well-Being and Virtue’ and Haybron, ‘Happiness, the Self, and Human Flourishing’. 4. Haybron, Pursuit, 179–182. 5. Haybron, Pursuit, 193, 156. 6. Haybron, Pursuit, 193. 7. Haybron, Pursuit, 157. 8. Haybron, Pursuit, 156.

Well-Being and Eudaimonia


9. Haybron, Pursuit, 156. 10. Haybron, Pursuit, 155. 11. As Haybron notes (Pursuit, 156, 158), surely what really matters is whether some account is a viable theory of well-being, not whether or not it is Aristotle’s. 12. Haybron, Pursuit, 164. 13. Haybron, Pursuit, 164. 14. This problem is also noted by Sumner, Welfare, Happiness, and Ethics, 212. For discussion, see LeBar, ‘Good for You’. 15. See Michael Ridge, ‘Reasons for Action: Agent-neutral vs. Agent-relative’. 16. Haybron, Pursuit, 169. 17. Haybron, Pursuit, 161–163. 18. Haybron, Pursuit, 168–170. 19. Haybron, Pursuit, 168. 20. Haybron, Pursuit, 173. 21. For clear examples of this sort of argument, think about the central arguments of Plato’s Republic or Gorgias. 22. See also Rosalind Hursthouse, Beginning Lives, 222, who also observes that in this context the ‘good life’ is not to be thought of as the ‘good moral life’. 23. Haybron seems to accept the idea that we are beings for whom deliberative goal-setting and -achieving are essential (Pursuit, 172). 24. Haybron (Pursuit, 163) claims that ‘there is no credible sense, non-moral or otherwise, in which Angela, or her activities, would exhibit more excellence on the whole if she retired’. This is too hasty: on the Aristotelian view, Angela’s decision to retire could exhibit precisely the excellence in deliberation that counts as practical wisdom. 25. Haybron, Pursuit, 156–157. This is not to say, please note, that the question is whether Angela thinks that something would or would not make her happy as the individual she is. It is possible, after all, to be quite mistaken about such things. Haybron, Pursuit, chap. 9, does an excellent job of separating these two issues. 26. Haybron suggests a similar notion in his discussion of Elsie’s option to become a cellist (Pursuit, 185). We thank him for pressing us to think more about the resources available to the internalist here. 27. Of course, the development of virtue has opportunity costs as well (as noted earlier: it forecloses the options of becoming a burglar or a tyrant). The point is that the Aristotelian conception of well-being on offer here provides the normative resources to explain why that is a tradeoff worth making, while internalist conceptions cannot. 28. Haybron, Pursuit, 193. 29. ‘Humanity’ here might be too narrow; the real issue is one’s ‘kindedness’, so to speak, not necessarily one’s specifically human kindedness. Even so, we shall generally speak of human kindedness for the sake of simplicity, which will be harmless provided that this point is kept in mind. 30. Haybron, Pursuit, 185, italics in original. 31. Haybron, Pursuit, 186. 32. Haybron, Pursuit, 190. 33. Haybron, Pursuit, 184. 34. Haybron, Pursuit, 186. 35. Such a stipulation might be able to cleave to our intuitions, but it cannot explain them—and just cleaving to them is not enough. See LeBar, ‘Good for You’, 200. 36. This is the idea behind Aristotle’s thesis that pleasure, which for Aristotle is a genus of certain human emotions (Rhetoric II.1), is what ‘completes’ activity (NE X.4).


Mark LeBar and Daniel Russell 37. Cf. Richard Kraut, ‘Two Conceptions of Happiness’, 167–197, and Rosalind Hursthouse, On Virtue Ethics, 174–177. 38. For the same reason, Aristotle rejects the suggestion that we could wish for our friends that they were gods (NE VIII.7, 1159a3–8); that amounts to a wish that they were beings of a very different kind than human. 39. Haybron, Pursuit, 160. 40. NE I.7, 1097b22–1098a20. 41. We would like to thank Dan Haybron and Jason Raibley for comments on an earlier draft of this chapter. We also thank Julia Peters for her excellent and patient editorial assistance.

REFERENCES Haybron, Daniel, ‘Happiness, the Self, and Human Flourishing’, Utilitas 20 (2008), 21–49. ———, The Pursuit of Unhappiness (New York: Oxford University Press, 2008). ———, ‘Well-Being and Virtue’, Journal of Ethics and Social Philosophy 2 (2007), 1–27. Hursthouse, Rosalind, Beginning Lives (New York: Blackwell, 1987). ———, On Virtue Ethics (New York: Oxford University Press, 1999). Kraut, Richard, ‘Two Conceptions of Happiness’, The Philosophical Review 88 (1979), 167–197. LeBar, Mark, ‘Good for You’, Pacific Philosophical Quarterly 85 (2004), 195–217. Ridge, Michael, ‘Reasons for Action: Agent-neutral vs. Agent-relative’, in Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta (Fall 2008 Edition), http:// Sumner, L. W., Welfare, Happiness, and Ethics (New York: Oxford University Press, 1996).


Virtue, Personal Good, and the Silencing of Reasons Julia Peters

I One of the most intriguing aspects of Aristotelian and neo-Aristotelian virtue ethics is that it promises to show that there is a link between the virtuous life and happiness or eudaimonia, or that virtue benefits its possessor. However, it is surprisingly difficult to find, both in Aristotle and in the neoAristotelians, a positive account of what exactly the benefit of virtue consists in. Neo-Aristotelians tend to be much more preoccupied with addressing and if possible refuting the corresponding negative claim: that the possession and exercise of virtue in the long run tends to conflict with or threaten the possessor’s personal well-being. The reason why this negative claim may sound initially plausible—and therefore in urgent need of being addressed and if possible refuted—is that it appears to be a commonplace that the strict adherence to virtue can require personal sacrifices.1 Aristotle’s very first example of a virtuous action in the Nicomachean Ethics (NE), for instance, is the example of the courageous soldier who is about to sacrifice his life in order to defend his city against its enemies.2 As Aristotle presents the case, the soldier finds great satisfaction and even elation in performing the noble deed of defending his country, but the question is how this satisfaction stands in relation to the harm or loss suffered for the sake of the virtuous action: can it fully compensate for the latter? And it is easy enough to come up with further examples in which sticking with virtue requires the virtuous person to make a sacrifice: a single mother, for instance, who has to choose between either pursuing her professional career or caring for her sick child; or a person who has the chance to boost his career by betraying his friend. For the moment, we can therefore note that it appears undeniable that the exercise of virtue can require the virtuous agent to sacrifice personal goods. The question arises, then, how this observation can be reconciled with the idea that the possession and exercise of virtue is in the best interest of the virtuous person.3 One might reply at this point that this challenge is not a very serious one. The mere fact that virtue may conflict with personal well-being is not enough to seriously undermine the claim that the virtuous benefit from the possession and exercise of virtue. For after all, the cases in which such


Julia Peters

conflicts occur may be extremely rare, and it may still be true that in general, in most cases, the virtues benefit their possessor. This seems to be the line of reply taken by Rosalind Hursthouse. According to Hursthouse, virtue is very likely to benefit its possessor, likely enough, in fact, to make it a safe bet for us to decide to lead a life of virtue: such a life is most likely to be in our best interest. Thus Hursthouse writes, ‘We think that (for the most part, by and large), if we act well, things go well for us. When it does not, when eudaimonia is impossible to achieve or maintain, that’s not ‘what we should have expected’ but tragically bad luck’.4 It seems to me, however, that this line of reply somewhat misses the force of the challenge. Genuine, full virtue is a disposition that implies a certain ‘unconditionality’ in one’s adherence to certain values or normative commitments. As Philippa Foot points out, using a rather drastic example, ‘It is perfectly true that if a man is just it follows that he will be prepared, in the event of very evil circumstances, even to face death rather than to act unjustly—for instance, in getting an innocent man convicted of a crime of which he has been accused. For him it turns out that his justice brings disaster on him, and yet like anyone else he has had good reason to be a just and not an unjust man. He could not have it both ways and while possessing the virtue of justice hold himself ready to be unjust should any great advantage accrue. The man who has the virtue of justice is not ready to do certain things, and if he is too easily tempted we shall say that he was ready after all’.5 A genuinely virtuous person is someone who will not easily flinch or shrink back when her adherence to a certain value—justice, say—requires her to risk or sacrifice her own good.6 But if this is true, the question arises why the possession of a disposition that is less demanding and less unconditional than genuine virtue should not always be more beneficial for its possessor than genuine virtue. Such a less demanding disposition would allow its possessor to abstain from virtuous action in cases where such action would require a significant personal sacrifice—as, for instance, in the case sketched by Foot, where someone has to face death in the name of justice. In short, it is not clear why, even if virtuous action is more often beneficial for the virtuous than not, one would not be better off by aspiring to less than virtue proper in order to save oneself from harm precisely in those (allegedly rare) cases where virtue and personal wellbeing diverge. A proper reply to the challenge sketched previously would therefore somehow have to argue that even where virtue is ‘demanding’, by requiring personal sacrifices, this does not undermine its benefit—for being potentially demanding in this way appears to belong to the essence of virtue.

II In a series of papers, John McDowell has presented an influential argument that implies a response to this challenge.7 In essence, McDowell’s argument states that the virtuous never consider it as a sacrifice when they forgo or

Virtue, Personal Good, and the Silencing of Reasons


miss a personal good in acting virtuously. For them, virtuous sacrifices are not genuine sacrifices, and losses virtuously endured are no genuine losses— somehow, the virtuous are safe from incurring suffering through the practice of virtue. In the following, I shall look at McDowell’s argument in more detail and consider to what extent it is successful in countering the challenge just outlined. At the core of McDowell’s argument lies the following moral-psychological thesis: (P1) A virtuous individual’s perception of requirements of virtue silences all opposing considerations, in particular, all opposing prudential reasons (i.e., reasons arising from concern for one’s own personal good). The notion of silencing is to be understood, more specifically, in contrast to the notion of outweighing or overriding.8 The courageous soldier, for instance, who decides to risk his life in a battle in defense of his city does not reach this decision by weighing countervailing reasons against each other: the reasons for dropping the weapon and running away on the one hand, and the reasons for staying on and continuing to fight on the other. Rather, if he is truly courageous, the moment he sees that dropping his weapon and running away would mean for him to abandon his city and fellow-soldiers, all his reasons for doing so have been silenced—they cease to be reasons. The courageous soldier no longer considers himself as having any reason whatsoever for running away. A second premise in McDowell’s argument is a claim about the relation between reasons and goods, or reasons and potential losses: (P2) If one misses something that one has no reason to pursue, then that is no loss. From this, McDowell draws the conclusion: (C) If a virtuous person misses (forgoes) a personal good for virtuous reasons, that is no loss for her. The virtuous person’s reasons for pursuing the personal good have been silenced by virtuous considerations. Accordingly, she has no reason to pursue the good in question. In missing something she has no reason to pursue, she suffers no loss. It follows that she suffers no loss in forgoing the personal good in question.9 Before looking at the premises of McDowell’s argument in more detail, we need to consider how exactly the argument can be directed against the challenge sketched previously. The core claim of the challenge is that the practice of virtue may demand sacrifices from the virtuous person, such


Julia Peters

that virtuous action may deprive the virtuous person of personal goods. In this way, the practice of virtue may undermine the virtuous person’s happiness. In contrast, McDowell’s argument attempts to show that where it may appear that a virtuous person endures a loss when forgoing a personal good for the sake of virtue, this is really not so: her reasons for pursuing the good have been silenced, that is to say, she has no reason to pursue it; when she misses something she has no reason to pursue, that constitutes no loss, no personal sacrifice for her. Forgoing personal goods for the sake of virtue is therefore not really painful for the virtuous person, or does not undermine her happiness. What, then, can be said about the premises of the argument? I take it that one important consideration underlying P1, the claim that reasons against acting virtuously are silenced rather than outweighed from the virtuous person’s point of view, is the following one. Unless we assume that for the virtuous person, considerations opposing a requirement of virtue are silenced, rather than outweighed, it would in certain extreme cases be impossible to explain why people act virtuously at all. Consider again the courageous soldier on the battlefield. If he was to weigh personal gains against losses, how could he possibly come to decide that the loss ensuing from his action—the loss of his life—is outweighed by the gain he expects from it? Or, similarly, how could the just person hope to outweigh the loss of her life by acting justly? McDowell’s point is that as long as we assume that the virtuous person registers potential goods forgone for the sake of virtue on the loss-side of a balance, we will not be able to account for her decision to embark on the virtuous course of action in extreme cases. Virtuous action in extreme cases can only be explained if reasons against acting virtuously are not expected to be outweighed, but are simply silenced.10 On this view, then, considerations of virtue are not so much understood as giving rise to certain reasons that can then be weighed against other reasons, but rather as putting a constraint on what can count as a reason in the first place. A second, ancillary thought underlying P1 may be that the notion of silencing is needed in order to draw a distinction between two types of moral agent, one of which is more excellent—deserves greater moral praise—than the other: the virtuous and the merely continent agent. Both the virtuous and the continent moral agent end up doing the same thing, but the structure of the practical consideration preceding their choice of action is significantly different in each case. Consider again the soldier on the battlefield. One important mark of his virtue—his courage—is that he is not tempted by the possibility of saving his life by acting cowardly: running away and abandoning his fellow soldiers is simply not an option for him. Accordingly, he stays on to fight. But he might also have reached the decision to stay and fight in a different, less straightforward way: he might have weighed the reasons for and against staying to fight, and he might have been seriously tempted to run away and save his life—while nevertheless deciding that in the end, he has stronger reasons to stay. One might say that in this case, his decision to

Virtue, Personal Good, and the Silencing of Reasons


stay is preceded by a period of inner division: he is torn between different options. If this was the case, he would not be truly virtuous, but merely continent. While reaching precisely the same practical decision as the continent agent, the virtuous agent knows no such inner division, and this constitutes part of his moral excellence, over and above the fact that he also ends up doing the right thing. The notion of silencing captures precisely this difference in practical deliberation between the two types of moral agent: the continent agent takes (prudential) reasons that speak against acting virtuously into consideration and is torn between different courses of action; for the virtuous agent, in contrast, countervailing reasons are silenced.11 The thought underlying P2, the thesis that if one misses something one has no reason to pursue, one does not suffer a loss, is a claim about the relation between goods and reasons. This becomes more obvious when one focuses on the reverse of P2: As long as one suffers a loss when missing or forgoing something, one also has a reason to pursue it. In other words, as long as something constitutes a good, such that losing it has to be considered a genuine loss, one also has a reason to pursue it. Accordingly, where one has no reason to pursue something—where one’s reasons to pursue it have been altogether silenced—it no longer constitutes a good, its goodness must have been ‘cancelled’. For the virtuous person, then, the goodness of a supposed personal good is conditional on its being in accord with considerations of virtue. Something can be a genuine personal good for her only insofar as its pursuit does not flout considerations of virtue.12 In spite of the initial plausibility of its premises, however, the conclusion of this argument appears problematic. Consider again the case of the soldier. Even if McDowell is right in pointing out that the truly courageous soldier will not consider himself as having any reason for saving his life instead of staying on the battlefield to fight for his city, and even if the fact that he has no reason ‘cancels’ the goodness that his life has for him, somehow the intuition lingers that the soldier nevertheless suffers a loss or makes a sacrifice in giving his life for his city. This intuition can be articulated more accurately by the following consideration. For all his willingness to give his life in fighting for his city, the soldier would certainly prefer a world in which he could have both, save his life and fight for the city, to the actual one, where the two are incompatible. For it is not the case that the soldier does not in general value his life; rather, it is merely under the present circumstances, the actual world being as it is, that the value of preserving it has been cancelled for him. But it is not clear how McDowell’s argument can account for this intuition and explain the soldier’s hypothetical preference for a world in which he could both save his life and fight for his city. According to McDowell’s account, the soldier loses nothing when forgoing the opportunity to save himself from personal harm by running away. Hence there seems to be no reason why he should prefer a world in which he could have both—both escape personal harm and fight for his city—to the actual one. If we accept McDowell’s argument, then, it seems to imply that the


Julia Peters

virtuous person necessarily has to have a kind of fatalist attitude: whether the exercise of his virtue conflicts with his pursuit of personal goods or not makes no difference to him, for even if he has to forgo a personal good in the name of virtue, this constitutes no loss for him. McDowell’s argument presents us with a problem: while its premises—or the considerations underlying them—appear plausible, its conclusion nevertheless seems to violate a strong intuition we have regarding the losses potentially suffered by a virtuous person as she acts in the name of virtue. In the remainder of this chapter, I want to suggest a way to solve this problem, by showing how it is possible both to do justice to the intuition just sketched and to accept, in essence, McDowell’s argument. On the whole, my suggestion aims not so much at criticizing McDowell’s argument, but rather at showing how it can be reconciled with countervailing intuitions. My overall conclusion will be that McDowell’s argument is successful in offering a reply to the challenge sketched in the introduction: it is successful in establishing that the exercise of virtue never makes the virtuous person unhappy. Nevertheless, the crucial point I shall try to make is that it does not follow from this that a virtuous person does not suffer any losses.

III What I have in mind can be best introduced by pointing out that there are (at least) two different ways in which the pursuit of a personal good can conflict with a requirement of virtue. Consider an example involving the virtue of temperance: A temperate person is invited by her friends to go out binge drinking with them. As she is temperate, she of course declines the invitation. McDowell’s claim that the virtuous person has no reason whatsoever to pursue the personal good in question seems very plausible here. If the person in the example was seriously tempted by the opportunity to go binge drinking, and had to weigh reasons in favor of and against seizing this opportunity, we would be hesitant to call her temperate, even if in the end she abstained and declined the opportunity—at best, she could be considered continent in this case. For the truly temperate person, an opportunity to go binge drinking does not constitute a good whose loss has to be weighed against the reasons speaking in favor of declining the invitation. Accordingly, the temperate person has no reason to accept her friends’ invitation and seize the opportunity to go binge drinking. At first sight, this example may appear to be parallel in all important respects to the one of the courageous soldier. Like the temperate person, the courageous soldier has no reason to run away and save his life instead of fighting for his city. His life does not constitute a good for him, as it can only be preserved by acting cowardly. Similarly, for the temperate person, going out binge drinking does not constitute a good, as it involves behaving intemperately. However, in spite of their similarity, there is an important dif-

Virtue, Personal Good, and the Silencing of Reasons 75 ference between the two cases. To go binge drinking means to go drinking with the explicit intent of drinking excessively, that is, intemperately. Hence binge drinking is an inherently intemperate activity—there is no such thing as practicing binge drinking in a temperate way. Accordingly, the temperate person never has a reason to go out binge drinking, for doing so is always in conflict with considerations of virtue. It follows, according to P2, that an opportunity to go binge drinking does not constitute a good for the temperate person under any circumstances. In contrast, for the courageous soldier, his own life is usually of great value, and he usually has a reason to preserve it. It is just that in the situation he is in, as he can only preserve his life by acting cowardly, that his reasons for preserving it are silenced, such that his life no longer constitutes a good for him. Only under these specific circumstances, in which his reasons for preserving his life have been silenced, does it no longer constitute a good for him. I want to suggest that it is crucial to pay attention to the difference between these two examples, for this difference holds the key for reconciling McDowell’s argument with the intuitions mentioned previously about losses suffered by the virtuous person. It seems correct that with regard to the temperate person who forgoes an opportunity to go binge drinking, it does not make sense to speak of her suffering a loss or making a sacrifice. If she is temperate, then not going binge drinking is not a sacrifice for her. We saw previously that the reason why one would be inclined to hold that even the courageous soldier suffers a genuine loss in giving his life for his city is that he would certainly prefer circumstances in which he could preserve his life and fight for his city to the actual ones, in which he can only have the latter. But the same consideration does not apply in the example of the temperate person. Since there are no circumstances under which she can both go binge drinking and be temperate— because binge drinking is intemperate under any circumstances—it makes no sense to hold that she would prefer such circumstances. In this case, then, there is no lingering intuition that even though the virtuous person has no reason to pursue a certain good, she may nevertheless be suffering a loss. In contrast, in the case of the courageous soldier, the fact that his life does not constitute a good worth pursuing for him is due to the circumstances he finds himself in. But this makes it possible to distinguish between two different senses in which he can be said to suffer a loss. On the one hand, he can be said to suffer a loss in failing to pursue something that constitutes a good in the situation he is in. According to McDowell’s thesis P1, the courageous soldier does not suffer a loss in this sense because his life, which he fails to preserve, does not constitute a good for him in his situation. It is not his failing to pursue something that constitutes a good in his situation, then, that makes the soldier suffer a loss. But on the other hand, the soldier can be said to suffer a loss in being deprived by the circumstances of the goodness of something that would usually, under different circumstances, constitute a good for him. If the situation was different, if the circumstances


Julia Peters

were more fortunate, his life would constitute a great good for the soldier. As the circumstances are, however, his life has lost its goodness for him, or the goodness of his life has been cancelled. But this must certainly be considered as a loss by him: something that would usually be of great value has lost its value, or has been made unavailable as a good. His life has been deprived of goodness. More generally, where it depends on the circumstances whether the virtuous person’s reasons for pursuing a certain good are silenced or not, there is always room for her to suffer the kind of loss just described with regard to the case of the courageous soldier. If her reasons to pursue the good in question are silenced, then it constitutes no loss for her not to pursue the good. But insofar as the fact that the good in question is no longer a good for her is due to unfortunate circumstances, it does constitute a loss for her that the circumstances have deprived something that usually constitutes a good for her of its goodness, or that they have made unavailable as a good what is usually a good for her. One can also express this thought in terms of the regret the virtuous person may feel and the reasons she has for feeling it. Where the virtuous person’s reasons for pursuing a good are silenced, she has no reason to regret not pursuing the good because it does not constitute a good for her (or, in hindsight, she has no reason to regret not having pursued the good because it did not constitute a good for her). But if it is due to unlucky circumstances that her reasons for pursuing the good are silenced, she has reason to regret that the circumstances are as they are (or that they were as they were). Because this is what makes the good in question unavailable to her as a good, by depriving it of its goodness. Thus the virtuous person’s regret is not directed at what she herself does or did under certain circumstances, but at the circumstances themselves. She experiences her loss as being incurred, not by herself or her own action, but by the circumstances—or more grandiosely, by fate. This makes a crucial difference. Because it means that from the point of view of the virtuous person, it is not her virtuous action that constitutes the cause or ground of her loss, but the circumstances. It is not her virtuous action that undermines her happiness or well-being, in short, but fate.13

IV We can now come back to the overall theme of the relation between virtue and eudaimonia. As sketched in the beginning, the thesis that the possession and practice of virtue is in the best interest of the virtuous looks dubious in light of the apparently undeniable fact that the virtuous may have to sacrifice personal goods in acting virtuously. I presented McDowell’s argument as implying a response to this challenge. McDowell argues that the virtuous person, when missing or forgoing a personal good in acting virtuously, does

Virtue, Personal Good, and the Silencing of Reasons


not make a genuine sacrifice and does not suffer a genuine loss. For what she misses or forgoes does not constitute a genuine good for her. However, the conclusion of McDowell’s argument contradicts the strong intuition that there is a sense in which the virtuous person, after all, does suffer a loss when she gives up a personal good in acting virtuously. Otherwise, we could not make sense of the fact that whenever possible, the virtuous person would prefer to be able to both act virtuously and pursue the personal good in question. In light of the previous discussion, we can now see how this tension can be dissolved. Giving up a personal good in acting virtuously, the virtuous person does not suffer a loss in the sense of making a sacrifice, that is, of depriving herself of a personal good that she could have secured for herself in the situation she is in, or that would have been available as a good in her situation. However, she does suffer a loss in the sense of being deprived of a good by the circumstances: something that usually constitutes a good for her has been deprived of its goodness or has been made unavailable as a good. The fact that she has been deprived of a good by the circumstances gives her a reason for regret. Accordingly, while she would not prefer acting or having acted in a different way than she acts or has acted—namely, non-virtuously—she would prefer a world in which it was possible for her to both act virtuously and pursue or secure the personal good in question. Hence she is precisely not a ‘fatalist’: while she does not regret acting or having acted virtuously, she is susceptible to regret directed at the circumstances in which she finds herself. McDowell’s argument thus implies a subtle way of refuting the challenge discussed in the beginning. The argument shows that for the virtuous person, the practice of virtue is never the ground or cause of her unhappiness or loss; acting virtuously is never what deprives her of a personal good, or what makes her suffer a loss. In this sense, it is true that she does not suffer any losses by or through acting virtuously. However, it does not follow that the virtuous person suffers no loss at all as long as she acts virtuously.14 If the pursuit of some personal good conflicts with a requirement of virtue in a certain situation, and the virtuous person forgoes the personal good in that situation, she does not sacrifice or give up anything that constitutes a good in the situation she is in. But she is nevertheless being deprived of a good, and in this sense she suffers a loss. It is important to bear in mind that the stronger thesis that the virtuous person literally suffers no loss at all cannot be established by means of the argument alone discussed previously. For McDowell is unfortunately often read as intending to argue for this stronger conclusion. According to Rosalind Hursthouse, for instance, McDowell holds that for the virtuous person, no personal good missed or forgone in meeting a requirement of virtue counts as a genuine loss because her own virtuous action is all that is of value to the virtuous person, or at least its value is so much more significant to her than that of any personal good that no ‘sacrifice’ of such a good


Julia Peters

for the sake of acting virtuously really counts as a sacrifice.15 Hursthouse argues against this position, pointing out that it only looks plausible—if at all—in a very limited number of cases. These are cases in which the virtuous person sustains the ‘loss’ of a personal good as she performs a noble, admirable action—for instance, sacrificing her life for a worthy cause. Here one might think that if only the cause is great and worthy enough, any personal losses endured in furthering it may appear insignificant. By contrast, Hursthouse argues, McDowell’s thesis looks less plausible with regard to cases in which the virtuous action consists merely in the avoidance of something base, rather than in the performance of a noble deed: for instance, ‘dying not to serve some noble cause but only because you have fallen into the hands of a mad tyrant and, despite his threats, refused to do something wicked’.16 Hursthouse’s thought seems to be that where losses are endured in the accomplishment of something noble or great, we can somehow make sense of the notion that from the point of view of the virtuous agent, the losses are compensated for by the good accomplished through virtuous action. But this does not seem to be possible in the case where losses are endured merely for the avoidance of something bad: for here nothing noble or great is accomplished that could outweigh the losses. However, this is an unfair reading of McDowell’s position. As I argued previously, one of the thoughts underlying McDowell’s thesis about the silencing of reasons, P1, is that if we took the virtuous person to weigh reasons for and against acting virtuously in cases where virtue and personal good conflict, this would sometimes make it impossible to explain why she ends up acting virtuously. On this view, the virtuous person would concede that she is suffering a personal loss in acting virtuously, but hold that this loss is compensated for by the good attained through virtuous action. ‘In suitably described cases’, McDowell writes, ‘any such claim would be implausible to the point of being fantastic’.17 Among such cases, presumably, are those in which nothing great or noble is to be attained through virtuous action but merely something base to be avoided. To escape such implausibility, McDowell suggests, we should instead assume that the personal goods in question cease to be personal goods for the virtuous person when they are seen to conflict with virtue, such that she does not consider herself as having any reason to pursue them. From McDowell’s point of view, then, it should make no difference to the virtuous person whether she forgoes a personal good in order to virtuously accomplish something noble, or in order to virtuously avoid something base. In both cases, the good in question simply ceases to be a good since its pursuit is in conflict with considerations of virtue, hence it is no question for her how its loss may be compensated for. Nevertheless, there is an important intuition underlying Hursthouse’s complaint. In a case such as the one involving the mad tyrant, we are inclined to insist that there is some loss that is endured by the virtuous person as she forgoes the personal good in question, or that she has a reason for regret.

Virtue, Personal Good, and the Silencing of Reasons


After all, would she not prefer circumstances to have been different—would she not prefer not to have fallen into the tyrant’s hands—such that she could have saved her own life and acted virtuously? In contrast, one might think that there may be cases in which a virtuous person even welcomes an opportunity to sustain a personal loss in acting for a great, noble cause. However, it is possible to do justice to this intuition in the way suggested previously. If a mad tyrant confronts the virtuous person with a choice between either dying or doing something wicked, the virtuous person will choose to die. Her life, saved through a wicked action, would have no value for her, hence she suffers no loss in not preserving it by acting wickedly. But she nevertheless suffers a loss, and has a reason for regret, since she is being deprived, by the circumstances, of some good: the circumstances are such that her own life has lost its goodness for her. For her, her loss is not induced by her virtuous action, but rather brought about by the circumstances.18 The strong thesis that, for the virtuous person, conflicts between virtue and personal good result in no personal loss at all could only be established on the basis of an assumption to the effect that the only thing that is of value for the virtuous person is her own virtuous or non-virtuous action. If this was the case, it would be true that the virtuous person was safe from suffering any losses at all as long as she only acted virtuously: the only thing that could potentially constitute a loss for her was her failure to act virtuously. But this is not McDowell’s view, at least not according to the argument discussed previously. This argument establishes a more modest claim: The virtuous person does care about her own personal good, in addition to her virtue or virtuous action. But from her point of view, the loss of a personal good is never suffered as a result of or due to her virtuous action. Rather, she experiences such losses as imposed on her by the circumstances. Hence it is the way in which they experience such losses, rather than the fact that they do not experience them at all, that is distinctive of the virtuous.19

NOTES 1. This is true at least as far as the traditional ethical virtues—courage, temperance, justice, generosity, perhaps also charity—are concerned. With regard to the intellectual virtues, the possibility of a conflict with personal well-being is not as obvious. 2. Aristotle, Nicomachean Ethics (NE), 1117a30–1117b20. 3. It is important to note the precise nature of this challenge: the point is not just that even the virtuous person’s life may fall short of being a good life, or that even the virtuous person may wind up unhappy. This could be readily conceded both by Aristotle and by most neo-Aristotelians, for they agree that the possession (and exercise) of virtue is not sufficient for the good life: external goods such as health, friends, or wealth are needed as well. Rather, the challenge is that the exercise of virtue itself may limit a person’s well-being, insofar as it requires her to make personal sacrifices, or in short, that virtue and its exercise may contribute to someone’s unhappiness.


Julia Peters 4. Hursthouse, On Virtue Ethics, 185. 5. Foot, ‘Moral Beliefs’, 130. 6. It is nevertheless not quite correct to state that a virtuous person’s adherence to certain values is unconditional. The reason is related to the unity or reciprocity of the individual virtues. If it is true, as many Aristotelian virtue theorists argue, that a person, in order to possess one virtue, needs to possess at least some others as well, then her adherence to one virtue—justice, say—will be conditional on whether this adherence does not conflict with her adherence to the other virtues she has. Thus conflicts with other virtues, rather than conflicts with her own personal good, may make it occasionally necessary for the virtuous person to abstain from performing an act in accord with a particular virtue. The issue of whether and to what extent the virtues have to be thought of as unified or reciprocal is a controversial and much-debated one. For a good discussion of the problem and overview of the current debate, see Halbig, ‘Die Einheit der Tugenden. Überlegungen zur Struktur eines Problems’. 7. See McDowell, ‘The Role of Eudaimonia in Aristotle’s Ethics’, 359–376; McDowell, ‘Eudaimonism and Realism in Aristotle’s Ethics’; McDowell, ‘Virtue and Reason’, 141–162. 8. See in particular McDowell, ‘The Role of Eudaimonia’, 369–370, and ‘Virtue and Reason’, 146. 9. McDowell nowhere presents his argument explicitly in this form; rather, the version given previously is a reconstruction of the argument as it occurs in different versions in the papers listed in footnote 7. 10. McDowell, ‘The Role of Eudaimonia’, 369. 11. See McDowell, ‘Virtue and Reason’, 146. 12. Michael Slote, in Goods and Virtues, challenges this thesis about the necessary interrelation between goods and reasons, while wishing to maintain McDowell’s claim about the silencing of reasons. Slote’s overall aim is to escape the conclusion of McDowell’s argument—that the virtuous suffer no loss when virtuously forgoing personal goods—while holding on to what he takes to be the important moral-psychological insight captured by P1. Slote’s main argument against P2 relies on a counterfactual consideration. He asks: If, per impossibile, a virtuous person was to flout a requirement of virtue in order to gain what looks to be a personal good, what would he do with the ‘good’ once gained—in contrast to a non-virtuous person who gains the good in the same way? His answer is that the virtuous person would somehow refuse to make use of the good, or to refuse to gain any advantage from having gained it contrary to virtue. For instance, if a virtuous, honest person was to act out of character on one occasion and steal money, she would not be capable of gaining any advantage from the money, once it was in her possession. Instead of using it for herself, she would give it away to charity. But such refusal is best understood, Slote argues, as a refusal to make use of an advantage that the virtuous person, therefore, does in fact consider as a genuine advantage (Goods and Virtues, 115). Consequently, even the virtuous, if pressed, would have to concede that such advantages constitute genuine goods. However, Slote’s argument is not fully convincing. Slote may be right that in the counterfactual case he describes, we would not expect a virtuous person simply to go ahead and reap the benefits she has acquired for herself contrary to virtue. But rather than thinking of her, in the counterfactual situation, as refusing to make use of a good, we might just as well think of the virtuous person as finding herself without any good that she could make use of. If the stolen money, or anything she could buy with it, simply constitutes no good for the virtuous person, she will also give it away to charity, rather

Virtue, Personal Good, and the Silencing of Reasons


14. 15. 16. 17. 18.



than using it for herself. Slote’s description of the counterfactual situation does not, therefore, by itself undermine McDowell’s conception of the essential connection between goods and reasons. In the following, I shall accept both P1 and P2 of McDowell’s argument as reconstructed previously. Philippa Foot probably has something similar in mind when she writes that the virtuous person ‘in sacrificing his life for the sake of justice would not have said that he was sacrificing his happiness, but rather that a happy life had turned out not to be possible for him’ (Foot, Natural Goodness, 97). McDowell himself, misleadingly, uses the formulation ‘no loss at all’ in ‘The Role of Eudaimonia’, 370. Hursthouse, On Virtue Ethics, 180–183. Hursthouse, On Virtue Ethics, 183. McDowell, ‘The Role of Eudaimonia’, 369. In chapter 3 of this volume, Christoph Halbig likewise criticizes McDowell for maintaining that in forgoing personal goods by acting virtuously, the virtuous person suffers ‘no loss at all’. He writes: ‘In thinking about what to do it might well be that possible losses for herself [for the virtuous person, JP] are indeed silenced—but that just means that they are not admitted to the deliberative process, in the sense of providing reasons that have to be weighed against other kinds of reasons. It does not mean that the fact that such a loss is undergone itself loses its prudential disvalue; what it does lose is just its normative impact on the deliberating agent’ (Halbig, ‘The Benefit of Virtue’), this volume, 42. However, for McDowell, the silencing of reasons to avoid something of prudential disvalue implies that avoiding it has no positive prudential value: where there are no reasons to pursue something, there is no good to be pursued. But again, we can grant McDowell this claim about the interrelation between goods and reasons without having to conclude, implausibly, that the virtuous person does not see herself confronted with any prudential disvalue in cases where virtue conflicts with the pursuit of personal good. What is crucial is that for her, such disvalue is not incurred through her virtuous action. Christine Swanton makes a similar criticism: see Swanton, Virtue Ethics: A Pluralistic View, 89. Like Halbig, Swanton wishes to question McDowell’s thesis that the absence of reasons to pursue a good in a situation implies that it has ceased to constitute a genuine good in that situation. In contrast, my aim was to argue that it is possible to hold on to both P1 and P2 and to do justice to the intuition that there is a sense in which the virtuous person suffers a loss in cases in which virtue and personal good conflict. I would like to thank the participants of the workshop ‘Agency, Reasons and the Good’ at Humboldt University, Berlin, in July 2011 for helpful comments on an earlier draft of this chapter, in particular Elif Özmen, Michael Smith, and Jay Wallace.

REFERENCES Aristotle, Nicomachean Ethics, transl. H. Rackham (Cambridge, MA: Harvard University Press, 2003). Foot, Philippa, ‘Moral Beliefs’, in Virtues and Vices (Oxford: Clarendon Press, 2002), 110–131. ———, Natural Goodness (Oxford: Oxford University Press, 2001). Halbig, Christoph, ‘Die Einheit der Tugenden. Überlegungen zur Struktur eines Problems’, in Foundations of Ancient Ethics, Grundlagen der Antiken Ethik, ed. Jörg Hardy and George Rudebusch (Göttingen: V & R unipress, 2011).


Julia Peters

Hursthouse, Rosalind, On Virtue Ethics (Oxford: Oxford University Press, 2001). McDowell, John, ‘Eudaimonism and Realism in Aristotle’s Ethics’, in Aristotle and Moral Realism, ed. R. Heinaman (Boulder, CO: Westview Press, 1995), 201–208. ———, ‘The Role of Eudaimonia in Aristotle’s Ethics’, in Essays on Aristotle’s Ethics, ed. A. Oksenberg-Rorty (Berkeley: University of California Press, 1980), 359–376. ———, ‘Virtue and Reason’, in Virtue Ethics, ed. R. Crisp and M. Slote (Oxford: Oxford University Press, 2007), 141–162. Slote, Michael, Goods and Virtues (Oxford: Oxford University Press, 1990). Swanton, Christine, Virtue Ethics: A Pluralistic View (Oxford: Oxford University Press, 2003).


Human Nature, Virtue, and Rationality John Hacker-Wright

Contemporary Aristotelian virtue ethicists appeal to human nature to tell us why character traits such as courage, honesty, benevolence, and justice should be regarded as virtues. Virtues make us good human beings, and yet what does it mean to say that a character trait is necessary to be a good human being? Philippa Foot and Rosalind Hursthouse argue that it means the same as when we say, for example, that a keen sense of smell is a part of what is necessary for a tiger to be a good tiger; it means that this quality, among others, makes the tiger a non-defective instance of its kind, able to carry out its characteristic predatory life.1 On the other hand, if the tiger lacks this sense, it would be a defective tiger, inasmuch as it would lack an important characteristic for carrying out a tiger’s life. Similarly, should a female tiger be indifferent to its cubs, it would also be defective. Though this defect may not impede the mother’s own well-being, the tiger falls short of a norm for tigers when caring for their cubs.2 Yet not just any discrepancy from a supposed normal life form constitutes a defect; as Foot points out, a blue tit lacking the patch of blue on its head is abnormal but not defective since the patch of color does not appear to play any essential role in the characteristic life of that species.3 To make a judgment that an organism is a defective instance of its kind, then, we must grasp how its parts and behaviors contribute to leading a life of the sort that is characteristic of its life form. Foot and Hursthouse argue that such a structure can be applied to human beings. A virtue enables one to live a life that is characteristic of human beings, and lacking virtue constitutes a defect that impairs one from living such a life. Of course, one immediately wonders whether these claims could possibly be true; after all, humans are capable of living a wide range of different lives, and in light of such diversity, one wonders whether there is any characteristically human life for which virtue is essential. Despite this diversity, Foot argues that humans are vulnerable to deprivations that parallel natural defects found in plants and animals. For example, humans need the mental capacity for learning language, understanding stories, joining in songs, and laughing at jokes.4 Yet humans are rational animals, and this introduces a ‘sea change’, as Foot puts it, in our description of humans as


John Hacker-Wright

animals; reasoning and the application of reasoning to action are features that are evaluated in human beings. As Foot puts it, ‘[W]hile [non-human] animals go for the good (thing) that they see, human beings go for what they see as good’.5 We are capable of responding to reasons in a distinctively explicit way, inasmuch as we act on some understanding of which things are good. In fact, it is precisely the application of reasoning to our action that interests us in ethical evaluation. Foot’s claim is that vices are defects in our responsiveness to reasons for action, a sort of natural defect in humans; specifically, individuals with vices have defective wills. Inasmuch as we have the capacity to reason about how to act, we are subject to a distinctive sort of evaluation; unlike other natural defects, which may be the result of bad luck, we are responsible for our conception of how to act and can answer to rational criticism of that conception. Thus far this view looks like an attempt to build the foundations of ethics on the very general characteristic of rationality, leaving aside other aspects of our nature. If so, the view will surely not get us very far in justifying the virtues, for such a view seems to lead us to the vague standard of acting in accordance with ‘right reason’. Nevertheless, such a path is pursued explicitly by Hursthouse when she writes, ‘Our characteristic way of going on, which distinguishes us from all the other species of animals, is a rational way. A rational way is any way we can rightly see as good, as something we have a reason to do’.6 In this manner she appears to reject the idea that nature is normative for us in the way that it is for non-human animals, inasmuch as humans do not adhere to a single manner of living. Yet she also explicitly holds that there are certain ends in some sense prescribed by nature that substantively constrain what we can rightly do. Specifically, she believes there are four ends that any character trait must promote in order to count as a virtue: individual survival, species survival, freedom from pain and experience of species-characteristic pleasures, and the good functioning of the social group. Appealing to these four ends, she argues that ethical naturalism provides grounds for rejecting a virtue of impersonal benevolence, which would tell us to maximize happiness across all sentient beings. This is because impersonal benevolence would not allow us to carry on our species or to have good functioning social groups. Yet this raises the question: why should we have to meet those ends? David Copp and David Sobel complain: How can Hursthouse reject the thought that nature determines how humans should be yet think that the same considerations that grounded the four ends in plants and animals also ground the normative status of the four ends in humans? She gives no new arguments to support such a status for the four ends in the case of humans.7 Copp and Sobel see Hursthouse’s argument as incomplete at best, and they are skeptical of anyone’s ability to complete it. Indeed, there is a missing component to this account of ethical naturalism, but it can be filled in, and was,

Human Nature, Virtue, and Rationality 85 at least in part, by Philippa Foot. The crucial missing piece is that the appeal to human nature serves not only to tell us that we are rational, but also to define what it is to reason well. Neo-Aristotelian ethical naturalism should be understood as a thesis about rationality, according to which practical rationality is species-relative. Our reasoning cannot ignore what we need as human beings and yet still claim to exhibit practical rationality. Hence, Hursthouse can claim that as rational animals we are freed from a certain kind of obedience to nature, while maintaining that nature has some normative role for us; nature is normative over our reasoning, but not directly over our action. When Foot states that human beings go for what we see as good rather than the good that we see, she adds that what we see as good is inevitably informed by a conception of our form of life. Making that conception explicit and subjecting it to criticism is an essential part of moral reform, for ethical naturalists. When Hursthouse’s claims about impersonal benevolence are placed against the background of a proper understanding of ethical naturalism, a version of her argument against impersonal benevolence can go through, or so I will argue here. I will first argue for a way of conceiving human nature that is crucial to the interpretation of ethical naturalism defended here.

A PRACTICAL CONCEPTION OF HUMAN NATURE The idea of human nature may enter our practical reasoning explicitly in the form of a major premise, such as ‘It befits a human to overcome fear for the sake of worthy goals’, or something as general as ‘It befits a human to attempt to further the survival of the human species’. Yet one might wonder what recommends such premises as starting points for practical reasoning. After all, if my identity is to enter into my reasoning at all, perhaps I would rather appeal to what befits a Buddhist or an American since these identities might mean more to me than my identity as human, which seems, after all, to be rather abstract and thin. Neo-Aristotelian ethical naturalists are not assuming that our reasoning must or should appeal to being human in this way. Instead, the view is that there is a sense in which we inevitably make an implicit appeal to some understanding of human nature inasmuch as we reflect and act. This understanding of human nature may be informed by scientific ideas that we have about human beings, but in principle it is separate from these ideas and instead shared with others who diverge from us in their scientific views on anthropology. It is a practical conception of human beings that we must employ in order to understand ourselves or anyone else as engaged in an action. Further, this conception of human nature supplies standards of evaluation or natural norms, including norms of conduct. Actions are done by living things, and so understanding a human action, on the naturalist’s view, is a matter of understanding a particular life form. As Michael Thompson argues, all living things exhibit, in various ways, a


John Hacker-Wright

special kind of agency.8 The growth of a fern is essentially different from the growth of a puddle of rainwater or a trash heap because it is the fern itself that brings about its own growth by a process of cell division. We can provide a description of how such a process is supposed to advance and recognize when it has gone awry, whereas there is no question of the growth of a trash heap going wrong. Further, the growth of a fern differs from that of a rhododendron, even though they both exhibit the same sort of agency that is characteristic of living things. We must, in principle, be able to discern the growth of the fern from a growth on the fern that is due to blight, for example, and such discernment requires a conception of how life is supposed to go for the fern. To characterize something as possessing the agency distinctive of living things requires bringing to light norms for the life form of that individual; this is no less true in our own case. Yet the case of human beings differs in that we obviously stand in a different relationship to our own life form. With other life forms our views are based entirely on observation. In the case of the human life form, we must apply an understanding of it whenever we describe ourselves. Hence, our understanding of the human life form is, in part, internal and not based on observation; I know something about the category ‘human being’ from my own case because it is in the background of the fact that I think and act. In this way, the category ‘human being’ is not merely used to gain an understanding of part of the world, but rather it is a practical concept. Consider a simple case of action: sawing a plank of wood. There are facts I must know and practical competencies I need in order to intentionally undertake this action. Among the facts I need to know are that by performing such and such actions, I will, if everything goes well, be sawing a plank of wood. Many such facts I have gathered by watching others perform similar actions, as well as through trial and error. But I must possess even more basic knowledge inasmuch as I have seen others perform such actions and am able to know that I am doing the same. To understand an intentional act as such is to understand a living thing as being engaged in a specific form of agency that involves responsiveness to reasons. This is not something that could simply appear in a ‘rogue individual’, as Thompson puts it; the capacity to respond to reasons to act must characterize my life form.9 To see this, consider a simpler case like that of eating. For an organism to be regarded as eating something, one must take it to be ingesting that which is normally nutritious for its kind as well as absorbing it in such a way as to derive nutrition from it. Otherwise, we could not say that we saw an organism eating something, but instead we would be watching a fortuitous occurrence whereby an organism took in some sort of material that happened to further its life. Likewise, to see an organism as engaged in an intentional action, we must posit a non-accidental concurrence of a specific kind of mental event that consists of forming an aim followed by an initiation of movement such that, if all goes well, I will bring something about related to fulfilling that aim. Further, to intentionally undertake an action, I must realize that I am

Human Nature, Virtue, and Rationality


doing so. As Anscombe argues, if I am asked why I am sawing a plank of wood, and my reply is that I did not realize that I was doing so, say because I was instead simply enjoying the rhythm of the sound produced when the saw is placed against the wood, then I was not sawing intentionally.10 It must be a norm for that life form to engage in such actions, otherwise, there will be a mere coincidence of mental events and bodily jerks. Action implies a sort of insight into human nature inasmuch as in acting I implicitly posit that it is normal to initiate movements in order to undertake the fulfillment of an adopted aim; in other words, in undertaking any action, we posit that we are members of a life form that normally possesses rational agency. This bit of knowledge concerning our life form is discerned from undertaking an action, rather than from observation. Indeed, we could never observe an action without first possessing this knowledge about our life form. For we must see someone as setting out to do something knowing what they are about, and I cannot learn about this capacity from observation. Rather, I must learn it from acting intentionally myself and take others to be doing the same. Hence, to see other human beings as acting requires us to frame them against the background of the same practical conception of the category human to which anyone adverts whenever he acts. There is some normative content to this aspect of our self-understanding since problems such as acting on false premises, weakness of will, and slips are defects of actions, and we must all regard them as such if we are rational agents. Presumably, if we take no steps to avoid defective actions, we are defective agents. This brings us to a conception of human nature that we all apply, at least implicitly, in our actions. But of course we need not coincide completely in our understanding of the category; there is surely a great deal more to being a non-defective agent than simply taking steps to avoid acting on false premises or to prevent weakness of will. Since we act on reasons, our actions are susceptible to an open-ended range of criticisms. Our conceptions of virtue and vice capture some of the reasons to which an agent must respond in order to attain non-defective agency. For example, generally one is seen as deficient in charity when she fails to respond to a plea for help for the sake of avoiding a mild inconvenience and delay. Does ethical naturalism help us validate this judgment? The following sections will demonstrate that it does.

RATIONALITY FOR HUMAN BEINGS If the case made in the preceding section is correct, imputing actions to ourselves or others places us against the background of our form of life. Now I will argue that our judgments about proper conduct are likewise inevitably framed as judgments with a certain practical generality, in that they characterize our form of life. As I pointed out previously, one may think that there are various ‘practical identities’ from which I could draw norms for my conduct, and even that


John Hacker-Wright

one could imagine norms applicable to oneself alone. So perhaps I could have the thought, ‘This particularly self-serving action befits JHW’. In thinking this, I could hold that humans are characteristically charitable, while being indifferent to this fact. In that case, I regard myself as a bad human being but an excellent individual, that is, I am excellent as JHW. At least, this appears to be an option, but is it really? Can I conceive of reasons that are reasons only for me? The argument can be extended to show that I cannot; the connection between the aim I hope to achieve and the movements I initiate is only conceivable as acting on a reason against the background of a norm for my form of life as shown previously. Whatever I am aiming at must be intelligible as an aim, and the movement must have some intelligible connection with the fulfillment of that aim in order for what I am doing to be comprehensible as an action. One might think that I can desire anything at all and have any sort of belief pertaining to the fulfillment of that desire, and therefore anything at all can be understood as an action given a certain condition of the individual. But the identification of a desire designates a process within a living thing. Again, to see an organism as having a desire one must employ some understanding of that organism’s form of life. As Anscombe pointed out, the primitive sign of wanting is trying to get.11 Outside the case of organisms possessing the ability to explicitly tell us that they want something, we would have to witness an organism trying to get something in order to attribute a desire to it, or else have found some other physiological signs of desire. In our own case, when we are aware that we desire something, it is a matter of conceiving ourselves as being disposed to attain that thing. In any case, we must appeal to an organism’s life form to attribute desires to it, for we must be able to see it as trying to get something or as having a disposition to do so. While I can have a wildly idiosyncratic desire, it must be like other desires in some ways, and to attribute such a thing to myself invokes some standards for my form of life. Similar assertions can be made about beliefs; after all, some species are evidently not capable of having certain beliefs. Wittgenstein presumably means to point this out when he asks, rhetorically, ‘A dog believes his master is at the door. Can he also believe his master will come the day after tomorrow?’12 By contrast, it may seem that, as language users, we can believe anything whatsoever; I can, for instance, believe that I am Louis XIV. Again, the point is not that wildly idiosyncratic and insane beliefs are not possible, but that at some level knowledge of the life form is involved in identifying the presence of a belief in an individual. For belief involves species-typical capacities for registering the way the world is, and even someone who believes he is Louis XIV has the relevant capacities for determining what it would be for him to actually be Louis XIV, even if those ideas are mistaken or if he is mistaken that those conditions are fulfilled. Taking someone to believe that he is Louis XIV is, among other things, attributing to him those relevant capacities.

Human Nature, Virtue, and Rationality 89 Still, one might wonder whether all of this sets any limits on what can be believed, desired, or understood as a reason for action. As I have just pointed out, the view does permit wildly idiosyncratic desires and beliefs, yet there exists some formal constraint on how I regard my desires and beliefs as giving me reasons. I might take a wildly odd desire and belief to give me a reason: for example, I might believe that by donning a certain necktie, Napoleon Bonaparte will slide a desired twig of mountain ash under the door, and I may think that this method will work only for me. So, I might think I alone have reason to put on this necktie here and now. Nevertheless, to see this desire as reason-giving is to see the desire as directed at something that humans sensibly want to acquire. This is not to say that certain desires are not perplexing; it would be difficult, as Wittgenstein puts it, to ‘find one’s feet’ with someone possessing the desire for a twig of mountain ash, and especially someone who thinks he can get it in such a magical way. One question that may arise is whether the person with such a desire is acting on a reason or under an irrational compulsion that is irresponsive to reasons. Although someone can have idiosyncratic desires and beliefs, to take a desire to provide reason is implicitly to judge that this is something a human can intelligibly try to attain and that someone can be rationally criticized for taking the desire as reason under given circumstances. At the very least, I cannot be perplexed about someone else contriving to get a twig of mountain ash if it is something I have taken myself to have reason to try to get. Reasons I take myself to have attach to an interpretation of reasons for humans. If I cannot understand my reasons to be unique to myself, perhaps I could accord a special weight to acting on the reason that something would benefit me, but what would justify such a weighting, if not reasons that are, in principle, applicable to everyone? Hence, when I act more selfishly than is generally considered acceptable, I demonstrate that my conception of a norm for human life is one with that degree of selfishness; that is, one in which considerations of benefiting oneself have a wider scope and greater weight. I have argued that I must frame my desiring and acting on reasons against the background of my form of life. This is not to say that any desire I have must be counted as a good for human beings, or that I must aim at some good in my desires; the limits of what I can intelligibly desire constitute only part of my conception of my form of life. Acting on a reason places me against a background of my form of life, but that then brings my acting on a reason into contact with a broader background of norms for human beings. I may have what I recognize as a bad desire, and understand myself as acting badly, against other reasons I know I have not to act on that desire. Now we are arguing about what is proper to human beings, and this is the level at which we must argue, ultimately, for what considerations constitute genuine reasons for action. Any argument concerning what reasons we have will have to come back to certain inescapable facts about human beings. These facts include the following: we are agents, we die, we are vulnerable to physical harm, we are entirely dependent on adults through


John Hacker-Wright

infancy, and many similarly obvious facts. I believe any viable norm will have to allow for our survival as individuals and as a species, our happiness, and our ability to live in groups; these are Hursthouse’s four ends. To add to her list, viable norms for human beings will also have to allow for the achievement of rational agency. These are ends that constrain conceptions of what it is for a human being to act well, as it is unreasonable to propose norms for a form of life that would not, under normal circumstances, allow an individual to live, or allow the species to survive, or would require suffering. Yet the fact that we are agents, I think, is particularly important since it is an essential part of our self-conception. Defensible norms must allow us to achieve that which we must rationally take ourselves to be. Human beings can only reliably become rational agents under certain conditions, and the norms we propose for our species must be responsive to facts about how we cultivate rational agency. For example, we need a significant input of care from one or more adults if we are to attain rational agency, and this point is pertinent to defending Hursthouse’s case against impersonal benevolence. Before I flesh out that claim, it is worth noting the connection of the argument thus far with Foot’s call for a ‘fresh start’ in moral philosophy.13 Foot argues that we should dispense with any idea of practical rationality that does not relate to goodness of the will. In her argument for this claim, she draws on Warren Quinn’s case that practical rationality could not be considered a virtue if it allowed us to be rationally shameless; the dominant maximizing conception of rationality surely does just that. Foot follows Quinn in asking what particular hold such a conception of rationality could have on us if it requires of us that we act badly. The argument I made previously strengthens this case by adding that those who think that an alternative conception of rationality exists are simply mistaken. Good reasons answer to the human good on this account, though something like a maximizing account could be sustained by arguing for a maximizing account of the human good. Yet a good human cannot simply maximize the fulfillment of her preferences since, as I will elaborate in the following section, organisms guided by such norms could only fortuitously reproduce themselves as rational agents. Foot is right to hold that morality does not have to answer to a default maximizing view of rationality, and indeed, maximizing desires or preferences cannot be a norm for human beings.

AGENCY AND THE RATIONALITY OF THE FOUR ENDS According to my argument, we are rationally required to be responsive to our needs as human agents because whatever else we think of human beings, we must think of ourselves as agents. This means that we have reasons to protect and promote human agency, and responding to such reasons is an essential part of acting well as a human being. On my view, Hursthouse’s four ends derive their normative authority for us from the necessity of protecting

Human Nature, Virtue, and Rationality


and promoting agency; we must regard the survival of the species as an end because we necessarily think of humans as agents, which they could not be if the species did not survive. So, the way I necessarily think of the human form of life commits me to accepting the norm of species survival. It is important to note that in the case of humans we accept this norm in a distinctively practical way. In the case of any form of life we must think that the survival of the species is a norm for it, as surely, any organism that appears to be bent on destroying its own kind is somehow defective. Yet, we may still think of this as a good thing for our purposes if, say, a mosquito somehow turns on its kind, and we have in fact undertaken to breed just such defective mosquitos.14 Such a position is not rationally open to us in the case of humans, however, because in this case we are thinking of our own life form, and the norms of our conduct are ultimately part of our conception of our own life form. We cannot rationally think of it as a good thing for human beings to perish. I may think of it as good for me that everyone else perishes, but unless such a thought is merely a transient flicker of misanthropy, it necessarily makes me a defective human. Obviously, norms pitting us against our own survival as individuals would similarly be contrary to our self-conception as agents. Yet, it is less obvious how the other two ends relate to our agency. In the case of characteristic enjoyments and freedom from pain, it should suffice to remember that although we take the ability to refuse a pleasure or endure a pain to be indicative of strength of will, it is nevertheless a most basic feature of agency to be able to pursue characteristic enjoyments and to avoid pain. In the case of the fourth end, the good functioning of the social group is also essential to our agency. Though we may choose to live a life in isolation, or indeed though we might be stranded in involuntary isolation due to such circumstances as transportation malfunction or international nuclear aggression, we are creatures who can (for now) anticipate that when we try to fulfill our desires, it will happen in a social environment. This means that we are creatures for whom it is a constant fact that our attempts at obtaining something may be interfered with or enhanced by others.15 As discussed previously, Copp and Sobel find that Hursthouse did not defend authority of the four ends, and it is indeed unclear what status the four ends are supposed to have within her account. Julia Annas takes the four ends to act as a sort of psychological limit to what we might choose to do.16 On that view, prescriptions that require that we go against the four ends should be rejected because the proposed actions are bound to be difficult and frustrating for us, given our psychological constitution. This reading takes the four ends to be inscribed in our nature in much the same way in which atavistic drives and behaviors are authoritative for non-human animals. This is a possible view, but it is not the best view that an Aristotelian ethical naturalist can advance, and Annas rightly argues against it. She claims that it makes it seem as though the ethical naturalist is appealing to the ‘old Adam part of us’.17 This is not compelling because there would appear


John Hacker-Wright

to be little evidence for the claim that we would be frustrated by all norms requiring actions in conflict with the four ends. Such frustration would follow surely only in extreme cases, such as being consistently required to forgo characteristic enjoyments or to be in constant pain. Another problem Annas points out is that this view does not challenge received views of rationality. The ends, on her reading of Hursthouse, place limits on our pursuit of what would otherwise be rational. Hence, this reading places the four ends in conflict with Foot’s call for a fresh start, with the result that aspiring to be virtuous would mean aspiring to be systematically irrational. Whether Annas is right to interpret Hursthouse in this way is not clear, but my view allows for a very different interpretation of the four ends. On the reading that I propose, the status of the four ends is derived from our rational self-interpretation. We must understand ourselves to be agents if we are to act, and the four ends set rational constraints on what we can rightly aim to do because only norms that require actions consistent with promoting the four ends will be consistent with taking ourselves to be agents. Further, my understanding of the four ends is consistent with Foot’s fresh start; the demands of protecting and promoting human agency define what it is to be a good human being, and the four ends help to flesh out that conception of human goodness. They are not barriers to rationality, on my reading, but specify what a rational human being should aim to attain. Hence, there is a response to the demand of Copp and Sobel for an argument supporting the four ends that avoids the pitfalls of Annas’s psychological reading.

NATURALISM AND IMPERSONAL BENEVOLENCE As noted previously, Hursthouse applies her views on ethical naturalism to the issue of whether impersonal benevolence can be considered a virtue. Her argument is quite tentative, but she concludes that the burden of proof lies on the advocate of impersonal benevolence to show that impersonal benevolence is compatible with the four ends. Further, she is doubtful that the ends of species survival and the good functioning of the social group can be secured by impersonally benevolent agents. Of course, it is doubtful whether agents pursuing those ends can actually achieve them, but it is a mistake to pose the question as though the vindication of the virtues is a matter of the results we will likely achieve through acting in accordance with a particular virtue or set of virtues. Instead, as I see the four ends, they set limits on what we can rationally aim to achieve. The question is not whether by aiming at maximizing the balance of happiness for sentient creatures we will likely maintain the survival of the species, but whether having the higher order aim of maximizing the balance of happiness would readily allow the adoption of more specific aims that, in principle, conflict with survival of the species. The goodness of the will is not determined by the outcomes it

Human Nature, Virtue, and Rationality


attains but on the basis of one’s actions, which is determined, under normal circumstances, by one’s intentions. Peter Singer raises this question: ‘[I]s the continuance of our species justifiable in the face of our knowledge that it will certainly bring suffering to innocent future human beings?’18 Though he confesses to a certain degree of optimism, it is not clear how he would justify such an opinion. Since there seems to be a great deal of suffering on the horizon, in light of the facts and his methods, the answer he should give to his own question could well be ‘no’. On the naturalist view as I am construing it, we cannot leave the matter of whether we should aim at the destruction of our species up to facts such as these. The rationality of adopting aims that are consistent with the preservation of the species cannot turn on whether our continued existence promises more pleasure than pain. We are rational when we respond to reasons that would move someone possessing dispositions to adopt aims consistent with the four ends under such conditions as we are likely to confront. Of course, there are readily imaginable apocalyptic situations in which none of the four ends is likely to be achievable, and in that case, there may not be any justification for the traditional virtues. Yet in circumstances such as we face, where ecological catastrophe and scarcity threaten to weigh hedonic scales strongly and perhaps inevitably to the side of suffering, the rational thing to do is not to end the species. Clearly, many people think that the value of life extends beyond the balance of pleasure and pain. Though intuition can be pushed too far, to the point of making life sacrosanct, the intuition has some connection with the four ends and the value of rational agency. One who embodies virtues consistent with the four ends will of course pursue a pleasant life and try to bring about a pleasant life for others. But if this proves impossible, it is not clear that the value of life has vanished; agency under such conditions is surely impaired, but good people can find value in life working for the amelioration of suffering even under very bad conditions. It is difficult to imagine any circumstance under which it would be rational to aim at the destruction of the species because it is difficult to imagine something for the sake of which it would be worth destroying our entire species. However, it is imaginable that an individual could embrace an action that is certain to bring about his own death. Still, this is not the same as aiming at one’s death. When such a death is rational, we aim at something that is worth doing for the sake of friends or our community, such as to protect their lives. The naturalist’s view suggests that suicide is irrational when it is done in the face of a life that promises more suffering than pleasure. Of course there are limits to this claim; for example, suicide would surely be rational if the suffering is so extreme that I am impaired from doing anything for myself or others. In such a case, unremitted suffering is all that is attainable. In other cases, hope and other virtues make demands on us that override reasons to take our own lives to avoid a balance of suffering. Another aspect of Hursthouse’s argument against impersonal benevolence is that virtues such as friendship, care, and loyalty are legitimate virtues. That


John Hacker-Wright

is to say, there are positive requirements to give special attention to friends, family, and our own communities. These claims are much more difficult to sustain than the claim that we ought to take special care to preserve our own species. Hursthouse seems to think that these virtues are required for the good functioning of our social groups, but they may only appear to be necessary because they are established social conventions and so we feel ruffled by any deficit of attention from friends and family. There is a deeper question of whether such conventions have any kind of rational foundation on the naturalist’s view. Oddly, Hursthouse’s discussion of impersonal benevolence does not appeal to the fact that long-term relationships are surely one of our characteristic enjoyments and that virtues such as friendship, loyalty, and care are surely essential to realizing that enjoyment, and not only as means.19 That is, these three virtues play a constitutive role in our relationships, leading us to spend time with other people and to pay attention to them in a way that yields the enjoyment characteristic of long-term relationships, which would surely be impossible if we distributed our efforts in accordance with the demands of impersonal benevolence. Further, care must be a virtue for organisms that attain rational agency in the way that human beings do, through a long period of physical and psychological dependency. As I have argued elsewhere, we need others to care for us through our infancy and childhood, and this cannot be based on a hope of future reciprocity or other self-interested motive.20 Such care also cannot be reduced to impersonally benevolent motives; the shifting possibilities of improving the balance of happiness would undermine the possibility of the devotion to one’s own charge in a way that would make it constantly a question whether to adopt some aim conflicting with the aim of imparting rational agency. If this approach to justifying the virtues is sound, then impersonal benevolence is not a virtue, whereas benevolence is one virtue among others. In the case of human beings, agency occurs in a form that is particularly susceptible to physical harm from violence, malnutrition, toxins, and other threats. Hence, we need to be committed to helping others avoid such harms and recover from them when they occur.21 This means that the virtuous agent will have a commitment to helping others that certainly overrides any concerns about inconveniences that such aid may bring. I have argued that Hursthouse’s four ends have rational authority over humans because they specify ends at which we must aim if we are to remain consistent with our nature as rational agents. Further, I have argued that it follows from the four ends, so conceived, that we must reject impersonal benevolence and embrace a view of benevolence as one virtue among others such as friendship, loyalty, and care. Neo-Aristotelian ethical naturalism supports these virtues by arguing for a view of rational action that is indexed to our species. Critics of this view, such as Copp and Sobel, have taken its advocates to ignore the important transition that occurs when we shift to the case of rational animals such as human beings. Navigating this transition is crucial to the success of this brand of ethical naturalism,

Human Nature, Virtue, and Rationality


and hopefully this chapter has helped clarify how natural norms apply to rational beings.

NOTES 1. The two major defenses of this naturalistic approach to justifying the virtues are Philippa Foot’s Natural Goodness and Rosalind Hursthouse’s On Virtue Ethics. See also Alasdair MacIntyre, Dependent Rational Animals. 2. An important qualification of this claim is that it is a norm for some species to withhold care under conditions of scarcity. Yet, this point can be folded back into our characterization of the normal life of that species; in other words, it is part of the characteristic life of that species to withhold care under such circumstances. See Andreou, ‘Getting On in a Varied World’. 3. Foot, Natural Goodness, 30. 4. Foot, Natural Goodness, 43. 5. Foot, Natural Goodness, 56. 6. Hursthouse, On Virtue Ethics, 222. 7. Copp and Sobel, ‘Morality and Virtue: An Assessment of Recent Work in Virtue Ethics’, 540. Note that Copp and Sobel are a bit mistaken in framing this criticism; plants do not have characteristic enjoyments or well-functioning social groups, and so there are fewer than four ends for some species, on Hursthouse’s view. 8. Thompson, Life and Action, 43. 9. Thompson, ‘Apprehending Human Form’, 71. 10. Anscombe, Intention, 49. 11. Anscombe, Intention, 68. 12. Wittgenstein, Philosophical Investigations, 183. 13. Foot, Natural Goodness, 5–24; see also Foot, ‘Does Moral Subjectivism Rest on a Mistake?’. 14. As I write this, genetically engineered male mosquitoes are being released that produce moribund larvae. Though defective mosquitoes, they are at least good from our standpoint insofar as they will cause a crash in mosquito populations and hence reduce diseases that are carried by their species. See ‘Sterile Males for Mosquito Control’. 15. Of course, this puts our social relations in a purely instrumental light, and as I will argue later, our interest in relationships must be more than instrumental. 16. See Annas, ‘Virtue Ethics: What Kind of Naturalism?’. 17. Annas, ‘Virtue Ethics’, 25. 18. Singer, ‘Should This Be the Last Generation?’. 19. Although Hursthouse does not connect the issue of characteristic enjoyments to the rejection of impersonal benevolence, she clearly recognizes the point in On Virtue Ethics, 234. 20. See my ‘Human Nature, Personhood, and Ethical Naturalism’. 21. A point made forcefully by Alasdair MacIntyre in Dependent Rational Animals, 99–118.

REFERENCES Andreou, Chrisoula, ‘Getting On in a Varied World’, Social Theory and Practice 32 (2006), 61–73.


John Hacker-Wright

Annas, Julia, ‘Virtue Ethics: What Kind of Naturalism?’, in Virtue Ethics Old and New, ed. Stephen Gardiner (Ithaca, NY: Cornell University Press, 2005), 11–29. Copp, David, and David Sobel, ‘Morality and Virtue: An Assessment of Recent Work in Virtue Ethics’, Ethics 114 (2004), 514–554. Foot, Philippa, ‘Does Moral Subjectivism Rest on a Mistake?’, in Moral Dilemmas (Oxford: Oxford University Press, 2002), 189–208. ———, Natural Goodness (Oxford: Oxford University Press, 2001). Hacker-Wright, John, ‘Human Nature, Personhood, and Ethical Naturalism’, Philosophy 84 (2009), 413–427. Hursthouse, Rosalind, On Virtue Ethics (Oxford: Oxford University Press, 1999). MacIntyre, Alasdair, Dependent Rational Animals (Chicago: Open Court Press, 1999). Singer, Peter, ‘Should This Be the Last Generation?’, New York Times, June 6, 2010. ‘Sterile Males for Mosquito Control’, Nature, 479 (03 November 2011), 9. Thompson, Michael, Life and Action (Cambridge, MA: Harvard University Press, 2008). Wittgenstein, Ludwig, Philosophical Investigations, trans. G.E.M. Anscombe, P.M.S. Hacker, and Joachim Schulte, rev. 4th ed. (Malden, MA: Blackwell, 2009).


Good (as) Human Beings Philipp Brüllmann

I Virtue ethical projects that draw their inspiration from Aristotle often come with a special kind of naturalism. This ‘Neo-Aristotelian Naturalism’ (NAN) is the subject of this chapter.1 What is NAN? In a first approximation, ignoring the varieties, one could say the following: (i) NAN is a form of ethical naturalism in that it claims that moral judgments can be grounded in judgments about human nature.2 (ii) NAN differs from most other forms of ethical naturalism by its functional, teleological perspective and its focus on the evaluation of something as ‘good of its kind’, especially when applied to living beings. (iii) NAN considers moral judgment a variant of this kind of evaluation. According to NAN, the morally good person is good as a human being. Recent attempts to assess NAN have mainly focused on its concept of (human) nature. Opponents of NAN doubt that the functional view of nature is compatible with the natural sciences, especially with biology ‘after Darwin’.3 But perhaps this doubt is beside the point, for proponents of NAN would respond that their aim is not to offer a biological account of animals but to spell out what is necessary for us to recognize something as a living being.4 It seems an open question to what extent NAN has to take the natural sciences into account. The debate has hit an impasse. In what follows, I suggest a different approach to an assessment of NAN. I will argue that if we follow its lines, we do not only have to accept a certain picture of ‘human nature’, adequate or not by whatever standards. We also have to accept a certain picture of ‘moral thinking’. The functional perspective that is characteristic of NAN imposes a number of constraints on our consideration of the morally good. These constraints turn out to be problematic as soon as we take NAN as a way of grounding moral judgments.5 Or so I will argue. NAN is interesting in its own right, but in an important sense, it is a ‘non-starter’.


Philipp Brüllmann

II The following approach to NAN will be based on Rosalind Hursthouse’s book On Virtue Ethics.6 This is not because I wish to criticize her account in particular, but because Hursthouse proposes NAN explicitly in the indicated way. For Hursthouse, the appeal to human nature plays a role in grounding moral judgments. Her book is therefore especially well-suited to illustrate the more general point at stake. In accordance with the general aim of this chapter, I will not attempt to do full justice to On Virtue Ethics. Instead, I will pick out those aspects that appear relevant for my present purposes. It should be noted, however, that Hursthouse’s argumentation is far more elaborate, and careful, than my summary suggests.

III In her 1999 book On Virtue Ethics, Rosalind Hursthouse pursues a normative project. She contradicts the opinion that virtue ethics cannot provide action guidance, that is, that it does not tell us what sorts of acts we should do (Part I), and she develops a virtue ethical account of right action that is conceived as an alternative to utilitarianism and deontology. The starting point of this account is the following, well-known premise: P.1. An action is right iff it is what a virtuous agent would characteristically (i.e., acting in character) do in the circumstances. (28) As Hursthouse correctly notes, this premise does not provide action guidance—it does not tell us what to do—unless we know who the virtuous agent is. So what we need is a specification of the virtuous agent. And to make sure that the virtue ethical account of right action offers a genuine alternative to deontology, this specification should not refer to moral rules or duties. It should not define the virtuous agent as someone ‘disposed to act in accordance with correct moral rules’ (28–9). Here is Hursthouse’s suggestion: P.1a. A virtuous agent is one who has, and exercises, certain character traits, namely, the virtues. (29) But of course this premise cannot be the whole story, either. Unless we know which character traits are the virtues, we do not know who the virtuous agent is, and thus have no answer to the question of what sort of acts we should perform. P.1a calls for an account of the virtues. In order to give this account, Hursthouse introduces the following premise, explicitly labeled as ‘neo-Aristotelian:’

Good (as) Human Beings


P.2. A virtue is a character trait that a human being needs for eudaimonia, to flourish or to live well. (167) Hursthouse’s explanation of P.2 (chs. 8–11) is long and complex, but in the present context it is not necessary to go into the details of her account. Instead, we will confine ourselves to two important aspects that should be kept in mind. First, according to Hursthouse, P.2 offers a basis for ‘validating’ beliefs about which character traits are the virtues. It helps us to find out whether our beliefs about the virtues are correct or not (164). Second, on Hursthouse’s account, the claim that human beings need the virtues to flourish or to live well encapsulates two interconnected theses (167). One of these, call it ‘P.2a’, is the following:7 P.2a. The virtues make their possessor a good human being. (167) Referring to the work of Philippa Foot, Hursthouse asserts that P.2a talks about human beings as living beings, and she develops an account of good human beings that starts from good plants and good (social) animals (ch. 9–10). It is with P.2a that NAN comes into play.

IV Hursthouse claims that P.2a ‘can get off the ground as a criterion for a particular character trait’s being a virtue’ (192). This criterion might be specified as follows (this is my own formulation): C: A character trait is a virtue iff it contributes to making its possessor a good human being. Since the notion of ‘contributing’ is extremely vague, C cannot be the last word on the matter. In fact, our picture of the virtues will differ decisively depending on how this notion is understood. But even in this provisional formulation, one can see—digressing for a moment from the account of Hursthouse’s theory in particular—what makes C attractive to the virtue ethicist in general (i.e., to someone who thinks of virtue ethics as a selfstanding, normative project). If we can show that P.2a is correct, and if we can give an account of good human beings that draws on something like ‘human nature’, then we seem to have gained an independent and objective basis for our judgments about the virtues.8 The basis seems independent because the concept of human nature does not presuppose any notion of morally good actions, intentions, choices, etc., which might perchance depend on a different account of morality. The appeal to human nature would help virtue ethics to avoid a vicious circle and to establish its own position as an alternative to other theories (which is exactly what


Philipp Brüllmann

Hursthouse intends). And the basis seems objective because (i) whether a certain character trait is a virtue would depend on the way things are, as opposed to, for example, the way we feel about them, and (ii) whether a certain character trait is a virtue would depend on the way things are in nature, as opposed to, for example, the way things are in relation to a certain cultural, historical, or social framework.

V The series of specifications that were introduced to develop a virtue ethical account of right action has led to an attractive result, as it seems. The concept of a good human being offers an independent and objective basis for judgments about the virtues, which in turn offers a basis for judgments about right action. On a closer look, however, the account faces problems, for there is a fundamental objection against the naturalism expressed by C. This objection is that C might show (morally) ‘unacceptable results’.9 One prominent example goes as follows. If we assume with NAN that C talks about human beings as living beings and that the evaluation of living beings refers to ends like individual survival and continuance of the species, then it seems inevitable to judge the childless (just as practicing homosexuals or people who practice celibacy for religious reasons) as vicious (214–216). Like other virtue ethicists of the Neo-Aristotelian camp, Hursthouse is well aware of this objection and deals with it thoroughly. And like other Neo-Aristotelians, she replies to the objection by pointing to the differences between human beings and (other) animals.10 While the determination of animal nature is a task for the natural sciences, Hursthouse asserts, the determination of human nature proceeds from within an ‘ethical outlook’ (187–191): [T]he beliefs and putative facts about who can and cannot be relied on, about whether you can fool most of the people most of the time, or whether they can easily be manipulated, about what can be discerned to be a pattern in life, what is to be attributed to good or bad luck and what is ‘just what is to be expected’—about in short, human nature and the way human life works—do not fall tidily under either classification [of empirical vs. moral ‘facts’, P.B.]. Neither side believes what they believe about how life works on the basis of even local, let alone worldwide, observation or statistical analysis. The beliefs are part and parcel of their ethical (or immoralist) outlook. (189) Yet, it is easy to see how this reply invites another objection (which Hursthouse deals with as well). If we enrich our concept of human nature with facts from within an ethical outlook, we are at risk of merely re-expressing our ethi-

Good (as) Human Beings


cal beliefs instead of grounding them. We get drawn, it seems, into that vicious circle that the appeal to human nature was meant to avoid (165). According to Hursthouse, this second objection operates with the wrong concept of validation. Referring to good human beings in order to validate beliefs about which character traits are the virtues does not mean to derive ethical conclusions ‘from a neutral point of view’. It rather means to follow a ‘Neurathian procedure’ (a term she borrows from John McDowell), that is, to start from within an acquired ethical outlook and to submit opinions that are part of this outlook to ‘reflective scrutiny’: ‘And those [particular judgments, P.B.] that were part of the outlook and survived the reflective scrutiny would not merely re-express it; they would now express, so to speak, that they had survived the scrutiny’ (166).

VI Hursthouse’s version of NAN is characterized by two interrelated specifications. The first specification concerns the concept of a human being. This concept is not to be understood as a scientific account of human nature but as an account given from within an ethical outlook. The second specification concerns the question of how beliefs about the virtues can be validated with reference to human nature. This validation has the form of a Neurathian procedure that submits parts of our outlook to reflective scrutiny. What Hursthouse does, in a word, is to combine naturalism with validation ‘from inside’.11 Both specifications no doubt provoke a number of interesting questions, but in the present context we should set them aside. As I said previously, my aim is not to criticize Hursthouse’s account but to explain how her way of proceeding throws light on the constraints NAN imposes on moral thinking. We should hence take the next step. Hursthouse’s idea seems to be something like the following. Although it is true that C provides a criterion for a certain character trait’s being a virtue, it is wrong to assume that we could use C to derive ethical conclusions (conclusions about which character traits are the virtues) from a neutral point of view. Validation with reference to C should rather be described as a Neurathian procedure that starts from within an ethical outlook and submits certain beliefs to reflective scrutiny. In this idea, and its elaboration, Hursthouse seems to separate two issues. On the one hand, there is the issue of explaining why an account of human nature might offer a criterion for a particular character trait’s being a virtue. This explanation is based on Philippa Foot’s NAN. On the other hand, there is the issue of explaining how certain beliefs about which character traits are the virtues can be validated with reference to this criterion. This explanation is based on John McDowell’s idea of a Neurathian procedure. So at least implicitly, Hursthouse seems to hold that these two issues can be treated


Philipp Brüllmann

separately, and she seems to hold that the explanations she gives are more or less independent of each other. It is this ‘independence assumption’ that is problematic as well as revealing.

VII I take the independence assumption to be problematic for two reasons. First, I think that the theory of NAN determines, to some extent at least, the way in which beliefs about the virtues can be justified with reference to human nature. For that reason it is generally problematic to separate the question of justification or validation from the Neo-Aristotelian framework. Second, I think that NAN, if taken literally, leaves little or no room for a Neurathian procedure. If we pursue a Neo-Aristotelian Naturalist strategy, we cannot take a place in Neurath’s boat. In the following, I will try to elaborate these two reasons a little further and argue that NAN, a theory that is based on a certain view of the grammar and meaning of ‘good’, is incompatible with a Neurathian procedure. The argument will be rather sketchy. Instead of giving a full account of the theories involved, I will only be able to point to some relevant aspects. Yet I hope to show that the problem at stake does not concern the details of those theories but some of their basic ideas.

VIII As mentioned previously (section I), NAN differs from most other forms of naturalism by its functional, teleological perspective and its focus on the evaluation of something as good of its kind. As will turn out in what follows, it is this functional perspective that stands behind the idea that the concept of human nature provides a criterion for the virtues. NAN’s functional perspective rests on one basic assumption: the assumption that ‘good’ is a logically attributive adjective.12 What does that mean? Briefly, it means that ‘good’, when used properly, always means ‘good relative to a certain kind’. This is taken to be true also of cases in which ‘good’ appears in the grammatically predicative way. When ‘good’ is used properly, the judgment ‘X is good’ is to be understood as elliptical for ‘X is a good F’. In this respect, the adjective ‘good’ is similar to adjectives like ‘big’ and ‘small’, but different from adjectives like ‘red’ and ‘yellow’. Usually, the following criterion is used to mark the difference between ‘good’ as attributive and, for example, ‘red’ as non-attributive: Whereas from ‘X is a red F’, you can infer ‘X is red’ and ‘X is an F’, you cannot infer ‘X is good’ and ‘X is an F’ from ‘X is a good F’.13 There is a debate on whether this criterion really does the job,14 but in the present context we can leave the intricacies aside. Instead, we should focus on the question of what the idea that ‘good’

Good (as) Human Beings


is attributive—provided its correctness—implies according to NAN. Two implications seem to be of special importance. First implication. There are cases in which it is unquestionable that ‘good’ is used attributively, as in the judgment ‘X is a good knife’. In cases like these, the relation between the description of X as a knife and the evaluation of X as a good knife seems unproblematic because if you know what a knife is, you already know (in some way or other) what a good knife is. In the present context, we cannot go into the question of why exactly this is the case, but obviously it has something to do with the fact that knives are tools, that is, artifacts designed to serve a certain purpose or function. Thus (i) ‘X is a good knife’ means ‘X is a knife that serves its purpose well’; and (ii) to know what a knife is, is to know the purpose it was designed for and, at least partially, the conditions under which it serves its purpose well.15 Apparently, good knives offer an example for an evaluation that is reducible to a description (to know that X is a good knife is nothing but to know that X cuts well). Now if it is true that the only proper use of ‘good’ is the attributive one, and if all the other cases have enough in common with the case of tools, then the relation between descriptive and evaluative judgments is never problematic because the latter are always reducible to the former. This, in short, is the reply that NAN offers to G.E. Moore’s ‘Open Question Argument’.16 Second implication. When Neo-Aristotelian Naturalists talk about ‘good’ as logically attributive, they usually talk about standards of goodness. While the standards of redness are the same no matter which object we call red, the standards of smallness change with, and depend on, the object that is called small. If ‘good’ is like ‘small’, then the same holds for the standards of goodness, that is, the standards of goodness change with, and depend on, the object that is called good. Why should we care? According to authors like Philippa Foot, we should care because this implication guarantees the objectivity of those standards in the following sense. The criteria for good knives cannot be chosen, so to speak (we cannot use any standard we like when evaluating knives) because partially those criteria are determined by what knives are, namely, tools designed to serve a certain purpose. This, in short, is the objection that NeoAristotelians raise against ethical non-cognitivism, which—in their view at least—makes consistency the only restriction to an application of evaluative terms.17 Thus Philippa Foot writes in ‘Goodness and Choice’: [T]he man who uses these words [‘a good knife’, P.B.] correctly must use them in conjunction with particular criteria of goodness: those which really are the criteria for the goodness of knives. No matter what he may do in the way of choosing knives which are M he cannot say ‘M knives are good knives’ unless M is a relevant characteristic, or unless he is prepared to show that M knives are also N knives, and N is a characteristic of the right kind.18


Philipp Brüllmann

The basic idea is that using the words ‘a good knife’ correctly implies using them in conjunction with those criteria that ‘really are’ the criteria for good knives, that is, in conjunction with objective standards. Let us sum up this idea by the following thesis: T 1: If someone uses the wrong criteria when evaluating knives as knives, she either does not know how to use the concept ‘X is a good F’ correctly, or she does not know what a knife is.19 So, according to NAN, there are two important implications of the assumption that ‘good’ is logically attributive. The first implication is that the relation between the descriptive and the evaluative is unproblematic. The second is that there are objective standards of goodness, determined by the object that is called good. IX The assertion that ‘good’ is attributive grounds the functional perspective that is characteristic of NAN. But it is not the whole story. You can subscribe to this assertion without being a Neo-Aristotelian Naturalist in a proper sense. For NAN is also, and mainly, a theory about the scope of the functional perspective. So to be a Neo-Aristotelian Naturalist in a proper sense, one has to make two further claims. First claim. The evaluation of living beings as living beings, that is, the evaluation of the behavior they display in order to survive and reproduce in a specific way, follows the pattern we have illustrated by the example of an evaluation of knives. It is an evaluation of something as good of its kind, based on objective standards, determined by what kind of living being we are dealing with. To be sure, since living beings are neither artifacts nor designed to serve a certain purpose, there has to be a particular reason for the claim that the functional perspective is applicable to (say) bees and wolves. The teleological story has to be different from the one that is told in the case of tools. And as mentioned in section I, it is this teleological story that many critics of NAN find problematic. But let us assume for the moment that NAN manages to provide such a story, declaring that while in the case of a knife, the function is built into its identification as a tool, in the case of a bee or a wolf, the function is built into their identification as ‘life-forms’ or organisms.20 Our thesis from the preceding paragraph would then take the following form: T 2: If someone uses the wrong criteria when evaluating bees as bees or wolves as wolves, she either does not know how to use the concept ‘X is a good F’ correctly, or she does not know what a bee or a wolf is. Second claim. Moral evaluation is the evaluation of a certain living being, namely, a human being, as the living being it is. This means not only that

Good (as) Human Beings


moral judgments follow the pattern we have illustrated by the example of an evaluation of knives (moral judgments are evaluations of something as good of its kind). It also means that these judgments are about a life-form and hence share some criteria with the judgments about other living beings.21 (This latter point is important. If there were only a structural similarity in the evaluation of all living beings, and no similarity in content, we could skip the animals and pass directly from knives to human beings.) It is no doubt crucial to the project of NAN in general to give a thorough account of the similarities and dissimilarities between human beings and (other) animals. Yet apart from that, there is a specific problem that has to be mentioned. This problem lies in a prima facie asymmetry between the two claims just outlined. While it seems fair to say that the first claim talks about something more or less obvious, the same does not seem to hold for the second claim. It is more or less obvious that judgments about good bees follow the pattern of judgments about good knives22 (although, as I said, the teleological story has to be different). But I think to most of us it is not obvious that moral judgments also follow this pattern. Put in a different way, NAN claims that the application of the functional perspective to the case of animals is justified by our mastery of the concept of a life-form. Can the claim that moral evaluation has the same pattern be justified in a similar way? What is its conceptual basis? As far as I can see, NAN does not provide an argument for the second claim. Instead, it follows the strategy of a reversal of the burden of proof, couched in the question why the meaning of ‘good’ should suddenly undergo a mysterious change when it comes to judgments about moral goodness.23 Though I doubt that this is convincing,24 I would like to circumvent the problem by making the following suggestion. Instead of talking about judgments of the form ‘X is morally good’, let us concentrate on judgments of the form ‘X is virtuous’ or ‘X is a virtue’, for in cases like these it seems more natural to assume that the relevant evaluations are evaluations of human beings as human beings. In a traditional reading, one could even say that the term ‘virtuous’ means nothing but ‘good of its kind’. On this basis, our thesis now takes the following form: T 3: If someone uses the wrong criteria when evaluating human beings as human beings, i.e., when evaluating someone in terms of the virtues, she either does not know how to use the concept ‘X is a good F’ correctly, or she does not know what a human being is.

X This summary of NAN is of course far from exhaustive, but it should suffice to illustrate two things: first, what Neo-Aristotelian Naturalists mean when they claim that there is a ‘common pattern’ to the evaluation of bees


Philipp Brüllmann

and wolves on the one hand and moral evaluation on the other; second, why they find this common pattern attractive. If NAN is correct, then moral evaluation is just as unproblematic and rests on standards as objective as the evaluation of knives, wolves, and bees. Let me develop a little further what this tells us about the theory of NAN. Apparently, the point of NAN is not that we should take into consideration what a human being is when deliberating on moral questions. This, I think, is hardly controversial. The point of NAN is that there is a special connection, a link, between judgments about the virtues and judgments about human nature. As shown in the preceding paragraphs, this link amounts to a reduction and is based on our mastery of certain concepts like ‘tool’, ‘lifeform’, and (as suggested) ‘virtue’. NAN is naturalism on a conceptual basis. By claiming that there is a specific, conceptual link between judgments about the virtues and judgments about human nature, NAN does not only presuppose a functional concept of human nature. It also determines how this concept grounds our judgments about the virtues. In a nutshell, it is because we know how to apply ‘X is a good F’ correctly, and because we also know the scope of this concept, that we consider the virtuous person as good qua human being. This, once again, is the point of NAN’s criticism of ethical non-cognitivism that Hursthouse reiterates herself.25 So there is a close connection between the features by which NAN has been characterized in section I. NAN claims that moral judgments can be grounded in judgments about human nature because it regards moral judgment as a variant of the evaluation of something as good of its kind. As we have seen in section IV, the reference to human nature is attractive for an ethics of virtue because it seems to provide an independent and objective basis for our judgments about the virtues. Now we can see that NAN determines the way that happens, namely, as I said, by a reduction. To understand the point of this remark, it is crucial to distinguish the question of how, according to NAN, virtue ethics can take advantage of the objectivity of human nature (a) from the question of how objective NAN’s concept of human nature actually is (b). It is the first question that I have been dealing with, and on the whole my suggestion is to characterize NAN rather by (a) than by (b).

XI Let us now take a look at the Neurathian procedure. As mentioned previously (section V), Hursthouse borrows the concept from John McDowell (who of course draws on W.V.O. Quine).26 But her remarks on the topic are too brief for us to decide how much of McDowell’s (let alone Quine’s) theoretical background she wants to adopt in addition to the basic idea of a Neurathian procedure. This basic idea brings together two aspects. First, there is the denial of the existence of an ‘external standpoint’. As Neurath’s mariner has to keep

Good (as) Human Beings


his ship in good order while at sea, we have to start the reflective scrutiny of (say) our ethical beliefs from within an acquired ethical outlook. There is no way of escaping this outlook to start from dry or neutral ground. All we can do is affirm or change this outlook bit by bit. In the words of Rosalind Hursthouse, ‘The general idea is that I take one of my ethical beliefs—say, that courage is a virtue—and, holding the rest of my ethical outlook intact, put it up for question’ (167). To be sure, the denial of the existence of an external standpoint is not in itself a theory of justification. It does not tell us what the standards of successful justification are. But usually, and this is the second aspect, the idea of a Neurathian procedure is considered as an idea about how justification works; unsurprisingly, it is considered as an alternative to foundationalist accounts of justification.27 Hursthouse follows this line but is quick to warn us against having exaggerated expectations toward the kind of validation the procedure offers. Even though our ethical outlook might on the whole survive reflective scrutiny, this might still not convince the ‘mafioso drug baron’ to change his life (165).28 Be that as it may. It is another aspect of the Neurathian procedure that I wish to highlight in the present context. Like John McDowell, Hursthouse emphasizes the wide scope of the procedure: every belief can be submitted to the process of reflective scrutiny, and every belief is in principle open to the risk of revision. As Hursthouse nicely puts it, it might turn out that Neurath’s boat is Theseus’s ship (166). The idea of a Neurathian procedure implies that there are no a priori fixed points in our outlook. The problem is to determine the exact scope of ‘every belief’ and ‘no fixed points’. There surely have to be some restrictions for the scrutiny to be recognizable as a rational procedure, restrictions imposed by what McDowell calls Logos.29 But within these boundaries there is still room for different options. Two of them appear relevant. (i) A radically wide scope: Except for those beliefs that guarantee the rationality of the procedure, none of our beliefs are a priori fixed, which explicitly includes all kinds of definitions. This option fits the idea that the Neurathian procedure is about our conceptual schemes and the Quinean origin, which is pointed to by Hursthouse herself (165). (ii) A narrower scope: There is no ethical belief (be it in the form of a particular judgment or a principle) that is a priori fixed. But there might be a fixed background of other beliefs (in addition to those that guarantee the rationality of the procedure) that stays intact while we submit our ethical outlook to reflective scrutiny. This option fits the idea that it is the ethical outlook to which the Neurathian procedure is applied.

XII I do not know for sure which is the correct interpretation of Hursthouse’s view, but it seems clear to me that NAN is incompatible with both.


Philipp Brüllmann

It should not come as much of a surprise, I think, that NAN is not compatible with a Neurathian procedure of a radically wide scope (i). For, as a matter of fact, NAN is characterized by a number of ‘a priori fixed points’, like the belief that there is a common pattern to the evaluation of knives, bees, and human beings, or the belief that moral evaluation is an evaluation of something as good of its kind. In other words, whereas the (radical) Neurathian procedure is directed against the idea that our conceptual schemes are invariably fixed, NAN rests on a conceptual basis (see section X). If the radically wide scope is what Hursthouse has in mind, then showing the incompatibility of NAN and Neurathian procedure is not very interesting. More interesting is the case of the narrower scope (ii), which restricts the procedure to our ethical outlook and allows for a stable background of other beliefs (I am inclined to ascribe this view to Hursthouse). Why is this option incompatible with NAN? In my view, the answer lies in the role that NAN ascribes to this background. As we saw in section X, NAN is essentially a theory about how judgments about the virtues can be grounded in judgments about human nature, namely, on the basis of appropriate conceptual knowledge. It is our conceptual ‘background-knowledge’ that allows us to judge the correctness of criteria for evaluative judgments in the way Foot and Hursthouse do against R. M. Hare (see note 25). Someone who calls a diseased cactus a good one does not have a different, perhaps depraved, view of the flora that is to be submitted to reflective scrutiny. He is not a ‘cactus-mafioso’, but just someone who does not know how to use certain concepts correctly. This is the whole point of the analogy illustrated by T 1–3. So, applying a Neurathian procedure to our ethical outlook does not only mean introducing a different concept of validation (‘X survived reflective scrutiny’). It also means ‘putting up for question’ beliefs that are already validated if we follow NAN.

XIII The result of the argument in sections VII–XII is that NAN is not compatible with a Neurathian procedure. It is already by submitting naturalism to this procedure that Hursthouse detaches it from its Neo-Aristotelian background. So it seems we are left with two alternatives. The first alternative is to give up the Neo-Aristotelian way of validating C (‘A character trait is a virtue iff it contributes to making its possessor a good human being’) and to regard the latter as one belief of our ethical outlook that may or may not survive reflective scrutiny. To my impression, this is what Hursthouse is actually doing;30 and if what I say is correct, her approach leads to a completely different kind of naturalism, a different way of taking nature into account when deliberating on ethical questions (which is true also if C should survive reflective scrutiny). To fully understand the consequences of this first option, we have to remember how NAN is connected to the advantages of naturalism for an

Good (as) Human Beings


ethics of virtue (section X). If applying a Neurathian procedure implies denying that C is sufficiently justified on the basis of a priori fixed points (on the basis of our mastery of certain concepts), it also implies doing without the specific way in which NAN provides virtue ethics with independence and objectivity, namely, by reducing judgments about the virtues to judgments about human nature.31 And this appears to me a decisive shift in the theory. Be that as it may. It is the second alternative that is more relevant in the present context, that is, the option to stick with NAN as a way of grounding judgments about the virtues and hence to give up the idea that C is to be validated by a Neurathian procedure. As I will try to show, this option tells us something about the constraints NAN imposes on our moral thinking.

XIV To begin with, remember what made Hursthouse introduce the Neurathian procedure in the first place (section V). It was the objection that the application of C might lead to ‘unacceptable results’ such as a condemnation of homosexuality or physical disablement. So what Hursthouse does is, basically, to compare C with ‘considered judgments’ about single cases, an approach that is characteristic for a view according to which no beliefs of our ethical outlook are privileged in comparison to others. It should be clear by now that NAN, by contrast, considers certain beliefs as privileged. The belief that a diseased cactus is not a good one (because a cactus is a living thing) cannot so easily be outweighed by considered judgments about single cases of good and diseased cacti, numerous as they may be. Mutatis mutandis, the same is to be expected for the ethical realm. The background assumptions of NAN grant privileges to some of our ethical beliefs in comparison to others and hence determine the changes our ethical outlook can undergo. Let me spell this out a little further. Following NAN, it is of course possible to revise the criteria for good knives, good bees, and virtuous human beings, but it is not possible to revise them independently of how we determine knives, bees, and human beings as such. Someone who does so either does not know how to apply the concept ‘X is a good F’ correctly, or does not know what knives, bees, and human beings are. Consider (again!) knives. For all those who have the appropriate background knowledge, the discussion of the criteria for good knives comes down to two questions: (a) What is the purpose of a knife? (b) What does it mean to serve this purpose well? Now, how can considered judgments about single cases enter that discussion? The answer is they can either inform (a) our concept of the purpose of a knife, or (b) our concept of what it means to serve this purpose well. However, they cannot be used to test the correctness of ‘a good knife is a knife that serves its purpose well’, for the correctness of this criterion is already established by our mastery of the relevant concepts.


Philipp Brüllmann

Accordingly, in the case of the virtues, our considered judgments can either inform (a) our concept of the specific human way of survival and reproduction, or (b) our concept of what it means to display well the kind of behavior that characteristically supports this way of survival and reproduction. But they cannot be used to test the correctness of ‘a virtuous human being is a human being who displays this kind of behavior well’. For, again, the correctness of this criterion is already established by our mastery of the relevant concepts. We have now gained some crucial information about the constraints NAN imposes on our moral thinking: NAN determines the way in which considered judgments about single cases can change our ethical outlook. And obviously, these judgments cannot be used in the way Hursthouse suggests: to test C as a ‘moral principle’. If we want to find out whether C gets off the ground as a criterion for the virtues (whether it yields correct results), we have to deny that the correctness of C is already established, which in turn means to give up the very point of NAN. So the problem is not that the reference to human nature offers no way of grounding moral judgments. The problem is that NAN makes it impossible for us to find out whether it does.32 XV NAN cannot be reduced to the claim that the virtuous person is good as a human being, for NAN is a theory about how this claim is grounded. It is a theory that invites (and commits) us to take thinking about good knives and good bees as a model for moral thinking, which means to impose some important constraints on the latter. It is conspicuous that Rosalind Hursthouse, a virtue ethicist who sees herself in the tradition of Elizabeth Anscombe and Philippa Foot, is willing to drop naturalism, if it should turn out that those character traits that make us good as human beings are completely different from those we call the virtues (194–195). But if we think it possible that naturalism might yield such a result, we have already dropped NAN, as I have tried to show. In this sense, NAN is a non-starter.33

NOTES 1. The most important proponents of NAN are Philippa Foot and Rosalind Hursthouse (see Foot, ‘Does Moral Subjectivism Rest on a Mistake?’ and Natural Goodness; Hursthouse, On Virtue Ethics) who are influenced in a general way by the work of Elizabeth Anscombe and Peter Geach (see especially Anscombe, ‘Modern Moral Philosophy’; Geach, The Virtues), and in a more specific way by Michael Thompson’s investigations into the logic of judgments about ‘life-forms’ (see his ‘The Representation of Life’). I set aside here the question of whether it is really adequate to call NAN an ‘Aristotelian’ approach. Though it is clear that Aristotle thinks that a virtuous

Good (as) Human Beings

2. 3. 4. 5. 6. 7.

8. 9. 10. 11. 12. 13. 14. 15.


17. 18. 19.

20. 21.


person is good as a human being, it is far from clear whether he intends to ground judgments about the virtues in judgments about human nature. For some reflections on how NAN transforms Aristotle, see Brüllmann, ‘Laster als natürliche Defekte?’. Taking ‘moral judgment’ and ‘to be grounded’ in a very wide sense. See, e.g., Millum, ‘Natural Goodness and Natural Evil.’ See, e.g., Hacker-Wright, ‘What is Natural about Foot’s Ethical Naturalism?’. This is true even if we take ‘grounding’ and ‘moral judgments’ in a very wide sense. All references in the main text are to this work. The second one is ‘The virtues benefit their possessor.’ Since Hursthouse asserts that the two claims are ‘interrelated’ (167), picking out one of them means to oversimplify her account. But, once again, my aim is not to offer a discussion of that account but to take some of its features as a basis for illustrating a more general point. See Gowans, ‘Virtue Ethics and Moral Relativism’ for an interesting discussion of this aspect. Cf. Woodcock, ‘Philippa Foot’s Virtue Ethics Has an Achilles’ Heel’. Cf. Foot, Natural Goodness, ch. 3, and Nussbaum, ‘Aristotle on Human Nature and the Foundations of Ethics’. In other words, Hursthouse tries to combine Foot and McDowell, which is interesting because the latter develops his ideas (in ‘Two Sorts of Naturalism’) starting from a criticism of Foot’s NAN. See Foot, Natural Goodness, 2–3; Hursthouse, On Virtue Ethics, 195. The classic text is Geach, ‘Good and Evil’. See Thomson, Normativity, 1–17 and 233–248 for an elaboration. See Rind and Tillinghast, ‘What Is an Attributive Adjective?’. I say ‘partially’ because there is room for disagreement, and it might take some deliberation to determine when exactly a knife serves its purpose well. But importantly, NAN defines the kind of disagreement to be expected and the kind of deliberation necessary for those who have the appropriate conceptual knowledge (see section XIV). See Geach, ‘Good and Evil’; Foot, ‘Moral Beliefs’ and Natural Goodness, 2–3; and (more explicitly) Thomson, Normativity, ch. 1. Roughly, the idea seems to be that someone who considers (say) ‘Is a knife that cuts well a good knife?’ an open question has a lack of conceptual knowledge. See, e.g., Foot, Natural Goodness, 7. Foot, ‘Goodness and Choice’, 133. This is an approximation. Depending on our background assumptions, we might for instance (i) add ‘she does not know how to apply the concept “X is a good F” to knives,’ or (ii) reduce the apodosis to ‘she does not know what a knife is’ (under the assumption that describing something as a knife already implies taking a normative stance). In the present context we need not decide these questions. The important point is to see that NAN defines the kind of mistakes that can be made in the evaluation of knives. See Thompson, ‘The Representation of Life’, and Foot, Natural Goodness, 30–31. See Foot, Natural Goodness, ch. 3, who associates human as animal good with the goals of survival and reproduction, but emphasizes how complex these goals turn out to be in the case of human beings. Cf. also the more explicit account of Hursthouse who spells out four ends to which evaluations of social animals as good of their kind refer (individual survival, continuance of the species, freedom from pain and enjoyment, good functioning of the social group; On Virtue Ethics, 202), and then assigns a ‘genuinely transforming effect of our [human, P.B.]


22. 23. 24.

25. 26. 27. 28. 29. 30. 31.



Philipp Brüllmann rationality’ (218) on this basic structure (instead of adding a fifth aim for the case of an evaluation of human beings). I say ‘more or less’ because in most contexts we do not judge animals as good of their kind but as good in relation to our needs and goals (‘Good dog!’). See Foot, Natural Goodness, 2–3, 38–39; Hursthouse, On Virtue Ethics, 195. It seems that the meaning of ‘good’ does undergo a change (mysterious or not) that is indicated by the fact that the opposite of ‘good’ in its non-moral use is ‘bad,’ whereas the opposite of ‘good’ in its moral use (or in one case of moral use) is ‘evil’. For a criticism of Geach’s position, see Pigden, ‘Geach on Good’. ‘Hare can call a cactus a good one on the grounds that it is diseased and dying, and choose it for that reason, but what he must not do is describe it as a good cactus, for a cactus is a living thing’ (On Virtue Ethics, 195). Hursthouse’s reference is McDowell, ‘Two Sorts of Naturalism’. For some remarks, see Harman, ‘General Foundation versus Rational Insight’. This illustration, which somehow undermines the distinction between epistemic and motivating reasons, seems problematic to me (see section XII), but I cannot pursue the issue here. See McDowell, ‘Two Sorts of Naturalism’. It thus seems that the transition from T 1 (good knives) and T 2 (good bees) to T 3 (virtuous human beings) is not as obvious on Hursthouse’s account as the theory of NAN seems to imply (see section IX). The point is, within the framework of NAN, judgments about the virtues are just as independent and objective as judgments about human nature. But the same does not seem to hold for the situation of a Neurathian procedure where the former judgments are not reduced to the latter. Here, judgments about the virtues are independent and objective to the extent that judgments that survive reflective scrutiny have these properties. Cf. Gowans, ‘Virtue Ethics and Moral Relativism’, 406–408, for some remarks on NAN and the Neurathian procedure. It is important to note that this result does not beg the question with respect to what counts as ‘moral’. It suffices to assume that there are certain content restrictions as to when a judgment is a moral judgment, whatever those restrictions may be. Cf. Hurthouse’s worry that naturalism might ‘yield far too many horrific [results, P.B.] for us to count it as validating ethical beliefs at all’ (On Virtue Ethics, 194). I would like to thank Anne Burkard, Benjamin Kiesewetter, and Thomas Schmidt for helpful comments on an earlier draft of this chapter.

REFERENCES Anscombe, G.E.M., ‘Modern Moral Philosophy’, Analysis 33 (1958), 1–19. Brüllmann, Philipp, ‘Laster als natürliche Defekte? Philippa Foot’s Natural Goodness und die Transformationen der aristotelischen Ethik’, in Transformation: Ein Konzept zur Erforschung kulturellen Wandels, ed. H. Böhme et al. (Munich: Wilhelm Fink, 2011), 213–238. Foot, Philippa, ‘Does Moral Subjectivism Rest on a Mistake?’, Oxford Journal of Legal Studies 15 (1995), 1–14. ———, ‘Goodness and Choice’, Virtues and Vices and Other Essays in Moral Philosophy (Oxford: Basil Blackwell, 1978), 132–147. Originally published in Proceedings of the Aristotelian Society 35 (1961, supp.), 45–60. ———, ‘Moral Beliefs’, in Virtues and Vices and Other Essays in Moral Philosophy (Oxford: Basil Blackwell, 1978), 110–131. Originally published in Proceedings of the Aristotelian Society 59 (1958/59), 83–104.

Good (as) Human Beings 113 ———, Natural Goodness (Oxford: Oxford University Press, 2001). Geach, Peter, ‘Good and Evil’, Analysis 17 (1956), 35–42. ———, The Virtues (Cambridge: Cambridge University Press, 1977). Gowans, Christopher W., ‘Virtue Ethics and Moral Relativism’, in A Companion to Relativism, ed. Steven D. Hales (Malden, MA: Blackwell, 2011), 391–410. Hacker-Wright, John, ‘What is Natural about Foot’s Ethical Naturalism?’, Ratio (new series) 22, no. 3 (2009), 308–321. Harman, Gilbert, ‘General Foundation versus Rational Insight’, Philosophy and Phenomenological Research 63 (2001), 657–663. Hursthouse, Rosalind, On Virtue Ethics (Oxford: Oxford University Press, 1999). McDowell, John, ‘Two Sorts of Naturalism’, in Virtues and Reasons: Philippa Foot and Moral Theory, ed. R. Hursthouse, G. Lawrence, and W. Quinn (Oxford: Clarendon Press, 1995), 149–179. Millum, Joseph, ‘Natural Goodness and Natural Evil’, Ratio (new series) 19, no. 2 (2006), 199–213. Nussbaum, Martha, ‘Aristotle on Human Nature and the Foundations of Ethics’, in World, Mind, and Ethics: Essays on the ethical philosophy of Bernard Williams, ed. J.E.J. Altham and R. Harrison (Cambridge: Cambridge University Press, 1995), 86–131. Pigden, Charles R., ‘Geach on Good’, Philosophical Quarterly 40 (1990), 129–154. Rind, Miles, and L. Tillinghast, ‘What Is an Attributive Adjective?’, Philosophy 83 (2008), 77–88. Thompson, Michael, ‘The Representation of Life’, in Virtues and Reasons: Philippa Foot and Moral Theory, ed. R. Hursthouse, G. Lawrence, and W. Quinn (Oxford: Clarendon Press, 1995), 247–296. Thomson, Judith Jarvis, Normativity (Chicago: Open Court, 2008). Woodcock, Scott, ‘Philippa Foot’s Virtue Ethics Has an Achilles’ Heel’, Dialogue 45 (2006), 445–468.


Attachment Theory, Character, and Naturalism Edward Harcourt

The subject matter of this chapter falls within the triangle marked out by three Aristotelian ideas—human nature, human excellence, and human well-being. The chapter is a highly programmatic attempt to introduce some material from developmental psychology—specifically, from attachment theory—and to explain why philosophers working somewhere within the Aristotelian triangle have reason to take more of an interest in it than they do now. I am assuming of course that attachment theory is true: if it isn’t, or doesn’t make some reasonable claim to be believed, there’s no reason for anyone to take an interest in it. But if you grant that much—and it is one of the leading theoretical orientations in developmental psychology1—there are lots of true theories: why should neo-Aristotelians be interested in this one? Attachment theory is a theory of child development. Indeed properly speaking it is a theory of human development, but—partly because children are easier to study than adults, partly because childhood experience may be especially important in making us the way we are—attachment theorists have taken a special interest in the early years. This is already one, rather general, reason why neo-Aristotelians should be interested in the theory, for Aristotle’s ethics is (in part) a developmental theory: it aspires to provide not only a theory about what human excellence is, but a theory about how we acquire or fail to acquire it. Aristotle himself says that because good character is produced by habituation, it makes no small difference . . . whether we form habits of one kind or of another from our very youth; it makes a very great difference, or rather all the difference.2 So simply insofar as it has much to say about ‘our very youth’, attachment theory stands a chance of telling us what Aristotle tells us very little about, namely how we get from there to here, as well as merely that we sometimes do. But there is a more specific reason than this. Ethics in the neo-Aristotelian mould sees itself as a naturalistic undertaking: that is, it seeks to locate ethical life in the world as made intelligible to us by natural science. Moreover there is reason to think that the proper form of such an undertaking is to

Attachment Theory, Character, and Naturalism


display the continuity between our second and our first, or between our ethical and our psychobiological natures.3 Now attachment theorists claim that some human traits described by the theory arise through natural selection. Those traits, then, would belong to our first natures, or to ‘an account of human beings which is to the greatest extent possible prior to ideas of the ethical’.4 But attachment theory is also a taxonomy of psychological dispositions, plus a theory about why people (children, adults) have the disposition(s) of that sort which they do. Furthermore, these psychological dispositions appear to stand in an explanatory relation to some traditional traits of character, that is, to some virtues and vices. Attachment theory therefore looks as if it is well equipped to put some empirical flesh on the bones of Aristotelian naturalism by making vivid the continuity I have said this variety of naturalism demands. So far it looks as if I envisage the flow of ideas as being entirely from attachment theory to neo-Aristotelian ethics. But that’s not so. As far as I know, attachment theorists are unaware of the ways in which different forms of naturalism are debated within ethics and might be surprised at the thought that their work had anything to do with ethics. But perhaps precisely because they think that what they are up to is just psychology or evolutionary biology (i.e., some sort of natural science), some of them have apparently signed up unawares to a version of ethical naturalism that goes much further than the continuity thesis I mentioned earlier. They are in distinguished philosophical company here—the late Philippa Foot’s, for example.5 But I shall argue that Foot’s ambitious version of neo-Aristotelian naturalism and the versions of attachment theory that unwittingly subscribe to it both run into difficulties. Thinking through the connections between attachment theory and neo-Aristotelian ethics, then, should be a way both of reining in some of attachment theory’s own more extravagant theoretical ambitions, and of demonstrating both the prospects (the continuity thesis itself) and some of the limits of naturalism in ethics.

I I shall return to the continuity thesis shortly, but let me focus for now on the ambitious version of neo-Aristotelian ethical naturalism, as exemplified by Foot. In this version of naturalism, claims about human virtues and vices, or excellences and defects, are identified with claims about the way human beings should be if they are to be good of their kind, or properly suited to lead our characteristic species life. Here—in contrast to the kind of naturalism that Moore is famous for opposing, a species of reductionism in which moral properties are, in one way or another, identified with ‘natural’ ones—it isn’t a matter of identifying one kind of property with another: only one kind of property, the excellences and defects themselves, is ever under discussion. Nor is it a matter of deriving surprising claims about which properties are


Edward Harcourt

excellences or defects in humans: it’s assumed we will more or less agree on that at the start. Rather it’s a matter of arguing that these interesting properties are the virtues and vices because they have a further property, that of playing a certain kind of role, namely, that they are necessary (in the case of the virtues) for us to lead the kind of life that, as members of that species, we’re supposed to lead. Just as it is an ‘Aristotelian necessity’6—a necessity that ‘determine[s] what it is for members of a particular species to be as they should be’—‘for plants to have water or birds to build nests’,7 so ‘for human beings the teaching and following of morality is something necessary. We can’t get on without it’; again, ‘getting one another to do things without the application of physical force [and which morality accomplishes] is a necessity for human life’. Correspondingly, a species member who does not do what it is necessary for the species to do is ‘naturally defective’, so immoral human beings are defective in just the same sense as birds who fail to build nests, or owls who cannot see in the dark. And because it’s a matter of plain fact that virtues are necessary for us, it is also a matter of plain fact that a human being with a given virtue is excellent, or with a given vice defective. Thus Foot’s ambitious naturalism is designed to fulfill the cognitivist meta-ethical ambition common to many ethical naturalisms. Contrast the continuity thesis, whose relation to cognitivism is looser and won’t be discussed further here. This ambitious form of naturalism ought to interest empirical investigators who address questions about the way human beings should be or— transposing the same idea into a developmental idiom—about optimal development,8 attachment theorists included. But there’s an interesting difference. There’s no difficulty in getting moral philosophers to recognize that what they’re working on are virtues and vices, that is, ethical notions. The controversy arises when ambitious naturalists try to get people to agree that these ethical notions are also ‘natural’ notions—that is, that virtue consists is a perfection of our first nature, or of our nature as we can make it intelligible to ourselves ‘to the greatest possible extent prior to ideas of the ethical’. With developmental psychologists the sticking point is different: many of them have no difficulty at all in agreeing that the notion of optimal development—human beings turning out as they should—is a psychobiological one. The surprise to them is that this notion is also ethical. But suppose they get over their initial surprise, and suppose the concept of optimal development, as deployed by attachment theorists, really is an ethical one. Isn’t the very fact that these developmentalists have been studying it unawares, using empirical methods, for all that time, evidence for the truth of ambitious naturalism? I shall argue (in section V) that despite appearances, there’s no support for ambitious naturalism to be derived from developmental psychology, at least in the form of attachment theory. Roughly, insofar as the dispositions in which attachment theory deals are ‘adaptive’—even if some theorists have claimed otherwise—this is not in the sense that they are the result of natural selection (and so belong to our first natures), but rather in the sense that they are favorably related to human social life—and perhaps

Attachment Theory, Character, and Naturalism


to a particular form of it rather than to human social life in general. Insofar as more is claimed, attachment theory overreaches itself, in a way that I shall suggest ambitious naturalism in ethics does, too.

II Attachment theory was first formulated by a dissenting psychoanalyst, John Bowlby.9 Partly inspired by the effects of maternal deprivation as a result of the wartime evacuation of children,10 he criticized classical Freudian ideas about human nature—he rejected, for example, the idea that infants are just in it for what they can get out of it (the ‘secondary-drive theory’), that is, that infants form attachments to particular others simply as a means to getting food or warmth.11 He also wanted to test ideas about infancy using some of the methods of science and social science (e.g., large sample sizes and multiple observers), rather than relying only on what goes on in the consulting room. The theory was developed by two North American women, Mary Ainsworth and Mary Main, and has developed further since. It now has its own journals and is part of the academic mainstream, though of course it has its enemies as well as its friends. As to its main claims, attachment theorists distinguish between the ‘normative’ and the ‘individual-difference’ components of the theory,12 the ‘normative’ component—that is, that part of the theory that deals with what is common to all (or almost all) human beings—being the more fundamental. The normative component (which is designed to apply also to other primates and perhaps to other mammals) notes that newborn offspring of various species, including our own, cannot survive unaided. Attachment theory’s hypothesis is that the ‘attachment system’ in infants serves when activated to maintain proximity to an attachment figure who is able to protect the infant from threats to its survival. Everyone has heard about the unignorable pitch of a baby’s screams, but don’t forget about the unignorable charm of a baby’s smile: the attachment system’s behavioral expressions are various and maintain proximity in correspondingly various ways, including smiling and vocalizing (enlisting the attachment figure’s interaction), clinging (the attachment figure can’t get away), crying, and approaching and following. Thus we attach ourselves in infancy to a special attachment figure and by doing so tie them to us, enlisting the attachment figure’s responses; if the infant responds in turn to those responses, a virtuous cycle is set in train that will help see to it that it survives. This is not to say that the attachment system is the only trait that serves to maintain proximity: a specific ‘caregiving system’ among parents, and non-behavioral characteristics of infants such as endearing ‘babyish’ features, may also play this role. Nonetheless maintaining proximity with the attachment figure enhances the offspring’s chances of making it through to reproductive age, and thus enhances their genes’ chances of being replicated. So it is plausible that each one of a cluster of


Edward Harcourt

traits that, working together, serve to maintain infants’ proximity to attachment figures has emerged under pressure of natural selection.13 If evolutionary considerations explain why humans are equipped with the attachment system, isn’t this back to the bit of Freud that Bowlby rejected, that is, that human infants form bonds with their special others in order to survive? No. If there’s an instrumentality here, it’s at the level of the unit on which selection operates: the gene. What helps genetic replication is that individual species members are ‘inherently motivated’14 to form attachment bonds to certain other species members. Dropping the jargon, the formation of such bonds is, like drinking or sleeping, something we are disposed to do for no reason (let alone any further reason)—it is simply our nature to do it, so looking for an individual’s reasons for doing it (which gives rise to the Freudian thought that the reasons are instrumental) is a mistake.15 Once exercised, however, the disposition gives rise to many goods (intimacy, reassurance, the pleasures of touch), which in turn are things we do ‘for their own sakes’ (i.e., for a reason but for no further reason). Thus the behavior characteristic of attachment bonds—exclusivity, but also physical contact for its own sake, staying close especially in times of threat or distress, heightened anxiety at separation, grieving when an attachment figure dies—shares or (better) shares the general outlines of what we recognize in humans as relatedness to another that is valued or pursued for its own sake.

III Although almost all human infants form attachments, not all attachments are alike in quality, and this fact and its explanation forms the subject-matter of attachment theory’s ‘individual-difference component’. The first measure of attachment quality was the ‘Strange Situation’, developed by Mary Ainsworth.16 This test is administered at either 12 or 18 months, to one parent–infant pair at a time—and note that at this age, at least, infants can fall into different attachment types with respect to different parents. To simplify, the Strange Situation proceeds as follows: the mother (let’s say) enters an unfamiliar room with the infant and settles it down to play with some toys. A ‘stranger’ (an unfamiliar research assistant) then enters who starts to play with the infant, and after a short time the mother tells the infant she is leaving, leaves for three minutes, and then comes back. The separation and reunion is repeated, with the stranger absent.17 The infant’s behavior is recorded throughout. Observed infant behavior falls into three recognizable patterns. (There is also a fourth pattern that was theorized later, but I omit that, again for simplicity’s sake.) In pattern B, the infant is overtly distressed when the mother leaves, then seeks proximity with her when she comes back and is comforted by it, and then resumes playing. In pattern A, the infant does not express distress when the mother leaves, though it displays other signs of distress such as more rapid breathing and heart rate, suggesting

Attachment Theory, Character, and Naturalism


it is suppressing the expression of distress rather than simply indifferent. The infant doesn’t show a preference in play as between the mother and the stranger and is then also relatively indifferent when the mother returns (e.g., looks or turns away from her) and ‘if picked up . . . makes no effort to maintain the contact’. In pattern C, the infant may be ‘clingy’ toward the mother and uninterested in the toys even before she leaves, is immoderately distressed when she does leave, but when she comes back doesn’t calm down and exhibits ‘furious clinging’—‘seeking contact, then resisting contact angrily once it is achieved’. These three attachment types are labeled ‘insecure-avoidant’ (type A; also ‘deactivating’), ‘secure’ (type B), and ‘insecure-ambivalent’ or ‘insecure-resistant’ (type C; also ‘hyperactivating’).18 But though infants were the first to be systematically classified into attachment types, attachment theory does not apply only to infants: on the contrary, it is supposed to apply across the life span. Accordingly, other tests, based both on observation and on interview data, have been developed for various later stages of life (e.g., the Cassidy-Marvin system and the Preschool Assessment of Attachment for preschool-age children,19 and the Adult Attachment Interview20), with roughly the same number of attachment classifications, often with similar names to those used in the Strange Situation. Though there is some debate about which age-specific test is the most reliable for a given age, and about the extent to which different agespecific tests keep track of the same characteristics, there is an evident family resemblance between the criteria for these age-specific tests. The second aspect of attachment theory’s individual-difference component that I want to draw attention to concerns the further characteristics with which secure and insecure attachment are associated. These associations have both a diachronic and a synchronic dimension: the diachronic dimension concerns the characteristics predicted, at a greater or lesser distance in time, by a secure infantile attachment history; the synchronic dimension concerns the characteristics contemporaneously associated with secure attachment as measured by the test(s) appropriate to the life-stage in question. On the whole the contemporaneous associations are stronger than the predictive ones,21 and the predictive associations are weaker the longer the distance in time and the further removed the ‘outcome domain’ is from quality of relations with the infantile attachment figure him- or herself. This is thanks to the fact that attachment classifications can shift quite early in life (e.g., an infant who is securely attached to its mother aged 12 months may become insecure if the mother suffers from post-natal depression following the birth of a second child; and an insecurely attached child may become securely attached to someone if it is fortunate in its adoptive or foster parents).22 And if they don’t shift, this is likely to be not only thanks to the infantile attachment history, but thanks to the persistence of the factors—such as warm relations with parents—that also explain the infantile attachment classification. Thus if there is a predictive relation to later characteristics, it is likely to be mediated by a variety of further factors.23


Edward Harcourt

To summarize some recent findings,24 infant attachment security predicts a good relationship at least a year later between the child and the attachment figure, considered in terms of ‘enthusiasm, compliance’, and ‘less frustration and aggression’ during shared tasks; secure attachment also predicts harmonious caregiver–child relations over longer periods in the presence of continued sensitive caregiver behavior. Secure attachment in adulthood, meanwhile (whether or not itself predicted by secure attachment in infancy), is correlated with greater sensitivity to one’s own children’s needs and ‘more warmth and appropriate structuring of learning tasks’25 and, in attachments to peers, a capacity inter alia to admit vulnerability and need for the other without ‘continually worrying about the attachment figure’s availability’.26 Moving to the next widest outcome domain, that of other relationships, children with secure attachment histories have less conflictual relationships with peers from preschool to 7 years,27 are less dependent on teachers in preschool,28 are less dependent on counselors at summer camp aged 10, and are more sociable with unfamiliar adults.29 By contrast, the insecurely attached 4-year-old boys exhibit more ‘aggressive, assertive, controlling and attentionseeking behavior than their securely attached counterparts’. Finally, attachment theory argues for a connection between attachment security and broader personality traits. The Minnesota Study of Risk and Adaptation from Birth to Adulthood30 argues for ‘significant associations between early attachment security and personality characteristics throughout childhood and adolescence . . . [including] self-esteem, agency and self-confidence, [and] positive affect’. Securely attached children aged 6 describe themselves generally in positive terms but are better at admitting flaws—insecurely attached either are more negative about themselves or do not admit flaws. There is also an important contemporaneous association between secure attachment and the capacity for emotion-regulation, including in adulthood,31 and between secure attachment and psychological understanding (they are more ‘proficient at identifying emotions in others, . . . especially . . . negative emotions and mixed feelings’). Thompson concludes that ‘children with a secure attachment history are capable of developing and maintaining more successful close relationships, especially with parents and with peers, than are insecure children; they develop a variety of desirable personality qualities in childhood and adolescence [including ‘social problem-solving skills’]; they are more likely to exhibit constructive forms of emotionality and emotion self-regulation; and they exhibit more positive self-regard’.32

IV I now want to use my sketch of attachment theory to explain why the theory is well placed to flesh out what I earlier called the continuity thesis. Attachment dispositions face both backward toward the psychobiological (thanks to attachment theory’s normative component) and (thanks to its individual-

Attachment Theory, Character, and Naturalism


difference component) forward toward the ethical. As a result, the more we get to understand attachment, including what explains it and what it explains, the better we should be able to bring the continuity idea to life and, thus, to take it beyond a picture of how some naturalistically minded philosophers think things ought to turn out. As regards the forward-facing connection, it is striking how generally the characteristics with which secure attachment is associated, whether predictively or contemporaneously, seem to be positive characteristics: the capacity to relate harmoniously to others, co-operativeness, ‘positive affect’, self-esteem, the capacity to be realistic about oneself and to tolerate one’s own imperfections, are all apparently more worth having than those typically possessed by the insecurely attached or by those with insecure attachment histories (dependence, attention-seeking, low self-esteem, limited capacity for symbolic play, and so on). It’s thus unsurprising that a great deal of effort is expended, by parents, educators, therapists, and others, on trying to get children into attachment category B and making sure they stay there. However, the characteristics in question are ill-assorted. Some seem either to be virtues or to imply virtues. Here I am thinking of the capacity for friendship; the capacity to form a realistic appraisal of one’s own excellences and defects, which is surely a virtue, though perhaps not an Aristotelian one;33 and the capacity to offer help and to ask for help when you need it. Cooperativeness, meanwhile, if it is not itself a virtue, surely implies the traditional virtues of trust and, unless the co-operation is very short-term, also honesty and fidelity to promises. Other characteristics I mentioned seem like more general character-traits that have sometimes been argued (emotional self-regulation34) to go with or (self-esteem35) to underlie the virtues, while some are traits that may be as it were adverbially related both to virtues and to vices (aggression—good in fighting the local authority to get your child a place at school, not so good in bullying a colleague into accepting an unfair workload; the same goes for ‘positive affect’, though one really needs to know more about what’s meant—if it means ‘the capacity to enjoy life’s goods to the full’, arguably it is itself a virtue). Empathy is another tricky one, depending, for example, on whether one thinks cruel people genuinely possess it. There is a lesson to be learned both from the evaluative asymmetry (as we might put it) between secure and insecure attachment and from the heterogeneity of the positive traits. The evaluative asymmetry seems to show that secure attachment stands in a privileged relation to the virtues (and insecure attachment to the vices). But how close is the relation? At one extreme, the answer would be that secure attachment is virtue; or perhaps that it’s the disposition that underlies and unifies the virtues. That would be a highly ambitious direction for neo-Aristotelian ethics to try to go in.36 On this view, attachment theory would not be an intermediate level of theory that merely mediates between the biological and the ethical, because the individual-difference component of attachment theory already is a theory about the ethical, that is, a theory that stands to virtue and vice as (say)


Edward Harcourt

Plato’s tripartite moral psychology does. On a more modest view, secure attachment would belong at a level intermediate between the biological and the ethical, with secure attachment occupying the place occupied (roughly) by Aristotelian ‘natural virtue’,37 a disposition that is not yet virtue but may turn into it if properly cultivated. That certainly fits the fact that attachment dispositions are relatively unstable in the early years. One might also take this more modest thought in a skeptical direction: if secure attachment is clearly in some sense privileged among the attachment dispositions, the fact that it’s not straightforwardly related to the virtues (after all the previous list leaves a great many out) could be used to explain why people are not typically unified in respect of the virtues (honest without being generous, or honest to colleagues but dishonest to lovers), or to challenge the Aristotelian thought that we’re ‘made for virtue’.38 I can’t develop these lines of thought here, let alone adjudicate between them, but this doesn’t matter for my present purpose: whichever way these lines of thought are developed, we are going to end up with a richer and more realistic version of the continuity thesis than we have so far.

V I want to comment now on the backward-facing connection, from attachment to biology, and here my comments are of a more cautionary sort. Attachment theorists can hardly fail to think that attachment, meaning the disposition simply to form attachments, is ‘adaptive’. But when they go on to specify that the ‘adaptive goals’ are ‘the facilitation of social integration . . . problem solving ability, flexibility, . . . and the ability to use adult assistance’, or that ‘toddlers of 12 to 18 months of age who experience an attachment relationship that supports mastery competence are more adaptive than children who experience an attachment relationship in which exploration apart from the parent is difficult to achieve’,39 it is clear that they don’t mean attachment simpliciter but rather secure attachment. Again, Londerville and Main write that it is ‘adaptive’ to form a secure attachment in year one because it increases the likelihood that a second ‘positive adaptation’, for example, ‘the capacity for cooperation with the mother to gain needed help in problem solving’, will develop in the second year of life.40 Now of course it is unhelpful to have too many people around who don’t comply with social or moral rules, who can’t cooperate, who can’t ask for help when they need it, and who are also poor at striking out on their own (all characteristics associated with secure attachment). But we must be careful: ‘adaptive’ is a word borrowed from evolutionary theory and appears (in the previous quotations) to be used to smuggle in the idea that secure attachment is selected for. Depending on how close one takes the connection between secure attachment and the virtues to be, it might indeed be smuggling in even more—virtue is selected for. Thus some attachment theorists appear to make a strong con-

Attachment Theory, Character, and Naturalism


nection between virtue and our biological natures analogous to that aspired to by ambitious ethical naturalists: we can get straight to the privilege of secure attachment by reflection on our first natures. However, it’s surely a mistake to say secure attachment is selected for. For one thing, if it were, it would be hard to explain why so many people fall into the various insecure attachment categories (between 35% and 45% of the population41)—why would selective pressure not have been better at rendering them extinct? One might think that secure attachment is more effective in enabling people to reach reproductive age, but this seems only very marginally to be so. It’s true that category C—insecure-ambivalent—is associated with certain sorts of risk-taking and self-destructive behavior. (The idea is that you can attract a reasonably attentive mother by saying ‘mummy’; with a distracted or neglectful mother you need to start climbing the bookshelves.) But almost all the children who display these traits make it to reproductive age anyway, presumably because their ‘maximizing’ (attention-seeking) strategy is effective. Indeed, one reason why categories A and C are perpetuated is that they are born to parents who occupy these categories themselves.42 The second point is that difference of attachment type is explained not by natural selection but by the interaction between the evolutionarily determined generalized attachment disposition and the parental environment— among other things. The difference between secure and insecure attachment is not like the difference between a picture that’s securely fixed to the wall and one that’s precariously hanging off a nail—whatever the terminology may suggest, secure/insecure isn’t a distinction between degrees of attachment. If it were it would be very hard to explain why the parents of insecurely attached children have such a big effect on them. Suppose the postman calls every day, he is a reliable and pleasant figure and he always gives you a smile—but if he is no more than that, why should the character of the postman have any effect on your character? And the same would be true of insecurely attached children, if insecurely attached meant ‘not very attached’. But attachments do not work like this. If you have a bad accountant, you can sack them and get another one. But children cannot change their parents, so even a cold or inconsistent parent gives you a better chance of making it out of infancy than no parent. So, instead of changing parents, you change. Thus for example, the insecure-avoidant type is an adaptation to indifferent or cold parents because, the thought is, such parents would be annoyed, and thus more rejecting, of a child who expressed its needs more overtly; the insecure-ambivalent child on the other hand has experience of parental interest, so when this is replaced by neglect it will go to extremes to get it back. But all attachment types are children’s adaptations to different parental environments43—adaptations the more complexly mediated the older the child and thus the more capable of complex forms of learning, comparison of goods and so on—aiming at the creation and ‘maintenance of a degree of proximity with the caregiver over time’.44 So it is a mistake to single out


Edward Harcourt

secure attachment as a better strategy for making it to reproductive age. If there is a sense in which secure attachment is ‘adaptive’ (or ‘optimal’), it is that the cost of adapting to a rejecting parental environment will be a sacrifice in intimacy, for example, and enjoyable mutual interactions: adaptations to a good environment are likely to be better, in ways that evolutionary biology and attachment theory itself are thoroughly unsuited to describe, than adaptations to a poor one.

VI I want finally to apply the foregoing reflections to raise some questions for Philippa Foot’s claim that vice is a natural defect in humans and for the familiar idea that the virtues stand in a privileged relation to well-being. Whatever its relation to virtue, secure attachment appears to be in some sense best for the person whose disposition it is. Perhaps this is most clearly argued in connection with infancy itself. The salient characteristics of secure attachment in infancy—the freedom to express distress when it is felt without fear of rejection and in the expectation of comfort, the capacity for the pleasures of ‘affective sharing’45 and warm physical contact, and the freedom to become absorbed in the environment—are real human goods. These both reflect the real goods of the kinds of relationship that give rise to secure attachment, and—especially in the case of the freedom to become absorbed—make available to the infant a great many other goods in their turn. They are also goods that are, to varying degrees and in varying combinations, unavailable to the insecurely attached infant. Thus quite independently of what, if anything, secure attachment predicts about characteristics later in life, to describe secure attachment in infancy is to describe a good infancy in the sense that parallels ‘a good childhood’ or ‘a good life’. But if the point is especially vivid in connection with infancy, the capacity for good close relations in later life (which, e.g., balance intimacy and autonomy) is also associated with secure attachment. One need only remind oneself of the number of people who refer themselves for psychotherapy because they find themselves unable to enjoy those goods to gauge the privilege of secure attachment in relation to well-being.46 This privilege plays well—as far as it goes—for the Aristotelian association between virtue and well-being if secure attachment also has a privileged relation to the virtues. I have also suggested that secure attachment enjoys another kind of privilege, in the sense that the characteristics associated with it seem desirable in a way those associated with insecure attachment are not. But there is surely an element of cultural relativity here. One only has to switch context to, say, ancient Sparta—of legend if not of fact—for it to be quite probably better (for me) to be insecurely attached: think of the oft-cited insecure-avoidant trait of precocious self-reliance, useful if one has to spend days on end on solitary sentry duty. One can make the same point for insecure-ambivalent

Attachment Theory, Character, and Naturalism


attachment: in a war zone, where real threats are more or less constant, insecure-ambivalent unwillingness to allow distance from the attachment figure is the more ‘adaptive’ characteristic.47 If secure attachment is privileged in respect of suiting us better to social life, the privilege is thus surely relative to the more or less stable circumstances in which we live. Note however that this is not to endorse cultural relativism: one of the awful things about ancient Sparta might be that it prized the reproduction of insecureavoidant types, thus leading many of its citizens to miss out on the real goods of warm personal relations; the same goes mutatis mutandis for war zones. But if secure attachment bears a privileged relation to some virtues, the relativity raises a problem for the idea that these virtues are necessary for our species life, or that lacking them is a ‘natural defect’, for it looks as if, as long as circumstances are imagined to be appropriately different, some version of our species life could be carried on just as well if the distribution of insecurely attached people in the population were the same as the distribution of securely attached people here and now. The point about relativity can be pressed further. If insecure attachment would be optimal for a majority of the population in radically challenging or threatening circumstances, it is surely useful in some of the population even in our circumstances. Although our circumstances are more or less stable, our social world is sufficiently complex to make it likely that some division of labor—made possible by the variety of attachment dispositions in a given population—is necessary to the form in which we, locally, carry on our species life. Thus it is surely good in our own fortunate though imperfect circumstances to have some people around who are risk-takers (and so presumptively insecure-ambivalent) and some who are precociously self-reliant (presumptively insecure-avoidant): it is not obvious that the full range of goods that are available to humans would be realized in a society in which everyone was cooperative, affectionate, and compliant. Thus it might not be that the only human analogue of Foot’s naturally excellent wolf who hunts with the pack is the cooperative person who sits attentively round the committee table: if the naturally excellent are those who have the characteristics necessary to sustain our species life, then granted the point about the division of labor, this description might net not only the good committee person but the odd person who angrily storms out of meetings (or simply, never attends meetings because they are hatching a plan on their own). The real human equivalent of the lone wolf or the night-blind owl would rather be the rare human being who has no disposition to form attachments—a defect indeed, but whose absence seems to leave just about everything open as far as virtues and vices are concerned. At the very least, the conclusion to draw is that since our species life can be carried on in circumstances that vary greatly in respect of stability and the presence of threats, there is no single attachment disposition (or single distribution of different attachment dispositions among a population) that is necessary for us to do so, and so no attachment disposition48 which is per se


Edward Harcourt

a natural defect. So far that skeptical point says nothing about virtue either way: the extent to which virtue comes in depends on the strength of the association between virtue and secure attachment. If secure attachment does have a privileged relation to virtue, it looks as if vices can’t be natural defects because insecure attachment, too, is an ingredient in the mix necessary for our species life—it would be as impossible to sustain if no one had that as if no one was securely attached. But perhaps secure attachment doesn’t have a privileged relation to virtue. If that is so, then granted the apparently privileged relation of secure attachment to well-being, that would make trouble for the association between virtue and well-being, though perhaps that connection is in trouble anyway.49 In any case, virtue could be underpinned psychologically by either a secure or an insecure attachment disposition (perhaps depending on the virtue, perhaps depending on the circumstances): the good committee person and the awkward individualist who doesn’t turn up for meetings might, though in different attachment categories, both be virtuous. Whether secure attachment does bear a privileged relation to virtue, however, awaits a proper investigation of what in the way of virtues secure attachment is and is not correlated with—a question that attachment theorists may not ask in so many words, but on which their data bear in a multitude of ways. NOTES 1. See Schaffer, Key Concepts in Developmental Psychology, 160. 2. Aristotle, Nicomachean Ethics, 1103b24, 1743. 3. For senses of ‘naturalism’, see McDowell, ‘Two Sorts of Naturalism’, and Bernard Williams, ‘Naturalism and Genealogy’, esp. 148–150. For the idea of second nature, which but for the term itself McDowell ascribes to Aristotle, see McDowell, ‘Two Sorts of Naturalism’, 184, and Mind and World, 84. I take it that to display the continuity of our second with our first natures would be to relate ‘human ethical life . . . to the rest of human nature’, that is, to answer the ‘recurrent naturalist question’ (‘can we explain . . . the phenomenon in question in terms of the rest of nature ?’) in the ‘special form’ it takes in the case of our ethical lives (Williams, ‘Naturalism and Genealogy’, 154, 150). See also Williams, ‘Replies’, 203 (‘a conception of ethics . . . continuous with our understanding of human beings in other respects’). 4. Williams, ‘Naturalism and Genealogy’, 154. 5. See Foot, Natural Goodness. 6. Foot, Natural Goodness, 15. 7. Foot, Natural Goodness, 15. 8. I take the phrase from Jay Belsky, ‘War, Trauma and Children’s Development: Observations from a Modern Evolutionary Perspective’. 9. For a very brief history, see Holmes, Exploring in Security, 3–5; for Bowlby and psychoanalysis, see Holmes, Attachment, Intimacy and Autonomy. 10. Robertson and Bowlby, ‘Responses of Young Children to Separation from Their Mothers’. 11. Freud, ‘Five Lectures on Psycho-Analysis’. 12. See Simpson and Belsky, ‘Attachment Theory within a Modern Evolutionary Framework’, 136.

Attachment Theory, Character, and Naturalism


13. In this paragraph I am indebted to Cassidy, ‘The Nature of the Child’s Ties’. 14. Cassidy, ‘The Nature of the Child’s Ties’, 5. 15. For discussion of things we do without doing them for a reason (whether for further reasons, or for their own sake), see Alvarez, Kinds of Reasons, 111 ff. Thanks to Luke Brunning for drawing my attention to this. 16. See for example Ainsworth and Wittig, ‘Attachment and Exploratory Behavior in One-Year-Olds in a Strange Situation’. 17. Solomon and George, ‘The Measurement of Attachment Security’, 386. 18. Further variant terminology is also used: see Holmes, Exploring in Security, 3. Quotations in this paragraph are from Weinfeld, Sroufe, Egeland, and Carbon, ‘Individual Differences in Infant-Caregiver Attachment’, to which I am more generally indebted. See also Solomon and George, ‘The Measurement of Attachment Security’. 19. Solomon and George, ‘The Measurement of Attachment Security’, 297, 299. 20. Feeney, ‘Adult Romantic Attachment’. 21. Thompson, ‘Early Attachments and Later Development’, 361. 22. Holmes, Exploring in Security, 4. 23. See Thompson, ‘Early Attachments’, 343. 24. See Thompson, ‘Early Attachments’, 348ff. 25. Simpson and Belsky, ‘Attachment Theory within a Modern Evolutionary Framework’, 145. 26. Mikulincer and Shaver, ‘Adult Attachment and Affect Regulation’, 507. 27. Thompson, ‘Early Attachments’, 355. 28. Sroufe, ‘Infant–Caregiver Attachment and Patterns of Adaptation in Preschool’. 29. Thompson, ‘Early Attachments’, 355. 30. Sroufe, The Development of the Person: The Minnesota Study of Risk and Adaptation from Birth to Adulthood. 31. Mikulincer and Shaver, ‘Adult Attachment’, 503–531. 32. Thompson, ‘Early Attachments’, 357–361. 33. Cp. Aristotle’s virtue of truthfulness (Nicomachean Ethics, 1779). 34. See Aristotle, Nicomachean Ethics, 1106b, 1747. 35. See, for example, Homiak, ‘Virtue and Self-Love in Aristotle’s Ethics’, and Chazan, The Moral Self. 36. Compare Chazan’s attempt to show that the Aristotelian virtues are underpinned by (roughly) good object relations, The Moral Self, esp. 63–154. There are clearly affinities between Bowlby’s work in attachment theory and the Kohutian psychoanalysis Chazan draws on. 37. Aristotle, Nicomachean Ethics, 1144b, 1807. 38. It would thus answer Bernard Williams’s complaint against Aristotle’s and Plato’s own moral psychologies that they do not answer to the demands of the continuity thesis because they build too much of what they are trying to explain into the psychology that supposedly explains our ethical lives. (‘Aristotle’s psychology, despite its richness and elaboration, can seem ethically superficial. . . . [We need] a psychology that is less moralized, less adapted already to the demands of the ethical’; Williams, ‘Replies’, 202.) But I take it that Williams also has in mind the thought that telling the continuity story in a non-circular way would display the fact that moral considerations do not have the privileged position in our practical lives that has sometimes been claimed for them: ‘A non-moralized, or less moralized, psychology . . . leaves it open, or even problematical, in what way moral reasons and ethical values fit with other motives and desires, and how far they are in conflict with them’ (Williams, ‘Replies’, 202). 39. Humber and Moss, ‘The Relationship of Preschool and Early School Age Attachment to Mother–Child Interaction’.


Edward Harcourt

40. Londerville and Main, ‘Security of Attachment, Compliance, and Maternal Training Methods in the Second Year of Life’, 290. 41. Magai, ‘Attachment in Middle and Later Life’, 533. 42. Fonagy, Steele, and Steele, ‘Maternal Representations of Attachment during Pregnancy Predict the Organization of Infant-Mother Attachment at One Year of Age’. 43. ‘Each attachment pattern reflects a different ‘strategy’ that could have solved adaptive problems presented by different kinds of rearing environments’ (Simpson and Belsky, ‘Attachment Theory’, 138). 44. Waters, Hay, and Richters, ‘Infant-Parent Attachment and the Origins of Prosocial and Antisocial Behavior’, 105. 45. Weinfeld, ‘Individual Differences’, 72. 46. See Holmes, Attachment, Intimacy and Autonomy. 47. See Belsky, ‘War, Trauma and Children’s Development’, 265. 48. Or organized attachment disposition? 49. In part for reasons well stated by Foot herself: Foot, Natural Goodness, 85.

REFERENCES Ainsworth, M., and B. Wittig, ‘Attachment and Exploratory Behavior in One-YearOlds in a Strange Situation’, in Determinants of Infant Behaviour, ed. B. M. Foss, vol. 4 (London: Methuen, 1969), 111–136. Alvarez, M., Kinds of Reasons (Oxford: Oxford University Press, 2010). Aristotle, Nicomachean Ethics, trans. W. D. Ross, rev. J. O. Urmson, in Aristotle: The Complete Works. Electronic edition. Bollingen Series LXXI, vol. 2. Belsky, Jay, ‘War, Trauma and Children’s Development: Observations from a Modern Evolutionary Perspective’, International Journal of Behavioral Development 32 (2008), 260–271. Cassidy, Jude, ‘The Nature of the Child’s Ties’, in Handbook of Attachment: Theory, Research and Clinical Applications, ed. J. Cassidy and P. Shaver (New York: The Guilford Press, 2008), 3–20. Chazan, Pauline, The Moral Self (London: Routledge, 1998). Feeney, J., ‘Adult Romantic Attachment’, in Handbook of Attachment: Theory, Research and Clinical Applications, ed. J. Cassidy and P. Shaver (New York: The Guilford Press, 2008), 456–481. Fonagy, P., H. Steele, and M. Steele, ‘Maternal Representations of Attachment during Pregnancy Predict the Organization of Infant-Mother Attachment at One Year of Age’, Child Development 62 (1991), 891–905. Foot, Philippa, Natural Goodness (Oxford: Oxford University Press, 2001). Freud, Sigmund, ‘Five Lectures on Psycho-Analysis’, in The Standard Edition of the Complete Psychological Works of Sigmund Freud, ed. J. Strachey, vol. 11 (London: Hogarth Press, 1957). Holmes, Jeremy, Attachment, Intimacy and Autonomy (Northvale, NJ: Aronson, 1996). ———, Exploring in Security (London: Routledge, 2011). Homiak, Marcia, ‘Virtue and Self-Love in Aristotle’s Ethics’, Canadian Journal of Philosophy 11 (1981), 633–651. Humber, N., and E. Moss, ‘The Relationship of Preschool and Early School Age Attachment to Mother–Child Interaction’, American Journal of Orthopsychiatry 75 (2005), 128–141. Londerville, Susan, and Mary Main, ‘Security of Attachment, Compliance, and Maternal Training Methods in the Second Year of Life’, Developmental Psychology 17 (1981), 289–299.

Attachment Theory, Character, and Naturalism 129 Magai, Carol, ‘Attachment in Middle and Later Life’, in Handbook of Attachment: Theory, Research and Clinical Applications, ed. J. Cassidy and P. Shaver (New York: The Guilford Press, 2008), 532–551. McDowell, John, Mind and World (Cambridge MA: Harvard University Press, 1994). ———, ‘Two Sorts of Naturalism’, in Mind, Value and Reality (Cambridge, MA: Harvard University Press, 1998), 167–197. Mikulincer, M., and P. Shaver, ‘Adult Attachment and Affect Regulation’, in Handbook of Attachment: Theory, Research and Clinical Applications, ed. J. Cassidy and P. Shaver (New York: The Guilford Press, 2008), 503–531. Robertson, J., and J. Bowlby, ‘Responses of Young Children to Separation from Their Mothers’, Courrier of the International Childrens’ Center, Paris, 2 (1952), 131–140. Schaffer, H. Rudolph, Key Concepts in Developmental Psychology (London: Sage, 2006). Simpson, Jeffrey A., and Jay Belsky, ‘Attachment Theory within a Modern Evolutionary Framework’, in Handbook of Attachment: Theory, Research and Clinical Applications, ed. J. Cassidy and P. Shaver (New York: The Guilford Press, 2008), 131–157. Solomon, J., and C. George, ‘The Measurement of Attachment Security’, in Handbook of Attachment: Theory, Research and Clinical Applications, ed. J. Cassidy and P. Shaver (New York: The Guilford Press, 2008), 383–418. Sroufe, L. A., The Development of the person: The Minnesota Study of Risk and Adaptation from Birth to Adulthood (New York: Guilford Press, 2005). ———, ‘Infant–Caregiver Attachment and Patterns of Adaptation in Preschool’, in Minnesota Symposia on Child Psychology, ed. M. Perlmutter, vol. 16 (Hillsdale, NJ: Erlbaum, 1983), 41–83. Thompson, Ross, ‘Early Attachment and Later Development’, in Handbook of Attachment: Theory, Research and Clinical Applications, ed. J. Cassidy and P. Shaver (New York: The Guilford Press, 2008), 348–365. Waters, E., D. Hay, and J. Richters, ‘Infant-Parent Attachment and the Origins of Prosocial and Antisocial Behavior’, in Development of Antisocial and Prosocial Behavior: Research, Theories, and Issues, ed. D. Olweus, J. Block, and M. RadkeYarrow(Orlando: Academic Press, 1986), 97–128. Weinfeld, Nancy S., L. A. Sroufe, B. Egeland, and E. Carbon, ‘Individual Differences in Infant-Caregiver Attachment’, in Handbook of Attachment: Theory, Research and Clinical Applications, ed. J. Cassidy and P. Shaver (New York: The Guilford Press, 2008), 78–101. Williams, Bernard, ‘Naturalism and Genealogy’, in Morality, Reflection and Ideology, ed. Edward Harcourt (Oxford: Oxford University Press, 2000), 148–161. ———, ‘Replies’, in World, Mind and Ethics, ed. J.E.J. Altham and R. Harrison (Cambridge: Cambridge University Press, 1995), 185–224.


Notes Toward an Empirical Psychology of Virtue Exploring the Personality Scaffolding of Virtue Nancy Snow

INTRODUCTION: SITUATIONISM AND ITS CHALLENGES Situationism is the view, now familiar in contemporary ethics, that virtue ethics is empirically inadequate. The central complaint is that virtues are global or robust traits, that is, traits that are deeply entrenched parts of personality manifested in regular behavior across different types of situations, and that a wealth of social psychological experiments show either that such traits do not exist, or are so scarce that they are not significant factors in producing behavior. Specific situationist complaints take a variety of forms. For example, Harman complains that social psychology gives us no reason to think that the kinds of traits that can be virtues exist and, thus, no reason to think that we can become the kinds of people that virtue ethics tells us to be.1 He is so negatively disposed to virtue ethics that he thinks even the use of the term virtue is harmful and should be abandoned.2 Doris is more moderate.3 He admits that small numbers of people might actually possess global traits but maintains they are insignificant in producing behavior. More recently, Doris has come closer to Harman’s position on the empirical impossibility of virtues and character, contending that the burden of proof lies with virtue ethicists to show that traditional philosophical conceptions of virtue and character are empirically possible.4 Recently, Merritt, Doris, and Harman extended their challenge to rationality, arguing that rationality itself is too fragmented to support the coherent character that virtue ethics extols as an ideal.5 Their argument relies on dual process theory from psychology, according to which the mind’s workings are explained by conscious or controlled and nonconscious or automatic processes. Conscious processes are those to which we devote attention; nonconscious operate below the level of conscious awareness.6 Human rationality, say the situationists, is fragmented, with controlled or conscious deliberation telling us to do one thing and nonconscious or automatic processes pulling us in other ways, often undermining our best laid plans, including plans to be virtuous.

Notes Toward an Empirical Psychology of Virtue


Elsewhere I argue against the situationist critique of global traits and maintain that the CAPS (cognitive-affective personality system) theory of social-cognitive psychologists Walter Mischel and Yuichi Shoda provides a promising empirical grounding for global traits and, thus, for virtue ethical theories that rely on such traits.7 Social-cognitivism is a school of psychological thought that explains the workings of personality in terms of the interactions of internal variables, structures, and processes, both with each other and with external stimuli. For social-cognitivists, people’s construals of their situations, that is, the meanings situations have for them, are crucial for understanding personality as well as behavior. The social-cognitivist critique of situationist experiments—that they ignore how people construe situations—identifies general grounds for thinking these studies inadequate reason for skepticism about the existence or influence of global traits. I argue that CAPS traits are the kinds of traits a subset of which are likely candidates for virtues. To be sure, Mischel and Shoda are careful not to draw conclusions beyond what is warranted by the empirical evidence. Consequently, they do not claim that the cross-situationally consistent behavior they’ve documented in their empirical studies furnishes evidence of global traits. I argue, however, that CAPS traits, though initially local, can be generalized to become global traits, though the process is not easy. In the rest of this chapter I use two complementary social-cognitivist approaches to personality to offer notes toward a more comprehensive account of empirically grounded virtue than that begun in my earlier work. Each approach is used to complement the others in order to sketch an overview of how different features of the moral psychology of virtue can be explained using the resources of empirical psychology. In other words, I supplement the work of using CAPS traits as a possible underpinning for Aristotelian (and other) virtues. In the following section, I examine aspects of Daniel Cervone’s theory of ‘knowledge and appraisal personality architecture’ (KAPA).8 Cervone’s approach illuminates what I call the ‘personality scaffolding’ of virtue—the psychological structures and mechanisms that can help or hinder the development, sustenance, and exercise of virtue. These structures and mechanisms provide forms of personality coherence and go some way toward answering the philosophical situationists’ challenge to the unity of cognition. In the section titled ‘Virtue Development: A Response to Merritt, Doris, and Harman’, I examine work on the development of virtue, especially by Darcia Narvaez and Daniel K. Lapsley.9 As with Cervone, their work furnishes insights into personality coherence, the unity of cognition, and elements of the personality scaffolding of virtue. Their focus on the use of nonconscious processes to cultivate virtue offers a more virtue-friendly picture of practical rationality and agential control than is given by Merritt, Doris, and Harman and, thus, gestures toward directions philosophers might take to counter this most recent salvo against virtue ethical, especially Aristotelian, conceptions of character and rationality.


Nancy Snow

KAPA STRUCTURES AND MECHANISMS: ELEMENTS OF THE PERSONALITY SCAFFOLDING OF VIRTUE Virtues do not exist in a personality vacuum, but they co-exist with a constellation of other personality structures and processes that can help or hinder the development, sustenance, and exercise of virtue. Insights into some of them are found by adverting to Cervone’s theory of ‘knowledge and appraisal personality architecture’ (KAPA). Cervone avers that two aspects of cognition are central for modeling the architecture of personality: knowledge and appraisal.10 According to him, knowledge is structural and appraisals are dynamic. He contends that ‘Knowledge consists of beliefs about actual or prospective attributes of persons or the environment. Elements of knowledge, then, are enduring mental representations of a feature or features of oneself, other persons, or the physical or social world’.11 In addition, knowledge varies in the extent to which it is generalized or domain-linked. Cervone’s account of knowledge is largely consistent with standard philosophical conceptions of knowledge as beliefs or mental representations. Appraisals are more complex: They are relational judgments, that is, evaluations of the relations between oneself and occurrences within particular encounters. Specifically, appraisals are relational judgments that concern the meaning of encounters for oneself; they are ‘continuing evaluation[s] of the significance of what is happening for one’s personal well-being’ (Lazarus, 1991, p. 144). In the appraisal process, people construct personal meaning by relating features of the self (one’s concerns, aims, and capacities) to features of an encounter (its opportunities, threats, and constraints).12 Cervone continues: ‘Appraisals, then, are not mere representations of information, but affectively significant evaluations of the personal implications of information’.13 Appraisals can clearly take a variety of forms. When I ask myself, for example, how well I did or am doing on a particular task or in an encounter, such as a job interview, I’m appraising my performance, and that evaluation can have implications for which actions I take or refrain from taking in the situation, how I view myself, my prospects, and so on. Suppose that I’m invited to participate in a research collaboration. When I ask myself, ‘What’s in it for me?’, ‘Am I the best qualified person to contribute to this?’, and so on, I’m appraising the ‘pros’ and ‘cons’ of my participation. Will this project take up too much time and yield too few dividends in terms of career advancement? Would someone else be a better ‘fit’ with the goals of the collaboration? Appraisals are relevant to virtuous action and the development of virtue, too, insofar as I can stop to ask myself how well I’m doing in exercising or developing a specific virtue. Suppose I intend to have a conciliatory

Notes Toward an Empirical Psychology of Virtue


conversation with someone with whom I’ve been rowing, but instead, find myself becoming angry and making cutting remarks. If I stop myself and recalibrate my emotions and words in accordance with my original goal of conciliation, I’m using appraisal to modify my reactions and behavior. Suppose I seek in general to become more generous and less self-centered. Appraisals of how well I am doing are parts of the self-regulatory processes involved in moving toward my goals. If we think appraisals relevant to virtue, we must amend Cervone’s reliance on Lazarus’s claim that appraisals focus on the relevance of a state of affairs for one’s personal well-being. Virtue-relevant appraisals take into account factors other than the personal well-being of the agent who is trying to be or become virtuous. This presents no special problem for our use of Cervone, though, for we can allow that virtue-relevant appraisals have a broader scope than the personal well-being of the virtuous agent while maintaining the core notion of an appraisal as a judgment about the meaning of an encounter for the self. We might also need to broaden other features of Cervone’s description, such as the specification of aspects of the self as ‘concerns, aims, and capacities’, and that of attributes of an encounter as ‘opportunities, threats, and constraints’, to include a richer panoply of features. Again, this seems consistent both with Cervone’s approach and with the central meaning of ‘appraisal’. KAPA uses knowledge structures and appraisal mechanisms to describe the architecture of personality.14 Knowledge structures are complex mental representations of the self and the world. My conception of myself, or self-schema, is an example of a knowledge structure. My self-schema is the constellation of beliefs and representations I have of myself—my conception of myself as a woman, as a philosopher, as a virtue ethicist, as a professor, and so on. Appraisal mechanisms are the processes by means of which I relate encounters with the world to myself. Cervone writes: ‘Appraisal processes function as proximal determinants of experience and action in a given encounter. Knowledge structures are more distal determinants that influence emotion and action through their influence on appraisals’.15 The general picture that emerges here is that knowledge—enduring representations of myself and of the world—forms the backdrop against which ongoing appraisals of my encounters with the world are made. We can amend this account with a proviso that, again, seems consistent with Cervone’s overall approach: not only do knowledge structures influence appraisals, but appraisals can also affect knowledge structures. This is especially important for understanding how knowledge structures and appraisal mechanisms help or hinder the development of virtue. If I am trying to become kind, for example, my appraisals of how kind I am being in certain situations should affect my self-conception, in particular, whether I am able to view myself as a truly kind person. Positive changes in self-conception can, in turn, support the development and exercise of my virtue. I will discuss this in more detail in a moment. For now, let us note that, working in tandem, the duo of


Nancy Snow

knowledge structures and appraisal mechanisms explains how we encounter the world and exercise personal control over our actions and reactions. We might find this to be a sketchy and partial account of how our engagement with the world works. To be sure, there is more detail to KAPA than can be described here. The basic picture, rough though it is, allows us to home in on a knowledge structure and an appraisal mechanism that form parts of the personality scaffolding of virtue: the self-schema and appraisals of self-efficacy. An added bonus of studying the roles of the self-schema and self-efficacy appraisals in Cervone’s work is that empirical studies have identified cross-situational coherence between specified domains of subjects’ self-schemas and their self-efficacy appraisals in certain contexts.16 These results provide evidence of personality coherence and counter the philosophical situationists’ view of personality as fragmented. Let us turn first to the argument that the self-schema and self-efficacy appraisals form scaffolding for the development and maintenance of virtue, then examine the empirical studies. In Aristotle’s account of virtue, the development of virtue in children begins early, with habituation into appropriate actions and response provided by caregivers. Ideally, the cultivation of virtue becomes more deliberate and self-reflective as children mature and continues throughout life, even in adulthood. Moreover, as recent Aristotelian-inspired accounts of virtue stress, virtue cultivation can be modeled on the acquisition of practical skills.17 This acquisition, as well as the habituation that partly constitutes it, are not matters of mindless routine, but intelligent, situation-sensitive efforts at virtue development. On this model, a person who is trying to become kind, for example, will pay intelligent heed to the circumstances in which she finds herself, and not offer ‘kind’ remarks as a matter of mere habit or rote, but take pains to focus her comments and actions appropriately on the intended recipient of her kindness. Consider a cashier at a café who is trying to develop the virtue of kindness. She will ask customers how they’re doing and will eschew mindless, automatic responses such as the oft-repeated, mechanical ‘have a good day’, in favor of more personalized remarks that genuinely engage with their conversation. In other words, in the course of cultivating her kindness, she will seek to leave her customers with the impression that she genuinely cares about them, is glad to see them, wants them to have a pleasant experience at the café, etc. She will treat them as individuals and not as mere bodies in an assembly line that she has to process. Parts of her personality other than kindness can help or hinder her efforts to develop her virtue. Her temperament has an influence. If she is naturally extroverted and interested in people, her efforts to engage her customers will seem more genuine than if she is introverted or shy and has to force herself to interact. Other virtues or the lack thereof could influence her development of kindness. If she is impatient, she could seem curt and abrupt, whereas, if she is patient, she could be more disposed to exhibit kindness with customers who are indecisive about placing their orders, or fumble for change. Crucial

Notes Toward an Empirical Psychology of Virtue 135 to the deliberate cultivation of her kindness are certain aspects of her selfschema or self-conception, as well as ongoing appraisals of her self-efficacy in becoming a kind cashier. If a part of her self-schema is that of a cashier who is committed to the kind and professional treatment of her customers, having that self-knowledge should be, to echo Cervone, a distal determinant that affects emotion and action by influencing appraisals. In less technical terms, how she thinks of herself affects how she interacts with customers and her alacrity in monitoring the nature of her interactions. Someone who conceives of herself as a kind cashier will be more likely to develop a kind and considerate attitude, to adopt a kind demeanor, and to treat her customers with kindness, than, say, someone who conceives of herself as just doing a job, as putting in the hours, as doing what is needed to earn a paycheck, and so on. Additionally, someone whose self-conception includes being a kind cashier is more likely to look for opportunities for kindness, to gauge the need for kindness in the treatment of specific customers, to judge what would count as kindness in specific interactions, and so on—in other words, to make the appraisals necessary for the successful exercise of kind acts and the concomitant development of the virtue of kindness. Appraisals of self-efficacy are especially important in efforts to develop and sustain virtue. Appraising self-efficacy is a process of self-monitoring, of noting and keeping track of how well one does in achieving a goal in certain kinds of situations. These appraisals can then be used to adjust affect and behavior to better achieve the goal, if need be. Suppose our cashier notices that she tends to be frustrated with elderly customers who fumble for change, and this causes her to treat them less kindly than she would like. This appraisal of her self-efficacy leads her to realize that she needs to be calmer and more patient with them. She then attempts to moderate her affect, to be more patient, to slow down and let them take their time. In doing so, she adopts and implements a self-regulatory strategy that helps her achieve her goal of treating customers kindly, is consistent with her self-conception as a kind cashier, and is a way of cultivating appropriately virtuous responses in the relevant situations. More complexity could be added to this sketch, of course. In particular, the questions of how and why the desire to develop virtue arises and comes to influence one’s self-conception and self-efficacy appraisals are important, yet unanswered, parts of this story. My aim, here, however, is not to trace the origins of our motivations to be or become virtuous, but to outline how features of personality that have been investigated in the empirical psychological literature can help or hinder virtue. The example of the kind cashier is meant to furnish an idea of how this can happen. What should we say about a person whose self-schema with regard to virtue is mistaken? Suppose that someone conceives of herself as a kind person, but in reality falls short of kindness. One response to this is to say that, if she is truly interested in appraisals of self-efficacy, they should reveal her deficiencies. But perhaps she does not bother to make self-efficacy appraisals, in


Nancy Snow

which case she is complacent and cannot be said truly to care about being a genuinely kind person. Alternatively, her appraisals could be as flawed as her self-conception. In this case, we might say she has a moral ‘blindspot’ about her kindness and that her self-schema as a kind person hinders, rather than supports, her virtue and its development. We might advise her to take seriously input from friends and colleagues who notice her lack of kindness and its inconsistency with her self-schema. This sort of case does not present a general problem for the notion that self-schemas and appraisals of self-efficacy are parts of the personality scaffolding of virtue, though it does reveal that sometimes this scaffolding can go awry and hinder virtue. Then, it needs correction. We might suspect that the operation of knowledge structures and selfefficacy appraisals offers a way of understanding cross-situational consistency in behavior and thus, personality coherence. The self-schema of the cashier as a kind cashier supports her desire to exercise her virtue in the workplace, motivates her self-efficacy appraisals, and gives rise to a variety of situation-sensitive actions expressing kindness. If we understand that she sees herself as a kind cashier, we can understand how and why she acts as she does in specific situations. This, of course, is consistent with social-cognitivists’ insistence that subjects’ construals matter. We can understand the cashier’s actions and personality once we see what she is doing from her perspective. We can see that her actions are, indeed, consistent across objectively different situation-types, in that she believes they express kindness. Provided that her conception of kindness and of kind acts are not too idiosyncratic, but are largely in tune with standard conceptions of virtue, such as that found in Aristotle, we can say that she genuinely has and exercises the virtue of kindness and can understand the personality structures and processes that support her commitment to, and cultivation of, that virtue.18 Now consider empirical studies supporting the hypothesis that aspects of the self-schema influence self-efficacy appraisals in certain domains and thereby support cross-situational personality coherence. Cervone addresses these questions.19 Empirical studies of college students examined coherence among aspects of the self-schema, situational beliefs, and self-efficacy appraisals. As part of their course requirements, 122 undergraduates took part in three sessions during a one-month period.20 During the first session, they completed self-schema measures, which identified each subject’s personal strengths and weaknesses. Some subjects’ self-schemas included generalized traits, such as ‘calm’, and ‘friendly’; others included domainlinked characteristics, such as ‘go off on tangents when talking to people’. Situational knowledge was gathered during the second session, especially subjects’ beliefs about the relevance of specific situations to personality attributes. Among these attributes were traits derived from subjects’ selfschemas, as well as aschematic traits, that is, traits not included in subjects’ self-schemas but provided by experimenters. The inclusion of aschematic traits allowed Cervone to test the hypothesis that cross-situational coher-

Notes Toward an Empirical Psychology of Virtue


ence in subjects’ appraisals of self-efficacy would be weak or not displayed in situations linked to aschematic traits. In session three, subjects completed a multi-domain self-efficacy scale. The study yielded several interesting results. First, subjects displayed higher self-efficacy appraisals across situations they deemed relevant to their personal strengths and personality traits they had judged to be most important, and lower appraisals in situations linked to personal weaknesses. Selfappraisals of aschematic positive and negative traits did not significantly differ, and data showed that subjects gave higher self-efficacy appraisals for personal strengths and traits they judged most important than for either positive or negative aschematic traits.21 These results support Cervone’s hypotheses that self-efficacy appraisals would show cross-situational coherence indexed to aspects of subjects’ self-schemas and beliefs about the relevance of situations to those self-schemas and that self-efficacy appraisals would not strongly relate to aschematic attributes. Second, Cervone hypothesized that having a more complex or nuanced knowledge system would lead people to see numerous possibilities and difficulties stemming from a situation and, thus, lead to less extreme appraisals of self-efficacy.22 This hypothesis was borne out: ‘Higher cognitive complexity was found to predict lesser situation-to-situation variability in appraisal’.23 Finally, though the studies recounted in Cervone rely on self-reports, they provide a way of testing for the underlying causes of surface-level behavioral tendencies.24 Data revealed that subjects linked the same behavioral tendencies to different traits. For example, on self-efficacy measures of agreeableness relating to specific dating situations, participant 63 linked her ability to be gracious to her partner’s parents to her being nice, whereas participant 118 linked his ability to be gracious to his ability to manipulate people. Cervone’s experimental approach, then, offers a more fine-grained way of discovering the traits people think guide, or would guide, their behavior in any given circumstance than the studies the philosophical situationists cite. Studies have been done on the relevance of self-schemas to smokers’ selfefficacy appraisals and on using KAPA architecture to predict consistency and variability in smokers’ self-efficacy appraisals in high-risk situations.25 Shadel and Cervone tested whether two smoking self-schemas, those of the abstainer-ideal possible self and the abstainer-ought possible self, affected the self-efficacy of smokers to resist temptation when exposed to provocative smoking cues.26 The former asked subjects to imagine the characteristics they would ideally hope to have as a nonsmoker, and the latter, to imagine the characteristics they would be obligated to have as a nonsmoker. Subjects were also asked to describe themselves as smokers, that is, to provide a smoker self-schema. Researchers then manipulated priming conditions to activate the three self-schemas under two craving conditions, background and episodic.27 They found that ‘priming the abstainer-relevant possible selves produced significant increases in self-efficacy compared with when the smoker self-schema was primed’.28


Nancy Snow

Cervone, et al. predicted that ‘smokers’ self-efficacy appraisals would vary across contexts and that these variations would be predictable from assessments of self-knowledge and situational beliefs’.29 The experimenters note that they did not simply document intraindividual consistency and variability in self-efficacy appraisals but found a way to predict and explain them. They did this by measuring response times for subjects to appraise the efficacy of their performance. They found systematic variations in response times across smoking contexts. For example, they found slower response times in subjects who had previously identified a situation as one in which they had the ability to avoid smoking, but also identified a personal quality as a hindrance to smoking avoidance in those circumstances. The experimenters interpreted slower response times as indicating a slower, more deliberate self-appraisal and contrasted such cases with those characterized by fast response times, in which subjects did not see their personal qualities as hindering their abilities to resist smoking in given situations. The researchers note that self-efficacy appraisals ‘may be the most consistent predictor of smoking outcomes among a range of possible variables’.30 They believe that linking these appraisals to aspects of subjects’ self-schemas provides a way forward in the search for smoking cessation treatments.31 Several of these empirical findings are of special relevance to the development and sustenance of virtue. First, findings from Cervone about the weak relation of self-efficacy appraisals to aschematic traits reinforce the general social-cognitivist critique of the studies philosophical situationists cite: subjects’ construals matter. If a subject does not think she possesses a trait, or does not find it important, she will be less interested in, and confident of, her self-efficacy in situations that call for its exercise. The finding suggests the need for further virtue-related empirical research. What are the virtues that people think they don’t have, or deem unimportant? Identifying gaps in the virtues possessed by the general population is one vital step forward in crafting education programs for virtue development. A second finding from Cervone is the linking of higher cognitive complexity to less extreme variability in self-efficacy appraisals across situations. That is, people who are able to discern the possibilities inherent in situation-sensitive responses are more moderate in predicting how successful they will be in acting in those contexts; this suggests that deliberation is at work in making self-efficacy appraisals. Surely this cognitive element reinforces the views of those virtue theorists who stress the need for deliberative excellence in the development and exercise of virtue and indicates the need for further empirical study. Finally, Cervone’s finding that different traits were linked to the same selfefficacy appraisals in the same situation-type surely merits further research, as it is an important step toward identifying the motivations that underlie appraisals, and, ultimately, behavior. As the findings indicate, some of the motivations that drive the same appraisals could be virtue-relevant, such as the desire to be ‘nice’, or aligned with vice, such as the ability to be manipulative.

Notes Toward an Empirical Psychology of Virtue


The smoking studies, too, have relevance for virtue ethics. The experimental design of Shadel and Cervone could provide a model for studies focusing directly on virtue. One can readily imagine studies comparing selfefficacy appraisals linked with the primed ‘compassion-ideal self-schema’ with those generated by the primed compassion self-schema of the ordinary self, or by the primed ‘indifferent self-schema’. The finding of slower response times in self-efficacy appraisals involving personal hindrances in Cervone et al. extends the finding about higher cognitive complexity in Cervone and reveals the need for further experiments probing the cognitive complexity of self-schemas and their relation to self-efficacy appraisals. To conclude this section, I have used Cervone’s KAPA model to make a preliminary case for a knowledge structure, the self-schema, and an appraisal process, self-efficacy appraisals, as parts of the personality scaffolding that can support or hinder the development, sustenance, and exercise of virtue. This extends the project begun in Virtue as Social Intelligence, where I argue that virtues are likely subsets of Mischel and Shoda’s CAPS traits. In the next section I take this project one step further by drawing on social-cognitivist psychologists’ work to explore the nature of virtue development. I motivate this discussion by showing how it responds to the most recent attack on the unity of cognition by Merritt, Doris, and Harman.

VIRTUE DEVELOPMENT: A RESPONSE TO MERRITT, DORIS, AND HARMAN32 In responding to the philosophical situationists’ challenge to global traits, some philosophers stress the integrative, unifying function of practical reason as a way of countering the fragmented conception of personality put forward by Doris.33 Merritt, Doris, and Harman are dubious, contending: ‘The empirical research suggests that reason is no less situationally susceptible than overt behavior; the suggestion we must consider is that notions of rationality operative in traditional understandings of character are themselves empirically inadequate’.34 To make their case, they invoke dual process theories of cognitive functioning, according to which the mind’s workings are explained in terms of both conscious and nonconscious processes. We are aware of conscious processes (I know that I am now typing), but unaware of nonconscious processes (I do not have to consciously think about where to place my fingers on the keyboard). Merritt, Doris, and Harman argue that numerous empirical studies show that nonconscious processes do the lion’s share of the cognitive work in our daily lives, often undermining our deliberative efforts and discrediting the Aristotelian picture that highly deliberative uses of reason can and should guide our actions and shape our moral lives. In some of these studies, Merritt, Doris, and Harman argue, subjects experience a phenomenon they call ‘moral dissociation’—subjects engage in behavior that contravenes norms they seem to endorse.35 The authors


Nancy Snow

maintain that this behavior is caused by ‘depersonalised response tendencies’: tendencies to respond to situational factors that operate below the level of conscious awareness.36 A particular form of moral dissociation is ‘incongruency’.37 Incongruency occurs when behavior is produced by nonconscious processes that the subject would not endorse as reasons for action that accord with his or her normative commitments.38 Were the person consciously aware that the behavior was produced by the nonconscious processes, she would condemn it. Incongruency, the authors write, ‘unsettles notions of well-integrated deliberation’.39 As always, the authors support their view with a wide range of empirical studies. I review these studies elsewhere and argue that they provide but a partial picture of the workings of the conscious and nonconscious mind.40 Indeed, numerous other studies offer a more integrated, unified perspective on how cognition works. Cervone’s studies of the self-schema and self-efficacy appraisals contribute to the psychological work that seeks to document and explain cognitive structures and processes at work in personality coherence. In a similar vein, Lapsley and Hill argue that moral personality is unified and explained by the chronic accessibility of a person’s moral schemas.41 These schemas afford epistemic receptivity for the processing of certain kinds of information. An individual with the appropriate moral schemas will be more disposed to notice and respond to a person in distress than someone who lacks them. Repeated processing of certain kinds of information reinforces the strength and salience of the relevant moral schemas. Elsewhere I argue that goal-dependent automaticity is another case in which a cognitive process that mediates moral action results from internalizing a knowledge structure—this time, a goal.42 In this kind of nonconscious processing, environmental stimuli activate representations of a person’s enduring goals. Upon encountering relevant situational features, the representation of the goal is activated and sets in motion behavior directed to goal attainment. These behaviors are intelligent, flexible responses to environmental stimuli with some of the same qualities as consciously chosen actions. Repeated activations of situation-stimuli links can result in behavior that eventually becomes habituated. Virtue-relevant goals are likely to be enduring and, thus, among the goals that can be nonconsciously activated and pursued in different types of situations. These lines of research suggest ways in which nonconscious processing can be aligned with the consciously chosen goals, schemas, and self-schemas of the agent to produce action in accordance with an agent’s moral values and beliefs. These knowledge structures, operating below the level of conscious awareness, focus the agent’s attention on morally salient features of the environment and activate appropriate moral responses. Narvaez and Lapsley articulate a robust account of how conscious and nonconscious processing can work in tandem in the development of virtue.43 Though their account has many dimensions, I focus here on their suggestions for using nonconscious processing for developing virtue. Not only does their view counter Merritt, Doris, and Harman’s fragmented picture of human cogni-

Notes Toward an Empirical Psychology of Virtue


tion, it also shows how nonconscious processing is an important personality scaffolding for virtue. A first point is that their account allows significant roles for tacit or nonconscious information processing in moral learning, including roles for various kinds of automaticity, such as post-conscious automaticity. Postconscious automaticity ‘operates after a recent conscious experience or recent deployment of attentional resources’.44 Post-conscious automaticity has been discussed by philosophical situationists, who describe experiments in which some subjects are primed with words likely to activate stereotypes, such as ‘elderly’ or ‘bingo’, and others, in a control group, are primed with neutral words.45 Subjects primed with words such as ‘elderly’ are then observed to walk more slowly toward an elevator than subjects in the control group, even though they subsequently claim not to have noticed the ‘priming’ words nor thought the words affected their behavior. Philosophical situationists use such studies, which document the existence of situational priming, to illustrate that we have less awareness of the factors that affect us and less conscious control over our behavior than we think. Situational priming is one temporary manifestation of post-conscious automaticity. By contrast, Narvaez and Lapsley discuss chronic priming as a tool that enables us to teach and learn virtue. The idea here is that once virtue-constructs have been ‘built into’ someone’s mind, they become available for information processing and can be accessed nonconsciously.46 As we teach virtue to children, we repeatedly expose them to virtue concepts and their meanings and applications in various social settings, with the hope that they will internalize the scripts or action sequences that show how to be virtuous, say, how to be kind or generous. The idea is that children’s learning of virtue through repeated exposure to scripts and action sequences can result in chronic or enduring manifestations of virtue so that children, internalizing guidance for how to act virtuously, will begin acting virtuously over time and eventually develop virtuous dispositions as parts of their emerging characters. A final point on nonconscious factors in moral learning has to do with tacit information processing. Drawing on the work of Ulrich Neisser, Narvaez and Lapsley note the existence of three systems of the unconscious: the basic, the primitive, and the sophisticated (which includes types of automaticity).47 Let me discuss but one feature of the sophisticated unconscious—Neisser’s notion of ‘affordances’. Narvaez and Lapsley explain that ‘An affordance is the reciprocity of the organism and the environment, that is, the offerings of the environment and the way the organism (through evolution and through experience) can use the resources’, and ‘Perceiving an affordance is to perceive the relationship between environmental support and personal capacity’.48 We perceive affordances nonconsciously; examples of perceiving affordances include understanding the drift of an argument, noticing the location of a fire exit, or picking up the emotion in a comment.49 Part of virtue cultivation includes teaching children to perceive affordances for when and how to act virtuously, and this implies influencing, through various techniques, the development of their nonconscious


Nancy Snow

minds. To be good at being generous, for example, one must be able to perceive affordances—to know when an offer of financial assistance would wound someone’s pride, or to know what kind of gift would strike the right tone for a special occasion. Narvaez and Lapsley build an account of virtue acquisition as the development of expertise on this understanding of the nonconscious mind. In their emphasis on expertise, their view meshes well with Annas’s recent account, which models virtue development on the deliberate acquisition of expertise in practical skills.50 Their account of moral expertise development is too detailed to be examined here. Suffice it to say, however, that it is similar to the aspects of their view already discussed in that it shows how nonconscious and conscious processing can be guided to function together in the acquisition and sustenance of virtue. Thus, their view offers a picture of cognition as more integrated and virtue-friendly than that outlined by Merritt, Doris, and Harman.

CONCLUSION The ideas offered in this chapter are sketches that require much further elaboration. They are offered here as notes toward the development of a more comprehensive empirical grounding for virtue than that begun in my earlier work. Here as there, I disagree with the specific conclusions and visions of personality, cognition, and virtue that philosophical situationists endorse. Yet, we have to thank them for prodding us to do the empirical work of grounding virtue.

NOTES 1. Harman, ‘Moral Philosophy Meets Social Psychology: Virtue Ethics and the Fundamental Attribution Error’, 316. 2. Harman, ‘Moral Philosophy Meets Social Psychology’, 327–328. 3. Doris, Lack of Character: Personality and Moral Behavior, 6. 4. Doris and Stich, ‘As a Matter of Fact: Empirical Perspectives on Ethics’, 121. 5. Merritt, Doris, and Harman, ‘Character’. 6. See Snow, Virtue as Social Intelligence: An Empirically Grounded Theory, 40ff; Narvaez and Lapsley, ‘The Psychological Foundations of Everyday Morality and Moral Expertise’, 144. 7. See Snow, Virtue as Social Intelligence. 8. See Cervone, ‘The Architecture of Personality’ and ‘Social-Cognitive Mechanisms and Personality Coherence: Self-Knowledge, Situational Beliefs, and Cross-Situational Coherence in Perceived Self-Efficacy’. 9. Narvaez and Lapsley, ‘Psychological Foundations’. 10. Cervone, ‘The Architecture of Personality’, 186. 11. Ibid. 12. Ibid., 186–187. 13. Ibid., 187.

Notes Toward an Empirical Psychology of Virtue


14. Ibid. 15. Ibid. 16. See Cervone, ‘The Architecture of Personality’, and ‘Social-Cognitive Mechanisms and Personality Coherence’; Shadel and Cervone, ‘Evaluating SocialCognitive Mechanisms That Regulate Self-Efficacy in Response to Provocative Smoking Cues: An Experimental Investigation’; and Cervone, Orom, Artistico, Shadel, and Kassel, ‘Using a Knowledge-and-Appraisal Model of Personality Architecture to Understand Consistency and Variability in Smokers’ SelfEfficacy Appraisals in High-Risk Situations’. 17. See Annas, Intelligent Virtue. 18. One might think it too high a standard to impose an Aristotelian conception of virtue on the kind cashier, but I think it is not too far-fetched to believe that folk conceptions of virtue express and reinforce central elements of the Aristotelian conception, such as the notion that virtue requires appropriate motivation, good deliberation, and regular success in achieving the targets of virtue. For example, someone who seeks to be kind only in order to ingratiate themselves with others is not commonly regarded as truly kind but as selfseeking. Someone whose thoughtlessness always causes her to miss the mark is regarded as a well-intentioned bumbler, not as a genuinely kind person. The ‘folk’ often have Aristotelian intuitions about virtue. Philosophical situationists decry this about the ‘folk’. 19. Cervone, ‘The Architecture of Personality’, 188–196. 20. See ibid., 190–191; see also Cervone, ‘Social-Cognitive Mechanisms and Personality Coherence’. 21. Cervone, ‘The Architecture of Personality’, 192, 194, figure 6. 22. Ibid., 193–194. 23. Ibid., 194. 24. Ibid., 194, 196, figure 8. In addition to the studies’ reliance on self-reports, another possible drawback is that the subject population was college students. For better or worse, that population seems to be the most readily available to empirical psychologists. 25. Shadel and Cervone; Cervone et al. 26. Shadel and Cervone, 91. 27. Ibid., 92. 28. Ibid., 93. 29. Cervone et al., 51. 30. Ibid., 51. 31. Ibid., 52. 32. This section draws on Nancy E. Snow, ‘Situationism and Character: New Directions’; ‘ “May You Live in Interesting Times”: Moral Philosophy and Empirical Psychology’; and ‘Intelligent Virtue: Outsmarting Situationism’. 33. See Annas, Intelligent Virtue; Kamtekar, ‘Situationism and Virtue Ethics on the Content of our Character’; and Doris, Lack of Character. 34. Merritt, Doris, and Harman, ‘Character’, 360. 35. Ibid., 263. 36. Ibid., 370–371. 37. John M. Doris, e-mail message to author, January 2, 2012. 38. Merritt, Doris, and Harman, ‘Character’, 375. 39. Ibid., 375. 40. See Snow, ‘ “May You Live in Interesting Times” ’ and ‘Situationism and Character’. 41. Lapsley and Hill, ‘On Dual Processing and Heuristic Approaches to Moral Cognition’, 322. 42. Snow, Virtue as Social Intelligence, 43–45.

144 43. 44. 45. 46. 47. 48. 49. 50.

Nancy Snow Narvaez and Lapsley, ‘Psychological Foundations’. Ibid., 144. See Merritt, Doris, and Harman, ‘Character’, 374. Narvaez and Lapsley, ‘Psychological Foundations’, 146–147. Ibid., 147–149. Ibid., 148. Ibid., 145. See Snow, ‘Intelligent Virtue: Outsmarting Situationism’.

REFERENCES Annas, Julia, Intelligent Virtue (New York: Oxford University Press, 2011). Cervone, Daniel, ‘The Architecture of Personality’, Psychological Review 111 (2004), 183–204. ———, ‘Social-Cognitive Mechanisms and Personality Coherence: Self-Knowledge, Situational Beliefs, and Cross-Situational Coherence in Perceived Self-Efficacy’, Psychological Science 8 (1997), 43–50. Cervone, Daniel, Heather Orom, Daniele Artistico, William G. Shadel, and Jon D. Kassel., ‘Using a Knowledge-and-Appraisal Model of Personality Architecture to Understand Consistency and Variability in Smokers’ Self-Efficacy Appraisals in High-Risk Situations’, Psychology of Addictive Behaviors 21 (2007), 44–54. Doris, John M., Lack of Character: Personality and Moral Behavior (Cambridge: Cambridge University Press, 2002). Doris, John M., and Stephen P. Stich, ‘As a Matter of Fact: Empirical Perspectives on Ethics’, in The Oxford Handbook of Contemporary Philosophy, ed. Frank Jackson and Michael Smith (New York: Oxford University Press, 2005), 114–152. Harman, Gilbert, ‘Moral Philosophy Meets Social Psychology: Virtue Ethics and the Fundamental Attribution Error’, Proceedings of the Aristotelian Society 99 (1999), 315–331. Kamtekar, Rachana, ‘Situationism and Virtue Ethics on the Content of our Character’, Ethics 114 (2004), 458–491. Lapsley, Daniel K., and Patrick L. Hill, ‘On Dual Processing and Heuristic Approaches to Moral Cognition’, Journal of Moral Education 37, no. 3 (2008), 313–332. Merritt, Maria W., John M. Doris, and Gilbert Harman, ‘Character’, in The Moral Psychology Handbook, ed. John M. Doris and the Moral Psychology Research Group (New York: Oxford University Press, 2010), 355–401. Narvaez, Darcia, and Daniel K. Lapsley, ‘The Psychological Foundations of Everyday Morality and Moral Expertise’, in Character Psychology and Character Education, ed. Daniel K. Lapsley and F. Clark Power (Notre Dame, IN: University of Notre Dame Press, 2005), 140–165. Shadel, William G., and Daniel Cervone, ‘Evaluating Social-Cognitive Mechanisms That Regulate Self-Efficacy in Response to Provocative Smoking Cues: An Experimental Investigation’, Psychology of Addictive Behaviors 20 (2006), 91–96. Snow, Nancy E., ‘Intelligent Virtue: Outsmarting Situationism’, paper presented at the Pacific Division Meeting of the American Philosophical Association, Seattle, Washington, April 6, 2012. ———, ‘ “May You Live in Interesting Times”: Moral Philosophy and Empirical Psychology’, Journal of Moral Philosophy (forthcoming). ———, ‘Situationism and Character: New Directions’, in Handbook of Virtue Ethics, ed. Stan van Hooft (Durham, England: Acumen Publishing, forthcoming). ———, Virtue as Social Intelligence: An Empirically Grounded Theory (New York: Routledge, 2010).

10 Natural Virtue and Proper Upbringing Candace Vogler

ZÔON LOGISTIKÓN For better and for worse human life is permeated by our sort of reason— the talking sort. To suppose that there is such a thing as right reason in the sphere of the practical—practical wisdom, one might say, or prudentia, or phrōnesis—is to suppose that we can act well and also that we can err. This is the sort of thing that Immanuel Kant pointed to in placing human beings among the addressees of the categorical imperative—the source of good action in us can, but need not, determine what we do. And it is what Aquinas marked in his initial discussions of those cultivated sources of action, habits, by remarking, ‘If the form be such that it can operate in diverse ways, as the soul; it needs to be disposed to its operations by means of habits’.1 That human beings can act ill is no surprise. That whatever a virtue might turn out to be, talk of virtue is talk of what Aquinas calls ‘habits’ that tend to practical wisdom, is a commonplace for contemporary Anglophone philosophers working in the area. But how it is that these habits tend to practical wisdom is less clear, even if we all agree that excellence in the exercise of practical reason, practical wisdom, shows itself in such habits as individual justice, courage, and temperance. Virtues are supposed to be beneficial powers or capacities that work in accordance with reason to make their bearers good human beings and to make what their bearers do count as acting well. Virtue gives reason appropriate direction in an individual human life, and this shows itself in both the particular actions of a virtuous adult and in the adult’s overall practical orientation. It belongs to neo-Aristotelianism to hold, further, that ethical virtue is not beyond reach for human beings with their wits about them; at least, it is normally and generally within reach even if many of us fail to get there. Aristotle certainly thought that it was rare. It is not clear whether he thought that it was generally possible for people. I will concentrate on two aspects of some contemporary Anglophone philosophical work on virtue—interest in corrective accounts of virtue (sometimes called, less helpfully, ‘remedial’ accounts) and a way of reading Aristotle that has it that a proper upbringing tends to virtue. In order to have a look at these,


Candace Vogler

I will rely on Anselm Müller’s discussion of Aristotle’s distinction between natural virtue and ethical virtue.2 Müller explains this distinction by way of offering a new interpretation of two still more perplexing aspects of Aristotle’s work on virtue: his treatment of virtue as a mean state—the notorious doctrine of the mean—and his claim that in order to have one virtue, I must have all of them—the equally notorious unity or reciprocity of the virtues thesis.

HABITUS OPERATIVUS BONUS I will not be concerned with catalogues of traits that may or may not count as virtues. A concatenation of psychological traits that happen to be agreeable or beneficial need not amount to virtue any more than a regime of musclebuilding that targets first one group of muscles, then another, and then another, need equip us for labor. Virtue is supposed to strengthen us for the challenges of acting well and living well in roughly the way that cultivating a strong, healthy body is supposed to equip us for physically demanding work. The topic at issue—virtue—turns on the cultivation of one’s practical powers for the sake of acting and living well, not the acquisition of a bundle of lovely psychological traits. I will accept all of this. I will accept two other Aristotelian commonplaces about virtue and one more recent commonplace as well: 1. No one is born virtuous [NE 1103a], but habituating oneself to virtue—building upon whatever help one has got from early childhood attachments and training—belongs to human nature [NE 1103a]. 2. However we understand virtue, the operation of virtue involves both particular actions done in such-and-such manner under specific circumstances involving such-and-such persons and things and the development of a general practical orientation that leafs out in many different kinds of actions [NE 1140a–b]. 3. If you are dissatisfied with how you have lived and how it has turned out you can set about changing your ways by cultivating new habits in the service of living a different kind of life. Such a change can, but need not, involve schooling yourself in virtue. The third commonplace was not one of Aristotle’s. Many commentators on Aristotle have operated on the assumption that one’s basic practical orientation is largely in place by the time one reaches, say, late adolescence. One of my former teachers stresses this point with repeated invocations of sentences like ‘The virtuous person was properly brought up’. Proper upbringing is defined as upbringing that tends to virtue. That is what gives it its propriety, and it is supposed to give us virtue as a second nature. Now, it is obvious (and I assume that this could not have come as news to Aristotle) that an appropriate upbringing is not sufficient for ethical virtue. Consider: if ever a boy laid claim to the best that ancient Athens had to

Natural Virtue and Proper Upbringing


offer its sons, that boy was Alcibiades. But if you make a study of the lives of famously ethically exemplary adults, you will quickly learn that an apparently appropriate upbringing seems not even to be necessary for virtue. On dark days, it can seem that things go the other way around: it is by overcoming upbringing by parents who act ill, or seeking experience with those who act ill if one’s parents have acted well, that ethically exemplary human beings came to be exemplary. This is often the case, for example, in conversion stories about the lives of Christian saints. Switching traditions dramatically, on one reading of the life of Siddhartha Gautama, ethically significant insight into an important source of human suffering seems to have required the young man’s deliberate pursuit of circumstances that exposed him to misery and vice that were no part of his life at home. In short, being raised by adults who act well at least appears to come apart from having something of the practical knowledge that is exercised in virtuous activity. And this should, likewise, come as no great surprise. It is one thing to work with a philosophical tradition that has it that there are such things as infused virtues—virtues that a human being cannot have unless something is given or put into the human being from a divine source. There are mysteries enough surrounding the question of what, exactly, is infused when one of us is blessed with faith, say, or hope, or caritas; there are mysteries enough surrounding how such a thing even is possible. It is quite another to suppose that adult human beings can, in effect, be like gods in this: they can infuse the young with something that somehow is virtue (if the adults act well) or vice (if the adults act ill). We can hold that giving children a proper upbringing is crucially important and that it matters very much whether adults model acting well around children, without thinking that the utter dependence of infants and small children on adults creates a site for human-to-human infused virtue. And if my justice, say, or temperance or courage was infused through my excellent upbringing, why does it make sense to praise me for acting well? Why not just be glad that I was lucky in youth when it comes time to partner with me in combat, say, or in business? Leave this question to one side. An apparently good upbringing seems to be neither necessary nor sufficient for virtue, and all of us know stories about people of good character who had bad parents, or else who faced the sorrow of bad children. If you meditate on the root of the word virtue—the vir—this is less puzzling. It points to strength, specifically to manliness and fortitude, as Cicero stressed,3 and it is not possible to build strength without finding, facing, and working with and against resistance. Other people and the world we share provide us with built-in points of resistance, of course. On some accounts of virtue, we provide ourselves with points of resistance as well. Now, it could be that an excellent upbringing just is one that challenges children appropriately, helping them to meet various kinds of resistance. If that is the case (which seems plausible), then, at least, something remains for the young to do—namely, use those opportunities well even if it takes a long time for the young to notice that these were opportunities


Candace Vogler

rather than gratuitous obstacles to happiness set up by adults incapable of understanding much of anything important. Perhaps that is what went wrong for Alcibiades—he failed to make good use of what Athens gave him. Perhaps the man who became the Buddha made better use of the opportunities he was given at home. I will return to this matter. For now, notice the suggestions that virtue is strength and that building strength happens through resistance. Both points, I think, are crucial to the deep motivation for corrective accounts of virtue—accounts that place me among the points of resistance I face in building and exercising ethical strength. Philippa Foot is the Anglophone atheist philosopher most notable for espousing a corrective account of virtue recently. In an early paper on the topic, she writes: [The virtues] are corrective, each one standing at a point at which there is some temptation to be resisted or deficiency in motivation to be made good.4 She illustrates the point in light of various specific virtues: ‘one may say that it is only because fear and desire for pleasure often operate as temptations that courage and temperance exists as virtues at all’,5 and: As with courage and temperance so with many other virtues: there is, for instance, a virtue of industriousness only because idleness is a temptation; and of humility only because men tend to think too well of themselves. Hope is a virtue because despair too is a temptation . . . With . . . justice and charity it is a little different, because they correspond not to any particular desire or tendency that has to be kept in check but rather to a deficiency of motivation that they must make good.6 Paula Gottlieb, the author of a recent book about Aristotle’s ethics, complains bitterly about Foot on this score.7 Gottlieb thinks that Foot’s view rests on a picture of human nature as essentially flawed and lacking—a dubious Christian or Stoical idea that has no place in a properly Aristotelian virtue ethics. I fear that Foot leaves herself open to Gottlieb’s complaint, partly by rejecting the reciprocity of virtues thesis. Foot writes: So far from forming a unity in the sense that Aristotle and Aquinas believed they did, the virtues actually conflict with each other: which is to say that if someone has one of them he inevitably fails to have some other.8 In short, unlike Aquinas, Foot rejects the version of the unity of the virtues thesis that rests upon the reciprocity of the virtues, and there is a hint that she rejects it because humans tend to be disorderly in their motives, aspirations, and actions and inadvertently at odds with themselves in their efforts to shape their practical orientations.

Natural Virtue and Proper Upbringing


The reciprocity of the virtues thesis is difficult. Anselm Müller gives the tidiest formulation, reminding us that, for all the attention it has generated, it is not a topic discussed at length by Aristotle. Here is Müller’s summary of the view: The possession of any one ethical virtue, for its application to particular situations, requires the possession of practical wisdom. But for wisdom the correct starting points for deliberation are needed, and so therefore all the virtues. Hence, by the transitivity of requirement, the possession of any one ethical virtue requires the possession of all the ethical virtues.9 Foot is not alone in finding the thesis implausible, but her rejection of it seems an outgrowth of her way of advocating a corrective account of virtue. That is certainly how Terrence Irwin understands some of Foot’s darkest remarks about virtue. Irwin points out that a corrective account of the virtues makes it reasonable to reject the conception of perfect virtue that is used to support [the reciprocity of the virtues thesis]. If different virtues are designed to counteract different dangerous tendencies, they need not be inseparable or free from conflict. . . . If the virtues are to be understood as piecemeal remedies for specific dangers and threats, we have no reason to suppose that the complete development of each virtue will result in the incorporation of other virtues; on the contrary, further development of any one virtue may include the growth of a tendency that actually makes it more difficult to acquire other virtues.10 I think that he is right to urge that Foot takes things in this direction. I do not think that defending a corrective account of virtue inevitably leads to a rejection of Aquinas’s position, however, partly because the sense in which virtues are corrective in Aquinas is not happily captured in the phrase ‘piecemeal remedies’. I don’t know whether one can square a corrective account of virtue with Aristotle. Aristotle certainly touches upon something connected to correction in discussing a man who knows the better and does the worse without becoming vicious. But corrective accounts of virtue hold that the ordinary and proper operation of fully-fledged ethical virtue is corrective, not just that virtue can keep us from spiraling toward vice when we are ungoverned. The corrective in virtue belongs to self-governance, rather than just kicking in when we fail to govern ourselves appropriately.11 PRACTICAL WISDOM, PHRŌNESIS, PRUDENTIA In thinking about the role that virtue has in right reason, it has become fairly common for philosophers like Foot to move very quickly between Aristotle’s work on phrōnesis and Aquinas’s work on prudentia. I do not think that it


Candace Vogler

is easy to treat the two as a single topic. Writ large, at least, there seems to be a divergence over the role of intellect in practice. Denis Bradley describes one pivotal controversy this way: Aristotelian morality seems bereft of a secure foundation in reason if we are content to accept what Aristotle baldly contends, that ‘we do not deliberate about ends, but about means’ and that choice ‘deals with means to ends.’ These and similar texts, which restrict the scope of phrōnesis, are the crux of contemporary scholars searching for a firm cognitive ground of Aristotelian morality. If they are interpreted narrowly, Aristotle’s explicit dicta would rob the agent of any rational choice about the rightness of his goals. Aristotle assumes that the virtuous man, through the good fortune of his upbringing and temperament, is, as a matter of fact, pursuing the right goals. But, then, is the Aristotelian phronimos making moral evaluations and pursuing goals that rest solely on conventional attitudes and sentiments?12 I will not take up the vexed question about how to widen the interpretation of the relevant passages in Aristotle in order to make the account look more like what contemporary Anglophone philosophers seek in seeking rational foundations for ethics. I will not venture into the fray over whether to handle the apparent insufficiency by delving more deeply into Aristotle’s account of desire as a source of human action, or by expanding the account of reason, or both. It does seem a bit much to argue, as Sarah Broadie does, that phrōnesis allows ‘continual re-evaluation’ of ends by way of the assessment and re-assessment of means and of means to means and their likely consequences,13 but, again, this is a question for scholars of Aristotle more able than I ever will be. I strongly suspect that if corrective accounts of virtue (as, I take it, Aquinas’s is) make contact with Aristotle, that contact is made by way of Aristotle’s stress on ethical virtue as a mean state. My thought is, roughly, this: given some plausible account of the sense in which the exercise of virtue aims at a mean, we will have found a sense in which Aristotelian virtue might have a corrective function in its ordinary operations. Finding something like a corrective function in Aristotelian virtue, in turn, might help us figure out whether an Aristotelian account of virtue is compatible with the thought that adults can change their ways in some manner that is regulated by reason, even if we do not think that the determined scoundrel can be argued into being a good human being.

THE DOCTRINE OF THE MEAN Müller also provides what I take to be the best account of virtue as a mean state. His interpretation requires rejection of some of Aristotle’s remarks about virtue and the mean, stressing that there are two very different senses

Natural Virtue and Proper Upbringing


in which Aristotle treats virtue as a mean state. Müller begins by outlining two aspects of virtue: (1) that virtues are distinguished from each other in terms of ‘the characteristic dimension of feeling and action in which they operate’14 and (2) that a fully-fledged ethical virtue is distinct from any ‘natural virtue’ that shares its characteristic dimension of feeling and action, and this distinction is a matter of the natural virtue having been ‘shaped by wisdom’.15 There are various ways in which a practical disposition may be ‘natural’. The uneasiness that somehow Aristotle’s virtue is nothing more than the child of a fortuitous marriage of temperament and pre-rational, conventionally sanctioned tendencies in thought and feeling—namely, those tendencies that happened to be attractive to the adults who governed us from infancy through adolescence—can be expressed as the worry that there is no clear line separating natural virtue and ethical virtue in Aristotle. Armed with this distinction between two aspects of virtue, Müller argues that there are two correspondingly different ways of treating virtue as a mean. On what he calls a ‘one-dimensional’ interpretation of the doctrine of the mean, the exercise of virtue involves feeling and acts that are neither excessive nor deficient. In short ‘this right feeling or action is said to occupy a position, within the characteristic dimension, that lies between too much and too little, considering the requirements of a good human life’.16 Müller contrasts this with what he calls a ‘many-dimensional conception of the mean’, which he associates with the second aspect of virtue. Müller writes: After saying, in a one-dimensional spirit, that ‘both fear and confidence and appetite and anger and pity and in general pleasure and pain may be felt both too much and too little, and in both cases not well,’ Aristotle continues: ‘but to feel them at the right times, with reference to the right objects, towards the right people, with the right motive, and in the right way, is what is both intermediate and best, and this is characteristic of virtue’ (NE, II, 1106b 18–23).—Here, a number of dimensions are indicated that have to be determined in the right way if you are to feel (or to act) well and ‘in the middle’.17 Müller calls these ‘critical dimensions’, and argues that, although Aristotle clearly does not treat these remarks as pointing to two different accounts of the mean, and although it is in no way obvious how to make sense of the one-dimensional account, these two ways of assessing intermediacy ought to be separated, and we ought to accept the many-dimensional account of the mean. There are problems enough brewing there. For example, it is unlikely that there will be a distinctive median in each of the many critical dimensions. It is correspondingly implausible to hold that the virtuous person is the one who tends to hit the mark of doing and feeling neither too much nor


Candace Vogler

too little dimension by dimension, even though some of Aristotle’s remarks seem to point in the dimension-by-dimension direction for the secondary assessment of the mean. Müller writes: quantification seems much less plausible where the mean position of a response depends, e.g., on getting objects, persons, and purposes right. The wrongness of these things need not correspond to, or come out in, quantitative deficiency or excess at all. . . . [Consider] the person who would stand his ground for some unjust causes and not for some (or any) just ones. Should we say that he is both too fearful and too fearless?18 And even in the examples where it is tempting to think that the idiom of too much and too little does have a clear point of application, whether one feels or does too much or too little ‘may depend on the object and other ‘circumstances’ of your response’, such that ‘even ‘where the critical dimension can itself be seen as a variant of the characteristic dimension’, an overall mean (and virtuous) response need not correlate with an intermediate degree in that critical dimension’.19 Müller explores various ways of salvaging something of the doctrine of the mean in the face of the many difficulties he discusses. He concludes that we should abide by the many-dimensional account of the mean and notice that the exercise of any one virtue finds its many-dimensional mean only when it is guided by the claims proper to other virtues: ‘Only then will you hit the many-dimensional ‘mean’, i.e., correctly decide when and how and because of what etc. to deploy, and not to deploy, the characteristic response of that virtue’.20 And that is how a reinterpretation of the doctrine of the mean sheds light on the reciprocity of the virtues.

BACK TO THE BUDDHA AND ALCIBIADES Having developed a plausible account of the many-dimensional mean and having given an account of how it turns on the reciprocity of virtues, Müller can draw a principled distinction between natural virtue and ethical virtue in Aristotle: The concept of a natural virtue is applied to a disposition, or inclination, to feel or act in a characteristic way in the sphere of that virtue. But it does not refer to further standards of evaluation of the kind alluded to by Aristotle’s mention of ‘right times, right objects, right people, right motive, right means and ways.’ So it does not refer to other virtues. The natural virtues do not delimit each other, the unity thesis does not apply to them. You can have some, perhaps any one, of them without having all the others.21

Natural Virtue and Proper Upbringing


That these are natural virtues, rather than just psychological traits that happen to please as often as not, suggests that they are, nevertheless, constrained. As Müller puts it: Natural virtues do to some extent share with their ethical counterparts constraints on the practice of their characteristic responses. These constraints are imposed, in the last resort, not only by considerations of practicality but, more importantly, by vague ideas of what serves human needs and ends.22 So, if friendliness is a natural virtue, then it is not just the abject need to please everyone all the time. If self-confidence is a natural virtue, it is not the same as infantile senses of omnipotence or more developed species of arrogance. If there is natural courage, we will not look for it in the tendency of some younger children to accept any and all dares issued by older children. In light of this insight, it is at least possible to counter the tendency on the part of some contemporary ethicists to suppose that proper upbringing can imbue human beings with ethical virtue—with fully-fledged, reciprocal, appropriately unified virtue—as though any human being, no matter how generous and fine, could effect that kind of alteration in any other human being, no matter how impressionable, needy, unguarded, or eager. What we can say, armed with Müller’s distinction between ethical and natural virtue, is that the propriety of a proper upbringing shows itself in this: proper upbringing contributes to the development of natural virtue in the young. Whether the youth will then become practically wise, whether they will be guided by right reason in their maturity, is not settled by looking to the advantages or disadvantages occasioned by their upbringing. This is not the conclusion that Müller draws. But it looks to be a conclusion that can be given some plausibility in light of his work, one that can help us understand how it is (1) that an apparently proper upbringing is not sufficient for ethical virtue, (2) that there can be virtuous adults whose upbringing looks to have been about as far from proper as you please, and (3) that it is nevertheless tremendously important to see to it that children are well brought up.

RETURNING TO FOOT In light of Müller’s work, it is also possible to revisit a vexed aspect of Foot’s cryptic remarks about the corrective operations of virtues. In Müller’s terms, it is pretty clear that Foot is taking it that the corrective in virtue operates one-dimensionally—the correction is entirely concerned with the characteristic sphere of inclination and action associated with particular virtues. Courage corrects undue fearfulness and recklessness alike. Industriousness corrects laziness. Hope is an antidote to the temptations of despair.


Candace Vogler

Temperance corrects excessive or deficient tendencies to seek the satisfaction of sense-appetite. And so on. If what we have in view is natural virtue as Müller teaches us to understand this, then it makes good sense to suppose that the development of one could positively impede the development of others, and also that there already is a corrective in the kind of virtue Foot discusses. Recall that Foot seems to take it that far from forming a reciprocal unity, virtues conflict so seriously that developing one of them will lead to failures to develop others. She does not give a full account of this aspect of her view. She could be thinking, for instance, that if I have something of generosity, humility, patience, and mercy in me—the soft, nurturing, sort of strengths—this aspect of my practical orientation may make it especially difficult to develop the qualities at issue in, for example, Aristotle’s picture of good temper, such as the tendency to be angry in the right way with the right people for the right amount of time in light of an appropriate assessment of the offense that warrants anger (a many-dimensional picture, notice). If I have vast inner reserves of sweetness, I may find that I am anger-challenged. I am off-balance. Foot will presumably agree. Nevertheless, Foot is committed to the view that my gentle disposition has something of virtue in it. Otherwise she could not argue that my virtues are in conflict with other virtues that look to be equally important in leading a good human life. Suppose, however, that the excellence of my gentle nature shows that I have natural virtue in me. Perhaps I was a violent and willful little girl constrained by the firm hand of wise and tender parents. Perhaps I was a frightened, beaten, and abused little girl whose willfulness was trained upon the need to be unlike my parents. Perhaps my circumstances were less extreme. It doesn’t matter. What matters is that I have the kind of constraint that belongs to a gentle disposition but do not yet have the corresponding ethical virtue. You can tell that I don’t because I cannot manage to get angry with anyone even when anger is needful and warranted. A properly developed corrective account of virtue (which is, I think, what we find in Aquinas) will presumably allow intellect and feeling to help take a human being from natural virtue to ethical virtue.

AN EXAMPLE Consider someone who wants to change when she sees that she has a taste for gossip about those of her colleagues who are not among her friends. She cultivates avid curiosity about matters that are none of her business. If the gossip is injurious, she is an avid participant in disgraceful chatter. We are imagining that she understands the problem and is disturbed by it. Having seen herself in a very unflattering light, and having forced herself to pay attention to what she saw, what does she need to do? Although one normally discusses temperance in connection with controlling sense-appetite, appetite can take many forms, and an appetite for gossip

Natural Virtue and Proper Upbringing 155 is no less appetitive for feasting on bits of information rather than slices of pie. The kind of guidance supplied by temperance is important here and will counsel discretion. Our unhappy agent might rely upon her temperance to help her, extending its scope from eating and such to conversation. Why is it important to curb one’s appetite for salacious talk? Any of several things might come into play. Although scintillating gossip tends to come cloaked as a kind of generosity—a sharing of information with those who take an interest—the mood of bad sorts of gossip is rarely generous. Normally, it is anything but generous to those whose lives are serving as fodder for the gossip mill, which is why it tends to circulate outside the hearing of those whose doings are under discussion. Our unhappy agent’s generosity might be at work in identifying and addressing the problem. Being in on the gossip may, at the very least, make it hard for her to feel comfortable interacting with a co-worker because she knows things that she has no business knowing, and this can lead to a well-grounded sense that she is being dishonest or unjust in her dealings with her colleague. The unhappiness of our unhappy agent, that is, could be grounded in her honesty or her justice. It is not that we can’t imagine someone disengaging from gossip for any number of reasons that have nothing to do with virtue. I could stop participating because no one at my workplace ever does anything interesting. However, in imagining somebody working to curb an appetite for gossip as part of a self-improvement effort, we were imagining someone who takes herself to have ethical reasons to change her ways. The problem isn’t boredom. The problem is flawed character. To so much as identify that problem correctly, and see its weight and scope in the right way, our agent requires the strength that she has built in the course of cultivating ethical virtue. She cannot do this on the basis of natural virtue alone, if Müller is right about natural virtue. Natural virtues do not delimit each other. It could be a natural tendency to generosity that is getting her into trouble, after all, and her challenge might be the challenge of developing properly ethical generosity given some insufficiently delimited natural generosity. What is called for in order to change this aspect of her practical orientation is precisely the kind of thing called for in cultivating any sort of virtue—taking the virtuous action (in this case, disengaging from gossip as an obvious first step).

CONCLUDING REMARKS Aristotelians hold that it belongs to mature human beings with their wits about them to pursue good and avoid evil. Aristotelians hold that this bit of practical direction belongs to our nature and is at work in vice and virtue alike. To that extent virtuous and vicious action alike are informed by reason. But right reason—practical wisdom—is charged with shaping and arranging our pursuits, or motives, and our responses to the end of acting and living well. In order to do this, practical wisdom works to realize the ends set by


Candace Vogler

specific virtues in light of an overarching concern with living a good human life. The need to continue the work of virtue in light of such matters as having a taste for gossip is the sort of need that can only be experienced as a need for ethical self-improvement by someone who already has virtue in her. If Müller is right, she will have to have ethical virtue, not just natural virtue, in her in order to undertake ethical self-improvement as an adult. Drawing on Müller’s work, I have urged that the way to understand the advantage given by a proper upbringing is in terms of the ways that it guides children to natural virtue. Developing natural virtue is genuinely helpful in cultivating ethical virtue since, on Müller’s account, natural virtues are virtues rather than just accidents of impulse and temperament. There is an element of constraint in the feelings and acts of those with natural virtue, and that constraint operates in the dimensions characteristic of the relevant corresponding ethical virtues. But a strong assortment of practical tendencies, each of which provides one-dimensional constraint, does not add up to an all-around good character. I took it that noticing this was the root of Foot’s insistence that the virtues do not form a reciprocal unity but instead conflict with one another. This may well be true of natural virtues as Müller teaches us to understand them. It should not be true of ethical virtue, partly because no one ethical virtue can operate well unless it is guided by other virtues. In the course of discussing these matters, I also came out on the side of the view that adult moral self-improvement is possible and that it requires guidance by ethical virtue. This, of course, entails a commitment to a weaker version of the reciprocity thesis than is common among Anglophone philosophers who accept the reciprocity thesis but take it that I must have perfect virtue in order to have any virtue at all. All I have to say in response to such views is: I hope that they are wrong.

NOTES 1. N.B.: English translations of passages from the Summa Theologiae of Saint Thomas Aquinas (hereafter, ST) are those provided by the Fathers of the Dominican Province. ST I–II, q. 49, a. 4, ad. 3: Si autum sit talis forma quae possit diversimode operari, sicut est anima; oportet quod disponatur ad suas operations per aliquos habitus. 2. Müller, ‘Aristotle’s Conception of Ethical and Natural Virtue: How the Unity Thesis Sheds Light on the Doctrine of the Mean’. 3. ‘Appelata est enim a viro virtus: viri autem propria maxime est fortitudo’ [‘The term virtue is from the word that signifies man; a man’s chief quality is fortitude’]; Cicero, Tusculanarum., I, xi, 18. 4. Foot, ‘Virtues and Vices’, 8. 5. Ibid., 9. 6. Ibid., 9. 7. See Gottlieb, The Virtue of Aristotle’s Ethics, 52–72; see also Gottlieb, ‘Are the Virtues Remedial?’. 8. Foot, ‘Moral Realism and Moral Dilemmas’, 397. 9. Müller, ‘Aristotle’s Conception of Ethical and Natural Virtue’, 19.

Natural Virtue and Proper Upbringing


10. Irwin, ‘Practical Reason Divided’, 197–198. 11. I am grateful to Jonathan Lear for suggesting that I think about Aristotle’s treatment of akrasia. 12. Bradley, Aquinas on the Twofold Human Good: Reason and Happiness in Aquinas’s Moral Science, 151. 13. Broadie, Ethics with Aristotle, 245. 14. Müller, Aristotle’s Conception of Ethical and Natural Virtue, 25. 15. Ibid., 26. 16. Ibid., 27. 17. Ibid., 28. 18. Ibid., 30–31. 19. Ibid., 31. 20. Ibid., 38. 21. Ibid., 43. 22. Ibid., 43.

REFERENCES Aquinas, The Summa Theologica of St. Thomas Aquinas, Second and Revised Edition (New York: Benziger Brothers, 1920). Bradley, Denis J. M., Aquinas on the Twofold Human Good: Reason and Happiness in Aquinas’s Moral Science (Washington, DC: Catholic University Press of America, 1997). Broadie, Sarah, Ethics with Aristotle (Oxford: Oxford University Press, 1991). Foot, Philippa, ‘Moral Realism and Moral Dilemmas’, Journal of Philosophy 80, no. 7 (1983), 379–398. ———, ‘Virtues and Vices’, in Virtues and Vices and Other Essays in Moral Philosophy (Berkeley: University of California Press, 1978), 1–18. Gottlieb, Paula, ‘Are the Virtues Remedial?’, The Journal of Value Inquiry 35 (2001), 342–354. ———, The Virtue of Aristotle’s Ethics (Cambridge: Cambridge University Press, 2009), Chapter 3, 52–72. Irwin, T. H., ‘Practical Reason Divided’, in Ethics and Practical Reason, ed. Garrett Cullity and Berys Gaut (Oxford: Oxford University Press, 1997), 189–214. Müller, Anselm, ‘Aristotle’s Conception of Ethical and Natural Virtue: How the Unity Thesis Sheds Light on the Doctrine of the Mean’, in Was ist das für den Menschen Gute/What is Good for a Human Being?, ed. Jan Szaif and Matthias Lutz-Bachmann (Berlin, NY: Walter de Gruyter, 2004), 18–53.

11 Kalou Heneka Timothy Chappell

Think of the glass itself, with its five grand colours stained right through. It was rougher than ours, thicker, fitted in smaller pieces. They loved it with the same fury as they gave to their castles, and Villars de Honnecourt, struck by a particularly beautiful specimen, stopped to draw it on his journeys, with the explanation that ‘I was on my way to obey a call to the land of Hungary when I drew this window because it pleased me best of all windows’.1 Kalou dê heneka ho andreios hypomenei kai prattei ta kata tên andreian.2 Say what you like, so long as it doesn’t stop you from seeing how things are. (And if you see that, there are plenty of things that you won’t say.)3 I This chapter is a meditation on the first two of these epigraphs, in the light of the third. Continuing a project I have begun elsewhere,4 of suggesting ways in which we might expand our repertoire of (recognized) ethical concepts, it explores an idea present in both the first two passages. This is the idea that tou kalou heneka, ‘for the sake of The Fine’, or ‘The Beautiful’, is a name for more than one important kinds of practical-rational intelligibility. One lesson of these explorations is that our practical reasons are much less structured, much more piecemeal, particular, and bitty, than moral philosophers generally like to think. Another lesson is a way to answer the familiar question ‘Why be moral?’. It is with that question that I begin. Typically, when people ask ‘Why be moral?’, they are asking for an explanatory reduction of The Moral to The Prudential, that is, for an argument that we have reason to do what is moral that shows that it is in our prudential interest. The question takes it as read that moral reasons are problematic in some way, whereas prudential reasons are not. Moral reasons require grounding, their force for us is somehow not obvious; whereas it is obvious how prudential reasons are reasons for us. Their force and applicability is self-evident, or something like self-evident.

Kalou Heneka


So far as I can see, we could just as well assume the inverse: that it is moral reasons that are unproblematic and prudential reasons that require explanatory reduction. Then the pressing question would not be ‘What (prudential) reason do we have to be moral?’. It would be ‘What (moral) reason do we have to be prudential?’. A third possibility: we could insist that moral and prudential reasons are both unproblematic so that no explanatory reduction is called for in either direction. Or indeed, to add the fourth permutation in the table, that both are problematic and unobvious—which is more or less the view I shall defend here. Problematic and unobvious

The moral

The prudential

The moral and the prudential

Obvious and unproblematic

The prudential

The moral

The moral and the prudential



(Quietist) dualism

(Skeptical) Dualism

The first of these presumptions, that the moral requires explanatory grounding in the prudential, we may call the prudentialistic presumption. It has probably been shared by enough moral philosophers to be worth calling the default presumption. It5 is there in Plato’s Republic in the mouths of his characters Thrasymachus, Glaucon, and Adeimantus; many interpreters think it is Plato’s own view. The same presumption is there in Hobbes, in Hume, in the classical utilitarians, in Rawls and Gauthier, in Philippa Foot and Rosalind Hursthouse, and in swarms of other writers. The second presumption, that on the contrary the prudential requires explanatory grounding in the moral, we may call the moralistic presumption. I myself think, though not everybody does, that this is Kant’s view; it is certainly Christine Korsgaard’s.6 Roughly, the idea is that moral reasons are the only real reasons there are. Prudential ‘reasons’, even if they come first in the order of discovery, are nowhere near primary in the order of explanation. They are no more than a prolegomenon to set us on the way to understanding what real reasons, moral reasons, are all about. In truth, until we grasp moral reasons, we will not grasp any genuine prudential reasons either. The third presumption can be called the dualistic presumption. (More specifically, the quietist dualistic presumption, in distinction, as in the previous Table, from the fourth possibility, skeptical dualism.) Quietist dualism assumes that moral reasons and prudential reasons are both in their different ways perfectly intelligible so that we do not need an explanatory reduction in either direction. Such is perhaps G. E. Moore’s view in Principia Ethica (or is one of Moore’s views in that testament of confusion) and Prichard’s in ‘Does Moral Philosophy Rest upon a Mistake?’. It is also, apparently, implied by Sidgwick’s ‘dualism of practical reason’, of which, here as elsewhere, both Prichard and Moore were (unwilling) heirs. The idea


Timothy Chappell

is especially clear in Sidgwick: it is that The Moral and The Prudential— capitalizations deliberate—are the two great categories into which all practical reasons exhaustively and exclusively divide. Neither sort of practical reason is reducible to, or fully intelligible in terms of, the other sort. But, Prichard and Moore might insist, that is no license for pessimism about the possibility of a complete and coherent moral system. After all, both sorts of reason are so well-known and familiar to us that there seems to be a sort of disingenuousness about both the prudentialistic and the moralistic presumption: a disingenuous philistinism on the prudentialistic side, and a disingenuous priggishness on the moralistic. (Disdain for philistinism, and for disingenuousness, is clearly one of Prichard’s key-notes in ‘Does Moral Philosophy Rest on a Mistake?’.) We might argue for the moralistic presumption, and against prudentialism, by pointing out that what prudentialists call ‘The Prudential’ is a loose and inchoate ragbag of all sorts of different reasons (or ‘reasons’). I think this is true, but it doesn’t go far enough because something similar is equally true of what moralists call ‘The Moral’. More about this in Section II. We might argue for the quietist dualistic presumption by appealing—as Prichard, Moore, and Sidgwick all do, in their different ways—to the intuitive immediacy of our understanding both of The Moral and The Prudential. I think this too is right but doesn’t go far enough; there aren’t just two categories of practical reason of which we have an intuitively immediate understanding. More about this in Sections IV and following. Many will want to argue for the prudentialistic presumption, and against moralism, on evolutionary grounds. Prudential reasons, they will say, are the currency of evolutionary explanation, and moral reasons aren’t. That’s why moral reasons need explanatory reduction—to prudential reasons— and prudential reasons don’t. Consideration of this argument for prudentialism I will not put off until later; it is easily separable from my main objectives, and it is worth pausing briefly to note just what a bad argument it is, influential though it unfortunately is in current moral philosophy. For one thing, even if prudential reasons were the currency of an evolutionary explanation of how in the past we came to be what we now are, it would still be the genetic fallacy, naked and unashamed, to infer that all our real reasons now must either be prudential reasons, or be smoothly reducible to prudential reasons. For another thing, and in spite of the way far too many otherwise intelligent people now tend to talk, prudential reasons just aren’t the currency of evolutionary explanation. Evolutionary explanation is not necessarily about reasons at all; it is about offering partial explanations of how creatures are the ways they are because selective pressures have adapted them to their environment so that they normally survive long enough to reproduce.7 Moreover, it follows from evolutionary theory itself that it cannot possibly explain everything. For once they are at work, evolutionary processes constantly produce all sorts of effects that are not themselves remotely explicable

Kalou Heneka 161 by appeal to ‘survival value’, but rather are either side-effects or incidental effects of what evolutionary pressure is directly responsible for (these are Dennett’s ‘spandrels’)—or else are simply pieces of genetic-mutational randomizing (‘free lunches’, as they are often called). But both spandrels and free lunches, once they exist, can have further effects and implications of their own, which, even if they do not have directly negative effects on survival value—though that can happen, too—are certainly capable of creating new possibilities that are broadly neutral with respect to survival value, and which appeal to survival value therefore plays little if any role in explaining. The science of anthropology is replete with simple and obvious examples of the sort of thing I mean. To take the broadest and most obvious example of all, perhaps having a culture is evolutionarily advantageous to hominids. If so, we can predict on evolutionary grounds that hominid populations will tend to develop cultures. However, evolutionary theory predicts little or nothing about what sort of culture hominids will have. Even what it does predict is frequently falsified since a culture’s characteristics can, in obvious ways, have negative survival value, and yet the culture survive. Once any sort of culture exists in a population, we have a whole new motor for the development of that population. That this motor could drive the population, not only in directions that make no difference one way or the other to survival value, but even in anti-evolutionary directions, is not just speculation. It is historical fact. Prudential reasons—at least as normally understood—are reasons of individual well-being, reasons that have to do with what promotes the individual’s health, wealth, and happiness. Adaptation for reproduction-facilitating survival has nothing directly to do with this, as any male spider could tell you.

II Suspicion about the moral as a category can take a number of forms. What I have labeled the prudentialistic presumption is itself one of them: here the suspicion is that moral reasons, to be made rationally intelligible, need to be explanatorily reduced to practical reasons in the other category, the category of the prudential. As will be clear already, I have my suspicions about this sort of suspicion. A different sort of suspicion about the moral is more to my present purposes. This is the sort of suspicion you get in Anscombe, MacIntyre, and Williams: the suspicion that ‘the moral’ just isn’t the name of any unitary category that does much if any interesting work in justifying and explaining what actual good agents characteristically do. At the deliberative level, there may be some things that such agents do in which such agents are motivated by the thought ‘Because it’s moral’. But such agents have lots of other motivating thoughts—‘Because fairness requires it’, ‘Because I promised’, ‘Because


Timothy Chappell

she’s my wife’, ‘Because we are friends’, ‘Because you are sinking’, ‘Because they are starving’, ‘Because otherwise he will be disgraced’, ‘Because it is suffering’, ‘Because my last delivery was a no-ball’, and so on indefinitely; the V-thoughts, as Rosalind Hursthouse calls them. Clearly none of these is just another way of thinking the ‘Because it’s moral’ thought; arguably many of them are not even consistent with that thought, and/or with the conscious entertainment of that thought. So it isn’t clear why the ‘Because it’s moral’ thought should be supposed to be a specially deep or important form of motivation for good agents. Similarly at the criterion of rightness level, there may be a few things that good agents characteristically do in which such agents are justified simply and directly by the consideration ‘Because it’s moral’. But such agents can have lots of other justifications, many of them very different from this justification and nearly all of them more informative. As before, what they do may be justified by considerations about fairness, or promises, or someone’s being my wife, or a friendship, or shipwreck, or famine, or disgrace, or pain—and so on indefinitely. Hence it is no clearer why the consideration ‘Because it’s moral’ should be a particularly special or basic justification than it is why the thought ‘Because it’s moral’ should be a particularly special or basic motivation for good agents. (Or should we take the domain of practical justification to be theoretically unified in a way the domain of practical deliberation conspicuously isn’t? It is hard to see why we would; more about that in section III.) At both the deliberative and the justificatory levels, to insist on the priority or basicity of appeals to The Moral looks like mere stipulation. There is no obvious explanatory gain in this redescription. If it is taken, as often, to be the right way to marshal the phenomena, to the exclusion of other ways of looking at them that may prove equally or even more fruitful, then we risk explanatory loss, too: this insistence may well obfuscate the real structure of our deliberative and justificatory practices. These phenomena about the multifariousness of the moral have been extensively studied by ethicists and moral psychologists. The thought is familiar—and I think true—that ‘the moral’ is not really the name of any wide-ranging and sharply defined category of practical reasons at all. What may be less familiar, but I think is equally true, is that something analogous applies to the prudential. ‘The prudential’ is supposed to be about what is advantageous, or beneficial, or in the agent’s own interest, or what furthers the agent’s well-being. There are serious ambiguities in all of these notions. ‘Advantage’ and ‘benefit’ and ‘interest’ and ‘well-being’ are concepts of which there are indefinitely many competing accounts, both philosophical and informal. These words can mean many different things, and there is no particular reason—aside from theory—to expect their extensions to converge at all neatly. In perfectly ordinary senses of the words, forgoing a pay-rise to impress my boss may be to my advantage but does not benefit me, while health-threatening and

Kalou Heneka


anxious dedication to my work as a novelist or a famine-relief worker may (in a way) be in my interest but does not further my well-being. ‘But that means I have an implicit notion of benefit, or advantage, or interest, or well-being, such that I value one of these more than the other according to it. It means I think that one of these scores higher on some scale than the other’. There is a strong inclination to say something like this. Yet there is no good reason why we should say something like this rather than saying, simply, that the agent prefers one of these options to the other. Philosophers are convinced that there must be some scale of well-being against which to calibrate such preferences. The fruitless, but not yet abandoned, search for that scale is not a search for something objectively ‘out there’ that, once found, could settle questions about the nature of ‘prudential wellbeing’ in favor of a unitary account of it. On the contrary, it depends on the prior assumption that prudential well-being has a unitary nature—the very assumption that I am questioning here. Nor is there any reason—again, aside from theory—to think that these ‘prudential’ notions can easily or conveniently be kept clear of what the proponent of the moral/prudential contrast would like to call moral connotations. This is especially obvious with ‘well-being’, but a similar ‘mixing of the prudential and the moral’ can be imaginatively effected just by adding real on the front of the other terms—‘real advantage’, ‘real benefit’, ‘real interest’. If the prudential is to be a category clearly distinct from the moral, having to do with the agent’s individual well-being, and fit to serve as an explanatory foundation for the moral, then this cross-infection of the moral and the prudential cannot be allowed. But suppose well-being turns out to be a notion that cannot be properly understood except when it is given a moral loading. This is likely to be how things turn out if, as seems plausible, inquiring what to count as well-being is not a value-neutral anthropological enterprise but a key part of constituting one’s own moral character.8 In this case the cross-infection of the moral and the prudential is inevitable. And so, to switch metaphors, the idea of an exclusive moral/prudential distinction is already holed below the waterline. Alongside these two problems, there is a third. The very idea of the prudential, as most commonly understood, appears to rest on an obvious falsehood. In its most typical form, the category of the prudential is meant to fit both of two criteria: (a) It is supposed to be about people acting in pursuit of their own interest (or welfare or advantage or whatever). (b) It is supposed to be definable by exclusion from the category of the moral: the moral and the prudential are supposed to be an exhaustive and exclusive pair of categories that between them cover every case of having a practical reason. But (a) and (b) together imply that whenever someone acts on a non-moral reason, he acts in pursuit of his own interest. Since ‘non-moral reason’ and ‘in pursuit of his own interest’ are both extremely vague phrases, it is difficult to conclusively refute this. But on any commonsensical understanding of the words, it is false. The man who works himself half to death to please his beloved, or to


Timothy Chappell

perfect his conducting of the Eroica Symphony, acts on a non-moral reason: his reason is romantic, or musical, or what you will. Does that mean that he acts ‘self-interestedly’? A natural thing to say is that what he is doing is against his own interest; he is making a great and dangerous sacrifice of his own interests for those of his lover or his art. In this sense it is obvious that not every non-moral reason is a reason of self-interest. So evidently not all practical reasons are either moral or else prudential—unless the prudential becomes so wide a category that it is not really a category at all.

III The moral has no clear structural unity; nor does the prudential, which is also very hard to articulate without the importation of moral elements that threaten the supposed exclusiveness of the moral/prudential distinction; the supposed exhaustiveness of that distinction seems highly questionable, too. These are the basic problems about ‘the moral’ and ‘the prudential’. These are the reasons that motivate me in adopting what according to the previous Table we may call skeptical dualism, though really the position is so much a skepticism about the moral and the prudential that it is not really a dualism at all—the whole point is that we can’t divide the phenomena at all neatly into just these two classes. As Plato in effect pointed out in the Republic, these problems are not well finessed by a Thrasymachean cynicism about the moral, precisely because of the instability of the notion of the prudential—especially when that notion is not allowed to ‘cross-infect’ with the moral in the way described previously. Nor are they well finessed by the kind of moralism we get in Kant, who will happily accept that the prudential is a shapeless mess, provided he is allowed to say that there is something that unifies the moral—namely, the move to universalizability. Familiarly, the trouble with this, as I have argued elsewhere,9 is that it isn’t at all clear that the universalizable and (what we might intuitively call) the moral coincide. Some other accounts of how to characterize ‘the moral’ succumb even faster to even more obvious problems. William Frankena, for example, stipulates in one well-known discussion that one is not ‘taking the moral point of view’ unless, inter alia, ‘one is willing to universalize’, and one’s ‘reasons for one’s judgments consists of facts about what the things judged do to the lives of sentient beings in terms of promoting or distributing non-moral good and evil’.10 Of course (see my third epigraph) we can use the word ‘moral’ however we like, but it is hard to see the point of using it in such a way that non-universalizable moral judgments, and moral judgments about the environment, are ruled out by definition. A similar objection can made to Catherine Wilson’s more recent suggestion that ‘There is an anonymity requirement on moral theorising’. If she wants to use the words this way, she is free to propose that we should see any endorsement of a partial norm

Kalou Heneka


as ‘ideology’, not as ‘moral theory proper’. The trouble with her proposal is that it just seems like a stretching of the sense of ‘moral’, and indeed a case of ideology.11 Wilson’s anonymity constraint entails that there can be no such thing as an individual’s moral style—no such thing as approaching the problems that arise for practical decision in line with any particularities of character. But this is no trivial loss to our ethical thought; it is a disabling deprivation.12 In reality, I suspect we barely know what it would mean to be her sort of moral reasoner. My point is not that, since neither notion can be cleanly and exclusively defined, we should just junk the notions of the moral and the prudential. On the contrary, I think we should rehabilitate them. There are some actions— not many, but some; perhaps handing in a lost tenner to the police station—that really are motivated and justified by nothing else but the thought ‘Because it’s moral’. With these actions we find the legitimate scope of the notion of the moral. Similarly, there are some things that people do—not everything, not perhaps even all that many things, but some: perhaps applying for a stop-gap job to pay the mortgage ‘just till something better comes up’—in which what motivates them, and perhaps even justifies them, really is solely and simply a concern with the agent’s own advantage or interest in some clearly non-moral sense. With these actions we find the legitimate scope of the notion of the prudential. The notions of the moral and the prudential work fine in these, their home territories. The question to the systematizing moral theorist is why we should feel any impulse to insist that these are the only two basic-level notions that we can deploy to think about our practical reasons in any territory.

IV Recall the sheer variousness of the things that can appear explanatorily basic to people whose rationality is, we would normally say, indisputable. The more we understand this variety, the less we will be tempted by the thought that, in ethics, we face a large and pressing task set for us by the opening question, ‘Why be moral?’: the task of providing an explanatory reduction of the moral to the prudential. As if the prudential were somehow the universal currency of practical reasoning; as if the prudential were an exchange for and a measure of every other sort of practical reason. If my argument so far is right, nothing of this sort can possibly be true. What we should pursue in the theory of practical reasons is not chimerical unificatory projects like this; it is an exploration and assessment of the diversity and disunity of our actual practical reasons. One example, the example I discuss in ‘Glory as an Ethical Idea’, is practical reasons having to do with glory. Another example is humor. A third—the one that I want to discuss in the rest of this chapter—is, What if someone were to say ‘I did it because it was beautiful’?


Timothy Chappell

One form of this is in effect what Villars de Honnecourt is saying in my first epigraph. The possibility of saying it, and saying it in many different ways, is arguably what Aristotle affirms in my second. ‘In many different ways’: of course there is a difference between Villars de Honnecourt, who does what he does because some object (the window) is beautiful, and Aristotle’s andreios, with whom the point is, I take it, that what he does is to-be-done because it, his action, is beautiful or fine, kalos. Different again is a case that Simone Weil describes: It is the beauty of the world which compels the man who is drained empty, the man who has spent all his inheritance, all his energy, to remember that his father’s slaves have more of a share in the good than he who is the son. The share that things have in the good, the wages of the slaves of the Father—it is beauty.13 Here someone changes his whole way of living (Weil is thinking of the gospel parable of the Prodigal Son, Luke 15.11–32) because he comes to see his present life in a new light that is cast over it by the illumination of something like the Platonic ideal of Beauty. Further cases, coming shortly, introduce still further differences. I do not want to discount these differences but, on the contrary, to emphasize them. I am not proposing that, alongside the two familiar monolithic categories of practical reasons, The Moral and The Prudential, we recognize a third monolithic, capitalized category, Reasons To Do With Beauty. Still less, if less is possible, do I mean to suggest that this category should be called The Aesthetic. I am not saying that if any practical reason is neither Moral nor Prudential, then it must be Aesthetic. As far as I am concerned, the more other categories of practical reason there are—and the less monolithic and indeed capitalized they are—the better.

V These caveats aside, the idea that the beautiful can give us reasons is—as my first and second epigraphs demonstrate—of course not new. In a classical Greek context, it is not even, so far forth, controversial. Aristotle, Pericles, and Plato disagree about many things; not about this. With Aristotle’s thesis, in my second epigraph, that the brave man does his brave deeds kalou heneka, ‘for the sake of the fine’ (or ‘the beautiful’ or ‘the noble’, as it is also sometimes translated), compare Socrates’s claim at Gorgias 477c8 that what is worst (aiskhiston) is also what is ugliest, and Pericles’s famous words in the Funeral Oration: They resigned to hope their unknown chance of happiness; but in the face of death they resolved to rely upon themselves alone. And when the moment

Kalou Heneka


came they thought it more beautiful (kallion hêgêsamenoi) to stand firm and die, rather than to fly and save their lives; they ran away from the word of dishonour (to men aiskhron tou logou), but on the battlefield their feet stood fast, and in an instant, at the height of their fortune, they passed away from the scene, not of their fear, but of their glory (doxês).14 The ideas that ‘I did it because it was beautiful’ can be a rationally intelligible motivation, and that the beauty of ways of acting can be an important aspect of their goodness are not exclusively pagan Greek ones either. If Plato speaks of the beauty of good people’s actions (Symposion, 210c), so too does Aquinas: Aquinas thinks the medium of claritas—light or resplendence—which enables us to see and take pleasure in beauty is a central feature of the form of the object itself . . . Aquinas follows Pseudo-Dionysius in holding that claritas is rooted in reason, which he describes as ‘the light that makes beauty seen’ [ST]. There is a ‘clarity of reason’ which gives a spiritual beauty to our actions when they are well directed towards reason.15 More recently—and for sure, very differently—from Malcolm Muggeridge’s significantly titled book Something Beautiful for God, here is Muggeridge’s description16 of his own reaction to meeting Mother Teresa of Calcutta. Arrestingly enough, this description invokes (whether knowingly or not, I have no idea) something like the Thomistic idea of claritas: When the train began to move, and I walked away, I felt as though I were leaving behind me all the beauty and all the joy in the universe. Something of God’s universal love has rubbed off on Mother Teresa, giving her homely features a noticeable luminosity; a shining quality. She has lived so closely with her Lord that the same enchantment clings about her that sent the crowds chasing after him in Jerusalem and Galilee, and made his mere presence seem a harbinger of healing. Outside, the streets were beginning to stir; sleepers awakening, stretching, and yawning; some raking over the piles of garbage in search of something edible. It was a scene of desolation, yet it, too, seemed somehow irradiated. This love, this Christian love, which shines down on the misery we make, and into our dark hearts that make it; irradiating all, uniting all, making of all one stupendous harmony. Momentarily I understood; then, leaning back in my American limousine, was carried off to breakfast, to pick over my own particular garbage-heap. I want to suggest that the answer to ‘Why be moral?’ is quite often ‘Because that is the beautiful thing to do’. It’s not that the moral act is itself prudentially disastrous, but just happens to be, unfortunately enough, one of a class


Timothy Chappell

to the whole of which we are somehow committed, if we are committed to any part of it—as theories like rule-consequentialism and Gauthier’s contractualism sometimes suggest. Nor is it that the moral act is prudentially advantageous in some way—just a very obscure way, one which is consistent with the fact that the moral act is attended with terrible penalties like those that Hans and Sophie Scholl faced, or those described by Callicles at Gorgias 486b. We need not think that there is any prudential advantage, in any sense, in the gravely sacrificial moral act. At least sometimes, advantage simply isn’t the point. It is rather that the moral act demands to be done even if it does involve a grave sacrifice—just because it is beautiful. Perhaps this appeal to to kalon, The Beautiful, is the answer to the puzzlement expressed by the person who said of Sophie Scholl and those who suffered with her that ‘the fact that five little kids, in the mouth of the wolf where it really counted, had the tremendous courage to do what they did, is spectacular to me. I know that the world is better for them having been there, but I do not know why’.17 Perhaps it is also the best answer to the difficulty that Philippa Foot is struggling with in the following two paragraphs: One may think that there was a sense in which the Letter-Writers18 did, but also a sense in which they did not, sacrifice their happiness in refusing to go along with the Nazis. In the abstract what they so longed for—to get back to their families—was of course wholly good. But as they were placed [facing imminent execution for involvement in the German resistance to Hitler] it was impossible to pursue this end by just and honorable means. And this, I suggest, explains the sense in which they did not see as their happiness what they could have got by giving in. Happiness in life, they might have said, was not something possible for them [. . .] Yet this is not the heart of the matter. For supposing that they had been offered a ‘Lethe-drug’ that would have taken from them all future knowledge of the action [of giving in to the Nazis]? They would not have accepted. And there would have been a way in which they would not have felt that happiness lay in acceptance. ((To see it as happiness they would have to have changed, and would not have accepted the prospect of such a change [. . .] one would not wish for the sake of friends one loved that ‘in the tight corner’ they would be able to forsake their virtue in time.))19 [. . .] Happiness isolated from virtue is not the only way in which the concept is to be found in our thoughts. The suggestion is, then, that humanity’s good can be thought of as happiness, and yet in such a way that combining it with wickedness is a priori ruled out.20 Throughout her philosophical career, Foot maintained an underlying commitment to the thesis that the virtues benefit their possessor. In the first of these paragraphs she questions that commitment: perhaps the Letter-Writers

Kalou Heneka


were not really after happiness at all since what they would have counted happiness was not accessible to them without surrendering something else that mattered more to them than happiness. Then in the second she apparently has a different thought: perhaps there is a sense in which the LetterWriters achieved happiness since for them happiness could not be isolated with virtue but is bound up with it, even if virtue means death in a Nazi prison.21 One problem with this second thought is that the most it shows is that the happiness of the good martyr is one sort of happiness, not the only sort. In which case you do not need to be Polus to wonder what the martyr’s sort of happiness has going for it, compared with other possible sorts: the happiness of a comfortable tyrant like Perdiccas, for example. When eudaimonism is forced into such extremes as these, it may seem time to ask whether eudaimonism is better not given up altogether. Maybe the point about the saintly martyr is not that he acts on an imperative generated by concern with happiness at all, but by concern with what is beautiful, or fine, or noble. I don’t of course mean that this concern must appear, in any crude and direct de dicto way, in the martyr’s reasoning (‘I must do what is beautiful; this is beautiful; so I must do this’). The problems with such crass pictures of the psychology of virtue—or for that matter with the hedonist’s psychology—are exceedingly familiar. On any more realistic picture, the concern with the beautiful is more likely to be part of the frame of the agent’s deliberations than an item in those deliberations: normally at least it will be, not so much another thing she deliberates about, as a condition on how she deliberates about anything. This is another use, perhaps, for the notion of styles of moral reasoning.

VI Suppose then that we can, sometimes, justify an action by appeal to to kalon: in English and capitalized (but not I hope monolithic), to The Beautiful. We justify it by saying that the action deserves to be done because it is concerned with a beautiful object, like Villars de Honnecourt’s, or because the action itself displays intrinsic beauty in one way or another, like the deeds of Foot’s Letter-Writers or Aristotle’s or Pericles’s andreioi. Then is that all we can say about the action? A dilemma seems to face me here, which Tom Hurka, in conversation, has helped me formulate. If I can say no more to justify the action than that it is beautiful, or is done for the sake of the beautiful, then my position looks obscurantist; it seems to run us very quickly into a philosophical and explanatory dead-end. If on the other hand I can say more to justify the action than this appeal to beauty, then the suspicion will be that it is this ‘more’ that does the real explaining, and then my alleged category of practical reasons to do with The Beautiful will turn out to be reducible to some other category after all.


Timothy Chappell

The dilemma is worth thinking about; when we think about it, however, we soon see that neither horn is really threatening. On the one side, there is often plenty more we can say to illuminate the justification of some action as beautiful. For instance: ‘it’s beautiful because of what it expresses’; ‘it’s a beautiful action because of its symbolic power’; ‘it’s a beautiful action because of the pure courage that it displays’; ‘it’s a beautiful action because of its appositeness, its wittiness almost’; ‘it’s a beautiful action because of its grace’. (What a fertile notion grace is, incidentally. Colloquially, being graceful is, as just said, one way among others of being beautiful; theologically, grace is a gratuitous generosity that transfigures its recipient. Colloquial grace and theological grace are not merely accidental homophones.) It is a mistake to think that remarks like these subsume the beautiful under some more basic justificatory category, rather than teaching us “our way about” the justificatory category of the beautiful itself. On the other side, sometimes that an action is beautiful is all that we need to say, simply because that justification applies, and none of its indefinitely possible defeaters has been activated. No doubt in such cases we could, in principle, rephrase ‘The deed was the thing to do because it was beautiful’ as ‘The deed was the thing to do because it was beautiful, and the justificatory power of the beautiful was not defeated by any other considerations in this context’. In cases where the justificatory force of the beautiful isn’t defeated, it will normally be superfluous to say so: just as, if you want to explain why something falls to the ground, it is typically enough to say ‘Because of gravity’, and superfluous to add ‘and because nothing overrides gravity in this case’. None of this is to say that the justificatory force of the beautiful is never defeated. Of course it is sometimes: an action can fail to be justified, beautiful though it is or seems, for indefinitely many reasons. Perhaps the action’s cost in some other currency is simply too high; or perhaps the action involves an aspect-blindness about cruelty or some other vice that cannot be ignored in the deed; or perhaps the action is not, in fact, beautiful at all—it is graceless or exaggerated or twisted in some way; or perhaps the action has the symbolic or expressive force of expressing something false, or something corrupting, or perhaps it rests on some sort of misunderstanding—and so on indefinitely. My thesis is not that the beautiful always justifies, only that it sometimes does. When the justificatory force of the beautiful isn’t defeated, we usually can add something about why it isn’t defeated, as and when occasion demands. One case where occasion so demands comes in a Gospel story: And being in Bethany in the house of Simon the leper, as he sat at meat, there came a woman having an alabaster box of ointment of spikenard very precious; and she brake the box, and poured it on his head. And there were some that had indignation within themselves, and said, Why was this waste of the ointment made? For it might have been sold for

Kalou Heneka


more than three hundred pence, and have been given to the poor. And they murmured against her. And Jesus said, Let her alone; why trouble ye her? she hath wrought a beautiful work on me. For ye have the poor with you always, and whensoever ye will ye may do them good: but me ye have not always. She hath done what she could: she is come aforehand to anoint my body to the burying. Verily I say unto you, Wheresoever this gospel shall be preached throughout the whole world, this also that she hath done shall be spoken of for a memorial of her. And Judas Iscariot, one of the twelve, went unto the chief priests, to betray him unto them.22 According to Jesus the justification of the woman’s action lies simply and straightforwardly in its beauty. (Kalon ergon êrgasato en emoi, influentially mistranslated in the King James Version as ‘she hath wrought a good work on me’. That the deed’s beauty is the point, as we would expect from the word kalon, is underlined by a detail that St John adds in his version of the story: ‘the house was filled with the fragrance of the perfume’, Jn 12.3.) The objection to the woman’s action is a form of the objection that its cost is too high; an objection that we will probably hear a lot against actions motivated by beauty. The objection is that the action—in which the woman was making a huge financial sacrifice—is wasteful and inefficient from the point of view of public utility. Jesus responds that there are cases where beauty overrides public utility. This is one of them because of the grace (both theological and colloquial) and the exactly apposite symbolic value of the woman’s action. The passage gives us a concise, but rather rich, example of how a debate about the relative justificatory powers of beauty and utility might be intelligible: for what Jesus says to justify the woman’s deed is perfectly intelligible, even if we do not accept it. (To many today Jesus’s justification of the action as a symbolic preparation for the arrest, torture, mock trial, and mob-justice execution that were about to happen to him will no doubt be a scandalous one, just as it seems to have scandalized some of his hearers at the time: Judas, for instance, for whom ‘the poor you have always with you’ seems to have been the last straw that provoked that good utilitarian into betrayal.)

VII On the picture of the theory of practical reasons that I am recommending, rationally intelligible practical reasons can take all sorts of forms that have nothing much to do with, and are not easily assimilable to, either The Moral or The Prudential. On this picture, there is no general philosophical pressure toward the unification of all our practical reasons under one, or two, or any small number of types. Certainly to understand any agent or practical reasoner as practically rational, we must be able to see her reasons


Timothy Chappell

as falling under some rationally intelligible type of reason. But to think that ‘For every practically rational action, there is some intelligible reason-type’ entails ‘There is one intelligible reason-type for every practically rational action’ is to commit a simple scope-fallacy. Setting that fallacy aside, we can see how to replace a ‘deep and narrow’ theory of practical reasons, such as the moralism or prudentialism described in Section I, with (if you like) a ‘wide and shallow’ account. On this sort of account—if a name is wanted, call it pluralism—our concern is still to find the intelligibility of practical reasons. It’s just that we deny the moralistic or prudentialistic assumption that there is only one place to find this intelligibility and that we must dig deep in order to find it. Rather, intelligibility can be found in all sorts of places in the landscape of practical rationality. Very often—more often than not, in fact—the intelligibility of a deed is right in front of our eyes, and no digging at all is necessary. NOTES 1. White, The Once and Future King, 577. 2. Aristotle, Nicomachean Ethics 1115b23: ‘It is for the sake of the fine that the courageous man stays at his post in battle, and does the things that are in accordance with courage’. Cp. NE 1120a23–24. 3. Wittgenstein, Philosophical Investigations I, 79, my own translation. 4. Chappell, ‘Glory as an Ethical Idea’. 5. Or something like it. Actually Plato’s concern in the Republic is with justice, not ‘the moral’, and this is not merely a verbal matter. It is more about the fissile nature of ‘the moral’ as a category in due course. 6. Korsgaard, Self-Constitution. 7. Agreeing with G. E. Moore makes me nervous, but here he seems right for once (he has Herbert Spencer in his sights—so perhaps this is easy shooting): ‘The survival of the fittest does not mean, as one might suppose, the survival of what is fittest to fulfil a good purpose . . . it means merely the survival of the fittest to survive . . . the value of the scientific theory [of evolution] . . . just consists in showing what are the causes which produce certain biological effects. Whether these effects are good or bad, it cannot pretend to judge’ (Principia Ethica, 48). 8. See Harcourt, ‘Self-Love and Practical Rationality’. 9. Chappell, ‘Intuition, System, and the ‘Paradox’ of Deontology’. 10. Frankena, Ethics, 113–114. 11. Wilson, Moral Animals: Ideals and Constraints in Moral Theory, 3–4. 12. For more about the notion of ‘moral styles’ see Williams, ‘Persons, Character and Morality’, and more recently Kekes, Enjoyment, e.g., 136: ‘the lasting enjoyment of life depends on developing a style of life that reflects one’s individuality’. 13. Weil, Cahiers XVI, 264. 14. Thucydides 2.42, tr. Jowett, with alterations. 15. Ramsay, Beyond Virtue, 135. See also Aquinas, ST 16. Muggeridge, Something Beautiful for God, 17–18. 17. Lillian Garrett-Groag, quoted in the Wikipedia article on Sophie Scholl. The remark is quoted—from the same source—and discussed by Stump, Wandering in Darkness, 149 and 549.

Kalou Heneka


18. That is, the contributors to Dying We Live, an anthology of letters written by members of the German resistance to Hitler in their last days before execution. 19. Foot puts what I have double-bracketed into a footnote, but I think it is an integral part of her main argument. 20. Foot, Natural Goodness, 95–96. 21. Compare my own remark, previously, that deciding what to count as happiness is not value-neutral anthropology, but an act of moral self-constitution. As we shall see, it does not follow from this remark that happiness is always what is in question in our key moral decisions. 22. Mk 14.3–10; KJV with one word altered—as explained in the main text, I have changed ‘good’ for kalon to ‘beautiful’. Cp. Mt 26.6–13, Lk 7.36–50, Jn 12.1–8.

REFERENCES Anscombe, G.E.M., ‘Modern Moral Philosophy’, Analysis 33 (1958), 1–19. Chappell, Timothy, ‘Glory as an Ethical Idea’, Philosophical Investigations 34 (2011), 105–134. ———, ‘Intuition, System, and the ‘Paradox’ of Deontology’, in Perfecting Virtue, ed. Wuerth and Jost (Cambridge: Cambridge University Press, 2011), 271–288. Foot, Philippa, Natural Goodness (Oxford: Oxford University Press, 2001). Frankena, William K., Ethics (Upper Saddle River, NJ: Prentice Hall, 1963). Harcourt, Edward, ‘Self-Love and Practical Rationality’, in Morality and the Emotions, ed. Carla Bagnoli (Oxford: Oxford University Press, 2011), 82–94. Hursthouse, Rosalind, On Virtue Ethics (Oxford: Oxford University Press, 1999). Kekes, John, Enjoyment (Oxford: Oxford University Press, 2008.) Korsgaard, Christine, Self-Constitution (New York: Oxford University Press, 2009). MacIntyre, Alasdair, After Virtue (London: Duckworth, 1981). Moore, G. E., Principia Ethica (Cambridge: Cambridge University Press, 2003 [1905]). Muggeridge, Malcolm, Something Beautiful for God (London: Harper One 1986 [1971]). Prichard, Harold, ‘Does Moral Philosophy Rest upon a Mistake?’ Mind 21 (1912), 21–37. Ramsay, Hayden, Beyond Virtue (London: Macmillan, 1997). Sidgwick, Henry, The Methods of Ethics (London: Macmillan, 1874). Stump, Eleonore, Wandering in Darkness (Oxford: Oxford University Press, 2010). Weil, Simone, Cahiers XVI, in Oeuvres Complètes, ed. André Devaux and Florence de Lussy (Paris: Gallimard, 2006). White, Terence Hanbury, The Once and Future King (London: Harper Collins 1996 [1958]). Williams, Bernard, ‘Persons, Character, and Morality’, in Moral Luck, ed. B. Williams (Cambridge: Cambridge University Press, 1981), 1–19. Wilson, Catherine, Moral Animals: Ideals and Constraints in Moral Theory (Oxford: Clarendon Press, 2004).

Part II

Beyond (Neo-)Aristotelian Virtue Ethics

12 A New Metaphysics for Virtue Ethics Hume Meets Heidegger Christine Swanton

There are two basic ways of placing Aristotelian ethics in contemporary critical perspective. One may subject the contemporary development of Aristotle’s virtue ethics to (continued) criticism, or develop an alternative account of virtue ethics that can be set alongside the dominant approach. In this way virtue ethicists have a choice: they may rebut or overcome the criticisms of contemporary Aristotle-inspired ethics, or offer new paradigms inspired by other thinkers who may be read as virtue ethicists (e.g., Hume) or whose metaphysics can be applied to virtue ethics. In this chapter I choose the second option, outlining an alternative metaphysics for virtue ethics. An attractive feature of neo-Aristotelianism is its naturalism. However, a broad range of virtue ethical theories suppose that naturalistic conceptions of human nature constrain the adequacy of conceptions of virtue and vice. Hume, for example, claims in the Enquiries Concerning the Principles of Morals: ‘Men are now cured of their passion for hypotheses and systems in natural philosophy, and will hearken to no arguments but those which are derived from experience. It is full time that they should attempt a like reformation in all moral disquisitions; and reject every system of ethics, however subtle or ingenious, which is not founded on fact and observation.’1 Nonetheless Hume believes that science alone cannot make the world of ethics intelligible. What else might be necessary? The problem is posed by McDowell: it is one thing to recognize that the impersonal stance of scientific investigation is a methodological necessity for the achievement of a valuable mode of understanding reality; it is quite another thing to take the dawning grasp of this, in the modern era, for a metaphysical insight into the notion of objectivity as such, so that objective correctness in any mode of thought must be anchored in this kind of access to the real.2 The problem then is this. How can ethics be understood as a ‘kind of access to the real’? What would anchor that mode of thought to the real if the ‘impersonal stance of scientific investigation’ cannot alone do the job? As


Christine Swanton

we explore further later in the chapter, the ‘facts and observations’ anchoring the ethical mode of thought’s ‘access to the real’ are for Hume emotional construals of the world, and feelings (‘passions’) as part of that construal.3 These construals are requisite for, indeed constitute, the very intelligibility of ethics. That is the heart of Hume’s ‘moral sentimentalism’. For example, without the construal of persons through the emotions of affection and tenderness there could be no such thing as a virtue of tenderness (one of Hume’s virtues). The empiricism characterizing Hume’s ethics is not fundamentally an empiricism grounded in sense impressions, though facts and observation of that type are necessary for an adequate ‘system of ethics’. Indeed Hume claims ‘the most abstruse speculations concerning human nature, however cold and uninteresting, become subservient to practical morality; and may render this latter science more correct in its precepts, and more persuasive in its exhortations’ (T 621). In short, Hume’s moral sentimentalism can be understood as an emotional response dependence ‘system’ of ethics that benefits from the input of scientific ‘speculations’ concerning human nature. In that sense, it is a ‘science’. In the contemporary literature, however, the way in which response dependence theories of ethics have been formulated has been subjected to several objections. I shall argue that construing such theories in a way suggested by the thought of Hume and Heidegger overcomes these objections. The next section provides the basic account that is subsequently elaborated through the medium of Heidegger’s ‘hermeneutic phenomenology’,4 and then applied in that form to Hume’s moral sentimentalism. VIRTUE ETHICS AND RESPONSE DEPENDENCE The core idea behind response dependence in ethics is this: sensibilities to ethical properties such as being virtuous or valuable are essential to the intelligibility and thereby the existence of those properties as ethical properties. That idea can be expressed in the following way: (I) An ethical property is response dependent if and only if the property is open to certain responses or construals in responders having appropriate sensibilities, and these responses or construals are what make the property intelligible as an ethical property. Without that mode of intelligibility, the property (such as being courageous, being generous, or being patient) could not exist as an ethical property (namely a virtuous trait), though it could exist as a property determined by other modes of intelligibility. Call (I) the Intelligibility Thesis. The Intelligibility Thesis is about the very existence of ethical properties as ethical properties. A similar thesis applied to redness is a thesis about the existence of redness as a phenomenological

A New Metaphysics for Virtue Ethics


‘secondary quality’: it is not a thesis about the existence of redness as a ‘microscopic textural property’ of the surface of an object.5 (I) has generally been construed as a claim that ethical properties are relational, rather than monadic, as in (R): (R) A person’s being virtuous consists in her virtuousness evoking some relevant response. Call (R) the Relational Thesis. (R) conceives of response dependent properties as relational rather than monadic. To answer the question what and whose responses are relevant the concept of qualified agent is invoked. So we have: (Q) A person’s being virtuous consists in her virtuousness evoking suitable responses in competent, qualified responders (for example, sentiments of approbation of those with an authoritative moral sense). Call (Q) the Qualified Agent Thesis. The Relational and Qualified Agent theses have been subjected to criticism. However, invoking the thought of Hume and Heidegger, I argue that Thesis (I) does not entail either (R) or (Q). The first problem concerns the Relational Thesis. Wiggins notes after all that ‘redness is an external, monadic property of a [British] postbox’.6 The second arises when we appreciate that understanding normative properties as response dependent is not sufficient for understanding them as normative. We want to say, for example, that a trait V is a virtue if and only if V merits the relevant response (and does not, for example, just cause it). The Qualified Agent Thesis is then invoked to give an account of what counts as a merited response. However, when we think of a property as meriting a certain type of response, it seems natural to think of it as a monadic property that makes certain responses justified. What makes it the case that a trait merits status as a virtue does not seem to depend on the responses of an agent, even a qualified one, but on certain facts, such as the trait actually systematically producing good consequences. Secondly, competent agents may be wrong about those key facts. Competence is not the same as omniscience. Competent, indeed virtuous, agents have access to at least some of the more important and relevant background theories of human nature and have wise, relatively informed views about long-term consequences, but some of these theories and beliefs may be false or incomplete. These objections to (Q) on my view are decisive. The solution is not however to reject response dependence but to reject (R) and (Q) as an analysis of the response dependent nature of ethical properties. ‘Virtue’ on the view outlined below is a response dependent concept in sense (I), but (I) is deemed to entail neither (R) nor (Q), which are false. These claims rely on three central


Christine Swanton

theses of Heidegger that I believe are shared by both Hume and McDowell. These theses are as follows: 1. We have ‘unmediated openness’7 to the world. Such openness (what Heidegger labels aletheia) reveals the world as a whole as it is (in one or other of its broad aspects). 2. This unmediated openness is constituted by what Heidegger calls ‘frameworks’ or ‘horizons of disclosure’, which allow for various ways in which the world is open to us. This is not to deny that mistakes in representing the world from within such a framework cannot be made. 3. Such frameworks are multiple. Like the scientific framework, the ethical framework can constitute ethical properties as monadic. This last claim is most important. Assume that in order to understand a monadic property of an agent (e.g., a virtuous trait of character) as a property that merits a certain evaluative response, we need to understand it as a property whose entire nature is accessible to scientific modes of ‘access to the real’. On this assumption this mode of access is both necessary and sufficient for such a response to be assessable as objectively correct. Similarly, we might say that dangerousness genuinely merits fear because to be dangerous is to be prone to cause harm, a proneness whose entire nature is accessible to scientific understanding. However, on scientific construals of normative demands, the power of a virtuous trait of character to exact demands such as cultivating it, training our children into it, acting according to its requirements, will seem mysterious. The solution to both of these problems is to understand (I) as construing moral properties as monadic, naturalistic, but as not wholly disclosed by the framework of science. In this chapter I shed light on (I) as a ‘new metaphysics’ for virtue ethics by first outlining Hume’s moral metaphysics, in particular his insight that the reality of ethics is not constituted through scientific modes of understanding alone.8 However, to see more exactly how the reality of ethics is constituted for Hume, I interpret him through Heidegger’s understanding of ‘openness’ to the world. On this deeper reading of Hume’s response dependence, he can be seen as subscribing to (I), avoiding subjectivism, and the problems of (R) and (Q).

HUME AND THE LIMITATIONS OF ‘REASON’ In recent Hume scholarship Hume’s moral theory has been rehabilitated through understanding his moral sentimentalism as a form of moral sense theory, in modern parlance a form of response dependence.9 On the view proposed, an understanding of Hume as moral sense theorist should be mediated through the complexities of Heideggerian phenomenology, not suggested by an overly literal understanding of the moral sense as something

A New Metaphysics for Virtue Ethics


closely analogous to sense impressions, and as something quite mysterious in the moral domain. For Hume the fundamental problem of philosophy is the tendency of reason (in a particular sense) not to know its limits. ‘Reason’ is here used in its ‘narrow’ sense and includes (1) demonstration (which itself includes (a) intuition not relying on sense impressions and (b) deduction) and (2) causal reasoning. Through demonstration we know of relations between ideas, and through causal reasoning we acquire beliefs about ‘matters of fact’. This is the reason proper to the ‘understanding’. The unbridled use of reason in this sense results in skepticism, as we find that we cannot justify all that we want to justify by its use alone. However, Hume himself, with some qualifications,10 should not be regarded as a skeptic so much as a resolute critic of the pretensions of reason in this narrow sense, a pretentiousness that leads to skepticism. In fact I would say that Hume shares with Heidegger an even more radical thesis. For Heidegger reason in the narrow sense should not even be seen as the primary mode of ‘disclosure’ of reality. That disclosure should be seen as fundamentally practical: one of absorption and engagement in a public social context. It is not primarily one of an isolated detached spectator, which in its extreme form is pathological rather than a recipe for maximal objectivity.11 Heidegger thus rejects a ‘spectator’ metaphysics dominated by perceptual analogies and paradigms12 in favor of what might be called a ‘metaphysics’ of engagement. To show this, he illustrates in Being and Time one important mode of engagement, that is engagement in the world of equipmentality—the world of competence with tools and other equipment. Let us illustrate the general problem of the pretensions of reason (in the narrow sense) in ethics with a well-known passage from Hume’s Treatise: ‘Take any action allow’d to be vicious: Wilful murder, for instance. Examine it in all lights, and see if you can find that matter of fact, or real existence, which you call vice. In which-ever way you take it, you find only certain passions, motives, volitions and thoughts. There is no matter of fact in the case. The vice entirely escapes you, as long as you consider the object. You never can find it, till you turn your reflexion into your own breast, and find a sentiment of disapprobation, which arises in you, towards this action’ (T 468–469). What Hume is claiming here is that if we do not appreciate that reason in the narrow sense has its limits, we will never find virtue or vice in the object, for vice does not consist merely in a ‘matter of fact’ in this sense, that is, something knowable merely through causal reasoning. Moral skepticism of some form is the only option if we avail ourselves of one or other of the following metaphysical postulates: a. Rationalistic metaphysics, which overplays the role of demonstration in the process of justification in such fields as ethics. As Hume claims: ‘There has been an opinion very industrially propagated by philosophers, that morality is susceptible of demonstration; and tho’ no one


Christine Swanton

has ever been able to advance a single step in those demonstrations; yet ’tis taken for granted, that this science may be brought to an equal certainty with geometry or algebra’ (T 463). b. What John McDowell calls empiricistic naturalism,13 which overplays the role of causal scientific reasoning in the process of justification in such fields as ethics, and makes such reason primary and the standard for all genuine rationality. If we cannot find vice ‘in the object’ understood in either of the two ways described in (a) and (b), where else should we look? Hume answers that not ‘till you turn your reflexion into your own breast’ will you find what you are looking for (T 468). Alas, it is easy to miss the point. It is tempting to read Hume as saying this: looking at the blood, etc., causes distress, disgust, horror, and so forth, so moral judgment is nothing but the following matter of fact: I am distressed at the willful killing; or if not a fact of this kind, then an expression of our emotion or a projection of them. But this kind of analysis misses the point because it is still trying to use reason in the narrow sense (in the form of causal reasoning) coming up with forms of subjectivism or skepticism about the genuine objectivity of ethics. In short at play here is ‘empiricistic naturalism’, one of the variants of the pretensions of reason in the narrow sense. Such an analysis has failed to grasp the central point of Hume’s philosophy: reason in the narrow sense should know its limits. What then should replace this notion of reason in the ethical domain? What does Hume mean by the need to turn your reflection into your own breast? In very general terms we need to turn our reflection to our passions in order to understand the nature of moral properties as response dependent.14 This is indeed how I have interpreted him,15 with the following major proviso. On my view he should be interpreted as a response dependence virtue ethicist: a form of virtue ethics not so far salient in the virtue ethical literature. This is important to secure the normativity of Hume’s response dependence as a form of substantive moral philosophy in a suitably objectivist tradition. However, in this chapter I wish to tie his response dependence not to his normative theory understood as a type of virtue ethics, but to a Heideggerian understanding of the very nature of Hume’s response dependence at a metaphysical level. Most importantly, we need to see how Hume’s response dependence, so interpreted, can be seen as realist (in Pigden’s sense)16 without being a species of ‘empiricistic naturalism’, where rationality in ethics has to be understood through the operations of reason in the narrow sense. Understanding the limits of reason in Hume’s narrow sense is the essential first step to understanding (I). To make ethics both intelligible and truth apt we need to understand the distinctive mode of intelligibility of ethics. Heidegger’s metaphysics of engagement, elaborated in his hermeneutic phenomenology, gives us explicit resources for securing that understanding.

A New Metaphysics for Virtue Ethics


ALETHEIA: UNCONCEALMENT It is well known that at the end of Book I of the Treatise Hume apparently lapses into skeptical despair being ready to ‘reject all belief and reasoning’ (T 268) as capable of furnishing answers to questions ranging from ‘from what causes do I derive my existence?’ to ‘whose anger must I dread?’ (T 269). He concludes that ‘since reason is incapable of dispelling these clouds, nature herself suffices to that purpose, and cures me of this philosophical melancholy and delirium’ (T 269). Hume was onto something here: not only is reason (in the narrow sense) entirely insufficient for practical competent engagement in the real world, but our very understanding of what it is to inhabit such a world must be furnished by nature much more broadly understood. Indeed Hume himself claims that ‘the understanding [reason in the narrow sense], when it acts alone, and according to its most general principles, entirely subverts itself’ (T 267). But it is Heidegger who provides at the very heart of his philosophy the metaphysical justification for such a view. Hume simply proceeds to act on the cure and writes Book II and III as if no skeptical despair had ever occurred. Hume’s problem is that he implicitly (at the end of Book I) privileges scientific modes of understanding understood through reason in the narrow sense so that other forms of practical engagement with the world do not have as much claim as the scientific to make reality intelligible. Heidegger resolves that problem by not privileging the scientific mode. On this view as McDowell suggests there are multiple forms of basic construals of the world as a whole, yielding multiple types of standard for ‘objective correctness’. Our task in this section is to offer a Heideggerian understanding of the metaphysics of (I). How can we understand Hume’s point that scientific understandings alone cannot bring the world of ethics into being, and yet claim that that world is both a world of nature and has objective normative reality? In Heideggerian terms, to determine what he calls the ‘worldhood’ or ‘world’ of ethics we must understand that particular mode of comportment to the world that brings its ethical aspect into being. This idea is at the core of Heidegger’s notion of aletheia, which we might call (following normal translations of Heidegger) openness, uncoveredness, unhiddenness, or unconcealment. Fundamental orientations or attunements are ways of what Heidegger calls unconcealing: opening or revealing the world as a whole in one or other of its aspects. These constitute backgrounds of understanding (comportments), including emotional understanding: forms of ‘disclosure’. Such a revealing is what Heidegger calls a ‘clearing’ (Lichtung), which is a way of securing intentional access to the world as a place of a certain type. Wrathall puts the point this way: ‘Unconcealment, when understood as a clearing, does not name a thing, or a property or characteristic of things, or a kind of action we perform on things, or even the being of things. It names, instead, a domain or structure that allows there to be things with properties or characteristics, or modes of being . . . It is something like the space of possibilities’.17


Christine Swanton

Such a domain or structure Heidegger calls a ‘worldhood’, even ‘world’ (Welt). Truth as correspondence, a property of judgments or assertions, being a relation of ‘agreement of the judgment with its object’, presupposes that there has been a ‘Lichtung’ within which we have a mode of intelligibility.18 This conception of truth Heidegger also calls truth as correctness. Crucially, truth as correspondence presupposes aletheia: ‘The correctness of seeing and viewing things, and thus of definition and assertion [truth as correspondence] is grounded in the particular manner of orientation and proximity to beings, i.e. in the way in which beings are in each case unhidden. Truth as correctness is grounded in truth as unhiddenness’.19 We are now in a position to see what is wrong with orthodox interpretations of the passage from Hume concerning ‘seeing the vice in the object’. Hume’s claim that it is not until you turn your reflection into your own breast that you will see the vice in the object is a claim concerning (I). Emotional construal is essential for the world of ethics to be intelligible. There is no ethical object that can be the object of our representation and epistemology if we only construe the object through the disclosures appropriate to forensics (the knife in the back, the heart stopped, the blood on the floor). There is no ‘space of possibilities’ for ethics at all. The Qualified Agent Thesis, in any version, cannot even apply till the ethical object is brought into being. If it is not, we are left only with our feelings toward the object forensically conceived; we are on the road to ethical subjectivism. It turns out that the process of adequate uncovering has many aspects. This complexity is explicit in Heidegger’s depiction of aletheia as a multiple unity having four aspects called by Heidegger ‘The space of the four-fold unitary openness’.20 In the remainder of this chapter I specify the four aspects or modes of the four-fold of openness and apply each of them to Hume. In so doing, I show the manner in which Hume’s moral sentimentalism can be understood as not in the least subjectivist, private, immune to critical analysis, and anti-realist. With the help of Heidegger’s analysis, I show how Hume’s moral sentimentalism provides a very rich account of (I). The four-fold of aletheia ‘uncovers’ ethics: the ‘mode’ or ‘character of being’, which is the ‘worldhood’ of ethics.

THE ‘THING’ MUST BE ‘OUT IN THE OPEN’ The first mode of the four-fold of openness is described by Heidegger as the ‘thing being out in the open’. This mode itself has two aspects: 1. The ‘thing’ must have a power to affect sensibilities. 2. The thing is open in virtue of the nature of the sensibilities constituting its nature as a kind of thing. For example, we might say, more or less controversially, if there were no sensibilities capable of experiencing color, there would be no color properties; if

A New Metaphysics for Virtue Ethics


there was no benevolence, love, or empathy, there would be no moral sense and thus no ethical properties; if there were no sensations and reason proper to the understanding (in Hume’s sense), there would be no scientific properties; if there were no aesthetic sensibilities, there would be no beauty. Without sensibilities and ‘Verstehen’ (understanding in a broad sense) responding to powers and creating frameworks of disclosure of aspects of the world, and thereby securing an intentional relation to them, there would only be, in Heidegger’s language, the mystery of Being. This is a reality conceived of prior to or independently of its powers to affect sensibilities and thereby allow for the possibility of being disclosed. Concealment, hiddenness, of the world of ethics is not to be confused with falsity or illusion, which come on board only with truth as correspondence. Rather, reference to the worldhood of ethics has not even been secured. Hume presupposes this distinction in his discussion of such beings as the ‘fancied monster’ of the Enquiries and the ‘perverse’ individual of the Skeptic: ‘where one is born of so perverse a frame of mind, of so callous and insensible a disposition, as to have no relish for virtue and humanity, no sympathy with his fellow-creatures, no desire of esteem or applause; such a one must be allowed entirely incurable, nor is there any remedy in philosophy’.21 Were the world to be entirely filled with such persons we would have a world of moral blindness analogous to a world of color blindness where no one had color vision. What is it not to be blind to the ethical realm? For Hume, the disclosure of the world of ethics requires a moral sense, a sentiment of approval; more particularly, a passion of ‘fainter and more imperceptible love’ (T 614; T, constituting our emotional sense of virtue properties. The conditions of possibility of such a sense (if it is to be understood as a moral sense disclosing that aspect of reality) are certain emotional responses and psychological capacities. The relevant responses and capacities are benevolence, and benevolent empathy and sympathy constrained by self-love. Without benevolence, we could not have an orientation to the world fundamental to ethics on Hume’s view: desire for another’s good. Without sympathy, we could not transmit this desire sufficiently widely: humanity toward strangers and those distant from us is also an orientation fundamental to ethics. Furthermore, if there were in addition no passion of self-love or ‘love of life’ (that is, one’s own life) we could not make sense of our approval of the self-regarding virtues (virtues that render people ‘serviceable to themselves, and enable them to promote their own interests’ (T 587)), and the world of ethics disclosed by the moral sense would then take on an entirely different cast. We would see ourselves wholly as instruments in the service of others’ needs. This would be a world without proper agency and not a world of morality as we would understand it for Hume. This is not to say that these are the only emotions relevant to ethics, but they constitute the background (horizon of disclosure) within which ethics makes sense. It is Hume’s fundamental contention that the sensibilities requisite for understanding the scientific nature of the world, namely sensations and the


Christine Swanton

reason proper to the understanding, are not sufficient to allow ethical properties to ‘be out in the open’ or unconcealed. What such sensibilities and reason aim to do is discover causal relations and ‘eternal’, ‘immutable’ fitnesses and unfitnesses of things that would impose obligations that would be ‘the same to every rational being that considers them’ (T 456), regardless of the kinds of sentiments (if any) that form their constitution. However, for Hume, there are no such eternal fitnesses in the moral, prudential, or aesthetic domains. This is the point of Hume’s claims that ‘ ’Tis not contrary to reason to prefer the destruction of the whole world to the scratching of my finger’ (T 416) and ‘ ’Tis not contrary to reason for me to chuse my total ruin, to prevent the least uneasiness of an Indian or person totally unknown to me’ (T 416). Here Hume means by ‘reason’ pure emotionless reason of the kind possessed by all rational beings regardless of their emotional constitutions. Human beings do possess such reason, provided we understand it as the operations of the understanding, a faculty ‘considered apart from any passions and any feelings of pleasure and pain’. However, reason in this narrow sense is simply not able to deconceal what Hume calls ‘natural fitnessess’ in the practical domains. Indeed the previous claims about what is not ‘contrary to reason’ are a reductio of the idea that reason in the narrow sense is fit to disclose such natural fitnesses.22

HUMANS MUST BE OPEN TO THE THING The possession of sensibilities necessary for us to be receptive to powers of things to be disclosed is not sufficient for them to be disclosed richly or fully. One might think that this aspect of openness is indistinguishable from epistemology understood as adequacy of representation within a world(hood), but Heidegger is thinking here of the problem of pervasive and deep-seated distortions that have large-scale consequences, such as seeing slaves as not fully human, patriarchal distorted views about woman as such, seeing animals as not sentient, and more latterly the environment as mere resource to serve our needs (a conception that for the later Heidegger precludes the fundamental mode of disclosure of ‘dwelling’, which requires taking the ‘poetic measure’).23 A desire for the good of another and some empathetic and sympathetic capacity allow us to conceive of the world as having ethical aspects, but a number of other properties are needed to avoid pervasive serious distortion, as well as providing an adequate epistemology. In ethics this aspect of openness requires a moral sensitivity, such as Iris Murdoch’s loving attention, Aristotle’s practical wisdom, and moral imagination. It may be wondered that if ethics is disclosed by an emotional moral sense, how can such a sense yield criteria of virtue that provide standards for whether or not a trait is a genuine virtue? The causes of the activation of the moral sense determine the range of properties that merit status as virtues.

A New Metaphysics for Virtue Ethics


For Hume something inaccessible to such a sense (such as God’s commands) cannot be such a property. For Hume there are two broad kinds of cause: pleasure or pain from reflections on traits’ characteristic consequences, and pleasure or pain from the immediate impact of their ‘species or appearance’ (T 589–590). These causes suggest then two general criteria of virtue relative to which the merits of traits can be assessed: (C1) A trait is a virtue if it tends to the happiness of mankind. (C2) A trait is a virtue if it has properties, not reducible to consequences for happiness, that make it ‘naturally fitting’ that its species or appearance causes ‘this immediate taste or sentiment’. (C2) is subdivisible into several criteria on the assumption that several types of feature, not reducible to consequences for the happiness of mankind, make it fitting that ‘immediate taste or sentiment’ be produced.24 Both these criteria can be applied in a warranted way by those who are ‘open to the thing’: that is, for Hume, those with doxastic virtue and an authoritative moral sense. THE THING MUST BE OPEN IN A ‘REGION BETWEEN THING AND MAN’ We have seen that, for both Hume and Heidegger, different sensibilities open up or disclose different aspects of the world: it must not be assumed that the reason of the understanding (in Hume’s sense) is competent to disclose all aspects of reality. This raises the question: what is the relation between the ‘region’ of ethics and others? Hume’s broad conception of ethics has the consequence that the worlds of natural fitnesses in the practical domains (aesthetic, ethical, prudential) should not be seen as sharply disjointed from each other. The moral and the prudential come together with the virtues useful or agreeable to self, virtues that inevitably have effects also on others. This is unsurprising since a condition of a moral sense proper to humans is benevolence constrained by selflove. Included amongst such virtues are ‘prudence, temperance, frugality, assiduity, enterprise, dexterity’ (T 587). Most importantly what is the relation between the region of ethics and that of science? I claimed previously that for Hume, emotional disclosure is necessary but not sufficient for the disclosure (unconcealment) of ethics. It is not sufficient, for as Hume makes clear, the cooperation of reason proper to the understanding is necessary. Such cooperation is necessary since the criteria of virtue cannot be ‘judiciously’ applied without it. Science provides background theories that help explain important moral practices and enrich our understanding of a good life for a human being.25 So, though moral truth depends on the possibility of a moral sense, and thereby on its conditions of possibility (notably benevolence and benevolent empathy),


Christine Swanton

its operations do not guarantee the discernment of moral truth. Even the warranted judgments of those with an authoritative moral sense may thus fail to be true, for status as a virtue or a vice depends on ‘matters of fact’, potentially discoverable by the understanding, which may not be picked up, even by wise and sensitive judges. Hence, Hume’s response dependent view of (virtue) ethics should be understood as a version of (I) (see section ‘Virtue Ethics and Response Dependence’), but not (Q).

HUMANS MUST BE ‘OPEN TO THEIR FELLOWS’ This fourth aspect of truth as aletheia brings into focus an essential aspect of Heidegger’s metaphysics—a rejection of a private, spectator metaphysics in favor of one constituted by an essentially social engagement. It may seem that Hume belongs to the former camp, but I hope to show that this is debatable, certainly where his ethics is concerned. We have seen that the fundamental idea of truth as aletheia is the intelligibility of a thing that, for Heidegger, presupposes engagement with and competence in a social world. The emotional background (attunement) essential for the disclosure of ethics, that is practical competence in ethics (as Hume also saw), thus has an essentially social dimension. Competent emotional engagement presupposes that we understand the significance of emotions within cultural and historical contexts: emotions are not to be seen as private or inner. What counts as naturally good and what counts as a defect is often a matter of interpretation.26 In a world of social meanings, the claim that something is a defect might be taken to be an insult. Hume recognized this feature of emotion in two ways: a. Hume understood the public nature of emotion at a very basic level; indeed he can be seen as a philosophical precursor of modern mirroring theory. He claims in the Treatise that emotion is transmitted to another as a ‘contagion’ or ‘infusion’ of sentiments. For example: ‘A chearful countenance infuses a sensible complacency and serenity into my mind; as an angry or sorrowful one throws a sudden damp upon me’ (T 317). ‘Empathy and sympathy are thus capacities enabling us to ‘receive by communication [others’] inclinations and sentiments’ (T 316). Now we may think that Hume’s talk of ‘infusion’ is committed to an asocial view of emotions of the sort criticized by Heidegger. For the latter, although ‘[emotional] attunements are feelings’, such an attunement is not a private emotional experience that is then transmitted to others, in the manner of ‘infectious germs’.27 An attunement, including an emotional one, is a way of being in the world and being with others. We say, for example, that persons of ‘good humour bring a lively atmosphere with them’.28 But this is exactly Hume’s view. The melancholy person of Hume’s description in a similar way brings a melancholy atmosphere with him: he ‘throws a damp’

A New Metaphysics for Virtue Ethics


on a person, or on the conversation in general. Compare Heidegger: ‘Or another person is in a group that in its manner of being dampens and depresses everything; no one is outgoing . . . Moods are not accompanying phenomena; rather, they are the sort of thing that determines beingwith-one-another in advance’.29 The four-fold of aletheia, which includes an irreducible social dimension, thus applies to emotional disclosure as to any other form of disclosure. b. Hume understood that emotional responses are inculcated within educational and social practices. This is clear in his discussion of justice. Although the natural materials for emotional response must be within our ‘frame and constitution’, it is education and upbringing that turns these into emotionally laden norms. Disgust, distaste, and hate are transmuted through education and custom into sentiments of seeing as disgraceful and ignoble; admiration into sentiments of seeing as noble and admirable, and so on.

CONCLUSION A Humean/Heideggerian response dependence theory of ethical properties, presented in this chapter as a new metaphysics for virtue ethics, has the following form. There is unmediated openness (constituted by the fourfold of aletheia) to monadic properties of virtue (being virtuous), through a form of disclosure described by Hume as a ‘moral sense’. Unlike Pacific salmon, we do not have unmediated openness to the world of directionality, which they possess in the form of detecting the earth’s magnetic lines by means of atoms of iron in their sensory cells. But we (or rather most of us) do have openness to the world of ethics where we construe qualities as virtues, emotions as fitting to their objects, actions as admirable permitted or required because in some way virtuous, or lacking in vice. A scientific understanding of reason and virtue is necessary but not sufficient for the disclosure of ethics in general, and the intelligibility of virtue and vice notions in particular. The complex nature of the disclosure of ethics enables us to make sense of Hume’s apparently conflicting accounts of the essential nature of virtue. I conclude by presenting these, indicating how they relate to the various aspects of the four-fold. (a) A virtue is a ‘power’ to affect human sensibility. This characteristic of virtue relates to the first feature of the first aspect of the four-fold of aletheia—the thing being ‘out in the open’. This feature is: things must have powers to affect sensibilities if they are to be disclosed. (b) The essence of virtue is to produce pleasure.


Christine Swanton

This characteristic of virtue relates to the second feature of the first aspect of the four-fold, namely the thing is open in virtue of the nature of the sensibilities constituting its nature as a kind of thing. This definition of virtue refers to the fact that the sensibility that opens up the world of ethics and virtue in particular is a form of pleasure, namely ‘fainter and more imperceptible love’ (and not, for example, sensation, or the reason of the understanding). (c) The moral sense most relevant to the disclosure of virtue is the authoritative moral sense of qualified persons. This characteristic of virtue is associated with the second aspect of the fourfold, being ‘open to the thing’. Not all moral senses are equally ‘open to the thing’. Notice, however, that (c) does not imply (Q) (section ‘Virtue Ethics and Response Dependence’). (d) Virtues are traits that tend to the good of mankind or are naturally fitted to produce immediate pleasure from ‘species or appearance’. This characteristic of virtue (the criteria of virtue) determines what traits merit the ascription of status as a virtue. These criteria are associated with the third aspect of the four-fold delineating the ‘region’ disclosed by the moral sense. It is a consequence of this double criterion of virtue that the ‘region’ of ethics is very broad for Hume. (e) Virtues are traits that are approved from the common point of view in social contexts. This feature of virtue relates to the fourth aspect of the four-fold, the essentially public nature of aletheia. Applying the account of the four-fold of aletheia to Hume shows that for him vice is not just a property like knives sticking out of backs, or a feeling of disgust in the contemplator of it; nor is it a queer property. Virtues and vices are powers but not queer powers. For talk of powers is simply a way of indicating the reality of virtue and vice: there is no suggestion of representational adequacy secured through scientific modes of understanding. An agent cannot just see such a power, or simply reason causally about it. Talk of virtue and vice is intelligible only in a background context of emotional comportment in a world of engaged agency, conducted by beings of a fundamentally benevolent empathetic disposition, steeped in contexts of education, politics and policy, friendship, family, justice, respect within (legitimate) social hierarchies, and so on. On these practices Hume has quite a lot to say, and much work by Hume scholars such as Annette Baier and Jackie Taylor has shown the practical, engaged, socially embedded nature of Hume’s moral philosophy.

A New Metaphysics for Virtue Ethics


NOTES I am grateful to the organizers of the International Hume Society meeting Antwerp 2012, where I presented a version of this paper, for inviting me to speak, and to the attendees for discussion and encouragement 1. Hume, Enquiries, 174–175. 2. McDowell, ‘Two Sorts of Naturalism’, 149–179, 164. 3. For an account of emotions as emotional construals, and one which I think fits Hume’s account of ethics in general and individual virtues in particular, see Roberts, Emotions: An Essay in Moral Psychology. For Hume, emotional construals of situations and objects are made possible through passions including desires. 4. See Dreyfus, Being-in-the-World: A Commentary on Heidegger’s Being and Time, Division I, 2. 5. McDowell, ‘Values and Secondary Qualities’, 112. At this point McDowell is usually (mis)read as confusing response dependence as a thesis about properties (redness) with our concept of redness (see further on in this section). 6. Wiggins, ‘A Sensible Subjectivism’, 107. 7. This useful term is used by Maximilian de Gaynesford to describe McDowell’s thought (see his de Gaynesford’s John McDowell). He claims ‘There is an unmediated openness between the experiencing subject and external reality: if our experience is not misleading, we are directly confronted by worldly states of affairs’ (34). Further, like Heidegger’s concept of aletheia (see the section ‘Aletheia: Unconcealment’) this is a thesis concerning intentionality (how our thoughts and so on are directed onto the world) rather than epistemology. However, in this chapter I do not explore McDowell’s own version of the response dependent nature of ethical properties. 8. Even where ‘science’ is broadly conceived: see Dupre, The Disorder of Things. 9. See Charles Pigden, ‘If Not Non-Cognitivism, Then What?’, 80–104, 95–96. 10. Hume was skeptical about our ability to have knowledge in certain domains (including ethics) in his restricted sense of knowledge as knowledge by intuition and demonstration of propositions that involve only relations between ideas and that cannot be false (A Treatise of Human Nature, 70. References to this work will henceforth be abbreviated to T in the text, followed by a page number. 11. See Sartre’s description of Roquentin’s pathological disengagement with a doorknob in his Nausea. (Discussed in Dreyfus, Being-in-the-World, 47.) 12. See further Rorty, Philosophy and the Mirror of Nature, 159. 13. In his ‘Two Sorts of Naturalism’. 14. In the words of Charles Pigden (in ‘If Not Non-Cognitivism, Then What?’, 95–96) ‘Hume was regarded both in his own day and for the next 200 years (roughly 1740–1940) as a dissident disciple of Frances Hutcheson and Hutcheson was a moral realist’ in the following sense: ‘moral judgments are “truth-apt,” true or false, and . . . some such judgments are (literally) true; and true . . .with respect to their distinctively moral contents’. In particular ‘in today’s jargon’ Hume should be read as following Hutcheson in being a response dependence theorist. By contrast Rachel Cohon describes the ‘reaction dependence’, ‘moral sensing’ interpretation of Hume as anti-realist in Hume’s Morality: Feeling and Fabrication. I think this is unfortunate since it would be odd to think of accounts of colors as secondary properties as “antirealist”. However, and, in my view, unfortunately, moral realism is often analogized to realism in science, itself conceived as a realism where scientific propositions are conceived as describing a reality ‘whose nature owes nothing to our natures’ and is ‘prior to and independent of’ it. (Railton, ‘Subjective


15. 16. 17. 18. 19.


21. 22. 23. 24. 25.


27. 28. 29.

Christine Swanton and Objective’.) Even if global response dependence were false, this kind of generic description and analogizing is rightly in my view criticized by Railton. See my ‘Can Hume be Read as Virtue Ethicist?’. See note 14. Mark Wrathall ‘Unconcealment’ in Heidegger and Unconcealment , 11–39, 14. Heidegger, Being and Time, sect. 44, 257. Heidegger The Essence of Truth: On Plato’s Cave Allegory and Theaetetus, 26. In earlier works, as we saw in the previous quote, Heidegger regarded aletheia as a form of truth. However this usage was criticized by Ernst Tugendhat (see Tugendhat, ‘Heidegger’s Idea of Truth’, 83–97), and Heidegger eventually ceased to call aletheia ‘truth’. The main issue as Malpas sees it is whether or not ‘truth’ is legitimately applicable to what makes truth (as correctness) possible, what Heidegger would regard as its essence (Malpas, ‘The Twofold Character of Truth: Heidegger, Davidson, Tugendhat’). He claims that ‘Heidegger’s later acceptance of Tugendhat’s claim that aletheia is not the same as truth has to be viewed as problematic, since it threatens to obscure the very twofold unity that is so important . . . there are, in an important sense, not two separated concepts here, but two aspects of a single structure’. Along the same lines Wrathall claims that ‘we could say that the being of truth lies in uncovering’ (Wrathall, ‘Introduction’, Heidegger and Unconcealment: Truth Language and History 1–8, 4). Wrathall, in a critique of Tugendhat, sides with Heidegger’s original nomenclature, and my sympathies lie with Wrathall and Malpas. For the purposes of this chapter we shall leave aside this debate, understanding aletheia as uncovering or deconcealing simply. Heidegger, Basic Questions of Philosophy, 18–19. Although aspects of ‘the four-fold unitary openness’ are separately described for the purposes of understanding them, they are not separate in re. Heidegger makes it clear that they are a unity. All are required together as it were for a successful intentional relation to the world to be secured: ‘this four-fold openness would not be what it is and what it has to be if each of these opennesses were separately encapsulated from the others. This four-fold openness holds sway rather as one and unitary’ (Heidegger, Basic Questions, 19.) Hume, ‘The Standard of Taste’, 222. See Hume, ‘The Standard of Taste’, 275; T 591. Discussion of the nature of dwelling elucidated largely in Heidegger’s essays collected in Poetry Language and Thought is beyond the scope of this chapter. These criteria of virtue and their pluralistic nature are discussed much more fully in my ‘What Kind of Virtue Theorist is Hume?’ It could be true both that what Don Garrett in his Critic’s comment on Rachel Cohon’s Hume’s Morality: Feeling and Fabrication calls ‘the categorical basis for the impression’ of vice is independent of our moral sensibilities and that the impression of vice as a moral property is not so independent. (See Rachel Cohon ‘Reply to Radcliffe and Garrett’, 277–288, 281). This point is the gist of Bernard Williams’s ‘representation problem’ in his ‘Evolution, Ethics, and the Representation Problem’, 100–110, where he points out that beings with a culture and language represent things in various possible ways, and ‘where there is culture, it affects everything’ (102). See also Hacker-Wright ‘What is Natural about Foot’s Ethical Naturalism?’. Heidegger, The Fundamental Concepts of Metaphysics: World, Finitude, Solitude, 67. Heidegger, Fundamental Concepts of Metaphysics, 66. Op. cit.

A New Metaphysics for Virtue Ethics 193 REFERENCES Baier, Annette C., Death and Character: Further Reflections on Hume (Cambridge, MA: Harvard University Press, 2008). ———, Progress of Sentiments (Cambridge, MA: Harvard University Press, 1991). Cohon, Rachel, Hume’s Morality: Feeling and Fabrication (Oxford: Oxford University Press, 2008). ———, ‘Reply to Radcliffe and Garrett’, Hume Studies 34 (2008), 277–288. de Gaynesford, Maximilian, John McDowell (Cambridge: Polity Press, 2004). Dreyfus, Hubert L., Being-in-the-World: A Commentary on Heidegger’s Being and Time, Division I (Cambridge, MA: The MIT Press, 1991). Dupre, John, The Disorder of Things (Cambridge, MA: Harvard University Press, 1993). Hacker-Wright, John, ‘What is Natural about Foot’s Ethical Naturalism?’, Ratio 22 (2009), 308–321. Heidegger, Martin, Basic Questions of Philosophy, trans. Richard Rojcewicz and Andre Schuwer (Bloomington: Indiana University Press, 1994). ———, Being and Time, trans. John Macquarrie and Edward Robinson (New York: Harper, 1962). ———, The Essence of Truth: On Plato’s Cave Allegory and Theaetetus, trans. Ted Stadler (New York: Continuum, 2002). ———, The Fundamental Concepts of Metaphysics: World, Finitude, Solitude, trans. William McNeill and Nicholas Walker (Bloomington: Indiana University Press, 1995). Hume, David, Enquiries Concerning the Principles of Morals, 3rd ed., ed. P. H. Nidditch. (Oxford: Clarendon Press, 1995). ———, ‘The Standard of Taste’, in Essays ,Moral, Political, and Literary, vol. 1, ed. T. H. Green and T. H. Grose (London: Longmans, Green, and Co., 1912). ———, A Treatise of Human Nature, ed. L. A. Selby-Bigge (Oxford: Clarendon Press, 1978). Malpas, Jeff, ‘The Twofold Character of Truth: Heidegger, Davidson, Tugendhat’, Divinatio (forthcoming). McDowell, John, ‘Two Sorts of Naturalism’, in Virtues and Reasons, ed. Rosalind Hursthouse, Gavin Lawrence and Warren Quinn (Oxford: Clarendon Press, 1995), 149–179. ———, ‘Values and Secondary Qualities’ in Morality and Objectivity, ed. T. Honderich (London: Routledge and Kegan Paul, 1985). Pigden, Charles R., ‘If Not Non-Cognitivism, Then What?’ in Hume on Motivation and Virtue, ed. Charles R. Pigden (Basingstoke: Palgrave McMillan, 2010), 80–104. Railton, Peter, ‘Subjective and Objective’, in Truth in Ethics, ed. Brad Hooker (Oxford: Blackwell, 1996), 51–68. Roberts, Robert C., Emotions: An Essay in Moral Psychology (Cambridge: Cambridge University Press, 2003). Rorty, Richard, Philosophy and the Mirror of Nature (Princeton: Princeton University Press, 1979). Swanton, Christine, ‘Can Hume be Read as Virtue Ethicist?’, Hume Studies 33 (2007), 91–113. ———, ‘What Kind of Virtue Theorist is Hume?’, in Hume on Motivation and Virtue, ed. Charles R. Pigden (Basingstoke: Palgrave McMillan, 2010), 226–248. Taylor, J., ‘Hume on the Standard of Virtue’, Journal of Ethics 6 (2002), 43–62. ———, ‘Virtue and the Evaluation of Character’, in The Blackwell Guide to Hume’s Treatise, ed. Saul Traiger (Oxford: Blackwell, 2006), 276–295.


Christine Swanton

Tugendhat, Ernst, ‘Heidegger’s Idea of Truth’, in Hermeneutics and Truth, ed. Brice R. Wachterhauser (Evanston, IL: Northwestern University Press, 1994). Wiggins, David, ‘A Sensible Subjectivism’, in Needs, Values, and Truth, ed. David Wiggins (Oxford: Clarendon Press, 1987). Williams, Bernard, ‘Evolution, Ethics, and the Representation Problem’, in Making Sense of Humanity and Other Philosophical Papers 1982–1993, ed. Bernard Williams (Cambridge: Cambridge University Press, 1995). Wrathall, Mark A., Heidegger and Unconcealment: Truth Language and History (Cambridge: Cambridge University Press, 2011).

13 A Kantian Plea for Virtues? Erasmus Mayr

I It is often thought that a Kantian approach to ethics is hostile to, if not directly incompatible with, the notion that the virtues ought to be accorded any independent and essential role in ethics. This is a feature Kantian moral theory is thought to share with other ‘rule’ or ‘law’ centered conceptions of ethics. According to these conceptions, what is crucial to acting morally is (only) to act in accordance with a set of rules, or in pursuance of a certain good, and (for most such theories, at least) to act so for the right reasons, that is, typically, because the action in question is prescribed by those rules or advances this good. In such a framework, virtue can still play a role as a person’s disposition to reliably comply with the demands of morality; however, this is only a derivative role in relation to the moral rules themselves, which exclusively determine what is morally right or wrong. The general move of ‘downgrading’ virtues, which is often seen as a natural consequence of a rule-centered approach to ethics, is nicely exemplified by the way in which John Rawls famously drew the distinction between what he considered to be the two main types of ethical theories:1 Claiming that moral theories primarily deal with the concepts of the right, of the good, and of moral worth, Rawls distinguished between theories that define the right in terms of the good—like utilitarianism—and theories that proceed in the opposite direction—like Kantian moral theory. On both these approaches, moral worth—the key concept for a theory of virtue—is defined in terms of one of the other two concepts and, therefore, is not itself a fundamental concept. Thus, Rawls himself considered virtues to be just ‘strong and normally effective desires to act on the basic principles of right’.2 Substituting ‘reliable disposition’ or ‘strength in the face of adverse desires’ for ‘desire’, this is pretty close to how Kant himself characterizes virtue in the Metaphysics of Morals (MM): ‘Virtue is the strength of a human being’s maxims in fulfilling his duty’.3 In the following, I am going to argue that, despite these appearances, virtues should be accorded an independent role on a Kantian view, even when it comes to determining what is morally right and wrong. For appealing to


Erasmus Mayr

the notion of a virtuous agent, where this notion cannot simply be defined in terms of acting in accordance with moral principles, allows us to fill a crucial lacuna in Kant’s account of moral decision-making: Namely how abstract principles like the Categorical Imperative (CI) can be applied to specific action-situations without themselves necessitating prior conceptualizations of these situations. Discussion of this lacuna goes back (at least) to Hegel’s ’empty formalism’ criticism of Kant’s moral philosophy, but has been especially prominent in the writings of Elizabeth Anscombe and Barbara Herman. I will begin by briefly sketching this lacuna before specifying more clearly which version of this problem will interest me here. I will go on to look at some unsuccessful attempts to provide a solution before showing how an appeal to the notion of a virtuous agent can be the key to developing an answer.

II Kant famously believed that the CI is a ‘formal’ test to be applied to maxims, or ‘general principles of action’.4 Though Kant does not explicitly state the general form of the content of maxims, it is plausible to think that their content must include descriptions of situation-types,5 action-types, and ends. The general form of maxims, on this understanding, will be something like ‘In a situation of type A, I will perform an action of type X with the aim of getting Z’.6 In all actions, Kant thinks, agents follow such maxims at least implicitly, and by subjecting these maxims to the CI test we can determine whether they act morally rightly or not. As many philosophers have realized, Kant’s conception of the CI as a ‘formal’ test faces a ianus-faced problem, which concerns partly the retrospective assessment of actions and partly the prospective determination of the morally required or permissible course of action. Applying the CI test already presupposes that there is a maxim of the agent available to be tested. So, retrospectively, in assessing an agent’s action, we have to know his maxim in order to decide whether he has acted rightly or not, and very often it will be very difficult, if not outright impossible, to establish the agent’s maxim, in particular to determine which features of the action-situation he was reacting to. This difficulty is particularly troublesome when there are different maxims we might equally well ascribe to the agent, but where those maxims fare differently on the CI test. Take the case of an officer executing barbaric orders under the Nazi regime:7 Both the maxims ‘I am going to do my duty as an official and follow the government’s orders’ and ‘I am going to kill Concentration Camp prisoners whenever I have been ordered to do so’ can equally well be ascribed to the agent, but only the latter obviously flouts the Categorical Imperative test, while the former doesn’t obviously do so. Given that the officer might well have had either maxim, how are we to

A Kantian Plea for Virtues?


decide which one to use as the basis of a moral assessment of his action? Obviously, we would want to say that in intentionally killing the prisoners the officer must have been following the second maxim, or at least a maxim whose content somehow reflected his killing the prisoners merely because he was ordered to do so, regardless of whether the order was justified. He simply could not disregard this aspect of the situation in forming his maxim. But where could this necessity, that some aspects of the situation must be relevant to the content of the agent’s maxim once the agent is aware of them, come from, within the framework of Kantian moral theory?8 Corresponding to this problem about retrospective assessment of actions, there is an analogous difficulty about prospective moral decision-making. When an agent deliberates about the morally right course of action, we would expect the CI to give him some guidance in finding an answer to this question (and Kant’s own examples in the Groundwork strongly suggest that the CI is intended to do so).9 As we have seen, applying the CI test presupposes that there is a maxim available to be tested, and in the first stage of pre-action deliberation this is not (yet) the case; the agent might form any one of a whole set of different maxims. Which of these he will choose will (typically)10 depend on his non-moral preferences. However, this cannot be the whole story, for maxims must include features of the situation to which one reacts, that is, conceptualizations of the action-situations. And some features of the situation—those that are morally ‘salient’—must be reflected in the maxim’s content. But how is the agent to know which features of a situation are morally salient ? The CI test cannot tell him what the right conceptualization of the situation is because it is only to be applied at the next stage when the maxim has already been formed—and, therefore, after the features the agent considers as salient have been picked out. As the CI appears to be the only genuine ‘morality’ test within Kantian ethics, this seems to imply that the features picked out as salient will be ‘random’ from the moral point of view. This would be innocuous if the result of the CI test did not depend on which features are specified in the maxim. But, as the case of the Nazi officer described previously shows, this is not the case; on the contrary, which features of the action-situation are picked out as ‘salient’ in the agent’s maxim often determines whether the maxim as a whole passes the Universalization Test or not. This means that in cases where different features of the situation can be treated as salient by the agent in forming his maxim, the CI does not generally give him an answer to the question how he should act and only provides very limited guidance for prospective decision-making. This is a considerable shortcoming of a Kantian moral theory: Even though it does not make the CI test completely futile, as Elizabeth Anscombe thought, who claimed that Kant’s ‘rule about universalizable maxims is useless without stipulations as to what shall count as a relevant description of an action with a view to constructing a maxim about it’,11 it undeniably leaves an important gap in the theory once it comes to applying the test to particular


Erasmus Mayr

action-situations.12 In order to fill this gap, we must, as Barbara Herman has argued, somehow enrich the CI test with ‘rules of moral salience’,13 which pick out those features of the action-situation that an agent must take into account when forming his maxims. But providing such rules isn’t as easy as it may appear, if, at the same time, we don’t want to give up on Kant’s core claim that the ultimate criterion for moral rightness must rest in the rational agent’s will. We cannot simply add rules of salience that may be arbitrary from the latter standpoint, for example, by simply adopting the list of aspects that conventional morality considers as salient. Also, some species-specific content will be needed for the rules of salience since specifically human vulnerabilities will plausibly contribute to determining what is morally salient for human agents, rather than, say, Martians. But we cannot just add further species-specific information about human agents, unless we can justify the moral relevance of the information we want to include. Even if the rules of salience cannot, generally, be derived from the CI, they must still be connected to it in the right way in order not to count as a completely independent source of morality. There are different aspects of accounting for the rules of moral salience that one might be interested in, in particular, (1) their content, that is, what the rules of salience are; (2) how they work in connection to the CI in the ‘mechanics of moral decision-making’; (3) their epistemology; and (4) the constitutive question, what makes something a rule of moral salience?, that is, what makes an aspect of a situation of the kind that it cannot be neglected when an agent forms a maxim? In the following, I will mainly be concerned with (4) since I consider it to be the most pressing question, given the worry raised in the last paragraph; but the suggested answer to (4) will also indicate ways of answering the other questions.

III There are different attempts to fill the gap in Kant’s account of moral rightness in the Groundwork, which differ with regard to the degree to which they correspond—or claim to correspond—to Kant’s own (supposed) views on how to apply the CI to particular cases. I’ll briefly look at some of these attempts before turning to the question of how an appeal to virtues could contribute to a solution. (a) First, there is what one might call the ‘Groundwork/Critique internal’ strategy, which attempts to fill the gap by only using resources from the Groundwork and the Critique of Practical Reason. The (unconscious) adoption of some such a strategy may, partly, explain the popularity of the third formula of the CI—the respect-for-humanity formula—in the debate since this formula seems to have a material content that the two other, lawformulas, lack. But, given that Kant thought the third formula to be equivalent to the other two, it would be very surprising if applying this formula

A Kantian Plea for Virtues?


should, ipso facto, yield substantially stronger results. Furthermore, while it appears plausible that we can reject some conceptualizations of actionsituations on the basis of the third formula, it seems implausible that this should work generally, for all illicit conceptualizations. While, for example, a discriminatory ‘racist’ conceptualization of an action-situation in which a person’s race is per se considered as relevant for how he is to be treated would fail to show the respect due to this person in virtue of his standing as a moral person, not all cases of ignoring morally salient features are illegitimate in precisely this way. For example, when the mistake in the conceptualization isn’t that fundamental, but ‘only’ involves neglecting some salient aspect of a person, such as her shyness, while otherwise acknowledging her moral standing. (b) Second, there is the ‘impure ethics’ strategy,14 according to which the CI test alone cannot tell us what to do in particular action situations—but neither did Kant ever intend it to. After all, the CI is only the expression of the moral law for finite rational beings in general, while in the MM Kant goes on to discuss species-specific duties for human beings as such, becoming even more specific—to differences of gender, race, etc.—in the Anthropology.15 And these latter specifications are not mere afterthoughts but had already been alluded to by Kant in the Groundwork when he stressed that ethics had an empirical part, ‘praktische Anthropologie’.16 So, it would seem, the worry about applying the CI to particular cases rests on a misunderstanding of Kant’s own conception of the CI. The impure ethics strategy can claim great exegetical plausibility, and it is very probable that Kant himself would have subscribed to one version of it. However, there remain grave doubts as to whether the ‘human-specific’ or ’empirical’ parts of Kant’s ethics can, by themselves, answer the constitutive question of what makes certain aspects of a situation morally salient. It is certainly true that these parts of Kant’s ethics will spell out some of the rules of moral salience, but the problem is not merely that the ‘impure’ part of Kant’s ethics has never been fully worked out by himself, but is, at best, fragmentary, as even philosophers who have followed this track, admit.17 The more basic problem is that describing features that are considered as salient for the application of moral laws in human life doesn’t by itself tell us why these features are salient and whence their normative significance for determining what is morally required in a particular situation derives. This is particularly difficult to see with regard to Kant’s primarily descriptive work, for example, in the Anthropology: while the factual information Kant presents here can, plausibly, give rise to further hypothetical imperatives once we know what our moral duties are, and can show us how agents naturally conceive of themselves and their action-situations, it remains unclear why this information should have a foundational role for ethics.18 Nor does an appeal to the Faculty of Judgment (‘Urteilskraft’), to which Kant ascribes the task of applying abstract principles to particular cases— also, explicitly, in the case of moral principles19—by itself fill the gap, or,


Erasmus Mayr

at least, not straightforwardly so. For the problem is not simply the general problem of subsuming particular cases under abstract principles—this general problem faces us even when we apply the CI to already formed particular maxims. The problem is rather to determine—or to ‘construct’— the particular objects to which the abstract principle is to be applied in the first place. What we need is therefore something like a ‘schematization’ of general moral principles, which will have to fulfill a similar task to the schematization of the categories in the Critique of Pure Reason.20 But while Kant, in the latter case, takes great care to argue how this schematization is supposed to work, in his ethics, he does not do so, nor does he tell us what makes one particular way of schematizing the right one.

IV These problems make it attractive to turn to virtues in order to fill the gap in Kant’s account of moral decision-making. The aspect of virtues that makes them appear particularly apt for playing this role is that having a specific virtue is, at least on the traditional, Aristotelian, conception, typically connected to a certain way of seeing situations in which this virtue is relevant. The virtuous man ‘sees’ the situation he faces in the required way, which allows him to emotionally and volitionally respond to it as it is appropriate. For example, the courageous man sees the attacking enemy in battle primarily as a danger to his own hometown, which must be averted, rather than as a menace to his own life, which must be escaped. If possessing a virtue is indeed connected to a special way of seeing action-situations and conceptualizing them, a virtuous person will, as such, be able to pick out those features of a situation that are morally salient and must be included within the content of his maxim. The insight that possession of a virtue has this cognitive component of perception of morally salient features has been defended, at some length, by John McDowell.21 In ‘Virtue and Reason’, McDowell argued that, in moral action, there is a complex two-stage interaction between general knowledge of how to live and particular knowledge about the action-situation in order for the general knowledge to be applied. His characterization is based on the model of the practical syllogism, and not of the CI, but I’ll quote it at some length here: ‘It is at the first stage . . . that knowledge of how to live interacts with particular knowledge: knowledge, namely, of all the particular facts capable of engaging with concerns whose fulfillment would, on occasion, be virtuous. This interaction yields, in a way essentially dependent on appreciation of the particular case, a view of the situation with one such fact, as it were, in the foreground. Seen as salient, that fact serves, at the second stage, as minor premiss in the core explanation’.22 So, in order for general knowledge of how to live to be applied to the action-situation, first some features of the situation are identified as salient in the light of this general knowledge,

A Kantian Plea for Virtues?


for example, ‘This food is healthy’. If one adds to this premise the general principle that healthy food should be eaten, one can, in the second step, construct a practical syllogism with the conclusion that this food should be eaten. It’s crucial to notice, though, that the real work is not constructing this practical syllogism, but identifying the relevant features of the situation that might be used in such a syllogism. And, as McDowell rightly insists, this work is not done by applying codified principles of moral relevance, but is a procedure that is best—perhaps, only—describable in terms of what features a virtuous man would pick out. Some analogous process of singling out morally salient features, it seems, must take place in the process of deciding what is morally right to do when conceived on the Kantian model. But this turns out to be much trickier than on the practical syllogism model used by McDowell. For, on the latter model, the major premise of the syllogism contains some substantial, not purely formal, ethical principle (like ‘healthy food should be eaten’), either constituting or being derived from a conception of how to live, whose content offers (some) guidance in picking out the relevant features of the situation. This is not the case with the CI, which is a purely formal principle. Furthermore, remember, our problem is not merely an epistemological one, but a constitutive one, about what determines which aspects are morally salient in this situation. This cannot be the CI test itself—for, as we have seen, applying the CI test already presupposes a maxim that picks out the aspects considered as salient, which means that making the CI test itself generally determine the morally salient aspects would lead to an infinite regress. So, do we have to accept additional and independent normative facts about which aspects are morally salient? Once we admit such facts, we seem to have given up on the central claim of the Kantian view that the imperatives of morality are obligatory for us qua rational beings simply in virtue of the structure of the will we have. For once we start to accept facts about moral relevance that are not themselves anchored in the structure of a rational will, it seems that we can, just as well, go along all the way to accepting moral facts about what to do that are thus independent. (And for McDowell himself, perception of morally salient features does also involve a perception of what these features demand of us.)23 But then, what role will be left for the CI test to determine whether an action is right or wrong? The test will have become dispensable, both in determining the rightness of actions and in explaining our grasp of it, given that we have to accept independent moral facts about the rightness and wrongness of actions, anyway, and that we must be aware of them in order to apply the test. So, while filling the gap between action-situation and applying the CI test requires some account of which features are morally relevant and of what makes them relevant, the Kantian account would be abandoned rather than completed if the sensitivity to morally relevant features was so comprehensive as to already include sensitivity to what was demanded of us. What the Kantian account needs to be supplemented by is only something that


Erasmus Mayr

excludes cases of moral ‘aspect-blindness’, such as the failure to see that the shyness of a person is a morally relevant factor for situations.24

V I am going to propose that by appealing to the notion of a ‘virtuous agent’ we can get just that: an account of moral salience in terms of the features a ‘virtuous agent’ would take into account, which does not introduce a source of moral rightness independent from the foundations of the moral law and, therefore, is compatible with Kant’s key claims about the source of moral obligation. At first glance, this task may appear an impossible one to complete since, after all, in order to give a workable account of moral salience we must go beyond the resources of the CI test itself. But, as I am going to argue, we can exploit some key features of the virtues—in particular, the fact that they are character-traits and have an ‘holistic’ aspect—to gradually enrich a Kantian ‘thin’ notion of a virtuous agent and a good life, so as to get substantial enough notions of both that can give us the criterion of moral salience we are looking for. A look at Aristotelian virtue ethics strongly suggests that once one possesses a substantive conception of a good, ‘flourishing’ human life, it is much easier, though certainly not a trivial matter, to determine concern for which features of situations human beings typically encounter within a certain cultural environment will be required to lead a good human life (e.g., if one thinks that a good human life includes engaging with one’s fellow humans and their concerns, one can explain why their feelings are relevant features in situations where one’s actions could hurt those feelings). However, how can we construct a sufficiently substantial conception of a good human life within a Kantian framework? Especially Kant’s own repeated criticisms of Eudaimonism in ethics seems to make such an attempt futile from the start.25 We should remember, however, that these criticisms are not directed against developing an account of human flourishing that could play a fruitful role in ethics, per se, but against deriving our moral duties from an independently developed notion of flourishing or happiness. In fact, there are several promising candidates for deriving a notion of human flourishing within the Kantian framework. One such candidate is the notion of the ‘Highest Good’,26 which combines moral rightness and human happiness, distributed in accordance with (and thus conditioned by) moral desert.27 Since the notion of the ‘Highest Good’ is defined, partly, in terms of morality, an account of human flourishing that is based on it escapes Kant’s own anti-Eudaimonistic arguments. The notion of the ‘Highest Good’ is a purely formal notion, though, given that ‘happiness’, for Kant, is to be spelt out in terms of desire-fulfilment and agents’ (even virtuous agents’) desires can vary wildly; so this notion can hardly be expected to yield substantial constraints

A Kantian Plea for Virtues?


on moral salience. Especially since all rational finite beings will aim at this good, it is not to be expected that this notion introduces the human-specific content for rules of moral salience that we need. I propose, therefore, not to start with the ‘Highest Good’ as the idealized ‘endpoint’ of human moral endeavor, but instead with the notion of a virtuous human agent and look at which characteristics this agent must possess and which features his life must have in order to count as a ‘good’ human life. What will characterize such an agent and his life (or, given that virtuous agents can have lives that significantly differ from each other, what will these agents and their lives have in common)? On a once-popular caricature of Kantian moral philosophy, this life will be one in which the agent faces and overcomes many situations of moral temptation and adversity—since this makes him more likely to perform actions that have moral worth than otherwise. Indeed, Kant’s remarks on matters such as the moral worth of self-preservation28 suggest that a life in which situations of moral temptation or adversity were completely missing would, in an important sense, be inferior to a life that contains some such ‘unhappy’ situations in which the agent resists temptation. But we shouldn’t infer that for Kant a life continually beset by strong temptations that the agent is always engaged in resisting is to be judged superior to a life that contained considerably less such ‘unhappy’ situations.29 In particular, we should distinguish these cases according to the source of these temptations: There would be no reason to consider a life as ‘superior’ that was full of such unhappy situations only through the fault of the agent himself, while it would be different when these situations arose from external circumstances. The following four kinds of characteristics seem, instead, more plausible candidates to describe the essentials of a ‘good’, that is, virtuous, human life on a Kantian conception. (1) First, leading a moral life must be part of leading a good life. This not only involves (a) reliably fulfilling one’s strict duties, but also (b) having adopted and pursuing the aims we are morally bound to adopt, according to the Tugendlehre in the MM, that is, one’s own self-perfection and the welfare of others, and fulfilling the duties of virtue that arise from these aims.30 In particular, a moral life will include an absence of the vices Kant discusses in the MM and the Religionsschrift, such as envy or insincerity. And (c) a human agent’s leading a moral life presupposes that he possesses those mental faculties that in the MM are described as necessary for the agent to be affected by moral demands in the first place.31 (2) There are additional features that, while not themselves strictly constitutive of a moral life, are directly connected to it because they are, for human beings, ‘practically’ necessary for reliably fulfilling their moral duties. (a) Partly, these are character-traits and cognitive capacities of the agent, such as the ability to control one’s desires and passions and the capacity for ‘cool thought’.32 (b) But also external circumstances are relevant, in particular a certain amount of happiness, understood as fulfillment of one’s desires, will be required because a continuous thwarting of one’s desires and a life of misery threaten to undermine one’s willingness for moral action in the long


Erasmus Mayr

run and to expose one to temptations for wrong-doing.33 (3) While (1) and (2) have included features that derive from the necessarily moral qualification of a virtuous life, there are further characteristics that arise from the trivial fact that a good human life is a life of a finite rational being at all. (a) First, the conditions of human agency, in particular the fact that we necessarily pursue our own happiness and do so through our own agency. Therefore, a good life will contain the pursuit of fulfillment of our desires and the possession—or continuous acquirement—of a sufficient set of capacities to fulfill a substantial set of these desires, or the ability to renounce those desires whose pursuit will make us unhappy. (There is an obvious connection to (1) (b) and the duties of self-perfection here.) The conditions of human agency also include our vulnerability vis-à-vis natural factors and the agency of other human agents. (b) A human life must be one that ‘can be lived as a whole’. This means, on the one hand, that a good human life must not put on the person in question unrealistically high requirements and that the life displays a sufficiently high degree of coherence and engagement in fulfilling projects not to ‘drive the person mad’, thus undermining his potential for agency. On the other hand, concerns and interests at each stage of the life must be considered. (4) Furthermore, given that we are social animals with other-directed desires and needs, a good life will be one in which we engage, by and large, successfully with others, evading an overly great measure of conflict. (If this component seems overly essentialist, one can also justify it by appeal to the morally prescribed aim of beneficence to others, whose pursuit will be either made possible at all, or at least significantly furthered, by this component.) These four elements provide us with the beginning of a conception of a ‘good human life’ within a Kantian theory. They constitute a ‘thin’ account of a good life and a virtuous agent, which has the advantage of only comprising features that either directly follow from our status as human agents, or are in an obvious way—either as constitutive parts or ‘practically’ necessary elements—connected to morality. Therefore, it introduces no features that are problematically independent from the sources of the moral law as it applies to us as human agents. However, this account is still too ‘insubstantial’ to generally allow us to determine concern for which features of action-situations are necessarily part of leading a good life. Especially, many of the features I’ve mentioned are still far too abstract for this, for example, the pursuit of fulfillment of our desires. Can we develop the very ‘thin’ account into a more substantial account of a good human life that allows us to determine which aspects of situations are morally salient without, at the same time, introducing independent normative facts? I think we can, once we notice that in order to achieve a good life even according to the very ‘thin’ account, a human being needs to possess a certain number of character-traits and intellectual capabilities; in particular, these are the character-traits included under (1) and (2). These character-traits, unsurprisingly, include many of those traits traditionally termed as ‘virtues’, such as moderation or courage, which will help one withstand temptations

A Kantian Plea for Virtues?


to immoral conduct, or to withstand threats that attempt to force one to immoral actions. (Something of which Kant himself was well aware since he explicitly discusses cases of the latter kind.34) In addition to these charactertraits, there are also a number of intellectual virtues, such as prudence, which is crucial to achieving a good human life, by enabling the agent to form plans that are not likely to be generally thwarted. But while these character-traits and intellectual virtues are crucial to achieving a good human life on Kant’s ‘thin’ account of a good life, this does not exhaust their effects on their bearers. Courage, for instance, is a character-trait that not only manifests itself in actions that are morally required but has a much wider range of manifestations. Courage will therefore make its bearer behave in a variety of ways that are not essentially connected to the ‘thin’ notion of a good life, by making him someone who generally does not shrink from taking risks when this seems reasonable, given the dangers he incurs and what he hopes to gain. Similarly, the virtue of prudence will prevent its bearer from taking unreasonable risks for gain even in cases where incurring a loss or a gain makes no difference to the achievement of a good human life on the ‘thin’ account. This means that we can enrich the notion of a good life on a Kantian account by using a recursive procedure of the following kind: (i) We start off from a ‘thin’ notion of a good life. In a first step, we identify those charactertraits and intellectual virtues possession of which is required for the achievement of a good life in this sense. (ii) In a second step, we look at how someone who possesses these character-traits and intellectual virtues would generally behave, what kind of life he would lead, and what kinds of maxims he would adopt. As the essential features of a good life included in the original ‘thin’ notion of a good life have to be maintained, we will ‘correct’ the resulting enriched description of a ‘good life’ of the agent so as to eliminate immoral behavior and maxims as well as internal inconsistencies that arise from the combination of the different character-traits, etc. As we have already seen, the notion of the life this agent would be leading will already be considerably more contentful than the ‘thin’ notion of a good life we have started with. (iii) In a third step, we can, again, enrich this notion, by adding further character-traits and intellectual capacities possession of which is strictly or ‘practically’ required for leading a life of the kind we have described in step (ii), but that were not yet included in the set of traits and capacities described in step (i).35 Again, we will have to correct for immoral behavior and ‘curtail’ the scope and extent of the character-traits and intellectual virtues so as to avoid inconsistencies. And so on. Plausibly, at one point in this series of expansions of the set of character-traits and corrective ‘curtailings’, the set of traits that have been picked out will stabilize, given that the overall set of potential human character-traits is finite. So, by completing a series of steps of the described kind we arrive at an enriched conception of a good life and at a corresponding set of character-traits and intellectual virtues possession of which is required for achieving this good life.


Erasmus Mayr

Once we have arrived at a ‘rich’ conception of a good life in this way, what remains to be done is only to derive those concerns that are relevant for the achievement of a good life from our conception of the latter. Given the enormous variability of situations a virtuous agent encounters, and as a consequence of the considerations McDowell has presented in his ‘Virtue and Reason’, it is not plausible to expect that there will be a codifiable list of morally salient features of actions-situations. But we can circumvent this obstacle by appealing to the figure of the ‘virtuous agent’ itself,36 that is, the figure of an agent who possesses the set of traits and capacities that we have identified in the course of developing the ‘enriched’ conception of a good life. The ‘virtuous agent’ possesses all the individual characteristics that are necessary, as far as the agent himself is concerned, for the achievement of a good life, both with regard to character-traits and with regard to intellectual virtues. He will have those concerns that are required for the achievement of a good life and will, consequently, consider those aspects that are rightly connected to these concerns as morally salient in an action-situation. We can therefore identify the morally salient features of an action-situation as just those features that a ‘virtuous agent’ would consider as morally salient and that would be reflected in the content of any maxim he would choose were he to find himself in this action-situation.

VI So here is, finally, the answer to our constitutive question about what makes it the case that a certain feature of an action-situation is morally salient: It is that a virtuous agent would take this feature to be morally salient, where the figure of the ‘virtuous agent’ has been constructed by the recursive process we have described. It is important to note four things about this answer: (i) First, since we have explained moral salience via the construction of the figure of the ‘virtuous agent’, we have not been forced to introduce normative facts about moral salience that are independent from the structure of a rational agent’s will, which, for a Kantian account, is the sole genuine source of normativity. (ii) Second, at the same time, there is a genuine and irreducible role for ‘virtue’ and the ‘virtuous agent’ in this account of moral salience. On the one hand, the notion of the ‘virtuous agent’ plays a genuine role in deciding what the agent should do in a particular situation—by determining which features of the situation are salient. On the other hand, since, during the recursive process, we have gradually enlarged the set of character-traits and intellectual capacities to be counted among the virtues, by exploiting the fact that virtues are character-traits that do not only manifest themselves in behavior of a certain kind, we cannot eliminate the notion of virtues from our account of moral salience in favor of other notions, such as a disposition to act on principles of right. So, the notion of virtue that we need in order to explain

A Kantian Plea for Virtues?


moral salience turns out to be tied to, but not reducible to, the notion of moral rightness. (And, indeed, had it turned out to be so reducible, the notion could not have helped us to explain moral salience at all—because, as we have seen, the CI Test alone was insufficient to determine which features of an action-situation are morally salient.) (iii) Third, during the procedure we have gradually fed more and more human-specific facts into our account of a good life and a virtuous agent, by exploiting knowledge about the conditions of agency and the constitution of character-traits that is specific for humans. For example, we have exploited the knowledge that courage does not only manifest itself in morally relevant situations, which is, presumably, a specific fact about courage in humans. Thus, we can escape the objection that the notion of the ‘virtuous agent’ is too general and abstract to determine what is morally salient for humans. (iv) While we have worked with one notion of a virtuous agent, it is crucial to notice that we haven’t presupposed that all virtuous agents must be the same—nor is this a result of the argument presented here. On the contrary, virtuous agents can be different and can, to a degree, consider different things as salient, for example, due to their diverging aims or due to different cultural settings. What we have been developing is only an account of which features all virtuous human agents will consider as salient—for these are the features that must be taken into account by an agent in a particular action-situation when he is forming his maxim. I therefore submit that appealing to virtues in the way described is indeed an attractive way to close a crucial lacuna in Kantian moral theory because it offers us a viable account of moral salience that is compatible with the key Kantian contention that morality is obligatory to us qua rational beings in virtue of the structure of our will. If this is correct, then, curiously enough, appealing to irreducible virtues might well be a Kantian’s best bet to save the ‘purity’ of the moral law, while ensuring its applicability.

NOTES I am grateful to Andree Hahmann, Franz Knappik, Julia Peters and Wilhelm Vossenkuhl for very helpful comments on earlier drafts of this paper. Cf. Watson, ‘On the Primacy of Character’, 229. Rawls, A Theory of Justice, 436. MM Akademie-Ausgabe (AA) 6: 394. Critique of Pure Reason (CPrR) AA 5: 19: ‘contain a general determination of the will’. 5. The situation-type description covers both external circumstances and the agent’s own self-conception at the time of his action. For different views on the content of maxims see Nell, Acting on Principle, ch. 1. 6. Some of Kant’s formulations in the Critique of Practical Reason suggest that only specification of action and end is needed; e.g., AA 6: 20: ‘prescribes action as a means to an effect’. But Kant’s treatment of the examples in Groundwork of the Metaphysics of Morals (GMM), AA 4: 421 ff., clearly indicates that also some specification of the situation is required.

1. 2. 3. 4.


Erasmus Mayr

7. See O’Neill, Constructions of Reason. Explorations of Kant’s Practical Philosophy, 87. O’Neill, however, believes that this difficulty can be resolved. 8. The same kind of problem arises for the required degree of specificity of maxims, with regard to the description of the circumstances and the action. 9. Pace Herman, The Practice of Moral Judgment, ch. 7. 10. Unless he acts in pursuance of one of the obligatory ends discussed in the Doctrine of Virtue of the MM. 11. Anscombe, ‘Modern Moral Philosophy’, 27. 12. Pace Esser, Eine Ethik für Endliche: Kants Tugendlehre in der Gegenwart, who argues that we always start off our deliberation either from a particular perspective or from an overarching moral perspective, not from a neutral description of the action-situation. ‘Moralisch relevante Erfordernisse der Situation erlangt man in Beschreibungen der Situation, die bereits unter Voraussetzung dieses umfassenden moralischen Anspruchs angestellt werden’ (271). 13. Herman, The Practice of Moral Judgment, 77 ff. 14. For this strategy in general see Louden, Kant’s Impure Ethics. 15. For an overview over the different degrees of ‘impurity’ in Kant’s ethics see Louden, Kant’s Impure Ethics, 10 ff. 16. GMM AA 4: 387. 17. Louden, Kant’s Impure Ethics, 6 ff. 18. This kind of worry about the role of ‘moral anthropology’ is raised by Gregor, Laws of Freedom, 8. 19. GMM AA 4: 389. 20. In MM 6: 468, Kant himself uses this comparison. 21. The locus classicus being McDowell, ‘Virtue and Reason’, 65 ff. Also Herman, Moral Literacy. 22. McDowell, ‘Virtue and Reason’, 69. 23. Cf. McDowell, ‘Are Moral Requirements Hypothetical Imperatives’, 80 ff. 24. Example from McDowell, ‘Are Moral Requirements Hypothetical Imperatives’, 85. 25. See, e.g., MM AA 6: 377. 26. For an attempt to develop a notion of human flourishing along these lines see Denis, ‘A Kantian Conception of Human Flourishing’. 27. CPrR AA 5: 110 ff. 28. GMM AA 4: 397. 29. Nor is it a life that necessarily exhibits more virtue, as Kant himself stresses: The temptations that are overcome merely allow us to (subjectively) ‘calculate’ the greatness of moral fortitude, but do not objectively determine this greatness, MM AA 6: 397. 30. MM AA 6: 385 ff. 31. MM AA 6: 399 ff. 32. GMM AA 4: 393. Also MM AA 6: 408. As Kant makes clear in the latter passage, there is a corresponding moral duty to cultivate self-control, but pursuing this duty is not the same as already possessing self-control. 33. See, e.g., MM AA 6: 388. 34. CPrR AA 5: 155 f. 35. Plausibly, there will be such further characteristics, and our procedure will not stop with step (ii). For example, leading a life that combines both courage and prudence will require intellectual capacities for assessing danger and possible gain that were not yet included in the set spelled out in step (i). 36. The ideal figure of the ‘virtuous agent’ has a parallel in the figure of Jesus in the Religionsschrift, who incorporates, for Kant, the ideal of moral perfection such as it is possible for human beings, AA 6: 61. (I owe this point to Andree Hahmann.)

A Kantian Plea for Virtues? 209 REFERENCES Page references to Immanuel Kant’s works are according to the pagination of the Akademieausgabe of Kant’s Gesammelte Schriften. English translations are from Gregor, Mary (transl. end ed.), and Allen Wood (general introd.), The Cambridge Edition of the Works of Immanuel Kant: Practical Philosophy (Cambridge: Cambridge University Press 1996). Anscombe, Elizabeth, ‘Modern Moral Philosophy’, in Virtue Ethics, ed. R. Crisp and M. Slote (Oxford: Oxford University Press, 1997), 26 ff. Denis, Lara, ‘A Kantian Conception of Human Flourishing’, in Perfecting Virtue. New Essays on Kantian Ethics and Virtue Ethics, ed. L. Jost and J. Wuerth (Cambridge: Cambridge University Press, 2011), 164 ff. Esser, Andrea, Eine Ethik für Endliche: Kants Tugendlehre in der Gegenwart (Stuttgart: Frommann-Holzboog, 2004). Gregor, Mary, Laws of Freedom (Oxford: Basil Blackwell, 1963). Herman, Barbara, Moral Literacy (Cambridge, MA: Harvard University Press, 2007). ———, The Practice of Moral Judgment (Cambridge, MA: Harvard University Press, 1993). Louden, Robert, Kant’s Impure Ethics. From Rational Beings to Human Beings (New York: Oxford University Press, 2000). McDowell, John, ‘Are Moral Requirements Hypothetical Imperatives’, in Mind, Value, and Reality, ed. J. McDowell (Cambridge, MA: Harvard University Press, 1998), 77ff. ———, ‘Virtue and Reason’, in Mind, Value, and Reality, ed. J. McDowell (Cambridge, MA: Harvard University Press, 1998), 50 ff. Nell, Onora, Acting on Principle (New York: Columbia University Press, 1975). O’Neill, Onora, Constructions of Reason. Explorations of Kant’s Practical Philosophy, (Cambridge: Cambridge University Press, 1989). Rawls, John, A Theory of Justice (Oxford: Oxford University Press, 1971). Watson, Gary, ‘On the Primacy of Character’, in Virtue Ethics, ed. S. Darwall (Oxford: Blackwell, 2003), 229 ff.

14 Toward a Humean Virtue Ethics Lorenzo Greco

I More and more scholars, for various and often contrasting reasons, have recently put Hume’s moral philosophy under the heading ‘virtue ethics’. Also, many contemporary philosophers are trying to elaborate a specific form of Humean virtue ethics to be contrasted with the more famous neo-Aristotelian alternatives. Hence, as occurred with the renaissance of Aristotelian virtue ethics, it appears that there is space to develop a full-fledged Humean version of it as well. My scope here, however, is more limited. After having presented the main reasons in favor of a classification of Hume among virtue ethicists, what I would like to do is to take into account some recent attempts at presenting a virtue ethical interpretation of Hume, with the aim of shedding some light on the theoretical direction I believe a project of a systematic Humean virtue ethics should take. I shall proceed by addressing some specific issues raised by the favorable reading of Hume provided by Christine Swanton1 and by the criticism moved against Hume by Rosalind Hursthouse.2 By doing that I’ll argue that Hume offers the philosophical tools to redefine some basic notions of virtue ethics in a more efficacious way compared to the opposing neo-Aristotelian model and that the strength of Hume’s version of virtue ethics is that he aims at the unity of character instead of the unity of the virtues. This makes it possible to develop a pluralistic and secularized morality that denies any supposed final cause or télos for human beings conceived as a species and instead upholds the individuality of the person as the fundamental value that should be pursued and promoted.

II To begin with, is Hume’s ethics a form of virtue ethics in all ways? What cannot be denied is that Hume himself, in his examination of morality, recognizes a crucial role to the notions of virtue and vice (EPM 1.10; SBN 173–174).3 Hume’s intent is to give a list of virtues and vices in accordance with the way human nature develops within particular contexts.4 Moreover,

Toward a Humean Virtue Ethics


Hume tells us that the objects of moral judgments are not people’s actions, but the motives that lie behind them; human actions may well be regarded in a positive or in a negative light, but only insofar as the motives that activated them are valued positively or negatively (T; SBN 477).5 In turn, these motives have to be related to the characters of people, and people are morally evaluated because they display characters of certain kinds (T–5; SBN 575 and T; SBN 584). This progress from actions to motives, and from motives to characters that make persons virtuous or vicious agents, brings Hume’s conception of morality very close to a virtue ethical model. Moreover, since agents are morally evaluated because of their characters, the way these characters are formed becomes an issue of the greatest importance for Hume (T; SBN 500–501).6 Hume appears to be concerned with that ‘ethical formation’7 that again occupies so much space in many virtue ethics discussions. It is important, however, to stress the original way in which Hume explains how characters develop, an explanation that is in line with his sentimentalism. Hume says that, by ‘custom and education’ (T; SBN 500) people can build up ‘calm’ passions, whereby it is possible to lead lives guided by stable principles of action. Often confused with reason because of their lacking of emotive violence, calm passions are in fact for Hume strong passions that organize one’s existence according to goals that in the end become firm and coherent. Thanks to calm passions, agents acquire ‘strength of mind’ by which they are able to persist in the realization of their projects, without being tempted by false ends—maybe more appealing in the short period, but in fact pernicious to their lives considered in their totality (T and 10; SBN 417–418).8 Only those who are properly educated and have adopted the correct habits will curb their passions and fortify those characters that will make them virtuous agents. But it is worth repeating that this moral learning, for Hume, works purely and solely at a sentimental level. Virtuous agents are those who come to be moved by calm passions, which correspond to traits of character regarded as virtues. This marks a difference between the Humean conception of ethics and other virtue ethical approaches—in particular some kinds of neoAristotelianism9—according to which being properly educated means being able to respond correctly to the moral features of a given situation. According to this neo-Aristotelian model, virtue should foremost be considered as a form of knowledge, and the virtuous person as someone who first of all gets things right and then acts accordingly. The phronimos is gifted with a perceptual capacity, usually explained in intellectual terms as a form of moral wisdom, by which he or she becomes sensible to the suitable requirements that the situation imposes on behavior. Conversely, Hume makes no reference to any intellectual faculty of any kind when he has to explain how a person becomes a virtuous agent; the Humean virtuous person does not act on the strength of such a faculty as ‘either desire-related intellect or thought-related desire’,10 which guarantees at the same time the right look


Lorenzo Greco

on things and the motivational force to move consequently. Besides, for Hume ‘morality [. . .] consists not in any matter of fact, which can be discovered by the understanding’ (T; SBN 468). Rather, values seem to work rather like secondary qualities (T; SBN 469). Whether the secondary quality comparison is the best way to explain Hume’s conception of the nature of values is still a much debated question, and I will not address it here. However, what can be observed is that, though for Hume the dimension of values is presented as a sort of projection onto the world due to the sentimental framework of human nature, this dimension does not require anything beyond this very sentimental framework to be stated. Taste, Hume affirms, moral and aesthetic, ‘has a productive faculty, and gilding or staining all natural objects with the colours, borrowed from internal sentiment, raises in a manner a new creation’ (EPM appendix 1.21; SBN 294). ‘To have the sense of virtue, is nothing but to feel a satisfaction of a particular kind from the contemplation of a character. The very feeling constitutes our praise or admiration. We go no farther; nor do we enquire into the cause of the satisfaction’ (T; SBN 471). This is because ‘there is just so much vice or virtue in any character, as every one places in it, and [ . . . ] ’tis impossible in this particular we can ever be mistaken’ (T; SBN 547; see also ‘The Sceptic’, 168). These passages in Hume’s texts seem to justify the conclusion that for Hume the evaluative dimension is a sentimental representation—not an intellectual one—that human beings cast on things as a result of the activity of their passions—not a state of affairs that is perceived, and with which the virtuous person becomes attuned. True, he says that in morality ‘reason and sentiment concur in almost all moral determinations and conclusions’ (EPM 1.9; SBN 172); but this ‘reason’ Hume refers to here is to be translated in terms of calm passions (T; SBN 583). In a sense human beings can sometimes be said to be ‘morally blind’ to the relevant ethical features of situations.11 If what has been said so far is correct though, reference to moral blindness (and, conversely, to moral vision) is to be taken figuratively. Human beings may be morally blind for Hume because they are primarily morally insensible, that is, because they are endowed with a poor sentimental equipment, incapable of being affected by sympathetic exchanges among people. There is not really anything to be seen out there; ‘seeing’ makes sense if taken as a metaphor for ‘feeling’ in a proper way, given a human nature described in sentimental terms, which presents itself as the benchmark for stating what the virtues and vices are.

III This interpretation of the way Hume conceives the sphere of value, and the role played by sentiment in it, does not go without criticism. Sentimentalism may be defined very broadly as ‘the thesis that evaluation is to be understood

Toward a Humean Virtue Ethics


by way of emotional response’,12 and some scholars have recently argued that Hume presents a kind of sentimentalist virtue ethics very close in its results to the neo-Aristotelian one that has been presented as non-Humean so far. For example, Christine Swanton considers Hume ‘as being part of both the sentimentalist and the virtue ethical traditions’13 in maintaining a responsedependent theory whereby ‘a virtue or a vice is a power in an object to elicit relevant responses in qualified actors’.14 In turn, a qualified actor is someone in possession of certain emotional dispositions that make him or her sensible and reactive to the powers in the object, which are the virtues. In this sense, in Swanton’s interpretation of Hume, ‘morality is not a matter of fact about our sentiments, it is a matter of fact about virtue and vice, which are in objects’;15 ‘virtues are response-dependent properties, and are therefore not projections as some commentators claim’.16 By appropriately exercising their moral sense, human beings can thus track the moral truths that allegedly compose the ethical reality.17 A partly similar conclusion has been recently given also by Michael Slote in his sentimentalist ethics of care.18 This response-dependent reading of Hume is a fascinating way of assessing his ethical sentimentalism in the light of virtue ethics, but doubts can legitimately be raised both about whether it corresponds to Hume’s own intentions and also, more generally, about whether this is the correct way to frame a Humean virtue ethics.19 It is indicative, for example, that both Swanton and Slote make reference to the work of David Wiggins. Wiggins presents a ‘sensible subjectivism’, according to which moral properties and appropriate human sentiments are mutually correlated in associations, so that ‘x is good/right/beautiful if and only if x is such as to make a certain sentiment of approbation appropriate’.20 By appealing to nothing more fundamental than human sentiments, Wiggins aims at giving a cognitivist account of the sphere of morality in which the claim to objectivity that appears to be deeply rooted in the very concept of morality finds its proper vindication.21 In developing his sensibilist model, Wiggins mentions Hume as one of the authors with whom he has a close affinity. But whereas Swanton says that Hume’s virtue ethics corresponds to a response-dependent theory matching Wiggins’ sensibilism, Wiggins, on his part, admits instead that his sensible subjectivism diverges from Hume’s ‘official theory’.22 Wiggins says that we can (and indeed we should) progressively move from ‘[The real] David Hume’ (who roughly corresponds to the projectivist description given above) to ‘[A possible] David Hume: x is good if and only if x is such as to arouse approbation’, and eventually end up with a ‘Refined Humean subjectivism: x is good if and only if x is such as to deserve (N.B.) or merit approbation’.23 But this is not what the real Hume does. So why should Hume (and those of us who want to develop a Humean virtue ethics) make this move? A revealing answer is given by Swanton herself: a Humean response-dependent virtue ethics can account for the reasongiving force of ethics, and in particular for our justification about the


Lorenzo Greco status of traits as virtues and vices. If we can show this, we will have refuted Philippa Foot’s view that Hume’s theory about moral sentiment ‘commits him to a subjectivist theory of ethics,’ and that for him there is no ‘method of deciding in the case of disagreement, whether one man’s opinion or another’s was correct’.24

The concern expressed by Foot,25 and echoed by Swanton, is that of ending up with a theory incapable of determining which moral answers are truly fitting, thus deserving authentic approbation. That is, Hume’s ethics, if it is seen as just focusing on nothing else but emotional states of the subjects, would be devoid of any stable criterion for determining what really deserves to be considered ‘moral’ (such as, for example, traits of character that are real virtues and vices) and what is just pleasant or unpleasant, but morally neutral (such as, for example, traits of character that end up being mere talents or defects).26 Nevertheless, it is odd that Foot’s blow against Hume should be warded off by adopting a strategy that is unlikely to be Humean and that finds a better, and maybe more natural, formulation from within a neo-Aristotelian perspective. Nowhere does Hume speak of ethics as an area where moral truths should be discovered. Nor does his ‘moral sense’ appear as a capacity to track moral truths of any kind.27 What is more, it is disputable that Hume’s purpose is that of providing an objective ethical theory, or that he has any specific problem with objectivity in ethics. As Rachel Cohon observes, ‘It is a little misleading to call Humean moral evaluation objective, since it is based on felt sentiment, but there is a very high degree of convergence in all moral assessments that are properly formed’.28 Hume is surely interested in explaining the convergence in moral judgments and, above all, in accounting for the practical dimension of morality (EPM 1.7–8; SBN 172),29 while he appears not to be concerned with giving an explanation in terms of the supposed objectivity of moral judgments. In this light, both Foot’s criticism of Hume’s subjectivism, and Wiggins’s proposal—taken up by Swanton—to reinterpret it in terms of a ‘sensible’ subjectivism, seem to be grounded in the worry that the lack of such an objectivist ethical criterion in Hume opens the door to ethical nihilism. A danger that is to be blocked either by rejecting Hume’s conception of morality as a whole (as Foot does), or by radically reformulating it (as Wiggins and Swanton do). Nonetheless, as I’ll try to argue, such a Humean convergence without objectivity provides in any case a canon of ethical correctness by grounding it in our human practices. A different interpretation of Hume as a virtue ethicist can be developed that does not focus on moral properties to be found in the world, but on individuals as owners of virtuous or vicious characters. To see how, let me address briefly Hume’s strategy for distinguishing between virtues and vices, in relation to a criticism moved against it by Rosalind Hursthouse.

Toward a Humean Virtue Ethics


IV For Hume, Every quality of the mind is denominated virtuous, which gives pleasure by the mere survey; as every quality, which produces pain, is call’d vicious. This pleasure and this pain may arise from four different sources. For we reap a pleasure from the view of a character, which is naturally fitted to be useful to others, or to the person himself, or which is agreeable to others, or to the person himself. (T; SBN 591) So human beings recognize as virtues those character traits that are useful to their possessors or to others, or immediately agreeable to their possessors or to others. Vices are the opposite. In turn, thanks to sympathy, which is considered by Hume as the principle of sentimental communication among human beings, we can approve those traits of character that produce pleasure or advantage for other people or for their possessors themselves and disapprove those traits of character that give pain or prove to be disadvantageous for other people or their possessors themselves. Specifically, we have a properly moral approval (or disapproval) when these sympathetic judgments on traits of character are given from what Hume calls a ‘steady and general’, or ‘common’, point of view (T–16 and 30; SBN 581–582 and 591. EPM 9.6; SBN 272), from which it is possible to determine a stable and as much as possible impartial perspective on virtues and vices. Now, like Swanton and Slote, Rosalind Hursthouse, too, takes into consideration Hume as a possible representative of virtue ethics, but she discards his moral theory as defective at the very root. In particular, she criticizes Hume’s four sources of pleasure and pain as a correct standard for defining which character traits should be appreciated and which not, since these four sources would correspond to a disjunctive claim, whose upshot is the impossibility of defending a single measure of virtue and vice. Justice and injustice, courage and cowardice, generosity and meanness would all turn out to be virtues.30 Moreover, the steady and general point of view cannot be a correct standard for moral judgments because it would be defined by Hume as ‘uninfluenced by distances in time: it can respond to the virtues of the ancient Greeks as competently as it can respond to those of its possessor’s contemporaries’.31 This would make the Humean steady and general point of view too abstract and distant from those who must endorse it for it to become a reliable standard in ethics. According to Hursthouse, to save Hume’s theory from collapsing, it has to be, so to speak, ‘Aristotelized;’ the steady and general point of view should be discarded as a reliable ethical yardstick and replaced with the good ‘critic’ in morals as it is expressed by Hume in his essay Of the Standard of Taste.32 Such a good critic is interpreted by Hursthouse as the well-trained person, who is immersed in a particular


Lorenzo Greco

reality of which he or she is able to recognize the relevant moral aspects, thus representing the closest approximation to the phronimos we can arrive at within a Humean framework. Yet what should be noted is that Hursthouse moves her objections against Hume while taking for granted from the very beginning the Aristotelian perfectionist conception of human nature she endorses as normatively sound. For Hursthouse, in fact, ‘the standard neo-Aristotelian completion claims that a virtue is a character trait a human being needs for eudaimonia, to flourish or live well’.33 And she defines what it means to ‘live well’ by making reference to those distinctive functions characteristic of human beings whose fulfillment allow human beings to live in the right way, as they are required qua human beings, and thus to obtain the real happiness, or the sort of happiness worth having.34 By presupposing such a unit of measurement—human nature as she conceives it—Hursthouse can present a notion of the phronimos that corresponds to somebody who shows practical wisdom, gathering coherently in himself or herself all the virtues at once, hence embodying in himself or herself the criterion for objectivity that has been looked for so far. However, Hume has never professed the need to single out a criterion of good and right that has to be valid in advance and that guarantees something like the unity of the virtues. Nor does the Humean steady and general point of view correspond, as Hursthouse seems to believe, to a timeless ‘point of view of the universe’, or a ‘view from nowhere’. It is, rather, a point of view that develops within human history as the result of people’s sympathetic exchanges, that is, of a moral sentiment where ‘is displayed the force of many sympathies’ (EPM 9.11; SBN 276). It is a reflective stance resulting from that moral conversation human beings entertain because of their sentimental constitution that assures a convergence in moral judgments, but does not provide that single, definitive measure of objectivity neo-Aristotelians are looking for. On the contrary, Hume’s steady and general point of view evolves through time and space, leaning on a fixed human nature whose constancy is nothing but the product of a generalization (EHU 8.7; SBN 83–84).35 Hume’s way of establishing what constitutes human flourishing is always an a posteriori operation, the consequence of empirical ascertainment. Which character traits happen to be agreeable or useful to their possessors or to others can be derived from ‘a cautious observation of human life’, and the list of virtues we will come up with is the outcome of ‘experiments [. . .] as they appear in the common course of the world, by men’s behaviour in company, in affairs, and in their pleasures’ (T Intro. 10; SBN xix). His quadripartite standard works contingently in the course of human history by considering how human nature expresses itself in the multiplicity of situations in which people find themselves. So Hume is far from presupposing a finalistic notion of human nature and then stating which character traits fully realize human nature’s peculiar ends. In turn, the Humean good critic is precisely someone who puts himself or herself, and reflects, from

Toward a Humean Virtue Ethics


the perspective of the steady and general point of view. That is to say, the good critic can be seen as the criterion for judging what is virtuous or vicious only insofar as he or she has embraced that very point of view. His or her practical wisdom does not reflect any phronesis whatsoever, but is the consequence of having adopted that contingent position that is the product of the continuous corrections human beings progressively bring to their moral assessments, thanks to their uninterrupted passional exchanges, and to their imaginative efforts to get in touch with their fellow men. In this sense the Humean good critic is a human being like anyone else, but one who has educated himself or herself to be more sentimentally attentive, less prone to prejudice, more willing to engage in specifically moral arguments, and thus to recognize moral distinctions established from the steady and general point of view as sound and to be pursued.

V From a Humean perspective, unlike from a neo-Aristotelian one, there is no ontological commitment regarding the nature of the virtuous agent. NeoAristotelians36 long for a unity of the virtues that can be stated only by presupposing an idealized notion of human nature, and hence by presupposing a notion of virtuous person—the one who is in possession of all the virtues—that presents an ideal of perfection that in fact is at risk of never being fulfilled by anybody real. Instead, what interests Hume is the determination of virtuous characters that are always specified a posteriori and can be referred back to the passions of the persons. Hume presents a picture according to which people become aware of themselves as particular persons insofar as they possess firm characters; being conscious of their own individual character is the element by which agents gain that unity that allows them to stand before others as identifiable individuals.37 And becoming conscious of one’s own character is possible for Hume due to the passion of pride (T; SBN 596–597 and T; SBN 599). What comes out from Hume’s explanation of pride is that the proud person ends up coinciding with the virtuous person. More specifically, the Humean virtuous person is someone endowed with a stable, ‘moralized’ pride, that is, with a stable sense of himself or herself as an individual who plays a role in the particular context he or she lives in and who is recognized and positively valued by those around him or her.38 But is not this the same as the Aristotelian phronimos? Not at all, for the Humean virtuous person is proud of precisely those character traits that are praised from that steady and general point of view that neo-Aristotelian perspectives like Hursthouse’s have excluded: a point of view that, even if it always reveals itself within human affairs, does not necessarily correspond to the point of view shared by a specific society. The Humean virtue ethics proposal is far from relativistic; by making reference to the sentimental


Lorenzo Greco

imagination of human beings, which is exercised in continual confrontations that take place within human history, Hume proves to possess the philosophical instruments for explaining how moral progress is possible. In contrast, the neo-Aristotelian model is stuck with a conception of the phronimos that comes to be rigid and hardly helpful for contemporary ethics. On the one hand, it is disputable whether the authentic notion of the Aristotelian phronimos is ever applicable to our contemporary liberal societies. On the other, by being defined through a notion of human nature whose proper goals are finalistically presupposed from the outside and not subject to any modification, the modern version of the phronimos ends up being relativized to the particular contingency in which he or she is able to exercise his or her perceptual capacity—with the result that it lays itself open to the criticism of having a skeptical outcome in ethics, and a communitarian one in politics.

VI The picture of the virtuous person as the proud person allows Hume to present his own peculiar notion of human excellence—a notion that competes with the one belonging to the classical, that is, ancient Greek tradition of virtue ethics.39 This Humean conception of human excellence takes the form of ‘greatness of mind’—which for Hume is nothing but a steady and well-established pride and self esteem—which displays traits of character such as courage, ambition, love of glory, magnanimity, explicitly presented by Hume as closely related to the classical world, and in opposition to the distorted values of the Christian tradition (T; SBN 599–600). Now, greatness of mind may well reveal itself also in the form of heroism and military glory. And even though ‘men of cool reflexion are not so sanguine in their praise of it’, because of the great damages it may cause to society, when we fix our view on the person himself, who is the author of all this mischief, there is something so dazling in his character, the mere contemplation of it so elevates the mind, that we cannot refuse it our admiration. The pain, which we receive from its tendency to the prejudice of society, is over-power’d by a stronger and more immediate sympathy. (T; SBN 601) Here Hume touches a point that has been acknowledged and accepted by present-day virtue ethicists such as Slote and Swanton, namely, the idea that there may exist an ‘admirable immorality’40 and that we frequently esteem virtues that do not bring any benefit to humankind.41 In doing that, Hume develops a virtue ethics that could be defined as ‘pluralistic’, to use Swanton’s expression.42 However, Hume’s ethics can be said to be pluralistic in a different way from Swanton’s. She conceives her pluralistic virtue ethics along with a response-dependent line, and the interpretation of Hume’s

Toward a Humean Virtue Ethics


ethics presented so far goes in another direction. Nonetheless, describing Hume’s virtue ethics in pluralistic terms makes sense if we take Hume as having as his core moral interest not so much an objective criterion to distinguish virtues and vices, but rather the individuality of persons. Individuality stands out as a value that should be pursued and promoted precisely because Hume has a pluralistic conception concerning virtues and vices, which does not look for the unity of the virtues, but instead for the unity of character.43 Hume does not have a problem of consistency among the virtues; consistency becomes a problem only if we decide to embrace an ‘absolute’ conception such as the neo-Aristotelian one. Rather, from Hume’s a posteriori perspective, we may well admit the existence of people whose characters are mixtures of a plurality of traits,44 some of which are virtues when seen from the steady and general point of view, while others turn out to be vices.45 What counts is character in its totality, as reflecting the individuality of a given person, not the determination of an objective perspective from which to label virtues and vices—a perspective that, Humeanly, runs the risk of being nothing but a philosophical chimera. In a sense, this allows Humeans to regain that notion of an end of human beings that the neo-Aristotelians are so fond of. But in a Humean perspective this notion— as with all the other fundamental notions of the virtue ethics vocabulary— acquires a new meaning. It ceases to stand for a télos of humanity taken as a species, but instead is always used in the plural form, to refer to the most different ends individuals pursue. This is not to be understood as an approximation of the ideal of the phronimos, but instead as the realization of a unified character in the light of the steady and general point of view. Finally, what should also be emphasized is that Hume mentions as an integral part of human excellence the virtue of benevolence (T 3.3.3; EPM 2). This is one of the differences between the Humean conception of a virtuous life and the classical one. What is peculiar to the alternative offered by Hume is that a life can be virtuous only if it is open to others, considered as different persons who deserve our respect. Such moral relevance of benevolence has nothing to do with Christian piety, but again is explained by Hume with reference to the sentimental mechanisms of human psychology. Greatness of mind and benevolence weigh each other out and are virtues insofar as they reveal the social nature of human beings, defining the virtuous person as someone who stands as a morally laudable individual because of his or her connections with other people (T; SBN 606). So it turns out that even though greatness of mind is indeed a virtue for Hume, it may not, in his own terms, be appropriately ascribed to common people. Greatness of mind suits soldiers or noblemen well; it represents an aristocratic way of being virtuous that is certainly accepted by Hume but that he does not consider to be the only or the best way of behaving virtuously. The peculiarity of Hume’s conception of the virtuous person is that it appears to be, as it were, ‘democratized;’ his virtue ethics is not addressed to heroes, even less to saints, but to people as they are commonly found in the world. That is, we do not need


Lorenzo Greco

to be heroes nor saints to be justly proud of ourselves since the steady and general point of view is set on that ‘middle station of life’ that, according to Hume, qualifies the condition of the greatest part of human beings, ‘affording the fullest security for virtue’, and giving opportunity ‘for the most ample exercise of it’.46 Hence according to Hume, virtue emerges as a process of continuous self-improvement in which people develop stable characters they can be proud of, thus conceiving themselves as unitary individuals, without having to presuppose an end-state of ideal or absolute perfection. In the end, it may well happen that, when regarded a posteriori, a certain virtuous person turns out to possess all the virtues. If so, this cannot be but a contingent result. But, from a Humean perspective, this is more than enough.47

NOTES 1. Christine Swanton, ‘Can Hume Be Read as a Virtue Ethicist?’. 2. Rosalind Hursthouse, ‘Virtue Ethics and Human Nature’. 3. Hereafter I shall refer to both Hume’s An Enquiry concerning the Principles of Morals: A Critical Edition and Enquiries concerning Human Understanding and the Principles of Morals, mentioned as ‘EPM’ and cited by section and paragraph number, followed by ‘SBN’ and page number in the Selby-Bigge’s edition. 4. The Enquiry concerning the Principles of Morals presents four sections (from 5 to 8) and one appendix (app. 4) explicitly dedicated to this enterprise. If we go back to A Treatise of Human Nature, we find something similar in the discussion about the difference between natural and artificial virtues. 5. Hereafter I shall refer to both the Selby-Bigge edition and the Norton and Norton edition of Hume’s A Treatise of Human Nature, mentioned as ‘T’ and cited by book, part, section, and paragraph number, followed by ‘SBN’ and page number in the Selby-Bigge’s edition. 6. See also David Hume, ‘The Sceptic’, 159–180, esp. 170. On the importance of the education of the virtuous character for Hume see Russell, ‘Moral Sense and Virtue in Hume’s Ethics’. 7. See Lovibond, Ethical Formation. 8. On Hume’s notion of strength of mind, see McIntyre, ‘Hume’s Passions: Direct and Indirect’, and Wright, ‘Butler and Hume on Habit and Moral Character’. 9. See Hursthouse, ‘Normative Virtue Ethics’; Hursthouse, ‘Virtue Ethics and Human Nature’; Hursthouse, On Virtue Ethics; McDowell, ‘The Role of Eudaimonia in Aristotle’s Ethics’; McDowell, ‘Virtue and Reason’; McDowell, ‘Values and Secondary Qualities’. 10. See Aristotle, Nicomachean Ethics, 1139b. 11. As Hume notes, ‘An action, or sentiment, or character is virtuous or vicious; why? because its view causes a pleasure or uneasiness of a particular kind’ (T; SBN 471, italics mine). On Hume’s moral blindness, see Abramson, ‘Hume on Cultural Conflicts of Values’; Taylor, ‘Humean Humanity versus Hate’. 12. D’Arms and Jacobson, ‘Sensibility Theory and Projectivism’, 187–188. 13. Swanton, ‘Can Hume Be Read as a Virtue Ethicist?’, 92. 14. Ibid., 95. 15. Ibid., 96. 16. Ibid., 97. 17. Swanton’s own position is in fact more complex than this. In Virtue Ethics: A Pluralistic View, discussing Christine Korsgaard’s theses, Swanton recognizes

Toward a Humean Virtue Ethics

18. 19.

20. 21. 22. 23. 24. 25. 26. 27.



30. 31. 32. 33. 34. 35.

36. 37. 38. 39.


that ‘the central practical task of ethics is not simply the search for truth. That search is constrained by an even more fundamental problem: of our needing to live together, solving our problems in ways consistent with this end. Dialogue does not just serve an epistemic truth-seeking goal, it serves also the social goal of solving problems’ (250–251; italics by Swanton). But notwithstanding the importance ascribed to dialogue, the search for (ethical) truths remains for Swanton an integral part of doing ethics. Slote, Moral Sentimentalism. Slote, for example, admits that Hume’s work gives way to different interpretations: ‘It seems to me that you can find large bodies of emotivism in Hume, of subjectivism, of projectivism, of error theory, of ideal observer theory, of response-dependence theory. You can find passages which support each of these forms of metaethics. But it is not clear to me which of these Hume actually believes’ (Slote, ‘Moral Sentimentalism’, 8–9). Wiggins, ‘A Sensible Subjectivism?’, 187. See Wiggins, Ethics: Twelve Lectures on the Philosophy of Morality, 331–333. Wiggins, ‘A Sensible Subjectivism?’, 194. Wiggins, Ethics, 371. Swanton, ‘Can Hume Be read as a Virtue Ethicist?’, 101. Philippa Foot, ‘Hume on Moral Judgement’. Note that Hume never draws a clear distinction between virtues and vices, on one side, and talents and defects, on the other (EPM app. 4). This reading of Hume makes him a ‘moral sense theorist’ very similar to Francis Hutcheson, and there are strong reasons to believe that Hume’s moral sentimentalism is to be framed differently. See Gill, The British Moralists on Human Nature and the Birth of Secular Ethics, chap. 19. Cohon, Hume’s Morality: Feeling and Fabrication, 242, italics mine. For attempts to interpret Hume’s ethics as objectivist, see Norton, David Hume: Common Sense Moralist, Sceptical Metaphysician; Swain, ‘Passionate Objectivity’. It is precisely this preference that is criticized by Foot. But perhaps we should take Hume literally when he tells us that ‘when you pronounce any action or character be vicious, you mean nothing, but that from the constitution of your nature you have a feeling or sentiment of blame from the contemplation of it’ (T; SBN 469). Hursthouse, ‘Virtue Ethics and Human Nature’, 73. Ibid., 78–79. Hume, ‘Of the Standard of Taste’. Hursthouse, ‘Normative Virtue Ethics’, 23. Hursthouse, On Virtue Ethics, part 3. I shall refer to both Hume’s An Enquiry concerning Human Understanding: A Critical Edition and Enquiries concerning Human Understanding and the Principles of Morals, mentioned as ‘EHU’ and cited by section and paragraph number, followed by ‘SBN’ and page number in the Selby-Bigge edition. See, for example, McDowell, ‘Virtue and Reason’. See also Foot, ‘Virtues and Vices’. Rosalind Hursthouse embraces a ‘limited’ or ‘weak’ view on the unity of the virtues in Hursthouse, On Virtue Ethics, 153–157. I discuss this point in Greco, L’io morale: David Hume e l’etica contemporanea, parts 2 and 3. On the notion of a moralized pride, see Baier, ‘Master Passions’; Herdt, Religion and Faction in Hume’s Moral Philosophy, chap. 2; Rorty, ‘The Vanishing Subject: The Many Faces of Subjectivity’. On Hume’s conception of human excellence, see Martin, ‘Hume on Human Excellence’.

222 40. 41. 42. 43. 44. 45. 46. 47.

Lorenzo Greco Swanton, Virtue Ethics, 102. Slote, From Morality to Virtue, chap. 14. See Swanton, ‘Profiles of the Virtues’; Swanton, ‘Nietzschean Virtue Ethics’. On the idea that Hume’s ethics is pluralistic, see Abramson, ‘Hume on Cultural Conflicts of Values’; King, ‘Hume on Artificial Lives with a Rejoinder to A. C. MacIntyre’. On ‘the person of mixed character’ see Cohon, Hume’s Morality, 149 ff. This aspect is well explained by Dees, ‘Hume on the Characters of Virtue’. Hume, ‘Of the Middle Station of Life’, 546. This paper was presented at the following conferences: New Perspectives on Virtues and Vices, Center for Advanced Studies, LMU Munich, Munich Competence Center for Ethics (MKE), February 4–5, 2011; Le legs de Hume dans la philosophie contemporaine, Institut Catholique de Paris, Faculté de Philosophie, September 13–14, 2011; Hume and the Virtues, International Hume Workshop, Oxford Brookes University, May 2, 2012, organized by Julia Peters, Ronan Sharkey, and Daniel O’Brien, respectively. A very early draft had originally been discussed at the 34th International Hume Conference, Boston University, August 7–12, 2007. I would like to thank all the participants at these events who contributed to this paper with their useful comments, and particularly Roger Crisp, Michael Gill, Eugenio Lecaldano, Alison McIntyre, Jacqueline Taylor, and David Wiggins.

REFERENCES Abramson, Kate, ‘Hume on Cultural Conflicts of Values’, Philosophical Studies 94 (1999), 173–187. Aristotle, Nicomachean Ethics, ed. Roger Crisp (Cambridge: Cambridge University Press, 2000). Baier, Annette C., ‘Master Passions’, in Explaining Emotions, ed. Amélie Oksenberg Rorty (Berkeley: University of California Press, 1980), 403–423. Cohon, Rachel, Hume’s Morality: Feeling and Fabrication (Oxford: Oxford University Press, 2008). D’Arms, Justin, and Jacobson, Daniel, ‘Sensibility Theory and Projectivism’, in The Oxford Handbook of Ethical Theory, ed. David Copp (Oxford: Oxford University Press, 2006), 186–218. Dees, Richard H., ‘Hume on the Characters of Virtue’, Journal of the History of Philosophy 35 (1997), 45–64. Foot, Philippa, ‘Hume on Moral Judgement’, in Virtues and Vices and Other Essays in Moral Philosophy (Oxford: Clarendon Press, 2002), 74–80. ———, ‘Virtues and Vices’, in Virtues and Vices and Other Essays in Moral Philosophy (Oxford: Clarendon Press, 2002), 1–18. Gill, Michael B., The British Moralists on Human Nature and the Birth of Secular Ethics (Cambridge: Cambridge University Press, 2006). Greco, Lorenzo, L’io morale: David Hume e l’etica contemporanea (Napoli: Liguori, 2008). Herdt, Jennifer A., Religion and Faction in Hume’s Moral Philosophy (Cambridge: Cambridge University Press, 1997). Hume, David, Enquiries concerning Human Understanding and the Principles of Morals, ed. Lewis A. Selby-Bigge, rev. Peter H. Nidditch, 3rd ed. (Oxford: Clarendon Press 1975). ———, An Enquiry concerning Human Understanding: A Critical Edition, ed. Tom L. Beaucham (Oxford: Clarendon Press, 2006). ———, An Enquiry concerning the Principles of Morals: A Critical Edition, ed. Tom L. Beauchamp (Oxford: Clarendon Press, 1999).

Toward a Humean Virtue Ethics 223 ———, ‘Of the Middle Station of Life’ [1742], in Essays, Moral, Political, and Literary, ed. Eugene F. Miller (Indianapolis: Liberty Fund, 1987), 545–551. ———, ‘Of the Standard of Taste’ [1757], in Essays, Moral, Political, and Literary, ed. Eugene F. Miller (Indianapolis: Liberty Fund, 1987), 226–249. ———, ‘The Sceptic’ [1742], in Essays, Moral, Political, and Literary, ed. Eugene F. Miller (Indianapolis: Liberty Fund, 1987), 159–180. ———, A Treatise of Human Nature, ed. Lewis A. Selby-Bigge and Peter H. Niddich, 2nd ed. (Oxford: Clarendon Press, 1978). ———, A Treatise of Human Nature: A Critical Edition, ed. David Fate Norton and Mary J. Norton, 2 vols. (Oxford: Clarendon Press, 2007). Hursthouse, Rosalind, ‘Normative Virtue Ethics’, in How Should One Live?, ed. Roger Crisp (Oxford: Oxford University Press, 1996), 19–36. ———, On Virtue Ethics (Oxford: Oxford University Press, 1999). ———, ‘Virtue Ethics and Human Nature’, Hume Studies 25 (1999), 67–82. King, James, ‘Hume on Artificial Lives with a Rejoinder to A. C. MacIntyre’, Hume Studies 14 (1988), 53–92. Lovibond, Sabina, Ethical Formation (Cambridge, MA: Harvard University Press, 2002). Martin, Marie A., ‘Hume on Human Excellence’, Hume Studies 18 (1992), 383–399. McDowell, John, ‘The Role of Eudaimonia in Aristotle’s Ethics’, in Mind, Value, and Reality (Cambridge, MA: Harvard University Press, 1998), 3–22. ———, ‘Values and Secondary Qualities’, in Mind, Value, and Reality (Cambridge, MA: Harvard University Press, 1998), 131–150. ———, ‘Virtue and Reason’, in Mind, Value, and Reality (Cambridge, MA: Harvard University Press, 1998), 50–73. McIntyre, Jane L., ‘Hume’s Passions: Direct and Indirect’, Hume Studies 26 (2000), 77–86. Norton, David Fate, David Hume: Common Sense Moralist, Sceptical Metaphysician (Princeton, NJ: Princeton University Press, 1984). Rorty, Amélie Oksenberg, ‘The Vanishing Subject: The Many Faces of Subjectivity’, History of Philosophy Quarterly 23 (2006), 191–209. Russell, Paul, ‘Moral Sense and Virtue in Hume’s Ethics’, in Values and Virtues: Aristotelianism and Contemporary Ethics, ed. Timothy Chappell (Oxford: Oxford University Press, 2006), 158–169. Slote, Michael, From Morality to Virtue (Oxford: Oxford University Press, 1992). ———, ‘Moral Sentimentalism’, Ethical Theory and Moral Practice 7 (2004), 3–14. ———, Moral Sentimentalism (Oxford: Oxford University Press, 2010). Swain, Corliss Gayda, ‘Passionate Objectivity’, Noûs 26 (1992), 465–490. Swanton, Christine, ‘Can Hume Be Read as a Virtue Ethicist?’, Hume Studies 33 (2007), 91–113. ———, ‘Nietzschean Virtue Ethics’, in Virtue Ethics, Old and New, ed. Stephen G. Gardiner (Ithaca, NY: Cornell University Press, 2005), 179–192. ———, ‘Profiles of the Virtues’, Pacific Philosophical Quarterly 76 (1995), 47–72. ———, Virtue Ethics: A Pluralistic View (Oxford: Oxford University Press, 2003). Taylor, Jacqueline, ‘Humean Humanity versus Hate’, in The Practice of Virtue: Classic and Contemporary Readings in Virtue Ethics, ed. Jennifer Welchman (Indianapolis: Hackett Publishing, 2006), 182–203. Wiggins, David, Ethics: Twelve Lectures on the Philosophy of Morality (London: Penguin, 2006). ———, ‘A Sensible Subjectivism?’, in Needs, Values, Truth: Essays in the Philosophy of Value (Oxford: Clarendon Press, 1998), 185–214. Wright, John P., ‘Butler and Hume on Habit and Moral Character’, in Hume and Hume’s Connexions, ed. Michael A. Stewart and John P. Wright (Edinburgh: Edinburgh University Press, 1994), 105–118.

List of Contributors

Philipp Brüllmann is Research Fellow at LMU Munich. His interests lie in ancient ethics (especially Aristotle and the Stoics) as well as contemporary ethics. He is the author of Die Theorie des Guten in Aristoteles’ Nikomachischer Ethik (De Gruyter, 2011). Timothy Chappell is Professor of Philosophy at the Open University, UK, and a Visiting Fellow in the Departments of Philosophy, St Andrews. His most recent books are Ethics and Experience: Life Beyond Ethical Theory (Acumen, 2009) and Knowing What To Do: Contemplation and Decision in Ethics (forthcoming). Lorenzo Greco is Marie Curie Fellow at the University of Oxford, Junior Research Fellow at Mansfield College, Oxford, and member of the Sapienza Università di Roma Research Unit on the British Enlightenment. In 2008 he published his first book-length study of Hume, L’io morale. David Hume e l’etica contemporanea (Liguori). John Hacker-Wright is Associate Professor of Philosophy at the University of Guelph, Guelph, Ontario, Canada. He works on contemporary virtue ethics, the metaphysics of value, and moral psychology. He is the author of Philippa Foot’s Moral Thought (forthcoming, Continuum Press) as well as ‘What is Natural about Philippa Foot’s Ethical Naturalism?’ (Ratio, 2009) and ‘Ethical Naturalism and the Constitution of Agency’ (The Journal of Value Inquiry, 2012). Christoph Halbig has held the Chair of Practical Philosophy at the University of Gießen since 2011; before that, he held the Chair of Practical Philosophy at the University of Jena from 2006–2010. In 2007 he received the ‘Award for outstanding research in the social sciences and the humanities’ from the Berlin-Brandenburg (former Prussian) Academy of Sciences and Humanities. His books include Objektives Denken. Erkenntnistheorie und Philosophy of Mind in Hegels System (Frommann-Holzboog, 2002) and Praktische Gründe und die Realität der Moral (Vittorio Klostermann, 2007).


List of Contributors

Edward Harcourt is Fellow & Tutor in Philosophy at Keble College, Oxford. His research interests include ethics (in particular moral psychology, neoAristotelianism and child development, ethical dimensions of psychoanalysis, meta-ethics, and Nietzsche’s ethics), literature and philosophy, and Wittgenstein, areas in which he has published a number of articles in journals and collections. Thomas Hurka holds the Chancellor Henry N. R. Jackman Distinguished Chair in Philosophical Studies at the University of Toronto. He is the author of Perfectionism (Oxford University Press, 1993), Virtue, Vice, and Value (Oxford University Press, 2001), and The Best Things in Life (Oxford University Press, 2011), as well as of numerous articles in moral and political philosophy, some of which are collected in Drawing Morals (Oxford University Press, 2011). Mark LeBar is Associate Professor at Ohio University. His research interests are in moral and political philosophy. His book The Value of Living Well is forthcoming from Oxford University Press. Erasmus Mayr is Junior Research Fellow in Philosophy at The Queen’s College, Oxford, and works mainly in philosophy of action and ethics. He is the author of Understanding Human Agency (Oxford University Press, 2011). Julia Peters is Lecturer in Philosophy at the University of Bonn. Her research interests include the philosophy of Hegel, aesthetics, and ethics, in particular contemporary virtue ethics. She has published several articles on the philosophy of Hegel and on aesthetics (Bulletin of the Hegel Society of Great Britain, 2009; Inquiry, 2010; European Journal of Philosophy, 2011). Anthony Price is Professor of Philosophy at Birkbeck College, University of London. He works on ancient and contemporary ethics and moral psychology. His main publications are Love and Friendship in Plato and Aristotle (Clarendon Press, extended edn. 1997), Mental Conflict (Routledge, 1995), Contextuality in Practical Reason (Clarendon Press, 2008), and Virtue and Reason in Plato and Aristotle (Clarendon Press, 2011). Daniel Russell is Professor of Philosophy at the Center for the Philosophy of Freedom, University of Arizona. His books include Plato on Pleasure and the Good Life (Oxford University Press, 2005), Practical Intelligence and the Virtues (Oxford University Press, 2009), Happiness for Humans (Oxford University Press, 2012), and Cambridge Companion to Virtue Ethics (editor, Cambridge University Press, 2012). Nancy Snow is Professor of Philosophy at Marquette University in Milwaukee, Wisconsin. Her research areas are virtue ethics and moral psychology.

List of Contributors 227 In addition to articles on virtue, she has published Virtue as Social Intelligence: An Empirically Grounded Theory (Routledge, 2010), is currently working on a book on hope, and is editing and co-editing forthcoming anthologies on virtue cultivation and virtue and psychology, respectively. Christine Swanton teaches in the Philosophy Department University of Auckland, New Zealand. She is currently working on the virtue ethics of Hume and Nietzsche. Her book on virtue ethics, Virtue Ethics: A Pluralistic View was published with Oxford University Press 2003, paperback 2005. Candace Vogler is the David B. and Clara E. Stern Professor of Philosophy and Professor in the College at the University of Chicago. Her research interests are in practical philosophy, practical reason, Kant’s ethics, Marx, and neoAristotelian naturalism. She is the author of Reasonably Vicious (Harvard University Press, 2002).


A achievement 45 acting badly 33–4 acting well see Aristotelian motivations for virtuous actions; Beautiful, Fine, or Noble acts action guidance 98 actions 86–90 action-situations 196–8 action-types 196 acts for their own sake 17–20, 30, 118 adaptations 122–4, 161 Adult Attachment Interview 119 Aesthetic, The (category) 166 affordances 141–2 agency 86; and the rationality of the four ends (Hursthouse) 90–2 agent-neutrality problem (in Perfectionism) 54, 56–9 agent-relative goods 54–9 agonistic Greek ethos view (Aristotle) 23 Ainsworth, Mary (attachment theory) 117; Strange Situation 118–19 Alcibiades 147, 152 Aletheia, four-fold of (Heidegger) 4, 183–91; see also Humean/ Heideggerian response dependence theory Alphone-and-Gaston routine 17 Anglophone philosophy 145–56 animal nature 100 Anscombe, Elizabeth 1, 9, 15, 87, 110, 196, 197 appraisals 132–9; see also knowledge and appraisal personality architecture (KAPA) Aquinas, Thomas 145, 148, 149, 154, 167 Aristotelian motivations for virtuous actions 27–36; complexity of 35; egocentricity and 33–4;

eudaimonia and 30–2, 35; external goals 27–30; individual agency and 34–5; instrumental 31; internal goals 29–30; for its own sake 30; kalon and 31; levels of justification 35; noble acts 32–3; obstacles to achieving goals 28; risking life for a good cause 31–3; self-concern and (the tremor) 34–5; see also virtue and happiness connection; virtuous sacrifice and personal loss; wellbeing and eudaimonia Aristotelian necessity 116 Aristotelian triangle (human nature, human excellence, human wellbeing) 114 Aristotle: De Anima (DA) 28; De Motu Animalium (DMA) 30; main catalogue of virtues and vices 13, 15; Nicomachean Ethics (NE) 11–21, 27–35, 69; Rhetoric 14, 21; see also Aristotelian motivations for virtuous actions; criticism of the Aristotelian approach to virtue; eudaimonia aschematic traits 136–7 assiduity 187 attachment simpliciter 122 attachment theory 4, 114–29, 146; adaptations 123–4; background of 117–18; backward-facing connection 122–4; claims 117; cultural relativity and 124–5; diachronic dimension 119; forward-facing connection 120–2; hypothesis 117; insecure attachment 118–21, 123–6; natural selection and 115, 116, 118, 122–4; Strange Situation



(Ainsworth) 118–19; synchronic dimension 119; type A: insecureavoidant 118–19, 123, 124–5; type B: secure 118–26; type C: insecure-ambivalent 119, 124–5; virtues and 121–6 authoritative moral sense 190 automaticity 140–1 autonomy, internalist welfare and 62–4

B backward-facing connection 122–4; see also attachment theory Baier, Annette 190 Beautiful, Fine, or Noble acts (kalos, kalon, kalou heneka) 4, 17, 20–3, 28–31, 158–73; evolutionary explanation 160–1; explanatory reduction of The Moral to the Prudent 159–60; grace and 170–1; Moral, The 158–68, 171–2; Philippa Foot (Letter Writers) and 168–9; practical reasoning 158–66, 169, 171–2; Prudential, The 158–66; 171–2; quietest dualistic presumption 159; Reasons To Do With Beauty (The Aesthetic) 166; skeptical dualism 159; vs. avoidance of something base 78; see also courageous soldier example; virtuous sacrifice and personal loss behavioral expressions 117 Being and Time (Heidegger) 181 Being There (film) 65 beliefs 88 benevolence 83, 219 best-agent relativity 58 Bonum est multiplex 37 Bowlby, John (attachment theory) 117 Bradley, Denis 150 brainwashing 62–4 bravery 27–8 Brentano, Franz 9 Broadie, Sarah 150 Brüllmann, Philipp 3–4, 97–113; see also Neo-Aristotelian Naturalism (NAN) Buddha 147, 148, 152

C calm passions 211–12 CAPS (cognitive-affective personality system) traits 131

caregiving system 117; see also attachment theory Cassidy-Marvin system 119 categorical imperative 145, 196–202; see also Humean virtue ethics causal reasoning 181 Cervone, Daniel, knowledge and appraisal personality architecture (KAPA) 131–9 change, internalist welfare and 61–2 Chappell, Timothy 4, 158–73; “Glory as an Ethical Idea” 165; see also Beautiful, Fine, or Noble acts character traits 4, 12–13, 83, 98–101, 108, 110; attachment theory and 115; Kantian 203–7; as virtues 187; see also attachment theory; Humean virtue ethics child development see attachment theory childhood training and education 11; see also natural virtue and proper upbringing children, welfare externalism and wellbeing of 65–6 chronic priming 141 circumstances 74–9; see also virtuous sacrifice and personal loss claritas 167 clinging 117, 119 cognitive psychology 2 cognitivist meta-ethical naturalism 116 Cohon, Rachel, on Hume 214 colloquial grace 170, 171 compassion 10, 11 conscientiousness 10 conscious processing 130, 140, 142 consequentialism 9 considered judgments 110 constitutively valuable 31 continent agents, vs. virtuous agents 72–3 continuity thesis 115, 116, 120–4 continuous self-improvement (Hume) 220 contractualism 168 Copp, David 84, 91 corrective virtue 148–50, 153–4 courage 14–15, 17, 47, 83, 153, 205; of Socrates 27–8, 33–4 courageous soldier example, of virtuous sacrifice and personal loss 69, 71–6 cowardly acts 71, 72

Index criticism of the Aristotelian approach to virtue 2–3, 9–26; choosing acts as kalon (noble, fine, beautiful) and 20–3; choosing acts for their own sakes and 17–20; dispositions vs. occurrent states and 12–13; doctrine of the mean and 13–14; egoistic motivation and 16–17; explanatory egoism and 14–16; praise and blame and 11–12; virtue as a higher-level good 9–12 Critique of Practical Reason (Kant) 198–9 Critique of Pure Reasons (Kant) 200 cross-situational personality coherence 136–7 cruelty 13 Crumb (film), internalist welfare and 59–61 Crumb, Robert 59–61, 65 cultural relativity 124–5

D Dancy, Jonathan 43 De Anima (Aristotle) 28 deduction 181 default value, prudential 43–4 deficiencies 3, 13–14 deliberating agents 42, 46–8, 57, 138 demonstration 181 De Motu Animalium (Aristotle) 30 deontology 9, 10, 18, 98, 195 depth, dimension of 44–5 developmental psychology 114, 116; see also attachment theory dexterity 187 disclosure 181, 183–91; see also Humean/Heideggerian response dependence theory dispositions vs. occurrent states 12–13 division of labor, attachment theory and 125 doctrine of the mean (Aristotle) 3, 13–14, 23, 146, 150–2 “Does Moral Philosophy Rest upon a Mistake?” (Prichard) 159–60 dog analogy, internalist welfare and 61, 63 domain-linked characteristics 136 Doris, John M. 130, 139–42 downgrading virtues 195 dual process theory of cognitive functioning 130, 139–42


E egoism, Aristotle and 3, 14–17, 23, 33–4 emotional fulfillment 52, 62, 65–6 emotional nature 52 emotional responses 188–9 emotion self-regulation 120 empirical psychology studies 4, 136–42 empiricistic naturalism 182 empty formalism criticism of Kant (Hegel) 196 Enquiries Concerning the Principles of Morals (Hume) 177, 185 enterprise 187 envy 10 ergon (human functioning) 66 ethical non-cognitivism 103 ethical non-naturalists 38 eudaimonia 2, 3, 99, 169, 216; background of 14–16, 22, 24; see also Aristotelian motivations for virtuous actions; criticism of the Aristotelian approach to virtue; virtue and happiness connection; virtuous sacrifice and personal loss; well-being and eudaimonia evil: good/bad asymmetry thesis 41–2; higher-level moral concept and 9–12 evolutionary explanation 160–1 excesses 3, 13–14 explanatory reduction of The Moral to The Prudential 159–60 explanatory egoism 14–16 external benefits 41, 43 external goals 28 externalism see welfare externalism; well-being and eudaimonia external values 15–16, 23–4, 27–30; see also welfare externalism

F Faculty of Judgment (Kant) 199–200; fate 74–9; see also virtuous sacrifice and personal loss flourishing 1, 37, 52, 202, 216 Foot, Philippa 70, 83–5, 99, 124–5, 156, 159; on corrective virtue 148, 153–4; “Goodness and Choice” 103; on Hume 214; on the LetterWriters 168–9; Natural Goodness 115–17; Neo-Aristotelian Naturalism (NAN) and 99, 101, 110; see also naturalism; NeoAristotelian Naturalism (NAN)



formal constraints 57–8 forward-facing connection 120–2; see also attachment theory four ends, agency and the rationality of the (Hursthouse) 84, 90–4 four-fold of aletheia 183–91; see also Humean/Heideggerian response dependence theory Frankena, William 164 freedom from pain and experience of species-characteristic pleasures (of the four ends; Hursthouse) 84 Freud, Sigmund, attachment theory and 117–18 friendship, virtue and 17, 41–2, 153 frugality 187 Funeral Oration (Pericles) 166–7

G generalized traits 136 general principles of action (maxims; Kant) 196 genetics 160, 161; attachment theory and 118 global traits 130–1, 139 “Glory as an Ethical Idea” (Chappell) 165 goal-dependent automaticity 140 good, higher-level moral concept and 9–12 good as logically attributive 102–4 good/bad asymmetry thesis 41–2 “good deeds” 55–6 good functioning of the social group (of the four ends; Hursthouse) 84 good life 57; see also eudaimonia good life simpliciter 47, 54 “Goodness and Choice” (Foot) 103 goods and reasons relationship claim 71, 73 goodwill 19 Gorgias (Socrates) 166, 168 Gottlieb, Paula 148 grace 170, 171 greater good 28 greatness of mind (Hume) 218–19 Greco, Lorenzo 4, 210–23; see also Humean virtue ethics grounding moral judgments 97–8, 109–10, 142, 158 Groundwork (Kant) 197–9 Groundwork/Critique internal strategy (Kant) 198–9

H habituation 114, 134, 140, 145–6

Hacker-Wright, John 3, 83–96; see also naturalism Halbig, Christoph 3, 37–51; see also virtue and happiness connection happy life: descriptive psychology and 52; other-regarding virtues and 38; see also eudaimonia; flourishing; virtue and happiness connection Harcourt, Edward 4, 114–29; see also attachment theory Harman, Gilbert 130, 139–42 has-accomplished-all metaphor (Philipps) 42 Haybron, Daniel 3, 52–68; Pursuit of Unhappiness, The 52; see also welfare internalism; well-being and eudaimonia Hegel, G. W. F. empty formalism criticism of Kant 196 Heidegger, Martin 4; Being and Time 181; four-fold of aletheia 183–91; see also Humean/Heideggerian response dependence theory Herman, Barbara 196, 198 hexis (state of character) 12–13, 56 higher-level good, virtue as a 9–12, 16–17, 19, 20, 22–4; see also criticism of the Aristotelian approach to ethics Highest Good (Kant) 202–3 Homeric conception 23 honesty 83 Honnecourt, Villars de 166, 169 horizons of disclosure 4; see also Humean/Heideggerian response dependence theory human development 114; see also attachment theory human excellence 114, 218–19 human functioning (ergon) 66 human nature 114; Neo-Aristotelian Naturalism (NAN) and 97; practical conception of 85–7 human selfhood 64 human sensibility 189 human social life 116–17 human well-being 63, 114; see also eudaimonia; well-being and eudaimonia Humean/Heideggerian response dependence theory 4, 177–94; aletheia (Heidegger): 1. openness 183, 184–6, 189; aletheia (Heidegger): 2. uncoveredness

Index 183, 186–7, 189–90; aletheia (Heidegger): 3. unhiddenness 183, 187–8, 190; aletheia (Heidegger): 4. unconcealment 183, 188–9, 190; empiricistic naturalism (McDowell) and 182; Hume and the limitations of “reason” 180–2; Intelligibility Thesis 178–9, 180; Qualified Agent Thesis 179–80; rationalistic metaphysics and 181–2; Relational Thesis 179, 180 Humean virtue ethics 210–23; benevolence and 219; calm passions 211–12; character traits and 211–20; compared to neoAristotelianism 210–20; human excellence (greatness of mind) 218–19; individuality and 219–20; limitations of reason 180–2; moral sentimentalism 178, 212–20; in pluralistic terms 218–19; proud person as virtuous person 218, 220; response-dependent theory and 213–14; see also Humean/ Heideggerian response dependence theory Hume, David 12, 29, 159; Enquiries Concerning the Principles of Morals 177, 185; Of the Standard of Taste 215–16; Treatise 181, 188; see also Humean/ Heideggerian response dependence theory; Humean virtue ethics Hurka, Thomas 2–3, 9–26, 169; Virtue, Vice and Value 2; see also criticism of the Aristotelian approach to virtue Hursthouse, Rosalind 3, 78, 83–94, 159; four ends, agency and the rationality of the 84, 90–2; Humean virtue ethics and 210, 214–16, 217; On Virtue Ethics 98; v-thoughts 162; see also naturalism; Neo-Aristotelian Naturalism (NAN)

I impersonal benevolence 85, 92–3 impure ethics strategy (Kant) 199 incongruency 140 individualism 3, 34–5, 210, 219–20 individual survival (of the four ends; Hursthouse) 84


inner division 73 insecure ambivalent attachment 119, 123, 124–5 insecure attachment 118–26 insecure-avoidant attachment 118–19, 123, 124–5 inside-out approach (Aristotelian) 23, 24 intrinsic good 9–11 instrumentally good acts 21 instrumentally valuable 31 intelligibility of practical reasons 158, 172 Intelligibility Thesis 178–9, 180 intentional acts 86–7 internal goals 29–30 internalism 38; see also well-being and eudaimonia intrinsic evil 9–11 intrinsic value 31 intuition 73–4, 78–9, 181

J Judas Iscariot 171 justice 83

K kakodaimonia (opposite of eudaimonia) 35 kalon/ kalos/ kalon heneka see Beautiful, Noble, or Fine acts Kantian good will 10 Kantian moral theory 1, 195–209; Categorical Imperative (CI) test 196–202; character traits for a virtuous life 203–7; Faculty of Judgment 199–200; flourishing and 202; Groundwork/Critique of Practical Reason internal strategy 198–9; Highest Good 202–3; impure ethics strategy 199; maxims or general principles of action 196–8; Metaphysics of Morals (MM) 195, 198–200, 203; moral salience 197–207; virtue defined, 195 Kant, Immanuel 9, 20, 145, 159, 164; Groundwork/Critique of Practical Reason internal strategy 198–9; Metaphysics of Morals 195, 199–200, 203; see also Kantian moral theory KAPA see knowledge and appraisal personality architecture kindness 134–6



knowledge 45 knowledge structures see knowledge and appraisal personality architecture (KAPA; Cervone) knowledge and appraisal personality architecture (KAPA; Cervone) 131–9; aschematic traits 136–7; knowledge structures 133–9; self-efficacy appraisals 134–9; self-schema appraisals 133–9; virtue cultivation 134–5; see also personality scaffolding of virtue; situationism Korsgaard, Christine 159

L Lapsley, Daniel K. 131, 140–2 LeBar, Mark 3, 52–68; see also wellbeing and eudaimonia Letter-Writers 168–9 logos 28, 107 loss, measuring amount of 47–8; see also virtuous sacrifice and personal loss lying 10

M mad tyrant example 78–9 Main, Mary (attachment theory) 117 malice 13 manipulation 62–4 many-dimensional account 151–2, 154 martyrs 169 maxims (Kant) 196 Mayr, Erasmus 4, 195–209; see also Kantian moral theory McDowell, John 3, 70–9, 177; empiricistic naturalism 182; Neurathian procedure 101–2, 106–9; no-loss-at-all thesis 39–44; Virtue and Reason 200–201, 206; see also virtuous sacrifice and personal loss megalophychos (proud person) 16–17, 22, 23 Merritt, Maria W. 130, 139–42 metaphysics 4; of engagement 181; rationalistic 181–2; see also Humean-Heideggerian response dependence theory; Kantian moral theory Metaphysics of Morals (Kant) 195, 198–9, 203 method of ethics 9 middle station of life 220 Mill, John Stuart 9, 12

modern mirroring theory 188 Moore, G. E. 9, 24; “Open Question Argument” 103; Principia Ethica 159–60 moral agents 72 moral blindness 212 moral dissociation 139–40 moral evaluation/judgments 104–6 moralized conception of happiness 40–4; see also virtue and happiness connection moral knowledge 45 moral salience (Kant) 197–207 moral schemas 140; see also selfschema appraisals moral sentimentalism (Hume) 178, 180–2, 184, 212–20 moral skepticism 181 Moral, The (category) 158–67, 171–2 Mother Teresa of Calcutta 167 motivation 3, 11; egoistic 16–17; theory of 38; see also Aristotelian motivations for virtuous actions Muggeridge, Malcolm, Something Beautiful for God 167 Müller, Anselm 146–56

N Narvaez, Darcia 131, 140–2 natural defects 116, 124, 126 Natural Goodness (Foot) 115–17 natural human perfection 1; see also Perfectionism naturalism 1–4, 83–96, 177; four ends of (Hursthouse) 84, 90–2; impersonal benevolence and 92–3; practical conception of human nature 85–7; rationality for human beings 87–90; virtue and happiness connection and 37–8; see also attachment theory; Foot, Philippa; Hursthouse, Rosalind; NeoAristotelian Naturalism (NAN) natural science 114–15 natural selection 115, 116, 118, 122–4 natural virtue and proper upbringing 4, 11, 145–57; Anselm Müller and 146–56 ; defined 146–7; the doctrine of the mean 150–2; example of 154–5; natural vs. ethical virtue 152–3; Paula Gottlieb and 148; Philippa Foot and 148–9, 153–4; practical wisdom and 145, 149–51,

Index 155–6; reciprocity of the virtues thesis 148–9 nature-fulfillment 53 Neisser, Ulrich 141 Neo-Aristotelian Naturalism (NAN) 2, 97–113; character traits and 98–101, 108, 110; claims 104–5; constraints of 97, 110; defined 97; difference between human nature and animal nature 100; good as logically attributive 102–4; grounding moral judgments and 97–8, 109–10; Neurathian procedure and 101–2, 106–9; point of 106; specifications of 101; theses 104–5; see also Hursthouse, Rosalind; naturalism neo-Aristotelian virtue ethics 1; compared to Humean virtue ethics 210–20; see also Aristotelian motivations for virtuous actions; criticism of the Aristotelian approach to virtue; eudaimonia; Neo-Aristotelian Naturalism (NAN) Neurathian procedure (McDowell) 101–2, 106–9 neutral goods 42 neutrality see agent-neutrality problem (in Perfectionism) new metaphysics for virtue ethics see Humean-Heideggerian response dependence theory Nicomachean Ethics (Aristotle) 11–21, 27–35, 69 Nietzsche, Frederick 37 noble acts see Beautiful, Fine or Noble acts no-loss-at-all thesis 39–44; see also McDowell, John; virtuous sacrifice and personal loss nonconscious processing 130, 131, 139–42 Nozick, Robert 35

O occurrent states vs. dispositions 12–13 Of the Standard of Taste (Hume) 215–16 one-dimensional account 151–2, 153, 156 On Virtue Ethics (Hursthouse) 98 openness (aletheia; Heidegger) 183, 184–6 “Open Question Argument” (Moore) 103 opportunity costs 54–6


optimal development 116; see also attachment theory outside-in approach (higher-level good) 9–12, 23–4

P pain: Aristotelian view and 18; higherlevel moral concept and 11, 12, 19 Parfit, Derek 48 Perfectionism 52–9, 66, 220; agentneutrality problem 54, 56–9; under-specification problem 53–6 Pericles 169; Funeral Oration, 166–7 personality scaffolding of virtue 130– 44; empirical psychology studies 136–9; knowledge and appraisal personality architecture (KAPA; Cervone) 131–9; situationism 130–1; virtue development: a response to Merritt, Doris, and Harman 139–42; see also knowledge and appraisal personality architecture (KAPA) personal loss see virtuous sacrifice and personal loss personal responsibility 33–4; see also Aristotelian motivations for virtuous actions Peters, Julia 3, 69–82; see also virtuous sacrifice and personal loss philautos (self-lover) 17, 22 Phillips, D. Z. 40–2 phrōnesis 149–50 phronimos (practical wisdom) 150, 211, 216–19 Plato 37, 166; Republic 20, 159, 164; Symposium 27, 167; tripartite moral psychology 122 Platonists 57 pleasure: good and evil and 9–10; virtue producing 189–90 pluralism 172 post-conscious automaticity 141 practical identities 87–8 practical morality 178 practical reasoning 28, 57–9, 63–4, 85–7, 158–66, 169, 171–2 practical wisdom 29, 145, 149–51, 155–6, 216–17; see also phronimos praise and blame (Aristotle) 11–12 Preschool Assessment of Attachment 119 Price, Anthony 3, 27–36; see also Aristotelian motivations for virtuous actions



Prichard, Harold 15; “Does Moral Philosophy Rest upon a Mistake?” 159–60 pride 16–17, 218, 220 Principia Ethica (Moore) 159–60 proper upbringing see natural virtue and proper upbringing proportionality 10–11, 15, 17, 18 prospective moral decision-making 196–7 proud person (megalopsychos) 16–17 prudence 187 prudentia 149–50 prudential disvalue 42–3 prudential evaluative valence 46 prudential goods/value 38–48; default values 43–4; see also virtue and happiness connection prudentially neutral 42 Prudential, The (category) 158–66, 171–2 psychic affirmation 52 psychic flourishing 52 psychobiology 4, 116; see also attachment theory psychological dispositions 115 public utility 171 Pursuit of Unhappiness, The (Haybron) 52

Q Qualified Agent Thesis 179–80 quietist dualistic presumption (category) 159–60 Quine, W.V.O. 106–7

R Rashdall, Hastings 9, 24 rashness 30, 47 rationalism 38, 130; agency and the rationality of the four ends (Hursthouse) 90–2; naturalism and impersonal benevolence 84, 92–3; practical conception of human nature 85–7; rationality for human beings 87–90; see also naturalism rationalistic metaphysics 181–2 Rawls, John 195 reason, limitations of Hume and 180–2; silencing of 71–3, 75–6, 78; see also practical reasoning; practical wisdom Reasons To Do With Beauty (category) 166 reciprocity of the virtues 148–9, 152–6 recursive theory of virtue 3

reductionism 115 regret 76–9; see also virtuous sacrifice and personal loss Relational Thesis 179–80 Republic (Plato) 20, 159, 164 resistance 147–8 respect-for-humanity formula (Kant) 198–9 response dependence theory, virtue ethics and 178–80, 182, 213; see also Humean/Heideggerian response dependence theory retrospective assessment 196–7 Rhetoric (Aristotle) 14, 21 richer life, value of 64–5 risking one’s life for a good cause 31–3 robust traits 130 Ross, W. D. 9, 24 rogue individual 86 rule-centered approach 195; see also Kantian moral theory rule-consequentialism 168 Russell, Daniel 3, 52–68; see also wellbeing and eudaimonia

S sacrifices 3; happiness and 43, 48; of opportunities 54–5; of prudential values 43; unbeneficial 46–7; see also virtue and happiness connection; virtuous sacrifice and personal loss scientific investigation 177–8, 182 scientific truth 31 scope-fallacy 172 secondary moral concept 10–11 secondary qualities 212 secure attachment 118–26 self-agency 35 self-centeredness 2, 22–3, 44–5 self-conception 133–6 self-concern 34–5 self-correction 4 self-depravation 44 self-development 55–6, 59 self-directed virtues 46 self-effacing 16, 48 self-efficacy appraisals 134–9; see also knowledge and appraisal personality architecture (KAPA) self-esteem 218 self-fulfillment view 52, 62 self-governance 149 selfhood 64 self-improvement 220

Index self-indulgence 22, 47–8 self-knowledge 135 self-lover (philautos) 17 self-monitoring 135 self-possession 45 self-reflection 134 self-regulation 135 self-schema appraisals 133–9; see also knowledge and appraisal personality architecture (KAPA) self-sufficiency 32 sensibilist mode (Wiggins) 213 sentimentalism see moral sentimentalism (Hume) sentimentalist ethics of care (Slote) 213 Shadel, William G. 137–9 shallow life 44–5 Siddhartha Gautama 147 Sidgwick, Henry 20–1, 34, 159–60 silencing of reasons 71–3, 75–6, 78 Singer, Peter 93 single standard measurement 28–9 situationism 2, 4, 130–1; see also knowledge and appraisal personality architecture (KAPA); personality scaffolding of virtue situation-types 196 skeptical dualism (category) 159 Slote, Michael 45, 213, 215 Snow, Nancy 4, 130–44; Virtue as Social Intelligence, 139; see also knowledge and appraisal personality architecture (KAPA); personality scaffolding of virtue Sobel, David 84, 91 social-cognitivism 131 social contexts, virtues and 190 social psychology 2; see also personality scaffolding of virtue Socrates, Gorgias 166, 168 Socrates’s courage 27–8, 33–4 Something Beautiful for God (Muggeridge) 167 Sparta, example of insecure-avoidant attachment 124–5 species survival (of the four ends; Hursthouse) 84 stable disposition 12–13 state of character (hexis) 12–13, 56 Stoic optimism 46 Stoics 37, 57 Stoic thesis 39, 40 Strange Situation (Ainsworth) 118–19 Swanton, Christine 4, 177–94, 210, 213–14, 218; Virtue Ethics:


A Pluralistic View 2; see also Humean/Heideggerian response dependence theory Symposium (Plato), 27, 167

T tacit information processing 141 Taylor, Jackie 191 teleios or telos (eudaimonia; ultimate end) 32, 57–8, 210 teleological perspective 97, 104 temperance 187 temperance examples: of gossip 154–5; of virtuous sacrifice and personal loss 74–5 theological grace 170, 171 Theory of Virtue, A (Adams) 2 third-person perspective 48 Thompson, Michael 85–6 thought experiment 54 Thrasymachean cynicism 164 tragic dilemmas 46, 48 Treatise (Hume) 181, 188 tremor, the 34–5 tripartite moral psychology (Plato) 122 trivial acts 17 truth, contemplating 31

U ultimate end (teleios or telos; eudaimonia) 32, 57–8, 210 unconcealment (aletheia; Heidegger) 183–91; see also Humean/ Heideggerian response dependence theory uncoveredness (aletheia; Heidegger) 183, 186–7 under-specification problem (in Perfectionism) 53–6 unhappiness 38, 77 unhiddenness (aletheia; Heidegger) 183, 187–8 unity of character 210, 220; see also Humean virtue ethics Universalization Test (Kant) 197–8 utilitarianism 1, 16, 98, 195 utility 22, 171

V vices 3, 9, 13–14, 181–2; attachment theory and 121, 123–6; pursuit of 40 vicious person 11, 46; sympathy for 44–5 virtue, characteristics of 189–91 virtue and happiness connection 3, 37–51; background 37–9; good/ bad asymmetry thesis 41–2;



i. being overall virtuous is both necessary and sufficient for happiness 39, 40; ii. being overall virtuous is a necessary condition on other goods being of prudential value for their possessor 40–4; iii. being overall virtuous is itself a basic kind of prudential value 40, 44–5; iv. possessing some virtue is a necessary condition on other goods being of prudential value for their possessor 40, 45; v. being overall virtuous is a contingent, personal project whose successful realization is of prudential value for its possessor 40; measuring amount of loss 47; no-loss-at-all thesis 39–44; reasons for forging a conceptual link between 37–8; virtuous acts and prudential value 46–8; see also virtuous sacrifice and personal loss; well-being and eudaimonia Virtue and Reason (McDowell) 200–201, 206 Virtue as Social Intelligence (Snow) 139 virtue benefit to the possessor 19–20; see also virtue and happiness connection virtue cultivation 134–5, 141–2 Virtue Ethics: A Pluralistic View (Swanton) 2 virtues: attachment theory and 121–6; downgrading 195; single or bundles of, 45–6 Virtue, Vice and Value (Hurka) 2 virtuous agents: defined 98; vs. continent agents 72–3 virtuous sacrifice and personal loss 40–4, 69–82; circumstances/fate and 74–9; courageous soldier example 69, 71–6; eudaimonia and 76–9; intuition and 73–4, 78–9; John McDowell’s argument and reconciling it with regret of personal loss 70–9; link between virtue and happiness 69; mad tyrant example 78–9; moral agent types 72–3; noble causes vs. avoidance of something base 78; regret and 76–9; silencing of reasons 71–3,

75–6, 78; temperance example 74–5; ways in which the pursuit of a personal good can conflict with a requirement of virtue 74–6; see also sacrifices; virtue and happiness connection; wellbeing and eudaimonia Vogler, Candace 4, 145–57; see also natural virtue and proper upbringing voluntariness 11–12 v-thoughts (Hursthouse) 162

W Weil, Simone 166 welfare externalism 53; defense of 59–66; human well-being 66; wishing for children’s well-being 65–6; see also well-being and eudaimonia welfare internalism 59–66; see also Haybron, Daniel; well-being and eudaimonia well-being, as a normative notion 52; see also well-being and eudaimonia well-being and eudaimonia (reply to Daniel Haybron) 52–68; Perfectionism: agent-neutrality problem 54, 56–9; Perfectionism: under-specification problem 53–6; welfare internalism cannot explain the importance of autonomy 62–4; welfare internalism cannot explain the value of a richer life 64–5; welfare internalism can yield no reasons to change one’s make-up 61–2; welfare internalism is not the relevant kind of welfare for people 59–61; welfare internalism must treat all emotional fulfillments as equal 65–6; see also virtue and happiness connection; virtuous sacrifice and personal loss White, Terence Hanbury 158 Wiggenstein, Ludwig 88, 158 Wiggins, David 213–4 Wilson, Catherine 164–5 wisdom see practical reasoning; practical wisdom Wollheim, Richard 34–5 worldhood (Heidegger) 183–4 Wrathall, Mark A. 183