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 dis-
cuss 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
2 Kant, Duty and Moral Worth
Philip Stratton-Lake
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
12 Challenging Moral Particularism
Edited by Mark Norris Lance,
Matjaž Potrč, and Vojko
Strahovnik
13 Rationality and Moral Theory
How Intimacy Generates Reasons
Diane Jeske
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

R~~J!~~~~"P
NEW YORK AND LONDON
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
Contents

Introduction: Aristotelian Ethics in Contemporary

Perspective 1
JULIA PETERS

PART I
Themes in (Neo-)Aristotelian Virtue Ethics

1 Aristotle on Virtue: Wrong, Wrong, and Wrong 9

THOMAS HURKA

2 Aristotle on Virtue: A Response to Hurka 27

ANTHONY PRICE

3 The Benefit of Virtue 37

CHRISTOPH HALBIG

4 Well-Being and Eudaimonia: A Reply to Haybron 52

MARK LEBAR AND DANIEL RUSSELL

5 Virtue, Personal Good, and the Silencing of Reasons 69

JULIA PETERS

6 Human Nature, Virtue, and Rationality 83

JOHN HACKER-WRIGHT

7 Good (as) Human Beings 97

PHILIPP BRÜLLMANN
vi Contents
8 Attachment Theory, Character, and Naturalism 114
EDWARD HARCOURT

9 Notes Toward an Empirical Psychology of Virtue: Exploring

the Personality Scaffolding of Virtue 130
NANCY SNOW

10 Natural Virtue and Proper Upbringing 145

CANDACE VOGLER

11 Kalou Heneka 158

TIMOTHY CHAPPELL

PART II
Beyond (Neo-)Aristotelian Virtue Ethics

12 A New Metaphysics for Virtue Ethics: Hume Meets

Heidegger 177
CHRISTINE SWANTON

13 A Kantian Plea for Virtues? 195

ERASMUS MAYR

14 Toward a Humean Virtue Ethics 210

LORENZO GRECO

List of Contributors 225

Index 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 Aristo-
tle 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 pri-
marily 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 develop-
ment 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 interpre-
tation 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 objec-
tive 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 flour-
ish 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 asso-
ciated conception of acting virtuously or acting ‘from’ virtue, play a crucial if
not fundamental role in moral philosophy. However, they also tend to interpret
2 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 assump-
tions that define the neo-Aristotelian virtue ethical school of thought today.
However, after this ‘golden age’ of neo-Aristotelian virtue ethics, its dis-
cussion 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, undis-
puted 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 neo-
Aristotelian 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 cli-
mate, moreover, several monographs have emerged that present powerful
alternatives to the neo-Aristotelian approach to virtue: for instance, Chris-
tine 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 cen-
tral 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
 Introduction 3
to virtue. Hurka challenges the Aristotelian tendency to blur the distinction
between the good and the right by making the virtues, which are constitu-
tive 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 chal-
lenges 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. Hal-
big looks at different versions of the claim represented in the contemporary
neo-Aristotelian literature and reaches the conclusion that the most plau-
sible 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 Aris-
totelian 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 Silenc-
ing of Reasons’, Julia Peters suggests a way of reconciling McDowell’s claim
that the virtuous never have to make genuine sacrifices when acting virtu-
ously 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 estab-
lishes the fact that humans are essentially rational agents, but also can func-
tion 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 Rosa-
lind Hursthouse—implies a problematic tension because it puts a conceptual
constraint on our conception of the moral good that is incompatible with the
4 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 Natural-
ism’ 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, Har-
court 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: Explor-
ing 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 correc-
tive accounts of the virtues. Proper upbringing, Vogler argues, facilitates the
development of natural virtue, which is a genuine, yet inchoate form of vir-
tue, 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 explor-
ing 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.
 Introduction 5
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’; Hurst-
house, On Virtue Ethics.
3. See for instance Hurka, Virtue, Vice and Value; Swanton, Virtue Ethics: A Plu-
ralistic 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
1 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 Aris-
totle 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 con-
trast 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 conse-
quentialist concepts of good and evil. One of their claims was that if some-
thing 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 intrin-
sically 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;
10 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 ori-
entation 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 atti-
tude 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. Some-
thing 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 plea-
sure 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 11
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 tor-
turer’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 instru-
mentally but in its own right. He also thinks virtue is a matter of your atti-
tudes 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 argu-
ments 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 virtu-
ously, 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, presup-
pose 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 pro-
duced it, and pain evil even when it’s no one’s fault. The higher-level account
12 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 oth-
ers’ 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 ascrip-
tions 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 particu-
lar 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 virtu-
ous character as one that tends to produce them?
The higher-level account takes the second view, ascribing virtue proper-
ties 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 char-
acter’ (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 pos-
session 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, indepen-
dently 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 commit-
tee deciding whether to give a medal for bravery to a soldier who threw
 Aristotle on Virtue 13
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. Cour-
age 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 cen-
tral 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
14 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 virtu-
ous (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 dispro-
portion’ 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 eudai-
monia 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 exer-
cise of virtue, which consists in part in acts expressing moral virtues such as
courage and liberality. Our ultimate reason to perform these acts is there-
fore 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 15
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 propor-
tionally 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 support-
ing 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 higher-
level account explains.
This points to the central flaw in Aristotle’s account of virtue: its underly-
ing 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 expla-
nation 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 funda-
mentally 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
16 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 ego-

istic 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 ‘self-
effacing’ because it tells people not to believe or be guided by its own foun-
dational 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 utili-
tarianism 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, espe-
cially 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 17
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 some-
what nakedly the self-absorption which is the bad side of Aristotle’s eth-
ics’.14 That self-absorption appears often in the NE.
In his discussion of courage Aristotle says ‘the end of every activity is con-
formity to the corresponding state of character’ (1115b20–21), as if a coura-
geous 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 oth-
ers’ 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 com-
petitive 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’
18 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 virtu-
ously 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 prop-
erties 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 distinc-
tive 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 citi-
zens’ (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 con-
sistent 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 identi-
fies 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 activ-
ity 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 inde-
pendent 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
20 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 vir-
tuous 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 con-
tributor 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 argu-
ment seems to value friends, like money, primarily as means to one’s own
exercise of virtue. A notorious argument in the Politics criticizes the pro-
posal 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 addi-
tional 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 with-
out 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 choos-
ing an act for its beauty not again choosing it for an inappropriately self-
centered 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 21
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 devot-
ing 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 con-
formity 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 virtu-
ous 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
22 Thomas Hurka
moral considerations such as gratitude and promise-keeping, it isn’t appro-
priate 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 sec-
ond, 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 self-
reflective, 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 moti-
vated 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 vir-
tuous 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 virtu-
ous 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 pri-
marily 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 unattrac-
tively 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 eudai-
monia. 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 23
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 virtu-
ous 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 uncon-
nected 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 self-
focused 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 per-
son’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 virtu-
ous motivation as likewise externally focused, so a virtuous person cares
most about his virtues’ objects, such as another’s pleasure or pain, and only
24 Thomas Hurka
secondarily about his own virtuous motives in pursuing them. The inside-
out 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 there-
fore 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 25
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 ‘self-
referentiality’ 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 Eth-
ics, 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 Aris-
totle, 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).
26 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 Acad-
emy 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 Aris-
totle 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 (Cam-
bridge: Cambridge University Press, 1981), 40–53.
2 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.

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 look-
ing 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 narra-
tive emphasizes) that of his companion, Laches. A different agent might risk
danger gratuitously in order to display his sang-froid; Socrates is a sensible
28 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 judg-
ment 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 impossibil-
ity, 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 neces-
sarily 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 infer-
ring 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 accord-
ing 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
30 Anthony Price
can be identified from case to case by reference to external goals, but in gen-
eral terms only by reference to internal goals. Thus Aristotle links the exer-
cise 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 cour-
age directs’ (III.7, 1115b20–24; cf. in general VI.5 1140a25–28).
In Aristotle’s view, therefore, virtues are not to be construed instrumen-
tally as dispositions that measurably maximize the good effects of action.
Rather, they are dispositions to act well in ways that can be roughly clas-
sified 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 Anima-
lium [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 chap-
ter 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 ethi-
cal 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 unlei-
surely 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 exam-
ple, military action can itself achieve nobility, how can it fail to be desir-
able 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 contempla-
tion above virtuous action is debatable; yet, so read, it is intelligible—and
confirms my understanding of Aristotle.
32 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 val-
ues 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 motiva-
tion 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 gener-
ously 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 side-
effect (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 self-
sufficiency: ‘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 informa-
tion he then possessed, he had decided and acted otherwise. I.7 must be
 Aristotle on Virtue 33
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 con-
tests’ (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). How-
ever, more needs to be said to explain what this unique value is that can pro-
vide 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 mani-
fested 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 Sala-
mis, 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
34 Anthony Price
government, powerful as it was, did not frighten me into any wrongdo-
ing. 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 35
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 unmedi-
ated. 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 eudai-
monism. 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 ele-
ments 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 rebut-
tal 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.
36 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 differ-
ent 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 chil-
dren 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 some-
thing 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).
3 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 virtu-
ous 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 vir-
tue 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). Vir-
tues 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 tradi-
tions3 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 pro-
vides 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
38 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 under-
stand 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 mat-
ters. Even if evidence came up that points to a considerable amount of covari-
ance 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 cru-
cial other-regarding virtues.

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 39
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 hap-
piness, 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 rela-
tively 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.
40 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 suc-
cessful 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 pru-
dential 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 can-
not 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 41
and so forth—according to which (for instance) no payoff from flouting a
requirement of excellence . . . can count as a genuine advantage; and, con-
versely, 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 coun-
terexamples: 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 pain-
ful, 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 exam-
ple: 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 friend-
ship 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 happi-
ness 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
42 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 steadi-
ness 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 her-
self 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 speak-
ing 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 43
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 situa-
tion 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 pun-
ish 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 circum-
stances, 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 pruden-
tial 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 axiologi-
cally 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 impor-
tant 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 conflict-
ing 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
44 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, par-
ents 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 vicious-
ness 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, sec-
ond, our reactions of sympathy to matters of virtue and vice are strongly
influenced by our moral attitudes that might confuse our prudential intu-
itions—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 funda-
mental 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 qual-
ity 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 45
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 consti-
tutive 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 pruden-
tial 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 perse-
verance. 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 vir-
tues.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 con-
nection 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 question-
able whether the virtues are really necessary for the prudential good as such:
virtues such as courage might prove instrumentally useful in acquiring knowl-
edge, but is the way it has been acquired essential to its status as prudential
46 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 connec-
tion. A fortune won in the lottery instead of having been amassed by relent-
less, perseverant, and courageous efforts of my own seems to enhance my
life considerably less even in prudential terms. In the same vein, being gener-
ous 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 coura-
geous) 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 per-
sonal relationships, achievement, or even knowledge. Being overall virtuous,
however, does not form in itself a basic category of prudential value. It con-
tributes 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, out-
ward appearances notwithstanding, does not benefit us at all. Such a posi-
tion 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 47
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 prob-
lem 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 per-
spective 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 coura-
geous 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 implau-
sible 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 some-
one 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
48 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 per-
son 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 perspec-
tive, 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 vir-
tue 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 dilem-
mas, 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 Selbstlo-
sigkeit’.
2. Cf. Hursthouse, ‘Virtue Theory and Abortion’, 226.
 The Benefit of Virtue 49
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 eudai-
monia/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 desir-
able 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 coura-
geous 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
50 Christoph Halbig
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.)
22. Cf. Dancy, Ethics Without Principles, 184–187.
23. 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.
24. 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.
25. For a discussion of the dimension of depth see Foot, Natural Goodness, 86ff.
26. 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.
27. Cf. Slote, ‘The Virtue in Self-Interest’, 276f.
28. 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.
29. Cf. Slote, ‘The Virtue in Self-Interest’, 277f., who holds that ‘knowledge con-
stitutes a distinctive form of personal good, and counts as wisdom, only when
it takes courage to acquire it’. Ibid., 278.
30. 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.
31. 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.
32. Cf. Parfit, Reasons and Persons, 24.
33. 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’.
34. The person and her character might of course be included in that situation.
35. 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 Aris-
totle’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 Vir-
tues, ed. T. Chappell (Oxford: Clarendon Press, 2006), 53–66.
4 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 con-
ception of happiness as a long-term, broadly positive emotional condition, a
‘stance of psychic affirmation’, which is ‘not merely a state of one’s conscious-
ness’ 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 Accord-
ingly, Haybron calls his account of well-being the ‘self-fulfillment’ view.
The connections that Haybron draws between self-fulfillment and well-
being 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 Aristote-
lian 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 Aristo-
telian 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 53
Second, Haybron’s self-fulfillment theory actually makes well-being wholly
dependent on individual make-up, and we find that implausible. While Hay-
bron agrees that well-being is ‘nature-fulfillment’, for Haybron this is the ful-
fillment 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 indi-
vidual’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 consid-
ered convictions about well-being than does Haybron’s recommended alterna-
tive. 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 well-
being 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 Hay-
bron 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
54 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 car-
ing 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 agent-
neutral, 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 well-
being 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 well-
being. 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 prob-
lems, 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 55
assume that one achieves the former by pursuing the latter.18 Since Perfec-
tionism 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 alter-
native 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 dif-
ferent 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 prob-
lem 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. Perfec-
tionism’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 virtu-
ous one. The idea seems to be that we can understand ‘virtue’ as ‘admira-
bility’, 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.
56 Mark LeBar and Daniel Russell
But that is not at all the sort of claim that Aristotle makes because Aristo-
tle 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 rec-
ognition (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 han-
dling 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, activi-
ties involved in developing a talent (say) and virtuous activity are not two
sets of activities. Rather, activities involved in developing talents are virtu-
ous activities when they are done with appropriate forms of practical rea-
soning 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 self-
development to be part of a good human life—including the kinds of self-
development 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, Aristo-
tle’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 57
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 virtu-
ous 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 well-
being 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 reason-
ing. 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 conversa-
tion about agent-relative goods. Aristotle observes that eudaimonia fig-
ures 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 agent-
relative 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. Aris-
totle 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 disagree-
ment 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
58 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 ulti-
mate 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 deliber-
ating 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 vir-
tuous activity is supposed to meet those constraints. But those questions are
not relevant to Haybron’s objections to the Aristotelian approach.23 What-
ever 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 Perfec-
tionism 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 Aristote-
lian eudaimonist that certainly does not settle the question of what Angela
has agent-relative reason to do with her life. Perfection of that sort there-
fore 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 59
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, Aristo-
telian 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 eudai-
monism 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 generaliza-
tion 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 nor-
mative 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
60 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 indi-
vidual 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 inca-
pable 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 comfort-
able, and leave it alone.
So internalism does capture one way of thinking about well-being—well-
being 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 well-
being. 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 capaci-
ties 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 coun-
tenance 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 make-
up—the make-up of Crumb, not Crumb*. On that view, therefore, there
62 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; internal-
ism 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 iden-
tify 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 well-
being. 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 reason-
giving, 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 neces-
sary 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 spe-
cifically, 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 63
Haybron’s endorsement of autonomy is sound, but his welfare internal-
ism 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 possibil-
ity of manipulation or brainwashing in the first place, and that possibility is
what gives autonomy its significance. The capacity for autonomy is there-
fore 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 thin-
ner 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 crea-
ture 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
64 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 make-
up 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 well-
being 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 hap-
piness ‘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 can-
not 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 65
The welfare internalist must understand a ‘richer’ activity as more com-
plex 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 fulfill-
ment 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 fulfill-
ment 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) well-
being 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 unnatu-
ral 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
66 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. How-
ever, 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 Aristo-
tle 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 well-
being 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 Perfection-
ism, 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 67
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 argu-
ments 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 excel-
lence 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 delibera-
tion 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).
68 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://
plato.stanford.edu/archives/fall2008/entries/reasons-agent/.
Sumner, L. W., Welfare, Happiness, and Ethics (New York: Oxford University Press,
1996).
5 Virtue, Personal Good, and
the Silencing of Reasons
Julia Peters

One of the most intriguing aspects of Aristotelian and neo-Aristotelian vir-

tue ethics is that it promises to show that there is a link between the virtu-
ous life and happiness or eudaimonia, or that virtue benefits its possessor.
However, it is surprisingly difficult to find, both in Aristotle and in the neo-
Aristotelians, 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 posses-
sion 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 pur-
suing 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
70 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 fol-
lows 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 sac-
rifice—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 well-
being 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 71
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 pur-
sue 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
72 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 happi-
ness. 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 per-
son’s point of view, is the following one. Unless we assume that for the vir-
tuous 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 sol-
dier 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 per-
son 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 abandon-
ing 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 73
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 con-
tinent. 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 differ-
ence in practical deliberation between the two types of moral agent: the con-
tinent 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 cir-
cumstances, 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 prefer-
ence 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
74 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 nev-
ertheless 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 sugges-
tion 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 establish-
ing 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 what-
soever 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 consid-
ered 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. Accord-
ingly, 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 temper-
ate 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 circum-
stances, 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 reconcil-
ing 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 suf-
fering 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 circum-
stances 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 cir-
cumstances 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
76 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 virtu-
ous 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 ques-
tion 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 circum-
stances 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 gran-
diosely, 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 sacri-
fice 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 77
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 act-
ing 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 hav-
ing 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 per-
son, 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 virtu-
ous 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 signifi-
cant to her than that of any personal good that no ‘sacrifice’ of such a good
78 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, admi-
rable 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, Hurst-
house 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 con-
cede 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, presum-
ably, 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 per-
sonal 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 79
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 oppor-
tunity 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 neverthe-
less 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 virtu-
ous 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 suffer-
ing 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, temper-

ance, 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: exter-
nal 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.
80 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 adher-
ence 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 pos-
sess 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 occasion-
ally 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 nec-
essary 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 some-
how 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 pos-
session. 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 genu-
ine 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 situ-
ation, 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 81
than using it for herself. Slote’s description of the counterfactual situation
does not, therefore, by itself undermine McDowell’s conception of the essen-
tial connection between goods and reasons. In the following, I shall accept
both P1 and P2 of McDowell’s argument as reconstructed previously.
13. 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).
14. McDowell himself, misleadingly, uses the formulation ‘no loss at all’ in ‘The
Role of Eudaimonia’, 370.
15. Hursthouse, On Virtue Ethics, 180–183.
16. Hursthouse, On Virtue Ethics, 183.
17. McDowell, ‘The Role of Eudaimonia’, 369.
18. In chapter 3 of this volume, Christoph Halbig likewise criticizes McDowell
for maintaining that in forgoing personal goods by acting virtuously, the vir-
tuous 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 nor-
mative impact on the deliberating agent’ (Halbig, ‘The Benefit of Virtue’), this
volume, 42. However, for McDowell, the silencing of reasons to avoid some-
thing 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 cru-
cial 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.
19. 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 Univer-

sity 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).
82 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).
6 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 liv-
ing 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 paral-
lel 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
84 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 explic-
itly 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 nat-
uralism 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 com-
ponent 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 rational-
ity 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 obedi-
ence 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 natu-
ralists. 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, per-
haps 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 reason-
ing 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 eval-
uation 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
86 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 sup-
posed 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 obser-
vation. In the case of the human life form, we must apply an understand-
ing 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 perform-
ing 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 simi-
lar 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 capac-
ity 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 eat-
ing 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 87
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 prem-
ises 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 our-
selves 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 charac-
terize our form of life.
As I pointed out previously, one may think that there are various ‘practi-
cal identities’ from which I could draw norms for my conduct, and even that
88 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 think-
ing 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 can-
not; 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 indi-
vidual. 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 organ-
isms 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 tomor-
row?’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 tak-
ing 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 prin-
ciple, applicable to everyone? Hence, when I act more selfishly than is gen-
erally 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 con-
stitute 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
90 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 suffer-
ing. 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 argu-
ment 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 consid-
ered a virtue if it allowed us to be rationally shameless; the dominant maxi-
mizing 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, organ-
isms 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 91
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 mos-
quitos.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 charac-
teristic 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 circum-
stances 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 crea-
tures 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 dif-
ficult and frustrating for us, given our psychological constitution. This read-
ing 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 Aristote-
lian 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
92 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 fol-
low 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 doubt-
ful 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 93
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 jus-
tifiable 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 preser-
vation of the species cannot turn on whether our continued existence prom-
ises 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 pleas-
ant 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 ratio-
nal 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 spe-
cies. 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 aim-
ing 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 any-
thing 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
94 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 neces-
sary 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 relation-
ships, 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 hap-
piness 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 benev-
olence 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 suscep-
tible 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 con-
cerns 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 oth-
ers such as friendship, loyalty, and care. Neo-Aristotelian ethical natural-
ism 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 95
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 Vir-
tue 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 Ani-
mals, 99–118.

REFERENCES

Andreou, Chrisoula, ‘Getting On in a Varied World’, Social Theory and Practice 32

(2006), 61–73.
96 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’, Philoso-
phy 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).
7 Good (as) Human Beings
Philipp Brüllmann

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 evalua-
tion. 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 propo-
nents 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 num-
ber of constraints on our consideration of the morally good. These con-
straints 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’.
98 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 norma-

tive 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 characteristi-

cally (i.e., acting in character) do in the circumstances. (28)

As Hursthouse correctly notes, this premise does not provide action guid-
ance—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 99
P.2. A virtue is a character trait that a human being needs for eudai-
monia, 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 ‘validat-
ing’ 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 par-
ticular character trait’s being a virtue’ (192). This criterion might be speci-
fied 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 self-
standing, 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 inde-
pendent because the concept of human nature does not presuppose any
notion of morally good actions, intentions, choices, etc., which might per-
chance 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
100 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.

The series of specifications that were introduced to develop a virtue ethi-

cal 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 liv-
ing 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 world-
wide, 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 Hurst-
house 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 101
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 judg-
ments, 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 specifica-

tions. 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 vir-
tue, 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 elabora-
tion, 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 vir-
tues 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
102 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 gram-
mar 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 assump-
tion that ‘good’ is a logically attributive adjective.12 What does that mean?
Briefly, it means that ‘good’, when used properly, always means ‘good rela-
tive 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 attribu-
tive 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 103
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 problem-
atic 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 Neo-
Aristotelians 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 pre-
pared to show that M knives are also N knives, and N is a characteristic
of the right kind.18
104 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 assump-
tion that ‘good’ is logically attributive. The first implication is that the rela-
tion 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 func-
tional 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 deal-
ing 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 teleologi-
cal 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 105
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 judg-
ments 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’ cor-
rectly, or she does not know what a human being is.

This summary of NAN is of course far from exhaustive, but it should suf-
fice to illustrate two things: first, what Neo-Aristotelian Naturalists mean
when they claim that there is a ‘common pattern’ to the evaluation of bees
106 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 consider-
ation 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 con-
nection, 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’, ‘life-
form’, 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 ethi-
cal 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 attrac-
tive 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 previ-

ously (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 107
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 suc-
cessful 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 rec-
ognizable 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 con-
ceptual 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 guar-
antee 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.
108 Philipp Brüllmann
It should not come as much of a surprise, I think, that NAN is not com-
patible 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) Neur-
athian 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 interest-
ing. 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 knowl-
edge. It is our conceptual ‘background-knowledge’ that allows us to judge the
correctness of criteria for evaluative judgments in the way Foot and Hurst-
house do against R. M. Hare (see note 25). Someone who calls a diseased cac-
tus 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 compat-
ible with a Neurathian procedure. It is already by submitting naturalism to
this procedure that Hursthouse detaches it from its Neo-Aristotelian back-
ground. 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 109
ethics of virtue (section X). If applying a Neurathian procedure implies deny-
ing 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 ground-
ing 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 appli-
cation 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 judg-
ments 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 back-
ground 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.
110 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 reproduc-
tion, 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 cor-
rectness 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 will-
ing 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 ‘Aris-
totelian’ approach. Though it is clear that Aristotle thinks that a virtuous
 Good (as) Human Beings 111
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?’.
2. Taking ‘moral judgment’ and ‘to be grounded’ in a very wide sense.
3. See, e.g., Millum, ‘Natural Goodness and Natural Evil.’
4. See, e.g., Hacker-Wright, ‘What is Natural about Foot’s Ethical Naturalism?’.
5. This is true even if we take ‘grounding’ and ‘moral judgments’ in a very wide
sense.
6. All references in the main text are to this work.
7. 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 illus-
trating a more general point.
8. See Gowans, ‘Virtue Ethics and Moral Relativism’ for an interesting discussion
of this aspect.
9. Cf. Woodcock, ‘Philippa Foot’s Virtue Ethics Has an Achilles’ Heel’.
10. Cf. Foot, Natural Goodness, ch. 3, and Nussbaum, ‘Aristotle on Human
Nature and the Foundations of Ethics’.
11. 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.
12. See Foot, Natural Goodness, 2–3; Hursthouse, On Virtue Ethics, 195. The
classic text is Geach, ‘Good and Evil’.
13. See Thomson, Normativity, 1–17 and 233–248 for an elaboration.
14. See Rind and Tillinghast, ‘What Is an Attributive Adjective?’.
15. 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 con-
ceptual knowledge (see section XIV).
16. 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.
17. See, e.g., Foot, Natural Goodness, 7.
18. Foot, ‘Goodness and Choice’, 133.
19. 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.
20. See Thompson, ‘The Representation of Life’, and Foot, Natural Goodness,
30–31.
21. 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 Eth-
ics, 202), and then assigns a ‘genuinely transforming effect of our [human, P.B.]
112 Philipp Brüllmann
rationality’ (218) on this basic structure (instead of adding a fifth aim for the
case of an evaluation of human beings).
22. 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!’).
23. See Foot, Natural Goodness, 2–3, 38–39; Hursthouse, On Virtue Ethics, 195.
24. 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’.
25. ‘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).
26. Hursthouse’s reference is McDowell, ‘Two Sorts of Naturalism’.
27. For some remarks, see Harman, ‘General Foundation versus Rational Insight’.
28. This illustration, which somehow undermines the distinction between epis-
temic and motivating reasons, seems problematic to me (see section XII), but
I cannot pursue the issue here.
29. See McDowell, ‘Two Sorts of Naturalism’.
30. 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).
31. 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.
32. 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 con-
tent 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).
33. 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 Kon-
zept 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 Philoso-
phy (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 Phe-
nomenological 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 Wil-
liams, 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.
8 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 tri-
angle 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 charac-
ter 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 eth-
ical 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 115
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 selec-
tion. 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 biol-
ogy (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 philo-
sophical 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 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
116 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 teach-
ing 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’. Cor-
respondingly, 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 cognitiv-
ism is looser and won’t be discussed further here.
This ambitious form of naturalism ought to interest empirical investiga-
tors 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 dif-
ference. 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 intelli-
gible 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 devel-
opment—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 develop-
ment, 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 psy-
chology, 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 117
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 ‘nor-
mative’ and the ‘individual-difference’ components of the theory,12 the ‘nor-
mative’ 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 sys-
tem 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 maintain-
ing 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
118 Edward Harcourt
traits that, working together, serve to maintain infants’ proximity to attach-
ment 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, reas-
surance, 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, height-
ened 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 mea-
sure 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 sim-
plify, 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 119
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 immod-
erately distressed when she does leave, but when she comes back doesn’t
calm down and exhibits ‘furious clinging’—‘seeking contact, then resist-
ing 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 ‘hyperactivat-
ing’).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 devel-
oped 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 attach-
ment 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 age-
specific 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 compo-
nent that I want to draw attention to concerns the further characteristics
with which secure and insecure attachment are associated. These associa-
tions 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 fol-
lowing 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 fac-
tors—such as warm relations with parents—that also explain the infantile
attachment classification. Thus if there is a predictive relation to later char-
acteristics, it is likely to be mediated by a variety of further factors.23
120 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 frustra-
tion 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, chil-
dren with secure attachment histories have less conflictual relationships with
peers from preschool to 7 years,27 are less dependent on teachers in pre-
school,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 attention-
seeking behavior than their securely attached counterparts’. Finally, attach-
ment 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 gener-
ally 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 attach-
ment 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. Attach-
ment dispositions face both backward toward the psychobiological (thanks
to attachment theory’s normative component) and (thanks to its individual-
 Attachment Theory, Character, and Naturalism 121
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 friend-
ship; 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. Co-
operativeness, meanwhile, if it is not itself a virtue, surely implies the tradi-
tional 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 het-
erogeneity 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)
122 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 Aristo-
telian 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.

I want to comment now on the backward-facing connection, from attach-

ment 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 integra-
tion . . . 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 explora-
tion apart from the parent is difficult to achieve’,39 it is clear that they don’t
mean attachment simpliciter but rather secure attachment. Again, Londer-
ville 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 quo-
tations) to be used to smuggle in the idea that secure attachment is selected
for. Depending on how close one takes the connection between secure attach-
ment 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 123
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 climb-
ing 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 deter-
mined 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 attach-
ment. If it were it would be very hard to explain why the parents of inse-
curely 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 paren-
tal 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
124 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 sacri-
fice in intimacy, for example, and enjoyable mutual interactions: adaptations
to a good environment are likely to be better, in ways that evolutionary biol-
ogy 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 with-
out 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 combi-
nations, unavailable to the insecurely attached infant. Thus quite indepen-
dently 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 privi-
lege, 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 125
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 privi-
leged 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 how-
ever that this is not to endorse cultural relativism: one of the awful things
about ancient Sparta might be that it prized the reproduction of insecure-
avoidant 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 distri-
bution 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 charac-
teristics 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
126 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 privi-
leged relation of secure attachment to well-being, that would make trouble
for the association between virtue and well-being, though perhaps that con-
nection 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 vir-
tue, 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 phenom-
enon 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 127
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 Behav-
ior 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 Car-
bon, ‘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 Frame-
work’, 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 Cha-
zan, The Moral Self.
36. Compare Chazan’s attempt to show that the Aristotelian virtues are under-
pinned 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. (‘Aris-
totle’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’.
128 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 Pro-
social 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-Year-

Olds 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 Mod-
ern 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 dur-
ing 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 (Lon-
don: 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 Mater-
nal 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 Hand-
book 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 Evolu-
tionary 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 Hand-
book 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. Radke-
Yarrow(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 Ideol-
ogy, 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.
9 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 psy-
chology 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 ratio-
nality, 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 pro-
cesses. Conscious processes are those to which we devote attention; noncon-
scious 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 131
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 psycho-
logical 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 cru-
cial 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. Conse-
quently, 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 gen-
eralized 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 over-
view 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 archi-
tecture’ (KAPA).8 Cervone’s approach illuminates what I call the ‘personal-
ity 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 Mer-
ritt, 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.
132 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 con-
stellation 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 per-
sons 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. Specifi-
cally, appraisals are relational judgments that concern the meaning of
encounters for oneself; they are ‘continuing evaluation[s] of the signif-
icance of what is happening for one’s personal well-being’ (Lazarus,
1991, p. 144). In the appraisal process, people construct personal mean-
ing 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 infor-

mation, 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 evalua-
tion 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 col-
laboration? 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 exer-
cising or developing a specific virtue. Suppose I intend to have a conciliatory
 Notes Toward an Empirical Psychology of Virtue 133
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 reli-
ance 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 try-
ing 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 mean-
ing of an encounter for the self. We might also need to broaden other fea-
tures 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 fea-
tures. 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 men-
tal 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 profes-
sor, 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 representa-
tions of myself and of the world—forms the backdrop against which ongo-
ing 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 impor-
tant 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, sup-
port 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
134 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 engage-
ment 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 apprais-
als 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 philo-
sophical situationists’ view of personality as fragmented. Let us turn first
to the argument that the self-schema and self-efficacy appraisals form scaf-
folding 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 pro-
vided by caregivers. Ideally, the cultivation of virtue becomes more delib-
erate and self-reflective as children mature and continues throughout life,
even in adulthood. Moreover, as recent Aristotelian-inspired accounts of vir-
tue 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 self-
schema 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 custom-
ers 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 cer-
tain 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 psycho-
logical 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 vir-
tue 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 defi-
ciencies. But perhaps she does not bother to make self-efficacy appraisals, in
136 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 pres-
ent 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 self-
efficacy appraisals offers a way of understanding cross-situational consis-
tency 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 differ-
ent 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 coher-
ence 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 ses-
sion, 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 domain-
linked 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’ self-
schemas, 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 137
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 impor-
tant, and lower appraisals in situations linked to personal weaknesses. Self-
appraisals 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 coher-
ence indexed to aspects of subjects’ self-schemas and beliefs about the rel-
evance of situations to those self-schemas and that self-efficacy appraisals
would not strongly relate to aschematic attributes. Second, Cervone hypoth-
esized 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 pre-
dict 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 differ-
ent 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 experi-
mental approach, then, offers a more fine-grained way of discovering the
traits people think guide, or would guide, their behavior in any given cir-
cumstance than the studies the philosophical situationists cite.
Studies have been done on the relevance of self-schemas to smokers’ self-
efficacy 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. Sub-
jects 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
138 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 experiment-
ers 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 experi-
menters interpreted slower response times as indicating a slower, more delib-
erate 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 develop-
ment 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: sub-
jects’ 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 vari-
ability 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 self-
efficacy 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 manipu-
lative.
 Notes Toward an Empirical Psychology of Virtue 139
The smoking studies, too, have relevance for virtue ethics. The experi-
mental design of Shadel and Cervone could provide a model for studies
focusing directly on virtue. One can readily imagine studies comparing self-
efficacy appraisals linked with the primed ‘compassion-ideal self-schema’
with those generated by the primed compassion self-schema of the ordi-
nary 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 Cer-
vone 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 rea-
son as a way of countering the fragmented conception of personality put for-
ward 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 ratio-
nality operative in traditional understandings of character are themselves
empirically inadequate’.34 To make their case, they invoke dual process theo-
ries 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 deliber-
ative 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
140 Nancy Snow
maintain that this behavior is caused by ‘depersonalised response tenden-
cies’: 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 non-
conscious processes that the subject would not endorse as reasons for action
that accord with his or her normative commitments.38 Were the person con-
sciously aware that the behavior was produced by the nonconscious pro-
cesses, 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 pro-
cessing, 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 sit-
uation-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 con-
scious 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 pro-
cessing 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 141
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. Post-
conscious 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 stereo-
types, 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 con-
trol 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 chil-
dren, 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 reciproc-
ity of the organism and the environment, that is, the offerings of the environ-
ment 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 affor-
dances nonconsciously; examples of perceiving affordances include understand-
ing 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
142 Nancy Snow
minds. To be good at being generous, for example, one must be able to per-
ceive 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 devel-
opment 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 Mecha-
nisms 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 143
14. Ibid.
15. Ibid.
16. See Cervone, ‘The Architecture of Personality’, and ‘Social-Cognitive Mecha-
nisms and Personality Coherence’; Shadel and Cervone, ‘Evaluating Social-
Cognitive 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’ Self-
Efficacy 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 self-
seeking. 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 situ-
ationists 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 Per-
sonality 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 stu-
dents. 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 Char-
acter’.
41. Lapsley and Hill, ‘On Dual Processing and Heuristic Approaches to Moral
Cognition’, 322.
42. Snow, Virtue as Social Intelligence, 43–45.
144 Nancy Snow
43. Narvaez and Lapsley, ‘Psychological Foundations’.
44. Ibid., 144.
45. See Merritt, Doris, and Harman, ‘Character’, 374.
46. Narvaez and Lapsley, ‘Psychological Foundations’, 146–147.
47. Ibid., 147–149.
48. Ibid., 148.
49. Ibid., 145.
50. 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 Jack-
son 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 Charac-
ter’, 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 Every-
day Morality and Moral Expertise’, in Character Psychology and Character Edu-
cation, 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 Exper-
imental 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 Aqui-
nas 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 philoso-
phers 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 appro-
priate 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 philo-
sophical 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,
146 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 doc-
trine 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 muscle-
building 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 vir-

tue—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 circum-
stances involving such-and-such persons and things and the develop-
ment 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 orienta-
tion 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 147
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 appar-
ently 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 sig-
nificant 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 chal-
lenges children appropriately, helping them to meet various kinds of resis-
tance. 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
148 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 oppor-
tunities 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 espous-
ing 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, com-
plains 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, aspira-
tions, and actions and inadvertently at odds with themselves in their efforts
to shape their practical orientations.
 Natural Virtue and Proper Upbringing 149
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 rea-
son to suppose that the complete development of each virtue will result
in the incorporation of other virtues; on the contrary, further develop-
ment of any one virtue may include the growth of a tendency that actu-
ally 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 ‘piece-
meal 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
150 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 nar-
rowly, Aristotle’s explicit dicta would rob the agent of any rational
choice about the rightness of his goals. Aristotle assumes that the virtu-
ous man, through the good fortune of his upbringing and temperament,
is, as a matter of fact, pursuing the right goals. But, then, is the Aristo-
telian 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 interpreta-
tion 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 assess-
ment 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 vir-
tue, 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 151
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 ‘nat-
ural 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 ‘natu-
ral’. The uneasiness that somehow Aristotle’s virtue is nothing more than
the child of a fortuitous marriage of temperament and pre-rational, con-
ventionally 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 doc-
trine 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 con-
ception 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,’ Aristo-
tle 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 dimen-
sions. 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
152 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 fear-
less?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 ‘circum-
stances’ 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 153
That these are natural virtues, rather than just psychological traits that
happen to please as often as not, suggests that they are, nevertheless, con-
strained. As Müller puts it:

Natural virtues do to some extent share with their ethical counterparts

constraints on the practice of their characteristic responses. These con-
straints 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, appropri-
ately 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 conclu-
sion 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 Mül-
ler’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. Industri-
ousness corrects laziness. Hope is an antidote to the temptations of despair.
154 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 under-
stand 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 think-
ing, 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 ten-
dency 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 par-
ents. 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 disposi-
tion 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 control-
ling 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 sev-
eral 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. Nor-
mally, 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 busi-
ness 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 par-
ticipating 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 bore-
dom. 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 generos-
ity 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
156 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 vir-
tues rather than just accidents of impulse and temperament. There is an ele-
ment of constraint in the feelings and acts of those with natural virtue, and
that constraint operates in the dimensions characteristic of the relevant cor-
responding 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 philos-
ophers 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 157
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 Philoso-
phy (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 Hon-
necourt, 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

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 con-
cepts, 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 philoso-
phers 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 explan-
atory 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 159
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 (pru-
dential) 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 The moral The prudential The moral and

and unobvious the prudential
Obvious and The prudential The moral The moral and
unproblematic the prudential
Prudentialism Moralism (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 call-
ing 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 presump-
tion. 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 pre-
vious Table, from the fourth possibility, skeptical dualism.) Quietist dual-
ism 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 Prin-
cipia 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
160 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 prac-
tical 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 prudential-
ism, 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 intui-
tive immediacy of our understanding both of The Moral and The Pruden-
tial. 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 unfortu-
nately is in current moral philosophy. For one thing, even if prudential rea-
sons 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 explana-
tion. 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 con-
stantly 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 ran-
domizing (‘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 exam-
ples 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 charac-
teristics 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 indi-
vidual well-being, reasons that have to do with what promotes the individual’s
health, wealth, and happiness. Adaptation for reproduction-facilitating sur-
vival 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 pur-
poses. 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 moti-
vating thoughts—‘Because fairness requires it’, ‘Because I promised’, ‘Because
162 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 suf-
fering’, ‘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 moti-
vation 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 jus-
tification 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 spe-
cial 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 prior-
ity 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 benefi-
cial, 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 ‘bene-
fit’ 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 163
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. Phi-
losophers 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 well-
being’ 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 con-
notations. 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 pru-
dential, as most commonly understood, appears to rest on an obvious false-
hood. 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 conclu-
sively 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
164 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 cyni-
cism 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, stip-
ulates 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 165
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 sta-
tion—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 apply-
ing 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 uni-
versal 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 practi-
cal 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’?
166 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 empha-
size 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 mono-
lithic and indeed capitalized they are—the better.

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 the-
sis, 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 167
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 intel-
ligible 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 hold-
ing that claritas is rooted in reason, which he describes as ‘the light
that makes beauty seen’ [ST 2.2.180.2]. 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 Gali-
lee, 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 irradi-
ated. 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 break-
fast, 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 pruden-
tially disastrous, but just happens to be, unfortunately enough, one of a class
168 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 con-
tractualism sometimes suggest. Nor is it that the moral act is prudentially
advantageous in some way—just a very obscure way, one which is consis-
tent 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 puzzle-
ment 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 com-

mitment 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 169
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 appar-
ently has a different thought: perhaps there is a sense in which the Letter-
Writers 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 eudai-
monism 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 hap-
piness 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 psychol-
ogy—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 delib-
erations 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 philo-
sophical 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.
170 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 grace-
ful 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 justifica-
tory 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, beauti-
ful 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 grace-
less 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 171
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 afore-
hand to anoint my body to the burying. Verily I say unto you, Where-
soever 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 mak-
ing a huge financial sacrifice—is wasteful and inefficient from the point of
view of public utility. Jesus responds that there are cases where beauty over-
rides public utility. This is one of them because of the grace (both theologi-
cal 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 recommend-

ing, 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
172 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 accor-
dance 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 shoot-
ing): ‘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 evolu-
tion] . . . 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 indi-
viduality’.
13. Weil, Cahiers XVI, 264.
14. Thucydides 2.42, tr. Jowett, with alterations.
15. Ramsay, Beyond Virtue, 135. See also Aquinas, ST 2.2.145.2c.
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 173
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 happi-
ness 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 Emo-
tions, 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. Wil-
liams (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 concep-
tions of human nature constrain the adequacy of conceptions of virtue
and vice. Hume, for example, claims in the Enquiries Concerning the Prin-
ciples 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 inves-

tigation is a methodological necessity for the achievement of a valu-
able 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
178 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 ten-
derness 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 ben-
efits 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 sug-
gested 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 intel-
ligibility 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 appro-
priate 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 vir-
tuous 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 179
‘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 proper-
ties 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 suit-

able responses in competent, qualified responders (for example, senti-
ments 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 norma-
tive. 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 quali-
fied one, but on certain facts, such as the trait actually systematically pro-
ducing 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 how-
ever 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
180 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 ethi-
cal 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 prop-
erty 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 suffi-
cient 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 require-
ments, will seem mysterious.
The solution to both of these problems is to understand (I) as constru-
ing 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 Hei-
degger’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 181
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 skep-
tic 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 prac-
tical: 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 engage-
ment. 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 philoso-
phers, that morality is susceptible of demonstration; and tho’ no one
182 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 objectiv-
ity 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 depen-
dent.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 objec-
tivist tradition. However, in this chapter I wish to tie his response depen-
dence 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 Pig-
den’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 essen-
tial 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 phe-
nomenology, gives us explicit resources for securing that understanding.
 A New Metaphysics for Virtue Ethics 183
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 sim-
ply 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 scien-
tific 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 comport-
ment 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 uncon-
cealment. Fundamental orientations or attunements are ways of what Hei-
degger calls unconcealing: opening or revealing the world as a whole in one
or other of its aspects. These constitute backgrounds of understanding (com-
portments), 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
184 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 intelligi-
bility.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 unhid-
den. Truth as correctness is grounded in truth as unhiddenness’.19
We are now in a position to see what is wrong with orthodox interpreta-
tions 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). Emo-
tional 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 epistemol-
ogy 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 sentimental-
ism 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 sen-
sibilities capable of experiencing color, there would be no color properties; if
 A New Metaphysics for Virtue Ethics 185
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 proper-
ties; if there were no aesthetic sensibilities, there would be no beauty. With-
out 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 dis-
position, 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 blind-
ness 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 disclo-
sure 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 3.3.5.1), constituting our emotional sense of virtue properties. The condi-
tions 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 psy-
chological 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 strang-
ers 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
186 Christine Swanton
reason proper to the understanding, are not sufficient to allow ethical prop-
erties to ‘be out in the open’ or unconcealed. What such sensibilities and
reason aim to do is discover causal relations and ‘eternal’, ‘immutable’ fit-
nesses 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 pos-
sessed 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 pas-
sions 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 ‘con-
trary 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 epis-
temology 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 ani-
mals 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 seri-
ous 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 187
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 condi-
tion of a moral sense proper to humans is benevolence constrained by self-
love. 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 pro-
vides 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),
188 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 debat-
able, certainly where his ethics is concerned.
We have seen that the fundamental idea of truth as aletheia is the intelligi-
bility of a thing that, for Heidegger, presupposes engagement with and com-
petence 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 pri-
vate 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 counte-
nance 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 [oth-
ers’] 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 man-
ner 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 189
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 accompany-
ing phenomena; rather, they are the sort of thing that determines being-
with-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 directional-
ity, 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.

190 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 eth-
ics 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 author-
itative moral sense of qualified persons.

This characteristic of virtue is associated with the second aspect of the four-
fold, 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 essen-
tially 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 feel-
ing 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 understand-
ing. 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 191
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, emo-
tional 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 proper-
ties (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 McDow-
ell’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 episte-
mology. 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 ‘reac-
tion 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 “anti-
realist”. 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
192 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.
15. See my ‘Can Hume be Read as Virtue Ethicist?’.
16. See note 14.
17. Mark Wrathall ‘Unconcealment’ in Heidegger and Unconcealment , 11–39, 14.
18. Heidegger, Being and Time, sect. 44, 257.
19. 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 ale-
theia as a form of truth. However this usage was criticized by Ernst Tugendhat
(see Tugendhat, ‘Heidegger’s Idea of Truth’, 83–97), and Heidegger eventu-
ally 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 ‘Hei-
degger’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.
20. 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 inten-
tional 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.)
21. Hume, ‘The Standard of Taste’, 222.
22. See Hume, ‘The Standard of Taste’, 275; T 591.
23. 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.
24. These criteria of virtue and their pluralistic nature are discussed much more
fully in my ‘What Kind of Virtue Theorist is Hume?’
25. 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).
26. 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 vari-
ous possible ways, and ‘where there is culture, it affects everything’ (102). See
also Hacker-Wright ‘What is Natural about Foot’s Ethical Naturalism?’.
27. Heidegger, The Fundamental Concepts of Metaphysics: World, Finitude, Soli-
tude, 67.
28. Heidegger, Fundamental Concepts of Metaphysics, 66.
29. 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. Nid-
ditch. (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. Hond-
erich (London: Routledge and Kegan Paul, 1985).
Pigden, Charles R., ‘If Not Non-Cognitivism, Then What?’ in Hume on Motiva-
tion 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 Vir-
tue, 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.
194 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 Wil-
liams (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

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; how-
ever, 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 natu-
ral 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 Claim-
ing 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 fun-
damental concept. Thus, Rawls himself considered virtues to be just ‘strong
and normally effective desires to act on the basic principles of right’.2 Sub-
stituting ‘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, vir-
tues 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
196 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 conceptu-
alizations 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 Bar-
bara 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 develop-
ing 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 con-
tent 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 ret-
rospective 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 max-
ims 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? 197
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 pre-
supposes 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
198 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 informa-
tion about human agents, unless we can justify the moral relevance of the
information we want to include. Even if the rules of salience cannot, gener-
ally, 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 indi-
cate ways of answering the other questions.

III

There are different attempts to fill the gap in Kant’s account of moral right-
ness 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 inter-
nal’ 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, law-
formulas, lack. But, given that Kant thought the third formula to be equiva-
lent to the other two, it would be very surprising if applying this formula
 A Kantian Plea for Virtues? 199
should, ipso facto, yield substantially stronger results. Furthermore, while
it appears plausible that we can reject some conceptualizations of action-
situations 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 illegiti-
mate in precisely this way. For example, when the mistake in the conceptu-
alization 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, becom-
ing even more specific—to differences of gender, race, etc.—in the Anthro-
pology.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 misun-
derstanding 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 constitu-
tive 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 natu-
rally 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,
200 Erasmus Mayr
at least, not straightforwardly so. For the problem is not simply the gen-
eral 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 con-
nected 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 appropri-
ate. 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 apprecia-
tion 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 knowl-
edge 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? 201
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 for-
mal, ethical principle (like ‘healthy food should be eaten’), either constitut-
ing 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 consti-
tutive 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 obliga-
tory 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 comprehen-
sive as to already include sensitivity to what was demanded of us. What
the Kantian account needs to be supplemented by is only something that
202 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

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 com-
plete 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 pos-
sesses 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 cer-
tain 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). How-
ever, 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 fruit-
ful 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 flourish-
ing 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 moral-
ity, 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? 203
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 miss-
ing 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 distin-
guish 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 presup-
poses 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 constitu-
tive 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
204 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 posses-
sion—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 con-
sidered. (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 com-
prising features that either directly follow from our status as human agents,
or are in an obvious way—either as constitutive parts or ‘practically’ neces-
sary 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? 205
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 character-
traits, 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 there-
fore 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 character-
traits and intellectual virtues possession of which is required for the achieve-
ment 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 consider-
ably 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 ‘cur-
tail’ 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.
206 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 ‘vir-
tuous 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 normativ-
ity. (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 intellec-
tual 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? 207
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 diverg-
ing 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 Vossen-
kuhl for very helpful comments on earlier drafts of this paper.
1. Cf. Watson, ‘On the Primacy of Character’, 229.
2. Rawls, A Theory of Justice, 436.
3. MM Akademie-Ausgabe (AA) 6: 394.
4. 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.
208 Erasmus Mayr
7. See O’Neill, Constructions of Reason. Explorations of Kant’s Practical Phi-
losophy, 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 perspec-
tive 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 Impera-
tives’, 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) ‘calcu-
late’ 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 pas-
sage, there is a corresponding moral duty to cultivate self-control, but pursu-
ing 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 cour-
age 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: Cam-
bridge 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 (Cam-
bridge: Cambridge University Press, 2011), 164 ff.
Esser, Andrea, Eine Ethik für Endliche: Kants Tugendlehre in der Gegenwart (Stutt-
gart: 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 (Cam-
bridge, 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 Philoso-
phy, (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

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 ver-
sion 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 shed-
ding 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 Swan-
ton1 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 con-
ceived 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, rec-
ognizes 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 211
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 3.2.1.2; SBN 477).5 In turn, these
motives have to be related to the characters of people, and people are mor-
ally evaluated because they display characters of certain kinds (T 3.3.1.4–5;
SBN 575 and T 3.3.1.19; 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 3.2.2.26;
SBN 500–501).6 Hume appears to be concerned with that ‘ethical forma-
tion’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 sentimental-
ism. Hume says that, by ‘custom and education’ (T 3.2.2.26; SBN 500) peo-
ple 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 lack-
ing 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 2.3.3.8 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 neo-
Aristotelianism9—according to which being properly educated means being
able to respond correctly to the moral features of a given situation. Accord-
ing 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 require-
ments 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
212 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 dis-
covered by the understanding’ (T 3.1.1.26; SBN 468). Rather, values seem
to work rather like secondary qualities (T 3.1.1.26; 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 dimen-
sion 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 stain-
ing 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 3.1.2.3; 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 3.2.8.8; 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 3.3.1.18; 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 sentimen-
tal 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 213
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 response-
dependent 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 assess-
ing 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 Swan-
ton and Slote make reference to the work of David Wiggins. Wiggins presents
a ‘sensible subjectivism’, according to which moral properties and appropri-
ate human sentiments are mutually correlated in <property, response> asso-
ciations, 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 cogni-
tivist 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 subjec-
tivism: x is good if and only if x is such as to deserve (N.B.) or merit approba-
tion’.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 reason-

giving force of ethics, and in particular for our justification about the
214 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 determin-
ing what really deserves to be considered ‘moral’ (such as, for example,
traits of character that are real virtues and vices) and what is just pleas-
ant 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 objec-
tive, 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 judg-
ments. 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 con-
ception of morality as a whole (as Foot does), or by radically reformulat-
ing 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 215
IV

For Hume,

Every quality of the mind is denominated virtuous, which gives plea-

sure 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 3.3.1.30; 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 plea-
sure or advantage for other people or for their possessors themselves and
disapprove those traits of character that give pain or prove to be disadvanta-
geous for other people or their possessors themselves. Specifically, we have
a properly moral approval (or disapproval) when these sympathetic judg-
ments on traits of character are given from what Hume calls a ‘steady and
general’, or ‘common’, point of view (T 3.3.1.15–16 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 con-
sideration 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 defin-
ing 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 pos-
sessor’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
216 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 per-
fectionist 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 mak-
ing 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 measure-
ment—human nature as she conceives it—Hursthouse can present a notion
of the phronimos that corresponds to somebody who shows practical wis-
dom, 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 some-
thing 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 sympa-
thetic 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 senti-
mental 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 ascertain-
ment. Which character traits happen to be agreeable or useful to their pos-
sessors 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 ‘experi-
ments [. . .] 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 presup-
posing 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 217
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 con-
sequence 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.

From a Humean perspective, unlike from a neo-Aristotelian one, there is no

ontological commitment regarding the nature of the virtuous agent. Neo-
Aristotelians36 long for a unity of the virtues that can be stated only by
presupposing an idealized notion of human nature, and hence by presup-
posing 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 particu-
lar 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 3.3.2.8; SBN 596–597 and T 3.3.2.11; 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 vir-
tuous 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 eth-
ics proposal is far from relativistic; by making reference to the sentimental
218 Lorenzo Greco
imagination of human beings, which is exercised in continual confronta-
tions 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 phron-
imos that comes to be rigid and hardly helpful for contemporary ethics. On
the one hand, it is disputable whether the authentic notion of the Aristote-
lian 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 com-
petes 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 3.3.2.13; 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 con-
templation of it so elevates the mind, that we cannot refuse it our admi-
ration. 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 3.3.2.15; 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 Swan-
ton’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 219
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 dis-
tinguish virtues and vices, but rather the individuality of persons. Individu-
ality 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 vir-
tues; consistency becomes a problem only if we decide to embrace an ‘abso-
lute’ 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 indi-
viduality of a given person, not the determination of an objective perspec-
tive 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 differ-
ent 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 3.3.3.9; 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, ‘democ-
ratized;’ 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
220 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, ‘afford-
ing 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 per-
son 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 3.1.2.3;
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 221
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.
18. Slote, Moral Sentimentalism.
19. Slote, for example, admits that Hume’s work gives way to different interpre-
tations: ‘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 actu-
ally believes’ (Slote, ‘Moral Sentimentalism’, 8–9).
20. Wiggins, ‘A Sensible Subjectivism?’, 187.
21. See Wiggins, Ethics: Twelve Lectures on the Philosophy of Morality, 331–333.
22. Wiggins, ‘A Sensible Subjectivism?’, 194.
23. Wiggins, Ethics, 371.
24. Swanton, ‘Can Hume Be read as a Virtue Ethicist?’, 101.
25. Philippa Foot, ‘Hume on Moral Judgement’.
26. 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).
27. 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.
28. 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 Objec-
tivity’.
29. 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 3.1.1.26; SBN 469).
30. Hursthouse, ‘Virtue Ethics and Human Nature’, 73.
31. Ibid., 78–79.
32. Hume, ‘Of the Standard of Taste’.
33. Hursthouse, ‘Normative Virtue Ethics’, 23.
34. Hursthouse, On Virtue Ethics, part 3.
35. 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.
36. 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.
37. I discuss this point in Greco, L’io morale: David Hume e l’etica contempora-
nea, parts 2 and 3.
38. On the notion of a moralized pride, see Baier, ‘Master Passions’; Herdt, Reli-
gion and Faction in Hume’s Moral Philosophy, chap. 2; Rorty, ‘The Vanishing
Subject: The Many Faces of Subjectivity’.
39. On Hume’s conception of human excellence, see Martin, ‘Hume on Human
Excellence’.
222 Lorenzo Greco
40. Swanton, Virtue Ethics, 102.
41. Slote, From Morality to Virtue, chap. 14.
42. See Swanton, ‘Profiles of the Virtues’; Swanton, ‘Nietzschean Virtue Ethics’.
43. On the idea that Hume’s ethics is pluralistic, see Abramson, ‘Hume on Cul-
tural Conflicts of Values’; King, ‘Hume on Artificial Lives with a Rejoinder
to A. C. MacIntyre’.
44. On ‘the person of mixed character’ see Cohon, Hume’s Morality, 149 ff.
45. This aspect is well explained by Dees, ‘Hume on the Characters of Virtue’.
46. Hume, ‘Of the Middle Station of Life’, 546.
47. 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 Confer-
ence, 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 Univer-
sity Press, 2008).
D’Arms, Justin, and Jacobson, Daniel, ‘Sensibility Theory and Projectivism’, in The
Oxford Handbook of Ethical Theory, ed. David Copp (Oxford: Oxford Univer-
sity Press, 2006), 186–218.
Dees, Richard H., ‘Hume on the Characters of Virtue’, Journal of the History of Phi-
losophy 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 Philoso-
phy (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: Clar-
endon 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 Metaphysi-
cian (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: Aris-
totelianism 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: Clas-
sic and Contemporary Readings in Virtue Ethics, ed. Jennifer Welchman (India-
napolis: 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’ Nikoma-
chischer 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 Sapi-
enza 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 Humani-
ties. His books include Objektives Denken. Erkenntnistheorie und Philoso-
phy of Mind in Hegels System (Frommann-Holzboog, 2002) and Praktische
Gründe und die Realität der Moral (Vittorio Klostermann, 2007).
226 List of Contributors
Edward Harcourt is Fellow & Tutor in Philosophy at Keble College, Oxford.
His research interests include ethics (in particular moral psychology, neo-
Aristotelianism and child development, ethical dimensions of psycho-
analysis, 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 Uni-
versity Press, 2011), as well as of numerous articles in moral and political
philosophy, some of which are collected in Drawing Morals (Oxford Uni-
versity 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 Col-

lege, 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 phi-
losophy 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 psychol-
ogy. 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 Milwau-

kee, 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 Intel-
ligence: 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 Auck-

land, 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 neo-
Aristotelian naturalism. She is the author of Reasonably Vicious (Harvard
University Press, 2002).
Index
A eudaimonia and 30–2, 35;
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;
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; well-
being and eudaimonia
Aristotelian necessity 116
Aristotelian triangle (human nature,
human excellence, human well-
being) 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
230 Index
(Ainsworth) 118–19; synchronic
dimension 119; type A: insecure-
avoidant 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 Chappell, Timothy 4, 158–73; “Glory
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 courageous soldier example, of virtuous
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
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 well-
being 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
sacrifice and personal loss 69,
71–6
cowardly acts 71, 72

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 virtuous sacrifice and personal
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
Index 231
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;
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 Letter-
Writers 168–9; Natural Goodness
115–17; Neo-Aristotelian
Naturalism (NAN) and 99, 101,
110; see also naturalism; Neo-
Aristotelian Naturalism (NAN)
232 Index
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 criticism of Kant 196
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 183, 184–6, 189; aletheia
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
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
(Heidegger): 2. uncoveredness

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 neo-
Aristotelianism 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 of Practical Reason internal
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
Index 233
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
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
234 Index
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 moral sentimentalism (Hume) 178,
Lapsley, Daniel K. 131, 140–2
LeBar, Mark 3, 52–68; see also well-
being 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 Müller, Anselm 146–56
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 self-
schema appraisals
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
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; Neo-
Aristotelian 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,

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 practical reasoning 28, 57–9, 63–4,
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
Index 235
optimal development 116; see also
attachment theory
outside-in approach (higher-level good)
9–12, 23–4
P
pain: Aristotelian view and 18; higher-
level moral concept and 11, 12, 19
Parfit, Derek 48
Perfectionism 52–9, 66, 220; agent-
neutrality 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
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
236 Index
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 values 43; unbeneficial 46–7;
Qualified Agent Thesis 179–80
quietist dualistic presumption
(category) 159–60
Quine, W.V.O. 106–7
R scope-fallacy 172
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 well-
being and eudaimonia
S
sacrifices 3; happiness and 43, 48; of
opportunities 54–5; of prudential
see also virtue and happiness
connection; virtuous sacrifice
and personal loss
scientific investigation 177–8, 182
scientific truth 31
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

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:
Index 237
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;
238 Index
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; well-
being 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