Virtue, Narrative, and Self: Explorations of Character in the Philosophy of Mind and Action (Routledge Studies in Ethics and Moral Theory) [1 ed.] 0367418207, 9780367418205

Virtue, Narrative, and Self connects two philosophical areas of study that have long been treated as distinct: virtue th

667 60 4MB

English Pages 268 [269] Year 2020

Report DMCA / Copyright


Polecaj historie

Virtue, Narrative, and Self: Explorations of Character in the Philosophy of Mind and Action (Routledge Studies in Ethics and Moral Theory) [1 ed.]
 0367418207, 9780367418205

Table of contents :
Half Title
Series Page
Title Page
Copyright Page
Dedication Page
1 Virtue, Narrative, and Self: An Introduction
2 Virtue Ethics and Narrative Virtue
3 Narrative Virtues and Second-Order Reasons
4 Narrative Virtue Ethics?
5 How Self-Narratives and Virtues Cause Actions
6 Virtue Ethics, Narrative, and Revisionary Accounts of Rightness
7 Virtuous Perception: A Gibsonian Approach
8 Virtue Ethics, Blameworthiness, and Role Failure
9 Well-Being, Narrative Value, and Virtue Ethics
10 On the Value of Moral Failure
11 Who Wrote Nietzsche’s Autobiography?
12 The Abnegated Self
13 Integrity and Messy Lives
List of Contributors

Citation preview

Virtue, Narrative, and Self

Virtue, Narrative, and Self connects two philosophical areas of study that have long been treated as distinct: virtue theory and narrative accounts of personal identity. Chapters address several important issues and neglected themes at the intersection of these research areas. Specific examples include the role of narrative in the identification, differentiation, and cultivation of virtue, the nature of practical reasoning and moral competence, and the influence of life’s narrative structure on our conceptions of what it means to live and act well. This volume demonstrates how recent work from the philosophy of mind and action concerning narrativity and our understanding of the self can shed new light on questions about the nature of virtue, practical wisdom, and human flourishing. This book will be of interest to scholars and advanced students working in virtue theory, moral philosophy, philosophy of mind and action, and moral education. Joseph Ulatowski is Senior Lecturer in Philosophy and Director of the Experimental Philosophy Research Group at the University of Waikato, New Zealand. He is the author of Why Facts Matter: Facts and Pluralism in the Age of Fake News (forthcoming) and Commonsense Pluralism about Truth: An Empirical Defence (Palgrave, 2017). Liezl van Zyl is Associate Professor of Philosophy at the University of Waikato, New Zealand. She is the author of Virtue Ethics: A Contemporary Introduction (Routledge, 2018) and Death and Compassion: A Virtuebased Approach to Euthanasia (Routledge, 2000).

Routledge Studies in Ethics and Moral Theory

An Intersectional Feminist Theory of Moral Responsibility Michelle Ciurria The Principle of Double Effect A History and Philosophical Defense David Černý Apologies and Moral Repair Rights, Duties, and Corrective Justice Andrew I. Cohen Kantian and Sidgwickian Ethics The Cosmos of Duty Above and the Moral Law Within Edited by Tyler Paytas and Tim Henning Cultivating Our Passionate Attachments Matthew J. Dennis Reason and Ethics The Case Against Objective Value Joel Marks Offense and Offensiveness A Philosophical Account Andrew Sneddon Virtue, Narrative, and Self Explorations of Character in the Philosophy of Mind and Action Edited by Joseph Ulatowski and Liezl van Zyl For more information about this series, please visit: Routledge-Studies-in-Ethics-and-Moral-Theory/book-series/SE0423

Virtue, Narrative, and Self Explorations of Character in the Philosophy of Mind and Action

Edited by Joseph Ulatowski and Liezl van Zyl

First published 2021 by Routledge 52 Vanderbilt Avenue, New York, NY 10017 and by Routledge 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2021 Taylor & Francis The right of Joseph Ulatowski and Liezl van Zyl to be identified as the authors 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 A catalog record for this book has been requested ISBN: 978-0-367-41820-5 (hbk) ISBN: 978-0-367-82330-6 (ebk) Typeset in Sabon by Apex CoVantage, LLC

For Ron Flood


Acknowledgements 1 Virtue, Narrative, and Self: An Introduction

ix 1


2 Virtue Ethics and Narrative Virtue



3 Narrative Virtues and Second-Order Reasons



4 Narrative Virtue Ethics?



5 How Self-Narratives and Virtues Cause Actions



6 Virtue Ethics, Narrative, and Revisionary Accounts of Rightness



7 Virtuous Perception: A Gibsonian Approach



8 Virtue Ethics, Blameworthiness, and Role Failure



9 Well-Being, Narrative Value, and Virtue Ethics N I C H O L A S RYAN SMITH




10 On the Value of Moral Failure



11 Who Wrote Nietzsche’s Autobiography?



12 The Abnegated Self



13 Integrity and Messy Lives



List of Contributors Index

246 251


We are grateful for the financial support we received from the University of Waikato’s Faculty of Arts and Social Sciences Dean’s Office to host a conference in 2017 on “Virtue, Narrative, and Agency.” The work that appears here is a consequence of conversations and discussions that arose during and after that conference. We would like to acknowledge the generous support of Allison Kirkman, the Dean’s Office staff, and the Philosophy Programme’s administrative assistant Paula Maynard. We also want to thank the University of Connecticut Humanities Institute. Much of the work associated with the volume’s assembly would not have been possible without the generous support Joseph received as a Visiting Faculty Research Fellow at the Institute. We also greatly appreciate the editorial team at Routledge, especially Andrew Weckenmann and Allie Simmons, who deserve special recognition for their unwavering support for this project and helping see to its completion during a global pandemic that has manifested in some selves the most virtuous conduct but in others the most vicious and callous behaviour. Finally, Liezl would like to thank her family, Paul, John, Alex, and Tessa, for their love and kindness and occasional technical support. Joseph is grateful to his family, Tiffany, Lucjan, and Anja, who were an immutable source of love, encouragement, and inspiration while he was drafting, revising, and finishing his contribution to this volume. Without them, no narrative merits to be told and no virtuous character worthy of cultivation.


Virtue, Narrative, and Self An Introduction Joseph Ulatowski and Liezl van Zyl

1. Introduction After a long dormant period, the late twentieth century saw a renewed interest in both moral and intellectual virtue. Foundational work on virtue ethics began with the works of Plato and Aristotle. However, from the Middle Ages to the middle of the twentieth century there were few contributions of original work that advanced the discussion of virtue theory. Much of the work looked at “virtues” as if they were intellectual artefacts created by Plato, Aristotle, and other Greek and Roman philosophers (cf. Foot 2002; Hursthouse 1999; MacIntyre 1981; Slote 1992; Williams 1985). In moral philosophy, especially during the nineteenth and early twentieth centuries, virtue ethics was eclipsed by utilitarianism and deontology. The pioneering work of Immanuel Kant in the Enlightenment and of Jeremy Bentham and John Stuart Mill during the Industrial Revolution diverted ethicists’ attention away from virtue and the important role it plays in moral philosophy. Instead, the focus of modern moral philosophy was on discovering a criterion of right action. The study and recognition of the importance of virtue returned to moral philosophy in the 1950s. It was ushered in by Elizabeth Anscombe and capitalised upon by Philippa Foot, Alasdair MacIntyre, Iris Murdoch, and Rosalind Hursthouse. Not only did the study of virtue resurface in debates of moral philosophy but it also became a part of the epistemological conceptual arsenal in the 1980s, following on work by Ernest Sosa. This volume aims to expand the reach of virtue theory beyond the confines of virtue ethics and virtue epistemology, while at the same time revealing how discussions of narrativity and self may inform work in virtue theory. The task we set before us is one that moves discussion forward into issues of philosophical psychology, action theory, and the philosophy of mind. Comparatively little work has been done to show how virtue may inform debates in these important areas of philosophical research. The exercise of virtuous character arises from a disposition to act in certain ways and those dispositions may be formed in part by what is constitutive of the self or the story we tell about ourselves. Similarly, what the


Joseph Ulatowski and Liezl van Zyl

self is and how we talk about the self may inform discussion of virtuous character and human flourishing. Virtue theorists have only begun to consider the ways in which research in action theory and the philosophy of mind can enrich theorising about virtue, right action, and human flourishing. Our hope for this volume is to begin a sustained and fruitful dialogue between virtue theorists, action theorists, and philosophers of mind. In this introduction, we intend to justify the expansion of virtue theory into the domains of philosophy of mind and philosophy of action. First, in Section 2, we provide some exegesis on the recent history of virtue theory. “The Virtue Turn” may be traced to two sources: Elizabeth Anscombe in ethics and Ernest Sosa in epistemology. This return to virtue theory has sparked some significant and fascinating debates. In Section 3, we defend expansion of virtue-theoretic talk into discussions of narrative and the self, and vice versa. Our arguments revolve around the broad connections we may make between virtue, narrative, and the self. For example, we show how living well involves the cultivation of good character, which makes telling a story of one’s self equally worthwhile. We believe that our understanding of the self and of the stories we tell about events that occur in a lifetime are enriched by the virtuous (or vicious) character that we nurture over a lifetime. The final section provides summaries of each contribution.

2. “The Virtue Turn” If there is a watershed moment in analytic philosophy marking the return to virtue theory, and virtue ethics in particular, it is the publication of Elizabeth Anscombe’s seminal paper, “Modern Moral Philosophy” (1958). Anscombe challenged the two dominant moral philosophies of the day, utilitarianism and deontology. She regretted the fact that virtues and vices, conative states, moral character and its education, as well as questions over how we should live, were no longer central in philosophical analyses of morality, and that debates in moral philosophy had become preoccupied with abstract moral principles that presuppose the existence of a supreme lawgiver. Anscombe argued for a return to the rich vocabulary of virtues and vices, which would be informed by an understanding of human nature and human flourishing. She also insisted that moral philosophy should turn to philosophical psychology to gain clarity on notions such as “action,” “intention,” and “desire.” Anscombe’s essay paved the way for work in virtue theory and moral psychology, as well as a renewed interest in questions about human nature and human flourishing. The literature on virtue in moral philosophy has broken new ground by showing how virtue figured prominently in the work of major historical figures in philosophy other than Aristotle, such as Immanuel Kant (Baxley 2010; Betzler 2008; Cureton and Hill 2014, 2018; Herman 1996; O’Neill 1996), David Hume (Slote 2001;

Virtue, Narrative, and Self


Swanton 2015), and Friedrich Nietzsche (Swanton 2003, 2011, 2015). Others have attempted to defend Plato as a virtue ethicist. While it is no doubt that Plato, and through his dialogues Socrates, was a virtue theorist, some ink has been spilt to revive him as a virtue ethicist (Annas 1999; Kamtekar 1998; Prior 1991; Reshotko 2006). Given how such studies of prominent philosophers have flourished, there is no doubt that such research will continue to expand within and beyond the confines of Western philosophy. On top of the historical work, new virtue-theoretic accounts began to arise in moral philosophy and epistemology to invigorate debates that had become stale. “Virtue” need not be thought of as a relic of the past but a fecund concept that may spawn new and interesting topics of philosophical research. The cultivation of good character through the habituation of acting virtuously may play an important contributory role in the development of moral and epistemic theories. By the middle of the twentieth century moral theorising had been belaboured by the overly systematised and reductivist styles of Kantian deontology and Millian utilitarianism (Darwall et al. 1992). Although there are various forms of virtue ethics, they have in common the view that the virtues play a central role in the evaluation of actions and agents. Instead of reducing morality to a single abstract principle, virtue ethicists focus on a plurality of virtues and vices and emphasise the complexity and specificity of moral reasoning. The most popular versions of virtue ethics include (i) eudaimonist, (ii) agent-based, and (iii) target-centred virtue ethics (cf. Hursthouse and Pettigrove 2018; Van Zyl 2018). Eudaimonist virtue ethicists follow ancient Greek and Roman philosophers by grounding virtue in the good or happy life (eudaimonia) and by emphasising practical wisdom as a requirement for virtue (Hursthouse 1999; Annas 2011; Badhwar 2014; Bloomfield 2014; LeBar 2009). Agent-based virtue ethicists have contended that forms of normativity, such as right action, justice, the value of eudaimonia, and practical rationality, may be explained by motivational or dispositional qualities of agents (Slote 2001; Zagzebski 2004). Finally, the target-centred approach to virtue ethics was developed by a contributor of this volume, Christine Swanton (2003). It defines virtue as a disposition to respond to or acknowledge items within a field in an excellent way. Although Swanton denies that the virtues are grounded in eudaimonia, she accepts that there is a strong link between them. Many virtue ethicists recognise the importance of narrative in our conception of what it means to be happy or live a good life as a human being. For example, in an essay on the relationship between virtue and human flourishing, Robert Roberts writes: Adult human persons have a strong bent to think of themselves and their fellow human beings as having a narrative in which they are a


Joseph Ulatowski and Liezl van Zyl continuous (though no doubt changing) character. A person does not just live; he or she lives a life. (Roberts 2015, 38)

As we will explain in more detail shortly, some of the contributors to this volume draw on the literature on narrativity to develop or enrich their account of role-specifc or “narrative” virtues, whereas others use it to develop their understanding of “the good life” or eudaimonia. Yet others consider the role of narrativity or narrative virtue in the evaluation of actions or agents. Outside of virtue ethics, the study of virtue has expanded other philosophical concepts, such as knowledge. In “Is Justified True Belief Knowledge?” Edmund Gettier (1963) had shown that there was no hope for a traditional justified true belief conception of knowledge—a view of knowledge that had been around since Plato. Gettier used a counterexample to show that one may have a justified true belief without also having knowledge. Suppose Valerie has a justified belief that “Mallory will win the Bachelor.” She also knows that she and Mallory like to ride horses, so she concludes that “the person who wins the Bachelor likes to ride horses.” In fact, Valerie wins the contest. Valerie’s belief “the person who wins the Bachelor likes to ride horses” was justified and true, but Valerie does not know that the person who wins the Bachelor likes to ride horses because her justification for that belief has to do with Mallory’s liking to ride horses. Following Gettier’s challenge, a cottage industry of articles offered replies to the counterexample in order to resurrect the orthodox conception of knowledge. Alvin Goldman, for example, proposed that justification for a belief had to be caused in the right way (Goldman 1967). In later work, he argued that justification had to be reliably formed, a view which he subsequently extended to all knowledge (Goldman 1979, 1986).1 None of the proposed replies fared any better because they suffered from their own deficiencies. In the 1980s, Ernest Sosa introduced the idea that an “intellectual virtue” may help to resolve the ongoing tension between foundational and coherentist theories of knowledge. Since that time, various veins of research on virtue epistemology have grown and blossomed not only in the work of Sosa but also Jonathan Kvanvig (1992, 2003) and Linda Zagzebski (1996).2

3. Expansion Beyond Ethics and Epistemology As we said at the outset, the aim of this volume is to expand the influence of virtue to action theory and philosophy of mind by exploring links connecting virtue with the notions of narrative and self. We have already spoken of the history of virtue and its expansion into territories and debates within moral philosophy and beyond that into epistemology. So, in this section our task is to persuade the reader that the notions of narrative

Virtue, Narrative, and Self


and self may be informed by virtue theoretic considerations, and vice versa. The concept of narrative and the concept of self tend to be points of discussion in metaphysics, philosophy of mind, philosophy of action, and practical reasoning. For example, since the 1980s, there has been a fruitful discussion of narrative in the philosophy of mind (e.g., Carr 1986; Dennett 1992; MacIntyre 1981; Ricoeur 1992; Schechtman 1996). Recently, around the middle 2000s, that discussion has expanded to include issues in moral philosophy. So, in this introduction, we would like to take stock of the state of these three subdisciplines and how it is that we have come to envision how they have become entangled. First, given that the discussion of virtue has infiltrated discussion of theories of knowledge, there seems to be no reason for us to rule out that such discussions will be fruitful for other areas of philosophical interest. Epistemology had been fatigued by its inability to make progress with respect to the Gettier problem. Sosa’s turn to questions of intellectual virtue spawned a new means of contending with the problem. There is no reason to believe that we should not expect the same to occur in action theory. For example, a relatively common line of thought for action theorists is to distinguish between what people do and what merely happens to them. What people do is called an “action,” which comes with a rich psychological structure or an intention usually ascribed to an agent. That intention may form from an agent’s disposition to behave in a certain way. This would suggest that a virtue an agent may have cultivated over a lifetime may figure into the formation of an intention. Moreover, action theorists have been for many years concerned with explaining an intentional action in terms of an agent’s reasons for action. Some have found a causal explanation of action in an agent’s desires, intentions, and meansend beliefs (cf. Goldman 1970), while others have stood firm against the causal view by suggesting a “non-causal bringing about” (Taylor 1964). More recently, George Wilson (1989) and Carl Ginet (1990) have proposed reason explanations as grounded in an agent’s intention, which resembles Anscombe’s view (1957/2000), while Michael Thompson (2008) has presented a case for jettisoning anything resembling a causal or non-causal approach to the explanation of action. Since the explanation of action is so interwoven with an agent’s intention and the formation of one’s character, this debate may play an informative role in thinking about virtues and its partnership with character. Injecting the study of virtue and character into action theory or thinking of ways in which issues in action theory may inform our appreciation of how one may lead a virtuous life, we believe, may yield new veins of research for us to mine. Second, at the same time, but independent of the developments in virtue ethics, there has been a growing interest among philosophers of mind and action, in the role of narrative in selfhood, identity, and agency.


Joseph Ulatowski and Liezl van Zyl

Broadly speaking, narrativity is the view that our lives have narrative shape and that this is of central importance to our sense of self and our view of what it means to live well (see, e.g., Dennett 1991, 1992; Flanagan 1996; Goldie 2012). Part of living well should involve the cultivation of good character and the virtues. So, we should expect that theorising about narrative will involve virtue. Likewise, the study of narrative may inform debates in the study of individual virtues. Since virtues are a characteriological disposition to behave in certain ways, and since the formation and maintenance of our character may involve our own reflection upon the events that comprise our life, we should expect that we have a story to tell of our virtues. This new research suggests that it could inform our theorising about virtue, and vice versa. Third, virtue as character is informed by the way we identify our selves. How we identify our selves is sometimes through the self being understood as having a narrative arc or a story to tell about our self. How we live our lives depends upon the character traits (virtues and vices) we have cultivated. So, there is a fluid connection between virtue as character and the narrative view of the self. Human beings have been endowed with the capacity to respond to reasons for action, and virtue ethicists of various stripes have been interested in the nature and role of practical wisdom as a disposition to reason well about what one is to do. For example, a truly courageous person acts from reasons consistent with the virtue of courage, such as “It is worth risking my safety to prevent many people from suffering great harm.” As acquired dispositions to act, feel, reason, and respond, the virtues (and vices) form a crucial part of the individual’s identity or sense of self. Further, in a way that Julia Annas has pointed out, given the differences in the circumstances of people’s lives, both the prominence and expression of particular virtues will be different. Hence, for example, a nurse’s courage will differ significantly from the courage of a firefighter (see Annas 2011, chapter 6). Although virtue theorists have increasingly turned their attention to the philosophy of education and moral psychology in their attempt to understand the nature of virtue and the process by which it is cultivated, the role of narrative—the story we tell of ourselves—in the development of character is only beginning to be explored. Finally, we should recognise a complex interplay in how a virtuous (or vicious) self figures as the author of a narrative. The author functions differently in different literary discourses. The author of novels engages readers differently than the short story author, and both engage the reader differently than the science textbook author. In any text, the author has their own perspective, and that perspective may be at least partly informed by their own virtuous or vicious character, as well as the virtuous or vicious characters of people who surround the author. Such a view has as its source the work of Michel Foucault:

Virtue, Narrative, and Self


We can conclude that, unlike a proper name, which moves from the interior of a discourse to the real person outside who produced it, the name of the author remains at the contours of texts—separating one from the other, defining their form, and characterizing their mode of existence. It points to the existence of certain groups of discourse and refers to the status of this discourse within a society and culture. The author’s name is not a function of a man’s civil status, nor is it fictional; it is situated in the breach, among the discontinuities, which gives rise to new groups of discourse and their singular mode of existence. (1969/1984, 113) Rich texts, such as novels, short stories, and essays, serve as a window into the author’s virtuous (or vicious) character. For example, witness Helen and Scott Nearing’s description of self-suffcient living in rural Vermont: We were in good health. We were solvent in that we had no debts. We were fairly hopeful of the future, but inexperienced in the ways of subsistence living and somewhat uncertain as to how we should proceed. After due consideration and in the spirit of the times, we drew up a ten year plan. (Nearing and Nearing 1954/1982, 29) It does not seem far-fetched to say that the Nearings exercised the virtues of humility, perseverance, and open-mindedness during their adventure in rural Vermont. Even the most sterile text, say that of a science textbook, may be riddled with the author’s virtuous (or vicious) character. For example, the author of a science textbook on evolutionary theory may lash out at Christianity in an otherwise hygienic scientifc text for giving a literal translation of the Bible. Narratives enable us to learn something about the authors themselves or the alter-ego that the author wants us to find in them. When we are confronted by a self-narrative, we have to decide whether the text is giving us an accurate picture of who the author is or it is giving us a picture of who the author wants us to think she is. This scepticism has led some to argue that the text and its author should be sharply distinguished. For example, Roland Barthes writes: [O]utside of literature itself (actually, these distinctions are being superseded), linguistics has just furnished the destruction of the Author with a precious analytic instrument by showing that utterance in its entirety is a void process, which functions perfectly without requiring to be filled by the person of the interlocutors: linguistically, the author is never anything more than the man who writes, just as I


Joseph Ulatowski and Liezl van Zyl is no more than the man who says I: language knows a “subject,” not a “person,” end this subject, void outside of the very utterance which defines it, suffices to make language “work,” that is, to exhaust it. (Barthes 1977, 144)

Unlike Foucault, Barthes wanted us to appreciate that the author is distinct from the person who wrote the text. Barthes felt that “To give a text an author is to impose a limit on that text, to furnish it with a fnal signifed, to close the writing” (Barthes 1977, 147). Barthes’ recommendation is for us not to read texts to uncover anything about the character of the author. Textual meaning is loosened, despite the authorial control we might otherwise think occurs in the writing process. We do not read texts to understand the author but to appreciate the text itself.3 Such instability in an author’s connection with her own writing has been amplifed by Jacques Derrida (2001) and Louis Althusser (1971).4 We imagine that none of the arguments we have just presented should be surprising to those who have worked in virtue ethics or virtue epistemology, but it may come as a surprise to those who work exclusively in topics of analytic metaphysics. Orthodox approaches to problems of analytic metaphysics only rarely consider anything beyond logic and philosophy of language. However, we foresee that discussions of virtue will expand further beyond the confines of the areas we have mentioned so far. For example, given some of the recent work in the history and philosophy of science, we believe that virtue theory will begin to make headway in discussions of scientific methods and their reliability, relevance, and effectiveness (cf. Pennock 2019). Let us now turn to an overview of each of the contributions to the volume and share how they fit the themes of this anthology.

4. Overview of the Contributions to This Volume Christine Swanton, in “Virtue Ethics and Narrative Virtue,” kicks off the volume by arguing that virtue ethics should move beyond basic virtue and develop an account of “narrative virtues,” that is, role differentiated forms of basic virtue, such as the kindness of a lawyer qua lawyer, or the generosity of a doctor qua doctor. As Anscombe pointed out, one of the advantages of virtue ethics is that it replaces the “thin” deontic concepts (“right” and “wrong,” “permissible,” “obligatory,” etc.) and the abstract moral principles favoured by modern moral philosophers with “thick” aretaic concepts and attention to the complexity of moral life. By evaluating an action as generous, for example, we do not merely approve of the action but also identify the features of the action that warrant approval (in this case, that it involves giving of one’s time or resources without demanding anything in return). Virtue theorists have given detailed analyses of specific virtues and vices, but for the most part they have focused

Virtue, Narrative, and Self


on the nature of what Swanton refers to as “basic virtue,” that is, “virtue understood at a high level of generality (e.g. kindness qua human being) or abstraction (kindness simpliciter).” She thinks we cannot settle debates about the requirements of specific virtues without considering the settings in which we act, such as cultural and historical location, role occupancy, and the narrative shape of our lives. Her focus in this chapter is on the narrative character of our ethical lives. Swanton draws on the work of Velleman, who defines “narrativism” as the view that “the meaning of a benefit depends on . . . the specific narrative relation between the goods and evils involved” (Velleman 2000, 63). She develops an Integrationist view about the link between basic virtue and narrative virtue, namely that the demands of ordinary morality and the complexities of social norms of narrative meaning and role demands are mutually constraining. Garrett Cullity further weaves together narrativity and virtues with second-order reasons, the content of which involves one’s having responded to first-order reasons, in “Narrative Virtues and Second-Order Reasons.” One approach to thinking about virtues conceives of them as dispositions to respond well to the reasons we have. This “reasonsresponsiveness” approach helps to illuminate what is distinctive about such virtues as loyalty and the kind of integrity that amounts to constancy in sustaining one’s allegiances to important goods. These virtues are ways of responding well to facts about the narrative shape that one’s relationships and allegiances have given to one’s life. Of particular interest are the forms that these narrative virtues take when they are responses to the way one’s life has been shaped by one’s previous reasons-responsive decisions. Narrative virtues with this feature are responses to secondorder reasons. Appreciating this helps us to distinguish different forms of loyalty and integrity from each other, to see the relationship between these virtues and the choices we face between the plurality of life-shaping goods, not all of which can be accommodated in a single life, to see how a virtuous sensitivity to the life-allegiances one has formed in resolving those choices need not be unduly self-regarding, and to appreciate what there is for someone who faces such choices to think about. Ramon Das, in his “Narrative Virtue Ethics?”, is more sceptical than Swanton and Cullity when it comes to the appreciation of narratives and whether they may be a source of normativity at all. More specifically, Das considers whether they are a serviceable substitute for dispositions as a normative basis for virtue ethics. The idea of a narrative has served a variety of philosophical purposes in the recent literature. It has been argued that narratives are an aid to understanding (De Bres 2018), constitute a distinct form of explanation (Velleman 2003), and are a source of meaning in life (Rosati 2013; Kauppinen 2015; De Bres 2018). Das uses as a foil a recent paper by Lisa Grover (2013), in which she argues that the “situationist critique” of virtue ethics can be avoided entirely by appealing to narratives, rather than dispositions as a basis for character


Joseph Ulatowski and Liezl van Zyl

traits. Even if narratives have the causative influence that Grover thinks that they have, they cannot provide the normative basis for right action that virtue ethics requires. In this respect, narratives do not serve as an improvement over dispositions because they cannot supply the requisite normative foundation for virtue ethics. While dispositions and narratives aim to provide a sound basis for character traits, character is ultimately unsatisfactory as a basis for right action. Das argues that right action must be sufficiently independent of character to allow for the possibility to act rightly (or wrongly) by acting out of character. Sometimes, persons with bad character nevertheless do the right thing and persons with good character nevertheless do the wrong thing. Insofar as the virtue-ethical appeal to narratives is in the end an attempt to provide a sound basis for character traits, Das argues, it cannot enable virtue ethics to meet this fundamental objection. While they appreciate Das’ claim that the nature of the virtues and their role in human action are controversial, in “How Self-Narratives and Virtues Cause Actions,” David Lumsden and Joseph Ulatowski wish to explore the thesis that virtues and self-narratives play a causal role in the production of action. One fruitful, though controversial, approach to understanding the nature of the self is through the notion of a narrative and in particular a person’s self-narrative or narratives. Lumsden and Ulatowski consider how virtues and self-narratives interrelate and, in particular, how they play a comparable role in the production of action. The basic ideas in the literature concerning reasons as causes of action provide us with a useful starting point even though the focus on reasons has tended to sideline potential causal roles for both virtues and selfnarratives. Without attempting to develop a new theory of causation, Lumsden and Ulatowski draw a picture of how virtues and self-narratives, in relation to each other, can be regarded as causally effective in producing an action. Not only ought we consider whether narratives play a causal role in the production of action but we should take up the question whether the narrative structure of an agent’s life influences the rightness of their actions. Jason Kawall, in “Virtue Ethics, Narrative, and Revisionary Accounts of Rightness,” does exactly this. Some virtue ethicists, most notably Rosalind Hursthouse and Daniel Russell, have proposed a revisionary account of right action, where “right action” is a matter of action assessment, and indicates that a given action is morally excellent and praiseworthy. The account emphasises both (i) an agent’s past and how she came to be in certain circumstances—is it a result of her own vice or wrong actions? and (ii) the agent’s own future happiness and well-being—will an action be so terrible that her life is marred and ruined? The narrative structure of an agent’s life thus plays a significant role in determining whether an action is right. Kawall thinks the revisionary account faces significant obstacles. In particular, he argues that non-virtuous agents can perform

Virtue, Narrative, and Self


actions that are far more praiseworthy and excellent than many of those characteristics of virtuous agents, even in circumstances that no virtuous agent would find herself in. Kawall also considers cases where virtuous agents face terrible circumstances through no fault of their own and must act in ways that will inevitably mar or ruin their lives. Hursthouse and Russell have argued that such circumstances make it impossible to perform a right, excellent action, given that it is a cause for sorrow and robs the agent of their ongoing peace of mind. Kawall thinks the revisionary account implausibly limits the scope for right action under difficult or oppressive conditions and concludes by arguing that it risks being excessively and implausibly egoistic by focusing on the peace of mind and flourishing of the agent herself in assessing the impacts of her actions. As we have seen, to be virtuous is to be constituted as a kind of rational agent, namely one who is exceptionally responsive to moral reasons. The virtuous person does not, however, learn to be responsive to moral reasons in a peculiarly moral way but rather she acquires a range of competencies associated with being a member of a community, such as the ability to recognise norms and conventions and to participate in conversation. In “Virtuous Perception: A Gibsonian Approach,” Richard Paul Hamilton argues that virtue represents a refined form of social competence. It proceeds from an unobjectionable assumption: the virtuous are exemplary because they are better able to reliably grasp the moral demands of a situation and act accordingly. We should follow these moral exemplars and recommend that others, particularly the young, emulate them. Hamilton begins by setting out Gibson’s original theory and introduces some objections and refinements to it. He then explores how the theory can be expanded to encompass social perception, as Gibson himself suggested (Gibson 1975/2015, 120). Broadly speaking, in Gibsonian terms, social perception involves detecting “affordances” presented by social settings, that is, seeing others’ actions as calling for us to act or refrain from acting in particular ways. Following on from this, Hamilton concludes that moral perception can be considered as a refined form of social perception. A Gibsonian view of perception makes this claim intelligible, protecting it against some common objections to direct moral perception. He links this Gibsonian analysis with Aristotle’s thoughts on training and habituation and Alfred Schutz’s insights regarding the habituated nature of mundane social knowledge. This is where he locates the idea of virtuous perception. He suggests, however, that times of moral crisis require more than merely habitual responses. The virtuous person has acquired the requisite skills to move beyond mere habit and is able to see more clearly than the non-virtuous how to resolve the situation satisfactorily. Hamilton concludes with a discussion of how the proposed approach addresses some common objections to naturalistic virtue ethics. We pivot now to three contributions whose main thrust is to better appreciate the application of virtue to the self and one’s self-narrative and


Joseph Ulatowski and Liezl van Zyl

how they make up the good life. Justin Oakley’s “Virtue Ethics, Blameworthiness, and Role Failure” focuses on the application of virtue ethics to problems in medical ethics, specifically the requirements of medical beneficence as a role-specific or narrative virtue. When doctors in nineteenth-century America instilled hope—false or otherwise—in dying patients, these doctors were acting in accordance with the advice of the American Medical Association code of ethics that prevailed at the time. The AMA no longer advises doctors to administer hope to dying patients, and contemporary doctors commonly regard the provision of false hope as unjustifiably paternalistic and judge practitioners as blameworthy for such behaviour. Virtue ethical approaches to medical practice typically evaluate such conduct as wrong because it is contrary to the role virtue of medical beneficence, and to the proper goals of medicine that this virtue aims to serve. Nevertheless, a nineteenth-century American doctor instilling false hope in their dying patient might seem to be less blameworthy than a contemporary doctor acting in such ways, given that this sort of conduct was being prescribed to nineteenth-century American doctors by their professional association at the time. This raises a more general question, which Oakley addresses in his chapter, namely whether virtue ethics is capable of accommodating judgements of diminished blameworthiness being made of role occupants who fail to act in accordance with the relevant role virtues, in circumstances where the professional, institutional, and regulatory environments they act in are not conducive to acting virtuously on those occasions? In “Well-Being, Narrative Value, and Virtue Ethics” Nicholas Ryan Smith argues in favour of a virtue-ethical account of what makes someone’s life good for them. First, he engages with the literature on the shape of a life and, following Velleman (2003) and Kauppinen (2015), argues that changes in the (dis)value of events and acts that are targeted by this literature are due to changes in narrative meaning rather than temporal location or order. Second, Smith outlines three constraints for a fully satisfactory account of well-being: adequacy in respect to narrative value, emotional fittingness, and practical reason. Finally, he argues that virtue ethics is superior to hedonism, desire-satisfaction theory, and standard objective list theory in respect to meeting these constraints. Emphasis is placed on how virtue ethics is better equipped to capture narrative value and disvalue than Kauppinen’s narrative calculus. The chapter by Damian Cox, “On the Value of Moral Failure,” further explores the role of narrativity in our conception of the good life. Cox puts forward the following thesis: moral success is sometimes born out of moral failure, and it is possible that a moral success that arises in this way can be of such great significance that it exceeds moral successes available to the agent without the failure. For example, if Oskar Schindler had not joined the National Socialists and had not made valuable contacts within it, it is highly improbable that he would have been able to

Virtue, Narrative, and Self


save so many lives. One should, in retrospect, be thankful for failures of this kind while also recognising them as failures worthy of remorse. Cox brings to light various ways in which moral failure is especially valuable. He uses his account of valuable moral failure to examine two contrasting forms that the moral assessment of a life may take. On the forensic model moral assessment of one’s life arises from a summative judgement of a person’s moral credits and debits, and on the narrative model moral assessment arises from the merit, success, and virtuous pursuit of significant moral projects. Cox argues that certain important kinds of valuable moral failure are best understood through a narrative model. This adds to the case for preferring the narrative model over forensic approaches to moral assessment. What Cox has shown is how an agent’s success may proceed from failure. The study of agency can involve many things including the ability to act intentionally, which itself leads to issues of moral responsibility but it requires postulating an agent, which could be identified with the self. Thus, a study of the self has the potential to illuminate some issues connected with agency while neglecting others. This broadly conventional approach tends to accept an agent as unified when agents act as they do within certain constraints (cf. Frankfurt 1988; Korsgaard 2008; Velleman 2000; and Williams 1981). For example, Michael Bratman explains agency on the basis of actions flowing from plans and policies that remain steadfast over time. For Bratman, figuring out what to do consists of designing, selecting, and deploying plans (Bratman 1987, 1999, 2001, 2007, 2009, 2014). Thus, one requirement of the conventional approach is the unity of the agent’s so-called psychic spine, something relatively stable, well structured, and action guiding. In his contribution, Elijah Millgram finds motivation in Nietzsche’s writings to rethink orthodox views of the unified agent. “Who Wrote Nietzsche’s Autobiography?” represents the continuation of a theme in Millgram’s recent work (e.g., Millgram 2014, 2015, 2019). He offers a Nietzschean anthropic argument for his version of the unity of agency. Agency is something akin to a spectral notion. An agent’s dominant priority usually falls somewhere between two extremes: the will to power, which is best described as a guide to action or choice, and amor fati, which is to accept things as they are. To not have will to power as a dominant priority is called “decadence” and the decadent cannot effectively control her own conative states. The decadent, therefore, is someone who has no priorities because the control structures that priorities presuppose are no longer in place. The will to power agent and decadent represent two unoccupied extremes of the spectrum. The Nietzschean anthropic argument shows us how someone who has his act together will not find the question of how we ought to live particularly urgent, but if one is a decadent, then that question is urgent given that the decadent is falling apart.

14 Joseph Ulatowski and Liezl van Zyl In “The Abnegated Self,” Nellie Wieland tells us about someone who has lost contact with her agency, either against or without her will, and who may be called a “self-abnegating person.” The loss of contact with her own agency does not mean that she cannot provide reasons for or a narrative about her actions. Quite the contrary, it’s just that those reasons and that narrative are someone else’s. People abnegate part of their agency regularly; for example, this is done within hierarchical institutions. This may be compared with what Millgram (this volume) has outlined as control and patterns of control in the priorities of the will to power agent. In other times, the self-abnegation is allencompassing; for example, this happens to victims of brainwashing. An agent in such a position can completely fail to understand herself or be understood as having a self. This can arise from internal and external forces. After describing the features of self-abnegation, Wieland concludes with a consideration of whether the self-abnegated person is owed respect. Finally, we turn to the question whether the messy lives we lead through the roles that we play need be absent of integrity and virtue in Tim Dare’s “Integrity and Messy Lives.” He notes that we should be surprised that Aristotle and Harry Frankfurt’s views of integration, integrity, and the self have been as influential as they have, since few of us obtain, or perhaps even aspire to, the peaceful coherence—or psychic harmony—of Aristotle’s virtuous person or to Frankfurt’s well-ordered self. Many, perhaps most, of us lead much messier lives than these views comfortably accommodate, and not because we are wicked or wanton. The messiness of our lives is explained at least in part by the inevitability that we will be forced to make choices, not merely within, but between, lives, and because we occupy social roles, the norms of which are likely to conflict with one another and with broader moral norms we endorse. However, what matters to integrity is not the successful integration of our desires and volitions, or the attainment of psychic harmony, but instead commitment to a certain kind of critical reflection and a willingness to embrace the recommendations of that reflection. So understood, integrity need not be denied to the many of us who live messy lives.

5. Conclusion Virtue theory will no doubt continue to expand into domains of philosophy beyond what we have described here and what the volume’s contributors exemplify. “The Virtue Turn” prompted by Anscombe in moral philosophy and Sosa in epistemology will not stop with philosophy of mind or philosophy of action. We envision that discussion of virtue may even enter debates in more formal areas of the philosophical enterprise, like logic and formal epistemology. We have defended expansion of virtuetheoretic talk into discussions of narrative and the self, and vice versa, in

Virtue, Narrative, and Self


this introduction. Our arguments have attempted to make broad connections between virtue, narratives, and the self, and the contributions—in one way or another—fill out each of these broad connections.

Notes 1. For more information on Gettier and responses to the infamous 1963 article, see Ichikawa and Steup (2017). 2. Heather Battaly has expanded work in virtue epistemology to include epistemic vices (2010, 2014; cf. Cassam 2019). 3. One may think that such interpretations of authorial intent only afflict continental philosophy. This would be a mistake. For example, resolute readings of Ludwig Wittgenstein recommend that we read the Tractatus LogicoPhilosophicus not to come to understand what nonsense is but to understand who he is. On a proper reading of Wittgenstein, so they claim, we are to read Wittgenstein to understand him and the kind of activity in which he engages. To read Wittgenstein is to evoke an inner change in the reader. The importance of Wittgenstein is therapeutic, and the point of understanding is inner change (cf. Conant 1990; Ulatowski forthcoming). 4. Space does not permit us to enter discussion with Derridean deconstruction or the poststructuralism of Althusser. Suffice it to say that we acknowledge the relevance of such material for endeavours such as we have embarked upon in this text.

References Althusser, L. 1971. “Ideology and State Ideological Apparatuses (Notes toward an Investigation).” In L. Althusser, ed., B. Brewer, trans., Lenin and Philosophy and Other Essays. New York: Monthly Review Press, pp. 127–186. Annas, J. 1999. Platonic Ethics, Old and New. Ithaca, NY: Cornell University Press. Annas, J. 2011. Intelligent Virtue. Oxford: Oxford University Press. Anscombe, G.E.M. 1957/2000. Intention. Cambridge, MA: Harvard University Press. Anscombe, G.E.M. 1958. “Modern Moral Philosophy.” Proceedings of the Aristotelian Society 58: 1–29. Badhwar, N. 2014. Well-Being: Happiness in a Worthwhile Life. Oxford: Oxford University Press. Barthes, R. 1977. “The Death of the Author.” In R. Barthes, ed., S. Heath, trans., Image Music Text. New York: Hill and Wang, pp. 142–149. Battaly, H. 2010.“Epistemic Self-Indulgence.” Metaphilosophy 41(1–2): 214–234. Battaly, H. 2014. “Varieties of Epistemic Vice.” In J. Matheson and R. Vitz, eds., The Ethics of Belief. Oxford: Oxford University Press, pp. 51–76. Baxley, A. 2010. Kant’s Theory of Virtue: The Value of Autocracy. Cambridge, UK: Cambridge University Press. Betzler, M. 2008. Kant’s Ethics of Virtue. New York: Walter de Gruyter. Bloomfield, P. 2014. The Virtues of Happiness: A Theory of the Good Life. Oxford: Oxford University Press. Bratman, M. 1987. Intention, Plans, and Practical Reasoning. Cambridge, UK: Cambridge University Press.


Joseph Ulatowski and Liezl van Zyl

Bratman, M. 1999. Faces of Intention. Cambridge, UK: Cambridge University Press. Bratman, M. 2001. “Taking Plans Seriously.” In E. Millgram, ed., Varieties of Practical Reasoning. Cambridge, MA: MIT Press, pp. 203–219. Bratman, M. 2007. Structures of Agency. Oxford: Oxford University Press. Bratman, M. 2009. “Intention, Practical Rationality, and Self-Governance.” Ethics 119/3: 411–443. Bratman, M. 2014. “Rational and Social Agency: Reflections and Replies.” In M. Vargas and G. Yaffe, eds., Rational and Social Agency: The Philosophy of Michael Bratman. Oxford: Oxford University Press, pp. 294–343. Carr, D. 1986. Time, Narrative, and History. Bloomington, IN: Indiana University Press. Cassam, Q. 2019. Vices of the Mind: From the Intellectual to the Political. Oxford: Oxford University Press. Conant, J. 1990. “Throwing Away the Top of the Ladder.” Yale Review 79: 328–364. Cureton, A. and T.E. Hill. 2014. “Kant on Virtue and the Virtues.” In N. Snow, ed., Cultivating Virtue. Oxford: Oxford University Press, pp. 87–110. Cureton, A. and T.E. Hill. 2018. “Kant on Virtue: Seeking the Ideal in Human Conditions.” In N. Snow, ed., The Oxford Handbook of Virtue. Oxford: Oxford University Press, pp. 263–280. Darwall, S., A. Gibbard, and P. Railton. 1992. “Fin de siècle Ethics: Some Trends.” Philosophical Review 101(1): 115–189. De Bres, H. 2018. “Narrative and Meaning in Life.” Journal of Moral Philosophy 15(4): 545–571. Dennett, D. 1991. “The Origin of Selves.” Cogito 2: 163–173. Dennett, D. 1992. “The Self as a Center of Narrative Gravity.” In F.S. Kessel, P. Cole, and D. Johnson, eds., Self and Consciousness: Multiple Perspectives. Hillsdale, NJ: Erlbaum Associates, pp. 103–115. Derrida, J. 2001. “The Deaths of Roland Barthes.” In J. Derrida, ed., The Works of Mourning. Chicago: The University of Chicago Press, pp. 31–68. Flanagan, O. 1996. Self-Expressions: Minds, Morals, and the Meaning of Life. Oxford: Oxford University Press. Foot, P. 2002. Virtues and Vices: And Other Essays in Moral Philosophy. Oxford: Oxford University Press. Foucault, M. 1969/1984. “What Is an Author?” In M. Foucault and P. Rabinow, eds., The Foucault Reader. New York: Pantheon Books, pp. 101–120. Frankfurt, H. 1988. The Importance of What We Care About. Cambridge, UK: Cambridge University Press. Gettier, E. 1963. “Is Knowledge Justified True Belief?” Analysis 23(6): 121–123. Gibson, J.J. 1975/2015. The Ecological Approach to Visual Perception. New York: Psychology Press. Ginet, C. 1990. On Action. Cambridge, UK: Cambridge University Press. Goldie, P. 2012. The Mess Inside: Narrative, Emotion, and Mind. Oxford: Oxford University Press. Goldman, A.I. 1967. “A Causal Theory of Knowing.” Journal of Philosophy 64(3): 357–327. Goldman, A.I. 1970. A Theory of Human Action. Englewood Cliffs, NJ: Prentice-Hall.

Virtue, Narrative, and Self


Goldman, A.I. 1979. “What Is Justified Belief?” In G.S. Pappas, ed., Justification and Knowledge. Dordrecht: D. Reidel, pp. 1–25. Goldman, A.I. 1986. Epistemology and Cognition. Cambridge, MA: Harvard University Press. Grover, L. 2013. “Narrative Virtues.” Theoretical and Applied Ethics 2(1): 67–82. Herman, B. 1996. “Making Room for Character.” In S. Engstrom and J. Whiting, eds., Aristotle, Kant, and the Stoics: Rethinking Happiness and Duty. Cambridge, UK: Cambridge University Press, pp. 36–60. Hursthouse, R. 1999. On Virtue Ethics. Cambridge, UK: Cambridge University Press. Hursthouse, R. and G. Pettigrove. 2018. “Virtue Ethics.” In E. Zalta, ed., Stanford Encyclopaedia of Philosophy (Winter 2018 edition). URL = . Ichikawa, J.J. and M. Steup 2017. “An Analysis of Knowledge.” In E. Zalta, ed., Stanford Encyclopaedia of Philosophy (Summer 2018 edition). URL = . Kamtekar, R. 1998. “Imperfect Virtue.” Ancient Philosophy 18(3): 315–339. Kauppinen, A. 2015. “The Narrative Calculus.” Oxford Studies in Normative Ethics, vol. 5. Oxford: Oxford University Press, pp. 196–220. Korsgaard, C. 2008. The Constitution of Agency: Essays on Practical Reason and Moral Psychology. Oxford: Oxford University Press. Kvanvig, J. 1992. The Intellectual Virtues and the Life of the Mind. Savage, MD: Rowman and Littlefield. Kvanvig, J. 2003. The Value of Knowledge and the Pursuit of Understanding. Cambridge, UK: Cambridge University Press. LeBar, M. 2009. The Value of Living Well. Oxford: Oxford University Press. MacIntyre, A. 1981. After Virtue. South Bend, IN: The University of Notre Dame Press. Millgram, E. 2014. “Segmented Agency.” In M. Vargas and G. Yaffe, eds., Rational and Social Agency: The Philosophy of Michael Bratman. Oxford: Oxford University Press, pp. 152–189. Millgram, E. 2015. The Great Endarkenment. Oxford: Oxford University Press. Millgram, E. 2019. John Stuart Mill and the Meaning of Life. Oxford: Oxford University Press. Nearing, H. and S. Nearing. 1954/1982. Living the Good Life. New York: Random House. O’Neill, O. 1996. Towards Justice and Virtue. Cambridge, UK: Cambridge University Press. Pennock, R.T. 2019. An Instinct for Truth: Curiosity and the Moral Character of Science. Cambridge, MA: MIT Press. Prior, W. 1991. Virtue and Knowledge: An Introduction to Ancient Greek Ethics. London: Routledge. Reshotko, N. 2006. Socratic Virtue: Making the Best of the Neither-Good-NorBad. Cambridge, UK: Cambridge University Press. Ricoeur, P. 1992. “Personal Identity and Narrative Identity.” In K. Blamey, ed., Oneself as Another. Chicago: The University of Chicago Press, pp. 113–139. Roberts, R. 2015. “Does Virtue Contribute to Flourishing?” In M. Alfano, ed., Current Controversies in Virtue Theory. New York, NY: Routledge. Rosati, C. 2013. “The Story of a Life.” Social Philosophy and Policy 30(1): 21–50.


Joseph Ulatowski and Liezl van Zyl

Schechtman, M. 1996. “Personhood and Personal Identity.” Journal of Philosophy 87: 71–92. Slote, M. 1992. From Morality to Virtue. Oxford: Oxford University Press. Slote, M. 2001. Morals from Motives. Oxford: Oxford University Press. Sosa, E. 1980. “The Raft and the Pyramid: Coherence Versus Foundations in the Theory of Knowledge.” Midwest Studies in Philosophy 5: 3–25. Swanton, C. 2003. Virtue Ethics: A Pluralistic View. Oxford: Oxford University Press. Swanton, C. 2011. “Nietzsche and the Virtues of Mature Egoism.” In S. May, ed., Cambridge Critical Guide to Nietzsche’s on the Genealogy of Morality. Cambridge, UK: Cambridge University Press, pp. 285–308. Swanton, C. 2015. The Virtue Ethics of Hume and Nietzsche. Oxford: WileyBlackwell. Taylor, C. 1964. The Explanation of Behaviour. London: Routledge and Kegan Paul. Thompson, M. 2008. Life and Action. Cambridge, MA: Harvard University Press. Ulatowski, J. Forthcoming. “Resolute Readings of Wittgenstein and Nonsense.” Journal for the History of Analytical Philosophy: 17 pp. Van Zyl, L. 2018. Virtue Ethics: A Contemporary Introduction. New York, NY: Routledge. Velleman, J.D. 2000. The Possibility of Practical Reason. Oxford: Oxford University Press. Velleman, J.D. 2003. “Narrative Explanation.” Philosophical Review 112(1): 1–25. Williams, B. 1985. Ethics and the Limits of Philosophy. Oxford: Oxford University Press. Wilson, G. 1989. The Intentionality of Human Action. Palo Alto, CA: Stanford University Press. Zagzebski, L. 1996. Virtues of the Mind: An Inquiry into the Nature of Virtue and the Ethical Foundations of Knowledge. Cambridge, UK: Cambridge University Press. Zagzebski, L. 2004. Divine Motivational Theory. Cambridge, UK: Cambridge University Press.


Virtue Ethics and Narrative Virtue Christine Swanton

1. Narrativism and Ethics A weakness of much contemporary normative ethical theory is that theorising is conducted at a very high level of abstraction and generality. As a consequence it is vulnerable to a complaint of Bernard Williams expressed thus by Elijah Millgram (2009, 142): the unintended lesson of Williams’ work is that we have made an astonishing mistake about who we are. The philosophical common sense of the past half-century has been suitable for impossibly simple-minded creatures, creatures competent to live only in impossibly simple environments. Consequently, the descriptive metaphysics and ethics that have been spun out of it are useless to us. It may be thought that virtue ethics with its apparent attention to the richness of our moral lives and language is immune from this criticism. This is not the case. For it generally speaks simply of virtue such as generosity simpliciter with little or no regard to the complexities afforded by the social facts that affect the nature of our generosity, facts such as the roles we occupy, the narrative shape of our lives in which “projects” have personal signifcance, and cultural norms. What is needed as both Williams and Alasdair McIntyre saw is attention to the context and settings in which we act: cultural and historical location, roles, and the narrative shape of our lives. Until now virtue ethics has largely focused on what I shall call basic virtues. By “basic virtue” is meant virtue understood at a high level of generality (e.g., kindness qua human being) or abstraction (kindness simpliciter) as opposed to the kindness of a lawyer or a doctor qua doctor or lawyer. Debates about for example whether the kindness of a doctor requires empathy with or detached concern for her patient (Halpern 2001) and how precisely we are to understand these possible requirements are underdetermined by appeal to basic virtue. Similarly, resolution of debates about what, on various views about social responsibility,


Christine Swanton

justice in business executives requires in relation to for example employees is left open. An ethics must appreciate that factors such as cultural location, role occupancy, and the narrative quality of our lives must form an integral part of an ethics if it is to apply to the real world. This chapter focuses on one class of social fact that structures our ethical lives: its narrative character. I show how virtue ethics and the notion of virtue can accommodate the truth of what has been called narrativism in our ethical theorising. Narrativism has been defined as the view that “the meaning of a benefit depends on . . . the specific narrative relation between the goods and evils involved” (Velleman 1991, 63). There are three aspects to narrativism in relation to ethics. 1. As Slote (1983) showed, the thesis of rejection of time preference is false. It can make a difference to the quality of our lives when an event occurs. He shows that the fact that a significant event occurs later rather than earlier can make a material difference to the goodness of a life. Call this feature “the narrativity of a life.” 2. What makes the difference to the goodness of a life, in respect of its narrative quality, is the significance or meaning of that event in relation to other events in that particular life. The fact that an event comes later may make that event redemptive. Alternatively, the fact that an event comes later may tarnish or even destroy earlier achievements. Call this feature “narrative meaning.” Narrative meaning is a broader notion than Williams’ (1973) “integrity.” At the core of integrity is connection with one’s sense of identity through commitment to and identification with personal projects or causes. But one may support a cause which has benefited someone one loves; one need not be committed to it as a personal project. Or a supported charity may simply be local: one knows about it, and one is focused on local matters. Integrity and connection with one’s sense of identity through identification is just one feature that may give something narrative meaning. 3. That the significance of an event affects the goodness of a life depends not just on its narrative meaning or its significance in relation to other events in a life but on its ethical features as understood in relation to a substantive view of ethics, for example some version of virtue ethics. Thus an event in a life may tarnish former events in that life because it is disgraceful and it is disgraceful because it exhibits egregious betrayal or dishonesty.

Virtue Ethics and Narrative Virtue


Call this feature “narrative ethical meaning.” What difference does the recognition of narrativity make to ethics? How, in short, can we construct a narrative ethics, and in particular a virtue ethics, sensitive to the narrative quality of our lives? More specifically, what is the relation between narrative meaning and narrative ethical meaning within a virtue ethical framework? This question raises the broader issue of the relation between social norms of various kinds and ethical norms. If we are to take Williams’ worry seriously, we need to recognise that social facts such as narrative meaning, cultural location, practices and beliefs, and role occupancy impact on our ethical lives. Narrative meaning may permit actions which may not be permitted from an impartial perspective. Role obligations may require actions that may not be permitted absent the role. Cultural norms may require actions not otherwise required. There are three basic approaches to the problem of the relation between social norms of this kind and ethical norms. The first I call the “Ordinary Morality” view. The common feature of such views is that they downplay the place social norms have on our ethical obligations. Our ethical obligations are entirely dependent on the norms of ordinary morality. What is required by ordinary morality will of course depend on one’s theoretical perspective. On this view there are no robust role obligations for example. Role obligations, virtues, and permissions are just “complex instances of ordinary morality” (Andre 1991, 73). Reasons to act as one’s role requires are basically reasons of ordinary morality: if ordinary morality prohibits something then one cannot take refuge in one’s role. In relation to narrative ethics, on the Ordinary Morality view the narrative shape of a life does not affect one’s ethical obligations. An example is the “effective altruism” movement. In charitable giving on this view one must always do the best one can from an impartial perspective, at least up to a certain limit. One does not have permission to favour charities that have narrative meaning for one. The second approach is at the other extreme. This approach I call “Reductivist.” Here ethical requirements are reduced to relevant social norms: those for example of cultural practices, role demands, and narrative meaning. In relation to culture, this position leads to cultural relativism; in role ethics, we have the view that roles go “all the way down” (cf. Dare 2020). On this view role obligations do not depend on a background normative ethical theory for their content and normative force (Dare 2020, 43). In narrative ethics we have the idea that we have narrative autonomy over our lives. A classic source of this view is the wellknown section 290 of The Gay Science, where Nietzsche (1974) claims that one must ‘“give style’ to one’s character” by fitting strengths and weaknesses of one’s nature “into an artistic plan.” Here it appears narrative virtue is differentiated not by appeal to the constraints of ordinary morality let alone impartialist morality. Rather it is virtue differentiated


Christine Swanton

according to the individual particularities of narrative meaning, especially a person’s individual mode of creativity in providing meaning to her life. As Nietzsche claims, there are constraints of style, and “‘giving style’ to one’s character” is done “under a law of [one’s] own.” The third approach, the one I favour, I call the “Integrationist Thesis.” This is the view that the demands of ordinary morality and social norms such as those of narrative meaning and role demands are mutually constraining. These social norms are taken seriously as norms that impact on ethical obligations and permissions. On this view cultural imperatives imposed on you by your belonging to a culture, role obligations imposed on you by your occupancy of a role, the narrative meaning of an action in the context of narrative features of one’s life often succeed in giving you reasons for action, reasons you would not have had without the cultural and role requirements or narrative meaning. The social norms of narrative meaning, role demands, and so on play a fundamental role in determining one’s ethical obligations. They are part of ethics, but for these norms to be genuinely ethical they must be constrained by the broader ethical norms of basic virtue. 1.1 Differentiated Virtue and the Integration Thesis As far as narrative ethics is concerned this chapter resolves the problem of the relation between the second and third feature of the narrative character of our lives (narrative meaning and narrative ethical meaning) by providing a version of the Integrationist Thesis within a virtue ethical framework. Here narrative virtue is understood as virtue that is both shaped by narrative meaning and constrained by the basic virtues of ordinary morality. Central to the Integrationist Thesis is an account of differentiated virtue: that is basic virtue “differentiated” according to for example norms of narrativity, role demands, and cultural imperatives. What is a differentiated virtue and what is the relation between basic and differentiated virtue? A differentiated virtue is a virtue that incorporates legitimate norms of the relevant institution, practices, or other social facts such as narrative meaning. Thus in role ethics the institution in which the role is embedded must pass ethical muster, and within that institution the roles themselves must serve the worthwhile goals of that institution. In narrative ethics narrative ethical meaning cannot be given by an agent’s identification with a completely vicious ideology which provides overarching narrative meaning to everything he does. The agent’s actions will have narrative meaning but not narrative ethical meaning, for they will not conform to narrative differentiated virtue. What is the relation between differentiated and basic virtue? A differentiated virtue has basic virtue at its core. Basic virtues provide anchors for our moral thought even when contoured by role demands, narrative

Virtue Ethics and Narrative Virtue


autonomy and integrity, and cultural imperatives. In this way, they provide constraints on the unfettered pursuit of institutional (in the case of role ethics) and personal goals and projects (in the case of narrative ethics) alerting us to possibilities of excess and other forms of wrongness when determining the scope of differentiated virtue. For example, zealous advocacy in a defence lawyer will be distinguished from what has been called hyper zeal (Dare 2004); narrative integrity and pursuit of one’s creative goals may permit differentiation of a basic virtue such as caring (for one’s family say) in a talented artist but not to the point of total neglect and abandonment; the hospitality required in a culture will not permit lavishness to the point of endangering the welfare of the group as a whole in future lean times. Such anchors are deeply embedded dispositions which provide constraints on contouring action in line with, say, role demands. Though accounts of basic virtues abstract away from role differentiation for example, then, they are not empty. The realisation that basic virtue anchors forms of differentiation reduces the potential for conflict between the demands of roles, narrative integrity, and cultural imperatives on the one hand, and “ordinary morality” conceptualised through basic virtues, on the other. Differentiation is not a form of unconstrained cultural demands, role demands, and narrative autonomy. It will not lead to the Reductivist approach. At the same time differentiation puts constraints on the unfettered pursuit of basic virtue in two ways. It will not lead to the Ordinary Morality approach. First, basic virtues leave courses of action indeterminate until differentiation by for example role requirements kicks in. A CEO cannot simply be generous with the firm’s funds treating them in accordance with Kant’s maxim of beneficence which allows latitude in accordance with inclination. On the contrary, she has to conform to constraints put in place by the board, the firm’s interests, and overseas bosses. Until a basic virtue is “contoured” by contextual features such as a person’s role, historical, and cultural location, we characteristically cannot form an accurate conception of the target of that virtue in a particular context. Second, the constraints imposed by the relevant social norms have to be taken seriously as constraints. In a context of role occupancy, the target of the basic virtue of generosity cannot be determined by a maxim of impartialist ordinary morality such as “do the best you can (with the firm’s money).” In these ways differentiation will not lead to the Ordinary Morality approach: ordinary morality does not have unfettered sway. I have basically described the Integrationist Thesis in a virtue ethical form. To summarise: according to this thesis the ethical norms of basic virtue and the social norms of e.g., narrative meaning are mutually constraining, resulting in conceptions of differentiated virtue. That is, basic virtue is differentiated by the relevant social norms to form a virtue sensitive to the ethical pull of those norms. In this way there is no subsumption


Christine Swanton

of narrative ethics within the norms of “ordinary morality” enshrined in basic virtue such as generosity simpliciter. By the same token there is no reduction of the norms of ordinary morality (basic virtue) to those of narrative meaning. If narrative permissions were based entirely on narrative meaning there would be no constraints enshrined in ordinary morality on the unfettered pursuit of what makes life meaningful for one, and indeed no apparent theoretical ethical base for the ethical critique of factors that make that life meaningful from a personal point of view.

2. Narrative Virtue Here I develop a conception of narratively differentiated virtue (narrative virtue for short) based on the distinction between basic and differentiated virtue. Broadly speaking narrative virtue is basic virtue differentiated by normative features pertaining to the narrative shape of one’s life. Narrative shape is given by an active shaping of one’s life in line with intentions and projects; in other words one’s life has significance and meaning by virtue of (inter alia) events in that life bearing narrative relations in that life (as opposed to say episodic satisfying of preferences or pleasurable experiences). The notion that the targets of virtues such as generosity and loyalty are differentiated by narrative features is grounded in the idea that a life has direction and shape constituted by narrative relations. The narrative ethics proposed then presupposes the truth of narrativism. Narrativism must be distinguished from two stronger theses, the thesis of the narrative constitution of the self, and an even stronger thesis that personal identity is determined by a form of psychological continuity—namely narrative continuity. The idea of narrative virtue assumes the truth of narrativism, but not the truth of the two stronger theses. Nonetheless I find plausible the thesis of the narrative constitution of the self, and one version of that thesis (which I also find plausible and whose truth I shall assume), that of the disunity of that constitution, poses a problem for traditional virtue ethics. According to this disunity thesis there is no overarching single narrative which unifies what Lumsden and Ulatowski (2017) call the various “narrative threads” that constitute the self, such as the different roles we occupy, the various careers and projects we have, our status as retired or worker. I shall later discuss the narrative virtues required for dealing with narrative disunity (Section 3). To investigate the special problems posed by narrativity we need to say more about the nature of narrative virtue. Narrative virtue may be divided into two broad types which I label global and specific. 2.1 Global Narrative Virtues These are virtues whose fields are the narrative qualities of one’s life as a whole, for example its overall goodness, and appropriate integrity,

Virtue Ethics and Narrative Virtue


coherence, and truthfulness across the disparate narrative threads. These virtues are interrelated and may be grouped according to what I shall call virtues of narrative accuracy, virtues of narrative integrity, virtues of narrative transition, virtues of narrative goodness, and virtues of narrative flexibility. Virtues of narrative transition, flexibility, integrity, and goodness will be discussed in Sections 4–5. Virtues of narrative accuracy enable one’s narrative self-conception to satisfy what has been called the accuracy constraint (Schechtman 1996) which is one of the norms of narrative meaningfulness (the second feature of narrative character in relation to ethics mentioned in Section 1). This constraint does not require accuracy in all one’s beliefs; it is rather a reality constraint in the sense that one’s narrative self-conception is not distorted by paranoid and egotistical fantasy, for example, resulting in major core falsehoods that impact on important features of that conception. 2.2 Specific Narrative Virtues The second broad way in which a life exhibits narrative virtue is that it exhibits basic virtues in general, but basic virtues that are differentiated by the narrative shape of one’s life. The fields of specific narrative virtue do not pertain to the shape of one’s own life as such, or aspects of that shape, but specific kinds of things in one’s life that need to be virtuously handled or responded to. These are the normal fields of a wide range of virtues: dealing well with danger or threat (courage); the welfare of others (compassion, generosity); delays, waiting and delayed gratification (patience); and so on. Hence the specific narrative virtue of generosity is generosity differentiated by the narrative shape of one’s life, the specific narrative virtue of courage is courage differentiated by the narrative shape of one’s life and so on. For example, the expectations of courage shown when rescuing people drowning in a riptide will vary according to whether one is a lifeguard or whether those threatened with drowning are one’s children. Notice then that narrative ethical meaning within the narrative threads, such as the roles one occupies, will be determined not just by narrative differentiation. The way that narrative meaning is expressed in virtuous behaviour will also be determined by role differentiated virtue for example. In short, given narrative threads what counts as conforming to narrative virtue is complex, for conforming to such virtue is determined by other forms of differentiation such as cultural and role differentiation. This suggests that the Integrationist Thesis applies not just to the relation between differentiated virtue in general and basic virtue but to the various forms of differentiated virtue. Norms of narrative meaning applying to the narrativity of a life must be integrated with norms of say role differentiated virtue if a narrative virtue is to be a genuine virtue.


Christine Swanton

In the next two sections I discuss two global narrative virtues which pose interesting problems for a narrative ethics: narrative goodness and virtues dealing with narrative disunity. 2.3 Narrative Goodness The global narrative virtue of narrative goodness requires that overall, one’s life must exhibit to a sufficient degree specific narrative virtue across the various narrative threads of that life. A virtue where narrativity has received philosophical attention is that of beneficence, particularly in relation to charitable giving, so I shall illustrate narrative goodness with the specific narrative virtue of beneficence. A good place to start in a discussion of the narrative dimensions of beneficence is Kant’s duty of beneficence since that duty has important similarities to narratively differentiated beneficence and has the same interesting problems. Not only is Kant’s duty of beneficence strongly virtue theoretical, but the duty has latitude in much the same way as the narrative virtue of beneficence has. Here are the chief features of Kant’s duty of beneficence. First, the duty of beneficence is a duty of virtue and of love. It is a duty of virtue in that it prescribes in its maxim an obligatory end: namely adopting the happiness of others as an end. It is a duty of love in that in its most abstract form it requires a “coming close.” Love is described by Kant as a “moral force” of “coming close,” as opposed to respect (a “keeping distant”) (Kant 1996, 6:449). In relation to beneficence love is “practical” love, not necessarily involving pleasure, delight, or affection (Kant 1996, 6:450). Indeed Kant even suggests that our species may not even be particularly lovable (Kant 1996, 6:402). Rather this form of love is an “unselfish benevolence” (Kant 1996, 6:401). Hence one violates the duty if through misanthropy for example one avoids others “(separatist misanthropy)” (Kant 1996, 6:402). Second, the duty of beneficence is a “wide” duty. According to O’Neill (2013, 116) Wide duties are duties to have certain ends; their wideness consists in the fact that often a variety of means to any ends is available to an agent, none of which is mandatory. In such cases an agent has some latitude in doing his duty. . . . Wide duties are expressed in maxims of ends. Kant describes feature 2 of the duty of benefcence thus: I ought to sacrifice a part of my welfare to others without hope of return, because this is a duty, and it is impossible to assign specific limits to the extent of this sacrifice. How far it should extend depends, in large part, on what each person’s true needs are in view

Virtue Ethics and Narrative Virtue


of his sensibilities, and it must be left to each to decide this for himself. .  .  . Hence, this duty is only a wide one; the duty has in it a latitude for doing more or less, and no specific limits can be assigned to what should be done—The law holds only for maxims, not for determinate actions. (Kant 1996, 6: 393) Third, notwithstanding the latitude involved in the duty of benefcence the duty requires that one perform benefcent acts on those occasions that demand such acts (due to urgency of need, proximity of the agent, etc.). I have claimed that narratively differentiated beneficence should be seen as having the same kind of latitude as Kant’s duty of beneficence. As a result problems posed by Kant’s duty of beneficence are worth discussing in the context of a conception of narrative goodness. The first problem relates to feature 3 of Kant’s duty of beneficence. According to this feature some beneficent acts are obligatory. This is controversial since it may be thought incompatible with Kant’s idea that beneficence is a duty having latitude (Stohr 2011). In short, Kant is thought to be faced with the following dilemma: (a) Beneficence is a wide duty. (b) Some beneficent acts are obligatory. However, the claim that “no specifc limits can be assigned to what should be done” does not imply that benefcence is everywhere optional. Rather Kant is claiming that specifc limits cannot be described in advance by a maxim or principle of action in terms of which the duty is framed. As O’Neill (2013) claims often a variety of means, none of which is mandatory, is available to an agent to further an obligatory end. This claim is compatible with the view that there are or could be specifc benefcent actions which are in the circumstances required of one. And indeed, there are cases where not being benefcent betrays egregious failure to act from practical love since such failure would be a mark of profound indifference, malice, callousness, or “separatist misanthropy” through hatred for example. We are not required by duty to have certain feelings (such as affection for our fellow humans) since this cannot be willed for Kant, but to act in a way which is describable as callous, indifferent, and so on is to act contrary to (practical) love. In this real world (b) is true since some acts are mandated by the general requirement to act from practical love. Here is a second problem. Kant makes it look as if latitude is governed simply by personal inclination. But what rationalises the latitude and what constraints if any are there on what each may “decide for himself?” Here I think a conception of narrative virtue can resolve the problem


Christine Swanton

in two steps. First, the ethical significance of narrativity rationalises the latitude since narrative meaning is particular to each agent. Kant speaks of “inclination” but what makes one inclined to give to one charity rather than another may well be the narrative significance of that charity in one’s life. Second, the idea of narrative virtue resolves the problem of ethical limits on inclination. What constraints if any are there on what each may “decide for himself?” Can one act well or badly in exercising that latitude or is there complete narrative autonomy? The Integrationist Thesis posits ethical constraints on the latitude rationalised by the ethical significance of narrative meaning. Narrative specific virtue is subject to the constraints of the relevant basic virtue. Notwithstanding latitude some acts of beneficence are required by narratively differentiated beneficence on the Integrationist view. This leaves us with the possibility that some beneficent acts, rationalised by personal narrative meaning, are criticisable though not prohibited because for example overly stingy or ineffective. One could have done better. Such acts are what are now called “suberogatory.” An example is given by McMahan (2018, 78) concerning the will of Leona Helmsley. Given the billions at stake the will is an example of highly ineffective altruism when considerably greater good could have been served by the bequest. If narrative autonomy in philanthropic bequest is unconstrained by the virtue requirements of basic charity then we could say that she has a right to leave her money to whomever she pleases, and that is all there is to the matter. On the other hand, if narrative virtue allows for ethical constraints then we can say that though she has a right, and has acted within her rights, she has acted impermissibly or badly in exercising her right the way she has. The beneficence has been too ineffective, or, if we follow Kant’s idea that beneficence is a duty of love, her beneficence, such as it is, betrays a lack of practical love towards humans even though the will meshes with what has given narrative meaning to her life. The concepts of virtue ethics have no problems with the idea that rights belong to a separate area of ethics: the point of them is to prohibit certain sorts of interference with actions within one’s rights even though they might be wrong. Nonetheless there will be much room for controversy about whether philanthropic acts within one’s rights conform to a narrative virtue of beneficence or not. It will be controversial whether acts are impermissible or merely suberogatory by the standards of ordinary morality (which are themselves controversial) when factored into norms of narrativity. Other acts within the area of latitude may not be required but are particularly meritorious. These are supererogatory. My account of the permissions of narrative virtue will allow for both supererogatory and suberogatory actions. To show this we need a brief foray into my target

Virtue Ethics and Narrative Virtue


centred virtue ethical account of right action (Swanton 2003, 2018, 2020, forthcoming). According to this view: An action is right if and only if it is overall virtuous, and an act is overall virtuous if and only if it hits the targets of relevant virtues to a sufficient extent. Hitting the targets of the virtues is what Aristotle calls hitting the “mean”: “virtue aims to hit the mean” (Aristotle 1976, 1106b16–24). Since the mean of a virtue constituting its target is multi-dimensional what counts as hitting the target is subject to incommensurability: one may act for the right reason but to the wrong extent; one may act to the right extent but in the wrong manner and so on. The mean is only partially hit. So what counts as overall virtuous in the circumstances depends on what dimensions of the mean are salient in the circumstances, degree of closeness to the target on any given dimension, how the various dimensions are weighted in that circumstance, and what virtues are in play. As a result of these features rightness suffers from both degree and combinatorial vagueness (see Swanton forthcoming, ch. 10). To account for the suberogatory we also need an account of wrongness. Following the general parameters of my target centred account of rightness Stangl (forthcoming) offers an account of wrongness according to which an act is vicious in respect of a virtue of generosity for example (it is stingy say) if it constitutes a serious failure to hit the target of that virtue. Thus, if a stingy act is not a serious failure to hit the target of generosity it is not vicious in that respect. An act is wrong if and only if it is overall vicious. My account differs from Stangl’s in that I would say that vicious actions constitute a sufficiently serious failure to hit the relevant target(s); they miss the targets of the relevant virtues to a sufficiently serious extent. The threshold for what counts as sufficiently vicious to be called overall vicious and thereby wrong tout court will vary according to context. Although I agree with Stangl that mundane acts can be suberogatory there are contexts where we might properly call a mundane undesirable act wrong. (Perhaps there has been a good reason to strive for greater virtue in this context.) It would be odd in these cases to describe such an act as a serious failure to hit the mean tout court. We can now understand supererogation and suberogation in the light of the target centred accounts of rightness and wrongness. Supererogatory acts are overall virtuous and are compatible with the existence of other overall virtuous actions available to the agent that are less meritorious (but still permitted) than the one performed. Hence, we have the common idea that supererogatory actions go beyond the call of duty. Suberogatory acts are acts that fail to hit salient dimensions of the mean,


Christine Swanton

salient dimensions of the mean sufficiently well, too few dimensions of the mean, and so on but this failure is not sufficiently serious as to make that act wrong because overall vicious. There are acts available to the agent that are better (that is closer to the targets of relevant virtues) than the one performed. Suberogatory acts are not meritorious (they are bad or undesirable to varying degrees) but still permitted. They are not so bad as to be wrong. They are forms of permissive ill doing but are not overall vicious. Indeed, such acts may be mundane: a minor donation to a charity may be somewhat stingy but does not constitute a failure to hit the mean that is sufficiently serious to be classed as wrong.

3. Virtue and Narrative Disunity Let us consider a problem for a virtue ethics which incorporates narrativity into its conception of virtue. Recall that according to the disunity thesis there is no overarching single narrative which unifies what Lumsden and Ulatowski (2017) call the various “narrative threads” that constitute the self. Now the disunity thesis apparently poses a problem for virtue ethics in that virtue is classically held to be both robust and stable over a mature life as a whole. Disunity may be thought to preclude what Chappell (1998, 106) and others term narrative integrity: To be a person is to live a life, and lives can have more or less integrity or order to them: part of living well is to live a life that has narrative sense. Virtue in relation to the disunity constraint raises questions about how a life with disunifed threads can be stable in a way required by virtue. On my view the notion of narratively differentiated virtue can solve the problem. More particularly it is solved by the distinction between global and specifc narratively differentiated virtue. It turns out that global narrative virtues show how disunity—a quality of the narrative shape of one’s life as a whole—can in various ways be virtuously handled. The cardinal virtues in this area are a virtue of narrative integrity, (overall) narrative goodness, and narrative virtues of transition and flexibility. Where one believes that the narrative shape of a life is constituted by disunified threads a good way to understand the virtue of narrative integrity is to think of it as a disposition to lead a life of appropriate disunity. A virtue of appropriate disunity is a virtue possessed by an agent who is disposed to lead a life that is neither so disunified that she cannot handle it or so unified that her life lacks variety or challenge. Where the mean lies in this respect will vary according to the individual and can be subject to controversy. Narrative goodness is constituted by dispositions to manifest basic virtue albeit narratively differentiated within the various narrative threads of one’s life (such as one’s life in each of one’s roles)

Virtue Ethics and Narrative Virtue


to a sufficient overall level to be deemed virtuous overall. The disunity thesis does not suggest that one can be overall virtuous if one is generous as a friend but mean as a colleague; the requirement of overall narrative goodness may be violated. Narrative virtues of transition consist of dispositions to transition relatively smoothly from one narrative thread to another (where possible) (Lumsden 2013). Smooth transitions may not always be easy or possible, such as with gender or racial transitions. For, such transitions may be ethically fraught or thought to be so, causing problems at several levels (Tuvel 2017). Here one avoids vices of excess and deficiency. A vice of deficiency is a tendency not to transition readily enough. For example one brings one’s work home like the teacher who treats his family as students, lecturing them regularly. Another vice of deficiency is a tendency to compartmentalise oneself too radically. Or one forgets or neglects some of one’s roles allowing one to dominate one’s life. At the extreme one’s life is unified to the point of narrative vice. (Consider the butler in The Remains of the Day, Ishiguro 1989.) A vice of excess is a tendency to transition too readily; for example while a university teacher one is overly ready to neglect academic demands in favour of a welfare role. One has a weak sense of the demands of one’s role. Finally, related to virtues of appropriate disunity and transition, are those of appropriate flexibility: dispositions to allow one’s life to be sufficiently flexible in accommodating appropriately several narrative threads. Narrative flexibility is an important virtue when one is juggling several roles: Perhaps perfectionist tendencies have to be mitigated, one has to be creative in accommodating multiple demands that are difficult to fulfil, one has to manage stress. Flexibility may seem to work against integrity, but so does rigidity. Further, some narrative virtues of flexibility may be targeted at dealing with narrative disruption such as trauma, retirement, forced relocation. To conclude: relevant global narrative virtue notably those of narrative goodness, virtues of transition and flexibility, and a virtue of appropriate disunity permit a life with disunified threads to be virtuous.

4. Narrativity and Virtue Ethics Let us now show how narrative virtue satisfies the three features of a narrative virtue ethics outlined in the opening paragraph. First, narrative virtue accommodates narrativism—it does justice to the fact that our lives extend over time in such a way that it can make a difference when in one’s life an event has occurred. Second, narrative virtue recognises the ethical significance of narrative meaning. The narrative quality of a life enshrines relationships of significance; it is not just quanta of pleasure and pain that has ethical significance for example. Narrative meaning is a norm governed notion: there are norms of narrative accuracy, coherence,


Christine Swanton

and integration. Such norms provide the contours for narrative differentiation of virtue. Third, a narrative ethics provides an account of narrative ethical meaning in the form of accounts of narrative virtue and narratively differentiated permissions and requirements. The Integration Thesis demands that narrative virtue is constrained by basic virtue for ethically acceptable expressions of narrative meaning. Norms of accuracy, coherence and so forth, features which provide narrative meaning, are not as such narrative virtues and thus are not as such constrained by norms of basic virtue. What are the practical implications of this virtue ethical view? If one takes narrative differentiation seriously then according to the Integrationist Thesis an important result occurs. Basic virtue is constrained by narrative virtue. For example, one will be permitted by virtue to refrain from or even fall well short of, maximising the good. A person’s philanthropy or charity may be shaped by narrative concerns, for example the roles specific local organisations have played in her personal life, such as helping one’s family or oneself through illness. Again, by contrast to one’s relations to a colleague one may be much more patient with the foibles of a friend who has played a significant role in the life of oneself and family. Because basic and narrative virtues are mutually constraining not all forms of giving narrative meaning to one’s life will be ethically acceptable. Not only is basic virtue constrained by narrative differentiation, permitted narrative differentiation is constrained by basic virtue. Forms of excess and deficiency are not mandated or permitted by narrative virtue. Another important implication is the interpenetration of narrative virtue with other forms of differentiation of virtue resulting in considerable legitimate variation in the way these virtues are manifested. Consider for example the global narrative virtue of narrative integration. In his The Subject of Virtue Laidlaw cites an example of the relation between the disparate narrative threads of the lives of gays in Indonesia. He claims that “the social scenes and arenas of gay life in Indonesia are not closeted or hidden, but they are distinct times and spaces, discontinuous for the most part from family and work” (Laidlaw 2014, 79). Integration of narrative threads or lack of it may be heavily dependent on role and cultural norms. The way the global narrative virtue of narrative transition is manifested will vary considerably according to culturally differentiated virtue. This chapter has discussed narrative ethics within a virtue ethical framework. A central theme in virtue ethics is what it is to “live well.” Indeed, much debate centres on what is required for living well in relation to the narrative quality of a life. Living well however is an ambiguous notion. In one sense a life lived well is a life that is good for the agent. In another sense a life lived well is a life that is good from an ethical point of view. Virtue ethics is an ethics, not as such a theory of agential

Virtue Ethics and Narrative Virtue


well-being. Hence a virtue ethical theory that does justice to narrativism is a theory that incorporates the narrative quality of a life into its virtue ethical view of the good life in the second sense. As I have argued before (2003), this sense is also to be distinguished from the successful and the admirable life. A life manifesting narrative as well as other forms of differentiated virtue need not be admirable; it may be quite mundane. Nor need it be successful. Such a life may be marred through ill luck or other cause of failure. Such failure may satisfy norms of virtue, indeed it may be honourable. Certainly, as is well known there are Herculean attempts to yoke together the notions of the good for (an agent) and the ethically good. For example in eudaimonist thought there is the idea that to be a virtue a trait must be characteristically good for the agent (see Swanton 2003, 2018). This chapter makes no such assumption. In summary, virtue ethics needs to move well beyond the virtues of orthodox virtue ethics. These are what I have called basic virtues which enshrine the tenets of ordinary morality in the language of virtue. It needs to theorise about the relationship between basic virtue and such pervasive features of our lives as role occupancy, narrative particularity, and cultural location. A classic problem is the relation between these complexities and “ordinary morality.” My view of this relation is a species of Integrationist view which relies on the distinction between basic and differentiated virtue. On this view the constraints of basic virtue are at the core of differentiated virtue, thereby reducing (but not eliminating) the potential for conflict between (differentiated) virtue and ordinary morality. Narrativity raises many interesting problems concerning the scope and nature of both global and specific narrative virtue. I have discussed but two narrative virtues: narrative goodness in relation to the narratively differentiated specific virtue of beneficence and virtues whose field is narrative disunity.

References Andre, J. 1991. “Role Morality as a Complex Instance of Ordinary Morality.” American Philosophical Quarterly 28: 73–79. Aristotle. 1976. Nicomachean Ethics, J.A.K. Thomson, trans., H. Tredennick, eds. Harmondsworth: Penguin. Chappell, T.D.J. 1998. Understanding Human Goods: A Theory of Ethics. Edinburgh: Edinburgh University Press. Dare, T. 2004. “Mere Zeal, Hyper-Zeal and the Ethical Obligations of Lawyers.” Legal Ethics 7: 24–38. Dare, T. 2020. “Roles All the Way Down.” In T. Dare and C. Swanton, eds., Perspectives in Role Ethics: Virtues, Reasons and Obligation. London: Routledge, pp. 31–44. Halpern, J. 2001. From Detached Concern to Empathy: Humanizing Medical Practice. Oxford: Oxford University Press.


Christine Swanton

Ishiguro, K. 1989. The Remains of the Day. London: Faber and Faber. Kant, I. 1996. The Metaphysics of Morals, M. McGregor, trans. and ed. Cambridge, UK: Cambridge University Press. Laidlaw, J. 2014. The Subject of Virtue: An Anthropology of Ethics and Freedom. Cambridge: Cambridge University Press. Lumsden, D. 2013. “Whole Life Narratives and the Self.” Philosophy, Psychiatry, and Psychology 20(1): 1–10. Lumsden, D. and J. Ulatowski. 2017. “One Self Per Customer? From Disunified Agency to Disunified Self.” The Southern Journal of Philosophy 55(3): 314–335. McMahan, J. 2018. “Doing Good and Doing the Best.” In P. Woodruff, ed., The Ethics of Giving: Philosophers’ Perspectives on Philanthropy. Oxford: Oxford University Press, pp. 78–102. Millgram, E. 2009. “D’Ou Venons-Nous . . . Que Sommes Nous . . . Ou AllonsNous?” In D. Callcutt, ed., Reading Bernard Williams. London: Routledge, pp. 141–165. Nietzsche, F. 1974. The Gay Science, W. Kaufmann, trans. New York: Vintage/ Random House. O’Neill, O. 2013. Acting on Principle: An Essay on Kantian Ethics, 2nd ed. Cambridge: Cambridge University Press. Schechtman, M. 1996. The Constitution of Selves. Ithaca: Cornell University Press. Slote, M. 1983. “Goods and Lives.” In M. Slote, ed., Goods and Virtues. Oxford: Clarendon Press, pp. 9–37. Stangl, R. Forthcoming. Neither Heroes Nor Saints. Oxford: Oxford University Press. Stohr, K. 2011. “Kantian Beneficence and the Problem of Obligatory Aid.” Journal of Moral Philosophy 8(1): 45–67. Swanton, C. 2003. Virtue Ethics: A Pluralistic View. Oxford: Oxford University Press. Swanton, C. 2018. “Eudaimonistic versus Target Centred Virtue Ethics.” Teoria 2: 43–54. Swanton, C. 2020. “Expertise and Virtue in Role Ethics.” In T. Dare and C. Swanton, eds., Perspectives in Role Ethics: Virtues, Reasons and Obligation. London: Routledge, pp. 45–71. Swanton, C. Forthcoming. Target Centred Virtue Ethics. Oxford: Oxford University Press. Tuvel, R. 2017. “In Defence of Transracialism.” Hypatia: A Journal of Feminist Philosophy 32: 263–278. Velleman, J.D. 1991/2000. “Well-Being and Time.” In J.D. Velleman, ed., The Possibility of Practical Reason. Oxford: Oxford University Press, pp. 56–84. Williams, B. 1973. “A Critique of Utilitarianism.” In B. Williams, ed., Utilitarianism: For and Against. New York: Cambridge University Press, pp. 75–150.


Narrative Virtues and SecondOrder Reasons Garrett Cullity

One approach to thinking about virtues conceives of them as dispositions to respond well to the reasons we have. According to this way of thinking, facts about the way things are and the way things could be provide you with normative reasons bearing on what you do, think, feel, and say. Those reasons are objective: you have them whether you are actually aware of them or not; and if you are mistaken about them, you need not be at fault for that. One kind of judgement that we can make concerns the overall balance of those reasons: which responses are best supported by them, overall. Another concerns the goodness or badness of how you respond in the light of the information available to you about what your reasons are. Judgements of rightness and wrongness belong to the first of these kinds; judgements of virtue to the second.1 As I see it, the case for this way of thinking about virtues comes from the illumination it can provide in helping us to understand various features of our normative and evaluative thought, and their relationship to each other. In particular, it can help us to understand the structure of the aretaic landscape: we can cast light on the relationships between different virtues by examining the different reasons to which those virtues respond. The aim of this chapter is to contribute to that case by examining the distinctive structure of what I shall call “narrative virtues.” Any response, and hence any virtuous response, to a reason is a temporally extended episode, and there is therefore a sense in which it has a narrative structure: there is the onset of the reason-giving fact, its recognition by you, your response to it, the impact of that response on others, and the way you react to that. More substantially, the attribution of a virtue to a person involves a generalisation about the pattern of responsiveness they sustain over time, in terms of which we understand their biography—the shape of the life they are leading. Every virtue is connected to narratives in these two ways. However, there is a further connection that some virtues have to narratives and others do not. Some are virtues of responding well to facts about the narrative structure of one’s own life: I reserve the name “narrative virtues” for these. When this narrative structure is one that I have shaped myself, we can have the following phenomenon. When


Garrett Cullity

I have responded to some reason-giving fact F throughout my life, then that fact—the fact that I have responded to F—can itself supply me with a reason to which I can respond well (that is, virtuously) or not. When a narrative virtue takes this form, the reasons to which it responds are constituted by narrative facts about one’s having responded to other reasons. These narrative facts are second-order reasons—reasons the content of which includes one’s having responded to other reasons. Two main examples of narrative virtues will be discussed in what follows. One comes from virtuous forms of loyalty; the other from a form of integrity that I call “integrity-as-constancy.” We come to these in Sections 5 and 6. The first four sections of the chapter set up a general framework for understanding the structure of the narrative virtues, the goods they help us to realise, and the second-order reasons to which they respond. In the last two sections (7 and 8), I draw out some implications of the treatment of narrative virtues that result from applying this framework, showing how this can help us to appreciate their special importance.

1. Reasons-Responsiveness To explain what is involved in conceiving of virtues as good forms of reasons-responsiveness, we can start with the example of kindness. Someone with this virtue responds well to other people’s welfare, in a range of ways: by being attentive to it, being perceptive about it, acting beneficently to further it, speaking up for its importance, and having feelings of sympathy and empathy. In order to be virtuous, these responses need to be discriminating, and non-accidental. A virtuously kind person sees facts about others’ welfare as counting in favour of the responses she makes, and when we admire her as virtuous, we are agreeing with her: these facts are indeed good reasons for responding in these ways. To turn this into a fully adequate account of virtue, many things need to be added. First, we need to allow for faultless mistakes about the objective facts: virtue, more carefully speaking, is a matter of responding well to the apparent reasons, not necessarily the actual ones.2 Second, virtuous responses need not be preceded by episodes of deliberative thought; so “reasons-responsiveness” has to be understood in a way that does not require that. Making a response for the reason that R requires that one is sensitive to the evidence that R, that one sees R as a reason for one’s response, and that one’s making the response is non-accidentally guided by these states of evidence-sensitivity and reason-attribution; however, the requisite forms of evidence-sensitivity, reason-attribution, and guidance need not involve any conscious reflection.3 Third, virtue requires responding well in all of the various dimensions in which one’s responses can be assessed—with respect to the time, the occasion, one’s manner, and so on.4 Fourth, when one responds to a reason, whether one does so well will depend on all of the various other reasons that bear on one’s

Narrative Virtues and Second-Order Reasons


response in the circumstances. Fifth, if holists about reasons are right that the reason-giving status of a fact can vary from one context to another, then virtue requires being appropriately sensitive to that.5 And sixth, one’s response need not be perfect in order to be virtuous: the many reasons bearing on how we conduct ourselves generate many ways in which one’s responsiveness to them can be better or worse, and therefore more or less virtuous. Filling in this outline properly, of course, requires a lot of work. Each of the six points I have just noted needs to be more fully developed and defended: I do not attempt that here.6 The overall task for proponents of the reasons-responsiveness conception of virtue is to account convincingly for the full range of virtues. What I attempt here is only one part of this larger task—giving an account of the narrative virtues.

2. Narrative Virtues A narrative, I take it, is a description of a sequence of events that shows the significance of their relationship to each other. There are therefore ways in which all virtues are connected to narratives. If you need help and I provide it, my action changes the narrative sequence of events: it becomes one in which your need is met with kindness. However, I reserve the name “narrative virtues” for those that are ways of responding well to a prior narrative to which one already belongs. In a narrative virtue, the reason by which the virtuous response is guided is a fact about the role that the response will play as a constituent in a pre-existing narrative. That is not true of the case of kindness just described. The reason to which the kind action responds is simply that you need help, not the self-regarding fact that my own action will help to constitute a narrative of kindness.7 Two narrative virtues will be the main focus of attention in what follows. One comes from those forms of virtuous loyalty that respond well to the history of a valuable relationship of special connection. The other comes from one of the virtues we refer to as “integrity.” My previous actions have given my life the distinctive shape it has had up until now. It has been shaped by the responses I have previously made to the reasons I then encountered. The fact of my having made those responses can now provide me with a reason to continue this narrative in a way that makes sense. However, the class of narrative virtues extends more broadly than these two. Other examples come from virtuous forms of apology, forgiveness, and solidarity. If I have wronged you, my apology can be an acknowledgement of the breach that exists between us, an acceptance of responsibility for repairing that breach, a declaration of willingness to do so, and a renunciation of any claim against you to repair it. These facts about my act of apologising can make it a vehicle for repairing the


Garrett Cullity

relationship between us. And that can provide me with a reason for apologising. In apologising, I change the narrative structure of our interaction from one in which I have breached our relationship to one in which I have breached it and then offered to repair the breach. In the same vein, the fact that your declaration of forgiveness will be an acceptance of my offer, transforming the situation into one in which we have both resolved to restore the relationship, can provide you with a reason to forgive. In both cases, a fact about the narrative significance that an action will have is a reason for performing that action—a reason to which the virtuous agent is responsive. Another example of this is provided by the reasons there can be to bear witness to a violation that one is powerless to remedy. Protesting at an unjust killing may do nothing to lessen the wrong that was done, to help the victim, or to bring the perpetrator to justice; but it can change the sequence of events into one in which an injustice has been opposed. Again, this fact about the narrative significance of the act of protest can be a reason for performing it and responding well to this reason can make one’s solidarity virtuous. Here, though, the focus will be on examining the virtues of loyalty and integrity, their common features and differences. To appreciate what makes these narrative virtues of special interest, there are two other ideas we need to introduce. The first is a thesis about the plurality of goods to which these two virtues are importantly related. Then I need to explain how responding to some of these goods rather than others can provide us with second-order reasons.

3. Berlinian Choices Of the various claims that travel under the label of “pluralism” in ethics, one of the most prominent is a thesis about value associated with Berlin. As he maintains, life presents us with forced choices between a plurality of different goods, which admit of no fully satisfactory resolution. Faced with a diversity of genuine goods, it is impossible to align ourselves with all of them fully: we must choose, and when we do so there is inevitably a loss. Choices of the kind that concern him cannot be treated as simple ties, like the predicament of Buridan’s Ass, in which we can arbitrarily pick one of two equally valuable alternatives without reasonably regretting that we have had to forgo the other. In contrast, when I face the choice between structuring my life around close family relationships or filling it with novelty and variety, or when we together face the choice between structuring our society around the values of freedom or equality, these are not properly treated as straightforward choices between goods of equal value, where either alternative is a fully satisfactory replacement for the other. We must choose, but whatever we end up choosing cannot adequately compensate for what we have forgone: whichever way we go, there is a loss, as well as a gain.8

Narrative Virtues and Second-Order Reasons


Berlin’s value-pluralism is naturally paired with a parallel thesis about reasons. There can be situations of the following kind. Sometimes, you can face a choice between A and B, where both A and B are important, but you cannot attain both of them together. One set of (one or more) reasons favours A; another, different set of reasons favours B. In situations of this kind, two things can be true. First, you can have decisive reasons to choose A-or-B rather than neither, but the reasons favouring A do not decisively outweigh those favouring B, and nor do the reasons favouring B decisively outweigh those favouring A. But secondly, whatever you choose, there is a loss: if you choose A, it is fitting to have negative attitudes towards your not having chosen B, and positive attitudes towards others’ choosing B; if you choose B, it is fitting to have negative attitudes towards your not having chosen A, and positive attitudes towards others’ choosing A. “Berlinian pluralism about reasons” is the name I shall give to the conjunction of these two claims. In situations with these features, I shall say that you face a “Berlinian choice.” (For clarity, I shall reserve talk of your “choice” for the range of alternative options between which you must choose; when you choose one of those options, I shall call that your “selection.” A selection need not be the outcome of a conscious decision.) Berlinian choices provoke an awkward question. When I face a choice of this kind, what is there for me to think about in deciding what to do? The corresponding question has a straightforward answer when my situation is like Buridan’s Ass, and the alternatives are equally desirable. Then I have no reason for preferring one alternative to the other: I must simply pick, without having any normative reason favouring the alternative I pick. But in a Berlinian choice, the alternatives are not equally desirable. The reasons favouring one alternative are not of exactly equal strength to those favouring the other. However, nor is either set of reasons decisively stronger.9 So how can I proceed? I can think as much as I like about the reasons favouring each alternative: but how can that help me to reach a decision? Should I just arbitrarily pick, as if my situation were like Buridan’s Ass? But how could that be a serious way of dealing with an important life-decision? This question needs an answer: I return to it later. Having noted it, we should now turn to another question. We do in fact manage to make selections when we are confronted with Berlinian choices. What is the significance of our having made them?

4. Allegiances and Second-Order Reasons Suppose I face a choice between helping one person and keeping a promise to others. I have promised to attend a committee meeting to organise a campaign against a local real estate development; on my way there, I see someone standing beside her broken-down vehicle, looking forlorn.


Garrett Cullity

She is not flagging me down, and although the road is not busy nor is it especially remote: there will be other motorists along shortly who will be able to help if I do not. But she does look as though she needs some help. We could continue this story in a way that makes it obvious what I should do. (The motorist is in tears, has a young child with her, and I can phone in my apology to the meeting. . . . I am chairing the meeting which I have pressed other people to attend, do not have a phone, and the motorist just looks bored, not upset . . .) However, there are versions in which this is not obvious. Then I could face a Berlinian choice—not a momentous or tragic one; but one that will leave me with something to regret, whatever I do. If I get to my meeting on time, there will be something regrettable about having ignored the needy motorist—something that cannot neatly be resolved by appealing to the appointment I have kept. (Suppose that later I meet the motorist by chance at a social event. Suppose I tell her that I saw her stranded beside the road. How do I continue that conversation? I can point out that I had a meeting to get to: should I expect her to find that satisfactory?) And if I stop to help the motorist, when someone else could have stopped instead, I owe the other committee members an apology. (It is an apology, rather than a fully satisfactory justification. I cannot present the situation as an emergency without stretching the truth.) The conflicting reasons in this situation are very straightforward. There is the fact that the motorist needs help (call this fact F) and the fact that I promised to attend the meeting. I have to act on one or the other: suppose I stop to help, acting on F. This new, further fact—the fact that I have responded to F in helping her—is itself normatively significant. Now, if the motorist and I later meet at a social event, this fact about what I have previously done makes our interaction different from meeting a stranger. The interaction we now have is part of a history that commenced earlier and can be continued in better or worse ways. She should be grateful for having been helped by me; but I should also be sensitive to the burdensomeness of obligations of gratitude and avoid smugness or the impression of patronage. What has significance here is not just my having performed the action of helping, but the reason to which I was responding in doing so. It matters greatly whether I did so for amusement, or under orders, or in order to attract attention to myself, or just because she needed it. The fact that I acted from the latter reason, F, is important. This further fact is a reason bearing on my future actions: a second-order reason, in the sense that its description contains an essential reference to the first-order reason F. With this in mind, we can turn to weightier cases. Suppose now that my choice is one with a more life-shaping significance. I am deciding whether to leave my close friends and elderly parents in Armidale to pursue a job opportunity in Brisbane. Again, this could be a Berlinian choice:

Narrative Virtues and Second-Order Reasons


one in which I lack decisive reasons to choose either alternative, and either way there will be some loss. But this time, there is a further aspect to the choice I face. Two values are in competition: the values of my close relationships and my career achievements. I am not being asked to betray one for the sake of another: I do not have an obligation to my parents to stay in Armidale with them, and nor will I have to make a fundamental sacrifice of my career aspirations if I do. But these are two goods around which a well-lived life can be structured—life-shaping goods, we can call them—and I am being required to choose between them. Given my circumstances, I have to compromise my engagement with one of them for the sake of the other: I must decide which of them to orient myself towards more fully. When I face a Berlinian choice with respect to lifeshaping goods, we can call it a “life-shaping choice.” It is important not to be misled by this label. When you face a lifeshaping choice, the selection you make need not be an irrevocable, life-defining watershed decision from which there is no turning back. If relocating to Brisbane turns out to be unsatisfying, I can return home again. If I stay in Armidale and end up regretting the lost opportunity, I can look around for another one. So the selection I make when faced with a choice between life-shaping goods need not end up actually reshaping my life fundamentally. Nonetheless, it remains true that the selection I have made in a situation like this has a significance that bears on the actions that come afterwards. It would make no sense to move to Brisbane, then immediately to settle into a routine of doing the minimum necessary to retain my job and spending the rest of my time having internet conversations with my parents and friends. The reason against doing that is again a second-order reason: the fact that I relocated for the sake of my career. The fact that I am here as a result of the action I took, for the first-order reason I did, is itself a reason not to act in that way.10 Why is that so? One thing that might have happened is that by moving to Brisbane, I have changed my circumstances in a way that has altered the balance of first-order reasons. My new situation may now be one in which my time would be better spent pursuing the excellent career opportunities wholeheartedly, then visiting my family and friends on holidays. I may no longer be facing a Berlinian choice: now that I am in Brisbane, the first-order reasons for devoting energy to my career might have become decisively stronger. However, if that is so, it is so independently of how I came to be in my current situation. The first-order reasons governing what I now do are the same as they would have been had I come to Brisbane through some conjunction of events over which I had no control. But surely how I have come to be in Brisbane is normatively significant. If I came to Brisbane for the sake of my career, that is an additional reason to be taking my career more seriously. I think this is to be explained as follows. There are various possible ways of structuring one’s life that amount to orienting oneself towards


Garrett Cullity

something good. Sharing the affection of friends and family is one of them; devoting oneself to a worthwhile career is another. To use Brentano’s language, these are ways of loving the good; and as such they are themselves good.11 When one sustains a pattern of orienting oneself, in action and/or attitude, towards a larger life-shaping good, we can call that an “allegiance” to that good. A person’s life can be better for including family relationships or a worthwhile career without an allegiance to them—an unappreciative member of a happy family may still be better off than he would be in an unhappy one—but fully realising the value of these life-shaping goods requires embracing them, and forming an allegiance. When I face a Berlinian choice between two competing lifeshaping goods and decide in favour of one of them, I only fully attain the good for the sake of which I am acting if I sustain a subsequent allegiance to it. In moving to Brisbane for the sake of my career, I am acting in a way that requires an allegiance in order to make full narrative sense. This is another way in which my circumstances have changed. It is a respect in which my circumstances are now different from the ones I would be in if I had been brought to Brisbane by events beyond my control. The fact that I arrived here through acting for the sake of my career is a secondorder reason because it now situates me in a new narrative, the successful development of which requires me to sustain an allegiance to the good for the sake of which I have acted. This gives me a reason to continue on the path I have embarked on.12 All Berlinian choices come with an associated cost—the cost constituted by the loss incurred in forgoing the alternative that is not taken. Life-shaping choices come with a further cost—the burden of realising the alternative good for the sake of which I have incurred that loss, by sustaining a successful allegiance to that good. The allegiance can be provisional, it is true. The reason I have to make a success of my allegiance is a pro tanto reason, which can be outweighed. Things may not work out for me in Brisbane: the career for the sake of which I moved may not be as fulfilling as I had hoped; the firm might restructure and eliminate my job. But the fact that, by acting on the first-order reason I did, I have created a narrative structure the value of which is realised by sustaining an allegiance is itself a reason to continue to direct my life in a way towards doing that.13 This explains how life gets in one way more complicated as you get older, but in another it can become simpler. It gets more complicated because as your life-narrative proceeds, you acquire further second-order reasons, given by the further reasons-responsive decisions you have made. However, in another respect, this can make it simpler. It can become simpler because the acquisition of these further reasons can give you a clearer (if not an easier) path forward. The Berlinian choices you faced when you were younger can cease to have that character now. Earlier, you faced a choice between different life-paths, none of which was decisively

Narrative Virtues and Second-Order Reasons


superior. Having embarked on one of them, you may now have acquired decisive reasons to continue on it.

5. Loyalty Equipped with this framework, we can now examine the narrative virtue of loyalty. “Loyalty” is our name for the virtue of responding well to the history of a valuable relationship of special connection.14 The connection in question can be to a person, a family, an institution, a group, a practice, or a tradition. It requires more than just having a close relationship that creates reasons of personal partiality: the dependence of my own new-born child on me creates such reasons, but it would be odd to talk of loyalty to a new-born infant.15 Loyalty requires not only a relationship of special connection, but that that relationship has a history. Loyalty involves seeing the fact that there is a valuable history of special connection between me and X as a reason for responses of action, attention, and feeling towards X, and responding well to that reason. So it qualifies as a narrative virtue. Among the various forms that loyalty can take, we can make a distinction. Suppose I appreciate the way in which the institutions of Australia have looked after me and my family. I recognise how these institutions have been shaped by the efforts of good people over generations, guided by a common set of values and shaped by an awareness of their contribution to a common project. This makes me proud to call myself an Australian, and if I were called on to fight to defend its people and values, I would do so. But suppose I have not been called on to do that. Indeed, suppose I have never done anything to shape my life as an expression of this allegiance. I have obediently paid my taxes and voted, like everyone else; but I have not shaped my life in any overt way to distinguish myself from a fellow citizen who found his country and all it stood for oppressive and distasteful. My attitudes would still suffice for one kind of loyalty. And this kind of loyalty, passive though it is, would (assuming my country merits it) amount to a narrative virtue. There are facts about my life-narrative—what I have received from my country, and the contribution this has made to my life-story—to which I am responsive in the attitudes I have towards my country, and my dispositions to serve it in extremity. We can call that uninvested loyalty. In uninvested loyalty one is responsive, in one’s attitudes and dispositions, to a history of special connection that has been shaped by others; but one has played no part in shaping it oneself. This is to be contrasted with invested loyalty, in which the history of my special connection to X has been shaped in some part by my own previous responsiveness to the reasons bearing on my relationship to X. When a history of special connection is shaped by my own reasonsresponsive actions, then the reason to which I am responding when I treat


Garrett Cullity

the fact that I have this history as significant is a second-order reason. All forms of virtuous loyalty exhibit a narrative virtue; invested loyalty is a virtue of responsiveness to a second-order reason. All forms of loyalty are allegiances; but only invested loyalties involve allegiances of an actively life-shaping kind. Uninvested loyalty towards a person is possible: you might have appreciatively loyal dispositions towards a benefactor you have never met. But typical loyalties to persons are invested. They develop through the accretion of second-order reasons on a prior foundation of responsiveness to first-order reasons. This is the usual trajectory of loyalties within families. One interacts with members of one’s family from childhood onwards; and all being well, these include interactions that are good. There is a history that includes consideration, cooperation, forgiveness, and respect, both given and received. Acting on those first-order reasons is not yet a manifestation of loyalty, though. Loyalty comes later, when you face choices between acting to express your family allegiance or not. It is when you catch a plane to get to the annual family gathering that you are acting from loyalty: now you are responding to the history of connection between yourself and your family that is constituted in part by your own actions when you were younger, performed for other, first-order reasons. Thus, all loyalties, whether invested or uninvested, are responses to a narrative; invested loyalties are those in which the narrative is shaped by one’s own reasons-responsive action. An uninvested loyalty is an allegiance: as soon as you act to express that allegiance in response to a Berlinian choice, large or small, your uninvested loyalty becomes an invested one. One might worry that this way of thinking about loyalty gives us too self-regarding a picture of the motivation of a loyal person. It may seem to locate the object of my loyalty in the wrong place. When I have a loyal friendship to Maynard, this account may seem to imply that the object of my loyalty is the effort I have put into my relationship with Maynard, rather than Maynard himself.16 However, this invites two responses. First, the challenge here is a general challenge for any attempt to provide a good account of loyalty. Since loyalty pertains only to those relationships of special connection with a history, any account of it faces the challenge of how to specify the connection that loyalty has to such relationships without mislocating its object. But second, the distinction our account makes between first- and second-order reasons provides the means of solving this problem: the problem is solved, not created, by our account, as follows. Suppose Maynard is my friend and Barbara is a stranger. My loyalty to Maynard could be expressed through simple actions of beneficence: actions motivated by recognising his welfare as a first-order reason for helping him. The same first-order reason, it is true, is available for helping Barbara instead: helping her might further her welfare just as much. So suppose we ask, “Why help Maynard rather

Narrative Virtues and Second-Order Reasons


than Barbara?” The succinct answer is: Because Maynard is my friend. This implicitly cites a second-order reason. My friendship with Maynard is our sharing a relationship whose history is constituted by our acting from concern for each other’s welfare. Maynard’s being my friend gives me a reason for helping him rather than Barbara because it locates my action within a valuable ongoing relationship. This is not to say that I lack a reason to help Barbara: on the contrary, I have the same first-order reason to help her. But I have an extra second-order reason to help my friend Maynard. The value of my relationship to him gives me a reason to sustain that relationship, and this is how I can do so. However, sustaining it requires that I help him for his sake, rather than for the sake of the relationship. The relationship is constituted by actions performed for the first-order reason. This is why my loyalty is properly described as being to Maynard, rather than to my relationship with him. It is expressed in actions directed towards his welfare. But what makes his welfare stand out as having a special claim on my attention is the history of our relationship. And my responsiveness to that is what makes my beneficent treatment of Maynard into an instance of loyalty. A further distinction is worth making. The success of my efforts to sustain a life-shaping allegiance is never solely up to me. I am always exposed to the possibility that future events obstruct me. The dance career I commit myself to may be ended by injury; the rural hospital I labour to build might be destroyed in war. But the success of my allegiances of loyalty are especially vulnerable. The success of my friendship with Maynard depends on him as well as me. This draws our attention to another, unhappier form of loyalty which can still be valuable. Ideally, in an invested loyalty there is a two-way relationship: the investment you make is reciprocated. But sometimes it is not. Even then, there can still be a reason for turning up to the awkward family gatherings, persevering with the loveless marriage, and singing the anthem of the country that is being ruined by a kleptocratic government. Loyalty under these circumstances, standing by one’s unreciprocated investment, can still make sense as a stubborn, ongoing statement of one’s allegiance to an ideal that has not been realised.

6. Integrity-as-Constancy “Integrity” is the name of several distinguishable virtues. A reasonsresponsiveness view of virtue helps us to distinguish them. Commonly, talk of integrity refers to moral fortitude. This is not a matter of acting for a particular kind of reason, but of living up to and speaking up for one’s moral commitments when there are strong incentives that make this costly in some way: it is a matter of not being deflected by those incentives—of responding well to the temptation to overrate their reasongiving strength.17


Garrett Cullity

However, there is another, different, good quality which it is natural to call by the same name. This is the quality of sustaining one’s allegiances— the quality lacked by someone who begins by embracing a life-shaping good, but then backtracks or loses interest. One way in which this might happen is through a failure of integrity of the first kind. I might idealistically commit myself to being an aid worker, or an environmental campaigner, or a parent, and then lose interest when the scale of the challenge I have taken on becomes apparent, settling for a more comfortable existence instead. But that need not be so. Another possibility is that of fortitude without constancy. I might always stand up for what is right despite the discomfort it involves, speaking out and making sacrifices— but do so serially, moving from one worthy cause to another. Faced with Berlinian choices between the many important goods that need to be actively defended, I might flit inconstantly from one to another while relishing the challenge of fighting for each, simultaneously displaying a high threshold for social discomfort and a low threshold for boredom. Then my life would display something valuable: namely, the fortitude involved in not being deterred from defending what is good. But it would also lack something valuable. In trying to stand up for too many things serially, overall I would stand for nothing in particular. Given the Berlinian choice between all of the competing goods around which to structure a human life, I might fail to form a stable commitment to any of them. My life would then lack a kind of unity or constancy that it makes sense to speak of as a lack of integrity. This gives us two distinguishable kinds of integrity: integrity-asfortitude and integrity-as-constancy. The latter, like loyalty, is a narrative virtue. My previous actions have given my life the distinctive shape it has had up until now. It has been shaped by the selections I have made in responding to some reasons rather than others. My having made those previous responses now provides me with further reasons—second-order reasons—to act in the ways that give my life coherence as an expression of the allegiances I have formed to life-shaping goods. Integrity-asconstancy is the narrative virtue of responding well to these reasons. It is a sensitivity to what I stand for—to my life’s significance as an expression of commitment to the values around which it is structured. (Taking a cue from Berlin, a similar point can be made about the political culture of a country. Polities face Berlinian choices too—between giving priority to the values of freedom and welfare-protection, for example. These choices, with their attendant gains and losses, have been resolved differently by different political communities. And those facts then provide current members of these communities with reasons that bear on the political decisions they make now. Americans and Swedes are part of different political projects: the projects of finding the best way of taking forward a political narrative that has declared its primary allegiance to the preservation of freedom and the protection of welfare, respectively.)

Narrative Virtues and Second-Order Reasons


Invested loyalty, as a sustained, active allegiance to a person, family, institution, group, practice, or tradition, is the most obvious vehicle for integrity-as-constancy. A “serially loyal” person would lack the quality we admire in loyalty, and the deficiency that deprives her of virtuous loyalty would be a deficiency in integrity-as-constancy. However, loyalty is not the only vehicle for integrity-as-constancy, since the allegiances of loyalty—allegiances sustained through the history of one’s relationship to a person, a family, an institution, a group, a practice, or a tradition— are not the only allegiances there are. Consider a roving freelance warcorrespondent, always moving on to the newest conflict, sustained by the vocation of bearing witness to the realities of warfare, and embracing the rootless life that this entails. A life of this kind, shaped by a strong allegiance to an important vocation and set of values, could be one of great integrity-as-constancy, but one that is unbound by personal loyalties (unless we stretch “loyalties,” unhelpfully, to cover all allegiances). Integrity-as-constancy, as I conceive of it, involves responding well to the shape my life has attained through my own previous responses to lifeshaping choices. It can take either an unreflexive or a reflexive form. In its unreflexive form, it simply amounts to embracing the goods to which I have formed allegiances and sustaining a constancy in those allegiances in the face of challenges. What is also possible, though, is having a conception of my own identity as someone who, faced with a choice between different possible allegiances, has shaped his life by forming the allegiances he has, and has acted to sustain this. Integrity-as-constancy takes a reflexive form when, in acting to sustain my allegiances, I am responding to that fact. In this reflexive form, it is actually responsive to thirdorder reasons. My children’s interests provide me with first-order reasons to help them; my having taken on the commitments of parenthood provides me with second-order reasons to stand by that commitment; and then, when I am aware that in acting to stand by that commitment I have given my life a shape in virtue of which it has the significance it does, allying me to one set of values amongst other alternative possibilities, I am recognising a third-order reason to sustain it in a way that preserves that significance. Does this recommend a self-centred approach to the values to which I profess an allegiance? No. A challenge of this form was addressed in connection with loyalties in the previous section; the corresponding challenge in connection with other allegiances invites a corresponding reply. Of the war correspondent (whose allegiances are not loyalties), we can ask two questions. One is, “Why endanger yourself to publish the truth about a war?” An answer to this will cite first-order reasons: it is a way of bearing witness to the victims’ suffering and of serving justice. If we then ask, “Why not drop this work, settle down somewhere, and raise a family?” he can answer this by saying: this is the allegiance I have chosen; these are the values around which my life is structured. These are second-order


Garrett Cullity

reasons—facts about the reasons to which he has already responded. They do not supplant his first-order reasons. On the contrary: they are reasons for continuing a narrative constituted by his responsiveness to the firstorder reasons. His integrity is not a way of caring about himself instead of caring about justice: it is his responsiveness to the (second-order) reasons he has to continue acting out of a concern for justice. It might still seem that the reflexive form of integrity-as-constancy is something that only a theorist of virtue could ever have. However, that is not so: recall Section 1. To be responding to my own allegiance-shaped identity as a reason for acting as I do, my action must be guided by my being sensitive to such facts and seeing them as reasons; but that does not require formulating explicit thoughts with these propositional contents. To say that integrity-as-constancy can take a reflexive form does not require that it must involve conscious reflection, any more than other virtues. I close this discussion of integrity by noticing that it has a further important form—one that cannot be identified with either integrity-asfortitude or integrity-as-constancy, although it contains elements of both. This is the quality possessed by someone who is passionate about finding, amongst the Berlinian plurality of values that present themselves to her as life-shaping options, the one to which she can commit herself most fully and successfully. People who are endowed with an especially intense and serious sense of life-mission of this kind can spend their youth searching for the vocation that provides it, making radical changes of direction as they do. At the extreme, someone might never find it, but without giving up the search. That, too, would be something it would make sense to describe as a kind of integrity; but it would be pursued at the cost of integrity-as-constancy. And someone could have integrity-as-fortitude, withstanding the pressure to behave dishonestly or timorously, without possessing that driven quality. It is a third kind of integrity.

7. Shaping One’s Life This gives us a description of two narrative virtues: loyalty, which is the virtue of responding well to the history of a valuable relationship of special connection, and integrity-as-constancy, which is the virtue of responding well to the shape one’s life has acquired through the allegiances one has formed. A tempting thought is to try to amalgamate the two, by treating integrity-as-constancy as a kind of self-loyalty—a fidelity to the narrative history you have created for yourself, through the relationships of special connection you have formed to some life-shaping goods rather than others. But we have seen that that is unconvincing. By insisting that the rootless, idealistic war correspondent is after all loyal (to his vocation), it misleadingly assimilates him to someone who is strongly attached to particular people or to a particular culture and

Narrative Virtues and Second-Order Reasons


way of life. “Loyalty” loses its usefulness as a name for what is distinctive about responsiveness to attachments of the latter kind if we stretch it to cover other forms of integrity. And thinking of integrity as self-loyalty also risks making it too self-regarding. The idealism that gives the war correspondent his steadfast commitment to his vocation may amount to a passionate attachment to ideals of justice, rather than to himself. So we should allow that there can be integrity-as-constancy without loyalty. And, in the other direction, there can also be loyalty without integrity. Loyalty can be uninvested: when it is, one has not done anything to shape one’s life around the loyalties one happens to have, and so one cannot be responding well to having done so. My treatment of the narrative virtues might seem committed to an overly rigid view of a virtuous life—making the best life-course into one in which a person commits herself to lifelong allegiances when young, then sticks to these unwaveringly, executing a neat and fully coherent Roycean or Rawlsian “life plan.”18 But that is not what I have been claiming. My claim is that there is a pro tanto reason to sustain one’s valuable relationships and remain constant in one’s allegiances. But this reason might be outweighed. I might find myself in a situation where something more important is at stake, and I must make a radical change of life-direction. There are various possibilities. One is that I come to realise that the allegiance I have formed was a mistake. What I thought was a Berlinian choice was not: one alternative was right, and I took the wrong one. The religion I have been following is untrue: I must redirect my life along a different path. Alternatively, my circumstances might change so that what was originally a Berlinian choice has ceased to be one. The political movement that was a vehicle for noble and achievable ideals back when I first joined it has lost its way, becoming hijacked by sectional interests. Or perhaps what has changed is not my circumstances, but me. I have lost the passion and drive that fuelled my activism, or the love that nourished my marriage. Another possibility is that I have not changed, but it turns out I never possessed the talents that were required to make a success of this path: the only way to find out if I had the gifts required for a musical career was to commit myself to it for years with all my energy—having done so, the answer turns out to be No. Finally, I might be unforeseeably confronted with some momentous new reason that demands that I abandon my old life completely. I thought I was a childless scholar, but I now discover that a brief relationship years ago produced a son who is languishing in a foreign jail, and I must get him out. Even here, though, the allegiances I have abandoned, as prominent parts of my narrative history, remain relevant to the course I shape for myself from here. I face questions that those with a different narrative past do not. What kind of ex-monk, or ex-soldier, or ex-husband am I going to be?


Garrett Cullity

Bearing all of this in mind, we can now finally return to the question left hanging at the end of Section 3. When I face a Berlinian choice, what is there for me to think about in deciding what to do? Must I simply pick one alternative, for no reason? But how could that be true to the gravity of a life-shaping decision? The intervening discussion has put us in a position to answer this question. When I face a choice between life-shaping goods, the selection I make carries with it important commitments concerning the future conditions that must be met if I am to sustain an allegiance to those goods. These conditions and the associated commitments are what I need to reflect on in deliberating about which alternative to choose. I need to ask myself: What will I need to do in future in order to sustain the loyalty or allegiance I am now contemplating? What challenges am I likely to face? What internal and external resources will I be able to draw on in meeting those challenges? If I take on these allegiances, what constraints will it place on me? To what extent will it constrain the other allegiances I can coherently form? How will I need to develop in order to live up to my commitments: what are the implications of this decision for the kind of life I am able to lead? How vulnerable will I be to the unreciprocated loyalties of others, how vulnerable will others be if I fail to reciprocate their loyalties to me? How dependent is the success of my commitment on future circumstances that are beyond my control, and what is the likelihood that such circumstances will arise? How risky is my decision, and where will I and others be left if the project I am embarking on does not work out? To understand what I am doing in asking these questions, we can make a distinction. “Berlinian choices,” as we defined them earlier, can refer to choices between types of goods—between close family relationships and career fulfilment, as such—or between particular token-instances of these goods that will be realised in my life, if I pursue them. Types of goods for which Berlinian choices arise are not hard to identify. But when I face a particular, situated choice between life-shaping goods within the circumstances of my own life, it will rarely be immediately obvious whether it is a Berlinian token-choice. It is by asking the previous questions that I must assess this. If a goal is worthy, but my abilities or circumstances make it unlikely that I will succeed and problematic if I fail, that can mean that, all things considered, there are after all decisive reasons not to commit myself to it. So one conclusion I might reach when asking these questions is that the token-choice I face is not Berlinian: my own life-circumstances and personal limitations give me decisive reasons to choose one alternative over the other. But what if that is not true? What if, after asking these questions, I conclude that my prospect of sustaining a successful allegiance is not clearly better if I make one selection than the other? Then my thinking has not been wasted. I am still going to be better equipped to make a

Narrative Virtues and Second-Order Reasons


success of the path I do take if I have a clear view of what this commits me to. Moreover, although in a Berlinian token-choice there is no reason that decisively favours one of my alternatives over the other, this does not reduce me to arbitrary picking. I can still choose one alternative for the reasons that favour it, without pretending that (at the time of my decision) they decisively outweigh the reasons for taking a different path. I think it is an attraction of the account I have offered of narrative virtues that these are the questions it gives us to ask when we make Berlinian choices. It makes these choices difficult, but recognisably so: they are difficult in the way we experience them as being. They are difficult because they depend in part on predicting the future, in part on the kind of objective self-scrutiny that is necessary to make a realistic assessment of one’s own capacities, and in part on having a realistic sense of how far allegiances to different goods can be combined. They are difficult, in short, because the best resolution of such choices is personality- and circumstance-dependent, the wisdom required to fully understand what is at stake comes from experience, and this is not available to the youthful selves who must decide for us.

8. Conclusions A reasons-responsiveness view of virtue—a view according to which virtues are forms of good responsiveness to the reasons bearing on what we do, think, feel, and say—helps to illuminate what is distinctive about virtues such as loyalty and the form of integrity I call integrity-as-constancy. These are ways of responding well to facts about the narrative shape that one’s relationships and allegiances have given to one’s life. Of particular interest are the forms that these narrative virtues take when they are responses to the way one’s life has been shaped by one’s previous reasonsresponsive decisions. Narrative virtues with this feature are responses to second-order reasons. Appreciating this helps us to distinguish different forms of loyalty and integrity from each other; to see the relationship between these virtues and Berlinian choices between life-shaping goods; to see how a virtuous sensitivity to the life-allegiances one has formed in resolving those choices need not be unduly self-regarding; and to appreciate what there is for someone who faces such choices to think about.19

Notes 1. For fuller presentations of the reasons-responsiveness conception of virtue, see Cullity (2017), Cullity (2018a, ch. 7), and Audi (2009). 2. Mistakes themselves can be evaluable too, of course: there are innocent mistakes and irresponsible ones. So the nature of one’s mistakes bears on the virtuousness of one’s responses too. 3. For a fuller attempt to spell out these conditions in a way that makes sense of unreflective reasons-responsiveness, see Cullity (2018b).

52 4. 5. 6. 7. 8. 9. 10.

11. 12. 13.



16. 17. 18. 19.

Garrett Cullity See Aristotle (1999, 1106b19–24). See Dancy (2004, 73). For my attempts to do so elsewhere, see the works cited in notes 1 and 3. See Williams (1985, 10), Foot (1972, 165), and Wallace (1978, 128). See Berlin (1991); also Stocker (1990). For discussion of the “incommensurability” that alternatives with these features present us with, see Chang (1997). Suppose that on arriving in Brisbane I find the new job unfulfilling and now recognise how important my Armidale relationships are. Then the focus on internet conversations might make sense as the best I can now make out of an unsatisfactory situation. But then what I have discovered is that my original choice was not Berlinian. See Brentano (1889/1969, 23), Chisholm (1986, ch. 6), and Chisholm (1976). For proposed sharpenings of Brentano, see Hurka (2001, 13–16), Lemos (1994, 74–76), and Zimmerman (2001, 199, 202). My claim here is not that deciding to do A is itself a reason to A: that claim is usually false. (For discussion, see Cullity (2008, 57–67.) Rather, the claim is that having already acted on a reason can be a reason for continuing to do so. Suspicions of a sunk cost fallacy might arise here; but they are unfounded. The proposal is not that we should treat sunk costs (the losses sustained in foregoing the option not chosen) as reasons for continuing on the path one has chosen. It is that, in a life-shaping choice, realising the good offered by either path requires sustaining an allegiance over time. When I set out on one of those paths, my having acted for the reasons I did situates me within a narrative whose value I can only fully realise by sustaining that allegiance. It is not the sunk costs, but the narrative significance of my past responses to reasons, that gives me a (second-order) reason to continue on the path I am on. I do not claim that everything that is correctly called “loyalty” is virtuous— only that there is a virtue of responding well to the history of a valuable relationship of special connection, and “loyalty” is our name for that. I offer an account of the relationship between virtues and aretaic properties in Cullity (2017) and Cullity (2018a, ch. 7). For the view that there is no virtue of loyalty, see Ewin (1992). A rule-proving exception: suppose you learn about a disability during pregnancy and promise your unborn child to do everything you can to help her through the world. Then the care you give her as a newborn may be responsive to reasons of loyalty—but that description now makes sense because of the prior relationship. On this point, compare Kolodny (2003). See Cullity (2018a, 146–47). See Royce (1908, 31–44) and Rawls (1971, 93, 528). For their help in improving this chapter I am grateful to the editors, as well as Damian Cox, Ramon Das, Richard Hamilton, Jeanette Kennett, Christine Swanton, Susan Pennings, Nic Southwood, Philip Pettit, Seth Lazar, Katie Steele, and an anonymous reader.

References Aristotle. 1999. Nicomachean Ethics, T. Irwin, trans. Indianapolis, IN: Hackett Publishing Co. Audi, R. 2009. “Moral Virtue and Reasons for Action.” Philosophical Issues 19: 1–20.

Narrative Virtues and Second-Order Reasons


Berlin, I. 1991. The Crooked Timber of Humanity. New York: Random House. Brentano, F. 1889/1969. The Origin of Our Knowledge of Right and Wrong, R.M. Chisholm and E.H. Schneewind, trans. London: Routledge and Kegan Paul. Chang, R., ed. 1997. Incommensurability, Incomparability, and Practical Reason. Cambridge, MA: Harvard University Press. Chisholm, R.M. 1976. “Brentano’s Theory of Correct and Incorrect Emotion.” In L.L. McAlister, ed., The Philosophy of Brentano. London: Humanities Press, pp. 160–175. Chisholm, R.M. 1986. Brentano and Intrinsic Value. Cambridge: Cambridge University Press. Cullity, G. 2008. “Decisions, Reasons and Rationality.” Ethics 119(1): 57–95. Cullity, G. 2017. “Moral Virtues and Responsiveness for Reasons.” In S. Braun and N. Birondo, eds., Virtue’s Reasons: New Essays on Virtue, Character, and Reasons. London: Routledge, pp. 11–31. Cullity, G. 2018a. Concern, Respect and Cooperation. Oxford: Oxford University Press. Cullity, G. 2018b. “Stupid Goodness.” In F. Schroeter, ed., The Many Moral Rationalisms. Oxford: Oxford University Press, pp. 227–246. Dancy, J. 2004. Ethics without Principles. Oxford: Clarendon Press. Ewin, R.E. 1992. “Loyalty and Virtues.” The Philosophical Quarterly 42(4): 403–419. Foot, P. 1972. “Morality as a System of Hypothetical Imperatives.” The Philosophical Review 81(3): 305–316. Hurka, T. 2001. Virtue, Vice, and Value. Oxford: Oxford University Press. Kolodny, N. 2003. “Love as Valuing a Relationship.” The Philosophical Review 112(1): 135–189. Lemos, N.M. 1994. Intrinsic Value: Concept and Warrant. Cambridge: Cambridge University Press. Rawls, J. 1971. A Theory of Justice. Cambridge, MA: Harvard University Press. Royce, J. 1908. The Philosophy of Loyalty. London: Macmillan. Stocker, M. 1990. Plural and Conflicting Values. Oxford: Oxford University Press. Wallace, J.D. 1978. Virtues and Vices. Ithaca: Cornell University Press. Williams, B. 1985. Ethics and the Limits of Philosophy. London: Collins. Zimmerman, M.J. 2001. The Nature of Intrinsic Value. Lanham, MD: Rowman & Littlefield.


Narrative Virtue Ethics? Ramon Das

The idea of a narrative has served a variety of philosophical purposes in the recent literature. It has been argued that narratives are an aid to understanding (De Bres 2018); that they constitute a distinct form of explanation (Velleman 2003); and that they are a source of meaning in life (Rosati 2013; Kauppinen 2015; De Bres 2018). In this chapter I sceptically consider whether narratives can also serve as a source of normativity. More specifically, I consider whether they are a serviceable substitute for dispositions as a normative basis for virtue ethics. I use as a foil a recent paper by Lisa Grover (2013), in which she argues that the “situationist critique” of virtue ethics can be avoided entirely by appealing to narratives, rather than dispositions, as a basis for character traits. Grover argues in particular that narratives have the power to cause behaviour in the agent in just the way that dispositions are commonly thought to do. In this chapter I’ll argue that even if narratives have the causative power that Grover thinks they have, they cannot provide the normative basis for right action that virtue ethics requires. In this normative respect, narratives are not an improvement upon dispositions, which, I’ll argue, cannot supply the requisite normative basis for virtue ethics either. The key problem in both cases is that while dispositions and narratives aim to provide a sound basis for character traits, character is ultimately unsatisfactory as a basis for right action. The latter must be sufficiently independent of character to allow for the truism that it is always possible to act rightly (or wrongly) by acting out of character. Sometimes, persons with bad characters nevertheless do the right thing and persons with good characters nevertheless do the wrong thing. Insofar as the virtue-ethical appeal to narratives is in the end an attempt to provide a sound basis for character traits, it cannot enable virtue ethics to meet this fundamental objection.1 The chapter is structured as follows. In Section 1, I argue that even if we assume that the situationist critique can be answered—hence that dispositions are in some sense real—dispositions cannot provide virtue ethics with the normative basis that it requires. I specifically argue that the idea of a “disposition to act for reasons” is inherently problematic.

Narrative Virtue Ethics?


In Section 2, I consider Grover’s argument that virtue ethics can sidestep the situationist critique by appealing to narratives rather than dispositions. Extending her argument, I explore some ways in which narratives might also be thought to improve upon dispositions as a normative basis for virtue ethics, particularly as they seem to provide agents with a rich source of reasons for action. In Section 3, I argue that that notwithstanding these potential advantages, narratives carry other disadvantages and remain, overall, no less problematic than dispositions as a normative basis for a virtue ethical standard of rightness.

1. Dispositions and Virtue Ethics There are different versions of virtue ethics. Here I focus on the leading Aristotelian conception that understands the virtues as dispositions with two important features. First, they are well-entrenched in their possessors, capable of reliably (perhaps invariably) guiding them to perform the right actions and to feel the appropriate associated emotions. Second, they are in some sense dispositions to act for reasons, enabling their possessors to respond, both internally and externally, in the right way to morally salient features of very different situations. These two features, I shall argue, are in considerable tension with one another. Although the idea of a well-entrenched disposition is conceptually unproblematic, it cannot provide a plausible normative basis for a standard of rightness. In contrast, although the idea of a “disposition to act for reasons” is more plausible as a standard of rightness, it is deeply problematic from a conceptual standpoint. The central problem is that if “acting for reasons” is to do the normative work that virtue ethics requires, it cannot be understood in dispositional terms at all. Rather, it must be understood as a kind of ability.2 Consider the dispositional feature of being “well-entrenched.” On its most natural understanding, this feature is unproblematic. The idea is that having a disposition leads its possessor to do (and feel) things “as a matter of course,” free from psychological division or conflict. Put differently, having a well-entrenched disposition leads its possessor to do what she is (most) naturally inclined to do. As I see things, this is by far the more distinctively virtue ethical of the two features of dispositions mentioned previously. It is worth considering why. Doing so will also help us to appreciate why a “disposition to act for reasons” is problematic. Perhaps the easiest way to see the point I wish to make is to draw a familiar contrast between Kantian and Aristotelian ethics. On the Kantian view, acting rightly is a matter of exercising one’s rational will, which, as experience teaches, often requires overriding one’s inclinations or desires. On the Aristotelian view, in contrast, being strong-willed in this way is not enough. Rather, truly right or excellent action requires that one’s inclinations be in harmony with one’s rational will.


Ramon Das

Rosalind Hursthouse (2006, 104) puts the point as follows: [A]ny account of virtue that does not depart too far from Aristotle preserves his distinction between enkrateia, continence or strength of will, and virtue. “Virtuous conduct gives pleasure to the lover of virtue” [Nichochamean Ethics 1099a12] because reason and inclination/desire are in harmony in such a person; they characteristically do what is virtuous desiring to do it. On this view, Kantian strength of will is not required for acting rightly— indeed, it is counter-indicated. Something different from rational willing, or at any rate something more, is required.3 Hursthouse suggests that this something more is the desire to do what is virtuous. More broadly, she has said that it includes virtuous feelings and emotions, motives, values, and much else. All told it adds up to the idea that virtues are dispositions which are well-entrenched in their possessor. The idea is that deep facts about a virtuous person’s character—about who she is—dispose her to behave and feel, reliably if not invariably, the morally correct or right way. Crucially, that she is so disposed—that she doesn’t feel internal confict when she does what is right—is a large part of what makes a virtuous person’s actions right on this view.4 Now to introduce the tension between the two features of dispositions mentioned previously. Notice how easy it is to substitute the word “inclined” or “incline” for “disposed” or “dispose” in the last two sentences of the previous paragraph. The meanings of inclination and disposition are evidently closely similar, to the extent that many dictionary definitions of the former list the latter as a synonym. In stark contrast, there is no apparent similarity between the idea of exercising one’s rational will and that of a disposition. Indeed, as we have seen, if we understand disposition as its near synonym inclination, then exercising one’s rational will, or (arguably equivalently) acting for reasons, often involves an internal conflict with one’s disposition(s), or with what one is disposed to do. It is not only on a Kantian conception that this is so. Common sense teaches that what we are inclined to do is quite often not what (reason tells us) we should do, and even virtue ethicists readily concede that this is true of non-virtuous persons. In short, if we understand “disposition” in the most natural way— namely, as “inclination”—then there is a clear tension between disposition and rational will—between what one is disposed to do and what one rationally should or ought to do. If this is right, then it should not be surprising that the idea of a disposition to act for reasons is problematic. I’ll now argue that a “disposition” in this sense is best understood as a complex ability that cannot be understood in dispositional terms at all, at least not if it is to do the normative work that Aristotelian virtue ethics requires of it.

Narrative Virtue Ethics?


What is the nature of this normative work? First and foremost, virtue ethics aims to provide an account of right action in terms of virtue, which is to say (given our present assumption), in dispositional terms. If the concept of a disposition is to provide this normative grounding, it cannot be understood as something mindlessly reflexive or habitual. Julia Annas (2006, 324) is especially clear on this point: A virtue . . . is not a habit in the sense in which habits can be mindless, sources of action in the agent that bypass her practical reasoning. .  .  . A virtue, unlike a mere habit, is a disposition to act for reasons, and so a disposition that is exercised through the agent’s practical reasoning; it is built up by making choices and exercised in the making of further choices. When an honest person decides not to take something to which he is not entitled, this is not the upshot of a causal buildup from previous actions but a decision, a choice that endorses his disposition to be honest. Annas is evidently keen to rebut the idea that a virtue is a mere habit that effectively bypasses one’s practical reasoning. The most signifcant and problematic aspect of the passage, however, is the way in which she characterises virtuous dispositions in a way that puts so much emphasis on decision, on choice. Given what has been said earlier, it may be clear why this is problematic. If dispositions are best understood as inclinations, they should guide their possessors to act and feel certain ways “as a matter of course.” But in that case “acting for reasons,” in the sense of making rational decisions or choices, should not be necessary for such dispositions to be effective. Indeed, as we have seen, on one standard understanding of Aristotle’s ethics, if virtuous dispositions are functioning properly then too much by way of rational willing may be undesirable.5 At this point one might object: my argument lays too much emphasis on strength of will as a kind of motivational counterweight to non-virtuous disposition, and not enough on rational will as a cognitive component of virtuous disposition. Just because the virtuous person doesn’t struggle to overcome vicious inclinations to act rightly (as the merely continent person does), that doesn’t mean she is acting reflexively or mindlessly; she may still be acting for reasons. If so, then why couldn’t someone be inclined, “as a matter of course,” to respond in the right way to (the right) reasons?6 Isn’t this precisely the picture of the virtuous person that Aristotelian virtue ethicists have long had in mind, from Aristotle himself, through to McDowell (1979), Hursthouse, and Annas? I concede that there is nothing inherently contradictory about the idea of a person so virtuous that her ability in any given situation to respond wholeheartedly to exactly the right reasons in exactly the right way, has been honed to such a degree that it appears to be effortless. In such a case, it would seem, she has a disposition to act for (the right) reasons in the


Ramon Das

sense of being strongly inclined to do so. Does the apparent possibility of this kind of case undermine the line of argument I have been pursuing? I don’t think so. The reason is that I don’t think that virtuous disposition qua inclination is the factor that makes the action right, that justifies it. To see this, however, the cases we need to consider are not those in which an ideally virtuous person effortlessly does the right thing but rather those in which disposition and action evaluatively come apart— those that reflect the truism that sometimes people act out of character. A good person may do the wrong thing, and a bad person may do the right thing, even if neither happens very often. A closely related truism is that a person may do “the right thing for the wrong reasons.” If virtue ethics is to respect these truisms, it must loosen the connection between virtuous disposition (qua inclination) and right action. It must allow for the possibility that even a vicious agent may in certain situations respond in the right way for the right reasons, thus acting rightly. Conversely, it must allow that even a virtuous agent may in certain situations fail to respond in the right way to the right reasons, thus acting wrongly.7 However, in allowing for these possibilities, virtue ethics risks abandoning the very idea that right action can be understood in dispositional terms. Instead, respecting the truisms just mentioned supports the idea that acting for reasons is a kind of complex ability, one shared by virtuous and vicious persons alike. Admittedly, the question remains whether this ability can itself be understood in dispositional terms (Vivhelin 2013). I have argued elsewhere that the relevant ability is nothing less than the ability to do otherwise, and that even if simpler abilities (such as the ability to ride a bicycle or to play the piano) can be understood in dispositional terms, the ability to do otherwise cannot be (Das, unpublished manuscript). I won’t delve into that argument in detail here, but its basic structure is roughly as follows. First, the “ability to act for reasons” evidently includes the ability to act for all sorts of reasons: good reasons, bad reasons, and any type of reason in between. Crucially, it includes the ability to act for reasons that are, for the agent, highly unusual. In short, it seems very much like the ability to choose (decide, intend) otherwise, where this means something like: the ability, at a given time t, to make any one of indefinitely many choices. Second, the most common way to understand abilities in dispositional terms is via a conditional analysis of the sort used to understand the dispositions of inanimate objects: one analyses dispositions in terms of conditionals in which (intuitively) if the antecedent or stimulus is satisfied, then the consequent follows, that is, the disposition is manifested. Now, in the case of even relatively simple abilities there is a problem with specifying the relevant stimulus conditions, which typically seem to involve “tryings.” In the case of riding a bicycle, the stimulus seems to be something like “try to ride a bicycle” and the full conditional: “If one tries to ride a bicycle, then one will ride a bicycle.” The problem is

Narrative Virtue Ethics?


that the stimulus—trying to ride a bicycle—doesn’t ensure that you will ride a bicycle (even setting aside mechanical failures such as the chain being jammed, etc.). Why not? Because one can always stop trying. Thus, in order to have any chance of working, the stimulus would have to be something like: “try to ride a bicycle and (after that initial trying) don’t stop trying.” That, I submit, is a very different sort of stimulus from “try to ride a bicycle.” The problem is brought into sharp relief when we consider attempts to understand the “ability to choose otherwise” in dispositional terms. To see why, return to the earlier formulation. The ability to choose otherwise, I said, means something like: the ability, at a given time t, to make any one of indefinitely many choices. More carefully, we can say that it involves the ability, holding fixed all facts and laws about the world up until time t, to make any one of indefinitely many choices. Now consider the view that this can be understood in dispositional terms via a conditional analysis and consider a candidate stimulus. Presumably, it would be something like: “try to make any one of indefinitely many choices” and the manifestation of the disposition would be that the agent actually makes the choice that she tried to make. However, to the extent that this even makes sense, it simply fails to capture the relevant ability. The ability to make any one of indefinitely many possible choices at a given time t cannot be understood simply as the sum of each particular one of the choices the agent would have manifested had she tried. The ability to choose otherwise cannot be understood in causal-dispositional terms. As we’ll see in the following section, narrative accounts of the virtues do not appeal to dispositions and thus have no need for conditional analyses. So, it is not surprising (as we’ll see in Section 3) that the problems faced by narrative accounts in providing a normative grounding for virtue ethics are somewhat different from those faced by dispositional accounts. In the coda to the chapter I’ll suggest that the difficulties facing the two accounts share an important similarity. At this point, I turn to consider the narrative account of virtue on its own terms.

2. The Narrative Account of Virtue In this section I consider Grover’s (2013) argument that virtue ethics can avoid talk of dispositions altogether and replace it with talk of narratives, yielding a narrative-based account of character traits or virtues. Her goal, recall, is to avoid the situationist critique, which holds that the experimental literature casts doubt upon the reality of stable traits of character that reliably produce certain types of behaviour. Instead, it supports the idea that behaviour is typically determined by contextual factors (“situations”) independently of any character traits that individuals allegedly possess. Grover aims to sidestep this critique entirely by focusing on narratives rather than dispositions. Her view is that a character


Ramon Das

trait or virtue is literally constituted by the narrative of that character trait and that the narrative does not describe an independent disposition. Her main task is thus to show that narratives qua character traits can cause behaviour in the agent, which she sees (correctly, I think) as central to virtue ethics. After outlining her argument, I’ll briefly consider some ways in which narratives might also be thought to provide a normative basis for virtue ethics. Grover takes as her point of departure a view defended by Stuart Hampshire (1953), in which he argued that there are different standards of appropriateness for attributing dispositions to inanimate objects as compared to persons. In the case of objects, purely conditional or hypothetical analyses suffice to make an attribution appropriate: one can attribute a disposition to an object without the object ever having displayed the relevant behaviour. For instance, it makes sense to say that a glass vase is fragile even if it has never broken (because never dropped); it is nevertheless true that it would break if it were dropped. In the case of character traits (qua dispositions) of persons, however, Hampshire argued that the person must have exhibited the relevant behaviour in the past for an attribution of the relevant disposition to make sense. For instance, it would not make sense to say that a woman is generous—to attribute the virtue of generosity to her—unless she has acted generously in the past. Hampshire thus rejects a purely conditional understanding of dispositions as they are attributed to persons. Grover calls this the “summary” view of character traits: attributions of traits qua dispositions to persons are nothing more than summaries of past behaviour. It is not hard to see such summaries as simple narratives, or perhaps elements of narratives, of a person’s life. If to call Susan generous (to attribute to her the disposition of generosity) is simply to summarise her past generous behaviour, then we have told a little story about her life. However, Grover argues that the summary view is not enough for narratives to replace dispositions within virtue ethics, because it does not answer the challenge posed by the situationist critique. As she puts it, “An answer is needed to this challenge so that the virtue ethicist can maintain the view that virtues (or character traits) cause behaviour in the agent, a central feature of the virtue ethical position.” (Grover 2013, 70–71, my emphasis). The summary view cannot do this and indeed, as Grover notes, accords with the situationist critique in holding that character traits lack the power to cause behaviour in the agent. In developing a view of narratives that can remedy this causal deficit, Grover draws upon work of Velleman (2003) and Goldie (2012) to emphasise two related aspects of narratives that enable them to causally affect us. The first is broadly evaluative and is a by-product of narrative construction itself. She writes: “The harmonization of fragments of past behaviour into a coherent narrative may lead individuals to evaluative themselves and motivate them to act similarly, or differently, in the

Narrative Virtue Ethics?


future” (Grover 2013, 72). The idea is that in creating stories that help us make sense of the disparate actions and events of our past, our evaluative reactions to such stories can have an impact on our behaviour going forward. I find this idea quite plausible, though I don’t think Grover would (or at any rate that we should) require that those who engage in such activity consciously think of themselves as constructing a narrative or story. Detecting patterns in one’s past behaviour and being motivated to change (or continue) those patterns based on one’s evaluative reactions to them, seems to be enough to exemplify what Grover has in mind. The second aspect of narratives that Grover emphasises is the emotional component of the first, evaluative aspect. This emotional aspect itself has two parts: the emotional reactions that may have been felt at the time by the agent and those that occur later in his or her role as protagonist/narrator and (especially) as audience. She uses as an example a passage taken from the autobiography of Tenzing Norgay, one of the first two persons (along with Edmund Hillary) to ascend Mt. Everest (Grover 2013, 73, citing Norgay and Ullman 1955): We were winding our way between the tall seracs, or ice-towers, when suddenly the steep snow under his [Hillary’s] feet gave way, and he fell into a crevasse. “Tenzing! Tenzing!” he shouted. But fortunately there was not too much rope between us, and I was prepared. Jamming my axe into the snow and throwing myself down beside it, I was able to stop his fall after about fifteen feet, and then, with slow pulling and hauling, managed to pull him up again. By the time he was out of the crevasse my gloves were torn from the strain; but my hands were all right, and, except for a few bruises, Hillary was unhurt. Grover notes that the emotional understanding to which this autobiographical narrative gives rise may be quite different from the emotions felt by the narrator at the time. Norgay may well have felt panic at the time the events occurred, but that is not the emotion refected in the subsequent narrative, in which he is portrayed as maintaining his composure. Signifcantly, the idea of maintaining one’s composure under pressure— the emotion, trait, or state of mind to which Norgay and others might well aspire—may be highly motivating going forward. Grover’s key point, again, is that there is something about constructing a narrative—both the raw doing and the structural details of that narrative—that has signifcant effects on how the agent sees and conducts herself in the future. Recapping Grover’s position, she argues that we should view narratives not merely as a summary of past actions and events, but as a constructive process that in turn has the power to cause future behaviour in the person who constructs it. This causal power is a by-product of narrative construction itself and has a markedly evaluative aspect with a key


Ramon Das

emotional component. Specifically, narratives have the power to induce emotional understanding in the narrator/protagonist when she reflects on her narratives later, in the role of audience. Again, Grover’s main concern is to enable virtue ethics to evade the situationist critique by replacing dispositions with a narrative account of virtue that has the power to cause behaviour in the agent.8 Her discussion doesn’t reflect any concern about our present question, which is whether narratives can serve as a normative foundation for virtue ethics. Although I shall ultimately argue for a negative answer to this question, I think certain aspects of Grover’s discussion suggest that a narrative account of virtue has some advantages over dispositional accounts with respect to it. I close this section by reviewing some of these potential advantages. Begin by recalling Annas’ point that virtues (qua dispositions) cannot be construed as mindlessly reflexive or habitual sources of action. Rather, she thinks, a virtuous disposition is in some sense a “disposition to act for reasons.” Yet this latter concept, as I argued in the previous section, is inherently problematic. Specifically, insofar as “disposition” is construed as its close synonym “inclination,” it is in tension with the idea of “acting for reasons,” which suggests an ability that is shared by virtuous and vicious persons alike. We can deepen the point by thinking about what dispositions must be like if they are to do the causal work that virtue ethics requires of them. First and foremost, they must be sufficiently fixed to underwrite stable traits of character. If dispositions were not able to underwrite stable character traits, there would be no reason to think they could underwrite reliable predictions of behaviour, and the situationist critique (which casts doubt on their existence) would be a much less serious problem for virtue ethics. However, the stability that makes dispositions qua inclinations so well suited for doing the causal work that virtue ethics requires is also what makes them problematic for doing the normative work that it requires. Again, character and action can come apart; it is always possible to act rightly (or wrongly) by acting out of character. From this critical perspective, a narrative account of virtue such as Grover’s arguably enjoys an advantage over dispositional accounts. Because it is not tied to relatively fixed dispositions, the narrative account seems to be more flexible. This apparent flexibility derives from two related factors. First, it is a by-product of the constructive aspect of narratives that Grover emphasises. Whereas character traits qua dispositions are relatively fixed, character traits qua narratives are comparatively fluid. They can be constructed, shaped, and (presumably) changed and interpreted in countless different ways. One upshot of this variety and fluidity is that the narrative account seems to provide agents with a rich source of reasons for action. There is, more specifically, no apparent tension between narrative virtue and acting for reasons in the way that, it was

Narrative Virtue Ethics?


argued previously, there is a tension between the latter idea and dispositional virtue. The second, related factor behind the narrative account’s greater flexibility is simply that an agent’s narrative is under his control in a way that his dispositions are not. It is a truism that changing one’s character qua disposition—changing what one is disposed to do and feel—happens very slowly and with great effort, if at all. In contrast, as we have seen, changing one’s character qua narrative is comparatively easy. (Whether it is too easy is a question that will be considered in the following section). However, it is undeniably easier and more under one’s control than changing one’s character in the (disposition-based) way that phrase is ordinarily understood.9 In short, Grover’s narrative account of virtue has much to be said in its favour. It presents a plausible alternative to the more familiar dispositional account of virtue with respect to the question whether character traits can be a cause of behaviour. If she is right, the narrative account may indeed enable virtue ethics to sidestep the situationist critique altogether. Further, we have seen that narrative accounts, due to their greater flexibility, may enjoy an advantage over dispositional accounts with respect to being able to provide virtue ethics with a sound normative foundation. In particular, narrative accounts do not face the same internal tension that dispositional accounts do with respect to the key idea of “acting for reasons.” Despite these advantages, I believe that narrative accounts of virtue ultimately fail to provide virtue ethics with the normative foundation it needs. In the final section I try to make this case.

3. Narrative Virtues and Normativity According to the narrative account of virtue, character traits are constituted by narratives which are themselves theoretically basic: the account makes no appeal to independent dispositions. As we have seen, this feature affords the narrative account a degree of flexibility that dispositional accounts lack, which makes it a better fit with the key concept of acting for reasons. However, this same flexibility also leads to problems for the narrative account. At first glance these problems seem quite different from those faced by the dispositional account. I’ll conclude by suggesting that they are fundamentally similar. Before getting into the argument it is necessary to clarify my target claim, which involves clarifying the relationship between a narrative account of virtue and a virtue ethical account of right action. As noted, Grover’s main concern in developing the narrative account is causal, not normative. She is interested in showing that character traits qua narratives can be a cause of behaviour in the agent. However, for the narrative account of virtue to provide a normative foundation for virtue ethics,


Ramon Das

there must be some sense in which character traits qua narratives provide the standard or criterion of ethical rightness. Perhaps, for instance, right action is grounded in virtuous or morally uplifting narratives; wrong action in vicious or morally perverted ones. I’ll now argue that there are good reasons to doubt that there is any such sense. I start with an issue that Grover discusses and that may have already occurred to the reader. This is the issue of narrative accuracy or truth. The narratives under consideration are self-narratives: narratives constructed by the agent herself. It is thus no surprise that some narratives are not an entirely (or even minimally) accurate rendition of past actions and events. Presumably, most narrative inaccuracies will reflect well (rather than badly) on their author, involving a familiar sort of rationalisation. In any case, the very possibility of inaccurate or untruthful narratives raises a problem for a narrative account of virtue, particularly insofar as such an account is taken to provide a normative foundation for virtue ethics. In fact, the possibility of narrative inaccuracies raises several related problems for a narrative-based virtue ethics. Some of these problems concern the content of the narrative itself: 1. An agent wittingly constructs an inaccurately (un)flattering ethical narrative. 2. An agent unwittingly constructs an inaccurately (un)flattering ethical narrative. Others concern the motivational effects of various narratives: 3. An agent (being vicious) reflectively endorses and is hence motivated by the more vicious of two competing narratives (for example, she endorses the narrative in which she is selfish over the narrative in which she is generous). 4. An agent (being ignorant) reflectively but mistakenly endorses the more vicious of two competing narratives. Setting aside the problems presented by the latter motivational cases (which one could argue are hardly unique to the narrative account of virtue), it is natural to suppose that the fundamental problem in the former cases is simply that the narratives are inaccurate. If so, an obvious remedy is to impose an accuracy or truthfulness constraint on narratives, which is effectively what Grover does.10 She distinguishes narratives in which one (merely) believes something about oneself from those in which one knows something about oneself, observing that “autobiographical narratives aspire to truth and objectivity, as does any historical narrative” (Grover 2013, 77). However, the requirement that narratives be accurate or truthful is not enough to make them an adequate normative basis for virtue ethics. As

Narrative Virtue Ethics?


we shall see, it leads to other problems. An initial worry is that any such requirement seems ad hoc in the context of Grover’s main argument, which is that character traits qua narratives can be a cause of behaviour. It is not obvious that factually inaccurate narratives are less motivating than accurate ones, and it is easy to imagine circumstances in which they are more motivating. A more fundamental problem, however, is that the very possibility of inaccurate or untruthful narratives raises the following question: What normative work is being done by the narrative itself, as opposed to the agent’s reaction to that narrative, or by something else altogether? I’ll approach this last question by considering a series of amendments to the basic narrative account of virtue, arguing that none of them yields an account that can provide virtue ethics with an adequate normative foundation. Let’s start by imposing an accuracy constraint on the account: agents construct (or are permitted to consider) only those narratives that are factually accurate. A less stringent alternative would require that narratives be mostly accurate. Either way, it should be clear why this won’t work. Some agents, after all, have lived ethically bad lives and have vicious characters. Presumably, their narratives will not be ethically uplifting and so will not provide their authors with rich sources of morally good reasons for action going forward. To meet this problem, let’s make another change. I have thus far tacitly assumed that the narrative account fits into a broadly “agent-based” version of virtue ethics, one which appeals to the agent’s own character to evaluate action. This is not, of course, the only option. “Virtuous-agent” forms of virtue-ethics, in which one looks to other, more virtuous persons for guidance, are arguably more popular, so let’s consider those.11 I assume that there are at least some virtuous persons whose narratives are accessible to others, providing them with a rich source of good reasons for action. Is this enough to provide a compelling normative ground for virtue ethics? I don’t think so. After all, even virtuous persons, assuming they’re human beings, aren’t perfect. Their narratives, though mostly virtuous, will (if accurate) presumably also contain vicious elements.12 We have already observed the truism that persons can act “out of character”: a good person can do the wrong thing and a bad person can do the right thing. Evidently this poses a problem for “agent-based” versions of virtue ethics, since it suggests that virtuous character and right action can come apart, hence that the former cannot be the ground of the latter. We can now see, however, that much the same problem arises for “virtuousagent” forms of virtue ethics, including those that understand right action in terms of virtuous agents’ narratives. If even a virtuous agent with a mostly good character (qua narrative) can nevertheless act wrongly, then again, it seems that good character cannot be the normative ground of right action.


Ramon Das

Let’s then consider one final amendment to the narrative account. I have been assuming that it is possible for even virtuous agents to act wrongly because I have been assuming that virtuous agents are actual human beings. Yet the most common form of “virtuous-agent” forms of virtue ethics appeals not to any actual agent but to the ideally virtuous agent, who presumably can be assumed always to act rightly. Can this do the trick? I have argued elsewhere that even “ideally virtuousagent” forms of virtue ethics cannot provide a distinctively virtue-ethical account of right action (Das 2015). Given the present focus on the narrative account of virtue, however, a different point is in order. If virtue ethics needs to appeal to an ideally virtuous agent to ground right action, then the relevant normative work seems to be done by the dispositions (virtues) of that ideally virtuous agent. Ideally virtuous agents, it would seem, have no need of self-narratives, which are useful, if at all, for actual human agents.13 In short, on an “ideally virtuous agent” version of virtue ethics, the importance of narrative in understanding virtue seems to have disappeared altogether. To meet the problem generated by the possibility of false narratives, the narrative account of virtue seems to have to write itself out of the picture. We can thus frame the possibility of false narratives as a dilemma for the narrative account of virtue. Either that possibility is an instance of the more general problem for virtue ethics discussed in Section 1, deriving from the fact that good character and right action can come apart, hence that the former cannot be the ground of the latter. Or, attempts to meet the problems caused by that possibility lead to a version of virtue ethics (based on the ideally virtuous agent) that has no need for a narrative account of virtue at all. Either way, the narrative account, whatever its success in enabling virtue ethics to meet the situationist critique, cannot provide virtue ethics with the normative foundation it requires.

4. Conclusion I close by suggesting that the apparently different problems facing the dispositional and narrative accounts of virtue share an important similarity. We can think about this similarity in terms of how each account must deal with the truism that it is possible for persons to act out of character. The dispositional account accommodates this truism by having to downplay the natural understanding of disposition as inclination, reflected especially in Annas’ idea of a “disposition to act for reasons” that I have criticised as inherently problematic. The narrative account, by comparison, encounters the truism in its attempts to meet the problem generated by the possibility of false narratives. It effectively accommodates the truism by moving toward an “ideally-virtuous agent” version of virtue ethics. In the process, however, it effectively undercuts the need for understanding virtues as narratives altogether.

Narrative Virtue Ethics?


In both cases, the fundamental problem is generated by the fact that the purported ground of virtue—dispositions in one case; accurate or true narratives in the other—are relatively fixed. That is no surprise: if there are to be stable traits of character, they must be relatively fixed. The normative ground of right action, however, cannot be fixed. It must be sufficiently flexible to accommodate the truism that persons can act out of character.

Notes 1. I follow established usage in understanding virtue ethics as specifically concerned with right action or the question what one should do. This is sometimes distinguished from virtue theory, which is more directly concerned with good character or the question what or how one should be. I make no claim here about how (or whether) the present argument bears on the latter question. 2. Although my focus here is on Aristotelian conceptions of virtue ethics, I believe the points just made apply equally well to non-Aristotelian conceptions that rely on dispositions, such as Slote (2001). 3. See Eylon (2009) and Roberts (1984) for related discussion. 4. Hursthouse (2006) distinguishes two senses of “right,” depending on the relevant interpretation of “what is done.” In some contexts, she suggests, disposition is relevant to rightness; in others it is not. Thanks to an anonymous referee. See also Das (forthcoming). 5. Hursthouse (1999, 103) challenges the standard contrast that I am presenting, arguing that Aristotle’s view is much closer to Kant’s. In particular, she writes that “the virtuous Aristotelian agent doesn’t act from inclination either, but from reason in the form of choice,” a position much closer to that of Annas’. Thanks to an anonymous referee. 6. Thanks to Simon Keller for pressing this point in conversation. 7. I have argued for these claims in more detail in Das (2015, 2017). In the former, I argue that defining rightness in terms of what a virtuous agent would “characteristically” do is unsatisfactory as a distinctively virtue-ethical solution to this problem. 8. A bit more carefully: If virtue ethics requires dispositions (qua character traits) and if the situationist critique shows that there are no dispositions in this sense, then virtue ethics has a problem. But if narratives can effectively replace dispositions, then virtue ethics has a satisfactory solution to the problem posed by the situationist critique. 9. Consider the following example. A man who has in the past avoided risky behaviour begins to see this avoidance, and hence himself, as cowardly. This self-narrative enforces in him the idea that he is too scared to take on the financial risks associated with buying a house, even though that would be best for his long-term financial health. He begins therapy and writes out his reflections in a diary. These practices help him to see that his past actions have actually been more prudent than cowardly. This reinforces the idea that although failing ever to take risks may be cowardly, avoiding unnecessary risks is not; it is wise. And that makes him realise that he really can buy a house, which he now sees as a risk well worth taking. Thanks to an anonymous referee for the example. 10. Cf. Rosati (2013), who effectively argues that false narratives are not in our self-interest in the long term.


Ramon Das

11. I set aside the worry that virtuous-agent versions of virtue ethics may be less congenial to Grover’s project, which is to show how narratives can cause behaviour in the agent. Presumably, this is more like when the author of the narrative is the agent himself. 12. Arguably, an accuracy constraint on narratives rules out the possibility of ideally virtuous narratives, at least on the reasonable assumption that there are no ideally virtuous persons. I discuss a different reason for rejecting “ideally virtuous-agent” agents of virtue ethics in the main text. 13. This point must not be confused with the importance of narratives to us: Arguably, we cannot really understand what an ideally virtuous person would be like without resorting to narratives (for example, about Jesus and other heroes). Thanks to an anonymous referee.

References Annas, J. 2006. “Virtue Ethics.” In D. Copp, ed., The Oxford Handbook of Ethical Theory. Oxford: Oxford University Press, pp. 515–536. Das, R. 2015. “Virtue Ethics and Right Action: A Critique.” In L. Besser-Jones and M. Slote, eds., Routledge Companion to Virtue Ethics. London: Routledge, pp. 331–343. Das, R. 2017. “Virtue, Reason, and Will.” In N. Birondo and S. Braun, eds., Virtue’s Reasons: New Essays on Virtue, Character, and Reasons. London: Routledge, pp. 91–106. Das, R. Forthcoming. “Virtue Ethics and the Concept of Action.” Revista Portugesa di Filosopie: 25 pp. Das, R. Unpublished Manuscript. “Virtue, Disposition, and the Ability to Do Otherwise.” Victoria University of Wellington, 15 pp. De Bres, H. 2018. “Narrative and Meaning in Life.” Journal of Moral Philosophy 15(4): 545–571. Eylon, Y. 2009. “Virtue and Continence.” Ethical Theory and Moral Practice 12(2): 137–151. Goldie, P. 2012. The Mess Inside: Narrative, Emotion, and the Mind. Oxford: Oxford University Press. Grover, L. 2013. “Narrative Virtues.” Theoretical and Applied Ethics 2(1): 67–82. Hampshire, S. 1953. “Dispositions.” Analysis 14(1): 5–11. Hursthouse, R. 2006. “Are Virtues the Proper Starting Point for Morality?” In J. Dreier, ed., Contemporary Debates in Moral Theory. Oxford: Wiley-Blackwell, pp. 100–112. Kauppinen, A. 2015. “The Narrative Calculus.” Oxford Studies in Normative Ethics 5: 196–220. McDowell, J. 1979. “Virtue and Reason.” The Monist 62: 331–350. Norgay, T. and J.R. Ullman. 1955. Man of Everest: The Autobiography of Tenzing as Told to James Ramsey Ullman. London: George G. Harrap. Roberts, R. 1984. “Will Power and the Virtues.” Philosophical Review 92: 227–247. Rosati, C. 2013. “The Story of a Life.” Social Philosophy and Policy 30: 21–50. Slote, M. 2001. Morals from Motives. Oxford: Oxford University Press. Velleman, J.D. 2003. “Narrative Explanation.” The Philosophical Review 112(1): 1–25. Vivhelin, K. 2013. Causes, Laws, and Free Will: Why Determinism Doesn’t Matter. Oxford: Oxford University Press.


How Self-Narratives and Virtues Cause Actions David Lumsden and Joseph Ulatowski

1. Introduction Virtue ethics provides an approach to understanding the moral value of actions that shifts focus away from following rules derived from our moral duties or the consequences of actions. Virtues are part of the person’s character and includes the notion that one should foster virtues in oneself and others. They are often understood as a certain kind of disposition to act. If dispositions are understood conditionally, capturing how a person does or would behave in certain circumstances, then what we want to emphasise gets lost, that a person’s virtues causally contribute to how they act. There is another understanding of dispositions, which does allow them to be causes, though. A self-narrative, or maybe multiple selfnarratives, plausibly contributes to, or reflects, a person’s character and indeed it has been argued that self-narratives define a person’s identity. Once again, it is natural to conceive of self-narratives causally contributing to the person’s actions. Thus the main topic of our chapter is how actions may causally flow from virtues and from self-narratives. Our expectation is that this pairing of topics is capable of generating insights and indeed we shall be showing some ways in which virtues and self-narratives are related, in particular through the notion of character. How we shall proceed is to first, in Section 2, outline some ideas about narratives and how they have been regarded by some authors to be central to, or indeed constitutive of, the self. In Section 3, we shall discuss some views about the nature of virtues with an emphasis on their influence on action, noting that they, too, form an integral part of the self. In Section 4, we shall introduce some of the basic ideas in the causal theory of action, as an initial strategy in the development of an account of the causal power of virtues and self-narratives. In Section 5, we shall address Elizabeth Anscombe’s opposition to causal theories of action. In Section 6, we shall develop our own account of the way that self-narratives and virtues causally contribute to action, recognising their position at the core of what it is to be a person.


David Lumsden and Joseph Ulatowski

2. Self-Narratives The notion of a narrative has a rich history going back to Aristotle with a focus on the structure of tragedies and epics. While the notion of a narrative remains an important one in literary theory, it appears also in many other academic disciplines, notably historiography (Mink 1970) and cognitive psychology (Bruner 1986, 1990). Oliver Sacks (1985, 105), a neurologist, viewed a person’s narrative as constituting their personal identity, saying, “Each of us is a singular narrative, which is constructed, continually, unconsciously, by, through, and in us—through our perceptions, our feelings, our thoughts, our actions; and not least, our discourse, our spoken narratives.” Such a notion of a self-narrative is central to the view we shall develop.1 In this section, we will argue that self-narratives are deeply ingrained and internalised structures that may play a contributory causal role in the formation of action. We are therefore interested in a self-narrative as an internal structure. But couldn’t the life of a person as they live it also have a narrative structure? Alasdair MacIntyre (1984) developed a narrative notion of the self, in the context of an examination of the virtues, a connection which reflects our project here. He takes a firm position on the way that narratives are important for understanding persons. It is because we live out narratives in our lives and because we understand our own lives in terms of the narratives that we live out that the form of narrative is appropriate for understanding the actions of others. Stories are lived before they are told. (MacIntyre 1984, 212) He contrasts his view with that of Mink (1970), who says that stories are not lived but told. There is something of a false dichotomy here. Our focus is initially on self-narrative as an inner construction, for we are interested in the way it can causally affect behaviour. Because we see the character of the self-narrative being played out in action to some extent and as circumstances permit, it is reasonable to suggest that the life of the person may also take the form of a narrative. The previously withdrawn high school student who “reinvents” herself as a fun-loving party animal upon going to university, and does so successfully, has both deliberately constructed an internal narrative and come to live it. We are not endorsing the priority MacIntyre places on the narrative as lived but we do allow that both the inner construction and the outer appearance may have a narrative form. This is to adopt a different position on the notion of a narrative from that found in Lumsden (2013, 2013/2014). We need to consider the metaphysical status of the internal narrative. Daniel Dennett’s approach brings that to the fore. He employs the notion of a narrative to characterise the self, saying, for example, “Our tales

How Self-Narratives and Virtues . . . 71 are spun, but for the most part we don’t spin them; they spin us” (Dennett 1993, 418). This challenging view sits against the background of his anti-realist, or as he would prefer to say “mild realist,” position on mental states. Dennett (1987) regards beliefs and desires to have a similar metaphysical status to the centre of gravity of an object: postulating them is useful practically speaking though they are not straightforwardly real. When he speaks of the tales spinning us, he is saying that the self is no more than the tales, or narrative, we tell of ourselves. But the tales too have a mild realist status. You might think that to accept that kind of metaphysical view of a self-narrative would remove it as a potential cause of action. In fact, it may provide the very clue to the kind of causal relationship we need to identify, a point to which we shall return in due course. What does the notion of a self-narrative bring to the table in developing a conception of the self? We can find one possible answer in Marya Schechtman (1990, 1996). One thread in her approach can be described as using the notion of a narrative to capture the holistic character of experience and memory. Schechtman (1990) employs an example from Casey (1987) of someone who recalls seeing a particular film with his family. That memory is not a discrete item that could be transported into the mind of someone else (in the manner of the Venetian Memories thought experiment of Parfit 1984, 220–221), for the memory is completely intertwined with other experiences of those family members and knowledge of all manner of things concerning the venue and the film. That a memory typically sits in a rich network of related memories does not require that they all be accurate. For example, a family member who would be typically present for “family movie night” was in fact absent for one but their presence had been mistakenly filled in the overall memory of that particular evening. The more that inaccuracies are involved that have no basis in reality, the more that there is not a proper setting to sustain a memory. The person who “remembers” seeing a certain Italian film sitting next to Luke Skywalker and Princess Leia in the Mos Eisley Cantina (a distinctively Star Wars setting) can barely be said to remember seeing the film. In Schechtman’s (1996, 119) narrative constitution view of selves, there is an explicit “reality constraint” which requires that the self-narrative “fundamentally cohere with reality,” thus ruling out such cases. Leaving such extreme cases aside, though, the moral is that there is a form of holism concerning memory: a particular memory cannot sit alone. How do we capture that kind of holism? One approach is to look to the notion of a narrative to describe the form of unity into which a particular memory or experience fits. Even though the notion of a narrative is a contested and open-ended one, it has the essential feature we need: it captures the way that represented events hang together in such a way as to create a larger unity, a unity that contributes to the significance


David Lumsden and Joseph Ulatowski

of those individual events. The narrative structure not only provides a setting for individual experiences and makes sense of them but also can play the creative role of producing, that is causing, actions, something that is central to this chapter. J. David Velleman (2006, Chapter 9) agrees with Dennett that we invent ourselves but in contrast says (p. 206) “we really are the characters whom we invent,” thus denying Dennett’s mild realist position. Velleman’s realism may, at first sight, make those characters seem better candidates for being causes of actions but our position is that a mild realist approach better captures the nature of the relevant causal relationship. Consider the puzzle raised by Dennett’s claim that our tales spin us rather than vice versa. Don’t we need a self to construct the tales? Dennett’s view is that it is the human being that creates the narrative, where the human being is not treated as an agent. That it is not an agent is made manifest by Dennett’s parallel with a robot that has the capacity to construct a narrative, in fact an autobiography, which is introduced by saying “Call me Gilbert.” Dennett’s account of the construction of the autobiography is that it records what the robot sensed and did. Velleman raises the other direction: how the narrative guides what is done. A robot smart enough to do all that Dennett describes should be able to “narrate ahead of himself” and then proceed to carry out that plan. In this way, “an autobiography and the behaviour that it narrates are mutually determining” (Velleman 2006, 211). This is an interesting way of articulating the role of a selfnarrative in producing action. Velleman has further things to say about that causal relationship, to which we shall return in Section 4.

3. Virtues Having seen how narrative approaches to the self can take a variety of forms and how self-narratives may form a part of the causal story of human behaviour, we need to say something about the nature of virtues. A person’s character surely involves her virtues and vices and her selfnarrative will surely also have points of contact with her character. It should be no surprise if the causal power of virtues and self-narratives to produce action runs in parallel. The virtues, a central feature of moral philosophy for Aristotle, took a back seat in the modern period until a revival in the twentieth century, so that virtue ethics is once more part of the standard suite of approaches in moral philosophy. The Aristotelian virtue ethics position is a dominant one and is summarised by Liezl van Zyl (2018, 14–15) as having five characteristics: (i) virtue is a human excellence, (ii) what makes a trait a virtue is that it allows its possessor to live a good (happy or flourishing) life, (iii) a virtuous person is motivated by the right feelings and the right reasons, (iv) practical wisdom is required for virtue, and (v) actions are to be evaluated in terms of virtue and vice.

How Self-Narratives and Virtues . . . 73 When a person embodies all of the excellences of human character, that person is good and virtuous. That position reflects Aristotle’s claim that there is a unity of the virtues, which depends on his views about practical wisdom. “[T]he man who is capable of deliberating has practical wisdom. . . . [I]t is a true and reasoned state of capacity to act with regard to the things that are good or bad for man” (Nicomachean Ethics 1140a30–1140b5). Van Zyl explains Aristotle’s reasoning for the unity of virtues thesis like this: Practical wisdom is the same for each virtue for they all require a broad or general conception of the ends of action, that is about what is good or worthwhile for human beings. It follows from this that someone who has practical wisdom will have all the virtues. (Van Zyl 2018, 84) Perhaps one of the most vigorous challenges to virtue ethics concerns the relationship between virtues and action. Virtuous character, so the critic may argue, fails to provide a clear guide to action. A response to that challenge is to be found in contemporary work in virtue ethics beginning with Elizabeth Anscombe (1958/2005). Her main aim in that work is to address a defcient moral view constructed from our zealousness for rules and obligations. In contrast to other ethical theories like utilitarianism and Kantian ethics, virtue ethics fnds its theoretical seat in the cultivation of a character that disposes people to behave in certain ways, ways that we recognise as expressions of virtue. Right action is defned in terms of virtue rather than virtue being defned in terms of right action. An action is right if and only if it is what a virtuous agent would, characteristically, do in the circumstances. (Hursthouse 1999, 79) Action-guidance, therefore, may be derived from what the agent considers proper motivational structures. Rosalind Hursthouse (1999), taking a lesson from Anscombe, argues that we can get perfectly serviceable rules for virtue ethics, what she terms the “v-rules.”“Do what is honest; do not do what is uncharitable” (Hursthouse 1999, 37). Of course, one should not expect possession of the relevant virtues to provide a clear guide to action in the sense of sharp defnitive rules. So, in this chapter, we want to avoid looking at the action from the outside in the manner of the Kantian or utilitarian and consider the issue from the perspective of the actor, one’s inner sense. It is an advantage of virtue ethics that it grounds the virtuous person’s decisions to act in the full complexity of the situation. Here is an example of what we mean. Suppose that Jones is considering whether to tell the truth. Even if Jones has lived a life in which she has cultivated the virtue


David Lumsden and Joseph Ulatowski

of honesty, the possession of that cultivated virtue may not absolutely compel her to tell the truth. Jones may, under the circumstances, remind herself: “Be honest!” Jones may even be compelled to ask herself, “What exactly is honesty and what are the circumstances under which I should be honest?” There is a trade-off between different virtues, such as being honest and being kind. If by telling the truth Jones must sacrifice being kind, such as when answering her friend’s question, “Do I look fat in this?” just before a significant social occasion, then maybe she spends time adjudicating between those virtues to uncover which should guide her action. Our interest in virtues, though, is not primarily as the basis of moral evaluation but as the inner cause of behaviour. To talk of virtues as causes deviates sharply from Anscombe’s picture.2 These inner causes of which we speak ought not to be understood as identifying specific physical structures that exemplify causal relations and we are proposing a mild realist metaphysical status for these inner causes. A central focus for us in this chapter is to uncover their nature. Reason to doubt that virtues are inner causes may stem from our understanding of virtues as dispositions. Dispositions can be understood merely in terms of conditional truths, counterfactual conditionals associated with a degree of probability. Being thirsty on this understanding is to be such that if offered a drink one is likely to accept it. Similarly, being generous is to be such that where someone is in need and one has the means to help one is likely to help. But we should not assume that this exhausts a proper understanding of dispositions. David Armstrong (1968/1993; cf. Crane et al. 1996/2002, 4) argues that dispositional statements are not made true by counterfactual states of affairs, for having a dispositional property needs to be explained in terms of a non-dispositional, categorical property. For example, if we say a glass is brittle, we know this implies what is likely to happen if it is dropped on a hard surface but we assume that there is some property of its microstructure that underpins those counterfactual properties. Armstrong takes a similar view of mental states in his functionalist-physicalist theory of mind. He considers them states “apt for bringing about a certain sort of behavior,” which makes them dispositions but that as a matter of empirical fact the state is always a brain state. The details of Armstrong’s view are controversial and U.T. Place challenges the distinction Armstrong makes between dispositional and categorical properties. Our moral for the moment, though, is simply that to identify virtues as dispositions is not to exclude them as inner causes. We have seen that the Aristotelian view of virtuous action involves motivation by the right feelings and the right reasons. It is natural to think of that motivation as causal: those feelings and reasons causally contribute to action. Other views of virtues give us different accounts of motivation but they mostly involve inner states that are best regarded as causes of action. The Kantian is motivated by a sense of duty such that

How Self-Narratives and Virtues . . . 75 feelings, other than feelings that are centred on carrying out one’s duty, are not what makes the actions of the virtuous person right. Michael Slote, a sentimentalist, thinks that “virtue consists in having and acting on warm motivating sentiments such as benevolence, gratitude, compassion and love” (Frazer and Slote 2015, 197, quoted in Van Zyl 2018, 25). Michael Slote (2001) has placed in the foreground an agent’s motivations and has argued that we need to look at what motivates a person’s actions to determine the action’s moral quality. Acting on motivating sentiments is best understood as those sentiments causally contributing to action. At first sight it would appear that the consequentialist would avoid any inner components of virtue that could be causal, for the virtuous person is the one whose actions tend to produce good consequences. But a consequentialist can opt for an internalist version in which a right action is one that arises out of the motivation to bring about good consequences and, correspondingly, a virtuous person is one who is motivated by a desire to bring about good consequences. In summary, most conceptions of virtue involve inner states that could causally contribute to action. Our focus, though, will not be primarily on the immediate motivating inner states but more on virtues as standing features of a person’s character, in line with the Aristotelian picture.

4. The Causal Theory of Action While there is not a well-developed literature on how virtues and selfnarratives cause actions, there is a significant body of literature concerning the way certain mental states describable in folk psychological terms may cause actions. On Donald Davidson’s view, we act from reasons, which is what makes us rational, and those reasons are causes of our actions.3 To provide a little more detail, a reason is composed of a belief and a desire, specifically the belief is that the action will bring about the fulfilment of the desire. Thus the causal theory of action specifies the connection between the desire and bodily movement when the latter is an intentional action that aims at the satisfaction of the person’s desire.4 For example, suppose I want to attract your attention and I believe that waving at you will attract your attention. This shows my reason for waving at you. Clearly, there could be various means to achieve the same end. If I had believed that lighting a flare would attract your attention, then the combination of my belief that lighting the flare attracts another person’s attention and my desire to attract your attention is my reason for lighting the flare to attract your attention. The causal theory allows us to distinguish between such an intentional action and a mere tic or indeed breathing (in normal circumstances). Tics and breathing are also caused internally, but not by a reason, so on Davidson’s account they are not intentional actions. The waving example is a simple example for which we would not normally bother to provide a reason. But there are more complex cases


David Lumsden and Joseph Ulatowski

where we seek the real reason. Why did the President’s press secretary resign? Various reasons may be offered by the person in question, for health reasons or to spend more time with family, and various others suggested by commentators, involving a conflict with a co-worker or as a result of pressure from the President. The causal theory allows us to ask the question, what really made him or her do it? Our topic is the nature of the causal relationship between virtues and self-narratives and action and it is useful to refer to the causal theory of action for guidance. Just as beliefs and desires can be causes of action so can virtues and self-narratives. The difference is that, while beliefs and desires may well be changeable aspects of a person, virtues and selfnarratives are more subtle, complex, and enduring features of a person and thus the causal powers of those things warrant special attention. At the time that Davidson first developed the causal theory of action, philosophical accounts of causation had not developed in the way they have since. We might say Davidson’s task was to reconcile a notion of causation based on a deterministic deductive-nomological account of scientific explanation with the intuitive idea that reasons are causes. In the deductive-nomological account, singular causal statements are taken to require the existence of a covering law of nature, so that the presence of the effect can be deduced from the law and the presence of the cause. This approach to causation has considerable merits, not least of which is the way that it grounds the popular notion of cause in an underlying scientific account of the world. But there are alternative ways of thinking about cause. For example, the counterfactual theory of causation claims that for a to cause b means that if it weren’t the case that a then b would not occur (Lewis 1973). This is a much broader notion that can be applied in situations that are not so obviously describable in the language of physical science. Thus we need not be tied to Davidson’s theory as a whole to propose that virtues and self-narratives cause actions. It is not our purpose here to endorse or develop a particular account of causation. Rather we want to paint a descriptive picture of the ways in which virtues and selfnarratives may be thought to exert their causal power. Virtues and self-narratives may well exert their influence on action by influencing beliefs and desires. A virtuous person, on seeing a little girl drowning, will form the desire to save her and believe he can do so by swimming out to her. He will very likely be displaying benevolence and courage, mediated by that belief and that desire.

5. Anscombe Against Causal Theories of Action Elizabeth Anscombe, in her disgust at modern moral theories and views, has called for the proper view of ethics, virtue theory, to be informed by a philosophy of psychology. She writes:

How Self-Narratives and Virtues . . . 77 In present-day philosophy an explanation is now required how an unjust man is a bad man, or an unjust action a bad one; to give such an explanation belongs to ethics; but it cannot even be begun until we are equipped with a sound philosophy of psychology. For the proof that an unjust man is a bad man would require a positive account of justice as a “virtue.” This part of the subject-matter of ethics is, however, completely closed to us until we have an account of what type of characteristic a virtue is. (Anscombe 1958/2005, 174) Anscombe continues: [I]s it not clear that there are several concepts that need investigating simply as part of the philosophy of psychology and—as I should recommend—banishing ethics totally from our minds? Namely—to begin with: “action,” “intention,” “pleasure,” “wanting.” More will probably turn up if we start with these. (Anscombe 1958/2005, 188) She sees ethics, virtue ethics in particular, as continuous with philosophy of psychology, though her proposed order of investigation is to “banish ethics totally from our mind” when we explore intention. Anscombe rejects the causal theory of action because, for her account, intentions need not always be prior to action. In contrast, on Davidson’s account, intention, which is comprised of the belief and conative states of an agent, must be fully formed for an action to be performed. She believes that intention and action sometimes overlap and an intention must come with a fully-formed mental act. As we have already mentioned, there have been numerous improvements upon or amendments to a theory of causation, whether in action theory or in philosophy of science, and our aim is to endorse Anscombe’s emphasis on the importance of the virtues while seeing them as causal in a way that need not entail a nomological account of causation. A second argument of Anscombe’s against the causal theory of action depends upon there not being a need for a strong introspective view of a felt desire to determine that one’s intention leads to action. She writes: The mistake is to think that the relation of being done in execution of a certain intention, or being done intentionally, is a causal relation between act and intention. We see this to be a mistake if we note that an intention does not have to be a distinct psychological state which exists either prior to or even contemporaneously with the intentional action whose intention it is. .  .  . But there was, we can suppose no prior formation of intention, nor is the intention a mental state that accompanies the action. That the action under this

78 David Lumsden and Joseph Ulatowski description, “applying a bit of force,” was intentional, comes out in his explanation, in what he says if someone asks him why. . . . So the application of the extra force was a means to the end which he mentions. But that this was so was formulated after the event. Was what was formulated not there before? All introspection or observation can tell him, we may suppose, is that it seemed jammed and then he acted. (Anscombe 2005, 95) On Anscombe’s account, there is no distinct mental state that exists prior to an action for which it is said to be a part of an intention that causes the action to occur. One may act intentionally without also having a felt or strong desire to do the act. We agree that there need not be a suitable distinct mental state prior to the action of which the actor is introspectively aware. But that leaves room for there being some kinds of prior mental states of the person, perhaps broader long-standing states, that cause the action and account for its intentional nature which are not open to introspection, or not always so. Anscombe’s view suggests that there is no state of an agent from which that person’s actions spring (cf. Anscombe 1957/2001, 1975; Hursthouse 2000; Thompson 2008). This is the point of one of the most vexing sections of her Intention: §19. Any feature which is supposed to be the defining or fundamental feature on the basis of which what someone does is an intentional action must have a particular content which relates suitably to the action, determining which intentional action is performed. Yet, nothing about the agent considered by the agent herself at the moment of acting could possibly determine that content. Perhaps a post hoc analysis may be deployed, but that would be an arbitrary assignment at best. The best possible explanation of our reasons for action, however, is that they are the agent’s own mental states. Anti-psychologistic practical justifications adduce the structure of actions (Thompson 2008, Part II). Why am I buying this book? Well, because I am writing my PhD thesis on Kahlil Gilbran. But the actions they gesture at are often to occur in the future. After all, I have three years to go before I am required to submit my PhD thesis for examination. The future is necessarily hidden from us, and “the best laid schemes o’ mice an’ men gang aft aglay” (Burns 1785). So, actually, my PhD thesis will turn out to be on the philosophical insignificance of Ayn Rand and Libertarian enthusiasts. Thus, anti-psychologistic practical justifications invoke something we do not necessarily know a lot about, and should feel very tentative and insecure about. In contrast, in some situations we are very much consciously aware of our reasons for action. It can be that I have a clear conscious thought that my thesis will be about Kahlil Gilbran and that the book purchase will contribute to that plan. Granted, conscious thought is sometimes unreliable, for we do sometimes deceive ourselves. Maybe, I am really buying this book

How Self-Narratives and Virtues . . . 79 to procrastinate. But our best hope for uncovering a coherent reason for action lies in my thinking at the time, in this case about my thesis topic, rather than what actually occurred later. Our reasons for action cannot really be the structures of our actual actions. Even so, the approach we shall develop does, in a certain way, de-psychologise causal theories of action, much in line with the calculative view of reasons in the structure of action that Candace Vogler proffers in her landmark work, Reasonably Vicious (2002). In the next section, we shift our focus from reasons for action to virtues and narratives as causes of action, while allowing that those causes may operate through the medium of reasons.5

6. Causation by Virtue and Narratives A person’s virtues are part of her character and her self-narrative is bound to embody aspects or manifestations of virtuous character, for virtues make themselves known in behaviour and the self-narrative will reflect prevailing patterns of behaviour to a significant extent, albeit that the narrative may introduce inaccuracies and distortions. The full relationship between virtues and self-narratives is a subtle matter, for different people will have different levels of awareness of their own virtues, or indeed vices. Our theme, though, is to elaborate on the common-sense view that virtues and self-narratives, which are interrelated, causally contribute to actions. We have used Davidson’s causal theory of action as a starting point to look at the way that psychological characteristics of a person may cause action. Davidson’s account rests on his theory of event causation. Applying that account to reasons requires acknowledgement that the beliefs and desires that make them up are often not momentary events; both beliefs and desires can be of long standing. That is not in itself a problem if you consider that the First World War is an extended event and certain beliefs and desires about it are long standing. When we look at the causal role of virtues and self-narratives, though, we are not merely dealing with extended events but with something more complex. We are dealing with the way that behaviour is a result of a complex system. Davidson’s thesis that reasons can be causes relied on a token identity between the reason and its underlying physical character, for it is only under that physical description that it will be covered by a law of nature. Our focus, though, is not on a single state, even a complex one, but a complex system, whose outputs will be sensitive to numerous inputs and internal states. While we endorse Davidson’s thesis that psychological states can cause actions without a reduction of the psychological to the physical, we do not rely in the same way on a token identity between a psychological state and a physical state. Recall Dennett’s mild realism about self-narratives, as well as about mental states generally. That position about self-narratives, which we


David Lumsden and Joseph Ulatowski

wish to apply to virtues too, removes the expectation that there will be some physical description of those things that connects with laws of nature which explain the action under its physical description. Rather, virtues and self-narratives sit within explanatory models that are justified by their usefulness. Virtues and self-narratives are part of a complex system, but not one whose parts systematically match up with physically definable features of the human body. The nature and level of description is not reducible to the language of physical science. Our thesis is that this need not prevent us from applying the notion of cause when employing the conceptual vocabulary of self-narratives and virtues. Consider a simpler system: a motor vehicle. How did Jack manage to avoid the oncoming truck by quickly pulling over, at speed, onto the gravel shoulder of the road and still return quickly to the roadway to avoid the power pole? While Jack’s driving skill is relevant, in this example if it had not been for the excellent handling qualities of the vehicle there would have been a serious, if not fatal, accident. With a cruder vehicle he would have spun off the road, but it is a subtle feature of his vehicle that it could be controlled in such a high-pressure situation. It is natural to say that the good handling of the vehicle causally contributed to Jack successfully avoiding a crash. Think of it as a motor vehicle virtue. “Good handling” is a handy descriptor that relates to vehicle performance but suggests a complex of internal characteristics of the vehicle that produce such performance. It presumably involves such things as: weight distribution, steering geometry, suspension, tyres, and braking systems. Even without knowing about all that, our use of the phrase “good handling” rests on there being some such complex of physical characteristics. As such it can be viewed as causally contributing to the success of the driving manoeuvre, even though good handling’s metaphysical status is very much mild realist. We have not returned to Davidson’s postulation of a token identity. A human being is a much more complex system than a motor vehicle. When we speak of self-narratives or virtues, we speak of subtle features of an extraordinarily complex system. Self-narratives and virtues are not mental states on a par with beliefs and desires but rather are high level features of the whole system. Further, we are taking a mild realist approach to such features. Speaking of them is reaching for an available descriptive terminology that assists explanation and prediction, the best we can do. Let us think of a case where the focus is on human action and capabilities. Consider a fast bowler in cricket, Tim Southee, who manages to bowl out an opposition batsman, Moeen Ali. Perhaps Southee was bowling well and took a number of other wickets, but what was the explanation of the success on that very ball? It will be a combination of Southee’s skill and temperament, Ali’s skill and temperament plus atmospheric and lighting conditions, and even the support of the crowd. Explaining that

How Self-Narratives and Virtues . . . 81 particular success will be elusive, to say the least, unless Ali made an uncharacteristic batting error or the umpires made a particularly bad call. We are a long way from causal relationships being grounded in underlying supporting laws, but this does not prevent commentators from pointing to aspects of the delivery and batsman’s shot choice as contributing causes of the loss of the wicket. Given that there are multiple explanations of contingent states of affairs that fail to provide us with the cause of an occurrence, we shouldn’t believe that any one of them would be a satisfactory account of causation. The truth of the matter is likely to lie in a nexus of causal connections. When we consider the messy causal nexus of human activity, the singular relationship we are interested in is unlikely to reflect a general pattern. If we attribute Southee’s skill as a bowler as the cause of the loss of the wicket, that may single out a salient contributing cause, without that as such being supported by an underlying law. The full scientific account of the event may be intractable in its complexity. We are used to the full explanation being elusive, which does not undermine the interest in, and viability of, a partial explanation such as cricket commentators (or, in another setting, political pundits) can provide. With the source of action in virtues and in narratives we are in much the same situation. We are not alone in wanting to go beyond immediate motivations as causes of actions by identifying some broader, higher level characteristics as the causes. David Velleman (1989) claims that, while human actions are often explained by motivation, they are sometimes explained by habit, by the agent’s character, or by emotion, and those other explanations do not need to always reduce to motivation. Annette Baier (2009) elaborates on that theme with particular reference to character.6 She points us to occasions where the motivations or beliefs themselves demand explanation. She puts it nicely when she writes, in the context of Hume’s discussion of Charles I’s actions, “The kind of person one is helps determine both the sort of desires one has and acts on, and the way one selects advisors and forms one’s beliefs” (Baier 2009, 246). This is not to regard character as fundamental in the explanation of behaviour. For one thing, she allows that character can also be explained, and indeed people will sometimes act out of character, which directs our attention back to particular beliefs and desires. Certainly, she sees a person’s character, while always present, as not the sole source of the action, even where the person is acting in character. A common human nature is also always present, such that often it does not need to be remarked upon to provide an explanation. Also, the different situations that people find themselves in play a major role in producing different life stories (p. 250). Elsewhere, Velleman (2006, Chapter 10) has had things to say about the causal relationship between self-narratives and action which draws on work by cognitive psychologists, with additional discussion in another context (Chapter 11). His starting point is Prescott Lecky’s (1945)

82 David Lumsden and Joseph Ulatowski posthumous volume Self-Consistency: A Theory of Personality. In short, Lecky’s view is that the fundamental principle guiding psychology is consistency, a principle that provides a theory of personality. In order to feel secure, a person needs to build an organised, consistent picture of the world, and central to that is her or his own self-image. This self-image, which captures a conception of the person’s place in the world, can be thought of as a self-narrative. An experience that challenges one’s own self-image needs to be accommodated in such a way as to maintain consistency. Velleman discusses the influence of Lecky’s ideas in subsequent cognitive psychology work concerning, notably, the notion of cognitive dissonance and different accounts of the effects that were found. Experimenters set participants a task they found boring, but then offered them $1 to tell others that it was enjoyable. Those who accepted the offer came to believe it was indeed enjoyable, for who would think of themselves willing to lie for $1? Where the offer is larger the outcome is different. We can say that participants who “sold out” for $1 needed to retrospectively attribute a feeling of enjoyment to make it consistent with what they did. Having made that attribution, it is likely that such a participant would be more inclined to undertake the task in the future. This is merely a suggestive example of how self-narratives might operate in a person’s cognitive structure. Lisa Grover (2013) shares a broadly similar position to ours, although with different emphases and a different presentation. She argues that character traits, virtues and vices are not dispositions to behave but are constituted by narratives. She further endorses the idea that they cause behaviour, citing Flanagan’s (1991, 279) support for the causal role of character traits in producing behaviour. Our reservations mainly concern the way Grover argues for her position. It is puzzling that she appears to cheerfully accept that dispositions can cause behaviour while her strategy for supporting a narrative constitution view of virtues is that it avoids two accounts of dispositions that could be applied to virtues. The first is a conditional view of dispositions and the second is Stuart Hampshire’s (1953) view of dispositions of persons as an indeterminate summary of past behaviour plus the assumption that it will continue into the future. Neither of those options seem to allow for a causal role for dispositions. Our perspective is that we can attribute causal power to dispositions by taking Armstrong’s view of them that they rest on underlying categorical properties. We don’t need to move to a thesis of the narrative constitution of virtue in order to allow for the causal role of virtues. But we see both virtues and self-narratives as causally contributing to behaviour and we are interested in the way that virtues are caught up in a person’s self-narrative. A person’s virtues and her self-narrative are bound to be related to some extent, and their status as causes of action show a strong parallelism, but there are still things to be said separately about the causal power of those two things. The relevance for us is that if we consider that

How Self-Narratives and Virtues . . . 83 a self-image can take the form of a self-narrative, then we have an illustration of how self-narratives may cause behaviour. We saw in Lecky’s experiment how our self-image can affect one’s beliefs and actions, in that case over a rather trivial matter. At a more serious level, our commitments motivate us to act from beliefs that have become firmly embedded in us, either by habituation or by sheer mental force. Virtues and self-narratives, which are closely bound up with such commitments, are standing features of the person that we wish to highlight on occasion as the source of an action, even though they are unlikely to contain within them the full explanation of the action at that particular moment. In his recent book, Michael Lynch (2019) has described a connection between self-narratives and conviction when he writes: Many of our beliefs become convictions because they have already been woven into the larger narrative of a tribe we aspire to remain a part of. They become morally entangled because those beliefs reflect who we think we want to be; and our emotional commitment remains, no matter what challenges come down the pike. Indeed, we take such challenges personally—literally, because we’ve expanded our self-identity, our self-narrative, to include the belief. (Lynch 2019, 71–72) Our own focus is on the more general character of such causal relationships. Whereas Lynch considers the narrative of a tribe dictating the actions of individuals, our focus has been how the self-narrative arises from the convictions of the individual, perhaps sometimes, though not always, independent of any tribe or community. Likewise, we believe that if we consider that a disposition or some inner sense can take the form of a character trait, we have an illustration of how virtues cause behaviour. Our job in this section is to defend a causal account of action that accommodates self-narratives and virtues. Let’s take a look at an example where virtue explicitly does bring about action. Suppose a US Special Forces unit happens upon a terrorist compound in Mali where it is believed that people are being held against their will. While not their primary objective, the unit has to decide whether to engage the enemy and how to do so. The least risky option available to them is to not engage, to make observations of the compound, and to report back to their superiors what it is that they witnessed in the field. Or, following an initial and quick assessment, they may mount a rescue operation, either a stealthy one or an overt one engaging the enemy in a surprise attack. The unit commander has a tricky decision to make and one that will very likely reflect his judgement of risks based on training and it will also reflect his character, including such virtues as courage and compassion. It is unlikely that the most enlightening account of the causes of his decision will be some explicit reasons.


David Lumsden and Joseph Ulatowski

It can be useful to consider cases of bad behaviour too. Consider an athlete behaving badly caught on camera fighting in a bar or a famous director assaulting young women. Again, it is unlikely that such behaviour is best explained in terms of some explicit contemporaneous reasons. Perhaps the explanation can be approached in terms of character, the presence of vice or absence of virtue. Also, it is striking how such behaviour contradicts popular narratives about such people. We’ve come to expect that the narrative one tells of the athlete or the Hollywood director concerns greatness, with respect to one’s athleticism or with respect to one’s being able to entertain the masses. Maybe there is an explanatory route to the behaviour in terms of those individuals’ self-narratives. Their own self-conception as important could be expressed as a sense of self-entitlement. They consider that they do not have to be bound by the rules that govern others. Virtues or a self-narrative, on our account, play a contributing role in the causal nexus of action. It would not be wrong to point to a soldier’s courage as a causal contributor to the soldier under fire rescuing a wounded comrade. The soldier’s heroism, so we might say, may have been partly generated by his relationship with the wounded soldier, by his sense of honour for his country, or (perhaps more pessimistically) by his not wanting to witness someone else’s death. Indeed, one might talk of various character traits such as loyalty, resolution, and so forth that may be highlighted in an explanation of the soldier’s acting so selflessly. This makes us recall Aristotle’s thesis that the virtues form a unity. We can still think of courage as an internal state and suggest that state as a causal contributor to the action even while imagining a fuller and more complete explanation, albeit one that in practical terms is beyond our reach. A similar account applies to self-narratives as causes of actions. Any example is likely to be simplistic as it will be hard to convey a sufficiently rich flavour of the self-narrative in a short space. But consider a person, George, whose self-narrative involves being someone with burgeoning musical talent but who is a victim of ethnic discrimination. When stopped by the police along a relatively desolate part of a Connecticut freeway, he answers all of the questions the officer has for him and adds that the large rectangular case in the car’s back seat contains his beloved vintage guitar, a 1958 Fender Stratocaster Sunburst. Under different circumstances, George may not have believed it important to volunteer such information; however, given that police action may have been guided by discriminatory attitudes or implicit biases that they may harbour toward ethnic minorities, such as George’s, he felt it necessary to alert them to the presence of the Stratocaster, thus establishing himself as a credible musician. Those two features of George’s self-narrative: being a credible musician and one subject to ethnic discrimination, are likely to have contributed to his reaction in those circumstances.7

How Self-Narratives and Virtues . . . 85 When we seek to explain actions as flowing from virtues or from narratives, we need to explore a space unlike both the Davidsonian account of the rationalisation of actions and the purely scientific, neural level, description. To exhibit a virtue in action is not to be persuaded by a practical syllogism, although no doubt virtuous people will from time to time consider such syllogisms. The virtue is a source of the action in a way that reaches down below the level of immediate awareness. What enters consciousness may be simply a gut feeling about what should be done. While virtues are graspable by the human mind, that is not what is effective in acting virtuously. When it comes to acting virtuously, conscious thinking about acting virtuously seems at least inessential and possibly a distraction. Acting in a virtuous way is distinct from following a rule, such as in the deontological approach, in which I consciously follow some moral rule. Acting virtuously is something typically done without conscious thought, though that does not rule out some conscious reflection on what one is doing as a side commentary or post hoc explanation. When I treat someone in trouble or distress with concern and compassion, I will be acting virtuously if that flows from my virtuous character without the need for conscious resolve to act that way. This is notwithstanding that a virtuous agent is bound to be an entity capable of conscious awareness and will almost certainly be consciously aware of what is happening and indeed what she is doing when acting virtuously. It is merely our conscious awareness of the virtuous nature of the action that is likely to be missing. The point concerning our lack of conscious awareness of acting virtuously suggests another option where one’s actions may be derived from habit. In this sense, one acts not from a sense of being compelled to do something or even having specific reasons for performing the action; rather, the action flows from one being habituated into doing it. Finding a lost wallet without yielding to the temptation of stealing its contents is something that has been ingrained in one’s mind. To steal lost property would not only be acting against one’s better interest but also be acting viciously. The person would have to part company from the ways in which she has been habituated to act. The action of returning property to its rightful owner is one of a mindless routine. Aristotle, in Nicomachean Ethics, for example, writes [I]t makes no small difference to be habituated this way or that way straight from childhood, but an enormous difference, or rather all the difference. (1103b, 23–25) The causal infuence of a virtue may indeed have the nature of a habit, generally speaking, but it is important to recall that the habit is making itself felt in a complex mental system which includes various kinds of


David Lumsden and Joseph Ulatowski

conscious awareness. Thus the nature and effectiveness of a virtue must be seen against that background. The capacity to narrativise is characteristic of the way that humans can see themselves and others as persons. But self-narratives are tricky beasts, particularly when seen as sources of action. One significant aspect of narratives is that they often involve emotional colouring, though the nature of that involvement can be complex. To take the example of anger, someone contemplating their own life story may always or typically feel angry about the events as they see them. There also can be a meta-level awareness that they are angry about those matters, which may or may not lead to a calming of the anger. When the angry self-narrative leads to a certain action it is unlikely that it is the more detached meta-level awareness that is doing the work. Rather, the action just springs from the anger, much the way that the virtuous action just springs from the virtue. And, further, one can act out of anger without being aware of that source. A narrative theory of the self has the tendency to operate in an overly simple picture of the mind as a source of action. The importance of emotional colour in the explanation of action has already been mentioned. We need to avoid a too bland conception of narrative as a kind of merely descriptive chronology. That would be to miss the explanatory power that is latent in the narrative approach. Also, there is a tendency to consider that a human life is governed by a singular narrative, such as the one alluded to by Sacks at the outset of this chapter. The better model is as a bundle of narrative threads. This is a view we have argued elsewhere (Lumsden and Ulatowski 2017, 2019), but we are not alone in seeing some kind of complexity along those lines. See Flanagan (1991) for his notion of the multiplex self. To briefly illustrate the theme of multiple narrative threads, consider the father who is a tyrant at work but is putty in the hands of his daughter at home. The work life and the home life have their largely separate descriptive narratives for the events from one day to the next at work tend to cohere and similarly the events from one day to the next at home tend to cohere. There are, of course, points of contact between the two, but there are fundamentally two narrative unities: one for home and one for work. But there is not merely a bland description of events in each case. There are different goals, ambitions, emotions, and senses of self in the two settings and these are integrated into the two narrative threads. And indeed there are likely to be different character traits associated with the two narrative threads. We now see that just as there is a danger in the narrative theory of the self operating with a simplistic model so there is a danger with virtue theory operating with a simplistic model of the self. Virtues may not always be wholly centrally located within a person. Rather, it is plausible that a certain narrative thread is imbued with a certain virtue, while it is absent from other narrative threads within the person. The generosity of spirit that the father displays at home is absent from his work life. A narrative

How Self-Narratives and Virtues . . . 87 thread model is helpful here too. Different character traits, including virtues or vices, can be associated with different narrative threads. Of course, some traits can be held in common. A father might exhibit patience both at work and at home. And even when there are largely different traits, there could be some bleeding of a virtue or vice from one thread to another. The father’s imperious manner from work may rear its head for a moment at home. Even so, it is often of explanatory value to align character traits with narrative threads. The guy out with his mates may not display the same restraint and consideration that he does at home. Or, conversely, the guy out with his mates may be amiable and generous, while taciturn and ungenerous at home.

7. Conclusion The central theme of this chapter is to show how self-narratives and virtues cause actions. We have noted that the cultivation of virtues amongst people, such as soldiers, may yield courageous action like saving the life of his comrade under the most horrendous conditions of war. We have also noted that the story that one tells of oneself may lead the person to perform or to avoid an action. The contributing role virtues and selfnarratives play, however, is different from the role of intentions. Davidson, remember, believed that the pair of beliefs and desires form a reason for someone to take action, and he argued that only after that intention has been fully formed will an agent be able to complete the action. To a certain extent, our view further de-psychologises causal theories of action, much in line with the calculative view of reasons in the structure of action. We have noted that intentions need not precede action for our account of a causal view of action, something that many notable virtue ethicists, such as Anscombe (1957/2001) and Hursthouse (1999), have thought seriously problematic. Moreover, there need not be a token identity between mental and physical events, à la Davidson. This is not how virtues and self-narratives may be in the service of one’s actions; instead, virtues and self-narratives are latent in the causal nexus of action and need not be fully formed and known to the agent. To understand the nature of the causal relationship we need to acknowledge that virtues and self-narratives are high level properties of a highly complex system. We need to take a mild realist stance to virtues and self-narratives, which allows us to provide an explanatory account of their influence on action.8

Notes 1. While we are broadly sympathetic with Sacks’ views about the role of a selfnarrative, we disagree with his emphasis on a singular narrative. In Lumsden and Ulatowski (2017), we argue for a plurality of self-narratives, based largely in the notion of parallel hyperspecialisation, a position we refer to here in Section 6.


David Lumsden and Joseph Ulatowski

2. Virtues as inner causes of action are also not a current focus of attention in the literature in virtue ethics and virtue epistemology. 3. The causal theory of action, sometimes referred to as ‘causalism,’ is very likely the most popular view in the philosophy of action. Davidson’s causal account (1963, 2001a, 2001b), originally written in response to the non-causal account of Melden (1961), served to fill a lacuna in the action theory literature. For good places to start reading on the causal account, see also Davidson (2004), Enç (2003), Mele (1992, 2003), and a recent anthology on it: Aguilar and Buckareff (2010). 4. A challenge for Davidson’s causal account of action is whether the intentional action has to be guided continuously for the duration of the bodily movement. Waving to a friend is an intentional action that takes little time to perform, but making a cocktail takes a little while longer. If the belief-desire pair merely initiates bodily movement but does nothing to guide that movement once it has begun, then it’s unclear that an intentional action is being performed in that bodily movement. The belief-desire pair may be available to guide the bodily movement, but at most we should say that the movements are congenial to the causal account, rather than being a consequence entailed by the belief-desire pair. For more on the challenge, see Bishop (1989). 5. The argument appearing in this paragraph stems from a conversation that Ulatowski had in a graduate seminar nearly a decade ago with Austin Booth, Meg Bowman, Dale Clark, Marissa Lelanuja, Elijah Millgram, Lex Newman, Matt Shockey, and Shelley VerSteeg. While it happened long ago, we are grateful for their discussion. 6. Incidentally, while David Hume is typically seen as an early source of the kind of causal theory of action that we have discussed, Baier argues that his discussions of character are in tune with Velleman’s claim. 7. In another work of ours, we explained how this suggests agents are parallel hyperspecialisers, in addition to being what Millgram (2015) called serial hyperspecialisers (see Lumsden and Ulatowski 2017, 2019). 8. We are grateful for constructive feedback from Paul Bloomfield, Damian Cox, Garrett Cullity, Tim Dare, Ramon Das, Richard Hamilton, Jeanette Kennett, Justine Kingsbury, Cei Maslen, Elijah Millgram, Christine Swanton, Martin Thursby, and Liezl van Zyl. Moreover, we appreciate the generous financial support of the University of Waikato’s Faculty of Arts and Social Sciences Dean’s Office to allow the Philosophy Programme to host a conference in 2017, and Joseph is especially grateful for the support provided by the University of Connecticut Humanities Institute to complete this project.

References Aguilar, J. and A. Buckareff, eds. 2010. Causing Human Actions. Cambridge, MA: MIT Press. Anscombe, G.E.M. 1957/2001. Intention. Cambridge, MA: Harvard University Press. Anscombe, G.E.M. 1958/2005. “Modern Moral Philosophy.” In M. Geach and L. Gormally, eds., Human Life, Action and Ethics: Essays by G.E.M. Anscombe. Exeter, UK: Imprint Academic, pp. 169–194. Anscombe, G.E.M. 1975. “The First Person.” In S. Guttenplan, ed., Mind and Language. Oxford: Oxford University Press, pp. 45–65. Anscombe, G.E.M. 2005. “The Causation of Action.” In M. Geach and L. Gormally, eds., Human Life, Action and Ethics: Essays by G.E.M. Anscombe. Exeter, UK: Imprint Academic, pp. 89–108.

How Self-Narratives and Virtues . . . 89 Aristotle. 1941. Nicomachean Ethics, W.D. Ross, trans. In R. McKeon, ed., The Basic Works of Aristotle. Oxford, UK: Oxford University Press. Armstrong, D.M. 1968/1993. A Materialist Theory of Mind. London: Routledge. Baier, A. 2009. “Acting in Character.” In C. Sandis, ed., New Essays on the Explanation of Action. Basingstoke, UK: Palgrave Macmillan, pp. 241–256. Bishop, J. 1989. Natural Agency: An Essay on the Causal Theory of Action. Cambridge, UK: Cambridge University Press. Burns, R. 1785. “To a Mouse, on Turning Her Up in Her Nest with a Plough.” (August 15, 2019). . Bruner, J. 1986. Actual Minds, Possible Worlds. Cambridge, MA: Harvard University Press. Bruner, J. 1990. Acts of Meaning. Cambridge, MA: Harvard University Press. Casey, E. 1987. Remembering: A Phenomenological Study. Bloomington, IN: Indiana University Press. Crane, T., D.M. Armstrong, and C.B. Martin, eds. 1996/2002. Dispositions: A Debate. Oxford: Oxford University Press. Davidson, D. 1963. “Actions, Reasons and Causes.” Journal of Philosophy 60(23): 685–700. Davidson, D. 2001a. “Causal Relations.” In D. Davidson, ed., Essays on Action and Events. Oxford: Clarendon Press, pp. 149–162. Davidson, D. 2001b. “Mental Events.” In D. Davidson, ed., Essays on Action and Events. Oxford: Clarendon Press, pp. 207–225. Davidson, D. 2004. Problems of Rationality. Oxford: Oxford University Press. Dennett, D.C. 1987. “True Believers.” In D.C. Dennett, ed., The Intentional Stance. Cambridge, MA: MIT Press, pp. 13–35. Dennett, D.C. 1993. Consciousness Explained. London: Penguin Books. Enç, B. 2003. How We Act: Causes, Reasons, and Intentions. Oxford: Oxford University Press. Flanagan, O. 1991. The Science of the Mind, 2nd ed. Cambridge, MA: MIT Press. Frazer, M.L. and M. Slote. 2015. “Sentimentalist Virtue Ethics.” In L. Besser-Jones and M. Slote, eds., The Routledge Companion to Virtue Ethics. New York: Routledge, pp. 197–208. Grover, L. 2013. “Narrative Virtues.” Theoretical and Applied Ethics 2(1): 67–82. Hampshire, S. 1953. “Dispositions.” Analysis 14(1): 5–11. Hursthouse, R. 1999. On Virtue Ethics. Cambridge: Cambridge University Press. Hursthouse, R. 2000. “Intention.” In R. Teichmann, ed., Logic, Cause and Action. Cambridge, UK: Cambridge University Press, pp. 83–105. Lecky, P. 1945. Self-Consistency: A Theory of Personality. New York: Island Press. Lewis, D. 1973. “Causation.” The Journal of Philosophy 70(17): 556–567. Lumsden, D. 2013. “Whole Life Narratives and the Self.” Philosophy, Psychiatry, and Psychology 20(1): 1–10. Lumsden, D. 2013/2014. “Narrative Construction of the Self.” Te Reo 56/57: 161–175. Lumsden, D. and J. Ulatowski. 2017. “One Self Per Customer? From Disunified Agency to Disunified Self.” Southern Journal of Philosophy 55(3): 314–335. Lumsden, D. and J. Ulatowski. 2019. “Casting Light Upon the Great Endarkenment.” Metaphilosophy 50(5): 729–742. Lynch, M.P. 2019. Know-It-All Society: Truth and Arrogance in Political Culture. New York: Liveright Publishing Corporation. MacIntyre, A. 1984. After Virtue. South Bend, IN: University of Notre Dame Press.


David Lumsden and Joseph Ulatowski

Melden, A.I. 1961. Free Action. London: Routledge and Kegan Paul. Mele, A. 1992. Springs of Action. Oxford: Oxford University Press. Mele, A. 2003. Motivation and Agency. Oxford: Oxford University Press. Millgram, E. 2015. The Great Endarkenment. Oxford: Oxford University Press. Mink, L.O. 1970. “History and Fiction as Modes of Comprehension.” New Literary History 1: 541–558. Parfit, D. 1984. Reasons and Persons. Oxford: Oxford University Press. Sacks, O. 1985. The Man Who Mistook His Wife for a Hat and Other Clinical Tales. New York: Touchstone. Schechtman, M. 1990. “Personhood and Personal Identity.” Journal of Philosophy 87(2): 71–92. Schechtman, M. 1996. The Constitution of Selves. Ithaca: Cornell University Press. Slote, M. 2001. Morals from Motives. Oxford: Oxford University Press. Thompson, M. 2008. Life and Action: Elementary Structures of Practice and Practical Thought. Cambridge, MA: Harvard University Press. Van Zyl, L. 2018. Virtue Ethics. London: Routledge. Velleman, J.D. 1989. Practical Reflection. Princeton: Princeton University Press. Velleman, J.D. 2006. Self to Self: Selected Essays. New York: Cambridge University Press. Vogler, C. 2002. Reasonably Vicious. Cambridge, MA: Harvard University Press.


Virtue Ethics, Narrative, and Revisionary Accounts of Rightness Jason Kawall

In response to prominent criticisms of virtue ethical accounts of right action, Daniel Russell (2008, 2009) has argued that these criticisms are misguided insofar as they rest on an incorrect understanding of what virtue ethicists mean by “right action,” drawing on Rosalind Hursthouse’s influential account of the term (1999). Liezl van Zyl (2011) has explored, though not fully endorsed, a similar approach. The response holds that virtue ethicists do not embrace a strong connection between (i) right action and (ii) what any given agent ought to do in a given set of circumstances. Rather, “right action” is a matter of action assessment, and indicates that a given action is morally excellent and praiseworthy. More generally, the proposed account of rightness emphasises both (i) an agent’s past and how she came to be in certain circumstances—is it a result of her own vice or wrong actions? and (ii) the agent’s own future happiness and well-being—will an action be so terrible that her life is marred and ruined? The narrative structure of an agent’s life thus plays a significant role in determining whether an action is right. This revisionary conception of right action is the focus of the current chapter. I first argue that the proposed change in the understanding of “right action” is significant enough that virtue ethicists would be discussing an altogether different concept from other theorists. I further argue that, even when understood as an account of moral praiseworthiness or excellence, this revisionary account of right action faces significant obstacles. In particular, I argue that non-virtuous agents can perform actions that are far more praiseworthy and excellent than many of those characteristic of virtuous agents, even in circumstances that no virtuous agent would find herself in. I then turn to cases where virtuous agents themselves must act while facing terrible circumstances, paying particular attention to the ongoing peace of mind and absence of sorrow that Hursthouse and Russell suggest are necessary accompaniments to right, excellent action. Drawing on work by Van Zyl, I argue that emphasising such peace of mind implausibly limits our scope for right action under difficult or oppressive conditions. I conclude by considering whether these worries may be grounded


Jason Kawall

in Hursthouse and Russell’s embrace of a eudaimonism that encourages a self-absorbed, egoistic account of rightness—a familiar worry for such theories.

1. Objections to the Standard Account Rosalind Hursthouse has developed perhaps the most influential contemporary virtue ethical account of right action: An action is right iff it is what a virtuous agent would, characteristically, do in the circumstances, except for tragic dilemmas, in which a decision is right iff it is what such an agent would decide, but the action decided upon may be too terrible to be called “right” or “good.” (Hursthouse 1999, 79) Given its prominence, we will refer to this account as the standard virtue ethical account of rightness (henceforth “SVAR”).1 The frst clause is well-known and generally treated as capturing the core of the account. The second clause, concerning tragic dilemmas, is intended by Hursthouse to capture scenarios where even virtuous agents are left with only terrible options; we will return to this clause in Section 5. We can begin by considering two broad lines of objection to SVAR that have sparked Van Zyl and Russell’s responses, and the further articulation of their conception of rightness. First, there are cases that suggest that, pace Hursthouse, there can be right actions that no fully-virtuous agent would perform. Consider Robert Johnson’s well-known case of a habitual liar trying to improve himself (2003). He might keep a journal of his lies with an eye towards improvement, he may pause before he speaks to reflect on who would be hurt by his lying and why he is tempted to do so, and so on. These all seem to be right actions, ones the liar ought to perform but also ones that no fully-virtuous agent would perform—after all, a virtuous agent would not have these temptations, and thus no need for such actions. The second line of objection suggests that there are cases where a flawed agent ought not to try to do what a virtuous agent would do in the circumstances; in other words, such agents should not try to do what is right. For example, suppose that a squash player with a foul temper has just lost a match. While the fully-virtuous agent in such circumstances would walk over to shake the victor’s hand to congratulate her, it seems that our flawed agent should not do this—if she did, she would surely give in to her temptation to punch the victor in the face. Thus, the flawed agent should not try to do what is right (according to SVAR) and should instead perhaps force a smile and walk away before the situation escalates.2

Virtue Ethics, Narrative, and Rightness


The two lines of objection can be summarised as follows: if we apply SVAR, it seems there are right actions that non-virtuous agents should not try to perform (which suggests that these are not in fact right actions for them), and there are non-right actions that non-virtuous agents should perform (which in turn suggests these are in fact right actions for them).

2. Revisionary Accounts of Rightness We can turn now to Van Zyl and Russell’s responses. As noted previously, Van Zyl herself is not committed to SVAR; instead, she simply explores what she takes to be promising lines of response to familiar objections to such accounts. Russell, on the other hand, embraces SVAR, embedded in his own richly developed virtue ethics. Hursthouse, Van Zyl, and Russell all draw on the prominent strand in virtue-ethical thought that such deontic notions as rightness, obligation, and permission do not fit comfortably within such theories. Hursthouse suggests that virtue ethics can provide an account of right action, But it does this under pressure, only in order to maintain a fruitful dialogue with the overwhelming majority of modern moral philosophers for whom “right action” is the natural phrase. “Right action,” with its suggestion of uniqueness, its implication of “if not right then wrong,” and its associations with “required/obligatory,” “forbidden/prohibited,” and “permissible,” is not a term it is happy with. It favours talking in terms of good action (eupraxia), of acting well (or badly), rather than in terms of right action. (Hursthouse 1999, 55) Consequentialists and deontologists typically think of right actions in terms of permissions or obligations. On such understandings, if an action is not right then it is not permissible to perform it; one should not perform an action if it is not right. On the other hand, if an action is right, it is at least permissible to perform it—and obligatory on some accounts. Russell and Van Zyl hold that virtue ethicists instead embrace a very different conception of right action, one where right action is a matter of moral excellence and praiseworthiness, having nothing to do with permissions or obligations. Furthermore, when this difference is recognised, the previous apparent counterexamples are seen to fail. According to Russell, virtue ethics offers not only a different account of right action but, indeed, a different conception of it. Unless we appreciate what is different about it, we risk simply begging all sorts of questions against it, wondering why the new account does not fit the old conception. The assumption


Jason Kawall that “ought” implies “right” is a case in point, since to make that assumption is to fail to engage virtue ethics on its own terms. (Russell 2008, 302)

The core of the alternative conception of rightness is captured by Hursthouse: [A right act can be understood as] an act that merits praise rather than blame, an act that an agent can take pride in doing rather than feeling unhappy about, the sort of act that decent, virtuous agents do and seek out occasions for doing. (Hursthouse 1999, 46; quoted by Van Zyl 2011, 83) Intuitively, the alternative conception holds that only certain actions warrant the status of “rightness,” of moral excellence. On the other hand, the mere fact that an action is not right does not mean that it is wrong or impermissible; rather, this simply shows that it does not yet reach the high standards demanded for rightness. As Van Zyl puts it, “if we use ‘right action’ in the sense of a ‘good deed,’ then it becomes obvious how there can be instances of actions that are not right and yet still ought to have been done” (Van Zyl 2011, 84). On this rival conception of rightness, then, rightness is an honourific that suggests praiseworthiness, and the absence of grounds for regret. An action could be permissible or even obligatory—and recommended to an ordinary agent—without yet being good enough to qualify as right. The previous putative counterexamples thus fail. The liar should engage in his remedial actions, but these do not yet qualify as right—they are not sufficiently excellent. And while congratulating the winner after a tough loss might be praiseworthy and right, it does not follow that a sore loser with a foul temper should attempt to do so. She should instead smile and walk away, an action that is appropriate given her flaws, but not excellent enough to qualify as right.

3. Different Conceptions or Different Concepts? An initial worry is that with this revisionary conception of rightness, our virtue ethicists have actually changed the subject and are no longer engaging with their consequentialist and deontologist rivals; they may no longer be discussing the same concept. Elizabeth Anscombe (1958) and others have, of course, argued that virtue ethicists would be wise to avoid appealing to such deontic notions as right action, obligation, permissibility, and so on (at least in the absence of a divine lawgiver). But it is one thing to reject the notion of right action altogether (given its associations with obligation and permissibility), and quite another to use the very same term to identify an entirely different concept from that shared by other theorists. This latter approach would seem to muddy the

Virtue Ethics, Narrative, and Rightness


waters and confuse discussion needlessly. If so, it would likely be best to introduce a new or different term altogether. But as it stands, the current revisionary conception of right action risks describing a new, very different concept from that being discussed by other theorists. One rather minor consideration supporting this worry is that typically praiseworthiness or excellence is treated as a matter of degree—an action can be more or less praiseworthy or excellent. On the other hand, obligatoriness and permissibility, the notions commonly at stake for consequentialists and deontologists, seem to be binary; it is unusual to speak of an act being more or less permissible. This may suggest that different concepts are at stake: one a matter of degree, the other not. Notice next that permissibility and obligatoriness are, on standard deontic logics, interdefinable; to hold that an action is permissible is to hold that it is not obligatory not to perform it, and so on. Insofar as deontologists and consequentialists are treating right action in such terms, they are plausibly seen as analysing and providing different conceptions of the same concept. But praiseworthiness is not related in the same ways to obligation or permission; indeed, Van Zyl and Russell rely on this very fact. But to the extent that there is such a difference, we may well wonder whether different concepts are at stake. Similarly, deontologists, consequentialists, and commonsense morality would hold that we ought to attempt to perform right actions; a key function of rightness is to provide action guidance. For example, “One should always do the right thing” is a familiar platitude of commonsense morality. The proposed virtue ethical understanding of rightness would strip away this function, with “rightness” serving only as a term of act assessment. Again, this suggests not merely different conceptions of a shared concept, but instead entirely different concepts. Most importantly, as noted previously, Van Zyl and Russell suggest that virtue ethicists mean something like praiseworthy or excellent by “right.” But then why speak of right actions at all, rather than simply speaking of praiseworthy or excellent actions? After all, deontologists and consequentialists recognise the concepts of praiseworthy and excellent actions; they simply treat these as distinct from rightness. Virtue ethicists muddy the waters by speaking of “right action” when the philosophical conversation among their rivals is in terms of permissibility and obligation, and with a concept of rightness that provides action-guidance. They could instead embrace a claim like the following: An action is morally praiseworthy or excellent if and only if it is what a virtuous agent would characteristically do in the circumstances. By making this change, Hursthouse, Van Zyl, and Russell could be understood as fully embracing the Anscombian thesis that we should eschew talk of permissibility, obligatoriness and related deontic notions (in the


Jason Kawall

absence of a divine lawgiver). They could argue that we should instead focus on appraisals of praiseworthiness or moral excellence; certainly this would clarify their view signifcantly. On the other hand, to do so would be to abandon Hursthouse’s original attempt to engage fruitfully with deontologists and consequentialists with respect to right action.

4. Non-Virtuous Agents and Excellent Actions While Hursthouse, Van Zyl, and Russell are perhaps correct to draw greater attention to morally excellent or praiseworthy action, it is worth assessing SVAR when understood on its own terms as an account of such actions. We can focus on instances where ordinary, flawed agents must act in circumstances that no virtuous agent could find herself in; all three authors would argue that such actions cannot be sufficiently or appropriately excellent to qualify as “right.” I will argue that such proposals ultimately lead to implausible assessments, even when treating rightness in terms of moral excellence. Recall the actions of Johnson’s self-improving liar. Both Van Zyl and Russell agree that there might be aspects of the liar’s actions that are truly morally excellent and praiseworthy—he is embracing a commitment to the virtues, he is demonstrating perseverance, prudence, and other virtues in maintaining his efforts (at least to some degree), and so on. But they also hold that such actions are not yet right. Why not? Van Zyl suggests that insofar as the habitual liar works on improving his character and strives to become more honest he does what a virtuous—courageous, determined, persistent— person characteristically does in facing difficult challenges (which is to seek professional help when needed, to persist until the problem is resolved, to not give up hope, etc.), thus giving his action a tick of approval. However, it cannot be an unqualified tick, for we must also remember that the challenge he faces is that of overcoming his own flawed character. (Van Zyl 2011, 11) There is something truly excellent about his behavior . . . his actions do reveal a certain amount of courage and determination. But the central virtue in question here is honesty, and by writing down his lies (and so on), the reforming liar does not act in a way that is characteristic of an honest person, at least not yet. (Van Zyl 2018, 107) Thus, while there are virtuous and praiseworthy aspects to the liar’s ameliorative actions, we cannot yet deem them right because they are made necessary by his own faws and errors; they are not the sorts of actions

Virtue Ethics, Narrative, and Rightness


a good person would seek out and desire to perform. Furthermore, the liar’s actions express only such secondary virtues as determination, rather than the key virtue at stake, honesty. Russell suggests that the liar’s actions are indeed morally good, but do not qualify as right because this term is reserved for “central” cases of moral excellence, where an action qualifies as a central case of moral excellence if and only if it is what a virtuous agent would characteristically do in the circumstances. [W]e need to understand the excellence of unqualified cases of excellent action before we can understand why certain remedial actions count as improvements and, thus, as having any sort of excellence at all. However, we do not need a view about remedial actions in order to understand why the unqualified excellent actions are excellent. Consequently, it seems clear both that the unqualified cases of morally excellent action are central cases of morally excellent action and that those cases are necessarily characteristic of the virtuous. (Russell 2008, 310; see also Russell 2009, 56) Russell thus posits an explanatory priority to the actions characteristic of the fully virtuous, which in turn justifes restricting the term “right” to morally excellent actions of this kind. Russell further defends this proposal, arguing that acts like those of Johnson’s habitual liar have real moral merit, even though they are acts that no virtuous person would do . . . to undertake remedial acts in good faith is admirable and praiseworthy—and even, in a sense, virtuous—because it is to (begin to) accept standards of virtue as standards of one’s own. (Russell 2009, 129) Thus, on Russell’s view a crucial element to the praiseworthiness of the remedial actions of non-virtuous agents is that they refect a broader acceptance of the standards of virtue by these agents. A key question for Van Zyl and Russell is whether non-right actions can ever be more morally excellent than morally right actions as they understand them. Notice how trivial are many of the characteristic actions of virtuous people, even when acting virtuously. They will honestly pay the correct bus fare, prudently brush their teeth, and justly refrain from taking exceptionally large slices of cake when others are yet to have any. These all seem to be decent, permissible, and in some cases obligatory actions. But do these truly strike us as central cases of moral excellence, worthy of moral praise?3 Compare such cases to the following. Imagine a troubled youth, Henri, becomes involved in a gang committing various petty crimes. Henri participates quite happily, working his way up within the group as a trusted


Jason Kawall

member. Because of his trusted status he is now taken onboard a new project. Unbeknownst to Henri, the gang leaders have decided to escalate to truly heinous crimes;4 he is enlisted to guard a terrified kidnapped family that the gang plans to kill later, regardless of any ransom paid. He is shocked. He had no qualms about petty theft and vandalism, but would never participate in a kidnapping, let alone murdering innocent people. As such he risks his own life and helps the family to escape; he will now need to run from the gang with his own life in grave danger, and perhaps seek help from the police while facing prison time. He takes on enormous risk and sacrifice to help the family. Of course, no virtuous persons would ever be in such circumstances, because they would never have been happily stealing from others and would never come to be in a trusted position in such a gang. As such, Henri’s actions do not qualify as right according to SVAR—there can be no characteristic action of a virtuous person in such a situation.5 But surely this is a morally praiseworthy and excellent action. We may judge harshly his previous criminal actions but saving the family at tremendous risk and cost to himself seems to warrant praise. Is Henri’s action less morally excellent and praiseworthy than a virtuous person refraining from taking too large a slice of cake? To say “yes” would seem to require a rather unusual understanding of praiseworthiness and excellence. Many of the actions of virtuous agents could be more praiseworthy than Henri’s action—but surely not every action where some minimal exercise of virtue is involved, no matter how trivial, must thereby be more admirable than these other actions simply because virtuous agents would characteristically perform them. True, Henri has made many mistakes in the past, and still has a flawed moral compass; he is no saint. But why should this so undermine the value of his action here and now, when in this moment his sense of justice, benevolence, and compassion all contribute to the extraordinary willpower and courage it takes to save the family? He is risking death at the hands of the gang, and probable lengthy prison time, but still he acts. To press a bit further—compare Henri’s action and courage with that of a virtuous agent who exercises a modicum of courage by not panicking when confronted by a cute, but growling, puppy. In which case is there a more demanding and admirable exercise of courage? For all the terrifying ferocity of angry puppies, presumably Henri’s action still demands rather more courage. The vast majority of people could, one hopes, muster the courage required to confront a growling puppy. On the other hand, we might expect most people to fall short when placed in Henri’s position—many would panic, others would obey the gang out of fear, and so on. The courage and cool-headedness demanded in Henri’s situation is exceptional. Consider also what is at stake. Henri is acting to save lives, while putting his own life at serious risk and facing years of imprisonment. The angry, growling puppy likely threatens a small

Virtue Ethics, Narrative, and Rightness


bite on a finger. Thus both (i) the demands placed upon the virtues and (ii) the goods (and harms) at stake seem far more significant in the case of Henri’s actions.6 How might defenders of SVAR respond? They might argue that we could, in fact, imagine a virtuous agent in similar (though perhaps not identical) circumstances to Henri—where she could rescue a kidnapped family, though at great risk to herself; if so, we might then see Henri’s actions as right even on SVAR, assuming that a virtuous agent would act in a similar fashion. But compare the following two situations. In both an agent’s family is kidnapped, and the agent is able to rescue them herself, with no enduring physical or emotional harm done to them. Next, imagine that in one case the kidnappers have chosen the agent’s family largely at random, or simply based on their apparent wealth. In the other, the family is deliberately targeted as an act of vengeance against the agent who is a key figure in a powerful gang—someone who has engaged in brutal violence herself and crossed many violent criminals over the years. At one level we could say both agents face similar circumstances of needing to save their kidnapped families. But I expect defenders of SVAR (and the revisionary account of rightness) would not want to treat the actions of both agents as right and praiseworthy; indeed, they would presumably hold that the criminal agent deserves little or no credit at all for rescuing her family when it is her own activities which have led to them being kidnapped in the first place; they would presumably not see the criminal’s rescue as worthy of being treated as right. If the previous is correct, it would in turn suggest that there would be a significant difference (in their eyes) between the case of Henri and the case of a virtuous agent in a situation where she could similarly save a kidnapped family, insofar as Henri (much like the criminal) arrived in this situation due to his own past bad actions—these are what led him to be in this position, and no virtuous agent would be in such a situation. On the other hand, I would argue that while both Henri and the violent criminal discussed earlier are led to their current situation by their past wrong actions, there is a tremendous difference in the badness of their past actions. Henri’s past wrongdoings could still leave him in a position where he can perform an admirable, morally excellent action (even if no virtuous agent could be in a truly similar situation, and even if we should properly blame him for his past wrongdoings). Perhaps Russell would claim that Henri’s case is not a central case of moral excellence, and as such is not right; nor would it be as praiseworthy as a right action. But then do we really believe that standing up to a puppy is such a central case? As noted previously, virtuous agents perform a wide variety of actions in their lives, from the morally extraordinary and excellent, to the quite mundane, where only a minimal exercise of virtue is required.7 It seems a mistake to treat all of the characteristic actions of virtuous agents as exemplifying moral excellence, given such


Jason Kawall

wide variation. Henri should feel guilt for his past pretty crimes. But surely this is not enough to render his current action less worthy than a virtuous agent’s not being scared by a growling puppy. Russell stresses that a theory of right action should answer to our practical interests: [w]e are interested in a theory of right action in the first place because we have serious concerns of a very practical nature—deciding how to act, assessing what we do, thinking about outcomes, having good reasons, and so on. (Russell 2009, 39) But then most moral agents are fawed and still aspiring to virtue; apologies, remediation, and making amends would seem to be key aspects of moral life. The world is a diffcult and often unjust place, and humans are fawed, fnite beings. If we are attempting to inculcate the virtues in our children, would not amelioration, amends, and remediation be essential—and thus “central”?8 Russell might claim that explanatory priority is crucial here, that we need “central” cases of moral excellence in order understand the lesser excellence of actions like those of Henri. But notice that to correctly explain and understand apologies, making amends, and so on—central elements of real-life morality—we will require reference to errors, mistakes, and the like. Ramon Das forcefully captures a related worry: I believe that severing the connection between rightness and action guidance is profoundly unsatisfying and a sign of theoretical desperation. In particular, it raises the following disturbing question. . . . What is the point of an ethical theory that does not even pretend to be useful to a person trying to decide what he morally ought to do? (Das 2015, 339)9 While our concern here is the concept of right action in particular, rather than virtue ethics as a whole, Das’ point seems apt: what are we to make of a narrow, non-action-guiding sense of “right action”—one that would apply only to a narrow range of actions where everything happens to go well, where the agent is left content and proud, and whose proponents stress is not intended to tell us what we ought to do?10 The question is particularly pressing given Russell’s claim that a theory of right action should address a wide range of important practical interests, including deciding how to act. An alternative response for Russell would be to hold that the characteristic actions of virtuous agents express a full commitment to the virtues, while remedial and other actions performed by more flawed agents merely express (at best) an aspiration towards, or the beginnings of such a commitment. Russell writes

Virtue Ethics, Narrative, and Rightness


for the virtue ethicist some cases of excellent action [i.e., those performed by virtuous agents acting in character] are actions that express a wholehearted or unqualified commitment to the standards of the virtues, such as honesty and charity, as standards of one’s own. Remedial actions, then, are assessed in terms of the extent to which they approximate to a commitment to such standards of the virtues. (Russell 2009, 56) [V]irtue ethics can take “what a virtuous person would do” as a necessary feature of any central case of moral excellence, expressing full commitment to standards of virtues, and extend its analysis of excellent action based on such central cases to less central cases, such as remedial and other self-improving actions. (Russell 2009, 57)11 We can focus on two concerns with such a proposal. First, it seems unlikely that all characteristic actions of virtuous agents express a full commitment to the standards of the virtues. When virtuous Claire prudently brushes her teeth does this truly—somehow—express a full commitment to the standards of prudence? Does she thereby brush in a particularly effective or heartfelt fashion, such that her toothbrushing better expresses prudence than Henri’s rescue of the kidnapped family expresses courage? Such claims would require signifcant defence. Second, why should we hold that a broad, full commitment to the virtues (or even just relevant virtues) would be required for right action in any particular instance? That is, it is unclear why Claire’s more general commitment to prudence and its demands across various situations would be required for the rightness of her particular toothbrushing here and now and to render it a central case of moral excellence. A rather different response would stress that while Henri’s action (or that of the brutal criminal who saves her kidnapped family) may seem strikingly excellent when taken in isolation, we must also recognise the broader context of Henri’s life and how he came to be in this position. Concerning a similar case (where an individual makes a decent decision in a situation that no virtuous person would fall into), Hursthouse writes In vain do I protest that now “I want to do what is right,” to undo past wrongdoing and start afresh with a clean slate. I should have thought of that before. The “satisfactory review of my own conduct” is not for me, or at least, not yet; remorse and guilt are my portion. (Hursthouse 1999, 50) Certainly, there are cases where such an appraisal seems apt, as with the brutal criminal. Or suppose Javier is meant to be closely watching as


Jason Kawall

several children play in the shallow end of a pool. He instead drinks too much and dozes off. He is awoken at the last minute by the panicked screams of the children, and by sheer luck is able to save a child who strayed into the deep end. Javier would hardly be considered a praiseworthy hero for saving a drowning child—indeed, it seems he should be ashamed and held blameworthy for his reckless, negligent actions. The history and background of how an agent came to be in a given situation can be highly relevant to assessing an action’s excellence (or lack thereof). Russell suggests that virtue ethics “views actions within the history of an agent,” while other theories “tend to focus on actions one at a time” (Russell 2009, 40fn3). Still, even if we often need to place an agent’s actions in the broader context and narrative of her life in order to properly assess them, we should not simply assume that if a virtuous agent could not find herself in certain circumstances, then there can be no right, morally excellent actions available. While a full discussion of this issue lies beyond the scope of this chapter, we can, for now, introduce two factors that would seem relevant to such assessments. First, we can consider to what extent the non-virtuous agent is responsible for the particular circumstances in which she finds herself. If, like Javier, it is her own negligence, laziness, and recklessness (or vice more generally) that puts her in a particular situation, it seems more likely that any actions she performs will be tainted, and simply making amends for the problems she herself has created. But in cases like that of Henri, even while a virtuous agent would never have ended up in his current circumstances, it is the gang leaders who create the circumstances at stake (they have kidnapped the family and established the gang’s plans). In such cases, where the flawed agent is not directly responsible for the immediate circumstances in which she acts, it seems more likely that genuinely excellent actions are possible, even if a virtuous agent would not be in such circumstances. Second, we might consider the proportionality (as it were) between an agent’s past errors and her current circumstances. Suppose you were to make an insulting comment to a rude stranger in a shopping mall in the United States—a comment that a virtuous person would never make. The stranger turns out to be violent and vicious; he furiously threatens you and other shoppers with a concealed gun. Yes, you played a key role in bringing about this very situation, and it was your own short temper that led you to make the triggering comment. But surely if you are then able to talk this person down, and to convince him to release the other shoppers before turning himself in to the police, your actions are both right and clearly morally excellent—even if no virtuous agent would be in such circumstances. Your own actions played a key role in creating the terrible circumstances you faced—but you were also extremely unlucky, and the goodness of what you then achieve seems easily to outweigh the minor peccadillo on your part that sparked the situation.

Virtue Ethics, Narrative, and Rightness


To summarise the preceding points: in considering the actions of agents in situations no virtuous person would be in, it is implausible to assume these must be lesser or non-central cases of moral excellence. The details of the cases will matter, but surely many of them would be more excellent than many of the quotidian, minimally virtuous actions of virtuous agents—contrary to the standard virtue ethical account of rightness and moral excellence. On the other hand, if Hursthouse, Russell, and Van Zyl were to accept that some actions performed by non-virtuous agents could be more morally excellent than right actions (even if “tainted” in some way), it becomes much harder to find a significant moral role for the proposed revisionary concept of “rightness.” This would be especially troubling given the central role played by notions of right action in everyday moral discourse.

5. The Role of Circumstances A difficult issue lurking in the background of our discussion is that of how to determine whether a virtuous agent might find herself in given circumstances. Relatedly, how do we determine whether two sets of circumstances are relevantly similar enough to qualify as the same circumstances for the purposes of SVAR? Consider two individuals who have both been in prison for several years, convicted of violent felonies. One is innocent and wrongly convicted, the other guilty and justly convicted. They both intervene one day to prevent an assault against a vulnerable new prisoner. In aiding the new prisoner, do they face the same circumstances? We might imagine an innocent virtuous agent being wrongly convicted and imprisoned, and thus in a position to aid a vulnerable prisoner. On the other hand, presumably no virtuous agent would be guilty of a violent felony and would not end up in prison in this way. If these suggestions are correct, then the innocent prisoner could perform a right action, while the guilty prisoner, even while helping to prevent an assault in exactly the same fashion, could not perform a right action—his action would be tainted because we can directly trace his current circumstances to his own past misdeeds, ones that no virtuous agent would have committed. Next consider the guilty felon following his release from prison. He struggles to find a job, can barely afford a cramped, dirty room in a rundown tenement, and his friends and family have largely abandoned him. Suppose that he still somehow manages to save money to visit his young son from a past relationship, tries to provide his son with needed school materials, etc. These actions require tremendous sacrifices on his part— he often goes without proper heating or meals in order to provide for his son. But in the end, his circumstances are due to his own past actions—it is because of the violence he committed many years ago (that no virtuous agent would have committed) that he ended up in prison, and it is


Jason Kawall

because of this violent past that he struggles to find work now. Suppose, further, that he is not yet virtuous—he still struggles with his temper, occasional dishonesty, and other vices. Do we need to consider his current actions as less than right, because even while we might admire his sacrifices and dedication to his son, no virtuous agent would have found herself in this position? And are his sacrifices less morally excellent than a virtuous person’s prudent flossing of her teeth because it is ultimately his own fault that he is in these difficult circumstances? One could make such claims, but surely to do so would be holding an agent’s past actions against him in deeply troubling ways. Must this individual be forever trapped by his past, such that his many (seemingly) admirable actions for the sake of his son and their relationship are deemed tainted and less than right because we can trace his struggles back decades ago, to something a virtuous agent would not have done? Yes, he should feel guilt and remorse for his past violent actions—but should these past actions also continue to prevent him from performing a wide range of good, right actions for decades after? The issues here are difficult, and it could well be that defenders of SVAR could develop plausible, fully-fledged accounts that provide clear guidance as to those instances where a virtuous agent could find herself in a given set of circumstances (and when not), and so on. But for now, such work remains to be done. For example, recall Van Zyl’s suggestion that Johnson’s habitual liar acts as a virtuous agent would when facing a difficult challenge (not giving up, seeking help as needed, etc.). This seems plausible enough, but at an abstract level we could describe almost any dilemma faced by a non-virtuous agent as “facing a difficult challenge” and thus circumstances a virtuous agent could face; presumably Van Zyl would not want her proposal to extend so far. We may wonder whether the innocent and guilty convicts face the same circumstances when aiding the vulnerable new prisoner, given the very different histories which led them to this point.12 More generally, we require further elaboration on these and related issues from proponents of SVAR in order to be able to more fully assess the view and its prospects.

6. Tragic Dilemmas In preceding sections, we have considered potential cases where the actions of non-virtuous agents seem both right and more morally excellent than many characteristic actions of virtuous agents, even if virtuous agents would never have followed such paths or allowed themselves to fall into such circumstances. We have also noted the looming background questions concerning how to determine whether a virtuous agent could find herself in a given situation, and how long past vicious actions might constrain the rightness of current actions.

Virtue Ethics, Narrative, and Rightness


In this section, we turn to cases in which virtuous agents act, but in terrible circumstances not of their own making—situations where virtuous agents would likely be left unhappy and haunted by what they must do. Hursthouse and Russell emphasise the impacts of such actions upon the agent’s own life and her future well-being in determining the rightness of her actions. In particular, they suggest that no right actions are available in tragic dilemmas. Van Zyl departs from this view, and we will draw upon her work to argue that such actions are in fact often right or even supererogatory, despite (or even in virtue of) the great demands placed upon the agents involved. To begin, consider the case of a military doctor working at the frontlines of a battlefield, needing to amputate both legs of a severely injured young soldier without anaesthetic. One could well imagine the doctor being haunted and distressed by the horrors he witnesses and the actions he must perform. But recall Hursthouse’s more general characterisations of right action. She holds that a right action is an act that merits praise rather than blame, an act that an agent can take pride in doing rather than feeling unhappy about, the sort of act that decent, virtuous agents do and seek out occasions for doing. (Hursthouse 1999, 46) She also holds that a right, morally excellent action is one that leaves an agent who performs it in those “circumstances (so) requisite to happiness,” namely “inward peace of mind, consciousness of integrity, (and) a satisfactory review of (her) own conduct” as Hume so nicely puts it. (Hursthouse 1999, 47) Consider again our battlefeld doctor, horrifed at the suffering he is witnessing, dismayed at the limits of what he can do to help and the further suffering he inficts while treating the injured, and so on. While I think most of us would see the actions of the doctor as morally right and strikingly excellent, it is not clear that Hursthouse can do so. The doctor is not happy about the actions he is performing; he would not seek out occasions to perform them—he would far rather hope that they were never necessary again. He will likely lack peace of mind as he is haunted by the cries of those upon whom he has operated, and the limits of the help he could provide. Or consider a veterinarian who euthanises an older dog that arrives at an animal shelter frail and severely neglected. There is very little chance of the dog being adopted and the shelter’s meagre resources are already stretched far too thin. Her action may well be right and require significant virtue to choose and to perform—but it hardly seems likely that she


Jason Kawall

would eagerly seek out opportunities to perform such actions, or feel a serene inner peace and pride as a result. I do not take cases like that of the battlefield doctor or the veterinarian to constitute clear counterexamples to Hursthouse or Russell’s claims concerning right actions. But they do raise significant questions concerning their more general understanding of right action and its grounding in the agent’s own well-being or flourishing. Hursthouse’s discussion of tragic dilemmas is relevant in this regard. Recall her account of right action: An action is right iff it is what a virtuous agent would, characteristically, do in the circumstances, except for tragic dilemmas, in which a decision is right iff it is what such an agent would decide, but the action decided upon may be too terrible to be called “right” or “good.” (Hursthouse 1999, 79) For Hursthouse and Russell,13 tragic dilemmas are cases in which an agent is faced with a choice between two truly dreadful options—perhaps as in Bernard Williams’ case of Jim and Pedro, where Jim must choose between shooting one innocent person himself thereby saving 19 lives, or refusing to shoot, and having Pedro, a vicious army commander, shooting all 20 (Hursthouse 1999, 59–60). Even if we think this dilemma is resolvable—perhaps there are clear moral reasons in favour of shooting the one individual in order to save the other 19—Jim is still left taking the life of another human being. Note that while ordinary non-virtuous agents can fnd themselves in such tragic dilemmas as a result of their own past mistakes and ill deeds, it is also possible for both them and virtuous agents to fnd themselves in such terrible circumstances through no fault of their own.14 Hursthouse holds that the lives of agents who face tragic dilemmas will be marred and filled with sorrow, even if they do what they should in the circumstances. Jim, as a virtuous agent, will be haunted by having killed an innocent human being—even if this was the best option available to him at the time (Hursthouse 1999, 73–74). She writes The actions a virtuous agent is forced to in tragic dilemmas fail to be good actions because the doing of them, no matter how unwillingly or involuntarily, mars or ruins a good life . . . simply in virtue of the fact that her life presented her with this choice, and was thereby marred, or perhaps even ruined. (Hursthouse 1999, 59) Tragic dilemmas, for Hursthouse and Russell, present further cases in which no right, morally good action is possible.

Virtue Ethics, Narrative, and Rightness


Van Zyl has argued—highly effectively to my mind—that Hursthouse and Russell are wrong to hold that agents in tragic dilemmas cannot act rightly, and that they in fact have the resources to treat the virtuous agent’s actions in such dilemmas as right, even excellent actions. Consider the following passage from Hursthouse: The charitable, honest, just agent, even when faced with a tragic dilemma, does not act callously, dishonestly, that is “as (in the manner) the callous, dishonest, unjust agent does.” She acts with immense regret and pain instead of indifferently or gladly, as the callous or dishonest or unjust one does. So we are not forced to say that virtuous agents faced with tragic dilemmas act badly. (Hursthouse 1999, 73–74; as cited by Van Zyl 2007, 53) Hursthouse goes further. We might be tempted to think that a virtuous agent who acts in a tragic dilemma cannot be acting in character (and thus her action would not be right). But Hursthouse writes it would not be correct to describe a virtuous agent who resolved a tragic dilemma rightly as thereby “not herself.” If anything, she might have to have been quite especially herself, calling on all her virtue and moral wisdom in order to resolve the dilemma rightly in the first place. (Hursthouse 1999, 62) But then, as Van Zyl observes, The difficulty for Hursthouse’s account is to explain the sense in which the act is “terrible,” given that she does not think it is morally wrong or blameworthy. If it is true that the agent is a virtuous person, that she finds herself in this situation through no fault of her own, and that she is blameless of wrongdoing, then we seem to have a strong case for assessing the act in question as right or virtuous, as warranting a tick of approval, regardless of the fact that the doing of the act will fill the agent with sorrow and could ruin her life. (Van Zyl 2007, 54) Indeed, given the previous claims from Hursthouse, we might expect a virtuous agent in a tragic dilemma to act in an especially praiseworthy fashion, drawing on “all her virtue and moral wisdom.” In a 2006 paper, Hursthouse provides a potentially attractive account of supererogation for her virtue ethics: If what makes it hard (on a particular occasion) for an agent to act virtuously is the circumstances, then the more virtue he shows if he


Jason Kawall acts well. . . . But: if what makes it hard (on a particular occasion) for an agent to act virtuously is (not the circumstances but) a flaw or imperfection of his character, then the less virtue he shows if he acts well. And this distinction between right action where virtue is not severely tested and right action where virtue is severely tested and comes through, is the virtue ethics account of the distinction between “the obligatory” and “the supererogatory,” a distinction that no account of the right can ignore. (Hursthouse 2006, 111)

Hursthouse has us imagine a virtuous agent who fnds a purse full of money. Suppose that the virtuous agent is reasonably well-off, and it is easy for her to return the purse to its owner—her virtue is not severely tested by the circumstances. In such a case the agent acts rightly and well but has done nothing exceptional. Now imagine the virtuous agent to be living in signifcant poverty, and facing large, unexpected bills that she worries she cannot pay. Here when she returns the purse, it is a particularly praiseworthy and excellent action—it would qualify as supererogatory insofar as the agent’s virtue would be severely tested, and many ordinary individuals would succumb to the temptation to keep the purse. This all seems quite plausible but consider how it ought to apply to tragic dilemmas. As Hursthouse herself notes, to be able to deliberate and find the strength to do something one finds so terrible—and yet so necessary—reflects virtue that has been put to a severe test. That is, it seems that many of the most demanding situations, where an agent’s virtue is most severely tested, will include tragic dilemmas and other cases where she might be haunted even by doing what is best; recall the battlefield surgeon and the veterinarian. Hursthouse (1999) holds that the actions in tragic dilemmas are not good enough to be deemed right. And she might say the same about the battlefield doctor and vet if their lives are sufficiently marred and haunted by the procedures they perform. But applying her proposed account of supererogation (2006) would instead suggest that these same actions are supererogatory; for the reasons discussed previously, I believe this is the more appropriate appraisal. Similarly, I would argue that the actions of Henri should also qualify as supererogatory, as an instance where a non-virtuous agent has gotten himself into a situation where extraordinary acts of virtue are now required. Recall the situation where you’ve insulted a rude individual who then threatens you and others with a gun. To stay calm and engage in dialogue to deescalate the situation would be very difficult and challenging even for very virtuous agents, let alone the rest of us. As such, it seems the door is open to agents performing supererogatory actions even in situations that no virtuous agents would have gotten themselves into. To summarise: Hursthouse and Russell ought to treat many of the actions of virtuous agents in tragic dilemmas, and in circumstances like

Virtue Ethics, Narrative, and Rightness


those of the battlefield doctor or veterinarian, as supererogatory. These are extraordinary, demanding circumstances where many would fail to act rightly, let alone act in a still more excellent fashion. Further, given the extreme demands of courage and other virtues required by Henri’s actions in saving the kidnapped family, his actions—and ones like them— should also qualify as supererogatory, even if no virtuous agent would find herself in such circumstances.

7. Right Action and the Good Life As we have seen, Hursthouse and Russell hold that right, morally excellent actions are ones that would leave an agent with peace of mind and a sense of integrity; they are the sorts of actions that they would be happy about and would seek occasion to perform. It is this broader understanding of right action that leads them to claim that virtuous agents cannot act rightly in tragic dilemmas. What grounds this emphasis on the agent’s own mental states and happiness in determining whether an action is right? One plausible answer would lie in their embrace of eudaimonism.15 Hursthouse writes “Good action” is so called advisedly, and although it is conceptually linked to morally correct (right) decision and to “action of the virtuous agent,” it is also conceptually linked to “good life” and Eudaimonia. (Hursthouse 1999, 59) In the end, it seems that Hursthouse and Russell want to hold that good, right actions must ultimately contribute to the agent’s own well-being and good life. Excellent action is grounded in the virtues—and in turn, the virtues are traits that reliably lead to the fourishing or eudaimonia of their possessors. As such, they may well hold that actions that leave an agent’s life marred or even ruined cannot be right, excellent actions. Still, notice that in cases where a virtuous agent would characteristically perform an action that might leave her unhappy or sorrowful, but not to the same extent as in tragic dilemmas, Hursthouse would hold that right action is still possible. She suggests that a resolvable dilemma which arises in circumstances in which a virtuous agent might well find herself will be resolvable by a morally right decision, and what is done, such as “x, after much painful thought, feeling deep regret, and doing such-and-such by way of restitution” will be assessed as morally right. (Hursthouse 1999, 41) Perhaps Hursthouse would hold that the cases of the battlefeld doctor and veterinarian are of this kind—these individuals can act rightly, so


Jason Kawall

long as they act with regret, after painful thought, and so forth. But this then raises several questions. First, why is right action available in these cases, but not in tragic dilemmas—why can virtuous agents not act virtuously, with regret, and rightly in tragic dilemmas? Second, how awful must a required action be in order to qualify a situation as a tragic dilemma; how severely marred or ruined must the agent’s life be, and what constitutes such ruining? Must the agent be constantly haunted and unable to think of anything else? Is a life sufficiently marred if the virtuous agent remembers the dreadful action every few days and is filled with sorrow for an hour or two, before bringing herself back to the present? Third, we might think that the marring or ruining of a life following a tragic dilemma is simply a descriptive fact—something that will often or inevitably happen. But Hursthouse suggests, more strongly, that such ruining should occur: For those who insisted on the appropriateness of guilt and remorse in these cases were surely right to insist that the mere fact that one had intentionally done x should haunt the rest of one’s life if x were very terrible, even granted that one was blameless. (Hursthouse 1999, 61) On one hand, this might seem appropriate—anyone with a decent character should surely be horrifed and haunted if they were put into a position where they had to torture another human being or perform an equally heinous act. But further questions arise: would Hursthouse allow that a virtuous agent might appropriately try to end such haunting—would it be wrong to try to move on with her life, or to consult with a therapist? Presumably Hursthouse would accept such steps, though this is not entirely clear.16 We might then say that in tragic dilemmas such sorrow would be justified or warranted, even if agents could properly take steps to help themselves move past what they’ve done. But what should we say of cases in which an agent’s life is marred and ruined by a lesser dilemma? For example, suppose our battlefield doctor is in fact haunted and tormented by what he has had to do, to such an extent that his life is severely marred. Are his actions no longer right? Do they remain right, even though his life is ruined, because the haunting and sorrow is excessive compared to what would be warranted, however this is established? That is, Hursthouse holds that a virtuous agent’s life would be marred by acting in a tragic dilemma and thus the action would not be right. But what should we say about cases where virtuous agents have lives that in fact are marred or ruined by their actions, but where the situations faced by these agents are not yet tragic dilemmas by her standards? More fundamentally still, why impose a concern with the agent’s own happiness or sorrow, or her own warranted pride (or lack thereof) into

Virtue Ethics, Narrative, and Rightness


our very conception of right or excellent action? There is a vast range of right actions with widely varying characteristics. Some are easy to perform but warrant little pride. Others arise in difficult situations and tragic dilemmas, ones that may lead to great sorrow on the part of the agent. But as Van Zyl (2007) has argued, they are still right, and quite possibly all the more admirable given the strength of character it would take to act rightly in such circumstances. Hursthouse might suggest that it is not the virtuous agent’s warranted pride or peace of mind that is fundamental to determining whether an action is right and excellent, but rather that these states reflect the nature of the actions that the agent has performed—and that those actions performed in tragic dilemmas are so terrible that they cannot be deemed right. The virtuous agent’s sorrow would simply serve as an indicator of how terrible the action was. But it is not clear that this response is open to her. Hursthouse explicitly focuses on how performing terrible actions would impact the agent herself; devastating outcomes and catastrophic failures are all filtered through their impact on the agent’s own life. The following passages from Hursthouse are representative: The actions a virtuous agent is forced to in tragic dilemmas fail to be good actions because the doing of them, no matter how unwillingly or involuntarily, mars or ruins a good life. (Hursthouse 1999, 59 [emphasis added]) [Following acting in a tragic dilemma] a virtuous agent’s life will be marred or even ruined, haunted by sorrow that she had done x. Here again, we arrive at a situation that deserves to be called “tragic”— not because the dilemma was irresolvable, but because, resolving it correctly, a virtuous agent cannot emerge with her life unmarred. (Hursthouse 1999, 61) It seems that what is crucial for Hursthouse in considering whether an action can be right in such circumstances is the impact upon the agent’s own well-being or fourishing. We may develop the current worry for Hursthouse and Russell in the form of a dilemma. On one hand, if the relevant object of the virtuous agent’s regret and sorrow are the circumstances that made a certain action necessary, then many—perhaps even most—of the actions of virtuous agents would not be right or excellent by their standards. If right actions are the kinds of actions that virtuous agents seek out opportunities to perform and leave such agents happy and with an inner peace of mind, then only a narrow range of actions could be right. Consider our battlefield doctor—he surely wishes that he would never have to perform another surgery on horrifically injured young people and may well be


Jason Kawall

haunted by the terrible procedures he had to perform. A virtuous agent may give generously to Amnesty International or other groups—but only because there is so much injustice and oppression in the world. She would regret the existence of such states and would wish her actions were not necessary. More generally, most morally right and excellent actions would seem to be made necessary by circumstances of suffering, injustice, dishonesty, oppression, and so on. Virtuous agents could feel truly happy, without a residue of sorrow or regret, only about a very narrow range of actions and circumstances—such things as providing gifts for friends, or spending time with loved ones—instances where one is not attempting to address bad or flawed circumstances. Surely this would restrict the range of right, excellent action too far, and on an implausible basis. On the other hand, perhaps the locus of sorrow and haunting in tragic dilemmas and other difficult situations should be only the agent’s own actions, rather than the broader circumstances which necessitated them. There is something attractive to this—we may think it inappropriate to focus on broader circumstances which are outside of the agent’s control, and in the end, are not part of her action itself. Thus, the question would be whether I did something of which I could be proud and happy, regardless of the broader circumstances. The battlefield doctor might think What I had to do was terrible . . . performing amputations without anaesthetic, trying to remove shrapnel while my patients screamed. They were pleading with me to stop and I had to listen to it all. This will haunt me forever. How will my life ever be the same? Presumably his actions would not leave him the peace of mind needed to be deemed right or excellent. But the worry is that such an approach to right action could well be highly self-absorbed and egoistic—charges that eudaimonism has long faced. Indeed, this approach might seem to confirm some of the worst worries of eudaimonism’s critics.17 The suffering of others, the terrible harms that may arise are not directly the reason why a given action would not be right in such circumstances—instead it is that the virtuous agent herself would now be haunted, that her life would be marred. We put aside the sorrow directed at the circumstances in which virtuous agents find themselves—the fact that soldiers are being injured or dying, that political prisoners are being tortured, and so on. All such sorrow and haunting would need to be filtered through their impacts on the agent herself. We would be concerned only with whether the agent should regret her own actions, or on the fact that things have gone so badly that her life is marred or ruined.18 But it is not at all clear why this should be a criterion of morally right action.19 Doing what is right and morally excellent can require diminishing or even sacrificing our own flourishing or eudaimonia, leaving us haunted

Virtue Ethics, Narrative, and Rightness


and scarred. Hursthouse is, of course, aware of this—she herself draws attention to cases where the right action is to act with regret, after inner struggle, and so on. Van Zyl draws attention to the following passage from Hursthouse: The claim is not that possession of the virtues guarantees that one will flourish. The claim is that they are the only reliable bet—even though, it is agreed, I might be unlucky and, precisely because of my virtue, wind up dying or with my life marred or ruined. (Hursthouse 1999, 172; as cited by Van Zyl 2007, 57) Hursthouse thus readily acknowledges that the virtues do not guarantee a good life; instead they only typically or reliably contribute to such a life. But if so, and as Van Zyl stresses, this allows us to accept that some right, virtuous actions may detract from an agent’s own personal well-being, leaving her exhausted and sorrowful, while remaining right, excellent actions. Indeed, such actions may be all the more admirable, precisely insofar as they demand such strength and sacrifce.

8. Conclusion Virtue ethics, particularly in its eudaimonistic forms, would have us situate and understand the actions of an agent within the broader course and context of her life. This understanding is taken to be essential to properly assessing her actions and determining what she ought to do. How did she come to face the circumstances she does? How will her actions influence the future course of her life? The revisionary account of rightness embraced by Hursthouse and Russell, and carefully explored by Van Zyl, embodies this concern. Is the agent acting in circumstances a virtuous agent could find herself in, or has the agent, through her own activities, placed herself in a situation the virtuous would avoid? When the agent acts, will she be left doing something that will leave the rest of her life marred or ruined by sorrow, or is it an action that will leave her with peace of mind, the kind of action that she would want to perform? I have attempted to show that this revisionary account of rightness requires significant refinement and fleshing out; indeed, we might do well to instead consider alternative virtue ethical positions. The revisionary account risks developing a concept of rightness entirely different from that at stake in other ethical theories; virtue ethicists would thus fail to engage with these others. Furthermore, the revisionary account assumes that if a virtuous agent could not find herself in certain circumstances, then no right, truly excellent action would be available. But we have considered a range of cases where the actions done in such circumstances seem both right, and more morally excellent than many characteristic actions of virtuous agents. Finally, and following work by Van Zyl, the


Jason Kawall

revisionary virtue ethical account of rightness risks being excessively and implausibly egoistic by focusing on the peace of mind and flourishing of the agent herself in assessing the impacts of her actions, and more generally risks failing to acknowledge that many of our most admirable, excellent actions may leave us sorrowful and haunted.20

Notes 1. Of course, other prominent virtue ethical approaches to right action have been developed, such as Christine Swanton’s important target-centred theory (2003). 2. Variations of the squash player example have appeared in several papers by different authors. Van Zyl has traced the original example back to Watson (1975, 210). 3. For discussion of such cases in the context of supererogation, see Kawall (2009). Note that there can be significant value to ongoing acts of maintenance and caring—cleaning, cooking for one’s loved ones and so on. Here we can focus on merely prudent or other less significant actions. Thanks to Kate Norlock for helpful discussion. 4. Indeed, it could be in part due to Henri’s own great success in committing various petty crimes, his loyalty and ingenuity, that the gang has grown larger and more successful, such that the leaders now feel emboldened to attempt far more ambitious crimes. 5. Svensson and Johansson (2018) argue that on the most plausible version of SVAR, and given a plausible understanding of the relevant conditionals, the problem in such instances is instead that all actions would be right. Very roughly: if no virtuous agent could be in a given set of circumstances, then anything follows. 6. Recall Van Zyl’s suggestion that the habitual liar’s remedial actions express secondary virtues, but not honesty, where this is the central virtue at stake (2018, 107). Notice this does not apply to Henri. When saving the family he is acting just as a courageous, benevolent agent would, where these are the central virtues at stake. 7. Many will require no virtue at all, though these are not relevant here. 8. See Das (2017, 103–104) for further worries concerning Russell’s proposal. 9. It is worth noting that Das’ critique applies most clearly to Russell’s particular approach; one could sever the link between rightness and action guidance without neglecting the latter (as Russell seems to do). 10. Van Zyl (2011, 2018, 107–108) presents an alternative proposal for action guidance for proponents of SVAR, in part suggesting that we ought not to act as a vicious person would. The approach is promising but would still leave unclear the role and value of the virtue ethical concept of “right.” 11. Strictly, Russell suggests only a necessary condition for central cases of moral excellence in this passage. But SVAR itself makes clear that being what a virtuous person would characteristically do would also constitute a sufficient condition for such excellence. 12. Svensson and Johansson (2018, 497) suggest, on behalf of virtue ethicists, that a virtuous agent could face almost any set of circumstances, no matter how awful or reflective of vice, if we assume that she faces these circumstances due to actions performed prior to her becoming virtuous. It is a clever proposal, though not one that they themselves endorse. It raises difficult questions concerning whether agents could be virtuous moments after committing a heinous crime, and so on.

Virtue Ethics, Narrative, and Rightness


13. Russell embraces Hursthouse’s approach to these issues, drawing extensively on her work in his own discussion (2009, 49–57); the concerns raised here should apply to both. 14. While there is a distinction between resolvable and irresolvable dilemmas (where in the latter case there is no rational basis for preferring one option to the other), this will not be relevant for our purposes. Hursthouse treats tragic dilemmas, whether or resolvable or not, in the same fashion. 15. Van Zyl (2007) explores the possibility that Hursthouse’s approach to tragic dilemmas is driven by the belief that certain actions—perhaps killing innocents for pleasure—are always prohibited. These actions would remain prohibited even when a virtuous agent must perform one of them when facing a tragic dilemma. But as Van Zyl notes, it seems clear that what virtuous agents would do (particularly in terms of their motives and reasons) would never be the same as any vicious, prohibited action. As such there seems no need for Hursthouse not to allow for right action in tragic dilemmas. 16. Van Zyl (2007, 58) raises similar questions. 17. See, for example, Solomon (1988). 18. It seems unlikely that we can so neatly distinguish between sorrow over one’s own actions and the situation that led to them. Can we truly separate the battlefield doctor’s sorrow and despair over the suffering he has caused while operating, or the limits to the help he could provide, from his sorrow and despair at what he has witnessed more generally? 19. Similarly, imagine the doctor reacted differently to his actions, such that he would not be haunted by his actions. It would seem strange that this difference, in particular, could change the status of his actions to being right, and that his own well-being should play such a key role in determining whether his actions were right and excellent. 20. Earlier versions of material in this chapter were presented at the Central Division Meetings of the American Philosophical Association, and the University of St Andrews Philosophical Society. I would like to thank the audiences on these occasions for much helpful feedback; many thanks in particular to Margaret Holmgren, Sam Iam, Lisa Jones, Brian McElwee, and Kate Norlock for their comments and suggestions, and especially to Liezl van Zyl for her careful, thoughtful comments on an earlier draft of this chapter.

References Anscombe, G.E.M. 1958. “Modern Moral Philosophy.” Philosophy 33(124): 1–19. Das, R. 2015. “Virtue Ethics and Right Action: A Critique.” In L.L. Besser and M. Slote, eds., Routledge Companion to Virtue Ethics. London: Routledge, pp. 331–344. Das, R. 2017. “Virtue, Reason, and Will.” In N. Birondo and S.S. Braun, eds., Virtue’s Reasons: New Essays on Virtue, Character, and Reasons. New York, NY: Routledge, pp. 99–114. Hursthouse, R. 1999. On Virtue Ethics. Oxford: Oxford University Press. Hursthouse, R. 2006. “Are Virtues the Proper Starting Point for Morality?” In J. Dreier, ed., Contemporary Debates in Moral Theory. Oxford: Wiley-Blackwell, pp. 99–112. Johnson, R.N. 2003. “Virtue and Right.” Ethics 113(4): 810–834. Kawall, J. 2009. “Virtue Theory, Ideal Observers, and the Supererogatory.” Philosophical Studies 146(2): 179–196.


Jason Kawall

Russell, D.C. 2008. “That ‘Ought’ Does Not Imply ‘Right’: Why It Matters for Virtue Ethics.” Southern Journal of Philosophy 46(2): 299–315. Russell, D.C. 2009. Practical Intelligence and the Virtues. Oxford: Oxford University Press. Solomon, D. 1988. “Internal Objections to Virtue Ethics.” Midwest Studies in Philosophy 13(1): 428–441. Svensson, F. and J. Johansson. 2018. “Objections to Virtue Ethics.” In N. Snow, ed., The Oxford Handbook of Virtue. Oxford: Oxford University Press, pp. 491–507. Swanton, C. 2003. Virtue Ethics: A Pluralistic View. Oxford: Oxford University Press. Van Zyl, L. 2007. “Can Virtuous People Emerge from Tragic Dilemmas Having Acted Well?” Journal of Applied Philosophy 24(1): 50–61. Van Zyl, L. 2011. “Right Action and the Non-Virtuous Agent.” Journal of Applied Philosophy 28(1): 80–92. Van Zyl, L. 2018. Virtue Ethics: A Contemporary Introduction. London: Routledge. Watson, G. 1975. “Free Agency.” Journal of Philosophy 72(8): 205–220.


Virtuous Perception A Gibsonian Approach Richard Paul Hamilton

1. Introduction To be virtuous is to be constituted as a kind of rational agent, namely one who is exceptionally responsive to moral reasons. The virtuous person does not, however, learn to be responsive to moral reasons in a peculiarly moral way but rather she acquires this capacity while acquiring a range of competencies associated with being a member of a community, such as the ability to recognise norms and conventions and to participate in conversation. This chapter argues that virtue represents a refined form of social competence. It proceeds from an unobjectionable assumption: the virtuous are exemplary because they are better able to reliably grasp the moral demands of a situation and act accordingly. We should follow these moral exemplars and recommend that others, particularly the young, emulate them. I will push the unobjectionable assumption in a potentially more controversial direction. It is, I believe, useful to think of this refined social competence in visual, or more broadly, perceptual terms.1 The virtuous person sees the demands of a situation more clearly than the non-virtuous.2 This may tempt us to think of such vision in a quasi-mystical fashion, for example, as Iris Murdoch appears to do in the following passage: Our activity of moral discrimination cannot be explained as merely one natural instinct among others, or our “good” identified with pleasure, or a will to live, or what the government says (etc.). The possession of a moral sense is uniquely human; morality is, in the human world, something unique, special, sui generis, “as if it came to us from elsewhere.” It is an intimation of “something higher.” The demand that we be virtuous. It is “inescapable and fundamental.” (Murdoch 1992, 26) This invitation to mysticism should be declined. If it were the case that the capacity for virtue represented “something higher” in the sense to which Murdoch alludes, then it would not make sense to talk about emulating the virtuous, as opposed to admiring or even envying them. In this


Richard Paul Hamilton

chapter, I will argue that the best way to resist the temptations of Platonism is to eschew traditional, representationalist theories of perception and adopt something like a Gibsonian (or 4E) approach to perception, that is, a direct (non-inferential) model. On such an account, the virtuous quite literally see the moral demands a situation. Their acting in accordance with virtue in turn provides a criterion by which we determine that they have in fact perceived the situation correctly. The form of moral perception supported in this chapter is a refined form of social perception. Social settings place numerous social demands upon the agent, such as the demands of etiquette and situational appropriateness. Moral demands or, more specifically, the requirements of virtue are those aspects of a situation that call upon us to act in ways which form and sustain an integrated moral character and enable us to lead the sort of life to which a virtuous person would aspire to live. They are of a greater level of gravity than many of the ordinary demands a social setting poses. In proper Aristotelian spirit, however, one should not seek to radically distinguish between the demands of etiquette and those of ethics. How one uses a knife and fork is not in and of itself ethically significant. Nevertheless, a boor, who goes out of their way to offend others or intentionally disregards propriety rules by misusing a knife and fork acts badly. If such behaviour is habitual then indeed the person should be considered vicious. By the same token, being virtuous is more than simply acknowledging, and acting upon, customary social norms. Someone suffering from an anti-social personality disorder may recognise social norms, including some ethical norms, because she sees how others acting in accordance with these norms may benefit from them. A conformist living in a corrupt society may follow the norms of that society unquestioningly and in doing so may become complicit in its corruption. Neither the person suffering from an anti-social personality disorder nor the conformist living in a corrupt society have cultivated a virtuous character. The virtuous person recognises that the requirements of virtue sometimes require the observance of customary norms and at other times setting them aside in pursuit of higher demands. I begin by setting out Gibson’s original theory and introduce some objections and refinements to it. I then explore how the theory can be expanded to encompass social perception, as Gibson himself suggested. (Gibson 1979/2015, 120) Broadly speaking, in Gibsonian terms, social perception involves detecting “affordances” presented by social settings, that is, seeing other’s actions as calling for us to act or refrain from acting in particular ways. Following on from this, I develop my central thesis that moral perception can be considered as a refined form of social perception and how a Gibsonian/4E view of perception makes this claim intelligible and protects it against some common objections to direct moral perception. I then link Aristotle’s thoughts on training and

Virtuous Perception


habituation with Alfred Schutz’s insights regarding the habituated nature of mundane social knowledge and locate the idea of virtuous perception in this setting. I suggest, however, that times of moral crisis require more than merely habitual responses. The virtuous person has acquired the requisite skills to move beyond mere habit and is able to see more clearly than the non-virtuous how to resolve the situation satisfactorily.3 I conclude with a discussion of how my proposed approach addresses some common objections to naturalistic virtue ethics.

2. Benefits and Caveats I will not be arguing in favour of a direct theory of perception per se as I believe that work has already been done better elsewhere, to the extent that such theories are now, if not mainstream, at least respectable contenders.4 I will be arguing, instead, that the reasons that might incline us to accept direct theories of perception in other areas should make us equally well inclined to accept an account of virtue as direct moral perception. At the very least, I hope to show that there is no radical separation between moral perception and other kinds of complex perception. The primary advantage of a virtue as direct moral perception approach is parsimony. As Sarah McGrath has argued, defenders of inferentialist accounts of moral knowledge have to expend considerable resources to render their view remotely plausible. In defending a non-inferentialist view she suggests that once one is sufficiently liberal about the possible contents of perceptual knowledge, to the point that we can literally have perceptual knowledge that someone is acting with a certain intention (and so on), it is hard to insist, in a principle, not ad hoc way, that moral facts must always remain on the outside looking in. (McGrath 2018, 165) Another benefit of the virtue as direct moral perception approach is that it allows for a form of moral realism which does not require us to posit bizarre entities such as “moral properties with objective prescriptivity” and thus defuses the so-called Placement Problem.5 The Placement Problem, which involves questions about how to place normativity within the supposedly disenchanted universe revealed by physics, has been with us since at least the Scientific Revolution but has received sustained attention in recent years. Since, on the Gibsonian account, perception is the detection of meaningful ecological wholes rather than an interaction with the objects described by classical physics, there is nothing anomalous or metaphysically queer about the perception of moral norms since they form part of the fabric of our social landscape. To paraphrase an argument made by Richard Norman (1997): if there is nothing odd about


Richard Paul Hamilton

directly perceiving “a friend” or a “goal scored according to the rules of Association Football,” then there is also nothing odd about directly perceiving “a friend in need” or “a friend behaving badly who needs making aware of that fact.”6 The virtuous person follows the right norms in the right circumstances and sets aside those norms that conflict with the demands of virtue. For instance, a visitor to South Africa, who is committed to animal welfare, may be invited to participate in a traditional Xhosa cattle slaughtering ceremony.7 While he recognises the need to be culturally sensitive and respectful to his hosts, he will also exercise practical wisdom to find a way of declining the invitation. He may take it as an opportunity to engage in a dialogue with them in which he explains his reasons. If both parties are operating in good faith, they may reach some sort of understanding which preserves, and may even strengthen, whatever relationship they may have. While the conformist may simply go with the flow and the sociopath will do whatever brings him the most advantage, the virtuous person will seek a just resolution which respects the dignity of his hosts and his own deeply held principles. One may accept all this but still question whether “seeing” moral reasons is an appropriate form of expression here.8 To this objection, I have two, admittedly quick, responses. The first is an ordinary language point. “Seeing” is a factive for which there are success criteria, e.g., one sees or fails to see something. If we were to choose some other term, such as “interpret,” then different criteria apply. Interpretations are not generally considered “correct” or “incorrect.”9 They are more or less plausible, more or less coherent, or more or less persuasive. In the virtue literature, there is a broad consensus that the virtuous person is recognised by their capacity to get things right. That is, most virtue ethicists are objectivists for whom it makes sense to talk about the virtuous person who gets moral judgements right. The language of seeing situations correctly conveys this much better than alternatives. Second, if we do not perceive moral character primarily through perceiving the actions and speech of others, then there is a distinctive nonperceptual means we possess to access them.10 Do we access to them by some process utterly different from how we access other features of the world, including the social world? Perhaps, like the Platonist may suggest, it is a faculty to access mathematical and logical truths. But this non-perceptual access has more than a whiff of the mystical about it. As I have suggested previously with regards to the quotation from Murdoch, it is difficult to comprehend why the virtuous agent would be worthy of emulation if she simply possessed some quasi-mystical gift for intuiting the moral features of a situation. Since this form of intuition is unavailable to the rest of us, then we would not possess the means of exercising it. If, however, the virtuous agent has trained herself by becoming a better observer, then this is something everyone may emulate. For these reasons,

Virtuous Perception


some form of perceptual account seems plausible. What remains is the question whether such a perceptual account should be construed in traditional representationalist, inferentialist terms or direct, non-inferential perception. This chapter will recommend the latter.

3. J.L. Gibson’s Ecological Theory of Perception vs. the Orthodox View There are two broad approaches to perception and cognition. The first is the computationalist and representationalist view, which may be called the “orthodox view” since it has been dominant in the perception literature since the decline of behaviourism in the late 1950s. The orthodox view posits a gap between perception and cognition which is filled by internal mental representation.11 On this view, action and speech are the outputs of a lengthy process which begins when inputs from the environment are received and then processed internally. Though theories of perception within this traditional approach have become increasingly more sophisticated in the last few decades they all retain a commitment to the idea that perception is something that takes place inside the head of the perceiving subject. The second is not just one approach but a set of approaches which have gathered under the banner of 4E Cognition. “4E” stands for “embodied, extended, enactive, and embedded.” Proponents of 4E Cognition reject representationalism, albeit in slightly different ways,12 and they regard perception as an active process in which the perceiving subject engages with the world. While 4E Cognition involves a range of diverse approaches, Gibsonian ecological psychology is central. For the purpose of this chapter, I will focus largely on the seminal work of J.L. Gibson since it is not possible to understand subsequent developments in 4E Cognition without understanding his contribution to it, and neither Gibson’s original work nor the refinements that have followed from it may be familiar to most people working in moral and political philosophy.13 Although the view has evolved from Pragmatism, has clear parallels with the phenomenology of Maurice Merleau-Ponty (1945/2012) and has affinities with Marxist theory, Gibson’s Ecological Theory of Perception developed largely as a response to the experimental inadequacies of conventional indirect theories of perception. During the 1950s, Gibson was struck by the fact that traditional approaches to perception were based upon the image of a photographic or movie camera. In the Introduction to his now classic work, Gibson suggests that: [t]he text books and hand books assume that vision is simplest when the eye is held still, as a camera has to be, so that a picture is formed that can be transmitted to the brain. Vision is studied by first requiring the subject to fixate a point and then exposing momentarily a


Richard Paul Hamilton stimulus or a pattern of stimuli around the fixation point. I call this snap shot vision. (Gibson 1979/2015, 1)

However, as Gibson points out, this is far from how visual perception operates in naturalistic settings. We are told that vision depends on the eye, which is connected to the brain. I shall suggest that natural vision depends on the eyes in the head on a body supported by the ground, the brain being only the central organ of a complete visual system. When no constraints are put on the visual system, we look around, walk up to something interesting and move around it so as to see it from all sides, and go from one vista to another. That is natural vision. (Gibson 1979/2015, 1) As Merleau-Ponty (1945/2012) was also arguing around the same time, traditional empiricist accounts of vision (which are Gibson’s primary target) regard visual perception as a largely passive phenomenon. In experimental settings popular at the time, the subject is restrained and prevented from moving his head or body. By contrast, vision that occurs when we walk through an environment and turn our heads from side to side is what Gibson calls “ambulatory.” When we are standing still or sitting, we scan the visual feld by moving our heads, and Gibson calls this “ambient” vision. According to Gibson, a flawed experimental method feeds a flawed theory of perception which then in turn generates further flawed attempts at testing it. A quick summary of that flawed view is that it fails to properly distinguish perception from sensation. According to this erroneous theory, what happens when we perceive is that our sensory faculties respond to such things as photons. From this input, we then mentally construct a model of the world. At the philosophical level, the view implies a commitment to a particularly crude form of physicalism. When we perceive the world, what entities we make contact with in the world are just as they are described by the physical sciences. Such physicalism notoriously generates numerous conceptual muddles, not least of which is the Placement Problem or how we fit back into this picture the normativity upon which even the physical sciences depend. That Gibson’s theory of perception avoids these problems means it is something that moral philosophers should at least consider. The force of anti-realist arguments, especially J.L. Mackie’s (1977), rests upon the idea that we should be suspicious of any form of moral realism that posits a mysterious perceptual faculty devoted to intuiting equally mysterious moral properties. As I will show in the next section,

Virtuous Perception


Gibson’s view removes much of the mystery here. On Gibson’s view, there is no chasm that separates moral perception from ordinary perception. If ordinary perception does not involve the sensing of the objects of physics, then the idea that moral norms cannot be found in the world described by physics should not trouble us unduly. We would be hard pressed to find friendships or football games there.

4. The Foundations of Gibson’s Ecological Approach Ask someone what to expect of a theory of perception and they will likely come out with something along the following lines: a theory of perception should explain how subjects make contact with objects which are in the world. Obviously, as it stands, this is rather sketchy so we require a bit of work to fill out the gaps. Who or what are the subjects? Well, obviously, they are human beings, a large range of animals, and perhaps some AI systems. The objects are the objects of classical physics, since we do not perceive quantum events. What would “making contact” with such objects look like? The traditional answer involves things like photons cascading onto the optic nerve, which are converted into signals transmitted to the brain. Once in the brain, the task is to then convert these signals into as accurate as possible a representation of the objects in the world. As Gibson suggests, this traditional answer presents itself as scientific common sense, yet it is based upon the strange idea that the environment “communicate[s] with the observers who inhabit it.” This, he suggests, is absurd: Why should the world speak to us? The concept of stimuli as signals to be interpreted implies some such nonsense as a world-soul trying to get through to us. The world is specified in the structure of the light that reaches us, but it is entirely up to us to perceive it. The secrets of nature are not to be understood by the breaking of its code. (Gibson 1979/2015, 57) The standard view generates a more immediate conundrum: we do not perceive the objects of classical physics, points extended in space; rather, we perceive meaningful wholes. This is a teapot, that is a cup, and so forth. Nominalist accounts since the Middle Ages would have us grouping the elementary inputs under conventional labels. After all, only someone who has learned the words “teapot” and “cup” would be able to recognise these objects as such. So, it seems obvious that perception is a rather elaborate process with cognition at the very end point.14 In a passage which Gibson quotes with approval, one of the founders of Gestalt Psychology, Kurt Koffka, writes “[e]ach thing says what


Richard Paul Hamilton

it is.  .  .  . [A] fruit says ‘Eat me’; water says ‘Drink me’; thunder says ‘Fear me’; and woman says ‘Love me’” (Koffka 1935 cited in Gibson 1979/2015, 129).15 Objects present themselves to us as things toward which we stand in actual or potential relations. Moreover, those relations are not the passive ones of “mere” observation. Though there are occasions when we merely observe things for the sake of observation. This is the exception rather than the rule, despite what standard theories of perception teach. Long before I learn that this is called a “cup,” I learn that one can drink from this kind of object. As Wittgenstein reminded us we do not typically learn words passively (Wittgenstein 1953/1999, §§1–20); rather, we learn them in the course of learning those activities in which they are characteristically embedded. For example, to learn cricket is pari passu to learn the language of cricket. Gibson suggests that Gestalt psychology correctly identified a problem with conventional theories of perception but failed to adequately solve it. While he credits them with beginning the process of undermining “sensation based theories,”“their own explanations of why it is that a fruit says ‘Eat me’ and a woman says ‘Love me’ are strained” (Gibson 1979/2015, 131). In part, the weakness of Gestalt theory lies in its thoroughgoing subjectivism. Rather than challenging the subject-object dichotomy, it emphasises the constitutive role of subjectivity in creating objects. This is a welcome antidote to the naive objectivism of conventional psychology but is ultimately inadequate for, not even the gestalt theorists, could think of [meaning and valence] as physical and, indeed, they do not fall within the province of ordinary physics. They must therefore be phenomenal, given the assumption of dualism. (Gibson 1979/2015, 130) To say that meaning and valence are physical sounds outrageous to many contemporary philosophers; to say that they are physical but “do not fall within the province of ordinary physics” will probably sound unhinged. Yet, this is precisely the radicalism of Gibson’s proposal. When animals perceive an environment, they do not perceive the objects described in classical physics. They perceive a world of challenges and opportunities or, in other words, a world that is replete with meaning. According to classical physics, the universe consists of bodies in space. We are tempted to assume, therefore, that we live in a physical world consisting of bodies in space and that what we perceive consists of objects in space. But this is very dubious. The terrestrial environment is better described in terms of medium, substances, and the surfaces that separate them. (Gibson 1979/2015, 12)

Virtuous Perception


Medium, substances, and surfaces are technical terms of art in ecological optics which Gibson spends a great deal of time explaining; roughly, the medium is that which supports the animal and permits it to carry out its characteristic life activities (air for avian animals, water for marine ones, and land for land-dwellers). Substances are things with which an animal interacts in various ways such as food or the bodies of other animals. The surfaces enable the animal to distinguish between different substances. Notice that the distinctions here are relative to the animal in question. For example, water is a medium for fsh but a substance for a thirsty horse. It might seem that Gibson is offering us yet another subjectivist theory of perception, albeit a more empirically rigorous one, but his starting point is the rejection of dualism. For our purposes, I will focus almost exclusively on his outline of his theory of affordances since it is most relevant to the question of the direct perception of value, and moral value in particular. It is worth noting that there was substantial empirical support for his theory at the time he wrote the book and that this support has continued to grow. Gibson’s theory was, however, controversial at the time of its publication and remains so. Even his supporters felt the need to refine it. In recent times, its insights have largely been incorporated into 4E Cognition. Nevertheless, the initial statement of the theory of affordances is clear and succinct enough for the purposes of explaining it to moral philosophers who are unfamiliar with it. The term “affordance” is Gibson’s own coinage. It is a nominalisation of the verb “to afford” and, as it implies, attempts to describe those features of an environment which afford the animal certain opportunities or, in a negative sense, pose a threat. It is worth bearing in mind that what an opportunity is for one animal may be a threat to another while remaining objectively speaking a feature of the same environment. He proposes that: the compositions and layout of surfaces constitute what they afford. If so, to perceive them is to perceive what they afford. This is a radical hypothesis for it implies that the “values” and “meanings” of things in the environment can be directly perceived. Moreover, it would explain the sense in which values and meaning are external to the perceiver. (Gibson 1979/2015, 117) The idea that we directly perceive values which are in some sense “out there” fies in the face of the subjectivism about value which has dominated much social thought and large swathes of moral philosophy for most of the last century. Simon Blackburn, for instance, contends that subjectivism about value is the price we must pay if we wish to be consistently


Richard Paul Hamilton

naturalistic. For philosophers like Blackburn, it seems obviously true that the real world is disenchanted and thus we project onto it rather than fnd value in it (Blackburn 1993, 168, 174). J.L. Mackie’s infamous “Argument From Queerness,” to which Blackburn is in part responding, rests upon a claim about perception: if there were queer moral properties of the sort moral realists endorse, then we would require an equally queer epistemic faculty to detect them. Gibson reminds us that we did not evolve to detect the objects of classical physics. We evolved precisely to reliably detect those elements of our environment both physical and social which presented threats or opportunities. i.e., affordances. For evolution to occur, affordances needed to remain relatively invariant. Recent developments in ecological niche construction theory have complicated matters here but I will leave this discussion for another paper. To return to the main point: it is only a dogmatic adherence to physicalism, and with it a highly tendentious conception of the unity of the sciences, that forces us to say that the objects an animal responds to in its natural habitat are somehow less real than those described by physics. The worry here, of course, is that of a descent into idealism. But Gibson argues that affordances are neither purely “out there” in the environment nor internal to the animal. The term emphasises “the complementarity of the animal and the environment” (Gibson 1979/2015, 127). As he earlier suggests, although the physical world long predates life, there is no such thing as an environment without living animate beings to inhabit it. The existence of such beings entails a continuous making, unmaking, and remaking of the physical world. Gibson suggests that an affordance is neither an objective property nor a subjective property; or it is both if you like. An affordance cuts across the dichotomy of subjective-objective and helps us to understand its inadequacy. It is equally a fact of the environment and a fact of behaviour. It is both physical and psychical, yet neither. An affordance points both ways, to the environment and to the observer. (Gibson 1979/2015, 121) It is little wonder that such talk baffed the few analytic philosophers that bothered to read Gibson, wedded as most were to the very subjectiveobjective dichotomy he castigates. There is, however, a genuine problem with Gibson’s formulation here and it reveals his own lingering adherence to an object-oriented ontology. His inability to express his point clearly stems from the persistence in his thought of the idea that animal and environment are both kinds of things. If we see the environment not as things which relate to one another externally but rather a complex interactive process, aspects of which may be isolated for clearly defned

Virtuous Perception


analytic purposes (provided that we never lose sight of the impossibility of their actual isolation) matters become much clearer. We can, however, forgive him the occasional solecism. The central point is clear enough: the separation between subject and object is an analytic fiction. In reality, subjects are always embedded in environments which shape them and which they in turn shape. Recognition of this fact has profound implications for axiology. Value is neither out there or in here. If something has value for an animal, it does so as an emergent property of the interaction between the animal and its environment. Most animals come endowed with a capacity for learning and thus, over the course of its development, an animal comes to distinguish features of its environment which offer opportunities, which pose threats and which can be safely ignored. Gibson highlights a significant difference between humans and other animals and that is the degree to which we are capable of changing our environment. He has made more available what benefits him and less pressing what injures him. In making life easier for himself, of course, he has made life easier for most of the other animals. Over the millennia, he has made it easier for himself to get food, easier to keep warm, easier to see at night, easier to get about, and easier to train his offspring. (Gibson 1979/2015, 122) These are not new observations but they are usually made in order to announce human beings’ radical separation from the rest of the natural world. By contrast, Gibson emphasises our continuity. It is a mistake to separate the natural from the artificial, as if these were two environments; artifacts have to be manufactured from natural substances. It is also a mistake to separate the culture environment from the natural environment as if there were a world of mental products distinct from the world of material products. There is only one world, however diverse, and all animals live in it, although we human animals have altered it to suit ourselves. We have done so wastefully, thoughtlessly, and, if we do not mend our ways, fatally. (Gibson 1979/2015, 122) The last line of this passage sounds remarkably prescient but let us focus for now on its central signifcance for moral and political theory. Projectivist accounts of value rest upon a series of dichotomies: between organism and environment; between mind and world; and between persons and organisms. However, to quote an aphorism from an anthropologist deeply infuenced by Gibson, Tim Ingold, the “person is the organism”


Richard Paul Hamilton

(Ingold 2000, 5). As he notes in making this claim, it would be easy to regard this as refecting a crudely reductionist sentiment if we do not also specify what it is we understand by “organisms.” They are not discrete bounded entities but nodes in an ongoing developmental relationship. When we turn our consideration to persons, we recognise that we act but we also interact. We interact with our physical surroundings, including the built environment, and when we do, we often do so using various tools and technologies that have an entire history and culture behind them. Most importantly, however, for our purposes: persons interact with other persons. The actions of others including their linguistic actions become affordances for me and vice versa. Gibson writes “[t]he richest and most elaborate affordances of the environment are provided by other animals, and, for us, other people” (Gibson 1979/2015, 126). Other animals (including people) move but they do not move in the way that physical objects move “which is to say their movements are animate” (Gibson 1979/2015, 127). Although animals are in a sense the subject of the laws of mechanics, “they are not governed by those laws” at least not in the way in which merely physical objects including plants are. Thus “infants learn almost immediately to distinguish [animals] from plants and nonliving things” (Gibson 1979/2015, 127). Turning then to the human realm: we recognise not only that persons are animals and move like animals but that persons qua persons have a distinctive aspect to their animal-like movement, which is to say that at least some person’s movements are made in the light of a reason. To recognise a certain kind of animal as a person is in part to recognise its characteristic bodily form and way of proceeding in the world but it is also to recognise that it is capable of acting in the light of a reason. This seems to present a problem, for while we might acknowledge that the person’s movements are visible, her reasons are not or at least not always. We are sometimes inclined to think of reasons as internal mental causes of our actions which we infer from their doings and sayings. There is, however, a long-established philosophical tradition which suggests that we are mistaken to think about reasons that way. It can be traced back to Aristotle and in recent times is associated with Wittgenstein (1953/2009) and Anscombe (1958). Briefly summarised, this view of reasons states that to see someone as acting in the light of a reason is to see that person and her action in a certain way. Suppose one sees John walking with an umbrella on a bright summer day and wonders what his reason might be for doing so, a perfectly acceptable response would be “because rain is forecast.” Now, it may turn out to be that John is in fact a Russian agent and the umbrella is laced with poison and he is walking about with homicidal intent. But then that would simply mean we have misperceived his actions. In neither variant are we positing some mysterious hidden mental mechanism and even were we to find one, it would merely embellish rather than replace, our original account of John’s reasons. Reasons are

Virtuous Perception


discursive and therefore necessarily public (cf. Hurley 1989). They are the sort of thing we might avow or disavow when prompted. The fact that we can dissemble or be mistaken about the reasons for our action does not thereby render them necessarily private. Learning personhood involves inter alia learning what sorts of things count as reasons for acting or refraining from acting. In order to do so, one needs to distinguish actions from other sorts of bodily movements but we can only learn this by distinguishing them in the environment in which they occur. To identify an action correctly one must also correctly pick out the relevant features of the environment which have occasioned it. To return to John and his umbrella. For the action to be identifiable as that of “carrying an umbrella” as opposed to say “participating in a religious ceremony” or “displaying Protestant bigotry in Glasgow,” one must know the kinds of situations in which an Englishman is liable to carry such an object, otherwise the action would be unintelligible (cf. Anscombe 1979). Taking things in a Gibsonian direction we would say that when I see a person’s reason for acting, I literally see something in the optic array which would count as that person’s reasons. I recognise them as such because I have acquired the skills of social interaction in that social setting. Gibson appears to acknowledge this fact when he writes that: Behavior affords behavior, and the whole subject matter of psychology and of the social sciences can be thought of as an elaboration of this basic fact. Sexual behavior, nurturing behavior, fighting behavior, cooperative behavior, economic behavior, political behavior—all depend on the perceiving of what another person or other persons afford, or sometimes the misperceiving of it. . . . The perceiving of these mutual affordances is enormously complex, but it is nonetheless lawful, [in the ecological and not physical sense] and it is based on the pickup of the information in touch, sound, odor, taste and ambient light. It is just as much based upon stimulus information as is the simpler perception of the support that is offered by the ground under one’s feet. For other animals and other persons can only give off information about themselves insofar as they are tangible, audible, odorous, tastable or visible. (Gibson 1979/2015, 127) There are some problems with Gibson’s choice of words: “behaviour” and “stimulus information” bear the hallmarks of the behaviourism which had until only recently been the dominant view in psychology and which he had already rejected. One can, however, recognise his central point, while discarding the behaviourist baggage. It is to be found here The other person, the generalized other, the alter as opposed to the ego, is an ecological object with a skin, even if clothed. It is an object,

130 Richard Paul Hamilton although it is not merely an object, and we do right to speak of he or she instead of it. But the other has a surface that reflects light, and the information to specify what he or she is, invites, promises, threatens or does can be found in the light. (Gibson 1979/2015, 127) If people’s actions were not identifable in this way, we would have to posit something utterly mysterious along the lines of telepathy to account for our ability to identify them. With respect to people’s reasons though we can misperceive them and they can misrepresent and dissemble if we were not similarly able to identify them in the course of identifying their doings and sayings we would not even have a conception of acting for a reason in the frst place.

5. An Ecological Approach to Social and Moral Perception Let me summarise how the fundamental features of Gibson’s theory of perception are relevant to moral philosophy. Virtue ethicists of a naturalist stripe seek a theory that is both objective16 and naturalistic but some have argued that it is not possible to have both: we must either reject objectivity or naturalism because the world, as revealed by science, is fundamentally disenchanted and any value we find there must have been placed there, if not by us then at least by other valuing subjects.17 Gibson’s ecological approach is an established scientific approach that permits us to retain objectivity and naturalism in virtue ethical theory. On his account, there is, of course, no value in the world described by physics, but the environment we and other animals inhabit except in the most trivial sense is the world described by physics. The social world is a world of meaning and value. The reciprocal relationships between us are based upon these values, and it is impossible to recognise the basis of our relationships for what they are without also picking these out. The standard response to this problem is to treat the social world as a mental phenomenon, but this makes us lose sight of the defining feature of the social world, which is its public, rather than private, nature. It is strikingly odd that Gibson had little to say about social perception and nothing to say about moral perception. If we return to the orthodox computational theory of mind and if we extrapolate from it to the context of the social and moral order as I have just explained with respect to Gibson’s ecological approach to perception, then we realise that we are asked to suppose that each of us somehow carries a representation of the social world and that each individual representation is subtly different from everyone else’s. Yet somehow these representations are sufficiently similar to allow for the possibility of social order. No one has yet produced a coherent story about how this computation is carried out by perception. An explanation is unlikely

Virtuous Perception


since the assumptions upon which the story is based are fundamentally flawed. We do not engage with the world, or with one another, in the way that computers do, for the simple reason that computers do not proactively engage with the world. As Gibson points out, a reason for the plausibility of the computational theory of mind is its explanation of perceptual information as a kind of language which implies a picture of us passively processing that information in the way that a computer does (Gibson 1979/2015, 57). Famously, from the 1960s onwards, Hubert Dreyfus challenged Artificial Intelligence research by suggesting the field was based upon faulty assumptions and thus headed in entirely the wrong direction (e.g., Dreyfus 1965, 1967, 1972, 1992). Partly in response to Dreyfus’ critique, an entire subdiscipline of embodied cognition has emerged, and it is challenging traditional approaches to Artificial Intelligence. Embodied cognition, or for short “embodiment,” is not something tacked onto to humans as an afterthought. Notably, in a recent Routledge Handbook, Shaun Gallagher (2014) acknowledges the affinity between embodiment and the phenomenological tradition, which Dreyfus represents, and Gibson’s ecological psychology. The confluence of traditions has been given a name: “4E Cognition.” The Es in question are: embodied, extended, enactive, and embedded. It is moot whether there should be more or less than four since in many respects what unites the approaches that fall under the banner is opposition to computational and representationalist approaches to cognition and rejection of a dichotomy between organism and environment. The embodied animal is embedded in its environment and its cognitive processes are distributed or extended into the environment rather than siloed inside the animal’s head. There is also no chasm that separates perception and action, since the two are aspects of the same process. Thus perception and cognition are enacted in the environment rather than being inner “mental” processes as traditionally understood. What would a Gibsonian/4E account of social and moral perception be like? When we focus upon relevant features of social situations, we find that they call for certain characteristic courses of action. Suppose that a society has a tradition of “buying rounds.” When my turn to buy “a round” for others in my group comes, I should oblige. A representationalist account would see a gap between my perception and cognition of the situation and my response. My sensory faculties would respond to the photons and sound waves which would then be interpreted by my inner cognitive processes and the final output would be me buying or failing to buy a round. To have grasped the situation successfully, I would need to know that I am in a culture like England or Ireland where round buying is the norm, rather than somewhere like Poland where buying a round would be perceived as insulting the hospitality of the person who invited


Richard Paul Hamilton

you for drinks.18 This distinction about Polish and Irish culture is a fact and to reduce this to some ineluctable inner process would be to misconstrue the necessarily public nature of societal norms. In light of this, a meritorious advantage of the alternative Gibsonian/4E is its parsimony. It is not necessary to propose endless elaborate mental representations of the fine-grained nuances of social life. Someone who has worked out what to do in any situation by directed attentiveness is a socially competent actor. Intuitively, we recognise the distinction between someone who is socially competent and someone who is socially awkward. Alfred Schutz, a sociologist of life, notes that it is not necessary for someone who is socially competent to have comprehensive knowledge of the social world; rather, navigation of the social world requires a person to have a sufficient grasp of social norms. He navigates the social world using a series of “recipes.” The knowledge correlated to the cultural pattern carries its evidence in itself—or, rather, it is taken for granted in the absence of evidence to the contrary. It is a knowledge of trustworthy recipes for interpreting the social world and for handling things and men in order to obtain the best results in every situation with a minimum of effort by avoiding undesirable consequences. The recipe works, on the one hand, as a precept for actions and thus serves as a scheme of expression: whoever wants to obtain a certain result has to proceed as indicated by the recipe provided for this purpose. On the other hand, the recipe serves as a scheme of interpretation: whoever proceeds as indicated by a specific recipe is supposed to intend the correlated result. Thus it is the function of the cultural pattern to eliminate troublesome inquiries by offering ready-made directions for use, to replace truth hard to attain by comfortable truisms, and to substitute the self explanatory for the questionable. (Schutz 1976, 95) Schutz refers to this recipe approach as “thinking as usual.” Crucially for our purposes, it normally suffces. That is, acting in the social world does not normally requires a complex interpretive approach: we are not feld anthropologists in our own culture. It holds so long as the following conditions apply: (1) that life and especially social life will continue to be the same as it has been so far; that is to say, that the same problems requiring the same solutions will recur and that, therefore, our former experiences will suffice for mastering future situations; (2) that we may rely on the knowledge handed down to us by parents, teachers, governments, traditions, habits, etc., even if we do not understand its origin and its

Virtuous Perception


real meaning; (3) that in the ordinary course of affairs it is sufficient to know something about the general type or style of events we may encounter in our life-world in order to manage or control them; and (4) that neither the systems of recipes as schemes of interpretation and expression nor the underlying basic assumptions just mentioned are our private affair, but that they are likewise accepted and applied by our fellow-men. (Schutz 1976, 96) To put things in simpler terms: I know what to do in a given situation because I know what to do in such situations. A socially incompetent actor does not properly grasp what to do in a given situation, because either they lack the relevant recipe or they mischaracterise the situation. To act competently in a social setting requires both the ability to correctly identify what type of situation one finds oneself in and the ability to know what recipe is called for in such situations. The standard model of perception would posit an extraordinarily laborious process of recognition and interpretation, but central to Schutz’s observations is the idea that social life operates precisely because of its largely unreflective nature. Some of the information we use to confront social life comes from personal experience but the bulk of it is handed down to us by others and forms the “unquestioned but always questionable sum total of things taken for granted until further notice” (Schutz 1962, 57). It is only in moments of difficulty or crisis that we engage in conscious deliberation about the way forward.

6. Habituation in Aristotle, Schutz, and the Gibsonian Ecological Approach I now wish to tie together Schutz’s view and Aristotle’s observations on habituation before I attempt to explain how these apply to a moral case and the Gibsonian ecological approach. Aristotle’s observations, like Schutz’s, begin with the recognition that deliberation is a relatively unusual activity. Much of our thought and action necessarily has a habitual character. The properly habituated person sees what to do in a given situation and does it. Aristotle’s view is based upon a qualified rejection of Platonic rationalism. We do not reason ourselves into a particular motivational state, but we can reason ourselves out of one, when, for instance, we recognise that our emotional response is excessive or unjustified. Aristotle’s view differs from Schutz’s view since Schutz seems to accept that we passively absorb our practical reason from our culture. Habituation for Aristotle is not a passive process. It is an interactive one. As the child matures, she is able to take greater charge of her own ethical formation. Most importantly, the truly virtuous are able to subject the mores of their culture to critical scrutiny, rejecting those which do not conform to


Richard Paul Hamilton

the demands of virtue. Nevertheless, it would be a mistake to think that one can dispense with the process of habituation and training altogether in favour of the explicit inculcation of correct moral principles. The capacity for ethical reflection rests upon the prior acquisition of good habits of action and thought. In order to be receptive to theories of virtue, one must first be receptive to the virtues. Receptivity to the virtues is necessary since Aristotle’s account of “cleverness” (deinotes) shows how it is possible to be clever and be a thoroughly despicable human being. In fact, certain species of vice seem to require the skilful misuse of reason. One cannot, for example, thoughtlessly engage in calculated cruelty and it is partly for that reason that we regard conscious malice in a different light from careless brutality. The idea of moral vision, particularly the view implied by Gibson’s ecological approach and Schutz’s view, actually clarifies this point. We distinguish between a trained and an untrained eye. Someone with a trained eye has a specialised skill for perceiving something that someone without a trained eye, without that specialised skill, misses, relative to the same context of activity. So, for instance, a trained eye with respect to the night sky may not necessarily be the most effective daytime birdwatcher, since non-nocturnal birds sleep at night. Nevertheless, the person with moral vision and a nocturnal birder have something in common, in that they are able literally to see things that the untrained observer does not. To have a trained eye, however, is not to have mastered some theory, although theoretical understanding can contribute to the refinement of one’s vision by directing one’s attention. A virtuous person has good basic dispositions, has refined these dispositions through a process of training and habituation, and has at a point begun the process of theoretical reflection upon the habits she has acquired. To return to our earlier point about the relationship between virtue and social competence: the merely competent social actor does in fact resemble the person described by Schutz who follows recipes rather than thinking carefully about what to do. The real test comes when someone faces an unfamiliar or particularly complex situation in which recipes no longer suffice. To pursue the cooking analogy: the difference between the virtuous and the merely competent is the virtuous person is like an excellent cook who uses recipes as an inspiration rather than an instruction and thus when confronted with a lack of a particular ingredient knows what to substitute because he understands the underlying principles of flavour. One of the worries about applying orthodox theories of representation to moral perception is that they seemed to downplay the rational component of ethical life. The view that I am defending here will, I hope, assuage some of those concerns. Moral demands are visible but only properly so to those who are well trained enough to detect them. Part of the process of acquiring a second nature is the acquisition of the rights habits of

Virtuous Perception


mind. Talk of habits often worries those philosophers committed to an intellectualist conception of acting well but it need not. Bill Pollard (2005, 2006, 2008) has persuasively argued in a number of papers that there is no inconsistency in the thought that an action can be both habitual and rational. Pollard’s views are articulated in engagement with the work of John McDowell (1992), who in turn was deeply influenced by Wilfrid Sellars’ (1956/1997) distinction between the “logical space of reasons” and the “logical space of causes.” Our conventional understanding of habits seems to locate them in the logical space of causes but that is largely based upon construing them in stimulusresponse terms.19 This seems to present a problem for McDowell since he has defined second nature as “habits of thought and action.” Yet, in his early work, McDowell was also committed to the view that one cannot have a reason for acting if one does not have a conception of such a reason and habits by definition operate at a pre-cognitive level (cf. McDowell 1978, 1979). They are, by definition, automatic and not the product of prior deliberation. But, as Pollard points out, we can be held responsible for certain of our habits in ways that we would not be for simple stimulus response behaviour. Someone in the habit of parking badly in ways that inconvenience other drivers is rightly held to account for this. Similarly, if, like Kant, I am in the habit of taking a daily walk, I could rightly be applauded for my good sense. Consider the martial arts, which as combative sports largely exist to overcome our bad habits and inculcate more efficiently defensive ones. The fact that the skill of a martial arts master is largely based upon habituation and training does not prevent us from finding him morally praiseworthy in his devotion to his art.

7. Conclusion To summarise, I have suggested that we regard moral competence as a refined form of social competence and that both in turn can be fruitfully regarded along the lines of a kind of vision albeit one not understood in traditional representationalist terms. The social agent does not perceive the objects of classical physics which she then must interpret in moral terms. The social agent inhabits a world of animate beings, most crucially among which are her fellow humans. The virtuous person is able to detect among the doings and sayings of her fellow humans, moral affordances, that is reasons for acting, or refraining from acting thus, but she does so in a way that is largely habitual. What makes her virtuous is that she habitually able to see the demands of virtue in a given situation, and ability she has acquired through a lengthy process of ethical formation. Numerous objections will no doubt have arisen by now. I have not, for example, made a sustained attempt at defending the Gibsonian/4E view of perception. This is because my case rests partly on the enormous amount


Richard Paul Hamilton

of work that has already been done to establish this view as a viable alternative to the Establishment view. I have argued that in the context of moral philosophy, adoption of something like a Gibsonian/4E view has the benefits of parsimony and also enables us to be consistently naturalistic while avoiding the “Placement Problem.” These considerations alone should make the alternative worthy of serious consideration.20

Notes 1. Such is the ferocity of resistance to the idea of moral perception that even those who defend the idea are at great pains to distance themselves from the thought that moral perception is very much like ordinary instances of perception. Robert Noordhof (2018) has an excellent discussion. I suspect that much of that resistance stem from assuming a traditional representationalist theory of perception to be the only option. 2. Susanna Siegel (2006) has argued for a sharp contrast between the ways in which novices and experts perceive. She does this, however, based upon the assumption that the debate turns on what properties can be represented in experience and investigate the influence of expertise in modifying perception. Thus she remains solidly within the orthodox representationalist framework. A more promising account for our purposes is provided by Gibson and Pick (2000). 3. This qualified, obviously, for genuine tragic moral dilemmas. 4. The most recent and most comprehensive survey of the field is Newen et al. (2018). The editor’s introduction gives an excellent overview of the historical development of 4E cognitive science and some of the key controversies. 5. The term was coined by Jackson (1998). The papers in De Caro and Macarthur (2004, 2010) offer an excellent overview of the contemporary debate. 6. Indeed, it is difficult to see how one could be said to correctly perceive “a friend without also perceiving this.” 7. In South Africa’s Eastern Cape the slaughters are carried out in a way that exacerbates the cow’s pain in the hope that the ancestors will hear their cries. 8. As Robert Noordhof (2018) notes such is the resistance to perceptual accounts of morality that even those who defend them take great efforts to distance themselves from the idea that moral perception is similar to ordinary perception. 9. The one obvious exception to this is the concept of legal interpretation where it does seem to make sense to talk about “correct” interpretations of the law but I would suggest that both “correct” and “interpretation” have specialised legal senses here meaning something like defensible in the face of appeal to the highest authority, e.g., the Supreme Court of the US. 10. There is, of course, a distinction between seeing an action in a moral way and seeing the morality of an action. I am making the stronger claim that the virtuous at least are capable of seeing the morality in an action. 11. While some within the “orthodox” tradition have simply rejected any attempt to offer a theory of direct perception, there have been some responses which have tried to incorporate some of the insights of the 4E tradition. For instance, many contributors to Bergqvist and Cowan (2018) suggest that “higher level” properties such as aesthetic and even moral evaluations may be represented in perception. For all their sophistication, however, such accounts remain solidly within a perspective which is representationalist and cognitivist.

Virtuous Perception


12. Members of the 4E community vary on the extent to which they reject representationalism. Andy Clark comes closest to the orthodox tradition in allowing quite a degree whereas the standard bearers for the radicals are Anthony Chemero, Alva Noë, and Daniel Hutto. See Chemero (2011), Hutto and Myin (2017), and Noë (2004, 2009). 13. Robert Audi (2013) has recently defended a perceptual view but he does so solidly within the orthodox representationalist account of perception. 14. Gestalt Psychology which emerged in the 1930s challenged that orthodoxy. 15. I acknowledge the sexist connotations of this example but am leaving them in since it is an important example of social perception. 16. Usually, virtue ethicists prefer a view of objectivity that is consistent with moral realism as opposed to moral constructivism. 17. See for instance Blackburn (1993). 18. Farennikova (2018) has made the intriguing suggestion that evaluative perceptions might be characteristically experiences of absence. In the round buying case what I perceive is the lack of beer in my fellow drinkers’ glass. Interestingly, she mentions Gibson in passing but does not develop the thought. 19. It is partly for that reason that Gibson’s adoption of certain behaviourist terms is unfortunate. 20. I would like to take this opportunity to thank the two anonymous reviewers and Andy Lamey, Dave Robinson, and Christian Mauri for their helpful comments on drafts of this work. The faults of course remain my own.

References Anscombe, G.E.M. 1958. “Modern Moral Philosophy.” Philosophy 33(124): 1–19. Anscombe, G.E.M. 1979. “Under a Description.” Nous 13(2): 219–233. Audi, R. 2013. Moral Perception. Princeton, NJ: Princeton University Press. Bergqvist, A. and R. Cowan, eds. 2018. Evaluative Perception. Oxford: Oxford University Press. Blackburn, S. 1993. Essays in Quasi-Realism. Oxford: Oxford University Press. Chemero, A. 2011. Radical Embodied Cognitive Science. Cambridge, MA: MIT Press. De Caro, M. and D. Macarthur, eds. 2004. Naturalism in Question. Cambridge, MA: Harvard University Press. De Caro, M. and D. Macarthur, eds. 2010. Naturalism and Normativity. Cambridge, UK: Cambridge University Press. Dreyfus, H. 1965. Alchemy and Artificial Intelligence. Cambridge, MA: Rand Corporation. Dreyfus, H. 1967. “Why Computers Must Have Bodies in Order to Be Intelligent.” Review of Metaphysics 21(1): 13–32. Dreyfus, H. 1972. What Computers Can’t Do. New York: Harper and Row Publishers. Dreyfus, H. 1992. What Computers still Can’t Do. Cambridge, MA: MIT Press. Farennikova, A. 2018. “Perceptions of Absence as Value-Driven Perceptions.” In A. Bergqvist and R. Cowan, eds., Evaluative Perception. Oxford: Oxford University Press, pp. 143–161. Gallagher, S. 2014. “Phenomenology and Embodied Cognition.” In L. Shapiro, ed., The Routledge Handbook of Embodied Cognition. London: Routledge, pp. 9–18.


Richard Paul Hamilton

Gibson, E.J. and A.D. Pick. 2000. An Ecological Approach to Perceptual Learning and Development. Oxford: Oxford University Press. Gibson, J.J. 1979/2015. The Ecological Approach to Visual Perception. New York: The Psychology Press. Hurley, S. 1989. Natural Reasons: Personality and Polity. Oxford: Oxford University Press. Hutto, D.D. and E. Myin. 2017. Evolving Enactivism: Basic Minds Meet Content. Cambridge, MA: MIT Press. Ingold, T. 2000. The Perception of the Environment: Essays on Dwelling, Livelihood, and Skill. London: Routledge. Jackson, F. 1998. From Metaphysics to Ethics: A Defense of Conceptual Analysis. Oxford: Oxford University Press. Koffka, K. 1935. Principles of Gestalt Psychology. New York: Harcourt Brace Publishing. Mackie, J.L. 1977. Ethics: Inventing Right and Wrong. London: Penguin Books. McDowell, J. 1978. “Are Moral Requirements Hypothetical Imperatives?” Proceedings of the Aristotelian Society, Supplementary Volumes 52: 13–29. McDowell, J. 1979. “Virtue and Reason.” The Monist 62(3): 331–350. McDowell, J. 1992. Mind and World. Cambridge, MA: Harvard University Press. McGrath, S. 2018. “Moral Perception and Its Rivals.” In A. Bergqvist and R. Cowan, eds., Evaluative Perception. Oxford: Oxford University Press, pp. 161–180. Merleau-Ponty, M. 1945/2012. The Phenomenology of Perception, F. Williams, trans. London: Routledge. Murdoch, I. 1992. “Fact and Value.” In I. Murdoch, ed., Metaphysics as a Guide to Morals. Hammondsworth: Penguin, pp. 25–56. Newen, A., L. De Bruin, and S. Gallagher, eds. 2018. The Oxford Handbook of 4E Cognition. Oxford: Oxford University Press. Noë, A. 2004. Action in Perception. Cambridge, MA: MIT Press. Noë, A. 2009. Out of Our Heads: Why You Are Not Your Brains and Other Lessons from the Biology of Consciousness. New York: Farrar, Straus, and Girroux. Noordhof, R. 2018. “Evaluative Perception as Response-Dependent Perception.” In A. Bergqvist and R. Cowan, eds., Evaluative Perception. Oxford: Oxford University Press, pp. 80–109. Norman, R. 1997. “Making Sense of Moral Realism.” Philosophical Investigations 20(2): 117–135. Pollard, B. 2005. “Naturalizing the Space of Reasons.” International Journal of Philosophical Studies 13(1): 69–82. Pollard, B. 2006. “Explaining Habits with Action.” American Philosophical Quarterly 43(1): 57–69. Pollard, B. 2008. Habits in Action. Berlin: Verlag. Schutz, A. 1962.“Common-Sense and Scientific Interpretation of Human Action.” In M. Natanson, ed., Collected Papers Volume I: The Problem of Social Reality. The Hague: Martinus Nijhoff, pp. 3–47. Schutz, A. 1976. “The Stranger: An Essay in Social Psychology.” In A. Brodersen, ed., Collected Papers Volume II: Studies in Social Theory. The Hague: Martinus Nijhoff, pp. 91–105.

Virtuous Perception


Sellars, W. 1956/1997. Empiricism and the Philosophy of Mind, reprinted with Study Guide by Robert Brandom. Cambridge, MA: Harvard University Press. Siegel, S. 2006. “Which Properties Are Represented in Experience?” In T. SzaboGendler and J. Hawthorne, eds., Perceptual Experience. Oxford: Oxford University Press, pp. 481–503. Wittgenstein, L. 1953/2009. Philosophical Investigations, 4th ed., G.E.M. Anscombe, P.M.S. Hacker, and J. Schulte, trans., P.M.S. Hacker and J. Schulte, eds., Oxford: Wiley-Blackwell.


Virtue Ethics, Blameworthiness, and Role Failure Justin Oakley

Some recent work in virtue ethics seeks to develop less individualistic forms of this approach than earlier versions, by demonstrating how hitting the targets of the relevant virtues sometimes requires support from institutional and regulatory structures, as much as efforts by individual agents strengthening their own dispositions involved in the virtue in question (Banks 2018; Niker 2018; Oakley 2018). These contributions to virtue ethics recognise the importance of the institutional and regulatory environments that professionals and various other role occupants work within for their successfully acting on the dispositions involved in role virtues—such as medical beneficence, which is arguably the central medical virtue for doctors. Highlighting the importance that those environments can have for a practitioner successfully acting on their virtuous professional dispositions also brings out how such environments can sometimes not be conducive to virtuous action by practitioners—and can indeed at times be altogether hostile to such action. Virtue ethics approaches to professional roles commonly evaluate practitioners as acting wrongly when their actions are not in accordance with the regulative ideal of the relevant role virtue, and thereby fail to hit the target of that virtue (Oakley and Cocking 2001). For example, a doctor whose concern for their decisionally capable patient leads them to withhold relevant information likely to alarm that patient would typically be judged as having acted wrongly, given that the doctor seems to have failed to serve the patient’s best interests here—which is the target of the central medical role virtue of medical beneficence. (This doctor’s actions could also be regarded as wrong on the grounds that they are contrary to the virtue of honesty here; in this chapter I shall focus on the wrongness of failing to adequately inform patients qua failures of medical beneficence, but my general arguments about virtue ethics and blameworthiness could also be made by considering such failures to inform as contrary to the virtue of honesty.) In this case the doctor seems to lack sufficient insight into how patients’ best interests are generally well served by their having relevant medical information—and many people would further judge such a doctor as blameworthy for their acting wrongly here. However,

Virtue Ethics and Blameworthiness


sometimes doctors with virtuous dispositions act wrongly not (or not only) from any lack of insight, but in part due to the absence of virtuesupporting (and sometimes the presence of virtue-undermining) institutional and regulatory environments where they work. In such cases, it is not clear whether a doctor who acts wrongly in those circumstances should be regarded as being equally blameworthy as the previous doctor is, where they both fail to hit the target of the relevant role virtue—here, medical beneficence. This raises a more general question about virtue ethics, which I address in this chapter. That is, is virtue ethics capable of accommodating judgements of diminished blameworthiness being made of role occupants who fail to act in accordance with the relevant role virtues, in circumstances where the professional, institutional, and regulatory environments surrounding their action are not conducive to their acting virtuously on those occasions? Before addressing this question, I want to acknowledge that the particular virtue ethics account of wrong action I draw on here is quite demanding, in holding that acting in ways which fall short of the regulative ideal involved in the relevant virtue in question is sufficient for acting wrongly (whether or not the agent’s action is one that a vicious person would do). Thus, on this account, the doctor who withholds alarming but relevant information from a decisionally capable patient thereby acts wrongly. A narrower and less demanding virtue ethics account of wrong action would not regard all actions falling short of right action as thereby wrong actions, but would instead hold that for an agent to act wrongly their action must be one that a vicious person would do in the circumstances.1 This narrower account of wrong action might seem to be an appropriate way of acknowledging that, as Swanton (2003, 241–242) puts it, hitting the relevant targets of a virtue “may reach beyond the strength of particular agents” who might be striving to act well and do not act as a vicious person would, and who it might therefore seem harsh to judge as having acted wrongly.2 Swanton (2003, 240) says that she prefers to “employ three categories: right actions. . . . ‘all right’ actions (which exclude actions which are overall vicious), and wrong actions (actions which are overall vicious).” But as it stands, that narrower virtue ethics account of wrong action seems to be too narrow and undemanding. For the actions of the doctor who withholds alarming but relevant information from a decisionally capable patient seem to be plausibly regarded as wrong (rather than “all right”), even if this doctor does not act from any vicious character trait in doing so—e.g., because it is a one-off case of their withholding this sort of information from a patient in such circumstances. Perhaps such an action might still be thought to count as one that a vicious doctor would themselves perform—e.g., as manifesting a vice of carelessness or dishonesty—even if this particular doctor withholding the relevant information does not themselves have a vicious character trait of carelessness or dishonesty. But in any case, I shall not try to settle


Justin Oakley

this debate here, as it would take us too far afield to develop an account of medical vices that might enable us to evaluate more fully the relative merits of this narrower virtue ethics account of wrong action, compared with various alternatives. Returning to the central question for this chapter, about whether virtue ethics can accommodate judgements of diminished blameworthiness where role occupants fail to act virtuously in environments unconducive to virtuous action, let us consider an historical example of this type of case. This is an example where doctors are plausibly seen as failing to hit the target of medical beneficence, and so as acting wrongly, in a context where an important feature of their professional environment seems to be undermining their acting on that role virtue here. In the first national professional code of ethics for medicine, the American Medical Association in 1847 advised that the physician should be the minister of hope and comfort to the sick; that, by such cordials to the drooping spirit, he may smooth the bed of death, revive expiring life, and counteract the depressing influence of those maladies which often disturb the tranquillity of the most resigned in their last moments. (AMA 1847, ch1, art1, no. 4, 9) In their role as healers, doctors were also expected to be skilled in what were seen as the spiritual aspects of consoling patients, especially as the medications available at the time were often somewhat ineffective. While healing patients remains central to the role of doctors, the role virtue of medical beneficence, which is crucial to serving patients’ health, is no longer standardly understood to include a preparedness to instil hope in the dying, nor a readiness to provide patients with spiritual advice where medications have failed. Traditional medical virtues like medical beneficence and medical courage are still key elements of what it is to be a virtuous doctor, but what these virtues amount to in clinical practice has changed significantly during the intervening years. Some of what was previously seen as medical beneficence is now regarded as misguided and unjustifiably paternalistic, and the medical profession has learned from the evidence that nurturing false hope in dying patients does not generally serve the best interests of patients in such circumstances. Contemporary doctors commonly regard the provision of false hope in dying patients (and in patients generally) as unjustifiably paternalistic, and typically judge as blameworthy practitioners who engage in such behaviour. Virtue ethics approaches to medical practice usually evaluate such conduct as wrong because it is (among other things) contrary to the role virtue of medical beneficence and to the proper goals of medicine that this virtue aims to serve (Oakley 2013, 2018; Pellegrino and Thomasma 1993). Although providing dying patients with false hope

Virtue Ethics and Blameworthiness


might conceivably benefit them in certain ways (by mitigating distressing ruminations about their impending death, and by averting despair), being given false hope is generally regarded as overall harmful to dying patients, as it often deprives them of opportunities to have the final conversations that they wish to have with people they love, and to prepare themselves adequately for their own death (Harrington and Smith 2008; Wrigley 2019). Nonetheless, a mid-nineteenth century American doctor instilling false hope in their dying patient might seem to be less blameworthy than a contemporary doctor who acts in such ways, given that this sort of conduct was being prescribed to American doctors by their own professional association at that time. Does virtue ethics have the resources to accommodate such judgements of diminished blameworthiness of agents, who act in circumstances which present countervailing influences to their acting virtuously and rightly? Broadly speaking there are two grounds on which an agent’s blameworthiness for their wrong action can be reduced or removed altogether: (i) the wrongness of their action (according to the relevant normative standard) can be reduced or overturned (e.g., the wrongness of breaking a promise to a friend to keep certain information confidential might be reduced if there was a very good reason for doing so—such as the need to avert a violent confrontation between two others) and (ii) the agent’s level of moral responsibility for their action is diminished or removed— for example, when the agent is acting under some duress (Milo 1984, 3–4, 219–225; Nozick 1981, 363). It is on this second ground for diminished blameworthiness that I will focus my discussion in this chapter. One way of addressing this central question could investigate the compatibility (or otherwise) of virtue ethics with a standard (though not uncontroversial) account of moral responsibility, such as an account that holds an agent A morally responsible for X-ing if and only if X-ing was reasonably avoidable for A—and holds A morally responsible for consequences C of X-ing if and only if C were reasonably foreseeable consequences of X-ing (D’Arcy 1963, 124–125; Glover 1970, 10–11; Van Inwagen 1999). However, in this chapter I wish to consider whether a plausible virtue ethical account of agent blameworthiness (and related concepts such as moral negligence) can be developed, drawing partly on a tentative sketch of what a virtue ethical account of moral responsibility could look like. I will argue that such an account could be developed by explicating the key concepts of avoidability and foreseeability in terms of what a virtuous agent would be able to avoid and foresee, where such explications can differ from what “reasonably avoidable” and “reasonably foreseeable” are usually understood to involve. Virtue ethicists are now beginning to address how virtue ethics might be able to provide the basis of a plausible account of the justifiability of judgements of an agent’s blameworthiness. For example, Van Zyl (2019) has recently offered a virtue ethics account of what qualities an


Justin Oakley

evaluator might plausibly manifest in their act of judging the agent to have acted wrongly (qua selfishly, uncompassionately, and so on), and of what blaming or other responses might be justifiably directed at agents in such circumstances.3 This seems to be a promising way to develop a virtue ethics approach to judging others’ conduct as being blameworthy, praiseworthy, or as warranting related evaluations. The approach I am investigating here differs from this, as I am considering whether appeals to a virtuous exemplar could plausibly be incorporated into the relevant normative standards underpinning moral responsibility and blameworthiness themselves. Virtue ethics commonly evaluates actions by asking, “what sort of person would do a thing like that?” Was the action, for example, generous or mean-spirited, courageous or cowardly? Is this the sort of thing that a kind person or a just person would do in the circumstances? Virtue ethicists often draw upon such considerations to provide the basis of a criterion of right action. Thus, a virtue ethics criterion of right action can be stated initially in broad terms as holding that an action is right if and only if it is what a virtuous agent, acting in character, would do in the circumstances (Hursthouse 1999, 28–31).4 This approach to right action has also been extended to various professional roles, such as those of doctors and lawyers. One version of such an approach in professional contexts highlights links between the proper goal of the profession in question— such as health and justice for medicine and law, respectively—and an Aristotelian conception of human flourishing, or eudaimonia. These proper goals can, in turn, provide the basis for an account of the role virtues for that profession. Thus in the case of medical practice, the role virtues for doctors can be understood as those character traits that enable doctors to serve the goal of health. These traits plausibly include dispositions such as medical beneficence, medical courage, and trustworthiness (Oakley and Cocking 2001, 93; see also Pellegrino and Thomasma 1993). So, on this account, a doctor’s action is right if and only if it is what a virtuous doctor, acting in character, would do in the circumstances. In considering the blameworthiness of various agents who do not meet such a criterion of right action and thus (I am assuming here) act wrongly, such a criterion could be taken as implying that all agents who, in a variety of circumstances, fail to act in a manner that this criterion prescribes are equally blameworthy for their conduct. For, considering the two elements of blameworthiness described previously, all such agents act wrongly in failing to do what this criterion prescribes, and they would all have equal levels of moral responsibility for their conduct—unless there are mitigating circumstances regarding some of those agents. Of course, there can indeed be various kinds of mitigating circumstances affecting an agent’s level of moral responsibility—and so their level of blameworthiness—for their action. As philosophers since Aristotle have long recognised, voluntariness, moral agency, and moral responsibility come in degrees. Taking

Virtue Ethics and Blameworthiness


voluntariness as a case in point, some actions are less voluntary than others, without being altogether non-voluntary or involuntary. For example, Aristotle (Nicomachean Ethics III, 1) argued that a ship’s captain who chose to throw his goods overboard in order to stop his ship sinking in a storm would be acting in a manner which is less than fully voluntary, though his doing so is not altogether involuntary. The captain would be acting under some (though not necessarily an overwhelming) level of compulsion or a form of duress here—Aristotle refers to these as “mixed” actions, since they seem to involve a mix of voluntary and non-voluntary aspects. There seems to be no reason why a virtue ethics criterion of right action would be incapable of recognising that strong (if not necessarily irresistible) circumstantial factors and pressures can mitigate an agent’s degree of moral responsibility, and thus blameworthiness, for acting wrongly. Likewise, virtue ethics could presumably also acknowledge that some circumstantial pressures are difficult for any agent to resist since it “overstrains human nature” (see Aristotle NE 1110a25) to withstand or alter them, and so the agent’s degree of moral responsibility and blameworthiness for their actions performed under such pressures could here plausibly be reduced to zero—or, as Aristotle puts it, such actions are to be “pardoned.”5 Let me therefore examine in more detail how a virtue ethical account of the grounds of an agent’s blameworthiness could be plausibly developed, in a way that is capable of recognising that various agents who fail to meet what a virtue ethics criterion of right action prescribes might have different degrees of moral responsibility—and thus blameworthiness— for their conduct there. This account can be developed further by considering the following sorts of cases. Compare two American doctors in 1847, both of whom act to instil what they realise is false hope in their dying patients, and both of whom are aware that their professional association prescribes this conduct: (a) Doctor A instils false hope in dying patients and wholeheartedly endorses their doing so, as A assumes that this serves the patient’s best interests here. (b) Doctor B doubts that instilling false hope in dying patients is in those patients’ best interests, but B proceeds (somewhat reluctantly) to do so here due to concerns about incurring significant professional opprobrium if B did otherwise. The failure of both of these doctors to meet the relevant virtue ethics criterion of right action in this context entails that both doctors have acted wrongly here. However, Doctor B might plausibly be regarded as less blameworthy in doing so than might Doctor A, because Doctor B’s moral responsibility for their action here could be seen as (at least marginally) somewhat diminished due to their responding to what they perceive as a signifcant level of professional duress here.6


Justin Oakley

Perhaps the main line of argument I am developing here might seem to be somewhat unobjectionable thus far. But how might the degrees of moral responsibility and blameworthiness of these doctors (and of agents generally) for their actions (and their consequences) be explicated in virtue ethical terms? I want to suggest that a virtue ethical account of moral responsibility and blameworthiness could perhaps be developed by explicating the relevant normative standards involved in terms of “What a virtuous person would do here,” and “What a virtuous person would foresee here,” in the following ways: Agent A is morally responsible for action X if and only if: (i) A performed action X. (ii) A virtuous person would have been able to avoid doing X. Agent A is morally responsible for consequences C (of action X) if and only if: (i) A performed action X. (ii) A virtuous person would have been able to avoid doing X. (iii) A virtuous person foresaw, or would have foreseen, C as a consequence of X. This approach ties moral responsibility (and thus blameworthiness) partly to intellectual virtue, about what a virtuous person would foresee and would choose to endure. The relevant Aristotelian intellectual virtue involved here seems to be practical wisdom (phronesis), and particularly those aspects of practical wisdom concerning judgements about what pressures are worth enduring for which goals (see NE III, 1), and those aspects concerning meticulously investigating potential risks and benefts of one’s actions. As Aristotle puts it in his discussion of practical wisdom (NE VI, 7), it is to that which considers well the various matters concerning itself that one ascribes practical wisdom, and it is to this that one will entrust such matters. This is why we say that some even of the lower animals have practical wisdom [or, as translator W.D. Ross notes, intelligence], viz. those which are found to have the power of foresight with regard to their own life. In the context of medical roles, this account would explicate the relevant normative standards for determining moral responsibility and blameworthiness in terms of what a virtuous doctor would have been able to avoid, and what a virtuous doctor would foresee. So, on this account, both Doctor A and Doctor B would be morally responsible and blameworthy for seeking to instil false hope in their dying patients (in

Virtue Ethics and Blameworthiness


mid-nineteenth century America) if and only if a virtuous doctor in that context would have realised that this is not in patients’ best interests, and would have been able to resist any concerns about perceived professional opprobrium were they to avoid seeking to instil false hope in their patients. Some might question whether appealing to what a virtuous person would be able to avoid and foresee in the circumstances improves upon the more familiar standard that appeals to what a reasonable person would avoid and foresee in the circumstances. One advantage of appealing to the standard of a virtuous person here is that this could provide a more enduring normative standard for moral blameworthiness and moral negligence, than does appealing only to what a reasonable doctor would avoid and foresee. Perhaps some might argue that Doctor A is not blameworthy (nor morally negligent) for instilling false hope in their dying patient, because a reasonable doctor in that historical and cultural context would not have foreseen that this was against patients’ best interests and would therefore not have refrained from doing so. However, relativising the normative standard to agents’ historical and cultural contexts in such ways seems to be problematic, as it deprives us of any more enduring normative standard by which to judge agents’ blameworthiness here. (We do, after all, regard the AMA’s abandonment of their advice to instil false hope in dying patients as a form of moral progress in the medical profession, and in doing so we seem to be invoking a more enduring normative standard here.) A virtuous doctor would continually reflect on what constitutes best evidence-based practice (Gruner 2000), which can enable them to grasp, even in nineteenth-century America, how providing dying patients with false hope is generally overall harmful to them in the ways mentioned earlier—by denying patients opportunities to adequately prepare themselves in various ways for their own deaths. Explicating the normative standard for blameworthiness and moral negligence in terms of what a virtuous doctor would avoid and foresee provides us with a more robust normative standard, which (without denying the inherent uncertainties in medicine) has the resources to help practitioners gain a better grasp of their overarching goal of serving patients’ best interests, and of how any given code of medical ethics may need to change so as to better serve that goal. Another attractive feature of this virtue ethics approach is that it connects moral responsibility, blameworthiness, and negligence neatly to discussions about the attainability of virtues (see e.g., Miller 2018). For example, suppose the level of practical wisdom for medical foresight is set at a very demanding standard. This might inculpate doctors for negligence in cases where that seems unjustifiable. For suppose only a brilliant doctor with the level of clinical acumen and insight displayed by the TV character Dr House would have foreseen that a particular medical procedure would likely harm a certain patient. Then, according to the


Justin Oakley

normative standard outlined previously, other doctors would be blameworthy and negligent for failing to foresee those harms if and only if Dr House would have foreseen them (using his skilful insight, rather than a lucky guess). But if most doctors are not able to inculcate Dr House’s level of clinical acumen, it might seem unjustifiable to hold them blameworthy and negligent for failing to foresee what only someone like Dr House was capable of foreseeing. The level of foresight involved in the normative standard might therefore be justifiably set a little lower than this—while still remaining appropriately demanding of doctors. That would seem to set at a suitable level the standard of moral responsibility and blameworthiness for the mid-nineteenth century American Doctor B instilling false hope, as this doctor would not be fully blameworthy for such actions (given the risks of strong professional opprobrium for failing to do this), and yet B would still remain somewhat blameworthy for doing this, given that B doubted that such action was in their patients’ best interests but went ahead with this action anyway. Proposing a standard of role virtue beyond mere reasonableness also better reflects how doctors (including Doctor A) can plausibly be expected to reflect on how well certain medical practices that are pervasive (and professionally recommended) might nonetheless not serve the proper goals of medicine very well (without, of course, expecting doctors to be omniscient). Of course, some degree of moral responsibility and blameworthiness for false hope being instilled in dying patients in mid-nineteenth century America should be allocated to the American Medical Association itself, for issuing and promulgating poor medical advice to doctors. This organisation bears some collective moral responsibility, and blameworthiness, for urging doctors to instil false hope in their patients. And, as I have argued, the doctors’ own failures of practical wisdom in following this advice was blameworthy, because even though their doing so was understandable (especially given the likely peer and professional pressure to follow this advice), they could have been expected to develop sufficient clinical insight to understand that such actions do not serve patients’ best interests.7,8

Notes 1. It is possible to read Hursthouse’s (1999, 79–81) brief discussion of wrong actions as endorsing such a claim, as she says (81) that “Terrible acts will in general . . . be . . . the very sort of thing that the most callous, dishonest, unjust . . . i.e., wicked characteristically do.” However, Hursthouse does not here go as far as saying that wickedness is necessary for an act to be wrong. Thanks to Liezl van Zyl for helpful comments here. 2. It should be noted that while the standard of right action that I am employing is demanding, in evaluating agents’ actions by reference to actions by someone with the relevant virtue to the level of excellence, it is less demanding than a standard that evaluates agents’ actions by reference to actions by someone who is perfect in regard to the relevant virtue.

Virtue Ethics and Blameworthiness


3. Van Zyl also considers such questions about making judgements of another’s praiseworthiness for having acted rightly (qua generously, bravely, and so on). J.J.C Smart (1961) considers some analogous questions about utilitarianism, which he argues can be applied directly to the utility of the act of judging an agent to be morally responsible, and (thus) praiseworthy or blameworthy (and of praising or blaming the agent here), for a particular action (that maximises or fails to maximise utility). Smart (1961, 302) argues that to ascribe responsibility to A for X is to attempt to influence them, and further, that “to praise a class of actions is to encourage people to do actions of that class. And utility of an action normally, but not always, corresponds to utility of praise of it” (304). Smart also argues (305), that we can justifiably blame A for X when it “is socially useful in spurring others” to behave better than A did here. 4. Articulating a virtue ethics criterion of right action in this way is not uncontroversial (see for example, the discussions by Swanton 2003, and Van Zyl 2018). Nevertheless, I will use this account in developing my arguments about blameworthiness as this formulation is often used in applying virtue ethics to evaluating professional practice. 5. Irwin (1980, 153) argues that Aristotle regards even these actions as a mix of voluntary and nonvoluntary aspects, in which case agents performing such actions could still be held blameworthy to some degree for performing them. 6. Note that a further doctor (who we could call Doctor C) in such circumstances, who also doubted that instilling false hope was in the best interests of dying patients but proceeded to do so anyway because they found this easier than revealing the truth to the patient, could plausibly be seen as more blameworthy than Doctor B (and indeed, possibly more blameworthy than Doctor A) here, given that this Doctor C was not acting under any professional duress here. 7. For a useful discussion of related issues about vices leading to cultural ignorance, and about when it is reasonable to expect people to overcome such ignorance, see Mason and Wilson (2017). 8. I wish to thank Liezl van Zyl for her detailed and helpful comments on previous drafts of this chapter. I presented an earlier version of this chapter at Australian Catholic University School of Philosophy (Melbourne), where I received valuable feedback from Stephanie Collins, Steve Clarke, Tony Coady, Simon Goldstein, Ole Koksvik, and Tyler Paytas. Research for this chapter was supported in part by Australian Research Council Discovery Project grant DP190101597, Religion, pluralism, and healthcare practice: A philosophical assessment.

References American Medical Association. 1847. Code of Ethics of the American Medical Association. Philadelphia: T.K. and P.G. Collins, printers. Aristotle. 1980. The Nicomachean Ethics, W.D. Ross, trans. Oxford: Oxford University Press. Banks, S. 2018. “Cultivating Researcher Integrity: Virtue-Based Approaches to Research Ethics.” In N. Emmerich, ed., Virtue Ethics in the Conduct and Governance of Social Science Research, vol. 3. Bingley: Emerald Publishing, pp. 21–34. D’Arcy, E. 1963. Human Acts: An Essay in Their Moral Evaluation. Oxford: Clarendon Press. Glover, J. 1970. Responsibility. London: Routledge. Gruner, J. 2000. “Complementary Medicine, Evidence Based Medicine, and Informed Consent.” Monash Bioethics Review 19(3): 13–27.


Justin Oakley

Harrington, S.E. and T.J. Smith. 2008. “The Role of Chemotherapy at the End of Life: ‘When Is Enough, Enough’?” Journal of the American Medical Association 299(22): 2667–2678. Hursthouse, R. 1999. On Virtue Ethics. Oxford: Oxford University Press. Irwin, T.H. 1980. “Reason and Responsibility in Aristotle.” In A.O. Rorty, ed., Essays on Aristotle’s Ethics. Berkeley, CA: University of California Press, pp. 117–156. Mason, E. and A.T. Wilson. 2017. “Vice, Blameworthiness, and Cultural Ignorance.” In P. Robichaud and J.W. Wieland, eds., Responsibility: The Epistemic Condition. Oxford: Oxford University Press, pp. 82–100. Miller, C.B. 2018. The Character Gap: How Good Are We? Oxford: Oxford University Press. Milo, R.D. 1984. Immorality. Princeton, NJ: Princeton University Press. Niker, F. 2018. “Policy-Led Virtue Cultivation: Can We Nudge Citizens towards Developing Virtues?” In T. Harrison and D.I. Walker, eds., The Theory and Practice of Virtue Education. London: Routledge, pp. 153–168. Nozick, R. 1981. Philosophical Explanations. Cambridge, MA: Harvard University Press. Oakley, J. 2013. “Virtue Ethics and Bioethics.” In D. Russell, ed., The Cambridge Companion to Virtue Ethics. Cambridge, UK: Cambridge University Press, pp. 197–220. Oakley, J. 2018. “Creating Regulatory Environments for Practical Wisdom and Role Virtues in Medical Practice.” In D. Carr, ed., Cultivating Moral Character and Virtue in Professional Practice. London: Routledge, pp. 83–95. Oakley, J. and D. Cocking. 2001. Virtue Ethics and Professional Roles. Cambridge, UK: Cambridge University Press. Pellegrino, E. and D. Thomasma. 1993. The Virtues in Medical Practice. Oxford: Oxford University Press. Smart, J.J.C. 1961. “Free Will, Praise and Blame.” Mind 70(279): 291–306. Swanton, C. 2003. Virtue Ethics: A Pluralistic View. Oxford: Oxford University Press. Van Inwagen, P. 1999. “Moral Responsibility, Determinism, and the Ability to Do Otherwise.” Journal of Ethics 3(4): 343–351. Van Zyl, L. 2018. Virtue Ethics: A Contemporary Introduction. London: Routledge. Van Zyl, L. 2019. “Does Virtue Ethics Allow Us to Make Better Judgments of the Actions of Others?” In E. Grimi, ed., Virtue Ethics: Retrospect and Prospect. Dordecht: Springer, pp. 99–110. Wrigley, A. 2019. “Hope, Dying and Solidarity.” Ethical Theory and Moral Practice 22(1): 187–204.


Well-Being, Narrative Value, and Virtue Ethics Nicholas Ryan Smith

Some lives, I assume, are better—more fulfilling, meaningful, enjoyable, desirable, and worthy of pride and gratitude—than others. In determining whether a life is good and how good or bad it is, it is natural to focus on what is done and what happens in that life. If, for example, someone creates something valuable, such as a well-received collection of songs, that (typically) makes their life better, for the creation is (typically) a fitting object of pride and satisfaction.1 If, by contrast, something unfortunate happens, such as losing a house due to a storm, that makes a person worse off. In recounting what took place in a life, it may not appear essential when they took place, provided that what all is done and what all happens does not depend on timing. Thus, Gabriel Richardson Lear, in discussing the requirement of a complete life in Aristotle’s conception of happiness, makes the related claim that “time does not have any positive value simply in itself. Time is valuable because of what happens during it” (2015, 131). After all, whether someone loses their house at one time or another does not appear to make a difference to how much worse off it makes them, assuming that nothing about the circumstances adds an additional misfortune to the original misfortune of losing the house—that is, as long as nothing significant changes in what happens. Such observations indicate that good-making and badmaking features of a life are temporally neutral: their contribution to the goodness or badness of a life does not essentially depend on when they take place. The literature on narrative value and the shape of a life calls into question the temporal neutrality of good-making and bad-making features of lives. The “shape” of a life refers to the temporal or narrative sequence, order, or arrangement of acts, events, moments, and episodes that comprise that life. A person who goes from “rags to riches” lives a differently shaped life than a person who goes from “riches to rags” even if they possess the same riches and the same rags for equal periods of time. The debate about the shape of a life centres on whether the shape of a life matters to how good or bad that life is and, if so, why.


Nicholas Ryan Smith

In this chapter, my ultimate aim is to argue in favour of a virtue-ethical account of well-being or of what makes someone’s life good for them.2 In the first two sections, I engage with the literature on the shape of a life and, following David Velleman and Antti Kauppinen, argue that changes in the (dis)value of events and acts that are targeted by this literature are due to changes in narrative meaning rather than mere temporal location or order. In the third section, I outline three constraints for a fully satisfactory account of well-being: adequacy in respect to narrative value, emotional fittingness, and practical reasons. In the remaining sections, I argue that virtue ethics is superior to hedonism, desire-satisfaction theory, and standard objective list theory in respect to these constraints. In the final section, as part of this argument, I focus on how virtue ethics is better equipped to capture narrative value and disvalue than Kauppinen’s narrative calculus.

1. Well-Being and Time In a formative paper, “Goods and Lives,” Michael Slote argues in favour of a “pure time preference” for goods, including achievements and beneficial events, that occur later in a person’s life: “When a benefit or good occurs may make a difference to how fortunate someone is (has been) . . . what happens late in life is naturally and automatically invested with greater significance and weight in determining the goodness of lives” (1982, 317). It is only by this pure time preference, according to Slote, that we can explain why it seems that, all else equal, people who achieve success after years of mediocrity or misfortune live better lives than those who achieve success early on while declining thereafter. Slote (1982, 317– 318) illustrates as follows: A given man may achieve political power and, once in power, do things of great value, after having been in the political wilderness throughout his earlier career. He may later die while still “in harness” and fully possessed of his powers, at a decent old age. By contrast, another man may have a meteoric success in youth, attaining the same office as the first man and also achieving much good; but then lose power, while still young, never to regain it. I share Slote’s evaluation that the life of the frst man, whom I, following Kauppinen (2015), call Uphill, seems more desirable than that of the second man, Downhill, even though, as we are meant to imagine, both lives feature similar achievements, struggles, failures, and misfortunes, which are arranged differently in time. A complementary thought experiment, which is designed to elicit a similar response, is presented by Dale Dorsey, who contrasts the life of O.J. Simpson with that of J.O. Nospmis, a fctional person who lives a similar life to Simpson but in reverse. Simpson was a highly

Well-Being, Narrative Value, and Virtue


successful and beloved athlete, actor, and sports commentator that fell into disrepute after being charged with the murder of Nicole Brown Simpson, his ex-wife, and Ronald Goldman, her friend. Though acquitted, he was widely thought to be guilty and, after being found liable for both deaths in a civil trial, was ordered to pay 25 million dollars to the victims’ families. Additionally, Simpson was later sentenced to 33 years in prison for burglary. By contrast, Nospmis was charged with—but not convicted of—murder early in life and later, at the age of 25, was sentenced to prison for armed robbery. After being released from prison, Nospmis turned her life around and eventually became a highly successful college basketball coach and, later, a popular sports commentator.3 Dorsey plausibly suggests that Nospmis’ life is more desirable than Simpson’s or at least has something desirable about it that Simpson’s life lacks, for “Nospmis’ life featured a dramatic upswing; Simpson’s a horrific downfall” (2015, 305). Since Simpson’s life is less desirable than Nospmis’ and an obvious difference between the two lives is when the good times and the bad times occur or the order in which they occur, the comparison provides some—though not conclusive—support for the view that the temporal location or temporal ordering of good-making and bad-making acts and events in a life matter to how valuable that life is. What explains our evaluation of these cases? On Slote’s view, the greater desirability of Uphill’s life and that of Nospmis should be explained by the fact that the good times occur later in their lives since, as the pure time preference maintains, good-making and bad-making features of a life that occur later are more significant to the value of that life than those that occur earlier. There are significant problems with this explanation. Although it gives the right result in the case of the politicians and in the case of Simpson and Nospmis, it gives the wrong result in many other cases. For instance, if we contrast two people whose lives contain the same episodes except that the first person experiences a very painful medical condition that debilitates him for a month in his twenties while the same happens to the second person in her fifties, there appears to be no reason to regard the second person’s life as any worse than the first although the unfortunate episode occurs later in life and thereby should be more significant to the value of the lives, according to Slote’s pure time preference. The same point can be made about valuable achievements. If someone learns to play guitar as a hobby and takes joy in it for a number of years, that makes their life better, but whether this period occurs in their twenties or forties does not make any apparent difference in the value of their life provided all else is equal. Such cases indicate that Slote’s pure time preference does not capture the essential difference between the lives of Uphill and Downhill. Furthermore, if later episodes in a life are more significant in virtue of being later, then there appears to be no reason why later events in a


Nicholas Ryan Smith

year, week, or even a weekend should not be more significant to the value of that year, week, or weekend. Yet, as Dorsey points out (2015, 316), attending a fun party on a Saturday and nursing a hangover on Sunday does not, in a typical case, seem to be any worse than a similar weekend in which one nurses a hangover on Saturday and attends a fun party on Sunday. Given that the timing of events and acts does not always make a difference in either shorter periods of time, such as weekends, or in longer periods of time, such as lifetimes, an alternative explanation is desirable for those cases in which the timing or something related to the timing apparently does make a difference. An alternative is provided by what Dorsey calls the Shape of a Life Hypothesis (SLH), which is the claim that the “[t]he temporal sequence of good and bad times in a life can be a valuable feature of that life as a whole” (2015, 305). The SLH is meant to capture the claim that an upward-sloping life—one in which good times occur after bad times— is, all else equal, more desirable than a downward-sloping life—one in which bad times occur after good times. This differs from Slote’s pure time preference because the SLH suggests that good times contribute additional value to a life only if they follow bad times whereas the pure time preference suggests that good times that occur later in a life contribute additional value whether they are preceded by bad times or not. Although, like Slote’s pure time preference, the SLH has intuitive implications in the case of the politicians and in the case of Simpson and Nospmis, it too is subject to counterexamples, for there are cases in which an upward or downward slope makes no apparent difference to the value of a life.4 Consider again the misfortune of suffering a painful medical condition for a month. If one suffers this misfortune after having an especially successful year at work, that does not by itself seem to make the misfortune any worse, unless it adds an additional misfortune, such as preventing one from enjoying some of the fruits of the successful work from the prior year. If the timing results in an additional misfortune, then, of course, the additional misfortune rather than facts about temporal ordering may explain why the subject is worse off. The problem with the SLH, I believe, is that it is framed in purely temporal terms. This is a problem because a good episode and a bad episode may follow or precede one another even if they have no meaningful relationship. If a good episode occurs after a bad episode, they may be meaningfully related, such as in the case of enjoying the public’s response to one’s enormously successful novel only after spending a number of years being roundly rejected by publishers. The good and bad times are meaningfully related in this case because one’s earlier struggles pay off in later success. However, temporally neighbouring episodes are not always meaningfully related, such as in the case of enjoying the success of one’s novel only after dealing with the months-long nightmare of eradicating bed bugs from one’s home. In a case of this kind, the earlier misfortune

Well-Being, Narrative Value, and Virtue


does not pay off in later success, nor do the events need to be meaningfully related in another way.

2. Meaning and Narrative Value David Velleman, who first marked the distinction between temporal and meaningful relations in the shape of a life debate, offers, in my view, a superior explanation to both the pure time preference and the SLH for the apparent significance of a life’s shape: The reason why later benefits are thought to have a greater impact on the value of one’s life is not that greater weight is attached to what comes later. Rather it is that later events are thought to alter the meaning of earlier events, thereby altering their contribution to the value of one’s life. (2000, 62)5 Velleman’s claim is that the timing of an event in a life matters for wellbeing only if the meaning of the event changes as a result of its place in the story of that life, when the “story” of a life is comprised by narrative relations between moments in a life rather than a story that is or might be told about a life. In Slote’s case, we imagine that the two politicians have different life stories. Uphill’s life is one marked by early struggles, which are later fulflled and thereby redeemed. In this life, the political victory is “a well-earned reward, and would prove his struggles to have been a good investment” (Velleman 2000, 64). By contrast, Downhill’s life is one of later struggles to return to the heights of early success, which are suffered in vain. In this life, the early political victory is “a windfall in relation to which his [later] struggles were superfuous” (Velleman 2000, 64). By being embedded in different life stories or by being meaningfully related to other events in the life, the “same” event or act can change character and its contribution to the value of a life can depend on this change of character. If so, the mystery of the greater desirability of Uphill’s life dissipates because what happens and what is done is, in fact, not the same in each life, for the events and acts have different meanings, and the meaning of an event or act partly determines what happens and what is done. Although Velleman does not provide necessary and sufficient conditions for when two acts or events are meaningfully related in the relevant way, he provides a number of illustrations. For instance, he distinguishes learning from one’s mistakes from learning an entirely new lesson which is not meaningfully related to one’s past mistakes. Although learning a lesson from one’s mistakes and learning that lesson after making an unrelated mistake may promote one’s good in similar ways, Velleman plausibly claims that learning from one’s mistakes contributes more value to


Nicholas Ryan Smith

one’s life than learning a new, unrelated lesson and, relatedly, that making the same mistake twice is worse than making two different mistakes (2000, 64–65). In each of the examples we have considered, it is important to note that Velleman’s meaning-based analysis (but neither the pure time preference nor the SLH) is compatible with Lear’s claim that time does not itself contribute value to a life but rather only what happens in time does. The compatibility is plain if one considers that what happens or what is done in a life may depend on what happens before or later. In one case, someone struggled in vain whereas, in another, someone did not struggle in vain although they struggled. One person learned from their mistake while another did not even if they learned something. Finally, Simpson might be said to have wasted his success and social status while Nospmis did not. A significant advantage of Velleman’s meaning-based analysis is that it avoids the previous counterexamples to Slote’s pure time preference and the SLH. If, for example, suffering a painful medical condition in one’s twenties has the same meaning as it does in one’s fifties—that is, if what happens in the lives is the same in extra-temporal terms—then its timing has no bearing on the value of the life. It might be objected that talk of meaningful relations avoids such counterexamples only at the cost of invoking an excessively mysterious concept of meaningfulness. This is not persuasive. The dependence of the character of an act or event on its relations to the past or future is, in reality, quite mundane. For instance, whether one repeats something depends on what happened before. Whether one assists in a basketball game depends on, though not only on, whether a score is made after one passes the ball. It is with such cases in mind that we should approach Alasdair MacIntyre’s claim that “[a]n action is a moment in a possible or actual history or in a number of such histories” (2007, 214). What all an agent is doing, on MacIntyre’s view, essentially depends on facts about what came before and what comes later; to isolate a piece of behaviour from the past or the future would be to obscure the full nature of what is done. MacIntyre’s point is that knowing the full nature or meaning of an act presupposes knowing its place in a narrative or history. Whether I am repeating something obviously depends on what happened before and, further, what all I am doing in or by repeating is rendered intelligible only by placing it in the context of other acts and events. If the stranger next to me on the train says to me, out of nowhere, “Four score and seven years ago,” I will not fully understand what he is doing although I will know that he is repeating what Abraham Lincoln once said. It should thus not perplex us that earlier or later events may partly constitute the meaning of an act or event provided that we understand the “meaning” of that act or event to be referring to the most apt description of it for the purposes

Well-Being, Narrative Value, and Virtue


of evaluation. Nor should it perplex us that the meaning of an act can bear on its value, for whether repeating Lincoln is valuable depends on its meaning—on what all the agent is doing by repeating Lincoln. For example, is the stranger mistakenly attempting to initiate conversation with an old friend by referencing an inside joke or is he rather using a coded phrase to identify himself as a member of a secret group to initiate some illicit activity?6 How and whether an act bears on the goodness of an agent’s life plausibly depends on the narrative facts that partly constitute what the agent is doing. For both MacIntyre and Velleman, the meaning of an act is partly determined by its narrative relations. How does narrativity differ from mere chronology or pure temporal relatedness? According to Antti Kauppinen, who surveys a number of definitions, theories of narrative “tend to point in the direction of goal-directed human agency extended over time” (2015, 203). That is, narratives concern an agent’s ends (their formation, specification, and modification), activity done in pursuit of such ends, acts and events related to the realisation of the agent’s ends, and the outcomes of activities as they relate to the agent’s ends. If so, an act or event has narrative meaning to the extent that it is directly related to the agent’s ends (and perhaps to what should be their ends). If, for example, an event causes (or should cause) an agent to modify or abandon her earlier ends, that event has narrative significance. Going beyond definitions, Kauppinen presents an account of how narratively meaningful acts and events contribute narrative value or disvalue to a life. According to Kauppinen, narrative value fully obtains in the case that (i) an agent is successful in meeting her goal(s), (ii) the agent’s goals are worthy of pursuit, and (iii) the agent deserves the success by exercising “the capacities that make us the kind of agents we are” in pursuit of the goal (2015, 206). Positive narrative value, on this view, is contributed to a life in the case that an agent deserves success for valuable achievements, and such value is diminished or rendered negative if the agent fails, pursues worthless goals, or does not deserve success. Kauppinen’s account provides an initially plausible analysis of at least some central cases of narrative value. For instance, it explains why the life of Uphill has more narrative value than that of the Downhill because Uphill is more deserving of the victory and is ultimately successful in meeting his worthwhile ends whereas Downhill is less deserving of the victory and is ultimately unsuccessful. Kauppinen’s emphasis on deserved success is especially attractive, for it illuminates why the narrative meaning of an event or act may alter the value of that event or act, considered in isolation. After all, deserved achievements are fitting objects of both pride and gratitude whereas windfalls are only fitting objects of gratitude.7 Furthermore, undeserved “benefits” sometimes contribute negative value to a subject’s life, such as in the case of receiving a reward for doing something that one did not in


Nicholas Ryan Smith

fact do. Considered in isolation, such a reward may appear to be good for the recipient, yet, given its full narrative meaning as an undeserved reward, it is revealed as a fitting object of shame and thereby a badmaking feature of a life, assuming that good-making features of a life are not fitting objects of shame, an assumption I attempt to make plausible in the next section. However, although Kauppinen’s account has a number of attractive features, I argue in the final section of the chapter that it fails to capture all cases of narrative value and disvalue. I contrast this failure with the success of a virtue-ethical account of well-being.

3. Constraints on a Satisfactory Account of Well-Being In the preceding section, I hope to have made plausible the claim that the full meaning of an act or event—that is, what I understand to be the most apt description of that act or event for the purpose of evaluating how it bears on the value of a subject’s life—(i) partly depends on that act or event’s relations to other acts and events and (ii) is significant to whether and how that act or event contributes to the value of a subject’s life. If this claim is plausible, then it places a constraint on a fully successful account of well-being. Namely, if a proposed account of well-being is not sensitive to narrative meaning and value—that is, if it implies that acts or events with narrative value do not contribute to a subject’s good or that acts and events with narrative disvalue do not make a subject worse off—I will say that that account is inadequate in respect to narrative value. Using this constraint and two others, which I motivate in what follows, I hope to support a virtue-ethical account of well-being by showing that it enjoys a number of advantages over hedonism, desire-satisfaction theory, and standard objective list theory. Before discussing rival accounts of well-being, consider how we might naturally arrive at the question that accounts of well-being attempt to answer: the question of what makes a person’s life good—that is, valuable or desirable—for her. The claim that some lives are better than others—along with our specific views about what makes some lives better than others—informs our deliberation. In deliberating about, for example, whether to propose marriage, accept a job offer, or move to a new place, a person characteristically considers the salient ways that her choice bears or may bear on the value of her life. Is getting married to this person at this time likely to make my life better? If I return to my hometown, would I be better or worse off? Such deliberation indicates, perhaps unsurprisingly, that people regard facts about their well-being as a source of reasons about what is choiceworthy: that what makes an act good for a subject provides a defeasible reason for that subject to perform that act. This is plausible simply because an agent has at least defeasible reason to do what is good for herself.

Well-Being, Narrative Value, and Virtue


Our view of well-being impacts not only our deliberation but also our emotional outlook on our lives. A person may feel a variety of emotions about her life, many, though perhaps not all, of which presuppose the idea that some lives are better than others. If, in reflecting on her life, a person feels regret, shame, or frustration, she regards her life as worse than it could have been. If, on the other hand, a person feels gratitude, pride, or a sense of fulfilment and achievement, she regards her life as better than it could have been. Such emotions are responsive to facts about the value of one’s life, which indicates that facts about the value of one’s life serve as reasons why such emotions are fitting or unfitting. A person only has reason to feel bad about her life in response to bad-making features of her life and only has reason to feel good about her life in response to good-making features of her life (or, with hope and fear in mind, goodmaking and bad-making possibilities). Of course, one may be mistaken about what one takes to be a fact about the value of one’s life; that is, one may believe that a fact about one’s life makes it more or less valuable even if it does not. Scrooge, for instance, may believe that wealth is the sole ingredient of well-being even if it is not. A person who obsessively counts blades of grass all day may believe that the counting enhances the quality of their life although it does not (cf. Rawls 1971, 432–433). Additionally, a person may not recognise facts that make his life more or less valuable. For instance, a person with imposter syndrome may not recognise the value of his achievements because he does not recognise them as achievements at all but rather as the product of luck. Or a person may feel quite satisfied with his comfortable albeit mediocre life because he does not recognise the value of developing his talents. Since one may be mistaken about the value of one’s life, the question arises: what constitutes a life that is good for the one who lives it? What makes someone’s life go better or worse for them? What is it to live and fare well? In accordance with the observation made just before, I assume that good-making features of an act for a subject contribute to the choiceworthiness of that act. As a result, if an account of well-being implies that feature F of an act is good for an agent although F provides no reason at all or an opposing reason for the subject to perform that act, that account of well-being is not fully satisfactory. For example, if an account of wellbeing implies that a desire to engage in a pointless activity, such as counting blades of grass for its own sake, contributes to the agent’s well-being, that is a defeasible reason to reject that account, since pointless desires do not appear to provide any reason at all for the subject to satisfy them and thus do not appear to make an agent any better off. An account of well-being that is unsatisfactory in this respect may be called inadequate in respect to practical reason in virtue of its failure to limit good-making features of an act for a subject to those features that provide practical reasons.


Nicholas Ryan Smith

Additionally, since a view of well-being informs and allows for the evaluation of our emotions, I assume that good-making features of an act or event for a subject contribute to the fittingness of positive emotions, including pride and gratitude, in response to that act or event, whereas bad-making features contribute to the fittingness of negative emotions, such as regret, frustration, or shame. Consequently, if an account of wellbeing implies that a feature F of an act or event is good-making for a subject but F does not contribute to the fittingness of positive emotions, then that is a defeasible reason to reject that account. For example, if an account of well-being implies that a person’s counting blades of grass in his yard every day last year makes his life better, that account is unsatisfactory for the reason that such activities do not appear to make that life any more fitting as an object of pride, gratitude, or other positive emotions. An account of well-being that is unsatisfactory in this respect may be described as inadequate in respect to emotional fittingness.8 Thus far, I hope to have motivated three constraints on a satisfactory account of well-being: adequacy in respect to narrative value, practical reason, and emotional fittingness. Do any accounts of well-being meet all three?

4. Hedonism and Desire-Satisfaction Theory In the contemporary literature on well-being, following Derek Parfit (1984, 493–501), three (kinds of) accounts of well-being take centre stage: hedonistic, desire-satisfaction, and objective list theories. According to hedonism, pleasures and pains, regardless of both their objects and causes,9 are the only factors that directly contribute to the (dis)value of a life. One life is better than another to the extent that it features more pleasure and less pain. Since hedonistic accounts imply that cruel and intemperate pleasures make a person’s life better, there is reason to doubt whether it is adequate in respect to practical reason and emotional fittingness. As holists about practical reasons often point out, the fact that one would derive pleasure from harming someone does not appear to provide any reason at all to harm them (see e.g., Dancy 1993, 61). Additionally, such pleasures seem clearly to be a fitting object of shame and regret. Thus, hedonism is evidently inadequate in respect to both practical reason and emotional fittingness. Hedonism is, in addition, inadequate in respect to narrative value, for the reason that deserved rewards need not be more pleasurable than undeserved ones and succeeding in worthwhile goals need not be more pleasurable than succeeding at worthless goals.10 Therefore, the full meaning of an action or event is not targeted by hedonism’s concerns. If we accept hedonism, the most apt description of an act or event for the purposes of evaluation is just a description of how much pleasure or pain it causes the subject; facts about what is done and what happens which

Well-Being, Narrative Value, and Virtue


do not bear on pleasure and pain are superfluous to hedonistic evaluation. Hedonism is thus at odds with narrative value. Desire-satisfaction theories claim that the only factors that directly contribute to the (dis)value of a life are the satisfaction or frustration of a person’s desires or preferences. One life is better than another to the extent that the subject of that life has more (or stronger) desires satisfied or fewer desires unsatisfied. Such a theory is susceptible to similar objections to hedonism, for not all desires appear to provide practical reasons. The satisfaction of pointless desires, such as counting blades of grass as something desired in itself, does not appear to make the subject any better off. Even when the satisfaction of desire is good for an agent, this is, in many cases, intuitively because the object of that desire is independently good for them rather than because it satisfies the agent’s desire.11 Furthermore, insofar as the satisfaction of some desires is a fitting object of regret and shame, desire-satisfaction theories are at least partly at odds with emotional fittingness. Do desire-satisfaction theories capture narrative value? Not adequately. On a desire-satisfaction theory, the most apt description of what is done or what happens for the purpose of evaluation is a description of what the agent desires and the features of acts or events that relate to the content of that desire. If the agent has no greater desire for deserved success over undeserved success and no desire to learn from his mistakes, then the narrative value of such acts is not captured by desire-satisfaction evaluation.12 Whether the narrative value of an act or event matters entirely depends on what the agent desires. (The same problems are present even if we shift to talking about what an agent would desire if she were fully informed and unimpaired.) Thus, desire-satisfaction theories do not capture the narrative value and disvalue of acts and events that fall outside the content of an agent’s desires.

5. Objective List Theory and Virtue Ethics Objective list theories are not as neatly definable as hedonism and desiresatisfaction theories. Such theories are pluralistic in holding that more than one kind of thing can make someone’s life better or worse and objectivist in holding that such things can be good or bad independently of whether they are desired. Yet these two commitments are not sufficient to distinguish objective list theories from virtue-ethical accounts. Standard objective list theories (which I will hereafter call “objective list theories”), however, are distinct from virtue ethics. Good-making items on an objective list theory include pleasure, achievements, friendship, knowledge, and the appreciation of beauty. One life is better than another, according to objective list theory, to the extent that it features items on the list of objectively good things and lacks items on the list of objectively bad things, such as pain, failure, isolation, and ignorance.


Nicholas Ryan Smith

Objective list theory is an improvement on hedonism and desiresatisfaction theory. For example, objective list theories allow that counting blades of grass as a final end does not benefit a person even if they desire to do it. They may also allow that events or acts involving pleasures can be evaluatively mixed. If someone takes pleasure in something that is bad for them, such as ignorance or bodily harm, objective list theory implies that the pleasure is good insofar as it involves pleasure but bad insofar as involves ignorance or bodily harm. Given that we have reason to avoid ignorance and bodily harm and that ignorance and bodily harm are fitting objects of negative emotions, objective list theory appears to better track practical reason and emotional fittingness than hedonism or desire-satisfaction theory. The paradigm of virtue-ethical accounts of well-being is provided by Aristotle, who conceives of eudaimonia, which I take to be equivalent to living and faring well, human flourishing, a good life, or well-being, as being constituted by rational activity in accordance with virtue in a complete life (1999, 1098a 3–18; see also previous fn. 2). With Aristotle’s characterisation as a paradigm, I take it that a minimal commitment of virtue-ethical accounts of well-being is the claim that living in accordance with virtue is at least part of what it is to live well: that virtuous activity contributes to the desirability of a person’s life whereas activities and decisions that are contrary to virtue contribute to the undesirability of a person’s life. According to Christine Swanton, objective list theory differs from virtue ethics insofar as the former but not the latter is committed to the Thesis of Non-Aretaic Value, which is, in part, the claim that there are objective values and “in relation to these values, certain actions and attitudes are required, permitted, or intrinsically good; and these actions and attitudes are not themselves virtue-based notions” (2003, 35).13 The idea is that activities such as acquiring knowledge and enjoying pleasures are activities that are possible to engage in without meeting a threshold of virtue and that, according to objective list theory, engaging in such activities always contribute to the good of a subject. Virtue ethics, by contrast, maintains that acquiring knowledge and enjoying pleasures do not always contribute to the good of a subject. Thus, Swanton claims that “pleasure, friendship, and failure, for example, are neither good nor evil in themselves. What is good, for example, is pleasure and friendship handled well [i.e., in accordance with virtue]—when ‘handling’ covers behaviour, motivation, and emotional response” (2003, 36).14 Note however that virtue ethics should and can acknowledge that some things, such as friendship (with the right people), knowledge (about important matters), and (bodily, tactile) pleasure are characteristically worthy of pursuit whereas other things, such as isolation, ignorance, and suffering are characteristically to be avoided. About such claims, virtue ethics does not disagree with objective list theory. Where virtue ethics

Well-Being, Narrative Value, and Virtue


differs is in the claim that the items on the list of the most plausible objective list theory should be construed as items in the field of a virtue— objects that a virtue is directly concerned with—and, consequently, that such items may be responded to, desired, loved, used, avoided, or pursued in a variety of ways, not all of which are in accordance with that virtue. For instance, bodily, tactile pleasures are items in the field of temperance. Yet, not all such pleasures are enjoyed or pursued (or avoided) in accordance with temperance. Objective list theories but not virtue ethics imply that such pleasures are good for the agent. Likewise, items in the field of friendship include opportunities to make friends, to help friends, and to bond with friends. Yet, not all such opportunities are responded to in accordance with virtue. To act in accordance with the virtue of friendship, one must, for example, bond with, make, and help friends for the right sort of reasons, with the right sort of feelings, and by doing the right sort of things. Additionally, one must choose the right sort of people as one’s friends. If one’s motivating reasons, acts, or feelings fall below the threshold of virtue and into the territory of, for example, merely going through the motions of friendship, using or being used by one’s friends, seeing one’s friends as beneath oneself, or participation in toxic friendships, then the contribution of the friendship to one’s well-being is diminished or even rendered negative. Consideration of the disvalue of intemperate pleasures and contraryto-virtue friendships reveals that virtue ethics is more adequate in respect to practical reason and emotional fittingness than objective list theory. Intemperate pleasures are fitting objects of negative emotions, such as regret and shame, and their prospect provides no reason to act, yet objective list theory counts them among the good-making features of a life. Contrary-to-virtue friendships are fitting objects of frustration, regret, and anger (at oneself or another), and the mere fact that they are nonetheless friendships provides no reason to participate in such friendships, notwithstanding that objective list theory counts them among the good-making features of a life. Moreover, these examples should not surprise us because acts that are contrary to virtue are, as a matter of course, fitting objects of negative emotions. Even if one misses the target of virtue in a way that is blameless, missing the target of virtue is a fitting object of frustration and regret if not shame, blame, or guilt.15 (Though forgiveness, love, and compassion toward a flawed self, as well as hope, are not to be forgotten.)

6. Virtue Ethics and Narrative Value In this final section, I hope to show how virtue ethics explains and illuminates the significance of narrative meaning and value. Recall the discussion of Slote’s case of the politicians. Uphill attains a high political office late in life and accomplishes many great things after being in the political wilderness throughout his early life. Downhill attains the same


Nicholas Ryan Smith

office early in life, but loses power while still young, never to regain it. The question is why the life of Uphill seems better, and, inspired by Velleman and Kauppinen, I argued that the difference in value does not result from a difference in timing but rather from differences in the meanings of the victories, struggles, and failures in the lives. Although both attained the same office and struggled for long periods, their achievements and struggles are not the same, for they have different narrative meanings. However, this solution is incomplete because the lives are, in fact, under-described. Essential to the meaning and (dis)value of the victories, struggles, and failures, according to virtue ethics, is how the victories were achieved, why the men failed, and why they persisted so long despite repeated failures. I show that integrating Velleman and Kauppinen’s solution with virtue ethics renders that solution more plausible and enriches our evaluation of the lives.16 Consider first the value of the victories. Both Velleman and Kauppinen explain the difference in meaning and value of the victories in terms of desert: Uphill deserves the victory more than Downhill. Kauppinen is explicit about what grounds the relevant kind of desert. We deserve success in a pursuit insofar as we “exercise the capacities that make us the kind of agents we are,” which is distinct from merely making an effort to succeed, since “mere effort is often stupid: you don’t do well as an agent if you expend a lot of physical effort instead of making use of intelligence and self-control, for example” (Kauppinen 2015, 206). This raises the question of what exactly an agent deserves by intelligently exercising his human capacities since, assuming that the political victory is a result of an election, it may be objected that the victor deserves the victory simply in virtue of winning the election, regardless of how intelligently he exercised his human capacities. The problem with this objection, of course, is that it fails to target the relevant object of desert. When Kauppinen says that Uphill deserves the victory more, I believe that he has in mind that, in virtue of the differing paths to victory, Uphill is more praiseworthy than Downhill. Since we only deserve praise for good things that we are responsible for, and we deserve more praise the more responsible we are, it is easy to see how intelligent use of our capacities contributes to our praiseworthiness. Thus, of the value of deserved success, Kauppinen (2015, 219) writes: You had luck, both in having opportunities for pursuing valuable goals, and in things turning out your way. But your success wasn’t due to sheer luck, or it wouldn’t be as richly deserved as it is. It is fitting for you to look back on your life with a mix of pride and gratitude, and for others to look on it with admiration and even wonder. Deserved success is thus revealed to be more valuable insofar as it makes one’s life more praiseworthy and insofar as pride and not merely gratitude

Well-Being, Narrative Value, and Virtue


is ftting,17 when the greater value of deserved over undeserved success is a key case of narrative value. However, praiseworthiness and pride are sensitive to a great deal more than intelligent exercise of human capacities, as virtue ethics reveals. Suppose, for instance, that Uphill won the election by coordinating a social media disinformation campaign, bribing and blackmailing dissenters and sources of unfavourable information, or through election interference or tampering. Nothing about this, as far as I can tell, needs to change whether and to what degree Uphill intelligently—that is, cleverly, not wisely (cf. Aristotle, Nicomachean Ethics, 1144a 25–8)—exercises his capacities. However, such facts certainly alter both the meaning and narrative value of the victory, for the victory is now aptly described as dishonourable, a fitting object of shame rather than pride. Virtue ethics, in its claim that the value of an achievement depends on whether it is achieved in accordance with virtue, perfectly captures this change in narrative value. Additionally, virtue ethics reveals important ways that failures may contribute positive narrative value to a life, for failures may be honourable. Christine Swanton (2003, 36–37) admirably illustrates this point in a passage that deserves to be quoted at length: Consider the supposed intrinsic evil of failure with Prada’s failure in the America’s Cup in February to March 2000. Prada, the Italian syndicate, lost the America’s Cup match in Auckland, New Zealand, 0–5. A spectacular failure on the surface—they were beaten by a better crew and a faster boat. But in what way was their failure intrinsically evil? One cannot separate the failure from the surrounding context, its history, and its manner. The context includes their defeat by a vastly more experienced match-racing crew and shore team with the home advantage in some of the most difficult sailing conditions in the world. The history must take account of the nature of the Prada campaign— their years of hard work mastering conditions on the Hauraki Gulf and boat and sail testing. The manner included the style, graciousness in victory and defeat, impeccable conduct of the Prada team, and the skipper in particular, throughout their incredibly tough five-month campaign in the Louis Vuitton challenger series and the match itself. Their conduct made the team stunningly popular in Auckland and New Zealand generally, so much so that they took part in the tickertape parade—normally a victor’s parade—where they received cheers almost as loud as those given Team New Zealand. Failure suffused with so much virtue and tinged with so little vice is not intrinsically evil in my view. . . . The failure was honourable. Since, for virtue ethics, the (dis)value of an act or event for an agent depends on whether it or the agent’s handling of it is in accordance with


Nicholas Ryan Smith

virtue, virtue ethics captures the narrative value and praiseworthiness of virtuous failures. Kauppinen’s narrative calculus, by contrast, does not fully capture the honourability of Team Prada’s failure because the praiseworthiness of Team Prada would simply be a result of the intelligent use of human capacities without reference to the graciousness and good sportsmanship that is displayed in their performance and in their handling of the failure. Building on Kauppinen’s model of narrativity as goal-directed human agency extended over time, a virtue-ethical alternative to Kauppinen’s narrative calculus may be developed as the claim that full narrative value obtains in the case that (i) an agent is successful in meeting her goals, (ii) the agent’s goals are worthwhile, when worthwhileness requires being in accordance with virtue, (iii) the agent is responsible for her success (her intelligent exercise of human capacities, as opposed to mere luck, is a significant cause of her success), and (iv) the agent’s pursuit is in accordance with virtue. The final requirement—that the agent’s pursuit is in accordance with virtue—is the distinctive claim of a virtue-ethical account of narrative value. It is this requirement, I suggest, that allows for more illuminating and specific diagnoses of the ways that acts can be narratively valuable or disvaluable. Since there are many virtues and each virtue has distinctive requirements, a virtue-ethical account offers a rich set of categories for diagnosing what is valuable and what is disvaluable in cases of goaldirected human agency extended over time. The specificity and richness of virtue ethics’ diagnoses is revealed in considering the case of Downhill’s struggles to regain political power. What, other than the fact the struggles do not pay off in the end, makes his struggles possess narrative disvalue? That depends on why he persevered so long. Was it out of an excessively rigid vision of the good life for himself? Was he cold to other ways that he might live a good life outside of a political career? Did he thereby neglect alternative opportunities? Such rigid perseverance is contrary to the virtue of hope, if, as I have argued elsewhere (Smith, forthcoming), the virtue of hope requires malleability in our specification of what a good life might amount to in our case. If so, Downhill’s struggles are regrettable not only because he struggled in vain but also because his struggles, being motivated out of excessively rigid hope, closed him off to other opportunities for achieving a good life. Alternatively, the narrative disvalue in Downhill’s struggles may not be due to excessively rigid hope but rather, say, an excess in filial piety. If Downhill was a member of a political dynasty family, he may have persevered so long in the pursuit to regain power out of inappropriate deference to his family’s expectations. If that were the case, then it may be that what is disvaluable about Downhill’s struggles is not merely that he struggled in vain but also that his struggles were a manifestation of

Well-Being, Narrative Value, and Virtue


his failure to adequately discover or create himself as an individual who is more than merely part of a family. What such possibilities reveal, I suggest, is that narrative disvalue has many possible sources and does not arise only out of failure, worthless goals, or lack of intelligence (conceived of as cleverness), at least to the extent that not all failures in virtue are reducible to either failures in meeting worthwhile goals or failures in intelligence. Although Kauppinen’s narrative calculus makes, on my view, a substantive and insightful contribution to understanding narrative value, virtue ethics more fully captures narrative value and disvalue. If so, virtue ethics is more fully adequate in respect to narrative value. In virtue of its adequacy in respect to narrative value, practical reason, and emotional fittingness, there is reason to regard virtue ethics as providing an account of well-being that is superior in important respects to its most salient rivals.18

Notes 1. According to virtue-ethical accounts of well-being, which I support later in the chapter, acts and events contribute to the goodness of a subject’s life only if they or the subject’s handling of them are in accordance with virtue. I omit this qualification throughout the chapter unless necessary for the point being made. Additionally, I hereafter omit “typically” qualifiers unless necessary. 2. In this chapter, I treat phrases such as “well-being,” “eudaimonia,” “flourishing,” “a good life,” and “living and faring well” as equivalent. Concerning their equivalence, I agree with Daniel M. Haybron (2007), who points out that reserving different phrases for different theories “obscures the issues, leaving it unclear how (say) Aristotelian and Utilitarian accounts of value are opposed” and that such theories “seem clearly to concern a common subject matter—what benefits a person, is in her interest, makes her life go better for her” (1fn3). However, Haybron later argues that there is a conceptual distinction between well-being and a good life, and he would likely see this chapter as advancing a theory of the latter while making the mistake of “dragoon[ing]” an Aristotelian theory of the good life “into service as [an account] of wellbeing.” (2007, 21). While there may be a conceptual distinction between well-being and a good life, such that it is not a conceptual confusion to claim that well-being is compatible with viciousness, I nonetheless believe that virtue is (nonanalytically) necessary for well-being, and that living in accordance with virtue contributes to—but is not sufficient for—well-being. My reasons are developed later in the chapter, starting in the third section. 3. For Dorsey’s description of the lives, on which I have relied, see (2015, 304– 305). After Dorsey’s article was published, Simpson was released after serving 9 years. 4. Strictly speaking, there can be no counterexamples to the SLH since it merely claims that upward slopes can be valuable. However, would-be counterexamples nonetheless indicate that upward slopes are not valuable simply in



6. 7. 8.


10. 11.

12. 13. 14.

15. 16. 17. 18.

Nicholas Ryan Smith virtue of being upward slopes: something else explains their value when they are valuable. Dorsey treats Velleman’s view as a proposed explanation of the SLH, but I think that it is best interpreted as an alternative to the SLH, for it locates value in the meaning of acts and events rather than in their temporal order. However, Dorsey (2015, 310) may ultimately agree since he interprets Velleman as holding that the shape of a life has “signatory value,” value as a sign of something else that is the true source of value. This example is derived from (MacIntyre 2007, 210). Compare (Kauppinen 2015, 219), which I discuss in the final section of the chapter. On my view, the fittingness of an emotion is determined by the accuracy of that emotion’s cognitive content, whether that cognitive content is construed as doxastic or, as I prefer, perceptual. If the cognitive content of fear is that its object is dangerous, then one fittingly fears x in the case that x is dangerous and one unfittingly fears x in the case that x is not dangerous. I allow that fitting emotions may be all-things-considered inappropriate or inadvisable for reasons that do not concern fittingness. Of course, I also allow that people may feel or be disposed to feel a certain way toward x even if their emotions are unfitting. Fittingness is a normative notion. This implies that Mill does not endorse a hedonistic account of well-being. Although Mill is commonly categorised as a hedonist, I accept the argument that if something’s pleasurableness and only its pleasurableness is what makes it good for someone, then the objects and causes of pleasures can matter only to the extent that they produce more or less pleasure in the subject. For discussion of this argument, with objections, see (Crisp 1997, 32–35). Dorsey (2015, 309–310) makes a similar point about what he calls the Instrumental View. T.M. Scanlon makes an analogous point about the relation between desires and practical reasons: “Having a desire to do something typically involves thinking, or imagining, that there is something to be said for doing it. And from the point of view of an agent it is the considerations that are seen as desirable rather than the fact of having the desire, that provide reasons for acting” (2014, 88). I am here indebted to Dorsey’s (2015, 310) evaluation of what he calls the Pro-attitude View. The other part of this thesis concerns the conceptual priority of virtue and value, which is not my focus. Although Swanton introduces this claim with the qualifier, “This allows for the possibility that . . .”, the context suggests that Swanton accepts the claim. It should also be noted that Swanton is attempting to account for “value” rather than “well-being” in her discussion of objective list theory and virtue ethics. Even if her view is not intended to be about well-being, I nonetheless believe that it is plausible as a view of well-being. Compare Liezl van Zyl’s (2014, 122) discussion of frustration in tragic dilemmas. I am indebted to Swanton’s discussion of how virtue ethics contrasts with Kauppinen’s view in the penultimate section of her chapter (this volume). Although, on Hume’s celebrated analysis of pride, we can fittingly feel proud of aspects of ourselves that we are not responsible for, it remains plausible that greater responsibility warrants greater pride. I thank Liezl van Zyl, Christine Swanton, and Eldar Sarajlic for helpful comments.

Well-Being, Narrative Value, and Virtue


References Aristotle. 1999. Nicomachean Ethics, 2nd ed., T. Irwin, trans. Indianapolis, IN: Hackett Publishing Co. Crisp, R. 1997. Mill on Utilitarianism. London: Routledge. Dancy, J. 1993. Moral Reasons. Oxford: Wiley-Blackwell. Dorsey, D. 2015. “The Significance of a Life’s Shape.” Ethics 125(2): 303–330. Haybron, D.M. 2007. “Well-Being and Virtue.” Journal of Ethics and Social Philosophy 2(2): 1–27. Kauppinen, A. 2015.“The Narrative Calculus.” In M. Timmons, ed., Oxford Studies in Normative Ethics, vol. 5. Oxford: Oxford University Press, pp. 196–220. Lear, G.R. 2015. “Aristotle on Happiness and Long Life.” In Ø. Rabbås, E.K. Emilsson, H. Fossheim, and M. Tuominen, eds., The Quest for the Good Life: Ancient Philosophers on Happiness. New York, NY: Oxford University Press, pp. 127–145. MacIntyre, A. 2007. After Virtue: A Study in Moral Theory, 3rd ed. South Bend, IN: The University of Notre Dame Press. Parfit, D. 1984. Reasons and Persons. Oxford: Clarendon Press. Rawls, J. 1971. A Theory of Justice. Cambridge, MA: Harvard University Press. Scanlon, T.M. 2014. Being Realistic about Reasons. Oxford: Oxford University Press. Slote, M. 1982. “Goods and Lives.” Pacific Philosophical Quarterly 63(4): 311–326. Smith, N.R. Forthcoming. “The Virtue of Hope.” In G. Pettigrove and C. Swanton, eds., Neglected Virtues. London: Routledge, 20 pp. Swanton, C. 2003. Virtue Ethics: A Pluralistic View. Oxford: Oxford University Press. Swanton, C. This volume. “Virtue Ethics and Narrative Virtue.” In J. Ulatowski and L. Van Zyl, eds., Virtue, Narrative, and the Self: Explorations of Character in the Philosophy of Mind and Action. London: Routledge, pp. 19–34. Van Zyl, L. 2014. “Right Action and the Targets of Virtue.” In S. van Hooft, ed., The Handbook of Virtue Ethics. Durham, UK: Acumen, pp. 118–129. Velleman, J.D. 2000. The Possibility of Practical Reason. Oxford, UK: Clarendon Press.

10 On the Value of Moral Failure Damian Cox

1. Introduction In the first episode of Michael Schur’s comedy television series The Good Place, an immortal, Michael (played by Ted Danson), explains the basis upon which one enters the “Good Place,” a place of heavenly reward. He says: You are all, simply put, good people. But how do we know that you are good? How are we sure? During your time on Earth, every one of your actions had a positive or a negative value depending on how much good or bad that action put into the universe. Every sandwich you ate, every time you bought a magazine, every single thing you did, had an effect that rippled out over time and ultimately created some amount of good or bad. You know how some people will pull into the breakdown lane when there is traffic, and they think to themselves “ah, who cares, no-one’s watching”? We were watching. Surprise! Anyway, when your time on Earth is ended, we calculate the total value of your life using our perfectly accurate measuring system. Only the people with the very highest scores—the true cream of the crop—get to come here, to the Good Place. What happens to everyone else you ask? Don’t worry about it. The point is, you are here because you lived one of the very best lives that could be lived. (2016, Episode 1) This is a consequentialist rendering of a familiar idea: that the moral achievement of a life can be calculated by summing up the moral worth of all of an individual’s actions. The idea is rarely conveyed in such an explicit and unadorned manner. In some versions of Catholic eschatology, for example, the image of a soul’s weight stands in for the sort of moral accountancy set out in The Good Place. When the dead rise from their tombs at the Second Coming of Christ, St Michael will hold the scales of judgement and weigh the souls of the resurrected. A righteous soul will

On the Value of Moral Failure


tip the balance of the scales downward and be sent to heaven. In ancient Egyptian traditions, the equation is reversed. A dead person’s mummifed heart is weighed by Anubis, the God of Death. A heart heavier than the feather of Ma’at, the goddess of order, truth, and righteousness, will die a second death and be annihilated. The Good Place reimagines this process for an audience at home with the idea of a digital spreadsheet. Many religious traditions, including those of Ancient Egypt, Christianity, Hinduism, and Islam, encode some version of a summative moral judgement: the judgement of the moral quality (or, in more explicitly religious terms, the righteousness) of one’s entire life. In secular moral philosophy, a corresponding idea persists. It is shorn of a commitment to actual post-mortem judgement—secular philosophers are unlikely to commit themselves to the existence of an angel of death swooping down to pass judgement upon the righteous and the wicked—but it persists in modal form. The possibility of such a final summative judgement is hinted at in much philosophical work. For example, Saul Smilansky (1997, 124) sets out a version of what he calls “moral accountancy.” I have called elsewhere what morality requires in the way of people doing good and not doing bad the “substantive” part of morality, and the aspect of morality where people get the good or bad credit for their actions the “accountancy” part. . . . “Negative accountancy” involves attitudes and practices such as condemnation, blame, punishment and the like. “Positive accountancy” includes attitudes and practices such as praise, gratitude and admiration. Kurt Baier (1970, 18) puts the point in terms of marks and how one gains them. In passing moral judgement on a person, the chief criterion is the person’s performance in satisfying what, morally speaking, can be expected of him. He gets good marks if he does more than can be expected of him, bad marks if he does less, and no marks at all if he does exactly that. Although summative moral judgement is a modal concept in contemporary ethics—that is, it is a claim about what would result if an overall moral life assessment were taken, even if it is very unlikely for most of us that such an assessment ever will be taken—the idea appears to play an important role in our picture of a good life and in our idea of moral exemplars. A moral exemplar, someone who we ought to seek to emulate, is someone who would receive a strongly positive summative moral judgement. A good life is, plausibly, a life that would earn a strongly positive summary moral judgement. The important question here is how to characterise summative moral judgement.


Damian Cox

In this chapter, I contrast two pictures of summative moral judgement. I begin by setting out a version of the accountancy model of summative moral judgement, improving upon the rough sketch offered by Michael in The Good Place. I call this “the forensic model.” I then contrast it with what I call “the narrative model.” In the final part of the chapter, I adduce reasons to prefer the narrative model. The key consideration in my analysis is the role of narratives of moral failure in summative moral judgement. I argue that only the narrative model of summative moral judgement can do full justice to the role that moral failure plays in our lives.

2. A Forensic Model of Summative Moral Judgement The Good Place model of summative judgement is straightforwardly consequentialist. Actions have consequences, both in the short term and the long term. They “ripple out over time and ultimately [create] some amount of good or bad.” Michael’s exposition doesn’t go into details (episodes of The Good Place are short, it is meant to be a comedy), but presumably the idea is that a single action’s various effects over time can be tracked, evaluated, assigned a numerical value (positive and negative), and summed. The graphic accompaniment to Michael’s exposition makes the digital nature of the process apparent. But it is no less mysterious for that. The effects of an action depend on context, including the performance of other actions, so the contribution of a single act will be hard to keep track of. And, of course, we will need a criterion of individuation for single actions. Perhaps the measure is counterfactual. Imagine that the total value state of the world is given a numerical measure. Consider what that total is given that action τ is performed. Now, consider what it would have been if τ had not been performed. The difference is the value contribution of τ. Unfortunately, this leads to a lot of double counting. If a good outcome depends on the performance of both τ and another action, φ, they will jointly contribute double points to the summation of one’s value-impact. The world without τ would be a poorer place; the world without φ would be an equally poorer place. Both τ and φ attract the full value of the difference they jointly make. Compare this to an equivalently good outcome that only depends upon a single action. Doing a good deed in two necessary stages would be twice as good as doing it in a single hit. Obviously, there are technical obstacles to be overcome if we are to make sense of the world-value contribution of singular actions and their summation. Any such technical difficulties are moot, however, because the very idea of arriving at a summative moral judgement in this way does not withstand scrutiny. Whether consequentialist metrics track moral rightness or wrongness is a long-standing issue in normative ethics, of course, but rightness is not being tracked in The Good Place. Desert

On the Value of Moral Failure


is being tracked. As a measure of desert, an objectively consequentialist metric like the one sketched in The Good Place is a non-starter. Consider the figure of the lucky villain. The lucky villain acts consistently to make the world a poorer place but fails at each turn. Every action intended to wreak havoc has a calming, benign effect on all around him. Every plan to humiliate an innocent person leads instead to their greater good. If lucky enough, the lucky villain would most certainly get into The Good Place. He would not deserve to be there. This point about The Good Place scenario generalises to all models of summative moral judgement that are metric-based. The relevant metric needs to track, not good outcomes per se, but moral desert. I call models of summative moral judgement that succeed in tracking desert “forensic models.” They are concerned to allocate and combine judgements of the deservingness of moral approbation or condemnation such that a forensically precise account of the overall moral achievement or failure of a life can be obtained, and different lives can be compared precisely, one to the other. I will propose in outline such a model in what follows. I do so, not to endorse it, but to set out an intuitive and coherent forensic model of summative moral judgement and allow a clear and fair contrast with the narrative model I will also describe. A basic feature of the forensic model I propose is that it is based on the moral credit or discredit of actions. Restricting judgement to actions (and to negligent failures to act—see the discussion in what follows) allows the forensic model to track what a person has done (or failed to do) over their lifetime, including what they have done to improve their character or diminish the potential harm of unbidden desires, temptations, or unwelcome fantasies. It tracks what we do in our lives, not what we have to put up with in them. How does an action (or failure to act) attract moral credit or discredit? What are the constituents of moral credit? I suggest they are: acting rightly; the significance of the action; the difficulty, risk, and cost of the action. And the constituents of moral discredit? I suggest they are: acting wrongly; the significance of the action; mitigation of the wrong (if any). Let me start with acting rightly. Moral credit attaches to us when we act rightly. I propose, for the purposes of the model at hand, to understand acting rightly in the following terms. We act rightly if and only if we respond to good reasons in an appropriate way. On this general picture of rightness, rightness is a property of reason-responsiveness. Actions are often done for reasons. (They needn’t be; one can choose an action without reason.) Reasons are considerations that, to the agent, count in favour of an action. Good reasons count strongly in favour of an action, independently of the judgement of the agent. (They don’t just seem like good reasons, they are good reasons.). A bad reason is either false or weak, independently of the agent’s judgement. An action is performed for a reason if it would not be performed at all, all else being equal, were the


Damian Cox

reason not apparent to the agent (i.e., were the relevant considerations things that didn’t appear to the agent to count in favour of the action, or didn’t occur to the agent at all). One acts rightly if and only if one acts for good reasons in the right way. This describes only the general structure of forensic models of summative judgement, but that allows us to remain neutral about what makes a reason good and what makes a reason-response right. But consider a few examples. Say that a person sees another in need and knows that only she can help her. She acts for these reasons: she wouldn’t have tried to help if she didn’t believe the other to be in need and believe that, in the circumstances, only she can furnish this help. Now a virtue theorist might say that the action of helping is right because it is a kind thing to do. They might say other things. They might say that the action is right because it is what a virtuous person would do in the circumstance, or because it hits the target of the relevant virtue. But consider the direct view that helping is right when it is kind and thus virtuous. What this means in the context of a reason-responsiveness account of acting rightly is that the help-giver’s reason-response exemplifies kindness and that ensures that she is acting rightly. She does not act because she wants to be kind. She acts on reasons (someone needs help now and only I can help), these are good reasons and she responds to them in such a way as to make her action kind. And that means she acts rightly. Consider the case of a utilitarian lying to the police to aid an innocent fugitive. He acts on reasons: that lying to the police in these circumstances maximises the chance of an optimal outcome. If what he does is right, it is right because he acts on good reasons in the right way. In his own terms, he lies because lying maximises the chances of the best outcome. Responding to utilitarian reasons in this way makes the lie right (assuming the utilitarian standard of rightness is correct). This is an added wrinkle on standard formulations of utilitarianism: actions aren’t made right by the maximally good outcomes themselves, but by their being done for the reason that they maximise chances of maximally good outcomes. The important point for the forensic model of summative judgement is that moral credit is earned through acting rightly, which is determined by the quality of the reason-responsiveness inherent in one’s actions. Moral credit is a function of acting rightly, but there must be more to it than this. Not all actions are created equal from a moral point of view. Some actions are more significant than others. Some actions are more difficult than others. Some actions are more sacrificial than others: more dangerous or costly. My suggestion is that moral credit factor in each of these additional factors. A dangerous, difficult, and yet significant act, for example, earns more credit than a safe, simple, and relatively inconsequential act. Consider two acts of kindness. In one, a person helps a stranger pick up a sheaf of papers he has dropped in a corridor. In the other, a person helps a stranger who has suffered a car accident on a

On the Value of Moral Failure


deserted stretch of road: she stays by the stranger’s side, comforting him, administering first aid, arranging rescue, signaling the hazard for other road-users, all the while missing an important job interview. It seems obvious that the latter action is more credit-worthy than the former. Consider two lives: those of A and B. A and B act rightly in equivalent proportion throughout their lives, let us say 91%. This is to say that 91% of the time both A and B act rightly. However, A acts rightly only when it is relatively easy to do so; he shirks the most difficult moral challenges he encounters. B invariably rises to difficult moral challenges but tends occasionally to slip up on insignificant matters even when they are unchallenging. Were we to measure the creditworthiness of a life merely in terms of the proportion of right acting in that life, A and B would be judged equally creditworthy. But they are not. B has led a morally better life than A. The task for a forensic model, therefore, is to discover a general metric to account for all of the factors that affect judgements of creditworthiness and the relative weights between them. Acting rightly is necessary for moral credit, but significance, cost, risk, and difficulty affect just how creditworthy it is to have acted rightly. Just what makes one action more morally significant than another? How much extra credit do we earn from highly significant actions? Just how much more credit do we earn when we not only act rightly, but sacrifice our own interests in the process? If an action is difficult, that is, if it is arduous and requires determination, skill, and inventiveness, then it earns more credit than an undemanding moral no-brainer. But how much more? Much could be said about the role of moral luck in a forensic model of this sort. For example, people living in morally challenging times possess a kind of moral luck that those living in unchallenging times do not possess: opportunity luck. In challenging times, people encounter opportunities for significant and self-sacrificial actions that those in unchallenging circumstances lack. If getting into heaven were a point-scoring competition, it wouldn’t be a fair one. In trading The Good Place’s consequentialism for a system that more plausibly tracks moral desert, we have made the task of constructing a metric of summative moral judgement very difficult indeed. Nonetheless, if the idea of a forensic model of summative judgement is to work at all, it will need to incorporate a metric that does justice to our informed judgements about the relative weights of moral credit across things such as significance and cost. Without such a metric, and the commensurability it presupposes, there would be no such thing as an accurate forensic judgement of one’s entire moral life. Getting into heaven would be a dice-game. That is the positive side of the leger. What about the negative side? How do people attract moral discredit? How do they get to the Bad Place? They do so either by acting on bad reasons or by negligently failing to act on good reasons. Consider an example of the first situation. A person is woken at night by her cat pawing at her bedroom door, clearly


Damian Cox

wanting to get out. She responds by grabbing the cat by its neck and tossing it, roughly, out of the room. The cat, she thinks, is wilfully annoying her, deliberately trying to deprive her of sleep. It deserves what it gets. These judgements constitute her reasons for tossing the cat. Her responding to them in the way she does (tossing the cat rather than, for example, saying to herself “get over it; it is just a cat; of course, it wants to escape into the night”) makes her action wrong. The reasons that she apprehends and responds to are bad reasons: her cat is not wilfully annoying, no matter how much it seems like it is; her cat does not deserve punishment, no matter how much it seems like it does. Her response to these reasons, acting on them rather than challenging them, makes her action wrong. If this is so, then one way to act discreditably is to respond to bad reasons. But this can’t be the only way. One can also act discreditably by negligently failing to act rightly. In such a case, one negligently fails to apprehend good reasons or one apprehends good reasons and fails to respond to them adequately. Consider the first case. Negligence tracks failures to fulfil standing obligations of moral attentiveness. If I should have seen that my words would cause offence, but don’t, I have failed to apprehend, and thus respond to, a good reason to keep my mouth shut. This is a failure of moral attentiveness, and it redounds to my discredit. (I lose points.) In the second case, I see that my words will cause offence, I really don’t wish to offend, but I have something to get off my chest and that’s all I am concerned with right now. I apprehend a reason to keep my mouth shut but fail to adequately respond to it. I shunt it aside in favour of other reasons. These other reasons—I have something to say, and it is important to me that to say it—are not bad reasons. In other situations (situations in which hurt and offence are not in prospect) acting on them is perfectly in order. In the example at hand, I apprehend both reasons—my reason to speak and my reason to remain silent—and fail to respond appropriately to the sum of them. I ignore the more pressing reason and elevate the less pressing reason. This constitutes my discredit. (I lose points again.) Discredit involves failures of reason-responsiveness. But other factors either increase or decrease discredit. Significance is again a factor. Consider two lives: C and D. Both lives are equally riven through with discreditable actions and discreditable failures of action, let us say 15%. However, C’s moral failures are about relatively insignificant affairs (forgetting to wave thank you when a driver lets them into a busy line of traffic) whereas D’s discredits are serious (defrauding the financially vulnerable). C and D have not lived equally discreditable lives. Mitigating factors are the other considerations affecting levels of discredit. A person with an excuse is less discreditable than a person without one. A person who has been under terrific emotional pressure earns less discredit when they act badly than someone who acts just as badly but who has sailed

On the Value of Moral Failure


tranquilly through life. Factors that mitigate against moral discredit are numerous, and the task of devising a metric for them has a degree of difficulty that is probably even greater than that of devising a metric for moral credit. I won’t attempt a model of mitigation at this juncture but only point out the dependence of any satisfactory forensic account of summative moral judgement on such a thing. Let me summarise the forensic account of summative moral judgement. The overall credit or discredit of a person’s life is the solution of an equation that is built from the following three things. 1. All instances of acting rightly, modified by the significance of the action; the difficulty of the action; the known agent-risk (i.e., perceived, potential, personal cost of acting); the agent-cost (i.e., risk realised). 2. All instances of acting wrongly, modified by the significance of the action; and any mitigation of the action. 3. All negligent failures to act rightly, modified by the significance of the failure; and any mitigation of the failure. The result of such a summative judgement would be a digital spreadsheet in which all of our deeds and all our failures of deed (there is going to be a lot of them) are displayed, one after the other, together with a calculation of their individual merit or demerit. The total down the bottom is the thing to look out for. It represents the exact degree to which we have had a good or a bad life.

3. A Narrative Model of Summative Moral Judgement The forensic model of summative moral judgement appears to be missing something. To illustrate this, let me introduce the figure of a “moral wanton.” The moral wanton is a distant relation of the wanton, a character described by Harry Frankfurt (1971, 11). Frankfurt’s wanton is permanently driven to act on his strongest first-order desires, no matter what they are. He lacks the ability to form and respond to second-order desires: desires that take first-order desires as their objects. The wanton cannot discriminate between his desires except on the grounds of their strength. He cannot dismiss some desires as unworthy of him; he cannot will that he acts on some desires and not others. He is buffeted through life by desires that control his actions, but which he cannot control, influence or integrate into a coherent life-plan. The moral wanton is made of better stuff than Frankfurt’s wanton but suffers a related disability. The moral wanton is permanently disposed to respond excellently to moral reasons. At every occasion for acting rightly, she acts rightly. She does the next right thing, no matter what that right thing is and no matter how it relates to the other right things she has


Damian Cox

done in her life. In other respects, she is like Frankfurt’s wanton. She lacks higher-order moral desires. That is, she lacks the ability to discriminate between moral options based on her desire to follow one path through life rather than another. She is buffeted around by extant opportunities to act rightly without possessing the resources to shape what those opportunities are. According to the forensic model of summative moral judgement, what is wrong with the moral wanton is that she cannot optimise her creditworthiness. She acts rightly 100% of the time, but moral credits are a function of more than acting rightly. The moral wanton lacks the resources (such as second-order moral desires) to shape her life so as to maximise her creditworthiness. She wanders blindly from one occasion of acting rightly to another without developing a plan for maximising the impact, difficulty, and risk of her actions. She is a suboptimal accumulator of credit points. It seems to me that this forensic critique of the moral wanton gets the flaw in the moral wanton’s life backwards. The moral wanton doesn’t lack a strategy for accumulating more points; the points system she works with fails to acknowledge the intrinsic value of a life strategy. The moral wanton’s life lacks narrative shape and overarching purpose. There is little she could say about the purpose of her life except that it is to do good. And this is too shapeless a purpose. I develop this thought by way of an alternative model of summative moral judgement: the narrative model. The narrative model abandons the metric-focused approach of the forensic model. In narrative terms, summative moral judgement is not a matter of evaluating and summing individual actions; it is a matter of constructing and morally evaluating a life story. What are such stories made of? Start with a more general question. What, in general, are stories made of? I think there are four basic elements of any story: character, action, event, and project. A plot consists of a series of significant events that happen to a set of characters, and a series of actions performed by a set of characters, that are organised around one or more projects. Characters in stories can be thinly or thickly described. They can be interesting or uninteresting, conflicted or resolute, admirable or despicable, and so on. Projects in stories are things that structure events and actions, both normatively and teleologically. A project may be a journey, i.e., structured around the goal of arriving at a valued place. It could be a quest, i.e., structured around rising to a valued challenge. A project might be a lesson learned (a valued acquisition of wisdom), a test passed (a valued achievement and the valued acknowledgement of it), an ordeal endured (valued survival against the odds), a relationship begun and nurtured, and so on. Our own projects are not always things we self-consciously enter into; they can be what, retrospectively, we recognise as normatively shaping many significant events and actions of our lives, tying them together in a coherent narrative package.

On the Value of Moral Failure


Stories thus have two basic ordering components: characters and projects. This is important when it comes to understanding life narratives. Life projects are at the heart of them, but life narratives do not consist solely of projects and their fates, and the moral evaluation of life narratives must consist in more than an evaluation of projects. Life narratives also encompass adventitious events and actions. Not everything we do of significance is bundled up in a project. That we do something terrible, seemingly at random, may be part of our story. But this is where character becomes narratively important. Even a random bad action will be related to our character in some way. Perhaps we act out of character in doing it. It is an “out of character” action and this becomes a key part of the narrative. How we respond to our action is also part of our story, and again it is related to character. That we refuse to apologise for our action, for example, may be very much in character for us. Character ensures that even adventitious actions and events are integrated in some way into an overarching life narrative. What is important in the story just told is that we act out of character when we act terribly and that we act in character when we refuse to apologise for it. Life narratives are thus stitched together by two things, in two different ways. They are stitched together teleologically by projects and qualitatively by character. What does a summative moral judgement of a person’s character look like? It will be complicated and messy. One’s character can become a project of its own: for example, we might adopt the goal of becoming more accepting of difference, less prejudicial, and less in-group focused. This is a moral project whose object is our character. Aspects of character change over time. Character can be severely tested or barely tested at all; tested in some ways and not in others. It is complex, with good, bad, and ambivalent aspects. It involves recognised virtues and vices but also a broader range of strengths and weakness, attitudes and dispositions, emotional temperaments, habitual moods, styles of friendship, of conversation, of work, of aesthetic attention, and so on. Aspects of character can be entrenched more or less deeply. They are often situational: we may manifest a certain character in one situation and an altogether different one in another situation. If the idea of developing a metric which tracks the creditworthiness of individual actions seems like an inordinately ambitious undertaking, a moral metric for character seems, on the face of it, to be an impossible thing. A thick, rich aretaic description of a person’s character and its development and drift over time does not lend itself to determinate summative judgement. And yet we do make summative judgements of the moral quality of a person’s character. Vespasian possessed a better overall character than Nero; John Curtain had a better character than Josef Stalin. We also make partial, overlapping judgements. Biographers and historians struggle with the task of summing up Winston Churchill’s character. Compare his character to that of Neville Chamberlain. It would be foolish to offer a determinate comparative


Damian Cox

judgement of them. In some respects, Chamberlain possessed a better character than Churchill. In other respects, he did not. There may well be no fact of the matter as to which of these respects matters more from a moral point of view. (As a practical matter, Churchill’s character suited him well to some circumstances; Chamberlain’s character suited him well to others. Chamberlain’s tragedy is that he encountered the wrong circumstance.) Summative moral judgement of character is thus a nuanced, uncertain, and partial affair. Perhaps this just reflects the facts of our moral lives. The moral quality of a life is not only about character, it is also about aspiration, achievement, and purpose; that is, it depends upon morally significant projects. Morally significant projects are life projects that have significant moral aspects or implications. Gauguin’s choice to abandon his family and travel to Tahiti to paint set him off on a life project. On the face of it, it is about painting; but, of course, it has serious moral aspects. The demarcation between moral life projects and non-moral ones is hard to draw precisely and generally. A tone-deaf person, someone without a shred of musical talent, dedicates herself to becoming a violinist. This is not a project focused on moral considerations, but, then, somebody has to listen to her. She has dedicated her life to doing something that, at best, will make others uncomfortable, at worst, cause them auditory pain. Does this make the project morally significant? Yes, it does, to a degree. What does summative moral judgement of a life project look like? We judge projects by their merit, their success, and the merit of their success. A project’s merit is a function of the value of its goal, the reasonableness of it as an ambition, the cost, including opportunity cost, of its pursuit. Other, more richly aretaic, judgements can be made of projects: their nobility, their generosity, their considerateness, their good sense; or their narcissistic self-indulgence, their manipulativeness, their impracticality, their cowardliness, and so on. Success of good projects is good; but the merit of success and the demerit of failure depend on familiar factors such as difficulty, risk, and cost. And luck. Bernard Williams (1981, 35–36) distinguished two kinds of project failure: internal failure and external failure. Internal failure is the deeper failure: it is a failure within the very idea of a project. Gauguin, to use one of Williams’ key examples, might have fallen ill as he sailed south to Tahiti. In this case, his abandonment of his family would have been to no avail; his project a failure. Alternatively, he could have arrived in Tahiti to discover that he has radically misjudged his talent. His work in Tahiti turns out to be trite and he soon abandons it. The former is an external failure and—unlucky as it is—would not lead to the sort of harsh judgement that would have been in store for him given the latter, internal, failure of his project. All in all, moral judgement of a life project will issue in a rich evaluative description of the project, its pursuit, and its results. It won’t be scored. We may be able to compare moral assessment of life projects in

On the Value of Moral Failure


a rough and ready way, as we compare character, but in adopting the narrative approach to summative moral judgement we abandon the idea of a score or a rank. Lives are too complex and judgement of them too unsettled to issue in such a thing.

4. The Value of Moral Failure Why prefer a narrative model of summative moral judgement to a forensic model? My discussion of the moral wanton introduced one consideration in favour of the narrative model: unlike the narrative model, a forensic approach to summative moral judgement fails to do justice to the significance of the integrity of a life: its shape, its consistency, its overarching purpose. I think this is right, but someone attracted to the forensic model is owed further argument. In the final section of the chapter, I offer an argument based on the value of moral failure and the merit of redemptive narratives. Moral failure can be a pervasive feature of even morally extraordinary lives. Cheshire Calhoun (2016) argues that moral failure is an inevitable feature of some ambitious lives. In particular, she argues that moral revolutionaries—those who struggle to transform our moral understanding—inevitably encounter moral failure. A necessary moral project for us all is to make ourselves legible and reasonable to others and moral revolutionaries inevitably fail in this task. They cannot but fail to make themselves fully understandable to the conservative minds they seek to affect; they cannot but appear unreasonable to those whose perspectives they seek to transform. They fail morally in this respect, at least initially and perhaps indefinitely. Lisa Tessman (2015) argues that moral failure is inevitable because of the conflict between non-negotiable obligations and what is possible for us. I agree with both Calhoun and Tessman, but the point I wish to emphasise is that moral failure is often a precondition of moral success; we may have reason to be thankful for some of our moral failures. Moral failure can be a precondition of moral success in at least two ways: by being a precondition of moral development and by being a precondition of a significant moral action or project. Consider the cases in turn. It seems that moral failure often functions as a precondition of moral self-improvement. In an ordinary life, given ordinary powers of moral discernment and ordinary qualities of moral character, moral failure is probably more than just one optional route to moral development amongst others. It may well be a precondition of this development. A person who never experiences significant moral failure—who, say, never gratuitously harms another and then witnesses the effect upon them—may be morally enfeebled as a result. Reading a treatise about gratuitous harm will likely not be enough to secure the appropriate moral development. Without failure, we fail to learn: or fail to learn enough, or fail to learn in the right way.


Damian Cox

Moral failure can also be an opportunistic precondition of moral success. If Oskar Schindler had not joined the Nazi party in 1939, if he had not been a committed and sincere Nazi both before the war and in its early stages, he would not have been in a position to save the lives of over a thousand Jewish people working in his enamelware factory. He would not have had the opportunity to save them. He would have had insufficient prestige within the party: his bribes would not have reached their targets; his bluffs would not have worked; he could never have convinced Amon Göth to move his factory to the Sudetenland; he would not have been in charge of a factory exploiting the labour of Jewish prisoners in the first place (Crowe 2004). It might seem that the relationship between Schindler’s moral failure and his moral success is a contingent one. We can imagine a situation in which Schindler had an opportunity to save 1200 lives that didn’t involve him joining the Nazi party and embarking on a career as a war profiteer. But the relevant judgement we make of Schindler is counterfactual. Moral failure is a precondition for moral success in Schindler’s case because of the truth of the following counterfactual: had, counter to the facts, Schindler not joined the Nazi Party and embarked on a career as a war profiteer, then he would not have saved 1200 Jewish lives in Nazi occupied Poland. His moral success counterfactually depends upon his moral failure. And his moral failure was great. A forensic approach to summative moral judgement would represent the value of moral failure in terms of net gains. The failures we learn from are cashed out in terms of future point gains. Schindler loses a number of points by joining the Nazi Party (probably a lot), but this furnishes him the opportunity—which he eventually takes—of gaining significantly more points down the track. Michael Schwartz and Debra Comer (2013) illustrate something very much like the forensic approach to Schindler by comparing him with another rescuer of Jews from the Nazis: Magda Trocme. Trocme sheltered thousands of Jews during the Nazi occupation of the French village Le Chambon. Trocme does not have Schindler’s backstory of moral failure. She did not join the Nazi party; she didn’t need to in order to accomplish her magnificent work. Lawrence Blum (1994, 74) concludes that “Magda Trocme does seem a more admirable character and a purer case of moral excellence than does Schindler.” Schwartz and Comer (2013, 157) counter with a more nuanced comparative judgement. A proper comparison of Oskar Schindler and Magda Trocme needs to take into account the nature of the situations they faced. An evaluation that regards the specific hazards and risks each experienced might recognise the former not only as a moral hero but also as more morally heroic than the morally saintly Magda Trocme. Magda Trocme sheltered Jews in an isolated region of France, a country in which the majority of French Jews survived the Holocaust. Oskar Schindler rescued Jews in Poland—in one case from Auschwitz itself. The Nazis

On the Value of Moral Failure


murdered nearly all of Poland’s Jews. Moreover, of the few Jews in Poland who survived and returned to their homes after the war, some were murdered by anti-Semitic Poles. Others, including children, died in pogroms. Although Magda Trocme’s world was also one of constant peril, Schindler had to cope with more and graver dangers. Schwartz and Comer offer a more convincing forensic analysis than Blum, but what each of them miss in their judgement of the two exemplars is the redemptive nature of Schindler’s story. From a narrative point of view, Schindler’s story has the structure of a redemptive narrative and Trocme’s story does not. A redemptive narrative is one in which moral failure is redeemed by future success. The connection and temporal relation between failure and success is fundamental to it. By contrast, the forensic model pays no attention to the temporal ordering. Comparison between Schindler and Trocme is invidious and misleading. So, consider a (fictional) case of a heroic surgeon who saved 1200 lives before the war in extraordinarily difficult and dangerous circumstances and then became a committed Nazi during the war. Forensic judgement considers cases equivalent if equivalent points are lost and gained. In Schindler’s case, points are lost and then gained. In the surgeon’s case, points are gained and then lost. If the total points are the same, the order of their arrival makes no difference. From the perspective of narrative, however, the cases are anything but equivalent. Schindler’s is a story of redemption and the fictional surgeon’s story is one of moral decline or degradation. Do redemptive stories have greater merit than stories of moral degradation, all else being equal? On the narrative model, the answer is yes. Redemptive stories are stories of successful transformative projects. The project has merit. Degradation stories are stories of overarching moral failure; they are failures of project. Intuition clearly favours the relative value of redemption over degradation and this is one clear reason to prefer the narrative conception of summative moral judgement to the forensic conception.

5. Conclusion Summative moral judgement is an odd thing. The idea that there is a summary judgement to be made of the moral quality of an entire life, and that this judgement can be both accurate and just, is one that is deeply entrenched in various religious traditions and is quietly assumed in much of the secular tradition of moral thought. Nonetheless, it is not an obviously coherent idea. In this chapter I have attempted to spell out what is involved in the very idea of summative moral judgement and I have done this by contrasting two conceptions of the task: the forensic and the narrative. The forensic conception has the advantage (if it is an advantage) of scaffolding interpersonal rankings. Each life can, in


Damian Cox

principle, be evaluated deed by deed, yielding a commensurable standing. In The Good Place, this is an actual number. We can thus be ranked according to our accumulated moral credits. By comparison, the narrative approach offers life stories whose summative judgement consists of a rich evaluative description. Precise ranking is all but inconceivable given the narrative material that these descriptions range over. This isn’t to say that comparative summative judgements cannot be made on the narrative model. Obvious comparisons abound. For the rest, we must be content with partial comparisons: good in one way, not so good in another, with no obvious way to prioritise them. Since the narrative model is less impressive in the kinds of determinative judgement it allows, why prefer it to the forensic model? I argue that it is preferable because it captures the sense of what is structurally important in a life: the integrity and purposefulness of a life; the value of redemption, and the disappointment of degradation. (Also, I admit, the idea of a moral points system is more than a little creepy.)

References Baier, K. 1970. “Moral Value and Moral Worth.” The Monist 54(1): 18–30. Blum, L. 1994. Moral Perception and Particularity. Cambridge: Cambridge University Press. Calhoun, C. 2016. Moral Aims: Essays on the Importance of Getting in Right and Practicing Morality with Others. Oxford, UK: Oxford University Press. Crowe, D. 2004. Oskar Schindler: The Untold Account of His Life, Wartime Activities, and the True Story behind the List. Cambridge, MA: Westview Press. Frankfurt, H. 1971. “Freedom of the Will and the Concept of a Person.” The Journal of Philosophy 68(1): 5–20. Schur, M. 2016. The Good Place, Season 1. NBC. Schwartz, M. and D. Comer. 2013. “The Difficulty of Being a Moral Exemplar When a Moral Exemplar Is Needed Most: The Case of Oskar Schindler.” In M. Schwartz and H. Harris, eds., Moral Saints and Moral Exemplars, Research in Ethical Issues in Organisations, vol. 10. Bingley: Emerald Group Publishing, pp. 153–168. Smilansky, S. 1997. “Moral Accountancy and Moral Worth.” Metaphilosophy 28(1/2): 123–134. Tessman, L. 2015. Moral Failure: On the Impossible Demands of Morality. Oxford, UK: Oxford University Press. Williams, B. 1981. Moral Luck: Philosophical Papers 1973–1980. Cambridge, UK: Cambridge University Press.

11 Who Wrote Nietzsche’s Autobiography? Elijah Millgram

In contemporary discussions of practical reasoning and moral theory, unity of agency receives, at any rate within the analytic tradition, a worshipful deference that ought to be uncharacteristic of philosophy. For instance, it is argued that whenever there is a choice to be made, a unified agent must be on hand, and so arguments about practical reasoning may treat unity of agency as a starting point; or again, a properly assembled agent is unified, and so unity is something that you can demand of an agent in something like the way you can demand of a house that it have no holes in the roof; or yet again, freedom of the will has been analysed as a form of practical unity, making agential unity an indispensable precondition for what is thought to be an indispensable good. And these are gestures at just a few of the more prominent discussions in which unified agency figures these days.1 But we can find in Nietzsche’s writings a line of criticism that is strong enough to motivate rethinking this cluster of received views about the unified agent, or so I would like to persuade you. I will commence by reconstructing a Nietzschean anthropic argument for his version of unity of agency.2 I call the argument “Nietzschean” as opposed to “Nietzsche’s” for more than one reason; since we will have to get to the deeper one only in due course, here is a stopgap: his writings present Nietzsche himself as a decisive objection to the Nietzschean anthropic argument. After presenting the problem for the argument that Nietzsche saw his own psychology to constitute, I will consider what could serve a disunified agent as a surrogate for practical reasons. I will conclude by suggesting that Nietzsche’s surrogate reasons can be seen as an objection to the broadly shared preference for unified over disunified agency.

1. Will to Power and an Anthropic Argument Let’s begin with Nietzsche’s portrait of agential unity and psychological health. To reconstruct it, and the Nietzschean anthropic argument that we will use to trace its contours, I will need first to gloss two related senses of the expression “will to power.”3 Without now doing textual


Elijah Millgram

footwork to substantiate my explication, I suggest that we think about will to power as follows. Control is a concept whose literal use does not nowadays require persons on either end of the relation; let me mention a couple of its more obvious features. (N.B.: I’m not going to try to list necessary and sufficient conditions for its application.) First, control provides a basis for explanation of facts about what is subject to control. For instance, if I control the temperature of the room, you can explain the temperature’s being such-and-such, and not otherwise, by adducing my control. Control involves the ability to effect change, and entails at least that there are true contrary-to-fact conditionals about such change being effected. Second, control implies monitoring, and monitoring, in all but the simplest feedback loops, requires memory. If you exercise control over someone by giving him orders, you have to be able to check on whether your orders have been carried out, and to do that, you have to remember what you told him to do.4 When you look around the world, you often see what it’s convenient to describe as patterns of control. I will take as my example the administrative structure of a university, because it is philosophically unmysterious and it will be familiar to many readers. In a university, the office of the president exercises control over the different colleges and schools; the dean of a college exerts control over its departments, centres, and programmes; a department’s director of undergraduate studies partially controls what courses of study the undergraduate majors adopt. Sometimes these patterns are relatively stable (as is often the case in an institution like a university), and sometimes they are not (as is also often the case with universities). When the pattern of control is reasonably extensive and relatively persistent, we can call it an organisation. Sometimes organisations are self-maintaining and self-perpetuating. A university maintains itself by pursuing and appropriating resources: soliciting money from donors and funding agencies, recruiting students, faculty, and so on. It also checks on whether its employees and subsidiary administrative units are executing the policies of the university, monitoring the performance of units within the university by conducting reviews. When an organisation is in this way self-maintaining, we will say that it is homeostatic. Sometimes such an organisation does not merely maintain itself as it is, but rather grows; here also universities provide familiar examples. Homeostatic and expanding organisations are a form taken by will to power that is of great interest to Nietzsche; the class of such centres of will to power would include, on his view, not just universities, states, companies, NGOs, and the like, but individual human beings and (at least some of) their subpersonal parts, as well as other organisms. Sometimes organisations are governed by priorities; that is, these priorities determine their choices. Nietzsche himself talks about “values,”

Who Wrote Nietzsche’s Autobiography?


and, for reasons we’ll come to, I mean “priority” to be somewhat thinner than “value”: “priority,” understood as a functional-role concept, is exhausted by its role in determining choice. For instance, the priorities of a dean’s office might include controlling the budget by reducing overall FTE—a priority, because it could serve as input to a decision, say, to attrit the faculty by 10%. Call such centres of will to power priority-guided. Now suppose that you are a priority-guided organisation, and you stop and ask what your priorities are. Just as this may be something we cannot find out by asking you, it may be something you cannot find out by asking yourself. There is, so far, no reason to suppose that you yourself know what your priorities really are: when we introduced the notion of a priority-guided centre of will to power, we said that the priorities direct the choices, but we did not say how they do that, and in particular we did not say that awareness of the priorities is a necessary part of the pathway. Let me use “will to power,” now as the name of a priority rather than for generic patterns of control, to mean this: the importance of maintaining and extending the very priority-guided pattern of control of which it is a priority. Then the first conclusion of the argument we are assembling is that the will to power is your highest priority. That is a dramatic and surprising claim, because on the face of it, different people and more generally different organisations differ in what matters to them and what they pursue, and their own will to power is not often advertised as the most important priority. University mission statements, to stick with our example, typically emphasise teaching, research, and service to the community, and while there is a good deal of variation, you will have to look far and wide to find a university (or a competent university administrator) proclaiming its most important mission to be: getting bigger. But now suppose that your top priority is not your own will to power. Then some other priority is—or perhaps your priorities fluctuate in such a way that we would be disinclined to say that you have a top priority at all. A pattern of control complex enough to be guided by priorities is a very complicated and consequently easily disrupted thing, and its maintenance is highly demanding. (This means that if you are a priorityguided organisation, you must also be homeostatic or expanding.) Over the relatively lengthy period you have existed, situations almost inevitably will have arisen in which your priorities were in competition with one another and, in particular, in which the will-to-power priority dictated one decision, and your hypothetical alternative priority dictated another. But since what will to power dictates is just maintaining and extending the pattern of control that constitutes yourself, and since by hypothesis will to power is not your highest priority, you must have been making decisions that disrupted and eroded that pattern of control. The cumulative effect of such decisions must have been the disintegration of the pattern of control that constitutes yourself. But in that case, you cannot now be asking what your priorities are. If you are now well


Elijah Millgram

enough organised to ask yourself that question, it will turn out (your most sincere protests notwithstanding) that your top priority is in fact will to power. Nietzsche famously emphasises anthropic bias for a special case: “the value of life cannot be estimated. Not by the living, for they are an interested party . . . not by the dead, for a different reason” (6:68/TI 2.2). But the point is more general, as becomes clear when we return to our example. Sometimes universities have to choose between opportunities for expansion and their official self-conception. America’s Ivy Leagues, in the early part of the twentieth century, understood themselves roughly as finishing schools for the offspring of the WASP elite. During and after the Second World War, enrolment growth and federal research money gave these universities an opportunity to expand rapidly and faced them with just such a choice. The Ivy League schools responded by changing their self-conceptions so as to allow them to exploit the new opportunities for growth, and, to make a long story short, you can be assured that if a university is here, healthy, and flourishing, then it has a well-entrenched corporate culture which makes such choices in only one way.5 There is a way of seeing a university as really just being a selfperpetuating pattern of control of the kind we have been discussing, and Nietzsche experiments with the idea that you could see everything this way—even what we think of as inanimate physical objects. (This last is a suggestion that has not gotten much uptake, probably for good reason.) I want to indicate how the argument we have just given can be pushed one step further in order to solve a persistent problem in Nietzsche exegesis: that on the one hand Nietzsche seems to endorse a metaphysics of will to power, while on the other he presents himself as a thoroughgoing perspectivist. There are indefinitely many ways to parse the world, and a perspective that renders its contents as patterns of control is no more than one of them. Perspectives are (among other things) ways of simplifying what one takes the world to be like in a way that makes representation cognitively tractable.6 Assume that you will be able to address your priorities more effectively if you see your world in terms of those priorities: someone who is trying to add to his butterfly collection will do better at it if the concept butterfly is part of his conceptual apparatus, if he is able to recognise butterflies, and if what he notices, when there are butterflies present, are the butterflies, rather than anything else. Then, given that your highest priority is will to power, you will do better to have a will-to-power rendering of your world. Such a rendering might be a will-to-power metaphysics (as in: “This world is will to power—and nothing besides!” [11:611/WP 1067]), or it might be somewhat less dramatic than that: A dean will do better at growing his piece of the university if he sees departments as repositories of lines, alumni as potential donors, changes in academic fashion as opportunities for starting up new programs . . . and all of these as features of or ways to change what I understand is now called the org

Who Wrote Nietzsche’s Autobiography?


chart.7 To be sure, Nietzsche does not think that the assumption I just asked you for is always, or even nearly always, true. Christian ascetics, he holds, have will to power as their highest priority, but they cannot afford to be aware of this, and they cannot afford to see the world from the will-to-power perspective. But when it is true, the will-to-power perspective will generally be, while not more than a perspective, nonetheless the perspective of choice.8 To reiterate, I want to attribute the argument I have just outlined to Nietzsche in only the thinnest sense. It is assembled out of Nietzschean materials, and I am fairly certain that the argument occurred to him. But I am about to claim that he did not endorse it, because he himself had to live through one of its biggest bugs.9

2. Forgetfulness on Display The alert reader will have had a complaint pending against the Nietzschean anthropic argument for some time now. Suppose it is true that you would not have come to be the complex and delicate priority-guided organisation that you are if (some variant of) will to power had not been your top priority. That does not show that will to power is still your top priority, because priorities shift. While the argument provides a reason— and let’s just allow for now that it’s a convincing reason—to think that a shift away from that priority will eventually prove your undoing, “eventually” can be a long way away. A while back, Yale started making expensive decisions that put its version of intellectual integrity ahead of will to power. But while Yale also went through its share of financial difficulties, it stayed with us; it takes a while to burn through that much accumulated institutional capital. (I understand that Yale is once again aggressively pursuing expansion-oriented policies, and, as far as prioritising will to power goes, is back on track.) So the argument does not show that your top priority is will to power, because you could be someone who (or some organisation which) has abandoned that priority but has not yet fallen apart. Now there are many ways to fail to have will to power as your top priority: you could have a different top priority; or you could have various priorities, none of which was ranked first; or you could fall short of having anything that amounted to priorities at all. The label for the way Nietzsche understands himself not to have will to power as a dominant priority is decadence, and to introduce the notion I will quote a muchquoted passage yet once more: What is the sign of every literary decadence? That life no longer dwells in the whole. The word becomes sovereign and leaps out of the sentence, the sentence reaches out and obscures the meaning of the page, and the page gains life at the expense of the whole—the


Elijah Millgram whole is no longer a whole. But this is the simile of every style of decadence: every time, the anarchy of atoms, disgregation of the will. (6:27/CW 7)10

The notion of a priority was introduced in terms of its guiding function within the organisation; priorities control the organisation’s choices, which entails control of the organisation’s constituent parts. The decadent is someone whose parts (which Nietzsche thinks of as drives) are not effectively controlled; so the decadent is someone who has no priorities because the control structures that priorities presuppose are no longer in place.11 Since I mean to treat Nietzsche himself as the most important problem facing the Nietzschean anthropic argument, I propose to use Ecce Homo as our point of entry into his complicated and difficult texts. As its title suggests, it is one of Nietzsche’s exercises in self-presentation, and his riff on the genre of autobiography; he uses it as a retrospective frame for his other published works.12 What Ecce Homo does is not just tell you about Nietzsche, but put him on display. And that means that what it says cannot simply be taken at face value; we will have to look at what it is showing us as well.13 Now Nietzsche does not do anything as straightforward as tell us that he is a decadent. On the contrary, we are given a flurry of pronouncements on the subject which include a number of direct and indirect denials. Let’s make a start on a pretty characteristic declaration: “Apart from the fact that I am a decadent, I am also the opposite.” (He glosses the remark almost immediately: “As summa summarum, I was healthy; as an angle, as a specialty, I was a decadent” [6:266/EH 1.2].)14 We need to stop and think about how to take assertions of this sort. After all, Nietzsche could be giving us a clear-headed and nuanced assessment of his state; but we would also expect a personality descending into chaos to give us mixed signals which would sound very much like this. Nietzsche follows this claim with a brief argument that he is the opposite of a decadent; one of his reasons is that he knows how to forget. Given his diagnosis of ressentiment as a type of sickness, that might mean merely that he does not hold a grudge. But it also might mean a good deal more than that: the second essay of the Genealogy describes how proto-human animals became full-fledgedly human by coming to have a memory. So the question is what function forgetting has here, and conveniently enough, Nietzsche repeatedly shows us. Let’s pick a representative passage; I apologise in advance for its length, but a stretch of text that will allow its author to exhibit an ability to forget will have to be longer than the standard-size quote.15 At this point, a large reflection is necessary. One will ask me why on earth I’ve been relating all these small things which are generally considered matters of complete indifference: I only harm myself,

Who Wrote Nietzsche’s Autobiography?


the more so if I am destined to represent great tasks. Answer: these small things—nutrition, place, climate, recreation, the whole casuistry of selfishness—are inconceivably more important than everything one has taken to be important so far. Precisely here one must begin to relearn. What mankind has so far considered seriously have not even been realities but mere imaginings—more strictly speaking, lies prompted by the bad instincts of sick natures that were harmful in the most profound sense—all these concepts, “God,” “soul,” “virtue,” “sin,” “beyond,” “truth,” “eternal life.”—But the greatness of human nature, its “divinity,” was sought in them. . . . All the problems of politics, of social organisation, and of education have been falsified through and through because one mistook the most harmful men for great men—because one learned to despise “little” things, which means the basic concerns of life itself. . . . Our present culture is ambiguous in the highest degree. . . . The German emperor making a pact with the pope, as if the pope were not the representative of deadly hostility to life! . . . What is being built today will no longer stand in three years.—When I measure myself against my ability, not to speak of what will come after me, a collapse, a construction without equal, then I more than any other mortal have a claim to the epithet of greatness. When I compare myself with the men who have so far been honoured as the first, the difference is palpable. I do not even count these so called “first” men among men in general,—for me they are the refuse of humanity, monsters of sickness and vengeful instincts: they are inhuman, disastrous, at bottom incurable, and revenge themselves on life. . . . I want to be their opposite: it is my privilege to have the subtlest sensitivity for all signs of healthy instincts. There is no pathological trait in me; even in periods of severe sickness I never became pathological; in vain would one seek for a trait of fanaticism in my character. There is not a moment in my life to which one could point to convict me of a presumptuous and pathetic posture. (6:295f/EH 2:10) To recap the progress of this stretch of Nietzsche’s prose: He starts out in a low key, evenly modulated tone, explaining that his longish discussion of the little things in life has had a philosophical point; those little things are what matters, not the delusory ideas that we associate with religion, or with what Peter Viereck once called metapolitics (1965). Then, with the most minimal transition, we get a brief outburst of Stammtisch metapolitics, in a register that is best described as fanatic raving. Then there is a moment of posturing—Nietzsche comparing himself to other “frst men”—that is both presumptuous and a bit pathetic. And then he tells us, almost immediately, that he never exhibits fanaticism, presumptuousness, or a pathetic posture. No wonder Elizabeth Förster-Nietzsche felt that she had to excise the middle of this block of text.


Elijah Millgram

Control, we pointed out, requires memory; here Nietzsche’s capacity for forgetting is exercising itself moment-to-moment, and functioning to paper over blatant inconsistency (both at the level of opinion, and at the level of tone of voice and character). There are many other passages that behave similarly, and one of the puzzles posed by Nietzsche’s writing is that they are so rarely noticed—that he induces the same moment-tomoment lapses of memory in most of his readers.16 Why is inconsistency being papered over, rather than rectified? By the time we see the longish passage I have just quoted, Nietzsche has given us an answer: his discussion of climate, reading, and eating is organised by the thought that one has to husband a scarce resource: the energy one spends in coping with an unsuitable climate, heavy food, and so on. And there is a further type of activity that is just as much, or even more of a drain on the resource in such short supply: an instinct of self-defense . . . commands us . . . to say No as rarely as possible . . . The reason in this is that when defensive expenditures, be they ever so small, become the rule and a habit, they entail an extraordinary and entirely superfluous impoverishment. Our great expenses are composed of the most frequent small ones. (6:291f/EH 2.8) The exercise of control, the effort of maintaining a pattern of control involving consistency of character and of doctrine, is to a great extent the effort of saying No to elements that won’t ft into a consistent pattern. And that effort is simply too great for Nietzsche (who has just fnished telling us that he can handle only one cup of tea in the morning) to make.17 At the end of the section from which we extracted our illustration of forgetfulness at work, Nietzsche tells us that his “formula for greatness in a human being is amor fati [love of fate or destiny]: that one wants nothing to be different” (6:297/EH 2.10). If we take it that this is what we have just had explained and displayed to us, we can now see that we need to distinguish two very different evaluative foci of Nietzsche’s thinking. The will to power, which we have already seen, is a priority, by which I mean to say that it guides action and choice, requiring an agent to take steps to maintain and extend the pattern of control that he is, by, in fact, exercising control: by forcing things to fit the pattern, which is to say, by not accepting them as they are. Amor fati, accepting things as they are, is a mode of acquiescence, and so is precisely not part of the exercise of control, which means that it is not in this sense a priority at all, and not compatible with the demands of will to power.18 A number of influential readings allow the ambiguity present in Nietzsche’s notion of “interpretation” to mask the tension between will to power and amor fati.19 But this is a mistake. One way of doing

Who Wrote Nietzsche’s Autobiography?


interpretation—call it active interpretation—is an exercise of will to power; recall his gesture at barbarian conquerors imposing social structure on a vanquished mass of humanity (5:325/GM 2:17). The other— call it passive interpretation—is precisely not such an exercise. Actual interpretations are inevitably mixtures of active and passive, but the difference between them is real, and of great practical importance. When an administrator announces a “Faster, Cheaper, Better” initiative (the slogan adopted by a former NASA head), this could have the effect of changing what the departments, centres, and so on actually do—or it could merely mean that the bureaucrats start describing whatever they were doing already as “Faster, Cheaper, Better,” while changing nothing. Decadence is the state in which the exercise of control over one’s constitution is no longer possible, and amor fati is an evaluative attitude that belongs to decadence: “Accepting oneself as if fated, not wishing oneself ‘different’—that is in such cases [of “Russian fatalism,” the desperate exhaustion of a soldier who finally just lies down in the snow] great reason itself” (6:272f/EH 1.6). Nietzsche apparently is a decadent, and the confusing pronouncements are the way he papers over what as a decadent he is no longer able to render unconfused. The Nietzschean anthropic argument claims that if a priority-guided organisation asks what its top priority is, it will turn out that the organisation is guided by will to power. Nietzsche is not exactly a counterexample to that claim (although we have seen in passing how to come up with what would be an exact counterexample). Rather, he illustrates a related objection: that you might turn out not to be addressed by the argument, because you could be asking the question even if you were not guided by priorities at all. Instead of priorities, Nietzsche has (and let’s introduce his own term at this point) a “value,” amor fati. (Warning: don’t let the sound of Nietzsche’s word, “value,” tempt you to import your understanding of the word in other popular or philosophical contexts. This is a technical term, one which we’ll be able to explain shortly.) But since amor fati does not control behaviour in the way that a priority does, what is it doing there? We evidently need to investigate the function of such an attitude in the decadent personality.20

3. Reasons for Disunified Agents Why go on? Why not commit suicide? Why not just let go, let your grip relax, and allow yourself to slide into a slacker’s life? Questions like these have not received a great deal of attention from philosophers in the analytic tradition, but a recent discussion exhibits the approach that comes to it most naturally. Christine Korsgaard has argued that action and agency are one’s plight; you act because you need to pull yourself together and do one thing or another.21 Normally I am committed to ongoing courses of action in whose service I must mobilise myself,


Elijah Millgram

when the occasion arises: maybe I can’t afford to slack off because I’m in the middle of writing an essay on Nietzsche, my cat has to be fed, the laundry has to get done, the exams graded, and vegan chicken soup brought over to a laid-up friend—although Korsgaard herself argues that one’s commitment to one’s ongoing agency does not depend merely on this sort of contingent commitment to particular objectives.22 Unity of agency is consequently reunity of agency. In something like the way that an animal remains alive in virtue of activities like eating and sleeping, activities which reconstitute it, from day to day, as a living creature, synchronic unity of agency is a product of—is the advancing wavefront of—diachronic unity of agency. That is, when a representative and mainstream analytic philosopher tries to explain your stake in your unified agency, the answers she finds it most natural to offer to questions like “Why go on?” assume that you are already an agent. Now, if you are an agent, the answer to this question will be a reason— a practical reason, since this is what agents consume in the course of making decisions, and so with content of a sort that you could express as a sentence.23 And whatever the merits of Korsgaard’s insistence that your stake in your own agency is not contingent, if the Nietzschean anthropic argument is on the mark, someone who has his act together will not find the question particularly urgent: remember, whatever the actual answer, your top priority already is going on. (And this seems to be the grain of truth in what strikes her readers as the wildly implausible claim that an agent cannot choose its own disintegration and collapse.) For the decadent, however, “Why go on?” is an urgent question: he is, after all, in the process of falling apart.24 A useful answer to this question cannot be a traditional practical reason, that is, a reason for a decision addressed to the agent, for the decadent does not have a centralised command post, as it were, that makes effective decisions. The decadent does not exercise much by way of control over his parts, and so an effective answer to the question, one that will keep the components flying in more or less their former configuration, will have to be addressed to the parts severally. If the parts of a decadent are themselves person-like agents—as they would be, say, in a university bureaucracy—then the answer (or answers, because we should not assume that the same answer will do for each component) might still take the form of a traditional practical reason. But when the parts are not themselves agents, as the parts of human beings ordinarily are not, then a suitable answer must take a form that appropriate parts can consume. Since Nietzsche thinks of himself as made up primarily of drives, we should not assume that the answer he seeks is necessarily to be identified with a well-defined sentence-like content at all. Rather, it will be a stimulus to which the drives respond. Nietzsche’s interest and background in music need to be emphasised at this point. An amateur composer, he published several books on primarily musical subjects, and his other works contain much discussion

Who Wrote Nietzsche’s Autobiography?


of musical matters.25 So it is not surprising that, time and again, we find Nietzsche thinking his way through some problem or other using musical analogies, and now that we have seen that what one offers to the drives by way of a response to the question, “Why go on?” might as well be music, we should ask what kind of music would do. Early engagements with a philosophical problem can be telling precursors to more mature attempts on it, and so a good place to take this query would be Nietzsche’s first book, where he argued that tragedy functioned as the ancient Greek answer to “Why go on?”26 Accordingly, in thinking about available responses to this question, we should expect to find him using tragedy as a model, and in fact, towards the end of the book, Nietzsche considered how one would achieve, in music, the experience produced by tragedy, that “of having to see at the same time that [one] also long[s] to transcend all seeing”: Quite generally, only music, placed beside the world, can give us an idea of what is meant by the justification of the world as an aesthetic phenomenon. The joy aroused by the tragic myth has the same origin as the joyous sensation of dissonance in music . . . we desire to hear and at the same time long to get beyond all hearing. (1:150,152/BT 24) Nietzsche never abandoned this idea, and dissonance ended up being his way of thinking about one kind of stimulus that would effectively coax his drives forward, once they could no longer be coerced. Shortly after the comparison of tragedy to dissonance, The Birth of Tragedy entertains the notion that we could “imagine dissonance become man.” It might not be immediately obvious that this is something we could imagine, but Ecce Homo unpacks the metaphor, and does its best to show how to do more than just imagine it. The drives—out of which the decadent personality is, to help myself to a nearby pun, composed—are lured onward by the very incoherence that, conveniently enough, is the immediate effect of decadence.27 The exercise of highlighting and contemplating these inconsistencies of tone, character, and doctrine—and of exaggerating them for artistic effect—is a self-conscious contribution that Nietzsche is able to make to his own continued psychological existence. Identifying this strategy confirms our answer to the question of whether Nietzsche himself was really a decadent: it would make no sense to avail oneself of such devices if one were not. As a decadent, he is riddled through and through with inconsistencies of all sorts, inconsistencies which he is too depleted to eradicate. His interest in his own continued existence—if you like, his will to power—expresses itself in the construction and contemplation of a representation of that very decadence, which turns out, rather neatly, to be one of the very things that will keep his drives in some semblance of order.28


Elijah Millgram

Here is the phrase I just quoted, with a bit more of its context: If we could imagine dissonance become man—and what else is man?—this dissonance, to be able to live, would need a splendid illusion that would cover dissonance with a veil of beauty. (1:155/BT 25) Amor fati shares with dissonance the function of “values” in the psychological economy of decadence—that is, to elicit a response from the drives.29 (We can now explain the contrast between the two functional concepts we have been developing. Priorities guide decisions. Values are personality glue.) Dissonance needs to be cut with harmony if it is not just going to be out-and-out discord; the requisite effect, remember, is that one both desires to hear, and not to. Amor fati is the veil of beauty thrown over the dissonance (or anyway, the spoonful of sugar that helps the medicine go down).

4. Whose Portrait Is It? The reader will by this point have another objection: doesn’t the evidence I have adduced for this interpretation in fact undercut it? Allow that Nietzsche is displaying himself as a decadent. Still, his presentation of his own incoherent character is so artful, so masterful, and so controlled as to belie the overt content of the display. A decadent is someone who cannot control the parts that make him up, but Nietzsche’s show of lack of control is so visibly intentional as to amount to precisely the presentation of a character that is in full control. And there is an additional objection that is best considered together with this one. On Nehamas’ reading, the Nietzsche we are being presented with is not the “writer,” the lonely man desperately scribbling away in Swiss and Italian pensions, but the “postulated author”: more or less, a character projected by the body of texts, on a par with the narrators of works of fiction. As a matter of method, to read a literary text is to reconstruct the text’s postulated author, which means in turn to reconstruct that author’s intention in producing such a text (Nehamas 1981, 1987).30 The exercise of reconstruction requires that the implicit author, Nietzsche in this case, be practically coherent and consistent; making the text out to be the product of a coherent intention is delineating a unified agent coordinate with that intention. So to take the Nietzsche presented by his texts to be a decadent, that is, a disunified agent, is on Nehamas’ view a methodological blunder. However, Nietzsche distinguishes the “driving force” of an action from its “directing force”: as when, on a large ship, the engine and current propel the vessel forward, while the helmsman can make only small adjustments in direction (and sometimes, Nietzsche suggests, none at all; 3:607f/GS 360). On the Nietzschean model of the workings of the mind,

Who Wrote Nietzsche’s Autobiography?


a decadent’s drives may express themselves in bursts of activity, whose occurrence or even general shape the agent as a whole cannot control. The agent’s purposes or goals, which Nietzsche analogises to the helmsman, can however make minor adjustments to activities like these, and here it is worth remembering that writing, for Nietzsche, was at least a two-pass activity: his published works are by and large lightly revised selections of raw material produced as journal entries.31 If I am reading this passage correctly, it tells us how writing produced by a decadent could come to exhibit the distinctive artifice and booby-trapping that we find in Nietzsche’s mature works (I mean, the sort of textual construction we saw in the lengthy excerpt from Ecce Homo).32 The appearance of control is managed by layering minor adjustments onto what amounts to an extended blurt, one emitted with hardly any control at all. Evidently, and overriding that first objection, one does not have to be master of oneself to write masterfully, and one does not have to be in control of oneself to write in a controlled style. Nietzsche concludes the passage we have just introduced with the pronouncement that “we still need a critique of the concept of ‘purpose’”— which in context amounts to a rejection of the widely shared and roughly Anscombian model of intentional action, on which actions are constructed by selecting subsidiary or component actions on the basis of their suitability for effecting a designated end.33 Nehamas’ postulated author is the agent projected by understanding a text as an Anscombian intentional action. Thus Nehamas’ exegetical methodology takes for granted a theory of action that Nietzsche is criticising. But if Nietzsche’s philosophical views are interesting and important enough to justify the effort of interpreting him, we should not proceed by interpreting him in a way that presupposes that those views can be dismissed.34,35 Let’s proceed, then, to take the Nietzsche we are discussing actually to be the living person that the portrait is of, rather than merely a literary artifact. (Again: the literary-artifact Nietzsche does not need the delicately drawn, self-undermining portrait of an incoherent personality; it is only in light of its living author’s need for it that the elaborate construction makes sense.36) Bear in mind, however, that we do not have to suppose that the representation is faithful. Nietzsche was probably not someone whose mental health could be kept up by doses of dissonance; remember, the therapy didn’t actually work. Although he is not best thought of as built up out of inadequately regimented drives, when he represents himself that way, he is representing himself. Because the commitments involved in belief and in authorial action require mental unity of various sorts, we can now see that if the representation of Nietzsche we have extracted from his texts is anything like substantively correct, attributing opinions and textual strategies to him is more problematic than I let on at the outset. That means that I need to bracket my earlier attributions retroactively, and raises the difficult


Elijah Millgram

question of whether they have been dispensable shorthand or structural elements of an argument that cannot be reconstructed without them.37 The question can only be fully addressed once we have worked our way through a great deal more of Nietzsche’s writing than it will be possible to do here, and in the interim, I want only to register a consequence of the view I am advancing for my own exposition. A reader familiar with contemporary Nietzsche scholarship is likely to have the sense that I am underquoting: the norm today is to support exegetical claims with a great deal of indiscriminately cited chapter and verse.38 If I am right, however, it is normally necessary to determine the role of a passage in its framing text before one can deploy it, and this is normally possible only once one has an account of how that text in particular demands to be read. (A bit of shorthand for this point: in reading Nietzsche, rather as in reading Wittgenstein, you always have to ask who—or what—is speaking.) With occasional exceptions that I think are safe, I am appealing to texts about whose circumstances of utterance I also have something to say.

5. Reconsidering Unified Agency Let us now return to my initial complaint about the use of the unified agent in contemporary philosophy. The new orthodoxy in moral philosophy is that if you are not a unified agent, then, from a practical point of view, you are not really there. (So there is not really anything for an anthropic or transcendental or other action-directing argument to be applied to, when we are considering the decadent.39) Nietzsche’s exercise in autobiography convincingly shows this rejoinder to be callous, practically irrelevant and unbelievable. It is callous in that people who are coming apart at the seams often end up on the street; when passersby step over street people as they enter the subway, they treat the street people as though they didn’t exist; the rejoinder amounts to ideology that underwrites this kind of shoddy behaviour: from the “practical point of view,” the street people really don’t exist.40 But, second, it’s no use to tell Nietzsche that since he isn’t, properly speaking, an agent, he doesn’t, from a practical point of view, exist. That won’t make his very difficult, very immediate, and very personal problems any less pressing for him. Finally, the sheer sustained ingenuity of the Nietzsche corpus makes it impossible simply to dismiss its author in the way this rejoinder would like; if he’s “not there,” who wrote all those books? Nietzsche, the patron saint and poster child of disunity, was not in a position to overlook the fact that disunified agency, disunified apperception, and disunified minds are, for some people, what it is like to be them. Now, a great deal of philosophising treats the unified self as a dialectical starting point: here you are, doing the philosophising, so we can begin the argument with your transcendental unity of apperception, or with your unified agency, or whatever. But Nietzsche couldn’t start with the

Who Wrote Nietzsche’s Autobiography?


assumption that he was a unified agent, because he just wasn’t one. One lesson to learn from our reconstruction of Nietzsche’s predicament, and of his response to it, is that it’s time to rethink the idea that the unity of the self is an Archimedean point in philosophy. Still, Nietzsche pined, loudly and frequently, for the unattainable attractions of unity.41 Will to power successfully asserting itself struck him as health, and more than anyone else he glorified that rendition of agency, with the yearning of those who have never actually tasted the object of their desire, and so with a passion that has inspired his more callow readers ever since. And here he is still in line with the received attitudes; even when it is allowed that we can fall short of unity, it is taken for granted that this is a bad thing for the agent, that the agent is falling short of the standards to which he should be held, and that this is a defect to be remedied as expeditiously as possible. But a second lesson to take away from our discussion is that unity of agency is overhyped. There is something it is like to occupy a perspective (in the sense that Nagel famously pointed out there is something it is like to be a bat), but during our brief encounter with the will-to-power perspective, we didn’t pause to get a sense of what it is like (1991).42 To help remedy that, here is David Wiggins musing over a representation of the perspective in question. Two or three years ago, when I went to see some film at the Academy Cinema, the second feature of the evening was a documentary film about creatures fathoms down on the ocean-bottom. When it was over, I turned to my companion and asked, “What is it about these films that makes one feel so utterly desolate?” Her reply was: “apart from the fact that so much of the film was about sea monsters eating one another, the unnerving thing was that nothing down there ever seemed to rest.” As for play, disinterested curiosity, or merely contemplating, she could have added, these seemed inconceivable. And the thought the film leads to is this. If we can project upon a form of life nothing but the pursuit of life itself, if we find there . . . no interest in the world considered as lasting longer than the animal in question will need the world to last in order to sustain the animal’s own life; then the form of life must be to some considerable extent alien to us. (Wiggins 1991, 102)43 The world as will to power is a disheartening and uninviting vision, and it is nihilistic: all priorities, and no values. It is not a coincidence that there is more to be said for and about the values that Nietzsche invented as psychological prostheses; just because they are meant to coax the drives onward, they are attractive. My own hypothesis is that a number of further famous Nietzschean notions—in particular, the Eternal Return (or Eternal Recurrence) and the Overman—are to be understood as having


Elijah Millgram

been produced to occupy the cognitive role of values. I am not going to argue for that now, but on the supposition that the suggestion is correct, I am going to help myself to a B-movie from a little while back in order to draw one last conclusion for our exegetical practice. The protagonist of Bubba Ho-tep (Coscarelli 2004) is a senior citizen who thinks that he is Elvis Presley. We see him marking time in his old age home, living an undignified, passive, and disintegrated life. When he comes to believe that he’s the only thing standing between the other residents of the old age home and Bubba Ho-tep—the eponymous redneck mummy, returned from the dead to inflict on them a fate too tacky to mention in mixed company—he mobilises himself. Or rather, he is mobilised by the emotional content of his value and transformed from a disunified agent into an agent capable of making effective decisions, and of course he defeats the mummy and saves the old-age home. The point I want to extract from my summary of this plot line is that, since what a value does is stimulate the drives, it is not necessarily required to be intellectually respectable. The idea that a redneck Egyptian mummy has come back from the dead is just plain silly (indeed, it is there precisely to provide the film with a comic dimension), but as long as it holds someone together as an agent, its suitability for that role is not impugned by its silliness. Neither is it impugned by its untruth: the movie happens to suggest that the character may really be who he thinks he is, and in addition that the mummy is real, but none of this matters for whether defeating the mummy can operate as a value for the person we are being shown. Psychological effectiveness as a value is one thing; other intellectual merits are another. Within Nietzsche scholarship, the Overman and the Eternal Return have been consistently treated as philosophical doctrines. Unsurprisingly, interpreters have worked to reconstruct them by producing one after another highly-articulated theoretical rendering, each fine-tuned to meet objections raised to earlier treatments.44 (Unsurprisingly: this is what historians of philosophy are after all trained to do, and the approach does work well enough when the figure in question is, say, Immanuel Kant.) However, a person’s relatively simpler and relatively unintellectual psychic parts are more likely to be successfully addressed by values that are in soft focus. Values have to be, more or less, inspiring; when there’s too much theoretical articulation, they stop being inspiring. If the cognitive function of values is to mobilise the elements of a personality, and if the Overman and the Eternal Return are meant to function as values, then the traditional exegetical approach to these Nietzschean ideas is no doubt a mistake.

6. Decadence and Inventing Values Returning to the problem with which we started, we can see the will-topower agent and the highly fragmented decadent as the two near-extremes

Who Wrote Nietzsche’s Autobiography?


of a spectrum.45 Because we have a place in our lives both for priorities and for values, most of us probably fall somewhere in the middle of it, and I imagine that that is where most of us want to—or ought to want to—remain. Nonetheless, we should not allow ourselves to be dismissive of personalities positioned towards the fragmented end of the spectrum. Recall that Nietzsche proposes that the right response to nihilism is to invent new values. Although this is a task he officially assigns to “philosophers of the future,” I have just been suggesting that the Eternal Return, the Overman, and amor fati are demos; and while they were meant in the first place for Nietzsche’s own consumption, they have proven inspiring to the psyches (and presumably to whatever is the actual counterpart of the Nietzschean drives) of generations of his readers as well. Nietzsche is showing you how it’s done, and he doesn’t do too badly at it. If his own example suggests that inspiring values are best invented by decadents, doesn’t that count, all on its own, as a weighty recommendation for the disunified self?46

Notes 1. Under the first and second headings, Christine Korsgaard has assembled an ambitious argument to the effect that being a unified agent means being a Kantian practical reasoner, with pieces of the argument distributed throughout Korsgaard (2008) and Korsgaard (2009). Frankfurt (1988) develops a wellknown version of the third move, and Bratman (2006) works up similarly motivated analyses of full-fledged action attribution and of self-government. 2. The ‘anthropic’ part of the label is meant to highlight structural similarities to the anthropic arguments deployed in cosmology and elsewhere. See Smart (1987), Barrow and Tipler (1986, 18, 251ff), and Millgram (2009, §4.5), for examples; for framing discussion, see Roush (2003). 3. See, e.g., 13:394/WP 46; 6:181/AC 14. References by volume and page to the Kritische Studienausgabe (1988), and where possible I also give book and section, using the standard North American Nietzsche Society conventions. (So, the first citation previously is from Will to Power, §46, which appears at the Studienausgabe, vol. 13, 394.) Except as indicated, the translations are those of Walter Kaufmann (1954, 1974, 2000) and R. J. Hollingdale (1982, 1996, 1997), or the both of them jointly (1968). 4. Etymologically, the word is derived from a special case of such monitoring: verification by duplicate register (Onions et al. 1966). Clark and Dudrick (2009) notice that what I am calling control is not just a matter of what does happen, but in part of what is supposed to happen: in an older vocabulary, that the notion straddles the distinction between descriptive and prescriptive. 5. Four points need to be registered before proceeding. Firstly, will to power, as a guiding priority, runs together the homeostatic and expansionist attitudes towards the organisation that has it, and it might be thought that the argument only establishes that a priority-guided centre of will to power must prioritise self-maintenance, but not that it must also be trying to expand. However, there are two things to be said in favour of keeping both aspects of will to power together. Organisations of this level of complexity do not come into being full-blown, like Athena from the head of Zeus, but bootstrap themselves from smaller and more modest kernels; if augmenting, rather than


Elijah Millgram

merely maintaining, the scope of the pattern of control were not part of the organisation’s highest priority, it would not have come into existence in the first place. Moreover, it is very unusual that merely attempting to maintain a steady state succeeds in doing even that; normally, trying to grow is necessary even to so much as stay as you are. (It’s interesting to speculate as to why this is so: perhaps, since downs are inevitable now and then, you need to be aiming for the ups that will even them out; perhaps it is that the steady state is a much narrower target than growth, and so harder to hit. And perhaps it is that control is in many circumstances a positional good.) Nietzsche often complains about supposing that life aims at self-preservation rather than will to power (Richardson 2004, 18 and esp. fn26, lists relevant passages), and I expect that we have just put part of an explanation for the complaint in place. Secondly, it is important not to confuse this argument with an appeal to natural selection. The argument operates at the level of the individual organisation; Darwinian selection requires replicators. The argument does not invoke anything like populations that reproduce themselves, or the evolutionary history of organisms (or organisations) of a given kind, though it is of course compatible with explanations that turn on natural selection. Here I am disagreeing with Richardson (2004) as to what is at the bottom of Nietzsche’s views: his “no other way” remark at p. 51 notwithstanding, not all selection is Darwinian selection, and in particular anthropic selection is not. Thirdly, notice incidentally that, in the case of a human being, we should not assume that the pattern of control that constitutes it is coterminous with its body. An expanding human being, in this sense, is not the character in David Foster Wallace’s The Broom of the System (1987), who intends literally to incorporate the universe, by eating it; accordingly, it is possible in special cases for human beings to augment their will to power by sacrificing their lives. Finally for now, you might think that priority-guided organisations are likely to persist because they have top priorities that are not will to power, but whose pursuit requires continued existence as a means. However, we can now see that if, realistically, one often achieves continued existence only by prioritising will to power, and if will to power typically trades off at the margin with most other priorities, the apparent counterexamples we are now considering will tend to be ineffective over the long run and will tend not to endure. Aiming merely at continued existence, rather than at will to power, is only too likely to put one in the position of being able to manage neither the continued existence, nor the further priority to which it was a means. 6. See, e.g., 5:24f, 35f/BGE 11, 21 (on the synthetic a priori and the concepts of ‘cause’ and ‘effect’), 3:471f, 477f/GS 111f, 121 (on logic, the geometrical perception of space, etc., as such cognitive optimisations—but see Poellner [1995, 193–195] for incredulity about the extent to which Nietzsche takes such simplifications to be responsible for the world as we encounter it). This gives us a quick way of resisting the mistaken but very common assimilation of perspectivism to garden-variety relativism. A relativism marks claims as true-in-___, or true-for-___, whereas Nietzsche thinks of the drastic simplifications that go to make up a perspective as falsehoods. Moreover, there is no self-contradiction in advancing, as an approximation or oversimplification, the claim that we think using approximations and oversimplifications. Consequently, the concern as to whether his alleged relativism is self-refuting, which appears repeatedly in the Nietzsche literature, should be treated as a symptom of inattentiveness to the text. 7. Obviously, the program of maintaining and extending the pattern of control that constitutes oneself does not actually amount to a guiding aim or

Who Wrote Nietzsche’s Autobiography?


a directive unless one can also tell what would count as a continuation or extension of the relevant pattern of control. (Compare Richardson [1996, 39]; during the twentieth century, Goodman [1979], and related work, heightened philosophical awareness of the problems that come with the notion of “same pattern,” but Nietzsche himself was likely to have been less edgy about this family of concerns.) So there must be additional content to the conception of will to power that guides an organisation. One sample such conception—just so that we have a sense of what these patterns can come to in practice—is the view attributed to Nietzsche by Nehamas (1985, 1988), on which a self is roughly the protagonist in a narrative, archetypally, the central character of a nineteenth-century novel. To maintain a self of this sort is to stay in character, that is, to act (but not just to act, because not everything that constitutes a character is an action) in a way that is consistent with one’s character. Notice that staying in character does not necessarily mean taking the character for granted, as opposed to exploring it, deepening it, and bringing out hitherto unexpected aspects of it; thus this notion of same pattern has some flexibility to it, which for this application is a desideratum. However, in what follows I will briefly explain why we should not follow Nehamas in treating the implicit protagonist of Nietzsche’s own texts as a highly coherent and highly consistent character of this sort. 8. The point gives us a low-key explanation for what might seem like an implausible Nietzschean doctrine, that a thing is the “sum of its effects” (13:275/WP 551). It is implausible if it is construed, with Nehamas (1985, ch. 3), as committing Nietzsche to a modally flat understanding of our world; when we cannot distinguish, roughly, essential from accidental properties, we lose the wherewithal to talk about change, and that is an intellectual competence that neither we nor Nietzsche are equipped to bypass. However, the imposition or construction of a perspective is in large part a matter of what one pays attention to, and what one ignores. Control exhibits itself in effects, and so if you are interested solely in control, when you think about an object, you are thinking pretty much solely about achieving effects. So if you are occupying the privileged will-to-power perspective, any feature of an object that is not exhausted by its effects is cognitively irrelevant to you and can be ignored. To revert to our ongoing example: perhaps there are features of scholarship— exquisiteness of treatment, say—that don’t make a difference to measures of academic productivity (such as rates of publication, or citation indexes), that don’t affect the national ranking of the author’s department, that don’t make a difference to enrolments, and so on—i.e., features that have, as far as an administrator is concerned, no effects. Then a dean doesn’t need to concern himself with such features as exquisiteness of treatment, and in fact he would do better to simply disregard them. 9. Since the argument I have sketched has a number of points of contact with the reading in Richardson (1996), this is a good place to register disagreements. Anthropic arguments are naturalised transcendental arguments; Richardson’s own view is that Nietzsche does not adopt his positions on the basis of transcendental arguments (288f). Richardson and I agree that the will-to-power metaphysics and perspectivism can be understood as compatible by taking each in a suitably modest sense (288–90); what he means by that, however, differs substantively from the modest renderings I have just sketched. In particular, Richardson holds that “Nietzsche replaces the bivalent notion of truth with a graded hierarchy of perspectives,” where the privileged “epistemic rank” of the will-to-power perspective is a matter of the perspective being, in the first place, “strong [i.e.,] (honest and courageous).” On the reading I have just given, the perspective’s privilege is a matter of







Elijah Millgram serving to advance the priorities one inevitably has. (Richardson [2004], e.g., 68, 105, 107, 115, 124f, raises the related question of how Nietzsche can think a value privileged, and attributes to him a variant of contemporary informed-desire theory: “these values have their status (‘higher’) from his making them in knowledge of the facts, since this is making them freely”; this seems to me anachronistic, and is not entailed by the argument I have just reconstructed.) Finally, Richardson reads the metaphysics of will to power through the lens of what used to be called process ontology, whereas my own view is that almost none of the arguments depend on this aspect of Nietzsche’s thought. Nietzsche gives a number of different characterisations of the term; for instance, an organism can also be described as decadent when it “chooses [which implies that it is capable of choice] . . . what is disadvantageous for it” (6:172/AC 6). Experiment shows that most of these characterisations of decadence are interderivable, given plausible auxiliary premises, but here our own interest is only in this one version of it. P. F. Strawson once argued that the notion of a disembodied soul was parasitic on the notion of an embodied soul (1971, 115f), and likewise we may be able to identify decadents only by understanding them as continuations of (or more generally deviations from) priority-guided homeostatic patterns of control—hence the “no longer.” Nietzsche seems to think of control structures as exhibiting top-down architecture, but notice that this stretch of argument goes through even in the face of models of unified agency which employ distributed control mechanisms. (Thanks to Diane Proudfoot for pressing me on this point.) The phrase is of course an allusion to John 19:5. Although I’m not the only reader to classify the book this way (e.g., Pippin 2006, 37), I have found that the genre does not always strike readers as obvious; recall, however, that the full range of autobiography includes not just straight birth-to-death narratives, but such classics as Cardano (2002). We could treat “Nietzsche” as the name of a literary artifact: the character depicted by Ecce Homo, along with the other writings that are, as one says in legalese, incorporated by reference. (This is Nehamas’ preferred method of reading.) Or we could identify this character with the flesh-and-blood author, Nietzsche, and treat the presentation as being about and of the once living, breathing individual. Later on I will argue for one of these over the other, but until that point I will develop the argument in a way that does not presuppose either. A quick sampling of the flurry: In the course of discussing his early essay, “Richard Wagner in Bayreuth,” he tells us that “in all psychologically decisive places I alone am discussed—and one need not hesitate to put down my name or the word ‘Zarathustra’ where the text has the word ‘Wagner’” (6:314/EH BT 4; for discussion, see Liébert [2004, 95–100]. In that essay we are given the picture of a character who vividly conforms to the willto-power characterisation we have already seen and are introduced to “the ruling idea of [Wagner’s] life—the idea that an incomparable amount of influence, the greatest influence of all in the arts, could be exercised through the theatre.” (We are also given his response to it: “Influence, incomparable influence—how? over whom?—that was from now on the question and quest that ceaselessly occupied his head and heart. He wanted to conquer and rule as no artist had done before” [1:475/UM 3.8].) Elsewhere he takes Wagner to task for being, precisely, a decadent, and admits to his reader at the outset that he is “no less than Wagner, a child of this time; that is, a decadent” (6:11/CW P).

Who Wrote Nietzsche’s Autobiography?


15. I’m using Kaufmann’s translation, with a handful of emendations. Kaufmann leaves a sentence off the beginning, which I’ve restored. As per usual, he introduces paragraph breaks into the English, but in this case it’s important to see that the flow of the passage is not broken up in this way; accordingly, ellipses are Nietzsche’s, and do not indicate that I am abridging his text. Finally, and most important, a short passage originally deleted by his sister is reproduced in a footnote, and I have returned it to the position it occupies in the original (and the currently standard German text). 16. Ecce Homo was written very near to Nietzsche’s collapse, and one might be tempted to dismiss the erratic tone and so on as a symptom of his illness, thus, not necessarily philosophically important signposting. However, the Genealogy of Morals similarly develops an equally elaborate and obviously contrived display of its author’s capacity to forget (for discussion, see Millgram 2007), and the Genealogy is widely, and in my view correctly, regarded as a masterpiece. If sudden swerves like these are a reason to dismiss Ecce Homo, they are a reason to dismiss the Genealogy, one of the canonical texts of moral philosophy, as well; moreover, recall that it is not all that long since all of Nietzsche’s work was dismissed as the product of dementia. So I do not recommend writing off Ecce Homo on these grounds. 17. Compare Nietzsche on the strength to discard (6:267/EH 1.2)—a strength he claims while taking pains to exhibit his inability to do so. To be sure, there are other reasons for not noticing discontinuities in our personalities, or for that matter our perceptual fields; see, e.g., Dennett (1991, 355f). 18. Formally, the contrast with will to power is striking: amor fati is all spin doctoring, all aspect, all desirability characterisation, and no guidance. The dramatic contrast between the function and content of these evaluative notions is part of a pattern; Müller-Lauter (1999) documents a number of similar contrasts, and argues that they are characteristic of Nietzsche’s work. That said, there is this connection between will to power and amor fati: effective action on the basis of priorities does require that one understand and accept the facts that form the basis for one’s judgements of effectiveness. 19. Probably the most important of these in the analytic tradition is Nehamas (1985), but in my own view Foucauldian readings of Nietzsche do so as well. 20. Anderson (2009) distinguishes two modes of interpretation that might be deployed in the service of Nietzsche’s famous demand, that one will one’s life, and the world as a whole, eternally to recur. On the Compensation Model, you judge that your life is worth reliving, on balance, because the good parts outweigh the bad parts. On the Transfiguration Model, the bad parts are “transfigured” by your reinterpretation of them, and you judge that your life is worth reliving because each part of it is worth reliving—as when you come to see an excruciatingly embarrassing experience as, in retrospect, hysterically funny. (The example is due to Joseph Jarone.) Anderson argues for attributing the Transfiguration Model to Nietzsche. We are now in a position to ask when these models can be expected to work. Who can decide to live with an ugly, painful, humiliating, experience now, in order to have a pleasant or beautiful or side-splitting experience later? The Compensation Model evidently requires a unified agent, someone who can command his psychic parts to take the hit, and take it because he says so: the drive sentenced to the painful experience (perhaps by being prevented from producing its characteristic mode of activity) is going to have to cooperate. And who can contemplate painful past experiences without being derailed by the responses of the drives that found them most frustrating? One again, it is the unified will-to-power agent that is able to suppress the unruly drive’s reaction to unpleasant memories. Decadent agents, however,






Elijah Millgram must have recourse to the Transfiguration Model: when they decide to go ahead with ugly, painful experiences, those experiences must be rewarding for the drives that undergo them, at the time they undergo them, because if the drive isn’t addressing its aim, it won’t chip in, and the agent as a whole can’t make the drive cooperate. And because decadent agents cannot suppress a drive’s reaction to what for the drive counts as past frustration, transfiguration is the only way for such an agent to avoid ressentiment. As a Kantian moral philosopher who hopes to use the inescapability of agency as an anchor for the moral duties, I imagine that Korsgaard counts it as auspicious that “plight” and “Pflicht” are etymological relatives. I reproduce here a few remarks that display the tenor of her view. She reminds us of “the things we say to people when it is time for them to stop dithering and bring deliberation to an end: Make up your mind, or even better, Pull yourself together.” Characterising failure to constitute oneself: “If you have a particularistic will, you are not one person, but a series, a mere heap, of unrelated impulses.” Motivating the specifically Kantian turn to principle: “a formal principle for balancing our various ends and reasons must be a principle for unifying our agency, since that is so exactly why we need it: so that we are not always tripping over ourselves when we pursue our various projects, so that our agency is not incoherent” (Korsgaard 2009, 58, 76, 126). Korsgaard’s argument is anchored by the claim that there are no alternatives to unified agency, and there has been some back and forth as to whether she is in a position to rule putative alternatives out (Millgram 2005; Enoch 2006; Ferrero 2009; and Tubert 2010). The present chapter is meant in part as an intervention in this debate; Nietzsche, as he presents himself, is alltoo-obviously such an alternative. However, in fairness to Korsgaard, participants in this discussion too often misconstrue her point as turning on the metaphysics of agency, and as bottoming out in a series of interlocking definitions. (Quickly: only agents act; so if you’re deciding what to do, you’re an agent; so anything that is true of agents must be true of someone who is making a choice.) The objections to her view are right about this much: a series of interlocking definitions, purporting to show that you have no alternative to acting—because nothing you choose will count as an alternative to action—don’t give anyone a reason not to live like a slacker. But the discussion is misconstruing Korsgaard’s position, which construes the necessity that sustains an agential constitution as practical rather than metaphysical. The apparently entailed regress—every instance of agency is reconstituted from a prior instance of agency—in turn means that Korsgaard still owes an account of how agency originates, which will presumably be a story about bootstrapping. For preliminary ground-clearing by a former student, see Schapiro (1999). For a critical reconstruction of Korsgaard’s argument that your commitment to your agency is not contingent, see Millgram (2011). But not necessarily a proposition, since we don’t know that the sentence would be assertoric: it might be, as Richard Hare thought, an imperative; or it might be something (like an expression of desire or intent) that looks like an assertoric sentence on the surface, but has, as Austin would have pointed out, another function entirely. Klossowski (1997) gives such an obscurely continental rendering of Nietzsche’s own sense of the urgency that I’m not sure I’ll be fairly representing it, but as I understand it, it does seem plausible enough. Nietzsche had symptoms that were either syphilis or would have been diagnosed that way, and so Nietzsche knew (or believed) that the spirochetes were eating his mind away from the inside. One day, the thought of the Eternal Return came to him, as an epiphany. The obvious explanation for the revelation must have

Who Wrote Nietzsche’s Autobiography?


26. 27.



immediately occurred to him: that he was finally going mad. On Klossowski’s view, Nietzsche’s sketches for a cosmological proof of the Eternal Return, along with his abortive plans to go back to school and get a science degree, were driven by the need to convince himself that the Eternal Return might be real physics, and not just a symptom of oncoming insanity. His best known composition is the Hymn to Life; for its early publication history, see Schaberg (1995, 140–149). The books are of course The Case of Wagner, Nietzsche Contra Wagner (which consists of selections from previous works), Richard Wagner in Bayreuth (now usually folded into the Untimely Meditations, but originally printed as a separate volume), and the Birth of Tragedy, whose 1872 title continues: Out of the Spirit of Music. His doctrine of the Eternal Return clearly has a musical model; Nietzsche sums it up with a “da capo,” i.e., with a bit of musical notation. For general discussion, see Liébert (2004), see also Perrakis (2011). See, e.g., the story of Midas and Silenus at 1:35/BT 3, and the remarks about “metaphysical comfort” at 1:56/BT 7. We can find Nietzsche expressing related views elsewhere also; he says, for instance, of the opposing demands of sensuality and chastity that “it is precisely such ‘contradictions’ that seduce one to existence” (5:341/GM 3.2 [=NCW 7.2]). Perrakis (2011, 53, 143) takes the intended dissonance to have to do with the clash between Dionysian and Apollonian artistic sensibilities, which seems to me to be a reach; but he does go on to attribute to Nietzsche a view that has points of contact with the reading I am developing: that insofar as man is dissonance incarnate, the human animal is infinitely interpretable. Since drives can have, as far as I can see, arbitrary foci, the decadent’s will to power might be the focus of one of numerous drives—in the terminology recently reappropriated from Freud for discussion of Nietzsche (Katsafanas 2013), its “aim.” Indeed, John Dewey argued that this does sometimes happen, attempted to characterise the conditions in which the will to power emerges as the aim of a distinct drive, and suggested that, when it does, it is normally a guise taken by ressentiment. (Dewey 1988, 97–99; the criticism which Dewey is making of Nietzsche is that he has the explanatory order backwards; instead of explaining ressentiment as an expression of will to power, he ought to be explaining will to power as an expression of ressentiment.) However, we do not have to assume that the agent does have such a drive. Earlier, I introduced the will-to-power perspective as the metaphysics of the org chart. Now, recall that Nietzsche devotes much of the early stretches of Beyond Good and Evil to arguing against atomism, which is in part a matter of treating the items in one’s ontology as lacking internal structure: in the will-to-power metaphysics, when you click on a box in the org chart, it opens up to reveal yet another org chart . . . presumably, all the way down. (5:31–34/BGE 19 gives us a first-pass sense of what that looks like, when we are considering volition in particular.) Thus if Nietzsche’s psychological ontology includes drives, we can ask what we see when we look inside a drive, and what holds the drive together; when we do, we recapitulate a variant of the Nietzschean anthropic argument. Drives are extremely complex patterns of control; so complex, indeed, that a recurring scholarly worry has been that drives look too much like intentional agents to be philosophically kosher. If drives themselves have something like governing priorities, those priorities must be centred on or include will to power: here, the priority of extending the scope of control of the drive. So we can also understand the decadent’s interest in his own




31. 32. 33. 34.



Elijah Millgram existence as the expression of one or another of his drives’ will to power. (I’m grateful to Margaret Bowman for helping me out with this point; the view I now think is correct is quite close to that of Richardson 1996, 26) Nietzsche objects to the ascetic priest’s insistence on prescribing the same medicine to everyone, so we do not need to suppose that dissonance will work for us, too. His treatment was tailored to his own “physiology”— not all that successfully, since he collapsed while completing and revising Ecce Homo itself. (For a description of the collapse, see Hollingdale [1999, 237–239]). The primary motivation for the distinction between “writer” and “author” seems to have been the desire to avoid committing the so-called Intentional Fallacy (introduced in Wimsatt and Beardsley 1954). The alleged Fallacy comes in two versions: the more obviously mistaken is that of taking what an author meant to say to be what he did say; the less obviously mistaken is that of taking biographical considerations of pretty much any kind to be relevant to the interpretation of an author’s texts. Because Nehamas’ postulated author is a “regulative ideal,” its intentions do not count as biography, and this allowed Nehamas to sidestep one of the academic McCarthyisms of his day. We will shortly consider whether these motivations are in place in a reading of Nietzsche. I’m grateful to Ian Anthony for reminding me of this. I am borrowing Bernard Williams’ remark that Nietzsche’s writing is often booby-trapped “not only against recovering theory from it, but, in many cases, against any systematic exegesis that assimilates it to theory” (2006, 300). Vogler (2002) is a recent and sophisticated exposition of this sort of view. I mentioned previously, in note 30, that Nehamas’ insistence on the distinction between “writer” and “author” is motivated by the need to avoid the Intentional Fallacy. So this is an occasion to make two related points about that motivation. First, recall that one of Nietzsche’s characteristic moves is to interpret religious, moral, and other doctrines as symptoms of underlying sickness— that is, to read texts through the biographies of their producers, and, often, through their unavowed intentions. Consequently, if you are convinced that the Intentional Fallacy really is a fallacy, you should find Nietzsche so thoroughly and persistently wrong-headed as not to be worth your exegetical time and attention. Conversely, if you find that Nietzsche is worth careful and close reading, you had better proceed on the working assumption that the Intentional Fallacy is not a fallacy after all. (I’m very grateful to Carolina Allen for this observation.) Second, notice that the techniques used for generating a Nehamasian postulated author are parasitic upon the methods used to interpret flesh-andblood “writers,” and that—as Grice once pointed out (1989)—these methods rely heavily on the intentions of the person producing the text or utterance being interpreted. So interpretive appeals to intentions of actual persons must make sense and be legitimate if the more exotic techniques modelled on them likewise make sense. It is hard to see what grounds—other than a fetishising of “literary texts”—could support ruling out the former in favour of the latter. Scholars have for the most part tried to understand Nietzsche’s drives, and Nietzschean valuing, teleologically (as, e.g., Richardson 2004, 37, 129). If I am right, we need to think of drives as Nietzsche’s attempt at an alternative to end-directed forms of explanation, rather than an implementation of end-directedness. For a related point, see Millgram (2002, 177).

Who Wrote Nietzsche’s Autobiography?


37. For thoughtful discussion of the problem, see MacIntyre (1990, ch. 2). 38. Scholars typically acknowledge that Nietzsche’s stylistic choices ought to complicate the use of his texts, but then go on to write as though those complications could be postponed until after Nietzsche’s position has been assembled, and assembled from such indiscriminately cited passages. The announcement, by Reginster (2006), at pp.  4f, that the book “is confined to an exploration of the substance of Nietzsche’s philosophy,” is a representative case, not least in that it assumes it to be possible to determine what that substance is before resolving the problems posed by Nietzsche’s “style.” (Reginster gestures at his take on those problems: the style is “a deliberate form of esotericism, an effort to conceal the truths [Nietzsche] reveals from those not worthy of them . . . who are not ‘entitled’ to his insights” (19), but the supporting passages he adduces are produced by authorial personae that do not allow them to be taken at face value.) 39. I am embarrassed to admit that I once made this move myself (1997, 175). It now rings hollow. 40. And what could be a more definitive condemnation of Kantian moral theory? So Korsgaard attempts a reply: the disunified agent is a defective agent (but still an agent), in something like the way a house with holes in the roof is a defective house, but a house nonetheless (2008, 112f). So far Nietzsche agrees; the decadent is sick. But she concludes that the standards we apply to the nondefective item apply to the defective item also. The problem with this strategy is not just that it’s obsolete metaphysics: that in the twenty-first century, nothing has a medieval essence determining it to be really a house, rather than, say, a gazebo. It’s that if the standards aren’t something you can use to guide your behaviour, then they’re practically irrelevant. There’s no point in telling a decadent to buck up, pull himself together, and conform to the standards we apply to fully functioning agents: he can’t. 41. Once we have contrasted the attitudes involved in amor fati with those of the will-to-power agent, we can see Nietzsche’s preferences on display in the selfmocking attitude he takes towards the former. The posture of amor fati has its ridiculous side, which Nietzsche goes out of his way to emphasise. He tells us that his painful eye problems, his severe (and he anticipates terminal) illness, his very disintegration (i.e., his decadence) were each, as we would colloquially put it, the best thing that ever happened to him (6:326/EH HTH.4; 6:265f/EH 1.1; for his “Dreiviertelblindheit,” see Janz 1981, vol. 2, 500). He evidently toyed with insisting that Christianity itself (formerly condemned as nearly two millennia of contemptible lies, self-mortification, and on and on) was justified as a necessary precondition of . . . himself (13:641/Nietzsche 2000, 799). And in a section that he did end up trying to replace with a vitriolic diatribe against his closest relatives, he implied that he had chosen the date of his own birth. (The former can be found as Kaufmann’s rendering of EH 1.3; the latter in the Studienausgabe version of the same section.) Nietzsche had a wonderful ear for style, and it is implausible that he was unaware of just how comic this sounded. 42. You might think that that there’s no need to wonder about it: if Nietzsche is right, this here is what it’s like to occupy the will-to-power perspective. But that is to assume that one is the sort of creature for whom the Nietzschean anthropic argument goes through, and we have just seen that that assumption is not always allowable. 43. The sea monsters seem to be a kind of icon for the world as will to power; here is Michael Walzer, quoting Edmund Wilson’s related picture of the international arena: “I think that it is a serious deficiency on the part of historians . . . that they so rarely interest themselves in biological and zoological


Elijah Millgram

phenomena. In a recent . . . film showing life at the bottom of the sea, a primitive organism called a sea slug is seen gobbling up small organisms through a large orifice at one end of its body; confronted with another sea slug of only slightly lesser size, it ingurgitates that, too. Now the wars fought by human beings are stimulated as a rule . . . by the same instincts as the voracity of the sea slug” (Walzer 1992, 60). Walzer comments: “There are no doubt wars to which that image might fit, though it is not a terribly useful image with which to approach the Civil War. Nor does it account for our ordinary experience of international society.” 44. For treatments of the Eternal Return, see, just for instance, Danto (1965, 212), Clark (1991, ch. 8), Nehamas (1985, ch. 5), Reginster (2006, ch. 5). Older interpretations include Zuboff (1980) and Soll (1980). 45. My own guess is that the extremes can be approached, perhaps quite closely, but not actually occupied. This is obvious enough at the fragmented end of the spectrum, but I think it is also, albeit less obviously, true at the will-to-power end. Imagine an agent that is fully governed by a dominant drive. When we open up the drive—in our earlier metaphor, when we click on that box in the org chart—we will either find structure that is in turn fully governed by one of its components, or we will find disorganisation and disunity. If we find the latter, we must expect the control of the drive to lapse sooner rather than later, and in any case, we have identified a residuum of decadence that prevents us from assigning this agent to the endpoint of the spectrum. But if we find the former, we can open up that governing component as well . . . and now it is obvious that we are embarked on a downward regress. 46. I’m grateful to Lanier Anderson, Chrisoula Andreou, Alyssa Bernstein, Sarah Buss, Leslie Francis, Brian Leiter, Rachel Shuh, Neil Sinhababu, Robert Solomon, and Lauren Tillinghast for comments on an earlier draft, and to audiences at Dartmouth College, the Hebrew University, Macquarie University, the University of Canterbury, the University of Melbourne, the Australian National University, the University of Arizona, the University of Washington, Simon Fraser University, the University of British Columbia, Ohio University, the University of Utah’s Tanner Humanities Center, Utah Valley State College, and the University of Texas at Austin for helpful feedback. I would also like to thank Lori Alward, Konstanze Ballueer, Luca Ferrero, Amy Johnson, Jonathan Lear, John Richardson, Sherri Roush, Valerie Tiberius, and Candace Vogler for discussion.

References Anderson, R.L. 2009. “Nietzsche on Redemption and Transfiguration.” In J. Landy and M. Saler, eds., The Re-Enchantment of the World: Secular Magic in a Rational Age. Stanford, CA: Stanford University Press, pp. 225–258. Barrow, J. and F. Tipler. 1986. The Anthropic Cosmological Principle. Oxford: Clarendon Press. Bratman, M. 2006. Structures of Agency. Oxford: Oxford University Press. Cardano, G. 2002. The Book of My Life, J. Stoner, trans., introduction by A. Grafton. New York: New York Review Books. Clark, M. 1991. Nietzsche on Truth and Philosophy. Cambridge, UK: Cambridge University Press. Clark, M. and D. Dudrick, D. 2009. “Nietzsche on the Will: An Analysis of BGE 19.” In K. Gemes and S. May, eds., Nietzsche on Freedom and Autonomy. Oxford: Oxford University Press, pp. 247–268.

Who Wrote Nietzsche’s Autobiography?


Coscarelli, D. 2004. Bubba Ho-tep, Produced by Don Coscarelli, Jason R. Savage, Ronnie Truss, and Mark Wooding. Metro-Goldwyn-Mayer. Danto, A. 1965. Nietzsche as Philosopher. New York: Columbia University Press. Dennett, D. 1991. Consciousness Explained. Boston: Little, Brown. Dewey, J. 1988. Human Nature and Conduct, J. Boydston and P. Baysinger, eds. Carbondale, IL: Southern Illinois University Press. Enoch, D. 2006. “Agency, Schmagency: Why Normativity Won’t Come from What Is Constitutive of Action.” The Philosophical Review 115(2): 169–198. Ferrero, L. 2009. “Constitutivism and the Inescapability of Agency.” Oxford Studies in Metaethics 4: 303–333. Frankfurt, H. 1988. The Importance of What We Care About. Cambridge, UK: Cambridge University Press. Goodman, N. 1979. Fact, Fiction and Forecast, 3rd ed. Indianapolis, IN: Hackett Publishing. Grice, P. 1989. “Logic and Conversation.” In P. Grice, Studies in the Way of Words. Cambridge, MA: Harvard University Press, pp. 22–40. Hollingdale, R.J. 1999. Nietzsche: The Man and His Philosophy, revised ed. Cambridge, UK: Cambridge University Press. Janz, C.P. 1981. Friedrich Nietzsche, Biographie. Munich: Deutscher Taschenbuch Verlag. Katsafanas, P. 2013. “Nietzsche’s Philosophical Psychology.” In J. Richardson and K. Gemes, eds., Oxford Handbook of Nietzsche. Oxford: Oxford University Press, pp. 727–755. Klossowski, P. 1997. Nietzsche and the Vicious Circle, D. Smith, trans. Chicago: The University of Chicago Press. Korsgaard, C. 2008. The Constitution of Agency. Oxford: Oxford University Press. Korsgaard, C. 2009. Self-Constitution. Oxford: Oxford University Press. Liébert, G. 2004. Nietzsche and Music, D. Pellauer and G. Parkes, trans. Chicago: The University of Chicago Press. MacIntyre, A. 1990. Three Rival Versions of Moral Enquiry. Notre Dame, IN: The University of Notre Dame Press. Millgram, E. 1997. Practical Induction. Cambridge, MA: Harvard University Press. Millgram, E. 2002. “How to Make Something of Yourself.” In D. Schmidtz, ed., Robert Nozick. Cambridge, UK: Cambridge University Press, pp. 175–198. Millgram, E. 2005. “Practical Reason and the Structure of Actions.” In E. Zalta, ed., The Stanford Encyclopedia of Philosophy. URL = . Millgram, E. 2007. “Who Was Nietzsche’s Genealogist?” Philosophy and Phenomenological Research 75(1): 92–110. Millgram, E. 2009. Hard Truths. Oxford: Wiley-Blackwell. Millgram, E. 2011. “Critical Notice: Christine Korsgaard, Self-Constitution and The Constitution of Agency.” Australasian Journal of Philosophy 89(3): 549–556. Müller-Lauter, W. 1999. Nietzsche: His Philosophy of Contradictions and the Contradictions of His Philosophy, D.J. Parent, trans. Urbana, IL: University of Illinois Press. Nagel, T. 1991. “What Is It Like to Be a Bat?” In T. Nagel, Mortal Questions. Cambridge, UK: Cambridge University Press, pp. 165–180.


Elijah Millgram

Nehamas, A. 1981. “The Postulated Author: Critical Monism as a Regulative Ideal.” Critical Inquiry 8: 133–149. Nehamas, A. 1985. Nietzsche: Life as Literature. Cambridge, MA: Harvard University Press. Nehamas, A. 1987. “Writer, Text, Work, Author.” In J. Cascardi, ed., Literature and the Question of Philosophy. Baltimore, MD: Johns Hopkins University Press, pp. 267–291. Nehamas, A. 1988. “Who Are ‘the Philosophers of the Future?’: A Reading of Beyond Good and Evil.” In R. Solomon and K. Higgins, eds., Reading Nietzsche. Oxford: Oxford University Press, pp. 46–67. Nietzsche, F. 1954. The Portable Nietzsche, W. Kaufmann, ed. and trans. New York: Penguin Books. Nietzsche, F. 1968. The Will to Power, W. Kaufmann and R.J. Hollingdale, trans. New York: Vintage Books/Random House. Nietzsche, F. 1974. The Gay Science, W. Kaufmann, trans. New York: Vintage Books/Random House. Nietzsche, F. 1982. Daybreak, R.J. Hollingdale, trans., M. Tanner, ed. Cambridge, UK: Cambridge University Press. Nietzsche, F. 1988. Sämtliche Werke. Deutscher Taschenbuch Verlag/de Gruyter, Berlin. 15 vols., edited by Giorgio Colli and Mazzino Montinari. Nietzsche, F. 1996. Human, All Too Human, R.J. Hollingdale, trans., R. Schacht, ed. Cambridge, UK: Cambridge University Press. Nietzsche, F. 1997. Untimely Meditations, R.J. Hollingdale, trans., D. Breazeale, ed. Cambridge, UK: Cambridge University Press. Nietzsche, F. 2000. Basic Writings of Nietzsche, W. Kaufmann, trans. and ed. New York: Modern Library/Random House. Onions, C.T., G.W.S. Friedrichsen, and R.W. Burchfield. 1966. The Oxford Dictionary of English Etymology. Oxford: Clarendon Press. Perrakis, M. 2011. Nietzsches Musikästhetik der Affekte. Munich: Karl Alber. Pippin, R. 2006. Nietzsche, moraliste français, I. Wienand, trans., M. Fumaroli, ed. Paris: Odile Jacob. Poellner, P. 1995. Nietzsche and Metaphysics. Oxford: Oxford University Press. Reginster, B. 2006. The Affirmation of Life: Nietzsche on Overcoming Nihilism. Cambridge, MA: Harvard University Press. Richardson, J. 1996. Nietzsche’s System. Oxford: Oxford University Press. Richardson, J. 2004. Nietzsche’s New Darwinism. Oxford: Oxford University Press. Roush, S. 2003. “Copernicus, Kant, and the Anthropic Cosmological Principles.” Studies in History and Philosophy of Modern Physics 34: 5–35. Schaberg, W. 1995. The Nietzsche Canon: A Publication History and Bibliography. Chicago: The University of Chicago Press. Schapiro, T. 1999. “What Is a Child?” Ethics 109: 715–738. Smart, J.J.C. 1987. “Philosophical Problems of Cosmology.” Revue Internationale de Philosophie 160(4): 112–126. Soll, I. 1980. “Reflections on Recurrence: A Re-Examination of Nietzsche’s Doctrine, Die ewige Wiederkehr des Gleichen.” In R. Solomon, ed., Nietzsche: A Collection of Critical Essays. Notre Dame, IN: The University of Notre Dame Press, pp. 322–342. Strawson, P.F. 1971. Individuals. London: Methuen.

Who Wrote Nietzsche’s Autobiography?


Tubert, A. 2010. “Constitutive Arguments.” Philosophy Compass 5: 656–666. Viereck, P. 1965. Meta-Politics: The Roots of the Nazi Mind. New York: Capricorn Books. Vogler, C. 2002. Reasonably Vicious. Cambridge, MA: Harvard University Press. Wallace, D.F. 1987. The Broom of the System. New York: Viking Penguin. Walzer, M. 1992. Just and Unjust Wars, 2nd ed. New York: Basic Books. Wiggins, D. 1991. “Truth, Invention, and the Meaning of Life.” In D. Wiggins, Needs, Values, Truth, 2nd ed. Oxford: Blackwell, pp. 87–137. Williams, B. 2006. “Nietzsche’s Minimalist Moral Psychology.” In M. Burnyeat, ed., The Sense of the Past. Princeton, NJ: Princeton University Press, pp. 299–310. Wimsatt, W.K. and M. Beardsley. 1954. “The Intentional Fallacy.” In W.K. Wimsatt, The Verbal Icon. Lexington, KY: University Press of Kentucky, pp. 3–18. Zuboff, A. 1980. “Nietzsche and Eternal Recurrence.” In R. Solomon, ed., Nietzsche: A Collection of Critical Essays. Notre Dame, IN: The University of Notre Dame Press, pp. 343–357.

12 The Abnegated Self Nellie Wieland

1st Gent. Our deeds are fetters that we forge ourselves. 2nd Gent. Ay, truly: but I think it is the world That brings the iron. (Eliot, Middlemarch IV)

Sometimes a person introspects and there is nothing there. Or, whatever is there is not a single coherent self to which can be attributed a narrative, reasons, or authentic life plan. In the most extreme case, the disunified mind is a kind of Humean theatre in which, “perceptions successively make their appearance; pass, re-pass, glide away, and mingle in an infinite variety of postures and situations” (Hume 1739/2000, 1.4.6).1 Oliver Sacks describes a patient with Korsakoff’s condition (severe memory loss and confabulation due to alcohol abuse). This patient is characterised as lacking a past or future, and stuck in a “constantly changing, meaningless moment.” Sacks said that the patient had lost part of himself but was in no position to know it. Unlike the loss of any other part of oneself (such as a leg), one cannot know that one has lost one’s self (Sacks 1985/2006). There are more ordinary cases of disunified selves that are less severe and may allow for eventual resolution. In an ordinary case, a person may be alienated from themselves and lack even internal coherence. In this chapter, I consider the problem of self-abnegation, both descriptively and normatively. This abnegation can be to another person, to external pressures, or even to a concept or an idea. The cases that I consider involve more than just deference, or institutionally circumscribed abnegation of agency. They involve something that looks like the transfer of agency to something or someone outside of oneself. Any narrative of one’s life or reasons given for action, in such a case, are properly those of another. For example, if I believe whatever B believes, if I want whatever B wants, and if I treat the reasons for my beliefs and desires as mine just in case they are B’s, then I have abnegated my self to B. In what follows, I look at two questions. The first is this: is there an account of selfhood that can reliably discriminate between a strong self and an abnegated self? The second is this: what are the reasons, if any,

The Abnegated Self 215 for treating a person with an abnegated self as if they were a person with a strong self? To answer these questions I begin with four features of a strong self. My goal is to identify the feature of a strong self that, when missing, is evidence of abnegation. I argue that abnegation differs from other ways in which a person’s self can be compromised. For example, if a self is lost to the psychiatric effects of isolation or to neurological degeneration from alcoholism, it will fail to evince the features of strong selfhood identified here. The case of self-abnegation is different because it is a kind of borrowed or externally imposed selfhood. In this case, an abnegated self may appear to satisfy each of the four features of strong selfhood on their own. In order to resolve this, I provide more detailed descriptions of abnegation in order to identify how it diverges. I speculate that these four features are all necessary features of a strong self and must be sourced autonomously from a first-person perspective. Even if an abnegated self putatively satisfies the four features of strong selfhood, those four features must also be derived from a first-person perspective. I then turn to the normative question of why the abnegated self deserves to be treated as an intact agent. I defend what I call a “conversational” and “ecological” account of the conditions of agential flourishing. In doing this I explicitly reject what I call a “counterfactual” account—which I take to be the primary alternative.

1. A Strong Self We can begin to think about what it might mean to abnegate your self, or your agency, by first providing an account of what it looks like when it is intact. A strong self has at least four features: 1. 2. 3. 4.

It is integrated with one’s reasons and desires. It has continuity over time. It is reflectively endorsed at a higher level. It is interpersonally engaged with or recognised by others.

Each of these features has been argued for extensively. Each on their own cannot reliably discriminate between a self that is autonomously and authentically one’s own and one that has been abnegated. Let’s see why this is the case. A strong self must be integrated with one’s reasons and desires.2 These three features of a person—selfhood, reasons, and desires—are interdependent. A person’s desires are authentically one’s own if they are coherently integrated with one’s reasons for action and if they come from one’s self. This can be described as having a “distinct perspective” (SayreMcCord and Smith 2014)3—that there is some way in which one acts in the world that comes from an integrated and self-determining stance. That stance is distinct insofar as a self is integrated with its own reasons


Nellie Wieland

and desires for acting. On its own, this condition for a strong self is fairly weak. An agent’s desires and reasons for acting could be fully integrated with one another and yet completely derived, as by a rule, from another agent. The self of that agent could be derived from the will of another and yet display full integration. A strong self has continuity over time. This kind of continuity is typically represented as narrative continuity.4 The standard view of narrative identity is one that allows for multiple narrators throughout the course of a life held together; Owen Flanagan (2007) calls this a “multiplex” self and contrasts it with a “multiple” self—which he describes as disordered. In this latter case, there is no self that is holding the narrative together between the sub-narrators. He calls this holding together “narrative connectedness.” This narrative connectedness comes about by active authorial work. This is itself interesting, and probably central (and not accidental) to any correct account of the self. It’s not just that a narrator can tell a coherent story, continuous across time, and from a firstperson perspective, but that this is something that is constructed by that first-person narrator. Telling a coherent story, and having a first-person perspective, is not a passive act. It requires autonomous action to tell (and be) the story of your life. Narrative continuity may demand that an agent be able to tell a connected story about who they are and how their life is going. This in turn should be held together from a unique first-person perspective on past events and a continuous thread to a present sense of self. A demand for narrative continuity can take different forms. It can ask for a whole life narrative: a single story that holds all of the parts of one’s life and identity together. Or, it can allow for segmented narratives: carving up either different parts of one’s life, or different aspects of one’s agency.5 Abnegation, on the other hand, is not always a short-term state. Long-term abnegation can still have all the marks of narrative continuity. A derivative self can be narratively continuous if it is derived from an agent who evinces narrative continuity. This will lack a unique, first-person perspective, properly understood, but is otherwise possible. A strong self is reflectively endorsed at a higher level. This is a simple statement of Harry Frankfurt’s (1971) account of second-order desires and volitions. A person is the kind of creature who possesses secondorder desires and volitions, and the self of that person is the kind of entity that is endorsed at the second-order. A person’s self is constituted by their beliefs and desires, which are endorsed as their own and arranged in some such way at the second-order. The volitions of this self just are those that emerge from this constitutional ordering.6 Some version of this concept of persons, and their selves, is undoubtedly both correct and essential to the possession of a strong self. However, on its own it still fails to distinguish the strong self from the abnegated self. The reason for this is that a person can possess first- and second-order beliefs and desires

The Abnegated Self 217 and not have come by them on their own. If I first-order desire whatever B desires, and if I second-order treat the reasons for my desires as mine just in case they are B’s, then I have abnegated my self to B. Second-order desires can still organise first-order desires and operate according to a rule such that they are whatever is derived from an external source. There is no particular way to ensure that they are authentically, autonomously, and internally generated. A strong self is interpersonally engaged with or endorsed by others. Without taking a position on whether this is a necessary condition, we can at least see that without such interpersonal recognition—in cases of isolation—selves disintegrate. There are reasonable descriptions of selfhood and autonomy that emerge from or are realised within relations.7 There are also theories of agency and selfhood that assume something about the significance of interpersonal engagement or recognition. Versions of this are embedded into compatibilist and conversational accounts of moral responsibility, and into interpersonal views of selfhood (McKenna 2012; Strawson 1962; Vargas 2013; and Westlund (2003).8 I will say more about this in the final section of the chapter. A strong self is integrated with one’s reasons and desires, has continuity over time, is reflectively endorsed at a higher-order, and is interpersonally engaged with or recognised by others. These sound like pretty high demands but I think that they are the steady state for selfhood. Before continuing to develop an account of self-abnegation in light of these four features of strong selves, let’s look at the conditions in which selves are abnegated.

2. Three Varieties of Self-Abnegation The self can be abnegated in a number of ways, some diffuse, some direct. I briefly describe three varieties of self-abnegation. The first example, the “diffuse” abnegated self, is a case where an agent does not develop, or fully realise, a self of their own due to social circumstance. This may happen over the course of a lifetime and may result in an under-developed conception of oneself as a source of reasons, narrative, and agency. The second kind of case I describe is of a self abnegated as a result of domination by another. The third kind of case I describe is of a self wilfully abnegated to another person or idea. I have primarily thought about self-abnegation as a bad thing. Perhaps it is a personal bias that I regard the giving up of one’s self as something that is worse for a person than anything else. But this isn’t obviously true. Subjective well-being is awfully important, and an abnegated agent can still score highly on measures of subjective well-being, even though their goal-fulfilment is inauthentic. There may be ways in which adopting and holding one’s own goals is antithetical to subjective well-being because that can be more difficult, isolating, and disappointing. Failing


Nellie Wieland

as a group, or failing at achieving goals inauthentically set, may be less painful than failing as an individual. So while I focus in this chapter on the harms that accompany that abnegation of self, a discussion of the joys will have to be had another time. In the previous section I introduced four features of strong selves. Selfabnegation works with these four features. Abnegation occurs because of some failure in the development of these features. For example, an agent who is systematically and uniformly unrecognised as an agent (a failure of the fourth feature), or an agent who does not arrive at second-order organisation of their first-order desires autonomously, is an agent who is partially or wholly abnegated. However, a self can be abnegated while one or more of the features is putatively realised. I return to this point in what follows. 2.1 Abnegation From Social Circumstance A person’s development of self can, pretty obviously, be shaped, limited, and structured by social forces. I think of this as diffuse because it cannot be tied to a limiting event or person (as in 2.2), but it is the result of chronic and structural suppression of the features of agency and selfhood. These features of agency and selfhood might include having a subjective point of view, permission to form life plans of one’s own, and recognition that one’s reasons and rational deliberations are taken as legitimate sources of these plans. They might also include being seen as someone who is capable of interpersonal engagement with their reasons and rational deliberations and of having the resources to organise one’s life around these processes. Suppression of these features of agency and selfhood take familiar forms: legal, societal, and cultural enforcements of suppressed personhood. They may be publicly instantiated in familiar ways: a lack of freedom of movement or livelihood, ineligibility to own property or represent oneself in court, denial of rights to form relationships with other people or to choose who one marries or with whom one is intimate, etc. In cases of chronic and structural oppression, these features of agency and selfhood can be suppressed, the abnegation of the self is assumed by the structure of society until it is assumed by the agent as well. The kinds of conditions in consideration here are, for example, slavery, and societies structured around sexist, racist, ethnic, or religious control. It is important to draw distinctions between the self-abnegation from social pressure that I have in mind and two other phenomena. Adaptive preferences are changes in one’s attitudes towards realising one’s desires given environmental limits (Bruckner 2009; Elster 1983; Zimmerman 2003). These can be conscious or subconscious. The kind of selfabnegating person I have in mind is not merely a person adapting their preferences, since adaptive preferences are compatible with self-affirming agents who are simply working within a world that doesn’t always give

The Abnegated Self 219 them what they want. The kind of structurally diffuse, oppressive abnegation I am describing is something that strips the authenticity and selfownership of an agent’s reasons, rational deliberations, and interpersonal engagement. So while adjacent to adaptive preferences, I am trying to describe a much stronger attack on the self. It’s also important to distinguish between the self-abnegation I have in mind and shared autonomy (Mackenzie and Stoljar 2000). The latter can be a positive phenomenon that does not identify an agent as having diminished or inauthentic agency, but rather as having agency or autonomy that is not located in a single self, isolated from their relationships with others. In cases of abnegation from social circumstances, a person is systematically immersed in the idea that their own agency and selfhood lacks value, authenticity, and autonomy, and ought to be substituted with the agency and selfhood of another person or group of people. This is the case when, for example, women or racialised persons are regarded as having deficient agency and selfhood and are not allowed to make decisions or have standing with respect to marriage, reproduction, finances, property ownership, legal representation, religious belief, enfranchisement, etc. The claim being made here is not that self-abnegation is determinate in these cases (obviously, it is not, which explains resistance against these very forms of oppression) but that it is not only possible, but likely, that self-abnegation will be internalised in the oppressed group. 2.2 Abnegation Under Domination The second kind of case I have in mind is of an agent who never develops a self of their own due to domination by a particular other. These are cases of self-abnegation to the point of internalising the domination or oppression and becoming pathologically deferential. Illustrative examples include brainwash, abuse, and other chronic and isolating oppressive circumstances. It may be the case that some cases of this kind of concentrated (in contrast with “diffuse” in 2.1) oppression do not result in self-abnegation. The kinds of cases I am interested in are those that do—those cases where a person objectifies themselves, and comes to believe that they lack subjectivity, rationality, and individuality as a result of persistent mistreatment. A person in this position cannot identify values, aspirations, or a self-reflective identity of their own and this can be seen as directly tied to delimited circumstances of autonomy-denial. For example, an isolated community or family that fosters an ideology of domination by one group over another, extreme deference, and provides no models or access to alternative conceptions of moral relationships is in a position to shape every aspect of a person’s self-conception. People may exist in such communities with no conception of themselves as oppressed, since they don’t conceive of themselves as having unrealised selves of their own. This can also be found in some abusive relationships,


Nellie Wieland

long-term kidnap victims, and, perhaps especially, in children raised in similar conditions. 2.3 Wilful Abnegation The third kind of case is of the agent who wilfully abnegates to another person or conception. Dean Zimmerman (2003) describes an agent in this circumstance who, in order to escape an overbearing work environment, renounces his self entirely (he might renounce the very idea of self) in order to be unflappable in the face of threats to his desires. He might simply reject the idea of his having a self and the beliefs and desires that come along with it. He might do this as a way of escaping negative experiences. But, as a result, he doesn’t have an authentic self to whom can be attributed reasons or interpersonal engagement. Or, consider a case of religious devotion that is so robust that an adherent regards themselves as “vehicles” of the will of God, or as “married” to Jesus, or as having “renounced the world” and themselves to their devotion. This kind of renunciation moves along several planes. The adherent may give up on their own self and its reasons, beliefs, and desires, but they may also give up on (moral) responsibility for themselves and for others. These cases differ from those in 2.2 because they don’t require express control or domination by another. They may not require the presence of another person—certainly renouncement of self to God is such a case. This doesn’t mean that there is always a clear line between wilful abnegation and domination by another. Some of the time we wilfully seek out those leaders, rulers, dominators, prophets who we enthusiastically follow until our selves are erased by control or abuse. However, these cases can come apart too, and it’s worth attending to the differences in these varieties of self-abnegation.

3. Identifying an Abnegated Self How can an abnegated self be identified, and what do the previous accounts of self-abnegation have in common? In each of the cases, the four features of a strong self may putatively be realised, or be realised in various degrees, but what is lacking is an integration of the four features from a first-person, autonomous position. This is why I argue that none of these criteria for selfhood is sufficient on its own. What does it mean for these four features to be integrated from a first-person, autonomous position? The remainder of this section is devoted to understanding this. Here I think it is helpful to think about moral responsibility. Accounts of moral responsibility have arisen out of two related problems: the (lack of) freedom of the will, and the problem of degrees of (diminished) agency. Very simply understood, if it’s the case that the freedom of the will is an illusion, then it seems that it must follow that there is no place for moral

The Abnegated Self


responsibility. One can only be responsible for that which one can control. Instead, we can look to how we hold one another responsible. We can look for how we take one another to be responsible, depending on our capacities, and that taking (or reacting) is what structures the norms of moral responsibility (cf. Strawson 1962). The taking of one another as responsible is a matter of assessment of the “quality of will” of the agent in question (cf. McKenna 2012). This can be minimally understood as the giving and taking of reasons. If someone is taken to be morally responsible, then presumably they are in a position to offer reasons for their behaviour, and to be responsive to questions. This is itself an interpersonal exchange, one where agents are taken to be responsible to the extent that they can provide reasons for their behaviour in exchange with other agents. This is a corollary to the criteria I use earlier of (1) integration with one’s reasons and desires, and (4) interpersonal engagement or recognition by others. Theories of moral responsibility can demand a richer understanding of criterion (4). They can also require the presumption of competence in the person taken to be responsible. For example, I take X reasons to be the kinds of reasons a competent agent takes to be governing. Insofar as I offer X reasons to you to explain your behaviour or to justify mine, I am claiming that each of us is such an agent. This can also be applied to agents who have a diminished set of competencies. It can make sense of agents with limitations in their competence (or perhaps willingness) to accept as governing those reasons that a strong agent accepts. These agents lack some quality of their will and could be considered morally responsible to some degree, even if not fully morally responsible.9 This enriched understanding, borrowed from accounts of moral responsibility, of the integration of one’s will with one’s reasons and desires, interpersonal engagement and recognition, narrative continuity, and second-order endorsement, still don’t seem to capture all that has gone wrong in the case of self-abnegation. What is missing is the source of the self that performs the integrative function. By definition, self-abnegation means that an agent’s self is borrowed from or derivative from another. The strong self must be sourced first-personally. Here I find it helpful to think about what it means to have beliefs of one’s own and to extrapolate from that to thinking about what it means to have a self of one’s own. The idea of borrowing the beliefs of another person is easier to understand. If I believe whatever my husband, boss, religious leader, or political party believes, then those beliefs are not mine. Or at least they aren’t mine in the same sense as if I arrived at them independently. Of course, it’s the case that all agents rely on authorities or external sources for their beliefs, so the question to ask is whether these borrowed beliefs are otherwise integrated with one’s reasons and desires and endorsed at a higher-level. But this too is still inadequate. What’s missing is the process whereby these beliefs are adopted autonomously.10 The autonomous


Nellie Wieland

holding of beliefs, as understood here, means that the belief can be identified with the agent, and the belief can be identified by the agent as their own (Sayre-McCord and Smith 2014, 135–136). This bi-directional identification is what constitutes the agent’s perspective on the world. The self is just the integrated set of autonomously derived and integrated beliefs and desires which are second-order endorsed (and result in second-order volitions), have narrative continuity, and are responsive to interpersonal engagement with others. At this point, it’s worth stepping back and looking at how far we’ve come. It seems at first that any one of the four criteria are adequate for a strong self (in fact, many people have argued just this). The problem of self-abnegation has led me to argue that all four criteria are—in the strongest version of the argument—necessary, but only jointly sufficient if they are sourced autonomously from a first-person perspective. This is perhaps too strong since a strong self is not a single end state: there can be degrees of achievement of a strong self, and degrees of waning. Who among us has such a strong self? Perhaps most of us do, but there are certainly grounds for doubt. The account of selves and agents defended here, and all mainstream theories of agency, assume that there is someone in charge. They assume that each of us has a “subjective motivational set,” or a “practical identity.” It is on this basis that we make decisions, fulfil our desires, and override this or that errant agency. What should we think, then, about the person who is not an “abnegated agent” but is experiencing a momentary lapse? We say that people like this are “not themselves” today or “out of character.” This assumes that there is some self underlying the errant agent being projected today, one who is, for some reason, merely resting? Hiding? Then who is the self that is here today? You are well enough constituted to direct your will, to organise your intentions and efforts, but that is not who you are. But perhaps it is the case that there is no you underneath the projection; there are just different projections (“projections from nowhere”) with varying endurance and recognisability.11 We are whoever we experience most of the time and whomever others take us to be. This might just be right; stable agency and strong selfhood might be illusions. If this is the case, then there may be no difference between the self that has been abgenated and the one who has not. But, rather than undermining the concern of this paper, it amplifies it.12 What I will consider in the next section is the nature of our obligations toward abnegated selves. If it turns out to be true that there’s no meaningful contrast with the abnegated self, then this worry just applies to everybody.

4. Respecting an Abnegated Self In this section I turn to questions about moral engagement with selfabnegating agents: should a person be treated as if their reasons and

The Abnegated Self


narratives are their own even though they are derived from the will of another? What kind of obligation exists to cultivate self-affirmation in the self-abnegating agent? Imagine a person who is profoundly deferential to her spouse now, and who was profoundly deferential to her parents before she was married. Imagine that she has never formulated a life plan of her own in any recognisable way. Her life plans were formed for her, and were not, in any substantive way, even identified as hers. Instead, she has been told to take, and has always taken, the goal of her agency to be to serve the life plans of someone else. We could think of her situation in two ways: we could say (a) that she has a life plan, but it is a plan that has been determined in generalities and specifics by someone else, or (b) that she does not even have a life plan since she is assumed to be a wholly instrumental agent, there to serve the life plan of someone else. Typically, the reason why it is wrong to undermine the life plan (or desires, autonomy, etc.) of an agent, is because that life plan is derived from her process of reason-giving and her values. When these reasons and values, and the goals and plans they produce, are derivative, or can only be understood instrumentally, does the obligation to respect those reasons and values change? I think that the answer to this is tricky, in part because it assumes quite a bit of confidence in ordinary agents to authentically determine and construct their selves. It also assumes a neat separation between different ways of harming the same person. For example, if the person in the previous example were to desire to undergo a medical procedure, but that desire was derivative on the will of her spouse, we might want to distinguish between harming her by blocking her authentic-desire and by blocking her derivative-desire. But, in both cases, this might involve us doing something that brings about pain or bodily control. Again, the reason why it is harmful to undermine or obstruct a person’s life plan is because it comes from that person in some meaningful way; it is one’s own in a way that matters. If this is not the case, then the basis of that harm changes in all of those cases that matter deeply to us—those which evinces values, beliefs, plans, and aspirations. 4.1 Counterfactual Arguments What do we owe the self-abnegating agent? We might think that this question can be answered by simply considering what we would owe them had they never abnegated their self. Consider the counterfactual status of this agent. The claim to consider is that respect for the selfabnegating agent should be grounded in the respect that would be owed that agent if they had lived life without the oppressive circumstances that brought about their self-abnegation. That is, ostensible agents are owed the regard of well-functioning agents because of the counterfactual possibility that they could have been well-functioning agents in the absence


Nellie Wieland

of mistreatment.13 This counterfactual argument is offered as a way to contrast the agentive qualities we have “naturally” and those we have that are socially imposed. We can acknowledge that we will have some difficulty separating natural and socially imposed agentive qualities. Still, when we apply the counterfactual argument in each of the previous cases, the outcomes are uneven. In the first case, the causal structure is too diffuse to locate the counterfactual reality. In the third case, the counterfactual argument is question-begging since the renunciation of the self, and its aspirations, was deliberate. Only in the second will the counterfactual argument be tractable. Even in this case it requires assumptions about ideal agency that are tricky. For example, return to the example of people living in the isolated community enforcing an ideology of maledomination. What is the counterfactual person like? How do we imagine her life? What is the model of agency that she realises in this counterfactual case? Perhaps she is self-aware and prone to interpersonal engagement; perhaps she is well-educated and checking all the boxes on indices of flourishing; perhaps she is respected and admired. But why? The model for this counterfactual person is herself a unicorn among us. This makes her no less a source of norms and expectations, but it means her magical qualities might be a matter of wishful thinking. 4.2 Self-Affirmation for the Self-Abnegated What would it look like to create the conditions of affirmation, and to forestall the conditions of abnegation? This is either to prevent selfabnegation or to recover agency in the abnegated self. It might involve conceiving of a place for treating abnegated selves as strong selves. And it might be virtuous to treat abnegated selves as strong selves (however that ends up looking). If so, are these virtuous acts for the performing agent, for the beneficiary, or are they part of structure of virtue that creates the conditions of agency? This final option is the most coherent. The obligation to affirm the life plans, respect the reasons, and interpersonally engage with the self-abnegated agent is itself weak when understood as a simple moral relationship. Instead, something like a structure of virtue surrounds agency among all of us—realised, potential, and counterfactual. The good of respecting even a derived or instrumental agency comes in putting that individual in a position to flourish, but more importantly it maintains the structure and expectation of agential flourishing that is a necessary condition of overcoming the oppressive conditions that lead to self-abnegation to begin with. The model for our responsibility toward self-abnegated agents cannot be predicated on a one-to-one relationship with the individual with, for example, derivative desires. Assuming that they may, counterfactually, have been a person who could have chosen for themselves, has to be grounded in another round of desire-derivation. If we regard the person

The Abnegated Self 225 who derivatively-desires a medical procedure as someone who does not authentically desire that medical procedure, then we are imposing a desire on them yet again. We can claim that it is counterfactual but we don’t know what that counterfactual looks like: we might be assuming that it looks something like us. Instead, the best model for responsibility toward self-abnegated agents should be thought of structurally, whereby each of us acts in such a way that creates the conditions of agential flourishing, interpersonal engagement, and authenticity. We do not know how things would have been in any given case, and we cannot always find the joints between interfering with derivative agency and other kinds of harm. Perhaps the best we can do is to expect the selves around us to be self-generated as a means of creating the structural conditions of overcoming the oppression that leads to self-abnegation to begin with. This puts us in an uncomfortable spot. On one hand, acting according to counterfactual assumptions of who the self-abnegating agent would have been had they retained or developed an authentic self of their own will not work. But the alternative cannot be to reinforce the abnegation. Return to the woman undergoing the medical procedure. We are imagining that she wants the medical procedure, and that she wants the medical procedure because her husband wants her to have the medical procedure. What would she want if her desires were not derivative? What would she want counterfactually? We cannot know this. There is not sufficient reason for us to think that she would not want the medical procedure. If, in our counterfactual imaginations, we ask what similarly situated women would want, we are again, abnegating her agency to those similarly situated women. The alternative is to respect her current desires—go forward with the medical procedure! We do this knowing that this desire is not authentically her own, even if it is one she is stating now. Modifying Michael McKenna’s golf example, all golfers play with handicaps. It’s just that, in order to have a level playing field, each golfer is allowed more or fewer strokes relative to the players they are playing with. A weak self is still a self, they are just acting given limitations, with compromised abilities to form beliefs, desires, and reasons of their own. Each golfer has different abilities, and has their strokes adjusted to match their abilities, but they are all still allowed to play. The weak self is not expected to have a unique, first-person stance from which they autonomously and authentically construct the narrative of their life, but they are not overlooked entirely. The woman’s request for the medical procedure is a minor act of agency, even if it is founded on a major substructure of abnegation. To respect it is unsatisfying, but it is still what we ought to do. 4.3 Moral Ecology14 I close with a slightly different consideration from that which is discussed in 4.2. If a person is accurately described as abnegated to the point of


Nellie Wieland

quasi-agency, then what obligations exist to cultivate or recover their agency? It is clear that we have obligations to respect agents, but do we have corresponding reasons to create the conditions of agency (or a structure of virtue) and the conditions of affirmation for that agency? If this is the case, it cannot be derived from respect for the abnegated self alone; instead it is derived from an understanding of the relationship between the abnegated self and others. For example, it could be grounded in whatever reasons we have for respecting one another’s existing agency. If A has an obligation to create the conditions of B’s self-flourishing under a scheme of flourishing decided by B, then A may, for the same reason, have an obligation to bring about the conditions where C could even formulate a scheme of flourishing to begin with. Seeing your life’s choices as aligned with a sense of self requires first that you have some sense of self, and so it is not unreasonable to suppose that we ought to help you realise that as a starting point. The ecological system within which the abnegated self lives is one where each of us in the position of acting as if we are all strong selves, and authentically integrated across the features of our agency. In doing so we project outward to others that they are the same. Conversely, we treat others—through our laws, our practices, our stances toward one another—as strong selves, and this is a way of affirming our own status within this moral community. While I do not think that we are “segmented” all the way down, I recognise that we all abnegate some of the time and in some spheres and that occasional deference is a normal part of social functioning. We are all in compromised states and these are usually opaque from the inside. The best we can do, for ourselves and others, is to treat ourselves and others as if our selves are intact through and through as our best hope of making it so.

5. Conclusion In this chapter, I have focused on two related problems. The first is whether there is a descriptive account of the self that can distinguish between what I have called a strong self and an abnegated self. I described four features of selfhood: (i) integration with one’s reasons and desires, (ii) continuity over time, (iii) higher-order reflective endorsement, and (iv) interpersonal engagement and recognition from others. Calling all four of these necessary conditions is too strong since a strong self is not all-or-nothing. Rather it is something that comes in degrees throughout one’s life. These four conditions provide a standard for determining the strength of a self and the extent to which it is lacking or distorted. I have argued that these four features themselves require autonomous integration from a first-person stance, or from a “distinct perspective.” I have compared the treatment of abnegated agency to certain ways of thinking about moral responsibility. If someone is regarded as having a

The Abnegated Self 227 strong self then we presume that they are in a position to demonstrate the integration of their self, demonstrate their higher-order endorsement, and demonstrate its narrative continuity with their previous self. This is similar to the giving and taking of reasons for action with those whom we take to be morally responsible. Finally, I turned from descriptive problems to normative problems. I considered what is owed to the abnegated self and whether those persons should be treated as if they have strong selves. I answered that, yes, they should be treated as if they have strong selves. I rejected that the reason for this is because of counterfactual arguments about what their selves could have been like had they not been oppressed. The counterfactual arguments introduce epistemic quandaries about how we could we know what an agent counterfactually would have wanted in other circumstances, and how that forces us into abnegating that self further by projecting idealised beliefs and desires onto the abnegated self. Instead I argue that we should treat the person with the abnegated self as if they had a strong self because that is how the conditions of agency are cultivated. We both make the world a place where each person can become a strong self, and we project our own selves into that world, through this kind of mutual respect. I acknowledge the counter-intuitiveness of this conclusion since it means respecting the will of a person knowing that that will is not her own.

Notes 1. Philosophers, literary authors, filmmakers, and psychologists have documented many ways in which a self can be lost. A self can be lost in utter isolation (cf. the main characters in David Markson’s Wittgenstein’s Mistress and in Lydia Davis’ The End of the Story, and accounts of the psychiatric effects of solitary confinement, as examples). It can also be brought upon by memory loss and memory fracture, as noted since Augustine. The filmmaker, Luis Buñuel, writes poignantly about this: “You have to begin to lose your memory, if only in bits and pieces, to realise that memory is what makes our lives. Life without memory is no life at all. . . . Our memory is our coherence, our reason, our feeling, even our action. Without it, we are nothing.” (Buñuel 2013, 4–5). In addition to isolation, memory loss, and alcoholism, similar effects are documented with other degenerative conditions. 2. This is a traditional account of selfhood, defended in some version by just about everybody since John Locke; paradigmatically by Michael Bratman, and challenged rarely, cf. David Hume and Elijah Millgram. 3. A fuller defence of this standard is in the text that follows. 4. A classic defence of this view can be found in Flanagan (1996). For another variation, see Schechtman (2007). 5. While there are critics of narrative views of the self (e.g., Lamarque 2004), there are also persuasive defences of narrative views that do not demand a whole-life narrative arc (e.g., Lumsden 2013). 6. Another variation on this view comes from Donald Bruckner (2009). In his discussion of reflective endorsement in the case of adaptive preferences, Bruckner distinguishes between autonomously acquiring and autonomously


7. 8. 9.

10. 11.




Nellie Wieland retaining a preference. They are one’s own once they are retained upon reflective endorsement. I think that this is a good distinction to make. The view defended in this chapter may be slightly stronger but is in a similar vein. For a sample of descriptions and defences of related views, see Westlund (2010), Mackenzie and Stoljar (2000), and Velleman (1997). See Strawson (1962), McKenna (2012), Vargas (2013), and Westlund (2003). See McKenna (2012, 84). Although this seems compatible with McKenna’s view, it’s not entirely clear that he endorses it. He mentions it in a footnote (2012, 82fn3) and claims that this articulation of the challenge comes from Michael Zimmerman. This is the conclusion of Sayre-McCord and Smith (2014), which I have found very helpful in thinking about the problem of self-abnegation. This argument is influenced by Elijah Millgram (2015) and his description of “segmented agents.” I take him to be arguing that all agents are segmented agents, or that there is nothing but “interfaces you conjure up to meet the needs of the moment” (Millgram 2015, 263). I take it that the challenge of segmented agency to the view I’m developing here parallels the challenge of determinism for moral responsibility. If the response to the challenge of determinism is to cast the norms of moral responsibility in the structure of how to act as if we are responsible—the reactive attitudes—then the response to the challenge of segmented agency is to cast the norms of mutual regard in the structure of the conditions of agency as argued in the final section of this chapter. I identify this kind of counterfactual argument with Langton (2009) and with Haslanger (2002). I don’t think either of them ever make this explicit argument, so my attribution should be taken with a grain of salt. Nevertheless, the counterfactual argument discussed here is still something that I think is at least implicit in their views. This is a term used by Manuel Vargas (2013). I may be using it in a different way, but it is meant to have the same spirit.

References Bruckner, D. 2009. “In Defense of Adaptive Preferences.” Philosophical Studies 142: 307–324. Buñuel, L. 2013. My Last Sigh. New York: Alfred A. Knopf. Eliot, G. 1871/2003. Middlemarch. London: Penguin Classics. Elster, J. 1983. Sour Grapes: Studies in the Subversion of Rationality. Cambridge, UK: Cambridge University Press. Flanagan, O. 1996. Self Expressions: Mind, Morals, and the Meaning of Life. Oxford: Oxford University Press. Flanagan, O. 2007. The Really Hard Problem: Meaning in a Material World. London: MIT Press. Frankfurt, H. 1971. “Freedom of the Will and the Concept of a Person.” Journal of Philosophy 68(1): 5–20. Haslanger, S. 2002. “On Being Objective and Being Objectified.” In Louise Antony and Charlotte Witt, eds., A Mind of One’s Own. Boulder, CO: Westview Press, pp. 209–253. Hume, D. 1739/2000. A Treatise of Human Nature. Oxford: Oxford University Press.

The Abnegated Self 229 Lamarque, P. 2004. “On Not Expecting Too Much from Narrative.” Mind & Language 19(3): 393–408. Langton, R. 2009. Sexual Solipsism: Philosophical Essays on Pornography and Objectification. Oxford: Oxford University Press. Lumsden, D. 2013. “Whole Life Narratives and the Self.” Philosophy, Psychiatry, & Psychology 20(1): 1–10. Mackenzie, C. and N. Stoljar, eds. 2000. Relational Autonomy: Feminist Perspectives on Autonomy, Agency, and the Social Self. New York: Oxford University Press. McKenna, M. 2012. Conversation and Responsibility. Oxford: Oxford University Press. Millgram, E. 2015. The Great Endarkenment. Oxford: Oxford University Press. Sacks, O. 1985/2006. “The Lost Mariner.” In O. Sacks, ed., The Man Who Mistook His Wife for a Hat. New York: Simon & Schuster. Sayre-McCord, G. and M. Smith. 2014. “Desires .  .  . and Beliefs .  .  . of One’s Own.” In M. Vargas and G. Yaffe, eds., Rational and Social Agency: The Philosophy of Michael Bratman. Oxford: Oxford University Press, pp. 129–150. Schechtman, M. 2007. The Constitution of Selves. Ithaca: Cornell University Press. Strawson, P.F. 1962. “Freedom and Resentment.” Proceedings of the British Academy 48(1): 1–25. Vargas, M. 2013. Building Better Beings. Oxford: Oxford University Press. Velleman, J.D. 1997. “How to Share an Intention.” Philosophy & Phenomenological Research 57(1): 29–50. Westlund, A. 2003. “Selflessness and Responsibility for Self: Is Deference Compatible with Autonomy?” The Philosophical Review 112(4): 483–523. Westlund, A. 2010. “Deciding Together.” Philosophers’ Imprint 9(1): 1–17. Zimmerman, D. 2003. “Sour Grapes, Self-Abnegation and Character Building: Non-Responsibility and Responsibility for Self-Induced Preferences.” The Monist 86(2): 220–241.

13 Integrity and Messy Lives Tim Dare

1. Introduction Aristotle thought the virtuous person’s life was wonderfully coherent: “his opinions are harmonious, and he desires the same things with all his soul.” The soul of the wicked person, by contrast, is “rent by faction . . . one element . . . draws them this way and the other that, as if they were pulling them in pieces” (Aristotle, 1166b 19–24). We see similar ideas in contemporary accounts of integrity. The “integrated-self” view proceeds from the idea that integrity is related to “integration, the unification of parts”—the elements of the soul in Aristotle’s terms—and so understands integrity as “the integration of ‘parts’ of oneself—desires, evaluations, commitments—into a whole”(Calhoun 1995, 235). The person of integrity is “undivided”(McFall 1987, 7); he “keeps his self intact” (Taylor and Gaita 1981, 148). The person who has achieved self-integration, on Harry Frankfurt’s account, has integrated their competing desires into a single hierarchy, abandoning desires that cannot be integrated, and wholeheartedly endorsing what remains (Frankfurt 1987). Such a person: no longer holds himself at all apart from the desire to which he has committed himself. It is no longer unsettled or uncertain whether the object of that desire—that is, what he wants—is what he really wants. (Frankfurt 1987, 38) For Frankfurt, this is more than a matter of achieving a harmonious soul. On his account, the “mere wanton” who fails to deliberate and decide which of their desires they want to be volitionally effective, simply acting on whichever desire happens to be psychologically strongest at a moment, lacks a self altogether. The creation of the self requires integrating competing desires into a single ordering, separating some desires from the self and relegating them to “outlaw” status and it is, Frankfurt writes, “these acts of ordering and of rejection—integration and separation— that create a self out of the raw materials of inner life” (Frankfurt 1987, 39). Hence the conditions that threaten integration, and so integrity, also threaten the very existence of a self.

Integrity and Messy Lives


These views of integrity, integration, and the self have been influential, but we should be puzzled by their favourable reception. Few of us obtain, or perhaps even aspire to, the peaceful coherence—or psychic harmony— of Aristotle’s virtuous person or Frankfurt’s well-ordered self. Many, perhaps most, of us lead much messier lives than any of these views seem to comfortably accommodate, and not (necessarily) because we are wicked or wanton. I will offer an explanation of these messy lives, and will argue that those who live them can secure something approaching and of equal value to Aristotle’s psychic harmony and Frankfurt’s integration-integrity by commitment to a certain kind of critical reflection upon their lives and a willingness to embrace the recommendations of that reflection. Integration and psychic harmony, I will argue, may be the targets of such reflection, but it is the reflection and engagement itself which grounds integrity, not whether agents in fact secure those outcomes.

2. Integrity and the Messy Life Why might we live messy lives? There may be as many explanations as there are messy lives, but I focus on two common features of contemporary lives: the likelihood that over the course of a rich life we will be forced to choose not merely how we will continue our current lives, but, in a sense to be explained, whether we should do so at all, and the likelihood that we will occupy roles which, inter alia, generate duties and permissions which conflict with those of our non-professional lives or with other roles we occupy. The need to make what I will call choices between lives, and the ubiquity of role-differentiated duties and permissions, place pressure on the idea that those who live rich lives might aspire to the psychic harmony or well-ordered lives described by Aristotle or Frankfurt. 2.1 Choices Between Lives Imagine Ann. She is a young New Zealand woman who came to the United States with her partner five years ago to begin a PhD. Her studies have been demanding, not least because she has had two children while in the US, but she has done well. Ann loves the US, is keen to remain, and is well placed to enter the US job market: her supervisor and PhD examiners are influential and enthusiastic supporters. Her partner, on the other hand, has never settled. While he has done his best to be supportive of Ann’s academic endeavour—he knows how important her studies have been to her—he now wants to return to New Zealand, restart his own career, and raise their children among extended family. He fervently hopes Ann will come—he does not wish to end their relationship—but, he tells her, he is not prepared to follow her as she works her way into a permanent position in the United States. He and the children are going home. Ann struggles with her options. She must choose: will she return to New

232 Tim Dare Zealand, keeping her family together, continuing the life she began with her partner, or will she separate from them, pursuing a new life as an academic in America, becoming a long distance mother, seeing her children in person only from time to time? Choices such as Ann’s are, we might say, not choices “within a life”—it is not about how she will continue with the life she now inhabits. Instead, it is a choice “between” lives: it calls into question whether she should continue with her current life at all, or whether she should abandon that life and begin another. It may seem that I have portrayed Ann’s choice implausibly starkly. She could return to New Zealand and hope to find an academic job there, for instance, and the distinction between choices within and between lives is probably not very clean: even quite fundamental choices may leave large parts of my current life intact while bringing an end to others. However, compromises will not always be possible: returning to New Zealand may in fact, for Ann, involve giving up her particular academic dream. Even allowing that the distinction between choices “within” or “between” lives may be a matter of degree, there will still be choices which clearly fall on the “between lives” side of the blurry line or which raise the same issues as such choices. Indeed, they are common: we confront such choices not merely when we decide whether to pursue a career in another hemisphere far from family, but also when we decide whether to have children (see Paul 2015); to abandon a religious tradition in which we have been raised; to pursue one career or another. Such choices are challenging not only because of their scope, because of what is at stake, but because they defy assessment or justification by reference to an endorsed goal. Choices within a life need not be easy, but they take place with an eye on an accepted end: the overall aim—to continue the life in which they occur—is not at stake, and that life will generate criteria by which to assess alternatives. One should do that thing that makes it most likely that the life one has committed to will go as well as possible. When faced with choices between lives, however, the criteria by which one might assess a choice within a life are themselves at stake. Treating any consideration (for Ann, maintaining her relationship with her children, or securing a good academic job in the US) as determinative is likely to beg the fundamental question: which life should she choose? Selecting and treating considerations from one of the alternatives as criteria for her choice assumes that she has already decided that that is the life she will pursue; that she will organise the desires relevant to that life into an integrated set, discarding the competing desires for the alternative. Ann, we might suppose, does feel “rent by faction”; some of her desires draw her one way while others pull in the opposite direction. She finds the suggestion that she should integrate her desires, outlawing some and endorsing others, intriguing but unhelpful: she sees it would be good to do so, but does not see how. She lacks both psychic harmony and self-integration. But Ann seems neither wicked nor wanton. If it seems

Integrity and Messy Lives


obvious it would be wicked for her to remain in the US while her partner and children returned to New Zealand, simply suppose that she decides to abandon her dream of an academic career and returns to New Zealand. Now she is surely not wicked, but still, she has faced and made a choice between lives. If you think she is wicked to even contemplate the possibility, imagine her making another choice between lives: an earlier choice whether to have children at all, or whether to pursue her dream to be an academic or to pursue a career in law. Good, decent people face such choices. They are tormented by them in part because they are good, decent people, who appreciate the significance of such choices for themselves and others. And Ann is surely not a mere wanton, someone who fails to deliberate and simply acts on whichever desire happens to be psychologically strongest at a moment. Ann agonises over her decision, trying to settle first on one option, and then on the other; trying to treat each settlement as a Razian decision—a second order reason to disregard the pull of the first order reasons which make each option attractive—but without success (Raz 1975, 485). If we care about Ann, we might hope that her current distress will pass. A decade after deciding to return to New Zealand, she might occasionally think about what things would have been like had she chosen the other path, but—if she is to have anything like psychic harmony—she had better not yearn for that life as she did at the moment of choice. As time passes, she may well have integrated her desires, if not outlawing the desire to have been an academic exactly, at least filing it away, like an old photo which piques regret when encountered in a box of old documents, but that hardly intrudes into ones consciousness between spring cleans. It is, I think, unclear how the accounts of integrity gleaned from Aristotle or Frankfurt—or the account of decisions offered by Raz—capture this aspect of the phenomenology of what I have called choices between lives. One problem is the apparent synchronicity of the accounts offered by these theorists. It seems, at least, that Frankfurt would have us integrate our competing desires in a moment and at each moment, just as Raz seems to think it as simple to stipulate that a decision is a second-order reason to disregard competing first order reasons as it is to recognise the institutional status of laws as second-order reasons. But even if we can, over time, come not to feel the demands of desires we chose not to follow, there is no reason to think we can do so simply, or instantly, or by stipulation. Similarly, we may be able to see that it would be psychologically disastrous to continue to strongly feel the pull of desires we have chosen not to follow long after we have given them up, but this is unlikely to be much help at or around the moment of the initial decision: in Ann’s position we are likely to feel drawn in different directions by our competing desires, and the calm recognition that she will (probably) not always feel as bad as she does now may not be available to Ann at the moment of choice.

234 Tim Dare Aristotle and Frankfurt’s focus on desires seems to generate other problems for the application of their views to choices between lives. The satisfaction of occurrent desires seems only one aspect of the object of such choices. Lives are rich: they come with commitments, practices, engagement with others, and will, in the future, present other choices which cannot be identified, let alone understood, in advance. Focusing on desires seems to trivialise the initial choice between lives. Second, we might think that we seek, not merely a life in which our desires are integrated or harmonious, but one in which that integration or harmony makes sense to us and to others. Suppose my desires are the result of a lifetime of indoctrination or more or less complete ignorance about the options. I may never choose to abandon the religious tradition in which I have been raised only because the alternatives never occur to me. I may have psychic harmony, my desires may be thoroughly integrated, but we should be reluctant to grant me integrity. One explanation for the existence of messy lives, lives in which it seems it would be difficult to achieve Aristotelian psychic harmony or Frankfurt’s integrated self, then, is the fact that those of us who live rich lives are likely to face transformative, identity defining choices, choices between and not merely within lives. 2.2 Roles Our lives are everywhere structured by the roles we occupy. We are friends, spouses, professionals, employees, committee members, citizens, and so on. Roles are comprised of role norms—rights, duties, liberties, and so on—that differ from role to role, and that make particular roles the roles that they are. The role of the lawyer and the role of the father simply are the clusters of norms that are distinctive of those roles. The norms of the roles we occupy may conflict with one another (my daughter has been my student, and my permission to favour her over the children of others qua father conflicted with my obligation not to do so qua teacher) and with the norms of what we might think of as ordinary or non-role morality (the parent’s role is widely thought to allow, or even require, parents to favour their children over the children of others, and moral theories that demand otherwise are often thought too demanding on that account; lawyers will occasionally be obliged to help clients obtain legally protected but immoral ends). Given these features of roles, it is not surprising that role occupants at least occasionally feel “rent by faction,” drawn one way and another by the competing demands of their roles. They may struggle to integrate the desires appropriate to conflicting roles into a single and coherent ordering: parents occasionally feel obliged to bend the rules to help their children, and lawyers battle with conflicting desires to be good people and good lawyers. Consider the extreme and tragic case of Alton Logan.

Integrity and Messy Lives


In 1983, 28-year-old Logan was sentenced to life imprisonment for the murder of a security guard at a Chicago McDonald’s. From the outset, lawyers for the actual offender, Andrew Wilson—charged with two unrelated murders—knew that Logan was innocent: their client admitted to them that he had killed the McDonald’s guard. However Wilson refused to waive the lawyer-client privilege which attached to his confession, and well-settled rules prohibited disclosure (and would have made it inadmissible as exonerating evidence) absent that waiver. Wilson’s lawyers, Dale Coventry and Jamie Kunz, swore an affidavit recording their client’s confession, and stored it away, revealing it only after Wilson’s death in prison in November 2007. By then, Logan was 54 and had spent 26 years in jail for a crime he did not commit (National Registry of Exonerations). Coventry and Kunz struggled to find ways to disclose the privileged information they held (for a time, it seems, even hoping Logan might be sentenced to death to trigger an exception to the rule) and there is no reason to doubt their reports that they were anguished by the part they played in an appalling injustice (Cohen 2008). They faced an appalling conflict between the well settled norms of their roles as lawyers and the norms of ordinary morality. Michel Montaigne, at least, may have thought Coventry and Kunz’s anguish showed merely that, notwithstanding their willingness to follow a demanding role-norm, they failed to appreciate the real significance of their role. The great French essayist was also Mayor of Bordeaux, but he insisted, “The Mayor . . . and Montaigne have ever been two by very manifest separation” (Montaigne 2003, 774). Montaigne here recommends dramatic “role-detachment”: just as Montaigne did not act when the Mayor acted, nor did Coventry and Kunz act when Wilson’s lawyers acted. It was Wilson’s lawyers—not Coventry and Kunz personally—who played a central role in a grave injustice. Why then, would they, personally, be distressed? However, the detachment strategy itself makes the lives of role occupants messy. If, on the one hand, role occupants can pull it off, they divide themselves, living by two sometimes inconsistent sets of moral constraints. No matter how committed they are to a particular moral view “at home,” they must be prepared to put aside that view when at work and commit themselves to the attainment of goals that, in another role persona, they find morally objectionable. If on the other hand, and as seems likely, they find it difficult to maintain Montaigne’s confidence that professional and lay personalities are “two different people, clearly separated,” they are left with the burden of knowing they have been involved with conduct which they find troubling: if a role occupant is unwilling or unable to detach they must bear “the costs of existing in a state of conflict between personal and professional self” (Yakren 2008, 141). Roles may also defy integration and engender psychic disharmony because of uncertainty about just what they require and how apparent

236 Tim Dare conflicts within or between roles are to be resolved. The norms of some roles—those within hierarchical organisations for instance—may specify the scope and priority of those roles over others. The norms of the role of Lieutenant, for instance, specify that role occupants must obey lawful orders from those who outrank them, (e.g., Captains and Majors) and the norms of those more senior roles will similarly specify the subordination of role occupants. But these clear institutional cases are the exception, and even when role norms address priority and scope, those norms may not remove uncertainty or be uncontroversial. Arguably the norms of law, and so the norms of the lawyer’s role, trump the norms of ordinary morality (so, again, it might seem that Coventry and Kunz’s choice was clear), but much of Western philosophy of law consists of arguments about the existence and precise nature of such priority: books have been written on both sides of the issue (see Simon 2009; and Dare 2016). Clear institutional specification of role- and role-norm priority are the exception not just because many institutional roles are silent or indeterminate as to priority, but because many—perhaps most—significant roles we occupy are social, not institutional. In the ideal case, at least, those creating institutional roles begin with a role-function in mind—provide health care, gather and disseminate news, educate children, and so on—and build roles by creating and attaching the role-norms required to fulfil that function to the role name: doctors are bound by patient confidentiality because that obligation makes it more likely patients will disclose information the doctor needs to provide care; journalists have role-differentiated permissions to be nosey because only by being nosey will they gather the information we want them to disseminate; and so on. In all of these cases the role-norms are attached to the role in order to allow the role-occupant to carry out the function of the role, and that derivative structure helps identify and interpret role norms. Many significant roles are institutional in this sense—the role of the lawyer, the physician, the policeman, and so on—but many are not. When I describe myself as a father, if I mean to say more than that I am a male biological parent, I place myself in a social role. There has been no process of authoritative institutional design, guided by a perception of the role of fathers, which has created the role “father” by attaching a set of role-norms to the role name “father.” My description of myself as a father generates expectations about how I ought to behave in those who know what the role “father” is in my community. If they know what “father” means they will have views about what I may and must do to be a good or bad father. The cluster of norms that constitute the role of the father are the product not of deliberate institutional design, but of social expectation produced by widespread beliefs about the function of father: “[T]hose who hold [these] position[s] are expected to act, and perhaps feel in certain ways. Deviations generate surprise, uneasiness, disappointment, or disapproval” (Andre 1991, 78). Although these social roles are everywhere, and may hold great sway over our lives, their

Integrity and Messy Lives


norms are much harder to predict and manage. We are more likely to be surprised, simply because of the greater opacity around the norms which constitute social roles. The distinction between institutional and social roles, furthermore, is not neat and tidy: Institutional roles are rarely purely institutional. The use of institutional role-names—lawyer, doctor, policemen, teacher, etc.—generate social expectations, and so the role-norms associated with such roles may go beyond those deliberately attached to the institutional role by deliberate design. A doctor is subject to all of the role-norms attached to that role by relevant regulation and legislation. In addition, when they hold themselves out as a doctor, they generate expectations among those who share common beliefs about what doctors ought to do and be. In a world in which doctors are widely expected to be compassionate, for instance, a doctor who meets all of the institutional requirements of the role, but who is not particularly compassionate—perhaps they are mainly concerned to maintain technical proficiency, or think it is important to maintain “professional distance” from their patients—may generate Andre’s “surprise, uneasiness, disappointment, or disapproval” in patients who come expecting compassion. Their role is partially constituted by role-norms imposed by social expectations, alongside those put in place by institutional designers, and so the “messiness” contributed by the social constitution of roles touches almost all roles. Roles are a central feature of the lives of us all. These features of roles— their ubiquity, the inevitability that those who live rich lives will occupy more than one at a time and over time, the likelihood that the norms of the roles they occupy will conflict with one another and perhaps with broader moral norms we endorse, uncertainty about role- and role-norm scope and priority—all contribute to the messiness of our lives. There seems no reason to suppose, furthermore, that the role occupants who struggle to find psychic harmony or to integrate the desires which attach to their various roles are wicked or wanton: it is the nature of roles themselves that make it difficult for role-occupants to meet the standards for integrity set by the likes of Aristotle and Frankfurt.

3. Why Does It Matter? I have suggested that accounts of integrity such as those offered by Aristotle and Frankfurt struggle to accommodate what I have called messy lives. Those who have made choices not merely within but between lives, or who occupy multiple or demanding roles, may well find themselves rent by faction, drawn first this way and then the other by competing desires, and may struggle to integrate desires associated with the lives between which they choose, or with the roles they occupy, into a single, coherent, ordering.

238 Tim Dare I have already given one reason to think we should be reluctant to accept that people who live messy lives lack integrity. Ann faced a striking—though I suspect not remarkably rare—decision. If I am right that much more common choices—whether to have children at all; what career to follow; whether to continue or abandon a religious tradition in which we have been raised; and so on—also involve choices between lives, then accepting those accounts will make integrity much rarer than we might have thought. We might make a similar point about roles. Our dealings with others are everywhere mediated by roles. Sometimes that mediation is obvious. When I go into a public hospital, I am unlikely to know anything about the person who comes into the treatment room and introduces herself as a doctor other than that she is holding herself out as a doctor, and what I know of those who legitimately occupy the doctor’s role; what I know, that is, about the norms of the role. What I know about the relevant role-norms will vary, but, in the imagined hospital situation, it is likely to include the knowledge that people who hold themselves out as doctors in respectable hospitals will have received training in medicine; that their competence and performance will have been assessed by medical experts; perhaps that there are norms around issues such as confidentiality, and so on. Whatever I know of the relevant role-norms, in the hospital setting, the mediating function of the doctor’s role is clear: it is all I have to go on. In other cases I might be tempted to think that the roles occupied by me and those with whom I engage are doing little or no work. In dealings with an old friend, I might think that what I know about my friend, and what I can expect from them, is gleaned entirely from what I have learned from them during our hours together, rather than on my knowledge of the norms which constitute the role of friendship and that my friend holds himself out as a friend of mine. Note how obvious it is, even in this case, however, to refer to the friendship role when, for instance, we are surprised that a friend has let us down or has disappointed some expectation. It seems natural to say “I thought you were a friend,” or “I asked you as a friend.” Those responses flow from the belief that the person in question is a friend and the speaker’s knowledge about what one can expect from friends, rather than from what one knows of the person as an individual. We might contrast this case with that in which I say to someone, even a friend, “I’m surprised you did that. Just last week you said you never would.” The mediation of roles seems apparent, then, even in these cases in which it may be tempting to think they are doing little or no work. For now, I simply want the modest claim that the more common messy lives are, the less attractive we should find an account of integrity which denies it to those who live them. There is another and perhaps more important worry about the inability of these influential accounts of integrity to accommodate those who live messy lives. It is not merely that neither Ann or her ilk, nor the vast

Integrity and Messy Lives


majority of role occupants, are wicked or wanton. Many, perhaps most, role occupants are positively good and committed to making careful decisions, and it is precisely that goodness and carefulness that leads them to be rent by faction in the face of groundless choices between lives or the requirements of demanding roles. Ironically, given the correlation Aristotle draws between the absence of psychic harmony and wickedness, their psychological distress is a symptom of their essential goodness and concern to make correct choices, not their wickedness or wantonness. They would enjoy more psychic harmony if they cared less. Note too that roles themselves often serve important social ends. Crucial social functions would be impossible to meet were it not for our capacity to create roles relative to those functions. Physicians need permissions (to examine patients, to administer potentially harmful medications, to give advice on matters of importance to patients, and so on) and they need obligations (to receive appropriate training, to maintain competence, to have proper regard to their patients, and so on) if communities and individuals are to engender and benefit from their skills. These sets of obligations and permissions—the role-differentiated norms of the physician role—simply are the role of the physician, and a similar account could be given for all significant social roles. If I am right that one cannot have complex networks of social roles without the challenges set out in the previous section—without the features of roles which make them contribute to messy lives—then, again, no account which sees the common markers of occupying roles in such contexts as evidence of wickedness or wantonness can be adequate.

4. A Path to Integrity for Those With Messy Lives On their face, the accounts offered by Aristotle and Frankfurt deny integrity to those who live messy lives. I will argue, however, that what matters to integrity is commitment to a certain kind of critical reflection and a willingness to embrace the recommendations of that reflection. Even though those who live messy lives may not secure psychic harmony or manage to integrate their desires into a single coherent ordering, they may be able to secure integrity so understood. This capacity for reflection and response—essentially a capacity to construct and deploy a narratable sense of self—allows us to make sense of our less than coherent lives, to construct and present an integrated account of ourselves to ourselves and to others.1 We are capable of thinking about the trajectory of our lives to date; about how we imagine they will proceed in the future; of interpreting and evaluating what we have done and what we think we ought to do; and of seeking to change our judgements, dispositions, and conduct in response to such reflection. The result of such reflection—the story of our lives that we tell—may not be seamless. It will acknowledge that there are points at which we have made identity conferring decisions for

240 Tim Dare which we cannot offer compelling explanations or justifications and that even good, careful, people may be troubled by the competing demands between and within the social roles they occupy. I begin by noting some potential difficulties with the views of integrity that are on the table. First, for Frankfurt, what matters is the integration of desires. He is silent both as to the content of the desires and as to how integration is achieved. Prima facie, someone might attain integration-integrity even though their desires and volitions are less than morally worthy: they may even be morally obnoxious. Claudia Koonz argues that at least some Nazis held coherent, integrated, belief-sets. The Nazi conscience, she writes, “is not an oxymoron. . . . The popularisers of anti-Semitism and the planners of genocide followed a coherent set of severe ethical maxims derived from broad philosophical concepts” (Koonz 2003, 1). If Koonz is right, then at least some Nazis may have been able to integrate their desires and volitions into a single coherent ordering. They could, it seems, have claimed integrity on the integration view. This objection has led some commentators to conclude that the integration view cannot give an adequate account of integrity; that whatever integration-integrity is, it is not really “integrity.” Damian Cox, Marguerite La Caze, and Michael Levine ask us to consider a less obnoxious case, a person who sells used cars for a living and is wholeheartedly dedicated to selling cars for as much money as possible. Such a person, they write: may well be perfectly integrated in Frankfurt’s sense, but we should feel no temptation at all to describe them as having exemplary integrity. (Cox et al. 2019) The integration view also appears to be indifferent to how agents came to have the desires and volitions they seek to integrate: it seems not to matter whether they have lived an examined life, or whether they have been carefully, and coherently, indoctrinated, never considering for themselves whether they endorse the views they hold. Cheshire Calhoun (1995) raises this objection against Bernard Williams’ suggestion that integrity rests on fdelity to identity conferring commitments. It seems equally relevant to the integration view. Williams deploys a fctionalised account of the life of the artist Gauguin who, on Williams’ account, abandons his family to poverty in France in order to pursue his art in Tahiti. Williams’ Gauguin “turns away from defnite and pressing human claims on him in order to live a life in which, as he supposes, he can pursue his art” (Williams 1981, 22). Does Gauguin have integrity? According to Williams, that depends upon whether following his muse to Tahiti is for him an identity conferring commitment: is this the thing that he “fnds his life bound up with,” that propels him forward, that is a condition of him going on as the same

Integrity and Messy Lives


self? Calhoun objects that it is not enough simply to identify with such commitments. We also need to ask how Gauguin came to have them. Sometimes, she points out, people are merely the passive recipients of identity conferring commitments. Such commitments may be something, in Williams’ phrase, I simply “fnd my life bound up with.” She writes “insofar as we imagine that Gauguin, in pursuing ‘what he found his life bound up with,’ acted merely on a psychologically deep impulse without critically refecting on the value of doing so, we may suspect him of not acting with integrity” (Calhoun 1995, 244). I think there is a solution to both of these objections which preserves at least the spirit of the accounts of integrity we have been discussing. Frankfurt says that integrity and, indeed, the creation of the self, rest upon the integration of desires. It seems that, for Frankfurt, it is the achievement of such integration which does the work. But is easy to think of what Frankfurt describes as a process, something we engage in over time, reflecting on our desires, testing them for strength, considering how we would feel if we abandoned or “outlawed” them. Integration, I propose, is relevant to integrity. It is something to which a person of integrity should have regard in a process of critical reflection, but integrity is not simply a matter of achieving or maintaining such integration. Integration is valuable only where it is the product of appropriate critical reflection. It is actually that reflection and readiness to act in response to it which grounds integrity, so, for Frankfurt, the self. The instances of integration that might plausibly ground integrity, on this account, are those—and only those— which have issued from a process of appropriate critical reflection. Armed with this understanding about what is really doing the work when we seek psychic harmony by ensuring our desires all pull in the same direction or Frankfurtian integration, let’s turn to those who have faced what I have called choices between lives. If they are honest, with themselves and others, there is little prospect of integrating into a single coherent ordering the desires that are attached to the lives between which they choose: there is no such ordering. They did not desire to have all the desires they had, at the moment of choice at least, since many of them were inconsistent. Any such ordering they could produce would misrepresent their actual lives, making them sound simpler—tidier—than they actually were. Often they will have found their choices vexing: psychic harmony will have been unavailable to them. Occasionally, their choices were of the utmost importance, defining their identity, and they faced them without the aid of determinative criteria by which to make or assess their choices: any criteria to which they might have appealed begged the prior question about which life they should choose. They do themselves a disservice by understating the attachment they felt to now forgone options. If integrity requires them to integrate their desires and to attain psychic harmony, there is little prospect of them attaining it, notwithstanding that they are neither wicked nor wanton.

242 Tim Dare I have suggested that we think of what Frankfurt describes as a process, something we engage in over time, reflecting on our desires, testing them for strength, considering how we would feel if we abandoned or “outlawed” them. Considered in this light, it seems a small step to say that whether or not someone attains integrity turns not on the outcome of this process—not on whether a person has actually succeeded in integrating their desires into a single coherent ordering—but instead on how carefully and thoroughly they approached the task. Mightn’t we ascribe integrity to someone who fails to produce an ordered set of desires, but who attempts to do so in good faith? Who reflects carefully upon their lives and can say why they have something other than a unitary ordering of desires and volition? Who acknowledges the difficulty of the task, the tensions they faced, and the desires put aside without good reason— perhaps only because they had to choose and could see no non-random way of doing so; who shows proper regard for themselves and others? Such a person need not be wicked, and surely, as described, is not wanton. Integration, we might say, is not necessary for integrity. What matters is the process—the sincere and genuine engagement—rather than its product. On such an account, Ann should, and can, have integrity. I have suggested that roles too contribute to the messiness of lives in ways which suggest role-occupants may struggle to claim psychic harmony or integrity, even though they are neither wicked nor wanton. The idea that integrity rests not on integration but on a willingness to engage in a certain kind of sincere critical reflection helps here too. With this in mind, consider the lawyers in the Alton Logan case again. They faced a striking and troubling choice. We can assume they battled conflicting desires, and they certainly lacked psychic harmony (we would surely think less of them if they had not). On the one hand they had a well settled role-norm which required them to keep information obtained from a client confidential. Client-lawyer privilege, they knew, was not a minor peripheral aspect of the lawyer’s role: it is frequently referred to as a cornerstone of the profession.2 There was, to be clear, no real doubt about the implications of the rule in the Logan case. Since Logan has been released, there have been many calls for the Chicago rules to be changed, but no serious argument has been advanced that the rule did not in fact require the lawyers to keep their client’s disclosure confidential (see, for instance, Rowe 2014). They were familiar with the background theoretical arguments both for and against the strong rule of client privilege: the rule is usually justified by appeal to the idea that without such guarantees clients would not create conditions in which they were prepared to be frank with their professionals (Upjohn Co. v. United States 1981). They knew the rule had been tested many times and upheld. They even knew of other cases very much like the one they were in, in which courts had disciplined lawyers who broke client confidentiality, and ruled that privileged information was inadmissible, even where it might have freed

Integrity and Messy Lives


an innocent person.3 They feared that if the confession were accepted as evidence against Wilson, their client, there was a real possibility he would be sentenced to death: he was serving life for the murders of two policemen and had narrowly escaped the death sentence for those crimes (CBS News 2008). They spoke to senior colleagues and their bar, without revealing Wilson’s identity or the details of the case, in order to verify their interpretation of the situation. They carefully extracted consent from Wilson to disclose his confession after his death (at the time it seemed likely he might be executed): without that release, the obligation to keep his disclosure confidential would have continued after his death (see Lee Wayne’s case at Upjohn Co. v. United States 1981, fn3). As soon as Wilson died, they came forward with the confession and worked for Logan’s release. If integrity turns upon whether a person engaged in sincere and serious critical reflection and displayed a willingness to embrace the recommendations of that reflection, rather than on whether they have been able to integrate the competing desires generated by their roles and the demands of ordinary morality, or upon whether they were able to attain psychic harmony given their role in a tragic injustice, it seems appropriate to attribute it to Coventry and Kunz. They considered their situation, including their role-norm, carefully, they consulted with senior colleagues, they mitigated the situation in so far as they could by obtaining the affidavit, by coming forward as soon as possible. They were not indifferent to the challenges of their situation: they were not merely and thoughtlessly following a rule. Their roles and the extraordinary situation in which they found themselves made it difficult to integrate their competing desires and volitions, and certainly to maintain psychic harmony. Their lives were messy, but any theory of integrity that denies it to Coventry and Kunz on that account is an inadequate theory. A theory of integrity must allow the possibility of it to those who lead rich and complex lives.

5. Conclusion We should be surprised that Aristotle and Harry Frankfurt’s views of integration, integrity, and the self have been as influential as they have, since few of us obtain, or perhaps even aspire to, the peaceful coherence—or psychic harmony—of Aristotle’s virtuous person or to Frankfurt’s wellordered self. Many, perhaps most, of us lead much messier lives than these views comfortably accommodate, and not because we are wicked or wanton. The messiness of our lives is explained at least in part by the inevitability that we will be forced to make choices, not merely within, but between, lives, and because we occupy social roles, the norms of which are likely to conflict with one another and with broader moral norms we endorse. However, what matters to integrity is not the successful integration of our desires and volitions, or the attainment of psychic

244 Tim Dare harmony, but instead commitment to a certain kind of critical reflection and a willingness to embrace the recommendations of that reflection. So understood, integrity need not be denied to the many of us who live messy lives.4

Notes 1. Peter Goldie makes a similar point in terms of the “narrative sense of the self”: “We are, as has often been remarked, reflective animals, capable of thinking about, and of seeking to change, our own thoughts, feelings, and dispositions of character and personality. A part of that capacity is the capacity to deploy in thought and feeling a narratable conception of oneself: with a narratable past, which one now remembers, interprets, and evaluates in various ways; with a present; and with a narratable future, about which one can make plans, have hopes and aspirations, and so on. This conception of oneself is the narrative sense of self”(Goldie 2011). 2. Legal professional privilege is “a fundamental condition on which the administration of justice as a whole rests.” Lord Taylor in R v Derby Magistrates’ Court, ex parte B [1996] AC 487 at 507. 3. In 1986, Staples Hughes, a North Carolina Lawyer, was told by his client Jerry Cashwell that he, Cashwell, had murdered two people in Fayetteville, N.C., crimes for which Lee Wayne was convicted and sentenced to life imprisonment: Wayne, Cashwell said, was not involved. Hughes stood by, just as the lawyers involved in Alton Logan’s case did, for over 20 years until Jerry Cashwell died in 2002. When Hughes gave evidence in court of Cashwell’s confession, the judge threw out his testimony and reported Hughes for violating attorney-client privilege. Wayne was never released. He died in prison in February 2019 (Rowe 2014, 292–293). And see State v. Macumber, 544 P. 2d 1084, 1086 (Ariz. 1976), holding that the privilege could not be abrogated after the client’s death, even where the deceased had confessed to his attorney and to law enforcement that he had committed the murder for which the defendant had been charged, and his federal public defender wanted to come forward to describe his confession. 4. As ever, Dr Justine Kingsbury has offered invaluable critical insight and assistance as I have written this chapter.

References Andre, J. 1991. “Role Morality as a Complex Instance of Ordinary Morality.” American Philosophical Quarterly 28(1): 73–80. Calhoun, C. 1995. “Standing for Something.” The Journal of Philosophy 92(5): 235–260. CBS News. 2008. “26-Year Secret Kept Innocent Man in Prison.” (March 6). URL = . Cohen, S. 2008. “A Killer’s 26-Year-Old Secret May Set Inmate Free.” Los Angeles Times (April 13). URL = . Cox, D., M. La Caze, and M. Levine. 2019. “Integrity.” In E.N. Zalta, ed., The Stanford Encyclopedia of Philosophy (Spring 2019 edition). URL = .

Integrity and Messy Lives


Dare, T. 2016. The Counsel of Rogues?: A Defence of the Standard Conception of the Lawyer’s Role. London/New York: Routledge. Frankfurt, H. 1987. “Identification and Wholeheartedness.” In F. Schoeman, ed., Responsibility, Character, and the Emotions. Cambridge: Cambridge University Press, pp. 27–45. Goldie, P. 2011. “Self-Forgiveness and the Narrative Sense of Self.” In C. Fricke, ed., The Ethics of Forgiveness. London: Routledge, pp. 89–96. Koonz, C. 2003. The Nazi Conscience. Cambridge, MA: Harvard University Press. McFall, L. 1987. “Integrity.” Ethics 98(1): 5–20. Montaigne, M. de. 2003. “Of Husbanding the Will.” In M. de Montaigne, ed., The Complete Works: Essays, Travel Journal, Letters. Everyman’s Library No. 259. New York: A.A. Knopf, pp. 766–784. National Registry of Exonerations. 2012. “Alton Logan.” The National Registry of Exonerations. URL = . Paul, L.A. 2015. “What You Can’t Expect When You’re Expecting.” Res Philosophica 92(2): 149–170. Raz, J. 1975. “Reasons for Action, Decisions and Norms.” Mind 84(336): 481–499. Rowe, G. 2014. “Potential Expansion, or Modification, to the Permissive Exceptions of Model Rule 1.6: Client-Lawyer Confidentiality in Criminal Law and the Gap.” Journal of Legal Profession 39: 291–303. R v Derby Magistrates’ Court, ex parte B [1996] AC 487. Simon, W.H. 2009. The Practice of Justice: A Theory of Lawyers’ Ethics. Cambridge, MA: Harvard University Press. State v. Macumber, 544 P. 2d 1084, 1086 (Ariz, 1976). Taylor, G. and R. Gaita. 1981. “Integrity.” Proceedings of the Aristotelian Society, Supplementary 55: 143–176. Upjohn Co. v. United States, 449 U.S. 383 (1981). Williams, B. 1981. “Persons, Charater, and Morality.” In Moral Luck: Philosophical Papers 1973–1981. Cambridge, UK: Cambridge University Press, pp. 1–19. Yakren, S. 2008. “Lawyer as Emotional Laborer.” University of Michigan Journal of Law Reform 42: 141–184.


Damian Cox is a philosopher who specialises in the areas of ethics, particularly virtue theory, and philosophy and film. He has co-authored three books and published over 40 journal articles. With Michael Levine, he is author of Thinking Through Film (Wiley-Blackwell, 2012); with Michael Levine and Saul Newman, he is author of Politics Most Unusual: Violence, Sovereignty and Democracy in the War on Terror (Palgrave, 2009), and with Marguerite La Caze and Michael Levine, he is author of Integrity and the Fragile Self (Ashgate, 2003). Cox has written on the virtue of integrity, as well as, more generally, the prospects of a theory of ethics based on virtues. His recent work in virtue ethics includes a defence of “vice ethics,” an account of right action framed in terms of the avoidance of vice. Cox’s recent work includes a defence of psychoanalytic approaches to film criticism and emotional self-knowledge. He uses film and television as case studies for philosophical work, emphasising the potential of doing philosophy through film. Recent work has also included examination of ethics in the academy. He has written on egalitarianism, competition and contest within university environments, and has outlined a theory of academic virtues. Garrett Cullity is Professor of Philosophy at the Australian National University. He taught previously at the universities of Oxford, St Andrews, and Adelaide. A moral philosopher whose work ranges across the theoretical and applied parts of the discipline, he has addressed topics including the nature of moral judgement, the sources of moral knowledge, the relationship between reasons for action and rationality, the content of the moral virtues, the moral emotions, human rights, fairness and collective action, and the ethical issues surrounding international aid, climate change, and our activities as consumers. He is the author of Concern, Respect and Cooperation (Oxford University Press, 2018) and The Moral Demands of Affluence (Oxford University Press, 2004), a co-editor of Ethics and Practical Reason (Oxford University Press, 1997), and an Associate Editor of the journal Philosophy and Public Affairs.



Tim Dare is Professor of Philosophy at the University of Auckland. He is also a lawyer and former research clerk to the New Zealand High Court. He is the author of The Counsel of Rogues? A Defence of the Standard Conception of the Lawyer’s Role (Routledge, 2016) and of articles and book chapters in the philosophy of law and applied and professional ethics, including pieces on vaccination, parental rights to consent to their children’s medical treatment, oncologists’ views about giving information about unfunded treatments beyond the means of their patients, Harper Lee’s To Kill a Mockingbird, and the ethics of use of predictive risk modelling in social policy contexts. He is committed to the practical value of philosophy. Dare works as a data ethics adviser to New Zealand’s Ministry of Social Development and provides ethical advice to other social service agencies in New Zealand and internationally. Ramon Das is Senior Lecturer in Philosophy in the School of History, Philosophy, Political Science, and International Relations, Victoria University of Wellington, New Zealand. Das works mainly in moral and political philosophy; his interests include foundational questions in ethics as well as more practical topics such as the ethics of humanitarian intervention and climate change. Other research interests include evolutionary debunking arguments, free will, and the philosophy of law. He is currently writing a book on “companions in guilt” arguments, which try to show that arguments for moral scepticism prove too much, undermining our scientific and ordinary perceptual beliefs if they work at all. Richard Paul Hamilton is Senior Lecturer in Philosophy and Ethics in the School of Philosophy and Theology, University of Notre Dame Australia. He completed a doctorate at Birkbeck College, University of London, under the supervision of Jennifer Hornsby and Susan James. In 2015–16 he held a visiting position as a Senior Research Fellow at the University currently known as Rhodes, in Makhanda, South Africa. Hamilton works on moral and political philosophy, with particular interests in virtue ethics, the theoretical foundations of professional ethics, and the contribution of the life sciences to our understanding of the human condition. He is currently completing a monograph that attempts to place contemporary naturalistic virtue ethics on a properly naturalistic trajectory while also pushing it in politically radical and emancipatory directions. Jason Kawall is the Carl Benton Straub ’58 Endowed Chair in Culture and the Environment and Professor of Philosophy and Environmental Studies at Colgate University. He has twice been a visiting fellow at the Centre for Ethics, Philosophy, and Public Affairs at the University of St Andrews. His research focuses on virtue ethics and epistemology, with



a particular emphasis on their application to environmental and sustainability issues, the nature of right action, and the epistemically virtuous life. Kawall has published extensively in these and related areas, with his work appearing in such journals as American Philosophical Quarterly, Canadian Journal of Philosophy, Environmental Ethics, Ethics, Policy and Environment, European Journal of Philosophy, and Philosophical Studies, and in a number of edited volumes, including The Handbook of Virtue Ethics and The Routledge Handbook of Virtue Epistemology. David Lumsden studied philosophy at the University of London (Bedford College and University College) and at Princeton University, where he was awarded his PhD. He taught at the University of Waikato, New Zealand, for 36 years. His research has mainly been in the philosophy of language with an emphasis on reference and pragmatics and in the philosophy of mind with an emphasis on intentionality, mental symbols, and narrative approaches to the self. His publications have appeared in Philosophia, Philosophical Studies, Erkenntnis, Dialogue: Canadian Philosophical Review, Philosophical Psychology, Journal of Pragmatics, Philosophy Psychiatry and Psychology, Southern Journal of Philosophy, Metaphilosophy, and elsewhere. Elijah Millgram is the E.E. Ericksen Distinguished Professor of Philosophy at the University of Utah. His research focuses on rationality, with attention both to theoretical and practical reasoning. He is the author of John Stuart Mill and the Meaning of Life (Oxford University Press, 2019), The Great Endarkenment (Oxford University Press, 2015), Hard Truths (Wiley-Blackwell, 2009), Ethics Done Right (Cambridge University Press, 2005), and Practical Induction (Harvard University Press, 1997), and editor of Varieties of Practical Reasoning (MIT Press, 2001). Millgram is currently working on a book on Nietzsche. He has held fellowships from the National Endowment for the Humanities, the Center for Advanced Study in the Behavioral Sciences, and the Guggenheim Foundation. Justin Oakley is Professor and Deputy Director of Monash Bioethics Centre, at Monash University. He is the author of Morality and the Emotions (Routledge, 1993, 2020), co-author (with Dean Cocking) of Virtue Ethics and Professional Roles (Cambridge University Press, 2001), editor of Bioethics (Ashgate, International Library of Essays in Public and Professional Ethics, 2009), and co-editor (with Steve Clarke) of Informed Consent and Clinician Accountability (Cambridge University Press, 2007). He has published articles on a variety of topics in ethics, applied ethics, and moral psychology, including virtue ethics, virtue attribution, shame, role-based evildoing, informed consent, surgeon report cards, surrogate motherhood, and



the ethics of pharmaceutical advertising. Oakley is also co-editor of the quarterly refereed journal Monash Bioethics Review and is currently working on a book-length project on policy applications of virtue ethics in professional practice. Nicholas Ryan Smith is Assistant Professor in the Department of Academic Literacy and Linguistics at the Borough of Manhattan Community College of the City University of New York. He holds a PhD in philosophy from the University of Auckland, under the guidance of Christine Swanton, Glen Pettigrove, and Rosalind Hursthouse. Smith’s research falls primarily in moral philosophy, especially virtue ethics and accounts of right action. His work has appeared in the Australasian Journal of Philosophy and the Journal of Value Inquiry. Christine Swanton is an Honourary Academic in Philosophy at the University of Auckland, New Zealand. She is the author of Target Centred Virtue Ethics (Oxford University Press, forthcoming), The Virtue Ethics of Hume and Nietzsche (Wiley, 2015), Virtue Ethics: A Pluralist View (Oxford University Press, 2003), and Freedom: A Coherence Theory (Hackett, 1992), and has contributed numerous articles to journals and reference works. Swanton is regarded as one of the leading academic figures currently working within the field of virtue ethics. Joseph Ulatowski is Senior Lecturer in Philosophy and Director of the Experimental Philosophy Research Group at the University of Waikato, New Zealand. His research covers three areas in metaphysics and philosophy of language, including truth, facts, and self, and he often employs empirical methods to explore traditional philosophical disputes. Ulatowski is the author of Why Facts Matter (forthcoming) and Commonsense Pluralism about Truth: An Empirical Defence (Palgrave Macmillan, 2017), and his work has appeared in Behavioral and Brain Sciences, Diametros, Erkenntnis, Mind and Language, Philosophia, and Southern Journal of Philosophy. His collaborative work with David Lumsden has investigated the topics of self-narrative, the plurality of selves, disunified agency, and virtue. He has had visiting fellowships with the University of Connecticut Humanities Institute and the University of Aberdeen and has held a number of visiting professorships. Liezl van Zyl is Associate Professor of Philosophy at the University of Waikato, New Zealand. She holds a PhD from the University of Stellenbosch, South Africa. Van Zyl is the author of Virtue Ethics: A Contemporary Introduction (Routledge, 2018) and Death and Compassion: A Virtue-based Approach to Euthanasia (Routledge, 2000), and co-author (with Ruth Walker) of Towards a Professional Model of Surrogate Motherhood (Palgrave Macmillan, 2017). She has published



articles on topics in moral philosophy and applied ethics, including virtue ethics and right action, euthanasia, abortion, and surrogate motherhood. She is an Associate Editor of the Journal of Philosophical Research. Nellie Wieland is Professor of Philosophy at California State University Long Beach. She received her PhD from the University of California San Diego in 2007. She works primarily in the areas of philosophy of language and social metaphysics. She has written on problems of agency, autonomy, and authority in both of those areas. She also works on quotation and other metalinguistic phenomena as a way of theorising about the nature of language.


abnegation see self-abnegation action(s): autonomous 216; bodily 75, 121, 129; causal theory of 10, 69, 71, 75–79, 83, 84, 88n3, 88n6; and character 2, 5, 10, 54, 56, 58, 62–67, 69–90, 107, 111, 120, 133, 141, 144, 178, 179, 202–203n7; and consequences 75, 146, 172–178; description of 7, 37, 40, 79–80, 86, 156, 158, 160–161; as events 5, 61, 128, 129, 178, 179; explanation of 5, 8, 9, 13, 23, 54, 76, 78, 80–81, 83, 86; free 218, 220; goaldirectedness 50, 86, 119–120, 142, 144, 146–147, 160; guided 36, 47–48, 65, 73, 76, 84, 88n4, 95, 100, 104, 114n10; habitual 11, 57, 62, 92, 96, 97, 104, 114n6, 118–119, 135; identity of 8, 79, 129–130, 131, 178; intentional 5, 10, 13, 42, 75–76, 77, 78, 87, 88n4, 118, 197; involuntary 144–145; justification for 1–3, 5, 41, 78, 232, 242; mental 71, 74, 75, 77, 78, 79–80, 109; and narratives 4, 10, 14, 21, 22, 37–38, 69, 79, 80, 156; reasons for 5, 6, 22, 37–38, 41, 43, 55, 62, 65, 78, 79, 85, 128–129, 173, 214, 215, 227; responsible 102, 135, 143–148, 149n3, 227; suberogatory 28–30; supererogatory 28–29, 105, 108–109; voluntary 144–145, 149n5; see also right action; virtuous action action theory 1–5, 14, 77, 88n3 affirming one’s life see self-affirmation agency 13, 14, 94–96, 157, 166, 185–213, 214, 215, 219, 222; agent(s) 3, 5, 10, 11, 12, 13, 14, 22, 27, 29–30, 33, 54, 58, 59, 62, 64,

65–67, 72, 73, 75, 78, 85, 87, 88n7, 91–93, 96, 99, 102, 106, 107, 109–114, 117, 141, 143, 149n3, 157, 164, 166, 173–174, 177, 185–213, 215, 217, 218, 219–227; animal 123–130, 189, 194; autonomous 215–218, 220, 222, 225; disunified 24, 30, 196, 198, 200, 209n40 (see also Nietzsche, Friedrich, decadent, the); goaldirectedness 157, 166–167, 179, 180, 197, 217, 223; morally responsible 140–150; and narrative 3, 4–10, 22, 30, 32–33, 38, 54–68, 72, 87, 91, 102, 157–158, 161, 166, 216–217, 222; priorityguided 187–189, 193, 201–202n5, 204n11; rational 11, 117, 218, 219; unified 13, 185, 193–194, 196, 198–200, 201n1, 205–206n20, 206n22; virtuous 11, 38, 58, 65, 66, 67n7, 85, 91–92, 96, 98–100, 102–104, 106–114, 114n12, 115n15, 120, 143 analytic philosophy 2 Anderson, R. Lanier 205n20, 210n46 Andre, Judith 21, 236, 237 Annas, Julia 3, 6, 57, 62, 66, 67n5 Anscombe, Elizabeth (G.E.M.) 1, 2, 5, 8, 14, 69, 73, 74, 76–79, 94, 95, 128–129, 197 anti-realism 71, 122 aretaic 8, 35, 52n14, 162, 179, 180 Aristotle 1, 29, 52n4, 56, 57, 67n5, 73, 84, 85, 118–119, 128, 133–134, 145, 146, 162, 165, 230–231, 234, 239, 243 Armstrong, David 74, 82 attitudes 39, 43, 84, 162, 179, 201n5, 209n41, 218, 228n12



Audi, Robert 51n1, 137n13 authenticity 219, 225 author 6–8, 64, 65, 68n11, 196–198, 204n13, 205n16, 208n30, 208n34, 209n38, 216, 221 autobiography 15, 61, 64, 72, 185–213; see also self-narrative autonomy 21, 23, 28, 217, 219; see also action, autonomous; agency, autonomous Baier, Annette 81, 88n6 Baier, Kurt 171 Barthes, Roland 7–8 behaviour 12, 25, 54, 59, 60–65, 68n11, 70, 72, 74, 79, 81, 82–83, 118, 121, 129, 137n19, 142, 156, 162, 193, 198, 209n40, 221 belief-desire model 75–76, 88n4 beliefs and desires 5, 71, 75, 76, 77, 79, 80, 87, 214, 216, 220, 221, 222 beneficence 12, 23, 26–28, 33, 44, 140–144 benevolence 26, 75, 76, 98 Bentham, Jeremy 1 Berlin, Isaiah 38–39, 46 Berlinian choices 38–51 Blackburn, Simon 125–126 blameworthiness 12, 102, 107, 140–150 Blum, Lawrence 182, 183 bodily control 223 bodily movements 75, 88n4, 129 Bratman, Michael 13, 201n1, 227n2 Brentano, Franz 42, 52n11 Bubba Ho-tep 200 Buridan’s Ass 38–39 Calhoun, Cheshire 181, 230, 240–241 callousness 27, 107, 198 Casey, Edward 71 causation 10, 76, 79–80 causes 10, 20, 69, 72, 74, 75, 79–84, 88n2, 128, 135, 157, 160, 168n10; structuring 77–79 chance 174 Chappell, Timothy 30 character traits 6, 10, 54, 59–64, 82–87, 144; see also vices; virtues cognitive science 117–139 Comer, Debra 182–183 common sense 56, 95, 123 compassion 25, 75, 237 compulsion 145

confidence 223, 235–236 conformity 22, 23, 25, 28, 118, 120, 133, 204n14, 209n40 consequentialism 75, 93, 94, 95–96, 170, 172–173, 175; see also utilitarianism counterfactuals 74, 76, 172, 182, 186, 215, 223–224, 227, 228n13 courage (bravery) 6, 25, 84, 96, 98, 101, 109, 142, 144 Coventry, Dale 235–236, 243 covering-law debate 76 cowardice 67n9, 144, 180 Cox, Damian 12–13, 170–184, 240, 246 creativity 22, 31 Crowe, David 182 cruelty 134, 160 Cullity, Garrett 9, 35–53, 246 Dancy, Jonathan 52n5, 160 Dare, Tim 14, 21, 23, 230–245, 247 Darwinism 202 Das, Ramon 9–10, 54–68, 100, 247 Davidson, Donald 75–77, 79, 80, 85, 87, 88n3, 88n4 De Bres, Helena 9, 54 deciding 39, 41, 50, 52n12, 99, 206n22 decisions 9, 42, 51, 57, 73, 162, 194, 196, 200, 219, 222, 233, 239 deference 214, 219, 226 Dennett, Daniel 70–71, 72, 79–80 deontology 1, 3, 85, 93–95; see also Kantian ethics Derrida, Jacques 8 desires: appetitive 55, 161, 168n12, 177, 214, 233; causal force of 5, 76, 79, 87, 177, 216, 230; firstorder 177, 217, 218; moral 178; non-rational 159, 161, 173, 220; occurrent 234; rational 14, 80, 81, 215, 217, 221, 230, 233, 237, 239–242; satisfaction of 161; second-order 177–178, 216, 222; see also beliefs and desires desire-satisfaction theory 12, 152, 158, 160–162 Dewey, John 207n28 dilemmas: moral 104; tragic 92, 105–112, 115n14, 115n15, 168n16 dispositions 58, 62, 85 disunified agent 185, 198, 200 disunified self 30, 61, 196–197, 201

Index disunity thesis, the 24, 30–31, 196–197, 198, 210n45 Dorsey, Dale 152–154, 167n4, 168n6, 168n11, 168n13 Dreyfus, Hubert 131 duties: Kantian 26–28, 74–75; roledifferentiated 231, 234 ecology, moral 225–226 egoism objection 11, 112–114 emotional fittingness 12, 22, 151, 152, 157–167 emotions 55, 56, 61–62, 81, 86, 159–160, 162–163, 179 eudaimonia 3, 109, 112, 144, 162, 167n2; see also flourishing (human) eudaimonism 3, 33, 92, 109, 112 events 2, 6, 12, 20, 24, 37, 38, 41, 42, 45, 61, 64, 71, 72, 79, 86, 87, 123, 151–158, 161, 167n1, 168n6, 178–179, 216; see also action(s), as events exemplars 11, 117, 144, 171; see also virtuous person experimental method 122 explanations 5, 80, 231, 240 facts 9, 19, 20, 36, 40; as reasons 35–53, 56, 156, 158, 159; descriptive 110, 132, 180; motivations as facts 35–36, 40, 46, 47; narrative 36, 38, 43, 51, 157; psychological 37; social 19–34 failure 12–13, 170–184 false hope 142–143, 145–148 Farennikova, Anna 137n18 fatalism 193; see also Nietzsche, Friedrich, amor fati (love of fate or destiny) feelings 28, 36, 43, 56, 70, 72, 74–75, 82, 85, 163, 244n1 Flanagan, Owen 82, 86, 216, 227n4 flourishing (human) 72, 106, 109, 111–113, 114; see also eudaimonia folk psychology 75–76 foresight 147–148 Foucault, Michel 6–8 Frankfurt, Harry 14, 177–178, 201n1, 216, 230, 231, 233, 234, 237, 239–243 Frazer, Michael 75 freedom 38, 46, 218 freedom of the will 185, 220 friendship 44–45, 161, 162–163, 238 functional-role concept 187, 196


Gallagher, Shaun 131 generosity 8, 19, 23–24, 25, 29, 31, 60, 74, 86, 112 Gettier, Edmund 4–5 Gibson, James 11, 118, 121–131, 134 Ginet, Carl 5 Goldie, Peter 60, 244n1 Goldman, Alvin 4, 5 good life 3, 4, 20, 32–33, 106, 109, 111, 113, 162, 166, 171; see also eudaimonia Good Place, The 170–173, 175, 184 gratitude 151, 157, 159, 160, 164, 171 group 23, 25, 43, 47, 97, 112, 123, 131, 157, 179, 218, 219 Grover, Lisa 9, 54, 55, 59–65, 68n11, 82 Gruner, John 147 guilt 100, 101, 110, 163 habit 57, 81, 85, 118–119, 133–135; see also action(s), habitual habitual liar case 92, 96–97, 104, 114n6 habituation 3, 11, 83, 133–135 Halpern, Joseph 19 Hamilton, Richard 11, 117–139, 247 Hampshire, Stuart 60, 82 happiness 91, 105, 109, 151; see also eudaimonia Haybron, Daniel 167n2 hedonism 152, 158, 160–162 honesty 73, 96–97, 140 hope 163, 166; see also false hope Hume, David 214 Hursthouse, Rosalind 56, 67n4, 67n5, 73, 91–96, 101, 105–111, 113, 115n14, 115n15, 144, 148n1 hyperspecialisers (serial or parallel) 87n1, 88n7 identity 5, 6, 20, 47, 48, 79–80, 216, 219, 234, 239, 241; narrative 216; personal 24, 69, 70 illusion 220, 222 Ingold, Tim 127–128 inner sense 73, 83 Integrationist Thesis 9, 22–25, 28, 32, 33 integrity 9, 14, 20, 36–38, 45–51, 105, 109, 181, 184, 230–245; see also narrative integrity intent 128, 206n23; authorial 15n3 Intention (Anscombe) 5, 78, 87



intentional action see action(s), intentional intentions 5, 24, 40–41, 77, 87, 208n30, 208n34, 222 Irwin, Terence H. 149n5 Johanson, Jens 114n5, 114n12 Johnson, Robert 92, 96, 97, 104 judgements 12, 35, 120, 141–143, 146, 149n3, 173, 175, 176, 179, 180, 184, 205n18, 239; see also moral judgement Kant, Immanuel 1, 2, 26–28 Kantian ethics 55–56, 73–75, 206n21, 209n40 Kawall, Jason 10–11, 91–116, 114n3, 247–248 Kauppinen, Antti 12, 152, 157–158, 164–167, 168n8 killing 38, 115n15 kindness 8, 9, 19, 36, 37, 174–175 knowledge 4–5, 11, 71, 119, 132, 161, 162, 238 Koffka, Kurt 123–124 Koonz, Claudia 240 Korsgaard, Christine 193–194, 201n1, 206n21, 206n22, 209n40 Kunz, Jamie 235–236, 243 Kvanvig, Jonathan 4 Laidlaw, James 32 Lear, Gabriel 151, 156 Lecky, Prescott 81–83 Lewis, David 76 Locke, John 227n2 love 20, 26–28, 45, 49, 124, 143, 163; see also Nietzsche, Friedrich, amor fati (love of fate or destiny) loyalty 9, 24, 36, 37, 43–51, 52n14, 84 luck 33, 102, 148, 159, 166, 173, 175, 180 Lumsden, David 10, 24, 31, 69–90, 248 Lynch, Michael P. 83 Mackie, John Leslie 122, 126 McDowell, John 57, 135 McFall, Lynne 230 McGrath, Sarah 119 MacIntyre, Alasdair 70, 156, 157, 168n7, 209n37 McMahan Jeff 28 meaning 8, 9, 12, 20, 22, 24, 28, 54, 56, 124, 130, 151, 155–158, 164–165

meaning in life 9, 54 mediocrity 152, 159 mental states 71, 74, 75, 78, 79, 109 Merleau-Ponty, Maurice 121, 122 Mill, John Stuart 1, 168n10 Millgram, Elijah 13, 14, 19, 88n7, 185–212, 228n11, 248 Mink, Louis 70 mistakes 36, 49, 51n2, 100, 106 “Modern Moral Philosophy” (Anscombe) 2, 73, 76–77, 94, 128 Montaigne, Michel 235 moral development 179, 181–182 moral failure 12–13, 170–184 moral judgement 22, 120, 171–184; forensic model of 172–178, 181–184; narrative model of 177–184; summative 171–177 moral knowledge 119 moral perception 117–139; see also perception moral responsibility 13, 143–148, 149n3, 217, 220–221, 226–227, 228n12 motivation 44, 64–65, 72–75, 81, 83, 133, 163, 222 Murdoch, Iris 117, 120 Nagel, Thomas 199 narrative(s) 7, 9, 35, 54, 55, 64–66, 157, 172, 179, 181, 216, 223; accuracy of 25, 31, 64–65, 68n11; as character trait(s) 54, 59–63; and the good life 3–4; and causation 10, 69–90; disunity 24, 30–31, 61; facts 36, 157; as reason(s) for action 22; shape of a life 6, 19–25, 30, 46, 51, 178 narrative ethical meaning 20–22, 25 narrative meaning 20–25, 28, 31–32, 152, 157–158, 163–164 narrative value 12, 151–152, 155, 157–158, 160–161, 163–167 narrative virtues 22, 24–33, 37–38, 82, 83; accuracy 25, 31–32; flexibility 25, 30, 31; global vs. specific 24–26, 30, 31, 32, 63–66; goodness 25, 26–27, 30–31; integrity 23, 25, 30; transition 25, 30, 31, 32 naturalism 119, 122, 126, 130 Nehamas, Alexander 196, 197, 203n7, 203n8, 208n30, 208n34 Nicomachean Ethics (Aristotle) 73, 85, 145, 146, 165

Index Nietzsche, Friedrich 185–212; amor fati (love of fate or destiny) 192–193, 196, 201, 205n18; anthropic argument 185–189, 190, 193, 194, 198, 201n2, 207n28, 209n42; Birth of Tragedy, The (BT) 195, 196, 204n14, 207n26; control 186; decadence 13, 189–190, 193, 194, 195–198, 200–201, 204n10, 205n20, 207n28, 209n40; decadent, the 190, 193–198, 200, 204n11, 204n14; Ecce Homo 190–191, 192, 193, 197, 204n14, 205n17, 208n29, 209n41; Eternal Return 199, 201, 206n24, 207n25, 210n44; Gay Science, The (GS) 21–22, 196, 202n6; On the Genealogy of Morals (GM) 190, 193, 205n16, 207n27; Overman, the 199–201; Twilight of the Idols (TI) 188; Wagner Case, The (CW) 189–190, 204n14; will to power 185–189, 199, 201n5, 202n7, 203n8, 203n9, 205n18, 207n28, 209nn41–43, 210n45; Will to Power, The (WP) 188, 201n3, 203n8 Non-Aretaic Value, Thesis of 162 non-virtuous agent 56, 91, 93–114 Noordhof, Robert 136n1, 136n8 Norgay, Tenzing 61 Norman, Richard 119–120 normativity 3, 9, 54, 63–66, 119, 122 norms 9, 11, 14, 19, 21, 22–25, 32, 117, 118–120, 132, 221, 224 Oakley, Justin 12, 140–150, 140, 142, 144, 150, 248 objectivity 64, 130 O’Neill, Onora 26, 27 ordinary morality 9, 21–24, 28, 33, 235, 236, 243 Parfit, Derek 71, 160 patients 12, 19, 140–143, 145–148, 214, 236, 237, 239 perception 117–137; direct theory of 119; ecological (4E) theory of 118, 121, 125, 131–132, 135–136; indirect theory of 121; representational theory of 118, 121, 131, 135 permissibility 93–97 Perrakis, Manos 207n27 persons 10, 54, 56, 58, 60, 62, 65, 70, 82, 86, 98, 127–128, 216, 219


phenomenology 121, 131, 233 Phenomenology of Perception (Merleau-Ponty) 121, 122 phronesis see practical wisdom Placement Problem, the 119, 122, 136 plans 13, 218, 223–224, 244n1 Plato 3, 11 pleasure 31, 56, 117, 160–163, 168n10; see also hedonism pluralism: in ethics 38, 39 Poellner, Peter 202n6 Pollard, Bill 135 powers 76, 181 practical reason 12, 57, 133, 152, 159–163, 168n12, 185, 194 practical wisdom 3, 6, 72–73, 120, 146–148 pragmatism 121 praiseworthiness 91, 93–98 pride 151, 157, 159, 160, 164, 165, 168n18 pro-attitudes see beliefs and desires; desires properties 74, 82, 87, 119, 122, 126, 136n2, 136n11, 203n8 psychological states 79 psychology: cognitive 70, 82; ecological 121, 131; Gestalt 123–124, 137n14; philosophy of 1, 2, 76–77 rationality: practical 3; subjective and objective 219; see also practical reason Rawls, John 49, 159 Raz, Joseph 233 reasoning 5–6, 14, 21–22, 35–53, 54–58, 62–66, 72, 170–184, 214–229; first- and second order 9, 35–53, 233 reasons: calculative view of 78–79; and causes 10, 69–90, 128–129, 159; “disposition to act for reasons” 54–68; first- and second-order 9, 35–53, 217, 233; moral 11, 21, 106, 117, 177; motivating 6, 39, 41, 55, 57, 163, 173, 174; and narratives 14, 22, 36, 48, 214–215, 219; normative 35, 72, 74; pluralism about reasons 39; space of 135; third-order 47 reasons-responsiveness 9, 11, 36–37, 40, 43, 44, 45–46, 51, 51n1, 51n3, 117–139, 174 reflection 6, 14, 36, 48, 85, 134, 190, 231, 239, 241–244



regret 38, 40, 94, 109–113, 159–160, 161, 163, 233 relativism 21, 202n6 responsibility 13, 19, 37, 143–148, 217, 220–221, 224–226, 228n12 Richardson, John 201–202n5, 202–203n7, 203–204n9 right action 10, 11, 28–29, 54, 55, 57–58, 63–67, 67n1, 73, 75, 91–114, 141; as moral excellence 91, 93–103; standard virtue ethical account (SVAR) 92–93, 96, 98–99, 103–104; target centred account of 28–29 risk 67n9, 83, 98–99, 113, 147–148, 173, 175, 177, 182 Roberts, Robert 3–4, 67n3 role failure 12, 140 role-norms 234–238 roles 14, 231, 234–239, 242–243; and narratives 21, 23; professional 140, 144, 146 Russell, Daniel 10, 91–97, 99–103, 105–109, 111, 113 Sacks, Oliver 70, 86, 87n1, 214 Scanlon, T.M. 168n12 Schechtman, Marya 25, 71, 227n4 Schindler, Oskar 12, 182–183 Schur, Michael 170 Schutz, Alfred 21, 119, 132–134 Schwartz, Michael 182–183 self: autonomous 215–220; integrated 14, 118, 179, 215–217, 226, 230–235, 237, 239–243; wellordered 14, 231, 243 self-abnegation 14, 214–215, 217–227; varieties of 217–220 self-affirmation 218, 223, 224–226 self-control 186, 188, 197 self-narratives 7, 10, 22, 64, 66, 67n9, 69–90; see also autobiography shame 158–161, 163, 165 Shape of a Life Hypothesis (SLH) 154–156 Siegel, Susanna 136n2 situationist critique 9–10, 54, 55, 59–60, 62, 63, 66, 67n8 skill 11, 80, 81, 119, 129, 134, 142, 175, 239 Slote, Michael 20, 75, 152–156, 163 Smart, J.J.C. 149n3 Smilansky, Saul 171 Smith, Nicholas R. 12, 151–169, 166, 249

Sosa, Ernest 1, 2, 4, 5 stability 13, 30, 46, 59, 62, 67, 186, 222 Stangl, Rebecca 29 Stohr, Karen 27 Strawson, Peter 204n11, 217, 221 subjectivism 124–126, 219 substances 125 Svensson, Frans 114n5, 114n12 Swanton, Christine 3, 8–9, 19–34, 141, 149n4, 162, 165, 249 Tessman, Lisa 181 Thompson, Michael 5, 78 time preference 152–154 trying 58–59, 176, 188, 233 Ulatowski, Joseph 1–15, 24, 30, 69–90, 249 uncertainty 235–237 utilitarianism 2, 149n3, 174; see also consequentialism Van Zyl, Liezl 1–15, 72, 73, 75, 91–97, 103–105, 107, 111, 113, 114n10, 115n15, 143–144, 149n3, 249 Velleman, J. David 9, 12, 20, 60, 72, 81–82, 152, 155–157, 164, 168n6; meaning-based analysis of life 155–156 vices 29, 31, 102, 118, 134 vicious action 29–30, 64, 79, 85, 104, 141 vicious person 6, 58, 64, 65, 114n10, 141 virtue epistemology 1, 4 virtue ethics 3, 20–21, 67n1, 72; and action-guidance 73, 95; Aristotelian 57, 66, 67n2, 72, 74, 144, 146; and blameworthiness 140–150; and narrative 21–24, 30–33, 54, 163–167; naturalistic 119; and right action 28–30, 54, 66, 91–114, 144–145; and rights 28; sentimentalist 75; target-centred account of 3, 28–29, 114n1; and well-being 151–167 virtue-rules (v-rules) 73 virtues: acquisition of 133–134; as dispositions 54, 55–59, 67n8, 85; basic 19, 22–23; contrary to virtue 12, 103, 140, 142, 162, 163, 166; intellectual (or epistemic) 1, 4, 5, 146; role differentiated 22–24, 140 (see also narrative virtues); and skill 119, 129, 134–135; as social

Index competence 117, 134, 135; targets of 23, 24, 28–30, 140–142, 163, 174 virtue theory 67n1, 76–77, 86 virtuous action 29, 74–75, 85–86, 94–113, 140, 142–143, 224 virtuous agent 66, 73, 75–76, 79, 85, 91–92, 94–113, 117–120, 133–135, 143, 146–147, 230–231 virtuous perception see perception Vogler, Candace 79, 208n33 voluntariness 145 wanton, the 14, 177–178, 181, 230, 233 wantonness 231, 232–233, 237, 239, 241–242 well-being 152–167, 217; Aristotelian view 162, 167n1; desire-satisfaction


theory of 152, 158, 160–162; hedonistic theory of 152, 158, 160–162; objective list theory of 152, 158, 161–163 wickedness 239 Wieland, Nellie 14, 214–229, 250 Wiggins, David 199 will, the 28, 55, 216, 223, 227 Williams, Bernard 19, 21, 106, 180, 208n32, 240–241 Wilson, George 5 Wittgenstein, Ludwig 15n3, 124, 128 wrongness 23, 29, 35, 140, 143 Yakren, Sofia 235 Zimmerman, Dean 220